Geopolymer and Geopolymer Matrix Composites [1st ed.] 9789811595356, 9789811595363

Citation preview

Springer Series in Materials Science 311

Dechang Jia · Peigang He · Meirong Wang · Shu Yan

Geopolymer and Geopolymer Matrix Composites

Springer Series in Materials Science Volume 311

Series Editors Robert Hull, Center for Materials, Devices, and Integrated Systems, Rensselaer Polytechnic Institute, Troy, NY, USA Chennupati Jagadish, Research School of Physics and Engineering, Australian National University, Canberra, ACT, Australia Yoshiyuki Kawazoe, Center for Computational Materials, Tohoku University, Sendai, Japan Jamie Kruzic, School of Mechanical & Manufacturing Engineering UNSW Sydney, Sydney, NSW, Australia Richard M. Osgood, Department of Electrical Engineering, Columbia University, New York, USA Jürgen Parisi, Universität Oldenburg, Oldenburg, Germany Udo W. Pohl, Institute of Solid State Physics, Technical University of Berlin, Berlin, Germany Tae-Yeon Seong, Department of Materials Science & Engineering Korea University, Seoul, Korea (Republic of) Shin-ichi Uchida, Electronics and Manufacturing, National Institute of Advanced Industrial Science and Technology, Tsukuba, Ibaraki, Japan Zhiming M. Wang, Institute of Fundamental and Frontier Sciences - Electronic, University of Electronic Science and Technology of China, Chengdu, China

The Springer Series in Materials Science covers the complete spectrum of materials research and technology, including fundamental principles, physical properties, materials theory and design. Recognizing the increasing importance of materials science in future device technologies, the book titles in this series reflect the state-of-the-art in understanding and controlling the structure and properties of all important classes of materials.

More information about this series at http://www.springer.com/series/856

Dechang Jia · Peigang He · Meirong Wang · Shu Yan

Geopolymer and Geopolymer Matrix Composites

Dechang Jia Institute for Advanced Ceramics Harbin Institute of Technology Harbin, Heilongjiang, China

Peigang He Institute for Advanced Ceramics Harbin Institute of Technology Harbin, Heilongjiang, China

Meirong Wang Harbin Institute of Technology Weihai, Weihai, Shandong, China

Shu Yan School of Metallurgy Northeastern University Shenyang, Liaoning, China

ISSN 0933-033X ISSN 2196-2812 (electronic) Springer Series in Materials Science ISBN 978-981-15-9535-6 ISBN 978-981-15-9536-3 (eBook) https://doi.org/10.1007/978-981-15-9536-3 © Springer Nature Singapore Pte Ltd. 2020 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Preface

Geopolymers have gained increasing attention over the past decades and are now promising candidates for producing heat-resistant coatings and high-temperature ceramics, encapsulating toxic and radioactive wastes, and fabricating sustainable building and construction materials. Following the surge for developing structural and functional geopolymers, geopolymer-matrix composites have also been widely investigated, endowing geopolymers with more fascinating properties and largely exceeding their applications. This book describes the recent progress in developing geopolymer and geopolymer-matrix composites from the following aspects: (i) geopolymerization mechanisms, (ii) preparation of geopolymer resins and binders, (iii) geopolymer matrix composites, and (iv) mechanical properties of geopolymer and geopolymermatrix composites. Moreover, popularizing geopolymer and geopolymer-matrix composites is further motivated by their eco-friendly and cost-efficient features, which are of dramatic importance to further promote sustainable development. In 2005, Dr. Jia Dechang learned about geopolymers during his visit in Prof. Waltraud M. Kriven’s group in the University of Illinois at Urbana-Champaign, supported by China Scholarship Council (CSC). This new inorganic non-metallic material aroused Dr. Jia’s research interest and further prompted him to open up a series of related research fields, such as investigating geopolymerization mechanisms, tailoring process control and preparation methods, and developing geopolymer-matrix composites targeted for flame- and heat-resistant lining for aircraft cabin. After returning to China in 2006, the follow-up studies were successively supported by Heilongjiang Provincial Science Fund for Distinguished Young Scholars, the first batch of excellent science and technology innovation teams of Harbin Institute of Technology, the excellent innovation group of National Natural Science Foundation of China, and the cultivation plan for internationally renowned scholars of Harbin Institute of Technology. Up to now, Dr. Jia has been leading his group to investigate geopolymer and geopolymer-matrix composites for 15 years, achieving a succession of original results. Aiming to summarize the past and recent development of geopolymer and geopolymer-matrix composites, this book first described the concept, classification, v

vi

Preface

and geopolymerization mechanisms of geopolymer from the perspective of materials science. This is followed by introducing geopolymer-matrix composites, in which the low strength and brittle failure nature of bare geopolymer was improved. Geopolymer-matrix composites with different second phases, including graphene, nanotube, particle, short fiber, and continuous fiber, were reviewed. We hope that it could play a positive role in promoting the research, development, and applications of geopolymer and geopolymer-matrix composites. This book could be used as a textbook or reference book for students who majored in materials science and researches who engaged in the field of geopolymer and geopolymer-matrix composites. We are glad to receive any helpful suggestions and constructive comments from the readers. August 2020

Authors In Science Park of Harbin Institute of Technology, Harbin, China

Acknowledgements

The authors gratefully acknowledge the financial supports of the National Natural Science Foundation of China (NSFC, Nos. 51225203, 51372048, 51621091, 51872063, 52072090), the Heilongjiang Touyan Innovation Team Program, the Science Fund for Distinguished Young Scholars of Heilongjiang Province, and the Natural Science Foundation of Heilongjiang Province (No. YQ2019E002).

vii

Contents

1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1 Geopolymerization Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2 Preparation and Properties of Geopolymers and Their Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Preparation of Geopolymers and Their Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Properties of Geopolymers and Their Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3 Applications of Geopolymers and Their Matrix Composites . . . . . . 2.3.1 Nuclear Waste Immobilization . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.2 Precursors for Advanced Ceramics and Ceramic Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.3 Heat-Resistant and Fire-Retardant Properties and Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.4 Next-Generation Building Materials . . . . . . . . . . . . . . . . . . . . 2.3.5 Adsorption of Heavy Metal Ions and Applications in Waste Water Management . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3.6 Biological Antibacterial Properties and Applications . . . . . . 2.4 Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Geopolymerization Mechanism of Geopolymers . . . . . . . . . . . . . . . . . . . 3.1 Mechanism of Transition from Kaolin to Metakaolin . . . . . . . . . . . . 3.1.1 Formation Process and Chemical Activity of Metakaolin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.2 Effect of Calcination Temperature on Thermal Transformation of Kaolin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1.3 Effect of Holding Time on Thermal Transformation of Kaolin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 4 7 8 11 11 14 25 25 26 26 26 28 28 29 30 35 35 35 37 46 ix

x

Contents

3.1.4 Transformation Mechanism of Kaolin to Metakaolin . . . . . . 3.2 Geopolymerization Mechanism of Geopolymer . . . . . . . . . . . . . . . . . 3.2.1 Driving Force of Geopolymerization . . . . . . . . . . . . . . . . . . . . 3.2.2 Geopolymerization Mechanisms . . . . . . . . . . . . . . . . . . . . . . . 3.3 Geopolymerization Mechanisms of Geopolymer Based on Synthesized Metakaolin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.1 Synthesis and Characterization of the Synthesized Metakaolin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.2 Microstructure of K-PSS Geopolymerization Products Based on Synthesized Metakaolin . . . . . . . . . . . . . . . . . . . . . . 3.3.3 Prediction of Hydrolyzed Products of Metakaolin in Alkaline Silicate Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3.4 Geopolymerization Mechanisms of K-PSS Geopolymer Based on the Synthesized Metakaolin . . . . . . . . 3.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

50 55 55 55

4 Graphene-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . 4.1 Reduction Mechanism of GO Under Alkaline Solution . . . . . . . . . . . 4.1.1 Effect of Alkaline Reduction on Functional Group and Valence Bond Structure of GO . . . . . . . . . . . . . . . . . . . . . 4.1.2 Defects and Microstructure Analysis of rGO . . . . . . . . . . . . . 4.1.3 Overview of the Reduction Mechanism of GO Under Alkaline Geopolymeric Solution . . . . . . . . . . . . . . . . . . . . . . . 4.2 Effect of GO on the Mechanism of Geopolymerization . . . . . . . . . . . 4.2.1 Effect of GO on Functional Groups and Valence Bond Structures of Geopolymerization Products . . . . . . . . . . . . . . . 4.2.2 Chemical Structure of 27 Al and 29 Si . . . . . . . . . . . . . . . . . . . . 4.2.3 Phase Analysis of rGO/geopolymer Geopolymerization Products . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4 Micromorphology Analysis of rGO/Geopolymer Geopolymerization Products . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.5 Geopolymerization Mechanism of rGO/Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Microstructure and Mechanical Properties of in Situ rGO/Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Microstructure of rGO/Geopolymer Composite . . . . . . . . . . . 4.3.2 Mechanical Properties and Toughening Mechanisms of rGO/Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . . 4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

81 81

122 127 128

5 Particles-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . 5.1 Carbon Nanotube-reinforced Geopolymer Matrix Composites . . . . 5.1.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.2 Phase Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

131 131 131 132

61 62 64 74 76 77 78

81 90 97 98 98 101 107 111 118 119 119

Contents

5.1.3 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1.4 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2 Graphite Powder-Reinforced Geopolymer Matrix Composites . . . . 5.2.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.2 Phase Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.3 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.4 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.5 Microwave Absorption Properties . . . . . . . . . . . . . . . . . . . . . . 5.3 Al2 O3 Particle-Reinforced Geopolymer Matrix Composites . . . . . . . 5.3.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3.2 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4 Chromium Powder-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.2 Phase Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.4.3 Density and Mechanical Properties . . . . . . . . . . . . . . . . . . . . . 5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.2 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5.3 Mechanical and Thermal Properties . . . . . . . . . . . . . . . . . . . . . 5.6 SiO2 Particle-Reinforced Geopolymer Matrix Composites . . . . . . . . 5.6.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.2 Phase Compositions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.3 Chemical Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.4 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.5 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.6.6 Chemical Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1 Random-Csf /Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.1 Phase Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.2 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.3 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1.4 Fracture Behavior of the Composites . . . . . . . . . . . . . . . . . . . . 6.2 2D-Csf /Geopolymer Composites with Different Fiber Length . . . . . 6.2.1 Phase Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.2 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.3 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2.4 Fracture Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xi

134 134 136 136 137 138 139 140 143 143 146 148 148 150 151 154 154 156 161 166 166 167 167 170 170 172 175 176 179 181 181 182 182 185 186 186 188 189 192 195

xii

Contents

6.3.1 6.3.2 6.3.3 6.3.4

Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fracture Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In Situ SEM Observation on Crack Initiation and Propagation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4 2D-Csf /Geopolymer Composites with Fiber Surface Treatment . . . . 6.4.1 Preparation Process of the Ni/P Coating on the Surface of Carbon Fiber . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.2 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.3 Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5 Effects of Thermal Load on the Microstructure and Mechanical Properties of 2D-Csf /Geopolymer Composites . . . . 6.5.1 Phase Composition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.2 Microstructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5.3 Thermal Shrinkage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.1 Flexural Strength Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.2 Modulus Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.6.3 Pulling-Out Energy Prediction . . . . . . . . . . . . . . . . . . . . . . . . . 6.7 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1 Short SiC Fiber-Reinforced Geopolymer Matrix Composites . . . . . . 7.1.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.2 Effect of SiCsf Content on Microstructure and Mechanical Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.3 Effects of SiCsf Length on the Mechanical Properties . . . . . . 7.1.4 In Situ Crack Growth and Fracture Behavior . . . . . . . . . . . . . 7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.1 Preparation Process Methods . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2.2 Microstructure and Mechanical Properties of Hybrid Csf and SiCsf Reinforced Geopolymer Composites . . . . . . . . 7.2.3 Microstructure and Mechanical Properties of Hybrid Csf –SiCsf –Al2 O3p Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

195 196 199 201 209 210 211 213 216 217 217 219 224 224 228 232 238 239 243 243 243 245 248 251 253 253 254

259 269 270

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites . . . . . 271 8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271 8.1.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 271

Contents

8.1.2 Unidirectional Carbon Fiber-Reinforced Geopolymer (Cuf /Geopolymer) Composites . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.3 2D Carbon Fiber-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1.4 Effects of High-Temperature Treatment on the Microstructure and Mechanical Properties of Cuf /Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . . . 8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.1 Preparation Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.2 Unidirectional SiC Fiber-Reinforced Geopolymer (SiCuf /Geopolymer) Composites . . . . . . . . . . . . . . . . . . . . . . . 8.2.3 High-Temperature Mechanical Properties of SiCuf /Geopolymer Composites . . . . . . . . . . . . . . . . . . . . . . 8.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

xiii

272 282

286 294 294 295 302 304 307

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309

Chapter 1

Introduction

Abstract This chapter is a brief introduction of geopolymer and geopolymer-matrix composites, mainly including the composition, structure, classification, and properties of geopolymer materials and the reason for why to develop geopolymermatrix composites. Among all the geopolymer-matrix composites, fiber-reinforced geopolymer-matrix composites show the best mechanical performances, and some typical results on fiber-reinforced geopolymer composites were reviewed.

Geopolymers are inorganic polymeric materials with three-dimensional network being composed of cross-linking [AlO4 ] and [SiO4 ] tetrahedral units and alkali metal cations. After more than 40 years development, especially in the past 20 years, geopolymer, as a new type of inorganic non-metallic material with outstanding performance characteristics, has attracted more and more attention of material scientists and technicians. As one kind of alkali-activated cementitious material, geopolymer is cured under certain conditions. The strength is obtained from the polymerization of –Si–O– and –Al–O– units, which is very similar to the organic polymer (i.e., organic polymer materials), and geopolymers are typically amorphous or partially crystallized. The concept of geopolymer was first proposed by Joseph Davidovits in the 1970s. In geopolymer, [AlO4 ] and [SiO4 ] tetrahedral units crosslink with each other by sharing oxygen atoms, and cations such as Li+ , Na+ , K+ , and/or Cs+ can distribute in the pores of the geopolymer network balancing the negatively charged [AlO4 ] units [1–3], and the analog structure of geopolymer is shown in Fig. 1.1. According to the activator, geopolymer can be divided into alkaline-activated geopolymer system and phosphate-activated geopolymer system. According to the open literatures, alkaline-activated geopolymer system is the most widely studied. In addition, alkaline-activated geopolymer systems can be classified as follows: (1) According to the type of raw materials. Minerals and industrial by-products, such as metakaolin, illite, quartz, fly ash, and slag, can all be used to prepare geopolymer, and the resulted geopolymer can be named as metakaolin-based geopolymer, quartz-based geopolymer, fly ash-based geopolymer, and slagbased geopolymer, accordingly. Metakaolin powders are commonly chosen for geopolymerization mechanisms study of geopolymer, due to its relatively © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_1

1

2

1 Introduction

Fig. 1.1 Analog structure of geopolymer [4]

Fig. 1.2 Structural unit model of geopolymer with different Si/Al ratios [5]

pure composition avoiding pollution of impurity which commonly coexisted in the industrial by-products. Metakaolin-based geopolymer also showed finer microstructure, higher mechanical and thermal properties than others. (2) According to the amount of AlO4 and SiO4 units (Si/Al ratio) in the geopolymer monomer. Davidovits divides the geopolymers into four categories: poly(sialate) when Si/Al = 1 (PS), poly(sialate-siloxo) when Si/Al = 2 (PSS), poly(sialatedisiloxo) when Si/Al = 3 (PSDS), and poly(sialate-multisiloxo) when Si/Al > 3 (PSMS), as shown in Fig. 1.2. Usually, with the increase in Si/Al, the

1 Introduction

3

mechanical properties of the geopolymer increased but both chemical stability and heat-resistance decreased. (3) According to the activation ions, geopolymer can be divided into Li+ , Na+ , K+ , Rb+ , Cs+ , NH4 + , Ca2+ , Ba2+ , or mixed alkali ions. The obtained geopolymer can be named as Na-based geopolymer, K-based geopolymer, Cs-based geopolymer, Csx K(1-x) -based geopolymer, and so on with the increase in radius of alkali ions, and the resulted geopolymer showed enhanced mechanical and thermal properties. For example, K-based geopolymer shows much high geopolymerization rate than Na-based geopolymer, and the melting points for leucite (derived from K-based geopolymer) is much higher than that of nepheline (derived from Na-based geopolymer), and pollucite (derived from Cs-based geopolymer) possessed the highest melting points as high as 1900 °C among all the glass ceramics. Geopolymers could be prepared at relatively low temperature but could serve at high temperature ranges up to 1000–1200 °C, besides their comparable mechanical properties with traditional ceramics, resins, and aluminum alloys (see details in Table 1.1), such as low thermal conductivity (0.24–0.38 W/m K), low density (2.2– 2.7 g/cm3 ), heat resistance, corrosion resistance, radiation resistance, insolubilization in organic solvents, and anti-leaking, and there are many virtues such as a wide range of raw material sources, simple technology, low-temperature preparation, less energy consumption, and low pollution of the environment. These unique properties make geopolymers as promising heat-resistant structural materials in the fields of construction, metallurgy, plastics, refractories, automobiles, and aerospace [6]. Moreover, the development and popularization of geopolymer is further motivated by its environmentally friendly synthetic method, which has low energy consumption and CO2 emissions [7]. Up to now many research institutions, including but not limited to Geopolymer Institute (French), University of Illinois at Urbana-Champaign (USA), the University of Melbourne (Australia), the University of Sheffield (England), Harbin Institute of Technology (China), Guangxi University (China), Southeast University (China), Lanzhou Institute of Chemical Physics (China), have shown their particular research interests in developing geopolymer materials. However, like other brittle materials such as cement and ceramic, neat geopolymer materials have the disadvantages of both low mechanical performance and catastrophic fracture behavior, which are the insurmountable obstacle and limit their wide applications, especially when they are used as structural components. Therefore, various strategies have been proposed, and typically, geopolymer-matrix composites have been developed with the presence of different kinds of second phases, including graphene [9–12], nanotube [13, 14], particle [15–19], epoxy [20–23], short fiber [24–31], and continuous fiber [19, 32–36], to address their brittle fracture and low reliability nature during service period. Among all the reinforcements, fibers showed much better strengthening effect on the geopolymer. For example, steel fiber-reinforced geopolymer-matrix composite showed a flexural strength of three times as high as that of neat geopolymer, as reported by Ranjbar et al. [37]. Flexural strength, compressive strength, and fracture toughness of flax fabric-reinforced

4

1 Introduction

Table 1.1 A comparison of the physical properties of geopolymer, Portland cement, glass, ceramic and aluminum alloy [8] Type

Density (g cm−3 )

Elastic modulus (GPa)

Tensile strength (MPa)

Flexural strength (MPa)

Work of fracture (J cm−3 )

Geopolymer

2.2–2.7

50

30–190

40–210

Portland cement

2.3

20

1.6–3.3

5–10

20

Glass

2.5

20

1.6–3.3

5–10

20

50–1500

Ceramic

3.0

200

100

150–200

300

Aluminum alloy

2.7

70

30

150–400

10,000

geopolymer composites were 411.1, 369.1 and 350.0% higher than those of pure geopolymer, respectively, as reported by Assaedi et al. [38]. Short carbon fiberreinforced geopolymer composites showed improved mechanical properties together with non-brittle failure mode, compared to pure geopolymer, as reported by Tiesong et al. [26]. He et al. [33] prepared continuous carbon fiber-reinforced geopolymer composites. Combined with the high-temperature treatment and sol-SiO2 impregnation, the resulted composites showed comparable mechanical performance to advanced ceramic matrix composites such as Cf /SiO2 [39], SiCf /SiC–Al2 O3 –Y2 O3 – CaO [40], and Cf /SiO2 –Si3 N4 [41] composites. Therefore, geopolymer technique provides an alternative route for the preparation of high-performance composites. Meanwhile, different kinds of reinforcements related to different preparation methods, including in situ reduction for graphene oxide, ball-milling for nanotube and particles, extrusion for short fibers, impregnation for short and continuous fibers, 3D printing for particles, or short fibers. This book aims to introduce the results by our research group at Harbin Institute of Technology, with a particular focus on geopolymerization mechanisms, and different kinds of geopolymer-matrix composites in order to overcome its disadvantage of low strength and low reliability. In this book, the geopolymer used is metakaolin-based geopolymer with alkali ion of K and Si/Al ratio of 2 (if not specified). It is hoped that this book can be helpful to the researcher working in the field of geopolymer.

References 1. J. Davidovits, M. Davidovics, Geopolymer: Ultra-High Temperature Tooling Material for the Manufacture of Advanced Composites (1991) 2. J. Davidovits, Geopolymers—inorganic polymeric new materials. J. Therm. Anal. 37, 1633– 1656 (1991) 3. D. Jia, P. He, Development of geopolymer and geopolymer-based composites. J. Chin. Ceram. Soc. 35, 157–166 (2007) 4. J. Davidovits, Geopolymer Chemistry and Applications (Institut Géopolymère, France, 2008) 5. J. Davidovits, Geopolymers and geopolymeric materials. J. Therm. Anal. 35, 429–441 (1989)

References

5

6. P. Duxson, A. Fernández-Jiménez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. Van Deventer, Geopolymer technology: the current state of the art. J. Mater. Sci. 42, 2917–2933 (2007) 7. P. Duxson, J.L. Provis, Designing precursors for geopolymer cements. J. Am. Ceram. Soc. 91, 3864–3869 (2008) 8. T. Lin, D. Liang, D. Jia, Mechanical properties and fracture behaviour of Csf(α-Al2 O3 p) reinforced inorganic polymer matrix composites. J. Synth. Cryst. 38, 283–287 (2009) 9. S. Yan, P. He, D. Jia, Z. Yang, X. Duan, S. Wang, Y. Zhou, Effect of reduced graphene oxide content on the microstructure and mechanical properties of graphene–geopolymer nanocomposites. Ceram. Int. 42, 752–758 (2016) 10. S. Yan, P. He, D. Jia, Z. Yang, X. Duan, S. Wang, Y. Zhou, In situ fabrication and characterization of graphene/geopolymer composites. Ceram. Int. 41, 11242–11250 (2015) 11. N. Ranjbar, M. Mehrali, M. Mehrali, U.J. Alengaram, M.Z. Jumaat, Graphene nanoplatelet-fly ash based geopolymer composites. Cem. Concr. Res. 76, 222–231 (2015) 12. M. Saafi, L. Tang, J. Fung, M. Rahman, J. Liggat, Enhanced properties of graphene/fly ash geopolymeric composite cement. Cem. Concr. Res. 67, 292–299 (2015) 13. K.J.D. MacKenzie, M.J. Bolton, Electrical and mechanical properties of aluminosilicate inorganic polymer composites with carbon nanotubes. J. Mater. Sci. 44, 2851–2857 (2009) 14. M. Saafi, K. Andrew, P.L. Tang, D. McGhon, S. Taylor, M. Rahman, S. Yang, X. Zhou, Multifunctional properties of carbon nanotube/fly ash geopolymeric nanocomposites. Constr. Build. Mater. 49, 46–55 (2013) 15. E. Kamseu, A. Rizzuti, C. Leonelli, D. Perera, Enhanced thermal stability in K2 O-metakaolinbased geopolymer concretes by Al2 O3 and SiO2 fillers addition. J. Mater. Sci. 45, 1715–1724 (2010) 16. Subaer, A. Riessen, Thermo-mechanical and microstructural characterisation of sodiumpoly(sialate-siloxo) (Na-PSS) geopolymers. J. Mater. Sci. 42, 3117-3123 (2006) 17. V. Medri, S. Fabbri, A. Ruffini, J. Dedecek, A. Vaccari, SiC-based refractory paints prepared with alkali aluminosilicate binders. J. Eur. Ceram. Soc. 31, 2155–2165 (2011) 18. V. Medri, A. Ruffini, Alkali-bonded SiC based foams. J. Eur. Ceram. Soc. 32, 1907–1913 (2012) 19. S.A. Bernal, J. Bejarano, C. Garzón, R. Mejía de Gutiérrez, S. Delvasto, E.D. Rodríguez, Performance of refractory aluminosilicate particle/fiber-reinforced geopolymer composites. Compos. B Eng. 43, 1919–1928 (2012) 20. F. Colangelo, G. Roviello, L. Ricciotti, C. Ferone, R. Cioffi, Preparation and characterization of new geopolymer-epoxy resin hybrid mortars. Materials 6, 2989–3006 (2013) 21. C. Ferone, G. Roviello, F. Colangelo, R. Cioffi, O. Tarallo, Novel hybrid organic-geopolymer materials. Appl. Clay Sci. 73, 42–50 (2013) 22. M. Hussain, R. Varely, Y.B. Cheng, Z. Mathys, G.P. Simon, Synthesis and thermal behavior of inorganic-organic hybrid geopolymer composites. J. Appl. Polym. Sci. 96, 112–121 (2005) 23. Y. Zhang, S. Li, D. Xu, B. Wang, G. Xu, D. Yang, N. Wang, H. Liu, Y. Wang, A novel method for preparation of organic resins reinforced geopolymer composites. J. Mater. Sci. 45, 1189–1192 (2010) 24. Z. Yunsheng, S. Wei, L. Zongjin, Z. Xiangming, Eddie, C. Chungkong, Impact properties of geopolymer based extrudates incorporated with fly ash and PVA short fiber. Constr. Build. Mater. 22, 370-383 (2008) 25. T. Lin, D. Jia, P. He, M. Wang, In situ crack growth observation and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 527, 2404–2407 (2010) 26. T. Lin, D. Jia, P. He, M. Wang, D. Liang, Effects of fiber length on mechanical properties and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 497, 181–185 (2008) 27. F.U.A. Shaikh, Deflection hardening behaviour of short fibre reinforced fly ash based geopolymer composites. Mater. Des. 50, 674–682 (2013) 28. F.U.A. Shaikh, Review of mechanical properties of short fibre reinforced geopolymer composites. Constr. Build. Mater. 43, 37–49 (2013)

6

1 Introduction

29. T.S. Lin, D.C. Jia, P.G. He, M.R. Wang, Thermal-mechanical properties of short carbon fiber reinforced geopolymer matrix composites subjected to thermal load. J. Cent. South Univ. Technol. (English Edition) 16, 881–886 (2009) 30. D. Sciti, L. Silvestroni, G. Saccone, D. Alfano, Effect of different sintering aids on thermo– mechanical properties and oxidation of SiC fibers—reinforced ZrB2 composites. Mater. Chem. Phys. 137, 834–842 (2013) 31. J. Yuan, P. He, D. Jia, S. Yan, D. Cai, L. Xu, Z. Yang, X. Duan, S. Wang, Y. Zhou, SiC fiber reinforced geopolymer composites, part 1: short SiC fiber. Ceram. Int. 42, 5345–5352 (2016) 32. M. Alzeer, K. MacKenzie, Synthesis and mechanical properties of novel composites of inorganic polymers (geopolymers) with unidirectional natural flax fibres (phormium tenax). Appl. Clay Sci. 75–76, 148–152 (2013) 33. P. He, D. Jia, M. Wang, Y. Zhou, Improvement of high-temperature mechanical properties of heat treated Cf/geopolymer composites by Sol-SiO2 impregnation. J. Eur. Ceram. Soc. 30, 3053–3061 (2010) 34. D. Pernica, P.N.B. Reis, J.A.M. Ferreira, P. Louda, Effect of test conditions on the bending strength of a geopolymer-reinforced composite. J. Mater. Sci. 45, 744–749 (2009) 35. Q. Zhao, B. Nair, T. Rahimian, P. Balaguru, Novel geopolymer matrix composites with enhanced ductility. J. Mater. Sci. 42, 3131–3137 (2007) 36. J.W. Ginacaspro, P.N. Balaguru, R.E. Lyon, Use of inorganic polymer to improve the fire response of balsa sandwich structures. J. Mater. Civ. Eng. 18, 390–397 (2006) 37. N. Ranjbar, S. Talebian, M. Mehrali, C. Kuenzel, H.S.C. Metselaar, M.Z. Jumaat, Mechanisms of interfacial bond in steel and polypropylene fiber reinforced geopolymer composites. Compos. Sci. Technol. 122, 73–81 (2016) 38. H. Assaedi, T. Alomayri, F.U.A. Shaikh, I.-M. Low, Characterisation of mechanical and thermal properties in flax fabric reinforced geopolymer composites. J. Adv. Ceram. 4, 272–281 (2015) 39. D.C. Jia, Y. Zhou, T.C. Lei, Ambient and elevated temperature mechanical properties of hotpressed fused silica matrix composite. J. Eur. Ceram. Soc. 23, 801–808 (2003) 40. K. Yoshida, T. Yano, Room and high-temperature mechanical and thermal properties of SiC fiber-reinforced SiC composite sintered under pressure. J. Nucl. Mater. 283–287, 560–564 (2000) 41. G.H. Zhou, S.W. Wang, X.X. Huang, J.K. Guo, Improvement of oxidation resistance of unidirectional Cf/SiO2 composites by the addition of SiCp. Ceram. Int. 34, 331–335 (2008)

Chapter 2

Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Abstract Geopolymers consisting of polymeric Si–O–Al networks are long-range, covalently bonded inorganic materials. In recent years, geopolymers and their composites have attracted considerable attentions in the fields of construction, metallurgy, and hazardous elements immobilization due to their unique properties such as low-temperature preparation, facile and cost-effective processing, excellent heat and corrosion resistance, and environmentally friendly feature. Due to the relatively low-temperature curing feature of geopolymers, additives with specific properties and functions can be easily incorporated or doped into geopolymers as needed, therefore, endowing them with electric and thermal conductive, electromagnetic wave absorbing, and luminescent features. In other words, geopolymers could be easily tailored as designed. Moreover, geopolymers and their composites can be converted into ceramic materials with controllable mechanical and thermal properties after being treated or holding at high temperature for a given period of time. This makes it possible to render geopolymers and their composites as promising precursors for the preparation of high-temperature ceramics with potential applications in aerospace, heat-resistant components, and stealth materials. This chapter is dedicated to providing an overview of recent advances empowering the development of geopolymers and their composites in the context of geopolymerization mechanisms, microstructure evolution, synthesis, characteristics, and potential applications (e.g., 3D printing and hazardous elements immobilization). Finally, the current challenges and future opportunities of geopolymers and their matrix composites are also addressed.

Geopolymers consisting of polymeric Si–O–Al networks are long-range, covalently bonded inorganic materials. In recent years, geopolymers and their composites have attracted considerable attentions in the fields of construction, metallurgy, and hazardous elements immobilization due to their unique properties such as lowtemperature preparation, facile and cost-effective processing, excellent heat and corrosion resistance, and environmentally friendly feature. Due to the relatively low-temperature curing feature of geopolymers, additives with specific properties and functions can be easily incorporated or doped into geopolymers as needed, therefore, endowing them with electric and thermal conductive, electromagnetic © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_2

7

8

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

wave absorbing, and luminescent features. In other words, geopolymers could be easily tailored as designed. Moreover, geopolymers and their composites can be converted into ceramic materials with controllable mechanical and thermal properties after being treated or holding at high temperature for a given period of time. This makes it possible to render geopolymers and their composites as promising precursors for the preparation of high-temperature ceramics with potential applications in aerospace, heat-resistant components, and stealth materials. This chapter is dedicated to providing an overview of recent advances empowering the development of geopolymers and their composites in the context of geopolymerization mechanisms, microstructure evolution, synthesis, characteristics, and potential applications (e.g., 3D printing and hazardous elements immobilization). Finally, the current challenges and future opportunities of geopolymers and their matrix composites are also addressed.

2.1 Geopolymerization Mechanisms Metakaolin is the most commonly used raw material for the synthesis of geopolymer, whereas other aluminosilicate minerals, such as fly ash [1–3], tailings [4, 5], perlite [6, 7], potash feldspar [5], and laterite [8, 9], can also act as silicon and aluminum sources for geopolymerization. As the chemical compositions of the other aluminosilicate materials are relatively complex, current works mostly employ geopolymers prepared by metakaolin to study the geopolymerization mechanisms. In their seminal works [10–17], Joseph Davidovits et al. divided geopolymerization into four steps: dissolution, diffusion, polymerization, and solidification. During the first step, decomposition of aluminosilicate powder containing silicon and aluminum takes place with the help of the alkaline solution. This is followed by a homogeneous diffusion of the decomposed solid particles into the liquid phase. Afterwards, polymerization reaction occurs between the Si–O and Al–O tetrahedron, leading to the formation of gel phase, which can further propagate in the capillaries, dissolute and diffuse with the remaining reactants, and remove the remaining water. Finally, geopolymer can be obtained by curing the resulting product in a suitable environment. Duxson et al. [18] found that the resulting geopolymer shows distinct microstructures at different regions, as shown in Fig. 2.1. The first typical microstructure is relatively homogenous and similar to those observed in pure alkali systems (Fig. 2.1a). In contrast, the second one shows the signature of phase-segregation (Fig. 2.1c), which may be attributed to the chemical inhomogeneity in either Si/Al ratio or alkali composition. Furthermore, Autef et al. [19, 20] found that: (1) curing temperature plays a critical role in the geopolymerization kinetics; (2) the activity of SiO2 affected the formation of geopolymer network, and the content of structural water in geopolymer decreases with the increasing temperature. Moreover, Zhang et al. [21, 22] recorded the geopolymerization processes, including generation, development, and evolution by in situ environment scanning electron microscope (ESEM). The results showed that metakaolin particles gradually

2.1 Geopolymerization Mechanisms

9

Fig. 2.1 TEM bright field images of geopolymer, showing a homogeneous regions, b transitional regions, and c phase-segregated regions of the microstructure, with permission from [18]

changed from loose accumulated structures to spongy colloids during the early stage, which is accompanied by the densification of the gel. At the later stage, the accumulation of K and Si was observed in the interfacial transition zone rather than the matrix, leading to compositional difference between these two regions. Such result is consistent with reference [18], where homogeneous, transitional, phase-segregated regions with distinct microstructures were observed. Additionally, it is also worth mentioning that the spongy gel was formed during the whole process and no crystalline phase was detected [21]. Weng et al. [23] studied the effect of aluminum source on the geopolymerization process by calculating partial charges of the aluminosilicate group, and found that aluminum source plays an important role in promoting the geopolymerization process. Geopolymer prepared by smaller metakaolin particles with more soluble aluminum component requires shorter curing time, and shows more homogeneous microstructure and higher mechanical strength. Nie et al. [24] used fly ash and kaolinite as raw materials to synthesize geopolymer and studied the mechanisms underlying precursor dehydration an amorphous phase formation. It has been found that the formation of oligomers (e.g., Si–O–Na and Al(OH)4 − ) is followed by the breaking of partial Si–O and Al–O bonds and is affected by K+ and OH− ions in the solution. The resulting oligomers then condense into gel-like substance and finally form amorphous zeolite-like precursors. In order to avoid the influence of impurities in natural metakaolin, Wang et al. [25] studied the geopolymerization mechanism using geopolymer prepared by synthetic metakaolin. The results showed that after geopolymerization, the five- and six-coordinated Al atoms in metakaolin can completely transform into four-coordinated structural units, whereas Si atoms exist in the form of Q4 (3Al) and Q4 (2Al) structural units. Figure 2.2 depicts the spatial structure model of the resulting geopolymer prepared by synthetic metakaolin. On top of that, Yan et al. [26] achieved the precise control of geopolymerization time for the first time by quenching the reaction at intermediate states, and characterized the geopolymerization process by means of Fourier-transform infrared spectroscopy, nuclear magnetic resonance, and scanning electron microscope. These results presented that after mixing metakaolin powder with alkaline silicate solution, Si atoms in the form of Q4 (1Al) dissolve together with four-, five- and six-coordinated

10

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.2 Spatial structure model of geopolymer prepared by synthetic metakaolin

Al atoms, leading to the hydrolysis of Si–O–Si and Si–O–Al bonds and the formation of monomers consisting of [Al(OH)4 ]− , [AlO(OH)3 ]2− , [Al(OH)4 (OH2 )]− , [Al(OH)5 ]2− , [Al(OH)4 (OH2 )2 ]− , [Al(OH)5 (OH2 )]2− , [SiO(OH)3 ]− , and a small amount of [SiO2 (OH)2 ]2− . Afterwards, the monomers undergo polycondensation reactions and form an amorphous network, during which the form of Al atoms gradually changes from a mixture of four-, five- and six-coordinated Al atoms to four-coordinated Al atoms. Such transformation can be detected by 27 Al NMR and the results are shown in Table 2.1. After 6 h, such transformation is almost complete and the resulting amorphous network is mainly composed of Si in the form of Q4(3Al) structural units and four-coordinated Al structural units. Moreover, the microstructure of the product changes from loose and porous structures at the beginning of geopolymerization to much denser structures after the completion of geopolymerization. In addition, Yan et al. [27] investigated the effect of graphene oxide (GO) additives on the geopolymerization mechanism, whose schematic is shown in Fig. 2.3. They found that GO can be in situ reduced to rGO in alkaline silicate solutions, which accelerates the conversion of Al–O sites into four-coordinated Al and Si in the form of Q4 (3Al) at the beginning of geopolymerization (0–30 min). Meanwhile, it was reported that the introduction of GO causes a larger aggregation and relatively uneven microstructures of the final geopolymer particles. Recent studies regarding the geopolymerization mechanism focus on discriminating and describing different reaction stages (e.g., dissolution, diffusion, polymerization, and solidification) and characterizing the products at different stages, but an in-depth understanding of the critical parameters affecting the geopolymerization process remains elusive. For example, the geopolymer network contains a large number of [SiO4 ] structural units with different coordination structures such as [SiO4 (3Al)], [SiO4 (2Al)], and [SiO4 (1Al)]. It remains unclear how to regulate the relative ratio of the different coordination structures, and insights into the transformation and evolution mechanisms between different coordination structures are still lacking. In addition, the effects of counterions (e.g., size and charge density) and water on the geopolymerization process remain the subjects of controversy. The existing questions and challenges require us to understand the geopolymerization

2.1 Geopolymerization Mechanisms

11

Table 2.1 Reaction products formed at different reaction times [26] Sample

Coordination of Al atom Chemical shift (ppm) Relative area Percentage (%)

MK

4

53.3

2.37

21.6

5

29.5

5.28

48.3

6

3.1

3.29

30.1

4

56.3

3.06

25.8

5

28.8

7.09

59.6

6

3.3

1.73

14.6

KGP-30 min 4

56.9

5.86

27.6

5

29.7

11.28

53.1

6

3.4

4.1

19.3

4

57.6

2.41

54.3

5

27.8

1.52

34.2

6

3.2

0.51

11.5

4

56.7

3.61

71.7

5

27.5

0.82

16.5

6

3.2

0.6

11.8

4

56.3

–

–

5

27.5

–

–

6

~2.3

–

–

KGP-6 h

4

56.3

10

100

6

~2.2

–

–

KGP-24 h

4

56.3

10

100

KGP-0 min

KGP-1 h

KGP-2 h

KGP-3 h

process from a more microscopic picture combining a variety of advanced characterization methods, and only after these are done, can the final structure and properties of geopolymers and their composites be well tailored or developed as expected.

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites 2.2.1 Preparation of Geopolymers and Their Matrix Composites Owing to the outstanding and tunable rheology, room temperature processing, and low curing temperature (40–100 °C) of the geopolymer slurry, geopolymers and their composites can be prepared by traditional casting processing, pressure molding, and 3D printing methods, which are also suitable for preparing cement- and resin-based materials. The advantages and shortcomings of each method are listed in Table 2.2.

12

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.3 Schematic of the geopolymerization process of rGO/geopolymer composites, with permission from [27]

Table 2.2 Preparation methods of geopolymers and their composites Method

Advantages

Shortcomings

Suitable system

Casting processing

Facile processing, suitable for preparing components with complex shapes

Low strength due to high solvent content

Geopolymers and their composites modified by nanotubes, particles, and short fibers

Pressure molding

The resulting materials have low defect density and high strength

Complicated processing and not suitable for preparing components with complex shapes

Geopolymers and their composites modified by nanotubes, particles, short fibers, and continuous fibers

3D printing

Suitable for preparing components with complex shapes

Complicated processing and high cost

Geopolymers and their composites modified by nanotubes, particles, and short fibers

(i) Casting processing Casting processing is the most commonly used method of preparing geopolymers. In this method, raw materials containing Si and Al, alkali-activated solutions, and other solvents are mixed together to form geopolymer slurry with a certain rheology. The obtained geopolymer slurry is then casted in the mold for molding. Such method is also suitable for preparing geopolymer-matrix composites containing particles, nanotubes, and short fibers. Jia’s group successfully developed pure geopolymers, and their matrix composites containing various kinds of reinforcements, such as

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites

13

carbon nanotubes [28], graphene oxide [27, 29–35], Al2 O3 particles [36], short fibers [37], and short SiC fibers [38], using this method. (ii) Pressure molding Pressure molding refers to molding the product obtained by mixing aluminosilicate solid with alkali-activated solutions under a certain pressure of 5–10 MPa usually. Similar to casting processing, pressure molding is also applicable for geopolymermatrix composites using particles, nanotubes, and short fibers as reinforcements. Moreover, when the reinforcement preform is a flake shape (e.g., flake-shaped preform consisting of short fibers or 2D fibers), one can first obtain geopolymer slurry using casting processing and then impregnate the flake-shaped preform with designed layer number or thickness into the geopolymer slurry, therefore, regulating the properties of the resulting composites. Jia et al. [39] fabricated geopolymers with high compressive strength up to 180–200 MPa based on the pressure molding method. Moreover, using this method, Jia and co-workers employed different reinforcements to prepare geopolymer-matrix composites, including short carbon fibers [40–43], short SiC fibers and continuous carbon fibers [44–46], continuous SiC fibers [47], stainless steel mesh [48], and so on. (iii) 3D printing As an emerging molding technique, 3D printing has advantages of high precision, high complexity, and rapid molding. 3D printing can produce components with complex shapes and structures that are very challenging for traditional methods. As current raw materials for 3D printing are relatively expensive, discovering printable materials with low cost toward high-performance components is the current hotspot. Exploring geopolymer for 3D printing is a brand-new research direction, and only powder-based and extrusion-based printing methods were reported so far. Xia et al. [49] developed a powder-based 3D printing methodology for formulating geopolymer-based material, as shown in Fig. 2.4. In this technique, a thin layer of geopolymer powder (~0.1 mm) is bound with the binder solution by the drop-on-demand (DoD) technique, after which the unbound powder is removed by an air blower and the binder solution is removed by follow-up processes. The results demonstrated that the resulting geopolymer shows good printability with high precision. However, pore structures appear after the removal of the binder solution (with a porosity of ~60%), which reduces the mechanical properties of the product (0.91 MPa in the X-orientation and 0.76 MPa in the Z-orientation). Zhong et al. [50] developed the extrusion-based 3D printing method (Fig. 2.5), in which the introduction of plasticizer can significantly improve the rheology and plasticity of geopolymer so that the modified geopolymers, even with different alkali metal cations (e.g., Na+ and K+ ) and Si/Al ratios, are suitable for 3D printing. As such extrusion-based 3D printing method does not involve removing the binder solution, the printed components exhibit denser structures and relatively high compressive strength ranging from 60 to 110 MPa. Based on this method, Zhong and co-workers further printed conductive GO/geopolymer nanocomposites, as shown in Fig. 2.5, with an electrical conductivity of 102 S/m after annealing.

14

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.4 Schematic of the powder-based 3D printing methodology: a 3D printing inkjet printing system, b Enlargement of the area in red rectangle: powder/binder interaction between adjacent layers, with permission from [49]

Furthermore, Zhang et al. [51] used similar extrusion-based 3D printing method to prepare PVA fiber-reinforced geopolymer composites, greatly improving the impact toughness of the geopolymer matrix. At present, the preparation of geopolymers and their composites mainly relies on casting processing and pressure molding methods, whereas other advanced forming methods (e.g., 3D printing and RTM) which can prepare geopolymer-based components with complex shapes are still in their infancy. Modifying the properties (e.g., rheology and viscosity) of geopolymers by tuning their chemical compositions and/or employing additives in order to meet the requirements of advanced forming methods will be the focus of further studies.

2.2.2 Properties of Geopolymers and Their Matrix Composites There are many parameters affecting the properties of geopolymers and their composites, which enable the tunability of their properties in a wide range. For instance, the mechanical and thermal properties of geopolymers can be tuned by regulating the chemical activity of Si and Al sources [52], Si/Al ratios [53], and the type of alkali metal cations [54]. Wang et al. [52] studied the effect of high-temperature treatment of

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites

15

Fig. 2.5 Schematic of the extrusion-based 3D printing method and 3D printed GO/geopolymer products, with permission from [50]

metakaolin on the mechanical properties of the resulting geopolymers. It was found that metakaolin being treated at 900 °C possesses better chemical activity and higher concentration of four-coordinate Al atoms. Flexural strength, compressive strength, and thermal conductivity of geopolymer prepared by 900 °C treated metakaolin are 269, 52, and 378% higher than that prepared by 800 °C treated metakaolin, respectively. Duxson et al. [18] compared the mechanical properties of geopolymers with different Si/Al ratios ranging from 1.15 to 2.15, and found that both flexural strength and elastic modulus increase with the increasing Si/Al ratio. Ren et al. [5] used sodium hydroxide and potassium hydroxide to activate gold mine for geopolymerization and demonstrated that geopolymer activated by potassium hydroxide shows higher compressive strength than that activated by sodium hydroxide. Kriven et al. [55] employed hot isostatic pressing to prepare geopolymers and studied the effect of relative molar ratio of sodium hydroxide and potassium hydroxide on the mechanical properties of the resulting geopolymers. They found that geopolymer reaches the maximum compressive strength when the mole fraction of potassium hydroxide is in the range of 30–50%. As the mole fraction of potassium hydroxide further increases, geopolymerization rate and curing rate both increases, leading to more cracks, larger pore size, and higher porosity in the resulting products. Yuan

16

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

et al. [56] investigated the effects of curing time and curing temperature on the mechanical properties of geopolymers and found that the mechanical properties of geopolymers first increase and then decrease with the increasing curing time and curing temperature. The maximum compressive strength (124.8 MPa) was obtained by curing geopolymers at 80 °C for 19 days. He et al. [57] further reported that the softening temperature and heat resistance of geopolymers both increase with radius of the alkali metal cations. With increasing Si/Al ratio from 2 to 4, although the mechanical properties of geopolymers greatly increase, their melting points and resistance to moisture significantly decrease. Geopolymers have been known as low-density, low-temperature molding, fireretardant, and low-cost components with potential applications in the fields of construction, transportation, aerospace, and defense. However, their inherent brittleness, low strength, and low toughness limit the further commercialization in some applications, which requires relatively high load capacity and reliability. Therefore, strengthening and toughening treatments are necessary to improve their mechanical properties. Thanks to the low-temperature geopolymerization process, many reinforcements can be easily introduced into geopolymers, including metal particles (e.g., chromium and tantalum), ceramic particles, short fibers, whiskers, carbon nanotubes, and continuous fibers. The additives can not only improve the mechanical properties of geopolymers but also optimize their thermal properties and high-temperature stability. Table 2.3 lists the mechanical properties of geopolymer composites with different reinforcements. It can be seen that fiber shows the most significant effect on improving the mechanical properties of geopolymers. Moreover, the introduction of short fibers and continuous fibers can also change the fracture mode from brittle fracture to pseudoplastic one. (i) Carbon nanotube-reinforced geopolymer matrix composites MacKenzie et al. [63, 64] studied the properties of carbon nanotubes-reinforced geopolymer composites and found that the addition of carbon nanotubes only improves the conductivity of geopolymers but shows no obvious effect on their tensile properties. Abbasi et al. [65] studied the influence of multiwall carbon nanotubes (MWCNTs) content on the mechanical properties of geopolymers. The results showed that MWCNTs play a role in bridging microcracks under uniformly dispersed conditions, and MWCNTs contribute to the enhanced mechanical properties by retarding the propagation of microcracks. The addition of 0.5 wt% MWCNTs increased the compressive and flexural strength of geopolymer by 32 and 28%, respectively. Bi et al. [66] made geopolymers with ultrahigh self-sensing performance by introducing carbon nanotubes with SiO2 coating. They found that the SiO2 coating can promote the dispersion of CNTs in the matrix, and the flexural and compressive strengths are increased by ~181.2 and ~21.7%, respectively. Khater et al. [67] investigated the effect of MWCNTs on the properties of alkali-activated geopolymers. The results showed that the addition of MWCNTs can enhance the mechanical properties of geopolymers to the greatest extent when the doping concentration is 0.1%. Meanwhile, the shrinkage and water absorption of the composites are also significantly reduced by MWCNTs doping. They proposed that MWCNTs can enhance the mechanical properties of geopolymers under uniform dispersion conditions by

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites

17

Table 2.3 Mechanical properties of geopolymer-matrix composites reinforced by different reinforcements Reinforcement

Content

Matrix [58]

/

12.3

8.3

54.2

Graphene [26]

0.3 wt%

17.9

8.6

NA

Carbon nanotubes [28]

3 wt%

17.5

12.5

NA

Particles

Cr [59]

50 wt%

24.5

NA

NA

Al2 O3 [36]

~25 wt%

25.8

11.3

NA

SiO2 [57]

40 wt%

85

37.5

NA

Fly ash ball [60, 61]

15 vol%

20.6

NA

NA

Basalt fiber [62]

10 wt%

19.5

NA

NA

Carbon fiber [41]

4.5 vol%

96.6

12.0 ± 0.5

6435.3

SiC fiber [38]

2 vol%

50

NA

1949.6

Short fibers

Continuous fibers

Three-point bending strength (MPa)

Young’s modulus (GPa)

Work of fracture (J m−2 )

Al2 O3 fiber

11.3 vol%

97

NA

NA

Carbon fiber

20 vol%

250

41

6056.2

SiC fiber

Stainless steel mesh

20 vol%

161.6

26.9

4421.4

NA

115.3

11.0

8220

forming strong interface bonding between MWCNTs and matrix. However, when the doping concentration is too high, agglomeration could take place, which has a negative effect on the mechanical properties of geopolymers. One of the biggest challenges in MWCNTs-reinforced geopolymer composites is that MWCNTs are difficult to disperse in the matrix. To improve their dispersion, Yuan et al. [28] modified MWCNTs with carboxyl and studied the effects of MWCNTs concentration on the mechanical properties of geopolymers and their high-temperature products. They proved that carboxyl-modified MWCNTs show a uniform distribution in the geopolymer matrix, and the maximum mechanical properties were obtained at a doping concentration of 3 wt%, as shown in Fig. 2.6. Furthermore, they found that geopolymer can convert into leucite phase upon treatment at 950 °C, and the mechanical properties of the high-temperature products reach the maximum after treatment at 1100 °C. Such enhancement was assigned to the densification of the matrix and proper interface bonding between MWCNTs and leucite matrix. (ii) Graphene-reinforced geopolymer matrix composites Compared to carbon nanotubes, graphene has smaller size and is more difficult to disperse into aqueous solutions. Saafi et al. [68] studied the influence of graphene

18

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.6 Mechanical properties of MWCNTs/geopolymer composites with different MWCNTs doping concentrations, with permission from [28]

oxide on the electrical properties of fly ash-based geopolymers and found that graphene oxide can be reduced to graphene with the help of NaOH solutions. When the doping concentration of graphene oxide is 0.35 wt%, the conductivity of geopolymer increases from 0.77 to 2.38 S/m. Considering that graphene oxide can be easily dispersed in aqueous solutions and reduced in alkaline and hightemperature environments, Yan et al. [32, 35] successfully in situ reduced graphene oxide to prepare graphene-reinforced geopolymer composites and their leucite ceramic composites, and studied the reduction mechanism and the effect of in situ reduction on the geopolymerization and mechanical properties of the products. The results showed that graphene oxide can be easily reduced under alkaline conditions, producing graphene-reinforced geopolymer matrix composites with good dispersibility of graphene and long-term stability. The reduction degree of graphene oxide increases with the increase of reaction temperature. The addition of graphene oxide has no obvious influence on the microstructure, and the introduced graphene oxide can uniformly disperse in the geopolymer matrix with a good bonding state. When the concentration of the reduced graphene oxide is 0.3 wt%, the flexural strength of the resulting composite reaches the maximum value of 17.9 MPa. In addition, the fracture toughness increases with the graphene oxide concentration. 0.5 wt% graphene oxide-doped composite shows 61.5% higher fracture toughness than that of pure geopolymer matrix. This can be attributed to the crack deflection and pullout of graphene, as shown in Fig. 2.7. Integrating geopolymer with multifunctional nanomaterials to endow geopolymer matrix with functional characteristics, such as electrical conductivity, thermal

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites

19

Fig. 2.7 Microstructure of the reduced graphene oxide/geopolymer matrix composites, with permission from [35]

conductivity, adsorption, and magnetism for potential applications (e.g., wave absorption, ion immobilization, and smart road) will be the development tendency in the future. However, detailed mechanisms underlying charge transport, heat transport, and wave adsorption still need to be further investigated. (iii) Particle-reinforced geopolymer matrix composites The preparation process of particle-reinforced geopolymer composites is relatively simple, with mechanical stirring and ball milling process being the two commonly used strategies. Bernal et al. [69] studied the mechanical properties of metakaolinbased geopolymers reinforced by Al2 O3 particles with and without fibers. In both cases, the introduction of Al2 O3 particles improves the compressive and flexural strengths of the geopolymer matrix. During thermal evolution, the geopolymer matrix shrinks, resulting in the formation of cracks and stress concentration. However, specimens with Al2 O3 particles reinforcement shows reduced shrinkage after exposed to high temperatures in the range of 600–1000 °C. They attributed the reduced shrinkage

20

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

to the improved volumetric stability through: (i) crack deflection as a result of the strong interaction between reinforcing particles and the matrix; and (ii) densification of the matrix at high temperatures. He et al. [70] doped geopolymers with amorphous silicon oxide ceramic particles and tuned the Si/Al ratio of the resulting composites by controlling the concentration of silicon oxide ceramic particles. The results showed that both Young’s modulus and bending strength of the resulting composites increase with the doping concentration, reaching a bending strength of 84.3 MPa with 40 wt% silicon oxide ceramic particles. Such high bending strength is comparable with traditional glass ceramics and is associated with the increase in the content of Si–O–Si bonds and the reinforcement effect of the remaining silica particles. Lin et al. [71] further reinforced geopolymer with α-Al2 O3 ceramic particles and found that the introduction of α-Al2 O3 ceramic particles can reduce cracks induced by the internal stress, thus improve the strength of the composites. When the doping concentration of α-Al2 O3 ceramic particles is 8 wt%, the resulting composite reaches the maximum strength of 0.75 MPa m1/2 , which is 1.7 times and 0.4 times higher than that of pure geopolymer matrix and geopolymer doped by the same volume content of short carbon fibers, respectively. Wang et al. [60, 61] developed geopolymer with fly ash balls and reported that although the addition of low-strength fly ash balls can reduce the mechanical properties of the composites, its low density and low thermal conductivity help to reduce the bulk density and thermal conductivity of the composites. Meanwhile, fly ash balls inhibit the heat shrinkage of the geopolymer matrix, therefore, improving the dimensional stability of the composites at elevated temperatures. Song et al. [72] fabricated polyvinyl chloride (PVC) reinforced geopolymer composites by hot pressing method and showed that the resulting composites possess both better mechanical properties and high-temperature stability than those of pure PVC materials. Zhang et al. [73] studied the effects of graphite powder on the mechanical properties and microwave absorbing properties of geopolymers, and demonstrated that a 40 wt% doping of graphite powder leads to the best mechanical properties and an electromagnetic wave reflection loss of −64.8 dB, as shown in Fig. 2.8. The heat-resistant temperature of the resulting composites is up to 1400 °C, indicating their potential applications in heat-resistant absorbing coatings. Particle-reinforced geopolymer composites exhibit comparable mechanical properties with fibers-reinforced geopolymer composites, but both of them cannot prevent composites from catastrophic damage. This limits their further usage as structural components can withstand tensile stress. However, on the other hand, particlereinforced geopolymer composites have facile preparation processes and particles with a small aspect ratio can avoid microcracks caused by thermal mismatch during the heating process. This renders them as promising heat-resistant and waveabsorbing coatings. However, the thermal evolution, high-temperature performance, and interfacial compatibility of particle-reinforced geopolymer composites need to be further studied. (iv) Short fiber-reinforced geopolymer matrix composites The preparation methods of short fiber-reinforced geopolymer composites mainly include single-layer impregnation method and mechanical ball milling and stirring

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites

21

Fig. 2.8 Microwave absorption curves of graphite/geopolymer composites, with permission from [73]

method. Lin et al. [40, 41] studied the effects of short fiber content and length on the mechanical properties of geopolymer-matrix composites prepared by the ultrasonicassisted single-layer impregnation method. The results showed that when the fiber content is 3.5 vol% and the fiber length is 7 mm, the bending strength, fracture toughness, and fracture work of the composites are 4.4 times, 10.6 times, and 118 times higher than those of pure geopolymer matrix, respectively. It is also worth mentioning that all the composites exhibit non-catastrophic fracture characteristics, as shown in Fig. 2.9. The strengthening and toughening mechanisms of the composites were assigned to microcrack toughening, fiber bridging, and pullout (Fig. 2.10). It has also been reported that Ni/P coating can effectively improve the interface bonding strength between the substrate and the fiber [37]. With increasing the thickness of Ni/P coating, the mechanical properties of the resulting composite first increases and then decreases, in line with the prediction model and the estimated threshold. Yuan et al. [38] prepared short SiC fiber-reinforced geopolymer composites by mechanical stirring and achieved the best mechanical properties by adding SiC fiber with a length of 5 mm and a content of 2.0 vol%. The fracture morphology shows remarkable ductile fracture characteristics. Yan et al. [74, 75] further fabricated geopolymer composites that reinforced by a mixture of carbon fiber (with a length of 7 mm) and SiC fiber, and studied their mechanical properties

22

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.9 Optical images of the composites after three points bending tests: a geopolymers, b short carbon fiber-reinforced geopolymer composites, with permission from [40]

Fig. 2.10 Microstructure of short carbon fiber-reinforced geopolymer composites: a–b surface morphology, c–d fracture morphology, with permission from [40]

after high-temperature treatments. The results showed that the combination of these two fibers can significantly improve the mechanical properties of the geopolymer matrix. During the high-temperature mechanical test at 1200 °C, the composites that reinforced by 3 mm silicon carbide fiber and 7 mm carbon fiber exhibited the highest fracture work of 808.7 ± 34.9 J m−2 . No catastrophic fracture was observed in the composites at 800 and 1000 °C, which was attributed to interface peeling, fiber breaking, and pullout. However, as the temperature further increases, the mechanical properties of the composites decrease due to the oxidation reaction at the fiber

2.2 Preparation and Properties of Geopolymers and Their Matrix Composites

23

surface. Ng et al. [76] improved the shear performance of geopolymers by introducing steel fibers, which affect the crack growth and crack width in the matrix. Zhang et al. [77] employed extrusion technique to prepare short polyvinyl alcohol (PVA) fiber-reinforced geopolymer composites, increasing the impact toughness of the composites up to 1833 mJ with a doping concentration of 2.0 vol%. In addition, thanks to the introduction of short PVA fibers, the fracture mode of the composites changes from brittle fracture to ductile one. Compared with continuous fiber-reinforced geopolymer composites, short fiberreinforced geopolymer composites require simpler preparation processes. As such, various geopolymer composites modified by different short fibers have been reported. Among them, geopolymer composites modified by carbon and SiC fibers exhibit broader application prospects than those modified by organic fibers, because carbon and SiC fibers can not only significantly improve the mechanical properties of the geopolymer matrix, but also optimize their thermal properties such as thermal shock resistance and thermal stability. However, at high temperatures, a large number of microcracks will appear in carbon and SiC fiber-reinforced geopolymer composites due to the asynchronous shrinkage of the geopolymer matrix and fibers. This leads to the decrease of material reliability. Exploring suitable strategies to optimize the thermal matching and interface compatibility between the matrix and fibers will be an important topic for further study. (v) Continuous fiber-reinforced geopolymer matrix composites The current preparation process of continuous fiber-reinforced geopolymer composites is hand lay-up, whose subsequent curing process involves vacuum curing, or laminated curing. Davidovits et al. [78] comparatively studied the mechanical properties of geopolymers reinforced by three different fibers (i.e., E-glass, SiC fiber, and carbon fiber). Figure 2.11 shows the comparison of the flexural strength of these fiber-reinforced geopolymer composites at different temperatures.

Fig. 2.11 Three points flexural strength of geopolymers reinforced with three different fibers (i.e., E-glass, SiC fiber, and carbon fiber) as a function of temperature. a F, M-PSDS, b K-PSDS, with permission from [79]

24

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.12 Mechanical performances of unidirectional carbon fiber-reinforced geopolymer composites. a Mechanical properties, b load–displacements curves, with permission from [46]

Lyon et al. [80] further reported that E-glass fiber can interact with the geopolymer matrix, which significantly reduces the strength of E-glass fiber. He et al. [46] prepared unidirectional continuous reinforced carbon fiber geopolymer composites with different Si/Al ratios and employed ultrasonic-assisted impregnation technique to make the fibers uniformly dispersed in the matrix slurry. The results showed that the flexural strength and elastic modulus of the composites increase with the increasing Si/Al ratio in the range of 2–4 (Fig. 2.12a), and all composites exhibit non-catastrophic fracture characteristics (Fig. 2.12b). Zheng et al. [81] used unidirectional SiC fibers as the reinforcement to prepare geopolymer composites, and found that the introduction of SiC fibers improves the mechanical properties of the composites. The mechanical properties of the composites reach the maximum when the SiC content is 20 vol%. Although the continuous fiber can strengthen and toughen the geopolymer matrix, current strategies for preparing continuous fiber-reinforced geopolymer composites are still far beyond being enough. Further studies are needed to produce continuous fiber-reinforced geopolymer composites with more complex shapes and structures. This requires advanced controls over the rheology and stability of the geopolymer slurry. In addition, the structure and performance evolution of the geopolymer composites under their service environments (e.g., high temperature, oxidation, corrosion, etc.) remain elusive. Systematic studies are necessary to reveal their evolution mechanisms under different environments.

2.3 Applications of Geopolymers and Their Matrix Composites

25

2.3 Applications of Geopolymers and Their Matrix Composites 2.3.1 Nuclear Waste Immobilization With the rapid development of nuclear energy, safe disposal of radioactive nuclear wastes generated in nuclear fuel cycles has become an urgent need. In recent years, immobilization has been regarded as an effective strategy to storage radioactive nuclear wastes. For the immobilization of intermediate-level nuclear waste (ILW) and low-level nuclear waste (LLW), traditional immobilization method usually involves cement immobilization [82]. However, properties such as high porosity, poor thermal stability, and corrosion resistance limit cement material as ideal long-term immobilization matrix [83] Recently, geopolymer has been considered an alternative for trapping radioactive nuclear wastes due to their high mechanical strength, stability, and immobilization capability. Qin et al. [84] reported that the leaching rate of 133Cs+ immobilized in fly ash-based geopolymer is only 1/15 of that immobilized in cement (Fig. 2.13a), indicating geopolymer exhibits much better immobilization performance than traditional cement materials. Our group [85] further studied the leaching kinetics of Sr+ and Cs2+ in metakaolin-based geopolymer under different temperatures and environments. The results showed that the leaching behaviors of Sr+ and Cs2+ are largely limited in metakaolin-based geopolymer compared to Portland cement during the long-term leaching tests (Fig. 2.13b), highlighting metakaolinbased geopolymer as promising salt-tolerant matrix for the immobilization of nuclear wastes.

Fig. 2.13 Leaching behaviors of simulated radioactive elements in cement and geopolymer. a Cs+ in fly ash-based geopolymer and cement, b Sr2+ in Na- and K-containing metakaolin-based geopolymer and Portland cement, c Cs+ in Na- and K-containing metakaolin-based geopolymer and Portland cement, with permission from [84, 85]

26

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

2.3.2 Precursors for Advanced Ceramics and Ceramic Matrix Composites Thanks to the compositional tunability of geopolymers and their composites, they can serve as precursors to fabricate advanced ceramics structures through the ionexchange method and high-temperature treatments. He et al. [58] obtained leucite ceramics and their matrix composites with different shapes (Fig. 2.14) by hightemperature treating geopolymers and carbon fiber-reinforced geopolymer matrix composites. They found that the thermal expansion coefficient of the resulting leucite ceramics can be tuned in the range of 7.0–5.4 × 10–6 /K, and the resulting leucite ceramics exhibit excellent mechanical properties and oxidation resistance in the temperature range of room temperature to 1200 °C. This indicates that geopolymer precursors provide a new way for green and low-cost synthesis of high-performance ceramic matrix composites. Furthermore, He et al. [86] ion-exchanged alkali metal cations in geopolymer (Na+ , K+ , and Cs+ ) with Ba2+ to get celsian precursors and obtained hexagonal celsian upon high-temperature treatment. Further replacing partial Ba2+ with Sr2+ into the ion-exchanged solution facilitates the formation of monoclinic celsian. Kriven et al. [87] first synthesized graphite/geopolymer composites and obtained silicon carbide ceramics by heating them to 1500–700 °C in an argon atmosphere. The same treatment temperature can also yield high melting point ceramics consisting of aluminum nitride and silicon nitride. Wang et al. [88] performed high-temperature treatments on Na-containing geopolymer, and found that Na-containing geopolymer can convert into porous mullite with a pore size of 10 μm at 1500 °C.

2.3.3 Heat-Resistant and Fire-Retardant Properties and Applications Geopolymers and their composites also show promising applications in the fields of offshore oil mining platforms, aircraft linings, defense, aerospace, and transportation [82, 83]. For instance, geopolymers reinforced by high-performance fibers can be prepared at room temperature and possess low thermal conductivity. This renders them as potential coating materials for the exhaust holes of internal combustion engines [89] and other refractory materials.

2.3.4 Next-Generation Building Materials Geopolymer has been considered as one of the candidates to develop next-generation building materials due to its high hardness and fast curing speed. Cement-like geopolymer has been successfully used for the construction of the University of

2.3 Applications of Geopolymers and Their Matrix Composites

27

Fig. 2.14 Leucite ceramics directly derived from geopolymer. a Plate-shaped sample, b bulletshaped sample, with permission from [86]

28

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

Fig. 2.15 The University of Queensland’s Global Change Institute—the first public building made from structural geopolymer materials, with permission from [90]

Queensland in Australia (Fig. 2.15) and the Wellcamp Airport in Brisbane. In addition, geopolymer shows excellent resistance to high temperature and low temperature cycles, good vacuum stability, low water consumption, and good mechanical properties, meeting all the specific requirements of lunar building materials [90].

2.3.5 Adsorption of Heavy Metal Ions and Applications in Waste Water Management The three-dimensional network of geopolymer holds great potentials in chemical adsorption, and it can be used to solidify and remove the heavy metal ions such as Cu2+ , Pb2+ , and Cd2+ contained in the sewage. In addition, geopolymer has also been used to prepare inorganic films, which can not only effectively filter heavy metals (such as Ni2+ ) in sewage, but also eliminate small molecular pollutants [91]. Such thin film prepared by geopolymer is featured with low cost, facile processing, exhibiting promising application prospects.

2.3.6 Biological Antibacterial Properties and Applications Since the alkali metal cations in geopolymer are weakly bound by the electrostatic attraction originating from the amorphous network, they can be easily replaced by other metal cations to get biological activity. Taking advantage of this feature, Shinobu et al. [92, 93] immersed the cured geopolymer in a copper chloride solution to partially replace K+ in the geopolymer matrix. They found that after 24 h of immersion, the displacement layer reached about 900 μm. The ion-exchange processes include: (i) form relatively dark green akamiite crystals on the surface and at a depth of 150–300 μm below the surface of the sample; (ii) copper ions

2.3 Applications of Geopolymers and Their Matrix Composites

29

diffused into the geopolymer matrix to replace the potassium ions. The resulting Cu2+ -exchanged geopolymer exhibits significant antibacterial activity.

2.4 Outlook With the rapid development of modern economy and society, global energy and environmental crises are becoming more and more serious. Energy storage, consumption reduction, and low carbon emission have become the inevitable trends of sustainable social development. Such requirements provide a good development opportunity for developing geopolymers and their composites from the following aspects: (i)

(ii)

(iii)

(iv)

(v)

Studies on revealing the mechanisms underlying geopolymerization and thermal evolution of geopolymers and their composites from a microscopic picture are needed. Further investigations regarding the effects of the type and content of alkali metal ions and reinforcements on the geopolymerization mechanism, thermal evolution, and crystallization kinetics are required, in order to provide theoretical basis and experimental guidance for preparing high-performance ceramics and their composites. Taking advantage of their good rheology and flexible composition, geopolymers can act as precursors for preparing high-performance ceramics with complex shapes and versatility. Systematic studies on tuning the crystalline phase, mechanical properties, thermal properties, electrical properties, and wave-absorbing properties are needed to achieve this goal. Further in-depth studies on the damage mechanism of geopolymers and their composites under service environments such as high temperature, air flow, particle flow, and loading will shed light on the macro and micro evolution of the materials. This helps to control the failure mode of the materials and further optimize their mechanical properties. Moreover, the establishment of performance database of geopolymers and their composites under different service conditions is also beneficial for designing the high-performance engine heat shield, large-area heat protection of super aircraft, and cabin partition. Further investigations on the effects of chemical composition, storage medium, and structure of geopolymer on their immobilization performance of radioactive ions and heavy metal ions are necessary for developing salt-tolerant immobilization matrices. Exploring printable geopolymers and their matrix composites by tuning their chemical composition, additives, and curing process are important for promoting 3D printing geopolymer-based components, which can be engineered with fine structures and high performance.

30

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

References 1. Z. Abdollahnejad, F. Pacheco-Torgal, T. Félix, W. Tahri, J. Barroso Aguiar, Mix design, properties and cost analysis of fly ash-based geopolymer foam. Constr. Build. Mater. 80, 18–30, (2015) 2. A.R. Sakulich, Reinforced geopolymer composites for enhanced material greenness and durability. Sustain. Cities Soc. 1, 195–210 (2011) 3. J. Zheng, W. Qin, The development of geopolymeric materials. New Build. Mater. 04, 11–12 (2002) 4. Y. Ren, H. Ma, G. Wang, W. Feng, Q. Ding, Experimental research on preparation of mineral polymeric materials from gold mine tailings. Geoscience 02, 171–175 (2003) 5. Y. Ren, M. Hongwen, G. Wang, W. Feng, Q. Ding, Preparation of gold tailings mineral polymer material and its influencing factors. Rock Miner. Anal. 02, 103–108 (2003) 6. V. Vaou, D. Panias, Thermal insulating foamy geopolymers from perlite. Miner. Eng. 23, 1146–1151 (2010) 7. S.T. Erdogan, Properties of ground perlite geopolymer mortars. J. Mater. Civ. Eng. 27, 04014210 (2015) 8. S.N.M. Hairi, G.N.L. Jameson, J.J. Rogers, K.J.D. MacKenzie, Synthesis and properties of inorganic polymers (geopolymers) derived from Bayer process residue (red mud) and bauxite. J. Mater. Sci. 50, 7713–7724 (2015) 9. K. Kaya, S. Soyer-Uzun, Evolution of structural characteristics and compressive strength in red mud–metakaolin based geopolymer systems. Ceram. Int. 42, 7406–7413 (2016) 10. J. Davidovits, Geopolymers and geopolymeric materials. J. Therm. Anal. 35, 429–441 (1989) 11. P.D. Silva, K. Sagoe-Crenstil, V. Sirivivatnanon, Kinetics of geopolymerization: role of Al2 O3 and SiO2 . Cem. Concr. Res. 37, 512–518 (2007) 12. D. Khale, R. Chaudhary, Mechanism of geopolymerization and factors influencing its development: a review. J. Mater. Sci. 42, 729–746 (2007) 13. X. Yao, Z. Zhang, H. Zhu, Y. Chen, Geopolymerization process of alkali–metakaolinite characterized by isothermal calorimetry. Thermochim. Acta 493, 49–54 (2009) 14. M.S. Muñiz-Villarreal, A. Manzano-Ramírez, S. Sampieri-Bulbarela, J.R. Gasca-Tirado, J.L. Reyes-Araiza, J.C. Rubio-Ávalos, J.J. Pérez-Bueno, L.M. Apatiga, A. Zaldivar-Cadena, V. Amigó-Borrás, The effect of temperature on the geopolymerization process of a metakaolinbased geopolymer. Mater. Lett. 65, 995–998 (2011) 15. A. Kumar, S. Kumar, Development of paving blocks from synergistic use of red mud and fly ash using geopolymerization. Constr. Build. Mater. 38, 865–871 (2013) 16. B.-H. Mo, H. Zhu, X.-M. Cui, Y. He, S.-Y. Gong, Effect of curing temperature on geopolymerization of metakaolin-based geopolymers. Appl. Clay Sci. 99, 144–148 (2014) 17. S.K. Nath, S. Mukherjee, S. Maitra, S. Kumar, Ambient and elevated temperature geopolymerization behaviour of class F fly ash. Trans. Indian Ceram. Soc. 73, 126–132 (2014) 18. P. Duxson, S.W. Mallicoat, G.C. Lukey, W.M. Kriven, J.S.J. van Deventer, The effect of alkali and Si/Al ratio on the development of mechanical properties of metakaolin-based geopolymers. Colloids Surf. A 292, 8–20 (2007) 19. A. Autef, E. Joussein, G. Gasgnier, S. Rossignol, Role of the silica source on the geopolymerization rate. J. Non-Cryst. Solids 358, 2886–2893 (2012) 20. A. Autef, E. Joussein, G. Gasgnier, S. Rossignol, Role of the silica source on the geopolymerization rate: a thermal analysis study. J. Non-Cryst. Solids 366, 13–21 (2013) 21. Z. Yunsheng, S. Wei, W. Lin, K. Zheng, J. Sha, S. Zheng, In-situ tracking of the hydration process of the K-PSDS geopolymer concrete interface zone with environmental scanning electron microscope. J. Chin. Ceram. Soc. 31, 806–810 (2003) 22. Z. Yunsheng, S. Wei, J. Sha, W. Sun, K. Zheng, S. Liu, Preparation, characteristics and mechanism of fly ash geopolymer concrete. J. Build. Mater. 03, 237–242 (2003) 23. L. Weng, S. Song, Novel geopolymer cementious materials. Mater. Rep. 02, 67–68+80 (2005) 24. Y. Nie, H. Ma, J. Yang, Y. Shu, R. Li, F. Gao, Study on the mechanism of geopolymerization in the curing process of geopolymer. Geoscience 02, 340–346 (2006)

References

31

25. M. Wang, Geopolymerization Mechanism of Geopolymer and the Microstructure and Properties of Their Composites Containing Floating Beads (Harbin Institute of Technology, 2011) 26. S. Yan, Geopolymerization and Ceramicization Mechanism of Graphene Oxide Reinforced Geopolymer (Harbin Institute of Technology, 2016) 27. S. Yan, P. He, D. Jia, X. Duan, Z. Yang, S. Wang, Y. Zhou, Effects of graphene oxide on the geopolymerization mechanism determined by quenching the reaction at intermediate states. RSC Adv. 7, 13498–13508 (2017) 28. J. Yuan, P. He, D. Jia, S. Fu, Y. Zhang, X. Liu, D. Cai, Z. Yang, X. Duan, S. Wang, Y. Zhou, In situ processing of MWCNTs/leucite composites through geopolymer precursor. J. Eur. Ceram. Soc. 37, 2219–2226 (2017) 29. S. Yan, P. He, D. Jia, X. Duan, Z. Yang, S. Wang, Y. Zhou, In-situ preparation of fully stabilized graphene/cubic-leucite composite through graphene oxide/geopolymer. Mater. Des. 101, 301– 308 (2016) 30. S. Yan, P. He, D. Jia, X. Duan, Z. Yang, S. Wang, Y. Zhou, In situ processing of graphene/leucite nanocomposite through graphene oxide/geopolymer. J. Am. Ceram. Soc. 99, 1164–1173 (2016) 31. S. Yan, P. He, D. Jia, X. Duan, Z. Yang, S. Wang, Y. Zhou, Crystallization kinetics and microstructure evolution of reduced graphene oxide/geopolymer composites. J. Eur. Ceram. Soc. 36, 2601–2609 (2016) 32. S. Yan, P. He, D. Jia, Z. Yang, X. Duan, S. Wang, Y. Zhou, In situ fabrication and characterization of graphene/geopolymer composites. Ceram. Int. 41, 11242–11250 (2015) 33. S. Yan, P. He, D. Jia, Z. Yang, X. Duan, S. Wang, Y. Zhou, Effect of reduced graphene oxide content on the microstructure and mechanical properties of graphene–geopolymer nanocomposites. Ceram. Int. 42, 752–758 (2016) 34. S. Yan, P. He, D. Jia, Z. Yang, X. Duan, S. Wang, Y. Zhou, Effect of fiber content on the microstructure and mechanical properties of carbon fiber felt reinforced geopolymer composites. Ceram. Int. 42, 7837–7843 (2016) 35. S. Yan, P. He, D. Jia, Z. Yang, X. Duan, S. Wang, Y. Zhou, Effects of treatment temperature on the reduction of GO under alkaline solution during the preparation of graphene/geopolymer composites. Ceram. Int. 42, 18181–18188 (2016) 36. T.S. Lin, D.C. Jia, P.G. He, M.R. Wang, Thermo-mechanical and microstructural characterization of geopolymers with α-Al2 O3 particle filler. Int. J. Thermophys. 30, 1568 (2009) 37. T. Lin, D. Jia, Mechanical properties and fracture behavior of electroless NI-plated short carbon fiber reinforced geopolymer matrix composites, in Fifth International Conference on Advanced Materials and Processing ICAMP-5 Harbin 2009, pp. 1371–1376 38. J. Yuan, P. He, D. Jia, S. Yan, D. Cai, L. Xu, Z. Yang, X. Duan, S. Wang, Y. Zhou, SiC fiber reinforced geopolymer composites, Part 1: Short SiC fiber. Ceram. Int. 42, 5345–5352 (2016) 39. Y. Jia, Study on Preparation and Performance of Na-Fly Ash Geopolymer (China University of Mining and Technology, Beijing, 2009) 40. T. Lin, D. Jia, P. He, M. Wang, D. Liang, Effects of fiber length on mechanical properties and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 497, 181–185 (2008) 41. T. Lin, D. Jia, M. Wang, P. He, D. Liang, Effects of fibre content on mechanical properties and fracture behaviour of short carbon fibre reinforced geopolymer matrix composites. Bull. Mater. Sci. 32, 77–81 (2009) 42. T. Lin, D. Jia, P. He, M. Wang, In situ crack growth observation and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 527, 2404–2407 (2010) 43. T. Lin, D. Jia, P. He, M. Wang, Thermal-mechanical properties of short carbon fiber reinforced geopolymer matrix composites subjected to thermal load. J. Centr. South Univ. Technol. 16, 881–886 (2009) 44. P. He, D. Jia, T. Lin, M. Wang, Y. Zhou, Effects of high-temperature heat treatment on the mechanical properties of unidirectional carbon fiber reinforced geopolymer composites. Ceram. Int. 36, 1447–1453 (2010)

32

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

45. P. He, D. Jia, M. Wang, Y. Zhou, Improvement of high-temperature mechanical properties of heat treated Cf/geopolymer composites by Sol-SiO2 impregnation. J. Eur. Ceram. Soc. 30, 3053–3061 (2010) 46. P. He, D. Jia, M. Wang, Y. Zhou, Preparation and mechanical properties of carbon fiber reinforced geopolymer composites. Rare Metal Mater. Eng. 40, 247–251 (2011) 47. P. He, D. Jia, B. Zheng, S. Yan, J. Yuan, Z. Yang, X. Duan, J. Xu, P. Wang, Y. Zhou, SiC fiber reinforced geopolymer composites, Part 2: Continuous SiC fiber. Ceram. Int. 42, 12239–12245 (2016) 48. J. Yuan, P. He, P. Zhang, D. Jia, D. Cai, Z. Yang, X. Duan, S. Wang, Y. Zhou, Novel geopolymer based composites reinforced with stainless steel mesh and chromium powder. Constr. Build. Mater. 150, 89–94 (2017) 49. M. Xia, J. Sanjayan, Method of formulating geopolymer for 3D printing for construction applications. Mater. Des. 110, 382–390 (2016) 50. J. Zhong, G.-X. Zhou, P.-G. He, Z.-H. Yang, D.-C. Jia, 3D printing strong and conductive geo-polymer nanocomposite structures modified by graphene oxide. Carbon 117, 421–426 (2017) 51. Z. Yunsheng, S. Wei, L. Zongjin, Z. Xiangming, Eddie, C. Chungkong, Impact properties of geopolymer based extrudates incorporated with fly ash and PVA short fiber. Constr. Build. Mater. 22, 370-383 (2008) 52. M.R. Wang, D.C. Jia, P.G. He, Y. Zhou, Influence of calcination temperature of kaolin on the structure and properties of final geopolymer. Mater. Lett. 64, 2551–2554 (2010) 53. R.A. Fletcher, K.J.D. MacKenzie, C.L. Nicholson, S. Shimada, The composition range of aluminosilicate geopolymers. J. Eur. Ceram. Soc. 25, 1471–1477 (2005) 54. P. Duxson, J.L. Provis, G.C. Lukey, S.W. Mallicoat, W.M. Kriven, J.S.J. van Deventer, Understanding the relationship between geopolymer composition, microstructure and mechanical properties. Colloids Surf. A 269, 47–58 (2005) 55. W.M. Kriven, J. Bell, M. Gordon, Geopolymer refractories for the glass manufacturing industry, in 64th Conference on Glass Problems: Ceramic Engineering and Science Proceedings (2004), pp. 57–80 56. J. Yuan, P. He, D. Jia, C. Yang, Y. zhang, S. Yan, Z. Yang, X. Duan, S. Wang, Y. Zhou, Effect of curing temperature and SiO2 /K2 O molar ratio on the performance of metakaolin-based geopolymers. Ceram. Int. 42, 16184–16190 (2016) 57. P. He, M. Wang, S. Fu, D. Jia, S. Yan, J. Yuan, J. Xu, P. Wang, Y. Zhou, Effects of Si/Al ratio on the structure and properties of metakaolin based geopolymer. Ceram. Int. 42, 14416–14422 (2016) 58. P. He, D. Jia, S. Wang, Microstructure and integrity of leucite ceramic derived from potassiumbased geopolymer precursor. J. Eur. Ceram. Soc. 33, 689–698 (2013) 59. P. Gao, Research on Mechanical Properties of Metal/K-PSS Geopolymer Composites (Harbin Institute of Technology, 2008) 60. M. Wang, D. Jia, P. Jia, Y. Zhou, Influence of size of fly ash cenosphere on the microstructure and property of 35% FAC/geopolymer matrix composite. Rare Metal Mater. Eng. 40, 257–261 (2011) 61. M.-R. Wang, D.-C. Jia, P.-G. He, Y. Zhou, Microstructural and mechanical characterization of fly ash cenosphere/metakaolin-based geopolymeric composites. Ceram. Int. 37, 1661–1666 (2011) 62. D. Ren, C. Yan, P. Duan, Z. Zhang, L. Li, Z. Yan, Durability performances of wollastonite, tremolite and basalt fiber-reinforced metakaolin geopolymer composites under sulfate and chloride attack. Constr. Build. Mater. 134, 56–66 (2017) 63. K.J.D. MacKenzie, M.J. Bolton, Electrical and mechanical properties of aluminosilicate inorganic polymer composites with carbon nanotubes. J. Mater. Sci. 44, 2851–2857 (2009) 64. K.J.D. MacKenzie, Inorganic polymers (geopolymers) as advanced materials, in Mechanical Properties and Performance of Engineering Ceramics and Composites, ed. by W.M.K.D. Singh (2010), pp. 251–261

References

33

65. S.M. Abbasi, H. Ahmadi, G. Khalaj, B. Ghasemi, Microstructure and mechanical properties of a metakaolinite-based geopolymer nanocomposite reinforced with carbon nanotubes. Ceram. Int. 42, 15171–15176 (2016) 66. S. Bi, M. Liu, J. Shen, X.M. Hu, L. Zhang, Ultrahigh self-sensing performance of geopolymer nanocomposites via unique interface engineering. ACS Appl. Mater. Interfaces 9, 12851–12858 (2017) 67. H.M. Khater, H.A. Abd el Gawaad, Characterization of alkali activated geopolymer mortar doped with MWCNT. Constr. Build. Mater. 102, 329–337 (2016) 68. M. Saafi, L. Tang, J. Fung, M. Rahman, F. Sillars, J. Liggat, X. Zhou, Graphene/fly ash geopolymeric composites as self-sensing structural materials. Smart Mater. Struct. 23, 065006 (2014) 69. S.A. Bernal, J. Bejarano, C. Garzón, R. Mejía de Gutiérrez, S. Delvasto, E.D. Rodríguez, Performance of refractory aluminosilicate particle/fiber-reinforced geopolymer composites. Compos. B Eng. 43, 1919–1928 (2012) 70. P. He, Research on Cf/Geopolymer and Its Conversion to Ceramic Matrix Composite (Harbin Institute of Technology, 2011) 71. T. Lin, D. Liang, D. Jia, Mechanical properties and fracture behaviour of Csf(α-Al2 O3 p) reinforced inorganic polymer matrix composites. J. Synth. Cryst. 38, 283–287 (2009) 72. X.-L. Song, X.-M. Cui, K.-S. Lin, G.-J. Zheng, Y. He, Hot-pressure forming process of PVC/geopolymer composite materials. Appl. Clay Sci. 71, 32–36 (2013) 73. Y. Zhang, P. He, J. Yuan, C. Yang, D. Jia, Y. Zhou, Effects of graphite on the mechanical and microwave absorption properties of geopolymer based composites. Ceram. Int. 43, 2325–2332 (2017) 74. S. Yan, P. He, D. Jia, J. Wang, X. Duan, Z. Yang, S. Wang, Y. Zhou, Effects of high-temperature heat treatment on the microstructure and mechanical performance of hybrid Cf-SiCf-(Al2 O3 p) reinforced geopolymer composites. Compos. B Eng. 114, 289–298 (2017) 75. S. Yan, P. He, Y. Zhang, D. Jia, J. Wang, X. Duan, Z. Yang, Y. Zhou, Preparation and in-situ high-temperature mechanical properties of Cf-SiCf reinforced geopolymer composites. Ceram. Int. 43, 549–555 (2017) 76. T.S. Ng, A. Amin, S.J. Foster, The behaviour of steel-fibre-reinforced geopolymer concrete beams in shear. Mag. Concr. Res. 65, 308–318 (2013) 77. Z. Yunsheng, S. Wei, L. Zongjin, Impact behavior and microstructural characteristics of PVA fiber reinforced fly ash-geopolymer boards prepared by extrusion technique. J. Mater. Sci. 41, 2787–2794 (2006) 78. J. Davidovits, M. Davidovics, Geopolymer: ultra-high temperature tooling material for the manufacture of advanced composites. SAMPE (1991) 79. J. Davidovits, Geopolymers: inorganic polymeric new materials. J. Therm. Anal. 37, 1633–1656 (1991) 80. A.J. Foden, P. Balaguru, R.E. Lyon, Mechanical Properties and Fire Response of Geopolymer Structural Composites (Society for the Advancement of Material and Process Engineering, Covina, CA (United States), United States, 1996) 81. B. Zheng, Unidirectional Continuous SiCf Reinforced Geopolymer Composites and Their Mechanical Properties (Harbin Institute of Technology, 2013) 82. S.B. Eskander, S.M. Abdel Aziz, H. El-Didamony, M.I. Sayed, Immobilization of low and intermediate level of organic radioactive wastes in cement matrices. J. Hazard. Mater. 190, 969–979 (2011) 83. S. Paria, P.K. Yuet, Solidification—stabilization of organic and inorganic contaminants using Portland cement. ChemInform 38 (2006) 84. Q. Li, Z. Sun, D. Tao, Y. Xu, P. Li, H. Cui, J. Zhai, Immobilization of simulated radionuclide 133Cs+ by fly ash-based geopolymer. J. Hazard. Mater. 262, 325–331 (2013) 85. P. He, J. Cui, M. Wang, S. Fu, H. Yang, C. Sun, X. Duan, Z. Yang, D. Jia, Y. Zhou, Interplay between storage temperature, medium and leaching kinetics of hazardous wastes in metakaolinbased geopolymer. J. Hazard. Mater. 384, 121377 (2020)

34

2 Geopolymers and Their Matrix Composites: A State-of-the-Art Review

86. P. He, S. Fu, J. Yuan, J. Rao, J. Xu, P. Wang, D. Jia, Celsian formation from barium-exchanged geopolymer precursor: thermal evolution. J. Eur. Ceram. Soc. 37, 4179–4185 (2017) 87. C. Bagci, G.P. Kutyla, K.C. Seymour, W.M. Kriven, Synthesis and characterization of silicon carbide powders converted from metakaolin-based geopolymer. J. Am. Ceram. Soc. 99, 2521– 2530 (2016) 88. H. Wang, H. Li, Y. Wang, F. Yan, Preparation of macroporous ceramic from metakaolinite-based geopolymer by calcination. Ceram. Int. 41, 11177–11183 (2015) 89. Z. Fang, C. Wu, T. Zhu, Development of high temperature resistant geopolymer cementing materials. J. Therm. Sci. Technol. 02, 178–185 (2007) 90. K.-T. Wang, Q. Tang, X.-M. Cui, Y. He, L.-P. Liu, Development of near-zero water consumption cement materials via the geopolymerization of tektites and its implication for lunar construction. Sci. Rep. 6, 29659 (2016) 91. Y. Ge, Y. Yuan, K. Wang, Y. He, X. Cui, Preparation of geopolymer-based inorganic membrane for removing Ni2+ from wastewater. J. Hazard. Mater. 299, 711–718 (2015) 92. H. Takeda, S. Hashimoto, S. Honda, Y. Iwamoto, The coloring of geopolymers by the addition of copper compounds. Ceram. Int. 40, 6503–6507 (2014) 93. S. Hashimoto, T. Machino, H. Takeda, Y. Daiko, S. Honda, Y. Iwamoto, Antimicrobial activity of geopolymers ion-exchanged with copper ions. Ceram. Int. 41, 13788–13792 (2015)

Chapter 3

Geopolymerization Mechanism of Geopolymers

Abstract Geopolymer is the reaction product of active aluminosilicate powders in alkaline solution, and the typical aluminosilicate powders include metakaolin, fly ash, slag, and so on. Metakaolin powders were commonly used for the study of geopolymerization mechanisms of geopolymer, due to its comparatively pure composition. In this chapter, transformation mechanism from kaolin to metakaolin was first studied, and then geopolymerization mechanism of natural metakaolin and synthesized metakaolin-based geopolymer were reported.

3.1 Mechanism of Transition from Kaolin to Metakaolin Metakaolin is an intermediate phase of kaolin after calcination, which can be directly used as a raw material and reactant for the synthesis of cement and geopolymer materials [1–4]. The mechanism of geopolymerization is studied by synthesis of metakaolin as raw material, which is helpful to reduce the influence of mineral and other impurities in natural kaolin on the geopolymerization. Therefore, how to make kaolin into metakaolin is the first important problem to be understood.

3.1.1 Formation Process and Chemical Activity of Metakaolin (1) The formation process of metakaolin Metakaolin is a metastable intermediate phase after calcination of kaolin. The basic structural unit of kaolin is composed of the tetrahedral [SiO4 ]4− layer and octahedral [AlO2 (OH)4 ]5− layer, which is connected by a common structure. The structure units are directly connected by hydrogen bonds, in which [AlO2 (OH)4 ]5− octahedron is composed of 1 unit of Al3+ , 2 units of O2− , and 4 units of OH− . Al3+ is located in the pores of the octahedron, O2− is located in the octahedron as a bridging oxygen connecting the tetrahedral [SiO4 ]4− layer and octahedron [AlO2 (OH)4 ]5− [1–5]. © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_3

35

36

3 Geopolymerization Mechanism of Geopolymers

Fig. 3.1 Structural models of kaolin (a) and metakaolin (b)

When kaolin is heated to above 450–550 °C, dehydration reaction can occur. The crystal structure of kaolin is shown in Fig. 3.1a. First, outside of the octahedral surface hydroxyl groups with weak binding force is taken off. Second, hydroxy of octahedral [AlO2 (OH)4 ]5− structure units are taken off and five coordinates are formed. Then, the dehydroxylation reaction occurs and the four coordinates are formed. The structural unit is transformed into [AlO4 ]− tetrahedron and the metakaolin is formed, as shown in Fig. 3.1 [5]. (2) Chemical activity of metakaolin The chemical activity of metakaolin refers to its ability to react with alkaline solutions (e.g., calcium hydroxide, sodium hydroxide, and potassium hydroxide) [6]. According to the principle of atomic compacting, if the packing density of the ball is higher, energy in the system is lower and the structure is more stable. Metakaolin is an amorphous material which is composed of four-, five-, and sixcoordinated Al atoms and four-coordinated Si atoms. In the structure of metakaolin, compared with the five- and six-coordinated Al atoms, the four-coordinated Al atom has a lower accumulation density and higher internal energy, resulting in more instability. Therefore, the chemical activity of metakaolin can be characterized by measuring the content of metastable four-coordinated Al atoms. Currently, metakaolin with different chemical activities has been obtained by various methods, including calcination, microwave treatment, and high-energy ball milling. For example, calcination temperature and time affect the generation rate and content of four-coordinated Al atoms in metakaolin, which is regarded as the most effective, economical, and practical method [7–11]. However, if the calcination temperature is too low, incomplete dehydroxylation will be caused, which can slow down the formation rate and reduce the content of four-coordinated Al atoms. Moreover, when the calcination temperature is too high, metakaolin is crystallized to be more stable mullite or alumina. The initial four-coordinated Al atom can transform into six-coordinated Al atom, causing the “death burn” of metakaolin. Similarly, calcination time also affects the formation of four-coordinated Al atom in metakaolin

3.1 Mechanism of Transition from Kaolin to Metakaolin

37

from the above two aspects. So, only with the appropriate calcination temperature and time, the highly active metakaolin can be obtained. The effects of calcination temperature and time are discussed in detail as follows.

3.1.2 Effect of Calcination Temperature on Thermal Transformation of Kaolin The thermal transformation of kaolinite is affected by many factors, such as order degree of kaolinite structure, calcination temperature, time, and particle size [12]. This section focuses on the effects of calcination temperature and holding time of kaolin in the calcination process.

3.1.2.1

Effect of Calcination Temperature on Phase of Calcined Products of Kaolin

Two groups of experiments are conducted on kaolin (the specific scheme is shown in Table 3.1): One group of kaolin is calcined at different temperatures with the same holding time, that is, at 500, 600, 700, 800, and 900 °C for 4 h. Calcination products are recorded as 500MK, 600MK, 700MK, 800MK, and 900MK, respectively. The other group is calcined at the same temperature with different holding time, that is, at 900 °C for 1, 4, 6, and 8 h. Calcination products are recorded as 900MK-1, 900MK-4, 900MK-6, and 900MK-8. The XRD patterns of the products obtained in the first group and raw natural kaolin are shown in Fig. 3.2. As shown in Fig. 3.2a, the phases of natural kaolinite are relatively complex. In addition to the main kaolinite, it also contains a small amount of accompanying minerals such as α-quartz and illite. As shown in Fig. 3.2a–f, the α-quartz phase remains unchanged before and after calcination. Zibouche et al. [14] Table 3.1 Two groups of calcination test for kaolin [13]

Test

Samples

Calcination temperature (°C)

Calcining holding time (h)

1

500MK

500

4

600MK

600

700MK

700

800MK

800

2

900MK

900

900MK-1

900

1

900MK-4

4

900MK-6

6

900MK-8

8

38

3 Geopolymerization Mechanism of Geopolymers kaolinite -quartz -aluminum oxide Illite

(f)

Relative intensity, a.u.

(e) (d) (c)

region 1

region 2

(b)

(a)

10

20

30

40

50

60

70

2theta, degree Fig. 3.2 XRD patterns of natural kaolin and products calcined at different calcination temperatures: a natural kaolin, b 500MK, c 600MK, d 700MK, e 800MK, and f 900MK [13]

reported that α-quartz in metakaolin did not participate in the polymerization of geopolymer. It can be seen from the comparison between Fig. 3.2a, b that the phase of kaolin (500MK) does not change after calcination at 500 °C. In Fig. 3.2c, the diffraction pattern of sample 600MK shows the typical peak of amorphous phase at 2θ = 23.6°, indicating that kaolin transforms from crystalline to amorphous when calcined at 600 °C. With the increase of calcination temperature, the calcination products of 700MK, 800MK, and 900MK are all amorphous, and the difference is not obvious. Gauss fitting is applied to analyze the amorphous width peak of curves in Fig. 3.2c– f, and the results are shown in Table 3.2. When kaolin is transformed into amorphous

3.1 Mechanism of Transition from Kaolin to Metakaolin Table 3.2 Gaussian fitting results of XRD patterns of 600MK–900MK [13]

39

Samples

Width-at-half height (°)

XRD amorphous center (°)

600MK

13.6

23.6

700MK

12.8

23.9

800MK

12.0

24.1

900MK

13.1

23.5

phase, the center of amorphous peak of the calcined products obtained at different calcination temperatures changes. With the increase of calcination temperature below 800 °C, the center of the amorphous peak moves from 23.6° to 24.1°. When the calcination temperature rises to 900 °C, it turns to be 23.5°, the width-at-half height of the peak also shows a similar change: 13.6°, 12.8°, 12.0°, and 13.1°, respectively. This indicates that in the range from 600 to 900 °C, the products after calcination have large difference in the degree of short-range ordering. In addition, in Fig. 3.2f, the diffraction peak of 900MK shows broadening tendency at the diffraction angle 2θ of both 45.61° and 67.45°; the two diffraction humps and a peak at 36.71° correspond to the peak of γ -Al2 O3 . Combined with TEM analysis (3.1.2.3), the precipitation of nanocrystal grain of γ -Al2 O3 in 900MK results in the diffraction peak broadening. Results from Sonuparkak et al. [15] showed that the generation of γ -Al2 O3 was the main reason for the transformation of kaolinite to mullite corresponding to the exotherm peak of 980 °C.

3.1.2.2

Effect of Calcining Temperature on Particle Morphology of Calcining Products of Kaolin

Figure 3.3 shows the particle morphology of natural kaolin and powders calcined at 500, 600, 700, 800, and 900 °C, respectively. As shown in Fig. 3.3a, most of the natural kaolin powder particles are lamellar, and a small amount is stick-shape. Among them, rod grains are corresponding to α-quartz [13, 16] by EDS analysis (results are shown in Fig. 3.4 and Table 3.2), which is consistent with the research results of Frost et al. [17]. Besides, it can be seen from Fig. 3.3b that the particles of 500MK powder are also flaky and rod-like, with little change compared with the original kaolin powder particles. It can be seen from Fig. 3.3c that the powder particles of 600MK are still flaky and rod-like, but the flaky particles increase and agglomerates occur between them. With the increasing of calcination temperature, the flake particles of 700MK, 800MK, and 900MK also agglomerate significantly, and a small amount of α-quartz bar-like particles can still be observed (Fig. 3.3d–f) [13].

40

3 Geopolymerization Mechanism of Geopolymers

Fig. 3.3 SEM micrographics of natural kaolin and its products calcined at various temperatures: a natural kaolin, b 500MK, c 600MK, d 700MK, e 800MK, and f 900MK [13]

3.1.2.3

TEM Microstructure of Calcined Products of Kaolin

Figure 3.4 is the TEM micrographs of natural kaolin and products calcined at 500, 800, and 900 °C, respectively. The kaolin particles have obvious lath-shaped characteristics and are arranged into bundles in an approximate parallel manner (Fig. 3.4a)

3.1 Mechanism of Transition from Kaolin to Metakaolin

41

Fig. 3.4 TEM structure and structural analysis of natural kaolin powder and its products calcined at different temperatures: a natural kaolin, b 500MK, c 800MK, and d 900MK [13]

[13]. After calcined at 500 °C, the products of sample 500MK still have a lamellar structure (Fig. 3.4b), which is thinner than that of kaolinite and almost feathery. However, it is still crystalline. After increased to 800 °C, the feather-like structure of products 800MK becomes rough [16], and amorphization has been realized. Compared with the sample of 800MK, sample 900MK becomes thicker. The featherlike characteristics disappeared. And nano-size γ -Al2 O3 grains with diameter of about 30 nm are precipitated on the amorphous matrix [18]. Generally, the phase and microstructure of kaolin remain unchanged after calcination at 500 °C. In contrast, after being calcined at 800–900 °C, coarsening behavior occurred and amorphous feature is exhibited. Especially after being calcined at 900 °C, it was found that nanocrystalline γ -Al2 O3 is precipitated from the amorphous matrix. Whether the presence of the nanocrystalline has any effects on the chemical activity of metakaolin remains to be further studied.

42

3 Geopolymerization Mechanism of Geopolymers

3.1.2.4

Functional Groups and Valence Bond Structure of Calcined Products of Kaolin

(1) FT-IR analysis Figure 3.5a–f shows the FT-IR spectra of natural kaolin and its products calcined at 500, 600, 700, 800, and 900 °C, respectively. In Fig. 3.5a, b, natural kaolin and 500MK show four O–H bonds located at 3,800–3,600 cm−1 . The three O–H groups on the outer surface are connected with the alumino-oxygen layer and located near 3696, 3670, and 3653 cm−1 , respectively, while the O–H groups on the inner surface of hydrogen bond formed by the connection with oxygen atoms in the silicon–oxygen layer are located near 3620 cm−1 [17, 19–22]. The characteristic peaks of the four Al–OH bonds are located at 937, 914, 797, and 755 cm−1 , respectively. Compared with natural kaolin, the characteristic peak of the O–H bond and the Al–OH bond is at the same location with 500MK, indicating

(f)

803

1095

563

468

Relative transmission

(e) (d) (c)

(b)

755 797

3653 3670 3623

(a)

937 914

3698

3653 3670 3696

1400

539

1200

472

790

755

1036

1114 1036

3600

431

1009

1114

3620

3800

694

697 937 914 1009

1000

431 539 472

800

Wave number, cm-1

600

400

Fig. 3.5 FT-IR spectra of natural kaolin and products calcined at various temperatures: a natural kaolin, b 500MK, c 600MK, d 700MK, e 800MK, and f 900MK [13]

3.1 Mechanism of Transition from Kaolin to Metakaolin

43

that they contain the same molecular vibration functional group of O–H and Al– OH bonds. However, the reduction of the characteristic peak intensity of 500MK indicates that the dehydroxylation of products after calcination at 500 °C due to the reduction of the content of O–H groups leads to the decrease of the characteristic peak intensity of O–H and Al–OH bonds [13]. When the calcination temperature increased to 600 °C, the vibration peaks of O–H and Al–OH bonds of the samples completely disappear (Fig. 3.5c), indicating that the dehydroxylation of kaolinite is over. With the calcination temperature increased to higher temperature, the vibration peaks of O–H and Al–OH groups appear in the 700MK, 800MK, and 900MK (Fig. 3.5d–f). It indicates that there is no dehydroxylation of kaolinite in calcined products at 600–900 °C. In the range of 1200–1000 cm−1 in Fig. 3.5, three Si–O vibrations of natural kaolin are located at 1008, 1032, and 1114 cm−1 , respectively [23, 24]. Compared with natural kaolin, the location of Si–O vibration of 500MK is unchanged, while the peak intensity changes. The Si–O vibration of 600MK moves to a high wave number of 1095 cm−1 , indicating that the Si–O groups of the product changed after calcination at 600 °C. With the temperature continues to rise, the Si–O vibrations of the calcined products 700MK, 800MK, and 900MK are all located at 1095 cm−1 , and the Si–O groups of the products remain stable. In the range of 1000–400 cm−1 in Fig. 3.5, the Al–O, Si–O–Al, and Si–O–Si peaks in hexagonal AlO6 of natural kaolin and 500MK are the same, which are located at 694, 539, and 472 cm−1 , respectively. Compared with natural kaolin and 500MK, the peaks of AlO6 of 600MK, 700MK, 800MK, and 900MK disappeared, the Si–O–Al peak changes to a higher wave number at 563 cm−1 , and the Al–O bond of AlO4 of new four-coordination appears at the 803 cm−1 [13, 25]; Si–O–Si bond moves to 468 cm−1 . This indicates that calcination process at 500 °C does not change the Al–O groups in the products. However, calcined at 600 °C, the Al–O group of the products is changed. When rising to 900 °C, the Al–O groups of the calcination products no longer change [13]. (2)

27 Al

NMR analysis

Figure 3.6 shows the 27 Al NMR spectra of natural kaolinite and its products calcined at different temperatures. The 27 Al NMR spectra of 600MK–900MK are fitted in Table 3.4. From Fig. 3.6a, b, in both natural kaolin and 500MK, Al atomic units are all composed of six-coordinated Al atoms (chemical shift = 2 ppm). This indicates that after calcination at 500 °C, the environment of Al atom in the calcined products remains unchanged, which is consistent with the results in Fig. 3.5. In Fig. 3.6c, the Al atomic unit in 600MK is composed of four-coordinated (60.8 ppm), five-coordinated (=28.2 ppm), and six-coordinated (=1.1 ppm) Al atoms, which are consistent with the reports by Fyfe et al. [26] and Klinowski et al. [27]. The results show that after calcination at 600 °C, the environment of Al atom in the product has been changed. With the increase of calcination temperature, the Al units in 700MK, 800MK, and 900MK samples are also composed of four-, five-, and six-coordinated Al atoms. Besides, it can be inferred from the results in Table 3.3

44

3 Geopolymerization Mechanism of Geopolymers 1.4

(a)

* 150

* 100

50

0

1.4

(b)

-50

-100

* -150 150

* 100

50

(c)

1.1

Fitting curve

(d)

100

50

0

-50

-100

-150 150

Fitting curve

59.9

-100

-150

100

50

0

Fitting curve

-50

-100

-150

Chemical shift, ppm

Chemical shift, ppm (e)

-50

57.6 29.4 1.1

60.8 28.2

150

0

Chemical shift, ppm

Chemical shift, ppm

(f)

59.4

Fitting curve 33.3

26.7 1.5

3.7

150

100

50

0

-50

Chemical shift, ppm

-100

-150 150

100

50

0

-50

-100

-150

Chemical shift, ppm

Fig. 3.6 27 Al NMR spectra of natural kaolin and products calcined at different temperatures: a natural kaolin, b 500MK, c 600MK, d 700MK, e 800MK, and f 900MK [13]

that, in the range of 600–900 °C, with the increase of calcination temperature, the content of four-coordinated Al gradually increases and the content of six-coordinated Al gradually decreases, while the content of five-coordinated Al shows no obvious change. This indicates that the activity of metakaolin gradually increases with the increase of calcination temperature. (3)

29 Si

NMR analysis

Figure 3.7a–d shows the 29 Si NMR spectra of natural kaolin and its products calcined at 500, 800, and 900 °C, respectively. As shown in Fig. 3.7a, b, the chemical shifts of 29 Si of natural kaolin and 500MK are both located at −89.8 ppm, which corresponds to the four-coordinated Si atom whose atomic structure environment is Q3 (0Al)

3.1 Mechanism of Transition from Kaolin to Metakaolin

45

Table 3.3 Results of Gaussian fit for metakaolin calcined at various temperatures [13] Samples

Coordination of Al atom

Chemical shift (ppm)

Relative area

Percentage

600MK

4

~60.8

10.00

30.43 46.97

700MK

800MK

900MK

5

~28.2

15.44

6

~1.4

7.43

22.60

4

~57.6

10.00

34.03 43.18

5

~29.4

12.69

6

~1.1

5.28

17.69

4

~59.9

10.00

42.31 31.48

5

~26.7

7.44

6

~1.5

6.20

26.21

4

~59.4

10.00

49.50

5

~33.3

7.45

36.87

6

~3.7

2.76

13.64

-89.8

(a)

(b)

*

*

* 0

-50

-100

-89.8

-150

-200

*

0

-50

Chemical shift (ppm) -105.9

(c)

0

* -50

-100

-150

Chemical shift, ppm

-150

-200

-106.2

(d)

*

-100

Chemical shift (ppm)

* -200 0

* -50

-100

-150

-200

Chemical shift, ppm

Fig. 3.7 29 Si NMR spectra of natural kaolin and products calcined at various temperatures: a natural kaolin, b 500MK, c 800MK, and d 900MK [13]

46

3 Geopolymerization Mechanism of Geopolymers

type. This indicates that calcination at 500 °C does not change the Si atom type of the product, which is the same as that of natural kaolin [13]. It can be seen from Fig. 3.7c, d that the chemical shifts of 29 Si of 800MK and 900MK are also the same, which are both located at −106 ppm. They are the fourcoordinated Si atom, whose atomic structure environment is Q4 (1Al). This further confirms the FT-IR results in Fig. 3.5: during the calcination range of 600–900 °C, the temperature does not change the Si atom structure environment in the calcination products of kaolin. Compared with 500MK, the coordinate number of Si atoms in 800MK and 900MK is the same. They are four-coordination, but the atoms in the sub-outer layer are different. Combined with the 27 Al NMR results in Fig. 3.6, the Al atom coordinate numbers in 800MK and 900MK are changed, which inferred that calcination temperature does not affect the Si atomic environment in the kaolin after calcination [13].

3.1.3 Effect of Holding Time on Thermal Transformation of Kaolin 3.1.3.1

Effect of Microstructure of Calcined Products

Figure 3.8a–d shows the SEM morphology of the products after calcined at 900 °C for 1, 4, 6, and 8 h. The powders of 900MK-1, 900MK-4, and 900MK-6 are lamellar, indicating that when the calcination time at 900 °C is lower than 6 h, the calcination time has no significant influence on the powder morphology. Besides, 900MK-8 has both block particles and lamellar powders. Combined with the EDS analysis (Fig. 3.9 and Table 3.4), it is inferred that this powder is mainly composed of Si, O, and Al elements, which is mullite.

3.1.3.2

Effect of Calcination Time on Product Phase

Figure 3.10 shows the XRD patterns of the products formed as a result of kaolin calcination at 900 °C for 1, 4, 6, and 8 h, respectively. 900MK-1 is mainly composed of amorphous phase and contains more α-quartz impurities. In addition, no obvious peak broadening occurs at diffraction angles 2θ = 45.61° and 67.45°. No γ -Al2 O3 is observed in the products after kaolin was calcined for 1 h at 900 °C. When the kaolin was calcined for a longer holding time, that is, 4 h, the diffraction peak at 2θ is 45.61° and 67.45° become broadening, indicating that nano γ -Al2 O3 is generated after 4 h. When the calcination holding time was extended to 6 h, the broadening/widening of diffraction peaks are more obvious (Regions 1 and 2), which shows that the content of the nanocrystalline γ -Al2 O3 increases obviously. When the calcination holding time is further extended to 8 h, mullite phase is also detected besides γ -Al2 O3 .

3.1 Mechanism of Transition from Kaolin to Metakaolin

47

Fig. 3.8 SEM micrographics of products calcined at 900 °C at various time: a 1 h, b 4 h, c 6 h, and d 8 h [13]

3.1.3.3

Effect of Holding Time on Functional Groups and Valence Bond Structure of Calcining Products

Figure 3.11 shows the FT-IR spectra of products of kaolin calcined for 1, 4, 6, and 8 h at 900 °C. With the increasing of holding time, the FT-IR spectra of calcined products all show the characteristic peaks of the same functional groups, which are located at 1095, 803, 568, and 468 cm−1 , corresponding to the vibration of fourcoordinate SiO4 , four-coordinate AlO4 , Si–O–Al, and Si–O–Si, respectively. Being different from the XRD analysis results in Fig. 3.10d, no mullite phase is observed in the FT-IR results in Fig. 3.11d. This is mainly due to the preponderant phase of 900MK-8 which is still amorphous; the vibration peak from mullite phase of such a small amount of crystal phase is too weak to be shown. Figure 3.12 indicates the 27 Al NMR spectra of calcined products after kaolin was held at 900 °C for 1, 4, 6, and 8 h, respectively. The Gaussian fit results of products

48

3 Geopolymerization Mechanism of Geopolymers

Intensity, cps

3000

Al

2000

1000 O

Si Au

Au 0 1

2

3 Energy, keV

4

5

Fig. 3.9 Energy of point A in Fig. 3.9d [13]

Table 3.4 Energy spectrum of point A in Fig. 3.8d [13]

Element O

Atom content (mol%)

48.76

61.87

Si

38.14

28.63

Al

13.10

9.50

100.00

100.00

Total

27

Content (wt%)

Al NMR spectra are listed in Table 3.5. When the holding time is less than 4 h, the characteristic peaks of four-, five-, and six-coordinated Al atoms are relatively obvious. However, when the holding time increased to 6–8 h, the peaks of four- and five-coordinated Al atoms become less obvious and mutually contained. Similar to Fig. 3.11, crystal resonance peak/resonant peak with narrow width and half height is not observed in Fig. 3.12d. According to Gaussian fitting, the content of four-coordinated Al atom in 900MK4 is greater than that of 900MK-1. This indicates that when the kaolin was calcined in less than 4 h, the six-coordinated Al atoms gradually transform to four-coordinated and five-coordinated Al atoms with the holding time increasing. When the holding time is between 4 and 8 h, the four- and five-coordinated Al atoms gradually transform to six-coordinated Al atoms with the holding time increasing. During calcination at 900 °C, the chemical activity of metakaolin increases with the extension of the holding time (1–4 h). However, if the holding time further increases (4–8 h), its chemical activity is decreased.

3.1 Mechanism of Transition from Kaolin to Metakaolin

49

Relative intensity, a.u.

-quartz Mullite -Aluminum Oxide

Region 1

Region 2

(d) (c) (b) (a) 20

30

40

50

60

70

2 theta, degree Fig. 3.10 XRD spectra of products from kaolin calcined at 900 °C at various time: a 1 h, b 4 h, c 6 h, and d 8 h [13]

Based on the XRD results in Fig. 3.10, the FT-IR results in Fig. 3.11, and 27 Al nuclear magnetic resonance (NMR) results in Fig. 3.13, when the calcination temperature is fixed at 900 °C with the holding time less than 6 h, the holding time does not show any effect on the powder morphology of calcined products of kaolin. When the holding time is 8 h, the existence of mullite is found in the powder particles. Besides, with the increasing of the time, the phase composition and structure of the product change gradually: calcined product has gradually presented the tendency of crystallization (generated mullite and γ -Al2 O3 ), and content of the fourcoordinated AlO4 in the structure reaches the maximum value for 4 h. In other words, the metakaolin treated at 900 °C for 4 h has the biggest chemical activity[13].

50

3 Geopolymerization Mechanism of Geopolymers

(d) 3428 564 804

(c)

468 3428

Relative transmission

1099

(b)

563 800 3428

471 568

(a)

1095 800 3430

468 568

1095 803

468 1096 400

600

800

1000

1200

1400

3000

3500

4000

-1

Wave number, cm

Fig. 3.11 FT-IR spectra of products from kaolin calcined at 900 °C at various time: a 1 h, b 4 h, c 6 h, and d 8 h [13]

3.1.4 Transformation Mechanism of Kaolin to Metakaolin There are three endothermic peaks and one exothermic peak on the TG-DTA curve of kaolin (Fig. 3.13). According to the extrapolation method, the starting and ending temperatures of the endothermic peaks on the DTA curve are calculated. The temperature ranges are: 35–423 °C, 423–557 °C, and 557–900 °C, respectively. Kaolin has an endothermic peak at 62 °C, which is caused by the loss of both the adsorbed water from the atmosphere and the inter-layer water of kaolin powder. The temperature of this endothermic peak is 25–250 °C, with a total weight loss of ~2.0%. The other two endothermic peaks in the DTA curve are located at 516 and 737 °C, respectively, which shows weight loss of about 10.6 and 1%, respectively. These results are different from that reported by Akolekar et al. [28], Suitch et al. [29], Meinhold et al. [30], and Chandraskhar et al. [31], and there are only two endothermic peaks in the kaolin thermal transition process below 900 °C. Besides, an obvious exothermic peak appears at 987 °C on the DTA curve due to the formation of mullite [18].

3.1 Mechanism of Transition from Kaolin to Metakaolin (a) 36.9

5.8

Fitting curve

51

(b)

Fitting curve

59.4 34.0

62.7

3.6

150

100

50

0

-50

-100 150

100

-1

(c)

50

0

-50 -1

-100

Chemical shift, cm

Chemical shift, cm

Fitting curve

(d)

Fitting curve

3.9

3.5

62.8 42.6 62.1

150

100

50

0

-50

-100 150

-1

100

42.0

50

0

-50

-100

-1

Chemical shift, cm

Chemical shift, cm

Fig. 3.12 27 Al NMR spectra of products from kaolin calcined at 900 °C at various time: a 1 h, b 4 h, c 6 h, and d 8 h [13] Table 3.5 Result of Gaussian fit for products from kaolin calcined at 900 °C at various time [13] Sample

Coordination of Al atom

Chemical shift (ppm)

Relative area

Percentage

900MK-1

4

~62.7

10.00

19.13

5

~36.9

19.92

38.11

900MK-4

900MK-6

900MK-8

6

~5.8

22.35

42.76

4

~59.4

10.00

48.89

5

~34.0

8.08

39.55

6

~3.6

2.38

11.62

4

~62.9

10.00

17.33

5

~42.6

6.82

11.86

6

~3.5

40.69

70.82

4

~62.1

10.00

24.82

5

~42.0

8.01

19.89

6

~3.5

22.28

55.30

52

3 Geopolymerization Mechanism of Geopolymers

100 o

423 C o

987 C

96

o

900 C

0.2 0.1

92 0.0

o

737 C DTA curve -0.1

88

o

62 C

-1

o

557 C

Endo DTA, uVmg Exo

TG curve

Weight loss, %

0.3

-0.2

o

516 C 84 0

200

400

600 800 o Temperature, C

1000

-0.3 1200

Fig. 3.13 TG-DTA curves of natural kaolin [13]

During the thermal transition of kaolin, different endothermic peaks can lead to the different understanding of the mechanisms of the transition from kaolin to metakaolin. Sayanam et al. [32] observed three endothermic peaks. They thought that the endothermic peak at 150 °C was due to the loss of interlaminar water, and the endothermic peak at 550 °C was due to the loss of structural water. The removal of hydroxyl can result in the endothermic peak centered at 850 °C. The chemical activity of metakaolin was attributed to the dehydroxylation. Chandrasekhar et al. [33] reported that the endothermic peak at ~550 °C on the DTA curve of kaolin was caused by the formation of silica or other phase transitions. Bich et al. [11] reported that the chemical activity of metakaolin was related to the degree of dehydroxylation. When the degree of dehydroxylation reached more than 95%, metakaolin with maximum chemical activity could be obtained. According to He et al. [34], the optimal calcination temperature of kaolinite was the time when the dehydroxylation of kaolinite just ended. So, here are a few questions to be paid attention to: (1) The questions about the six-coordinated Al atom both in 600MK–900MK and the original kaolin The 27 Al NMR results of Fig. 3.6 show that the six-coordinated Al atoms in 600MK, 700MK, 800MK, and 900MK are the same as those in natural kaolin and 500MK. However, according to the FT-IR results of Fig. 3.5, the –OH groups have been completely removed from 600MK, 700MK, 800MK, and 900MK. Theoretically,

3.1 Mechanism of Transition from Kaolin to Metakaolin

53

they are composed of four-coordinated Al atoms, instead of five-coordinated and six-coordinated Al atoms. Based on the analysis by Heide et al. [35], Wang Meirong [13] reported that this is mainly due to, in addition to dehydroxylation, interaction of hydroxyl and hydroxyl to form H2 during the thermal transformation of kaolin, as shown in (3.1). So, there are six-coordinated Al atoms in 600MK, 700MK, 800MK, and 900MK. Besides, it is also possible that during the thermal transformation of kaolin, a dehydrated reaction between hydroxyl and hydroxyl groups may form a negatively charged vacancy around O2– , as shown in (3.2), leading to the formation of four-, five-, and six-coordinated Al atoms. Therefore, the environment of sixcoordinated Al atoms after dehydroxylation is different from that of the original kaolin. → OH− + OH− H2 + 2O−

(3.1)

→ OH− + OH− H2 O + O− +−

(3.2)

(2) The chemical activity of metakaolin As mentioned above, metakaolin can be obtained after calcination for 4 h at the temperature of 600–900 °C, and metakaolin can react with alkaline solution. This ability to react with alkaline solution is called chemical activity or pozzolanic activity. According to the principle of atomic dense heap, the higher the density of the ball is, the less is the energy in the system and the more stable is the structure. Metakaolin is an amorphous material composed of four-, five-, and sixcoordinated Al atoms and four-coordinated Si atoms. In metakaolin structure, compared with the five- and six-coordinated Al atoms, the four-coordinated Al atom has a lower accumulation density and higher internal energy. Thus, the four-coordinated Al atom is more unstable. Wang Meirong et al. [13] proposed using the content of four-coordinated AlO4 as one of the factors to measure the chemical activity of metakaolin. The results show that 737–900 °C is the conversion temperature of activity of metakaolin. During this temperature range, the chemical activity increases as the heat treatment temperature increases. The activity of metakaolin after heat treatment is mainly due to the structural transformation of aluminum and the presence of the Si–O–Al bridging oxygen bond. Whether the nanocrystalline γ -Al2 O3 in 900MK improves the chemical activity of metakaolin or not, leading to the condensation degree (polymerization degree/degree of polymerization), is still needed to be further studied. (3) The mechanism of metakaolin formation In the past, most scholars proposed that the dehydroxylation and structural reorganization of kaolin was a process, when the hydroxyl group is removed, and the structure is also changed. Hindar [36] put forward a different view that the two processes were two independent stages. In the first stage, the dehydroxylation occurs, which

54

3 Geopolymerization Mechanism of Geopolymers

contains both the loss of structural water and the destruction of the kaolin lamellar structure. In the second stage, Al and Si structures are changed and recombined to form metakaolin structures. Combined with the XRD, TEM, FT-IR, 27 Al and 29 Si NMR results of the calcined products of kaolin after thermal transformation process, it can be inferred that the products are still crystalline after calcined at 500 °C (within the temperature range of the second endothermic peak). However, the dehydroxylation leads to the release of H2 O and H2 . The calcined products within the temperature range of the second endothermic peak did not transform into metakaolin. While after calcination at 600 °C (within the temperature range of the third endothermic peak), the product is completely transformed into amorphous phase, indicating the formation of metakaolin together with changed structure units of Al–O and Si–O. With the calcination temperature increasing to 900 °C, the structure of calcination products keeps amorphous, but the chemical environment of these amorphous phase is different. This kind of different amorphous phase structure environment mainly refers to that in the short-range ordered, the contents of different coordinate number of Al atoms contained in the structure are different (the six-coordinated Al atom is gradually transformed to four-coordinated Al atom), while the Si–O structure unit remains unchanged (the four-coordinated Si atom does not change). During the transformation process from kaolin to metakaolin, for example, when the kaolin was calcined at 600 °C, the original Si–O vibration peak is transformed into amorphous wide peak. It is mainly caused by the change of the coordinate number of Al atoms in the sub-outer layer of Si atoms which also changes the structural environment of Si atoms, but the coordinate number of Si atoms remains unchanged, with four coordinates. Therefore, the formation of metakaolin is mainly based on the change of Al–O structure unit in kaolinite structure rather than the change of Si–O structure unit. The change of Al–O structure eventually leads to the decrease of the order degree in kaolin and its disordered state, resulting in the formation of amorphous metakaolin. To sum up, the process of transformation from kaolin into metakaolin can be concluded as follows: after calcination at 500 °C, the phase of kaolin does not change and remains crystal kaolinite phase. Only hydroxyl groups on the layer of hexagonal AlO6 are removed. During calcination at 600 °C, the phase of kaolin begins to show structural transformation, turning into a disordered structure. Its short-range ordered structural units are composed of four-coordinated Si atoms and four-, five-, six-coordinated Al atoms. Only at this point, could the kaolin can transform into metakaolin. Though the dehydroxylation and structural transformation of kaolin are two independent processes, the dehydroxylation provides a prerequisite for the structural transformation.

3.2 Geopolymerization Mechanism of Geopolymer

55

Fig. 3.14 Free energy evolution of geopolymers

Free Energy

Raw Materials

Geopolymer

Zeolites

Ceramics

3.2 Geopolymerization Mechanism of Geopolymer 3.2.1 Driving Force of Geopolymerization Geopolymerization reaction is an exothermic dehydration process, in which water is used as the medium and then being removed after polymerization. A small amount of structural water replaces an O in [SiO4 ]. The polymerization process is a chemical reaction; in other words, a chemical equilibrium process between various aluminosilicate and strong basic silicate solution. The process of forming geopolymer corresponds to a process of reduction of Gibbs free energy (Fig. 3.14). This is just the driving force for geopolymerization. During the polymerization reaction, the bond of inorganic raw materials becomes broken and depolymerization occurs at first. Then, the aluminate, silicate, and aluminosilicate with different status aggregate to form the geopolymer eventually. During the process, the chemical states of the elements are also changed. In terms of metakaolin, geopolymerization process also contains the change of aluminum and silicon chemical environment. A higher unstable chemical state changes to a lower relatively stable chemical state, which is closely related to the chemical activity of metakaolin.

3.2.2 Geopolymerization Mechanisms According to the research by Davidovits [37] and Xu [5], the formation of geopolymer can be divided into four stages: (1) Aluminosilicate mineral powder dissolves in alkaline solution to generate many units containing aluminum and silicon, as shown in (3.3);

56

3 Geopolymerization Mechanism of Geopolymers

(2) Silicon–oxygen tetrahedron and aluminum–oxygen tetrahedron diffuse from the surface of solid particles to the liquid phase; (3) The silicate added in the reactant makes the alkaline silicate solution react with the silicon–oxygen tetrahedron and aluminum–oxygen tetrahedron to form a gel phase; (4) The solution diffusion occurs between the gel phase and the remaining reactants. The capillary movement of the gel phase excludes the remaining water, and solidifies to form the geopolymer. Ma et al. [38] simply summarized the above four stages as follows: dissolution complexation, dispersion migration, concentration polymerization, and dehydration hardening. The rationality of the above geopolymerization mechanisms have been confirmed by leaching test, NMR and SEM observations. However, it still has some limitations. For example, it cannot explain the true three-dimensional structure of geopolymer, nor can it explain the existence of the remaining unreacted substances in geopolymer. − −   Al2 Si2 O7 (MK) + 5H2 O + 4MOH → 2 Al(OH)4 + 2 OSi(OH)3 + 4M+ (3.3) The dissolution of aluminosilicate minerals can lead to the formation of gels, while the silicon–oxygen tetrahedron is polymerized with the aluminosilicate tetrahedron to form amorphous or semi-crystalline three-dimensional structures. During the geopolymerization process, the initial four-, five-, and six-coordinated Al atoms transform into four-coordinated AlO4 and are combined with [SiO4 ] to form a network structure. When the gel solidifies, some of the water evaporates, while the structural water is adsorbed in the nanopores of the material [39]. Yun-sheng Zhang and Wei Sun, and so on [40, 41] used in situ environmental scanning electron microscope, FT-IR, and X-ray diffraction (XRD) to study the generation-developmentevolution process of geopolymer. They found that in the early stage of polymerization, metakaolin is loose. There are several large spaces. As the geopolymerization is going on, many sponge-like colloids form on the surface of particle and fill the pores, which make the matrix dense. By energy dispersive X-ray analysis (EDXA), it is found that the mole ratio of K2 O, Al2 O3 , and SiO2 is close to the theoretical value. The FT-IR peaks of [SiO4 ] shift to the low wave number region, and six-coordinated Al units transform into four-coordinated ones. Besides, there are no crystallized products with regular shape appeared during the geopolymerization process. Only a homogeneous spongy colloid is formed. Based on the calculated group of aluminum silicate ion charge, Weng Lvqian et al. [42] investigated the mechanism of the role of aluminum components in geopolymer synthesis. They found that the aluminum components played a significant role in geopolymerization reaction, which promoted the process of the reaction. Thus, aluminum components with small size metakaolin is easy to be dissolved. The synthetic process of geopolymer has a shorter cure time, more uniform microstructure, and higher mechanical strength. Weng et al. [43, 44] found that the dissolution and reaction

3.2 Geopolymerization Mechanism of Geopolymer

57

polymerization processes of Si/Al were not the same for different ratios. When Si/Al ratio is low (Si/Al < 1), the forming of bonds between [AlO4 ] and [SiO4 ] is slow. However, when Si/Al ratio is high (Si/Al > 3), [SiO4 ] participates in the polymerization reaction once formed. According to the above geopolymerization mechanisms, many kinds of natural aluminosilicate minerals, such as metakaolin, blast furnace slag, construction waste and fly ash, can be used to synthesize geopolymer. Their polymerization ability has been studied previously [45]. The experimental study on the preparation of geopolymer by 16 kinds of natural aluminosilicate minerals shows that the aluminosilicate minerals with island, chain, layer, and framework structure can react with alkaline solution. The polymerization rate is very different. They can produce geopolymer with different strength. The geopolymer with frame structure has the highest reactivity in the process of geopolymerization, while the geopolymer with both frame structure and high calcium content has the highest compressive strength. Nie Yimiao and Ma Hongwen [46] studied the geopolymerization mechanism of the fly ash and kaolin-based geopolymer. They found that the Si–O and Al–O bonds are broken first. Then, the Si and Al components under the action of alkaline metal ions K+ , OH− form the Si and Al low polymers (Si–O–Na, Si–O–Ca–OH, Al(OH)4− , Al(OH)5− , Al(OH)6− ). As the solution composition and ion concentration change, oligomers form gelatinous zeolite-like precursors. Finally, the precursor is dehydrated to form an amorphous phase. In addition, they also prepared geopolymers with excellent properties by using natural minerals such as gold mine tailings, expanded perlite, potassium feldspar, and quartz sand as raw materials [45, 47–49]. However, the mechanism of geopolymerization still needs to be further studied. Because the impurities in natural minerals affect the properties of materials, scholars try to replace natural minerals by chemically active pure aluminum silicate as raw materials, such as synthetic metakaolin. Synthesized metakaolin has high reactivity, so it can produce intense polymerization reaction with alkaline solution. In order to control the reaction rate in the experiment, ice bath is employed. Compared with the geopolymer prepared using natural metakaolin, the geopolymer prepared using synthesized metakaolin also has a similar amorphous structure, but no residual lamellar metakaolin exists in the product, and the material has a more compact and uniform microstructure [50]. Schott and Oelkers [51] suggested that the Si–O–Si–O and Si–O–Al bonds in kaolin or metakaolin were gradually broken under alkaline conditions, in order to form the (OH)3 –Si–O–Al–(OH)3 monomer, which was gradually condensed into dimer, trimer, and polymer. Yang [52, 53] used the theory of covalent bond to explain the polymerization mechanism of geopolymer. She holds that the aluminum silicate vitreous affected by alkaline process, resulting in Si in [SiO4 ]4– combined with O atoms, forms three kinds of hybrid orbitals: sp3 , sp2 , and sp, and form a σ bond. The filled p track by O atoms can be reacted with 3 d track of Si atoms to form dπ–pπ, π and σ bonds to reinforce the Si–O bond and decrease the length of this bond. When reacted with the

58

3 Geopolymerization Mechanism of Geopolymers

OH− , Si4+ can pull O around it to break the Si–O bond. The impact of the Al–O–Al bond is the same. Nie Yinmiao et al. [46] reported that when metakaolin is mixed with alkali solution, the interlayer high energy state Al–O bond in metakaolin first shifts the charge distribution under the action of OH- , causing the Al–O bond broken. At the same time, the Si–O bond in the structure also undergoes similar changes, resulting in the disintegration of the aluminosilicate structure in metakaolin and the formation of an irregular network structure like vitreum. Silicate, aluminate, and silicoaluminate of different polymerization states are repolymerized with broken bonds and depolymerized. Finally, in the geopolymer, Si is in the form of Q4 (2Al), and Al is mainly in the form of four-coordination. Cheng et al. [54] proposed a three-stage viewpoint: overall, the geopolymerization process using silicate as raw material, for example, using kaolin and KOH or NaOH, contains three main stages, successively: (1) use the alkaline metal solution to dissolve Si and Al in the layered structure of kaolinite; (2) form ortho-sialate molecule, namely (OH)3 –Si–O–Al–(OH)3 , which is the monomer in geopolymerization; and (3) the ortho-sialate forms oligomers or larger polymers. North and Swaddle [55] used the ion theory to explain the mechanism of Nageopolymerization, which can be divided into seven steps: In the first step, under the alkaline condition, the four-coordinated Al is connected with Si–O in the form of side chains to form (OH)3 –Si–O–Al–(OH)3 , as shown in + the (3.4), which is equivalent to the monomer of Al(OH)− 4 Na .

(3.4)

In the second step, OH− connects with the Si atom to increase the force of valence bond to the five-coordinated state, as shown in (3.5).

(3.5)

In the third step, the subsequent reaction process can be explained by the division of Si–O bond (electrons transferred to O through Si) to form the intermediate silanol Si–OH group and the basic product of Si–O− , as shown in (3.6).

3.2 Geopolymerization Mechanism of Geopolymer

59

(3.6)

In the fourth step, the basic structural units in the aluminosilicate polymerization process further formed the Si–OH group and the positive aluminosilicate molecule, as shown in (3.7).

(3.7)

In the fifth step, Si–O− reacts with the Na+ to form the Si–O–Na. As shown in (3.8)

(3.8)

In the sixth step, with the reaction process with NaOH, the polycondensation reaction between Si–O–Na and OH–Al is gradually carried out to form the circular structure of aluminosilicate monomer. NaOH is released and participated in the reaction again. As shown in (3.9).

(3.9)

In the seventh step, three circular aluminosilicate monomers are condensed into a network structure of hydroxysodalite (mainly composed of quadrilateral and hexagonal aluminosilicate monomers). As shown in (3.10).

60

3 Geopolymerization Mechanism of Geopolymers

(3.10)

To sum up, the formation process of geopolymer mainly contains three steps: (1) The dissolution of metakaolin or similar precursor of aluminosilicate (precursor) can provide the Al component and all or part of the Si component; (2) Aluminate and silicate components are formed through hydrolysis; (3) These components and silicate from the activator form geopolymer network through condensation and polymerization. Weng and Sagoe-crentsil [43, 44] used the local charge model to predict the hydrolyzed monomers of metakaolin under alkaline solution. Alkaline solutions with different pH values may have different Al and Si components. When the pH value is 14, there may be three monomers [Al(OH)4 ]− , [SiO2 (OH)2 ]2− , and [SiO(OH)3 ]− . When the pH value is 11, there may be three monomers [Al(OH)4 ]− , [Al(OH)3 (OH)2 ], and [SiO(OH)3 ]− , which are consistent with the nuclear magnetic resonance (NMR) results. The dissolved monomers first form complexes with each other and then release the polycondensation components, as shown in formula (3.11).

(3.11)

The generated phase is amorphous and has a three-dimensional network structure with random distribution of SiO4 and AlO4 tetrahedron. Alkaline metal ions are distributed among the network pores to balance the electric charge. The basic structural units of geopolymer network are silicon aluminosiloxic units of (–Si– O–Al–O–), (–Si–O–Al–O–Si–O–), and (–Si–O–Al–O –Si–O–Si–O–), and so on [56, 57].

3.2 Geopolymerization Mechanism of Geopolymer

(a)

(b)

61

(c)

Fig. 3.15 Basic structural units of geopolymer materials [37]. a Poly(sialate)-(PS); b Poly(sialatesiloxo)-(PSS); c Poly(sialate-disiloxo)-(PSDS)

Davidovits [37] predicted the spatial basic structural unit model of some specific geopolymers (Fig. 3.15). The distribution of AlO4 and SiO4 was random, but there were only four SiO4 around each AlO4 and no binding form of Al–O–Al existed.

3.3 Geopolymerization Mechanisms of Geopolymer Based on Synthesized Metakaolin Natural metakaolin (NMK) contains very small amounts of impurities such as magnetite and perovskite; it also contains a lot of high content of Si, O, and/or Al element composition of α-quartz, illite, and mica and other associated mineral impurities. The elements of these impurities are the same with kaolinite. So, it is difficult to accurately analyze quantitatively the content of kaolinite, leading to the inaccurate determination of K+ ions content in alkaline exciting agent. Thus, the Al3+ /K+ mole ratio of the final geopolymer is not equal to 1. When Al3+ /K+ molar ratio is >1, the content of K+ is insufficient and the reaction is insufficient. So, the charge in the final reaction product is unbalanced and the product is in a metastable state. When the Al3+ /K+ mole ratio is 6 h). The ratio of C = O/C–C decreases sharply at 72 h, while the ratio of O–C = O/C–C increases slightly after reduction. After reduction, the two bonds related to the carboxyl group of C = O and O– C = O present an obvious downward tendency. Combined with the FT-IR results (Fig. 4.8), the reduction may be attributed to the decarboxylation reaction of the oxygen-containing functional groups on the GO surface under alkaline conditions. After reduction most of the oxygen-containing functional groups have been removed from the surface of rGO, especially for the C = O.

4.1.2 Defects and Microstructure Analysis of rGO Raman spectroscopy is an important characterization method in the research field of graphene. Structures of graphene can be reflected in the Raman spectra, such as D band, G band of sp2 carbon atoms, and G band (also called 2D peak) of stacking ways between layers of carbon atoms and other information [12, 13]. Figure 4.11 provides the Raman spectra of GO and rGO reduced at different temperatures. The G band reflects the first-order scattering of the E2g phonon of sp2 carbon atoms. The G band of raw GO is located at 1595 cm−1 . With the increase

Intensity, a.u.

D-Band G-Band

80˚C

ID/IG=0.89

60˚C

ID/IG=0.86

40˚C

ID/IG=0.92 ID/IG=0.89

RT GO 1000

ID/IG=0.88

1500

2000 2500 -1 Raman Shift, cm

3000

Fig. 4.11 Raman spectra of GO and rGO obtained after being reduced under geopolymeric solution for 3 h at different temperatures, with permission from [7]

4.1 Reduction Mechanism of GO Under Alkaline Solution

91

of temperature, the G band shifts gradually to lower position. At 80 °C, the G band moves to 1590 cm−1 . The D band reflects the breathing mode of k-point phonons of A1g symmetry, which corresponds to the structural defects and the size of the in-plane sp2 domain. The D band of GO is located at 1355 cm−1 . At 80 °C, it moves to 1343 cm−1 , similar to the position of natural graphite (1576 cm−1 ), indicating that the structure of GO has been changed after alkaline reduction at different temperatures. There are a certain number and size of defects in the sp2 region of samples before and after reduction [14]. In addition, the ratio between the intensities of the D and G bands (ID /IG ) is used to illustrate the relative disorder in graphitic structures [12, 15]. As shown in Fig. 4.11, the ID /IG ratio of GO is about 0.88, which fluctuates with the increase of reduction temperature. The ID /IG ratio reaches 0.92 at 40 °C, and reaches 0.89 when reduced at 80 °C. The ID /IG ratio of them are higher than that of the original GO, indicating that the GO has been reduced. The defects of rGO increase after reduction. Figure 4.12 gives the Raman spectra of rGO after 60 °C reduction at different times. With the extension of reaction time, the G band of rGO shifts to the left (from 1595 to 1578 cm−1 (72 h)), which is very close to the position of G peak of the perfect graphite (1576 cm−1 ), indicating that the reduction of GO is relatively complete at this time. Compared with GO, there is no significant change in the position of D band of rGO reduced for different times, and the band of rGO is still located at 1355 cm−1 for 72 h. With the prolongation of reaction time, the ID /IG ratio also changes, which slowly increases from the initial value of 0.88 to 0.93 for 72 h. It is attributed to the increase

D-Band G-Band

Intensity, a.u.

ID/IG=0.93

1000

1500

72h

ID/IG=0.90

6h

ID/IG=0.86

3h

ID/IG=0.89

0.25h

ID/IG=0.88

GO

2000 2500 -1 Raman Shift, cm

Fig. 4.12 Raman spectra of rGO reduced at 60 °C, with the permission from [4]

3000

92

4 Graphene-Reinforced Geopolymer Matrix Composites

in defects of graphene caused by reduction. Luo et al. [12] showed that the ID /IG ratio of graphene obtained by reduction with sodium hydroxide was close to 1.2. The value (0.93) obtained here is relatively low, which is caused by the difference of the original GO, reducing agent, and the reduction conditions. In geopolymeric solution, the low ID /IG ratio also indicates fewer defects in the rGO made here. Together with the FT-IR results (Fig. 4.8), it demonstrates that the in situ reduction method removes most of the oxygen-functional groups coupled with good defect repair. The AFM morphologies of the original GO and rGO after the reduction under alkaline geopolymeric solution for 0.25 h are shown in Fig. 4.13. The original GO sheet is smooth and the thickness of the single layer is about 1.76 nm, which is caused by its high content of oxygen-containing functional groups. When the reduction reaction occurs after mixing, the thickness of the rGO sheets decreases to 1.52 nm. The sheet is relatively complete and flat, which is still at the micron level, and the size is not significantly different from the original GO, indicating that GO can be in situ reduced in a short time in the alkaline solution and maintains a good and complete shape [1, 16]. The AFM morphologies of rGO after reduction for 3 h at different temperatures are shown in Fig. 4.14. The sheet of rGO keeps a relatively complete shape, and the size is at the micron level, indicating that GO can be reduced in alkaline solutions during this temperature range. Rourke et al. [2] showed that under strong alkaline conditions, GO sheets in solution were easy to adhere to the surface of the separated reduced graphene after reduction. As shown in Fig. 4.13, the thickness of rGO sheets after reduction is about 3 nm, indicating that the number of layers is relatively small [17, 18].

Fig. 4.13 AFM images and cross-section analysis of GO and rGO, with the permission from [6] a GO, b rGO

4.1 Reduction Mechanism of GO Under Alkaline Solution

93

Fig. 4.14 AFM images and cross-section analysis of the rGO after reduction for 3 h [3] a-b RT, c–d 80 °C

The reduction temperature not only affects the content of functional groups on the surface of rGO but also affects the microstructure of rGO. Figure 4.15 demonstrates the SEM photos of GO and rGO after reduction at different temperatures. As shown in Fig. 4.15a, the original GO is laid on the film, showing a translucent state, with many folds on the surface. When reduced at room temperature, there is no significant change in the degree of folding of rGO. The wrinkled and folding degree of the rGO sheets increases gradually with the reaction temperature. When the temperature is 80 °C, the crimp folds become more pronounced.

94

4 Graphene-Reinforced Geopolymer Matrix Composites

Fig. 4.15 SEM micrographs of GO and rGO reduced for 3 h at different temperatures, with the permission from [7] a GO, b room temperature, c 40 °C, d 60 °C, e 80 °C

Figure 4.16 shows the TEM morphology and the selected area electron diffraction (SAED) patterns of rGO reduced at 60 and 80 °C, respectively. The rGO obtained after reduction at 60 °C is relatively less wrinkled than the sample obtained at 80 °C. When the reduction temperature rises to 80 °C, there are obvious folds and crimps on the surface of the rGO. After reduction, the rGO still showed typical hexagonal symmetry diffraction spots, which indicates that the rGO sheets are partially crystalline with several layers.

4.1 Reduction Mechanism of GO Under Alkaline Solution

95

Fig. 4.16 Typical TEM images and the selected area electron diffraction (SAED) patterns of rGO reduced at different temperatures, with permission from [7] a–b 60 °C, c–d 80 °C

Figure 4.17 provides the SEM images of rGO reduced at 60 °C for different times. After reduction, wrinkled and folding degree of the rGO increases with the extension of reduction time. When the reduction time extends to 72 h, the folds are relatively obvious. In order to observe the morphology and structural changes of rGO, TEM is also used to observe the morphology changes before and after GO reduction. The TEM morphology and SAED patterns of rGO before and after reduction are shown in Fig. 4.18. The original GO sheet layer is flat at the micron level about a few microns, with relatively few wrinkles (Fig. 4.18a). After reduction, the scrolled and folded features of rGO increase, which may be due to the defects formed during reduction process (Fig. 4.18b). The insets at upper right corner of Fig. 4.18 show the SAED patterns of GO before and after reduction.

96

4 Graphene-Reinforced Geopolymer Matrix Composites

Fig.4.17 SEM micrographs of rGO reduced at 60 °C for different times [3] a 0.25 h, b 6 h, c 72 h

Fig. 4.18 Typical TEM images and the selected area electron diffraction (SAED) patterns (insets at upper right corner) of GO and rGO, with the permission from [4] a GO, b rGO for 72 h

4.1 Reduction Mechanism of GO Under Alkaline Solution

97

The typical hexagonal symmetry diffraction spots reveal that both GO and rGO sheets are at least partially crystalline. The distinct point patterns indicate the presence of single crystalline domains composed of sp2 -hybridized carbons arranged in a hexagonal lattice. Besides, wrinkling and folding of single graphene layers or overlapping with different graphene layers can lead to the SAED ring patterns due to the various orientations of graphenes. Thus, the weak characteristic ring observed in Fig. 4.18 reveals that the primary GO is made of nearly single layers and the rGO may comprise several layers. After reduction, GO can be reduced to crystalline graphene (rGO) with fewer layers.

4.1.3 Overview of the Reduction Mechanism of GO Under Alkaline Geopolymeric Solution GO can undergo in situ reduction during the preparation of geopolymer, as shown in Fig. 4.19. The alkaline reduction process of GO in the geopolymeric solution is discussed as follows: The introduction of oxygen-containing functional groups plays an obvious role in the separation of GO in water-soluble solvents, providing good condition for the formation of GO aqueous solution. The removal of these oxygen-containing functional groups can lead to the preparation of graphene. In general, the basic reduction of GO is a deoxidation process, which removes oxygen-containing functional groups from its surface. Studies by Luo et al. [12] and Rourke et al. [2] have shown that sodium hydroxide can be used to reduce GO, but the reduction mechanism is still controversial. At the same time, the presence of OH− in the alkaline solution can neutralize and decarboxylate the –COOH bands at the edge of GO. The possible decarboxylation reaction is as follows: R − COOH + KOH → RH + K2 CO3 + H2 O Here, based on the above results, although some C–O bands are still on the surface of rGO after reduction, the appropriate alkali reduction conditions can reduce most of C = O and C-O bands on the surface of the GO.

Fig. 4.19 Schematic diagram of reduction of GO [3]

98

4 Graphene-Reinforced Geopolymer Matrix Composites

4.2 Effect of GO on the Mechanism of Geopolymerization 4.2.1 Effect of GO on Functional Groups and Valence Bond Structures of Geopolymerization Products The in situ reduction of GO can occur after GO mixed with alkaline solution for a short time. Considering the degree of reduction, integrity of rGO, and the dispersion, metakaolin is introduced after GO mixed with alkaline solution for 0.25 h. During the preparation process, geopolymerization of geopolymer and in situ alkaline reduction of GO can occur at the same time. Therefore, in this section, mainly the studies on the effect GO addition on geopolymerization of the rGO/geopolymer were studied. Besides, the mutual combination of GO and geopolymer products and the phase transformation of the geopolymerization products were also studied and investigated. Figure 4.20 provides the FT-IR spectra of reaction products of metakaolin and geopolymer formed at various reaction times during the geopolymerization process. As shown, the FT-IR spectrum of metakaolin mainly includes the following functional groups: the band at 3500 cm−1 is associated with O–H in free water. The band at 1090 cm−1 corresponds to the Si–O from SiO4 , because it includes the influence of the subshell atoms of Si atoms in different environments on the Si–O bond. The peak centered at 803 cm−1 corresponds to the stretching vibration of Al–O in AlO4 . The peak centered at 570 cm−1 is related to the vibration of Si–O–Al, while the peak centered at 463 cm−1 represents the Si–O vibration. For geopolymer, the peaks of FT-IR spectra of the geopolymerization products obtained at different times during the curing process are shifted and changed accordingly. According to the literatures [20–24], with the extension of reaction time (0 min to 7 d), the peak strength of the Si–O–Si structure unit at 463 cm−1 gradually increases, indicating that its content continuously generates and increases with the extension of reaction time. When the reaction lasts for 6 h, the peak of Si–O–Al at 593 cm−1 and the four-coordinated AlO4 structure at 717 cm−1 increases significantly. These represents the structural changes and rearrangement occurring during geopolymerization of Al units. Different from the FT-IR spectra of the metakaolin, weak shoulder peaks of geopolymer products that reacted at different times appear at 880 cm−1 , indicating that Al–OH still exists in the polymerization process. Once the reaction proceeds, the peak of Si–O in metakaolin shifts from 1090 cm−1 to the lower position of 1022 cm−1 . When the reaction lasts for 7 days, the Si(Al)–O units are formed. The peak of Al– O at 803 cm−1 of the original metakaolin decreases, indicating the dissolution of metakaolin.

4.2 Effect of GO on the Mechanism of Geopolymerization

99 593

7d 880 1659

717 1022

5d 3d 24h 12h 6h

Transmittance, a.u.

3h 2h 1h 30min 0min metakaolin

803

463

1090

4000

3500

2000 1500 -1 Wave number, cm

1000

500

Fig. 4.20 FT-IR spectra of metakaolin and geopolymer reaction product formed at various reaction times in geopolymerization samples [19]

From the above, it can be concluded that the main geopolymerization reaction is the silicon substitution by aluminum. However, after a certain time of geopolymerization, Si in the reactants is gradually added into the Si–O–Al structure unit, and forms the Si–O–Al–O–Si or Si–O–Al–O–Si–O–Si structure unit [25]. Figure 4.21 gives the FT-IR spectra of the rGO/geopolymer geopolymerization products formed at various reaction times. The types of reaction products after the addition of GO are consistent with the rGO/geopolymer reaction results of the geopolymer matrix polymerization products (Fig. 4.20). The absorption peaks at 3500 and 1650 cm−1 correspond to the peaks of free water. The peak at 463 cm−1 corresponds to the Si–O–Si structure unit. When the polymerization reaction exceeds for 6 h, the vibration of Si–O–Al occurs in the vibration centers at 593 cm−1 . The

100

4 Graphene-Reinforced Geopolymer Matrix Composites

593

7d 880

1659

5d

717 1022

3d 24h

Transmittance, a.u.

12h 6h 3h 2h 1h 30min 0min metakaolin

803

1090

4000

3500

2000 1500 -1 Wave number, cm

463

1000

500

Fig. 4.21 FT-IR spectra of metakaolin and rGO/geopolymer reaction product formed at various reaction times in geopolymerization samples [19]

vibration centered at 717 cm−1 shows the Al–O vibration in AlO4 unit. During the reaction, the peak of Si–O band in metakaolin shifts from the original 1090 cm−1 to the lower wave number of 1022 cm−1 . No characteristic peaks of rGO are found in the FT-IR spectra of rGO/geopolymer geopolymerization products, which is caused by its low content and are undetected. The effect of rGO addition on the geopolymerization process of geopolymer can be further obtained by the 27 Al and 29 Si NMR analysis.

4.2 Effect of GO on the Mechanism of Geopolymerization

101

4.2.2 Chemical Structure of 27 Al and 29 Si According to the 27 Al NMR spectra and Gaussian fit, results of the geopolymer reaction products at different times during the geopolymer curing process are summarized in Fig. 4.22 and Table 4.4. The Al atoms of metakaolin are mainly composed of four-, five- and six-coordinated units, among them, the Al atom is mainly composed of four-coordinated units. When metakaolin is mixed with alkaline geopolymeric solutions, within a short time (0–30 min), the product of the geopolymerization is still composed of three Al atom structure units of four, five and six coordination. When the reaction time is longer (>1 h), the Al atoms in the geopolymer products are mainly composed of four-coordination units. The peak intensity of four- and six-coordinate significantly decreases, indicating that they transform to four-coordinated Al atoms. As the reaction progresses to 6 h, the intensity of the peaks representing four- and six-coordinated Al atoms decreases further. After holding for 24 h, the structural units of Al atoms in the geopolymer are completely transformed into four coordinates. For the geopolymerization process, it can be found that the coordination of Al atoms with different chemical shift have changes in a short period of time, and gradually shift to the lower chemical shift. Especially for the four-coordinated Al atoms, the peaks of metakaolin at 53.3 ppm shift to 56.3 ppm after geopolymerization. This comes from the effect of Si atoms of the subshell on the Al at different times. Eventually, Al atoms completely change into four coordinates. Figure 4.23 and Table 4.5 demonstrate the 27 Al NMR spectra and Gaussian fit results of the rGO/geopolymer geopolymerization products at different times. At the initial stage of 0–30 min (Fig. 4.23a-b), the Al atomic structure units in the reaction products are still mainly composed of four-, five- and six-coordination units. The proportions of the three coordinates change little and are still dominated by the five coordination. Compared with the fitting results of the 27 Al NMR spectrum of the matrix (Fig. 4.22, Table 4.4), after adding rGO, the atom content of four-coordinated Al in the products is relatively higher, which may be attributed to the acceleration of rGO on the initial products early in the geopolymerization process. The increase of the content of four-coordinated Al indicates the transformation of other coordinates to four coordination. When the reaction lasts for 1 h (Fig. 4.23c), the proportion of four-, five- and sixcoordinated Al atoms in rGO/geopolymer products is 46.8, 37.5, and 15.7% (Table 4.5), respectively. The proportion of four-coordinated Al atoms continues to increase to 69.8% when the geopolymerization reaction is carried out for 2 h (Fig. 4.23d). At 3 h (Fig. 4.23e), the peak of four-coordinated Al atoms at 56.6 ppm is dominant. When the reaction reaches 6 h (Fig. 4.23f), the peak of four-coordinated Al atoms disappears. At 24 h (Fig. 4.23g), the Al units in the rGO/geopolymer geopolymerization products are completely transformed into four coordination. During the polymerization process up to 24 h, the chemical shift of Al atoms gradually shifts

102 Fig. 4.22 27 Al NMR spectra of metakaolin and geopolymer reaction product formed at various reaction times in geopolymerization samples [19] a metakaolin, b 0 min, c 30 min, d 1 h, e 2 h, f 3 h, g 6 h, h 24 h

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(h)

to a lower position, and the four-coordinated Al shifts from 53.3 ppm of the original metakaolin to 56.3 ppm (reaction for 24 h). Comparing different coordinations of Al atom in the process of geopolymerization, the addition of rGO accelerates the conversion of four-coordinated Al in the

4.2 Effect of GO on the Mechanism of Geopolymerization

103

Table 4.4 Result of Gaussian fit for geopolymer reaction products formed at various reaction times [19] Sample

Coordination of Al atom

Chemical shift (ppm)

Relative area

Percentage (%)

4

53.3

2.37

21.6

MK

5

29.5

5.28

48.3

6

3.1

3.29

30.1

4

56.3

3.06

25.8

5

28.8

7.09

59.6

6

3.3

1.73

14.6

4

56.9

5.86

27.6

5

29.7

11.28

53.1

6

3.4

4.1

19.3

4

57.6

2.41

54.3

5

27.8

1.52

34.2

6

3.2

0.51

11.5

4

56.7

3.61

71.7

5

27.5

0.82

16.5

6

3.2

0.6

11.8

4

56.3

–

–

5

27.5

–

–

6

~ 2.3

–

–

geopolymer-6 h

4

56.3

10

100

6

~ 2.2

–

–

geopolymer-24 h

4

56.3

10

100

geopolymer-0 min

geopolymer-30 min

geopolymer-1 h

geopolymer-2 h

geopolymer-3 h

rGO/geopolymer products at the early stage (0–30 min) during the geopolymerization process. However, the effect of rGO is not obvious at the later stage (1–24 h). This is due to the geopolymer particles likely prior formed, attached, and grown on the surface of the rGO sheets. As a result, once the geopolymer matrix is solidificated, surface of the rGO is all coated with the geopolymerization products. The results show that the addition of rGO has a certain effect on the Al atomic structure environment during the geopolymerization, but has little effect on the final type of rGO/geopolymer geopolymerization products. For the geopolymer materials, Qn (mAl) can be usually used to represent the polymerization state of silica anion in geopolymer materials [25–27]. Different n values in Qn represent different structural units. Figure 4.24 gives the 29 Si NMR spectra of metakaolin and geopolymer reaction product formed at various reaction times in geopolymerization samples. As shown in Fig. 4.24a, metakaolin contains only tetra-coordinated Si atoms. Two peaks at −99.3 and −90.8 ppm correspond to fourcoordinated Si atom(Q4 (1Al) and Q4 (3Al)). Q4 (1Al) is the major species in the metakaolin.

104

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

(g)

Fig. 4.23 27 Al NMR spectra of rGO/geopolymer reaction product formed at various reaction times in geopolymerization samples [19] a 0 min, b 30 min, c 1 h, d 2 h, e 3 h, f 6 h, g 24 h

4.2 Effect of GO on the Mechanism of Geopolymerization

105

Table 4.5 Results of Gaussian fit for rGO/geopolymer reaction products formed at various reaction times [19] Sample

Coordination of Al atom

Chemical shift (ppm)

Relative area

Percentage (%)

rGO/geopolymer-0 min

4

57.5

3.16

35.2

5

32.1

3.41

37.8

6

3.3

2.43

27

4

58.5

2.25

34.3

5

29.3

3.13

47.6

6

3.1

1.19

18.1

4

56.6

2.33

46.8

5

27.8

1.87

37.5

6

2.9

0.78

15.7

4

56.6

4.34

69.8

5

27.8

1.39

22.3

6

2.9

0.49

7.9

4

56.6

–

–

5

27.5

–

–

6

2.9

–

–

rGO/geopolymer-6 h

4

56.3

10

100

6

2.6

–

–

rGO/geopolymer-24 h

4

56.3

10

100

rGO/geopolymer-30 min

rGO/geopolymer-1 h

rGO/geopolymer-2 h

rGO/geopolymer-3 h

On initiation of geopolymerization of geopolymer (Fig. 4.24b), the 29 Si NMR spectra shift and two 29 Si resonances at −79.7 and −87.6 ppm appear, assigned to Q4 (4Al) and Q4 (3Al) structure units, indicating that the geopolymerization begins. At 30 min (Fig. 4.24c), 29 Si NMR of the geopolymerization products shows peaks at −79.7, −87.6, and −93.7 ppm representing structural units of Q4 (4Al), Q4 (3Al,) and a new Q4 (2Al) species in the geopolymer. The 29 Si chemical shifts of the geopolymer products show no obvious change within 1–2 h, while the relative intensity of −80.6 ppm increases regularly (Fig. 4.24d-e). When the reaction lasts for 3 h (Fig. 4.24f), the 29 Si chemical shifts move to −81.2, 87.6, and −95.1 ppm, representing Q4 (4Al), Q4 (3Al), and Q4 (2Al) species, respectively. The peak shifts in a negative direction relative to the previous moment indicate an increase in the degree of silicon(aluminum)–oxygen anion polymerization of the geopolymer. When the geopolymerization time reaches 6 h (Fig. 4.24g), the chemical shifts are at −90.1 and 102.6 ppm, which are assigned to Q4 (3Al) and Q4 (1Al), respectively. When the reaction lasts for 24 h (Fig. 4.24h), the chemical shifts completely move to −90.1 ppm, indicating that Si is present mainly as Q4 (3Al) structural units. During the geopolymerization process, the chemical shift of SiO4 units gradually moves toward the low magnetic field as the reaction progresses, indicating that the number of Al connected to the SiO4 structure continues to increase.

106

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

(g)

(g)

(h)

(h)

Fig. 4.24 29 Si NMR spectra of metakaolin and geopolymer reaction product formed at various reaction times in geopolymerization samples [19] a metakaolin, b 0 min, c 30 min, d 1 h, e 2 h, f 3 h, g 6 h, h 24 h

4.2 Effect of GO on the Mechanism of Geopolymerization

107

With the reaction, the Si atoms gradually changed from Q4 (1Al) structural unit to Q4 (3Al) structural unit. This also shows that Si and Al units combine with each other as the geopolymerization reaction proceeds. The geopolymer prepared from synthetic metakaolin as raw material described in literature [25] finally generated Q4 (3Al) structural unit, without residual Q4 (1Al) and six-coordinated Al structural unit of metakaolin. Here, we use natural metakaolin as raw material preparation of geopolymer which also generate Q4 (Al) structure unit. The result is consistent with the result of using synthetic metakaolin. When the geopolymerization time reaches 6 h, the structural units of Si and Al almost all change into a relatively stable four-coordinated state, which is similar to the result of the geopolymerization of synthetized metakaolin used in the literature [25]. This indicates that the geopolymerization with natural metakaolin as the raw material is complete, although the changes of Si and Al structural units are different in the process. Figure 4.25 shows the 29 Si NMR spectra of the rGO/geopolymer reaction product formed at various reaction times in geopolymerization samples. As can be seen from Fig. 4.25a, when metakaolin is added to the alkaline mixture solution mixed with rGO and stirred for curing (0 min), the 29 Si NMR spectra moves to the low magnetic field, showing the sharp peaks of the new structural unit Q4 (4Al), Q4 (3Al), and Q4 (2Al) at the chemical shifts of −80.7, −87.8, and −95.2 ppm, respectively. Compared with the geopolymer matrix, the structure unit Q4 (2Al) appears. When the reaction lasts for 30 min (Fig. 4.25b), the peaks representing the structural unit Q4 (4Al), Q4 (3Al), and the newly emerged Q4 (2Al) are shown at −80.6, −87.8, and − 95.2 ppm of the product. At this point, there is little difference from the matrix. At 1 h (Fig. 4.25c), the peak of Q4 (2Al) is divided into two peaks at −92.2 and −94.6 ppm. At 2 h (Fig. 4.25d), the chemical shift of the rGO/geopolymer is not significantly different from that of the geopolymer matrix. When the geopolymerization lasts for 3 h (Fig. 4.25e), a peak of Q4 (1Al) appears at −106.2 ppm. When the reaction time reaches for 6 h (Fig. 4.25f), the type is almost unchanged compared with that at 3 h. At 24 h (Fig. 4.25g), the peaks at −87.0 and −90.1 ppm of rGO/geopolymer are observed, indicating that Si is present mainly as Q4 (3Al) structural units. Compared with the pure geopolymer, the addition of rGO accelerated the conversion of Si structure unit from Q4 (1Al) to Q4 (4Al), Q4 (3Al), Q4 (2Al), and Q4 (1Al) species, and it also promotes the formation of the final structural unit Q4 (3Al). Thus, the addition of rGO affects the Si structure during the geopolymerization and has a positive influence on the generation of Q4 (3Al) species.

4.2.3 Phase Analysis of rGO/geopolymer Geopolymerization Products Figure 4.26 indicates the XRD patterns of geopolymer products formed at various reaction times during the curing process of geopolymer. At the initial stage of reaction

108

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

(g)

Fig. 4.25 29 Si NMR spectra of rGO/geopolymer reaction product formed at various reaction times in geopolymerization samples [19] a 0 min, b 30 min, c 1 h, d 2 h, e 3 h, f 6 h, g 24 h

4.2 Effect of GO on the Mechanism of Geopolymerization

109

Relative intensity,a.u.

Quartz,low,65-0465

7d 5d 3d 24h 12h 6h 3h 2h 1h 30min 0h 10

20

30 40 2theta,degree

50

60

Fig. 4.26 XRD patterns of geopolymer reaction product formed at various reaction times in geopolymerization samples [19]

(0–6 h), the typical geopolymer broad amorphous humps around 17–32° 2θ are relatively low, and the degree of amorphous is relatively high. When the time exceeds 12 h, the position of the amorphous peak does not change, but the relative strength of the amorphous peak increases significantly, indicating that the amorphous content decreases with the increase of reaction time.

110

4 Graphene-Reinforced Geopolymer Matrix Composites

Figure 4.27 provides the XRD patterns of rGO/geopolymer reaction product formed at various reaction times. After adding rGO, the XRD patterns of the composites are not significantly different from that of the matrix, indicating that the rGO does not affect the formation of amorphous structure of the geopolymer products. No obvious peak representing rGO can be detected in the XRD pattern, due to that its content is low and not detected. The addition of GO does not affect the formation of amorphous structure of geopolymer.

Relative intensity,a.u.

Quartz,low,65-0465

7d 5d 3d 24h 12h 6h 3h 2h 1h 30min 0h 10

20

30 40 2theta,degree

50

60

Fig. 4.27 XRD patterns of rGO/geopolymer reaction product formed at various reaction times in geopolymerization samples [19]

4.2 Effect of GO on the Mechanism of Geopolymerization

111

4.2.4 Micromorphology Analysis of rGO/Geopolymer Geopolymerization Products So far, there are few reports on the observation of the microscopic morphology of the reaction products at different times during the geopolymer geopolymerization process. During this process, the morphology evolution of the geopolymer particles at different reaction times can be observed by SEM. According to the changes in the curing of geopolymer observed by eyes during the geopolymerization process, it can be divided into two stages: initial section (0–6 h) and last section (>6 h). Figures 4.28 and 4.29 show the micrographs of geopolymer particles for reaction products formed at various reaction time in geopolymerization process. As shown in Fig. 4.28a, at first (0 min), the particle size of the products is small and loose, and there are a lot of spongy morphology on the surface of particles, accompanied by irregular pores. As the reaction proceeds, the number of pores on the particles significantly decreases, and the density increases compared with the previous time. At 2 h (Fig. 4.28d), the particle size increases greatly whereas the pores decrease significantly. When increases to 3–6 h (Figs. 4.28e–f and 4.29e–f), the surface is obviously dense, and the pores can hardly be seen. As shown in Fig. 4.30a, when curing to 12 h, the surface of geopolymer is clearly denser than before. At 7 days (Fig. 4.30e), the surface of the products becomes quite denser. Compared with the geopolymerization process using synthetic metakaolin as raw materials, the geopolymer with nature metakaolin as raw material has a lower reaction rate and viscosity during the stirring process at room temperature, which make it more easier to introduce the reinforcing phase in the preparation process of composites. After curing for 6 h, the particle pores are extremely reduced. At 24 h, it is very dense. These provide sufficient time for the original reduction of GO in the preparation of geopolymer composite. Figures 4.31 and 4.32 show the micrographs of particles for the rGO/geopolymer reaction products formed at various reaction times. Compared with the pure geopolymer (Figs. 4.29 and 4.30), particle size of the rGO/geopolymer products is significantly larger, which is caused by the connection and aggregation of dispersed rGO and matrix geopolymer. There is no obvious separation between rGO and matrix, indicating that the rGO bonds well with the products of geopolymer matrix. At the starting point (0 min), both geopolymerization and reduction of graphene oxide occur at the same time, as shown in Fig. 4.31a. The products show smaller particle size, and there are many irregular pores. When the reaction lasts for 2 h (Fig. 4.32d), the rGOs are still combined well with the geopolymer particles. At 3 h (Fig. 4.32e), the rGO is wrapped around by the particles of geopolymer matrix. When the reaction reaches 6 h (Fig. 4.32f), the surface of the particles has become very dense and almost no voids can be seen. At this point, compared with the matrix of the same period, the particles are full of the edges and corners. RGO/geopolymer particles transform into a regular block when the geopolymerization process is in the last stage (12 h–7 days), as shown in Fig. 4.33. From Fig. 4.33a, rGO is well combined with the geopolymer matrix. A large number of

112

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 4.28 Micrographs with low magnification of geopolymer particles for reaction products formed at various reaction times in geopolymerization process [3] a 0 min, b 30 min, c 1 h, d 2 h, e 3 h, f 6 h

4.2 Effect of GO on the Mechanism of Geopolymerization

(a)

(b)

(c)

(d)

(e)

(f)

113

Fig. 4.29 Micrographs of geopolymer particles for reaction products formed at various reaction times in geopolymerization process [19] a 0 min, b 30 min, c 1 h, d 2 h, e 3 h, f 6 h

114

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

Fig. 4.30 Micrographs of geopolymer particles for reaction products formed at long reaction times in geopolymerization process [3, 19] a 12 h, b 24 h, c 3 days, d 5 days, e 7 days

4.2 Effect of GO on the Mechanism of Geopolymerization

(a)

(b)

(c)

(d)

(e)

(f)

115

Fig. 4.31 SEM micrographs with low magnification of particles for rGO/geopolymer reaction products formed at various reaction times in geopolymerization process [3] a 0 min, b 30 min, c 1 h, d 2 h, e 3 h, f 6 h

116

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 4.32 SEM micrographs of particles for rGO/geopolymer reaction products formed at various reaction times in geopolymerization process [19] a 0 min, b 30 min, c 1 h, d 2 h, e 3 h, f 6 h

4.2 Effect of GO on the Mechanism of Geopolymerization

(a)

117

(b)

rGO (c)

(d)

(e)

Fig. 4.33 SEM images of particles for rGO/geopolymer reaction products formed at long reaction times in geopolymerization process [3, 19] a 12 h, b 24 h, c 3 days, d 5 days, e 7 days

118

4 Graphene-Reinforced Geopolymer Matrix Composites

particles of geopolymer geopolymerization products are attached to the surface of the rGO sheets. At 7 days (Fig. 4.33e), the rGO/geopolymer composites manifest a relatively homogeneous and dense microstructure.

4.2.5 Geopolymerization Mechanism of rGO/Geopolymer Composites The schematic diagram of the geopolymerization process of geopolymer and rGO/geopolymer composite is shown in Fig. 4.34. The geopolymerization mechanism can be rationally expressed according to the experimental analysis as follows. First, nature metakaolin particles dissolve from the surface after mixing with alkaline silicate solutions, then the Si–O and Al–O bonds hydrolyzation, Si and Al monomers diffusion, polycondensation, and rearrangement occur. During these processes, fourand six-coordinated Al–O sites conversion to four coordination happens, resulting in the condensation of Si species with Al species mainly in the form of Q4 (3Al) and Al in four coordination. GO can be in situ reduced to rGO in alkaline silicate solutions and has positive effects on the geopolymerization during the reaction process [28]. As presented in Fig. 4.34, the rGO accelerates the conversion of Al–O sites into four coordinates and Si atoms in the form of Q4 (3Al). However, it does not change the final network structure of the rGO/geopolymer. The products bond well with the rGO sheets and show denser microstructure and lower amorphous degree with the increases of reaction time. The addition of GO is proper in preparing rGO/geopolymer composites.

Fig. 4.34 Schematic diagram of geopolymerization process of geopolymer and rGO/geopolymer [19]

4.3 Microstructure and Mechanical Properties of in Situ rGO/Geopolymer Composites

119

4.3 Microstructure and Mechanical Properties of in Situ rGO/Geopolymer Composites 4.3.1 Microstructure of rGO/Geopolymer Composite (1) Phase composition The XRD patterns of rGO/geopolymer composites with different GO loadings are shown in Fig. 4.35. The XRD patterns of the rGO/geopolymer composites with different GO contents are not greatly different from those of the geopolymer matrix, showing typical amorphous XRD characteristic refers to a broad amorphous hump near 28°. In addition, a small amount of quartz phase is from the raw material metakaolin. There is no obvious phase structure change of the rGO/geopolymer samples, suggesting little effect of the addition of rGO sheets on the structure of final composites. The comparison photographs of rGO/geopolymer composites with different rGO contents are shown in insets in Fig. 4.35. The matrix of the original geopolymer is white. The samples become darker gradually with the increasing amounts of rGO, which is caused by the reduction of GO to rGO [28]. The existence of rGO in rGO/geopolymer composites with different GO contents can be obtained by Raman analysis [29]. Figure 4.36 shows the Raman

Quartz,syn

Intensity, a.u.

(a) rGO/KGP0 (b) rGO/KGP0.5 (c) rGO/KGP1 (d) rGO/KGP3 (e) rGO/KGP5 (f) rGO/KGP10

f e d c b a 20

40 60 2theta, degree

80

Fig. 4.35 XRD patterns of rGO/geopolymer composites with different rGO contents. Insets are the photographs of the corresponding samples, with permission from [6]

120

4 Graphene-Reinforced Geopolymer Matrix Composites

D-Band G-Band

Intensity, a.u.

rGO/KGP10 rGO/KGP5 rGO/KGP3 rGO/KGP1 rGO/KGP0.5 rGO/KGP0

GO 1000

1500

2000 -1 Raman Shift, cm

2500

3000

Fig. 4.36 Raman spectra of GO and rGO/geopolymer composites with different GO contents, with permission from [6]

spectra of rGO/geopolymer composites with different GO contents. The peak of rGO/geopolymer composite is similar to that of the original GO, showing D band and G band at 1335 and 1599 cm−1 , respectively, which conforms the rGO in the composites. The intensity ratio of ID /IG is higher than in as-received GO, illustrating the reduction of GO in the composites. (2) Microstructure Figure 4.37 illustrates the surface microstructure of rGO/geopolymer composites with different GO contents. The surface of the composites is relative density, and there is no significant difference because of the small rGO size. So, it is difficult to accurately observe the rGO sheets from polished surface, especially when the rGO sheets are perpendicular to the surface. Therefore, TEM observation is needed to determine the existence of rGO. Figure 4.38 shows the TEM morphology of rGO/geopolymer5 composites. Consistent with other geopolymer materials [30], the matrix comprises fine sphere particles. The flat rGO has a wrinkled surface, surrounded by geopolymer particles. The well-bonding state can be observed between rGO and geopolymer matrix (Fig. 4.38b). Ring patterns along with point patterns can be observed from the selected area electron diffraction (SAED) patterns of rGO (insets in Fig. 4.38), which originated from various orientations of crystal rGO. Figure 4.39a-d shows the TEM images of rGO/geopolymer10 composites. The flat rGO sheet has a crumpled surface. The matrix presents a honeycomb-like structure

4.3 Microstructure and Mechanical Properties of in Situ rGO/Geopolymer Composites

(a)

121

(b)

5μm (c)

5μm (d)

5μm (e)

5μm (f)

5μm

5μm

Fig. 4.37 Typical surface microstructure of rGO/geopolymer composites with different GO contents [3] a 0 wt%, b 0.05 wt%, c 0.1 wt%, d 0.3 wt%, e 0.5 wt%, f 1 wt%

and the Debye ring in the selected area electron diffraction (SAED) pattern (inset in Fig. 4.39a, top) indicates its amorphous nature. However, the SAED pattern (inset in Fig. 4.39a, bottom) of the rGO covered with geopolymer shows the Debye ring with three distinct diffraction spots, which might be due to the hexagonal pattern of graphene. Besides, the rGO sheet also shows well dispersion and good adhesion with the geopolymer matrix (Fig. 4.39b), which is beneficial to the improvement of mechanical properties of composites. Since the rGO sheet has large width, geopolymerization can occur on the surface of rGO sheets, resulting in the formation of fine geopolymer particles attached with rGO sheets (Fig. 4.39c). As shown in the HRTEM image of Fig. 4.39d, the thickness of rGO in rGO/geopolymer is about 5 nm with few layers. Thus, combination of rGO and geopolymer matrix interface is good.

122

4 Graphene-Reinforced Geopolymer Matrix Composites

(a)

(b)

Wrinkled surface

rGO

Geopolymer particles Fig. 4.38 TEM images of rGO/geopolymer5; insets display selected electronic diffraction patterns, with permission from [6] a wrinkled rGO with matrix, b bonding between rGO and matrix

4.3.2 Mechanical Properties and Toughening Mechanisms of rGO/Geopolymer Composites (1) Mechanical properties Figure 4.40 shows the mechanical properties of the rGO/geopolymer composites with different rGO contents. After the addition of rGO, the elastic modulus of rGO/geopolymer composite decreases a little compared with the matrix. This is because although the modulus of graphene alone is higher, the added GO modulus is similar to that of multilayer graphite. So, the modulus decreases after adding rGO. The flexural strength of geopolymer matrix is 12.3 MPa. When the rGO content increases to 0.3 wt%, the flexural strength reaches the maximum value of 17.9 MPa, which shows an increase of 45% compared with the matrix. While further addition of rGO in the composites results in a decrease of flexural strength, this is attributed to the processing defects caused by uneven distribution of overlapped rGO sheets. The processing defects of geopolymer matrix are pores and voids that originated from the residual bubbles introduced during the stirring and curing process. After the addition of rGO, the fracture toughness of the composites is improved. The fracture toughness of rGO/geopolymer5 increases from 0.13 to 0.21 MPa·m1/2 with a 61% increase over the matrix. The improved fracture toughness is based on the good combination of flat rGO and the matrix. After stress, the cracks proliferate and deflect across rGO, avoiding the growth of the crack length. This process absorbs more energy and plays a buffering role on the crack [31, 32]. (2) Fracture morphology In order to observe the dispersion and interface combination of rGO in the composites, Fig. 4.41 presents the fractographs of the rGO/geopolymer with different GO contents. Figure 4.41a shows the geopolymer matrix. The fracture is relatively flat

4.3 Microstructure and Mechanical Properties of in Situ rGO/Geopolymer Composites

(a)

(b)

(c)

(d)

123

Fig. 4.39 TEM images of rGO/geopolymer composites, with permission from [4] a rGO sheets, the insets are SAED pattern of geopolymer matrix (up) and rGO sheet (bottom), respectively, b-c rGO sheets covered with geopolymer matrix, d HRTEM image of rGO sheet in geopolymer

with no obvious ridge. With the increase of rGO content, the amount of edges protruding from the fracture surface increases with increasing rGO content. These edges are caused by the pull out of the rGO sheets from the bonded geopolymer particles under stress. The detailed observation of interface microstructure of the rGO/geopolymer5 composites is shown in Fig. 4.42. As indicated in Fig. 4.42a and b, the thin dispersed rGO sheets are flexible, nearly enveloped in the matrix, combining well with the geopolymer matrix to prevent cracks. Crack deflection and propagation are also observed around the rGO sheets. As shown in Fig. 4.42 c, geopolymer particles in nano-size are also found closely wrapped on the surface of the rGO sheets. Crack deflections are observed which are attributed to the pulling-out effect of rGO sheets with larger size (Fig. 4.42b and d). Therefore, it can be inferred that the rGO sheets

124

4 Graphene-Reinforced Geopolymer Matrix Composites

14

35

0.1

12

30

10

25 8 20 6

15

Fracture toughness,MPa· m Flexural strength,MPa Elastic modulus,GPa

5 0.0

4

1/2

10

Elastic modulus,GPa

0.2

Flexural strength,MPa

Fracture toughness,MPa· m

1/2

40

2 0

0

0.0

0.1

0.2 0.3 0.4 0.5 Weight fraction of GO, %

1.0

Fig. 4.40 Mechanical properties of rGO/geopolymer composites with different GO contents [3]

can prevent the crack propagation under external stress. This good interfacial bonding can result in enhanced mechanical properties of the composites, which attributes that stress can be effectively transferred with the addition of rGO. (3) Mechanism of strengthening and toughening The strength and toughness of composites are influenced by both the content and dispersion of rGO. When the GO content is very low (95% and the content of −COOH was about 3.86 wt%, outer diameters (OD) 400 m2 /g. For convenience of expression, MWCNTs were still used to represent −COOH functionalized multiwall carbon nanotubes. A mixture of pristine MWCNTs and sodium dodecyl sulfate (SDS) solution was ultrasonically dispersed in an ice water bath for 10 min, and repeated six times to get homogeneous solution. The metakaolin (MK) was prepared by calcining kaolin at 800 °C for 2 h and the potassium silicate solution was prepared by dissolving amorphous silica (Shanghai Dixiang Indus., China) into KOH (Tianjin Fuchen Indus., China) solution and stirring for 48 h in order to dissolve the silica completely. Chemical compositions of metakaolin and sol-SiO2 are shown in Tables 5.1 and 5.2. Geopolymer resin with composition of SiO2 /Al2 O3 = 4, SiO2 /K2 O = 4, and H2 O/K2 O = 11 (mole ratio) was obtained by mixing metakaolin powder with potassium silicate solution. For the preparation of the MWCNTs/composites, metakaolin and MWCNTs solution were mixed in ethanol with zirconia balls in a plastic bottle for 24 h, then dried at 80 °C in a rotary evaporator and screened through a 120-mesh sieve. Then the mixed powders were gradually added into the potassium silicate solution under ultrasonic-assisted high-shear mixer. After that, the slurry was cast into polystyrene containers, sealed, © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_5

131

132

5 Particles-Reinforced Geopolymer Matrix Composites

Table 5.1 Chemical composition of metakaolin (wt%) Content

SiO2

Al2 O3

Fe2 O3

CaO

CuO

TiO2

SrO

wt%

51.906

40.404

0.921

0.113

0.018

0.763

0.011

Content

K2 O

P2 O5

SO3

ZrO2

PbO

Ga2 O3

ZnO

wt%

0.455

0.155

0.103

0.034

0.074

0.003

0.052

Table 5.2 Performance parameters of sol-SiO2 SiO2 (wt.%)

pH value

Specific gravity (g·mL−1 )

Viscosity (MPa·s)

Na2 O (wt%)

40–41

9.33

1.29–1.295

5–12

0.25

and cured at 60 °C for 5 days. After demolding, the hardened geopolymer was further cured at 60 °C for another 2 days. The content of MWCNTs is 0, 0.5, 1, 2, 3, and 5 wt%, respectively.

5.1.2 Phase Composition As depicted in Fig. 5.1, the crystalline kaolinite, with main component of natural kaolin, could transform into amorphous metakaolin associated with a hump located between 20 and 28° of 2s after 2 h calcined at 800 °C. The dehydroxylation process and the loss of crystal water accounted for the collapse of the kaolinite’s ordered structure in this transformation process [1]. After doping MWCNTs, there were no obvious phase change in the MWCNTs/geopolymer composites, indicating that such small additions were not sufficient to influence the phase composition of the final product. Besides, it could be found that α-quartz in kaolin keep the original state despite undergoing the high temperature processing. The XRD patterns of MWCNTs/geopolymer composites were shown in Fig. 5.2. After the geopolymerization reaction, the resulting geopolymer still showed a typical amorphous character, that is, a broad amorphous hump at ∼28° 2θ, which was different with that of metakaolin. Compared with metakaolin, the center of hump shifted toward the high angle region in geopolymer, which suggested that the geopolymerization reaction occurred in this process. The characteristic peak of αquartz could still be observed in the resulting geopolymer matrix, indicating α-quartz could not be solved in the geopolymerization reaction.

5.1 Carbon Nanotube-reinforced Geopolymer Matrix Composites Fig. 5.1 XRD patterns of kaolin, metakaolin (MK) and the mixture of MK with different content of MWCNTs, with permission from [2]

Fig. 5.2 XRD patterns of metakaolin (MK), geopolymer matrix, and MWCNTs/geopolymer composites, with permission from [2]

133

134

5 Particles-Reinforced Geopolymer Matrix Composites

Fig. 5.3 Fracture morphology of MWCNTs/geopolymer composites with different content of MWCNTs, a 0.5 wt%, b 1 wt%, c 2 wt%, d 3 wt%, e 5 wt%, with permission from [2]

5.1.3 Microstructure To evaluate the distribution of carbon nanotube in the geopolymer matrix, typical microstructures of fractographs of MWCNTs/geopolymer composites were studied, as shown in Fig. 5.3. As can be seen, when the content of MWCNTs was less than 3%, carbon nanotubes could basically evenly disperse in the composites. It was well known that the primary demerit of CNTs during the fabrication of composite was their tendency to aggregate due to the presence of van der Waals forces and smooth surfaces [3], and the higher the content of carbon nanotubes, the more likely it is to agglomerate. Therefore, when the content of MWCNTs came to 5 wt%, the carbon nanotubes showed obvious agglomeration, as indicated in Fig. 5.3e. The agglomerate phenomenon would undoubtedly weaken the strengthening and toughening effect from MWCNTs, which could be reflected by changes in mechanical properties.

5.1.4 Mechanical Properties Figure 5.4 provides the mechanical performances of geopolymer matrix and MWCNTs/geopolymer composites. With the increase in MWCNTs content, the mechanical properties increased first and then decreased after reaching the peak value. Flexural strength, elastic modulus, and fracture toughness achieved their maximum values when the content of MWCNTs reached 3 wt%, 17.5 MPa, 12.7 GPa, 0.18 MPa·m1/2 , which were 42.3, 29.6, and 38.5% higher than those of

5.1 Carbon Nanotube-reinforced Geopolymer Matrix Composites

135

Fig. 5.4 Mechanical properties of MWCNTs/geopolymer composites with different content of MWCNTs, with permission from [2]

geopolymer matrix, respectively. The performance improvements could be attributed to the strengthening and toughening effect caused by the introduction of the carbon nanotube. However, the mechanical performance declined with further increase in MWCNTs content to 5 wt%. The reason could be ascribed to the agglomeration of MWCNTs in the composites, as discussed in the above SEM results. Well-dispersed CNTs in a geopolymer matrix could improve the mechanical performance of composites, otherwise the agglomeration would reduce its strengthening and toughening effect. Allowing for the characteristics of carbon nanotubes, the mechanical performance of the MWCNTs/geopolymer composites was limited to larger extent if too much carbon nanotubes were introduced as the 5 wt% MWCNTs/geopolymer composites. Therefore, the optimal content of MWCNTs should be 3 wt% in this serial composites. In order to further evaluate the effect of curing time on mechanical properties, the mechanical performances of 3 wt% MWCNTs/geopolymer composites which underwent different curing time were compared. The microstructure of the composites did not change obviously after experiencing different curing time. In contrast, the mechanical performance had improved significantly in the early stage of curing process, as shown in Fig. 5.5. With the extension of the curing time, the performance of MWCNTs/geopolymer composites slowed down gradually. Compared with the performance after 3 days curing, the performance of the sample being cured for 28 days, including flexural strength, elastic modulus, and fracture toughness, increased by 44, 52, and 23%, respectively. The performance improvement could be attributed to the increase in the degree of geopolymerization reaction, especially during earlier stage.

136

5 Particles-Reinforced Geopolymer Matrix Composites

Fig. 5.5 Mechanical properties of 3 wt% MWCNTs/geopolymer composites with different curing time, with permission from [2]

5.2 Graphite Powder-Reinforced Geopolymer Matrix Composites 5.2.1 Preparation Process Graphite powder (purity of 99%) was obtained from Huatai Machinery Factory, China. Figure 5.6 gives the size distribution of the graphite powder and the median diameter is 4.6 μm. Preparition of metakaolin, potassium silicate solution and Fig. 5.6 Diameter distribution of graphite powder, with permission from [4]

5.2 Graphite Powder-Reinforced Geopolymer Matrix Composites

137

Table 5.3 Graphite contents and corresponding Gr/GP ratios Gr/GP ratio (mol)

0

3:1

6:1

9:1

12:1

15:1

18:1

Gr content (wt%)

0

14.3

25.0

33.3

40.0

45.5

50.0

Fig. 5.7 Samples of geopolymer matrix and Gr/geopolymer composite, with permission from [4]

geopolymer are the same to that in the aforementioned part. The composition of geopolymer matrix used here was SiO2 /Al2 O3 = 4.0, K2 O/SiO2 = 0.25, and H2 O/K2 O = 10 (mole ratio). Graphite powder was first ball-milled with metakaolin for 24 h to obtain a homogeneous dispersion. Then the obtained mixture was added to potassium silicate solution and stirred ultrasonically for 45 min. After that, the slurry was cast into polystyrene containers (70 mm in diameter and 8 mm in height), sealed, and cured at 70 °C for 7 days to accomplish the synthesis of graphite/geopolymer composites. Table 5.3 shows the composition of graphite/geopolymer composites. Figure 5.7 is the photo of pure geopolymer and Gr/geopolymer composite. The left one is white geopolymer matrix and the right one is Gr/geopolymer composite which has turned black because of the presence of graphite. Gr/geopolymer composites are further treated at 600 °C in N2 atmosphere, with heating rate of 5 °C/min and soaking time of 120 min.

5.2.2 Phase Composition The XRD patterns of geopolymer and Gr/geopolymer composites with different graphite contents are shown in Fig. 5.8. All the composites show a broad amorphous hump between 27° and 29° of 2θ , which is a typical feature for geopolymer materials. The patterns prove that the composites consist of amorphous aluminosilicate inorganic polymer and graphite, and it could be observed that with increasing graphite contents, the diffraction peak intensity of graphite also increased.

138

5 Particles-Reinforced Geopolymer Matrix Composites

Fig. 5.8 XRD patterns of geopolymer and Gr/geopolymer composites, with permission from [4]

5.2.3 Microstructure Figure 5.9 shows the fractographs of geopolymer and Gr/geopolymer composites. The composites show homogeneous microstructure when graphite content is not higher than 40%. The fractograph of geopolymer matrix is relatively flat, yet the surface of the composites becomes obviously rougher, with ridges protruding frequently, and the degree of such roughness gradually intensified with increasing graphite contents. This may be attributed to the graphite sheets inserted into the matrix

Fig. 5.9 Microstructure of Gr/geopolymer composites of different graphite contents: a and f 0, b and g 33.3%, c and h 40.0%, d and i 45.5%, c and j 50.0%, with permission from [4]

5.2 Graphite Powder-Reinforced Geopolymer Matrix Composites

139

that could ramify and deflect the main crack during crack growth, thus increasing the crack growth path and leading to such a coarse fractograph. The interfacial bonding between the graphite and the matrix seems stable, and graphite pulling-out is observed, which would enhance both the flexural strength and fracture toughness of composites. In Fig. 5.9d, e, graphite agglomerations around the fractograph should be noted. It could be inferred that a graphite content of 45.5 or 50.0% is too high for graphite to disperse into geopolymer and that this kind of agglomerations becomes almost inevitable. With such inhomogeneity, defects and cracks could be produced, jeopardizing the mechanical property and microwave-absorbing function of the composites.

5.2.4 Mechanical Properties The flexural strengths of geopolymer and Gr/geopolymer composites are compared in Fig. 5.10. With graphite content increasing, the flexural strength of composites showed an increasing trend at first, reaching a peak value of 17.4 MPa at graphite content of 40.0%, which has risen by 41.5% compared to matrix, and then declined at graphite content of 45.5 and 50.0%. This reinforcement effect could be explained by the mixing rule of composites. Meanwhile, graphite pulling-out and crack deflection can be found in Fig. 5.9, which could increase the crack tolerance of the composites, and this may also contribute to the enhancement of flexural strength. The decline of flexural strength of high-graphite composites may have resulted from the agglomeration of graphite sheets, which could produce defects and cause microstructural heterogeneity, and thus lead to a passive impact on the flexural strength of the material. 20 16 Flexual strength, MPa

Fig. 5.10 Flexural strength of Gr/geopolymer composites, with permission from [4]

12 8 4 0 0

33.3% 40.0% 45.5% Graphite contents

50.0%

140

5 Particles-Reinforced Geopolymer Matrix Composites

5.2.5 Microwave Absorption Properties The variations in complex permittivity and permeability of Gr/geopolymer composites are shown in Fig. 5.11. The variations of the real part ε and imaginary part ε of complex permittivity for Gr/geopolymer composites of different graphite contents are presented in Fig. 5.11a, b, respectively. From the figure it could be seen that the values of ε and ε for Gr/geopolymer composites are higher as compared to geopolymer matrix, and with the Gr/GP ratio increasing, both ε and ε show a rising trend. On the other hand, as shown in Fig. 5.11c, d, the values of real part μ and imaginary part μ of complex permeability for Gr/geopolymer composites have not found obvious change. The real part μ basically fluctuates around 1.1 and the imaginary part μ fluctuates around 0. Therefore, it could be inferred that the main microwave-absorbing mechanism of the Gr/geopolymer composites would be dielectric loss, which could be mainly due to the interfacial polarization and intrinsic electric dipole polarization of graphite. For a single-composite layer backed by metal plate, the reflect loss (RL) can be computed according to the theory of absorbing wall. The wave impedance of the layer is given by the formula (5.1)

Fig. 5.11 Dielectric constants and magnetic constants of the composites a ε , b ε , c μ , d μ , with permission from [4]

5.2 Graphite Powder-Reinforced Geopolymer Matrix Composites

 Z1 =

  εr 2π √ f d1 εr μr tanh j μr c

141

(5.1)

where Er and μr are the relative complex permeability and permittivity of the composite, respectively; f is the frequency of incident electromagnetic wave; d 1 is the thickness of the layer, and c is the speed of light in free space. Thus, the RL of the incident electromagnetic wave can be expressed as (5.2):    Z1 − 1    RL = 20 lg Z1 + 1 

(5.2)

It could be found that the RL of a single-composite layer is determined by six independent parameters: f , d 1 , ε , ε , μ , and μ . The RL/frequency curves of Gr/geopolymer composites with different thickness could be obtained using the data in Fig. 5.11 according to the above formulas. Figure 5.12 presents the absorption property of Gr/geopolymer composites of different Gr/GP ratios with various thicknesses. Table 5.4 shows the value of peak, peak frequency, valid wave-absorbing range (RL less than −10 dB), and corresponding layer thickness of the RL/frequency curves. It can be observed in Fig. 6 that the electromagnetic wave absorption of geopolymer matrix is weak, of which the RL is no lesser than −6 dB in the full frequency range of 2–18 GHz. With graphite introduced, all the composites showed enhanced wave-absorbing ability. The maximum attenuation of the incident wave is observed on the composite with Gr/GP ratio of 12:1 and thickness of 4.15 mm, reaching −64.8 dB at 5.1 GHz. With the increase of Gr/GP ratios, the valid wave-absorbing range kept augmenting, reaching 4.0 GHz when the Gr/GP ratio is 18:1 with thickness of 1.25 mm. However, the maximum wave reflection loss of the composites does not show such monotone increasing trend. The value of absorption peak declines when the Gr/GP ratios reach 15:1 and 18:1, which could be caused by the agglomeration of graphite sheets, as mentioned before. Figure 5.13 gives the RL/f curves of Gr/geopolymer composites after 600 °C heat treatment. Table 5.5 shows the absorption property of Gr/geopolymer composites after 600 °C heat treatment. For these post-heat treatment composites, the general RL showed similar trend alone with a slight decline compared to those that did not undergo heat treatment. The maximum wave attenuation appeared when Gr/GP ratio is 9, reaching −55.4 dB at 6.4 GHz. However, it could be observed that the composite thickness corresponding to the maximum wave reflection loss decreased significantly after heat treatment, which is very helpful for the application of GR/geopolymer composites as microwave-absorbing coating. After 600 °C heat treatment, free water and part of condensed hydroxyl groups in the composite evaporated [5, 6], leaving porous structure in the composite, thus the porosity of the composite increased. According to some researches [7, 8], a low porosity will lead to a low dielectric

142

5 Particles-Reinforced Geopolymer Matrix Composites

(b) 0

0

-20

-20

RL/dB

RL/dB

(a)

-40

1.25 2.65 2.89 4.15 4.65 4.69 4.97

-40

-60

-60

2

4

6

8

10

12

14

16

2

18

4

6

8

14

16

18

12

14

16

18

12

14

16

18

0

0

-20

RL/dB

-20

RL/dB

12

(d)

(c)

-40

-40

-60

-60

2

4

6

8

10

12

14

16

2

18

4

6

8

10

f GHz

f GHz

(f)

(e) 0

0

-20

-20

RL/dB

RL/dB

10

f GHz

f GHz

-40

-40

-60

-60

2

4

6

8

10

12

14

16

2

18

4

6

8

f GHz

10

f GHz

(g) 0

RL/dB

-20

-40

-60

2

4

6

8

10

12

14

16

18

f GHz

Fig. 5.12 RL/frequency curves of Gr/geopolymer composites with different thickness a 0, b 14.3%, c 25.0%, d 33.3%, e 40.0%, f 45.5%, g 50.0%, with permission from [4]

5.2 Graphite Powder-Reinforced Geopolymer Matrix Composites

143

Table 5.4 Absorption property of Gr/geopolymer composites of different graphite contents, with permission from [4] Graphite contents (%)

Absorb peak (dB)

f of peak (GHz)

Valid wave-absorbing range (GHz)

Thickness (mm)

0

−5.3

15.9

0

2.89

14.3

−7.8

13.9

0

2.65

25.0

−49.1

17.3

2.0

4.97

33.3

−43.6

16.2

1.6

4.65

40.0

−64.8

5.1

1.9

4.15

45.5

−26.0

4.2

3.8

4.69

50.0

−22.8

15.0

4.0

1.25

tangent loss, so the higher porosity of the composite after 600 °C treatment would contribute to a better microwave absorption performance. Besides, the loss of water will also cause a volume shrinkage, which would increase the graphite concentration spatially. Therefore, heat treatment at 600 °C could make the composite a more effective microwave-absorbing system with the reduction of the thickness corresponding to the maximum wave attenuation.

5.3 Al2 O3 Particle-Reinforced Geopolymer Matrix Composites 5.3.1 Preparation Process Preparition of metakaolin, potassium silicate solution and geopolymer are the same to that in the aforementioned part. The α-Al2 O3 powder is bought from China building Materials Research Institute, with a purity of 98% and an average particle size of 0.75 nm. The SEM photos and XRD results of the α-Al2 O3 powder are shown in Fig. 5.14. The Al2 O3p /geopolymer composite is prepared according to the process with the compositions in Table 5.6. The steps are listed as follows: (1) An appropriate amount of KOH is added to the sol-SiO2 solution and stirred magnetically at room temperature for 3 days to obtain the mixed solution. (2) Add the mixed powder of metakaolin and Al2 O3 to the mixed solution, stir with a high-shear mixer for 30 min to get the geopolymer mixture. In order to ensure that the whole process is carried out at a low constant temperature, an ice water bath is applied.

144

5 Particles-Reinforced Geopolymer Matrix Composites

(b)

(a)

0

0

1.33 1.44 1.50 2.48 2.85 3.36 3.64

-40

-60

-20

RL/dB

RL/dB

-20

-40

-60 2

4

6

8

10

12

14

16

18

2

4

6

8

10

12

14

16

18

f GHz

f GHz

(d)

(c)

0

-20

-20

RL/dB

RL/dB

0

-40

-40

-60

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

-60

18

2

4

6

8

f GHz

10

12

14

16

18

12

14

16

18

f GHz

(e)

(f) 0

0

-20

RL/dB

RL/dB

-20

-40

-60

-40

2

4

6

8

10

12

14

16

-60

18

2

4

6

8

10

f GHz

f GHz

(g) 0

RL/dB

-20

-40

-60

2

4

6

8

10

12

14

16

18

f GHz

Fig. 5.13 RL/frequency curves of Gr/geopolymer composites after 600 °C heat treatment a 0, b 14.3%, c 25.0%, d 33.3%, e 40.0%, f 45.5%, g 50.0%, with permission from [4]

5.3 Al2 O3 Particle-Reinforced Geopolymer Matrix Composites

145

Table 5.5 Electromagnetic wave-absorption property of Gr/geopolymer composites after 600 °C treatment, with permission from [4] Graphite content (%)

Absorb peak (dB)

0

f of peak (GHz)

Valid wave-absorbing range (GHz)

Thickness (mm)

3.6

14.3

0

3.36

14.3

11.8

13.3

2.8

2.85

25.0

54.8

12.5

2.2

2.48

33.3

55.4

6.4

2.3

3.64

40.0

46.9

17.5

2.5

1.33

45.5

33.0

15.3

4.1

1.50

50.0

17.0

14.7

4.0

1.44

Intensity(cps)

4000

-α -Al2O3

3000

2000

1000

0

20

40

60

80

2theta (degree)

Fig. 5.14 SEM image and XRD pattern of the raw α-Al2 O3 powder a SEM image, b XRD pattern

Table 5.6 Composition of the geopolymer composites reinforced by Al2 O3 particles Sample

KOH solution mass/g

Sol-SiO2 volume /mL

Metakaolin /g

Mass ratio of Al2 O3p to metakaolin

The volume content of Al2 O3p /vol%

Matrix

9.37

21.8

16.6

0

0

A1

9.37

21.8

16.6

1/4

4.6

A3

9.37

21.8

16.6

3/4

6.3

A5

9.37

21.8

16.6

5/4

9.2

(3) Remove bubbles in the slurry by ultrasonic vibration. (4) The mixture is poured and sealed into plastic containers, and then placed in the oven at 50 °C for curing for 7 days.

146

5 Particles-Reinforced Geopolymer Matrix Composites

Fig. 5.15 XRD analysis of raw material metakaolin and Al2 O3 -reinforced geopolymer composite

5.3.1.1

Phase Composition of Al2 O3p /Geopolymer Composites

The phase of the composites (Fig. 5.15) shows that the addition of the Al2 O3 is still α-Al2 O3 phase, which is consistent with its original state, indicating its good stability. The quartz phase is the impurity phase of the raw kaolin. The amorphous hump of the geopolymer (the center position is at 28°) is transferred from 22° to 28° after preparation of geopolymer, which is consistent with the aforementioned geopolymers.

5.3.2 Mechanical Properties Table 5.7 lists the flexural strength, elastic modulus, fracture toughness, and hardness of the composites at room temperature. The mechanical properties of the composites significantly increase with the addition of Al2 O3p , indicating that the Al2 O3p plays a good role in both strengthening and toughening. The tendency is shown in Figs. 5.16, 5.17, 5.18 and 5.19. Except for the monotonous increase of elastic modulus, other properties show the tendency of increasing first and then decreasing. Compared Table 5.7 Mechanical properties of the geopolymer matrix and composites reinforced by Al2 O3p Mechanical Flexural Elastic Fracture toughness/MPa·m1/2 Vickers hardness/MPa properties strength/MPa modulus/GPa Matrix

15.3 ± 1.4

8.3 ± 0.6

0.28 ± 0.05

284 ± 9.3

A1

23.6 ± 0.9

9.2 ± 0.2

0.58 ± 0.03

335 ± 9.1

A3

25.8 ± 1.1

9.8 ± 0.2

0.48 ± 0.01

289 ± 5.9

A5

23.0 ± 0.9

11.3 ± 0.4

0.45 ± 0.02

284 ± 6.5

5.3 Al2 O3 Particle-Reinforced Geopolymer Matrix Composites Fig. 5.16 Effect of mass ratio of Al2 O3p to metakaolin in raw powders on the flexural strength of composites

147

28 26

Flexural strength (MPa)

24 22 20 18 16 14 12

0

1/4

3/4

5/4

Mass ratio of Al2O3p/metakaolin

11

Young's Modulus (GPa)

Fig. 5.17 Effect of mass ratio of Al2 O3p to metakaolin in raw powders on elastic modulus of composites

10

9

8 0

1/4

3/4

5/4

Mass ratio of Al2O3p/metakaolin

with the matrix, the flexural strength of the composites improved by 50–70%. The maximum increase of fracture toughness is more than twice. However, the elastic modulus and Vickers hardness did not improve too much. Due to the addition of Al2 O3p , the geopolymer is strengthened. However, the fracture characteristics of the composites did not improve, and it still shows a typical brittle fracture (Fig. 5.20). From the fractograph and crack propagation characteristics (Figs. 5.21 and 5.22, respectively), the difference between the composite and the matrix after the addition of Al2 O3p is obvious, and with the addition of Al2 O3p the fractography became much coarser and crack propagation path became more zigzag, which contributed to the increase in fracture toughness.

148

0.60 0.55

Fracture toughness (MPa·m1/2)

Fig. 5.18 Effect of mass ratio of Al2 O3p to metakaolin in raw powders on fracture toughness of composites

5 Particles-Reinforced Geopolymer Matrix Composites

0.50 0.45 0.40 0.35 0.30 0.25 0.20

0

1/4

3/4

5/4

Mass ratio of Al2O3p/metakaolin

Fig. 5.19 Effect of mass ratio of Al2 O3p to metakaolin in raw powders on hardness of composites Hardness (MPa)

340

320

300

280

260

0

1/4

3/4

5/4

Mass ratio of Al2O3p/metakaolin

5.4 Chromium Powder-Reinforced Geopolymer Matrix Composites 5.4.1 Preparation Process Preparition of metakaolin, potasssium silicate and geopolymer are the same to that in the aforementioned part. The chromium powder (Crp ) is purchased from Shanghai Panyin Chemical Co., Ltd, with a purity of >99.9% and a density of 7.15 g cm−3 . Powders of 200 mesh (53.54 μm), 400 mesh (20.64 μm), and 1000 mesh (12.68 μm)

5.4 Chromium Powder-Reinforced Geopolymer Matrix Composites

149

25

20

Load (N)

A5 15

10

matrix

5

0 0.00

0.04

0.08

0.12

Displacement (mm) Fig. 5.20 Load–displacement curve of matrix and composite(A5) (a)

(b)

(c)

(d)

Fig. 5.21 SEM fractograph of the geopolymer matrix and composites reinforced with Al2 O3p matrix, b A1, c A3, d A5

150

5 Particles-Reinforced Geopolymer Matrix Composites

Crack propagation

Crack propagation

(a) matrix˗

(b) A3

Fig. 5.22 SEM observation on crack propagation path of matrix and Al2 O3p /geopolymer composite a matrix; b A3

are selected, respectively. Crp is added into the geopolymer slurry, and the mixture is stirred for another 20 min followed by molding and curing. This part reported the effects of Crp content, water addition, curing time, and curing temperature on the properties of geopolymer composites through orthogonal experiments.

5.4.2 Phase Composition Phase analysis shows that the chromium powders do not react with the geopolymer. The XRD pattern of Crp /geopolymer composite is shown in Fig. 5.23, which indicates that the Crp is stable under alkaline solution. It is a foundation for the strengthening effect of geopolymer matrix. Besides, the characteristics of XRD patterns are Fig. 5.23 XRD patterns of Crp /geopolymer composites

Ƶ--SiO2

2000

ͩ-Cr

Intensity

1500

1000

500

0

20

40

60

2 theta(degree)

80

100

5.4 Chromium Powder-Reinforced Geopolymer Matrix Composites

151

consistent with the geopolymer matrix, which shows an amorphous hump at 26.8° corresponding to the geopolymer matrix and a peak of α-SiO2 from raw metakaolin.

5.4.3 Density and Mechanical Properties Table 5.8 lists the density of the Crp /geopolymer composites prepared under different factors. Crp are selected with the total mass of 40, 50, 60, and 70%, respectively. The water content is 5, 6, 7, and 8 g, respectively. The curing time is 5, 7, 9, and 11 days, respectively. The curing temperature is 80 and 60 °C, respectively. However, 80/30 and 60/30 represent the samples curing at 80 or 60 °C for 3 h, followed by curing at 30 °C. Since the density of chromium (7.15 g/cm3 ) is significantly higher than that of geopolymer (~2 g/cm3 ), the density of composites increases with the content of Crp . In addition to the content of Crp , water addition, curing temperature, and time also have an impact on the density of composites: when the water addition is higher, the curing is more sufficient, which is more conducive to obtain a higher density. When curing at 80 °C, the density of the composites is higher than those with the same Crp content. The density of the composites cured at 60 °C is also higher than that of materials cured at 30 °C. The reason is that the high-temperature curing process can Table 5.8 Density and flexural strength of the Crp /geopolymer composites Sample Crp mass/w% Water Time/d Curing Density/g/cm3 content/g temperature/°C

Flexural strength/MPa

1

15.9 ± 1.0

5

5

80

1.86

2

40

6

7

60

1.88

20.7 ± 0.6

3

7

9

80/30

1.92

20.5 ± 0.8

8

11

60/30

2.02

20.2 ± 0.6

5

7

80/30

2.25

18.8 ± 1.5

6

6

5

60/30

1.93

18.9 ± 1.1

7

7

11

80

2.21

24.5 ± 0.5

8

8

9

60

2.22

18.7 ± 0.2

5

9

60/30

2.07

21.4 ± 0.5

10

6

11

80/30

2.05

20.1 ± 0.1

11

7

5

60

2.30

18.8 ± 0.5

12

8

7

80

2.40

21.5 ± 0.7 24.3 ± 0.7

4 5

9

13

50

60

5

11

60

2.23

14

70

6

9

80

2.47

22.0 ± 0.5

15

7

7

60/30

2.13

20.3 ± 0.4

16

8

5

80/30

2.16

21.8 ± 0.2

152

(a)

5 Particles-Reinforced Geopolymer Matrix Composites

(b) Crp

Geopolymer

Fig. 5.24 SEM observations of the surface of the Crp /geopolymer composite

improve the activity of metakaolin, which is beneficial for the geopolymerization. Increasing the curing time is also beneficial for the pore exclusion and hydration reactions, which can make the material denser. Figure 5.24 is the observation of the surface of the Crp /geopolymer composites. Under a lower magnification (Fig. 5.24a), the Crp is evenly distributed in the geopolymer matrix, and there are some pores in the matrix. Under a higher magnification (Fig. 5.24b), the interface between the Crp and the geopolymer matrix is observed, which shows that the combination is good without obvious defects and pores. The flexural strength of the Crp /geopolymer composites with various factors is also shown in Table 5.8. The influence of various factors on its mechanical properties is also shown in Fig. 5.25. The flexure strength of the composites increases first with the amount of water, and then decreases when the amount of water is 7 g. The hydration process of the geopolymer is as follows: aluminum and silicon oxide particles in metakaolin are dissolved and polymerized in alkaline solution (K2 SiO3 •xH2 O) to form a gel phase [Kx (AlO2 )Y •nKOH•mH2 O], which is finally condensed and hardened to remove the remaining water. The whole process is an exothermic dehydration process. When the water content is low, the flow of slurry is poor, and the entrapped air cannot be well exhausted, resulting in the high porosity of the composites. However, when the water content is too high, the geopolymerization rate becomes slow, which is not beneficial for geopolymer forming. Besides, increasing curing time and curing temperature are also beneficial to the improvement of the flexural strength of the composites, which is consistent with the density change of the composites. Figure 5.26 shows the fractograph of the Crp /geopolymer composites. It can also be noted that there are still defects in the geopolymer matrix, which consist of the observation results on the surface of samples. Besides, Crp pull-out is noted. In the high-magnification SEM observation (Fig. 5.26b), there are still some defects such as microcracks in the matrix, which are detrimental to the mechanical properties of the composites.

5.4 Chromium Powder-Reinforced Geopolymer Matrix Composites

153

Fig. 5.25 Effects of various factors on mechanical properties of the Crp /geopolymer composites a Crp content, b water addition, c curing time, d curing temperature

(a)

(b)

Fig. 5.26 SEM fractograph of the Crp /geopolymer composites. a low magnification, b high magnification

154

5 Particles-Reinforced Geopolymer Matrix Composites

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites The fly ash cenosphere (FAC) has thin hollow walls, low volume density (0.4– 0.8 g/cm3 ), low thermal conductivity (about 0.065 w/(mK)), high fire resistance (>1500 °C), high strength, good chemical stability, low cost, and other advantages, which makes it to be one of the ideal additive materials for preparing heat insulation, sound absorption, and high temperature resistance materials. Wang Meirong (2011) added it into geopolymer to produce the cenosphere/geopolymer composites containing different volume fractions and sizes of cenosphere, and discussed the volume density, microstructure, interface, mechanical properties, thermal conductivity, and thermal dimensional stability, respectively.

5.5.1 Preparation Process The preparation of metakaolin, potassium silicate solution, and geopolymer is the same to the aforementioned part. Fly ash cenosphere (FAC) (Pindingshan Yaodian fly ash Co., China) was used as the starting material. Except for α-quartz and mullite, the main phase of FAC is amorphous (Fig. 5.27). Its chemical composition and parameters are listed in Tables 5.9 and 5.10, respectively. In addition to Si, O, and Al elements, there are also Fe, Ti, Ca, and K elements in the FACs. From the SEM morphology of the FAC (Fig. 5.28), the surface of most FAC is smooth and dense. Fig. 5.27 XRD pattern of fly ash cenospheres [9]

Table 5.9 Chemical compositions of fly ash cenosphere (FAC) [9] Composition

SiO2

Al2 O3

Fe2 O3

TiO2

CaO

K2 O

Balance

Amount (wt.%)

62.212

29.704

3.531

1.201

0.902

1.702

0.749

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites Table 5.10 Physical properties of FAC [9]

Properties Stacking

density/kg·m–3

155 Performance index 367 ~ 432

Compressive strength/MPa

10.3

Vickers hardness/MPa

895

Melting point/K

1900

Ignition loss /%

1.2

Thermal conductivity /W/(m·K) 373 K

0.065

473 K

0.082

773 K

0.123

1273 K

0.163

1573 K

0.173

a) appearance of fly ash cenospheres; (b) section of fly ash cenospheres

Fig. 5.28 SEM microscopic morphology of fly ash cenospheres from Pindingshan Yaodian Fly Ash Co., China [9] a Appearance of fly ash cenospheres; b section of fly ash cenospheres

However, there are many bubbles of different sizes on the hole wall. Effects of FAC content and sizes on the microstructure and properties of FAC/geopolymer composites were reported in this part. The average diameter of FAC is about 306 μm, and volume contents of FAC in the FAC/geopolymer composites are 15, 20, 25, 35, and 40%, respectively. FACs with average diameters of about 131, 180, 215, 246, and 306 μm, wall thickness of about 6.2, 6.8, 7.4, 7.8, and 8.3 μm, and volume density of 0.432, 0.394, 0.381, 0.378, and 0.367 g·cm–3 are obtained, respectively. The volume content of FAC is 35 vol%.

156

5 Particles-Reinforced Geopolymer Matrix Composites

Geopolymer slurry with theoretical composition of SiO2 /Al2 O3 = 4.0, K2 O/Al2 O3 = 1.0, and H2 O/K2 O = 11.0 (molar ratio) was obtained by mixing the as-calcined metakaolin powder with potassium silicate solution, and FAC was added into the asprepared geopolymer slurry and cured at 80 °C for 6 days to get the FAC/geopolymer composites.

5.5.2 Microstructure (1) FAC/geopolymer composites with different FAC contents Figure 5.29 shows the bulk density of the FAC/geopolymer composites with different volume contents of FAC. The bulk density of the FAC/geopolymer composites shows a significant linear decline tendency. When the content of FAC is 40 vol%, the bulk density of the composites drops to 0.82 g·cm–3 , which is about 60% of that of matrix 1.37 g·cm–3 . The effect of FAC on reducing the bulk density of the composites is due to its hollow structure. Figures 5.30 and 5.31 provide the SEM images of surface and fractograph of the FAC/geopolymer composites with different FAC content, respectively. When the FAC addition varies from 15 to 40 vol%, it is distributed within the matrix. Besides, many microcracks are observed in the matrix, from both surface and fractograph. The FAC is rarely pulled out and the fracture is smooth. The main crack tends to directly pass through the FAC when it meets the FAC, indicating that the interface between the FAC and matrix is strong. High-magnification SEM images are conducted on the FAC/geopolymer composites on the fractograph of FAC (Fig. 5.32a). The interface between FAC shell wall and matrix is clear and the combination is in good state. EDS analysis of element distribution at the interface (Fig. 5.32b) shows that element K drops from about 11% Fig. 5.29 Bulk density of the composites versus FAC content [9]

1.5

Bulk density, g cm

-3

1.4 1.3 1.2 1.1 1.0 0.9 0.8 0

10 20 30 Volume fraction of FAC, %

40

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites

157

(b)

(a)

400 µm

(c)

400 µm

(d)

400 µm

400 µm

(f)

(e)

400 µm

400 µm

(a) 15 vol.%, (b) 20 vol.%, (c) 25 vol.%, (d) 30vol.%, (e) 35 vol.%, (f) 40 vol.% Fig. 5.30 SEM micrographs of the FAC/geopolymer composites with various volume content of FAC [9] a 15 vol%, b 20 vol%, c 25 vol%, d 30 vol%, e 35 vol%, f 40 vol%

158

5 Particles-Reinforced Geopolymer Matrix Composites

(a)

(b)

Region A

(c)

(d)

(e)

(f)

Fig. 5.31 SEM fractograph of the FAC/geopolymer composites with various volume contents of FAC a 15 vol%, b 20 vol%, c 25 vol%, d 30 vol%, e 35 vol%, f 40 vol% [9]

of the matrix to about 1% of the shell wall of FAC, which indicates that element K in alkaline-activated solution does not enter into the FACs, and confirms that there is no obvious chemical reaction between the FAC and the geopolymer matrix. TEM analysis was carried out on all specimens to further investigate the interface between the geopolymer and the FAC. Figure 5.33 represents a TEM micrograph of a FAC/MK-based geopolymeric composite specimen. An interfacial layer of about 100 nm in thickness was clearly observed between the geopolymer matrix and the FAC, as indicated in Fig. 5.33a, b. Figure 5.33c demonstrates the high-resolution transmission electron micrograph of interfacial layer and geopolymer matrix. The

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites

159

Fig. 5.32 SEM fractographs of the FAC/geopolymer composite (a) and linear scanning spectrum of K (b) [9]

typical amorphous halos (left inset in Fig. 5.33c) implied that the interface layer was amorphous. The geopolymer matrix (right inset in Fig. 5.33c) was a glassy phase. However, a few scattered lattice fringes with a size of 2–5 nm were found in the matrix. The ordered regions by the fast Fourier transformation (FFT) filtered images determined that the lattice constant was 0.212 nm, which corresponded to the α-quartz phase. Energy dispersive spectrum (EDS) analysis (Fig. 5.33d–f) showed that the composition of the interfacial layer was Si19.33 Al9.67 O69.96 K1.02 , which was different from those of geopolymer matrix (Si21.08 Al10.59 O57.85 K10.49 ) and FAC (Si26.47 Al6.90 O65.29 K0.44 ). These results proved unambiguously that the interfacial layer was formed by the diffusion of elements rather than reactions between the FAC and the MK-based geopolymer. It could be concluded that the geopolymer, interfacial layer, and FAC bonded very well to each other. (2) FAC/geopolymer composites with different FAC size Figure 5.34 shows the bulk density of the composites with different sizes of FACs. When the size of FACs increases, the bulk density of the composites decreases linearly from 1.15 to 0.99 g·cm–3 , from the result of the increased porosity with FAC size. Figure 5.35a–e shows the microstructure of the FAC/geopolymer composites with a volume content of FAC of 35 vol% and an average size of 131, 180, 215, 246, and 306 μm, respectively. No matter the size of the FAC, FAC distributed in the matrix uniformly. However, the quantity of FAC per unit area decreased due to its volume content in the composites. From Fig. 5.35f, the interface between the matrix and the shell wall of FAC is clear, which confirms the stable existence of FAC under the condition of alkaline solution. An amorphous boundary layer with a thickness of about 300 nm is also observed between the FAC and the matrix (Fig. 5.36), and the interfacial layer is well combined with the matrix. According to the EDS analysis of the interface layer, the composition of the non-crystalline compounds in the interfacial layer is KSi19 Al9.9 Fe2.2 O69.9 (at.%).

160

5 Particles-Reinforced Geopolymer Matrix Composites

(d) 5000

O

Intensity, cps

4000 3000

Si

2000

Al

1000 K 0 0

2

3 4 Energy, keV

5

6

5

6

(f)

(e) 5000

5000

O

4000

Si

3000 Al

2000

O

4000 Intensity, cps

Intensity, cps

1

K

1000

3000 Si

2000 1000

Al

K

0 0

K

0

1

2

3

4

Energy, keV

5

6

0

1

2

3

4

Energy, keV

Fig. 5.33 TEM analysis of the FAC/geopolymer composite. a The interfacial layer between the FAC and the geopolymer matrix, b Magnification micrograph of Region A, c HRTEM of the interfacial layer and geopolymer matrix, d EDS of interfacial layer, e EDS of geopolymer matrix, f EDS of FAC, with permission from [10]

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites

161

Fig. 5.34 Effect of FAC with various sizes on the FAC/geopolymer composites [9]

5.5.3 Mechanical and Thermal Properties (1) FAC/geopolymer composites with different FAC contents The flexural and compressive strengths of the composites (Fig. 5.37) exhibit a decreasing trend with the volume content of FACs in the range of 15–40 vol%. Specifically, due to the low equivalent strength caused by the FACs, flexural strength of the composites decreases from 28.5 MPa (matrix) to 5.6 MPa (with FACs of 40 vol%). The compressive strength of the composites reduced from 106.2 to 36.5 MPa. Nevertheless, this compressive strength of the composites can still meet the requirements of thermal insulation materials. With the presence of FACs, the thermal conductivity of the composites (Fig. 5.38) also manifests a decreasing trend from 0.361 W/(m K) (matrix) to 0.173 W/(m K) (FAC: 40 vol%) with the increase in FAC volume content, which shows a decline of more than 50%. The results also demonstrate that the addition of FACs of low thermal conductivity and low volume density can effectively reduce the thermal conductivity of geopolymer composites. According to the work by Duxson et al., when the diameter of the pore is greater than 1 cm, the effect of convection in the pore is very important. Thermal radiation also contributes to the continuous materials for large particles when the temperatures are >200 °C. In this part, the inner diameter of FACs used is about 160 μm, and the testing environment is room temperature. So, convection and thermal radiation can be ignored, and only heat conduction should be considered. Therefore, the low thermal conductivity of the FAC/geopolymer composites is mainly attributed to the FACs with low thermal conductivity. Besides, the thermal resistance at the FAC/matrix interface also contributed to the reduction of the thermal conductivity. Figure 5.39 shows the effect of FACs with different volume contents on the thermal shrinkage of the FAC/geopolymer composites. The thermal shrinkage of the composites with the addition of FACs is like that of the geopolymer. When

162

5 Particles-Reinforced Geopolymer Matrix Composites

(a)

(b)

500 µm

500 µm (c)

(d)

500 µm

500 µm (e)

(f)

500 µm

5 µm

Fig. 5.35 SEM micrographs of the 35 vol% FAC/geopolymer composites with various sizes of FAC: a matrix, b 180 μm, c 360 μm, d micrograph of interfacial layer between the matrix and the FAC

the volume content of FACs increases from 20 to 50 vol%, the thermal shrinkage of the composites decreases gradually, indicating FACs inhibit thermal contraction from geopolymer upon heating. The decrease in thermal contraction of the FAC/geopolymer composites is beneficial to improve their dimensional stability. (2) FAC/geopolymer composites with different FAC contents The influence of FAC size on the flexural strength of the FAC/geopolymer composites is depicted in Fig. 5.40a. As indicated, with the increase of the size of FACs, the flexural strength of the composites decreased from 14.8 MPa (the average diameter of FACs is 131 μm) to 6.4 MPa (the average diameter of FACs is 306 μm) in a

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites

163

5000

(c)

O Intensity, cps

4000 3000 Si 2000 Al 1000 K

Fe

0 0

1

2

3

4

5 6 7 Energy, keV

8

9

10

Fig. 5.36 TEM analysis of FAC/geopolymer composite a The interfacial layer between the FAC and the geopolymer matrix, b HRTEM of the interfacial layer, c EDS corresponding to interfacial layer [9] 120 Compressive strength, MPa

Flexural strength, MPa

30

20

10

100 80 60 40 20

0

0

10 20 30 Volume fraction of FAC, %

40

0

10 20 30 Volume fraction of FAC, %

40

Fig. 5.37 Relationship between strength of composites and volume content of FACs. a Flexural strength, b compressive strength, with permission from [10]

164

5 Particles-Reinforced Geopolymer Matrix Composites

-1

Thermal conductivity, W m K

-1

0.4

0.3

0.2

0.1

0

10 20 30 Volume fraction of FAC, %

40

Thermal shrinkage, dL/L0

Fig. 5.38 Thermal conductivity of the composites versus volume fraction of FAC, with permission from [10]

50% FAC 40% FAC 30% FAC 20% FAC Matrix

0.0

-0.1

-0.2 0

200

400

600

800

1000

o

Temperature, C

Fig. 5.39 Effect of FAC with various volume contents on the thermal shrinkage of the FAC/geopolymer composites. The arrow indicates increasing volume content of FAC [9]

very good linear way. The increase of FAC size is not beneficial to the improvement of strength of the composites. Similarly, with the increase of the average diameter of FACs, the compressive strength of composites also shows decreasing tendency (Fig. 5.40b). When the average diameter of FACs increases from 131 to 306 μm, the compressive strength of the composites gradually declined from 58.9 to 16.2 MPa with a drop of 72.5%. Figure 5.41 shows the thermal conductivity of the FAC/geopolymer composites versus sizes of FACs with 35 vol.%. When the average size of the FACs increases, the thermal conductivity of the composites near linearly decreases from 0.22 to

5.5 Fly Ash Cenosphere-Reinforced Geopolymer Matrix Composites 24 18

20

12

15

6

10

0 -6

5 0 100

-12

150

200 250 Diameter of FAC,

300

80

Compressive strength, MPa

Flexural strength Reciprocal volume percent

25

Reciprocal volume percent, %

Flexural strength, MPa

30

165

60

40

20

0 100

350

150

m

(a) flexural strength

200 250 300 Diameter of FAC, m

350

(b) compressive strength

Fig. 5.40 Effect of FAC with various sizes on the mechanical properties of the 35 vol% FAC/geopolymer composites [9] a flexural strength; b compressive strength

-1

0.25

-1

Thermal conductivity, W m K

Fig. 5.41 Effect of FAC with various sizes on the thermal conductivity of the 35 vol% FAC/geopolymer composites [9]

0.20

0.15 100

150

200

250

Diameter of FAC,

300

350

m

0.178 W m–1 K–1 , which is mainly due to the increase in the volume of the hollow part inside the FAC (the air or inert gas introduced by the hollow part of the FAC). Figure 5.42 exhibits the influence of FAC sizes on the thermal shrinkage of the FAC/geopolymer composites. The presence of FACs significantly inhibits the thermal shrinkage of the matrix during the whole temperature range (1), as shown in formula (1), which consumed alkalinity of the solution. For reaction between metakaolin and alkaline solution, first was the dissolution of metakaolin including releasing of silicate and aluminate species in the solution of high alkalinity, with the breakage of the Si–O–Al and Si–O–Si bonds; and then the reaction was between these released species, as shown in formula (2). Due to low reactivity of metakaolin compared with fused silica, most alkaline reacted with fused silica, with unreacted metakaolin left in the final geopolymer product, which were consistent with the aforementioned XRD, FT-IR, and NMR results.

170

5 Particles-Reinforced Geopolymer Matrix Composites

K2 O · SiO2 · xH2 O(l) + (n − 1)SiO2 (g) → K2 O · nSiO2 · yH2 O(l)

(5.3)

[Al(OH)4 ]− + [SiO(OH)3 ]− → [(OH)3 Al - SiO(OH)2 ]2− + H2 O

(5.4)

5.6.4 Microstructure Figure 5.46 shows the microstructure of G2 and G3. According to the above analyses, geopolymers are known to contain amounts of unreacted solid metakaolin. In G2 the voids were produced during the polishing process as the soft, plate-like metakaolin particles remaining unreacted were torn from the binder phase. However, G3 showed a much denser microstructure although content of residual metakaolin in it was much higher than G2. It can be explained by comparing the strength of binder phase in G2 and G3. G3 also indicated multiphase microstructure as observed in the BSE micrograph (Fig. 5.46c). It was noted that most silica dissolved in the alkaline solution, but large silica particles only partially dissolved. To get further information, element distributions were recorded in Fig. 5.46d. As shown in Fig. 5.46d, the diffusion of K and Al atoms into silica took place in the binder phase/silica interface zone. EDS analyses on areas A, B, and C proved that the core area (dark area) corresponded to silica. Therefore, it can be deduced that the actual SiO2 content in G3 was slightly lower than the designed value of 3 due to the residual silica. Meanwhile, the strength of silica was much higher than that of geopolymer, which contributed to the enhancement in mechanical properties considering their strong interfacial bonding strength through chemical reaction.

5.6.5 Mechanical Properties With the increase in SiO2 content, both flexural strength and Young’s modulus increased simultaneously, as shown in Fig. 5.47. G2 showed relatively low flexural strength of 13.5 MPa, as high as that of traditional Portland cement. However, the flexural strength of G4 was 84.3 MPa, which was comparable with some glass ceramics such as silica, spodumene, cordierite, feldspar, among others. Young’s modulus also showed similar trends to flexural strength. Much higher mechanical performance of G4 than those of G2 should be attributed to the following two factors. On the one hand, higher SiO2 content resulted in higher amount of Si–O–Si bonds in G4, and it was well known that Si–O–Si bonds were stronger than Si–O–Al bonds. On the

5.6 SiO2 Particle-Reinforced Geopolymer Matrix Composites

171

Fig. 5.46 Microstructure of G2 and G3: SEM images of a G2 and b G3, c BSE image of G3, d linear EDS analysis in the interface area in c. (A), (B), and (C) correspond to the EDS analysis of areas A, B, and C in (c), with permission from [11]

other hand, residual silica particles played the role of reinforcement due to the strong interface bonding between binder phase and silica. Figure 5.48 indicates the crack propagation path in G4 and the cracks resulted from capillary stress when G4 was immersed in deionized water for 1 min. It was evident that when cracks encountered residual silica they were deflected and bridged, leading to a crack propagation path in zigzag shape, which greatly extended the effective crack length and absorbed more fracture energy with an improvement of

5 Particles-Reinforced Geopolymer Matrix Composites 100

40

Flexural strength (MPa)

Flexural strength Young's modulus

80

30 20

60

10

40

Young's modulus (GPa)

172

0 20 -10 0

2.0

2.5

3.0 Si/Al ratio

3.5

4.0

Fig. 5.47 Flexural strength and Young’s modulus of geopolymer with different SiO2 contents, with permission from [11]

Fig. 5.48 Crack propagation path in G4, with permission from [11]

fracture resistance. Figure 5.49 gives the SEM fractographs of G2 and G4 after threepoint bending test. With respect to G2, G4 showed much denser microstructure with residual silica homogeneously dispersed, which also contributed to the increase in its mechanical properties.

5.6.6 Chemical Stability Figure 5.50 provides the XRD patterns of G2 and G3 after being placed in air for 30 days. It was noted that compared with their respective phase compositions as shown in Fig. 5.5.43, G2 kept unchangeable, but G3 showed great difference with significant refraction peaks corresponding to KHCO3 caused by efflorescence. Visible surface efflorescence was observed on the surface of G4 sample, as shown in Fig. 5.51a, b. All KGP-II samples showed similar efflorescence phenomenon.

5.6 SiO2 Particle-Reinforced Geopolymer Matrix Composites

Fig. 5.49 Typical SEM fractographs of G2 (a–b), and G4 (c–d), with permission from [11] Fig. 5.50 XRD patterns of the surface area of the G2 and G3 placed in air for 30 days, with permission from [11]

173

174

5 Particles-Reinforced Geopolymer Matrix Composites

Fig. 5.51 SEM and EDS analysis of the G3 after being placed in air for 30 days a Low magnification, b high magnification, c energy dispersive analysis of area B, with permission from [11]

However, no efflorescence was observed in KGP-I samples. The presence of efflorescence in KGP-II samples was clear evidence that excess potassium was used to create the geopolymer. EDS spectra of the efflorescence is shown in Fig. 5.51c. The spectra also strongly suggested the presence of potassium carbonate, which was consistent with the XRD results in Fig. 5.50. The reason for this behavior is believed to arise from residual metakaolin. The present work suggests that the presence of SiO2 has significant effect on the geopolymerization reaction between alkaline solution and metakaolin. It has been widely accepted that geopolymer consists of cross-linked AlO4 − and SiO4 tetrahedral, where charge balance from AlO4 − is provided by hydrated alkali metal cations. According to the aforementioned FT-IR and NMR analyses, silica would prevent dissolution of the aluminate and silicate species from metakaolin by the activating alkali due to their different reactivity. Residual metakaolin increased with the SiO2 contents, which meant that AlO4 - in geopolymer with higher SiO2 content was much

5.6 SiO2 Particle-Reinforced Geopolymer Matrix Composites

175

lower than that with lower SiO2 content. Therefore, free K+ also increased with SiO2 contents. Consequently, the excess alkali was able to migrate to the surface of the geopolymer during curing which then crystallized as potassium hydrate carbonate. Chemical stability would influence the service of geopolymer materials and should be paid attention to before real application.

5.7 Summary 1.1.1.1. The mechanical performance of geopolymer could be improved with the presence of MWCNTs, and reached the peak value when the MWCNTs content was 3 wt%. The mechanical performance was closely related with the dispersion of carbon nanotube. Carbon nanotubes can be uniformly dispersed in the geopolymer matrix when its content was below 3 wt%. The agglomeration of carbon nanotube became more obvious when content rose further, which decreased the mechanical performance of MWCNTs/geopolymer composites. 2.2.2.2. The presence of graphite particle is also beneficial to the increase in mechanical performance of geopolymer. The flexural strength rose when the graphite content was not higher than 40.0% and then started to decline. The reinforcement effect could be explained by mixing rule of composites, graphite pulling-out and crack deflection, while the fall on flexural strength of the high graphite composites should be attributed to agglomeration of graphite sheets. With graphite contents increasing, dielectric constants of the composite increased gradually while the magnetic constants stayed almost unchanged, and dielectric losses were the main microwave-absorbing mechanism of the composites. Similar to the flexural strength, a trend of first rising then falling was observed on the maximum wave reflection loss, for which the passive effect of graphite sheets agglomeration should also be blamed. 3.3.3.3. Crp can effectively strengthen the geopolymer composites. The flexural strength of the Crp /geopolymer composites increased with the addition of Crp . When adding water in the mixing process, the flexural strength of the composites rose first and then decreased. The increase of curing time and temperature was also beneficial to the increase of density and flexural strength of the composites. 4.4.4.4. Al2 O3p was stable in the geopolymer composites. The addition of Al2 O3p improved the flexural strength and fracture toughness. The strengthening mainly depended on the mass ratio of Al2 O3p to metakaolin in raw powder. But the reinforced composites still showed brittle fracture.

176

5 Particles-Reinforced Geopolymer Matrix Composites

5.5.5.5. The FAC/geopolymer composites with low density, low thermal conductivity, good heat resistance, and high strength can be obtained by adding FACs into geopolymer matrix. An interface amorphous layer of about 100– 300 nm in thickness was formed between the FACs and the matrix, which was beneficial to increase the thermal resistance of the interface. When FAC content increased (15–40 vol%), both density and thermal conductivity showed decreasing trends, and the composites showed much better thermal stability. 6.6.6.6. Fused silica (SiO2 ) can partially react with geopolymer matrix to increase Si/Al ratio and partially played the role of reinforcement. With the increase in SiO2 contents, SiO2 /geopolymer composites showed much higher mechanical properties than pure geopolymer, which was due to the increased Si–O–Si bonds and residual silica as reinforcement. Geopolymer with higher SiO2 content showed worse chemical stability in air than those with lower SiO2 content, with the presence of efflorescence on the surface, which was attributed to their higher residual-free K+ .

References 1. I. Ozer, S. Soyer-Uzun, Relations between the structural characteristics and compressive strength in metakaolin based geopolymers with different molar Si/Al ratios. Ceram. Int. 41, 10192–10198 (2015) 2. Jingkun Yuan, Peigang He, Dechang Jia, Shuai Fu, Yao Zhang, Xuzhao Liu, Delong Cai, Zhihua Yang, Xiaoming Duan, Shengjin Wang, Yu Zhou, In situ processing of MWCNTs/leucite composites through geopolymer precursor. J. Eur. Ceramic Soc. 37 (2017) 2219–2226 3. S.-O. Lee, S.-H. Choi, Soo Han Kwon, Kyong-Yop Rhee, Soo-Jin Park, Modification of surface functionality of multi-walled carbon nanotubes on fracture toughness of basalt fiber-reinforced composites. Compos. B Eng. 79, 47–52 (2015) 4. Y. Zhang, P. He, J. Yuan, C. Yang, Yu. Dechang Jia, Zhou, , Effects of graphite on the mechanical and microwave absorption properties of geopolymer based composites. Ceram. Int. 43, 2325– 2332 (2017) 5. V.F.F. Barbosa, K.J.D. MacKenzie, Thermal behaviour of inorganic geopolymers and composites derived from sodium polysialate. Mater. Res. Bull. 38, 319–331 (2003) 6. Peigang He, Dechang Jia, Tiesong Lin, Meirong Wang, Yu Zhou, Effects of high-temperature heat treatment on the mechanical properties of unidirectional carbon fiber reinforced geopolymer composites. Ceramics Int. 36 (2010), 1447–1453 7. D.H. Kuo, C.C. Chang, T.Y. Su, W.K. Wang, B.Y. Lin, Dielectric properties of three ceramic/epoxy composites. Mater. Chem. Phys. 85, 201–206 (2004) 8. S.J. Penn, N.M. Alford, A. Templeton, X.R. Wang, M.S. Xu, M. Reece, K. Schrapel, Effect of porosity and grain size on the microwave dielectric properties of sintered alumina. J. Am. Ceram. Soc. 80, 1885–1888 (1997) 9. Meirong Wang, Geopolymerization mechanism of aluminosilicate geopolymer and microstructure and properties of fly ash cenosphere/geopolymer composite (in Chinese), Harbin Institute of Technology, 2011

References

177

10. Mei-Rong Wang, De-Chang Jia, Pei-Gang He, Yu Zhou, Microstructural and mechanical characterization of fly ash cenosphere/metakaolin-based geopolymeric composites. Ceramics Int. 37(2011), 1661–1666 11. Peigang He, Meirong Wang, Shuai Fu, Dechang Jia, Shu Yan, Jingkun Yuan, Jiahuan Xu, Pengfei Wang, Yu Zhou, Effects of Si/Al ratio on the structure and properties of metakaolin based geopolymer. Ceramics Int. 42(2016), 14416–14422

Chapter 6

Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Abstract Short fibers are effective reinforcements in strengthening and toughening geopolymer materials. In this chapter, random-dispersed short carbon fiber (fiber length ≤1 mm) or sheet-like carbon fiber preform (fiber length = 2, 7, 12 mm) was developed as starting materials and are used to prepare geopolymer matrix composites. Mechanical properties, fracture behavior, microstructure, and toughening mechanisms of the as-prepared composites were investigated by three-point bending test, optical microscope, and scanning electron microscopy. Effects of fiber surface treatment and high-temperature treatment on the mechanical properties of the composites were also studied. The results show that Csf /geopolymer composites exhibit apparently improved mechanical properties and an obvious non-catastrophic failure behavior. The predominant strengthening and toughening mechanisms are attributed to the apparent fiber bridging and pulling-out effect based on the weak fiber/matrix interface.

The geopolymer resin used in this part has a composition of SiO2 /Al2 O3 = 4, K2 O/SiO2 = 0.3, and H2 O/K2 O = 11 (mole ratio). Kaolin powder was calcined at 800 °C for 2 h to obtain metakaolin powder with amorphous structure. A typical processing route for geopolymer resin is as follows: a potassium silicate solution was made by dissolving silica sol into a KOH solution with a magnetic stirrer. The metakaolin powder was then added to the potassium silicate solution and mixed for 30 min with a high-shear mixer. TX-3 carbon fiber is bought from Jilin Carbon Indus., China. The main parameters are shown in Table 6.1. The surface morphology of carbon fiber is shown in Fig. 6.1. For random-Csf /geopolymer composites, the length of the starting carbon fibers is within 1 mm. The chopped short carbon fiber is mixed with metakaolin through ball-milling with acetone first. Then, the acetone was evaporated to get short fiber/metakaolin mixture with different fibers content. The mixture was used for preparation of random-Csf /geopolymer composites. For 2D-Csf /geopolymer composites, starting short carbon fibers used for preparing preforms have a diameter of 6–8 μm and an average length of 2, 7, and 12 mm, respectively. They were first separated by an ultrasonic vibrator in ethanol, then filtered out by a wire sieve to get sheet-like short carbon fiber preforms (as shown © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_6

179

180

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Table 6.1 Parameters of TX-3 carbon fiber Type

Size

Volume density (g/cm3 )

Carbon content (%)

Monofilament diameter (μm)

Tensile strength (GPa)

Elastic modulus (GPa)

Elongation (%)

TX-3

1K

1.76

≥92

6–7

≥3.0

210–230

≥1.4

Fig. 6.1 SEM surface morphology of carbon fiber

Fig. 6.2 Optical images of as-prepared sheet-like short carbon fiber preforms with a 2 mm, b 7 mm, and c 12 mm starting short carbon fibers in length, respectively, with permission from [1]

in Fig. 6.2) with a thickness in the range of 0.15–0.2 mm. The as-prepared preforms were impregnated with geopolymer resin in a plastic container and laid together one by one to get a stack consisting of 30-layer preforms. In order to avoid the formation

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

181

Fig. 6.3 Fabrication process of short carbon fiber reinforced geopolymer matrix composites, with permission from [1]

of small pores between the sheet preforms during the lamination, a vacuum-bag technique was used at 80 °C for 48 h, followed by curing at 120 °C for 24 h. The volume fraction of short carbon fibers in the as-prepared composites is about 3.5, 4.5, 6, and 7.5 vol%. The fabrication process of 2D-Csf /geopolymer composites is illustrated in Fig. 6.3.

6.1 Random-Csf /Geopolymer Composites The composition of the random-Csf /geopolymer composites is provided in Table 6.2.

6.1.1 Phase Composition As shown in Fig. 6.4, the XRD pattern of composite (1C) has typical amorphous structure, which contains the amorphous hump of geopolymer matrix and peaks of a small amount of impurity quartz phase. Carbon fiber is not shown in the XRD because of its low content and amorphous state. Table 6.2 Composition ratio of the random-Csf /geopolymer composites Type

KOH solution (g)

SiO2 volume (mL)

Metakaolin mass (g)

Csf (length ~1 mm) volume content (%)

05C

9.37

21.8

16.6

0.5

1C

9.37

21.8

16.6

1.0

2C

9.37

21.8

16.6

2.0

3C

9.37

21.8

16.6

3.0

182

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.4 XRD pattern of the 1.0 vol% random-Csf /geopolymer composites (1C)

6.1.2 Microstructure Figure 6.5 shows the microstructure of random-Csf /geopolymer composites with different fiber contents. The distribution of carbon fiber in the matrix is relatively uniform, indicating that the preparation method results in effective fiber dispersion when fiber length is within 1 mm. There are also some obvious microcracks in the geopolymer system, which showed an increasing trend with fiber contents. These cracks might be due to the thermal mismatch between fiber and matrix because of matrix shrinkage during the curing process.

6.1.3 Mechanical Properties Table 6.3 lists the mechanical properties of the geopolymer matrix and randomCsf /geopolymer composites. The curves of flexural strength, fracture toughness, and elastic modulus with the content of Csf are given in Fig. 6.6. Both flexural strength and fracture toughness of random-Csf /geopolymer composites decrease first and then increase sharply, and decrease slowly after reaching the maximum values. When the fiber content is 2.0%, the flexural strength and fracture toughness of the composites reached the highest values, 25.6 MPa and 0.57 MPa m1/2 , respectively, which were 67.3 and 103.6% higher than those of neat geopolymer. However, both Young’s modulus and Vickers hardness of the composites showed deceasing tendency, when compared to the pure geopolymer, which might be due to the low fiber content and short fiber length. When the fiber content is 0.5 vol%, the strength and toughness of the composites are the lowest, which may be due to the fiber content lower than the critical volume

(d)

(c)

Fig. 6.5 Microstructure of the random-Csf /geopolymer composites with different short carbon fiber content a 05C; b 1C; c 2C; d 3C

(b)

(a)

6.1 Random-Csf /Geopolymer Composites 183

184

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Table 6.3 Mechanical properties of the random-Csf /geopolymer composites Mechanical properties

Flexural strength (MPa)

Fracture toughness (MPa m1/2 )

Elastic modulus (GPa)

Vickers hardness (MPa)

Matrix

15.3 ± 1.4

0.28 ± 0.05

8.3 ± 0.6

284 ± 9.3

05C

11.3 ± 1.2

0.24 ± 0.04

4.7 ± 0.3

271 ± 7.8

1C

14.8 ± 0.7

0.29 ± 0.05

5.6 ± 0.1

262 ± 8.5

2C

25.6 ± 1.6

0.57 ± 0.03

6.0 ± 0.4

245 ± 6.9

3C

22.1 ± 1.3

0.56 ± 0.03

4.2 ± 0.3

243 ± 5.5

1/2 (MPa·m

(MPa)

)

0.7

25

20

0.6 0.5 0.4 0.3

15

0.2 10 0.0

0.5

1.0

1.5

2.0

2.5

0.1

3.0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

(%)

(%)

(GPa)

9 8 7 6 5 4 0.0

0.5

1.0

1.5

2.0

2.5

3.0

(%)

Fig. 6.6 Relationship between mechanical properties and carbon fiber volume fraction of the random-Csf /geopolymer composites a flexural strength, b fracture toughness, c Young’s modulus

fraction in the composite system. Besides, although there are obvious cracks in the matrix, the flexural strength and fracture toughness of the composite can maintain a high level, indicating that when the fiber content is increased, the composites are less sensitive to these cracks.

6.1 Random-Csf /Geopolymer Composites

185

6.1.4 Fracture Behavior of the Composites Figure 6.7 exhibits the load–displacement curves of the Csf /geopolymer composites and the geopolymer matrix. Different from the typical brittle fracture of the geopolymer matrix, after the initial elastic deformation stage, the load–displacement curve of the Csf /geopolymer composites begins to deviate from the elastic deformation slowly and enters the typical nonlinear elastic or pseudoplastic deformation stage. After the maximum load is reached, the load remains high. The fracture of the composites is though quasi-static crack growth/sub-critical crack propagation, then the load gradually decreases without catastrophic failure. After unloading, the sample is still not completely fractured (Fig. 6.8). The direct effects of the above deformation and fracture behaviors are as follows: (1) The fracture strain increases significantly, which is several times or ten times higher. (2) The work of fracture (WOF) increases significantly, which was shown in Fig. 6.9. The WOF of the composites is calculated according to the method described in the literature [1]. Different from the bending strength, fracture toughness, Young’s modulus, and Vickers hardness, the WOF increases first, reaches the peak value of 624.5 J/m2 when the fiber addition is 2.0 vol%, which is increased by 14 times 60 50

(4)

(5)

(N)

40

(1) (2)05C (3)1C (4)2C (5)3C

30 (3) 20 (2) 10 0 0.0

(1)

0.2

0.4

0.6

(mm)

Fig. 6.7 Load–displacement curve of the random-Csf /geopolymer composites and matrix

10mm Fig. 6.8 Digital image of carbon fiber disordered distribution of the 3D-Csf /geopolymer composites after three-point flexural strength test

700 600 2

Fig. 6.9 Relationship between WOF and fiber contents of the random-Csf /geopolymer composites

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

(J/m )

186

500 400 300 200 100 0

0.0

0.5

1.0

1.5

2.0

2.5

3.0

(%)

compared with the matrix (38.6–56.4 J/m2 ). When the fiber increases to 3.0 vol%, the WOF decreases significantly. By analyzing the load–displacement curve and the fractograph of the composites, it is found that during the fracture process, the composites consume much energy by fiber debonding, pulling-out and bridging, and so on, which makes it present nonlinear elastic deformation and non-brittle failure mode. As shown in Fig. 6.10, the fractograph shows that the length and number of pullingout fibers are increasing with fiber content. This indicates that during the fracture process of the sample, the carbon fiber shows obvious debonding, bridging, and pulling-out, which is important to the toughness of the matrix. Figure 6.11 shows the crack propagation path of the composites. Besides the aforementioned fiber bridging and pulling-out, there is also obvious crack deflection and branching when cracks encountered fibers.

6.2 2D-Csf /Geopolymer Composites with Different Fiber Length 6.2.1 Phase Composition Figure 6.12 indicates the XRD patterns of the 2D-Csf /K-PSS composites, metakaolin, and geopolymer matrix. Metakaolin is mainly amorphous with a hump at about 22°. After geopolymerization, the typical peak moves to about 28°, which is consistent with the characteristic peak of geopolymers [2]. As mentioned above, the presence of Csf showed little effect on the amorphous geopolymer matrix. Besides, there are some peaks of α-SiO2 before and after geopolymerization as well as in the

6.2 2D-Csf /Geopolymer Composites with Different Fiber Length

(a)

(b)

(c)

(d)

187

Fig. 6.10 SEM fractographs of the random-Csf /geopolymer composites a 05C; b 1C; c 2C; d 3C

(a)

(b) Cf Cf Crack branching

debonding

Crack deflection

Fig. 6.11 SEM images of crack propagation path: a crack deflection and branching, b interfacial debonding

composites, indicating that the α-SiO2 in the raw material did not take part in the geopolymerization process. Table 6.4 lists the apparent density of the 2D-Csf /geopolymer composites. With the presence of carbon fiber, the apparent density of the composites does not only increase but also slightly decreased, although the density of carbon fiber (1.76 g/cm3 )

188

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.12 XRD patterns of metakaolin (a), geopolymer (b), and 2D-Csf /geopolymer composites [3]

Table 6.4 Apparent density of the 3.5 vol% Csf /K-PSS composites [3] Material Volume

density/g·cm−3

M

C2

C7

C12

1.42

1.42

1.40

1.41

is much higher than that of the geopolymer matrix. This is mainly due to the low fiber volume content (3.5 vol%). The presence of fiber makes the geopolymer slurry degassing difficult, leading to the increase in the number of defects such as pores and cracks.

6.2.2 Microstructure Short carbon fibers possess a high aspect ratio (330–2000), so they tend to twist during mixing. In this part, the short carbon fiber preform preparation method with the help of the ultrasonic scattering treatment can effectively prevent fibers from aggregation. As a result, the short carbon fibers have a relatively uniform distribution in the geopolymer matrix from two directions (x and y directions, as illustrated in Fig. 6.13) as shown in Figs. 6.14 and 6.15. Figure 6.16 presents the SEM crack propagation path in the 2D-Csf /geopolymer composites. Fiber bridging and deflection have played an important role in preventing the crack propagation, indicating high crack tolerance of the composites.

6.2 2D-Csf /Geopolymer Composites with Different Fiber Length

189

y

x Fig. 6.13 Diagrammatic sketch of 2D-Csf /geopolymer matrix composites [3]

(b)

(a)

200µm

200µm

(c)

200µm Fig. 6.14 Micrographs of 2D-Csf /geopolymer composites along y direction with starting fibers having an average length of a 2 mm, b 7 mm, and c 12 mm, respectively, with permission from [1]

6.2.3 Mechanical Properties The flexural strength of the composites in both x and y directions is collected in Table 6.5. In both directions, the flexural strength of the composites has significantly

190

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

(a)

(b)

200µm

200µm (c)

200µm Fig. 6.15 Micrographs of 2D-Csf /geopolymer composites along x direction with starting fibers having an average length of a 2 mm, b 7 mm, and c 12 mm, respectively [3] Fig. 6.16 SEM photograph of crack propagation path in the 2D-Csf /geopolymer matrix composites [3]

Table 6.5 Mechanical properties of the 2D-Csf /geopolymer composites [3] Material

Flexural strength (MPa)

Fracture toughness (MPa m1/2 ) x direction

Young’s modulus (GPa) x direction

y direction

x direction

M

16.8 ± 0.7

16.8 ± 0.7

0.28 ± 0.05

8.6 ± 0.4

C2

45.9 ± 6.2

61.5 ± 2.8

1.81 ± 0.08

6.7 ± 0.6

C7

52.2 ± 8.9

91.3 ± 1.3

3.26 ± 0.13

6.6 ± 0.2

C12

51.6 ± 12.9

84.6 ± 2.4

2.82 ± 0.13

6.7 ± 0.3

6.2 2D-Csf /Geopolymer Composites with Different Fiber Length

191

improved compared with that of the matrix, especially in the x direction. The flexural strength of the Cf /geopolymer composites increases with the increase of the starting carbon fiber length from 2 to 7 mm. The addition of short carbon fiber increases the geopolymer matrix strength from 16.8 to 91.3 MPa (4.4 times higher), which is regarded as a great strengthening effect for such a low volume percentage (3.5%) of short carbon fibers. However, the strengthening effect of fibers unexpectedly lowers a little with a further increase in the starting fiber length from 7 to 12 mm, and 84.6 MPa is obtained for the composite with 12 mm starting fiber in length. The highest fracture toughness of composite is more than 10 times higher than that of the matrix, reaching a value of 3.26 MPa m1/2 . Overall, both strengthening and toughening effects from fibers have significantly improved compared to the composites with 1 mm Csf . At the same time, the Young’s modulus decreased compared with the matrix. So, under the same condition of the fiber volume content, the elastic modulus is not sensitive to the fiber length. Figures 6.17 and 6.18 clearly manifest the change in tendency of flexural strength and fracture toughness with the length of carbon fiber. 120

Flexural strength, MP

Fig. 6.17 Flexural strength of the 2D-Csf /geopolymer composites reinforced by different fiber lengths [3]

100

Geopolymer matrix Csf/geopolymer composite(x direction) Csf/geopolymer composite(y direction)

80 60 40 20 0

C2

C7

C2

C7

C12

-2

3.5

Fracture toughness, MPa·m

Fig. 6.18 Fracture toughness of the 2D-Csf /geopolymer composites reinforced by different fiber lengths (x direction) [3]

M

3.0 2.5 2.0 1.5 1.0 0.5 0.0 M

C12

192

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

6.2.4 Fracture Behavior The bar specimens of the 2D-Csf /geopolymer composites deform significantly but without breaking completely even after a three-point bending test (Fig. 6.19b), suggesting a non-catastrophic fracture behavior. In contrast, the geopolymer matrix shows a typical brittle failure mode (Fig. 6.19a). The apparent failure difference between the composites and the matrix can also be clearly seen from their corresponding load–displacement curves as shown in Fig. 6.20. The low fracture energy of geopolymer matrix results in an apparent catastrophic failure. In comparison, the Cf /geopolymer composites all exhibit a nonlinear elastic increase of load until the maximum load was reached; then the load gradually decreases with increasing displacement, which is a typical pseudoplasticity behavior. The higher fracture

(a)

100µm (b)

100µm Fig. 6.19 Images of the 2D-Csf /geopolymer specimens after three-point bending test a matrix, b composites, with permission from [1]

90

c)

80 70

d)

60

Load, N

Fig. 6.20 Load– displacement curves for the matrix and 2D-Csf /geopolymer composites. a geopolymer matrix, b–d composites reinforced by 3.5 vol% short carbon fiber with b 2 mm, c 7 mm and d 12 mm in length (x direction), with permission from [1]

50 40 30

b) a)

20 10 0 0.0

0.5

1.0

1.5

Displacement, mm

2.0

6.2 2D-Csf /Geopolymer Composites with Different Fiber Length 7000

-2

6000

Work of fracture, J·m

Fig. 6.21 Work of facture of the 2D-Csf /geopolymer composites with different fiber lengths (x direction) [3]

193

5000 4000 3000 2000 1000 0 M

C2

C7

C12

energy of the composite shows a progressive fracture behavior rather than a catastrophic behavior through the propagation of cracks proceeding in a progressive and controlled manner with increasing displacement of the crosshead [15]. The work of fracture of the composites shows a similar trend as that of the flexural strength, as shown in Fig. 6.21. Incorporation of the preforms with 7 mm starting short carbon fibers increases the work of fracture of geopolymer matrix from 54.2 to 6435.3 J/m2 (nearly 118 times increase), indicating that the composite reinforced by the stacked sheet-like short carbon fiber preform can also absorb much energy to avoid the catastrophic fracture behavior probably due to the friction between fibers and geopolymer matrix during fiber pulling-out. Therefore, geopolymer matrix composites reinforced with stacked sheet-like preforms of short carbon fibers exhibit an obvious non-catastrophic fracture behavior. The toughening mechanisms of 2D-Csf /geopolymer composites can be deduced by observing the tensile side and fracture surface of specimens. A lot of cracks were formed on the tensile side of composite beam as shown in Fig. 6.22a. They are generated with nearly homogeneous spacing and propagate relatively perpendicular to the direction of tensile force. There is a high distribution density of cracks on tensile side of composites, which can consume a lot of energy and make the composites show the pseudoplasticity behavior. In addition, many holes formed on the surface of tensile side of the composites beam, as shown in Fig. 6.22a, b. They resulted from the fiber/matrix interfacial debonding and fiber “pulling-out” from the geopolymer matrix during the loading process. The fiber/matrix interfacial debonding and fiber pulling-out can consume much higher fracture energy and thus prevent catastrophic failure during service. Meanwhile, Fig. 6.22c proved that the fiber was not damaged during the failure process and played the role of bridging matrix together, which was the main reason for the obvious difference between the matrix and the composites, as shown in Fig. 6.19. Typical fracture surfaces of the obtained composites after flexural strength tests are shown in Fig. 6.23. Lots of fiber pulling-outs are found because the bonding strength of fiber/matrix is relatively weak as shown from the clean surface of the pulling-out

194

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.22 Low (a) and high (b)–(c) magnified scanning electron microscopy micrographs of the tensile side of 2D-Csf /geopolymer composites after flexural strength tests, with permission from [1]

fibers, and the strength of carbon fibers is far higher than that of geopolymer matrix. The pulling-out length is so large that it ensures substantially effective toughening effect from carbon fibers and therefore prevents catastrophic fracture of the composites. In addition, a lot of cracks (Fig. 6.23a) distribute in the geopolymer matrix on the fracture surfaces of the obtained composites. These cracks are generated because lots of fibers are pulled out from the geopolymer matrix in different direction during testing to failure. And they are helpful to obtain higher work of facture value and make the composites show the pseudoplasticity behavior.

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

195

(a)

100µm (b)

(c)

50µm

50µm

Fig. 6.23 SEM image of perpendicular (a) and parallel (b) to the fracture surface of 2DCsf /geopolymer composites after flexural strength tests, with permission from [1]

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents 6.3.1 Microstructure The surface morphologies of the 2D-Csf /geopolymer composites with different 7 mm-Csf contents in y and x directions are shown in Figs. 6.24 and 6.25. When the fiber content is from 4.5 to 7.5%, fiber dispersion in the matrix is still relatively uniform, and no obvious fiber agglomeration phenomenon is observed. Therefore, with the help of the ultrasonic scattering treatment, fiber agglomeration or uneven distribution in the composites can be effectively avoided which is the main problem using conventional methods, such as ball mill or mixing. The apparent density of the composites with different fiber contents (Fig. 6.26) manifests increasing trend with fiber content. This is because carbon fiber has a higher density than geopolymer matrix does; on the other hand, density of the geopolymer

196

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

(b)

(a)

200µm

200µm

(c)

200µm Fig. 6.24 Optical surface micrographs of the 2D-Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents (y direction). a 4.5 vol%; b 6.0 vol%; c 7.5 vol%, with permission from [4]

matrix in Cf /geopolymer composites improved under an increased forming pressure during curing. It is well known that the microstructure of geopolymer matrix has particulate features and pores which are of the order of 5–10 nm [3]. Increasing of forming pressure during curing will undoubtedly reduce the size and scale of the pores.

6.3.2 Mechanical Properties Table 6.6 lists the flexural strength and elastic modulus of the 2D-Csf /geopolymer composites with different fiber contents. The elastic modulus increases significantly when fiber content is within 6.0 vol%, which is more than two times higher than that of composites with fiber of 3.5 vol%, as shown in Fig. 6.27. When the fiber content increases to 7.5 vol%, the elastic modulus of the Csf /geopolymer composites decreases slightly, which is most likely caused by the decrease of density of the composites. Besides, during the curing process of the composites in the vacuum bag, the improvement of pressure resulted in the excessive water extruding from the matrix and insufficient matrix geopolymerization. High stress also increases the possibility of the fiber damage, which may have partly offset its enhancement effect.

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

(a)

197

(b)

200µm

200µm (c)

200µm Fig. 6.25 Optical surface micrographs of the 2D-Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents (x direction). a 4.5 vol%; b 6.0 vol%; c 7.5 vol% [3]

1.7

-3

1.6

Density, g·cm

Fig. 6.26 Density of the Csf /geopolymer composites reinforced by 7 mm short carbon fiber with permission form [4]

1.5 1.4 1.3 1.2

0

3.5

4.5

6

7.5

Volume fraction of fiber, %

The three-point bending strength of the composites in both x and y axial directions as a function of volume fraction of short carbon fiber is compared in Fig. 6.28. With the increase of volume fraction of carbon fiber from 3.5 to 4.5 vol%, the flexural strength of the composites in both directions increases from 16.8 to 96.6 MPa attaining the peak value (4.75 times higher), which is regarded as an unexpectedly

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Table 6.6 Mechanical properties of the Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents [3]

Fig. 6.27 Young’s modulus of the 2D-Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents (x direction) [3]

Fiber content (vol%)

Flexural strength (MPa)

Young’s modulus (GPa)

x direction

y direction

x direction

3.5

91.3 ± 1.3

44.2 ± 8.9

6.6 ± 0.6

4.5

96.6 ± 4.9

76.6 ± 1.6

12.0 ± 0.5

6.0

87.4 ± 14.5

69.7 ± 11.8

20.5 ± 1.6

7.5

42.0 ± 6.1

25.5 ± 2.7

17.8 ± 0.8

20

Young's modulus, GPa

198

16 12 8 4 0

3

4

5

6

7

8

Volume fraction of fiber, %

120

Flexural strength, MPa

Fig. 6.28 Flexural strength of the 2D-Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents [3]

x axial direction y axial direction

100 80 60 40 20 3

4

5

6

7

8

Volume fraction of fiber, %

great strengthening effect for such a low volume percentage of short carbon fiber. However, the strengthening effect of short carbon fiber lowers with a further increase in the fiber volume fraction to 6 and 7.5%, which could be attributable to the decrease of the density of the composites. This should be the same cases for the decrease in the elastic modulus mentioned above.

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

199

6.3.3 Fracture Behavior Typical load–displacement curves of matrix and composites are shown in Fig. 6.29. The low fracture energy of geopolymer matrix results in apparent catastrophic failure. In contrast, the failure of the Csf /geopolymer composites is quite different. The composites extend elastically at the beginning of the test. Beyond the elastic limit, the applied load produces plastic deformation until maximum load is reached, and then the load drops with the increase of displacement and formed a long tail due to the fiber debonding and pulling-out. With the increase of volume fraction of short carbon fibers, the displacement of the crosshead shortens where the maximum applied load is reached and the load reduces fast during the unload process. This implies that the deformation and safety factors of the Csf /geopolymer composites during service both decrease with the increase of volume fraction of short carbon fibers. According to the load–displacement curve, the relationship between the work of fracture of composites and the fiber content is manifested in Fig. 6.30. The work of fracture shows the maximum value (118 times higher than that of geopolymer matrix) at 3.5 vol% Csf /geopolymer composite, and then it gradually reduces with the increase of fiber volume content. The fracture behavior of the investigated composites can be seen clearly from the observation of the micrographs perpendicular to the fracture surface of Csf /geopolymer composites, as shown in Fig. 6.31. A lot of pulled-out carbon fibers are clearly observed on the fracture surface. The pulling-out length is so big that it ensures the reinforcement of short carbon fibers effective in preventing catastrophic fracture. The pulling-out length of fiber decreases from 600 to 100 μm with the increase of volume fraction of short carbon fibers from 3.5 to 7.5 vol%. The decrease of pulling-out length might be due to higher shear stresses at the intersection between fibers and interface cohesion strength of fiber/matrix formed by higher forming pressure that makes the fibers favor breakage rather than pulling-out during 100

80 a)

Load, N

Fig. 6.29 Load– displacement curves of the Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents (x direction). a 3.5 vol%; b 4.5 vol%; c 6 vol%; d 7.5 vol%, with permission from [4]

60 b)

40 c) d)

20

0

0.2

0.4

0.6

0.8

1.0

1.2

Displacement, mm

1.4

1.6

1.8

200

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites 7000

Work of fracture, J·m

-2

6000 5000 4000 3000 2000 1000 0

3

4

5

6

7

8

Volume fraction of fiber, %

Fig. 6.30 Work of fracture of the 2D-Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents (x direction) [3]

a)

b)

100μm

100μm d)

c)

100μm

100μm

Fig. 6.31 SEM micrographs of parallel to fracture surface of the 2D-Csf /geopolymer composites reinforced by 7 mm short carbon fiber with different fiber contents. a 3.5 vol%; b 4.5 vol%; c 6 vol%; d 7.5 vol%, with permission from [4]

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

201

load. Fiber pulling-out absorbs more energy than fiber breakage [5]. This indicates energy consumption and the work of fracture decreases with the increase of volume fraction of short carbon fibers, which matches the results mentioned above.

6.3.4 In Situ SEM Observation on Crack Initiation and Propagation Over the past years, many researches on the failure mechanisms of short fiberreinforced composites have been conducted, which demonstrated that the cracks play an important role in the mechanical properties and fracture behavior of the composites [6–12]. In addition, some macroscopic and theoretical relations between the cracks and the fracture mechanism of the composites have been proposed [7–10]. However, the details of failure mechanism, especially for the effects of microcracks propagation and distribution on fracture behavior of the composites, are unspecified up to now. In this part, the bending test of short fiber-reinforced composites was first employed on an environmental scanning electron microscope (ESEM) to determine the crack growth and the fracture behavior with increasing displacement of the crosshead. Relations between the crack growth and the fracture behavior of the composites were also reported. The typical load–displacement curve and in situ crack evolution for the Csf /geopolymer composites is given in Fig. 6.32. The composites exhibit a significant deformation and an obvious non-catastrophic fracture behavior during the bending test, which is regarded as a great toughening effect for the short carbon fiber with such a low volume percentage (3.5 vol%). The composites exhibit a nearly elastic response in the initial stages (stages I and II) though a change appears at a load of about 6 N, which is similar to that of unidirectional continuous fiber reinforced

30 I

II

III

25

d

D

c

Load, N

Fig. 6.32 Analysis of typical load/displacement curve of 2D-Csf /geopolymer composite (fiber length of 7 mm and content of 3.5 vol%), with permission from [13]

C

20 15

e

10 5 0 0.0 A

b

B

E

a

0.5

1.0

Displacement, mm

1.5

2.0

202

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

composites [13]. Beyond the elastic limit, the applied load produces plastic deformation until the maximum load is reached. Then the load gradually decreases with the increasing displacement and forms a long tail (stage III). Figure 6.33 presents a series of ESEM images of crack initiation and propagation process on the side of the beam sample of the composites, which corresponds to the test points in the load/displacement curve in Fig. 6.32. At the first elastic stage (stage I), no crack is found on the beam sample as shown in Fig. 6.33a. However, at the beginning of the second elastic stage (stage II), a lot of microcracks (Fig. 6.33b) appear on the side of the beam sample. To encourage a crack growth, an increasing energy is required. When the bending stress is higher than the strength of the geopolymer matrix, a microcrack will initiate first in the geopolymer matrix on sample beam surface. With higher load applied, the microcrack will propagate and meet with the reinforced fibers inevitably. Due to their high mechanical strength, the reinforced fibers will try to keep the composite integrity instead of being broken. Hence, the microcrack growth will greatly slow down, and an internal stress will be cumulated between the matrix and the reinforced fibers. When the increasing cumulation internal stress in the matrix is high enough (the mechanical strength of the reinforced fiber is much higher than that of the matrix), other new microcracks will occur on the beam surface, as shown in Fig. 6.33b. This interesting phenomenon indicates that the stress distribution in the matrix has changed well due to the enhancement effect of the reinforced fibers. Though the formation of these microcracks reduces the matrix elastic modulus, as indicated by the load/displacement curve slopes in Fig. 6.32, the sample keep a nearly elastic deformation behavior companying with the propagation of the microcracks (Fig. 6.33c). Under the increasing bending load, the microcracks have grown up with similar rates. This unconventional fracture behavior is supposed to be attributed to the following reasons. The short carbon fibers used in this study have a length of 7 mm and the gap lengths are 300–500 μm, as shown in Fig. 6.33c. Hence, the fibers are long enough to bridge several microcracks together. As discussed above, the fibers have a far higher mechanical strength than that of the matrix. Thus, the bridging fibers in the microcracks are difficult to be fractured, which is helpful to keep the composites integrity and to retard the formation of a main crack. The significant deformation of the composites can be attributed to the large number of microcracks during the bending test. It can be seen from Fig. 6.34a, the microcracks are generated nearly on the whole span side surface of the beam sample. The gaps become broader as they are far away from the main crack, which means that the stress will be less as the stress locations are far from the crosshead. It is also found that the fracture path is unstraight, as shown in Fig. 6.34b, c. A lot of crack deflection (Fig. 6.34d) and the crack branching (Fig. 6.34e) are found, which undoubtedly lead to the increase of the fracture toughness. Similar crack formation phenomenon was also noted from Csf /geopolymer with fiber contents of 4.5, 6, and 7.5 vol%, as shown in Figs. 6.35, 6.36 and 6.37. A schematic drawing of the fiber bridging cracks is provided in Fig. 6.38, assuming the fiber is rigid and its elastic deformation during the bending test can be neglected.

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

203

(a)

250μm

(b)

microcracks 250μm

(c)

250μm

(d)

250μm

(e)

250μm

Fig. 6.33 Series of ESEM images (a)–(e) of crack initiation and propagation process on the side of a beam sample of the Csf /geopolymer composites corresponding to the position A–E of the load/displacement curve in Fig. 6.32, separately, with permission from [13]

204

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.34 The side of a bar specimen of the Csf /geopolymer composites after a three-point flexural test (a) and series of images of fiber pulling-out (b), fiber bridging (c), crack deflection (d), and crack branching (e) corresponding to the zones 1–4 in the image (a) separately, with permission from [13]

The fibers which bridge more than two microcracks are under the effects of the tension force, F t , and the holding forces, F HS , which are from the matrix blocks. For point A, with the increasing bending load, the fiber is undergoing an increasing F t and FH2 , which is from Block 2. It will not be pulled out from the matrix if F t < F H2 . When F t > F H2 , F H1 , which is supplied by Block 1, starts to effect on the same fiber. As FH2 < Ft < FH1 + FH2 , the fiber will still be arrested in the matrix. Only when F t > F H1 + F H2 (assuming only two matrix blocks in the left side of point A), the fiber will be pulled out from the matrix. If there are lots of matrix blocks (the number is N) in the left side of point A, the fiber will be broken instead of being pulled out when F b < F t < F H1 + F H2 + … + F HN . (F b is the ultimate breaking load of the fiber.) As the F b is very high, the bridging fibers show significant “arrest effects” on the microcracks and prevent them from further opening and propagation. It is supposed that the “arrest effects” from the reinforced fibers have changed the stress distribution in the sample beam during the three-point bending test. For usual rigid materials, their stress distribution on the tension side of the beam samples is triangle like, as shown in Fig. 6.39b. Normally only one main crack will form at the central area, as shown in Fig. 6.39a. However, there are lots of similar microcracks nearly homogeneously distributed on the surface of the Cf /geopolymer composites, as shown in Fig. 6.33c. It is indicated that the stress distribution in this deformation stage is different from that in Fig. 6.39b. As they have been effectively transferred from the central area to the edge area by the bridging fibers, the stress in the edge

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

205

(a)

500µm (b)

500µm (c)

500µm (d)

500µm

Fig. 6.35 Series of ESEM images of crack initiation and propagation process (a)–(d) on the side of a sample beam of the Csf /geopolymer composites reinforced by 4.5 vol% and 7 mm short carbon fibers during flexural strength test [3]

206

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

(a)

500µm (b)

500µm

(c)

500µm

(d)

500µm

Fig. 6.36 Series of ESEM images of crack initiation and propagation process (a)–(d) on the side of a sample beam of the Csf /geopolymer composites reinforced by 6 vol and 7 mm short carbon fibers during flexural strength test [3]

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

207

(a)

500μm (b)

500μm

(c)

500μm

(d)

500μm

Fig. 6.37 Series of ESEM images of crack initiation and propagation process (a)–(d) on the side of a sample beam of the Csf /geopolymer composites reinforced by 7.5 vol% and 7 mm short carbon fibers during flexural strength test [3]

208

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.38 Analytical models of a fiber bridging cracks and the forces appearing in the fiber, with permission from [13]

Fig. 6.39 Schematic figures of the elastic deformation as well as their stress pattern of usual rigid materials (a), (b) and the Csf /geopolymer composites (c), (d) under the three-point bending test, with permission from [13]

area will be a little bit lower, as illustrated in Fig. 6.39d. The maximum stress is still in the center point. The increasing deformation amount of the beam sample is attributed to the homogeneous growth of the cracks during stage II. The beam sample has presented a pseudoelastic deformation behavior until a main crack is formed, as shown in Fig. 6.32. With the increase of the bending load, the stress level distributed in the tension surface will be improved accordingly. The cracks will continue to grow up until the stress in the central point reach their critical value. The fibers will be pulled out from the matrix or be broken when the Ft is high enough. As a result, the bridging effect from the pulled out or broken fibers will disappear and the stress in the tensile surface will be redistributed. At this time, the deformation steps into stage III. Most of the loading stress will gather in the crack front in the central area. The stress in other areas will be decreased gradually. Hence, a main crack is formed in the central area and the growth of the microcracks in other areas slow down till arrested or closed to some degree as shown in Fig. 6.33d. With more gathering stress, more fibers are

6.3 2D-Csf /Geopolymer Composites with Different Fiber Contents

209

Fig. 6.40 The evolution of the stress on the tensile surface of the composite beam sample under bending load: a after the initiation of the microcracks; b during the extension of the microcracks; c after the formation of the main crack, with permission from [13]

pulled out or broken. The main crack is gradually broadened, and the pulled out and broken fibers are easily found in the main crack, as shown in Fig. 6.33e. The left fibers still show the bridging effect, which prevents the beam sample from a catastrophic fracture. The evolution of the stress on the tensile surface of the beam sample can be described by Fig. 6.40. In summary, in situ crack growth observation during flexural test proved that a lot of microcracks are formed on the whole surface of the beam sample. The propagation of the microcracks ceases, and they tend to close to some degree while the main crack forms because the stress in the microcrack area is somewhat relaxed and the stress in the main crack area greatly increased. The fiber bridging effect in the micro and main cracks effectively keeps the composites integrity and still have high ability to bear the load during the bending test, which makes the composites exhibit a non-catastrophic fracture behavior.

6.4 2D-Csf /Geopolymer Composites with Fiber Surface Treatment The interfacial properties of fiber and matrix are very important to the mechanical properties of composites. Too weak interface bonding make it impossible to achieve load transfer which is necessary for the improvement of mechanical properties of composites, whereas too strong interface bonding is not conducive to the toughening of composites [14]. From the previous discussion of the Csf /geopolymer composites, the interfacial bonding strength between the Csf and the matrix is in low state. Therefore, the interfacial properties of the composites are modified by the Ni/P coating on the surface of short carbon fibers by electroless plating method, and the influence of coating thickness on the mechanical properties and fracture behavior of composites is investigated.

210

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

6.4.1 Preparation Process of the Ni/P Coating on the Surface of Carbon Fiber The treatment process for the electroless plating Ni/P coating on the carbon fiber surface is as follows [15–17]. First, Csf is successively soaked in the SnCl2 and PdCl2 solutions for sensitized and activated treatment. Then, electroless Ni/P plating is applied in an acid bath. Acid bath has the advantages of stable, easy to control, and simple. The composition of the plating solution is shown in Table 6.7. The prepared Csf is placed into the bath for 0.5–3 min and then fished out. After washing with distilled water for several times, they are put into the oven at 80 °C for 12 h before use. The weights of carbon fiber before and after electroless plating are recorded to calculate the mass increase ratio of G/G; then the plating thickness r is calculated according to (6.1).  r =

ρNi G 1+ ρc G



 −1 r

(6.1)

where G is the weight of the coating (g), G is the weight of the fiber before coating (g), r is coating thickness of fiber (μm), ρ Ni is the density of Ni(g cm−3 ), and ρ c is the density of carbon fiber (g cm−3 ). Figures 6.41 and 6.42 show the SEM images of the surface morphology of the carbon fiber with and without Ni/P coating. The surface of untreated carbon fiber is clean with obvious grooves (see high magnification image). After electroless Ni/P plating, the grooves on the fiber surface are gradually filled up. As the thickness of the coating increases, more Ni/P particles adhere to the coating surface, which make the fibers coarser and coarser. Ni/P particles with diameter of about 200–300 nm are closely accumulated together on the fiber surface, which makes the surface rough. Such structure is conducive to the realization of interconnecting and interlocking between fiber and matrix, resulting in the increase in the interfacial bonding strength. EDS analysis of the coating region shows that the mass ratio of Ni/P in the coating is about 43/7 (Fig. 6.41f). Table 6.7 Components and conditions of chemical bath [3] NiSO4 · 6H2 O (g L−1 )

NaH2 PO2 · 2H2 O (g L−1 )

C3 H6 O3 (g L−1 )

KIO3 (mg L−1 )

NaAC (g pH L−1 )

Temperature (°C)

Plating time (min)

25

25

25

1

5

70

0.5–3

4.8

6.4 2D-Csf /Geopolymer Composites … (a)

211 (b)

20µm (c)

20µm (d)

20µm (e)

20µm (f)

20µm

Fig. 6.41 SEM images of Ni/P coated short carbon fiber preforms with different coating thicknesses. a 0 μm, b 0.05 μm, c 0.08 μm, d 0.15 μm, e 0.27 μm and f EDS analysis of obtained coating [3]

6.4.2 Microstructure The apparent density of the composites with carbon fiber of Ni/P coating is shown in Table 6.8. With the presence of Ni/P coating, the density of the composites increases significantly compared with that of the untreated ones, while the apparent density of the composites increases slightly with the increase of the thickness of the fiber surface coating.

212

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites (b)

(a)

2µm

2µm (c)

(d)

2µm

2µm (f)

(e)

500nm

2µm

Fig. 6.42 SEM images of Ni/P coated short carbon fiber with different coating thickness. a uncoated Ni/P; b 0.05 μm; c 0.08 μm; d 0.15 μm; e 0.27 μm; f high magnified image [3]

Table 6.8 Apparent density of Csf /geopolymer composites with Ni/P coating (fiber length of 7 mm and content of 3.5 vol%) [3] Material Apparent density

(g/cm3 )

C7

C7N05

C7N08

C7N15

C7N27

1.40

1.56

1.58

1.59

1.58

The original carbon fiber surface is hydrophobic, which cannot be well impregnated by the water-based geopolymer slurry and resulted in residual gaps between fiber and matrix in the final Csf /geopolymer composites. However, with the presence of Ni/P coating, the fiber surface changed from hydrophobic to hydrophilic, which

6.4 2D-Csf /Geopolymer Composites …

213

(b)

(a)

Ni C Al Si K 50µm

10µm

Fig. 6.43 BSE images of Ni/P plated 3.0 vol% Csf /geopolymer composites (a) and EDS line analysis of the fiber/matrix interface (b) [3]

is conducive to decrease the interfacial gap between the fiber and the matrix. Meanwhile, the density of Ni/P is much higher than both fiber and geopolymer, which also contributed to the increase in apparent density. Figure 6.43 presents the BSE image of the surface microstructure of the composite and the linear elemental scanning analysis of the element distribution at the interface area. Ni/P coating on the fiber surface is clearly visible. The high magnification image (Fig. 6.43b) shows that the Ni/P coating is continuous but not uniform in thickness, and the granular-like bumps of the coating extend into the matrix. This is consistent with the previous observation of the fiber surface coating. Besides, the matrix can fully impregnate into the particle gap on the surface coating, and the interfacial bonding between fiber and matrix is in good state.

6.4.3 Mechanical Properties The flexural strength of the Csf /geopolymer composites with different Ni/P coating thickness on the fiber surface is shown in Fig. 6.44. At the beginning, the flexural strength increases slightly with the increase of coating thickness, which is increased from 51.5 to 55.2 MPa when Ni/P coating thickness is from 0 to 0.15 μm. However, compared with the peak value at coating thickness of 0.15 μm, when the Ni/P coating thickness increases to 0.27 μm, the flexural strength decreases to 46.3 MPa, which decreases by 16.0% and is also 10.1% lower than composites without coating. Young’s modulus of the composites shows a similar trend with that of flexural strength, as shown in Fig. 6.45. Therefore, the thickness of the fiber surface coating has a significant influence on the flexural strength of the composites, and only with appropriate coating thickness, the fiber can play better strengthening effect. Different from the flexural strength and Young’s modulus mentioned above, the work of fracture is more sensitive to the interface coating, which generally presents a monotonically decrease with the increase of the coating thickness on the fiber

214

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.44 Variation of flexural strength of the 3.0 vol% Ni/P plated short carbon fiber reinforced Csf /geopolymer composites as a function of coating thickness (x direction) [3]

Fig. 6.45 Variation of Young’s modulus of the 3.0 vol% Ni/P plated short carbon fiber reinforced Csf /geopolymer composites as a function of coating thickness (x direction) [3]

surface, as shown in Fig. 6.46. Specifically, it drops sharply from 3656 J m−2 of the uncoated composite to 2309 J m−2 of composite with coating thickness of 0.05 μm. Fig. 6.46 Variation of work of fracture of the Ni/P plated 3.0 vol% Csf /geopolymer composites as a function of coating thickness (x direction) [3]

6.4 2D-Csf /Geopolymer Composites …

215

Fig. 6.47 Load– displacement curves of fracture of the Ni/P plated 3.0 vol% Csf /geopolymer composites as a function of coating thickness (x direction) [3]

After that, the work of fracture changes slowly with the increase of the thickness of the fiber coating to 0.15 μm. The above-mentioned changes of flexural strength, Young’s modulus, and work of fracture are also fully reflected in the load–displacement curves of the composites (Fig. 6.47). The reason for the significant decrease of work of fracture with the Ni/P coating is due to the changed failure modes: composites with uncoated fibers show a pseudoplastic fracture mode, and both fiber bridging and pulling-out are the mail toughening mechanisms. However, the composites with Ni/P coating show increased interfacial bonding strength, and toughening effects from fibers are substantially weakened, which result in the difference in tendency from pseudoplastic failure to brittle failure. Figure 6.48 shows the fractographs of the composites with different fiber coating thickness. With the thickness of the Ni/P coating increasing to 0.27 μm, the average fiber pulling-out length decreases from 600 to 100 μm, indicating the bonding strength between the fiber and the matrix increased significantly after Ni/P coating. Together with the above analyses on flexural strength, Young’s modulus, and work of fracture, it can be deduced that the increased interfacial bonding strength lead to the increase in both strength and modulus, but failure mode changes from ductile to brittle one, as shown in the obvious decrease in work of fracture. To sum up, the fiber coating treatment reflects the contradiction between the strengthening and toughening in the fiber/geopolymer composites. Therefore, the coating thickness should be optimized according to the specific service conditions in the practical engineering application of this kind of composites.

216

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

(b)

(a)

50µm

50µm (d)

(c)

50µm

50µm

(e)

50µm Fig. 6.48 Fractographs of Csf /geopolymer composites with different Ni/P coating thicknesses. a 0 μm, b 0.05 μm, c 0.08 μm, d 0.15 μm, e 0.27 μm [3]

6.5 Effects of Thermal Load on the Microstructure and Mechanical Properties of 2D-Csf /Geopolymer Composites In this section, 2D-Csf /geopolymer composites were prepared as described at the beginning of this chapter. Short carbon fibers of length 7 mm and content 3.5 vol% were used. In order to increase the thermal stability, Al2 O3 particles were added into the geopolymer slurry before sheet-like fiber preforms impregnation. The compositions of the mixture for sample preparation are listed in Table 6.9. The heating of the samples to the predetermined temperature was carried out at a rate of 5 °C/min in a vacuum atmosphere with soaking time for 2 h, and finally, they

6.5 Effects of Thermal Load on the Microstructure and Mechanical Properties …

217

Table 6.9 Composition of 2D-Csf/geopolymer composites [18] Material no

Fiber volume content (%)

Mass of KOH (g)

Water volume (mL)

Mass of silica sol (g)

Mass of metakaolin powder (g)

Mass of α-Al2 O3 (g)

C7A0

3.5

25

2

60

44.4

0

C7A1

5

10.2

C7A3

7

30.6

C7A5

10

50.1

were slowly cooled to room temperature. The target treatment temperatures were 400, 600, 800, 1000, and 1200 °C, respectively.

6.5.1 Phase Composition Typical XRD traces of Csf /geopolymer composites after treatment at different temperatures are indicated in Fig. 6.49. Composites C7A0 (Fig. 6.49a) show a characteristic amorphous hump 2θ of about 28° and a sharp major reflection of the quartz in the original kaolinite in the temperature range of 25–800 °C. The diffraction lines corresponding to leucite (KAlSi2 O6 ) are found at 1000 °C and become sharper at 1200 °C as the broad amorphous background feature disappears, which indicates that the amorphous geopolymer matrix is crystallized and leucite is formed at 1000 °C. However, for composites C7A3 (Fig. 6.49b), the leucite phase is not found until 1200 °C. The onset crystalline temperature of composites C7A3 is obviously higher than that of composites C7A0. Therefore, the addition of α-Al2 O3 particle filler into geopolymers clearly increases the onset crystalline temperature. The distribution of α-Al2 O3 particles in geopolymer matrix probably generates “inserted effect” in amorphous structure of geopolymer and reduces the viscosity of geopolymer at high temperatures, which may results in the retardance of onset crystalline temperature.

6.5.2 Microstructure Csf /geopolymer composites showed featured reticular cracks after being treated at high temperature, as shown in Fig. 6.50. Starting from 400 °C, there were a large number of uniform reticular microcracks in the surface; at 600 °C, the crack width seems to increase slightly, which was greatly due to different thermal shrinkage between the fiber and the geopolymer matrix caused by evaporation of free water and chemical bonded water. After 800 °C high temperature treatment, the number of surface crack was reduced, which may be due to the softening of geopolymer

91.3

72.4

73.0

61.8

C7A1

C7A3

C7A5

51.3

68.4

61.6

55.1

19.4

22.8

31.0

18.3

21.1

10.6

9.3

8.2

800 °C

24.7

32.6

28.0

29.6

1000 °C

15.0

27.7

29.5

26.4

1200 °C

3.7399

4.4074

4.4464

6.4353

4.0801

6.2737

4.8645

4.4723

400 °C

25 °C

600 °C

25 °C

400 °C

Fracture work (kJ m−2 )

Flexural strength (MPa)

C7A0

Material no

Table 6.10 Mechanical properties of Csf /KGP composites after treatment at different temperatures [18]

0.4427

0.8275

2.2530

0.7856

600 °C

2.0985

0.8078

0.4665

0.5498

800 °C

2.8583

4.3244

3.9578

4.9746

1000 °C

0.2397

0.6811

0.6194

0.5570

1200 °C

218 6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

6.5 Effects of Thermal Load on the Microstructure and Mechanical Properties …

219

Fig. 6.49 XRD patterns of composites C7A0 (a) and C7A3 (b) after treatment at different temperatures, with permission from [18]

matrix. With further increase in treatment temperature, the reticular cracks become prominent again, which was the result of matrix sintering and crystallization, leading to the further increase in the thermal mismatch between the fiber and the matrix. From the surface morphology of specimen treated at 1200 °C (Fig. 6.50e), it was clear that the cracks were well bridged by carbon fibers and no fiber fracture was observed.

6.5.3 Thermal Shrinkage Figure 6.51 collects the thermal shrinkage curves of the composites after treatment at different temperatures. It can be observed that the overall thermal shrinkage of specimens is greatly influenced by treatment temperature. With the increase of treatment temperature, the shrinkage in the direction perpendicular to the lamination of composites (Fig. 6.51a) gradually increases until 1000. Beyond 1000, the composites exhibit a larger degree of shrinkage. However, in the direction parallel to the lamination (Fig. 6.51b), the composites show an expansion behavior in the temperature range of 400–1000 °C. At 1200 °C, a larger degree of shrinkage is noted which was the same to that in perpendicular to the lamination direction of the composites. Furthermore, the volume change of the composites gradually decreases with the increase of the content of α-Al2 O3 particles at the same temperature. This indicates that the addition of α-Al2 O3 is helpful to keep volume of the composites stable. Geopolymer matrix after curing retains about 15% water, including free water and hydration water. The free water is lost at low temperatures. However, the hydration water is either bounded tightly or less able to diffuse to the surface, and continues to evolve gradually until about 500 °C and beyond [19]. The water evolution will undoubtedly result in the decrease of mass. In addition, the microstructure of geopolymer matrix is of nanoparticulate features and pores with size of 5–10 nm [20]. The densification and crystallization of geopolymer matrix also result in reduction

220

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

(b)

(a)

500µm

500µm (c)

(d)

500µm

500µm

(e)

250µm Fig. 6.50 Microstructure of C7A3 after being treated at different temperatures. a 400 °C; b 600 °C; c 800 °C; d 1000 °C; e 1200 °C[3]

and disappearance of nanopores. The presence of α-Al2 O3 particle into the composites indirectly reduces the content of water and nanopores because α-Al2 O3 has no water and nanopores. Furthermore, the thermal resistance of α-Al2 O3 is far larger than that of geopolymer matrix. Therefore, the increase in content of α-Al2 O3 particles (composites C7A3 and C7A5) clearly reduces the shrinkage and keeps the volume

6.5 Effects of Thermal Load on the Microstructure and Mechanical Properties …

221

Fig. 6.51 Thermal shrinkage of Csf /geopolymer composites in perpendicular (a) and parallel (b) to lamination directions at different heat treatment temperatures, with permission from [18]

stable at high temperatures, especially in the temperature range of 800–1200 °C. In addition, the onset crystalline temperature of geopolymer matrix increases by almost 200 °C due to α-Al2 O3 filler, as shown in Fig. 6.49. This also favors to reduce the shrinkage and keep the composites stable at high temperatures. The expansion behavior in the direction parallel to the lamination of the composites is probably since the shrinkage of matrix results in debonding between short carbon fiber preforms and matrix because the interlaminar shear strength of the composites is relatively low. However, the densification and crystallization of the composites at 1200 °C lead to larger degree of shrinkage. Table 6.10 presents the mechanical properties of the as-prepared Csf /geopolymer composites after treatment at different temperatures. The flexural strength of the composites decreases with the increase of treatment temperature until 600–800 °C, as shown in Fig. 6.52a. Geopolymer network consists of SiO4 and AlO4 tetrahedral linked alternately by sharing oxygens. And positive ions (K+ , H3O+ ) should be present in the framework cavities to balance the negative charge of AlO4 [21]. The

Fig. 6.52 Flexural strength (a) and work of fracture (b) of Csf /geopolymer composites after treatment at different temperatures, with permission from [18]

222

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

balance will be destroyed once the free water and hydration water are gradually lost as the composites are subjected to the high temperatures. And the degree of polycondensation of the geopolymer matrix is lowered, which will weaken the strength of the composites. In addition, the thermal loads lead to the appearance of macrocracks as the composites are subjected to the high heat treatment temperatures. These cracks negatively affect the mechanical properties of the composites in quite a remarkable way. The hydration water of geopolymer matrix is lost at about 600. At the same time, the flexural strength of the composites shows the minimum in the temperature range of 600–800 °C. The strength increases at 1000. This is probably because the geopolymer matrix is sintered at 1000 and the content of nanopores is reduced. The sharp decrease of strength at 1200 °C is likely due to the clear shrinkage (as shown in Fig. 6.51) that destroys the structure of short fiber preforms and weakens the interface of fiber/matrix. The fracture work of the composites shows almost a similar trend as the flexural strength, as demonstrated in Fig. 6.52b. The addition of α-Al2 O3 particle into geopolymer matrix cannot improve the mechanical properties of composites at room temperature, but the mechanical properties of the composites subjected to high temperature treatment can be improved to a certain extent, because the addition of α-Al2 O3 particles effectively reduces the shrinkage and keeps the volume of the composites stable for high temperature treatment. The typical load–displacement curves for the geopolymer composites are given in Fig. 6.53. The composites reinforced with short carbon fiber preform show a noncatastrophic fracture behavior. The composites extend elastically at the beginning of the test. Beyond the elastic limit, the applied load produces plastic deformation until the maximum load reaches, and then the load drops with the increase of displacement and forms a long tail due to the fiber debonding and pulling-out. Although the ultimate strength of the composites reduces after heat treatment in comparison with that of the composites at room temperature, the head displacement prior to final fracture increases after treatment at 400, 800, and 1000 °C. This indicates that the interfacial

60 RT

o

400 C

50

Load, N

Fig. 6.53 Load– displacement curves of composites Csf /geopolymer composites after treatment at different temperatures, with permission from [18]

40 30 o

1000 C

o

20

600 C

10

1200 C

o

0 0.0

o

800 C 0.5

1.0

1.5

2.0

2.5

Displacement, mm

3.0

3.5

4.0

6.5 Effects of Thermal Load on the Microstructure and Mechanical Properties …

223

structure between fiber and matrix is ideal and suitable for the fiber pulling-out mechanism. The fracture behavior of the investigated composites can be demonstrated clearly from SEM images by observing the fracture surface of the composites, as shown in Fig. 6.54. A lot of pulling-out fibers are found on the fracture surface, because the strength of fiber/matrix is relatively weak as shown from the clean surface of the pulling-out fibers, and the strength of carbon fibers is far higher than that of geopolymer matrix. The pulling-out length is long to ensure substantially effective toughening effect from carbon fibers and prevent catastrophic fracture of the composites. In addition, a lot of matrix cracks (as shown in Fig. 6.54b) distribute in the geopolymer matrix on the side of the composite beam near the fracture surface. These cracks are generated because a lot of fibers are pulled out from the geopolymer matrix in different directions during testing to failure. And they are helpful to obtain more facture work and make the composites show the pseudoplasticity behavior.

(b)

(a)

200µm

200µm

Fig. 6.54 SEM images of composites Csf /geopolymer after treatment at 1000 °C for 2 h. a Parallel to fracture surface; b Perpendicular to fracture surface, with permission from [18]

Fig. 6.55 Relationship between orthoaxis and off-axis

224

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites In this section, we use the mesoscopic mechanics theory to investigate the relationship between fiber content, length to diameter ratio, distribution, and fiber properties of short fiber and the mechanical properties of the composites. Furthermore, the strength, modulus, and pulling-out energy of short carbon fiber (Csf ) reinforced geopolymer (Csf /geopolymer) composites are predicted, which can provide a theoretical reference for further mechanical analysis and optimal design of this kind of composites.

6.6.1 Flexural Strength Prediction (1) Strength theory When the fiber-reinforced composite is under load, the fiber does not contact the load directly. The load directly acts on the matrix, and then was transferred to the fiber through the interface between the matrix and the fiber. Therefore, the length and arrangement of fiber and the interface between the fiber and matrix have important effects on the mechanical properties of the final composites. When the length of the fiber is much longer than the length of the interface region which transmits the stress, the effect of the load at the end of the fiber can be negligible. At the same time, the fiber can be considered as continuous phase. In the short fiber-reinforced composites, with the decrease of the length/diameter ratio of fiber, the stress and strain between the fiber and surrounding matrix is changed, which is due to the discontinuity of the fiber, leading to the decrease in strengthening effect from fibers. Therefore, the stress distribution at the top of short fiber plays a very important role in the performance of short fiber-reinforced composites. Cox [22] reported that there was a minimum length for the short fibers to reach the maximum fiber stress (σ f )max . If the stress of fiber is along the direction of the load, the minimum length of the short fiber is defined as the critical fiber length L c as follows (6.2): Lc =

σfu df 2τi

(6.2)

where the σ fu is fiber tensile strength, d f is the fiber diameter, τ i is the interfacial shear stress. The critical fiber length L c is the maximum value of the load transfer length. It is one of the important parameters of short fiber-reinforced composites, and also can affect the ultimate properties of composites. If the actual length of the fiber L

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

225

is less than or equal to L c , all the fibers of the composites can be pulled out during the fracture process. If the actual length of the fiber L is longer than L c , when the composite is broken, parts of the fibers will be pulled out, while the other parts will be considered as fracture. In the fiber-reinforced composites, only when the fiber is completely in the same direction as the load, it can fully play the strengthening effect. When the fiber is perpendicular to the load, no strengthening effect occur. When the angle between fiber direction and loading direction is at certain angle, θ, fiber oriental coefficient ηθ (or the orientation factor) was suggested by Cox [22] and Bowyer [23]. Thus, they define the L cos(θ) to illustrate the effective fiber length, ηθ L. When the fibers in composites are in different directions, ηθ is the mean value of the cosθ, with scope of 0 ≤ ηθ ≤ 1. Assuming the yield strain of the fiber is greater than that of the matrix (εfu > εcu ), the predicted strength of the composites can be obtained as follows: ➀ If L < L c , the maximum stress of the fiber cannot reach the ultimate tensile strain of the matrix (σf )εmu . Therefore, no matter how much the stress is applied, the fiber is broken before reaching its average strength σ fu . The failure of composite is caused by the failure of matrix or interface. Considering the arrangement of fiber, the flexural strength can be calculated as follows: σcu = ηθ

τL Vf + σmu Vm df

(6.3)

➁ If L > L c , when the matrix stress reaches σ mu , the maximum fiber stress (σ f )max reaches the ultimate tensile strain εcu of the corresponding matrix (σf )εmu . When the volume of the fiber is low, as the matrix cracks, the fiber is pulled off immediately, leading to the fracture of the composites. At this point, the prediction formula of flexural strength is: σcu = ηl ηθ (σf )εmu Vf + σmu Vm

(6.4)

Here, ηl is the fiber length coefficient, which is the ratio between the mean stress and maximum stress of the fiber considering the influence of the fiber tip on the stress transferring. When the volume of fiber is high enough, most fiber did not fracture under load from matrix, causing more cracks in the matrix under counteraction load from fibers. Till the composite fracture, the load is entirely on the fibers. In this case, the composite showed complex fracture performance, and the prediction formula of its flexural strength is: σc = ηl ηθ σfu Vf

(6.5)

When σ cu = σ c , the critical fiber volume ratio of the two failure forms can be obtained as follows:

226

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Vfcr =

σmu ηl ηθ [σfu − (σf )εmu ] + σmu

(6.6)

Based on the simplified stress transfer theory provided by Cox, the ηl is as follows: ηl = 1 −

Lc 2L

(6.7)

(2) Prediction and analysis Including carbon fiber, silicon carbide fiber, steel fiber, quartz fiber, PVA fiber, and other kinds of mineral fibers (such as basalt fiber), the maximum strain of fiber is much higher than that of geopolymer. Thus, the strength of the Csf /geopolymer composites can be predicted by formula (6.3), (6.4), and (6.5). As for the 2D planar short fiber-reinforced composites, assuming that the planeoriented short fiber exhibits uniform distribution in the 2D direction and the distribution probability of the fiber is equal on the circular plane, the angle between the fiber direction and the load direction is θ, so the ηθ can be calculated from (6.8). 

π 2

ηθ = 0

2 2 cos θ dθ = ≈ 0.637 π π

(6.8)

Physical and mechanical properties of the carbon fiber and silicon carbide fiber are shown in Table 6.11. The bonding strength between carbon fiber and geopolymer matrix is ~3.7 MPa [3]. According to the formula (6.2) and the basic performance parameters of this material, the critical length of the short carbon fiber in this composite is 2.92 mm. Therefore, when the fiber length is 2 mm which is below the critical length, the strength of the composite can be predicted according to (6.3). When the fiber length is 7 and 12 mm, according to the (6.7), the fiber length coefficients ηl of the two composites are 0.793 and 0.878, respectively. And the critical fiber volume ratio is 1.46 vol%. Since all the calculated fiber volume are lower than the actual fiber content of 3.5 vol%, the strength of composite can be predicted based on the formula (6.5). Similarly, the bonding strength between SiC fiber and geopolymer is ~6.9 MPa, and the calculated critical length of SiC fiber is 2.82 mm. The strength of the Table 6.11 Performance parameters of the fibers Material

Strength (MPa)

Elastic modulus (GPa)

Shear modulus (GPa)

Poisson’s ratio

Diameter (μm)

Length (mm)

Carbon fiber

2700

230

88.46

0.3

8

2, 7, 12

Silicon carbide fiber

3000

200

–

13

2, 5, 8

–

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

227

composite is predicted when the fiber length is 2 mm according to (6.3). Similarly, when the length of SiC fiber is 5 and 8 mm, the critical fiber volume ratio is lower than the actual content of 2.0 vol%. Thus, the strength of the composites can also be predicted based on the formula (6.5). As for the random short carbon fiber-reinforced geopolymer composite described in the 3D-Csf /geopolymer composites, all the short fibers can be regarded as evenly distributed in three-dimensional directions. Therefore, the fiber oriental coefficient is expressed as formula (6.9).  ηθ =

π 2

cos θ sin θ dθ =

0

1 2

(6.9)

Since the length of the short carbon fiber in the composite is only 1 mm, it is shorter than the critical fiber length of 2.92 mm. Thus, the strength prediction of the composite should be calculated according to the formula (6.3). The predicted flexural strength of the composites is shown in Table 6.12 based on the above-mentioned formulas, and was compared with the experimental values. The prediction model assumes that the matrix is perfect without any defects. However, it is difficult for the geopolymer to be free of defects such as microcracks and pores. Meanwhile, microgaps would appear between the fiber and the geopolymer matrix, which would weaken the bonding strength. Both factors make most of the predicted values higher than the real ones, and the deviation increases with increase of the defects in the composites. Table 6.12 Predicted flexural strength of the composites

Composite

Composite strength Real value (MPa)

Predicted value (MPa)

Relative error (%)

61.5

64.3

+4.6

C7V35

91.3

79.5

−12.9

C12V35

84.6

88.3

+4.2

C1V05

11.3

19.6

+73.4

C1V10

14.8

20.9

+41.2

C1V20

25.6

23.6

−7.8

C1V30

22.1

26.2

+18.5

C2V35

Note C means short carbon fiber and the following numbers represent the length of the fiber; V and the following numbers are the volume content of the fiber. Conversion factor between tensile and flexural strength for random and 2D short fibers are 1.2 and 1.66

228

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

6.6.2 Modulus Prediction Halpin–Tsai, Mori–Tanaka, and Tengdon–Weng models are the commonly used models for the stiffness prediction of the composites. Halpin–Tsai model is a simple and accurate model based on both experiment and theory, and η can be obtained based on the theoretical and experimental calculation. The Mori–Tanaka model is simple, and the interaction between different fibers in the composites is considered. When the fiber content in the composite is high, the calculated value is accurate. The Tengdon-Weng model is an improved model of the Mori–Tanaka model. (1) Prediction of unidirectional short fiber-reinforced composites The Halpin–Tsai model is a semi-empirical model to predict short fiber-reinforced composites for a long period. Charles [24] proposed the following assumptions for the Halpin–Tsai model: (1) Both the fiber and matrix are linearly elastic. The matrix is isotropic. However, the fiber can be isotropic or transversally isotropic. (2) All the fibers have the same shape and size. Besides, they are axisymmetric. (3) Interfacial bonding between fiber and matrix are good, which should be maintained during the process of deformation. In this way, sliding, separation, and microcracks at the interface between the fiber and the matrix can be ignored. The Halpin–Tsai model is as follows: 1 + ς ηvf P = Pm 1 − ηvf

η=

Pf /Pm − 1 Pf /Pm + ς

(6.10)

Here, P is the elastic constant of composite, Pf is the elastic constant of the fiber, Pm is the elastic constant of the matrix, ς is a parameter which is determined by both the Poisson’s ratio and elastic properties of the matrix, as shown in Table 6.13. The unidirectional short fiber-reinforced composite is orthotropic. Thus, we denote 1 as the direction of the fiber, 2 as the direction perpendicular to the fiber, and 3 as the direction of thickness, respectively. When the stress is in-plane, it is under a plane stress state. There is only surface stresses σ 1 , σ 2 , τ 12 , and other stress is zero, that is, σ 3 = τ 23 = τ 13 = 0. Gu Zhenlong [25] provided the orthoaxis macroscopic constitutive equation when the fiber was parallel or perpendicular to the external stress direction: Table 6.13 Parameter values in the Halpin–Tsai equation [24]

P

Pf

Pm

ς

E 11

Ef

Em

2(L/d f )

E 12

Ef

Em

2

G12

Gf

Gm

1

G23

Gf

Gm

(1 + υ m )/(3 − υ m − 4υ m 2 )

υ 12

–

–

V fυf + V mυm

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

⎧ ⎫ ⎛ ⎞⎧ ε ⎫ ⎪ ⎪ Q 11 Q 12 0 ⎨ 1 ⎪ ⎨ σ1 ⎪ ⎬ ⎬ ⎠ ⎝ σ2 = Q 21 Q 22 0 ε2 ⎪ ⎪ ⎪ ⎩ ⎪ ⎭ 0 0 Q 66 ⎩ γ12 ⎭ τ12

229

(6.11)

Namely, {σ } = [Q]{ε}, and [Q] is the orthoaxis stiffness coefficient. During the mechanical analysis, stress–strain relationship of off-axis macroscopic constitutive equation is used. In Fig. 6.55, 1O2 is the principal axis, xOy is the offaxis (actual direction of external stress), and the angle between the axis x and 1 is θ. After coordinate conversion, the equation is as follows: ⎧ ⎫ ⎛ ⎞⎧ σ ⎫ ⎪ sin2 θ −2 sin θ cos θ ⎪ cos2 θ ⎬ ⎬ ⎨ 1⎪ ⎨ σx ⎪ 2 2 ⎝ ⎠ σy = σ2 cos θ 2 sin θ cos θ sin θ ⎪ ⎪ ⎪ ⎭ ⎩ ⎪ sin θ cos θ − sin θ cos θ cos2 θ − sin2 θ ⎩ τ12 ⎭ τx y

(6.12)

make ⎛

[T ]−1

⎞ sin2 θ −2 sin θ cos θ cos2 θ = ⎝ sin2 θ cos2 θ 2 sin θ cos θ ⎠ sin θ cos θ − sin θ cos θ cos2 θ − sin2 θ

(6.13)

In order to represent the strain vector more accurately, the Router is used: ⎛

⎞ 100 [R] = ⎝ 0 1 0 ⎠ 002

(6.14)

⎧ ⎫ ⎧ ⎫ ⎪ ⎪ ⎨ εx ⎪ ⎬ ⎬ ⎨ σx ⎪ −1 −1 σ y = [R][T ] [R] [Q][T ] ε y ⎪ ⎪ ⎩ ⎪ ⎭ ⎭ ⎩ ⎪ τx y γx y

(6.15)

⎧ ⎫ ⎛ ⎞⎧ ε ⎫ ⎪ Q 11 Q 12 Q 16 ⎪ ⎬ ⎬ ⎨ σx ⎪ ⎨ x ⎪    ⎝ ⎠ σ y = Q 12 Q 22 Q 26 εy ⎪ ⎪ ⎪ ⎭ ⎩ ⎪ Q 16 Q 26 Q 66 ⎩ γx y ⎭ τx y

(6.16)

It can be proved:

If

Thus, the off-axis stiffness coefficient of the value of [Q ] can be calculated as follows: Q 11 = Q 11 cos4 θ + Q 22 sin4 θ + (2Q 12 + 4Q 66 ) sin2 θ cos2 θ

230

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.56 Representative unit and model simplification of plane-oriented fiber-reinforced composites [26]

Q 22 = Q 11 sin4 θ + Q 22 cos4 θ + (2Q 12 + 4Q 66 ) sin2 θ cos2 θ Q 66 = (Q 11 + Q 22 − 2Q 12 − 2Q 66 ) sin2 θ cos2 θ + Q 66 (sin4 θ + cos4 θ ) Q 12 = (Q 11 + Q 22 − 4Q 66 ) sin2 θ cos2 θ + Q 12 (sin4 θ + cos4 θ ) Q 16 = sin θ cos θ [Q 11 cos2 θ − Q 22 sin2 θ + (Q 12 + 2Q 66 )(sin2 θ − cos2 θ )] Q 26 = sin θ cos θ [Q 11 sin2 θ − Q 22 cos2 θ + (Q 12 + 2Q 66 )(cos2 θ − sin2 θ )] (6.17) (2) Prediction of 3D short fiber-reinforced composites Zhao [26] used the mechanical model of unidirectional fiber-reinforced composites to predict the modulus of the composites. As shown in the coordinate system in Fig. 6.56, if we need to calculate the elastic modulus in the axis 1, we consider axis 1 as the main direction. θ (0 ≤ θ ≤ π/2) represents the angle between fiber orientation and axis 1, and ϕ(0 ≤ ϕ ≤ 2π) represents the orientation angle of each layer. Assuming ϕ = 0, the composite is a 2D composite. As shown in Fig. 6.56a, the 3D short fiber-reinforced composite can be decomposed into the layered 2D short fiber-reinforced composites with layer angle of ϕ(0 < ϕ < 2π). Furthermore, each layer can be divided into several different unidirectional fiber-reinforced composite layers with orientation angle of θ (0 < θ < π/2) (Fig. 6.56c), which is the angle between the fiber direction and the selected main direction. Therefore, the mean fiber orientation θ mean can be obtained as:  θmean =

θmax θmin

θg(θ )dθ

(6.18)

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

231

Here, θ min is the minimum value of fiber orientation angle, and θ max is the maximum value of fiber orientation angle. After simplifying the 3D short fiber-reinforced composites according to the above method, we can first calculate the modulus of the 2D short fiber-reinforced composites based on the model of unidirectional fiber-reinforced composites. Then, according to similar methods, the modulus of 3D composite can be obtained. The detailed calculation steps are as follows: (1) Modulus E 11 , E 22 , G12 , G23 , and υ 12 can be calculated based on the modulus prediction model of unidirectional short fiber-reinforced composite in Fig. 6.56. (2) According to the elastic modulus in step (1), (6.11) is used to calculate the orthoaxis stiffness coefficient [Q] of the unidirectional layer with angle θ. (3) Transformation of [Q] in 1–2 coordinate system into [Q ] in the x–y coordinate system according to formula (6.17). (4) Take the average value of [Q ] of any fiber direction as the plane stiffness matrix [C ij ] according to formula (6.19). [Cij ] =

1 π

 0

π

Q ij dθ

(6.19)

Meanwhile, the stress–strain relationship of isotropic materials under the plane stress can be written as: ⎧ ⎫ ⎛ E− ⎪ 2 ⎬ ⎜ 1−υ ⎨ σ11 ⎪ − − υ ⎜ E σ22 = ⎝ 1−υ 2 ⎪ ⎭ ⎩ ⎪ τ12 0

− −

υE 1−υ 2 − E 1−υ 2

0

0 0 −

E 2(1+υ)

⎞⎧ ⎫ ⎪ ε11 ⎪ ⎬ ⎟⎨ ⎟ ε22 ⎠⎪ ⎪ ⎩ ⎭ 2ε12

(6.20)

(5) The elastic modulus of two-dimensional planar composite can be obtained according to (6.16) and (6.17). (6) Similar with the above method, the Young’s modulus of 3D short fiberreinforced composites can be calculated. (3) Prediction of Csf /geopolymer composites The modulus of the composite was predicted based on the Halpin–Tsai equation. The Poisson’s ratio of the geopolymer was ~0.227. It is also assumed that the pores in the composites are closed spherical pores. According to the effect of porosity on the modulus of composites, elastic modulus of composites is obtained as follows: E = E t (1 − 1.9 p + 0.9 p 2 )

(6.21)

Here, E t is the predicted elastic modulus of composite, and p is the porosity. Table 6.14 lists the predicted modulus of the 2D-Csf /geopolymer composites with different lengths (fiber volume ratio of 3.5 vol% and porosity of 16.5%). It can be seen that the predicted modulus has a good matching with the experiment value,

232

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Table 6.14 The predicted elastic modulus of different composites Composites

Elastic modulus Real value (GPa)

Predicted value (GPa)

Relative error (%)

C2V35

6.7

6.5

−2.9

C7V35

6.7

6.6

−1.5

C12V35

6.7

6.6

−1.5

Note C is the short carbon fiber and the following number is the length of the fiber, and V and the following number is the content of the fiber by volume

and the relative error is only 2.9–1.5%. This indicates that using the Halpin–Tsai equation and the simplified model of plane-oriented fiber composite and considering the effect of the porosity of the composite on the elastic modulus, the elastic modulus of the short fiber-reinforced geopolymer matrix composite can be predicted relatively accurately.

6.6.3 Pulling-Out Energy Prediction The strengthening effect from fiber on the composite is not only reflected in strength improvement, but also in the high service reliability during application. There are many mechanisms to improve the fracture energy of the fiber toughened composites, such as the debonding at the fiber and matrix interface, the friction after debonding, the deformation of the matrix, the fracture between the fiber and the matrix, and the pulling-out of the fiber. As for a certain composite system, certain toughening mechanisms play the most important role in improving the fracture energy of the composite, while others may only play a supplementary role. For the Csf /geopolymer composites, due to the great different strength between the fiber and the matrix, fiber pulling-out is the main toughening mechanism during the fracture process and absorbs most of the fracture energy. Therefore, the calculation of the pulling-out energy during the fracture process can represent the fracture behavior of composites. (1) Pulling-out energy of a single fiber in short fiber-reinforced composites As for the unidirectional short fiber-reinforced composites, the critical fiber length L c is discussed before. As for the non-unidirectional short fiber-reinforced composites, there is a certain angle between fiber and loading direction. So, the critical fiber length is changed. According to the results by Fu and Lauke [27], the critical fiber length L cθ of non-unidirectional short fiber is: L cθ =

σfuθ df 2τi exp(μθ )

(6.22)

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

233

Fig. 6.57 Schematic diagram of short fiber pulling out of matrix [27]

Here, σ fuθ is the effective stress on fibers, which is called the fiber inclined strength. σfuθ = σfu (1 − A tan θ )

(6.23)

L c (1 − A tan θ ) exp(μθ )

(6.24)

L cθ =

Here, A is the fiber inclined strength coefficient, μ is the friction coefficient. For the non-unidirectional short fiber-reinforced composites, the friction between the fiber and the matrix must be considered when the fiber is pulled out from the matrix at a certain angle. Thus, the interface shear stress is shown as (6.25). τ (δ) = a0 + a1 δ + a2 δ 2

(6.25)

Here, a0 , a1 , and a2 are empirical constants, δ is the fiber pulling-out length. As shown in Fig. 6.57, it is assumed that there is angle θ between the crack propagation direction in the composite and the fiber distribution direction. When the cracks tip encounters the fiber, the shorter fiber would be pulled out for a certain length. Considering the friction between the fiber and matrix, the force used to pull out the fiber is: P(l, s, θ ) = τf (s)π df (l − s) exp(μθ )

(6.26)

If the fiber is non-deformable, the length of crack propagation is equal to the slip distance of the fiber, that is, s = δ, as follows: P(l, δ, θ ) = τf (δ)π df (l − δ) exp(μθ ) when l ≥ δ P(l, δ, θ ) = 0 when l ≥ δ Therefore, the fiber pulling-out energy is:

(6.27)

234

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

 Wpo (l, θ ) =

l δ=0

P(l, δ, θ )dδ when l ≤ L cθ

Wpo (l, θ ) = 0

when l ≥ L cθ

(6.28)

(2) Pulling-out energy of fibers in short fiber-reinforced composites Assume that the composite is of certain length, width, and height of c1 , c2 , and c3 . c3 is parallel to the loading direction. Ac and Af are the fracture areas of composite and fiber, respectively. V f and L  are the volume content and average length of fiber, respectively. Thus, the number of fibers in the sample is: N=

Ac c3 Vf Af L 

(6.29)

Assuming distribution functions of length and angle of fibers in the composite are independent, the number of fibers in the range of L ~ L + dL, θ ~ θ + dθ is: d N = N f (L)g(θ )dLdθ

(6.30)

If the fiber distribution in the composite is even, then the number of fibers within the range of L ~ L + dL, θ ~ θ + dθ is: dNc =

dN L cos(θ ) c3

(6.31)

Therefore, the number of fiber length in the range of l ~ l + dl is:  dNc (l) = dNc

2dl L

 (6.32)

In the above section, we have discussed the fiber critical length at an angle between the fiber and loading direction in the composite. Affected by the interfacial friction between fibers and the matrix, as well as the fiber inclined strength, the fibers would fracture when l ≥ L cθ / 2. Formula (6.33) can be used to describe the fiber fracture:  U (l) =

1 l < L cθ /2 0 l ≥ L cθ /2

(6.33)

And the fiber pulling-out energy of the composite can be expressed as: Wpo

1 = Ac



π 2

θ=0



L=L max L=L min

 2 cos(θ ) dldLdθ Wpo (l, θ )U (l)N f (L) × g(θ ) c3 l=0 (6.34)



L 2



If the length of the fiber is fixed, the above formula can be changed into:

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

Wpo =

8Vf πdf2 L





π 2

θ=0

L 2

Wpo (l, θ )U (l)g(θ ) cos(θ )dldθ

235

(6.35)

l=0

Substitute (6.28) into the above equation: Wpo =



8Vf df L



π 2

θ=0

L 2

l=0



α0 l 2 α1l 3 α2 l 4 + + 2 6 12

 exp(μθ )U (l)g(θ ) cos(θ )dldθ (6.36)

In order to ensure σ fuθ = σ fu (1 − Atanθ ) ≥ 0, there should be a maximum angle between fiber and the loading direction: θmax = arctan

  1 A

(6.37)

Thus, Wpo =

8Vf df L



θmax θ=0



L 2



l=0

α0 l 2 α1l 3 α2 l 4 + + 2 6 12

 exp(μθ )U (l)g(θ ) cos(θ )dldθ (6.38)

Assuming all the fibers showed a constant angle to the loading direction, the fiber pulling-out energy can be: Wpo1

Wpo2

8Vf = df L 8Vf = df L



L 2

l=0



L cθ 2

l=0



α0 l 2 α1l 3 α2 l 4 + + 2 6 12





α0 l 2 α1l 3 α2 l 4 + + 2 6 12

exp(μθ ) cos(θ )dl L < L cθ

(6.39)

 exp(μθ ) cos(θ )dl L > L cθ

(6.40)

Meanwhile, fiber pulling-out energy of the composite can also be expressed as:  Wpo =

θmax θ=0

(Wpo1 + Wpo2 )g(θ )dθ

(6.41)

(3) Prediction of Csf /geopolymer composites The angle distribution function of short fiber-reinforced composites can be obtained by (6.41), as proposed by Suemasu et al. [28]:  Wpo =  g(θ ) =

2 π

θmax θ=0

(Wpo1 + Wpo2 )g(θ )dθ

Short fibers in two dimensions sin θ Short fibers in three dimensions

(6.42)

(6.42)

236

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Similarly, in order to calculate the work of fracture of 2D-Csf/geopolymer composite, g(θ ) = π2 is used. Assuming that the interfacial shear stress between the fiber and matrix is constant, a1 = a2 = 0, τ f = a0 , the fiber pulling-out energy of the composite can be expressed as: Wpo

 arctan( A1 )  L2 8Vf α0 = l 2 exp(μθ ) cos(θ )dldθ π df L θ=0 l=0   arctan( 1 )  L cθ A 2 2 + l exp(μθ ) cos(θ )dldθ θ=0

(6.43)

l=0

It can be seen from the above formula that V f and d f are not involved in the integral calculation. Therefore, according to the relationship of the formula (6.43), it can be concluded that the fiber pulling-out energy of the composite is proportional to the fiber volume content, but inversely proportional to the fiber diameter. For the 2D-Csf /geopolymer composites, we have V f = 0.035, d f = 8 μm, a0 = 3.7 MN m−2 . Assume μ = A = 0.25 in formula (6.43). Mathematics 7.0, from Wolfram Research Company in the United States is adopted (the software is used for the integral calculation in this chapter). After the integral calculation, Fig. 6.58 can be obtained. It can be noted that when the critical fiber length is fixed, with the increase of fiber length, the fiber pulling-out energy increased first, reached the maximum value, and then gradually decreased. Therefore, it can be concluded that increasing fiber length does not necessarily have positive effect on the fiber pulling-out energy of the composite. When the fiber length is longer than the critical fiber length, fiber is likely to be fractured rather than pulling-out. Since the energy consumed during fiber pulling-out is greater than that of fiber fracture, the work of fracture does not necessarily increase with fiber length. In 2D-Csf /geopolymer composites, the critical fiber length is 2.92 mm. When the short fiber length is 2 mm, shorter than the critical length, fiber is most likely to be Fig. 6.58 Fiber pulling-out energy of the composites with different fiber lengths [3]. a L c = 0.4 mm, b L c = 1 mm, c L c = 1.6 mm, d L c = 2.2 mm, e L c = 2.8 mm, f L c = 3.4 mm, g L c = 4 mm

6.6 Mechanical Property Prediction of the Csf /Geopolymer Composites

237

Fig. 6.59 Fiber pulling-out energy of the composites with different fiber lengths [3]. a μ = 0, b μ = 0.25, c μ = 0.5, d μ = 0.75, e μ = 1.0

pulled-out but the highest pulling-out energy cannot be obtained. When the short fiber length increased to 7 and 12 mm, it was much longer than the critical fiber length. Although fiber pulling-out energy can reach the highest value, fiber was more likely to be fractured, which resulted in the decrease of work of fracture when fiber length was 12 mm. In formula (6.43), assuming A = 0.25, V f = 0.035, d f = 8 μm, a0 = 3.7 MN m−2 , L c = 2.92 mm, Fig. 6.59 shows the fiber pulling-out energy with different fiber length and friction coefficient. When the fiber length is relatively short (≤1.5 mm), the higher the friction coefficient between the fiber and the matrix is, the higher the fiber pulling-out energy becomes. However, when the fiber is longer than the critical fiber length, with the increase in friction coefficient, bonding strength between the fiber and the matrix increased, which makes the fiber more likely to be fractured, leading to a decrease in pulling-out energy. In formula (6.43), assuming μ = 0.25, V f = 0.035, d f = 8 μm, a0 = 3.7 MN m−2 , L c = 2.92 mm. Figure 6.60 shows the fiber pulling-out energy with different fiber length and fiber inclined strength coefficient. As shown in Fig. 6.60, when the interfacial shear strength and the critical fiber length are fixed, with the increase of value A, the fiber pulling-out energy of the composite under the same fiber length is increased. According to the (6.35) and (6.36), with the increase in A, the critical fiber length increased, resulting in the higher fiber pulling-out energy in the composite. However, when the fiber length is much longer, the corresponding composites showed similar fiber pulling-out energy, which depended mainly on the interfacial bonding strength and friction coefficient between the fiber and the matrix.

238

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

Fig. 6.60 Fiber pulling-out energy of the composites with different fiber lengths [3]. a A = 0.25, b A = 0.5, c A = 0.75, d A = 1.0

6.7 Summary Compared with graphene, nanotube, and particles, short carbon fiber showed much enhanced strengthening and toughening effects on the related composites, and the following conclusions can be drawn: (1) For the random distribute short carbon fiber-reinforced geopolymer composites, when fiber content is 2.0 vol%, both flexural strength and fracture toughness reached the highest values, which were 67.3 and 103.6% higher than that of matrix. (2) 2D-Csf /geopolymer composites was prepared by impregnating sheet-like short carbon fiber preform using geopolymer slurry. The composites with 7 mm starting short carbon fiber length and 3.5 vol% fiber contents show dramatic strengthening effect from the Csf , as a result, much high flexural strength and work of fracture values are achieved, which increased by 4.4 times and 118 times, respectively, compared to that of the geopolymer matrix. The Csf /geopolymer composites show a typical pseudoplasticity behavior rather than brittle failure mode. And the main strengthening and toughening mechanisms are attributed to the fiber bridging and pulling-out. (3) Failure process of the Csf /geopolymer composites studied by in situ crack growth and propagation observation show that with the increase of the bending load, the beam sample keeps a nearly elastic deformation behavior at initial stages and exhibits an obvious displacement. The beam sample produces pseudoplastic deformation as the maximum load is reached. The propagation of the microcracks is arrested, and they tend to close to some degree accompanied by the main crack formation; as a result, the stress in the microcrack area is somewhat relaxed and the stress in the main crack area greatly increased. The fiber bridging the cracks effectively keeps the composites integrity and load bearable, which

6.7 Summary

239

makes the composites exhibit a clear non-catastrophic fracture behavior similar to that of the continuous fiber-reinforced ceramics. (4) Ni/P coating was successfully prepared on the carbon fiber surface by electroless plating method. Csf /geopolymer composites with Ni/P coating on fiber surface showed increased apparent density due to the fiber surface changing from hydrophobic to hydrophilic. When the thickness of Ni/P coating is 0.15 μm, the composite showed increased flexural strength and Young’s modulus, which was 7 and 60% higher than those of the uncoated counterpart. However, work of fracture of the composites with Ni/P coating showed significantly decreasing trend, and failure modes transferred from the initial non-catastrophical mode to brittle one, which can be attributed to the increased interfacial bonding strength between the fiber and the matrix. (5) When exposed at a temperature within 1000 °C, Csf /geopolymer composites showed high linear thermal stability. Beyond 1000 °C, the composites showed substantial shrinkage due to sintering. The presence of α-Al2 O3 particles in the composites clearly increases the volume stability particularly at high temperatures, thus improves the mechanical properties of the composites subjected to high temperature treatment. With the increasing thermal treatment temperature from 400 to 1200 °C, both flexural strength and work of fracture first decreased and reached the lowest value for the composites exposed at 800 °C, and then increased slightly when exposed at 1000 °C. (6) Based on the conventional strength theory and Halpin–Tsai model of composite, predictions for both flexural strength and modulus of short fiber-reinforced geopolymer composites were given considering the effect of fiber orientation, volume, and length. The predicted values were comparable to the experimental results with acceptable deviation, and both models can be used for the structural design for this kind of composites. Meanwhile, the fiber pulling-out energy of the Csf /geopolymer composites is not only proportional to the fiber volume content and inversely proportional to the fiber diameter, but also increases with the increase in critical fiber length L c and fiber inclined strength coefficient A of the composite. The increase in the interfacial friction coefficient u in the composite is beneficial to the fiber pulling-out energy when the fiber is short. The fiber pulling-out energy of composite increases first and then decreases with the increase in fiber length.

References 1. T. Lin, D. Jia, P. He, M. Wang, D. Liang, Effects of fiber length on mechanical properties and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 497, 181–185 (2008) 2. J. Davidovits, Geopolymers and geopolymeric materials. J. Therm. Anal. 35, 429–441 (1989) 3. T. Lin, Mechanical Properties and Fracture Behavior of Csf(Al 2 O3 p) Reinforced Geopolymer Matrix Composites (in Chinese) (Thesis for the Doctor Degree, Harbin Institute of Technology, 2009)

240

6 Short Carbon Fiber (Csf )-Reinforced Geopolymer Matrix Composites

4. T. Lin, D. Jia, M. Wang, P. He, D. Liang, Effects of fibre content on mechanical properties and fracture behaviour of short carbon fibre reinforced geopolymer matrix composites. Bull. Mater. Sci. 32, 77–81 (2009) 5. D.R. Mulligan, S.L. Ogin, P.A. Smith, G.M. Wells, C.M. Worrall, Fibre-bundling in a shortfibre composite: 1. Review of literature and development of a method for controlling the degree of bundling. Compos. Sci. Technol. 63, 715–725 (2003) 6. J.M.L. Reis, A.J.M. Ferreira, Assessment of fracture properties of epoxy polymer concrete reinforced with short carbon and glass fibers. Constr. Build. Mater. 18, 523–528 (2004) 7. S. Kumaria, S. Kumar, R.N. Singh, The first matrix cracking behavior of fiber-reinforced ceramic matrix composites. Acta Mater. 45, 5177–5185 (1997) 8. J. Tsai, A.K. Patra, R. Wetherhold, Numerical simulations of fracture-toughness improvement using short shaped head ductile fibers. Compos. A Appl. Sci. Manuf. 34, 1255–1264 (2003) 9. H. Huang, R. Talreja, Numerical simulation of matrix micro-cracking in short fiber reinforced polymer composites: Initiation and propagation. Compos. Sci. Technol. 66, 2743–2757 (2006) 10. S. Sirivedin, D.N. Fenner, R.B. Nath, C. Galiotis, Matrix crack propagation criteria for model short-carbon fibre/epoxy composites. Compos. Sci. Technol. 60, 2835–2847 (2000) 11. J. Li, R. Luo, Y. Bi, Q. Xiang, C. Lin, Y. Zhang, Na. An, The preparation and performance of short carbon fiber reinforced adhesive for bonding carbon/carbon composites. Carbon 46, 1957–1965 (2008) 12. Y.L. Zhang, Y.M. Zhang, J.C. Han, Y.Y. Han, W. Yao, Fabrication of toughened Cf/SiC whisker composites and their mechanical properties. Mater. Lett. 62, 2810–2813 (2008) 13. T. Lin, D. Jia, P. He, M. Wang, In situ crack growth observation and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 527, 2404–2407 (2010) 14. J.P. Singh, D. Singh, M. Sutaria, Ceramic composites: roles of fiber and interface. Compos. A Appl. Sci. Manuf. 30, 445–450 (1999) 15. S.J. Park, Y.S. Jang, K.Y. Rhee, Interlaminar and ductile characteristics of carbon fibersreinforced plastics produced by nanoscaled electroless nickel plating on carbon fiber surfaces. J. Colloid Interface Sci. 245, 383–390 (2002) 16. Z. Shi, X. Wang, Z. Ding, The study of electroless deposition of nickel on graphite fibers. Appl. Surf. Sci. 140, 106–110 (1999) 17. V.S. Mironov, M. Park, Electroflocking technique in the fabrication and performance enhancement of fiber-reinforced polymer composites. Compos. Sci. Technol. 60, 927–933 (2000) 18. T.-S. Lin, D.-C. Jia, P.-G. He, M.-R. Wang, Thermal-mechanical properties of short carbon fiber reinforced geopolymer matrix composites subjected to thermal load. J. Central South Univ. Technol. 16, 881–886 (2009) 19. Subaer, A. van Riessen, Thermo-mechanical and microstructural characterisation of sodiumpoly(sialate-siloxo) (Na-PSS) geopolymers. J. Mater. Sci. 42, 3117–3123 (2007) 20. W.M. Kriven, M. Gordon, J.L. Bell, Geopolymers: nanoparticulate, nanoporous ceramics made under ambient conditions. Microsc. Microanal. 10, 404–405 (2004) 21. F. Guan, H. Zhong, G.-y Liu, S.-G. Zhao, L.-Y. Xia, Flotation of aluminosilicate minerals using alkylguanidine collectors. Trans. Nonferr. Metals Soc. China 19, 228–234 (2009) 22. H.L. Cox, The elasticity and strength of paper and other fibrous materials. Br. J. Appl. Phys. 3, 72–79 (1952) 23. W.H. Bowyer, M.G. Bader, On the re-inforcement of thermoplastics by imperfectly aligned discontinuous fibres. J. Mater. Sci. 7, 1315–1321 (1972) 24. C.L. Tucker Iii, E. Liang, Stiffness predictions for unidirectional short-fiber composites: review and evaluation. Compos. Sci. Technol. 59, 655–671 (1999) 25. Z. Gu, Mechanics of Short Fiber Reinforced Composite (in Chinese) (National Defense Industry Press, 1987) 26. Y. Zhao, Prediction Study on the Mechanical Properties of Short Fiber Reinforced Composites (in Chinese) (Zhengzhou University, 2004)

References

241

27. S. Fu, B. Lauke, The fibre pull-out energy of misaligned short fibre composites. J. Mater. Sci. 32, 1985–1993 (1997) 28. H. Suemasu, A. Kondo, K. Itatani, A. Nozue, A probabilistic approach to the toughening mechanism in short-fiber-reinforced ceramic-matrix composites. Compos. Sci. Technol. 61, 281–288 (2001)

Chapter 7

Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced Geopolymer Matrix Composites

Abstract Although short carbon fiber-reinforced geopolymer composite exhibited good mechanical properties, it is well recognized that carbon fiber is susceptible to be oxidized to CO2 at a temperature as low as 500 °C in air atmosphere, which is detrimental to the mechanical performance of the composites when it is exposed to high temperature in air. SiC fiber showed similar mechanical performance with carbon fiber but much higher oxidation resistance, and it can retain 80% room-temperature strength even at 1200 °C in air, which makes it a prominent high-temperature reinforcing candidate. In this chapter, both short SiC fiber and hybrid short SiC and carbon fiber-reinforced geopolymer composites were prepared, and the effects of both fiber contents and fiber length on the microstructure, interfacial state, mechanical properties, and fracture behavior of the composites were systematically reported.

7.1 Short SiC Fiber-Reinforced Geopolymer Matrix Composites In this part, we first report a facile approach to prepare short SiC fiber (SiCsf ) reinforced geopolymer composites (SiCsf /geopolymer), followed by investigating the effects of SiCsf content and length on the microstructures and mechanical properties of the composites. Furthermore, their in situ crack growth behaviors during the bending test are recorded to reveal the fracture behavior and toughening mechanism of the composites.

7.1.1 Preparation Process Geopolymer with the chemical composition of SiO2 /Al2 O3 = 5, K2 O/SiO2 = 0.2, and H2 O/K2 O = 11 (mole ratio) is obtained by mixing metakaolin powder with potassium silicate solution. The metakaolin is prepared by calcining kaolin at 800 °C for 3 h. The potassium silicate solution is prepared by dissolving amorphous silica (Shanghai Dixiang Indus., China) into KOH (Tianjin Fuchen Indus., China) solution © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_7

243

244

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Table 7.1 Properties of SiCsf Brand

Density (g cm−3 )

Diameter (μm)

Tensile strength (GPa)

Modulus (GPa)

Elongation (%)

SLFC1

2.36

13 ± 0.5

1.5–1.6

140 ± 10

1.3–1.4

108 105

Mass loss, %

Fig. 7.1 Thermal gravimetry (TG) results of SiCsf in air, with permission from [1]

102 99 mass change:-1.02%

Mass change: 5.95%

96 93 90

0

200

400

600

800

Temperature, oC

1000

1200

and stirring for 48 h in order to dissolve the containing silica. SiCsf used in this study (Sailifei Ceramic Fiber Co., Ltd., China) has a diameter of 13 μm and an average tensile strength of 1.5–1.6 GPa, as shown in Table 7.1. The raw SiCsf is coated with a thin epoxy layer to form a bundle containing 1 K fibers. According to the TG results shown in Fig. 7.1, the coated epoxy layer can be removed from the fiber surface by high-temperature treatments, corresponding to a mass loss of ~1% when heating from room temperature to 370 °C. Therefore, in order to achieve a relatively homogeneous dispersion, specimens are first treated at 370 °C in air for 2 h, followed by dispersing the products in ethanol and performing ball milling with the metakaolin powder for 6 h. The dried mixture is then gradually added into the potassium silicate solution under ultrasonic-assisted stirring. After that, the resulting slurry is cast into polystyrene containers (70 mm in diameter and 8 mm in height), sealed, and cured at 70 °C for 48 h. Afterwards, the resulting SiCsf /geopolymer composite is further cured at 70 °C for another 72 h. To study the dependence of fiber content and fiber length on the properties of the resulting composites, we set the fiber contents as 0, 0.5, 1, 2, 3 vol%, and select the fiber lengths as 2, 5, and 8 mm. The chemical composition of SiCsf /geopolymer composites are listed in Tables 7.2 and 7.3.

7.1 Short SiC Fiber-Reinforced Geopolymer Matrix Composites

245

Table 7.2 Chemical composition of SiCsf /geopolymer composites with different fiber contents Materials no

KOH (g)

M

12.50

S05

12.50

S10

12.50

S20 S30

Sol-SiO2 (g)

Metakaolin (g)

SiCsf volume fraction (vol%)

SiCsf length (mm)

30.00

24.25

–

–

30.00

24.25

0.5

2

30.00

24.25

1

2

12.50

30.00

24.25

2

2

12.50

30.00

24.25

3

2

Table 7.3 Chemical composition of SiCsf /geopolymer composites with different fiber lengths Materials no

KOH (g)

Sol-SiO2 (g)

Metakaolin (g)

SiCsf volume fraction (vol%)

SiCsf length (mm)

M S2

12.50

30.00

24.25

–

–

12.50

30.00

24.25

2

2

S5

12.50

30.00

24.25

2

5

S8

12.50

30.00

24.25

2

8

7.1.2 Effect of SiCsf Content on Microstructure and Mechanical Properties (1) Phase composition The XRD patterns of pure geopolymer (labeled with 0 wt%) and SiCsf /geopolymer composites with different fiber contents are shown in Fig. 7.2. In line with early reports and our expectation, pure geopolymer has a halo peak around 27° 2θ originating from its amorphous network. Meanwhile, the resulting SiCsf /geopolymer Fig. 7.2 XRD patterns of pure geopolymer and SiCsf /geopolymer composites, with permission from [1]

246

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

composites also show similar XRD patterns to that of pure geopolymer regardless of SiCsf content (up to 3 wt%), exhibiting an amorphous peak around 27° 2θ and several crystalline peaks due to the presence of remaining quartz. This indicates that geopolymer network is well-remained after SiCsf doping and the introduction of SiCsf has no obvious effect on the phase composition of the resulting products. (2) Microstructure Microstructures of SiCsf /geopolymer composites with fiber contents of 0.5, 1.0, 2.0, and 3 vol% are shown in Fig. 7.3. No fiber agglomeration is observed and it is evident that all the composites show a relatively homogeneous distribution of SiCsf . Figure 7.4a exhibits the interface of SiCsf /geopolymer composites. It can be seen that SiCsf embedded in the geopolymer matrix shows similar morphology with that of pure SiCsf , as shown in Fig. 7.4b. This suggests no chemical reaction takes place between SiCsf and geopolymer matrix and SiCsf in the composites is mainly encapsulated by mechanical interlocking. Meanwhile, interfacial gaps between SiCsf and geopolymer matrix are noticed, implying the presence of detrimental defects induced by the introduction of SiCsf .

Fig. 7.3 Microstructures of SiCsf /geopolymer composites with different fiber contents. a 0.5 vol%, b 1 vol%, c 2 vol%, d 3 vol%, with permission from [1]

7.1 Short SiC Fiber-Reinforced Geopolymer Matrix Composites

(a)

247

(b)

Fig. 7.4 Interfacial state of SiCsf and geopolymer. a SiCsf in the composite, b pure SiCsf , with permission from [1]

(3) Mechanical properties and fracture behavior All the composites show non-brittle failure mode as observed in the load–displacement curves in Fig. 7.5a. With the increase in fiber contents, flexural strength, compressive strength, and work of fracture of the composites show similar trends, that is, increasing first, reaching the highest value at 2.0 vol%, and then decreasing, as shown in Fig. 7.5b. Taking the flexural strength as an example, it increases from 8.9 MPa for the pure geopolymer to 15.8 MPa for the composite with a fiber content of 2.0 vol%. It is also worth mentioning that the composite with a fiber content of 2.0 vol% possesses nine times higher work of fracture than that of the pure geopolymer. This further highlights the strengthening and toughening effect of SiCsf on enhancing the mechanical properties of geopolymer. However, further increasing the fiber content to 3.0 vol% leads to a slightly decrease in both the flexural strength

Fig. 7.5 Load–displacement curves (a) and mechanical properties (b) of SiCsf /geopolymer composites with different fiber contents, with permission from [1]

248

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

and the work of fracture. This might be attributed to a larger number of microscopic defects and interfacial gaps as a result of the higher SiCsf content. Therefore, we demonstrate that SiCsf /geopolymer composites with a fiber content of 2.0 vol% possesses the best mechanical performance in all the investigated samples in our study.

7.1.3 Effects of SiCsf Length on the Mechanical Properties In this part, we fix the SiCsf content as 2.0 vol% and investigated the effects of fiber length, ranging from 2 to 8 mm, on the mechanical properties of the composites. Microscopic images in Fig. 7.6 show that SiCsf evenly disperse in the geopolymer matrix regardless of the fiber length. Figure 7.7 compares the mechanical properties of geopolymer and SiCsf /geopolymer composites with different fiber lengths. With increasing the fiber length from 0 to 5 mm, the flexural strength of the composite gradually increases, reaching the maximum value of 49 MPa at the fiber length of 5 mm. Formula 7.1 estimates the critical fiber length for obtaining the optimal strengthening effect, L c = σfu df /(2τi )

(7.1)

where σ fu is the fiber tensile strength (1.5 GPa), d f is the fiber diameter (13 μm), and τ i is interfacial shear strength (3.45 MPa). The estimated fiber length is 2.83 mm, indicating that only if the introduced fiber length is over 2.83 mm, fiber pullout and fracture start to play a role in strengthening the geopolymer matrix. Such threshold is perfectly in line with our experimental observations that the composite with a fiber length of 5 mm shows the highest strength. However, with further increasing the fiber length to 8 mm, the strength of the composite falls to 34.6 MPa, which is 29.4% lower than the maximum value obtained by using 5 mm SiCsf. This might be caused by the aggregation or winding of SiCsf with the increased fiber length. Meanwhile,

Fig. 7.6 Microstructures of SiCsf /geopolymer composites with different fiber lengths: a 2 mm, b 5 mm, c 8 mm, with permission from [1]

7.1 Short SiC Fiber-Reinforced Geopolymer Matrix Composites

249

Fig. 7.7 Mechanical properties of SiCsf /geopolymer composites with different fiber lengths, with permission from [1]

work of fracture shows similar trend as the flexural strength vs. fiber length, reaching the peak value of 1888.1 J m-2 with a fiber length of 5 mm, which is 63 times higher that of the pure geopolymer. The substantial enhancements in the flexural strength and work of fracture implied significant strengthening and toughening effects by introducing SiCsf. Figure 7.8 shows the fracture morphologies of the composites with different fiber lengths after the flexural strength tests. In all cases, considerable fiber pullout is noted due to the weak bonding strength between the fiber and the matrix. Such weak bonding is consistent with the aforementioned interfacial state of SiCsf and geopolymer, as shown in Fig. 7.4. The fiber pullout length is so long that it ensures the toughening effect of SiCsf , therefore, preventing the composites from catastrophic fracture. Meanwhile, the side view of the fracture (Fig. 7.8d) shows signatures of fiber debonding and microcracks, which contribute to the high work of fracture and pseudoplasticity behavior, respectively. Load–displacement curves of geopolymer and SiCsf /geopolymer composites with different fiber lengths are shown in Fig. 7.9. Geopolymer exhibits a typical brittle failure mode (Fig. 7.9a), where the load decreases dramatically after reaching the maximum value. However, for SiCsf /geopolymer composites, no catastrophic failure is observed and they all show elastic regions and nonlinear regions. After reaching the peak value, the load declines slowly with long tails, corresponding to remarkable fiber debonding and pullout, as shown in Fig. 7.8.

250

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Fig. 7.8 Fracture morphologies of SiCsf /geopolymer composites with different fiber lengths: a 2 mm, b 5 mm, c 8 mm, d side view image, with permission from [1] Fig. 7.9 Load–displacement curves of SiCsf /geopolymer composites with different fiber lengths, with permission from [1]

7.1 Short SiC Fiber-Reinforced Geopolymer Matrix Composites

251

7.1.4 In Situ Crack Growth and Fracture Behavior We further studied the in situ crack growth and fracture behavior of SiCsf /geopolymer composites with a fiber length of 5 mm during the three-point bending test by employing an environmental scanning electron microscope (ESEM). Load–displacement curve together with a series SEM images at different loads are provided in Figs. 7.10 and 7.11, respectively. As indicated in Fig. 7.10, the load–displacement curve can be divided into three stages: (i) the initial elastic region (load < 6 N), (ii) nonlinear region (6 N < load < 40 N), and (iii) falling region (load > 40 N). From the corresponding SEM image (Fig. 7.11a), it was deduced that the elastic deformation of the geopolymer matrix occurs during stage (i) with no obvious crack formation. However, at the beginning of stage (ii), apart from the main crack locates below the loading head, a large number of microcracks in parallel to the loading direction appear on the side of the sample (Fig. 7.11b). With the increase in load, these microcracks further grow up and propagate along the loading direction (Fig. 7.11c). Meanwhile, crack deflection is also noticeable. The formation of these microcracks is assigned to the low strength nature of the geopolymer matrix and the weak interfacial bonding strength between the matrix and the SiCsf . The formation and propagation of these microcracks is accompanied by fiber debonding, which consumes most of the fracture energy (as shown by the nonlinear deformation in the load–displacement curve) and retards stress concentration of the main crack. It is also worth noting that the width of the gap between these microcracks is in the range of 200–500 μm, which is much shorter than the fiber length (5 mm). As a result, these microcracks can be

Fig. 7.10 In situ load–displacement curves of SiCsf /geopolymer composites with a fiber content of 2.0 vol% and a fiber length of 5 mm, with permission from [1]

252

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Fig. 7.11 In situ SEM images of the composite at different loads as shown in Fig. 7.8, and (f) is the enlarged image of area F labeled in (e), with permission from [1]

bridged together by SiCsf to maintain the integrity of the composite. Whereas, with further increasing the load, the stress reaches the threshold for fiber pullout or fracture. This results in the rapid broadening of the main crack coupled with significant fiber pullout behaviors, as shown in Fig. 7.11d–f.

Fig. 7.12 Preparation of hybrid Csf and SiCsf reinforced geopolymer composites, with permission from [2]

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

253

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites Early studies have proved that the introduction of Csf is beneficial for improving the mechanical properties of geopolymer composites, in which non-brittle failure behaviors have been observed. However, Csf can easily undergo oxidation reactions at high temperatures in air, which limits the usage of Csf -reinforced geopolymer composites at elevated temperatures. Although SiCsf possesses better oxidation resistance, it is also much more expensive than Csf . In order to obtain low-cost and oxidationresistant geopolymer composites with superior mechanical properties, we developed hybrid Csf and SiCsf -reinforced geopolymer composites, where Csf and SiCsf are both introduced to the geopolymer matrix. On top of that, α-Al2 O3 particles are also incorporated into the hybrid system. The effects of α-Al2 O3 particles on the microstructure and mechanical properties of the composites are also discussed.

7.2.1 Preparation Process Methods The chemical composition of the geopolymer matrix is designed as SiO2 /Al2 O3 = 4.0, K2 O/SiO2 = 0.3, and H2 O/K2 O = 11 (mole ratio). The kaolin powder is heated at 800 °C for 2 h to get the metakaolin powder. Csf (Jilin Carbon Indus., China) used in this study has a diameter of 6–8 μm and an average tensile strength of 3 GPa. SiCsf (Jiangsu Sailifei Ceramic Fiber Indus., China) has a diameter of 13 μm and an average tensile strength of 1.5–1.6 GPa. Csf is cut into 7 mm and SiCsf is cut into 1, 3, and 5 mm, respectively. SiCsf used in this study is first treated at 370 °C in air for 2 h in order to remove the epoxy coating on the surface. Both Csf (7 mm) content and SiCsf (1, 3, and 5 mm) content are selected as 4.5 vol%. As shown in Fig. 7.12, we followed the typical preparation method to prepare the composites as reported in Chap. 6. Specifically, the alkaline solution is obtained by mixing a solution of KOH and the silica-sol under magnetic stirring for 3 days. Afterwards, the metakaolin powder is added into the alkaline solution and mixed for 45 min to get the geopolymer slurry. α-Al2 O3 powders with the chemical composition of Al2 O3 /metakaolin = 75 wt% are added into the geopolymer slurry, mixing for another 30 min to get α-Al2 O3 modified geopolymer composites. Hybrid Csf and SiCsf sheets are prepared by mixing Csf and SiCsf together under ultrasonic separation in ethanol, which is followed by filtering them out using a wire sieve. They are then added into the geopolymer slurry layer by layer (up to 20 layers) with the help of ultrasonic infiltrating. The resulting product is then degassed at 70 °C for 3 days using the vacuum-bag technique after mechanical pressing, in order to remove the pores in the composites. Finally, hybrid Csf and SiCsf reinforced geopolymer composites are heated to 800, 1000, and 1200 °C for 60 min in air. The composites with and without Al2 O3 particles are denoted as ACS and CS, respectively. The untreated geopolymer composites with different SiCsf lengths are denoted

254

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

as (A)C7S1, (A)C7S3, and (A)C7S5, respectively. The composites treated at given different temperatures are denoted as (A)CS-800, (A)CS-1000, and (A)CS-1200, respectively.

7.2.2 Microstructure and Mechanical Properties of Hybrid Csf and SiCsf Reinforced Geopolymer Composites (1) Thermal evolution Figure 7.13 presents the TG/DTA curves of hybrid Csf and SiCsf reinforced geopolymer composites that heated from room temperature to 1500 °C in air. The overall trend of the TG curve is similar to that of pure geopolymer [3–7], exhibiting mass loss of ~16.7 wt% upon heating to 760 °C. The mass loss in the temperature range from room temperature to 760 °C is attributed to the evaporation of free water, dehydration of OH groups, and oxidation of the fibers (250–760 °C). Meanwhile, the DTA curve shows an exothermic peak centered at 713 °C, which corresponds to the oxidation of Csf in air [8]. The broad exothermic shoulder observed at 760–1100 °C is associated with the crystallization process of the leucite phase. Note that the reaction between the fiber and the matrix is also possible in such Si–C–O system, as given by the formulas (7.2) and (7.3) [3]. C(s) + O2 (g) → CO2 (g)

(7.2)

SiO2 + C → SiC + CO(g)

(7.3)

Fig. 7.13 TG/DTA curves of Csf and SiCsf reinforced geopolymer composites that heated from room temperature to 1500 °C in air with a heating rate of 10 °C/min, with permission from [2]

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

Leucite α-SiO2

Relative intensity

Fig. 7.14 XRD patterns of Csf and SiCsf reinforced geopolymer composites after being treated at different temperatures, with permission from [2]

255

CS-1200 CS-1000 CS-800 CS-RT 10

20

30

40

50

60

70

80

90

2 theta,degree

(2) Phase composition and density Figure 7.14 provides the XRD patterns of Csf and SiCsf reinforced geopolymer composites after being treated at different temperatures. As it is shown, the untreated Csf and SiCsf reinforced geopolymer composites present the typical halo peak of geopolymer around 27° 2θ. Several crystalline peaks are also detected due to the remaining SiO2 in the raw materials such as kaolin. After high-temperature treatment at 800 °C, Csf and SiCsf reinforced geopolymer composites maintain their amorphous nature without any obvious change in the phase composition. However, after being treated at 1000 °C, characteristic peaks of leucite (K2 O·Al2 O3 ·4SiO2 ) emerge. This is consistent with the DTA curve, where a broad exothermic shoulder centered at 1000 °C is noticed (Fig. 7.13). Note that the characteristic peaks of the introduced fibers are invisible due to their low contents. The densities of Csf –SiCsf /geopolymer composites after being treated at different temperatures are depicted in Fig. 7.15. The densities of C7S1, C7S3, and C7S5 are 2.26, 2.16, and 2.21 g cm−3 , respectively. It is found that the densities of the composites slightly increase after being treated at 1200 °C, probably due to the crystallization of the matrix [6, 9], as inferred by the XRD patterns and the DTA curve. (3) Mechanical properties and fracture behavior Figure 7.16 and Table 7.4 present the mechanical properties of Csf and SiCsf reinforced geopolymer composites after high-temperature treatments. It can be seen that the flexural strength of all the samples decrease after high-temperature treatments, which may be attributed to the degradation of the fibers upon heating, especially for Csf [10]. Taking C7S3 as an example, after being treated at 1200 °C, the flexural strength of C7S3 decreases by ~22%.

256

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Fig. 7.15 Densities of Csf and SiCsf reinforced geopolymer composites after being treated at different temperatures. a 800 °C, b 1000 °C, c 1200 °C, with permission from [2]

Compared with untreated Csf and SiCsf reinforced geopolymer composites, Young’s modulus of the samples first increases after being treated at 800 °C and then decrease after being treated at 1000 °C. In addition, Young’s modulus of C7S1 and C7S3 reach the maximum values after being treated at 1200 °C, which may be due to the crystallization of the matrix and the strong interfacial bonding between the fibers and the matrix. The work of fracture of each composite tends to decrease with the increasing treatment temperature. After being treated at 1200 °C, the work of fracture of C7S1, C7S3, and C7S5 are only 150.7 ± 22.3, 808.7 ± 34.9, and 700.2 ± 24.7 J m−2 , respectively. The decline of the work of fracture can be rationalized by the degradation of the fibers and the failure of the interfacial bonding. Figure 7.17 displays the surface morphologies of C7S5 after different hightemperature treatments. The fibers show a relatively homogeneous distribution in the geopolymer matrix for all the cases. It could be seen that some small reticular cracks form on the surface after being treated at 800 °C. Further increasing the treatment temperature to 1000 °C, an increasing crack density is observed on the surface, which may be caused by the crystallization of the matrix and thermal mismatch between the matrix and the fibers during heating. Although it is challenging to distinguish Csf and SiCsf , it has been reported that Csf is easier to be oxidized upon heating [10, 11].

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

257

Fig. 7.16 Mechanical properties of Csf and SiCsf reinforced geopolymer composites after being treated at different temperatures. a Flexural strength, b Young’s modulus, c work of fracture, with permission from [2] Table 7.4 Mechanical properties of Csf and SiCsf reinforced geopolymer composites after being treated at different temperatures Samples

Flexural strength (MPa)

Young’s modulus (GPa)

Work of fracture (J m−2 )

C7S1

55.6 ± 1.1

15.7 ± 1.3

3359.1 ± 245.8

C7S3

41.3 ± 2.7

13.3 ± 0.7

2860.9 ± 178.2

C7S5

41.8 ± 2.3

12.5 ± 1.4

2399.7 ± 211.3

C7S1-800

55.3 ± 1.7

25.0 ± 2.4

3324.5 ± 221.7

C7S3-800

44.7 ± 1.3

24.8 ± 1.5

2343.4 ± 164.5

C7S5-800

30.6 ± 1.1

19.9 ± 0.9

2184.3 ± 207.2

C7S1-1000

19.3 ± 0.7

10.1 ± 0.3

1197.4 ± 109.2

C7S3-1000

29.6 ± 2.1

16.8 ± 0.8

1742.2 ± 89.3

C7S5-1000

30.0 ± 1.5

19.9 ± 1.1

1227.5 ± 104.2

C7S1-1200

12.8 ± 0.7

25.7 ± 2.5

150.7 ± 22.3

C7S3-1200

32.3 ± 3.2

27.4 ± 1.6

808.7 ± 34.9

C7S5-1200

25.0 ± 2.6

17.5 ± 1.3

700.2 ± 24.7

258

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Table 7.5 Mechanical properties of Csf –SiCsf –(Al2 O3p )/geopolymer composites [19] Samples

Flexural Young’s Samples strength (MPa) modulus (GPa)

Flexural Young’s strength (MPa) modulus (GPa)

C7S1

55.6 ± 1.1

15.7 ± 1.3

AC7S1

62.4 ± 2.1

24.1 ± 1.6

C7S3

41.3 ± 2.7

13.3 ± 0.7

AC7S3

50.6 ± 3.2

23.5 ± 1.8

50.1 ± 4.3

21.9 ± 1.2

C7S5

41.8 ± 2.3

12.5 ± 1.4

AC7S5

C7S1-800

23.8 ± 1.2

12.4 ± 0.7

AC7S1-800

34.28 ± 3.4

7.5 ± 0.6

C7S3-800

31.33 ± 2.2

12.5 ± 0.4

AC7S3-800

25.71 ± 1.7

6.5 ± 0.5

C7S5-800

AC7S5-800

17.83 ± 1.3

9.3 ± 0.3

26.45 ± 2.4

8.4 ± 0.2

C7S1-1000 52.56 ± 1.3

21.1 ± 2.3

AC7S1-1000 35.79 ± 1.9

41.5 ± 4.4

C7S3-1000 42.49 ± 2.7

27.5 ± 3.1

AC7S3-1000 28.26 ± 1.4

30.7 ± 2.3

C7S5-1000 43.89 ± 2.4

21.4 ± 2.9

AC7S5-1000 26.76 ± 3.1

31.1 ± 1.9

C7S1-1200 18.06 ± 1.5

24.1 ± 2.9

AC7S1-1200 15.47 ± 0.6

28.6 ± 2.1

C7S3-1200 22.58 ± 1.2

16.7 ± 1.2

AC7S3-1200 17.82 ± 0.9

21.6 ± 1.2

C7S5-1200 21.75 ± 2.1

19.2 ± 1.4

AC7S5-1200 13.89 ± 0.3

31.3 ± 2.3

Figure 7.18 exhibits the fracture morphologies of C7S5 after being treated at 800, 1000, and 1200 °C. In all the cases, fiber fracture and pullout originating from the weak bonding strength between the fibers and the matrix are noticed, leaving a large number of holes on the surface. Moreover, the length of the pulled fibers decreases with the increasing treatment temperature. Samples after being treated at 1200 °C show a relatively smooth surface with few pulled fibers. Figure 7.19 shows the load–displacement curves of Csf and SiCsf reinforced geopolymer composites. For the composites treated at 800 and 1000 °C, noncatastrophic fracture behaviors are observed. Two distinct regions, that is, elastic and nonlinear regions are noted, as a result of fiber bridging and sliding from the weak geopolymer matrix after debonding and pullout (with higher elastic modulus). There are several steep-like decrease after reaching the maximum load, especially for the samples that treated at 1000 °C. These steps correspond to the interface debonding, fiber fracture, and pullout [12–14]. The load–displacement curves of the composites treated at 1200 °C are similar to that of pure geopolymer, which follows the brittle failure mode. This can be explained by the degeneration of the fiber and the debonding at the interface due to the formation of leucite and the oxidation reaction at high temperatures, respectively. They are consistent with the fracture morphology of C7S5 after 1200 °C treatment, as indicated in Fig. 7.18.

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

259

Fig. 7.17 Surface morphologies of C7S5 after high-temperature treatments. a 800 °C, b1000 °C, c1200 °C, with permission from [2]

7.2.3 Microstructure and Mechanical Properties of Hybrid Csf –SiCsf –Al2 O3p Reinforced Geopolymer Matrix Composites In Chap. 6, it is evident that geopolymer-matrix composites modified by Al2 O3 particles possess much higher thermal stability than the pure geopolymer matrix. Therefore, in this section, in order to further enhance the thermal stability of hybrid Csf and SiCsf reinforced geopolymer composites, Al2 O3 particles were used to regulate the thermal stability of the composites and study the effects of Al2 O3 particles and treatment temperature on the microstructure and mechanical properties of the composites. (1) Thermal evolution TG/DTA curves of the Csf –SiCsf –(Al2 O3p ) reinforced geopolymer composites in Ar with a heating rate of 10 °C/min are compared in Fig. 7.20. These curves are similar to those of Cf /KGP that reported in our previous work [3]. As seen from the TG

260

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Fig. 7.18 Fracture morphologies of C7S5 after high-temperature treatments. a 800 °C, b 1000 °C, c 1200 °C, with permission from [2]

curves, the major mass loss before 750 °C are both ~10 wt% in both cases, which are mainly due to the evaporation of free water (RT ~200 °C), the dehydration of –OH groups (200–750 °C), respectively [6, 15]. Correspondingly, the DTA curves in the range of 200–750 °C of the composites exhibit broad exothermic peaks. From 1100 to 1500 °C, the composites show another weight loss of ~6 wt%. The wide exothermic peaks in the range from 1000 to 1350 °C in the DTA curves can be assigned to the crystallization process and the oxidation reaction of the fibers [16–18]. The weight loss in the range from 1100 to 1500 °C is mainly attributed to the interfacial reaction between the fiber and the matrix. Interfacial reaction between Si–O species in the matrix and Csf is shown in following formula (7.4), leading to the formation of SiC and CO [18]. The Gibbs free energy for reaction (7.4) can be calculated using the formula (7.5). SiO2 + C → SiC + CO(g)

(7.4)

G 0 = 605250 − 339.61T

(7.5)

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

261

Fig. 7.19 Load–displacement curves of Csf and SiCsf reinforced geopolymer composites, with permission from [2]

Fig. 7.20 TG/DTA curves of a Csf –SiCsf /geopolymer and Csf –SiCsf –Al2 O3p /geopolymer composites in Ar with a heating rate of 10 °C/min, with permission from [19]

According to the calculation results, the theoretical temperature to activate this reaction is 1509.2 °C. In our previous investigations regarding the Cf /geopolymer system [3], such interfacial reaction is observed above 1200 °C. Moreover, by increasing the temperature to 1400 °C, the thickness of β-SiC layer (confirmed by SEM and TEM analysis) increases to 1000 nm. In other words, high-temperature treatment leads to the occurrence of the interfacial reaction, which is accompanied

262

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

by the release of CO [18]. This leads to the weight loss above 1200 °C as shown in the TG curves. Meanwhile, due to the presence of a small amount of oxygen on the surface of SiCf (4.5 vol%), SiCx Oy may also form. However, when the treatment temperature is above 1200 °C, SiCx Oy phase can decompose, resulting in the weight loss of the composite. The possible reaction can be written as formula (7.6) [16, 17]: SiCx O y → SiC + C + CO + SiO

(7.6)

The DTA curve for Csf –SiCsf /geopolymer composites shows an exothermic peak centered at ~1400 °C, suggesting that Cf has been degraded upon heating. While the Csf –SiCsf –Al2 O3p /geopolymer composites exhibit an exothermic peak at ~1425 °C, slightly higher than the composites without Al2 O3 particles. This indicates that the addition of Al2 O3 particles may paly a role in postponing the degradation of fibers during the heating process. (2) Phase composition Figure 7.21 presents the XRD patterns of Csf –SiCsf /geopolymer composites with and without Al2 O3 particles after being treated at different temperatures. The composites are dominated by amorphous phase at 800 °C. Several characteristic peaks of quartz are present due to the remaining SiO2 in the raw kaolin. However, after being treated at 1000 °C, kalsilite phase emerges. Thus, it can be inferred that the fibers may shift the crystallization process of the geopolymer matrix to a higher temperature [10, 20]. After being treated at 1200 °C, leucite (K2 O·Al2 O3 ·4SiO2 ) crystallizes, but amorphous phase still remains. Compared with the Csf –SiCsf /geopolymer composites, crystalline Al2 O3 peaks can be observed in Csf –SiCsf –Al2 O3p /geopolymer composites, indicating that Al2 O3 does not react with the matrix at elevated temperatures.

(a)

(b)

Leucite

Leucite Kalsilite ¦ -ÁSiO2 ¦ -ÁAl2O3

Kalsilite

CS-1200 CS-1000 CS-800

Relative intensity

Relative intensity

¦ -ÁSiO2

ACS-1200 ACS-1000 ACS-800 ACS-RT

CS-RT 10

20

30

40

50

60

2 theta,degree

70

80

90

10

20

30

40

50

60

70

80

90

2 theta,degree

Fig. 7.21 XRD patterns of the a Csf –SiCsf /geopolymer composites and Csf –SiCsf – Al2 O3p /geopolymer composites before and after high temperature treatments, with permission from [19]

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

263

Fig. 7.22 Shrinkage in the direction of a perpendicular and b parallel to the lamination of Csf – SiCsf –(Al2 O3p )/geopolymer composites, with permission from [19]

(3) Microstructure Figure 7.22 manifests the thermal shrinkage of Csf –SiCsf /geopolymer composites and Csf –SiCsf –Al2 O3p /geopolymer composites during the high-temperature treatments at different temperatures. As shown in Fig. 7.22a, the thermal shrinkage in the direction perpendicular to the lamination of Csf –SiCsf –(Al2 O3p )/geopolymer composites range from −1% to ~2%, showing a little expansion behavior due to the fiber debonding during the crystallization of the matrix. However, in the direction parallel to the lamination, the composites exhibit much larger thermal shrinkage, exceeding 25% at 1200 °C. But Csf –SiCsf –Al2 O3p /geopolymer composites show lower thermal shrinkage than Csf –SiCsf /geopolymer composites, indicating that the addition of Al2 O3 is helpful to increase the thermal stability and stabilize the volume of the composites [21]. The microstructures of the polished surfaces of Csf –SiCsf –(Al2 O3p )/geopolymer composites before and after high temperature treatments are shown in Fig. 7.23. After high temperature treatments, a large number of short cracks and pores are formed in the composites. On increasing the treatment temperature, the thermal mismatch between the fiber and the matrix becomes more obvious, so that the cracks become wider and the fibers are more severely degraded. This can be explained by the thermal shrinkage during the viscous sintering and the large thermal mismatch between the matrix and the fibers. (4) Mechanical properties and fracture behavior Figures 7.24, 7.25, and Table 7.5 show and summarize the flexural strength and Young’s modulus of the Csf –SiCsf –(Al2 O3p )/geopolymer composites after being treated at different temperatures. It is obvious that the flexural strength and Young’s modulus of the composites decrease after the high-temperature treatments. For instance, the flexural strength of the composites treated at 800 °C is slightly lower than the untreated ones. Before sintering, the positive ions (K+ , H3O+ ) are present in the geopolymer network to balance the negative charge [AlO4 ] units. However, after being treated at high temperatures, the loss of free and hydration water in the network

264

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Fig. 7.23 Microstructures of the polished surfaces of the Csf –SiCsf –(Al2 O3p )/geopolymer composites after high-temperature treatments at different temperatures. a C7S5, b AC7S5, c C7S5-800, d AC7S5-800, e C7S5-1000, f AC7S5-1000, e C7S5-1200, f AC7S5-1200, with permission from [19]

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites 70

70

a)

60

Flexural strength, MPa

Flexural strength, MPa

60 50 40 30 20 10

b)

50 40 30 20 10

0 C7S1

C7S3

0

C7S5 AC7S1 AC7S3 AC7S5

C7S1

C7S3

70

70

c)

60

Flexural strength, MPa

Flexural strength, MPa

C7S5 AC7S1 AC7S3 AC7S5

Samples

Samples

60

265

50 40 30 20 10

d)

50 40 30 20 10

0 C7S1

C7S3

C7S5 AC7S1 AC7S3 AC7S5

Samples

0 C7S1

C7S3

C7S5

AC7S1 AC7S3 AC7S5

Samples

Fig. 7.24 Flexural strength of Csf –SiCsf –(Al2 O3p )/geopolymer composites after being treated at different temperatures. a Untreated, b 800 °C, c 1000 °C, d 1200 °C, with permission from [19]

leads to the imbalance of the internal charges and thus a decrease in the degree of polycondensation. The high temperature also leads to the formation of macrocracks, which decreases the mechanical properties of the composites. The flexural strength of the treated composites achieves the maximum values at 1000 °C. For example, C7S1 shows flexural strength of 52.56 ± 1.3 MPa after 1000 °C treatment, which is attributed to the formation of kalsilite and densification of the geopolymer matrix. Furthermore, the overall shrinkage (dL/L0 ) of the geopolymer matrix can be divided into four stages during the high temperature treatment, namely loss of free water (I, 25–100 °C), capillary strain (II, 100–300 °C), physical contraction of –OH groups (III, 300–800 °C), and viscous sintering (IV, 800–950 °C). Viscous sintering leads to the closure of pores and channels in the geopolymer network. According to previous studies [6], ~80% of the shrinkage takes place in the fourth stage. Meanwhile, the surface area reduces by ~40%, which is due to the reduction of specific surface area, densification, and elimination of large pores. After being treated at 1200 °C, the shrinkage is completed, and the flexural strength of the composites starts to decrease. This is because further shrinkage damages the fiber sheets and weakens the interaction between the fiber and the matrix. The Young’s modulus of the treated composites also shows the maximum values after 1000 °C treatments. It can be

266

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced … 50

30

a)

45

b)

40

Young's modulus, GPa

Young's modulus, GPa

25 20 15 10 5

35 30 25 20 15 10 5 0

0 C7S1

C7S3

C7S1

C7S5 AC7S1 AC7S3 AC7S5

C7S3

c)

45

d)

40

Young's modulus, GPa

40

Young's modulus, GPa

AC7S1 AC7S3 AC7S5

50

50 45

C7S5

Samples

Samples

35 30 25 20 15 10

35 30 25 20 15 10 5

5

0

0 C7S1

C7S3

C7S5

AC7S1 AC7S3 AC7S5

Samples

C7S1

C7S3

C7S5 AC7S1 AC7S3 AC7S5

Samples

Fig. 7.25 Young’s modulus of Csf –SiCsf –(Al2 O3p )/geopolymer composites after being treated at different temperatures. a Untreated, b 800 °C, c 1000 °C, d 1200 °C, with permission from [19]

deduced that the addition of Al2 O3p obviously increases the Young’s modulus of the composites. Fracture morphologies of Csf –SiCsf –(Al2 O3p )/geopolymer composites after hightemperature treatments are shown in Fig. 7.26. High-resolution fracture morphologies are given in Fig. 7.27 to identify the bonding between the fibers and the matrix. Obviously, a large number of pulled fibers can be found on the fracture surface of the composites after being treated at 800 °C, (Fig. 7.26a, b), which is attributed to the weak bonding strength between the fiber and the geopolymer matrix. After being treated at 1000 °C, although pulled fibers are also visible, the length is much shorter than that after being treated at 800 °C. Moreover, upon heating to 1200 °C, the amorphous geopolymer converts into leucite with more smooth fracture surfaces and less pulled fibers, suggesting an enhanced bonding strength between the fibers and the matrix. Table 7.6 summarizes the energy-dispersive X-ray spectroscopy (EDS) results of Csf –SiCsf /geopolymer composites after being treated at different temperatures. No obvious change in the chemical composition is seen during high-temperature treatments. Note that similar results have been also reported in the Cs-based geopolymer after high-temperature treatments. Figure 7.28 shows the load–displacement curves of Csf –SiCsf – (Al2 O3p )/geopolymer composites after being treated at different temperatures. After high-temperature treatments, all thecomposites reinforced by Cf and SiCf show

7.2 Hybrid Csf and SiCsf Reinforced Geopolymer Matrix Composites

(a)

(b)

(c)

(d)

(e)

(f)

267

Fig. 7.26 Fracture morphologies of Csf –SiCsf –(Al2 O3p )/geopolymer composites after high temperature treatments. a C7S5-800, b AC7S5-800, c C7S5-1000, d AC7S5-1000, e C7S5-1200, f AC7S5-1200, with permission from [19]

268

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

Fig. 7.27 Fracture surfaces of Csf –SiCsf –(Al2 O3p )/geopolymer composites after high-temperature treatments. a C7S5-1000, b AC7S5-1000, c C7S5-1200, d AC7S5-1200, with permission from [19]

Table 7.6 Chemical composition ofCsf –SiCsf –(Al2 O3p )/geopolymer composites after being treated at different temperatures [19] Temperature (°C)

At.% K

Al

Si

O

25

15.25

11.17

34.70

38.88

800

10.26

10.17

35.18

44.39

1000

10.79

11.00

32.95

45.26

1200

10.67

11.87

32.55

44.91

non-catastrophic fracture behaviors with both elastic regions and nonlinear regions. Owing to the fiber pullout and debonding, the load slowly decreases with the increasing displacement after reaching the maximum values.

7.3 Summary

269

Fig. 7.28 Load–displacement curves of Csf –SiCsf –(Al2 O3p )/geopolymer composites after being treated at different temperatures. a 800 °C, b 1000 °C, c 1200 °C, with permission from [19]

7.3 Summary 1. The mechanical properties of the geopolymer matrix can be greatly enhanced by introducing and controlling the content and length of SiCsf . The resulting SiCsf /geopolymer composites all exhibit non-catastrophic fracture behaviors and the best mechanical properties can be obtained by adding SiCsf with a fiber content of 2.0 vol% and a length of 5 mm. They show flexural strength of 49 MPa and work of fracture of 1888.1 J m−2 , which are 5.6 and 63 times higher than those of pure geopolymer, respectively. In situ observation demonstrates that a large number of microcracks forms in the direction parallel to the loading direction during the bending test, and the microcracks can further propagate with the increasing load until reaching the maximum load, consuming most of fracture energy. Thus, the toughening mechanisms of SiCsf /geopolymer composites can be summarized as the formation and propagation of microcracks, crack deflection, fiber debonding, and pullout. 2. Hybrid Csf and SiCsf reinforced geopolymer composites are successively prepared, and the resulting composites show improved mechanical properties with non-catastrophic fracture behaviors. Meanwhile, although the oxidation of Csf occurs at 1000 °C, the flexural strength of the composites is still slightly higher than that of the untreated one. Importantly, the presence of Al2 O3 leads to much enhanced thermal stability, with potential usages as high-temperature materials (under 1000 °C).

270

7 Short SiC Fiber and Hybrid SiC/Carbon Fiber Reinforced …

References 1. J. Yuan, P. He, D. Jia, S. Yan, D. Cai, L. Xu, Z. Yang, X. Duan, S. Wang, Y. Zhou, SiC fiber reinforced geopolymer composites, Part 1: Short SiC fiber. Ceram. Int. 42, 5345–5352 (2016) 2. S. Yan, P. He, Y. Zhang, D. Jia, J. Wang, X. Duan, Z. Yang, Y. Zhou, Preparation and in-situ high-temperature mechanical properties of Cf -SiCf reinforced geopolymer composites. Ceram. Int. 43, 549–555 (2017) 3. P. He, D. Jia, T. Lin, M. Wang, Y. Zhou, Effects of high-temperature heat treatment on the mechanical properties of unidirectional carbon fiber reinforced geopolymer composites. Ceram. Int. 36, 1447–1453 (2010) 4. V.F.F. Barbosa, K.J.D. MacKenzie, Synthesis and thermal behaviour of potassium sialate geopolymers. Mater. Lett. 57, 1477–1482 (2003) 5. J.L. Bell, P.E. Driemeyer, W.M. Kriven, Formation of ceramics from Metakaolin-based geopolymers: Part I: Cs-based geopolymer. J. Am. Ceram. Soc. 92, 1–8 (2009) 6. J.L. Bell, P.E. Driemeyer, W.M. Kriven, Formation of ceramics from metakaolin-based geopolymers: Part II: K-based geopolymer. J. Am. Ceram. Soc. 92, 607–615 (2009) 7. P. Duxson, G.C. Lukey, J.S.J. Deventer, Physical evolution of Na-geopolymer derived from metakaolin up to 1000 °C. J. Mater. Sci. 42, 3044–3054 (2007) 8. Z. Wangxi, L. Jie, W. Gang, Evolution of structure and properties of PAN precursors during their conversion to carbon fibers. Carbon 41, 2805–2812 (2003) 9. P. He, D. Jia, B. Zheng, S. Yan, J. Yuan, Z. Yang, X. Duan, J. Xu, P. Wang, Y. Zhou, SiC fiber reinforced geopolymer composites, Part 2: Continuous SiC fiber. Ceram. Int. 42, 12239–12245 (2016) 10. B. Liang, Z. Yang, Y. Li, J. Yuan, D. Jia, Y. Zhou, Ablation behavior and mechanism of SiCf/Cf/SiBCN ceramic composites with improved thermal shock resistance under xyacetylene combustion flow. Ceram. Int. 41, 8868–8877 (2015) 11. T. Shimoo, F. Toyoda, K. Okamura, Oxidation kinetics of low-oxygen silicon carbide fiber. J. Mater. Sci. 35, 3301–3306 (2000) 12. T. Lin, D. Jia, P. He, M. Wang, In situ crack growth observation and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng, A 527, 2404–2407 (2010) 13. T. Lin, D. Jia, P. He, M. Wang, D. Liang, Effects of fiber length on mechanical properties and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 497, 181–185 (2008) 14. T. Lin, D. Jia, M. Wang, P. He, D. Liang, Effects of fibre content on mechanical properties and fracture behaviour of short carbon fibre reinforced geopolymer matrix composites. Bull. Mater. Sci. 32, 77–81 (2009) 15. P. He, D. Jia, M. Wang, Y. Zhou, Thermal evolution and crystallization kinetics of potassiumbased geopolymer. Ceram. Int. 37, 59–63 (2011) 16. N. Jia, R. Bodet, R.E. Tressler, Effects of microstructural instability on the creep behavior of Si-C-O (nicalon) fibers in argon. J. Am. Ceram. Soc. 76, 3051–3060 (1993) 17. D.J. Pysher, K.C. Goretta, R.S. Hodder Jr., R.E. Tressler, Strengths of ceramic fibers at elevated temperatures. J. Am. Ceram. Soc. 72, 284–288 (1989) 18. P. He, D. Jia, Interface evolution of the Cf /leucite composites derived from Cf /geopolymer composites. Ceram. Int. 39, 1203–1208 (2013) 19. S. Yan, P. He, D. Jia, J. Wang, X. Duan, Z. Yang, S. Wang, Y. Zhou, Effects of high-temperature heat treatment on the microstructure and mechanical performance of hybrid Cf -SiCf -(Al2 O3 p) reinforced geopolymer composites. Compos. Part B Eng. 114, 289–298 (2017) 20. J. Wang, X. Duan, Z. Yang, D. Jia, Y. Zhou, Ablation mechanism and properties of SiCf/SiBCN ceramic composites under an oxyacetylene torch environment. Corros. Sci. 82, 101–107 (2014) 21. T.-S. Lin, D.-C. Jia, P.-G. He, M.-R. Wang, Thermal-mechanical properties of short carbon fiber reinforced geopolymer matrix composites subjected to thermal load. J. Cent. South Univ. Technol. 16, 881–886 (2009)

Chapter 8

Continuous Fibers-Reinforced Geopolymer Matrix Composites

Abstract Geopolymer with its advantages of low density, low cost, mesoporous structure, and excellent thermal properties shows great promise as gas adsorption, aircraft cabin, and heat-resistant materials, and has been investigated as an alternative to polymer composites. However, pure geopolymer has not been widely used in industrial applications as structural material yet because of its overall low toughness and other mechanical properties in comparison to metals or ceramics. Over the past years, various kinds of geopolymer-matrix composites, including particulate, continuous fiber, and short fiber-reinforced geopolymer composites have been extensively investigated. Continuous fiber-reinforced geopolymer composites have generated a great deal of attention due to their adaptability to conventional polymer composites manufacturing techniques. Meanwhile, the high strength and modulus of the fibers can prevent catastrophic brittle failure in composites. In this chapter, continuous carbon fiber and SiC fiber reinforced geopolymer composites were prepared, and the mechanical properties of the obtained composites were reported.

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites 8.1.1 Preparation Process Geopolymer slurry with compositions of SiO2 /Al2 O3 = 4, K2 O/SiO2 = 0.25 and H2 O/K2 O = 11 (mole ratio) was obtained by mixing metakaolin powder with a potassium silicate solution. Metakaolin powders were prepared by calcining kaolin powders at 800 for 2 h. The main phase of metakaolin was amorphous with a characteristic hump centered at about 23° in 2θ and its minor phase was α-quartz. The potassium silicate solution was prepared by dissolving amorphous silica (Shanghai Dixiang Indus., China) into a KOH (Tianjin Fuchen Indus., China) solution. The solution was then allowed to mature under stirring for 48 h to dissolve the silica completely. Different contents of amorphous fused SiO2 powders (99.9% in purity, 16.3 μm in d 50 , China Building Materials Academy, Beijing, China) were also added into the geopolymer slurry for matrix modification. © Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3_8

271

272 Table 8.1 Typical properties of carbon fiber [1]

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites Diameter (μm)

Density (g cm−3 )

Tensile strength (MPa)

Tensile modulus (GPa)

6–8

≥1.76

2930

200–220

The carbon fiber (Jilin Carbon Indus., China) has a diameter of 6–8 μm, and its properties are summarized in Table 8.1. The composite was prepared by infiltrating geopolymer resin into the PAN-based carbon fiber preform with the help of the ultrasonic vibration treatment and stacked one by one to get a green sample. To remove the pores in the green compact, degassing was applied at 80 °C for 48 h using a vacuum-bag technique. After demolding, the composites were cut into specimens for tests.

8.1.2 Unidirectional Carbon Fiber-Reinforced Geopolymer (Cuf /Geopolymer) Composites (1) Microstructure Cuf /geopolymer composites with different carbon fiber contents (10, 20, 25, and 30 vol%) are obtained by controlling the pressure during the curing process, and they are labeled as 10C, 20C, 25C, and 30C, respectively. Figure 8.1 presents the microstructure of Cuf /geopolymer composites with different fiber contents. It can be noted that most fibers maintained unidirectional and all the fibers were well impregnated by the geopolymer matrix, indicating ultrasonic-assistant impregnation can get fully impregnated composites. Meanwhile, with the increase in fiber contents visible fiber number in the images also increased. Figure 8.2 shows the apparent density of the composites with different fiber contents. With the increase in fiber contents, the apparent density of the composite increased linearly. Apparent density of the composites was much lower than their theoretical value. Since the surface of fiber was hydrophobic, wetting between fiber and water-based geopolymer would be difficult. Therefore, defects (such as pores) were inevitably in the interface area, which makes the apparent density deviate from the theoretical value. Moreover, when the fiber content is higher, the deviation is much more obvious. (2) Mechanical properties Table 8.2 and Fig. 8.3 shows the mechanical properties of Cuf /geopolymer composites vs. fiber contents. Compared with the geopolymer matrix, the mechanical properties of the Cuf /geopolymer composites, including flexural strength, elastic modulus, and work of fracture in the x direction (the direction along fiber axial direction), have been greatly improved. And when fiber content was 20 vol%, they reached the highest

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

(a)

273

(b)

50μ m

50μ m

(d)

(c)

50μ m

50μ m

Fig. 8.1 Microstructure of Cuf /geopolymer composites with different fiber contents. a 10C, b 20C, c 25C, d 30C [1]

-3

1.52

Apparent density, g·cm

Fig. 8.2 Apparent density of the Cuf /geopolymer composites with different fiber contents [1]

1.50 1.48 1.46 1.44 1.42 0

5

10 15 20 25 Volume fraction of fiber, %

30

35

values of 143.5 MPa, 36.5 GPa, and 3874.5 J/m2 , which was 10.6, 2.6, and 70.5 times of those of pure geopolymer matrix, respectively. With the increase in fiber contents, flexural strength, Young’s modulus, and work of fracture of the Cuf /geopolymer composites increased greatly first and then decreased slightly. And when fiber content is 20–25 vol%, the composite showed the best mechanical performance. Flexural strength in y direction (the direction perpendicular to fiber axial direction) was much lower than that of pure geopolymer and showed decreasing trend with fiber contents. Because the fiber preform in this part

274

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

Table 8.2 Mechanical properties of the Cuf /geopolymer composites with different fiber contents [1] Flexural strength (MPa) x direction K-PSS matrix

12.3 ± 1.2

Young’s modulus (GPa)

Work of fracture (J m−2 )

y direction

(x direction)

(x direction)

12.3 ± 1.2

10.3 ± 1.2

54.2 ± 8.1

10C

109.8 ± 7.6

9.4 ± 1.7

23.1 ± 2.9

2146.7 ± 184.7

20C

143.5 ± 8.5

8.6 ± 2.3

36.5 ± 3.4

3874.5 ± 266.8

25C

136.0 ± 6.1

8.8 ± 1.5

37.8 ± 4.8

3793.1 ± 257.4

30C

128.5 ± 9.5

7.6 ± 1.9

32.8 ± 4.2

3511.4 ± 218.6

Fig. 8.3 Flexural strength and Young’s modulus of the Cuf /geopolymer composites (x direction) [1]

60

160

Flexural strength, MPa

140

50

120 40

100 80

30

60

20

40 10

20 0

Young's modulus, GPa

Materials

0

5

10 15 20 25 Volume fraction of fiber, %

30

0

is unidirectional, in y direction, fibers cannot play the strengthening effects. On the contrary, the presence of carbon fiber would lead to the formation of defects in the geopolymer matrix, leading to the decrease in flexural strength in y direction. The stress–displacement curve of Cuf /geopolymer composites is shown in Fig. 8.4. It can be noted that pure geopolymer showed typical brittle failure mode, that is, after load reaching the maximum value it decreased sharply. With the presence of carbon fiber, all the composites showed non-catastrophic failure mode and they all exhibited elastic region and nonlinear region. The initial elastic region is associated with elastic deformation of geopolymer matrix, carbon fiber and their interface, while the nonlinear region is associated with the microcrack formation, fiber-matrix interface debonding, and fiber pulling-out. After the load reached the highest value, they tended to decrease slowly, and then, the strength–displacement curves spread more widely indicating the shear fracture mode. The fractograph of the Cuf /geopolymer composites with fiber content of 20 vol% is shown in Fig. 8.5. There are many residual geopolymer detritus on the fiber surface, and the matrix breaks into debris rather than maintaining fine integrity. Meanwhile,

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

275

Fig. 8.4 Flexural stress–displacement curves of the Cuf /geopolymer composites (x direction) [1]

Fig. 8.5 Fractograph of the 20 vol% Cuf /geopolymer composites [1]

both fiber breaking and pulling-out can hardly be observed. This indicates that after the load reaches the highest value, the matrix failure first happens and turned into debris, which cannot effectively transfer the load from interface to the fiber, leading to the failure of the composites. Therefore, strong matrix should be developed in order to get composites with high strength. (3) Effects of SiO2 contents on the microstructure and mechanical properties of Cuf /geopolymer composites In Chap. 5, we found that the presence of SiO2 particle would enhance the mechanical performance of geopolymer by increasing Si/Al ratio and strengthening through residual particle. Therefore, in this part, SiO2 was used to get strong

276

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

geopolymer matrix. And its effect on the microstructure and mechanical performance of Cuf /geopolymer composites was reported. The 20 vol% Cuf /geopolymer composites with principle Si/Al ratios of 2, 2.5, 3, 3.5, and 4 are marked as CG2, CG2.5, CG3, CG3.5, and CG4, respectively, by adding different fused SiO2 particles. The SEM images of the microstructure of Cuf /geopolymer composites with different Si/Al ratios are shown in Fig. 8.6. It can be noted that fibers are fully impregnated by the geopolymer matrix. And the fiber distribution is relatively uniform. With the increase of the Si/Al ratio, the microstructure of the Cuf /geopolymer composites shows no significant difference. Figure 8.7 shows the mechanical properties of the Cuf /geopolymer composites with increasing Si/Al ratio. With the increase of Si/Al ratio, both flexural strength and Young’s modulus of composites increase. Especially, the Young’s modulus increases linearly. When Si/Al ratio increases from 2 to 4, the flexural strength of the composites increased from 143.5 to 250.0 MPa, and the Young’s modulus increased from 36.5 to 56.5 GPa. The work of fracture increased with the Si/Al ratio first, reached the top value with the Si/Al ratio of 3, and then gradually reduced. With a certain carbon fiber content, the Si/Al ratio can significantly affect the mechanical properties of the Cuf /geopolymer composites by influencing the properties of the matrix. When the Si/Al ratio is low, both the matrix and interface are weak, which cannot effectively transfer load from matrix to fibers, and thus fibers cannot fully play the role of strengthening and toughening. After increasing the Si/Al ratio, the matrix strength also increases, and the interfacial bonding strength between the fiber and the matrix gradually increases, which is more conducive to the transfer of the load from the matrix to the reinforced carbon fiber. Thus, it is beneficial to improve the flexural strength and modulus of the Cuf /geopolymer composites. The trend of work of fracture is different from that of both strength and modulus. In CG2 and CG2.5 composites, most of the fibers did not fracture, and work of fracture of the composites was mainly contributed to the interfacial debonding between fibers and matrix. However, work of fracture of CG3, CG3.5, and CG4 composites was mainly due to the fiber broken and pulling-out, which was much higher than debonding. Therefore, work of fracture of CG3, CG3.5, and CG4 composites was higher than that of CG2 and CG2.5 composites. When the Si/Al ratio of the matrix increased from 3 to 4, crack propagation along the interface and the debonding between the fiber and the matrix became difficult. During the fracture process of the composites, length of fiber pulling-out became shorter. So, the work of fracture of the composites decreased. Figure 8.8 provides the stress–displacement curves of composites with different Si/Al ratios. When the Si/Al ratio of the matrix increases, the composites show ductile fracture characteristics. From the macroscopic photos of the fractured sample (Fig. 8.9), the main crack of CG2.5 composite is not obvious, and the material still maintains the state of “lotus root fracture and thread connection” after failure, and the lamination damage on the compressive side of the sample is serious, indicating that the failure of composite is shear failure mode. In contrast, when the Si/Al ratio increases to 3, the fiber broken and pulling out are obvious, and the fracture surface

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

277

Fig. 8.6 Microstructure of the Cuf /geopolymer composites with different Si/Al ratio. a, b CG2; c, d CG2.5; e, f CG3; g, h CG4 [1]

278

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

400

7000

250

40

200

30

150 20 100

Work of Fracture, J·m

50 Youngs modulus, GPa

Flexural strength, MPa

300

-2

60

350

6000

5000

4000

10

50

0

0 2

2.5

3 3.5 Si/Al ratio

3000

4

2

2.5

3 Si/Al ratio

3.5

4

Fig. 8.7 Mechanical properties of the Cuf /geopolymer composites with different Si/Al ratios [1] Fig. 8.8 Flexural stress–displacement of the Cuf /geopolymer composites with different Si/Al ratios [1]

250

Flexural stress, MPa

200 150 100 right to left CG2, CG2.5, CG3, CG3.5, CG4

50 0 0.0

Fig. 8.9 Digital images of Cuf /geopolymer composites with different Si/Al ratios after bending test. a CG2.5, b CG3, c CG4 [1]

0.2

0.4 0.6 0.8 Displacement, mm

1.0

(a)

3mm (b)

3mm (c)

3mm

1.2

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

279

becomes smooth with the increase of Si/Al ratio, indicating that the fracture mode changes from shear failure to flexural failure mode. Moreover, the surface of carbon fiber is hydrophobic, which is inert in alkali solution and has poor compatibility with water-based geopolymer [2]. Meanwhile, there was no interfacial reaction between the fiber and the matrix at the current preparation temperature. Therefore, interface between the fiber and the matrix was in a weak mechanical bonding state. With the increase in Si/Al ratio, geopolymer became much denser than those with low Si/Al ratios. On the one hand, the density of matrix increased, and on the other hand, gripping force between the matrix and the fibers also enhanced. Under a certain friction coefficient between matrix and fiber, interfacial bonding strength would increase. Figure 8.10 presents the SEM fractographs of the Cuf /geopolymer composite with different Si/Al ratios. For the composite with Si/Al ratio of 2, most of the fibers in the composites did not fracture, and the matrix debris distributed among the fiber surface,

(a)

(b)

(c)

(d)

Fig. 8.10 SEM fractographs of Cuf /KGP composites with different Si/Al ratio. a CG2; b CG3; c CG3.5; d CG4 [1]

280

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

indicating that the interface between the matrix and the fiber was weak. When the Si/Al ratio increases from 3 and 3.5 to 4, the matrix of the composites becomes denser, and the fiber pulling-out is obvious. However, the fiber pulling-out length gradually decreased with Si/Al ratio from 3 to 4, indicating the increased interfacial bonding strength. This was in good agreement with the change of mechanical properties of the composites. When Si/Al was 4, fiber pulling-out was obvious, but pulling-out length was short. So, the fiber plays the best role in strengthening. And the strength and modulus of the composites CG4 were the highest. When Si/Al ratio was 3, the composite showed the longest pulling-out length, which seemed to consume the highest level of fracture energy, corresponding to the maximum damage tolerance and the highest work of fracture. Figures 8.11 and 8.12 provide the schematic diagram of crack growth during the fracture process of CG2 and CG4 composites, together with the stress distributions on the tensile surfaces, respectively. Since the tensile strength of CG2 matrix are much lower than that of carbon fiber, the stress on the tensile surface of the composite easily exceeded its ultimate tensile strength. So, a lot of microcracks can be formed on the tensile surface. The crack width was wider when the crack was closed to the main crack. On the contrary, the crack width is narrow when it is far away from the main crack [3, 4]. The formation of these microcracks resulted in the stress changes from linear to curvilinear distribution on the tensile surface of the composite. Due to the high strength of fiber, it did not fracture under low stress. Thus, microcrack

(a)

(b)

I

I

II

IV

III

II

IV

III

Fig. 8.11 Schematic fracture processes of the CG2 composite. a cracks evolution, b stress evolution on the tensile surface [1]

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

281

I

II

IV

III

(a)

(b)

I

II

IV

III

Fig. 8.12 Schematic fracture processes of the CG4 composite. a cracks evolution, b stress evolution on the tensile surface [1]

pinning would take place when it encounters the fiber, which can prevent further crack propagation. During the crack propagation along the interface, the stress near the fiber was redistributed, and the weak matrix could not sustain such stress and broke into fragments coupled with fiber debonding. Meanwhile, microcracks would appear on the compressive surface of the composite under compressive stress. Because of the unidirectional distribution of fibers, the matrix is easy to be broken under low stress in the direction vertical to fiber axial, resulting in the shear failure of whole composites. With the further increase of the displacement, the main crack propagation region is formed near the maximum stress on the tensile surface. When enough matrix became fragment, matrix cannot maintain the integrity of the composites, leading to the failure of the composite. And during the failure process, most fibers did not fracture and its strengthening effect cannot be fully adapted. The failure process of CG4 sample was in an obvious different way. Because the strength of the matrix is high, microcracks did not appear under a low stress state. When the main crack meets the fiber, the crack is blocked. And the applied stress must be increased for further crack propagation. With the increase of stress, the crack propagation along the interface led to the interface debonding between the matrix and the fiber. The stress concentration at the interface resulted in the fiber fracture and pulling-out from the matrix. Besides, due to the great difference of the matrix

282

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

strength, the interlaminar bonding strength of CG4 was much higher than that of CG2 and thus avoid the interlaminar shear failure. During the failure process of CG4, most of the fibers are pulled off. Thus, the strengthening effect from the fibers was more sufficient thus the composites got high flexural strength.

8.1.3 2D Carbon Fiber-Reinforced Geopolymer Matrix Composites (1) Microstructure Figures 8.13, 8.14, and 8.15 show the microstructure of 2D continuous carbon fiberreinforced geopolymer composites from both perpendicular and parallel to the layingup directions. The fiber volume contents in the composites are 30–35%. It can be noted that defects of such residual pores and microcracks coexisted in the three composites, and these defects were visible at low magnification images. Residual pores resulted from the insufficient impregnation of geopolymer matrix to the 2D fiber fabrics, and these residual pores were more likely to exist at the intersection of fiber bundles from two directions. Microcracks were also noted which should be attributed to the thermal shrinkage of geopolymer matrix during the curing process. During the curing process, geopolymer with Si/Al ratio higher than 2.5 showed high thermal shrinkage. In this part, the Si/Al ratios were 3, 3.5, and 4, which were higher

Fig. 8.13 SEM images of 2D-Cf /geopolymer composite with Si/Al ratio of 3: Perpendicular (a), (b) and parallel (c), (d) to the laying-up directions

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

283

Fig. 8.14 SEM images of 2D-Cf /geopolymer composite with Si/Al ratio of 3.5: Perpendicular (a), (b) and parallel (c), (d) to the laying-up directions

Fig. 8.15 SEM images of 2D-Cf /geopolymer composite with Si/Al ratio of 4: Perpendicular (a), (b) and parallel (c), (d) to the laying-up directions

284

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

Table 8.3 Properties of 2D-Cf /geopolymer composite with different Si/Al ratios Si/Al ratio of the Composite

Apparent density (g cm3 )

Flexural strength (MPa)

Work of fracture (kJ m−2 )

Young’s modulus (GPa)

3

1.63

216.7 ± 3.1

6.09 ± 0.51

40.3 ± 3.2

3.5

1.71

271.8 ± 21.3

7.77 ± 0.62

45.7 ± 2.6

4

1.76

304.5 ± 36.4

8.54 ± 0.16

48.0 ± 4.7

than 2.5. However, during the curing process, carbon fiber kept volume stability. Therefore, microcracks appeared due to the significant different thermal shrinkage between the fiber and the matrix. With the increase in Si/Al ratio, residual pores and microcracks showed decreasing trend, which might be due to increased residual SiO2 contents to release thermal shrinkage from geopolymer matrix. (2) Mechanical properties and fracture behavior Table 8.3 summarizes the properties of 2D-Cf /geopolymer composite with different Si/Al ratios. With the increase in Si/Al ratio, apparent density, flexural strength, work of fracture, and Young’s modulus increased simultaneously, and reached the highest values of 1.76 g/cm3 , 304.5 MPa, 8.54 kJ/m2 , and 48.0 GPa, for composite with Si/Al ratio of 4, which was 8.0, 40.5, 40.2, and 19.1% higher than those of composite with Si/Al ratio of 3. Meanwhile, the flexural strength of 2D-Cf /geopolymer composite with Si/Al ratio of 4 was also 21.8% higher than that of unidirectional carbon fiber-reinforced geopolymer composite with the same Si/Al ratio, which might be attributed to the higher fiber content in 2D-Cf /geopolymer composite. The stress–displacement curves of 2D-Cf /geopolymer composite with different Si/Al ratios are shown in Fig. 8.16. It can be noted that all the composites showed non-catastrophical failure mode. After the stress reaches the highest value, there are several significant steep drop-steps and the curves extended evident tails. The fractographs of the composites with different Si/Al ratios (Fig. 8.17) proved that obvious

Stress (MPa)

Fig. 8.16 Stress– displacement curves of 2D-Cf /geopolymer composite with different Si/Al ratios

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

285

Fig. 8.17 SEM fractographs of 2D-Cf /geopolymer composite with different Si/Al ratios. a, b 3, c, d 3.5, e, f 4

fiber bundles and fiber pulling-out were observed, indicating the great strengthening and toughening effects from carbon fibers. High Si/Al ratio of geopolymer is beneficial for the improvement in mechanical properties of the geopolymer composites, and the strengthening and toughening effects from the carbon fiber are also obvious. However, when the Si/Al is ≥3, chemical stability and thermal properties of the geopolymer matrix become worse [5, 6], which would inevitably limit the application fields of this kind of composites.

286

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

So, further work should also be carried out to explore other method to get a strong geopolymer matrix with good chemical stability and thermal properties.

8.1.4 Effects of High-Temperature Treatment on the Microstructure and Mechanical Properties of Cuf /Geopolymer Composites Upon high-temperature treatment, geopolymer can be converted into ceramics, like leucite, nepheline, kalsilite, pollucite, and so on, based on the chemical composition of geopolymer [5, 7–11]. In this part, in order to get a strong matrix in the Cuf /geopolymer composites, the composites were high-temperature treated in Ar atmosphere and the effects of treatment temperatures on the phase composition, interfacial state between fiber and matrix, mechanical properties, and fracture behavior are reported. The Si/Al ratio of geopolymer was below 2.5, and fiber contents were 20 and 25 vol% for untreated and high-temperature-treated composites. The hightemperature treatment process is shown in Fig. 8.18. The composites before and after treatment at 1000, 1100, 1200, 1300, and 1400 °C were denoted as KC-B, KC-1000, KC-1100, KC-1200, KC-1300, and KC-1400, respectively. (1) Thermal shrinkage After being treated at 1100 °C, the Cuf /geopolymer composite is taken as digital image (as indicated in Fig. 8.19) on the surface to compare with the composite without being treated. It can be noted that after high-temperature treatment, composite showed

1000∼1400

600 60min

90min

5 /min

5 /min

Fig. 8.18 High-temperature treatment procedure of the Cuf /geopolymer composites [1]

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

287

(a)

y (b)

x

Fig. 8.19 Morphologies of Cuf /geopolymer composite a before and b after treatment at 1100 °C for 90 min in Ar atmosphere [1]

evident shrinkage in y direction, while in x direction there is no obvious shrinkage. Thermal shrinkages in both x and y directions are shown in Fig. 8.20. Upon heating from 900 to 1200 °C the thermal shrinkages (dL/L0 ) in y and x directions are 12.4 and 0.9%, respectively, which was caused by the viscous sintering of geopolymer. Matrix shrinkage in x direction was highly hindered by the carbon fibers, while in y direction, thermal shrinkage was not curbed. Figure 8.21 gives the variation of apparent density of Cuf /geopolymer composites after treatment at different temperatures. The apparent density showed an increasing trend with the treatment temperature, which was due to the densification of matrix caused by viscous sintering at high temperature. Fig. 8.20 Thermal shrinkage of the Cuf /geopolymer composite in x and y directions, with permission from [12]

//

1.8 1.6 -3

Fig. 8.21 The apparent density of the composites treated at different temperatures for holding 90 min [1]

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

Apparent density, g·cm

288

1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0 RT

//

1000 1100 1300 1200 o Heat treatment temperature, C

1400

Fig. 8.22 XRD patterns of Cuf /geopolymer composites before and after heat treatment, with permission from [12]

(2) Phase composition The XRD patterns for composites heated treated at a variety of temperatures are collected in Fig. 8.22. KC-B displayed typical amorphous XRD characteristic. In addition, several sharp characteristic peaks were also observed, which were identified as α-quartz introduced by metakaolin. For KC-1000 sample, the XRD patterns remained predominately amorphous, indicating the non-crystallization of matrix. After composite was heat treated at 1100 °C, leucite (K2 O·Al2 O3 ·4SiO2 ) and a little kalsilite (K2 O·Al2 O3 ·2SiO2 ) appeared. For the samples being heat treated from 1100 to 1300 °C, leucite became a major phase. When heat treatment temperature was further increased to 1400 °C, a broad amorphous hump between 20° and 30° of 2θ appeared, which revealed the partial melting of matrix. Simultaneously the peak intensity of leucite decreased indicating the crystal phases dissolve. At 1400 °C it was also revealed that a little amount of β-SiC phase appeared in the composites,

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

289

Fig. 8.23 TGA curves of geopolymer and Cuf /KGP composites, with permission from [12]

indicating that carbon fiber reacted with Si–O units into SiC phase at an enough high temperature. Figure 8.23 shows the TG curves of the geopolymer and obtained Cuf /geopolymer composite, which further confirms that interface reaction occurred at high temperature. The major weight loss before 700 °C from geopolymer and composite resulted from water loss by evaporation of both free water and condensed hydroxyl groups [7, 10]. From 700 to 1400 °C, geopolymer showed little weight change. In contrast, for the composite there was another fast weight loss stage above 1170 °C which might be due to the interface reaction between carbon fiber and matrix. Considering the XRD analysis results, we can deduce that interface reaction was as in the following formula (8.1) [13]: Si−O + C → SiC + CO(g)

(8.1)

Table 8.4 shows the standard Gibbs free energy formation of each reaction in Si–C–O system, and the G° for formula (8.1) can be calculated according to the formula (8.2): G ◦ = 605250 − 339.61T

(8.2)

Here, the equilibrium partial pressure of CO(g) was ignored, and when G° ≤ 0 the above reaction could occur. The calculated interfacial reaction temperature Table 8.4 The reactions and standard Gibbs free energy

Reaction

Standard Gibbs free energy (J mol−1 )

(1) C(s) + 0.5O2(g) = CO(g)

−14,400 − 85.77 T

(2) Si(s) + C(s) = SiC(s)

−73,050 + 7.66 T

(3) Si(s) + O2(g) = SiO2(s)

−907,100 + 175.73 T

290

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

should be 1509 °C; however, as indicated by the TG results, the interfacial reaction between Si–O and C started at 1170 °C, which was much lower than the theoretical result and that in the Cf /SiO2 system [14]. It might be attributed to the existence of K+ in the matrix. Because the bond energy of K–O is much lower than that of the Si–O, the oxygen which is connected with the K+ can be easily captured by the Si4+ at high temperature, and thus the bridge oxygen bonds to the Si4+ in the Si–O groups were disconnected. Therefore, the frame poly-structure of the Si–O groups was decomposed to be a variety of oligomers, which results in the decreased viscosity of the leucite matrix at high temperature as compared with the pure SiO2 matrix. Although the low viscosity was helpful to the densification of the composite, it was also associated with the much more accelerated atom movement, which made the interface reaction occur at low temperature. (3) Microstructure characterization Figure 8.24 presents typical microstructures of composites after heat treatment. After the composite was heat treated at 1000 and 1100 °C, many short oval cracks in radial direction of carbon fiber were formed. These oval cracks resulted from the large volume shrinkage caused by viscous sintering of geopolymer matrix at high temperature, as shown in Fig. 8.20. Matrix shrinkage in x direction was highly hindered by the carbon fibers, which led to the formation of evenly distributed oval cracks in the softened matrix. For KC-1100 sample, Fig. 8.24c indicated that the cracks were fully bridged by fibers which maintained fine integrity. With increasing heat treatment temperature, the thermal mismatch between the fiber and the matrix increased and the interfacial reaction aggravated, so the thermally mismatched cracks opened wider and fibers were damaged more seriously, as shown in Fig. 8.24d, e. Cf /matrix interfacial structure is an important factor to determine mechanical properties of composites. And it can be evaluated by the morphology of fracture surfaces [15, 16], as shown in Fig. 8.25. For the samples KC-B and KC-1000 (Fig. 8.25a, b), most fibers didn’t fracture and fragmentation of geopolymer matrix distributed in the intra-bundles of fibers. Fiber debonding could also be observed, indicating the weak Cuf /matrix interface. After composites were heat treated at higher temperature, for the great shrinkage of matrix, frictional restraint stress at Cuf /matrix interface would increase and interface bonding was enhanced. Figure 8.25c showed that long fiber pullout dominated the fracture surface and all fibers maintained fine integrity. For KC-1200 and KC-1300, fiber pull-out was still very clear on the fracture surface, though most of the pulled-out fibers appeared to be a little shorter than that of the KC-1100 sample. However, for the sample KC-1400, the fracture surface was much flatter and few fibers pullout could be observed. Extensive fiber pullout of composites usually indicated a relatively weak Cuf /matrix interfacial bonding, while little fiber pullout and short pullout length indicated a strong Cuf /matrix interfacial bonding [15, 17]. Therefore, the Cuf /matrix interfacial bonding of KC-1100 was more desirable than those in other samples and great reinforcement could be achieved. For the KC-1400, this was not the case. To get further information, element distribution was analyzed and higher magnification SEM fractograph was recorded in Fig. 8.26. As shown in Fig. 8.26a, the diffusion of Si atoms into carbon fiber takes place

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

(a)

291

(b)

400μm

(c)

400μm

(d)

(c) 80μm

80μm (e)

80μm Fig. 8.24 Microstructure of polished cross-section of: a KC-1000, b KC-1100 (low magnification), c KC-1100 (high magnification), d KC-1300, e KC-1400, with permission from [12]

in the Cuf /matrix interface zone. As indicated in Fig. 8.26b, an interface reaction layer of about 1 μm in thickness was clearly observed. According to the aforementioned XRD analysis results in Fig. 8.22, it can be concluded that the reaction layer should be β-SiC. Due to formation of the thick interfacial SiC layer, too much strong Cuf /matrix bonding formed and the carbon fiber was seriously degraded; as a result, the reinforcement effect from the carbon fibers must be decompensated. (4) Mechanical properties and fracture behavior Figure 8.27 presents the mechanical properties of geopolymer, leucite ceramic (derived from geopolymer after treatment at 1100 °C, polycrystalline material), and composites. It was found that the leucite formation effectively strengthened the composite matrix from the starting 12.3–70 MPa. The effects of heat treatment

292

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

(b)

(a)

80μm

(c)

80μm

(d)

80μm

(e)

80μm

(f)

80μm

80μm

Fig. 8.25 Fractographs of Cf /KGP composites treated at different temperatures: a KC-B, b KC1000, c KC-1100, d KC-1200, e KC-1300, f KC-1400, with permission from [12]

on mechanical properties of composites were quite obvious. As compared with the KC-B sample, after being treated at 1000 °C, the mechanical properties of composite (KC-1000) decreased by 30%. While after composite was heat treated at 1100 °C for ?? min, flexural strength, work of fracture, and Young’s modulus attained their peak values at 234.2 MPa, 4445.7 J m−2 , and 63.8 GPa, increasing by 76, 15, and 75%, respectively, compared with that of KC-W. With increasing heat treatment temperature from 1100 to 1300°C, the mechanical properties tended to decrease gradually but flexural strength and Young’s modulus were still much higher than that of KC-W and retained approximately 70–80% of the peak strength and 80–90% of the peak modulus. However, when heat treatment temperature was further increased to 1400 °C, flexural strength, work of fracture, and Young’s modulus of composite

8.1 Continuous Carbon Fiber-Reinforced Geopolymer Matrix Composites

(a)

293

(b) Si

Matrix Reaction layer Carbon fiber

Fig. 8.26 Characteristic X-ray line profiles of Si atom (a) and interface reaction layer (b) in sample KC-1400, with permission from [12]

Fig. 8.27 Variations of flexural strength and work of fracture of Cuf /KGP composites treated at different temperatures, with permission from [12]

decreased sharply to 54.6 MPa, 366.6 J m−2 , and 41.5 GPa, respectively, and it was in well agreement with the aforementioned effect of the SiC interface reaction layer on the mechanical properties. It should be mentioned that after heat treatment the fiber volume fraction for the composites increased relatively from the starting 20– 25 vol% due to the shrinkage of the matrix. According to composite mixing rules, the calculated strength should be around 164.5 MPa, which was still far lower than that of KC-1100 sample. So, the increased fiber fraction had only a minor effect on mechanical properties of the composites.

294

Stress (MPa)

Fig. 8.28 Typical load–displacement curves of geopolymer and Cuf /geopolymer composites before and after treatment, with permission from [12]

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

Figure 8.28 manifests typical curves for stress vs. displacement of the geopolymer and Cuf /geopolymer composites before and after heat treatment. Geopolymer indicated a typical brittle failure mode (Fig. 8.28a). For the composites except KC-1400 (Fig. 8.28g), no catastrophic failure was observed and they all exhibited elastic region and nonlinear region. For KC-1100, KC-1200, and KC-1300 (Fig. 8.28d–f), there were several significant steep drop-steps after reaching the maximum, indicating the composites fractured in tensile fracture mode. However, the strength–displacement curves of KC-B and KC-1000 (Fig. 8.28b, c) were different from the above three. The strength tended to reduce slowly after the maximum load, and then the strength– displacement curves spread more widely, indicating the shear fracture mode rather than tensile fracture mode. In the case of KC-1400, the fracture behavior was similar to that of the geopolymer, showing a catastrophic fracture behavior. These were in good consistent with the fractograph observations in Fig. 8.25.

8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites 8.2.1 Preparation Process Geopolymer resin with a composition of SiO2 /Al2 O3 = 4, K2 O/SiO2 = 0.25, and H2 O/K2 O = 11 (mole ratio) was obtained by mixing metakaolin powder with potassium silicate solution. SiC fiber used in this study (Sailifei Ceramic Fiber Co., Ltd., China) has a diameter of 13 μm and an average tensile strength of 1.5–1.6 GPa, as shown in Table 8.5. SiC fiber first treated at 370 °C in an air atmosphere for 2 h to remove the surface epoxy coating. The composite was prepared by infiltrating geopolymer resin into the unidirectional continuous SiC fiber preform with the help

8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites

295

Table 8.5 Properties of SiC fiber [18] Brand

Density (g cm−3 )

Diameter (μm)

Tensile strength (GPa)

Modulus (GPa)

Elongation (%)

SLFC1

2.36

13 ± 0.5

1.5–1.6

140 ± 10

1.3–1.4

Table 8.6 High temperature mechanical properties of the SiCuf /KGP composites [19]

Sample

Flexural strength (MPa) Work of fracture (J m−2 )

HS20

158.7 ± 7.7

3724 ± 310

HS-700

152.1 ± 17.7

4565 ± 231

HS-800

143.1 ± 10.0

3737 ± 457

HS-900

137.1 ± 10.6

2755 ± 88

HS-1000 111.5 ± 14.7

1523 ± 270

HS-1100 110.1 ± 12.4

1290 ± 55

of the ultrasonic vibration treatment, and stacked one by one to get a green sample with 16 layers. To control fiber contents in the composites, the different mechanical pressures of 0.5, 1.0, 1.5, and 2 MPa were applied to the green compact in a hydraulic machine. To remove the pores in the composites, degassing was applied at 70 °C for 24 h using a vacuum-bag technique after mechanical pressing.

8.2.2 Unidirectional SiC Fiber-Reinforced Geopolymer (SiCuf /Geopolymer) Composites In this part, the effects of fiber contents of 10, 15, 20, and 25 vol% on the microstructure and mechanical properties of the SiCuf /geopolymer composites were reported. The composites with different SiC fiber contents of 10, 15, 20, and 25 vol% are labeled as S10, S15, S20, and S25. (1) Phase composition Figure 8.29 compares the XRD patterns of SiC fiber, metakaolin, geopolymer, and SiCuf /geopolymer composites. SiC fiber is mainly composed of β-SiC (Fig. 8.29a). Metakaolin is mainly amorphous with a slight amount of α-quartz. The characteristic amorphous hump of geopolymer shifted to a higher position, 27–29° (2θ). However, with the presence of SiCuf , the amorphous hump in geopolymer became weak, and hump between 22 and 24° became stronger, indicating SiCuf retarded the geopolymerization process of the matrix. (2) Microstructure Figure 8.30 demonstrates that most fibers maintained unidirectional and all the composites showed homogeneous microstructure. The magnified images in Fig. 8.31

296

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

Fig. 8.29 XRD patterns of SiC fiber (a), metakaolin, geopolymer and SiCf /geopolymer composites (b), with permission from [18]

Fig. 8.30 Microstructure of SiCuf /KGP composites with different fiber contents: a 10 vol%, b 15 vol%, c 20 vol%, d 25 vol%, with permission from [18]

showed that the fiber bundles were impregnated with geopolymer matrix. Meanwhile, microcracks were also observed which were perpendicular to the fiber axial direction. During the curing process, fiber maintained unchangeable, but geopolymer matrix showed shrinkage due to both evaporation of waters and condensation of –OH

8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites

297

Fig. 8.31 Magnified microstructure of SiCuf /geopolymer composites: (b) and (c) correspond to the areas B and C, respectively, with permission from [18]

groups in the geopolymer system. Matrix shrinkage in the axial direction was highly hindered by the fibers, leading to the formation of the microcracks in the matrix. (3) Mechanical properties and fracture behavior Flexural strength and Young’s modulus of SiCuf /geopolymer composites in the x direction vs. fiber content are depicted in Fig. 8.32. With the increase in fiber content, flexural strength increased first, reached the maximum value when fiber content was 20 vol%, and then decreased for fiber content of 25 vol%. At 10 vol% fiber content, the flexural strength of the composite was 147.5 MPa, which was 12.2 times higher than that of pure geopolymer. According to the strength mixing rule of composite, critical fiber content (V fc ) can be calculated based on formulas (8.3) to (8.6), where σ mu , σ fu , σ f , and σ f  , are the ultimate stress of matrix, ultimate stress of fiber,

30

160

25

120

20 15

80

10 40

Young's modulus (GPa)

Flexural strength (MPa)

200

5 0

0

5

10 15 Fiber content (vol %)

20

25

Fig. 8.32 Flexural strength and Young’s modulus of SiCuf /geopolymer composites in x direction, with permission from [18]

298

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

additional fiber stress, and fiber stress when matrix fractured, respectively. V f and V m are the volume contents of fiber and matrix, respectively. σmu Vm Vf

(8.3)

Vf + Vm = 1

(8.4)

σf =



σf = σfu −σf Vfc =

(8.5)

σmu  σmu + σfu − σf

(8.6)

σmu σfu

(8.7)

Vfc ≈

In SiCuf /geopolymer composite system, σ fu >> σ f  >> σ f , effective V fc can be calculated according to the formula (8.7). It explained the increase in flexural strength when fiber contents were lower than 25 vol%. With the fiber content increasing from 10 to 20 vol%, the flexural strength of the composites increased slowly, from 147.5 to 161.6 MPa, rising by 9.6%. The SiCuf /geopolymer composite showed higher flexural strength than Cuf /geopolymer composite prepared through similar process [12], indicating a better interfacial state in SiCuf /geopolymer than that in Cf /geopolymer. However, when SiC fiber content was 25 vol%, the flexural strength showed a 17.5% drop to 133.3 MPa, compared to the highest value at fiber content of 20 vol%. Higher fiber content indicated higher fiber-matrix interfacial area, where defects such as microcracks and micropores would gather as indicated in Fig. 8.31, resulting in the decrease in flexural strength. With the increase in fiber content, Young’s modulus of the SiCuf /geopolymer composites showed a linear rising trend. Young’s modulus can be calculated according to the Halpin–Tsai model as shown in formulas (8.8) and (8.9), where E c , E f , and E m were Young’s modulus of the composites, fiber and matrix; ξ was a parameter related to fiber morphology. Because modulus of SiC fiber (200 GPa) was much higher than that of matrix (8 GPa), it resulted in the increase in modulus of composites with fiber content. 1 + ξ ηVf Ec = Em 1 + ξ ηVf η=

E f /E m − 1 + ξ ηVf E f /E m + ξ

(8.8) (8.9)

Figure 8.33 shows the flexural strengths of SiCuf /geopolymer composites in both x and z directions. The composite showed anisotropic characteristic, that is, higher flexural strength in the x direction than that in the z direction. But with the increase

8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites

299

200

x direction z direction

180

Flexural strength,(MPa)

160 140 120 100 80 60 40 20 0 0

10 20 Fiber content (vol %)

25

Fig. 8.33 Flexural strength of SiCuf /KGP composites in x and z directions, with permission from [18]

in fiber contents, the strength difference between two directions decreased. In the z direction, the gap between two layers was the matrix-rich area of weak and stress concentrated, resulting in decreasing of the strength in this direction. With increasing pressure, the gap between two layers (along z direction) decreased; therefore, the strength difference between two directions became not so evident. Figure 8.34 provides fracture toughness and work of fracture vs. fiber content of SiCuf /geopolymer composites, which showed similar trends to flexural strength. When fiber content was 20 vol%, the composite exhibited the highest fracture toughness and work of fracture, 4.26 MPa m1/2 and 4421.4 J m−2 , which were 15.2 and 81.6 times as high as those of pristine geopolymer. This implied that geopolymer provides a low-cost route to prepare advanced composites of high reliability. Figure 8.35 shows typical stress–displacement curves of geopolymer and SiCf /geopolymer composites. Geopolymer exhibited the brittle failure mode, whereas all the composites showed non-catastrophic one. For the composites, the strength tended to reduce slowly and several significant steep drop-steps were noted after the stress reached the maximum value. And then, the curves spread more widely with long tails, corresponding to interface debonding, fiber fracturing, and pulling-out. In both x and z directions, SiCuf /geopolymer composites deformed but without breaking completely after flexural tests (Fig. 8.36a, b). In contrast, the geopolymer matrix showed a typical brittle failure mode by breaking into two parts (Fig. 8.36c). The significant different failure highly depended on their different fracture energy, as analyzed in the aforementioned part. It was also observed that composites fractured in different failure modes in x and z directions, as shown in Fig. 8.37. In the x direction, the main crack was along with fiber radial direction (Fig. 8.37a), resulting in significant fiber debonding, fracture, and pulling-out (Fig. 8.37b, c). Therefore, composite failed in tensile fracture mode in the x direction. However, in the z direction, crack

300

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

6

6000 -2

4

4000

3

3000

2

2000

1

1000

0

0

20

15

10

-2

Fracture toughness (MPa· m

5000

Work of fracture(J·m )

Work of fracture (J· m )

5

1/2

)

Fracture toughness,MPa· m

1/2

0

25

Fiber content (vol %) Fig. 8.34 Fracture toughness and work of fracture versus fiber content of SiCuf /KGP composites, with permission from [18]

160

Flexural stress,MPa

140 120

S20

100 80

S10

60

S15

40 20 0 0.0

S25 M

×

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Displacement,mm Fig. 8.35 Stress and displacement curves of SiCuf /KGP composites inx direction, with permission from [18]

8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites

301

Fig. 8.36 Images of bar specimens of SiCf /KGP composites in a x direction, b z direction, and c geopolymer matrix, with permission from [18]

Fig. 8.37 Fractograph of SiCf /KGP composites in different directions. a–c x direction, d–e z direction, with permission from [18]

302

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

propagated mainly along fiber axial direction (Fig. 8.37d), and composites showed interface debonding between fiber and matrix, rather than fiber fracture, indicating the shear failure mode.

8.2.3 High-Temperature Mechanical Properties of SiCuf /Geopolymer Composites High-temperature flexural test was carried out on SiCuf /geopolymer composite with fiber content of 20 vol%. The target testing temperatures are 700, 800, 900, 1000, and 1100 °C, respectively, with heating rate of 10 °C/min and holding time of 10 min. The composites tested at room and high temperature are denoted as HS-20, HS-700, HS-800, HS-900, HS-1000, and HS-1100, respectively. (1) Mechanical properties Figure 8.38 presents the flexural strength and work of fracture of the SiCuf /KGP composites at high temperature in air atmosphere. Basically, the flexural strength of the composites decreases with increasing temperature linearly. However, the flexural strength decreases slightly at 700 °C. The flexural strength remains 86% at 900 °C and 70% at 1000 °C of its strength at room temperature, respectively. When the temperature is 1100 °C, the strength decreases significantly. These data are listed in Table 8.6. Different from the flexural strength at high temperature, the work of fracture at 700 °C is 22.6% higher than that of room temperature. As the temperature rises, the work of fracture of the composites also shows a linear decline tendency. The load–displacement curves of the SiCuf /KGP composites at high temperature are shown in Fig. 8.39. The curves of samples tested at room temperature, 700 and 800

(b)

(a) 180

5000

170

4500 -2

//

150

Work of Fracture, J •m

Flexural strength,MPa

160

140 130 120 110 100 90

4000 3500 3000 2500 2000 1500 1000

0

20

700

800

900

Temperature, °C

1000

1100

020 50

700

800

900

1000

1100

Temperature, °C

Fig. 8.38 High-temperature (a) flexural strength and work of fracture (b) of SiCuf /KGP composite [19]

8.2 Continuous SiC Fiber-Reinforced Geopolymer Matrix Composites Fig. 8.39 Load– displacement curve of the SiCuf /KGP composites at different testing temperatures [19]

303

140 120

Load,N

100 80

HS-20 HS-700

60 40

HS-800

HS-900 HS-1000

HS-1100

20 0

Displacement,mm

°C showed relatively typical “ladder-like” curves, including both elastic deformation and nonlinear deformation stage. These composites fail in non-catastrophical mode. However, the curves of the samples at 900, 100, and 1100 °C only have elastic deformation stage, showing brittle fracture mode. The macroscopic picture of the sample is shown in Fig. 8.40. At room temperature, 700, and 800 °C, the samples still maintain good integrity. When the test temperature increased, the specimen breaks into two parts, which is consistent with their brittle failure mode as observed in their load–displacement curves. Figure 8.41 shows the surface morphology of the SiCuf /KGP composite after high-temperature testing. The surface of the composite is relatively smooth with little microcracks before the high-temperature test. When the testing temperature is increased, cracks on the surface of the composite are perpendicular to the axial direction of SiCuf . These cracks are caused by the viscous sintering of geopolymer matrix. At high temperature, thermal shrinkage would take place during the sintering process of geopolymer. However, SiC fiber kept unchanged. Therefore, matrix was in tensile state, while fiber is in the compressive state. When tensile stress on the matrix exceeded its ultimate tensile strength, microcracks formed in the matrix perpendicular to the fiber axial direction. The fibers tend to bridge the cracks in the composite. The crack density kept almost not changed with the further increase of test temperature, and the crack width increased slightly. The fracture mechanism of the composite is related not only to the strength of both matrix and fiber but also to the interfacial bonding between the matrix and the fiber. The fractographs of the SiCuf /KGP composite is shown in Fig. 8.42. As indicated in Fig. 8.42a–c, the crack growth path is very circuitous and shows a “sawtooth” feature for HS-20, HS-700, and HS-800. Obvious fiber debonding and bridging can be noted, but most fibers do not fracture. However, when testing temperature is 900, 1000, and 1100 °C, as shown in Fig. 8.42d–f, for HS-900, HS-1000, and HS-1100, obvious fiber fracture was noted. Combining with the analysis of the load–displacement curve of the composite, brittle fracture exits in the samples tested at 900–1100 °C, which is

304

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

Fig. 8.40 Macro morphology images of the SiCuf /KGP composites after high-temperature testing [19]

mainly due to the high interface bonding strength between the fiber and the matrix due to the matrix shrinkage and the diffusion reaction at high temperature. It should also be pointed out that the microcracks on the composite surface did not result in the sharp decrease in mechanical performance of the composites, implying that the composite has a good tolerance for microcracks. This is also one of the characteristics of continuous fiber-reinforced composites.

8.3 Summary In this chapter, continuous carbon fiber and SiC fiber-reinforced geopolymer composites were reported, and the effects of fiber contents and post treatment on the microstructure and mechanical properties of the composites were studied. The following conclusion can be drawn:

8.3 Summary

305

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 8.41 Surface morphology of the SiCuf /KGP composites after high-temperature testing. a HS20; b HS-700; c HS-800; d HS-900; e HS-1000; f HS-1100 [19]

(a)

(b)

(c)

(d)

(e)

(f)

Fig. 8.42 Fracture surface of the SiCuf /KGP composites after high-temperature testing. a HS-20; b HS-700; c HS-800; d HS-900; e HS-1000; f HS-1100 [19]

306

8 Continuous Fibers-Reinforced Geopolymer Matrix Composites

(1) The obtained Cuf /geopolymer composites show a full infiltration and homogenous distribution of carbon fibers, and the fibers are undamaged during the preparation process. Compared with the geopolymer matrix, the Cuf /geopolymer composites exhibit drastically improved mechanical properties due to the incorporation of carbon fiber, especially to the composites reinforced by 20 vol% carbon fibers. The flexural strength, Young’s modulus, and work of fracture increase by 11.6 times, 3.6 times, and 71.5 times, separately, and all the composites fractured in non-brittle mode. (2) With increasing Si/Al ratio of the geopolymer matrix, flexural strength and Young’s modulus of the composites increased gradually due to the increased strength and stiffness of geopolymer resin. Considerable fiber pull-out was observed in composites with Si/Al ≥ 3, indicating the significant reinforcing effects of carbon fibers in these samples. Mechanical properties of both unidirectional and 2D carbon fiber-reinforced geopolymer matrix composites showed increasing trends with the increase in Si/Al ratios. (3) After proper high-temperature treatment, the Cuf /geopolymer composite can be converted into the carbon fiber-reinforced leucite ceramic matrix composites, and machinal properties were greatly enhanced. For the composite treated at 1100 °C, flexural strength, work of fracture, and Young’s modulus reach their highest values increasing by 102.3, 29.1, and 84.7%, respectively, relative to their original state before treatment. The property improvement can be attributed to the densified and crystallized matrix, and the enhanced fiber/matrix interface bonding based on the fine-integrity of carbon fibers. In contrast, for composite heat treated at higher temperature, the mechanical properties lowered substantially and it tended to fracture in a very brittle manner owing to the seriously degraded carbon fibers. (4) SiC fibers are fully impregnated by geopolymer matrix and the composites show homogeneous microstructure. When SiC fiber content is 20 vol%, the related composite presents the highest flexural strength, fracture toughness, and work of fracture, which are 14.2, 15.2, and 81.6 times as high as those of geopolymer matrix, respectively, indicating significant strengthening and toughening effect from SiC fiber. All the SiCuf /geopolymer composites showed non-brittle failure modes rather than the catastrophic failure as observed in neat geopolymer. Meanwhile, the composites failed in tensile failure mode in the x direction and shear failure mode in the z direction. (5) SiCuf /geopolymer composites show high strength retention rate (>87%) at high temperature range up to 900 °C. Even at 1100 °C, they still can retain 69.3% of their room temperature flexural strength. When tested at a temperature below 800°C, the composites fractured in a non-catastrophic failure mode. In contrast, when tested at 900 °C or a higher temperature, they fractured in a brittle failure mode.

References

307

References 1. P. He, Processing and characterization of the Cf/geopolymer and derived ceramic matrix composites (in Chinese). Thesis for the Doctor Degree, Harbin Institute of Technology, 2011 2. T. Lin, Mechanical properties and fracture behavior of Csf(Al2 O3 p) reinforced geopolymer matrix composites (in Chinese). Thesis for the Doctor Degree, Harbin Institute of Technology, 2009 3. T. Lin, D. Jia, P. He, M. Wang, In situ crack growth observation and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 527, 2404–2407 (2010) 4. T. Lin, D. Jia, P. He, M. Wang, D. Liang, Effects of fiber length on mechanical properties and fracture behavior of short carbon fiber reinforced geopolymer matrix composites. Mater. Sci. Eng. A 497, 181–185 (2008) 5. V.F.F. Barbosa, K.J.D. MacKenzie, Synthesis and thermal behaviour of potassium sialate geopolymers. Mater. Lett. 57, 1477–1482 (2003) 6. V.F.F. Barbosa, K.J.D. MacKenzie, Thermal behaviour of inorganic geopolymers and composites derived from sodium polysialate. Mater. Res. Bull. 38, 319–331 (2003) 7. J.L. Bell, P.E. Driemeyer, W.M. Kriven, Formation of ceramics from metakaolin-based geopolymers: Part I: Cs-based geopolymer. J. Am. Ceram. Soc. 92, 1–8 (2009) 8. J.L. Bell, P.E. Driemeyer, W.M. Kriven, Formation of ceramics from metakaolin-based geopolymers. Part II: K-based geopolymer. J. Am. Ceram. Soc. 92, 607–615 (2009) 9. P. Duxson, A. Fernández-Jiménez, J.L. Provis, G.C. Lukey, A. Palomo, J.S.J. Van Deventer, Geopolymer technology: the current state of the art. J. Mater. Sci. 42, 2917–2933 (2007) 10. P. Duxson, G.C. Lukey, J.S.J. van Deventer, The thermal evolution of metakaolin geopolymers: Part 2: Phase stability and structural development. J. Non-Cryst. Solids 353, 2186–2200 (2007) 11. R.A. Fletcher, K.J.D. MacKenzie, C.L. Nicholson, S. Shimada, The composition range of aluminosilicate geopolymers. J. Eur. Ceram. Soc. 25, 1471–1477 (2005) 12. P. He, D. Jia, T. Lin, M. Wang, Y. Zhou, Effects of high-temperature heat treatment on the mechanical properties of unidirectional carbon fiber reinforced geopolymer composites. Ceram. Int. 36, 1447–1453 (2010) 13. N. Klinger, E.L. Strauss, K.L. Komarek, Reactions between silica and graphite. J. Am. Ceram. Soc. 49, 369–375 (1966) 14. Q.-S. Ma, Z.-H. Chen, W.-W. Zheng, H.-F. Hu, Effects of pyrolysis processes on microstructure and mechanical properties of Cf/SiOC composites fabricated by preceramic polymer pyrolysis. Ceram. Int. 31, 305–314 (2005) 15. D. Koch, K. Tushtev, G. Grathwohl, Ceramic fiber composites: experimental analysis and modeling of mechanical properties. Compos. Sci. Technol. 68, 1165–1172 (2008) 16. G.H. Zhou, S.W. Wang, J.K. Guo, Z. Zhang, The preparation and mechanical properties of the unidirectional carbon fiber reinforced zirconia composite. J. Eur. Ceram. Soc. 28, 787–792 (2008) 17. D.C. Jia, Y. Zhou, T.C. Lei, Ambient and elevated temperature mechanical properties of hotpressed fused silica matrix composite. J. Eur. Ceram. Soc. 23, 801–808 (2003) 18. P. He, D. Jia, B. Zheng, S. Yan, J. Yuan, Z. Yang, X. Duan, J. Xu, P. Wang, Y. Zhou, SiC fiber reinforced geopolymer composites. Part 2: Continuous SiC fiber. Ceram. Int. 42, 12239–12245 (2016) 19. B. Zheng, Mechanical properties of unidirectional SiC fiber reinforced potassium-based geopolymer composites (in Chinese). Thesis for the Master Degree, Harbin Institute of Technology, 2013

Index

A Al2 O3 particle, 13, 19, 131, 145, 216, 217, 219, 220, 222, 239, 253, 259, 262 Alkaline silicate solutions, 9, 10, 56, 74, 76, 118, 127 Aluminosilicate powders, 8, 35 Amorphous, 1, 9, 10, 20, 28, 36, 38, 39, 41, 46, 47, 53, 54, 56, 57, 60, 63, 68, 77, 109, 110, 118, 119, 121, 127, 131, 132, 137, 146, 151, 154, 159, 166, 167, 176, 179, 181, 186, 217, 243, 245, 246, 255, 262, 266, 271, 288, 295

B Brittle failure mode, 186, 192, 238, 247, 249, 258, 274, 294, 299, 303, 306

C Carbon fiber, 4, 13, 17, 21–24, 26, 179–182, 184–188, 191, 194, 195, 197, 199, 200, 210–212, 219, 223, 226, 239, 243, 271, 272, 274, 276, 279, 280, 282, 284, 285, 287, 289–291, 304, 306 Carbon nanotube, 13, 16, 17, 124, 131, 134, 135, 175 Chemical activity, 14, 15, 35, 36, 41, 48, 49, 52, 53, 55, 77, 78 Chromium powder, 148, 150 Composites, 1, 3, 4, 7, 8, 11–14, 16–26, 29, 81–84, 110, 111, 118–122, 124, 126, 127, 131, 132, 134–167, 175, 176, 179, 181–209, 211–228, 230–239, 243–269, 271–306

Continuous fibers, 3, 4, 12, 16, 17, 23, 24, 201, 239, 271, 304

D Dehydroxylation reaction, 36

F Fiber pulling-outs, 193, 201, 204, 215, 223, 233–239, 274, 280, 285 Fly ash cenosphere, 154, 155 Fracture behavior, 3, 179, 185, 192, 193, 199, 201, 202, 209, 222, 223, 232, 239, 243, 247, 251, 255, 258, 263, 268, 269, 284, 286, 291, 294, 297 Fracture morphologies, 249, 250, 258, 260, 266, 267

G Geopolymer, 1–4, 7–29, 35, 38, 55–58, 60, 61, 64, 65, 69, 71, 73, 74, 76–78, 81– 84, 97–103, 105–111, 114–127, 131– 147, 149–170, 172, 174–176, 179– 209, 211–217, 219–224, 226–228, 231, 232, 235, 236, 238, 239, 243– 259, 261–269, 271–279, 282–291, 294–299, 301–304, 306 Geopolymerization, 1, 3, 4, 7–10, 12, 15, 16, 18, 29, 35, 55–58, 61, 64–73, 77, 78, 98–112, 114–118, 121, 127, 132, 135, 152, 167, 168, 186, 187, 196, 295 Graphene oxide, 4, 10, 13, 18, 19, 82, 111 Graphite powder, 20, 131, 136

© Springer Nature Singapore Pte Ltd. 2020 D. Jia et al., Geopolymer and Geopolymer Matrix Composites, Springer Series in Materials Science 311, https://doi.org/10.1007/978-981-15-9536-3

309

310

Index

H High-temperature treatments, 4, 14, 22, 26, 244, 255, 256, 259, 260, 263, 266, 268, 286, 306

I Inorganic polymeric materials, 1 Interface, 17, 21–23, 121–123, 152, 154, 156, 158, 159, 171, 176, 179, 199, 209, 224, 225, 228, 232, 233, 272, 274–276, 279–281, 293, 299, 302, 304, 306

126, 127, 161, 170, 213, 222, 246, 258, 289–291,

K Kaolin, 35–54, 57, 58, 61, 63, 77, 81, 131– 133, 146, 179, 243, 253, 255, 262, 271

L Leucite, 17, 18, 26, 27, 217, 254, 255, 258, 262, 266, 286, 288, 290, 291, 306 Load-displacement curve, 149, 185, 186, 192, 199, 201, 215, 222, 247, 249– 251, 258, 261, 266, 269, 294, 302, 303

M Mechanical properties, 3, 4, 13, 15–24, 26, 28, 29, 119, 121, 122, 124, 126, 127, 134–136, 139, 146, 151–153, 165, 170, 172, 176, 179, 182, 184, 189, 190, 196, 198, 201, 209, 213, 216, 218, 221, 222, 224, 226, 239, 243, 245, 247–249, 253–255, 257–259, 263, 265, 269, 271, 272, 274–276, 278, 280, 284–286, 290–293, 295, 297, 302, 304, 306 Metakaolin, 1, 2, 4, 8–10, 15, 19, 25, 35–38, 41, 44, 45, 48–50, 52–58, 60–64, 69, 71, 73, 74, 76–78, 81, 82, 98–103, 106, 107, 111, 118, 119, 127, 131– 133, 137, 143, 145–148, 151, 152, 156, 166–170, 174, 175, 179, 181, 186, 188, 217, 243–245, 253, 271, 288, 294–296 Microstructure, 2, 7–10, 18, 19, 22, 40, 41, 46, 56, 57, 61, 63–65, 81, 90, 93,

118–121, 123, 125–127, 134, 138, 154–156, 159, 170–172, 182, 183, 188, 195, 196, 211, 216, 217, 219, 220, 243, 245, 248, 253, 254, 259, 263, 264, 273, 275–277, 282, 286, 290, 295–297, 304, 306 Microwave adsorption properties, 140

135, 179, 213, 246, 272, 291,

N Non-catastrophic fracture, 21, 24, 192, 193, 201, 209, 222, 239, 258, 268, 269

R Reduction, 4, 18, 29, 43, 55, 81–87, 89–95, 97, 98, 111, 119, 120, 126, 127, 143, 219, 265 Reinforcement, 3, 4, 12, 13, 16, 17, 19, 20, 24, 29, 126, 139, 171, 175, 176, 179, 199, 290, 291

S SEM observation, 56, 150, 152, 201 Short carbon fiber, 4, 13, 20, 22, 179–181, 183, 188, 191–193, 196–202, 205– 207, 209, 211, 212, 214, 216, 221, 222, 224, 226, 227, 232, 238, 243 Short SiC fiber, 13, 21, 243 Si/Al ratio, 2, 4, 8, 13–16, 20, 24, 57, 166, 176, 275–280, 282–286, 306 SiC fiber, 13, 17, 21, 23, 24, 226, 227, 243, 271, 294–296, 298, 303, 304, 306 SiO2 , 4, 8, 16, 17, 56, 62, 82, 131, 132, 137, 143, 145, 151, 154, 156, 166–168, 170, 172, 174–176, 179, 181, 186, 187, 243, 245, 253–255, 262, 271, 275, 276, 284, 289, 290, 294 Strengthening, 3, 16, 21, 124, 126, 134, 135, 146, 150, 175, 179, 191, 198, 213, 215, 224, 225, 232, 238, 247–249, 274–276, 280–282, 285, 306

T Thermal conductivity, 3, 15, 19, 20, 26, 154, 155, 161, 164, 165, 176 Thermal evolution, 19, 20, 29, 254, 259 Toughening mechanisms, 21, 122, 179, 193, 215, 232, 238, 243, 269