The Chemistry and Technology of Magnesia [1st edition] 0471656038, 9780471656036, 9780471980568

Citation preview

THE CHEMISTRY AND TECHNOLOGY OF MAGNESIA

MARK A. SHAND Premier Chemicals, LLC Findlay, Ohio

A JOHN WILEY & SONS, INC. PUBLICATION

THE CHEMISTRY AND TECHNOLOGY OF MAGNESIA

THE CHEMISTRY AND TECHNOLOGY OF MAGNESIA

MARK A. SHAND Premier Chemicals, LLC Findlay, Ohio

A JOHN WILEY & SONS, INC. PUBLICATION

Copyright # 2006 by John Wiley & Sons, Inc. All rights reserved. Published by John Wiley & Sons, Inc., Hoboken, New Jersey. Published simultaneously in Canada. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/ permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Shand, Mark A. The chemistry and technology of magnesia/Mark A. Shand. p. cm. Includes index. ISBN-13: 978-0-471-65603-6 (cloth) ISBN-10: 0-471-65603-8 (cloth) 1. Magnesium oxide. 2. Magnesium oxide—Industrial applications. 3. Magnesium oxide—Safety measures. I. Title. TP889.S53 2006 6610 .0932—dC22 Printed in the United States of America 10 9 8 7 6

5 4 3

2 1

2005032305

CONTENTS

Preface Acknowledgments 1

History of Magnesia

xv xvii 1

1.1 History of Magnesia, 1 Bibliography, 4 2

Formation and Occurrence of Magnesite and Brucite

5

2.1 Introduction, 5 2.2 Sedimentary Magnesite—Basis for Carbonate Deposition, 7 2.2.1 Secondary Nodular Magnesite, 8 2.2.2 Biogenic Carbonate, 8 2.3 Serpentine Alteration by Hydrothermal Processes, 10 2.4 Cryptocrystalline Magnesite Formation by Infiltration, 11 2.5 Crystalline Magnesite—Replacement of Limestone and Dolomite, 11 2.6 Brucite, 12 2.7 Worldwide Occurrence of Magnesite and Brucite, 12 2.7.1 United States and Canada, 13 2.7.2 Brazil, 17 v

vi

CONTENTS

2.7.3 Australia, 17 2.7.4 China, 22 2.7.5 North Korea, 24 2.7.6 Nigeria, 24 2.7.7 South Africa, 24 2.7.8 India, 24 2.7.9 Saudi Arabia, 26 2.7.10 Iran, 26 2.7.11 Greece, 26 2.7.12 Turkey, 27 2.7.13 Serbia and Bosnia, 28 2.7.14 Austria, 28 2.7.15 Russia, 29 2.7.16 Slovakia, 29 2.7.17 Spain, 30 2.8 Physical and Chemical Properties of Magnesite, 30 2.9 Chemical and Physical Properties of Brucite, 33 Bibliography, 35 References, 35 3

Synthetic Magnesia 3.1 3.2

3.3

Introduction, 39 Composition of Seawater and Brines, 41 3.2.1 Seawater Chemistry, 41 3.2.2 Brine Extraction, 42 3.2.3 Sump Leaching Phase, 42 3.2.4 Preparation Phase, 43 3.2.5 Production Phase, 44 3.2.6 Evaporite Production, 44 3.2.7 Subsurface Brines, 44 Process Description, 45 3.3.1 Precipitation Reaction, 45 3.3.2 Influence of Reaction Conditions on Mg(OH)2 Particle Morphology, 46 3.3.3 Dolime/Lime Requirements, 47 3.3.4 Seawater Pretreatment, 48

39

CONTENTS

vii

3.3.5 Precipitation Process, 50 3.3.6 Settling and Compaction, 51 3.3.7 Washing, 52 3.3.8 Filtration, 52 3.3.9 Brine Precipitation, 54 3.4 Calcination, 55 3.5 Grinding, 55 3.6 Packaging, 56 3.7 Sampling and Testing and In-Process Quality Control, 56 3.7.1 Dolime, 56 3.7.2 Seawater, 57 3.7.3 Reactor, 57 3.7.4 Settling/Thickener, 57 3.7.5 Washing, 58 3.7.6 Filtration, 58 3.7.7 Calcining, 58 3.7.8 Grinding, 58 3.7.9 Finished Product, 58 3.8 Aman Process, 59 3.8.1 Pyrohydrolysis of Magnesium Chloride Hexahydrate, 59 3.9 General Properties of Synthetic Magnesia, 60 Bibliography, 60 References, 60 4

Mining and Processing Magnesite 4.1 Mining Operations, 63 4.1.1 Overburden Removal, 63 4.1.2 Drilling, 63 4.1.3 Bench Height, 64 4.1.4 Hole Diameter, 64 4.1.5 Burden and Spacing, 64 4.1.6 Subdrilling, 65 4.1.7 Hole Stemming, 66 4.1.8 Blast Hole Pattern, 66 4.1.9 Blast Timing, 67

63

viii

CONTENTS

4.1.10 Blasting Agents, 68 4.1.11 Secondary Blasting, 68 4.1.12 Chemical Contour and Muck Maps, 68 4.2 Processing Magnesite, 69 4.2.1 Ore Removal and Primary Crushing, 70 4.2.2 Gyratory and Cone Crushers, 71 4.2.3 Jaw Crushers, 72 4.2.4 Roll Crushers, 72 4.2.5 Size Separation, 73 4.2.6 Screening, 74 4.2.7 Pneumatic (Air) Classification, 75 4.2.8 Hydroclones, 76 4.3 Gravity Concentration, 76 4.3.1 Float –Sink Separation, 76 4.3.2 Froth Floatation, 77 4.3.3 Floatation Reagents, 79 4.3.4 Floatation Machines, 80 4.4 Tertiary Crushing, 81 4.5 Postcalcination Screening and Grinding, 81 References, 81 5

Calcination of Magnesium Hydroxide and Carbonate

83

5.1

Calcination of Magnesite, 83 5.1.1 Energy Requirement for Calcination Process, 85 5.1.2 Effect of Time and Temperature, 85 5.1.3 Kinetics of Calcination, 85 5.1.4 Stone Size, 88 5.2 Calcination of Magnesium Hydroxide, 88 5.2.1 Energy Requirement for Calcination Process, 89 5.2.2 Decomposition Mechanism, 90 5.2.3 Kinetics of Decomposition, 93 5.2.4 Effect of Time and Temperature, 94 References, 96 6

Furnaces and Kilns 6.1

Introduction, 97

97

ix

CONTENTS

6.2 Multiple-Hearth Furnaces, 98 6.2.1 Single Progressive Rabble (Four Arms per Hearth), 101 6.2.2 Full Progressive Rabble (Four Arms per Hearth), 102 6.2.3 Back Rabble (Four Arms per Hearth), 102 6.2.4 Full Progressive Rabble (Two Arms per Hearth), 102 6.2.5 Refractory Linings, 102 6.3 Horizontal Rotary Kilns, 103 6.4 External Water Coolers, 105 6.5 Shaft Kilns, 107 6.5.1 Ore Charging, 107 6.5.2 Discharge, 107 6.5.3 Modern Shaft Kiln, 109 6.5.4 Double-Inclined Kiln, 109 6.5.5 Multichamber Kiln, 109 6.5.6 Annular Shaft Kiln, 111 6.5.7 Parallel-Flow Regenerative Kiln, 113 7

Postcalcination Processing

115

7.1 Introduction, 115 7.2 Grinding, 115 7.2.1 Ring-Roller Mills, 116 7.2.2 Ball Mills, 117 Bibliography, 119 Reference, 119 8

Physical and Chemical Properties of Magnesium Oxide 8.1 Introduction, 121 8.2 Physical Properties of Magnesium Oxide, 121 8.3 Chemical Properties of Magnesium Oxide, 125 8.3.1 Dissolution of Magnesium Oxide, 126 8.4 Surface Structures of MgO, 127 8.5 Molecular Adsorption on MgO, 129 8.5.1 Chemisorption of Various Molecules on MgO, 129 Bibliography, 130 References, 131

121

x

CONTENTS

9

Other Magnesia Products

133

9.1 Production of Hard-Burned Magnesia, 133 9.2 Production of Dead-Burned Magnesia, 133 9.2.1 Sintering, 139 9.2.2 Sinter Aids, 142 9.2.3 Production Methods, 144 9.3 Fused Magnesia, 144 9.3.1 Refractory-Grade Fused Magnesia, 145 9.4 Magnesium Hydroxide Slurry, 146 9.4.1 Production of Synthetic Magnesium Hydroxide Slurry, 146 9.4.2 Hydration of Magnesium Oxide, 148 9.4.3 Hydration Kinetics and Mechanisms, 150 9.4.4 Testing and Quality Control of Magnesium Hydroxide Slurry, 151 9.5 Purification by Carbonation of Magnesium Hydroxide Slurry, 151 9.5.1 Precipitation of Magnesium Carbonate from Bicarbonate Solution, 153 References, 153 10 Water and Wastewater Applications for Magnesia Products 10.1 Introduction to Applications, 155 10.2 Industrial Wastewater Treatment, 155 10.3 Advantages of Magnesium Hydroxide in Wastewater Treatment, 157 10.3.1 Safety, 157 10.3.2 pH Control, 157 10.3.3 Metals Removal, 158 10.3.4 Sludge Volume and Dewatering, 159 10.3.5 Treatment Methods, 161 10.3.6 Handling Requirements, 161 10.3.7 Environmental Impact, 163 10.4 Adsorption of Dyes on Magnesium Hydroxide, 163 10.5 Biological Wastewater Treatment, 163 10.5.1 Aerobic Processes, 164

155

CONTENTS

xi

10.5.2 Nitrification, 164 10.5.3 Anaerobic Digestion, 165 10.6 Bioflocculation and Solids Settling, 166 10.7 Phosphorus Removal from Wastewater and Struvite Formation, 167 10.8 Odor and Corrosion Control in Sanitary Collection Systems, 168 10.8.1 Addition to Raw Sewage, 170 10.8.2 Crown Spraying, 172 10.9 Acid Mine Drainage, 172 10.10 Silica Removal from Industrial Plant Water, 173 10.10.1 Mechanism of Silica Removal, 174 10.10.2 Factors Controlling the Removal of Silica, 174 References, 176 11

Magnesia in Polymer Applications

179

11.1 Magnesium Hydroxide as a Flame Retardant for Polymer Applications, 179 11.2 Flame-Retardant Mechanisms, 180 11.3 Properties Required of Magnesium Hydroxide for Flame-Retardant Applications, 181 11.3.1 Surface Treatment, 182 11.3.2 Stearic Acid, 183 11.3.3 Silanes, 184 11.4 Novel Applications for Magnesium Hydroxide as a Flame Retardant, 184 11.5 Polymer Curing and Thickening, 184 11.5.1 Sheet Molding Compound (SMC), 184 11.5.2 Synthetic Rubber, 185 Bibliography, 186 References, 186 12

Environmental Applications 12.1 Flue Gas Desulfurization, 189 12.2 Regenerative Process, 189 12.2.1 Process Description, 190 12.2.2 Once-Through Process, 192

189

xii

CONTENTS

12.2.3 Kawasaki Process, 192 12.2.4 Dravo Thiosorbic Process with Magnesium Hydroxide Recovery, 194 12.2.5 Sorbtech Process, 194 12.3 Remediation Applications, 194 12.4 Nuclear Waste Disposal, 196 12.5 Hazardous Spill Cleanup, 196 12.6 Antibacterial Activity of Magnesium Oxide Powder, 197 12.7 Carbon Dioxide Sequestration Using Brucite, 197 Bibliography, 198 References, 198 13 Role of Magnesium in Animal, Plant, and Human Nutrition

201

13.1 Role of Magnesium in Plant Nutrition, 201 13.1.1 Uptake of Magnesium from the Soil, 202 13.1.2 Functions of Magnesium in Plant Growth, 202 13.2 Magnesium Fertilizers, 203 13.3 Magnesium in Animal Nutrition, 203 13.3.1 Ruminant Animals, 204 13.3.2 Magnesium Oxide Requirements for Animal Nutrition, 205 13.3.3 Factors Affecting Magnesium Utilization, 205 13.3.4 Magnesium Bioavailability, 206 13.3.5 Preventing Grass Tetany by Magnesium Fertilization, 207 13.3.6 Preventing Grass Tetany by Oral Supplementation, 207 13.3.7 Magnesium Requirements of Swine, 207 13.3.8 Magnesium Requirements of Poultry, 208 13.3.9 Magnesium in Dairy Ration Buffers, 208 13.4 Magnesium in Human Health and Nutrition, 209 13.4.1 Health Benefits of Magnesium, 209 Bibliography, 210 References, 211 14 Magnesium Salts and Magnesium Metal 14.1 Magnesium Acetate, 215 14.2 Magnesium Alkyls, 215

215

CONTENTS

xiii

14.3 14.4 14.5 14.6 14.7 14.8 14.9

Magnesium Chloride, 216 Magnesium Nitrate, 217 Magnesium Sulfate, 217 Magnesium Soaps, 218 Magnesium Overbase Sulfonates, 218 Magnesium Peroxide, 219 Magnesium Metal Production, 220 14.9.1 Thermal Processes, 220 14.9.2 Electrolytic Production, 221 Bibliography, 221 15

Pulp Applications

223

15.1 Sulfite Pulping, 223 15.2 Magnefite Pulping Process, 223 15.2.1 Pulping Liquor Preparation, 225 15.2.2 Pulping Process, 226 15.2.3 Chemical Recovery, 226 15.3 Pulp Bleaching, 226 15.3.1 Oxygen Bleaching and Delignification, 227 15.3.2 Hydrogen Peroxide Bleaching, 228 15.4 Deinking, 229 Bibliography, 229 References, 230 16

Magnesia Cements 16.1 Introduction, 231 16.2 Magnesium Oxychloride Cement, 231 16.2.1 Phase Formation, 232 16.2.2 Water Resistance of MOC Cement, 233 16.3 Magnesium Oxysulfate (MOS) Cement, 234 16.4 Thermal Insulative and Fire Resistance Properties of Sorel Cement, 234 16.4.1 Thermal Insulation, 234 16.4.2 Fire Resistance, 235 16.5 Magnesium Phosphate Cement, 235 16.5.1 Reaction Mechanism, 235

231

xiv

CONTENTS

16.5.2 Magnesium Phosphate Cement Derived from Ammonium Dihydrogen Phosphate, 236 16.5.3 Magnesium Phosphate Cement Derived from Diammonium Phosphate, 237 16.5.4 Magnesium Phosphate Cement Derived from Ammonium Polyphosphate, 237 16.5.5 Magnesium Phosphate Cement Derived from Potassium Dihydrogen Phosphate, 238 Bibliography, 238 References, 238 17 Miscellaneous Magnesia Applications

241

17.1 Sugar Manufacture, 241 17.1.1 Clarification and Filtration, 243 17.1.2 Reduction in Scaling, 243 17.1.3 Additional Benefits, 243 17.2 Chrome Tanning of Leather, 244 17.3 Magnesia as a Catalyst Support, 245 17.3.1 Example of Magnesium Oxide as Catalyst, 245 17.4 Fuel Additives, 246 17.4.1 High-Temperature Corrosion, 247 17.4.2 Low-Temperature Corrosion, 248 17.4.3 Types of Additives for High-Temperature Corrosion, 248 17.4.4 Types of Additives for Low-Temperature Corrosion, 249 17.5 Well-Drilling Fluids, 250 17.6 Nanoparticulate Magnesia, 251 17.6.1 Synthesis of Nanoparticulate Magnesia, 251 17.6.2 Chemical and Catalytic Properties of Nanocrystals, 251 17.7 Transformer Steel Coating, 254 Bibliography, 254 References, 254 Appendix

257

Index

263

PREFACE

This book attempts to encompass “all things magnesia.” Although the magnesia industry is similar in many respects to the much larger lime industry, there is to my knowledge no text concentrating solely on magnesia. There are, however, several excellent texts covering the lime industry. Unfortunately, in these books magnesia is practically a footnote. Although lime and limestone production far exceeds that of the magnesia industry (total lime production in the United States in 2004 was 20.4 million metric tonnes, compared with a total magnesia production of 280,000 metric tonnes), magnesia is still an important chemical and maintains many niche applications. By far, the largest consumer of magnesia worldwide is the refractory industry, which consumed about 56% of the magnesia in the United States in 2004, the remaining 44% being used in agricultural, chemical, construction, environmental, and other industrial applications. This text starts with the geological occurrences of magnesite and brucite, followed by the processing of magnesite to the end product magnesium oxide. The production of magnesium hydroxide and magnesium oxide by precipitation from seawater and brine sources is also introduced, along with details on the wide range of applications in which magnesia is utilized. These applications span animal feed to wastewater treatment, catalyst support and fertilizers to the production of pulp and paper. Like many industries, a certain amount of jargon arises as the industry matures, and the magnesia industry is no exception to this. However, there may be some confusion as to which compound magnesia really applies. xv

xvi

PREFACE

My definition is that the term magnesia is a generalization for magnesium oxide, whether it is derived from natural magnesite or extracted from seawater or brine. The term magnesite, in the strictest sense, refers to the mineral consisting of magnesium carbonate, but the same term is often used for the oxide, that is, dead-burned magnesite, the term even being used when the oxide has been produced from seawater or brine sources. The major products produced by the magnesia industry are magnesium carbonate (magnesite); magnesium hydroxide, both natural (brucite) and that derived from seawater and brine; magnesium oxide, which in itself has a number of categories, namely, light-burned or caustic-calcined MgO, hard-burn MgO, and dead-burn MgO, or as it is otherwise known, periclase; and the last category, fused magnesia. Light-burn or caustic-calcined MgO refers to a product that has been calcined at the lower end of the temperature spectrum, typically 1500 –17008F. This product typically has the highest reactivity and greatest specific surface area of the entire magnesium oxide category. Hard-burn MgO is calcined at a higher temperature, 2400–28008F, and has a correspondingly lower reactivity and surface area. Dead-burn MgO, or periclase, is produced at temperatures above 28008F, which having a very small surface area, makes it unreactive. Finally, fused magnesia, produced at temperatures above the fusion point of magnesium oxide (28008C), is the least reactive.

ACKNOWLEDGMENTS

First and foremost, I am indebted to my wife Karen, whose encouragement kept me working during periods when my enthusiasm waned, and also to my kids, Alex, Victoria, and Madalyn, whose persistent question “haven’t you finished that book yet?” also provided the needed impetus to get the text finished. I also owe a debt of gratitude to John Gehret and the board of directors of Premier Chemicals, LLC for allowing me the opportunity to write the book. I would also like to especially thank my long-time friend and mentor, Dr. Ronald Wardle, along with John Noble of Vesuvius USA and Mark Wajer of Martin Marietta Magnesia Specialties for diligently correcting the text and providing helpful suggestions. I would also like to thank Lynden Johnson for increasing my knowledge of the process of mining magnesite. Finally, I would like to thank Gerry Spoors, Dr. Theofilos Zampetakis, and John Turner for their affirmation that this book should be written, and to John Wiley & Sons for agreeing to publish said text.

xvii

1 HISTORY OF MAGNESIA

1.1

HISTORY OF MAGNESIA

Magnesia alba, otherwise known to alchemists as white magnesia or mild magnesian earth, is known today as magnesite or magnesium carbonate, MgCO3. Magnesia nigra, however, refers to black manganese oxide, MnO2. Both of these names are derived from Magnesia, Ma´gvh´sıa´, which is a prefecture in Thessaly, Greece. Manganese and magnesium, as well as iron, are abundant in the form of oxides and carbonates in this region, and these minerals were referred to as “stones from Magnesia.” The iron oxides present in magnesia were in the form of magnetic magnetite or lodestone, and both magnesia alba and magnesia nigra contain large amounts of magnetite, thus making them magnetic. This explains why magnesium and magnet are both derived from the place name Magnesia. In alchemical terms, magnesia meant “a stone shining like silver” and was purported to be an ingredient of the philosopher’s stone. In the more modern sense of the word, it is thought to have originated from magnes carneus, which means flesh magnet from the way it stuck strongly to the lips. Bergman’s essay (Bergman et al. 1784– 1791) “De Magnesia” claimed that the Roman Count di Palma prepared a white powder that he claimed was a panacea for all diseases. The white powder was called “magnesia alba,” or Count Palma’s powder, and its origin was a closely guarded secret. The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 1

2

HISTORY OF MAGNESIA

In 1701 M.B. Valentini (Valentini, 1707) prepared magnesia alba by the calcination of the residue remaining after evaporating the mother liquor to dryness from the preparation of niter or potassium nitrate. To add even more to the confusion, it is apparent that at least three minerals of different chemical composition were called magnesia: (1) magnesius lapis, which referred to magnetic magnetite, (2) magnesia nigra, which refers to pyrolusite (MnO2), and (3) a silver-white mineral that was probably steatite or talc. At the beginning of the eighteenth century the term manganes was employed for the manganese mineral and magnesia for the white mineral. However, the difference between magnesia nigra and magnesius lapis was not demonstrated until the middle of the eighteenth century. Hoffman (Hoffman, 1729) was the first to recognize the differences between magnesia and lime. He stated that an alkaline earth prepared by the reaction of a bitter salt (Epsom salts) with a fixed alkali differed from lime. Whereas lime gave a sparingly soluble salt with sulfuric acid that was nearly without taste, magnesia alba gave a bitter soluble salt. However, it was not until 1754 that magnesia was finally distinguished from lime by Joseph Black. Black was the first person to recognize that magnesium was an element. Black, a prominent professor of anatomy and chemistry at Edinburgh, showed that magnesia alba (magnesium carbonate), when heated, evolved “fixed air” (carbon dioxide). The residue from this heating, calcined magnesia (magnesium oxide), was lighter and more alkaline than the basic carbonate. Limestone (calcium carbonate) was found to behave in the same manner. Black also demonstrated that magnesia alba produced a soluble sulfate in contrast to lime. He gave the alkaline earth the name magnesia. Black’s thesis (Black, 1777) presented in June 1754, “On the Acid Humour Arising from Food and Magnesia Alba” dealt primarily with the value of magnesia as an antacid. In 1808 Humphrey Davy (Davy, 1808) proved definitively that magnesia is the oxide of a metal, which he named magnium. At this juncture, the term magnesium was being used by some to define metallic manganese. Davy’s technique involved mixing moistened alkaline-earth oxide with mercuric sulfide (cinnabar) and placing the paste onto a platinum plate. A drop of mercury was dropped into a depression made in the paste and the whole covered with naphtha. The platinum plate and drop of mercury were then connected to the poles of a voltaic pile. The resultant amalgam that formed on the mercury pole was then transferred to a glass tube and the mercury distilled off. Davy described the characteristics of magnesium as follows (Davy, 1808) The metal from magnesia appears to react with the glass, especially before all the mercury has distilled off. In one experiment, in which I interrupted the distillation before the mercury had been completely removed, the metal appeared

1.1 HISTORY OF MAGNESIA

3

as a solid body, which exhibited the same white color and the same luster as the other metals of the alkalide-earths. It immediately sank to the bottom of the water although surrounded by gas bubbles, formed magnesia. It changed quickly in the atmosphere, a white crust forming, and finally it disintegrated into a white powder, which proved to be magnesia.

Eventually, much to the consternation of Davy, the name magnesium was adopted for the metal in magnesia alba and manganese for the metal in pyrolusite. Michael Faraday produced magnesium metal in 1833 by the electrolysis of fused anhydrous magnesium chloride. The mineralogical term magnesite was first applied to a series of magnesium salts (carbonate, sulfate, nitrate, and chloride) by J.C. Delame´thrie (Delame´thrie, 1795) in 1795. The same term was also being applied to magnesium carbonates and silicates by A. Brongniart (Brongniart, 1807). Deposits of natural magnesium carbonate were discovered at Hrubschu¨tz in Moravia, which is now Hrubsˇice in the Czech Republic, and named Kohlensaurer Talkerde by W.A. Lampadius in 1800 (Lampadius, 1800). C.F. Ludwig described these minerals as talcum carbonatum in 1803. The use of the term magnesite was first restricted to the carbonate minerals by D.L.G. Karsten in 1808 (Karsten, 1808). The name magnesite gradually grew in acceptance. Deposits of magnesite were found in Austria and Greece during the later half of the nineteenth century, and around the same time magnesite mines were opened in Canada. In 1886 magnesite was discovered in California and commercial mining commenced around 1900, and in 1913 production of magnesia commenced in Pennsylvania. The development of the magnesite industry was accelerated in 1914 by the outbreak of World War I as supplies from Austria and Greece were cut-off by the blockade of central European powers. Magnesite was found in Stevens County, Washington, in 1916 and mining started in 1917. During World War I, California magnesite came from small deposits in the Porterville district, which were operated in a crude fashion by small owners or contractors. The deposits situated at Magnesite, California, which is a short distance from Portersville, were developed in a more systematic manner, and after the war formed the nucleus of a sound mining and production operation. However, much of the early magnesite technology was developed in the Portersville district, such as the mechanical beneficiation of magnesite and the calcining of magnesite to develop a product with specific and controllable characteristics. It was also at Portersville that the first high-purity crystalline magnesium oxide was produced in a rotary kiln. Commercial production of refractorygrade magnesia was also in operation in the Livermore district of California by Western Mine and the Bald Eagle mine operation by Westvaco Chlorine products in Stanislaus County, California.

4

HISTORY OF MAGNESIA

An immense deposit of medium- to low-quality magnesite exists in Steven County, Washington, which was exploited by the Northwest Magnesite Company during World War I. Operations were centered on the towns of Chewelah and Valley. It was here that the first use of froth flotation to beneficiate magnesite was employed to reduce silica and lime content. A very large deposit of dolomite, magnesite, and brucite in the Paradise Range, Nye County, Nevada, has been known since about 1927 when brucite was discovered by Harry Springer. Drilling by U.S. Brucite in 1930 and 1931 revealed the presence of considerable quantities of magnesite adjacent to the brucite. From 1931 to 1933, the U.S. Geological Survey mapped the deposit and estimated that there were 71 million tons of magnesite and brucite bearing rock. Both magnesite and brucite were mined by Basic Ores, Inc., and the Sierra Magnesite Company. Currently, the only magnesite deposit being exploited in the United States is the one at Gabbs, Nevada.

BIBLIOGRAPHY Black, J. (1777). Experiments upon Magnesia ALBA, Quick-Lime, and other Alkaline Substances. W. Creech: Edinburgh. Bergman, T. O., Cullen, E. and Beddoes, T. (1784 – 1791). Physical and Chemical Essays. J. Murray: London. Brongniart, A. (1807). Traite´ E´le´mentaire de Mineralogic, 2. 489, Paris. Davy, H. (1808). Electro-chemical researches on the decomposition of the Earths; with observations on the metals obtained from the alkaline Earths, and on the amalgam procured form ammonia. Philosophical Transactions of the Royal Society of London, 98, pp. 333– 370. Delame´theric, J. C. (1975). The´orie de la Terre, 2. 93. Paris. Greenberg, A. (2000). A Chemical History Tour: Picturing Chemistry from Alchemy to Modern Molecular Science. Wiley-Interscience: New York. Hoffman, F. (1729). Dissertationum physico-chymicarne Haloe Magdeburigicae. Karsten, D. L. G. (1808). Mineralogische Tabellen, 48. 92, Berlin. Knibbs, N. (1924). Lime and Magnesia. D. Van Nostrand: New York. Lampadius, M. A. (1800). Sammlung Praktisch-Chemischer Abhandlungen, 3. 241. Dresden. Mellor, J. W. (1960). A Comprehensive Treatise on Inorganic and Theoretical Chemistry, Vol. IV. Wiley: New York. Seaton, M. Y. (1942). Production and Properties of the Commercial Magnesia. Mining Tech., 1496. Valentini, M. B. (1707). Relatio de Magnesia alba, novo, genuino, polychresto et innixio pharmaco purgante Roma nuper advecto. Giessa Hassoram: Italy. Weeks, M. E. (1968). Discovery of the Elements. J. Chem. Ed., 7th ed. 495– 502.

2 FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

2.1

INTRODUCTION

Magnesium is the eighth most abundant element in the solar system and constitutes about 2% of Earth’s crust. It is also the third most abundant element in solution in seawater, with a concentration of about 1300 ppm. The element magnesium is composed of three stable isotopes: 24Mg (78.6%), 25Mg (10.1%), and 26Mg (11.3%), with an average atomic weight of 24.31. More than 60 magnesium-containing minerals are known. The most important rock-forming minerals containing magnesium are the chlorites, the pyroxene and amphibole group minerals, dolomite, and magnesium calcite (calcite with some of the Ca replaced by Mg). Magnesium is also present in magnesite (MgCO3), and the hydrated carbonates such as nesquehonite (MgCO3.3H2O) and lansfordite (MgCO3.5H2O) as well in brucite [Mg(OH)2]. In addition there is a series of basic magnesium carbonates having the empirical formula xMgCO3.yMg(OH)2.zH2O. These include hydromagnesite [4MgCO3.Mg(OH)2.4H2O] and artinite [MgCO3. Mg(OH)2.3H2O]. Magnesium also occurs in salt deposits such as carnallite (KMgCl3.6H2O), epsomite (MgSO4.7H2O), and kieserite (MgSO4.H2O). It has been estimated that the total mass of the sedimentary layer in Earth’s crust is 1.4  1018 tonnes (Wedephol, 1968). Magnesium is present in the sediment mainly in dolomite and phyllosilicates, such as chlorite and The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 5

6

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

glauconite. Carbonate rocks constitute about 8 wt % of the total sediments (Goldschmidt, 1954). Dolomite makes up about 30 wt % of the carbonates in the sedimentary layer and contains 4.6  1015 tonnes Mg. Magnesite is the magnesium end member of an isomorphous series of carbonates. An increase in calcium content results first in dolomite and then calcite as the Ca end member. Since the difference between the ionic ˚ , Fe2þ ¼ 0.79 A ˚ ), is not as diameters of Mg2þ and Fe2þ (Mg2þ ¼ 0.65 A 2þ ˚ ) and Mg2þ, Fe and Mg substitute great as that between Ca (0.99 A for each other and form the isomorphic series from magnesite through breunnerite (5–30% FeCO3) to siderite (FeCO3), and from dolomite to CaCO3.FeCO3; see Figure 2.1. This isomorphic miscibility results in the presence of both calcite and dolomite in magnesite deposits as mechanical admixtures. However, iron is bound within the magnesite crystal lattice. Pure magnesite is rarely found, and the natural mineral tends to occur part way along an isomorphic series. Four main types of magnesite deposits have been described to date: 1. 2. 3. 4.

Magnesite as Magnesite as Magnesite as Magnesite as

Figure 2.1

a sedimentary rock an alteration of serpentine a vein filling a replacement of limestone and dolomite

Isomorphic series substitution of Fe for Mg to form brunnerite.

2.2 SEDIMENTARY MAGNESITE—BASIS FOR CARBONATE DEPOSITION

7

There are two physical forms of magnesite: cryptocrystalline or amorphous magnesite and crystalline, macrocrystalline, or bone magnesite. 2.2 SEDIMENTARY MAGNESITE—BASIS FOR CARBONATE DEPOSITION Carbonate compounds are relatively insoluble, and Table 2.1 lists the solubility constants for a number of geologically important carbonates. Since these minerals are relatively insoluble, carbonates are precipitated at relatively low carbonate and counterion concentrations. As an example, a solution containing 1024 M Ca2þ will be precipitated by a concentration 24.32 M. This occurs because the product of the of CO23 in excess of 10 two ionic concentrations exceeds the solubility constant for calcium carbonate; see Equation (2.1): ½Ca2þ Š½CO23 Š ¼ Ksp ¼ 10

8:32

(2:1)

Since the solubility of calcium carbonate is considerably less than that of magnesium carbonate, evaporation and concentration of salt lakes and lagoons must have produced calcium carbonate deposits initially. The brine would gradually be depleted of calcium ion and enriched with magnesium. Eventually, a condition would be reached where the brine concentration of the magnesium ion and carbonate exceeds the solubility constant of magnesite, and precipitation would proceed; see Equation (2.2): ½Mg2þ Š½CO23 Š ¼ Ksp ¼ 10

5

(2:2)

Sedimentary deposits of cryptocrystalline magnesite occur either in lagoons, salt lakes, or freshwater lakes (lacustrine). The genesis of magnesite in saltwater requires specific conditions for it to occur: a reducing alkaline environment, a high concentration of magnesium sulfate, and a concentration TABLE 2.1 Carbonates Compound CaCO3 MgCO3.3H2O MgCO3 CaMg(CO3)2

Solubility Constants of Geologically Important (Ksp)

Ref.

1028.32 1025 1024.59 10216.7

Latimer and Hildebrand (1942) Latimer and Hildebrand (1942) Weast and Astle (1982) Stumm and Morgan (1981)

Source: Adapted from H. L. Lutz, Geomicrobiology, Marcel Dekker, New York, 2002, p. 191.

8

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

of CO2 above 380 mg/L, a low Ca content (below 50 mg/L), and the presence of H2S, ammonia, or organic salts, along with an elevated temperature. It is thought that magnesium precipitates as the hydroxide first [Mg(OH)2], which is subsequently altered by reaction with carbonate ion to form MgCO3.xH2O. Dehydration then follows to form MgCO3. Massive magnesite formation in freshwater lacustrine environments proceeds from magnesium derived from either hot solutions emanating from magma or from weathering of ultrabasic rock or serpentinites. 2.2.1 Secondary Nodular Magnesite A variation of the sedimentary process occurs when the host ultrabasic rock is weathered by water, the eroded material being transported and deposited in a distant lacustrine environment. Here, magnesite was precipitated as concretions along with impurities in a mud matrix. The recrystallization of the magnesite would then have occurred through additional uptake of carbon dioxide from the atmosphere. Wave movement on the lake during crystallization carried away noncrystallized impurities such as SiO2, Fe2O3, Al2O3, and CaO from the newly formed magnesite nodules. These impurities are then deposited on the exterior and in the pores of the magnesite nodule. The resultant sand–clay matrix with embedded magnesite nodules is covered with soil and compacted, and over time a silica-rich encrustation forms over the nodule. The percolation of bicarbonate-rich subsurface waters through the deposit results in raising the magnesite density (Schmid, 1987). 2.2.2 Biogenic Carbonate A significant portion of the insoluble carbonate deposits at Earth’s surface is of biogenic origin, while the remainder is the result of magmatic and metamorphic activity. The biological fixation of carbon as carbonate involves bacteria, fungi, and algae, and the carbonate can be deposited both extraand intracellular. A part of magnesium present in the oceans is withdrawn by marine CaCO3-secreting organisms, and this Mg replaces a part of the Ca in the hard parts of the organism. The quantity of Ca replaced depends upon the phylogenic position of the organism. Algae have the highest (5 wt %) and barnacles the lowest Mg concentration (0.9 wt %) (Chave, 1954). Sedimentary deposits of magnesite have been described as having either a biogenic origin or a chemical route via direct precipitation. These deposits are cryptocrystalline in form. Actinomycetes bacteria, which belong to the genus Streptomyces, are thought to have played a role in the formation of

2.2 SEDIMENTARY MAGNESITE—BASIS FOR CARBONATE DEPOSITION

9

hydromagnesite (Can˜averas et al., 1999). These bacteria were found associated with moonmilk deposits containing hydromagnesite in Altamira in northern Spain. Moonmilk is a soft, white, claylike substance present on the walls of many caves. It consists mainly of calcite but can also contain appreciable quantities of magnesium. The magnesium-containing minerals present in moonmilk consist of any of the following: hydromagnesite, magnesite, huntite, nesquehonite, and dolomite. When the minerals constituents of moonmilk are dissolved in a weak acid, there remains a large quantity of organic matter, which consists mainly of such bacteria as Macromonas bipunctata, along with Actinomycetes, and algae. The larger mineral bodies in calcite moonmilk consist of rods with a diagonal grain impressed on the surface of the rods, the combined effect resembling an ear of corn. In many moonmilk samples these rods are enmeshed in a net of calcite filaments about 0.1 mm in width. It is believed that the filaments are associated with microbes that serve as nuclei for growth of the calcite bodies. The hypothesis of microbial mediation in the formation of magnesium carbonate minerals have been reported by Pontoizeau et al. (1987) who have suggested the involvement of sulfate-reducing bacteria in the genesis of huntite and magnesite in Sabkhat el Melah (Tunisia). The relationship between magnesium carbonates and sulfate-reducing bacteria have often been noticed (Hardie, 1987; Lasemi et al., 1989; Middleburg et al., 1990). Sulfate-reducing bacteria (SRB), such as Desulfovibrio species, which are obligate anaerobes, are able to oxidize organic matter by using the oxygen of sulfates as the terminal electron acceptor, see reaction (2.3). SRB produce energy by consuming sulfate and oxidizing the carbon of organic matter. During this metabolic process, bicarbonate (HCO2 3 ) and carbonate ions are produced that may combine with magnesium in high-Mg2þ/Ca2þ brines to form magnesite (Pontoizeau et al., 1987). C106 H291 O124 N16 P1 þ 53SO24 þ 14H2 O

! 53H2 S þ 106HCO3 þ HPO24 þ 16NH4 þ 14OH (2:3)

This reaction produces the accumulation of the bicarbonate ion along with an increase in the alkalinity due to the formation of the hydroxyl ion, which also favors carbonate precipitation. Therefore, the precipitation of high-Mg carbonates in sediments will be favored by: (1) anoxic conditions allowing the growth of SRB, (2) the presence of organic matter, (3) the presence of sulfate, and (4) a high Mg2þ/Ca2þ ratio. These conditions are encountered in continental salt lakes and flats. During the course of sedimentation, calcium ion is depleted from solution by precipitation of the carbonates

10

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.2 Carbonate – bicarbonate speciation as a function of pH.

and sulfates, which lead to a magnesium-enriched residual brine. Dissolved and deposited sulfates are the energy source necessary for the growth of sulfate-reducing bacterial populations. The presence of the cyanobacteria Synechococcus sp. in alkaline, mineralrich lake water has been implicated in the biogenic precipitation of magnesite as well as calcite (Thompson and Ferris, 1990). The photosynthetic metabolism of Synechococcus results in the alkalization of the microenvironment close to the exterior of the cell due to its ability to use HCO2 3 as its primary source of inorganic carbon; see reaction (2.4): HCO3 þ H2 O

! CH2 O þ O2 þ OH

(2:4)

2 The exchange of HCO2 3 /OH across cell membrane results in the accumulation of the hydroxyl ion in the exterior microenvironment, which causes a local pH increase and shifting of the bicarbonate equilibrium toward carbonate; see Figure 2.2. The bacterial cell wall may also act as a nucleation site for the precipitation of magnesite by binding the Mg2þ ion.

2.3 SERPENTINE ALTERATION BY HYDROTHERMAL PROCESSES Deposits of massive cryptocrystalline magnesite occur in serpentized ultrabasic rock that has undergone a hydrothermal leaching of magnesium from the serpentine; see reaction (2.5). The hydrothermal solution

2.5 CRYSTALLINE MAGNESITE

11

contains dissolved carbon dioxide necessary for the dissolution process (Bain, 1924): H4 Mg3 Si2 O9 þ 2H2 O þ 3CO2

! 3MgCO3 þ 4H2 O þ 2SiO2

(2:5)

(Serpentine) The magnesium carbonate is deposited in veins and the silica carried away in solution. Magnesite may also form through the serpentization of ultrabasic rock, under low-temperature and low-pressure conditions, which favor the exosolution of hydrothermal cryptocrystalline magnesite. Reaction (2.6) shows serpentization of olivine, (Mg, Fe)2SiO4, to form antigorite, H4(Mg, Fe)2Si2O9, and magnesite: 2(Mg, Fe)2 SiO4 þ 2H2 O þ CO2

! H4 (Mg, Fe)2 Si2 O9 þ MgCO3 (2:6)

The magnesite occurs as massive bodies, lenticular (lens-shaped) masses, and veins of varying thickness.

2.4 CRYPTOCRYSTALLINE MAGNESITE FORMATION BY INFILTRATION The weathering of serpentine by CO2-laden waters can result in the deposition of cryptocrystalline magnesite in cracks and fissures. Following dissolution of serpentinite, the Al3þ and Fe3þ rapidly precipitates to form layers at the surface of the serpentine. This results in a relative increase of Mg, Ca, and SiO2 in solution, as well as a rise in solution pH. If the solution pH reaches 11, magnesium precipitates either as Mg(OH)2 or MgCO3.xH2O, dependent upon the partial pressure of CO2 present. Calcium is still left in solution since it precipitates at a higher pH than Mg. Subsequent carbonation of brucite and dehydration of the hydrated magnesite results in a porous structure, where pores can occupy up to 25% of the volume.

2.5 CRYSTALLINE MAGNESITE—REPLACEMENT OF LIMESTONE AND DOLOMITE Crystalline magnesite forms through the hydrothermal metasomatic replacement of older dolomite or limestone formations. The mechanism whereby magnesite replaces the extant rock is thought to involve magmatic magnesium-rich waters entering the body though fissures, cleavage plains,

12

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

and capillary openings; see reactions (2.7), (2.8), and (2.9): 2CaCO3 þ MgCl2 CaMg(CO3 )2 þ MgCl2 CaCO3 þ MgCl2

! CaMg(CO3 )2 þ CaCl2

(2:7)

! 2MgCO3 þ CaCl2

(2:8)

! MgCO3 þ CaCl2

(2:9)

The magnesium present in the hydrothermal waters is thought to be derived from underlying dolomites. Contacts are gradual, magnesite passing into dolomite and dolomite into the host limestone rock. An intermediate area of dolomite is always present so that limestone never abuts magnesite. The degree of replacement by magnesite is variable and depends on such factors as host rock porosity and purity. The process may be arrested at any stage, and replacement dolomite deposits are much more common than magnesite. The shape of the deposit is controlled by the original dimensions of the host rock along with the degree of metasomatism. The deposits usually form lenticular bodies and nests, and deposit size may range from hundreds to thousands of meters in length and tens to hundreds of meters in width. 2.6 BRUCITE Brucite forms through dedolomitization by thermal metamorphism; see reaction (2.10): CaMg(CO3 )2

! CaCO3 þ MgO þ CO2

(2:10)

It can also form through thermal decomposition of magnesite to form periclase (MgO), reaction (2.11), which in both cases is subsequently hydrated to form the hydroxide, reaction (2.12): MgCO3

! MgO þ CO2

MgO þ H2 O

! Mg(OH)2

(2:11) (2:12)

Brucite is a rare mineral, and commercial deposits are only found in China, Russia, and the United States. 2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE Magnesite deposits are natural occurrences of concentrations of magnesium carbonate with purity levels of about 90 – 95%, and they typically have

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

13

TABLE 2.2 Worldwide Reserves of Magnesite (Million Tonnes) Country

Reserves

Reserves Base

9 91 14 59 780 27 50 408 589 37 27 59 354

13.6 109 18

United States Australia Austria Brazil 41 China 345 Greece India 13 North Korea Russia Slovakia Spain 9 Turkey Other

27 680 622 289 145 399

Source: Adapted from D. A. Kramer (2002).

reserves in excess of several millions of tonnes of exploitable material. The estimate of known global reserve base is about 13 billion tonnes, and as of 2002 worldwide production is about 10.8 million tonnes (Mt) per year. Table 2.2 gives a breakdown of the major world reserves by country. In addition to those deposits listed in Table 2.2, magnesite is also found in Spain, Pakistan, and the Sudan. Deposits with reported reserves of less than one million tonnes occur in Mexico, Iran, the Philippines, Australia, Egypt, and the Republic of South Africa. Small deposits have also been found in Cuba, Norway, Sweden, Poland, Scotland, France, Italy, Kenya, and Tanzania. The dominant producers of magnesite are North Korea, China, Slovakia, Turkey, Russia, Austria, and India, which together account for 75% of world output. For a more complete listing of other minor occurrences of magnesite and brucite the reader is referred to www.mindat.org. 2.7.1 United States and Canada Nevada Mining of natural magnesite formerly took place in California, New Mexico, Texas, and the state of Washington. Today, however, the sole operational magnesite mine in the United States is located at Gabbs, Nevada; see Figure 2.3. Gabbs is located in Nye County, which is about 208 km southeast of Reno, on the western slopes of the Paradise Range. The mine is operated by Premier Chemicals LLC, and concentrates on the production of caustic calcined magnesia. During World War II, the magnesite deposit at Gabbs was the source of ore for the world’s largest magnesium metal plant. This government-owned, but privately operated, plant, at Henderson, Nevada, drew its source of power from Boulder Dam and

14

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.3

Locations of magnesia production in North America.

produced 73,713 tonnes of magnesium metal from 834,440 tonnes of ore in the period from September 1942 to November 1944. Current reserves at Gabbs are estimated to be about 22,600,000 tonnes of high-grade magnesite with a CaO content of less than 5%. There are much larger tonnages of lower grade material. Brucite reserves are considerably less than that of magnesite and are estimated at approximately 2.7 Mt, which is the largest deposit in the United States. Only five other occurrences of brucite have been listed in the United States (Gildersleeve, 1962). The Gabbs deposit

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

15

is based on a series of magnesite–brucite deposits contained in the upper section of the Late Triassic Luning Formation. The magnesite bodies are scattered over an area of approximately one square mile around a prong of granodiorite that extends north from the stock. The brucite is principally located in two bodies in contact with the granodiorite; at the surface and along the deeper cracks the brucite has weathered to hydromagnesite and artinite. The magnesite and brucite bodies of commercial-grade ore are located in an area of approximately 5.1 km2. The most productive area of magnesite is 1525 m long and 915 m wide and at least 198 m in depth. The distribution of various grades of magnesite is irregular and complex. There appears to be no simple pattern to the ore bodies but rather a random intermixing of dolomite and magnesite interrupted by numerous dikes. The major impurities aside from the dikes include carbonaceous matter, dolomite, talc, tremolite, forsterite, grossularite, garnet, chalcopyrite, calcite, and clay minerals. Interestingly, callaghanite [Cu2Mg2(CO3)(OH)6.2H2O], named after the geologist Eugene Callaghan who made the first detailed study of the deposit, was first described from the Gabbs body (Beck and Burns, 1954) and remains the only occurrence of this mineral. Most of the magnesite is of the crystalline type and occurs as coarse to very fine grained, and individual crystals are either magnesite or dolomite with no apparent solid solutions of one in the other having formed. The magnesite and marmorized dolomite are white to gray when fresh. The gray color is due to the presence of finely divided carbonaceous matter. Magnesite also occurs as dense nodules of hard white crystalline “bone” magnesite at the contact between granodiorite and magnesite or dolomite. Most of the brucite is massive and ranges in color from white through gray or pale-yellow to brown and has the greasy luster of soap. The brucite has weathered and oxidized to a depth of 5 m forming chalky white fibrous hydromagnesite [4MgCO3.Mg(OH)2.4H2O] and small amounts of white acicular artinite [MgCO3.Mg(OH)2.3H2O]. Very small amounts of periclase are scattered throughout the brucite, and more rarely in the magnesite and type 2 dolomite. It is thought that the magnesite was formed by the replacement of dolomite of the Luning Formation by magnesium-rich hydrothermal solutions moving upward through the dolomite. Hydrothermal solutions were heated by granitic magma, and the magnesium required for mineralization was provided by dedolomitization caused by the magma-heated solutions. Brucite was formed by the thermal loss of carbon dioxide from magnesite followed by hydration of the resultant MgO (periclase) to Mg(OH)2. Mining is by the open-pit method, and, because of the chemical irregularity of the deposit, selective mining methods have to be employed. Visual distinction between the various grades of ore is difficult, so identification has to be made by determining the two principal contaminants, lime

16

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

(present as dolomite) and silica (as dyke and contact minerals). This is achieved by chemical analysis of drill hole cuttings. Production at this site is primarily caustic calcined magnesia, with minor quantities of both brucite and magnesite being produced. The annual production capacity is 127,000 tonnes. Texas and Arizona Brucitic marble production from the Marble Canyon deposit is located in the Diablo Mountains, Culberson County, west Texas, and is currently being exploited by Applied Chemical Magnesias Corporation (ACM). This brucite deposit formed through the alteration of the Bone Springs and Hueco dolomitic limestone formations. The brucite ore zone is about 130 m wide and 330 m thick and grades from 60 wt % brucite near the igneous contact to 20 wt % near the unaltered dolomite. The ore is fine-grained and bluish-white to bluish gray marble containing on average 40% brucite, the balance being mainly calcite and about 1% traces of forsterite and grossularite. Proven reserves are put at 2.3 Mt and probable reserves at 23 Mt. Mining is conducted using the room-and-pillar underground method. A deposit of cryptocrystalline, massive, white brucite is also being mined by ACM in Mohave County, Arizona, just north of Oatman. This ore body contains much less brucite than the Gabbs deposit and is of a lower grade. The brucite deposits were studied in 1943 (Campbell, 1944) as part of the U.S. Geological Survey assessment of the magnesium resources of the United States, and again in 1953 when a small amount of brucite-bearing material was extracted. However, the grade of this material proved to be too low in quality to be of any commercial value and was abandoned soon after. Analysis of the deposit shows that the grade of the ore body varies considerably from place to place and contains considerable amounts of calcite and dolomite or magnesite in addition to brucite. The brucite occurs in layers as much as 9 m thick and ranges in color from green to light gray to almost white. Canada Baymag Mines Company, Ltd., is the sole magnesite producer in Canada and exploits the Brussilof deposit near Radium Hot Springs in the Rocky Mountains of southeastern British Columbia. The magnesite is hosted by Middle Cambrian aged dolomite of the Cathedral Formation. The purity of this deposit is such that selective mining eliminates the need for beneficiation. Drilling has outlined the deposit at least 790 m long and 500 m wide and 120 m thick. The two preferred theories discussing the origin of the sparry-magnesite are either replacement of dolomite by magnesite due to interaction with metasomatic fluids or the recrystallization of finegrained sediments of magnesite or hydromagnesite in a marine or lacustine environment. In 1980, proven and probable reserves were estimated to be 9.5 Mt of 95% MgO purity or better in the calcined product, and 13.6 Mt

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

17

of 93 –95.5% MgO in the calcined product. Possible reserves were estimated at 17.6 Mt of calcined product averaging 92.44% MgO (Schultes, 1986). These reserves are sufficiently large to give the mine approximately 150 years working life. Open-pit production at Brussilof is approximately 118,000 tonnes of magnesite per year. The ore is shipped to Exshaw, Alberta, where it is calcined either in rotary or Herreschoff-type kilns producing 54,000 tonnes of caustic-calcined magnesia per year. Other closely located occurrences of magnesite also occur in the Anzac, Driftwood Creek, Jab, Marysville, Topaz, Cleland Lake, and Botts Lake deposits. The overall chemical composition of the deposits has been found to be similar to the Mount Brussilof deposit. However, the Mount Brussilof deposit has higher overall MgO content and lower concentrations of iron, silica, and manganese-bearing impurities. The median lime content of all the deposits is approximately the same. 2.7.2 Brazil Magnesita SA is South America’s leading producer of dead-burn and caustic-calcined magnesia as well as refractory products. Production of magnesite is almost entirely centered on the deposits around Brumado, located in southwestern Bahia State; see Figure 2.4. A second producer, Industrias Brasileiras de Artigos Refractarios (IBAR), is also located at Brumado. Magnesita has the capacity to produce around 2.7 Mt of unburnt magnesite a year as well as 272,000 tonnes/year of dead-burned magnesite, although current production levels are about 231,000 tonnes per annum. Causticcalcined magnesite output is around 41,000 tonnes per year. The major commercial deposit of crystalline magnesite is in the area of Serra das E´guas, where the magnesite occurs as bladelike aggregates and euhedral crystals. Magnesite also occurs as massive aggregates, which comprises the commercial ore. This ore ranges in color from white to a red-orange owing to the presence of a finely divided hematite. The most typical is fine- to medium-grained white or light gray magnesite. The formation of this deposit is thought to have involved magnesium-bearing solutions depositing magnesite as a replacement of dolomite and altering quartz to enstatite and talc. 2.7.3 Australia In 2001 Australia had an estimated 366 Mt of economically demonstrated resources (EDR) (see National Atlas of Mineral Resources website listed in References) defined as that which could be economically extracted at current prices with existing technology). Most production activity is located around the Kunwarara deposit, which is located 37 miles northwest of Rockhampton, Queensland; see Figure 2.5. This low-iron, nodular

18

Figure 2.4

Locations of magnesia production in South America.

19

Figure 2.5 Locations of magnesia production in Australia.

20

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

cryptocrystalline deposit is mined by Queensland; Magnesia a wholly owned subsidiary of Australian Metals Corporation Ltd. (AMC) and represents one of the world’s largest magnesite operations. This deposit has 85.6 Mt of economically demonstrated reserves. In 2001 AMC mined 2.7 Mt of crude magnesite ore at Kunwarara from open-pit mines, which was beneficiated to 594,674 tonnes of magnesite. This in turn produced 117,872 tonnes of dead-burned magnesia, 54,346 tonnes of caustic-calcined magnesia, and 21,781 tonnes of fused magnesia. Due to the high purity of the magnesite nodules, the beneficiation process is fairly straightforward and consists of washing, sorting, and screening processes. The deposit covers an area of some 61 km2, which is overlain by black clay up to 12 m thick with an average thickness of 4 m. Most of the magnesite is in the form of irregularly sized and shaped nodules ranging from 1 mm to 50 cm in diameter and are generally confined to mudstone and weakly indurate sandstone units; see Figure 2.6. Magnesite is usually found to a depth of 20 m but rarely deeper. Two principal types of cryptocrystalline nodular magnesite have been identified; a bone magnesite, which is a white, dense, compact, and extremely hard porcelaineous variety. It is distinguished by its high density and lack of porosity and exhibits a rough gnarled surface and conchoidal fracture; see Figure 2.7. A medium-density magnesite is more commonly found throughout the deposit and is much softer and more porous than the bone magnesite.

Figure 2.6 Cross section of nodular magnesite deposit. (Reproduced by permission of Queensland Magnesia Pty Ltd.)

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

21

Figure 2.7 Photograph of a cryptocrystalline magnesite nodule. Note the distinguishing conchoidal fracture pattern.

There is no visual or chemical distinction between the two different types of magnesite. Examination of the magnesite under an electron microscope reveals individual magnesite crystals in the 1- to 5-mm range, with minor minerals present in the nodules being dolomite, quartz, clays, and iron and manganese oxides. In the neighboring area of Yaamba, which is some 35 km north of Rockhampton, a nodular magnesite deposit with reserves of 170 Mt has been identified. The Oldham deposit 5 km southeast of the Kunwarara deposit has an estimated 112 Mt of ore of which 53 Mt has an average MgO content of 93.7% on the oxide basis. The Triple Four magnesite deposit lies 27 km west of Kunwarara and contains about 32 Mt of magnesite with an average MgO content of 92.3%. The largest deposit of magnesite in Australia has been identified by SAMAG (a wholly owned subsidiary of Pima Mining NL) in the south in the Willouran Ranges, northwest of Leigh Creek, which has a global resource of 579 Mt of magnesite. Approximately 235 Mt of this resource is defined as economically demonstrated resources in the Mount Hutton and Witchelina deposits. Minor EDR occurs in the Winchester deposit, near Batchelor, Northwest Territories, and at Thuddungra, which is located 38 km northeast of Young, New South Wales. The Thuddungra deposit occurs as veins and nodules

22

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

formed from serpentinization by hydrothermal action with the sequence of amphibolites and ultrabasic rock and migration of magnesium bicarbonate groundwater along paleochannels to lake systems where evaporation precipitated the magnesite stratum. The ore is of exceptionally high purity and contains 95–99% MgCO3. The deposit varies in thickness from 2 to 10 m and is being mined by Causmag International. The Thuddungra mine has been in operation since 1935. In 2001– 2002 New South Wales produced about 35,000 tonnes of crude magnesite from the Thuddungra and Lake Cargelligo deposits. Conventional open cut mining methods involving prestrip topsoil and stockpiling, overburden removal, and ore excavation are employed. Beneficiation of the ore comprises crushing, screening, weathering, washing, and sorting. Weathering of the ore is required to break down clay adhering to the magnesite, which disintegrates the matrix and laterite clay (exfoliation), after shrinkage and expansion occurs due to exposure to the elements. After weathering and washing and further screening the ore is subjected to optical sorting using a monochromatic laser light source. Ore that does not meet a preset brightness is rejected by being air blasted off the belt. The magnesite is then calcined in vertical shaft kilns. Other deposits located in New South Wales are found at Fifield, Cobar, Nyngan, Warialda, and Attunga; however, these are smaller and less pure than the Thuddungra deposit. The third largest inventory of EDR magnesite is in Tasmania where the Arthur River deposit has an estimated resource of 29 Mt of fine-grained, massive magnesite formed by the replacement of limestone and dolomite. The deposit grades at 42.8% MgO and is part of a much larger global resource of 180 Mt in the Arthur-Lyons River area, which is located 85 km south of Burnie. 2.7.4 China China is rich in magnesite resources and is second only to North Korea in crude magnesite production. Chinese reserves account for about 30% of the total world reserves with most of the magnesite being concentrated in the Liaoning Province of Northern China. The largest deposits in Liaoning Province are located about 43 miles to the south and southeast of Anshan; see Figure 2.8. Other reserves are located in Hepei, Shandong, Henan, Inner Mongolia Autonomous Region, Sichuan, Shaanxi, Gansu, and Xinjaing provinces, and Tibet. The explored reserves of magnesite are approximately 4 billion tonnes (Hongchuan, 1983; Tainyou et al., 1979) of which about 95% reside in Liaoning Province. Magnesite minerals exposed in the hillsides have been identified in a 124-mile stretch from Dashiqiao into Haicheng and the Xiuyan Manchu region. The magnesite is

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.8

23

Locations of magnesia production in Asia.

medium to coarsely crystalline masses and grades in color from white to pale pink and shades of gray. The main contaminant is a fine-grained, white to pale gray dolomite. There is strong evidence to suggest that the magnesite is sedimentary in nature, formed by the precipitation of both magnesite and dolomite in a Precambrian lagoon or shallow sea. Accurate Chinese production figures are difficult to obtain; however, it is estimated that the total production capacity of magnesia in China is about 2.7 Mt per annum. This comprises of 1.8 Mt of dead-burned magnesia and 900,000 tonnes of caustic-calcined magnesia. The purity of the magnesite ranges from 91 to 98% loss-free MgO content. Two varieties of brucite occur in China: fibrous and massive. Fibrous brucite is only found in Ningqiang, Shaanxi Province, where it is mined

24

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

by Shaanxi Shaannan Asbestos Industry Group Company at a rate of about 2270 tonnes per annum. Massive brucite is found in KuandianFengcheng, of Liaoning Province, where a number of producers exploit the deposits. 2.7.5 North Korea North Korea possesses one of the largest and some of the best quality magnesite deposits in the world. Reserves are estimated at 435 Mt, and in 1998 the country was reported to have produced 800,000 tonnes of magnesite. The magnesite is mined near Tanchon in the Kankyo Province in the east of the country. 2.7.6 Nigeria A large deposit of magnesite is located near the village of Sakasimta, which is about 20 km west of Garkida Town in Gombi. A reserve of 500,000 tonnes has been confirmed and calculated reserves are in excess of 4 Mt. A typical analysis of the magnesite shows the deposit contains 86% magnesite, along with high levels of phosphorus (2.2– 5.0%). The occurrence of the magnesite is thought to be derived from tertiary basalt flows (Industrial Minerals Magazine, September 2004). 2.7.7 South Africa The main occurrence of magnesite in South Africa is in the Mpumalanga and Limpopo provinces, which are in the vicinity north of Soutpansberg (Apiesboomen deposit) and in the Burgersfort and Guyani districts; see Figure 2.9. Total magnesite resources have been estimated to be approximately 18 Mt. Mining is by open-pit methods, the ore being beneficiated by crushing, screening, washing, followed by heavy-media separation. The principal use for magnesite in South Africa is for the production of dead-burned magnesite for refractory manufacture. The Strathmore mine at Malelane, Mpumalanga, and the Syferfontein magnesite mine in the Soutpansberg district, Limpopo, are the only South African magnesite producers operating in 2001. South Africa’s production of magnesite was recorded as being 34,000 tonnes in 2001 (Agnello, 2001/2002). 2.7.8 India Recoverable reserves of magnesite are estimated as 245 Mt, which, at the current rates of usage give a reserve life of 542 years. A major location

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.9

25

Locations of magnesia production in South Africa.

for magnesite production in India is around Salem, Tamil Nadu, which is located about 174 miles southwest of Madras (Chennai). The primary producers in the area include Dalmia Magnesite Corporation, Tamil Nadu Magnesite Ltd., Tata Refractories Ltd., Pon Kumar Magnesite Ltd., Salem Refractories, Ramakrishna Magnesite, and Badrinath Refractory. The magnesite in the area occurs in veins as a white, compact, cryptocrystalline magnesite. Combined capacity of this region is about 18,000 tonnes per year of caustic-calcined magnesite and 154,000 tonnes of dead-burned magnesite. Purity of the product ranges from 75 to 93% MgO. In the Almora area of Uttar Pradesh State, coarsely crystalline magnesite occurs in association with dolomite, talc, cherty limestone, and

26

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

pyrite-bearing shale. Producers in this region are Almora Magnesite and Himalayan Magnesite, which have a combined capacity of about 41,000 tonnes per year of dead-burned magnesite. Magnesite production also occurs at Chandak, Pithoragarh. The mine is operated by Magnesite and Minerals, Ltd., and has a capacity of 41,000 tonnes/year of dead-burned magnesite. The purity of the product ranges from 76 to 89% MgO. The Kharidhunga deposit in the Dolkha District of Nepal has an estimated ore reserve of 180 Mt, of which 66 Mt is refractory-grade magnesite (Wu, 1995). The deposit is a crystalline, breunnerite type, in which the MgO content varies from 88 to 96%, with a maximum of 4.5% SiO2 and Fe2O3, and 1% each for Al2O3 and CaO. The estimated output of dead-burned magnesite in 2001 was approximately 25,000 tonnes. 2.7.9 Saudi Arabia A cryptocrystalline magnesite deposit was discovered in 1962 near Zarghat in the Hijaz Mountains. Although relatively pure, the deposit is small. Another crystalline, high-lime silica magnesite deposit was also discovered in 1962 at Jabal Al Rokhan, which is located about 59 km northeast of Mahd adh Dhahab. Both of these deposits are being evaluated for possible exploitation. 2.7.10 Iran The only Iranian magnesia producer is situated near Birjand in the Khorasan Province; however, another commercial deposit is also located in SistanBlouchestan Province. Iranian Refractories Procurement and Production Company owns the majority of mines in the Khorasan province. The current capacity is 30,000 tonnes per annum of dead-burned magnesite with a purity range of 86–96% MgO and a density range of 3.30–3.38 g/cm3. Total Iranian reserves of magnesite are estimated at 3,542,000 tonnes of mainly low-grade ore (Industrial Minerals Magazine, July 2004). 2.7.11 Greece The main deposits of magnesite in Greece are on the island of Euboea in the Aegean Sea and on the Chalkidiki Peninsula in the north; see Figure 2.10. The ore in these deposits consists of 97 – 99 mol % magnesium carbonate with the remainder made up of calcite, siderite, and rhodochrosite (MnCO3). The deposits are thought to have been formed by the precipitation of magnesite from CO2-rich hydrothermal solutions. Grecian Magnesite is the sole large-scale producer and operates a 400,000 tonnes/year capacity

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.10

27

Locations of magnesia production in Europe.

open-pit mine at Yerakini in the Chalkidiki. As of 2000, the combined output of both caustic-calcined and dead-burned magnesite was about 200,000 tonnes. 2.7.12 Turkey A belt of ultrabasic rock stretching from former Yugoslavia through Albania, Greece, and Turkey and then east to Pakistan is a major source of cryptocrystalline magnesite. Turkey has now exceeded Greece in magnesite production at an annual rate of about 900,000 tonnes. The most important magnesiteproducing region in Turkey is in the provinces of Eskisehir, Bilecik, and Ku¨tahya. Ku¨tahya-based Kumas produced about 180,000 tonnes of dead-burned magnesite in 2003. Magnesit AS (MAS) production sites are located on the Anatolian plateau near the city of Eskisehir. MAS operates three mines in the Tutluca region: Tutluca, Kocbal, and Komurluk. It also have mining operations in Beylikova and Bahtiyar, which are located about 150 km east of Eskisehir. The combined capacity of MAS is about

28

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

140,000 tonnes per annum. Calmag, a subsidiary of Austrian Magnesia producer, Styromag, which is based in Tavsanli, produces magnesite from underground mines and sells about 15,000 tonnes per year of causticcalcined magnesite. In southern Turkey, near the town of Yes¸ilova cryptocrystalline magnesite occurs mixed with mud in dunes and lumps on the shores of Salda Lake. The other major producer is Konya Krom Magnezit, which exploits the deposits at Konya, using open-pit mining, some 240 km miles south of Ankara, and produced approximately 45,000 tonnes of dead-burned magnesia in 2000 (Industrial Minerals Magazine, May 2004). Akdeniz Mineral Kaynaklari (AMK) produced 26,000 tonnes of highpurity raw magnesite in 2003. Its mining operations are located in the region between the cities of Eskisehir, Kutahya, and Bozuyuk. 2.7.13 Serbia and Bosnia Magnesite occurs in Serbia and Bosnia as sedimentary and ultrabasic deposits. A major group of deposits occurs within the inner Dinarides, extending from Skopje, northwards toward Sarajevo and Belgrade. The areas of production include Bosnia, Zlatibor, Kapaonik, and Kosovo. Small magnesite deposits also occur in the Gomsique Ultramafic complex east of Shkodra, Albania. The combined production capacity of this region is about 200,000 tonnes of dead-burned and 40,000 tonnes of caustic-calcined magnesite. 2.7.14 Austria Magnesite was first commercially exploited in 1881 with the opening of the Gros Veitsch Mine. With the development of the Basic Oxygen Furnace in Linz, Austria, in 1952, this major advance in steelmaking required a basic refractory lining using magnesia, which tremendously boosted production at the Veitsch Mine. Production is centered in the southern part of the country and mining occurs in a belt of metamorphosed sedimentary rocks known as the Graywacke Zone. The mines operated by Veitsch-Radex AG are Hochfilzen, Radenthein, and Breitenau and have a combined capacity of about 250,000 tonnes of dead-burned magnesite. The Breitenau Mine is the largest magnesite mine in Austria, the deposit being about 200 m thick and 500 m long. Radenthein is the second largest mine, this deposit being 600 m long and 450 m high and is mined by block caving. Styromagnesit Steirische Magnesitindustrie GmbH (Styromag) is the only other magnesite producer in Austria. The mine at Oberdorf an der Laming, which is about 15 km northwest of Bruck an de Mur, exploits the so-called Wieser deposit. Raw magnesite is extracted from both underground and

2.7 WORLDWIDE OCCURRENCE OF MAGNESITE AND BRUCITE

29

open-pit mines. Styromag has an annual production capacity of about 35,000 tonnes of caustic-calcined magnesite.

2.7.15 Russia The majority of magnesite production in Russia is situated around Satka, Ufa Province, in the southern Urals. The magnesite is contained in 14 bodies within two parallel series of lenses forming a belt some 5 miles long. The belt is composed of late Precambrian to Cambrian dolomite, marl, sandstone, slate, and phyllite some 300–500 m thick. The individual magnesite bodies are up to 2000 m in length. Deposits also exist in the Savin Mountains of eastern Siberia and have reserves reported to be in excess of 1.8 billion tonnes of low-grade magnesite ore. Other areas of magnesite deposits are the Orsk district, which is located about 400 km south of Satka, near Lake Baikal, and the Kuban district of the Caucasus. JSC Kombinat Magnezit, located in Satka, is the largest dead-burned magnesia producer in the Commonwealth of Independent States (CIS). The company has a production capacity of 2.1 Mt per annum of dead-burned magnesia and 90,000 tonnes per annum caustic-calcined magnesia. There are 11 reported brucite-bearing deposits in Russia. The Kuldurskoye deposit is the world’s second largest brucite deposit and the only one currently being exploited in Russia. Proven ore reserves are about 38 Mt with magnesium hydroxide content of up to 65%.

2.7.16 Slovakia Magnesite ore is mined from a number of deposits between the towns of Kosˇice and Luc´enec in the extreme east near the border with Hungary. The deposits are Podrecany, Burda, Lubenik, Amag, and Ddbrava-Mikova´. The deposits are thought to have originated through the hydrothermal replacement of fine-grained dolomite. The magnesite lenses vary between 20 and 400 m in thickness. Ddbrava-Mikova´, which is the largest deposit in Slovakia, is 400 m thick and underlies an area of 324 hectares (800 acres). The ore ranges from 40.5 to 43.5% MgO. Slovenske´ Magnezitove´ Za´vody AS Jelsava (SMZ Jelsava) operates an underground mine in the Jelsava area and has the capacity to produce 330,000 tonnes per annum of deadburned magnesia, the bulk of which is used in the refractory industry. Slovmag AS Lubenik produces exclusively dead-burned magnesia for internal production of basic refractories, producing about 90,000 tonnes per annum of magnesite brick.

30

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

2.7.17 Spain Raw magnesite production was 637,000 tonnes in 2002 from which 140,000 tonnes of caustic-calcined magnesite and 70,000 tonnes of dead-burned magnesite were produced. The major portion of Spanish magnesite is produced in the province of Navarra, in the foothills of the Pyrenees. The magnesite in this region is crystalline and coarse grained. Magnesitas Navarras, S. A., operates a mine at Eugi and a processing plant at Zubini, both near Pamplona. The plant has an annual capacity of 170,000 tonnes of both caustic-calcined and dead-burned magnesite. Magnesitas de Rubian, S. A., exploits the Rubian deposit, which is situated along the northwest coast of Spain in Incio (Lugo). The macrocrystalline magnesite is mined underground using the room-and-pillar method and has a magnesium oxide content ranging from 68 to 85% on a calcined basis. The company has a processing plant at Monte Castelo, which has a production capacity of 125,000 tonnes per annum. Here the magnesite is calcined in a rotary kiln at 700–9008C using a fuel mixture of 80% coke and 20% kerosene (Industrial Minerals Magazine, June 2004).

2.8 PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESITE Table 2.3 displays physical properties of a variety of magnesium carbonates and basic carbonates. Color/Luster In its pure form, magnesite is white often with an opaque cast. Generally though, in its native form it is colored yellow-brown or gray. Crystal System/Habit Hexagonal/rhombohedral crystal system; see Figure 2.11. Crystals are rare. Commonly massive, coarse to fine granular, or compact, earthy or chalky, or lamellar or coarsely fibrous. Specific Gravity 2.98–3.44 g/cm3 for magnesite and somewhat lower for the hydrates and basic carbonates. The X-ray density for magnesite is 3.0095 g/cm3 at 268C. Hardness 3.5–4.5 on the Mohs hardness scale. Crystalline magnesite tends to be softer than the cryptocrystalline variety. Index of Refraction Magnesite is uniaxial negative and when pure has refractive indices, for sodium light of: h0 ¼ 1.700 and he ¼ 1.509. It fluoresces on irradiation with X rays.

31

2.8 PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESITE

TABLE 2.3

Physical Properties of Magnesium Carbonates MgCO3 Magnesite

Property CAS number Mol. wt. Melting point, 8C Density, g/cm3 Index of refraction Hardness Crystal system Solubility, g/100 mL

MgCO3.Mg(OH)2. 3H2O Artinite

MgCO3.3H2O Nesquehonite

4MgCO3 Mg(OH)2.4H2O Hydromagnesite

546-93-0 84.31 – CO2 (900)

12143-96-3 196.68

14457-83-1 138.36 – 3H2O (100)

39409-82-0 467.67

2.98 – 3.44

2.02

1.85

2.236

1.700, 1.510

1.489, 1.534, 1.557 3.5 Monoclinic

1.495, 1.501, 1.526 — Monoclinic

1.527, 1.530, 1.540 — Monoclinic

3.5 – 4.5 Trigonal 0.016—cold water. Soluble in acids and aqueous CO2

—

0.179—cold 0.04—cold water. Soluble water 0.011— in acids and hot water. aqueous Soluble in CO2 acids and aqueous solns. NHþ 4 salts

Source: Data adapted from Lide (1992).

Thermal Conductivity Figures not available. Specific Heat Capacity 75.45 J/mol K (18.05 cal/deg mol) at 258C. The heat capacity equation, which is valid from 298 to 1000 K, is as follows: C8 p ¼ 81:119 þ 5:2254  10 2 T

1:8320  106 T

2

Heat of Formation MgOsolid þ CO2gas

! MgCO3solid

(2:13)

The standard enthalpy of formation of magnesite from the elements at standard temperature and pressure (STP) has been measured as 21113.28 kJ/mol (2266.33 kcal/mol;, see Equation (2.13). Thermal Expansion The coefficients of expansion of magnesite along the principle axis has been measured as 2.13  1025.

32

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.11

Crystal structure of magnesite.

Chemical Properties of Magnesite Magnesium carbonate is practically insoluble in CO2-free water. The presence of dissolved CO2 in water increases magnesium carbonate solubility, however, most anion is present as bicarbonate rather than carbonate; see Equation (2.14). MgCO3 þ CO2 þ H2 O

! Mg(HCO3 )2

(2:14)

The solubility of magnesium carbonate is displayed in Table 2.4 at various temperatures and partial pressures of carbon dioxide. Magnesite solubility increases with an increase of CO2 partial pressure and decreases with increasing temperature. MgCO3 is about 15 –20 times more soluble than CaCO3 on an equivalent basis, whereas dolomite is only about 2 – 3 times more soluble (Mitchell, 1923). The presence of dissolved salts affects the solubility of magnesite with or without the presence of dissolved carbon dioxide. Solubility is increased by chlorides, sulfates, and nitrates.

2.9 CHEMICAL AND PHYSICAL PROPERTIES OF BRUCITE

33

TABLE 2.4 Solubility of Magnesium Carbonate (g/100 g Saturated Solution) under Various Temperatures and Partial Pressures of CO2 Temperature (8C) 0 5 10 30 40 50 60

Partial Pressure CO2 (kPa) 101.3

507

3445

— 3.40 2.85 1.58 1.18 0.95 —

— — 3.57 — 1.37 — —

8.58 8.32 7.93 6.88 6.44 6.18 5.56

Source: Adapted from Seidel (1958).

2.9

CHEMICAL AND PHYSICAL PROPERTIES OF BRUCITE

Brucite, which was named after Archibald Bruce (1777–1818) in 1824, who discovered it in Hoboken, New Jersey, occurs typically as tabular crystals. Less commonly it can occur in acicular, fibrous, and scaly form. Its color can range from white, pale green, gray, gray-blue, and blue. It can also have a transparent, pearly, waxy, or vitreous appearance; see Table 2.5, which displays a variety of physical properties of brucite. Color/Luster White, pale green, gray, gray-blue, blue. Transparent, pearly, waxy, or vitreous. Crystal System/Habit Hexagonal/rhombohedral. Crystals are usually broad tabular. Commonly, foliated massive; fibrous with fibers separable and elastic. TABLE 2.5

Physical and Chemical Properties of Brucite

Property CAS number Mol. Wt. Melting point, 8C Density, g/cm3 Index of refraction Hardness, Mohs Crystal system Solubility

1309-42-8 58.32 – H2O (350) 2.38 1.559, 1.580 2.5 Hexagonal/rhombohedral g/100 mL 9  1024 cold water (188C) 0.004 hot water (1008C) Soluble in acids and aqueous soln. of NHþ 4 salts

34

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

Figure 2.12

Crystal structure of brucite.

Crystal Structure Brucite has a layer structure in which each layer, parallel to f0001g, consists of two sheets of OH in hexagonal close packing, with a sheet of Mg atoms between them; see Figure 2.12. Each Mg atom is in sixfold coordination between OH, and each OH fits into three OH of the next layer. The layers are held together by weak secondary forces between adjacent OH sheets. Density The specific gravity of brucite has been given as 2.38–3.40. The X-ray density has been calculated as 2.38 g/cm3. Hardness The Mohs hardness is 2.5. Index of Refraction Brucite is uniaxial positive and has refractive indices, for sodium light of: h0 ¼ 1.559 and he ¼ 1.580. Brucite is pyroelectric. Thermal Conductivity Figures are not available. Specific Heat Capacity 76.95 J/mol K (18.41 cal/deg mol). Heat of Formation The standard enthalpy of formation from the elements has been measured as 2924.54 kJ/mol (2221.18 kcal/mol). Thermal Expansion Figures are not available.

REFERENCES

35

Chemical Properties of Brucite Magnesium hydroxide’s solubility in water is very small, on the order of 0.009 g/L at 188C. The solubility product has been determined as 5.61  10212. Magnesium hydroxide can be prepared either from the oxide by hydration or through the precipitation from a magnesium salt with sodium hydroxide or lime. In general, magnesium hydroxide behaves in a similar manner to calcium hydroxide, but it is a much weaker base owing to its much lower solubility than calcium hydroxide. The equilibrium pH of a saturated solution of pure magnesium hydroxide is 10.5. Brucite forms the oxide on heating to 3508C and above.

BIBLIOGRAPHY Callaghan, E. (1933). Brucite Deposit, Paradise Range, Nevada — A Preliminary Report. Bulletin of Nevada State Bureau of Mines and Mackay School of Mines, University of Nevada, Reno, Nevada. Cleveland, J. H. (1963). Paragenesis of the Magnesite Deposit at Gabbs, Nye County, Nevada, PhD Thesis, Indiana University. Geoscience Australia (2002). Australia’s Identified Mineral Resources 2002. Geoscience Australia, Canberra. Harben, P. W., and Kuzˇvart, M. (1996). A Global Geology. Industrial Minerals Information Ltd., Metal Bulletin PLC, London. Houssa, C. E. (2000). Magnesia and Magnesium Compounds — A Global Producers and Market Review. Industrial Minerals Information Ltd, London. Kramer, D. A. (2002). Minerals Year Book 2002. Magnesium Compounds, USGS, Reston, Virginia. Lide, D. R. (1992). Handbook of Chemistry and Physics, 72nd ed. CRC Press: Boca Raton, FL. Schilling, J. H. (1966). The Gabbs Magnesite-Brucite Deposit, Nye County, Nevada. In Ore Deposits of the United States, 1933 – 1967, Vol. 2, Part II, pp. 1607 – 1622, Nevada Bureau of Mines, Reno, NV.

REFERENCES Agnello, V. N. (2001/2002). South Africa’s Mineral Industry Review. South African Minerals Bureau, Pretoria, South Africa. Bain G. W. (1924). Types of Magnesite Deposits and Their Origin. Econ. Geol. 19, 412 – 433. Beck, C. W., and Burns, J. H. (1954). Callaghanite, a New Mineral. Am. Min. 39, 630 – 635. Campbell, I. (1944). USGS, unpublished report, Reston, Virginia.

36

FORMATION AND OCCURRENCE OF MAGNESITE AND BRUCITE

Can˜averas, Hoyos, M., Sanchez-Moral, S., Sanz-Rubio, E., Bedoya, J., Soler, V., Groth, I., Schumann, P., Laiz, L., Gonzalez, I., Saiz-Jimenez, C. (1999). Microbial Communities Associated with Hydromagnesite and Needle-Fiber Aragonite Deposits in a Karstic Cave. Geomicrobiol. J. 16, 9 –25. Chave, K. E. (1954). Aspects of the Biochemistry of Mg.1. Calcareous Marine Organisms. J. Geol. 62, 266 –283; Calcareous Sediments and Rocks. J. Geol. 62, 587 – 599 (1954). Gildersleeve, B. (1962). Magnesite and Brucite in the United States, USGS, Mineral Inv. Resource Map MR-27, Reston, Virginia. Goldschmidt, V. M. (1954). Geochemistry. Clarendon: Oxford. Hardie L. A. (1987). Perspectives Dolomitization: A Critical Review of Some Current Views. J. Sed. Petrology 57(1), 166– 183. Hongchuan, Yu. (1983). The Distribution of Magnesite in China. Wuhan Iron & Steel Institute, Wuhan, China. Industrial Minerals Magazine, July 2004, p. 48. Industrial Minerals Magazine, June 2004, p. 37. Industrial Minerals Magazine, May 2004, pp. 31– 37. Industrial Minerals Magazine, September 2004, p. 83. Kramer, D. A. (2001). Magnesium, Its Alloys and Compounds. U.S. Geological Survey Open-File Report 01-341, Reston, Virginia. Lasemi Z., Boardman M. R., and Sandberg P. A. (1989). Cement Origin of Supratidal Dolomite, Andros Island, Bahamas. J. Sed. Petrology, 59(2), 249 –257. Latimer, W. M., and Hildebrand, J. H. (1942). Reference Book of Inorganic Chemistry. McMillan: New York. Lide, D. R. (1992). Handbook of Chemistry and Physics, 72nd ed. CRC Press: Boca Raton, FL. Middleburg J. J., De Lange G. J., and Kreulen R. (1990). Dolomite Formation in Anoxic Sediments of Kau Bay, Indonesia, Geology 18, 399 – 402. Mitchell, A. (1923). J. Chem. Soc. 123, 1887– 1904. National Atlas of Mineral Resources, Mines and Processing Centers (Australia), http://nationalminesatlas.gov.au/info/aimr/magnesite.jsp. Pontoizeau P., Castanier S., and Perthuisot J. P. (1987). International Workshop on Microbial Mediation in Carbonate Diagenesis. Abstract book. Publication ASF: Paris. Schmid, H. (1987). Turkey’s Sada Lake. A Genetic Model for Australia’s Newly Discovered Magnesite Deposits. Ind. Min. 239, 19 –31. Schultes, H. B. (1986). Baymag—High-Purity Magnesium Oxide from Natural Magnesite, CIM Bull. 79(889), 43– 47. Seidel, A. (1958). Solubilities of Inorganic and Metal Organic Compounds, 4th ed. D. Van Nostrand: New York. Stumm, W., and Morgan, J. J. (1981). Aquatic Chemistry. An Introduction Emphasizing Chemical Equilibia in Natural Waters. Wiley: New York.

REFERENCES

37

Tainyou, Xi, et al. (1979). The Production of Refractories. Anshan Iron & Steel Institute, Anshan, China. Thompson, J. B., and Ferris, F. G. (1990). Cyanobacterial Precipitation of Gypsum, Calcite and Magnesite from Natural Alkaline Lake Water. Geology 18, 995– 998. Weast, R. C., and Astle, M. J. (1982). CRC Handbook of Chemistry and Physics, 63rd ed. CRC Press: Boca Raton, FL. Wedephol, K. H. (1968). Composition and Abundance of Common Sedimentary Rocks. In Handbook of Geochemistry, Vol. 1. Springer: Berlin, Chapter 8. Wu, J. C. (1995). The Mineral Industry of Nepal. Ministry of Industry, Department of Mines and Geology, Lainchaur, Nepal.

3 SYNTHETIC MAGNESIA

3.1

INTRODUCTION

Today a significant portion of the world magnesia supply is produced by approximately 20 manufacturers by precipitation of magnesium hydroxide either from seawater or brine sources, which amounts to about 14% of the world magnesia production; see Table 3.1. The first synthetic magnesia production occurred along the French coast of the Mediterranean Sea, near Aiques-Mortes, operating sometime before 1885 (Wicken, 1949). Magnesium hydroxide was precipitated from seawater by reacting it with milk of lime in a continuous fashion using three sequential agitated reactors. The magnesium hydroxide suspension was then filtered on a shallow sand bed some 304 m long and 5 m wide. Once the filter had collected sufficient “mud,” the inflow was stopped and the filter cake dried by solar evaporation. The plant treated about 1000 m3 of seawater a day, from which was obtained approximately 1 kg of magnesia per square meter of filter surface area per day, which amounted to about 6 tonnes a week. Unfortunately, this plant had a short operational life and shut down after a few months, presumably due to competitive pressure from natural magnesite then being exploited in Austria and Greece.

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 39

40

SYNTHETIC MAGNESIA

TABLE 3.1 Location

Synthetic Magnesia Manufacturers Company

Yearly Production Capacity 10,000 tonnes dead-burnt MgO 3,000 tonnes caustic-calcined MgO

France

Manchurian Seawater Works Jiaozhou Guhe Magnesium Salt Factory Scora

Ireland Israel

Premier Periclase Dead Sea Periclase

Jordan

Jordan Magnesia Company Ltd.

South Korea Japan

Sam Hwa Chemical Co. UBE Material Industries Co., Ltd. Shin Nihon Salt Co., Ltd.

China

Naikai Salt Ind. Co., Ltd. Ako Kasei Co., Ltd. Tateho Chemical Ind. Co., Ltd. Konoshima Chemical Co. TMG Corp. Nihon Kaisui Kako United States

Premier Chemicals, LLC Martin Marietta Rohm & Haas SPI-Pharma

Brazil

Buschle & Lepper

Mexico

Pen˜oles, S.A. de C.V.

,10,000 Mt caustic-calcined MgO (dolomitic purification) 90,000 tonnes dead-burnt MgO 10,000 tonnes caustic-calcined MgO 60,000 tonnes dead-burnt MgO 50,000 tonnes dead-burnt and 10,000 tonnes caustic-calcined MgO and Mg(OH)2 50,000 tonnes dead-burnt MgO 250,000 tonnes dead-burnt MgO 50,000 tonnes caustic-calcined 40,000 tonnes 35– 40% Mg(OH)2 slurry 20,000 tonnes Mg(OH)2 slurry 2,000 tonnes Mg(OH)2 powder Mg(OH)2/MgO Fused MgO/Mg(OH)2 12–15,000 tonnes Mg(OH)2 10,000 tonnes Mg(OH)2 8,000 tonnes Mg(OH)2 slurry 24,000 tonnes MgO/Mg(OH)2 powder 50,000 tonnes Mg(OH)2/ caustic-calcined MgO 80,000 tonnes caustic-calcined MgO 10,000 tonnes caustic-calcined MgO Pharmaceutical-grade magnesium hydroxide High-purity seawater-grade magnesium oxide and hydroxide Brine precipitation Approx. 40,000 tonnes caustic, fused, hydroxide, and dead burn

The now defunct Steetley Company built the first large-scale plant to produce a refractory grade of magnesia from seawater at Hartlepool, England, during the late 1930s. A local supply of dolomite was used as the alkali source and the plant had an initial capacity of 10,000 tons per year. Considerable effort was made in developing the seawater process in the United Kingdom during the 1930s because there was no indigenous

3.2 COMPOSITION OF SEAWATER AND BRINES

41

source of refractory-grade magnesia or suitable raw material for magnesium metal production.

3.2

COMPOSITION OF SEAWATER AND BRINES

3.2.1 Seawater Chemistry Magnesium is the third most abundant element in seawater, behind sodium and chorine, and has an average concentration of approximately 1300 ppm. Table 3.2 displays the major and some minor elemental constituents of seawater. Eleven major constituent ions account for 99.5% of the total solutes present in seawater. These 11 are chloride, sulfate, bicarbonate, bromide, fluoride, sodium, magnesium, calcium, potassium, strontium, and boron, and they largely determine the chemistry of seawater. Seawater is slightly alkaline with a pH between 7.8 and 8.3 and is buffered primarily by the carbonate system. The equilibrium reactions between CO2 gas in the atmosphere and seawater are shown in reactions (3.1)–(3.5): CO2 þ H2 O

! H2 CO3

H2 CO3

! HCO3 þ Hþ

HCO3

! CO23 þ Hþ

(slow) (very rapid) (very rapid)

HCO3 þ H2 O

! H2 CO3 þ OH

Ca2þ þ CO23

! CaCO3

TABLE 3.2 Element

(3:1)

(rapid)

(slow)

Elemental Composition of Seawater Abundance (ppm)

Cl Na Mg S Ca K Br C

19,353 10,760 1,294 812 413 387 67.3 28

Sr B Si F

8.0 4.6 3 1.3

Source: McIlhenny and Zeitoun (1969).

Principal Species 2

Cl Naþ Mg2þ, MgSO4 SO22 4 Ca2þ, CaSO4 Kþ Br2 22 HCO2 3 , H2CO3, CO3 Organic compounds Sr2þ, SrSO4 B(OH)3, B(OH)2O2 Si(OH)4, Si(OH)3O2 F2, MgFþ

(3:2) (3:3) (3:4) (3:5)

42

SYNTHETIC MAGNESIA

When equilibrium with the atmosphere is reached, approximately 87% of ionic carbonate is present as bicarbonate ion, the remainder being carbonate. In many places, especially close to the surface, seawater is saturated with respect to calcium carbonate, which will precipitate slowly from solution, thus regulating the amount of carbonate in solution. This process is perhaps the most important of all the geological systems since it regulates the amount of carbon dioxide in the atmosphere. Seawater also contains a wide variety of dissolved organic compounds, the total amount being about 2 ppm. More than 100 different compounds have been identified in solution in seawater. These include a wide variety of organic acids, carbohydrates, amino acids, polypeptides, as well as various vitamins. 3.2.2 Brine Extraction Brine wells and lakes are another usable source of magnesium salt solution for the production of magnesia. The composition of the brine depends upon the chemistry of the surrounding salt-bearing rocks. Recovery of magnesium salts from brine lakes simply involves pumping the solution into the plant. However, the underground recovery is a more complicated operation. Extraction usually involves a leaching process, although in some instances the salts are already in solution. Leaching or solution mining of the salt deposit uses drilling and recovery technology similar to that of the oil and gas industry. A vertical borehole, anywhere from 600 to 1500 m, is first drilled into the deposit from the surface. Concentric tubes or casings are run into the borehole and cemented into place. These casings provide three separate paths for liquid flow to and from the salt deposit; see Figure 3.1. Freshwater is pumped into the borehole via the central tube. The water dissolves the salt from the walls, and the resultant brine is then forced back to the surface through the inner annulus. To ensure that the roof of the cavern remains intact and is not dissolved, a blanket of compressed air or a pad of oil, typically diesel, is injected into the cavern through the outer annulus. This pad oil or air blanket floats on top of the brine and acts to control upward dissolution, thus ensuring the structural integrity of the cavern roof. 3.2.3 Sump Leaching Phase During the initial leaching phase, it is usually necessary to construct a sump cavern to act as a trap for the insoluble components contained in the salt, such as clay and anhydrite. These insoluble materials sink to the cavern floor where they accumulate. The sump is constructed using direct leaching,

3.2 COMPOSITION OF SEAWATER AND BRINES

Figure 3.1

43

Solution mining of subterrain brine deposits.

so as not to cause blockage of the inner leaching tube. The direct leaching method involves pumping freshwater into the deepest part of the borehole through the tubing. Since freshwater is lower in density than brine, the freshwater flows upward along the walls toward the blanket –brine interface, becoming saturated on the way up. The brine is then withdrawn from the cavern via the inner annulus.

3.2.4 Preparation Phase After the sump leaching phase is complete, the system is converted to indirect leaching in order to undercut the deposit. Freshwater is pumped into the cavern via the inner annulus. As it rises to the roof of the cavity, it mixes with the brine. This partly saturated brine then sinks to the bottom of the cavity while becoming more saturated as it dissolves salt from the cavern walls. Brine with the greatest concentration is withdrawn from the bottom of the cavity via the tubing. This leaching process creates a disk-shaped space in the salt deposit, which is further raised vertically during the production phase.

44

SYNTHETIC MAGNESIA

3.2.5 Production Phase The leaching phase is normally carried out by indirect leaching. The leaching tubing is raised in height to allow the continuation of leaching into the body of the deposit, thus the deposit is mined from bottom to top. During this process, the level of the blanket is raised correspondingly after each section has been leached out to the required diameter. After the required quantity of salt has been extracted, the blanket is again withdrawn and a new circular area exposed for leaching. 3.2.6 Evaporite Production Evaporite salt deposits occur when bodies of saline water are concentrated by evaporation. This occurs when natural barriers isolate seawater or other bodies of water containing dissolved salts and the rate of evaporation exceeds the rate of influx of seawater. When continued evaporation reaches the supersaturation point of any salt, that salt is precipitated. The least soluble salts are precipitated first and the most soluble last. Evaporation of seawater, which contains 3.5% by weight of salts, to 53% of its original volume, will begin to precipitate calcite. When evaporation has reached one-fifth of the original volume, gypsum or anhydrite is deposited. Common salt, halite, will precipitate once the volume is reduced to onetenth of the original, followed by magnesium sulfate and chloride. Further evaporation will result in the deposition of sodium bromide and the bittern salts (the chlorides of K, Mg, and Na) and some magnesium sulfate. Table 3.3 details the order of precipitation. In salt deposits, magnesium is present in a number of minerals, of which the most abundant are carnallite

TABLE 3.3 Chemical Species Water NaCl MgCl2 MgSO4 CaSO4 NaBr K2SO4 CaCO3 Fe2O3 MgBr2

Precipitation Sequence of Seawater Evaporitic Salts Percent by Weight 96.24 2.94 0.32 0.25 0.14 0.06 0.01 0.0003

Percent of Total Solids as Salts

Volume Reduction

Salt Precipitated

0 77.76 10.88 4.74 3.60

1.00 0.53

Fe2O3 CaCO3

0.19 0.095

CaSO4.2H2O NaCl MgSO4 MgCl2

2.47 0.35

0.039 0.016

NaBr Bittern salts

0.22

3.3 PROCESS DESCRIPTION

45

(KMgCl3.6H2O), kieserite (MgSO4.H2O), kainite (MgSO4.KCl.3H2O), and polyhalite [K2MgCa2(SO4)4.2H2O]. 3.2.7 Subsurface Brines Naturally occurring subsurface brines occur in porous sandstone or other porous rocks and are regarded as connate or buried seawater. Some brines form locally by solution of rock salt beds. The most important subsurface brines of the United States are those in Mississippian and Pennsylvanian beds and also in Michigan, Ohio, New York, and West Virginia. The chemistry of brine varies from source to source depending on the original source of the salts; however, typical Manistee, Michigan brine, has the following composition: CaCl2 MgCl2 NaCl (KCl, LiCl, SrCl) Br B pH Specific gravity

230 g/L 122 g/L 50 g/L 14 g/L 4 g/L 50 ppm 4–5 1.28

The magnesium chloride content of this brine could theoretically yield about 195 kg (430 lb) of magnesium oxide equivalent for every 3785 L (1000 gal) of brine. However, freshwater may be added to the brine at the well head in order to prevent salting out in the well casings.

3.3

PROCESS DESCRIPTION

3.3.1 Precipitation Reaction Figure 3.2 displays a generalized flow diagram for a seawater/brine precipitation process. The recovery of magnesium hydroxide from seawater or brine is a fairly straightforward process. A strong base is added to the magnesium salt solution to raise the pH to the precipitation point of magnesium, which occurs around pH 10.5; see reaction (3.6). The standard practice used is to employ calcium hydroxide or sodium hydroxide if a low-calcium product is desired. The calcium hydroxide is generally derived from calcined limestone (lime) or dolomitic limestone (dolime). The use of dolime holds some advantages over the use of high-calcium lime, in that half the volume of seawater or brine is required per ton of magnesia since half of

46

SYNTHETIC MAGNESIA

Figure 3.2

Generalized flow diagram of a seawater/brine precipitation process.

the magnesia is derived from the dolime; see reactions (3.7) and (3.8). CaO þ H2 O þ MgCl2 ! Mg(OH)2 # þ CaCl2

(3:6)

CaO
MgO þ 2H2 O þ MgCl2 ! 2Mg(OH)2 # þ CaCl2

(3:7)

CaO
MgO þ 2H2 O þ MgSO4 ! 2Mg(OH)2 # þ CaSO4

(3:8)

As seawater contains about 1294 ppm Mg2þ, 3785 L (1000 gal) of seawater would theoretically yield approximately 8 kg (18 lb) of MgO. If lime is used as the precipitating base, then the reaction would require approximately 1.4 kg of lime to produce 1.0 kg MgO. However, when dolime is used, approximately 1.2 kg is required to produce 1.0 kg of MgO. Since the concentration of magnesium ion in brine is generally much higher than that in seawater, it can be seen that a much lower volume of brine is required to produce an equivalent quantity of MgO from seawater. This implies that a plant utilizing brine should require smaller tank capacities. 3.3.2 Influence of Reaction Conditions on Mg(OH)2 Particle Morphology In general, crystal shape can be modified by changing the conditions of crystallization such as the solvent used, supersaturation, temperature, pH,

3.3 PROCESS DESCRIPTION

47

impurity levels, and degree of agitation. The precipitation of magnesium hydroxide from a solution of magnesium chloride, using either ammonium or sodium hydroxide, has been studied at elevated temperature (Phillips et al., 1977a). In the pH range 8.75– 12.5 it was found that the morphology, average diameter, diameter-to-thickness ratio, and surface area all varied with pH. At an elevated pH level of between 10.5 and 12.5, the precipitates were platelike in shape and tended to be small and thin along with a high diameter-to-thickness ratio. Higher pH favors the edgewise growth of the plates in a direction lying in the basal crystallographic plane, rather than growth along the c axis. However, at a lower pH near 8.75, growth was equal both in the basal and c axis resulting in the formation of equiaxed particles. At pH 9.5, hexagonal plate morphology was evident, which indicates that excess OH2 anion or Mg2þ cation affect the growth rate in a particular direction preferentially causing changes in the crystal shape. Phillips et al. (1977b) also studied the precipitation of magnesium hydroxide by the continuous stoichiometric reaction between MgCl2 and Ca(OH)2 in stirred 3.46 M CaCl2 solution. At 608C using an excess of Ca(OH)2 varying between 0 and 4.5 g/L, sheetlike crystals of up 4.0 mm diameter were formed. This formation was attributed to the adsorption of the Ca2þ ion on the basal surface of the growing crystal. It was argued that this would favor edge growth due to the larger calcium ion blocking adjoining kink growth sites. It was also theorized that some Ca2þ ions may become cationically substituted in the surface and in the bulk of the crystal, which would cause an alteration in crystal morphology. An excess of MgCl2 at a pH of about 8.5 (43 –778C) caused a transition from plates to a “stacked-plate” (pyramidal) layered morphology. At 918C and a 4.5-g/L excess of Ca(OH)2, trigonal plates formed that grew preferentially along the c axis forming triangular prisms. This particular growth pattern was attributed to desorption of Ca2þ due to the higher temperature. Regardless of the quantity of excess reagent present, it was found that an increase in temperature lowered pH and significantly altered particle morphology. 3.3.3 Dolime/Lime Requirements The purity of the limestone or dolomite is crucial to the purity of the end product, MgO. The major impurities derived from this source are silica, Fe2O3, Al2O3, and CaO. Low levels of other impurities may also be present in the raw stone, such as MnO and B2O3, and may result in some degree of contamination of the MgO end product. Many of the impurities present in the raw stone are difficult to reduce economically and, therefore, a high-purity limestone or dolomite source is essential to producing a high-purity product.

48

SYNTHETIC MAGNESIA

Apart from the chemical purity of the dolomite, calcination conditions are also critical for the production of a dolime suitable for the subsequent reaction with seawater. If the dolomite is calcined at too low a temperature, the magnesium hydroxide precipitated has poor settling, compaction, and filtering characteristics. In addition, some dolomite is lost in the form of an uncalcined core, since it does not react with seawater or brine. However, if the calcination is conducted at too high a temperature, the reactivity of the dolime is too low for sufficient reaction with seawater, which results in a significant quantity of the material being classified out of the wet system and going to waste. In addition, high calcination temperatures result in the formation of excessive amounts of brownmillerite, dicalcium ferrite, and calcium silicate, which contaminates the magnesium hydroxide. The objective when burning dolomite or limestone for magnesium hydroxide precipitation is to produce a burned oxide with a low-residual CO2 content, preferably less than 0.5% to minimize contamination of the magnesium hydroxide with calcium carbonate. Dolomite and limestone are typically burned in a rotary kiln with a burning zone temperature between 2300 and 28008F. 3.3.4 Seawater Pretreatment Seawater is pretreated before reaching the reactor and generally involves screening and filtration to remove suspended particles such as silts, sand, and marine creatures followed by decarbonation. Decarbonation is achieved by adding concentrated sulfuric acid to the seawater to lower the pH to 4; see reaction (3.9). The seawater is then passed over a wooden (creosote- and tartreated) desorption tower where it is aerated to remove carbon dioxide. Ca(HCO3 )2 þ H2 SO4 ! CaSO4 þ CO2" þ H2 O

(3:9)

It is necessary to remove the CO2 from the system to prevent calcium carbonate from being precipitated along with the magnesium hydroxide, which would result in the magnesia having a high CaO content. Another, but more antiquated method of softening the incoming seawater is to react it with a quantity of lime sufficient to precipitate the soluble bicarbonate as insoluble calcium carbonate without precipitating any magnesium hydroxide, as in reaction (3.10): Ca(HCO3 )2 þ Ca(OH)2 ! 2CaCO3# þ 2H2 O

(3:10)

The disadvantage of this method is that it tends to form supersaturated solutions of calcium carbonate that make removal of the carbonate difficult.

3.3 PROCESS DESCRIPTION

49

A variety of methods are used to speed up crystal formation, such as passing the seawater through a sand filter where the grains of sand act as numerous nucleation sites for the precipitation of calcium carbonate. Another method involves the use of contacting the seawater with suspended seed particles that act in a similar manner to the grains of sand. As calcium carbonate precipitates, the particles grow in size to a point where they can be removed and sent to waste. In addition to softening the incoming seawater, some operations use chlorine gas to kill any living organisms present in the raw seawater. This helps prevent the clogging of pipe work by marine growths. In some instances, it is also desirable to reduce the level of boron in the incoming seawater, especially when the magnesia is to be used in refractory brick applications. Seawater contains boron as boric acid (H2BO3) to the extent of 4.6 ppm as boron (see Table 3.1) and can be reduced by the addition of excess alkali or so-called overliming at the Mg(OH)2 precipitation stage. This method of boron reduction was first disclosed by Robinson et al. in 1943. The presence of an excess of alkali reduces the adsorption of borate anion onto the surface of the magnesium hydroxide precipitate. Presumably, the reaction works by raising the reactor pH to well above the precipitation pH of magnesium ion and close to the zero-point charge (zpc) of the surface of the magnesium hydroxide particle, which occurs around pH 12. With the normally positive surface potential reduced to near zero at the zpc, the affinity of the borate anion for the surface of the hydroxide particle is reduced. Figure 3.3 displays the variation of the

Figure 3.3 Variation of boron contamination in seawater-precipitated magnesium hydroxide as a function of precipitation pH.

50

SYNTHETIC MAGNESIA

Figure 3.4 Lime contamination in precipitated magnesium hydroxide as a function of precipitation pH.

B2O3 contamination level with the pH at which precipitation of Mg(OH)2 takes place. However, overliming also has the effect of increasing the CaO content of the precipitate. This is demonstrated in Figure 3.4, which shows a linear increase in CaO content with increasing reaction pH. Magnesium hydroxide precipitate produced by overliming is more difficult to settle but does exhibit improved compacting and filtration characteristics. It is also possible to remove boron prior to precipitation by treatment with a boron specific ion-exchange resin. A resin with N-methyl glucamine functionality is capable of reducing boron levels from 100 to ,10 ppm in a brine of 12% MgCl2. An example of this type of ion-exchange resin is Amerlite IRA743 manufactured by Rohm and Haas. However, this treatment option is only commercially viable when working with concentrated brine since the amount of resin needed to adequately treat seawater would be prohibitively expensive. The overliming reaction is therefore a delicate balance between B2O3 content and CaO content. Using this method, it is possible to reduce the B2O3 level in dead-burned seawater magnesia to below 0.03%. 3.3.5 Precipitation Process Following pretreatment, the seawater is pumped into an agitated reactor vessel where it is contacted with a strong alkali, typically either lime or dolime. The dolime is withdrawn from storage by means of a weigh-belt feeder so that precise control over the rate of addition can be maintained.

3.3 PROCESS DESCRIPTION

51

The dolime is then crushed and added directly to seawater in the reactor. The dolime may also be slaked first by reacting with sufficient recycle water using a continuous mixer. The slaking operation converts most of the calcium and magnesium oxides to their respective hydroxides, although some of the magnesium oxide may be slow to hydrate since calcination temperatures for dolime are far above the dissociation temperature for magnesium carbonate. This results in reduced activity of the MgO portion of dolime. Rake arms in the reactor move any unreacted dolime and impurities to the center where it is pumped out and then either hydrocloned or screened out. This material can then either be rejected from the system or returned back to the reactor for further contact with seawater. Milling the unreacted fraction, before returning to the reactor, can further enhance its reactivity. The precipitation is generally controlled so as not to remove all Mg2þ from seawater, as doing so would result in the possibility of co-precipitating calcium sulfate. The theoretical solubility of calcium sulfate is reached when about 75% of the magnesium has been precipitated from seawater. Fortunately, supersaturated solutions of calcium sulfate are quite stable and precipitation does not normally represent a problem of contaminating the Mg(OH)2 precipitate and extraction of magnesium from seawater can be safely carried out to 95% removal. To enhance precipitation and compaction of magnesium hydroxide, especially from seawater, a source of seed crystal is also fed into the reactor. This is generally obtained by recycling a portion of the thickener (settling tank) underflow back to the reactor. The total density of Mg(OH)2 in the reactor is varied usually somewhere about 0.091 kg (0.2 lb) MgO per gallon. This density is approximately three to five times greater than would be present from the magnesium hydroxide obtained from seawater alone. As can be seen, the seed recycle ratio back to the reactor in a seawater plant can be quite high. The recycle ratio is controlled to obtain a “grainy” magnesium hydroxide precipitate that settles readily and compacts well. Magnesium hydroxide precipitated in the absence of any seed crystals results in slurry that has poor settling and filtering properties. 3.3.6 Settling and Compaction The overflow from the reactor carries the precipitated magnesium hydroxide into a large settling tank, which typically range from 46 to 98 m in diameter and can also range from a few meters to 12 m in depth. This large area and capacity is necessary to provide sufficient settling area and time for compaction of the magnesium hydroxide precipitate before it is removed from the settling tank and passed through the washing system. A flocculating polymer, based on polyacrylamide, is normally added to the settling tank

52

SYNTHETIC MAGNESIA

in the range of 1–10 ppm to enhance the rate at which the precipitate settles. The tank is equipped with centrally pivoted rake arms that bring the settled solids to the tank center. Spent seawater from the settling tank overflow is returned to the sea. Thickener underflow solids at this stage are about 20– 25%. Before the slurry is passed onto the washing stage, the thickener underflow generally undergoes a degritting step to remove any unreacted dolime. This can take the form of passing the slurry over a screen or through a hydroclone. The grit is then either rejected from the system or it can be reduced in size and reintroduced back into the reactor. Slurry compaction is a critical factor in the magnesia precipitation process. Compaction directly affects filtration solids, which in turn affects plant production rates. Slurry compaction can be influenced by a number of factors in the precipitation process; however, it is primarily dependent upon seed density in the reactor as well as reactant concentrations. 3.3.7 Washing Concentrated magnesium hydroxide solids collected at the tank center are then pumped to countercurrent washers. A large portion of the solids (typically 85%) are cycled back to the reactor tank for seeding the precipitation. Typically two to three countercurrent washing stages are employed to remove dissolved salts such as CaCl2 and NaCl from the settled solids. The washing is performed countercurrently by pumping the slurry uphill through each successive wash stage, while allowing a larger volume of freshwater to flow by gravity downhill through each stage; see Figure 3.5. This method of washing is economical on freshwater usage, as the spent wash water from each successive stage is used again to wash the slurry in the prior step. Wash water rates are approximately 33 L/kg of MgO. Since this wash water is used two to three times, the effective wash rate is 66– 99 L/kg of MgO. Again, rake arm thickening mechanisms are employed in each wash tank. 3.3.8 Filtration After the final washing stage, the underflow from the wash tank is pumped to the filtration system, where large vacuum filters are employed; these may be of the leaf, rotary disk (see Fig. 3.6), or the rotary drum type (see Fig. 3.7). Owing to the physical nature of magnesium hydroxide precipitate, it is difficult to dewater, so a filter cake with solids content greater than about 55% is difficult to obtain. To obtain a filter cake with higher solids content, pressure filters are usually employed, such as a pressure belt filter (see Fig. 3.8), which is capable of producing a filter cake in excess of 70% solids. One advantage

3.3 PROCESS DESCRIPTION

Figure 3.5

53

Three-stage countercurrent wash flow schematic.

of using a rotary drum filter is that washing the filter cake can be carried out on the filter itself. The filter cake has the consistency of thick paste that can present significant material handling problems, and it has been found that screw conveyers are an appropriate method of transporting the filter cake. However, once the filter cake is increased in solids beyond about 60%, the consistency changes to a semidry solid with a dilatant rheology, which even a screw conveyer would have trouble handling. A filter cake with the minimum possible water content is desirable, as this reduces the drying burden in the furnace and thus reduces gas consumption.

Figure 3.6

American rotary disk filter. (Reproduced by permission of GL&V/Dorr-Oliver.)

54

SYNTHETIC MAGNESIA

Figure 3.7

Rotary drum filter. (Reproduced by permission of GL&V/Dorr-Oliver Eimco.)

3.3.9 Brine Precipitation A brine precipitation plant works on the same principles as a seawater magnesia plant. However, there are some differences in operations. As subsurface brine sources have a higher concentration of magnesium salts than

Figure 3.8 Atlantis high-pressure belt filter press. (Reproduced by permission of US Filter, Dewatering Systems.)

3.5 GRINDING

55

seawater, the volume of liquid having to be pumped into the plant is smaller, thus making the tankage capacity requirements smaller. The brine may require heating and/or dilution to prevent premature crystallization. Seed crystal recycle to the reactor may not be required since the concentration of precipitated magnesium hydroxide in the reactor is much higher than in the seawater process and thus provides plenty of nucleation sites. Disposal of spent process water can be problematic in brine operations since it contains high levels of dissolved salts, which may be restricted from disposal in surface water bodies. Reinjecting wastewater into nonproducing brine wells is one method of overcoming this problem. However, a balance has to be made between what volume is pumped out of brine wells and what is reinjected subsurface; this is known as the “up–down ratio,” and it is desirable to keep this as close to 1.0 as possible, otherwise overdilution of the upbrine may occur. The plant effluent is filtered through a sand filter before deep-well injection to remove suspended solids that could eventually plug the well. It is also acidified to a pH compatible with the natural brine pH, so as to not cause any subsurface precipitation and also to remove any solids not captured by the sand filters. The spent brine may also be pumped to a calcium chloride plant.

3.4

CALCINATION

The production of chemical-grade magnesia or light-burned MgO requires careful control of the calcination temperature to achieve the required specific surface area of the finished product. A furnace well suited to this requirement is the multiple-hearths Herreshoff-type. For the production of dead-burned magnesia typically shaft or rotary kilns are employed. See Chapter 5 for furnaces used in MgO production.

3.5

GRINDING

A variety of mills can be employed to reduce the magnesia to the desired sizing. Light-burned magnesia tends to already be in fine state of division just by the nature of the precipitation process. The milling process for light-burned magnesia generally involves air classification to first remove the fines. The oversize material is then sent to the milling system, which usually employs either ball mills or ring roller mills. Hard-burned and dead-burned magnesia require the use of jaw crushers, roll mills, and ball mills to obtain the desired sizing. For a more complete description of mills and milling circuits see Chapter 6.

56

SYNTHETIC MAGNESIA

3.6 PACKAGING Since magnesium oxide is a hygroscopic material, it is liable to react with atmospheric moisture over a period of time and hydrate to form magnesium hydroxide. It is, therefore, important to ensure the finished product is well protected from atmospheric moisture. The finished product is generally packaged in 50-lb polyethylene lined multiwall paper bags, bulk sacks, pressurized tank truck, or railcar. When pneumatically conveying magnesia, the use of dried compressed air is recommended to reduce moisture pickup, which helps maintain the activity of the oxide and does not contribute to increasing the loss on ignition.

3.7 SAMPLING AND TESTING AND IN-PROCESS QUALITY CONTROL Adequate in-process quality control is essential for producing consistent magnesia that meets industry requirements. There are numerous points along the process train, from incoming dolime quality to finished product, that require periodic quality control checks. In-process quality control chemical analysis, as well as that conducted on magnesium oxide, can be performed using either an X-ray fluorescence spectrometer according to ASTM C1271-94a Standard Test Method for X-ray Spectrometric Analysis of Lime and Limestone or using ASTM C25-95b Standard Test Methods for the Chemical Analysis of Limestone, Quicklime, and Hydrated Lime. Details of chemical analysis are outside the scope of this text; however, there are numerous texts on analytical chemistry that should serve as useful references (Jeffery et al., 1989; Dean, 1995). Dry and wet sieve analysis can be conducted according to ASTM C11095b, sec. 5, Standard Test Methods for Physical Testing of Quicklime, Hydrated Lime and Limestone. Sieve analysis of magnesia powders finer than 100 mesh (U.S. Standard Sieve Series) can be performed using ASTM C110-95b, sec. 20. Test methods that are unique to the magnesia industry will be detailed in the Appendix. 3.7.1 Dolime As discussed above, the quality of the dolime has a large impact on the purity of the finished product. Impurities present in the dolime have a tendency to

3.7 SAMPLING AND TESTING AND IN-PROCESS QUALITY CONTROL

57

end up in the finished product. Silica can be removed, to some extent, by screening and hydrocloning grits from the system. The reactivity of the dolime is normally determined by slaking the dolime and measuring the temperature rise at 30 s and 3 min; see ASTM C110-95b Standard Test Method for the Physical Testing of Quicklime, Hydrated Lime and Limestone. The appropriate reactivity is important; underburned dolime results in large losses of dolime from the system since the unburned cores are screened out and rejected. This can also result in an increased presence of carbonates in the finished product. Underburned dolime can also result in a light precipitate that is hard to settle and compact. Overburned dolime results in a less reactive product that does not react as well with seawater. The unreacted dolime captured by the screening process can be ground to a finer sizing and returned to the reactor. This should not result in increased carbonates in the final product, as the residual CO2 in the dolime should be at a very low level. 3.7.2 Seawater The acidified seawater pH is usually monitored using a calibrated inline pH probe. The dissolved CO2 content of acidified seawater should be around 10 mg/L, see Appendix for test method. 3.7.3 Reactor This test method is used for determining the solids content expressed as MgO per gallon in the reactor overflow; see Appendix for test method. 3.7.4 Settling/Thickener The thickener overflow is sampled for excess Mg2þ. This test measures the quantity of magnesium ion (expressed as lb MgO/1000 gal) in the reacted seawater. This test is critical for ensuring that adequate removal of magnesium from seawater is occurring; see Appendix for test method. Thickener underflow is also tested for compaction. This test monitors the ability of the slurry to undergo settling and compaction in the thickener, which affects the filterability of the slurry; see Appendix for test method. The underflow from the thickener is also tested for silica content before and after any degritting operations.

58

SYNTHETIC MAGNESIA

3.7.5 Washing The washing stages are monitored for chloride removal efficiency. The chloride content of both the washer underflow solids and the overflow water is measured for chloride content. The input of freshwater is gauged from the washing efficiency. 3.7.6 Filtration The filtration stage is monitored for percent solids content of the filter cake, which can be determined by a simple loss on drying at 1058C. 3.7.7 Calcining Magnesium oxide exiting from the furnace should be checked for loss on ignition performed at 10008C and specific surface area (SSA). SSA is performed using the BET nitrogen absorption method. The single-point Brunaver, Emmett and Teller (BET) method is detailed in ASTM D456786; however, for a more accurate determination of specific surface area the multipoint method is preferred; see ASTM D3663. There are commercially available instruments that automate this process; such instruments are available from Quantachrome Corp., Micromeritics Instrument Corp., and Horiba Instruments, Inc. Other tests that measure the activity of the magnesia can also be used to control the calcining process, such as the Caustic Magnesia Activity Test (see Appendix for test method); however, SSA is the preferred method since it is a quick and reliable test. 3.7.8 Grinding Specific surface area, loss on ignition, and screen analysis are measured during the milling operation. 3.7.9 Finished Product A complete chemical analysis is performed on the finished product: MgO, CaO, SiO2, Fe2O3, Al2O3, SO3, Cl2, as well as carbonate content. The total carbon content of the magnesia can be measured using a Leco Carbon Analyzer, which is manufactured by the Leco Corporation (see ASTM C25-95b, sec. 22). The carbon content of magnesia is present primarily as carbonate. A direct determination of the CO2 content of magnesia can be made using ASTM C25-95b, sec. 32. The SSA and loss on ignition of the finished product should also be determined.

3.8 AMAN PROCESS

3.8

59

AMAN PROCESS

The Aman process for producing high-purity MgO by steam pyrohydrolysis of hydrous magnesium chlorides was developed by Joseph Aman in the early 1950s. The process involves decomposing hydrous magnesium chloride obtained from a brine source in a furnace into which steam is injected. The overall reaction can be represented by MgCl2
6H2 O

! MgO þ 2HCl þ 5H2 O

Dead Sea Periclase utilizes the Aman process for producing high-purity magnesium oxides and hydroxide from magnesium chloride obtained from the Dead Sea. It currently operates three Aman reactors (spray roasters) that have an operating capacity of about 100,000 tonnes of high-purity magnesium oxide each year, as well as producing 170,000 tonnes of byproduct hydrochloric acid. The magnesium oxide produced in the pyrohydrolysis process contains significant quantities of water-soluble salts that are removed through an extensive washing operation. During the washing stage the oxide is hydrated to the hydroxide, which then has to be calcined back to the oxide again. This process is capable of producing a magnesium oxide with a purity .99.0%. 3.8.1 Pyrohydrolysis of Magnesium Chloride Hexahydrate Laboratory investigations have shown that the dehydration reactions occurring, when magnesium chloride hexahydrate is heated in the temperature range of 958 to 9008C, follows the equations displayed in the Table 3.4.

TABLE 3.4 Pyrohydrolysis Reaction Sequence of Magnesium Chloride Hexahydrate Occurring During the Aman Reaction Step 1a 1b 2a 2b 3a 3b 4a 4b 5 6

Temperature Range (8C)

Reaction

95– 117

MgCl2.6H2O ! MgCl2.4H2O þ 2H2O MgCl2.6H2O ! MgOHCl þ HCl þ 5H2O MgCl2.4H2O ! MgCl2.2H2O þ 2H2O MgCl2
4H2O ! MgOHCl þ HCl þ H2O MgCl2.2H2O ! MgCl2.H2O þ H2O MgCl2.2H2O ! MgOHCl þ HCl þ H2O MgCl2.H2O ! MgCl2 þ H2O MgCl2.H2O ! MgOHCl þ HCl MgOHCl ! MgO þ HCl MgCl2 þ H2O ! MgO þ 2HCl

135 – 182 185 – 240 þ240 þ240 þ240

60

SYNTHETIC MAGNESIA

Steps 1 –3 are thought to occur sequentially at the indicated temperatures, however, above 2408C, where reactions 4–6 predominate, specific decomposition temperatures are not available. The introduction of steam into the process results in both an acceleration of the rate of reaction and also an increase in the extent of reaction. For example, when MgCl2.6H2O is heated to 6008C in an atmosphere composed of air for 20 min, the extent of decomposition is about 37%. Under identical conditions, the introduction of steam produces decomposition in excess of 70%. At 8008C, the steam pyrohydrolysis of magnesium chloride hexahydrate results in 98% MgO.

3.9 GENERAL PROPERTIES OF SYNTHETIC MAGNESIA In general, the purity of synthetic magnesia is superior to that derived from magnesite; and given that there can be large differences in the quality of magnesite from different locations, the following are the typical differences between light-burned synthetic and natural magnesia: .

. . .

.

Synthetics typically contain .97 wt % MgO. Typical impurity levels are as follows: Fe2O3 , 0.2, Al2O3 , 0.3, SiO2 , 0.5, and CaO , 1.0 wt %. Particle and crystallite size are smaller for synthetics. Bulk density is greater for calcined magnesite. Owing to its greater purity, synthetic magnesia is “whiter” than calcined magnesite. Higher surface areas and, hence, greater activity are more readily obtained with synthetic MgO. Specific surface areas greater than 170 m2/g are currently marketed.

BIBLIOGRAPHY Bateman, A. E. (1956). Economic Mineral Deposits. Wiley: New York.

REFERENCES Dean, J. A. (1995). Analytical Chemistry Handbook. McGraw-Hill: New York. Jeffery, G. H., Bassett, J., Mendham, J., and Denney, R. C. (1989). Vogel’s Textbook of Quantitative Chemical Analysis, 5th ed. Wiley: New York.

REFERENCES

61

McIlhenny, W. F., and Zeitoun, M. A. (1969). A Chemical Engineer’s Guide to Seawater. Chem. Eng., Nov 17, p. 251. Phillips, V. A., Kolbe, J. L., and Opperhauser, H. (1977a). J. Crystal Growth 41, 229 – 234. Phillips, V. A., Kolbe, J. L., and Opperhauser, H. (1977b), J. Crystal Growth 41, 235 – 244. Robinson, H. E., Friedrich, R. E., and Spencer, R. S. (1943). U.S. Patent 2,405,055. Wicken, O. M. (1949). Production of Seawater Magnesite, Electric Furnace Steel Proceedings.

4 MINING AND PROCESSING MAGNESITE

4.1

MINING OPERATIONS

The majority of magnesite mining is carried out using the open-pit method, and as such this process can be divided into five operations—overburden removal, drilling, blasting, loading, and hauling to the processing plant. 4.1.1 Overburden Removal Overburden is a layer of intervening material, primarily soil and rock, between the surface and the ore body. The thickness of the overburden can vary considerably from less than one foot to tens of feet. If the overburden thickness is too great, the cost of removal can be considerable and may force the use of subsurface mining methods. 4.1.2 Drilling The purpose of drilling or coring is twofold: to obtain a detailed map of the quality of the underlying ore body by taking samples of the drill hole cuttings and chemically analyzing them and also to provide holes with which to prime with explosives during blasting operations.

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 63

64

MINING AND PROCESSING MAGNESITE

4.1.3 Bench Height There are several empirical rules given in the literature for calculating bench height. The Atlas Powder Company recommends a bench height of twice the burden distance (Atlas Powder Co., 1987). However, in practice, most open-pit operations have a bench height equal to about 1.6 times the burden distance (Hustrulid, 1999). The generally accepted safe bench height is a maximum of 15 m; however, the quality of the ore in selective mining practices also contributes to the bench height, and usually ore quality along with acceptable fragmentation overrides any other rule of thumb.

4.1.4 Hole Diameter Once the bench height has been determined, the diameter of the blast hole is next for consideration. The suggested ratio between bench height and blast hole diameter (Konya and Walter, 1990) is given by Equation (4.1): DE ¼ L=5

(4:1)

where L is the bench height in feet and DE is the blast hole diameter in inches. The Atlas Powder Co. gives an alternative formula [Eq. (4.2)]: D ¼ H=10

(4:2)

where D is the blast hole diameter in inches and H is the bench height in feet.

4.1.5 Burden and Spacing The blast hole pattern used at any particular site will be the result of a certain amount of trial and error. Konya and Walter (1990) suggest using Equation (4.3): B ¼ 3:15DE (SGE =SGR )0:33 where B ¼ burden distance, ft SGE ¼ specific gravity (g/cm3) of the explosives SGR ¼ specific gravity (g/cm3) of the rock DE ¼ diameter of the explosive charge, in.

(4:3)

4.1 MINING OPERATIONS

65

However, the Atlas Powder Co. suggests the following empirical relationship; see Equation (4.4): B ¼ 25

35DE =12

(4:4)

The value for the ratio of burden to hole diameter, for average rock using an explosive such as ammonium nitrate –fuel oil (ANFO), has been suggested as 25 (Hustrulid, 1999). The spacing of the blast holes will be related to burden by an empirical ratio. The Atlas Powder Co. has suggested that the spacing should be between 1.0 and 1.8 times the burden. The pattern and spacing of drill holes depends upon the uniformity of the deposit. A deposit of uniform thickness and composition may allow widely spaced holes of 100–500 ft, whereas the nonuniform deposit may require closely spaced holes of every 3 – 20 ft. Inclined holes (up to 208 to the vertical) generally have the effect of producing a safer and more stable face, thus allowing higher benches to be blasted safely and also reduce the incidence of “toes,” or rock stumps, at the foot of the face, which result in a better rock pile profile for loading; see Figure 4.1. Alternatively, toe holes can be drilled along the base of the bank 908 to the face. 4.1.6 Subdrilling To ensure adequate breakage at the intended toe of the bench, the blast hole should be drilled some distance below the depth of the bench to the correct toe elevation. This process is termed subdrilling and should be between 0.2 and 0.5 times the burden distance.

Figure 4.1

Cross section of a blast hole.

66

MINING AND PROCESSING MAGNESITE

4.1.7 Hole Stemming After the blast hole is charge with explosives, the top of the hole is stemmed (filled) with fine stone, typically drill hole cuttings. This acts to contain the explosive force within the hole and reduce the occurrence of air overpressure (noise level). The stemming length (see Fig. 4.2) is the vertical height of the bench minus the blasting agent height. The relationship between stemming length T (feet) and burden distance (feet) is given by Equation (4.5) (Atlas Powder Co., 1987): T ¼ 0:7B

(4:5)

4.1.8 Blast Hole Pattern Blast hole patterns are laid out either in a staggered or square arrangement; see Figure 4.3. The merits of the two arrangements are concerned with the distribution of the explosive energy within the pattern. Examination of the two blasting patterns reveals that the total area not directly influenced by any particular blast hole is the same for each arrangement. The size of individual areas not in the zone of influence from adjacent holes is smaller when using a staggered pattern. If staggered blast patterns are used, then the assay from each blast hole can be assigned to a square or rectangular area around the hole in the same manner as that used for square blast patterns.

Figure 4.2

Cross section of a charged blast hole.

4.1 MINING OPERATIONS

Figure 4.3 pattern.

67

Comparison of the energy distribution for a square vs. a staggered blasting

4.1.9 Blast Timing The main principle of blast sequencing is to allow adequate dynamic relief of individual blast holes as the pattern is detonated, but at the same time maintaining confinement and controlling air blast and vibration. To control these aspects of the blast, a delay is built into the firing sequence between adjacent holes in the same row and also between adjacent rows. A rule of thumb for the timing of hole-to-hole delays is given by Equation (4.6) (Konya and Walter, 1990): t h ¼ TH S

(4:6)

where th ¼ delay in milliseconds between rows or holes TH ¼ constant S ¼ spacing distance, ft The constant TH (ms/ft) is dependent upon rock type and has a value of 1.2– 1.5 ms/ft for compact limestone while for magnesite rock 2.5 ms/ft is used between holes.

68

MINING AND PROCESSING MAGNESITE

The timing of interrow delays will depend upon the desired shape of the muck pile (blasted rock pile). A rule of thumb for this delay is given by Equation (4.7): tr ¼ TR B

(4:7)

where tr ¼ interrow delay, ms TR ¼ interrow time constant, ms/ft B ¼ burden distance, ft Since the objective of the blast is to break the rock but to minimize forward movement and mixing of the muck pile, it is important to use an appropriate interrow time delay, which is typically 3– 4 ms/ft of burden. 4.1.10 Blasting Agents One of the most common blasting agents used in open-pit mining is a blend of ANFO. The bottom of the hole is typically primed with a 1.0-lb cast booster, which has an explosive velocity of 26,000 fps, using nonelectrictype detonators. The usual round is about 400–700 holes drilled on 10 ft centers and 18 ft deep, and each hole is filled with 45 lb ANFO. The powder factor is kept at between 0.35 and 0.40 lb/ton, depending upon the density of ANFO being used. The best fragmentation would require the ANFO to be loaded to within 7– 8 ft from the collar. Less height in a larger hole would produce larger stone that the crusher could not crush, while more height in a smaller hole would create fly rock and wasted energy. A 3-in. hole drilled to a depth of 18 ft with 8-ft spacing and a burden of 6.93 ft equals 77 ton of rock. A 3.5-in. hole on a 10-ft spacing and a burden of 8.66 ft equals 120 tons/hole. This is calculated keeping the 7 – 8 ft of stemming and a powder factor of 0.36 lb/tons. 4.1.11 Secondary Blasting If the fragmentation of the rock is good on the first blast, then secondary blasting of oversized pieces should not be necessary. Rock size produced during the initial blasting is important since all the ore has to go through some sort of primary crushing system. 4.1.12 Chemical Contour and Muck Maps Chemical contour maps (or bench maps) are made for the purpose of selective mining of the ore. The chemical contour map is formed by correlating the drill

4.2 PROCESSING MAGNESITE

69

Figure 4.4 Chemical contour map.

hole cutting chemical analysis with the uniquely identified drill hole from which the cutting sample came. Nowadays, this information is entered into a computer program that automatically produces a map with the corresponding chemical contour lines; see Figure 4.4. The numbers within the dark lines of the contour map denotes the CaO content of the ore within the dark-line boundaries; while the numbers within the lighter colored lines indicates the acid-insoluble content of the magnesite ore. The muck map is used to determine which areas of the blasted rock is to be mined or put to waste and is the “in-field” map. As the blasted ore does not change its position appreciably, the chemical contour map can be translated directly onto the blasted area. 4.2

PROCESSING MAGNESITE

The processing of magnesite can be divided into the following operations: crushing, sizing, and beneficiation, and most of these operations are

70

MINING AND PROCESSING MAGNESITE

Figure 4.5 Schematic flow diagram of magnesite processing.

common to the production of sized aggregates and ores; see Figure 4.5 for a generalized processing flow diagram for magnesite. This section seeks to give an overview of the methods involved in the sizing and the beneficiation of magnesite; and since most magnesite processing operations are somewhat unique, only generalities will be discussed. 4.2.1 Ore Removal and Primary Crushing Once the ore has been broken by blasting and the chemical contour map overlaid onto the broken rock, the usable ore is marked by staking or by some other visible marker. Unusable ore is removed and transported to a waste site and dumped. The usable ore is then transported to the primary crusher where it is reduced in size, the crushed ore then being screened and any oversized material returned to a secondary crusher. The fines from the screening process are rejected if chemical analysis shows them to be of inferior quality; otherwise they may be returned to the processing circuit. Samples from the primary crushing circuit are taken at regular interval for chemical analysis to ensure the maintenance of purity. The crushed and sized ore is then stockpiled over a belt tunnel feeder system that has numerous access ports that allows loading of material onto the belt from anywhere in multiple stockpiles. This arrangement allows the blending of

4.2 PROCESSING MAGNESITE

71

magnesite from different piles to achieve the desired chemistry in the final product. Ore that requires beneficiation is generally put into a stockpile separate from the ore that is of sufficient purity to use without further purification (direct-haul ore). 4.2.2 Gyratory and Cone Crushers Once the magnesite rock has been blasted, it is generally loaded onto trucks and transported to the primary crushing operation. Primary crushers are heavy-duty machines and handle dry run of mine feed material as large as 1 m. There are two principle types of primary crushers in operation— gyratory crushers and jaw crushers, with the gyratory type being the most common. A gyratory crusher consists of an inverted conical crushing chamber inside of which is suspended a conical mantle; see Figure 4.6. As the

Figure 4.6 Nordberg Superior MK-II primary gyratory crusher. (Reproduced by permission of Metso Minerals.)

72

MINING AND PROCESSING MAGNESITE

conical mantle oscillates, it causes a narrowing of the gap between the mantle and the chamber walls, thereby compressing the ore between the two faces. As the mantle continues to oscillate, the narrowed gap opens up allowing material to fall farther into the crusher, where it is again crushed. This cycle continues until the ore is of the correct size to exit the crusher. These types of crushers have a reduction ratio of 4 : 1 to 8 : 1. Gyratory crushers can also be used in second-stage crushing, during which ore from the primary crusher is further reduced in size. The typical top feed size for secondary crushers is about 15 cm. Other types of crushers employed in the second stage are cone crushers, roll crushers, and impact crushers. Cone crushers are similar to gyratory crushers except that the angle of the inverted cone is steeper than that of the mantle. The mantle also rotates more rapidly, which results in rock breakage by both impact and compression. The preferred type of second-stage unit is the cone crusher because of its high size reduction ratio (4 : 1 to 6 : 1) and low wear rate. 4.2.3 Jaw Crushers Jaw crushers consist of a fixed jaw and one that moves; see Figure 4.7. When the moving jaw is driven toward the fixed jaw, it crushes the ore in contact with both jaws. When the jaw moves backward, then the ore moves downward. The top feed size of the ore is limited by the opening at the top of the unit, while the crushed size is determined by the gap setting at the lower edge of the jaws. The crushing surfaces are lined with manganese steel wear plates. Reduction ratios of 4 : 1 to 6 : 1 are obtainable with this type of crusher. Impact Crushers One type of impact crusher is the hammer mill. This consists of swinging hammers rotating at high speed on either a vertical or horizontal axis inside of a lined crushing chamber. Ore is broken by impact with the hammers and breakage plates that reside on the periphery of the crusher. Reduction ratios as high as 40 : 1 can be obtained; however, because of their very high reduction ratios, they have a tendency to produce more fines than other crushers. 4.2.4 Roll Crushers Roll crushers are comprised of two counterrotating rolls. This actions draws material into a “nip zone” between the rolls, producing high compression forces that fracture the ore. Reduction ratios of only 4 : 1 are obtainable, with a low production of fines.

4.2 PROCESSING MAGNESITE

73

Figure 4.7 Nordberg C Series jaw crusher. (Reproduced by permission of Metso Minerals.)

4.2.5 Size Separation Size separation in mineral processing circuits are used for a number of reasons, such as the beneficiation of the ore by rejecting a certain size fraction that may contain the majority of the impurities or to avoid the transfer of a certain size of material that may be unsuited for a subsequent processing step. Other reasons include removal of the final sized material from the grinding circuit to improve grinding efficiency. Devices that are used in size separation steps may be screens or classifiers. Screens are a discrete pass device that only allows particles of a certain size to pass through screen apertures. Classifiers act on particles suspended in a medium such as air or water and separate based on a physical property difference such as particle size or density. Size separation equipment is used for a specific range of particle sizes; screening is used to make coarse separations, while classifiers are used to produce a finer separation.

74

MINING AND PROCESSING MAGNESITE

4.2.6 Screening Screening is one of the oldest methods of industrial-size separation techniques. There is a multitude of different varieties of screening devices available in the marketplace such as stationary, vibrating, shaking, revolving, and sifter screens. Table 4.1 lists the various modes of operations of these devices along with their typical uses. The screen deck material is generally made from three categories of material: woven wire screen or cloth, perforated screen plate, and profile wire or bar. The shape of the aperture can be square, circular, rectangular, slotted, or even hexagonal. Square mesh screens are often used for coarse sizing applications if accurate sizing is necessary. However,

TABLE 4.1

Industrial Screening Methods (Adapted from Mathews, 1985)

Screen Type Grizzly, stationary

Grizzly, moving Vibrating screen, horizontal Vibrating screen inclined Shaking, oscillating Shaking, reciprocating

High-speed screen Revolving screen

Sifter screen Centrifugal screen

Sieve bend Dutch State Mines (DSM)

Mode of Operation Inclined or level parallel rails, bars, or rods with definite spacing Vibrating/moving disks, rollers spaced bars Horizontal screen; deck motion pulsed up/forward, then backward/down Inclined screen, deck motion circular to stratify bed Large-stroke, slow-speed linear oscillation Horizontal linear motion, 1 – 4 in. stroke, 30– 200 rpm, and deck slightly inclined Speeds of 3000 rpm or cpm; inclined deck Inclined trommel, scrubber, or barrel; cylindrical, rotating wire cloth, and perforated plate; 15– 20 rpm; open at ends Circular, gyratory, or spiral motion at screen plane Rotating movement; cylindrical screen; fines pass through wall, coarse moves to bottom Parallel bars or wires at right angle to flow; inclined up to 508

Typical Uses Scalping of coarse rock

As above Dewatering and close sizing

Crushing circuits, scalping, and high capacity 0.5 in. þ60 mesh, light and free flowing Conveying, size separation, sizing large lumps Fine and ultrafine screening Scrub, wash, rough size; placer mining; low capacity and efficiency

4 – 200 mesh and finer Wet/dry, 20.5 in. to 35 mesh

Wet scalping and dewatering 10 mesh.

4.2 PROCESSING MAGNESITE

75

when used on an incline, the capacity may be reduced. Rectangular apertures of comparable sizing have a higher capacity since the proportion of open area is greater, as well as having a reduced susceptibility to blinding. Perforated screen plates are useful for coarse separations. Moreover, they resist wear, are less susceptible to blinding, have a high degree of accuracy in sizing, and have a long life. Profile wire, rods, or bars are used for coarse screening and dewatering applications. 4.2.7 Pneumatic (Air) Classification Air classification is widely used in dry grinding to control the degree of fineness of the product. Particles that are larger than the cut size are returned to the grinding circuit, while the ultrafine particles are removed in cyclones or in a baghouse. Pneumatic classifiers are capable of size separations in the 0.1- to 1000-mm range. The forces acting on the entering particles are gravity, aerodynamic drag, centrifugal force, and collision forces, and a suitable balance of these forced are employed in sizing. Figure 4.8 displays a diagram of a Sturtevant side-draft air classifier.

Figure 4.8 Sturtevant side draft air classifier. (Reproduced by permission of Sturtevant, Inc.)

76

MINING AND PROCESSING MAGNESITE

4.2.8 Hydroclones Hydroclones use centrifugal forces to classify particles in a fluid that undergo vortex movement inside the device. The feed slurry is either pumped or flows by gravity into a feed pipe and flows at a tangent to a cylindrical feed chamber under pressure. Fine particles leave through the vortex finder and travel to an exit pipe, while the coarse particles travel downward in a spiral path and are discharged at atmospheric pressure though a spigot that connects to an underflow pipe. 4.3 GRAVITY CONCENTRATION Gravity concentration involves the separation of particles of mixed size, shape, and specific gravity from each other in a fluid medium by the force of gravity or centrifugal force and is used to separate minerals from associated gangue material. Gravity concentration is most effective in the range of less than 5 in. to 200 mesh. Below 200 mesh, the separation of particles by specific gravity becomes increasingly difficult until about 15 mm, where it becomes inapplicable. Below 15 mm froth floatation is the most effective method to use. 4.3.1 Float–Sink Separation The current process of heavy media separation (HMS; sometimes called dense media separation) is the most widely used sink–float process today and is very efficient at reducing the acid-insoluble content of magnesite; however, it is not as effective at reducing the lime content. This means that magnesite undergoing HMS can have a much higher insoluble content than that of lime and still achieve a chemical purity equivalent to that of direct-haul ore. The practical limit for the lime content is approximately 6 wt %, while the limit for acid-insoluble material can be as high as 20 wt %. Heavy solutions, heavy liquids, and ferrofluids have all been used as the dense media. However, the most popular are aqueous suspensions of fine particles of magnetic solids such as ferrosilicon (density, 6.8 g/cm3) or magnetite (5.2 g/cm3). The basis of the process is the preparation of a suspension that has a specific gravity somewhere in between the specific gravity of the mineral to be beneficiated and that of the impurities. The aqueous suspension of dense media is typically adjusted to have a specific gravity in the range 3.0– 3.10 g/cm3, close to the density of magnesite itself (3.0 g/cm3). Gangue minerals having a density lower than 3.0–3.10 g/cm3, such as quartz and numerous silicates, will float in the cone separator, while the

4.3 GRAVITY CONCENTRATION

Figure 4.9

77

Schematic flow diagram of the heavy media separation process.

magnesite will sink. Cone separators can process up to 300 tph (tons per hour), whereas drums and trough or drag tank vessels can treat 700– 800 tph. Feed size to a cone separator unit is typically in the range 1/4 to 5 in. The ore fed into the HMS plant is first sized before entering the cone separator, the fines being rejected from the system to ensure that the suspension media is not contaminated with this material. The floats and sinks are removed from the cone and onto separate screen belts and are thoroughly washed using overhead sprays to remove any adhering suspension media. The suspension media is reclaimed from the washings using a wet magnetic drum separator. The media is then fed into a dewatering unit and the ferrosilicon or magnetite is then demagnetized before being recycled back to the media makeup tank. Figure 4.9 displays a schematic of the HMS process. 4.3.2 Froth Floatation Another common method of ore beneficiation is to use the froth floatation process; see Figure 4.10. However, this process is more costly to operate than HMS. The separation of minerals using froth floatation utilizes the

78

MINING AND PROCESSING MAGNESITE

Figure 4.10

Schematic flow diagram of the froth floatation process.

difference in the wettability of the different types of mineral particles present in the feed. Particles that are to be floated must be either naturally hydrophobic or made hydrophobic by the addition of selective surface-active agents, called collectors. These chemicals selectively adsorb onto the surface of the desired mineral, usually with the assistance of other chemicals, called auxiliary reagents. The now hydrophobic particle can attach to air bubbles and be floated, while the hydrophilic particles sink. The auxiliary

4.3 GRAVITY CONCENTRATION

79

reagents function by selective adsorption on the particles or by complexing with chemical species that interfere with the adsorption of the collector molecule on the mineral surface. 4.3.3 Floatation Reagents Collectors The principle role of the collector is to selectively adsorb and impart hydrophobicity to the mineral particles to be floated. These can be cationic in nature, such as amines, or anionic in nature, such as carboxylates, sulfonates, xanthates, or alkyl sulfates. Fatty acids, such as oleic and stearic, are also commonly used. Frothers The purpose of a frothing agent is to produce air bubbles in the floatation cell, which must then remain intact until they are skimmed off to collect the floated mineral particles. If the collector cannot act as a frothing agent on its own, then additional agents are added, such as pine oil, polyglycols, and cresylic acid. Activators If a mineral surface will not absorb a collector molecule, then floatation of that mineral will not occur. In this circumstance, special reagents, called activators, are added that allow the collector to bind to the mineral surface. An activator normally works by adsorbing onto the mineral surface, thereby providing sites for the adsorption of the collector species. Depressants The purpose of depressants is to inhibit or retard the floatation of a desired type of particle. Its mode of action is to either inhibit the collector molecule from adsorbing onto the particle surface or to mask the adsorbed collector species from the bulk solution so that the particle does not display a hydrophobic exterior surface in contact with bulk solution. Examples of depressants include starch, tannin, silicates, aluminum salts, and dextrin. Dispersants The presence of very fine particles, called slimes, in the floatation process can coat the larger particles and consume excessive quantities of reagents because of their high surface area. The addition of a dispersant, such as silicates, phosphates, or carbonates, clears the surface of the larger particles of slimes. pH Modifiers The pH of the floatation pulp must be carefully controlled to maximize recovery and selectivity. Common reagents that are used for pH control are sodium hydroxide, lime, sodium carbonate, and hydrochloric and sulfuric acids.

80

MINING AND PROCESSING MAGNESITE

4.3.4 Floatation Machines There are two basic designs of floatation cells: mechanical and pneumatic. Mechanical cells are the most common design and generally use a stator with an overhead-driven rotor to circulate the slurry medium (pulp) and also to provide air dispersion and bubble shear; see Figure 4.11. Aeration may be self-induced or introduced from an external air source, and the froth is removed using peripheral launders. Pneumatic cells make use of aeration and hydrodynamics to circulate and suspend the pulp as well as achieving froth separation. Further recovery of magnesite from HMS floats can be achieved by the use of froth floatation. Since floatation beneficiation requires fine particles, the HMS floats are milled to a 100-mesh powder and first fed into the insolubles cell, where a major reduction of silicates occurs. The concentrates from the insolubles cells are then fed into the rougher cells where large reduction in dolomite, calcite, and chlorite occurs. From the rougher cells, the concentrates enter into the cleaner cells where dolomite is again reduced along with plagioclase. The recleaner cells remove dolomite from

Figure 4.11 WEMCO SmartCell Froth flotation machine. (Reproduced by permission of GL & V/Dorr-Oliver Eimco.)

REFERENCES

81

the cleaner concentrates and the scavenger cells again reduce the levels of dolomite. The concentrates from the recleaner cells are the final concentrate and represent the purified magnesite. The tailings from insolubles, rougher, and cleaner cells are put to waste. The following combination of reagents has been found to be effective in the froth floatation beneficiation of magnesite: Sodium hexametaphosphate Aluminum sulfate Sodium silicate Sulfuric acid Oleic acid Dow 9N9 surfactant

4.4

TERTIARY CRUSHING

Before magnesite is calcined, it undergoes a third crushing step (tertiary), where it is generally reduced in size to less than 1 in. using short head cone crushers. 4.5

POSTCALCINATION SCREENING AND GRINDING

After calcining the magnesite, the resultant magnesium oxide may undergo a combination of screening and grinding treatments. The exact treatment would depend upon how the magnesite decrepitates in the kiln; macrocrystalline magnesite tends to decrepitate more than the cryptocrystalline variety. The purpose of the screening process would be to either produce a coarsesized product or to perhaps improve the product purity by rejecting certain size fractions of material that may contain a higher level of impurities. REFERENCES Atlas Powder Co. (1987). Explosives and Rock Blasting. Maple Press: p. 647, Lebanon, PA. Hustrulid, W. (1999). Blasting Principles for Open Pit Mining, Vol. 1. A. A. Balkema: Rotterdam, Netherlands, p. 382. Konya, C. J., and Walter, E. J. (1990). Surface Blast Design. Prentice-Hall: Englewood Ciffs, NJ, p. 303. Mathews, C. W. (1985). General Classes of Screens, SME Mineral Processing Handbook, N. L. Weiss, Ed. AIME: New York.

5 CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

5.1

CALCINATION OF MAGNESITE

The calcination of magnesium carbonate has not been as extensively studied as limestone. However, much of the theory of limestone calcination can be applied to magnesite. The essential reaction that occurs during heating is the loss of carbon dioxide from magnesite, with the corresponding formation of magnesium oxide; see reaction (5.1): MgCO3 þ Heat ! MgO þ CO2 "

(5:1)

The full decomposition of 1.0 kg of pure magnesite yields 0.48 kg magnesium oxide and 0.52 kg of carbon dioxide. It will, therefore, require 2.08 kg of pure magnesite to yield 1.0 kg of magnesium oxide. There are three essential factors in the kinetics of magnesite decomposition. 1. The magnesite rock has to be heated to the decomposition temperature. 2. This minimum temperature must be maintained for a certain duration, that is, until the magnesite has been essentially decomposed to magnesium oxide.

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 83

84

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

3. The carbon dioxide gas evolved during the decomposition has to be removed from the kiln to increase the rate of decomposition. If the temperature and pressure are in equilibrium, then dissociation is static. However, if there is a small change in either of these variables, such as a decrease in CO2 pressure or concentration, or an increase in temperature, dissociation proceeds immediately with the evolution of CO2 gas and the simultaneous formation of magnesium oxide. The temperature at which the dissociation pressure of MgCO3 reaches one atmosphere has been reported to be between 402 and 7508C (Boynton, 1980; Knibbs, 1935). These discrepancies in the dissociation temperature are probably due to the differing sources of the magnesium carbonate used in these determinations, since it is known that differences in the impurity level, crystallographic structure, and microstructure of the magnesium carbonate can have a marked effect on its calcination properties. The standard heat of dissociation of MgCO3 to the oxide has been reported to be 3027 kJ/kg (723 kcal/kg) of MgO formed (Knibbs, 1924). The relationship between the dissociation pressure of a natural magnesite and temperature is shown in Figure 5.1. The pressure of carbon dioxide in the atmosphere above the magnesite reaches one atmosphere in pressure (760 mmHg) at a temperature of about 7508C. This dissociation temperature is considerably higher than other determinations, so one must bear in mind that this measurement was made on a natural magnesite and the figure reflects the more practical considerations of its production.

Figure 5.1 Relationship between dissociation pressure of natural magnesite and temperature. [Data adapted from J. Chem. Soc. 123, 1055 (1923).]

5.1 CALCINATION OF MAGNESITE

85

5.1.1 Energy Requirement for Calcination Process The theoretical energy requirement to achieve the above process can be calculated as follows. First, the magnesite rock has to be raised from ambient temperature to the temperature of decomposition. Second, once at the temperature of decomposition, there is another energy component, the enthalpy of decomposition, required to decompose the magnesite to magnesium oxide and carbon dioxide. The energy required for the first step can be calculated from the specific heat capacity (Cp) of magnesite using the formula: Cp  increase in temperature (K); we have, in going from 298 K (258C) to 1023 K (7508C) 1:56 kJ  725 K ¼ 1131 kJ/kg

(5:2)

The enthalpy of decomposition of magnesite at 1000 K is 1284 kJ/kg, which is the energy required for the second step. This gives a total energy requirement of 2415 kJ (2289 Btu) to decompose 1.0 kg of magnesite, or 2.42  106 kJ per metric tonne (2.3 MBtu), or 2.1 MBtu/short ton. The kinetics of the calcination process can be described by a “shrinking shell” model, where the process of decomposition proceeds gradually from the outside surface inward to the center. 5.1.2 Effect of Time and Temperature If the minimum dissociation temperature was precisely maintained in the kiln, then the rate of calcination would be extremely slow, in fact too slow to be of practical benefit. In practice, to accelerate the dissociation of magnesite, temperatures significantly above the minimum dissociation temperature are maintained until all of the CO2 has been expelled. Increasing the calcination temperature has a much greater influence over the dissociation rate than does residence time in the kiln. However, a balance has to be maintained between a high production rate, which can be achieved through use of higher calcining temperatures and product quality; temperatures that are too high can result in an “overburned” and hence a lower reactivity product than desired. 5.1.3 Kinetics of Calcination The progress of the calcination of magnesite can be divided into five stages. Stage 1, Preheating Zone The magnesite is preheated from ambient temperature to between 700 and 9008C by the kiln gases. If firing in

86

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

a 14-hearth multiple-hearth furnace (MHF), the top four hearths are typically not fired and preheating occurs from the combustion gases from the lower burners. Stage 2, Calcination Zone When the magnesite reaches a temperature of about 7508C, the pressure of carbon dioxide produced by dissociation of magnesite equals the partial pressure of CO2 in the combustion atmosphere. As the magnesite progresses through the calcination zone, the temperature further rises and the surface layer of the ore begins to decompose. Stage 3, Once the temperature of the magnesite has exceeded the decomposition temperature, the partial pressure exceeds one atmosphere and the process of dissociation can proceed beyond the surface of the particles. Stage 4, Sintering If all of the magnesite has decomposed to magnesium oxide before it leaves the calcination zone, then the process of sintering begins. For further discussion of sintering see Section 9.2.1. Stage 5, Cooling The calcined magnesite leaves the calcination zone and starts cooling. A typical temperature profile for calcining magnesite to produce causticcalcined magnesia in a MHF is shown in Figure 5.2.

Figure 5.2 Temperature profile of a multiple-hearth furnace calcining magnesite. Numbers inside the bars indicate the number of burners being fired on each hearth.

5.1 CALCINATION OF MAGNESITE

87

Figure 5.3 Stepwise processes involved in thermal dissociation of magnesite.

Dissociation of Magnesite Figure 5.3 illustrates the processes occurring during the dissociation of magnesite above its decomposition temperature. Step 1 Heat is transferred from the combustion gases to the surface of the magnesite particle. Step 2 Heat is then conducted away from the surface to the reaction interface. The heat channels through micropores present in the layer between the particle surface and the reaction interface. Step 3 The heat arriving at the interface causes the dissociation of MgCO3 into MgO and CO2. Step 4 Carbon dioxide created during the decomposition migrates from the interface to the surface of the particle through the porous calcined layer. During its migration to the surface, the CO2 is gradually heated to the same temperature as the particle surface. Step 5 Carbon dioxide diffuses away from the surface into the kiln gases. Steps 2 and 4 are strongly influenced by the microstructure of the MgO layer surrounding the reaction interface, and the diffusion of carbon dioxide to the particle surface appears to be the rate-determining step. The porosity of this layer is determined by the crystal size of the magnesite; macrocrystalline magnesite decrepitates more readily than does the cryptocrystalline type. Heating causes an expansion of the crystal matrix, which stresses the individual crystals and causes them to fracture. Finer-grain crystals are more resistant to the thermal stress caused by crystal expansion and, therefore, have less resultant crystal breakage and hence less decrepitation of the magnesite ore. Intercrystalline fissures also help resist crystal

88

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

breakage during thermal expansion. These minute fissures act as expansion joints and tend to relieve expansion pressures. 5.1.4 Stone Size As displayed in Figure 5.3, the dissociation of magnesite progressively occurs from the outer surface to the interior of the stone. Therefore, the larger the stone size the more difficult it is to achieve a uniform degree of burn. To expel carbon dioxide from large magnesite pieces, high temperatures have to be employed to generate sufficient CO2 pressure in the interior of the crystal lattice for the gas to escape. This has a tendency to overburn the surface layer causing excessive shrinkage and also results in the narrowing and closure of pores and fissures. This phenomenon can completely cutoff the escape route for CO2 and result in a product that has a high proportion of magnesium carbonate still present at the core of the stone. The optimum size of stone should be less than ½ in. and be of uniform size and shape, which, of course, is practically impossible to achieve. To recap, the major variables influencing the rate of calcination of magnesite are: . . . . . . .

Stone feed size Stone shape Microstructure of the magnesite Particle size Particle shape Temperature profile of the calcining zone Rate of exchange of heat between the kiln gases and the particles of magnesite

5.2 CALCINATION OF MAGNESIUM HYDROXIDE Calcining magnesium hydroxide, such as that produced in either a brine or seawater process, involves heating a filter cake that contains between 50 and 72% magnesium hydroxide solids, the balance being water. The thermal decomposition involves the following reaction: Mg(OH)2 þ Heat ! MgO þ H2 O

(5:3)

The decomposition reaction begins to take place around 3508C and increases rapidly above this temperature. One kilogram of pure magnesium hydroxide

5.2 CALCINATION OF MAGNESIUM HYDROXIDE

89

when fully decomposed will yield 0.69 kg magnesium oxide and 0.31 kg water vapor. Filter cake entering a kiln undergoes a number of processes, in not an entirely sequential manner: . . .

Dehydration of the filter cake Decomposition of dry magnesium hydroxide into MgO and water vapor Sintering of magnesium oxide

In reality, all three processes are occurring to some degree at the same time. It is very difficult to remove the last vestiges of chemically bound water from magnesium hydroxide, unless the kiln temperature is raised above 10008C. It is believed that this residual water is adsorbed onto the nascent magnesium oxide surfaces in a monolayer (Gregg and Packer, 1955).

5.2.1 Energy Requirement for Calcination Process The processes involved in calcining magnesium hydroxide filter cake are the removal of free water from the filter cake, followed by the decomposition of magnesium hydroxide solids. For the purposes of this calculation, the calcination process can be separated into two different sequences, that involving magnesium hydroxide solids and that involving free water. Assuming that we are interested in calcining 1.0 kg of 55% magnesium hydroxide filter cake, the balance being free water, the energy requirement for the hydroxide solids can be calculated as follows. The energy required to raise the temperature of the hydroxide solids from ambient temperature (298 K) to the temperature of decomposition, 3508C (623 K) can be calculated using the specific heat capacity of magnesium hydroxide at 600 K (1.78 kJ/kg K) as follows: 1:78 kJ=kg K  0:55 kg  325 K ¼ 318 kJ

(5:4)

Once the solids are at decomposition temperature, the energy required to effect the decomposition can be calculated from the enthalpy of decomposition, which at 600 K is 1304 kJ/kg: 1304 kJ=kg  0:55 kg ¼ 717 kJ

(5:5)

Therefore the total energy requirement to decompose 0.55 kg magnesium hydroxide to magnesium oxide and water vapor is 717 kJ þ 318 kJ ¼ 1035 kJ (981 Btu).

90

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

To effect the removal of free water from the filter cake the following processes have to take place; first water has to be raised from ambient to boiling point (373 K); then the liquid water has to be vaporized (enthalpy of vaporization); the resultant steam then has to be heated from 373 to 623 K. The energy required to heat 0.45 kg of water from 298 to 373 K can be calculated from the specific heat capacity of water (4.18 kJ/kg K): 4:18 kJ=kg K  0:45 kg  75 K ¼ 141 kJ

(5:6)

The enthalpy of vaporization of water is 2283 kJ/kg; therefore, the energy required is 2283 kJ/kg  0.45 kg ¼ 1027 kJ. The energy required to heat the resultant steam from 373 to 623 K requires the use of the heat capacity of water vapor (1.86 kJ/kg K): 1:86 kJ=kg K  0:45 kg  250 K ¼ 209 kJ

(5:7)

The energy required to remove free water from the filter cake is therefore 141 þ 1027 þ 209 kJ ¼ 1377 kJ (1305 Btu). The total energy requirement to calcine 1.0 kg of 55% solids filter cake, which will yield 0.38 kg magnesium oxide is the sum of 1035 þ 1377 kJ ¼ 2412 kJ (2287 Btu). The energy needed to be expended to produce one metric tonne of MgO will therefore be 6.35 MBtu or 5.77 MBtu for one short ton. The decomposition of magnesium hydroxide is one of the most frequently studied reactions since it represents one of the simplest examples of this type. Both reactant and product posses only one known crystal structure, and the conversion of the hexagonal hydroxide lattice to the cubic oxide is a fairly simple transformation. The decomposition occurs with a high degree of topotaxy (maintenance of crystallographic orientational relationship between precursor and product), the hexagonal close-packed structure of O22 ion rearranging to a cubic close-packed structure; see Figure 8.1 (Ball and Taylor, 1961; Bernal, 1938; Brett, 1969; Moodie et al., 1966). Thus, the remnants of the brucite hexagonal platelet structure is retained in the Mg(OH)2 to MgO conversion.

5.2.2 Decomposition Mechanism There have been two different models proposed for the decomposition of magnesium hydroxide. The Goodman “homogeneous” model assumes that water is lost in the same manner throughout the crystal. This occurs by the hydroxyl groups from adjacent layers combining to form water, which

5.2 CALCINATION OF MAGNESIUM HYDROXIDE

91

then diffuses away (Goodman, 1958). The reaction commences at an outer surface and advances toward the crystal center. A second model, known as the “inhomogeneous” model, has been proposed by Ball and Taylor (1961) and Brindley (1963). The reaction is thought to occur simultaneously throughout the crystal by the development of “donor” and “acceptor” regions during the decomposition. In this model magnesium ions are thought to migrate from the donor to the acceptor regions, while the OH2 ions combine with protons to form water, which then diffuses away, leaving behind in its process intercrystalline pores. The experimental evidence purported to support the homogeneous theory are the observed topotaxy, the production of a porous MgO and the formation of “spinel-like” or defect structures detected in X-ray powder and single-crystal photographs. However, it was later shown that the spinel was derived from impurities present in the brucite, since pure samples of magnesium hydroxide did not show the presence of spinel (Brett and Anderson, 1966). The experimental evidence for the inhomogeneous theory was thus invalidated. The decomposition of magnesium hydroxide has been evaluated as a process of nucleation and growth of MgO crystals within the brucite matrix. The formation of a defect layer of hydroxide structure, which suddenly recrystallizes to the cubic MgO structure when the fracture stress is exceeded in the defect layer, has been postulated (Freund et al., 1975; Guilliat et al., 1970; Garn et al., 1978; Lønvik, 1978). The MgO crystals formed have an expanded cubic lattice that upon increasing calcination temperature gradually decreases in size until the equilibrium unit cell dimension is reached. The sudden recrystalliztion leads to cracking of the precursor/product crystal and the formation of MgO crystals with a size in the range of 5 – 10 nm, when the facture stress is exceeded by the contraction strains (Anderson and Horlock, 1962). Cryptocrystalline precursors, when the crystal size is less than 60 nm, experience less strain, and the decomposition may proceed to completion without cracking. Most commercially produced precipitated magnesium hydroxides have crystal sizes greater than 60 nm so cracking is expected to occur during decomposition. Infrared studies of the decomposition have observed the broadening of the infrared spectrum of the O–H band at just slightly below 3008C (Garn and Freund, 1975). The broadening has been interpreted as the start of the OH – OH interaction between adjacent groups at the crystal surface. A proton tunneling mechanism between the adjacent hydroxyl groups results in the formation of water that then diffuses away (Na¨gerl and Freund, 1970). When the energy levels of the adjacent hydroxyl groups overlap at low temperatures, tunneling can take place. However, at higher temperatures when the groups are further apart, proton jumping leads to the removal of

92

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

hydroxyl groups with the loss of hydrogen (Martens et al., 1976; Freund et al., 1977). The interpretation of the dehydroxylation process is consistent with the advancing interface concept rather than the inhomogeneous mechanism since the tunneling of protons explains the reaction of hydroxyl groups in adjacent layers. This mechanism also maintains electroneutrality, so the postulate of Mg2þ countermigration is no longer required. Measurements on hydrogen evolution during decomposition of Mg(OH)2 indicates that approximately 0.05–0.1% of all the hydroxyl groups present in the starting Mg(OH)2 will form hydrogen gas instead of following the normal decomposition path to give water (Freund et al., 1977). Two distinct hydrogen evolution peaks have been observed, one with its maximum at 4508C, the other at 7508C, which is also coincident with the peak of oxygen evolution. However, evolution of hydrogen was only observed after the start of dehydroxylation of the hydroxide. It is postulated that the formation of the defect hydroxide layer and subsequent recrystallization to MgO is necessary for the formation of a high concentration of defect sites on the MgO surface. The cubic MgO formed in the recrystallization includes numerous defects, mainly as residual OH2 groups and cation vacancies. It is postulated that the formation of molecular hydrogen is connected with the formation of these defects. The proposed mechanism for hydrogen formation is as follow: OH þ OH $ 2O þ H2 " (5:8) This reaction is thought to take place within a cation vacancy [Mg]00 in MgO substrate neighbored by two residual OH2 groups. For each hydrogen molecule formed, two O2 ions are produced, which are positive holes with respect to the surrounding O22 sublattice. The diffusion of three O2 ions to form a cluster around a cation vacancy leads to the formation of a V1 on a (111) surface plane. It takes a relatively long time for V1 centers to build up in concentration, typically several hours at 5508C, a process limited by the diffusion of O2 ions. The evolution of hydrogen at the higher temperature, i.e. 7508C, is due to thermal dissociation of residual OH2 groups: 1 OH ! O þ H2 (5:9) 2 This is followed by the evolution of oxygen due to the thermal decomposition of O2 ions between 300 and 7008C: 2O

! O2 þ O †

O 1 þ O 1 ! O2 "

(5:10a) (5:10b)

5.2 CALCINATION OF MAGNESIUM HYDROXIDE

93

5.2.3 Kinetics of Decomposition The dehydroxylation reaction has been extensively studied using differential thermal analysis (DTA) or differential thermogravimetry (DTG). However, there has been little agreement over the activation energy Ea of this process. A review article by Sharp (1973) concluded that many methods developed for kinetic analysis of DTA traces of solid-state reactions produce unreliable values for the order of reaction and Ea. It was also concluded that the kinetics of the reaction were strongly influenced by the ambient water vapor pressure and that, under vacuum, the activation energy is about 85 kJ mol21. Bouvier et al. (1980) carried out thermogravimetric analysis (TGA) of the decomposition of natural brucite and seawater magnesium hydroxide and concluded that the rate constant was dependent on the inverse of the square of the size of the Mg(OH)2 particle. The wide variation of Ea values can be explained by postulating that the ratelimiting step is the diffusion of water away from its site of formation and that pressure dependence is in effect. Decomposition in a vacuum permits the rapid removal of water vapor from the crystals and prevents diffusion outward being the rate-limiting step. If the surrounding gaseous pressure increases, then the rate of diffusion is slowed, due to hindered diffusion of water vapor along a lowered pressure gradient between the reaction site and the exterior of the particle. The presence of water vapor in the system during decomposition of magnesium hydroxide greatly influences both the nucleation and growth kinetics of MgO. The water molecules appear to enhance the initial nucleation of magnesium oxide, resulting in MgO crystals of smaller size than of those produced in a vacuum under the same conditions of time and temperature (Anderson et al., 1965). Further heat treatment of the magnesium hydroxide results in further interaction of water molecules with the product oxide surface according to the postulated reaction mechanism; see reactions (5.11) and (5.12) (Anderson and Morgan, 1964): Mg2þ þ O2 þ H2 O ! Mg2þ þ 2OH þ Vm

(5:11)

Mg2þ þ O2 þ H2 O ! Vm þ V0

(5:12)

where Vm represents a cation vacancy and V0 represents an anion vacancy. This mechanism provides a possible process for O22 ion movement by desorption of water composed of an original oxygen atom present in the crystal lattice as an O22 and hydrogen atom from an adsorbed hydroxyl group. This adsorption/desorption mechanism gives rise to rapid crystal growth and agglomeration, with a corresponding decrease in surface area and reactivity

94

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

of the magnesium oxide product. Riddel (1972) found that the rate of crystal growth was proportional to the partial pressure of water vapor above a threshold level of 610 N m22. The above process has a great influence on the sintering of MgO compacts during the production of dead-burned MgO. The enhancement of the initial and intermediate stages of sintering has been observed in the presence of water vapor when the porosity of the compact is still open and accessible to gaseous diffusion (Eubank, 1951). Most workers have reported a sintering enhancement in the presence of water vapor; however, others have reported retardation (Anderson and Morgan, 1964; Hyde and Duckworth, 1962). The densification of MgO compacts is also dependent on other factors as discussed below and as such are manipulated to control the speed of movement of crystal surfaces (grain boundaries). The relative movement of the pores in relation to the grain boundaries is also an important factor during densification. If the pores remain attached to the grain boundaries during sintering, densification is facilitated. If, however, the pores break away from the boundaries, they tend to become trapped in the crystal interior, thus making densification more difficult. 5.2.4 Effect of Time and Temperature The rate of crystal growth and sintering of the MgO, formed by decomposition of the hydroxide, is controlled primarily by the temperature and time of heat treatment. Crystal growth has been shown to fit an equation of the following form (Lindner and Parfitt, 1957; Riddell, 1972): Gn

Gn0 ¼ kt ¼ A exp ( E=RT)t

(5:13)

where G and G0 are the X-ray crystal sizes at time t and 0 and n is a constant between 5 and 7. This relationship was found to be valid up until a crystal size of 50 –67 nm, after which there is an apparent change in mechanism and the crystal growth kinetics were found to fit the following equation: dG=dt ¼ Mb (A=G

C)

(5:14)

where A and C are constant and Mb is the grain boundary mobility. The change in kinetics generally occurs when the density reaches 90 –95% of theoretical. Calcining of magnesium hydroxide to produce caustic-calcined magnesia takes place under a specific set of temperature conditions, generally in the range 500–8008C. Figure 5.4 displays graphically the effect calcination time and temperature have on the resultant surface area of the oxide. The specific surface area of the starting magnesium hydroxide is 10–20 m2g21, and it is

5.2 CALCINATION OF MAGNESIUM HYDROXIDE

95

Figure 5.4 Effect of calcination time and temperature on surface area of MgO using seawater magnesium hydroxide precursor.

apparent that the surface area increases dramatically over a short period of time, reaching a maximum at 15 min. The highest surface area is produced using a temperature close to the decomposition temperature of magnesium hydroxide (3508C). However, higher calcining temperatures produce an MgO with a progressively lower surface area at 15 min. It is also noticeable that an increase in

Figure 5.5 Variation of crystallite size and surface area for magnesium oxide produced from seawater magnesium hydroxide.

96

CALCINATION OF MAGNESIUM HYDROXIDE AND CARBONATE

calcination time also effectively decreases surface area, although its effect is not as dramatic as the effect of temperature. Figure 5.5 displays the relationship between MgO crystallite size and specific surface area, and, as one would expect, a magnesium oxide with a small crystallite size has a correspondingly greater specific surface area.

REFERENCES Anderson, P. J., and Horlock, R. F. (1962). Trans. Faraday Soc. 58, 1993. Anderson, P. J., and Morgan, P. L. (1964). Trans. Faraday Soc. 60, 930. Anderson, P. J., Horlock, R. F., and Avery, R. G. (1965). Proc. Brit. Ceram. Soc. 3, 33. Ball, M. C., and Taylor, H. F. W. (1961). Min. Mag. 32, 754. Bernal, J. (1938). Trans. Faraday Soc. 34, 834. Bouvier, A., Griesser, H. J., and Herzog, G. W. (1980). Radex Rundsch. 231. Boynton, R. S. (1980). Chemistry and Technology of Lime and Limestone, Wiley: New York. Brett, N. H. (1969). Min. Mag. 37, 244. Brett, N. H., and Anderson, P. J. (1966). Trans. Faraday Soc. 63, 2044. Brindley, G. W. (1963). In Progress in Ceramic Science, Pergamon: London. Eubank, W. R. (1951). J. Am. Ceram. Soc. 34, 225. Freund, F., Martens, R., and Scheikh-OL-Eslami, N. (1975). J. Therm. Anal. 8, 525. Freund, F., Martens, R., and Gentsch, H. (1977). Sci. Ceram. 9, 308. Garn, D. C., and Freund, F. (1975). Trans. Brit. Ceram. Soc. 74, 23. Garn, D. C., Kawalec, B., and Chang, J. C. (1978). Thermochim. Acta 26, 375. Goodman, J. F. (1958). Proc. Royal Soc. 247, 346. Gregg, S. J., and Packer, R. K. (1955). J. Chem. Soc. p. 51. Guilliatt, I. F., and Brett, N. H. (1970). Trans. Brit. Ceram. Soc. 69, 1. Hyde, C., and Duckworth, W. H. (1962). AD 266 735 U.S. Government Research Report 37, 69. Knibbs, N. V. S. (1924). Lime and Magnesia. D. Van Nostrand Company, New York. Lindner, R., and Parfitt, G. D. (1957). J. Chem. Phys. 26, 182. Lønvik, K. (1978). J. Chem. Phys. 27, 27. Martens, R., Gentsch, H., and Freund, F. J. (1976). Catal. 44, 366. Moodie, A. F., Warble, C. E., and Williams, L. S. (1966). J. Am. Ceram. Soc. 49, 676. Na¨gerl, H., and Freund, F. J. (1970). Therm. Anal. 2, 387. Riddell, D. (1972). PhD thesis, Sheffield University, United Kingdom. Sharp, J. H. (1973). Trans. Brit. Ceram. Soc. 72, 21.

6 FURNACES AND KILNS

6.1

INTRODUCTION

A variety of furnaces are used in the magnesia industry to calcine either magnesite or magnesium hydroxide filter cake. The most commonly used are the multiple-hearth furnace (MHF) and rotary and shaft kilns. MHFs are primarily used to produce reactive lightly calcined magnesia, while rotary and shaft kilns produce hard and dead-burned grades of magnesia. Heat transfer in kilns occurs in a variety of ways. By far the largest transfer of heat from the hot flame to the kiln burden occurs by radiation; a much smaller quantity of heat is also transferred by radiative processes from the refractory lining. The hot refractory lining can also transfer heat to the burden by conduction. Heat transfer from the hot combustion gases can also occur by convective processes as well. The heat balance of a furnace at steady state is described by Equation (6.1); see also Figure 6.1: Enthalpy of material input þ heat of reaction

¼ enthalpy of material output þ convective and radiative losses

(6:1)

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 97

98

FURNACES AND KILNS

Figure 6.1 Furnace heat balance.

6.2 MULTIPLE-HEARTH FURNACES Multiple-hearth furnaces basically consist of a vertical cylindrical steel shell lined with refractory that encloses a series of horizontal circular hearths; see Figure 6.2. The number of hearths is variable depending upon its application; the magnesia industry, however, generally employs MHF with 10–14 hearths firing in the temperature range 1500–19008F. Energy usage is about 6330 –8440 MJ (6–8 MBtu) per ton of MgO produced. A central shaft extends the full height of the furnace and typically supports four cantilevered rabble arms above each hearth. Attached to the rabble arms are rabble blades or ploughs that move with the rotation of the central shaft, ploughing material on the hearth and conveying it to the lower hearth. The material actively being moved by the plough is called the “live bed” and through the action of the ploughs creates windrows of material. Beneath the bottom of the ploughs and the hearth there builds up a static layer of material that is not moved by the ploughs; this is known as the “dead bed.” See Figure 6.3. Both the central shaft and rabble arms are of a double-wall design that allows forced cooling air to travel up though the center shaft, through the

6.2 MULTIPLE-HEARTH FURNACES

Figure 6.2

Figure 6.3

Cross section of a multiple-hearth furnace.

Windrows formed by multiple-hearth furnace ploughs.

99

100

FURNACES AND KILNS

rabble arms, and back into the shaft annular space to exit at the top of the furnace. Feed material is introduced continuously at the top of the furnace and is rabbled in a spiral path across each hearth to a drop hole located alternately at either the periphery or the inner most part of the hearth. The drop hole allows material to fall to the hearth below. This process of mechanical transfer of material through the furnace results in a highly uniform product since all of the material has had the same retention time. Control of retention time is determined by furnace size (number of hearths and diameter of kiln), shaft rotation speed, number of arms, rabble plough-type, plough spacing, and configuration of the rabble ploughs. Proper raking action is an important factor in the control of the calcination process. The ploughs that move material from the center of the hearth to the perimeter are called the out ploughs, while the ploughs that move material in the opposite direction are called the in ploughs. The angle at which the plough encounters the ridge of material is called the plough angle; see Figure 6.4. The plough angle and dimension is a function of the desired

Figure 6.4

Multiple-hearth furnace plough angles for an “in” or “out” plough.

6.2 MULTIPLE-HEARTH FURNACES

Figure 6.5

101

Various arrangements of MHF ploughs or rabble teeth.

rabbling configuration, nature of material being processed, and desired retention time. Some rabbling configurations may use a spacer between ploughs, especially if the material is prone to bridging between the ploughs, such as magnesium hydroxide filter cake. Rabble ploughs are typically arranged into four basic configurations dependent upon the desired degree of mixing and retention time; see Figure 6.5. 6.2.1 Single Progressive Rabble (Four Arms per Hearth) This system moves the material in a steady inward spiral on an in-hearth and an outward spiral on an out-hearth. The rabble ploughs have spacers between them to prevent “bridging” of sticky material between the ploughs.

102

FURNACES AND KILNS

6.2.2 Full Progressive Rabble (Four Arms per Hearth) This rabble system produces a more uniform retention time than the single progressive method by reducing the windrow size. Sticky material may pose a bridging problem since there are no spacers between the ploughs. 6.2.3 Back Rabble (Four Arms per Hearth) The back rabble system is used to increase mixing and surface exposure and also to increase retention time. The material will advance three steps and then regress one step, for a net advance of two steps per shaft rotation. 6.2.4 Full Progressive Rabble (Two Arms per Hearth) This system uses two rabble arms and accomplishes the same mixing as the four-arm single progressive method. However, the retention time is doubled compared with the four-arm configuration. Heat is normally provided by firing oil or gas burners on certain hearths. The most common configuration is for direct flame to be introduced into the hearth, although it can be designed for just the hot gases to be introduced onto the hearth as well. The latter involves combusting the fuel external to the kiln and then piping the hot gases through ductwork to the hearths. Normally, the hot gases are exhausted at the top of the furnace, flowing countercurrent to the flow of material. 6.2.5 Refractory Linings Multiple-hearth furnaces are lined with an appropriate refractory backed by one or more layers of insulating material. The type of refractory lining is chosen to match the firing temperature conditions, required external shell temperature, and resistance to processing conditions. Generally, high-duty fireclay brick or superduty high alumina is used in magnesia production; see Figure 6.6. Castable refractory is used to form required shapes around burner ports. For the production of light-burned grades of magnesia, close temperature control is important. MHFs are capable of closely controlled firing, typically +108C of the set point at all stages of the calcining. The firing conditions on each hearth can be varied independently. Automatic draft and oxygen monitoring equipment provide close control over power draw and fuel demand, which assures clean combustion. Processing time in the furnace is controlled by the speed of the shaft, which is altered using a variable speed drive, and feed rate to the furnace controls the bed depth on the hearth.

6.3 HORIZONTAL ROTARY KILNS

103

Figure 6.6 Cutaway section of an MHF displaying the various types of refractories used in construction. (Reproduced by Permission of Metso Minerals.)

6.3

HORIZONTAL ROTARY KILNS

A rotary kiln consists of a long cylinder between 110 and 140 m long, inclined to horizontal at an angle of 38 to 48 with a typical diameter of 2.4– 2.7 m. The refractory-lined cylindrical shell is motor driven using a variable-speed drive and rotates slowly around the inclined axis at about 1.5 rpm. Magnesite is fed into the upper back-end countercurrent to the combustion gases, while the flame is generated at the lower front-end. The hot zone of the kiln is lined with a high-temperature basic magnesia brick, while on either side of the hot-zone high alumina brick is used. The magnesia is discharged from the kiln into a cooler. Equation (6.2) can be used to calculate the speed with which stone may flow through a rotary kiln: p 1:77 Q  L t¼ SDN

(6:2)

104

FURNACES AND KILNS

where t ¼ time of stone in the kiln, min L ¼ length of kiln, ft D ¼ interior diameter of kiln, ft S ¼ slope of kiln, deg N ¼ revolutions per minute Q ¼ angle of repose of stone There are a number of designs that can be used, but generally the magnesia industry utilizes a directly fired kiln using an internal burner. Firing temperatures can be as high as 30008F with residence times of 3–6 h. Kilns can have internal adaptations that help improve the efficiency of the calcinations process. These include: .

. .

Chains located at the back-end help to break up and deagglomerate a wet filter cake or sludge. This arrangement is typically used when calcining seawater or brine precipitated magnesia. Lifters that cause the ore to cascade through the hot combustion gases. Internal refractory dams that increase residence time of the burden.

Burner design is important for fuel efficiency and reliable kiln operations. The flame must be of the correct length in relation to the length of the kiln. If the flame is too long, radiant heat transfer is insufficient in the calcining zone, which results in an elevated back-end temperature and a subsequent decrease in thermal efficiency. If the flame is too short, it can cause excessive temperatures and premature refractory failure in the front-end. The flame should also be centered along the inclined axis and not impinge upon the refractory lining. The rotary kiln can accept a wide range of feed sizes from about 3 in. downward. The tumbling burden inside the kiln causes the larger size material to move toward the outside of the bed, while the finer fractions concentrate at the center of the bed. This results in the larger pieces of ore being exposed to higher temperatures and avoids overburning of the finer material. A rotary kiln can produce a wide range of activities from a light burn to deadburned product; however, the rotary kiln tends to introduce more variability in product activity than does a multiple-hearth furnace or shaft kiln. The rotary kiln can be fired using gas, oil, or solid fuel such as petroleum coke or coal. Figure 6.7 displays a diagram of a gas-fired burner system. Energy consumption of the rotary kiln depends upon the type of product being fired and the required firing temperature as well as radiative and convective losses from the kiln. The energy consumption when producing hard-burned magnesia at a firing temperature of about 2600–28008F is about 10,550–12,660 MJ (10 –12 MBtu) per ton of product.

6.4 EXTERNAL WATER COOLERS

Figure 6.7

105

Rotary kiln gas burner system. (Reproduced by permission of Metso Minerals.)

One unique feature of the rotary kiln is its propensity to develop rings of accumulated material on the refractory lining of the kiln, which have to be removed periodically so they do not interfere with material flow through the kiln. The removal process consists of firing a large gauge iron slug, using a shotgun, aiming down the kiln at the ring. The impact of the slug on the ring generally has the desired effect of causing it to collapse back into the center of the burden. Rotary kilns can also be fitted with a kiln-feed preheater; see Figure 6.8, which enables a shorter kiln to be used, approximately one-half the length of a conventional rotary. The exhaust heat is used to preheat the burden entering the kiln thus increasing heat recovery and improving the thermal efficiency of the furnace. 6.4

EXTERNAL WATER COOLERS

These are of simple design and consist of a slightly inclined rotating cylinder wall that is cooled externally either by a water spray or by partial submersion in a water bath. Heat transfer occurs from the product to the cylinder walls and then is transferred to the water whereby some of the heat is lost in the generation of steam. See Figure 6.9, which displays a diagram of an external

106

FURNACES AND KILNS

Figure 6.8 Rotary kiln fitted with a preheater. (Reproduced by permission of J.A.H. Oates and Wiley-VCH.)

water-cooled cooler. Another water-cooled system uses a cooling jacket, through which cold water is circulated. This design ensures that the hot material is in continuous contact with the water-cooled walls of the cylinder. A cascade cooler operates in a countercurrent manner in which the cooling air flows in direct contact with the hot material. Lifter flights on the interior wall of the cylinder lift distribute and transport the hot material.

Figure 6.9

Water cooler.

6.5 SHAFT KILNS

107

The cascading of the hot material through the flow of cooling air results in the transfer of heat to the air. This type of system is only applicable to coarser materials.

6.5

SHAFT KILNS

Three aspects are common to all shaft kiln designs, namely charging, drawing of the ore, and combustion. A shaft kiln is essentially a vertical refractory lined cylinder or ellipse. The ore is charged in at the top of the furnace, along with, in some cases, a solid fuel such as coke or anthracite coal. Other fuels such as natural gas and oil can also be employed. There are a number of different variants of shaft kilns, such as the mixed feed, traditional type and modern basic design, annular, parallel-flow regenerative, double inclined, and multichamber. 6.5.1 Ore Charging The problem with charging the kiln feed at a single point is that it produces a conical surface similar to that of a bed of stone. This results in the larger pieces rolling down the heap toward the outside of the bed, while the finer material remains concentrated toward the axis. This size gradation can result in an uneven flow of gases through the burden. The finer material at the center tends to restrict gas flow, while the larger material toward the outer limits of the kiln allows less resistance to the flow of gases. This results in the finer material being undercalcined. A number of devices have been developed to overcome this problem. The fixed plate and cone system (see Fig. 6.10a), where the arrangement of the cone and striker plate, relative to feed chute, is adjusted to produce a uniform size profile around the kiln. The rotating hopper and bell system (see Fig. 6.10b) produces a more uniform profile. Movement of the bell up and down controls the flow of magnesite ore into the kiln. 6.5.2 Discharge The drawing of the kiln feed should be such as to produce a uniform movement of the burden across the kiln. It can be a simple system, using a single discharge point and a conical table; see Figure 6.11. A more reliable system uses four discharge points, which allows for compensations to be made if a blockage occurs in the kiln by speeding up the discharge rate of the free ports. This allows the same rate of material flow to occur in the calcination zone without overburning occurring.

108

FURNACES AND KILNS

Figure 6.10 (a) Plate and cone charging system for a shaft kiln, and (b) bell and hopper charging system.

6.5 SHAFT KILNS

109

Figure 6.11 Discharge system for a shaft kiln. (Reproduced by permission of J.A.H. Oates.)

6.5.3 Modern Shaft Kiln Outputs can range from 100 to 800 tonnes per day. The minimum feed size is 0.8 in. to a top size of about 7 in. Figure 6.12 shows a cutaway section of a modern pressurized shaft kiln along with a diagram showing burner arrangement and the flow of gases and materials through the kiln. 6.5.4 Double-Inclined Kiln The double-inclined kiln incorporates two inclined sections in the calcining zone; see Figure 6.13. The offset arches created by the inclines create a void on both sides of the kiln where fuel and preheated combustion air can be fired in combustion chambers. Cooling air is drawn into the bottom of the kiln, and a portion of this is withdrawn further up the kiln and is reinjected in the combustion chambers. The double incline creates a circuitous route for both the burden and combustion gases, which ensures a uniform distribution of heat that allows the feeding of material as small as 0.8 in. The top feed size is about 4 in. 6.5.5 Multichamber Kiln The multichamber shaft kiln is a variation of the double-inclined kiln. It consists of 4 or 6 alternating inclined sections in the calcining zone, which

110

FURNACES AND KILNS

Figure 6.12 (a) Cutaway section of a Maerz pressurized shaft kiln, and (b) burner arrangement and gas flow through furnace. (Reproduced by permission of RCE Industrieofenbau Engineering GmbH.)

6.5 SHAFT KILNS

111

Figure 6.13 Double-inclined shaft kiln. (Reproduced by permission of J.A.H. Oates and Wiley-VCH.)

creates offset arches that serve the same purpose as those in the doubleinclined kiln. The unique feature of this kiln is that the temperature of the lower combustion chambers can be varied, enabling this kiln to produce magnesia with a wide range of activities. 6.5.6 Annular Shaft Kiln The annular shaft kiln consists of a central cylinder that restricts the effective thickness of the burden; see Figure 6.14. This ensures a good distribution of

112

FURNACES AND KILNS

Figure 6.14 Annular shaft kiln. (Reproduced by permission of J.A.H. Oates and Wiley-VCH).

heat in the burden itself. The burden is drawn through the annulus between the walls and the central cylinder of the kiln and passes by two sets of burners. The majority of the fuel is fired in the upper burners using a deficit of air. This deficit is satisfied by the gases from the lower burners, which are operated with an excess of air. The temperature in the lower zone of the kiln is moderated by mixing of cooling air. This kiln can utilize feed sizing in the range of 0.5–10 in.

6.5 SHAFT KILNS

113

6.5.7 Parallel-Flow Regenerative Kiln The parallel-flow regenerative kiln consists of two interconnected vertical cylindrical shafts. Burden is charged alternatively to each shaft and drawn downward through a preheating heat exchange zone and into the calcining zone. In the first stage, burden is charged into shaft 1 and fuel is injected through the lances into the shaft and burns in the combustion air blown down the shaft. The cooling air in shaft 1, together with the combustion gases, pass through the interconnecting shaft and into shaft 2, where they heat the refractory lining in the preheating zone of shaft 2. After a period of time suitable for calcination of the burden, the second stage commences, which starts off with the fuel and air to shaft 1 being cutoff. Burden is charged into shaft 2 along with the injection of fuel and air. The exhaust gases are finally vented from the top of shaft 1. Feed size to this kiln is in the range of 12 to 112 in. and is capable of producing an active product with low residual carbonate content.

7 POSTCALCINATION PROCESSING

7.1

INTRODUCTION

Once the magnesite has been calcined and has exited the kiln, there are a number of processing steps that it can undergo to produce the finished product. These include processes such as screening to produce coarse fraction products, or to increase the purity of the finer fraction by separating a lower purity coarse fraction, or grinding to produce fine powders. These various processing steps are discussed below.

7.2

GRINDING

The two mills most commonly used to fine-grind magnesium oxide are the ball mill and the ring-roller mill. If the mill feed already contains an appreciable quantity of the desired size fraction, then the mill feed is typically air classified first, the fine fraction being separated from the oversized fraction. This helps to improve grinding efficiency by removing a large quantity of the fines before entering the mill, where they would tend to cushion the coarse fraction from the mill grinding surfaces. The mill is continually swept with air to remove fines, which then reenter the classification circuit. Alternatively, if the mill feed is coarse and does not contain any significant The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 115

116

POSTCALCINATION PROCESSING

Figure 7.1

Schematic of a typical milling circuit.

quantity of fines, then the material can be fed directly to the mill. Figure 7.1 displays a schematic diagram of a typical milling circuit.

7.2.1 Ring-Roller Mills Ring-roller mills are equipped with rollers that operate by grinding against a grinding ring. The ring or the roller may be stationary and may be arranged in either a horizontal or vertical position. Grinding occurs between the surfaces of the roller and ring. The pressure between the rollers and ring is applied either through springs or centrifugal force. Mills may have a system of internal particle size classification that consists of either screens or some arrangement of rotational classifier or “whizzer.” The Alstom ring-roller mill (see Fig. 7.2), which is typically used in synthetic magnesia plants, is of the internal classification type. The grinding ring is located in the base of the mill in the horizontal plane. Beneath the grinding ring are located tangential air ports that allow sweep air to enter into the grinding chamber. The pivoted rollers are attached to a driven vertical shaft and move outward against the grinding ring under the influence of centrifugal force. The individual rollers also rotate on their own bearings while traveling around the ring. The feed material drops between the rolls and ring and is subsequently crushed. The sweep air entrains fines and conveys them away from the grinding zone and into the classifier. Oversize material is returned back to the mill for regrinding. The classification method used in the mill depends upon the degree of fineness required. If a medium fineness material is required (100 mesh), then a single-cone air classifier is used. If a finer ground material is required, or frequent changes in fineness are

7.2 GRINDING

Figure 7.2

117

Ring roller mill. (Reproduced by permission of Alstom Power.)

employed, then a whizzer-type classifier is used. This type of ring-roller mill is known as the “high-side mill.” Grinding capacities range from 0.5 to 50 tons per hour. 7.2.2 Ball Mills A ball mill is used for intermediate and fine grinding applications and consists of a horizontally rotating cylinder; see Figure 7.3. The length of the drum is typically 1–1.5 times the shell diameter and is filled to 40 –50% capacity with grinding media. The largest mills in operation today are about 8 m in diameter. Three types of discharge arrangements are commonly used, namely overflow, grate, and end peripheral discharge. The material to be ground fills the voids between the grinding media, and the rotation of the shell results in a tumbling or cascading motion of the grinding media, which captures particles in ball-ball or ball-liner contacts

118

POSTCALCINATION PROCESSING

Figure 7.3

Ball mill cutaway. (Reproduced by permission of Metso Minerals.)

and loads them to the point of fracture. Typical feed size for ball mills is less than ½-in. while output size is about 20 mm and greater. Ball mills typically have replaceable liners that are constructed of cast alloy steels, cast iron, or polymer. Some commonly used mill liners are the double-wave variety, which has a corrugated surface and a wedgeshaped shiplap liner. The purpose of these liner surface profiles is to help prevent slippage of the charge during tumbling, which typically occurs in smooth-lined mills. Lifters fitted to the liners may also help prevent this problem from occurring. Grinding balls can be made of forged steel, cast steel, cast iron, ceramic, or natural stone. Heat-treated forged-steel balls have superior wear characteristics. The rotational speed of the mill is crucial for optimal operation with the most efficient grinding occurring when the balls roll or cascade from the top to the bottom of the pile. If mill speed is too fast, a “cataracting” motion occurs where the balls, rather than cascading down the pile, are literally thrown through the air from the top to the bottom of the pile. The optimum mill speed is about 65 –80% of the critical speed, which is defined as the theoretical speed at which the centrifugal force on a ball in contact with the mill shell at the height of its path equals the force on it due to gravity; see Equation (7.1): Nc ¼ 76:6=(D)1=2

(7:1)

REFERENCE

119

where Nc is the critical speed in revolutions per minute and D (in feet) is the diameter of the ball mill, assuming that the ball diameter is small compared with the diameter of the mill. The chief factor that determines grinding media size is the feed size of material to be ground. A coarser feed requires a larger diameter ball than a finer sized feed. The recommended optimum size makeup for balls (or rods in the case of a rod mill) (Bond, 1958) is calculated as follows: pffiffiffi Db ¼ (200nr =X p Ei )1=2 ( D=rs )1=2

(7:2)

where Db is the ball diameter (inches), D is the mill diameter (feet), Ei is the work index for the feed, nr is speed expressed as percent of critical speed, and rs is feed specific gravity. The grinding media charge can be expressed as a percentage of the mill volume, that is, a mill half-full of media contains a 50% charge. The void space present between a static ball charge is approximately 41%. The quantity of material in a mill can be expressed as the ratio of mill volume to the volume of the voids in the ball load, otherwise known as the material-to-void ratio (M/V ratio). In practice, the M/V ratio is kept close to 1, while grinding media charging varies from 20 to 50%.

BIBLIOGRAPHY Fuerstenau, M. C., and Han, K. N. (2003). Principles of Mineral Processing, Society for Mining, Metallurgy, and Exploration: Littleton, Colorado, USA. Perry, R. H., Green, D. W., and Maloney, J. O. Perry’s Chemical Engineers’ Handbook, 6th ed., (1984). McGraw-Hill: New York.

REFERENCE Bond, F. C. (1958). Min. Eng. 10, 592– 595.

8 PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

8.1

INTRODUCTION

The occurrence of periclase, or magnesia as it is commonly called, in its natural form is rare and is generally produced either by the calcination of magnesium hydroxide or magnesium carbonate. Periclase was discovered by A. Scacchi in 1840 at Monte Somma, Vesuvius, Italy, as a mineral contaminated with ferrous oxide. The word periclase is derived from the Greek word peri, which means “around,” and the word klao, which means “to cut,” presumably alluding to the way periclase cleaves. There is a large amount of data concerning the physical properties of magnesia owing to its importance as a refractory oxide. Some of the more useful properties are summarized here.

8.2

PHYSICAL PROPERTIES OF MAGNESIUM OXIDE

Color/Luster Colorless to grayish white, also yellow to brown or black due to iron or foreign inclusions. Its luster is vitreous and transparent. Crystal System/Habit Isometric. Usually occurs as anhedral to subhedral crystals in a matrix.

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 121

122

PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

Figure 8.1

Magnesium oxide crystal structure.

Crystal Structure Periclase has a cubic face-centered crystal lattice isomorphous with that of sodium chloride and calcium oxide; see Figure 8.1. Density 3.55–3.68 g/cm3. X-ray density 3.581 g/cm3. Hardness 5.5 on the Mohs scale. Index of Refraction The refractive index of periclase (polycrystalline) is 1.7350 at 633 nm. Thermal Conductivity The thermal conductivity of magnesia has been determined by a number of workers. The figures for polycrystalline sintered magnesia at lower temperatures (Francl and Kingery, 1954) and sintered magnesia at a higher temperature range (McQuarrie, 1954) are given in Tables 8.1 and 8.2, respectively.

TABLE 8.1 Thermal Conductivity of Polycrystalline Sintered Magnesia Temperature (8C) 0 100 300 500 600

Thermal Conductivity (cal s21 cm22 8C21 cm) 0.10 0.083 0.067 0.031 0.026

8.2 PHYSICAL PROPERTIES OF MAGNESIUM OXIDE

123

TABLE 8.2 Thermal Conductivity of Sintered Magnesia at High Temperatures Temperature (8C)

Thermal Conductivity (cal s21 cm22 8C21 cm)

1000 1200 1400 1600 1700

0.015 0.014 0.014 0.016 0.018

Electrical Resistance The electrical insulating power of magnesia is very high, which makes for an excellent high-temperature electrical insulator. The specific resistance of crystals of pure periclase is the highest and has been measured as 2.3  109 V.cm at 7008C (Rochow, 1938). The relationship between specific resistance r of magnesia and temperature T (K) can be expressed by

r ¼ AeB=T

(8:1)

where A and B are specific constants. Very high purity magnesia, sintered in a vacuum at 19508C, has an extremely high electrical resistivity of about 1016 V.m. In the region of lower temperatures, the specific resistivity of magnesia is a function of its chemical purity. However, at 20008C there is virtually no difference in the resistivity of pure magnesia as compared to an impure one. The dielectric constant for periclase has been measured as 3.2–9.8 at 258C and 1 MHz, while its dielectric loss at 1 MHz is 3  1024 (tan d). Tables 8.3 and 8.4 display the specific resistivity for sintered and high-purity MgO, respectively.

TABLE 8.3 Magnesia Temperature (8C) 950 1000 1300 1500

Specific Resistance of Sintered Specific Resistance (103 V) 120 95 9 1.5

Source: Data from Diepschlag and Wulfesting (1929).

124

PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

TABLE 8.4 Magnesia

Specific Resistance of High-Purity

Temperature (8C)

Specific Resistance (V) 9  107 2.5  105 5.5  103 4.4  102

900 1300 1700 2100 Source: Data from Foe¨x (1942).

Melting Point Periclase has a very high melting point of 28008C and a boiling point of 36008C. Thermal Expansion The thermal expansion of periclase is the greatest of all pure refractory oxides and approaches the expansion of metals. Exact expansion measurements have been carried out on single crystals of periclase and high-purity sintered magnesia, the results of which are tabulated in Tables 8.5 and 8.6. Heat Capacity Values for the specific heat capacity from ambient temperature to 3600 K are shown in Table 8.7. TABLE 8.5

Average Thermal Expansion Coefficient (L) of Single-Crystal Periclase

Temperature (8C)

L  1026/8C

50 100 300 600

6.7 9.1 11.6 13.0

Temperature (8C) 700 800 900 1000

13.2 13.5 13.7 13.8

Source: Data from Austin (1931).

TABLE 8.6 Thermal Expansion of High-Purity Sintered Magnesia Temperature (8C) 300 500 700 900 1100 1300 1500 1700 1800 Source: Data from Ebert and Tingwaldt (1936).

L  1026/8C

L  1026/8C 12.0 12.6 13.2 13.7 14.7 14.5 15.0 15.6 16.0

8.3 CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

125

TABLE 8.7 Specific Heat Capacity for MgO Temperature (K)

Specific Heat Capacity (J K21 mol21)

298.15 600 1000 1800 2800 3600

37.106 47.430 51.208 54.898 58.491 61.214

Source: Data from NIST-JANAF (1998).

Heat of Formation The standard enthalpy of formation is 2601.241 kJ/mol. Magnetic Susceptibility 20.25 (k) 1026 emu/g at 258C (Lide, 1992). Structural Properties (Data Obtained from General References) Elastic modulus 30.5–46  106 psi (210– 317 GPa) Modulus of rupture 14–15  106 psi (96.5–103.4 GPa) Tensile strength 13.9  106 psi (95.8 GPa) Shear modulus 11 –19  106 psi (75.8– 131 GPa) Flexural strength 13  106 psi (89.6 GPa) Compressive strength 1.2–2.1  105 psi (0.83–1.44 GPa)

8.3

CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

The physical and chemical properties of magnesium oxide are primarily governed by the source of the precursor, that is, derived from magnesite or precipitated from brine or seawater. Other important factors include time and temperature of calcination and the presence of trace impurities. Electron microscope studies have revealed that the precursor particle morphology has a large impact on the morphology of the final MgO particle. It has been shown that when brucite and magnesite crystals are thermally decomposed at low temperatures, pseudomorphs of a size and shape similar to the parent crystal are formed. Caustic-calcine magnesias are calcined below 9008C and are characterized by moderate to high chemical reactivities. They are readily soluble in dilute acids and hydrate rapidly in cold water. They also slowly react with atmospheric moisture and carbon dioxide to form the basic carbonate, 5MgO . 4CO2 . xH2O. Common test methods for determining the reactivity

126

PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

of caustic-calcine magnesia are specific surface area and the caustic magnesia activity test (see Appendix for test methods). Magnesium oxide has very limited solubility in water. However, an accurate determination is complicated by trace quantities of dissolved carbon dioxide in the water and the purity of the sample, where the presence of lime can introduce error. Another complicating factor is the source of the periclase and what heat treatment it had received. All these factors influence the rate at which equilibrium is reached in solution and hence solubility. Many solubility measurements have been made that have produced a wide variation in results. The most accurate determination to date has produced a result of 8.6 mg/L at 308C. Caustic-calcined magnesia readily hydrates to the hydroxide on exposure to moisture. 8.3.1 Dissolution of Magnesium Oxide The dissolution of magnesium oxide in acidic media has been extensively studied by Segall et al. (1978, 1981, 1984, 1985, 1988). Its dissolution rate in pH 1 acid at 308C has been estimated by them as 1.2  1024 mol m22 s21, which is very close to diffusion control and corresponds to a penetration rate of 1023 – 1024 cm s21. They speculate that the initial rate is controlled by the speed with which water molecules formed from the surface hydroxyl ions and protons can be transported out through the Helmholtz layer; see Figure 8.2. Segal et al. also speculate that there are four stages in the process of dissolution of MgO. As soon as the oxide is immersed in acid solution, a potential difference and a solution double layer containing no dissolved cation is formed. The surface morphology rapidly alters to a castellated structure by surface diffusion before any dissolution takes place. Next cations dissolve and a gradual approach to saturation of the interphase region between oxide and solution takes place, again with an alteration of potential difference. At this point water removal as a cation solvation sheath is the rate-determining step. Finally a stable double-layer potential is formed and reaction (8.2) becomes the rate-determining step: O2solid þ Hþ aq ! OHsolid

(8:2)

Fruhwirth et al. (1985) studied the dissolution of and hydration kinetics of MgO single crystals and powder samples with regard to Hþ and Mg2þ concentrations. Several pH-dependent rate-controlling processes were found to be present at room temperature. 1. At pH , 5 the rate-controlling step was Hþ attack followed by desorption of Mg2þ and OH2. The rate was proportional to either 2pH or

8.4 SURFACE STRUCTURES OF MgO

127

Figure 8.2 Representation of the dissolution of magnesium oxide in acid solution. (Reproduced by permission of the Royal Australian Chemical Institute.)

pMg–pH, and these processes are part of the overall neutralization reaction (8.3): MgO þ 2Hþ ! Mg2þ þ H2 O

(8:3)

2. At pH  5 the rate-determining step was proton diffusion, the rate being proportional to pH. 3. At pH . 7, the rate-controlling step was OH2 adsorption followed by Mg2þ and OH2 desorption. These processes are part of the overall dissolution reaction (8.4): MgO þ H2 O ! Mg2þ þ 2OH

8.4

(8:4)

SURFACE STRUCTURES OF MgO

Magnesium oxide possesses the rock salt crystal structure, which is the simplest oxide structure from the standpoint of the bulk arrangement of atoms, and as a consequence has been one of the most studied oxides; see Figure 8.1.

128

PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

Surface defects are a very important feature of many metal oxides since these structures are the key to their chemical reactivity. Even the best method of attempting to prepare nearly perfect surfaces, which involves cleaving an MgO crystal in ultrahigh vacuum, still results in a variety of surface defects being formed. These tend to arise due to the large amount of energy that is dissipated in the cleaving of an MgO crystal, which can cause disruption of the surface. Studies have shown that photons, electrons, atoms, and ions can be released during fracture. Once the crystal has been cleaved, a variety of techniques can then be used to characterize the new surface. These include, but are not limited to, scanning tunneling (STM) and atomic force microscopy (AFM), low-energy electron diffraction (LEED), ion scattering spectroscopy (ISS), and extended X-ray absorption fine-edge structure (EXAFS) Figure 8.3a displays a simplified block model of a metal oxide surface showing one-dimensional defects in the form of steps. These steps in turn give rise to point defects in the form of kink sites. Terraces can also possess a variety of different surface sites such as adatoms, vacancies, and substitutional and interstitial point defects. Figure 8.3b shows that the

Figure 8.3 (a) Magnesium oxide surface structures, and (b) block diagram showing the coordinate unsaturation of surface ions (numbers represent coordination number). [Diagram (a) reprinted with permission from Chem. Rev. 99(1), 83 (1999). Copyright 1999 American Chemical Society. Diagram (b) from Chemical Dissolution of Metal Oxides, Miguel A. Blesa. Copyright 2006 by CRC Press. Reproduced with permission of CRC Press in the format Textbook via Copyright Clearance Center.]

8.5 MOLECULAR ADSORPTION ON MgO

129

coordination number of the surface ions is also decreased making them coordinatively unsaturated. The (100) face of MgO crystals has been the subject of many studies and is energetically the most stable face in rock salt crystals (Coluccia et al., 1979; Colburn and Mackrodt, 1983; Moodie and Warble, 1971). The reason for its stability is that it is electroneutral (nonpolar), containing equal numbers of cations and anions. MgO cleaves extremely well along the (100) plane, yielding a relatively flat and defect-free surface. The anions and cations on the (100) surface are pentacoordinated, four in the surface plane and one directly below in the second atomic plane. The missing bond lies directly above the surface ion. Two simple types of structural defects that have been studied in MgO are the O vacancy, or F center, and the cation vacancy, or V center. A neutral O vacancy has two electrons remaining, which may be trapped by the unbalanced Coulombic potential associated with the vacancy. If one of these two electrons is removed, this results in the formation of an F þ center. Both F and F þ centers can be characterized spectroscopically by their optical absorption bands at 5.01 and 4.95 eV, respectively. The V 2 center is an Mg2þ vacancy, with one associated electronic hole. The hole may also be regarded as an O2 ion. Studies of the V 2 center show that the electronic hole is not delocalized over the oxide ions surrounding the vacancy, but is trapped at a single O2 site (Hayes and Stoneham, 1985). This configuration is stabilized by a polarization and distortion of the lattice. 8.5

MOLECULAR ADSORPTION ON MgO

Chemisorption on an MgO surface will be primarily an acid/base interaction. Cation sites are Lewis acids and may interact with donor molecules such as H2O through a combination of electrostatics (ion–dipole attraction) and orbital overlap. Oxide ions also act as basic sites and can interact with acceptors such as Hþ. In fact one of the most common dissociative reactions is the deprotonation of an adsorbate to produce surface hydroxyl groups. 8.5.1 Chemisorption of Various Molecules on MgO Hydrogen Experimentally no evidence for interaction with H2 has been observed at room temperature. Carbon Dioxide Lattice oxygen is essentially basic toward acidic molecules such as CO2, forming surface CO22 3 (Ochs et al., 1998); see reaction (8.5): CO2 þ O2(solid) ! CO23(solid)

(8:5)

130

PHYSICAL AND CHEMICAL PROPERTIES OF MAGNESIUM OXIDE

Water and Methanol The interaction of H2O with the (100) surface of MgO prepared by polishing and ion-bombarded single-crystal surfaces has been studied using X-ray photoelectron spectroscopy (XPS) both before and after annealing at 900 K (Onishi et al., 1987). It was observed that at room temperature water did not adsorb on the annealed surfaces. However, ion-bombarded surfaces did display features of OH2, presumably due to the creation of surface defect sites allowing water adsorption. At 200 K, both the polished and ion-bombarded surfaces showed evidence of OH2 after exposure to H2O, the ion-bombarded surface being covered with a monolayer of OH2 (Peng and Barteau, 1990, 1991). Like H2O, methanol (CH3OH) adsorbs only at defect sites on the (100) surface, undergoing heterolytic dissociation to methoxide (CH3O2) and OH2 (Onishi et al., 1987). Carboxylic Acids Both formic and acetic acid dissociate at room temperature and below on (100) terraces giving surface hydroxide and the formate (HCOO2) and acetate (CH3COO2) species (Peng and Barteau, 1991). Hydrogen Sulfide Hydrogen sulfide is similar to water and may adsorb either as a molecule or by proton loss (OH2 formation) through an acid/ base type interaction; see reaction (8.6): H 2 S þ O2

! H2 O þ S 2

(8:6)

BIBLIOGRAPHY Blesa, M. A., Morando, P. J., and Regazzoni, A. E. (1993). Chemical Dissolution of Metal Oxides. CRC Press: Boca Raton, FL. Carniglia, S. C., and Barna, G. L. (1992). Handbook of Industrial Refractory Technology: Principles, Types, Properties and Applications. Noyes: Park Ridge, NJ, p. 417. Henrich, V. E., and Cox, P. A. (1994). The Surface Science of Metal Oxides. Cambridge University Press: Cambridge, United Kingdom. Kingery, W. D. (1959). Property Measurement at High Temperature. Wiley: New York, p. 44. Kingery, W. D., Bowen, H. K., and Uhlmann D. R. (1976). Introduction to Ceramics. Wiley: New York, pp. 777, 791, 805. Samsonov, G. V., Ed. (1973). The Oxide Handbook. IFI/Plenum: New York. Vlack Van, L. H. (1964). Physical Ceramics for Engineers. Addison-Wesley: Reading, MA, Chapter 8, p. 118.

REFERENCES

131

REFERENCES Austin, J. B. (1931). J. Am. Ceram. Soc. 14, 795– 810. Chase, M. W., Ed. NIST-JANAF Thermochemical Tables (1998), 4th ed., ASC/AIP. Colburn, E. A., and Mackrodt, W. C. (1983). Solid State Ionics 8, 221. Coluccia, S., Segal, R. L., and Tench, A. J. (1979). J. Chem. Soc., Faraday Trans. I 75, 1769. Diepschlag, E., and Wulfesting, F. (1929). Iron and Steel Ind. No.3 (Sept.) 4. Ebert, H., and Tingwaldt, C. (1936). Physik. Z. 37, 471– 474. Foe¨x, M. (1942). Compt. Rend. 214, 665 –666. Francl, J., and Kingery, W. D. (1954). J. Am. Ceram. Soc. 37, 80 – 84. Fruhwirth, O., Herzog, G. W., Hollerer, I., and Rachetti, A. (1985). Surf. Tech. 24, 301 – 317. Hayes, W., and Stoneham, A. M. (1985). Defects and Defect Processes in Nonmetallic Solids. Wiley: New York. Lide, D. R., Ed. (1992). CRC Handbook of Physics and Chemistry, 73rd ed. CRC Press: Boca Raton, FL, pp. 9 –57. McQuarrie, M. (1954). J. Am. Ceram. Soc. 37, 84 –88. Moodie, E. F., and Warble, C. E. (1971). J. Cryst. Growth 10, 26. Ochs, D., Brause, M., Braun, B., Maus-Friedrichs, W., and Kempter, V. (1998). Surf. Sci. 397, 101 – 107. Onishi, H., Egawa, C., Aruga, T., and Iwasawa, Y. (1987). Surf. Sci. 191, 479– 491. Peng, X., and Barteau, M. A. (1990). Surf. Sci. 233, 283– 292. Peng, X., and Barteau, M. A. (1991). Langmuir 7, 1426 –1431. Rochow, E. G. (1938). J. Appl. Phys. 9, 664 –669. Segall, R. L., Smart, R. St. C., and Turner, P. S. (1978). J. Chem. Soc. Faraday Trans. I 74, 2907. Segall, R. L., Smart, R. St. C., and Turner, P. S. (1988). In Surface and Near-Surface Chemistry of Oxide Materials, Nowotny, J. and L.-C. Dufour, eds. Elsevier: Amsterdam. Segall, R. L., Jones, C. F., Smart, R. St. C., and Turner, P. S. (1981). Proc. Roy. Soc. London, A 374, 141. Segall, R. L., Jones, C. F., Smart, R. St. C., and Turner, P. S. (1984). J. Mater. Sci. Lett. 3, 810. Segall, R. L., Jones, C. F., Smart, R. St. C., and Turner, P. S. (1985). Radiat. Eff. 60, 167.

9 OTHER MAGNESIA PRODUCTS

9.1

PRODUCTION OF HARD-BURNED MAGNESIA

Hard-burn magnesia is normally produced in either a rotary or shaft kiln at temperatures between 2600 and 28008F. The feed to the kiln can be either raw magnesite or a magnesium hydroxide filter cake. Predensification steps are usually not employed. When hydroxide filter cake is used as the feedstock, chains and dams are used in the back end of the kiln. The dams retard the flow of the filter cake into the calcination zone allowing it to be dried with the exit gases. The chains, which flay around with the rotation of the kiln, help to break up any large agglomerates of dried filter cake before being calcined. The absence of these chains would result in large “boulders” rolling out of the exit of the kiln.

9.2

PRODUCTION OF DEAD-BURNED MAGNESIA

Dead-burned magnesia is used almost exclusively for refractory applications, primarily in brick and basic granular refractories. The essential qualities of dead-burned magnesite that make it suitable for refractory applications are that it has a very high melting point (28008C), it does not oxidize, and it has excellent resistance to attack by iron oxide, alkalis, and The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 133

134

OTHER MAGNESIA PRODUCTS

high lime fluxing agents. Its high density means that it has low porosity, which reduces slag penetration into the magnesia by capillary action. The lower quality grades have MgO contents ranging from 80 to 90% and bulk densities of 3.10–3.35 g cm23. The major impurities present are SiO2, Fe2O3, Al2O3, CaO, and B2O3. These grades are manufactured by feeding sized raw magnesite or magnesium hydroxide slurry directly to rotary kilns. Additives such as Fe2O3, SiO2, TiO2, and ZrSiO4 may be added to promote densification during the sintering process. The higher quality grades of dead-burned magnesia have MgO contents ranging 90–99% and bulk densities between 3.40 and 3.45 g cm23. The crystallites are larger and better developed than those of lower quality grades. The main impurities are the same type as for the lower dead-burned grades. These grades are produced usually by a double-burning process using initially a high-purity caustic-calcined MgO. The caustic-calcined magnesia is then pelletized under high pressure prior to firing in a rotary kiln at a temperature above 16508C or in a shaft kiln above 20008C. Figure 9.1 displays a diagram of a high-pressure pelletizing press.

Figure 9.1 Hosokawa Bepex MS 300 high-pressure briquetter. (Reproduced with permission of Hosokawa Bepex.)

9.2 PRODUCTION OF DEAD-BURNED MAGNESIA

135

The high boron content of seawater-grade magnesia (which can be as high as 0.3%), even if it is present in a few tenths of a percent, can be detrimental to the hot strength of magnesia refractories and markedly affect their service performance under extreme conditions (Davies and Havranek, 1965; Taylor et al., 1969). Tables 9.1 and 9.2 display the effect of firing temperature on density, crystal size, and hydration tendency of a low-boron magnesia and a seawater magnesia containing 0.2% B2O3. It can be seen that while both materials show an increase in density and crystal size with increased firing temperature, and also a decrease in hydration tendency, the B2O3 content has a marked affect on sintered density. B2O3 is known to exert a powerful fluxing action on the dicalcium silicate bonds that form during the dead burning of magnesia. It is these dicalcium silicate bonds that give deadburned (DB) magnesite its strength. In the system MgO/B2O3 three compounds were identified; MgO.B2O3, 2MgO.B2O3 (which has a melting point of 13408C), and 3MgO.B2O3 having a melting point of 13508C (Davis and Knight, 1945). This is why it is important to reduce the boron content of the magnesium hydroxide filter cake before firing. B2O3 contamination tends to volatilize to some extent at the temperatures used for producing dead-burned product, and magnesia with a B2O3 content in the range of 0.1– 0.2% can readily be produced. Increased volatilization of boron during firing can be achieved by the addition of alkali, such as caustic soda or sodium carbonate to the magnesium hydroxide filter cake prior to dead burning (Gilpin et al., 1967, 1971). It is possible to achieve B2O3 levels as low as 0.01% in the final product by this method. Another important criterion in the manufacture of high-purity DB MgO is the CaO/SiO2 ratio. It is desirable to have a lime/silica ratio .2 : 1 as this reaction leads to the formation of dicalcium silicate bonds; see Figure 9.2.

TABLE 9.1 Effect of Firing Temerature on the Densification, Crystal Size, and Hydration Tendency of 0.2% B2O3 Containing Magnesia Grain Firing Temp. (8C) 1500 1550 1600 1650 1700 1750

Density (g/cm3)

Crystallite Size (mm)

ASTM Hydration Tendency (%) (1 h at Temp.)

2.96 3.00 3.05 3.09 3.13 3.18

23 25 27 29 31 32

98 80 60 40 20 0

Source: Data adapted from Chesters (1973).

136

OTHER MAGNESIA PRODUCTS

TABLE 9.2 Effect of Firing Temperature on the Densification, Crystal Size, and Hydration of Low B2O3 (0.03%) Pelletized Magnesia Firing Temp (8C) Tendency (%) (1 h at Temp.)

Density (g/cm3)

Crystal Size (mm)

ASTM Hydration

3.11 3.19 3.26 3.32 3.37 3.40

14 17 23 26 30 40

91 68 41 27 21 16

1600 1650 1700 1750 1800 1850 Source: Data adapted from Chesters (1973).

Figure 9.2 Phase diagram for the binary system CaO – SiO2. [Reprinted from B. Phillips and A. Muan, J. Am. Ceram. Soc. 42(9) 414 (1959). Reprinted with permission of the American Ceramic Society, www.ceramics.org. Copyright 1964. All rights reserved.]

9.2 PRODUCTION OF DEAD-BURNED MAGNESIA

137

In general the properties of particular dead-burned magnesia will depend largely on the amount and distribution of the impurities present. Table 9.3 displays the effect of the CaO/SiO2 ratio in the magnesia on the mineral phases formed and their melting temperatures. Figure 9.3 shows that the microstructure of periclase is affected by the CaO/SiO2 ratio. Figure 9.3a is a dead-burn 95% MgO, while Figure 9.3b is a lower purity 87% dead-burn MgO. In this figure the light gray crystals are MgO and the silicate location (darker gray areas) is indicated. As the CaO/SiO2 ratio increases not only are different silicate phases formed, but they tend to concentrate into pockets rather than coating the MgO crystals. The consequence of this is that more direct bonding of the magnesia crystals can occur with a resultant increase in grain strength. The formation of fosterite is an important reaction that dead-burned magnesia undergoes in magnesia-based refractories, and it has been studied extensively by chemists and ceramists (Bowen and Andersen, 1914; Grieg, 1927). The MgO–SiO2 phase diagram (see Fig. 9.4), shows that refractories containing magnesia with silica are very stable with respect to high temperatures, when their composition corresponds to the field between fosterite and periclase. The purer grades of seawater magnesia have proved more difficult to densify during calcination owing to the lower concentration of mineralizing impurities. To overcome this difficulty, high-pressure pelletization of caustic-calcined magnesia is used. It has been demonstrated that there is an optimum calcination temperature for maximum fired density after TABLE 9.3 Mineral Phases in Equilibrium with MgO in the MgO– CaO – SiO2 System Weight Ratio CaO/SiO2 Less than 0.93 0.93 0.93– 1.40 1.40 1.40– 1.86 1.86 1.86– 2.80 2.80 More than 2.80

Minerals Present

Composition

Approx. Melting Temp. (8F)

Forsterite Monticellite Monticellite Monticellite Merwinite Merwinite Merwinite Dicalcium silicate Dicalcium silicate Dicalcium silicate Tricalcium silicate Tricalcium silicate Tricalcium silicate Lime

2MgO.SiO2 CaO.MgO.SiO2 CaO.MgO.SiO2 CaO.MgO.SiO2 3CaO.MgO.2SiO2 3CaO.MgO.2SiO2 3CaO.MgO.2SiO2 2CaO.SiO2 2CaO.SiO2 2CaO.SiO2 3CaO.SiO2 3CaO.SiO2 3CaO.SiO2 CaO

3450 2710 2710 2710 2870 2870 2870 3865 3865 3865 3450 3450 3450 4650

Source: Data Adapted from Alper (1970).

138

OTHER MAGNESIA PRODUCTS

Figure 9.3 (a) Micrograph of a dead-burn 95% MgO. (b) Micrograph of a dead-burn 87% MgO. Light gray crystals are MgO; darker gray areas are silicate phases.

9.2 PRODUCTION OF DEAD-BURNED MAGNESIA

139

Figure 9.4 Phase diagram of the binary system MgO– SiO2. [Reprinted from Grieg (1927). Reprinted with permission of the American Ceramic Society, www.ceramics.org. Copyright 1964. All rights reserved.]

compaction (Livey et al., 1957). Examination of active magnesia by X-ray line broadening and specific surface area measurements by nitrogen adsorption have shown that the “active” magnesia particles consist of porous aggregates of much smaller crystallites, which decreased in size with decreasing calcination temperature (Baker and Thomas, 1956). Optimum densification has been found to occur when the crystallites are of uniform size and the crystallite aggregates fairly weakly bound. This allows them to collapse during compaction, thus minimizing intra- and interparticle voids. 9.2.1 Sintering Sintering is essentially a process in which fine particles, which are in contact with each other, agglomerate when heated to a suitable temperature roughly one-half to two-thirds of the melting temperature. At this temperature, the crystallite faces display disorder, which enables rapid surface diffusion. The body of the powder remains solid while the surfaces become “slippery.” This agglomeration is accompanied by a decrease in the porosity and an increase in the bulk density of the mass. Simultaneously, there is a reduction in surface area and surface free energy, and hence a reduction in the total free energy of the system itself. During normal grain growth, some of the grains grow larger while other grains disappear, but the range of grain sizes remains in a normal distribution

140

OTHER MAGNESIA PRODUCTS

as the average grain size increases with time. The initial grain growth is rapid; however, as grain size increases over time the rate of grain growth slows. The main driving force for grain growth is the reduction in surface free energy as the grains grow larger since the interfacial free energy decreases during this process. The characteristic phenomena that occur during sintering can be classified into three phases. See Figure 9.5 for a schematic diagram of this process. During the initial phase of sintering, the main feature is an increase in interparticle contact area through neck formation as time progresses. This is accompanied by a rounding-off of the sharp contact areas formed at the points of contact. Mass transport is by surface diffusion and possible gasphase evaporation and condensation; see Figure 9.6. The surface atoms, which tend to be very mobile, are trapped at the necks between the particles because of the difference in average atomic coordination number between convex and concave surfaces. The average coordination number for a concave surface is slightly higher than a convex surface, so the atoms tend to accumulate at the concave surface, which results in continued neck growth. The kinetics of neck growth is modeled by the following

Figure 9.5

Particle sintering pathway.

9.2 PRODUCTION OF DEAD-BURNED MAGNESIA

Figure 9.6

141

Transport pathways during grain sintering. [After Van Der Put (1998), Figure 5.6.]

expression (Van der Put, 1998): (x=r)n ¼ (B=r m )t

(9:1)

The relative neck size is defined by x/r, where x is the diameter of the neck and r is the diameter of the particles. The values for n, m, and B are known for different sinter mechanisms. The linear shrinkage DL/L at this stage follows the kinetic law: (DL=L0 )n=2 ¼ (B=2n rm )t

(9:2)

in which n has a value between 2.5 and 3. In this expression, B depends exponentially on the temperature: B ¼ B0 exp½ Q=kTŠ

(9:3)

As the sintering process continues, the growing necks merge and the so-called intermediate stage of sintering begins. In this stage, the total surface area and porosity decrease along with a decrease in the distance between grain centers, and shrinkage occurs because of bulk diffusion. Atoms present at the grain boundaries diffuse to the concave sites on the surface. The increase in density at this juncture follows logarithmic time dependence; see Equation (9.4):

r(t) ¼ ri þ B ln (t=ti )

(9:4)

where rt is the sintered fractional density at time t . ti, ti is the time of the onset of the second stage, ri is the green density, and B is a

142

OTHER MAGNESIA PRODUCTS

temperature-dependent parameter as in the first stage of sintering. The average size of grain radius R increases with time according to: kRln

kRln0 ¼ kt

(9:5)

where k is a temperature-dependent parameter, R is the average grain radius at time t, kRl0 is the initial value of kRl at t ¼ 0, and n is the grain growth exponent. Numerous studies of normal grain growth using different growth models have predicted an n value of 2. At this point in time the pore structure is still interconnected; however, the exposed surface area is now less than half of the initial surface area of the powder. During the final stage of sintering, the porosity begins to break up and become enclosed bubbles. The growth of the grains depends upon the individual sizes. The large ones tend to grow at the expense of the smaller ones, which shrink and disappear. Grain growth at this stage, as expressed as a function of the radius r, increases as a function of the following expression: r(t)

r0 ¼ (2k)1=2 tm

(9:6)

where m is usually in the range 0.1– 0.5. Maximum densification of the sintered body occurs when the pores are able to move with the grain boundaries during grain growth. As long as the pores are on the boundary, they are able to shrink by diffusion of atoms through the attached boundaries. If grain growth occurs too rapidly, the fast moving grain boundaries tend to leave the pores behind in the interior of the grain, which is deleterious to shrinkage of pores. In this situation, pore shrinkage has to occur by bulk diffusion, which is a much slower process than boundary diffusion. Sintering conditions are adapted to the grain size of the powder to enhance densification. During the sintering process, grain growth should be restrained as long as large pores are present. When these pores have shrunk sufficiently, some growth in grain size can be tolerated as long as the pores remain attached to the moving grain boundaries.

9.2.2 Sinter Aids Additions of small quantities of additives are often used to improve the densification of a powder compact during sintering. Table 9.4 shows the

9.2 PRODUCTION OF DEAD-BURNED MAGNESIA

TABLE 9.4

143

Effect of Metal Oxide Impurities on the Densification of MgO

Sintering Temperature (8C) Ionic Impurity Valency Radius (nm) 1225– 1720 1700– 1800 1300– 1500 1500– 1600 1300 MgO Li2O Na2O CaO BaO ZnO FeO B2O3 Al2O3 Fe2O3 Cr2O3 Mn2O3 SiO2 MnO2 TiO2

þ2 þ1 þ1 þ2 þ2 þ2 þ2 þ3 þ3 þ3 þ3 þ3 þ4 þ4 þ4

0.066 0.068 0.097 0.099 0.134 0.074 0.074 0.023 0.051 0.064 0.063 0.066 0.042 0.060 0.068

þ 2 2 2

þ/ 2 2

þ þ þ

þ

2 þ

þ/2 þ þ/2

2 þ 2

þ þ

þ þ þ þ

þ/2 þ þ

þ/2 þ þ

þ

þ/2

þ

þ

þ indicates a positive effect on densification; a 2 indicates a negative effect, and a +, effect is dependent upon impurity concentration. Source: Data adapted from Green (1983).

effect that metal oxide additives have on the sintering of MgO at a variety of different temperatures. There are five main categories of sinter aids: 1. Sinter aids that temporarily form a small liquid phase at the sintering temperature by reacting with the major phase. This is termed liquid-phase sintering. A number of processes occur during liquidphase sintering; viscous creep of the liquid between the particles due to capillary forces induces shrinkage and pore filling. The grains may be attacked and solubilized by the liquid phase, for example, molten salts, which are very good solvents. The grains may also grow by ripening, which consumes the smaller particles and results in the growth of the larger ones. Examples of this type of sinter aid are sodium chloride and lithium fluoride. 2. Inhibition of grain boundary motion by the formation of a second phase at the grain boundaries. 3. Those that affect diffusivity at grain boundaries by modification of boundary defects. 4. Inhibition of crystallite growth by blocking growth sites. Examples of this type of sinter aid are Cr2O3 and Fe2O3.

144

OTHER MAGNESIA PRODUCTS

5. Sinter aids that inhibit grain growth by enriching grain surfaces with very small particles. This process lowers the surface energy and removes the driving force for physical sintering. 9.2.3 Production Methods Dead-burned magnesite is produced by the predensification of light-burned magnesia, followed by firing in either a rotary or shaft kiln. Raw magnesite and magnesium hydroxide are not suitable for direct densification since they both have poor briquetting characteristics, which lead to inadequate green density. Prefired densities should be above 2.0 g/cm3 to optimize fired density. Predensification is achieved using briquetting presses exerting a pressure in the range of 21,000 (144 MPa) to 50,000 psi (344 MPa). Dependent upon the characteristics of the light-burned magnesia, the briquettes may require a second-stage densification step to obtain maximum prefired or green density. The briquettes are then screened to remove broken or improperly formed pieces, which are then returned to a mill for grinding and reintroduction into the material feeding the presses. Correctly formed briquettes are stored in a bin before firing. Briquettes or pellets are then fed into the kiln using a variable-speed feeder where they experience temperatures of 3100–33008F in a rotary kiln and temperatures up to 38008F in a shaft kiln.

9.3 FUSED MAGNESIA Fused magnesia is manufactured using three-phase electric arc melting, typically by way of a Higgins furnace. Raw material feed can consist of causticcalcined magnesia, dead-burned magnesia, or magnesite. The temperatures generated in these electric arc furnaces (EAF) are in excess of 27508C and typically 12 h are required to complete the fusion process. The fusion process is very energy intensive, and consumption of electricity varies between 3500 and 4500 kWh/tonne. Fusion promotes the growth of very large crystals of periclase, in excess of 1000 mm, compared with deadburned magnesia, which has a crystal size in the range of 50–100 mm. Periclase density approaching the theoretical maximum of 3.58 g cm23 is also achieved. During the fusion process the silicate impurities tend to migrate to the surface of the magnesia ingots, where upon cooling, they can be removed by decrusting and crushing. Fused magnesia is superior to deadburned magnesia in strength, abrasion resistance, and chemical stability. Major applications for fused magnesia are essentially divided between two markets: magnesia-based refractories and the electrical insulating market

9.3 FUSED MAGNESIA

145

for heating elements. Some fused magnesia is now being used in friction materials such as a replacement for asbestos in brake pads and disks. 9.3.1 Refractory-Grade Fused Magnesia The incorporation of fused magnesia grains greatly enhances the performance and durability of basic refractories, such as magcarbon brick. This is a function of a higher specific gravity, large crystal size, and low porosity, which all help to slow the dissolution of magnesia by high-temperature slag. Refractory-fused magnesia has the following requirements: . . . .

High magnesia content, ranging from 96 to 99% MgO. Low silica and silica-to-lime ratio of 2 : 1. A specific gravity of 3.50 g cm23 or greater. Large crystal sizes of 1000 mm or greater; see Figure 9.7.

Due to its excellent corrosion resistance, refractory-grade fused magnesia is used in high wear areas in steelmaking, such as basic oxygen and electric arc

Figure 9.7 Micrograph of fused magnesia.

146

OTHER MAGNESIA PRODUCTS

furnaces, converters, and ladles. Very high purity grades also have application in optical equipment, nuclear reactors, and rocket nozzles. Fused magnesia for the electrical market is used mainly as a temperatureresistant electrical insulating material in ceramic sheaths for heating elements. This grade of fused magnesia has the following characteristics: low levels of boron, sulfur, iron, and trace metals and a lime–silica ratio of 1 : 2. Three grades of fused magnesia are generally produced. A hightemperature grade of fused magnesia for temperatures up to and in excess of 9508C. This application requires an MgO purity level of between 94 and 97%. A medium-temperature grade for temperatures up to 8008C requires a minimum MgO purity of 93%. Low-temperature application (up to 6008C) can utilize magnesia with a purity of less than 90%. All of these grades require high electrical resistivity and thermal conductivity, and these properties are measured routinely.

9.4 MAGNESIUM HYDROXIDE SLURRY Magnesium hydroxide slurry consists of an aqueous suspension of particulate magnesium hydroxide. The principle sources of slurry are from seawater- or brine-produced magnesium hydroxide, natural brucite, or from the slaking of magnesium oxide powder. Magnesium hydroxide slurry is gaining in popularity as a replacement for caustic soda and lime in wastewater treatment applications. The solids content of the slurry can range from 50 to 65% solids and have a viscosity in the range of 100–800 cP. To produce a marketable product, the magnesium hydroxide suspension has to have a high solids loading (preferably above 50%) and be reasonably stable in order to prevent the premature settling of solids. Unless the suspension is stable, the solids will have a tendency to settle to the bottom of the transportation vessel or storage vessel. If such settling does occur, it is also desirable that the settled solids can be easily resuspended and do not form a hard cake at the container bottom. The rheology of the slurry should be such that it is readily pumped using either centrifugal, peristaltic, or progressive cavity pumps. Much research and development effort has been expended by many companies seeking the “magic” formula that produces a stable suspension with all the desired rheological characteristics. 9.4.1 Production of Synthetic Magnesium Hydroxide Slurry Synthetic magnesium hydroxide slurry is produced using either seawater- or brine-precipitated magnesium hydroxide filter cake that has the desired solids content of the finished slurry. To produce pumpable slurry, the

9.4 MAGNESIUM HYDROXIDE SLURRY

147

addition of a dispersant is typically required to lower slurry viscosity. Seawater-precipitated hydroxide requires the use of an anionic dispersant, such as a sodium polycarboxylate (Zertuche-Rodriguez et al., 1998; Falc¸ione et al., 1980; Zupanovich et al., 1983), while brine-precipitated hydroxide requires cationic dispersants (Richmond and Gutowski, 1996; Rey, 1984). The difference in the charge of the dispersant used in a seawater magnesium hydroxide suspension versus that used for a brine-produced slurry is a reflection of the surface charge of the hydroxide particle in contact with water. Seawater hydroxides tend to develop a positive surface charge, while brine products tend to produce a negatively charged surface. To reduce the viscosities of these slurries, the dispersant has a charge opposite to that of the particle surface and the dispersant is adsorbed onto the surface by electrostatic attraction. Since the adsorbed dispersant is either a polyanion or a polycation, it has a charge in excess of that required to neutralize the particle surface charge. This residual charge imparts an overall negative or positive charge to the magnesium hydroxide particle surface. The individual particles then undergo electrostatic repulsion, which minimizes their interaction with one another, thus reducing “interparticle friction” and hence slurry viscosity (Conley, 1996). Methods of stabilizing the suspension from settling of solids have spawned numerous patents on the subject. These methods of stabilization include using a suspending or thickening additive in the slurry, particle size reduction, or a combination of the two. Many thickeners and suspending agents have been utilized in magnesium hydroxide slurries, such as sodium caboxymethylcellulose, hydroxyethyl cellulose, clays, and xanthan gums (Pomrink and Fillipo, 1998). However, many of these additives are costly and are difficult to disperse properly into the slurry. Milling is another commonly used method of stabilizing a suspension. Many different types of mills can be used for this process, such as ball, Sweco, Vibro-Energy, and Palla mills. Stirred media mills, also commonly known as attritors, can also be used to reduce the slurry particle size. These mills are employed in the paint industry where effective milling is critical to producing a stable paint. There are two basic types of media-stirred mills—vertical and horizontal; see Figures 9.8 and 9.9, respectively. Vertical stirred media mills consist of a cylindrical shell with a central shaft onto which are attached agitator disks or arms. When the central shaft is rotated, typically at 100– 1500 rpm, these disks or arms move and agitate the grinding media. The grinding media ranges from 3– 13 mm in diameter and can be composed of steel or a ceramic (Patton, 1979). Horizontal media mills consist of a horizontal cylindrical shell, again with a central agitation shaft. These mills operate at higher rpms than vertical mills, up to 3000 rpm, and can use much smaller grinding media (as small as 0.4 mm) and also operate at

148

OTHER MAGNESIA PRODUCTS

Figure 9.8

Vertical stirred media mill. (Reproduced with permission of Union Process, Inc.)

elevated pressures. These factors combined lead to a much greater grinding efficiency than the vertical mill variety. During the milling process, a rise in viscosity is experienced due to the generation of freshly exposed surface and also to particle breakages, which results in an increase in the number of particles per unit volume. To control viscosity increase, additional dispersant is added during the milling cycle. 9.4.2 Hydration of Magnesium Oxide Another common method of producing magnesium hydroxide slurry is by reacting magnesium oxide in water in excess of that require for theoretical hydration: MgO þ H2 O

! Mg(OH)2 þ Heat

(9:7)

9.4 MAGNESIUM HYDROXIDE SLURRY

Figure 9.9

149

Horizontal stirred media mill. (Reproduced with permission of Union Process, Inc.)

The hydration reaction is exothermic, producing about 9 kcal/mol in contrast to the hydration of lime, which produces 15 kcal/mol. This heat is actually beneficial to the hydration since it heats the water/oxide slurry accelerating the rate of hydration. Some recovery of heat can be gained from the reactor during the cooling phase after the hydration is complete by circulating cold water through a cooling jacket of a jacketed reactor vessel. This warmed water can then be stored and used in subsequent hydrations. The reaction can be carried out in large, well-agitated tanks that are open to the atmosphere to allow the escape of steam and mitigate any pressure buildup. The MgO and water are mixed together in the reactor and allowed to boil. Atmospheric pressure reactions are best suited to a lightly calcined MgO with a specific surface area of about 20–40 m2 g21 and having a narrow range of activity. It is important to use magnesium oxide with a narrow range of activity since this will allow the majority of the MgO to be hydrated at the same rate. Hydration of less reactive oxides, or those with a less homogenous activity, results in differing rates of hydration between the more active oxide fraction, which hydrate first, and the less active fraction. This situation, where the less active fractions continue to slowly hydrate over a longer period of time, can cause posthydration effects, which often result in a large upward creep in viscosity. Less active magnesium oxides, or those with a wide range of activities, are more suited to being hydrated above atmospheric pressure (25–150 psi, temperature .2128F) using a pressure hydration vessel (Witkowski et al., 1996).

150

OTHER MAGNESIA PRODUCTS

Figure 9.10

Schematic diagram of magnesium oxide slurrying plant.

A schematic of the basic setup of a hydration plant is shown in Figure 9.10. Water is first added to the reactor, then magnesium oxide powder to produce a water –oxide slurry. Any dispersants or additives required to maintain control of viscosity during the hydration phase can be added at this point. Good agitation is maintained to achieve a thorough wetting of the oxide powder. The slurry is allowed to come to a boil and then cooled below 1808F before being pumped out of the reactor. At this point the solids content of the slurry should be checked. The slurry is pumped into a premill tank and can then be milled using any of the mills mentioned previously. After the milling stage, the magnesium hydroxide slurry is stored in agitated tanks prior to shipment. At this final stage, testing should be conducted to ensure that the viscosity and solids content are in the desired range. 9.4.3 Hydration Kinetics and Mechanisms Smithson and Bakhshi (1969) investigated the kinetics of hydration of MgO and found that the rate of hydration is directly proportional to the surface area contained in a shell at the surface of the particles and that the activation energy for this reaction is 14.1 kcal/mol. They concluded that the value of the activation energy indicated that the reaction rate is limited by the chemical reaction. Their proposed reaction mechanism involved the reaction of

9.5 PURIFICATION BY CARBONATION OF MAGNESIUM HYDROXIDE SLURRY

151

water molecules with solid MgO to form an unstable surface layer of magnesium hydroxide, which subsequently dissolves exposing a fresh oxide layer thus allowing further hydration. The rate-determining step is not the hydration reaction, but rather the slow removal of the Mg(OH)2 layer as slow dissolution of the surface Mg(OH)2 and the diffusion of the resulting ions Mg2þ and OH2 into solution already saturated, or nearly so, with the same ions. This mechanism has been supported by other workers (Khangaonkar et al., 1990; Fruhwirth et al., 1985). In contrast to his work on the hydration of MgO Glasson (1963) concluded that the reaction mechanism was one of an advancing hydration interface into the core of the particle. 9.4.4 Testing and Quality Control of Magnesium Hydroxide Slurry The minimum basic quality control (QC) testing procedures that should be conducted on the finished magnesium hydroxide are as follows: .

.

Viscosity There are many different methods of determining the viscosity of a fluid, however, the quickest method for running routine QC checks is to use a Brookfield rotational viscometer; either the RV or LV models are suited to the range of viscosities normally encountered for magnesium hydroxide slurry. Solids Content The solids content can be checked by determining the loss on drying (LOD): %Solids ¼ 100

.

%LOD

Alternatively, a precalibrated density cup (lb/gal or g/L) can also be used to determine solids content. Particle Sizing If a milling step is employed in slurry production, then a check on the milling efficiency (to ensure the mill is operating optimally) should be employed. This can be as simple as determining a wet sieve analysis (325 mesh, U.S. Standard Sieves) or as sophisticated as using an instrumental method, such a laser scattering or X-ray Sedigraph.

9.5 PURIFICATION BY CARBONATION OF MAGNESIUM HYDROXIDE SLURRY The carbonation of magnesia slurries has been extensively studied for the best way to leach magnesia from low-grade magnesite and dolomite, as well as producing very high purity MgO from higher grade magnesia

152

OTHER MAGNESIA PRODUCTS

(Doerner et al., 1946; Horiguchi, 1952; Horiguchi and Atoda, 1951a, 1951b). The advantage of using CO2 leaching as opposed to common acids, such as sulfuric and hydrochloric acids, is that CO2 is a more selective leaching agent that reduces the overall amount of impurities in the magnesium bicarbonate solution. Reactions (9.8) to (9.11) are the sequence of reactions occurring in this process (Ferna´ndez et al., 1999): MgO þ H2 O

! Mg(OH)2

Mg(OH)2 þ 2CO2

! Mg2þ þ 2HCO3

(9:8) (9:9)

MgO þ CO2 þ xH2 O

! MgCO3
xH2 O

(9:10)

MgCO3
xH2 O þ CO2

! Mg2þ þ 2HCO3

(9:11)

Calcium does not readily form the bicarbonate but instead remains as the insoluble carbonate. Iron is also solubilized to some degree by carbon dioxide, which is represented by reaction (9.12) (Canterford et al., 1983): Fe2 O3 þ 6CO2 þ 3H2 O

! 2Fe3þ þ 6HCO3

(9:12)

Other impurities remain unaffected by carbon dioxide and can be filtered off leaving a purified magnesium bicarbonate solution. The parameters that control the leaching process are as follows: temperature, carbon dioxide pressure, agitation, slurry density, particle size, and leachant composition. In general the effects of these parameters on a slurry of Mg(OH)2 are as follows: . .

. .

.

An increase in CO2 pressure increases the rate of solution of Mg(OH)2. Temperature has a great influence on the initial reaction rate. Under conditions of low slurry density, where the concentration of magnesium bicarbonate does not reach its solubility limit, the rate of dissolution of MgO increases with increasing temperature. However, under conditions where the concentration of magnesium bicarbonate reaches its metastable limit, the maximum solution of Mg(OH)2 is reached at about 158C, after which magnesium carbonate starts to precipitate. The metastable limit for slurry density is about 10 g/L MgO in solution. Both the rate of dissolution of particulate MgO and the absorption rate of carbon dioxide in water are influenced by the degree of agitation. The maximum rate of dissolution is observed at an agitator speed of 1000 min21. The smaller the MgO particle size, the greater the rate of dissolution.

REFERENCES

153

9.5.1 Precipitation of Magnesium Carbonate from Bicarbonate Solution The precipitation of hydrated magnesium carbonate, such as hydromagnesite, from the filtered leachate, can be achieved using air sparging and/or heating to remove excess carbon dioxide. Since magnesium bicarbonate is not stable in the absence of dissolved carbon dioxide, the following precipitation occurs at temperatures above 558C; see reaction (9.13): 5Mg(HCO3 )2 ! Mg5 (CO3 )4 (OH)2
4H2 O þ 6CO2 "

(9:13)

At temperatures below 208C, and using no air sparging to assist in removing CO2 from solution, results in extremely long precipitation times; even after 144 h, 60% of magnesium still remains in solution (Canterford and Moorrees, 1985). At 458C, and using air sparging, 95% of magnesium is precipitated from solution after 2 h. Boiling the solution speeds up precipitation; however, it still takes 1 h to remove the final 2% of soluble Mg from solution. The form of magnesium carbonate precipitated from solution is dependent upon temperature, and the following transformation between the forms of magnesium carbonate can occur (Langmuir, 1965; Davies and Bubela, 1973); see reaction (9.14): 108C

558C

(MgCO3
3H2 O)

(Mg5 (CO3 )4 (OH)2
4H2 O)

Lansfordite ! Nesquehonite ! Hydromagnesite MgCO3
5H2 O

(9:14)

High-purity MgO (.99%) can be reclaimed from the carbonates by calcination, liberating carbon dioxide and water in the process. REFERENCES Alper, A. M. (1970). High Temperature Oxides, Part I. Academic: New York, p. 13. Baker, T. W., and Thomas, D. K. (1956). Proc. Phys. Soc. 74, 673. Bowen, N. L., and Andersen, O. (1914). Am. J. Sci. 37(4), 488. Canterford, J. H., and Moorrees, C. (1985). Bull. Proc. Austras. Inst. Min. Metall. 290, 67 –70. Canterford, J. H, Everson, P. T., and Moorrees, C. (1983). Proc. Australas. Inst. Min. Metall. 288, 29 –36, December. Chesters, J. H. (1973). Refractories Production and Properties. Iron and Steel Institute: London, pp. 144– 145. Conley, R. F. (1996). Practical Dispersion. John Wiley and Sons, New York. Davies, B., and Havranek, P. H. (1965). U.S. Patent 3,275,461.

154

OTHER MAGNESIA PRODUCTS

Davies, P. J., and Bubela, B. (1973). Chem. Geol. 12, 289 – 300. Davis, H. M., and Knight, M. A. (1945). J. Am. Ceram. Soc. 28, 100. Doerner, H. A., Holbrook, W. F., and Fortner, O. W. (1946). U.S. Bur. Mines Tech. Paper 648. Falc¸ione, R. J., McManis, R. R., and Aufman, J. A. (1980). U.S. Patent 4,230,610. Ferna´ndez, A. I., Chimenos, J. M., Segarra, M., Ferna´ndez, M. A., and Espiell, F. (1999). Hydrometallurgy 53, 155– 167. Fruhwirth, O., Herzog, G. W., Hollerer, I., and Rachetti, A. (1985). Surface Tech. 24, 301 – 317. Gilpin, W. C., Lythe, T. W., and Woodhouse, D. (1967). Brit. Patents 1,085,841 and 1,115,386. Gilpin, W. C., Lythe, G. C. P., and Woodhouse, D. (1971). Brit. Patent 1,115,386. Glasson, D. R. (1963). J. App. Chem. 13, 119– 124. Green, J. (1983). J. Mater. Sci. 18, 637– 651. Grieg, J. W. (1927). Am. J. Sci. 13(5), 133 –154. Horiguchi, Y. (1952). J. Sci. Res. Inst. (Tokyo) 46, 258– 267. Horiguchi, Y., and Atoda, T. (1951a). J. Sci. Res. Inst. (Tokyo) 45, 144– 153. Horiguchi, Y., and Atoda, T. (1951b), J. Sci. Res. Inst. (Tokyo) 45, 193 –198. Khangaonkar, P. R., Othman, R., and Ranjitham, M. (1990). Minerals Eng. 3, 227– 235. Langmuir, D. (1965). J. Geol. 73, 730 –754. Livey, D. T., Wanklin, B. N., Hewitt, M., and Murray, P. (1957). Trans. Brit. Ceram. Soc. 56, 217. Patton, T. C. (1979). Paint Flow and Pigment Dispersion, 2nd ed., Wiley: New York. Pomrink, G. J., and Fillipo, B. K. (1998). U.S. Patent 5,788,885. Rey, P. R. (1984). U.S. Patent 4,430,248. Richmond, A., and Gutowski, R. J. (1996). U.S. Patent 5,514,357. Smithson, G. L., and Bakhshi, N. N. (1969). Can. J. Chem. Eng. 43, 508. Taylor, M. I., Ford, W. F., and White, J. (1969). Trans. Brit. Ceram. Soc. 68, 173. Van der Put, P. J. (1998). The Inorganic Chemistry of Materials. Plenum: New York. Witkowski, J. T., Smith, D. M., and Wajer, M. T. (1996). U.S. Patent 5,487,879. Zertuche-Rodriguez, C. E., Benavides-Perez, R., Bocanegra-Rojas, J. G., and Santoy-Alvarez, G. A. (1998). U.S. Patent 5,811,069. Zupanovich, J. D., Myers, J. G., and Walker, J. L. (1983). U.S. Patent 4,375,526.

10 WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

10.1 INTRODUCTION TO APPLICATIONS Apart from lime, there are not many other products that have such a diverse range of uses as magnesia. It is utilized in many industries ranging from refractory applications, building construction as a cementitious material, in agriculture as a plant nutrient, to wastewater treatment as an alkaline source to neutralize acidic wastewater. It is also utilized in the manufacture of catalysts, rubber, and plastics and is also used for flue gas desulfurization for air pollution control. In some applications, such as in the manufacture of neoprene rubber, it is a small component of the overall batch size. However, despite this, it does plays a vital role in preventing premature curing of the rubber; and, if the MgO is not produced to exacting specifications, this can result in substantial losses to the rubber compounder owing to the premature cure of the neoprene before being able to process it correctly.

10.2 INDUSTRIAL WASTEWATER TREATMENT In the United States, many industrial facilities are required to pretreat wastewater before discharge to publically owned treatment works (POTW) or to surface waters. Discharge pH is regulated to be between 6.0 and 9.0; so The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 155

156

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

wastewater that falls outside this range must be adjusted using either an alkali or acid. If the wastewater also contains any regulated dissolved metals (among other pollutants), such as lead and cadmium, these must also be removed before the water can be discharged since these metals are toxic and may have adverse effects on any biological treatment performed at the POTW. The most common method for removing metals from solution is by precipitating the metal as its hydroxide. Typically, this was achieved almost exclusively by the application of either caustic soda or lime to the wastewater until the pH reached the minimum solubility of the metal; see reaction (10.1): Mnþ þ nOH

! M(OH)n #

(10:1)

When this pH is reached, small particles of the metal hydroxide are formed. Figure 10.1 displays graphically the solubility of various metal ions as a function of pH. Once the metal hydroxides have precipitated, the small particles must be coagulated, or flocculated, into larger heavier particles so that they settle more readily. Although metal hydroxide precipitates often flocculate by themselves, there are cases where the flocculated particles do not completely settle, and the addition of polymer-based flocculants may be necessary. Separation of the solids from the liquid phase is then carried out in a sedimentation pond or solids contact clarifier, plate settler, or filter. After the solids have been separated from the liquid, the sludge, which still contains a large quantity of water, is transferred to a drying bed, vacuum filter, or filter press for dewatering. This sludge can now be disposed of in an appropriate landfill.

Figure 10.1

Solubility of various metal hydroxides as a function of pH.

10.3 ADVANTAGES OF MAGNESIUM HYDROXIDE

157

10.3 ADVANTAGES OF MAGNESIUM HYDROXIDE IN WASTEWATER TREATMENT While caustic soda and lime have traditionally been used for many years to achieve the above precipitation reaction, magnesium hydroxide is gaining popularity as a replacement for these two alkalis. Magnesium hydroxide has a number of advantages over lime and caustic soda, and these are further discussed below. 10.3.1 Safety Magnesium hydroxide can be regarded as an industrial grade of “milk of magnesia.” It will not cause chemical burns and is essentially nonhazardous and nontoxic. Lime and caustic soda, if contacted with the eyes, however, can cause permanent damage. Lime dust can cause severe irritation to the nose and throat, and the breathing of caustic soda mist may cause damage to the upper respiratory tract and lungs. 10.3.2 pH Control Magnesium hydroxide produces a much lower pH spike when added to wastewater. This phenomena is due to the lower solubility of magnesium hydroxide; see Table 10.1, comparing the physical properties of magnesium hydroxide with caustic soda and lime. Due to the limited solubility of magnesium hydroxide in water, the maximum pH obtainable in any reasonable overdosing situation is approximately 9; this also happens to coincide with the Clean Water Act maximum discharge pH. This limited solubility also results in a more gradual increase in pH; see Figure 10.2, which displays pH versus time curves for neutralization of acid using magnesium hydroxide, lime, and caustic soda. Overdosing with lime and caustic soda can lead to the wastewater reaching pH

TABLE 10.1 Comparison of Properties of Magnesium Hydroxide with Calcium and Sodium Hydroxide Property Percent hydroxide (dry basis) Solubility in water, g/100 mL Reactive pH Freezing point, 8F Dry basis neutralization factor compared to Mg(OH)2

50% NaOH

30% Ca(OH)2

63% Mg(OH)2

42.5 42 14 57 1.37

45.9 0.185 12.5 32 1.27

58.3 0.0009 9.0 32 1.0

158

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

Figure 10.2 Neutralization curves for magnesium, calcium, and sodium hydroxides in strong acid.

12 and 14, respectively. If this occurs, then an acid must be added to reduce the pH back down to between 6 and 9. If either lime or caustic soda is being used to control pH and alkalinity in biological digesters, then overdosing could result in the active bacteria being killed off if the pH rises high enough. This could cause a major treatment plant upset and result in it going out of compliance on discharge limits. 10.3.3 Metals Removal The lowest concentration of metal ions in solution is achieved when the pH of the solution reaches the minimum solubility for that particular metal ion. The practical maximum working pH of magnesium hydroxides is about 8.5. Attempting to raise the pH above this level will result in overdosing, and any unreacted magnesium hydroxide contributes to the sludge. There are two mechanism at work when Mg(OH)2 is being used to neutralize metal-bearing acidic waste. First, the supply of hydroxyl ion by dissolution for metal precipitation as the hydroxide; see reactions (10.2) and (10.3): þ 2þ MgOHþ surface þ Haq ! Mgaq þ H2 O at pH , 5 2þ MgOHþ surface þ OHaq ! Mgaq þ 2OH

(10:2) (10:3)

10.3 ADVANTAGES OF MAGNESIUM HYDROXIDE

159

Figure 10.3 Zone of higher pH surrounding a magnesium hydroxide particle as compared to bulk solution pH.

Second, magnesium hydroxide can also remove metals through surface absorption. The localized pH close to the surface of a hydroxide particle can be much higher than that of the bulk solution (Frost et al., 1990), see Figure 10.3. When a metal ion enters this zone of higher pH (ca. 10.5), the localized concentration of the hydroxyl ion causes it to precipitate on the surface of the particle, even though the bulk solution pH may not be sufficiently high to cause the metal ion to precipitate. This phenomenon can result in the removal of metal ions from solution one full unit of pH below that achievable using caustic soda (Teringo, 1987). Magnesium hydroxide can effectively remove most metal ions from solution down to levels that satisfy most discharge requirements. However, nickel and cadmium can be more problematic metals to remove from solution. Effective removal necessitates raising the pH to between 10 and 11, which is beyond the reach of magnesium hydroxide. In this case, the majority of the neutralization can be achieved using magnesium hydroxide until pH 8.5 is reached, the remainder of the precipitation can be carried out by “topping off” with caustic soda. 10.3.4 Sludge Volume and Dewatering Magnesium hydroxide produces a faster settling metal hydroxide floc as well as producing a more compact sludge (Marshall and St. Armand, 1992). This occurs because of magnesium hydroxides lower solubility in solution, which results in a slower release of the hydroxyl ion into solution, which in turn causes a more gradual increase in pH. This has the effect of promoting the development of larger metal hydroxide particles similar to that seen in

160

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

Figure 10.4 Settling cones demonstrating differences in metal hydroxide sludge generated during neutralization of acidic wastewater.

crystal growth, that is, the slower the crystal growth the larger the crystal. These large particles contribute to the formation of a more compact sludge that does not entrain a great amount of water. In contrast, lime and caustic soda have much greater solubility in solution and, therefore, provide a hydroxyl ion much more quickly than magnesium hydroxide. This results in a faster rate of reaction and a finer precipitate. Although the reaction has occurred more rapidly, the precipitate formed tends to be gelatinous, more difficult to settle, and entrains a considerable quantity of water. In fact, the volume of sludge produced can be as high as 10 times the volume of that produced when using magnesium hydroxide; see Figure 10.4. This in turn makes dewatering of the sludge much more difficult. Table 10.2 illustrates sludge properties and filtration cycle times for magnesium hydroxide, lime, and caustic soda neutralizing a typical chrome plating solution. The lower volume of sludge produced by using magnesium hydroxide can have a large positive impact on filter cake disposal costs (Louchart, 1987). Treatment of acidic wastewater containing hydrochloric or sulfuric acids, with magnesium hydroxide, produces the soluble salts magnesium chloride and sulfate, respectively; see Tables 10.3 and 10.4. However, treatment of TABLE 10.2 Sludge Volumes Generated by Wastewater Neutralization Using Magnesium Hydroxide Compared with Calcium and Sodium Hydroxides Treatment Alkali Caustic soda Lime Magnesium hydroxide

Solids (%)

Filter Cycle Time (h)

30 35 55

7–9 7–9 1.5– 2

Sludge density (lb/ft.3) 80 85 100 – 110

10.3 ADVANTAGES OF MAGNESIUM HYDROXIDE

TABLE 10.3 Acid Neutralized

161

Salts Produced During Neutralization Reactions Mg(OH)2

Ca(OH)2

NaOH

HCl

Mg(OH)2 þ 2HCl ! MgCl2 þ 2H2O

Ca(OH)2 þ 2HCl ! CaCl2 þ 2H2O

NaOH þ HCl ! NaCl þ H2O

H2SO4

Mg(OH)2 þ H2SO4 ! MgSO4 þ 2H2O

Ca(OH)2 þ H2SO4 ! CaSO4 # þ 2H2O

2NaOH þ H2SO4 ! Na2SO4 þ 2H2O

sulfuric acid waste with lime generates large quantities of gypsum (CaSO4.2H2O) sludge, which will create extra sludge dewatering and handling issues. Another problem that can occur when neutralizing sulfuric acid with lime is that calcium sulfate can precipitate on the surface of the lime particle, thus creating a barrier against effective dissolution of the lime; this can result in having to overdose to achieve the desired rise in pH. Magnesium hydroxide also produces the lowest quantity of dissolved solids per ton of acid neutralized. 10.3.5 Treatment Methods The application of magnesium hydroxide requires a slightly different approach than when using lime or caustic soda. When Mg(OH)2 is added to an acidic metal-bearing waste, the pH does not change immediately as it does when using lime or caustic soda. This again is due to its lower solubility. The continued addition of magnesium hydroxide to the wastewater will cause the pH to slowly rise only to the precipitation pH of the metal ion. When this stage is reached, the pH will remain relatively constant as the hydroxyl ion released from Mg(OH)2 is reacted with the metal ion to produce the metal hydroxide precipitate. Only when the precipitation of the metal ion is complete, will the pH start to rise again; see Figure 10.5. Most of the metal ions are removed from solution in the first 20 min of reaction. If additional reaction time is not available, the precipitation may be completed by the addition of a small quantity of caustic soda without significantly impacting the volume of sludge formed. 10.3.6 Handling Requirements Magnesium hydroxide is supplied either as aqueous slurry or as a dry powder. The slurry is required to be stored in agitated storage tanks to prevent settling of solids. Since the product freezes at 328F, there is no need for specialized heat tracing, unless it is used in very cold climates.

162

Ratio to Mg(OH)2 1.0 1.37 0.96 1.27

kg Required to Neutralize 1000 kg H2SO4 (100% basis)

594 814 569 753

1228 1446 1751a 1751a

Total Dissolved Solids in Effluent Acid (kg) per 1000 kg Acid 800 1095 765 1014

kg Required to Neutralize 1000 kg HCl (100% basis)

Quantity of Salts Produced During Neutralization of Hydrochloric and Sulfuric Acids

1.0 1.37 0.96 1.27

Ratio to Mg(OH)2

1305 1605 1520 1520

Total Dissolved Solids in Effluent Acid (kg) per 1000 kg Acid

a

The CaSO4.2H2O produced in this reaction would be produced as sludge and not a dissolved solid. Generally, one ton of H2SO4 neutralized produces 3 tons of wet sludge.

Mg(OH)2 NaOH CaO Ca(OH)2

Alkali (100% basis)

TABLE 10.4

10.5 BIOLOGICAL WASTEWATER TREATMENT

Figure 10.5

163

Metal hydroxide precipitation curve.

10.3.7 Environmental Impact Magnesium ion poses no threat to the environment and in fact may have a beneficial impact as it is a minor nutrient essential for plant growth. Accidental spills of magnesium hydroxide will not create any significant damage since the material is not caustic. Spills of lime, hydrated lime, and sodium hydroxide are very caustic chemicals and can cause immediate environmental damage, both to plant and animal life. 10.4 ADSORPTION OF DYES ON MAGNESIUM HYDROXIDE The presence of dyes in wastewater can present the wastewater treatment plant with a difficult task of removing them. Apart from the wastewater being colored, dyes are generally toxic and have the tendency to interfere with their biological breakdown in the treatment plant. Magnesium hydroxide has been investigated as an adsorbent for dyes (Soldatkina et al., 2001). Acid dyes, such as Acid Red and Acid Orange were found to be effectively removed from solution by adsorption onto the particle surface of magnesium hydroxide via electrostatic attraction. 10.5 BIOLOGICAL WASTEWATER TREATMENT Many wastewater treatment plants use some form of biological treatment to reduce the levels of biodegradable organic substances. These biological

164

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

processes convert the organic compounds into gases that can escape to the atmosphere and also become incorporated into biomass. Another purpose of biological treatment is also to remove nutrients, such as nitrogen and phosphorus from the wastewater. The major biological treatment processes can be divided into aerobic processes, which require the presence of oxygen, and anaerobic processes, which require the complete absence of oxygen. 10.5.1 Aerobic Processes Aerobic treatment process works by introducing organic waste into a reactor where an aerobic bacterial culture is maintained in suspension. Here the bacteria carry out the conversion of organic compounds in accordance with the following equations. 1. Oxidation and synthesis [reaction (10.4)]: COHNS þ O2 þ nutrients ! CO2 þ NH3 þ C5 H7 NO2 (new bacterial cells)

(10:4)

COHNS represents organic matter present in wastewater. 2. Endogenous respiration [reaction (10.5)]: C5 H7 NO2 þ 5O2 ! 5CO2 þ 2H2 O þ NH3 þ energy

(10:5)

To maintain effective biological treatment, digester pH is carefully monitored and controlled. Most aerobic systems work optimally in a pH range of 6.0– 8.5; excursion of pH outside of this range due to insufficient alkalinity (buffering) in the wastewater can inhibit bacterial activity. It is also important to maintain a stable pH in the digester as bacteria are stressed by sudden changes in pH. Magnesium hydroxide is able to maintain a more stable pH than caustic soda and lime, which are the two most commonly used alkalis in wastewater treatment plants. Magnesium is also an essential microbial nutrient that activates many enzymes, including those involved in phosphate transfer and binding enzymes to substrates. The addition of magnesium by way of magnesium hydroxide to a Mg-deficient wastewater should enhance microbial activity. 10.5.2 Nitrification Under aerobic conditions, ammonia is removed by a two-step biological oxidation process; see reaction (10.6). Ammonia is first oxidized to nitrite

10.5 BIOLOGICAL WASTEWATER TREATMENT

165

primarily by Nitrosomonas sp. bacteria, although Nitrosococcus, Nitrosospira, and Nitrosolobus are also capable of carrying out this reaction: NH3 þ 1:5O2 ! NO2 þ Hþ þ H2 O

(10:6)

The oxidation of nitrite to nitrate is carried out by Nitrobacter, according to: NO2 þ 12O2 ! NO3

(10:7)

As can be seen in reaction (10.6), acid is generated in the nitrification process, and a subsequent decrease in pH can occur in poorly buffered systems. For every 1.0 mg of ammonia nitrogen converted to nitrate nitrogen, 7.14 mg of alkalinity (expressed as CaCO3 equivalent) is consumed. The optimal pH range for nitrification is between 7.5 and 8.5, and below pH 6.1 the bacteria cease to function. Magnesium hydroxide can be metered into the digester to control pH during nitrification. Since magnesium hydroxide has a much lower solubility than either caustic soda or lime, it does not produce pH “hot spots” in the system that can result in bacterial inactivation.

10.5.3 Anaerobic Digestion The products of anaerobic digestion are methane, carbon dioxide, trace gases, and stabilized solids. The microbial populations involved in this digestion process can be divided into three groups, each responsible for a separate function: solubilization, acid formation, and methane formation. Carbohydrates, proteins, and lipids are first solubilized by hydrolysis. These products are then converted into short-chain organic acids, such as acetic, propionic, and lactic acid. The last step involves the conversion of these acids into methane, carbon dioxide, and other gases by methanogens, or methane-forming bacteria. The acid-forming bacteria are tolerant to changes in pH; however, the methanogens, which have an optimal pH range of 6.8–7.2, are not. If the methanogens are inhibited, organic acids start to accumulate and cause the pH to drop, which if continued will eventually lead to digester failure. Because of this sensitivity, alkalinity and pH control are very important parameters for the optimal functioning of anaerobic digesters. This condition makes magnesium hydroxide an ideal alkali for use in anaerobic digestion since it provides a stable and easily controllable pH. It has also been demonstrated that the magnesium ion, along with other light-metal cations can have a stimulatory effect on the digestion process within a certain range of concentration (150 mg/L for Mg), which is demonstrated by a significant increase in gas production (Yarnell, 2000). However, these cations can also have an inhibitory effect if the

166

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

concentration reaches a critical level, which is 1000 mg/L for Mg (Kugelman and Chin, 1971). 10.6 BIOFLOCCULATION AND SOLIDS SETTLING It has been demonstrated that increased levels of Mg2þ improve bioflocculation according to the divalent cation bridging theory (DCBT). This theory states that negatively charged sites on exocellular biopolymers are bridged by divalent cations such as Ca2þ and Mg2þ (Higgins and Novak, 1997). This bridging helps to stabilize the microbe–biopolymer floc matrix; see Figure 10.6. The þ ratio of the monovalent (Naþ, NHþ 4 , K ) to divalent cations (M/D) can be used as an indicator of potential sludge settling and dewatering problems. When this ratio is above approximately 2, which can occur when using caustic soda for pH control, settling and dewatering problems occur. The settling problems can be overcome by reducing the M/D ratio: 1. Increase the divalent cation concentrations. 2. Decrease the monovalent cation concentrations. 3. Do both by replacing monovalent cation with divalent alternatives. Higgins et al. (2004) demonstrated that replacement of sodium hydroxide by magnesium hydroxide at an industrial wastewater treatment plant

Figure 10.6

Divalent cation bridging biopolymer.

10.7 PHOSPHORUS REMOVAL FROM WASTEWATER

167

significantly improved the settling and dewatering properties of the floc. While using NaOH, the M/D ratio was 48, which decreased to 0.1 when using Mg(OH)2. During treatment with magnesium hydroxide, the sludge volume index, effluent total suspended solid, and effluent chemical oxygen demand were all reduced by up to 63%. Alum and polymer dose used for clarification were also reduced by approximately 50 and 60%, respectively. The dewatering properties of the activated sludge also improved, which resulted in an increase in filter cake solids. These combined improvements led to a savings of $30,000–$155,000 per year by switching from caustic soda to magnesium hydroxide.

10.7 PHOSPHORUS REMOVAL FROM WASTEWATER AND STRUVITE FORMATION Many wastewater treatment plants are regulated on phosphate discharge from treated plant effluent since excessive phosphate in surface waters can foster undesirable algae blooms. This can subsequently result in depletion of dissolved oxygen levels in the surface waters and ultimately result in fish kills. The problem is particularly acute in swine and dairy operations that typically handle large numbers of animals where the animals typically excrete 70% of their P intake (Barnett, 1994). This leads to the problem of having to dispose of a large quantity of phosphorus-rich waste. The manure is normally applied to crop pastures; however, this practice has led to a buildup of P on many farm lands since manure contains higher levels of P than N. Many states are now limiting manure application rates based on both nitrogen and phosphorus requirements; consequently, these limited application rates lead to an increased need to store manure on site. In the United States, it is estimated that approximately 52 Mt (dry weight) of manures are collected annually from animal houses (Sutton et al., 1996). When manure is applied to meet the crops nitrogen needs, this leads to an overapplication of P. The more soluble forms of P present in the manure have a tendency to be leached during rainfall and consequently can lead to eutrophication of surface waters. One possible approach to removing P from manure prior to land application is by the forced precipitation of magnesium ammonium phosphate hexahydrate, otherwise known as struvite (MgNH4PO4.6H2O). Struvite is frequently formed in the recycle flush of animal waste management systems, building up in pipes and creating blockages (Booram et al., 1975). Most research into struvite formation has aimed at trying to predict struvite formation in digesters and to prevent the deposits from forming in the first place (Ohlinger et al., 1998; Mamais et al., 1994). However, struvite has been found to be a good plant fertilizer since all three components of struvite are essential plant nutrients,

168

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

and it is slow to release them into the soil because of its low solubility in water (Buchanan, 1993). Some research is now concentrating on the controlled precipitation and removal of struvite from animal and human waste, with the prospect of turning the struvite into a resource instead of a nuisance (Greaves et al., 1999). Magnesium oxide and hydroxide can be employed in the process of controlled struvite precipitation and removal from wastewater and sludge. The process is based on the following reaction: 3 Mg2þ þ NHþ 4 þ PO4 þ 6H2 O ! MgNH4 PO4
6H2 O

(10:8)

The solubility of struvite is dependent upon pH and becomes increasingly less soluble at higher pH (Buchanan et al., 1994). Magnesium oxide and hydroxide are added to the wastewater to raise the pH closer to struvite minimum solubility (ca. pH 9.0). They also provide a source of Mg2þ since this ion is commonly the limiting factor during precipitation; phosphate and ammonium ions are generally abundant in these waste sources. This methodology has been applied both to animal wastes (Beal et al., 1999; Miles and Ellis, 2001) and to human waste anaerobically digested in wastewater treatment plants (Dunseth et al., 1970; Munch and Barr, 2001; Jaffer et al., 2002). Both have achieved a level of 90% phosphate removal. Other approaches have used sintered magnesia granules acting as seed crystals for struvite precipitation by passing wastewater through a bed of the magnesia granules in a reactor column (Spohr and Talts, 1971; Kaneko and Nakajima, 1988).

10.8 ODOR AND CORROSION CONTROL IN SANITARY COLLECTION SYSTEMS Odor and corrosion in sanitary collection systems due to hydrogen sulfide gas has become increasingly more prevalent over the past 20 or so years. A number of factors have contributed to this increasingly prevalent problem. The EPA, as part of the Clean Water Act, required point-source treatment of industrial wastewater, which led to the removal of the majority of metals from being discharged into sanitary collection systems. It was eventually discovered that these metals were in fact not only binding dissolved sulfides in the wastewater, thus preventing its escape into the atmosphere, but were also acting as biocide keeping sulfate-reducing bacteria in check. Other factors, such as longer retention times of the wastewater in the collection system, along with the introduction of low-flow water fixtures, have only exacerbated the situation in allowing sewage to become anaerobic, due to biological activity depleting dissolved oxygen.

10.8 ODOR AND CORROSION CONTROL

169

Organic sulfur compounds are present in domestic sewage at about 1 – 3 mg/L along with the sulfur contribution from background sulfate concentration present in potable water. The major contributors to the generation of sulfide in sewage are the sulfate-reducing bacteria (SRB), which reside in an anaerobic slime layer that builds up on the surfaces of the pipe submerged under water (USEPA, 1974); see Figure 10.7. The primary SRB responsible for sulfate reduction is Desulfovibrio desulfuricans, which uses inorganic sulfate as the oxygen source and organic matter as the food supply; see reaction (10.9): SO24 þ 2C þ 2H2 O ! 2HCO3 þ H2 S "

(10:9)

The release of hydrogen sulfide gas into the collection system headspace [reaction (10.10)] is governed primarily by wastewater pH and turbulence in the waste stream. The equilibrium governing the speciation of sulfide in solution is as follows: H2 Sgas $ H2 Saqueous $ Hþ þ HS $ Hþ þ S2

(10:10)

The equilibrium progresses to the right with increasing pH; see Figure 10.8. Therefore at pH 7, which is fairly typical of most sewage, approximately 50% of the sulfide is present as H2S and 50% as HS2. Increasing the pH

Figure 10.7 Cross section of a concrete sewer pipe depicting colonization of sulfate reducing and sulfur-oxidizing bacteria.

170

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

Figure 10.8

Speciation of hydrogen sulfide and hydrosulfide ion as a function of pH.

to 8.0 decreases the dissolved H2S concentration to about 10% of the sulfide species present in the wastewater. The corrosion of concrete sewer pipe is not a direct result of the presence of hydrogen sulfide gas in the headspace above the water; rather the primary mechanism of corrosion is caused by the presence of Thiobacillus bacteria, which reside on the invert of the pipe; see Figure 10.7. This genus of bacteria has the ability to oxidize hydrogen sulfide gas to form sulfate, which results in the formation of sulfuric acid; see reaction (10.11): H2 S þ 2O2 ! H2 SO4

(10:11)

The generation of acid causes a rapid depression in surface pH of the pipe and reacts with the concrete to form gypsum and anhydrite, which degrades the structural integrity of the pipe. The depression of pH can progress to values as low as 0.5. If this corrosion is allowed to progress unchecked, it can eventually lead to complete pipe failure and collapse. Figure 10.9 displays graphically the rate of corrosion of concrete sewer pipe as a function of surface pH. 10.8.1 Addition to Raw Sewage Magnesium hydroxide slurry can be used to help combat both the generation and release of H2S gas into the headspace, as well as mitigate the effect of Thiobacillus-generated sulfuric acid on the sewer crown and is sold under the

10.8 ODOR AND CORROSION CONTROL

171

Figure 10.9 Rate of concrete sewer pipe corrosion as a function of surface pH. (Data generated by the Los Angeles County Sanitation District.)

trade name Thioguard.1 Magnesium hydroxide slurry can be added to the sewage in the collection system in order to raise the water pH to about 8.5 (Few et al., 1999; Miller and Shand, 1996; Higgins et al., 1997). This shifts the sulfide equilibrium shown in reaction (10.10) to the right, which results in HS2 becoming the predominant sulfide species in solution. Since HS2 cannot escape into the gaseous phase, it remains in solution, thereby significantly reducing H2S gas odor. Once the quantity of hydrogen sulfide gas escaping into the headspace is reduced, this has the effect of depriving Thiobacillus bacteria of the source of sulfide needed for the biological generation of sulfuric acid. This can ultimately lead to a general rise in sewer crown pH significantly lessening the rate of corrosion. The presence of magnesium hydroxide in the sewage also has beneficial effects in the wastewater treatment plant (WWTP) as well; these effects include an increase in alkalinity and also biomass activity (Jefferson et al., 2002), which can enhance aerobic, anaerobic, and nitrification reactions as well as increase solids settling and water clarification. Increasing the pH of sewage also has an effect on the rate of sulfide generation in the slime layer by SRB. The optimum environmental pH for SRB is around pH 7.0 and the rate of sulfate reduction follows an inverted U-shaped curve with the maximum rate centered on pH 7.0 (Pomeroy and Bowlus, 1946). Raising the pH beyond 7.0 has the effect of significantly 1

Premier Chemicals, LLC. Thioguard is a registered trade mark, protected under U.S. patent numbers 5,683,748, 5,834,075, 5,833,864, and 6,056,997.

172

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

decreasing SRB ability to generate sulfide, thereby resulting in an overall decrease in the quantity of sulfide generated by SRB. The Mg2þ ion can also act as a catalyst for the aqueous oxidation of dissolved sulfide. Although the increase in the rate of oxidation is not as great as that seen with transition metal ions, it is still significant and a figure of a five- to sixfold increase has been reported (Chen and Morris, 1972). 10.8.2 Crown Spraying Spraying magnesium hydroxide slurry onto the sewer crown can also be used to combat acid corrosion. Since Mg(OH)2 is an alkali, it neutralizes sulfuric acid generated by Thiobacillus and raises the pH of the pipe surfaces (Lorenzo and Miller, 1996). Since Thiobacillus are acidophilic bacteria, increasing the environmental pH beyond their tolerance inactivates them and prevents further formation of biogenic sulfuric acid (Sydney et al., 1996). Magnesium hydroxide also reacts with H2S gas to form magnesium polysulfide, which eventually oxidizes to magnesium sulfate; see reaction (10.12): Mg(OH)2 þ xH2 S ! MgSx þ 2O2 ! MgSO4

(10:12)

This reaction of the layer of Mg(OH)2 sprayed over Thiobacillus effectively deprives the bacteria of their source of sulfide, which again will inactivate them. Film forming and adhesion additives can be added to the magnesium hydroxide slurry to make the layer sprayed onto the sewer crown surfaces more resistant to surcharging so that the protection is not washed away as soon as full pipe flow occurs. 10.9 ACID MINE DRAINAGE Acid mine drainage (AMD) from abandoned coal and metal mining is a serious environmental problem that has affected thousands of miles of streams and waterways. AMD is created via the oxidation of sulfide minerals, such as pyrite; see reactions (10.13) and (10.14): FeS2 þ 3:5O2 þ H2 O ! Fe2þ þ 2SO24 þ 2Hþ

(10:13)

FeS2 þ 14Fe3þ þ 8H2 O ! 15Fe2þ þ 2SO24 þ 16Hþ

(10:14)

This process is greatly enhanced by the presence of ferric ion, which occurs by both bacterial and air oxidation of ferrous ion. Bacteria such as

10.10 SILICA REMOVAL FROM INDUSTRIAL PLANT WATER

173

Thiobacillus thiooxidans also accelerate the process. Low-pH waters from tailings and acid mine waters have been reported in the range of 0.1–2.1 (Alpers et al., 1992), with the lowest value being reported as 23.6 at the Richmond Mine at Iron Mountain, California, which is now a Superfund site (Nordstrom et al., 2000). Dissolved metal concentrations as high as 200 g/L have also been recorded in this AMD. The Minnesota Department of Natural Resources has investigated the suitability of treating AMD using magnesium hydroxide, hydrated lime, and sodium hydroxide (Eger et al., 1993). While all three alkalis were successful at raising pH and removing dissolved metals from the drainage, magnesium hydroxide was found to have the additional benefit of producing a maximum pH of 9.5, while the other two alkalis resulted in pH values higher than 12. In addition, magnesium hydroxide also produced the lowest volume of metalbearing sludge. The Santa Fe Mine, located in west central Nevada, has utilized magnesite and brucite, mined from the nearby Gabbs magnesite deposit, as a chemical cap. These were placed on the tops and side slopes of the sulfidic mineral dumps before cover soil application. The chemical cap consisted of placement of 20 tons per acre, followed by 8 – 12 in. of cover soil. The magnesite and brucite provided an alkalinity of 1920 mg/kg of soil. The use of magnesium oxide in experimental permeable reactive barriers has also been investigated. Magnesium oxide was found to be more effective than limestone in removing metals from AMD. However, the metal hydroxides were found to precipitate in the porosity of the magnesium oxide column, which diminished its permeability. This problem was overcome by mixing limestone and silica gravel with the magnesium oxide (Cortina et al., 2002).

10.10 SILICA REMOVAL FROM INDUSTRIAL PLANT WATER Silica removal from boiler feed water is necessary to prevent or minimize silicate boiler scales, such as calcium and magnesium silicate, and also to prevent the formation of alumino-silicate scales such as analcite (Na2O.Al2O3. 4SiO2.2H2O). Hot process silica removal is carried out employing a conventional hot process lime or lime-soda softener. The magnesium compounds used may be Epsom salt (magnesium sulfate), dolomitic lime, calcined magnesite, magnesium carbonate, or magnesium oxide. Magnesium oxide is usually employed as it has the greatest removal efficiency and does not contribute to increase solids content of the water. Magnesium oxides possessing high surface areas have been found to have the greatest silica removal efficiency.

174

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

10.10.1 Mechanism of Silica Removal The removal of silica from solution by magnesium compounds is a nonstoichiometric reaction and proceeds by a surface adsorption mechanism with the formation of magnesium silicate; see reaction (10.15): Na2 SiO3 þ Mg(OH)2 þ xH2 O ! MgSiO3
xH2 O

(10:15)

The quantity of magnesium oxide required to remove silica is not proportional to the amount of silica removed. This action is characteristic of adsorption reactions, and the Freundlich equation closely fits the data; see Equation (10.16): x=m ¼ KC1=n

(10:16)

where K and n are constants, x is the quantity of silica adsorbed, m is the quantity of adsorbent, and C is the residual concentration of adsorbed material at equilibrium. 10.10.2 Factors Controlling the Removal of Silica Temperature Temperature has a dramatic effect on silica removal efficiency; the higher the temperature the greater the removal; see Figure 10.10.

Figure 10.10 Effect of temperature on silica removal. (Reproduced by permission of GE Infrastructure Water and Process Technologies.)

10.10 SILICA REMOVAL FROM INDUSTRIAL PLANT WATER

175

Figure 10.11 Effect of retention time on silica removal. (Reproduced by permission of GE Infrastructure Water and Process Technologies.)

Retention Time At a temperature of 958C silica removal is almost complete in 15 min; see Figure 10.11. A lower temperature of 238C results in a more gradual removal of silica. Effect of pH The effect of pH on silica removal is illustrated in Figure 10.12. It can be seen that in each of the three cases the optimal pH is approximately 10.1.

Figure 10.12 Effect of pH on silica removal. (GE Infrastructure Water and Process Technologies.)

176

WATER AND WASTEWATER APPLICATIONS FOR MAGNESIA PRODUCTS

Recirculation of Sludge The recirculation of partially reacted magnesium oxide sludge into fresh incoming water high in silica can affect a considerable increase in silica removal. It is possible, using sludge recirculation, to reduce magnesium requirements by up to 60%.

REFERENCES Alpers, C. N., Nordstrom, D. K., and Burchard, J. M. (1992). USGS WaterResources Invest. Report 91– 4160. Barnett, G. M. (1994). Biores. Technol. 49, 139– 147. Beal, L. J., Burns, R. T., and Stalder, K. J. (1999). ASAE Annual International Meeting Presentation, Toronto, Canada. Booram, C. V., Smith, R. J., and Hazen, T. E. (1975). Trans. ASAE 340– 343. Buchanan, J. (1993). Struvite Control in Flush Water Recycle Components of Livestock Waste Management Systems. Masters Thesis, The University of Tennessee, Knoxville. Buchanan, J. R., Mote, C. R., and Robinson, R. B. (1994). Am. J. Agr. Eng. 37(2), 617– 621. Chen, K. Y., and Morris, J. C (1972), J. Sanitary Eng. Div. Proc. A.S.C.E. 215 –227. Cortina, J. L., Pablo, J de, Ayora, C., Lagreca, I., and Hoffman, C. (2002). Recycling and Waste Treatment in Mineral and Metal Processing: Technical and Economic Aspects, June 16 – 20, Lulea, Sweden. Dunseth, M. G., Salutsky, M., Ries, K. M., and Shapiro, J. J. (1970). Ultimate Disposal of Phosphate from Wastewater by Recovery as Fertilizer for the Federal Water Quality Administration. Department of the Interior, Program #17070 ESJ. Eger, P., Melchert, G., Antonson, D., and Wagner, J. (1993). Paper Presented at the 1993 American Society for Surface Mining and Reclamation Meeting. Spokane, Washington, May 16– 19. Few, R., Chase, R., and Shand, M. (1999). Raising pH Helps Solve City’s H2S Problems. WaterWorld, October. Frost, M. T., Jones, M. H., Flann, R. C., Hart, R. L., Strode, P. R., Urban, A. J., and Tassios, S. (1990). Trans. Instn. Min. Metall. Sect. C, 99, September – December, pp. c117 – c124. Greaves, J., Hobbs, P., Chadwick, D., and Haygarth, P. (1999). Environ. Tech. 20, 697– 708. Higgins, M. J., and Novak, J. T. (1997). Water Environ. Res. 69, 215– 224. Higgins, M. J., Myers, R. D., Sprague, N. M., and Barron, K. (1997). Controlling Hydrogen Sulfide in Wastewater using Base Addition. Proc. 70th Annual. Conf. Water Environ. Fed., Chicago.

REFERENCES

177

Higgins, M. J., Sobeck, D. C., Owens, S. J., and Szabo, L. M. (2004). Water Environ. Res. 76(4), 353 – 359. Jaffer, Y., Clark, T. A., Pearce, P., and Parsons, S. A. (2002). Wat. Res. 36, 1834 – 1842. Jefferson, R., Hurst, A., Stuetz, R., and Parson, S. A. (2002). A Comparison of Chemical Methods for the Control of Odours in Wastewater. Trans. IChemE 80, Part B, March, pp. 93 –99. Kaneko, S., and Nakajima, K. (1988). J. WPCF 60, 1239 –1244. Kugelman, I. J., and Chin, K. K. (1971). Adv. Chem. Series, Am. Chem. Soc. 105, 55 – 90. Lorenzo, V., and Miller, T. (1996). L.A. Tests Magnesia Slurry to Combat Corrosion. WaterWorld 11(1). Louchart, G. W. (1987). Plating and Surface Finishing. 26 –30, November. Mamais, D., Pitt, P. A., Cheng, Y. W., Loiacono, J., and Jenkins, D. (1994). Wat. Environ. Res. 66(7), 912– 918. Marshall, J. W., and St. Armand, D. E. (1992). TAPPI Proceedings Environmental Conference, pp. 441 –447. Miles, A., and Ellis, T. G. (2001). Wat. Sci. Tech. 43(11), 259– 266. Miller, T. M., and Shand, M. A. (1996). U.S. Patent 5,833,864. Munch, E. V., and Barr, K. (2001). Wat. Res. 35(1), 151– 159. Nordstrom, D. K., Alpers, C. N., Ptacek, C. J., and Blowe, D. W. (2000). Environ. Sci. Technol. 34, 254– 258. Ohlinger, K. N., Young, T. M., and Schroeder, E. D. (1998). Wat. Res. 32(12), 3607 – 3614. Pomeroy, R., and Bowlus, F. D. (1946). Progress Report on Sulfide Control Research. Sewage Works J. 18(4), 597 – 640. Soldatkina, L. M., Purich, A. N., and Menchuk, V. V. (2001). Adsorp. Sci. Technol. 19(4), 267– 272. Spohr, G., and Talts, A. (1971). Public Works, 58 – 60, February. Sutton, A. L., et al. (1996). Integrated Animal Waste Management, Task Force Report No. 28, Council for Agricultural Science and Technology. Ames, Iowa. Sydney, R., Esfandi, E., and Surapaneni, S. (1996). Wat. Environ. Res. 68(3), 338 – 347. Teringo, J. (1987). Pollution Engineering Magazine, April. U.S. Environmental Protection Agency (1974). Center for Environmental Research Information, Cincinnati, OH. Process Design Manual for Sulfide Control in Sanitary Sewerage Collection Systems, NTIS No. PB-260479. Yarnell, E. M. (2000). Effect of Mg(OH)2 Addition on Odor and Corrosion Associated with H2S and the Effect on Wastewater Treatment Processes. MSc. Thesis, Bucknell University.

11 MAGNESIA IN POLYMER APPLICATIONS

11.1 MAGNESIUM HYDROXIDE AS A FLAME RETARDANT FOR POLYMER APPLICATIONS Every year in the United States alone there are approximately 5000 deaths as a direct result of fire. Aside from the loss of life, the cost associated with fire damage approaches 0.3% of the gross national product (GNP). These facts illustrate the importance of efficient fire protection and for mineral flame retardants, which play an important role in this issue. The European and North American market for mineral flame retardants are both approximately 340,000 tons per annum, with projected growth rates of 3 and 5%, respectively (Weber, 2000; Hornsby and Watson, 1989, 1990). In both Europe and the United States, aluminum trihydrate, or ATH [Al(OH)3)], has by far the largest share of the mineral flame-retardant market; however, magnesium hydroxide presently has the highest growth rate. To date, most of the research using magnesium hydroxide has focused on thermoplastics, including ethylene–vinyl acetate copolymer (EVA), polypropylene, acrylonitrile – butadiene–styrene (ABS) copolymer, and modified polyphenylene oxide (Hornsby and Watson, 1986).

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 179

180

MAGNESIA IN POLYMER APPLICATIONS

11.2 FLAME-RETARDANT MECHANISMS Being organic substances, polymers provide an excellent fuel source for the propagation of fire. The combustion of polymers occurs via a free-radical mechanism: the heat from the fire vaporizes and ionizes the constituents of the gas forming a “cloud” of free radicals. The spread of combustion occurs via the well-known free radical mechanisms of propagation: chain branching and termination. There are a number of ways to interrupt the chain of events involved in fire propagation: .

. .

Cool the polymer to lower the surface temperature and quench the formation of volatiles. Interfere with the free-radical chain propagation mechanism. Alter the degradation process of the polymer to produce nonflammable volatiles.

Traditional flame retardants are halogenated organic compounds, such as hexachlorocyclopentadiene (HEX) derivatives, chlorinated paraffin, and halogenated trialkyl phosphate esters. These compounds are commonly used in combination with antimony trioxide or other compounds, such as borates, which have a synergistic effect with the halogenated organics. The halogenated flame retardants work by inhibiting the free-radical propagation mechanism. However, there a number of issues associated with the use of halogenated flame retardants and synergistic additives such as antimony trioxide. First, it is now being recognized that these flame retardants have significant health risks associated with them, and, during a fire, acidic and toxic fumes are created that cause injury to people and cause secondary damage to electronic equipment, which can result in more damage than caused by the fire itself. The flame-retardant effect of magnesium hydroxide and ATH is based on the endothermic decomposition to magnesium or aluminum oxide, see reactions (11.1) and (11.2). This decomposition effectively acts as a heat sink cooling the surface of the polymer. Mg(OH)2 ! MgO þ H2 O 2Al(OH)3 ! Al2 O3 þ 3H2 O

(11:1) (11:2)

The decomposition of Mg(OH)2 occurs at about 3508C and absorbs 1.3 kJ/g of heat with the resultant loss of 30.9 wt % as water vapor. ATH, on the other hand, decomposes at about 2008C, requires 1.05 kJ/g with the resultant loss of 34.6 wt % as water vapor. During these processes, only nontoxic and

11.3 PROPERTIES REQUIRED OF MAGNESIUM HYDROXIDE

181

noncorrosive decomposition products are formed. The water vapor formed during the decomposition displaces oxygen and fuel vapors and functions as a protective gas. A heat-resistant (char) layer is left on the polymer surface consisting of magnesium oxide (or aluminum oxide) and carbonized polymer, which inhibits further combustion. This layer also acts to reduce smoke density by preventing soot particles from leaving the surface (Keating et al., 1986; Ingham et al., 1978). Although ATH and Mg(OH)2 have similar heats of decomposition and water vapor loss, the major difference between the two fillers is the temperature of decomposition. Since ATH decomposes at a lower temperature than magnesium hydroxide, this precludes its use in some polymers that require higher processing temperatures and throughput rates, such as polypropylene or nylons.

11.3 PROPERTIES REQUIRED OF MAGNESIUM HYDROXIDE FOR FLAME-RETARDANT APPLICATIONS There are three sources of magnesium hydroxide suitable as flame-retardant fillers: high-purity material derived from natural deposits, seawater grade, and synthetically produced material. Synthetic products have the advantage of very high purity, tailored particle size distribution, and ease of incorporation into the polymer; however, their significant disadvantage is high cost. Synthetic products have been produced using a hydrothermal method. This involves precipitating a fine agglomerated magnesium hydroxide from magnesium chloride solution using ammonia or lime. The hydrothermal conversion forms magnesium hydroxide particles of about 1 mm in size, which have well-defined hexagonal platelet morphology (Miyata et al., 1978); see Figure 11.1. Seawater grade and brucite have the advantage of being less expensive than synthetic material, but they do tend to cause discoloration in the polymer, which is especially noticeable in lighter color formulations. If color is not an issue, then these grades may make more economic sense. The development of the different methods for the production of flameretardant-grade magnesium hydroxide has been recently reviewed (Hancock and Rothon, 1995). One drawback to using mineral filler flame retardants is that they have to be used at very high loadings to achieve a sufficiently high degree of fire retardancy. Typically, on the order of 60 wt % magnesium hydroxide needs to be added to the polymer to achieve a UL 94 V0 rating. At these loadings, incorporating this amount of magnesium hydroxide into the polymer can present significant mixing problems, as well as causing deterioration in polymer physical properties. This can be overcome, in part, by the

182

MAGNESIA IN POLYMER APPLICATIONS

Figure 11.1 hydroxide.

Electon micrograph of Kisuma 5J flame-retardant-grade magnesium

magnesium hydroxide having a small median particle size, 0.5–10 mm, a surface area between 4 and 10 m2 g21 and surface treating the hydroxide particles. 11.3.1 Surface Treatment Surface modifiers are used for a number of reasons, and these include reduction of dustiness and moisture content, improved processing, and enhanced composite properties. Table 11.1 shows that filling polypropylene with uncoated magnesium hydroxide results in a significant decrease in impact and flexural strength. However, surface treatment of the filler can counteract these effects to some degree. The principal types of surface modifiers are fatty acids, such as stearic acid, and organosilane coupling agents. Since magnesium hydroxide has TABLE 11.1 Effect of Stearic Acid Coating on Physical Properties of Magnesium Hydroxide Filled Polypropylene (50%) Coating Level (wt %) Unfilled Impact strength (J 6mm) Flexural modulus (GPa) Flexural strength (MPa)

5.9 1.2 35.0

No SurfaceTreatment 1.8 3.4 30.2

Source: Data adapted from Hornsby and Watson (1995).

3

6

10

3.2 3.6 32.2

3.3 3.6 30.6

3.4 3.6 36.3

11.3 PROPERTIES REQUIRED OF MAGNESIUM HYDROXIDE

183

a polar nature, its surface has poor compatibility with the hydrocarbon polymer, which is nonpolar. This can result in processing problems such as extended processing times and poor mechanical properties. The aim of the surface treatment process is to react the surface of magnesium hydroxide with the coating agent to form a monomolecular layer that is itself nonpolar. 11.3.2 Stearic Acid The reaction between stearic acid and the magnesium hydroxide particle surface results in the formation of a strong ionic bond forming a metal carboxylate (Watson et al., 1987; Cook and Harper, 1996) in the process, while the saturated hydrocarbon tail provides compatibility with the polymer, although the interaction is relatively weak. The theoretical quantity of stearic acid required to form a monolayer on the hydroxide surface can be calculated from the area occupied by the vertically oriented carboxylic ˚ 2 (Ogino group of the stearic acid molecule, which is approximately 21 A et al., 1990), which equates to approximately 2.2 mg of stearic acid per square meter of surface area. A number of laboratory test, such as diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) and XPS can be used to determine the level of coating present on the hydroxide surface (Gilbert et al., 2000; Liauw et al., 1995). The purity of stearic acid is an important consideration when coating magnesium hydroxide; not only can it influence the color of the polymer blend but also the aging performance of the polymer as well. Liauw et al. (2000) have found that pure grades of stearic acid (99%) gave better elongation at break after 3 months of aging compared with a commercial stearine blend containing about 50% stearic acid. The presence of unsaturated components in the stearic acid can have detrimental effects on color stability as well as the completeness of coating coverage. Incomplete coverage of the hydroxide surface with the coating agent can result in an interaction between the hydroxide surface and antioxidants present in the polymer blend, resulting in the formation of various hues ranging from beige to gray (Titelman et al., 2002). The application of stearic acid and stearates to magnesium hydroxide surface can be achieved by the following methods: .

.

In the dry coating method the stearic acid is liquefied by heating to 808C and sprayed onto the magnesium hydroxide while undergoing intensive mixing. Co-milling both the stearic acid and hydroxide can often generate enough heat to liquefy the acid and achieve an acceptable surface treatment.

184 .

MAGNESIA IN POLYMER APPLICATIONS

React magnesium hydroxide in an aqueous solution of either sodium or ammonium stearate. This method can yield a very uniform coverage of the hydroxide powder.

11.3.3 Silanes Organosilane surface modifiers are more expensive than fatty acids, but unlike them they can significantly improve filler-to-polymer bonding. They are multifunctional molecules containing reactive groups capable of forming bonds with both the filler surface and the polymer matrix (Rothon, 1995). The functionality of the reactive groups can be specifically tailored for different types of polymers. The reactive group typically employed for metal hydroxides is an alkoxysilane, which reacts with filler surface hydroxyl groups. Magnesium hydroxide treated with organosilane has been found to result in a significant improvement in the impact strengths of polyolefin (Godlewski and Heggs, 1989). As with fatty acids, organosilanes can be applied using both wet and dry coating techniques.

11.4 NOVEL APPLICATIONS FOR MAGNESIUM HYDROXIDE AS A FLAME RETARDANT Cigarettes generate what is known as side-stream smoke and as such can be objectionable to people nearby, especially those who do not smoke. A number of patents have been issued that address this problem by incorporating magnesium hydroxide alone or in combination with magnesium carbonates into cigarette paper wrappers or the tobacco filler (Bensalem et al., 1999). Another object of this approach has been to reduce the peak burning temperature of the filler so as to reduce the incidence of fires caused by carelessly discarded cigarettes and tobacco ash. It has been found that a combination of magnesium hydroxide and hydromagnesite produces a superior “flavor” than magnesium hydroxide on its own. Clearly, the same flame-retardant mechanisms are operating in these cigarettes as described above for polymers.

11.5 POLYMER CURING AND THICKENING 11.5.1 Sheet Molding Compound (SMC) Magnesium oxide and hydroxide are incorporated into polyester resins for the purpose of modifying the viscoelastic response of the compound. The

11.5 POLYMER CURING AND THICKENING

185

compound is typically formulated using a low-molecular-weight polymer, on the order of 2000 average molecular weight, which is diluted with styrene. This low viscosity is required in subsequent processing and aids in wetting out fillers and glass fibers also present in the compound. The introduction of MgO or Mg(OH)2 into the resin [typically at about 2 – 4 parts per hundred (pphr) of polymer] induces an increase in viscosity, which helps reduce separation of glass fiber from the resin and also converts the resin into a semisolid nontacky sheet. This sheet is then subjected to heat and pressure during the molding process to affect a cure. The mechanism of the thickening reaction is believed to be due to Mg2þ ion complexing with dicarboxylic acid species of the resin; see reaction (11.3), resulting in formation of a higher molecular weight species (Gruskiewicz and Collister, 1982): R1 COO . . . Mg2þ . . . OOCR2 COO . . . Mg2þ . . . OOCR3

(11:3)

It has been demonstrated that the Brunaver, Emmett and Teller (BET) specific surface area of magnesium oxide correlates well with its thickening activity, while for magnesium hydroxide particle size correlates with thickening activity (Casey et al., 1989).

11.5.2 Synthetic Rubber Natural rubber is a polyterpene synthesized naturally by the enzymatic polymerization of isopentylpyrophosphate, and the repeating unit is the same as that of 1,4-polyisoprene [22CH2C(CH3)55CHCH22]. One form of synthetic rubber, polychloroprene, also known as neoprene rubber, is based on the monomer chloroprene [CH255C(Cl)22CH55CH2]. Magnesium oxide is added to the polymer compound, typically at a level of 2–4 pphr, where its primary function is to neutralize trace levels of hydrogen chloride that may be liberated during processing, vulcanization, and heat aging or service. The removal of HCl prevents autocatalytic decomposition of the polymer and hence greater stability. Magnesium oxide also takes part in the vulcanization (cross-linking) process, where it acts as a cross-linking agent. It is generally added along with zinc oxide to give a good balance between processing safety (premature curing) and cure rate. A suitable grade of magnesium oxide is one that has a high surface area (greater than 100 m2 g21) and a small particle size (less than 5 mm). Surface activity indicates the ability of the MgO to react with HCl; hence, the higher the surface area, the greater the processing safety and vulcanizate properties. To prevent loss of surface activity during storage due to reaction with

186

MAGNESIA IN POLYMER APPLICATIONS

atmospheric moisture and carbon dioxide, the oxide can be purchased in hermetically sealed polyethylene sachets. For the MgO to be fully effective it should be dispersed uniformly throughout the polymer matrix, and, owing to this fact, many compounders have adopted the predispersed forms of MgO and/or zinc oxide (Vickery, 1974; Vickery and Mozynski, 1976). BIBLIOGRAPHY Jancar, J., Ed. (1999). Advances in Polymer Science, Mineral Fillers in Thermoplastics I. Springer – Verlag, New York. Rothon, R. N. (2002). Particulate Filler for Polymers, 12(9), RAPRA Report 141, Shrewsbury, United Kingdom.

REFERENCES Bensalem, A., Caang, W., Fournier, J. A., Kallianos, A. G., Paine, J. B., Podraza, K. F., Schleicn, D. M., and Seeman, J. L. (1999). U. S. Patent 5,927,288. Casey, R. B., Lewis, K. A., and Sopp, S. W. (1989). 44th Annual Conference, Composites Institute, The Society of the Plastics Industry, February 6– 9, Session 5-D. Cook, M., and Harper, J. F. (1996). Plast. Rubber Compos. Process. Appl. 25, 99. Gilbert, M., Sutherland, I., and Guest, A. (2000). J. Mater. Sci. 35, 391 –397. Godlewski, R. E., and Heggs, R. P. (1989). Coupling Agents. In Thermoplastic Polymer Additives, Lutz, J. T., Ed., Marcel Dekker: New York. Gruskiewicz, M., and Collister, J. (1982). Polymer Composites, 3(1), 6– 11. Hancock, M., and Rothon, R. N. (1995). Principle Types of Particulate Fillers, Particulate-Filled Polymer Composites. Rothon, R. N., Ed., Longman: Harlow, p. 77. Hornsby, P. R., and Watson, C. L. (1986). Plast. Rubber Proc. Appl. 6, 169 – 175. Hornsby, P. R., and Watson, C. L. (1989). Plast. Rubber Proc. Appl. 11, 45 – 51. Hornsby, P. R., and Watson, C. L. (1990). Polymer Degradation Stability 30, 73– 87. Hornsby, P. R., and Watson, C. L. (1995). J. Mater. Sci. 30, 5347 –5355. Ingham, J. D., Mosesman, M., Chung, S. Y., and Lawson, D. D. (1978). NASA Technical Brief, 3, Item 35. Keating, L., Petrie, S., and Beekman, G. (1986). Plast. Compounding, July/August. Liauw, C. M., Lees, G. C., Hurst, S. J., Rothon, R. N., and Dobson, D. C. (1995). Plast. Rubber Compos. Process Appl. 24, 211. Liauw, C. M., Rothon, R. N., Lees, G. C., Dumitru, P., Iqbal, Z., Khunova, V., and Alexy, P. (2000). Filler Surface Modification with Organic Acids and

REFERENCES

187

Derivatives. The Manchester Metropolitan University Website, www. Mmu.ac.uk. Miyata, S., Kuroda, M., Okada, A., Okazaki, T., and Takasu, M. (1978). U.S. Patent No. 4,098,762. Ogino, K., Abe, M., Goto, Y., Goto M., Takanata, Y., Furada, T., and Hirnao, J. (1990). Yukugaku 39, 398. Rothon, R. N. (1995). Surface Modification and Surface Modifiers. ParticulateFilled Polymer Composites. Longman: Harlow. Titelman, G. I., Gonen, Y., Keidar, Y., and Bron, S. (2002). Polymer Degradation Stability 77, 345 – 352. Vickery, G. C. (1974). U.S. Patent 3,850,845. Vickery, G. C., and Mozynski, D. M. (1976). U.S. Patent 3,951,849. Watson, C. L., Hornsby, P. R., and Holloway, L. R. (1987). European Patent Application 0.243.201. Weber, M. (2000). Industrial Minerals Magazine, Issue 389, February, 2000, pp. 19 –27.

12 ENVIRONMENTAL APPLICATIONS

12.1 FLUE GAS DESULFURIZATION In the 1970s, environmental controls began to include the removal of oxides of sulfur from the flue gas of power stations burning fossil fuels. While lime and limestone-based processes are by far the most widespread desulfurization technology, magnesia can also be utilized in flue gas desulfurization (FGD); however, it only has a minor presence in the United States compared with the use of lime and limestone based scrubbers.

12.2 REGENERATIVE PROCESS The Chemico-Basic magnesium oxide flue gas desulfurization process, developed in early 1970, not only removes sulfur oxide pollutants but is also a regenerative process that produces sulfuric acid as a salable byproduct (Essex Chemical, 1974). There are three series of reactions that occur in the magnesia scrubbing-regeneration process. In the first series, the following reactions predominate. Sulfur dioxide absorption is represented by the formation of magnesium sulfite hexahydrate followed by its further reaction with sulfur dioxide to form magnesium bisulfite; see

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 189

190

ENVIRONMENTAL APPLICATIONS

reactions (12.1)–(12.3): SO2 þ Mg(OH)2 þ 5H2 O ! MgSO3
6H2 O #

(12:1)

SO2 þ MgSO3
6H2 O ! Mg(HSO3 )2 þ 5H2 O

(12:2)

Magnesium bisulfite neutralization is represented by the following reaction: Mg(HSO3 )2 þ Mg(OH)2 þ 4H2 O ! 2MgSO3
6H2 O #

(12:3)

Magnesium sulfite oxidation is represented by reaction (12.4): 2MgSO3 þ O2 ! 2MgSO4

(12:4)

In the first set of reactions magnesium sulfite hexahydrate is precipitated, and the crystals are removed from the scrubbing solution and either sent directly to a drier or thermally converted to the trihydrate, MgSO3.3H2O, and then dried. The reactions occurring in the drier are as follows: MgSO3
3H2 O þ Heat ! MgSO3 þ 3H2 O "

(12:5)

MgSO3
6H2 O þ Heat ! MgSO3 þ 6H2 O "

(12:6)

MgSO4
7H2 O þ Heat ! MgSO4 þ 7H2 O "

(12:7)

The dried crystals of magnesium sulfite are then calcined at 900–11008C in the presence of coke to regenerate MgO and SO2: C þ O2 þ MgSO3 ! MgO þ SO2 " þ CO2 "

(12:8)

The magnesium oxide formed in the regeneration process is recycled back to the scrubber system and the sulfur dioxide is sent to the sulfuric acid plant. 12.2.1 Process Description Figure 12.1 displays a schematic diagram of the FGD MgO recovery process. Flue gas is first prescrubbed to remove ash and the cooled humidified gas is then contacted with the slurry of magnesium hydroxide and sulfite solids in a buffered solution (containing magnesium sulfite, bisulfite, and magnesium sulfate), which absorbs SO2 gas. The advantage of magnesium salts are that they are much more soluble than the corresponding calcium salts, which translates into a higher scrubbing efficiency. The scrubbing solution is continually replenished with fresh and regenerated magnesium hydroxide.

191

Figure 12.1 Schematic diagram of the MgO recovery flue gas desulfurization process. [Reprinted from R. R. Lunt and J. D. Cunic (2000), Profiles in Flue Gas Desulfurization, p. 81; reproduced with permission. Copyright # 2000 AIChE.]

192

ENVIRONMENTAL APPLICATIONS

The spent slurry is continuously withdrawn from the system, where it is dewatered and dried. The sulfite solids are stored independently of the scrubbing system, from where they are fed to the regeneration circuit. The solids are calcined with coke, coal, fuel oil, or natural gas to provide reducing conditions and are usually calcined using a fluidized bed. The sulfur dioxide released during the regeneration of MgO, usually on the order of 7–10%, is suitable for conversion to sulfuric acid. Regenerated MgO from the calciner is returned to the scrubbing system, via a slaker, prior to introduction to the scrubber. The absorption tower is of the Venturi type, and the gas flows co-currently with the absorption slurry, which has the effect of breaking the liquid into fine droplets that maximizes transfer surface. Sulfur dioxide removal efficiencies of 90–95% have been achieved in commercial operation. 12.2.2 Once-Through Process The once-through process is akin to once-through sodium hydroxide scrubbing. It involves scrubbing with a recirculated solution of magnesium sulfite/bisulfite, which is refreshed by the addition of magnesium hydroxide to maintain pH. The absorption tower uses both spray and open-type, selfdraining sieve trays to avoid blockages from the presence of particulates that would present problems in a packed tower. The spent scrubbing solution can be dealt with in the following manner; it can be reused, as is generally the case in pulping plants employing the magnefite process, or it can be neutralized with magnesium hydroxide to form magnesium sulfite followed by oxidation to soluble magnesium sulfate. The magnesium sulfate solution can then be sent to a wastewater treatment plant. 12.2.3 Kawasaki Process The Kawasaki process (see Fig. 12.2), which is principally used in Japan, involves absorption of SO2 in spray towers using slurry composed of magnesium and calcium sulfite and magnesium hydroxide, which forms soluble magnesium and calcium bisulfite. The spent slurry is continuously withdrawn and sent to an air oxidizer where the sulfites and bisulfites are converted to their corresponding sulfate. This results in the production of gypsum and magnesium sulfate. The gypsum is removed by thickening and centrifugation for sale in commercial applications such as drywall manufacturing. Some of the recovered liquor is returned to the scrubbing tower, while the remainder is sent to the magnesium hydroxide regeneration tank where it is reacted with lime to convert MgSO4 back into Mg(OH)2, along with the precipitation of additional gypsum. The regenerated slurry is then

193

Figure 12.2 Schematic diagram of the Kawasaki flue gas desulfurization process. [Reprinted from R. R. Lunt and J. D. Cunic (2000), Profiles in Flue Gas Desulfurization, p. 69; reproduced with permission. Copyright # 2000 AIChE.]

194

ENVIRONMENTAL APPLICATIONS

fed back into the scrubber. The Kawasaki process has an operating SO2 removal capability of between 85 and 95%. 12.2.4 Dravo Thiosorbic Process with Magnesium Hydroxide Recovery The Thiosorbic process utilizes quicklime that contains a small amount of MgO. This lime is slaked to form calcium hydroxide and magnesium hydroxide, and the scrubbing slurry is then contacted with flue gas in an absorber module. The scrubbing liquor, which now contains magnesium sulfite and bisulfite, is air oxidized to form magnesium sulfate. The MgSO4 in solution is then reacted with slaked lime to produce gypsum (CaSO4.2H2O) and magnesium hydroxide. The precipitated gypsum ranges in size from 60 to 100 mm, while the magnesium hydroxide particles are very small, ranging from 4 to 6 mm. Due to large differences in size, the mixture of the two can be separated using hydroclones (Stowe and Benson, 1991; Stowe, 1991; College and Benson, 1992). Dravo has attempted to market this by-product magnesium hydroxide slurry with limited success owing to the low solids content (ca. 30%) and the presence of significant quantities of calcium sulfate, which indicates that the hydrocloning separation process is inefficient. 12.2.5 Sorbtech Process The Sorbtech process uses MgO-coated vermiculite sorbent that is contacted with humidified flue gas using a specially designed moving radial-panel bed. The sorbent is capable of removing greater than 90% of the SO2 along with a large portion of NOx (40 –80%) and other acid gases such as HCl. The sorbent mechanism is a combination of absorption, reaction, and capillary condensation. Spent sorbent is continuously removed from the system and regenerated. Two types of regeneration can be used; the first is heating under a reducing atmosphere at 7508C to decompose magnesium sulfite and sulfate, which forms SO2, H2S, sulfur vapor, nitrogen, and water from the decomposition of NOx. The second regeneration method involves heating in an oxidizing atmosphere at 5508C, which release SO2; however, this method does not completely decompose NOx. Loss of regenerated sorbent is in the order of 5– 20%, which can be used as a fertilizer and soil amendment.

12.3 REMEDIATION APPLICATIONS The leaching of heavy metals into groundwater from sites that have been used to dispose of these elements, or from areas that have been contaminated

12.3 REMEDIATION APPLICATIONS

195

with them, is of great concern because of the possibility of drinking water contamination. Solid wastes are classified hazardous by the U.S. Environmental Protection Agency (EPA) for a number of reasons. Certain wastes are classified as hazardous because they contain chemicals listed by the U.S. EPA as hazardous. Others are classified hazardous because of the characteristic of the waste, which includes ignitability, corrosiveness, reactivity, and toxicity. The characteristic of toxicity is determined using the Toxicity Characteristic Leaching Procedure (TCLP) test. This test subjects the waste to a leachate to determine whether it has unacceptable levels of hazardous substances, such as heavy metals, which can be leached from the waste. The current limit on leachable cadmium and lead is less than 1.0 and 5.0 mg/L, respectively. Magnesium oxide, magnesium hydroxide, and magnesium carbonate used alone or in combination with phosphate salts and other compounds are currently being used in remediation of heavy-metal-contaminated soil and treatment of electric arc furnace dust (Stanforth, 1989; Stanforth and Chowdhury, 1993). The treatment additives can be introduced into contaminated soil by spreading the remediation additive on top of the soil or waste followed by mechanical mixing, such as using a rotary tiller. Other methods of addition include injecting the additive in slurry form into the ground through an infiltration gallery or injection well. The treatment of electric arc furnace dust (EAFD) is achieved by injecting the treatment additive into the dust stream prior to the dust collection system in order to achieve an intimate mixing of the two. The combination of magnesium oxide and phosphate salts, such as calcium triple superphosphate, is particularly useful for the remediation of lead and zinc. Lime or Portland cement is contraindicated for the remediation of these metals as the higher pH generated by lime will have a tendency to resolubilize both lead and zinc compounds. The mode of action of the additive is to chemically react with the metals in such a manner that they will not leach at unacceptable levels under naturally occurring conditions. The magnesium oxide, hydroxide, or carbonate acts as a pH control additive by neutralizing acidic leachate and buffering the solution to a pH of about 9.0. The phosphate acts to precipitate any solubilized metals as the insoluble metal phosphate, thus preventing the metals from dissolving and migrating into the surrounding groundwater. For a specific waste to be treated, the appropriate ratio of the treatment additive (phosphate) and pH control agent (magnesium oxide) that gives the best performance in terms of maximum contaminant level (MCL) and the secondary maximum contaminant levels (SMCL) for drinking water of the EPA’s National Primary and Secondary Drinking Water Regulations is determined experimentally. One hundred grams of the waste is blended

196

ENVIRONMENTAL APPLICATIONS

with the various MgO/phosphate ratios and subjected to the TCLP test until the appropriate ratio is found that meets the necessary treatment standards.

12.4 NUCLEAR WASTE DISPOSAL Low-level nuclear waste is currently being disposed in a repository located in a salt formation in Carlsbad, New Mexico. The repository, known as the Waste Isolation Pilot Plant (WIPP), stores items such as tools, articles of clothing, and equipment contaminated with actinide compounds that were used during the Cold War to manufacture the U.S. nuclear arsenal. The storage vaults are located some 2200 ft below the surface where workers have dug an 8-mile series of tunnels that lead to the containment rooms, which are mostly 100 yards deep by 33 ft wide and 13 ft in height. Here the radioactive waste is stored in specially designed stainless steel drums 8 ft in diameter and 10 ft in height. Once the drums are in place in the containment vault, they are surrounded with specially designed bags containing a granular hard-burnt MgO and the vault is then sealed. Modeling studies of the waste under the ambient condition of the deposit demonstrated that the partial pressure of carbon dioxide in the repository atmosphere could be substantially increased due to microbial-induced decay of organic materials present in the waste. This could potentially lead to a lowering of brine pH (should there be any water intrusion into the repository) and result in an increase in actinide solubility and possible leaching into areas outside of the repository. The purpose of the magnesium oxide is threefold: (1) to raise the pH of any brine intrusion into the repository to prevent solubiliztion of the actinide elements, (2) absorb carbon dioxide from the atmosphere, thereby lowering the potential for forming low pH brines with the resultant formation of either nesquehonite or hydromagnesite, and (3) to form a cementitious mass thereby consolidating the waste and reducing its mobility. The cementing phases are predicted to be produced by brine reaction with MgO to form Sorel cement (Wronkiewicz and Lee, 1999; Krumhansl et al., 2000).

12.5 HAZARDOUS SPILL CLEANUP The accidental spillage of acid and alkalis in manufacturing plants and laboratories occur on a daily basis. An amphoteric buffer composed of magnesium oxide or hydroxide and magnesium sulfate can be used to not only absorb the liquid spill and suppress the emission of vapors but also neutralize it (McGillivray et al., 2004). This product is sold under the trade name

12.7 CARBON DIOXIDE SEQUESTRATION USING BRUCITE

197

AmphoMag (a registered trademark of Premier Chemicals, LLC), and the buffer is produced by mixing an excess of magnesium oxide together with sulfuric acid. This produces a mixture of magnesium oxide, hydroxide, and partly hydrated magnesium sulfate. The mixture is amphoteric, as it is capable of reacting with acids as well as bases; see reactions (12.9) and (12.10). Magnesium oxide present in the mixture reacts with acid spills to form the magnesium salt of the acid: MgO þ 2HCl ! MgCl2 þ H2 O

(12:9)

The magnesium sulfate portion of the buffer is capable of reacting with alkaline solutions, which have a pH above the precipitation point of that of the magnesium ion, for example, sodium hydroxide, lime slurry, and ammonia solution: MgSO4 þ 2NaOH ! Mg(OH)2 þ Na2 SO4

(12:10)

This buffering effect produces a neutralized spill with a resultant pH from about 9 to 10, which is significantly less hazardous than the unneutralized spill. The incorporation of pH indicator dyes into the product produces a visual key as to whether the spill is neutralized or requires the addition of further buffer agent.

12.6 ANTIBACTERIAL ACTIVITY OF MAGNESIUM OXIDE POWDER The antibacterial properties of MgO powder has been studied by Sawai et al. (2000) and has been found to have antibacterial activity toward Escherichia coli and Staphylococcus aureus. It was found that contact between the bacterial cells and the MgO powder was necessary for the bactericidal activity. It is speculated that the generation of active oxygen species, such as O2 2 , which was found to be produced by MgO powder, may be the primary factor in its antibacterial activity rather than any effect due to increasing the growth medium pH caused by the MgO.

12.7 CARBON DIOXIDE SEQUESTRATION USING BRUCITE Fossil fuels are the major worldwide energy source and will continue to be so for the foreseeable future. This has raised environmental and climatic concerns over increasing levels of anthropogenic CO2 emissions. Brucite has

198

ENVIRONMENTAL APPLICATIONS

been examined as a prototype for permanent mineral CO2 sequestration in the form of thermodynamically stable alkaline earth carbonates, namely magnesite (Butt et al., 1996, 1997; Jung et al., 2004). Studies on the simultaneous dehydroxylation/carbonation of brucite have revealed that the carbonation reaction is governed by several factors (Be´rat et al., 2002). First, the carbon dioxide pressure must exceed the critical pressure needed for direct carbonation to yield magnesite at the prevailing reaction temperature. Further, increasing CO2 pressure above the critical pressure slows both the dehydroxylation and carbonation processes, which has been attributed to the formation of a passivating carbonate layer. This layer forms physical barriers that inhibit further reaction. The effect of the passivating layers is countered by translamellar cracking, delamination, and morphological reconstruction of the lamella structure of brucite that accompanies the dehydroxylation process. This process is known to produce high surface area products (100–200 m2 g21), which provides excellent gas –solid contact areas. It is also proposed that transient intermediate products are formed that exhibit enhanced carbonation reactivity in the form of oxyhydroxides. At a constant CO2 pressure, increasing the reaction temperature increases carbonation reactivity until a temperature is reached where MgCO3 becomes thermodynamically unstable.

BIBLIOGRAPHY Lunt, R. R., and Cunic, J. D. (2000). Profiles in Flue Gas Desulfurization. AIChE: New York.

REFERENCES Be´rat, H., McKelvy, M. J., Chizmeshya, A. V. G., Sharma, R., and Carpenter, R. W. (2002). J. Am. Ceram. Soc. 85(4), 742 – 748. Butt, D. P., Lackner, K. S., Wendt, C. H., Conzone, S. D., Kung, H., Lu, Y.-C., and Bremser, J. K. (1996). J. Am. Ceram. Soc. 79(7), 1892 – 1898. Butt, D. P., Lackner, K. S., Wendt, C. H., Benjamin, A. S., Currier, R., Harradine, D. M., Holesinger, T. G., Park, Y. S., and Rising, M. (1997). World Resources Rev. 9(3), 324– 336. College, J. W., and Benson, L. B. (1992). U.S. Patent 5,084,255. Essex Chemical (1974). The Production and Marketing of Sulfuric Acid from the Magnesium Oxide Flue Gas Desulfurization Process. USEPA Flue Gas Desulfurization Symposium, Atlanta, GA, November 4 –7.

REFERENCES

199

Jung, K. S., Keener, T. C., and Green, V. C. (2004). Int. J. Environ. Tech. Manage. 4(1/2), 116– 136. Krumhansl, J. L., Papenguth, H. W., Zhang, P. C., Kelly, J. W., Anderson, H. L., and Hardesty, J. O. E. (2000). Mat. Res. Soc. Symp. Proc. 608. McGillivray S., Shand, M. A., and Miller, T. M. (2004). U.S. Patent 6,692,656. Sawai, J., Kojima, H., Igarashi, H., Hashimoto, A., Shoji, S., Hakoda, A., Kawada, E., Kokugan, T., and Shimizu, M. (2000). World J. Microbiol. Biotech. 16, 187 –194. Stanforth, R. R. (1989). Method and Mixture for Treating Hazardous Waste, U.S. Patent No. 4,889,640. Stanforth, R. R., and Chowdhury, A. K. (1993). In Situ Method for Decreasing Heavy Metal Leaching from Soil or Waste, U.S. Patent No. 5,202,033. Stowe, D. H. (1991). U.S. Patent 5,039,499. Stowe, D. H., and Benson, L. B. (1991). U.S. Patent 4,996,032. Wronkiewicz, D. J., and Lee, J. H. (1999). Mat. Res. Soc. Symp. Proc. 556.

13 ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

13.1 ROLE OF MAGNESIUM IN PLANT NUTRITION The concentration of magnesium in soils generally lies in the range between 0.5 g/kg for sandy soils and 5 g/kg for clay soils. The levels of magnesium are higher in clay soils due to the presence of weatherable ferromagnesian minerals, such as biotite, serpentine, and olivine and also the carbonate mineral dolomite. It is also present in secondary clay minerals, such as chlorite and vermiculite. The speciation of magnesium in soils can be divided into nonexchangeable, exchangeable, and water-soluble forms, which are in equilibrium. The largest fraction of magnesium present in soil is present as the nonexchangeable form, which includes all the Mg present in the primary and most of the Mg present in the secondary clay minerals. The exchangeable magnesium represents about 5% of the total magnesium content; this fraction along with soluble magnesium is of the greatest importance in the supply to plants. The fraction of Mg complexed with soil organic matter is usually less than 1% of the total Mg, and magnesium has the lowest affinity for soil organic complexes of the various divalent cations present in soil. Magnesium ions are more readily leached from the upper layers of soil than Ca2þ owing to the more extensive hydration sphere surrounding Mg2þ, which causes it to be less strongly bound to soil colloids. Also, The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 201

202

ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

Mg2þ is not specifically bound to clay minerals as is the case for Kþ. Magnesium is, therefore, very prone to leaching and leaching rates are on the order of 2–30 kg Mg/ha/yr. 13.1.1 Uptake of Magnesium from the Soil Generally, the uptake of Mg2þ by root cells is much lower that the uptake rate of Kþ. In the presence of an excess of other cationic species, especially Kþ and NHþ 4 , the uptake rate of magnesium can be greatly depressed. This competition can lead to Mg2þ deficiency in plants. In acid soils, magnesium uptake by plants is frequently low due to the increased levels of cationic Al species, which also depresses the uptake of Mg2þ. It has also been found that the appearance of Mn toxicity can be prevented by increasing the Mg2þ supply. 13.1.2 Functions of Magnesium in Plant Growth In plant tissues, a large proportion of the total Mg is diffusible and associated with inorganic and organic ions such as citrate and malate. A well-described function of Mg is its occurrence at the center of the chlorophyll molecule. The quantity of total plant Mg associated with chlorophyll is on the order of 15– 20%. A vital function of Mg2þ in photosynthesis is the activation of ribulose bisphosphate carboxylase. The ribulose bisphosphate carboxylase enzyme is responsible for the assimilation of carbon dioxide and its conversion into organic carbon such as sugar and starch. When chloroplasts are illuminated with light, the import of Mg2þ into the stoma in exchange for Hþ is triggered. This provides conditions of high pH and high Mg2þ concentration necessary for the carboxylase reaction. A major function of Mg2þ is adenosine triphosphate (ATP) bridging with enzymes, which catalyze phosphorylation. This biochemical pathway is the basic mechanism by which energy is transferred and enzymatic processes are enabled. Another fundamental function of Mg2þ is that it stabilizes the structure and conformation of nucleic acids (Travers, 1989). Ribozymic enzymes, which are located in the ribosome, require divalent cations, particularly Mg2þ. In ribozymes, high concentrations of Mg2þ are required for binding the tRNA (transfer ribonucleic acid) strand to the ribosome (Ahsen and Noller, 1995). Therefore, a deficiency of Mg2þ will have a detrimental impact on polypeptide formation and hence protein synthesis, which then has a dramatic effect on plant growth. It has been found that in Mg-deficient plants, the proportion of protein N decreases while nonprotein N increases (Haeder and Mengel, 1969). It is thought that magnesium ions stabilize the ribosomal particles in the conformation necessary for protein synthesis.

13.3 MAGNESIUM IN ANIMAL NUTRITION

203

The symptoms of Mg deficiency always begins in the older leaves and then moves into the younger leaves as Mg2þ is very mobile in the phloem and is translocated from the older to younger leaves when supply is insufficient, which results in chlorosis or interveinal yellowing. In leaf tissue, the threshold value for the onset of deficiency symptoms is in the region of about 2 mg Mg/g dry matter.

13.2 MAGNESIUM FERTILIZERS Intensive farming can deplete soil of essential nutrients such as magnesium, which then can adversely influence crop quality and yield. The most severe and widespread magnesium-deficient soils in the United States are in the Atlantic and Gulf coastal plains, which therefore require the use of magnesium soil amendments. Magnesium has seen widespread use in citrus fertilizers, especially in Florida, for many years. It has been shown that the magnesium requirement for citrus trees is fairly small until the tree bears fruit (Spencer and Wander, 1960). Spencer and Wander also demonstrated that there is a definite relationship between the quantity of extractable magnesium in the soil and the amount taken up by the tree and hence the leaf magnesium content. They also determined that MgO derived from seawater magnesia was superior to Mg derived from olivine, dolomite, or kieserite, especially on acid sandy soil. On neutral soil, the use of MgSO4 is indicated and provides Mg rapidly to the plant; see Table 13.1 for a comparison of mineral magnesium fertilizers.

13.3 MAGNESIUM IN ANIMAL NUTRITION Magnesium is an element that has numerous functions in biochemical and physiological pathways. The animal body contains about 0.05% Mg (Rook and Storry, 1962), of which about 70% is contained in the skeleton with

TABLE 13.1

Comparison of Various Magnesium-Containing Fertilizers

Mineral Magnesian limestone Magnesite Epsom salts Kieserite (MgSO4.H2O) Potassium magnesium sulfate (K2SO4.MgSO4) Magnesium oxide

Magnesium (g/kg) 30– 120 270 96 160 66 540

204

ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

the remainder in the soft tissues. Only about 1% of the total Mg is found in the extracellular fluids. Disturbances in Mg metabolism can have serious health effects on the animal, and Mg deficiency (hypomagnesaemia) can be a life-threatening condition. Magnesium was first shown to be an essential nutrient by Leroy in 1926. Later investigations of the effects of induced magnesium deficiency in rats found the following manifestations of the deficiency: retarded growth, hyperirritability, peripheral vasodilatation, and decline in serum Mg, anorexia, muscular in-coordination, convulsions, coma, and death, along with an excessive calcification of the bone (Kruse et al., 1933, 1934). It was also found that despite normal calcium levels being present in the blood during hypomagnesaemia, aberrations in Ca metabolism were observed. These were manifest as calcification of the kidneys (Greenberg et al., 1938), liver (Hajj and Sell, 1969), aorta (Britton and Stokstad, 1970), and heart muscle (Forbes, 1965). Disturbances in phosphorus metabolism characterized by hypophophatemia have also been reported, as well as intracellular abnormalities. In rats, the mitochondria became swollen and oxidative phosphorylation by liver tissue decreased (Chutkow, 1964). Magnesium also plays an essential role in protein and amino acid metabolism. The synthesis of proteins in rat liver decreased during magnesium deficiency (Sharma, 1975), and magnesium-deficient rats also exhibited aminoaciduria and increased excretion of total urinary nitrogen (Mazzacco et al., 1966). Magnesium deficiency has also been implicated in both animal and human reproduction problems. A review of the importance of Mg in reproduction has been given by Stolkowski (1977). 13.3.1 Ruminant Animals A common example of hypomagnesaemia in ruminant animals is a disease known as grass tetany, which primarily afflicts dairy cows. Grass tetany, or the grass staggers as it is commonly called, is caused by a decrease in the plasma Mg concentration and was first described by Sjollema (1930). The normal blood serum level of magnesium in cows has been reported to be between 1.2 and 3.8 mg/100 mL (Rook and Storry, 1962). Magnesium homeostasis is not regulated by a hormonal system. The concentration of magnesium in the plasma depends upon the continuous absorption of Mg from the gastrointestinal tract inflow and the requirements for maintenance, milk production, growth, and pregnancy (outflow). The plasma concentration of Mg remains normal as long as the Mg inflow exceeds the outflow, the difference being excreted in the urine. If the reverse situation is in effect, then Mg deficiency will occur, and, since the mobilization of

13.3 MAGNESIUM IN ANIMAL NUTRITION

205

bone Mg does not occur in hypomagnesaemic animals, the only way for the body to redress the problem is to stop urinary Mg excretion. Indeed, this is what is seen in hypomagnesaemic animals, the concentration of urinary Mg drops to almost zero (Martens and Schweigel, 2000). 13.3.2 Magnesium Oxide Requirements for Animal Nutrition The Mg requirement and absorption site for cows has been well researched. The primary site for Mg absorption in cows is the rumen (Khorasani et al., 1995). Four milligrams per kilogram of body weight per day is required to compensate for endogenous magnesium loss into the gut. For a cow with a body weight of 700 kg, 2.8 g/day Mg is required. If the cow is producing milk, which contains about 0.12 g Mg/L, then the daily requirement is increased considerably, since Mg secretion into the milk is maintained regardless of the plasma Mg concentration. Therefore, a cow producing say 60 L of milk a day has an additional requirement of 7.2 g Mg per day over and above that required for maintenance; so the daily requirement will be 10 g/day Mg. These figures, however, do not account for the availability of the magnesium supplementation used, so in reality dosing will be somewhat higher. 13.3.3 Factors Affecting Magnesium Utilization The nutrient having the greatest adverse impact on Mg absorption is an excess of potassium in the feed ration. The first study made on the possible effect of K on Mg was conducted by Fontenot and co-workers (1960). Their study, using sheep, found that an increase in oral administration of K significantly reduced Mg digestibility. Other more recent studies have confirmed these findings (Greene et al., 1983). It has also been noted that the incidence of grass tetany appears to be higher in animals grazing on pastures that had previously been fertilized with potassium (Nicholson and Shearer, 1938). Kemp and t’Hart (1957) demonstrated that the K/(Ca þ Mg) ratio in the grass was correlated to the occurrence of grass tetany. Their work indicated that, when the ratio was less than 2.2, there was a reduced incidence of tetany; however, when the value rose above 2.2, there was a significant increase in its occurrence. Butler (1963) also found that the K/(Ca þ Mg) ratio was also correlated with the incidence of grass tetany. It is postulated that soils containing high levels of K lower the Mg uptake by the plant (Ellis, 1979). High levels of added fat in dairy rations can also have a negative effect on Mg availability. Minerals such as Ca and Mg tend to form Ca and Mg soaps, which can possibly lead to reduced availability. Many animal nutritionists recommend feeding an additional 0.05% Mg when feeding higher fat rations.

206

ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

The feeding of ionophores such as monensin and lasalocid was found to increase Mg absorption in some tests. For example, steers on a high forage diet were fed different levels of the ionophore lysocellin or monensin. It was found that the absorption of Mg, Ca, K, and P were all higher (Spears et al., 1989). 13.3.4 Magnesium Bioavailability There are various methods that can be used to measure the biological availability of magnesium supplements; these can involve in vitro or laboratory studies such as the rate of solubilization in rumen fluid or weak acid solution. This method of study is less time consuming and less costly than pursuing animal studies, although in vivo methods are preferred. One study compared three commercial-feed-grade magnesium oxides with differing reactivity’s and particle sizes by measuring the solubiliztion of Mg in acid solution and rumen fluid (Xin et al., 1988). Xin found that the finer, more reactive MgO was more readily soluble in both acid solution and rumen fluid. Another in vitro experiment compared the Mg solubility of commercial MgO sources in rumen conditions for 48 h and then in abomasal conditions for a further 2 h. Magox magnesium oxide was found to be more soluble than the other sources of magnesium oxide (Beede et al., 1989). The results of this study are shown in Table 13.2. In another availability test, again using lambs, five feed-grade magnesium oxides were compared to magnesium sulfate. One MgO source was derived from seawater and the others were calcined magnesite products. Urinary Mg excretion data revealed that the seawater source was 85.3–86.3% as available as magnesium sulfate, while the magnesite sources ranged from 77.5 to 81.8% in Mg availability. Magnesite had essentially zero availability. Davenport et al. (1990) found that magnesium oxide and magnesium hydroxide had similar bioavailability in beef cattle. TABLE 13.2

Comparison of Various Magnesium Oxide Sources for Animal Feed % of Total Mg Solubilized

MgO Source Magox Chinese Turkish Spanish Baymag Greek 1 Greek 2

Ruminal Stage

Abomasal Stage

Total

Relative

22.6 11.7 14.6 14.5 14.2 7.6 6.5

51.1 48.2 45.1 33.8 33.2 33.4 30.7

73.7 59.9 59.7 48.3 47.2 41.0 37.2

100 81 81 66 64 56 50

13.3 MAGNESIUM IN ANIMAL NUTRITION

207

13.3.5 Preventing Grass Tetany by Magnesium Fertilization One of the simplest methods of preventing grass tetany is to fertilize the grazing pastures with magnesium. If this results in an increase in Mg in forage to levels considered safe, tetany should be prevented. Bartlett et al. (1954) showed that the application of 2500 lb calcined magnesite (87% MgO) per acre to a heavily nitrogen-fertilized sward resulted in a large increase in magnesium content of the forage. Parr and Allcroft (1957) compared application of 2.5 tons of magnesian limestone and 1000 lb calcined magnesite per acre to light and sandy soil pasture. Both treatments increased herbage magnesium levels; however, the calcined magnesite gave the greater response. 13.3.6 Preventing Grass Tetany by Oral Supplementation The addition of magnesium oxide to mineral mixes is helpful in alleviating the fall of serum magnesium in cattle. Allowing cows free access to a 1 : 1 mixture of salt and magnesium oxide has been useful under field conditions for preventing grass tetany (Fontenot et al., 1965). If magnesium intakes are too low, then supplementation can be provided using a mixture of magnesia and molasses compressed into mineral blocks (Horvath et al., 1967). The blocks should be soft enough to allow the consumption of about 6 oz per day to ensure sufficient supply of magnesium (Horvath, 1974). Self-hardening molasses blocks use magnesium oxide in the range of 4 – 20% as a “blocking agent,” whereby a reaction between the MgO and molasses acts to harden the liquid mixture into a block. This reaction has not been entirely elucidated but is thought to be a combination of MgO hydrating in the presence of free water, the formation of magnesium saccharates, and cross-linking reactions between Mg2þ and the molasses sugars. Other ingredients present in these blocks are typically; an ammonium phosphate, attaplugite clay, and edible vegetable and animal fats. The molasses, water, phosphate salt, and clay are dispersed under high shear mixing conditions first followed by the fats and finally the magnesium oxide. The viscous liquid is then poured into an appropriate mold and allowed to harden, which can take up to 48 h. There are a number of patents covering this process; see Skoch and Hodge (1979), Schrodeder and Sawhill (1979), Skoch et al. (1981), and Schroeder and Findlay (1984). 13.3.7 Magnesium Requirements of Swine Information concerning the magnesium nutrition of swine is not as extensive as that for ruminant animals. Magnesium deficiency in swine is not as common as it is in cows since the feedstuffs making up the ration typically have well over 1000 ppm Mg.

208

ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

Growing pigs fed magnesium-deficient diets developed signs of hyperirritability, a reluctance to stand, reduced growth, tetany, and death (Mayo et al., 1959). It was concluded from this research that newly weaned pigs required 400–500 ppm per day Mg. Miller et al. (1965) observed that young pigs fed 225 ppm or more exhibited no overt signs of Mg deficiency. They also estimated that 325 ppm Mg was needed to support efficient growth in pigs and also maintain normal tissue levels of magnesium. Harmon et al. (1976) found no effect of magnesium level during gestation on the number of pigs produced per litter or pig weight when fed diets containing either 400 or 900 ppm magnesium. Sows fed 150 ppm Mg during lactation were found to be in severe negative Mg balance, while those fed 605 ppm magnesium maintained magnesium balance. The National Research Council recommendations for swine suggest that at least 400 ppm Mg should be fed daily to baby pigs, growing-finishing pigs, and sows (National Research Council, 1979). 13.3.8 Magnesium Requirements of Poultry Feedstuffs commonly used for poultry generally contribute 1100–2400 ppm to the ration, so typically Mg supplementation is not necessary. The National Research Council (NRC) suggests that 600 ppm Mg be provided to chicks from hatch to 8 weeks of age, and 400 ppm thereafter (National Research Council, 1977). The NRC also recommends 500 ppm dietary Mg for growing turkeys and also 500 ppm for growing ducks. Published data indicate that dietary magnesium levels of up to 6000 ppm can be tolerated by growing chicks. Growing turkeys are less tolerant and suffer at 2000–4000 ppm per day. However, laying hens are able to tolerate quite high levels of Mg up to 12,000 ppm. 13.3.9 Magnesium in Dairy Ration Buffers Buffers in dairy rations are compounds that neutralize excess acid within the animal’s digestive system. The increased emphasis on production and efficiency in dairy cows has lead to an increased use of high concentrate or high-energy rations. Since these rations have more easily fermentable starch in them, this results in increased acid production in the rumen and subsequent acidosis. Higher acidity can result in reduced feed intake, lower milk production, and decreased butterfat. It can also endanger the animal’s health and cause liver abscesses, fatty liver syndrome, rumenitis, and laminitis. Buffering the ration with sodium bicarbonate can help reduce the effects of excess acidity. However, tests have demonstrated that feeding a combination of magnesium oxide and sodium bicarbonate can help maintain a

13.4 MAGNESIUM IN HUMAN HEALTH AND NUTRITION

209

more desirable rumen pH and improve milk yield and butterfat content than sodium bicarbonate alone. It is generally agreed that buffers should contain both sodium bicarbonate and magnesium oxide in a ratio of about 2 : 1, the most appropriate combination being about of 0.45 lb bicarbonate and 0.25 lb MgO.

13.4 MAGNESIUM IN HUMAN HEALTH AND NUTRITION Apart from the neutralizing properties of magnesium hydroxide (milk of magnesia) used in antacid preparations, the magnesium ion plays a critical role in human health and nutrition and is essential in numerous biochemical pathways. Divalent magnesium is the fourth most abundant metal ion found in cellular metabolism. About 90% of the intracellular magnesium ion is bound to the ribosome, which is an RNA/protein complex that mediates protein synthesis. Its biological functions include structural stabilization of proteins, nucleic acids, and cell membranes. Magnesium ion is also required to promote specific structural or catalytic activities of proteins, enzymes, or ribozymes. The recommended daily allowance (RDA) for an adult male is 420 mg of magnesium per day, and 320 mg for an adult female (Clinical Nutrition Service, NIH). Magnesium deficiency is rare if a well-balanced diet is consumed that is rich in fruits, vegetables, whole grains, and nuts. Potable water can also contribute significant amounts of Mg to the diet especially in hard-water areas; however, softened water is depleted in magnesium. The symptoms of magnesium deficiency in humans follow a similar pattern to that in animals and include confusion, disorientation, loss of appetite, depression, muscle cramps and spasms, tingling, numbness, arrhythmia, and seizures. Extra magnesium may be required by individuals with conditions that cause excessive urinary loss of Mg, such as chronic malabsorption or severe diarrhea. Certain medications can also cause the kidneys to excrete excessive Mg, such as diuretics, and Cisplatin, which is used as a chemotherapeutic agent for cancer, as well as certain antibiotics. 13.4.1 Health Benefits of Magnesium Magnesium may play and important role in regulating blood pressure (Institute of Medicine, 1999). Diets that provide plenty of potassium and magnesium are consistently associated with lower blood pressure (Appel, 1999; Simopoulos, 1999; Appel et al., 1997). Magnesium deficiency can cause metabolic changes that may contribute to heart attacks and strokes

210

ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

TABLE 13.3 Food Chemical Codex (FCC) and U.S. Pharmacopoeia (USP) Requirements for Magnesium Oxide and Hydroxide Powders Test

FCC/USP MgO

USP Mg(OH)2 Powder

MgO ignited basis Calcium, as Ca Loss on ignition Loss on drying Iron, as Fe Acid insolubles Heavy metals (as Pb) Free alkali and soluble salts Lead (Pb) Arsenic (As) Tamped density

96.0 –100.5% 1.1% max. 10.0% max. N/Ra 0.05% max. 0.1% max. 20 ppm max. 2.0% max. 4 ppm max. 3 ppm max. 0.2– 0.65 g/mL

95.0 – 100.5% 1.5% 30.0– 33.0% 2.0% N/R N/R 20 ppm max. 2.0% max. 2.0 ppm N/R N/R

a

N/R ¼ no requirement.

(Altura and Altura, 1995), as well as increased risk of abnormal heart rhythms. Population surveys have associated higher blood magnesium levels with lower risks of coronary heart disease (Ford, 1999; Liao et al., 1998; Gartside and Glueck, 1995). Dietary surveys have suggested that a higher magnesium intake is associated with a lower risk of stroke. Magnesium is important in carbohydrate metabolism, and it may influence the release and activity of insulin (Paolissio et al., 1990). Elevated blood glucose levels increase the urinary magnesium loss, which in turn lowers blood magnesium levels, which explains why hypomagnesemia is seen in poorly controlled type I and type II diabetes (Tosiello, 1996). If magnesium supplementation is required over and above that obtained through a balanced diet, then the use of oral magnesium oxide tablets may be indicated. Magnesium supplementation is gaining popularity in the nutraceutical area and can be provided by taking either a U.S. Pharmacopeia (USP) or food-grade magnesium oxide. The requirements for a pharmaceutical-grade magnesium oxide and magnesium hydroxide powder are displayed in Table 13.3. Further details on magnesium’s role in biological chemistry can be found in the books by Cowan (1995) and Sigel and Sigel (1990).

BIBLIOGRAPHY Mengel, K., and Kirkby, E. A. (2001). Principles of Plant Nutrition, 5th ed. Kluwer Academic: New York.

REFERENCES

211

REFERENCES Ahsen, von, U., and Noller, H. F. (1995). Identification of Bases in the 16S rRNA Essential for tRNA binding at the 30S Ribosomal P Site. Science 267, 234– 237. Altura, B. M., and Altura, B. T. (1995). Cell. Mol. Biol. Res. 41, 347– 359. Appel, L. (1999). J. Clin. Cardiol. 22, 1111 –1115. Appel, L. J., Moore, T. J., Obarzanek, E., Vollmer, W. M., Svetkey, L. P., Sacks, F. M., Bray, G. A., Vogt, T. M., Cutler, J. A., Windhauser, M. M., Lin, P.-H., Karansa, N., Simons-Morton, D., McCullough, M., Swain, J., Steeie, P., Evans, M. A., Miller, E. R., and Harsha, D. W. (1997). N. Engl. J. Med. 336, 1117 – 1124. Bartlett, S., Brown, B. B., Foot, A. S., Rowland, S. J., Allcroft, R., and Parr, W. H. (1954). Brit Vet. J. 110, 3. Beede, D. K., Hirchert, E. M., Lough, D. S., Sanchez, W. K., and Wang, C. (1989). Proc. Florida Dairy Production Conf., Gainsville, FL. Britton, W. M., and Stokstad, E. L. R. (1970). J. Nutr. 100, 1501 –1506. Butler, E. J. (1963). J. Agr. Sci. 60, 329. Chutkow, J. G. (1964). J. Lab. Clin. Med. 63, 71 –79. Clinical Nutrition Service, Office of Dietary Supplements, National Institutes of Health, Washington, DC. (http://ODS.nih.gov/factsheets/magnesium.asp) Cowan J. A., Ed. (1995). The Biological Chemistry of Magnesium. John Wiley and Sons, New York. Davenport, G. M., Boling, J. A., and Gay, N. (1990). J. Animal Sci. 68, 3765. Ellis, R., Jr. (1979). Am. Soc. Agron. Spec. Publ. 35, 79. Fontenot, J. P., Gerken, H. J., Kalison, S. L., Watson, D. F., and Engel R. W. (1965). Feedstuffs 37(12), 66. Fontenot, J. P., Miller, R. W., Whitenair, C. K., and McVicar, R. J. (1960). Anim. Sci. 19, 127 – 130. Forbes, R. M. (1965). J. Nutr. 86, 193– 200. Ford, E. S. (1999). Intl. J. Epidem. 28, 646 –651. Gartside, P., and Glueck C. (1995). J. Am. Coll. Nutr. 14, 71 – 79. Greenberg, D. M., Lucia, S. P., and Tufts, E. V. (1938). Am. J. Physiol. 121, 424 – 430. Greene, L. W., Webb, K. W., and Fontenot, J. P. (1983). J. Anim. Sci. 56, 1214–1221. Haeder, H. E., and Mengel, K. (1969). The Absorption of Potassium and Sodium in the Dependence on the Nitrogen Nutrition Level of the Plant. Landw. Forsch. 23/ I, Sonderh., 53– 60. Hajj, R. N., and Sell, J. L. (1969). Can. J. Physiol. Pharmacol. 53, 311– 316. Harmon, B. G., Lui, C. T., Jensen, A. H., and Baker, D. H. (1976). J. Animal Sci. 42, 860 – 865. Horvath, D. J. (1974). Better Crops 3, 16.

212

ROLE OF MAGNESIUM IN ANIMAL, PLANT, AND HUMAN NUTRITION

Horvath, D. J., Dozsa, L., Kidder, H. E., Warren, J. E., Bhott, B., and Croushore, W. (1967). J. Animal Sci. 26, 875. Institute of Medicine (1999). Food and Nutrition Board, Dietary Reference Intakes: Calcium, Phosphorus, Magnesium, Vitamin D and Fluoride. National Academy Press: Washington, DC. Kemp, A., and t’Hart, M. L. (1957). Neth. J. Agr. Sci. 5, 4. Khorasani, G. R., Janzen, R. A., McGill, W. B., and Kennelly, J. J. (1995). J. Animal Sci. 73(Suppl. 1), 336 (Abstr.). Kruse, H. D., Orent, E. R., and McCollum, E. V. (1933). J. Biol. Chem. 100, 603– 643. Kruse, H. D., Schmidt, M. M., and McCollum, E. V. (1934). J. Biol. Chem. 106, 553– 572. Leroy, J. (1926). Necessite du magnesium pour la croissance de la souris. Cempt. Rend. Soc. Biol. 94, 431 –433. Liao, F., Folsom, A., and Brancati, F. (1998). Am. Heart J. 136, 480 –490. Martens, H., and Schweigel, M. (2000). Vet. Clin. N. Am-Food A 16, 339 –368. Mayo, R. H., Plumlee, M. P., and Beeson, W. M. (1959). J. Animal Sci. 18, 264– 274. Mazzacco, V. E., Lizarralde, G., Flink, E. B., and Jones, J. E. (1966). Proc. Soc. Exp. Biol. Med. 123, 403– 408. Miller, E. R., Ullrey, D. E., Zuant, C. L., Baltzer, B. V., Schmidt, D. A., Hoefer, J. A., and Luecke, R. W. (1965). J. Nutr. 85, 13– 20. National Research Council (1977). Nutrient Requirements of Poultry, 7th ed. National Academy of Science, Washington, DC. National Research Council (1979). Nutrient Requirements of Swine, 7th ed. National Academy of Science: Washington, DC. Nicholson, M. A., and Shearer, G. D. (1938). Vet. J. 94, 388. Paolisso, G., Scheen, A., D’Onofrio, F., and Lefebvre, P. (1990). Diabetologia 33, 511– 514. Parr, W. H., and Allcroft, R. (1957). Vet. Rec. 69, 1041. Rook, J. A. F., and Storry, J. E. (1962). Nutr. Abstr. Rev. 32, 1055. Schroeder, J. J., and Findley, J. E. (1984). U.S. Patent 4,431,675. Schroeder, J. J., and Sawhill, J. W. (1979). U.S. Patent 4,160,041. Sharma, S. (1975). Proc. Indian Acad. Sci. 82, 63 – 76. Sigel, H., and Sigel, A., Eds. (1990). Metal Ions in Biological Systems, Vol. 26, Marcel Dekker: New York. Simopoulos, A. P. (1999). Compr. Ther. 25, 95– 100. Sjollema, B. (1930). On the Nature and Therapy of Grass Staggers. Vet. Rec. 10, 425– 430. Skoch, L. V., and Hodge, D. E. (1979). U.S. Patent 4,171,385.

REFERENCES

213

Skoch, L. V, Harmon, B. G., Dickerson, C. W., and Chou, N. S. (1981). U.S. Patent 4,265,916. Spears, J. W., Schricker, B. R., and Burns, J. C. (1989). J. Animal Sci. 67, 2140. Spencer, W. F., and Wander, I. W. (1960). Proc. Florida State Horticultural Soc. 73, 28 – 35. Stolkowski, J. (1977). Rev. Canad. Biol. 36, 135 –177. Tosiello, L. (1996). Arch. Intern. Med. 156, 1143 – 1148. Travers, A. A. (1989). DNA Conformation and Protein Binding. Annu. Rev. Biochem. 58, 427 –452. Xin, Z. X., Tucker, W. B., and Hemken, R. W. (1988). J. Animal Sci. 66(Suppl. 1), 464 (Abstr.).

14 MAGNESIUM SALTS AND MAGNESIUM METAL

14.1 MAGNESIUM ACETATE The solubility of magnesium acetate at 258C approaches 40 g magnesium acetate per 100 g water (Rivett and Packer, 1927). Aqueous solutions of magnesium acetate display high viscosities thought to be due to the association of acetate ions in solution. Anhydrous magnesium acetate occurs in two forms: a-Mg(C2H3O2)2 (orthorhombic crystal system), which is formed by the reaction of MgO with concentrated acetic acid (13 –33%) in boiling ethyl acetate, and b-Mg(C2H3O2)2 (triclinic crystal system), which is formed with 5–6% acetic acid. The commercially available product is the tetrahydrate, which crystallizes from aqueous solution and is the only stable phase below 688C. See Table 14.1 for a listing of physical and chemical properties. The largest use of magnesium acetate is in the production of viscose rayon fiber for the manufacture of cigarette filter tow, and it is also used in dye fixing and textile printing.

14.2 MAGNESIUM ALKYLS Magnesium dialkyls, R1R2Mg, and the alkyl magnesium halides, RMgX (where R is an alkyl groups such as n-butyl or n-hexyl and X is a halogen) The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 215

216

MAGNESIUM SALTS AND MAGNESIUM METAL

TABLE 14.1

Physical Properties of Various Magnesium Salts

Property Molecular weight Crystal system Density, g/cm3 Melting point, 8C DH8298, kJ/mol DG8298, kJ/mol

a-Mg(C2H3O2)2

MgCl2.6H2O

Mg(NO3)2.6H2O

MgSO4.7H2O

142.36 Orthorhombic 1.507 323 dec — —

203.31 Monoclinic 1.56 116 – 118 dec 22499 22115.0

256.38 Monoclinic 1.636 898C 22613.3 22080.7

246.48 Orthorhombic 1.677 26H2O 1508C 23388.6 22871.9

are white, crystalline solids that are pyrophoric. They also react vigorously with water, alcohols, and other reactive hydrogen-containing compounds. Magnesium alkyls are used as polymerization catalysts for olefins and dienes, such as butadiene.

14.3 MAGNESIUM CHLORIDE Magnesium chloride is one of the most important magnesium compounds and forms hydrate with 2, 4, 6, 8, and 12 molecules of water. Both the anhydrous and hexahydrate salt are deliquescent and need to be stored under dry cool conditions. Magnesium chloride is very soluble in water 54.6 g/100 mL, and the hexahydrate is the only stable hydrate between 0 and 1008C. The anhydrous salt is also soluble in methanol (20.4 g at 608C) and ethanol (15.9 g at 608C). On cooling, these solutions crystallize the alcoholate addition compound, such as magnesium chloride hexamethanolate, MgCl2.6CH3OH, and magnesium chloride hexaethanolate, MgCl2 . 6C2H5OH. Anhydrous magnesium chloride can also form addition compounds with ammonia, that is, MgCl2.6NH3, MgCl2.2NH3, and MgCl2.NH3. Magnesium chloride can be produced by reacting magnesium oxide, carbonate, or hydroxide in hydrochloric acid and cooling the solution to crystallize MgCl2.6H2O (bischofite). It can also be produced from naturally occurring brines and seawater by solar evaporation, using fractional crystallization of other salts such as potassium chloride, sodium chloride, and magnesium sulfate. Anhydrous MgCl2 can be made by the direct chlorination of magnesium oxide in the presence of a reducing agent such as coke or coal. Magnesium chloride is used primarily in the production of magnesium metal and as a gauging solution for Sorel cement. It is also used as a fireproofing agent for wood, as a de-dusting agent for roads and in mines, road deicing, in textiles, water treatment, and sugar production from sugar beets.

14.5 MAGNESIUM SULFATE

217

14.4 MAGNESIUM NITRATE Magnesium nitrate hexahydrate, Mg(NO3)2.6H2O, is prepared by dissolving magnesium oxide, hydroxide, or carbonate in nitric acid followed by evaporation and crystallization at room temperature. Magnesium nitrate is soluble in methanol and ethanol and also forms addition compounds with pyridine, urea, and aniline. Dehydration of the hexahydrate above its melting point results in hydrolysis and the formation of basic nitrates such as Mg(NO3)2.4 Mg(OH)2. Heating the hexahydrate salt to above 4008C results in the complete conversion to MgO and oxides of nitrogen. Magnesium nitrates main use is as a foliar fertilizer, and it is also used as a prilling and coating aid in the manufacture of ammonium nitrate fertilizer. Ammonium nitrate crystals undergo a phase change upon heating, and it has been reported that cycling crystals through the 908F transition zone, where the crystal system changes to the rhombic form, produces a 3.6% volume expansion (Russo, 1968). After a few cycles through this transition temperature, the prills start to fracture, chip, and may even turn into a fine powder. Magnesium nitrate is believed to function by raising the transition temperature or stabilizing the rhombic phase. Magnesium nitrate is also utilized in the production of concentrated nitric acid. In this process a 72% Mg(NO3)2 solution is contacted with 60% nitric acid. Since magnesium nitrate has a greater affinity for water, it removes water from the nitric acid stream until the concentration of HNO3 is greater than the 66% HNO3 azeotrope. This allows a 90 –95% nitric acid to distill over, the diluted magnesium nitrate being recovered at the bottom of the column. The magnesium nitrate solution is then reconcentrated to 72% in an evaporator and recycled. At 258C, it has solubility in water of about 42 g/100 g saturated solution.

14.5 MAGNESIUM SULFATE Magnesium sulfate forms a number of hydrates and double salts, the main commercial products being epsomite (MgSO4.7H2O) and kieserite (MgSO4.H2O). Technical-grade Epsom salt is produced by reacting magnesium oxide, hydroxide, or carbonate with sulfuric acid followed by crystallization; however, United States Pharmacopeia (USP) grade has the impurities precipitated out before crystallization using an excess of MgO. The precipitated impurities are then removed by filtration, the filtrate then being evaporated to crystallize Epsom salt.

218

MAGNESIUM SALTS AND MAGNESIUM METAL

The MgSO4 – H2O system is complex due to a number of metastable phases that can coexist at the same time within an aqueous solution at a given temperature. The heptahydrate is stable from about 25 to 48.28C, the hexahydrate from 48.2 to 67.58C, and the monohydrate above 67.58C; all the other hydrates are metastable. The thermal decomposition of the hydrated salts begins at about 1508C and is accompanied by hydrolysis, which gives rise to oxysulfates. In the presence of carbon, decomposition occurs at about 7508C according to reaction (14.1): MgSO4 þ C ! MgO þ SO2 þ CO

(14:1)

Magnesium sulfate is used as a fertilizer, as a dietary supplement in animal feed, and in medicine as a cathartic. It is also used to make magnesium oxysulfate cement and in the textile industry as a conditioning agent for cotton and wool and also as a mordant in dyeing. It is also used in the rubber and plastic industry as a coagulating agent, as well as in the manufacture of citric acid and magnesium stearate.

14.6 MAGNESIUM SOAPS Some common magnesium soaps are the stearate, sulfonate, laurate, oleate, and naphthenate. They are manufactured by reaction of the fatty acid with either magnesium oxide or magnesium chloride or citrate. Magnesium soaps are used as: 1. Lubricants and stabilizers in the production of polymers to improve flow properties 2. Lubricants and as an additive to oils and greases 3. Mold-release agent 4. Surface treatments of fillers

14.7 MAGNESIUM OVERBASE SULFONATES Magnesium overbase sulfonates are used primarily as a motor oil additive where they act to neutralize acidic by-products formed during the combustion process. This helps to reduce corrosion of the internal surfaces of engine components and also reduce the formation of sludge, which are also highly acidic (Spearot, 1974).

14.8 MAGNESIUM PEROXIDE

219

Overbase magnesium sulfonates are a transparent colloidal dispersion (0.1 mm or smaller) of magnesium carbonate in oil containing protective colloids, which are usually magnesium derivatives of petroleum sulfonic acids or alkyl benzene sulfonic acids. An extensive patent literature exists on production methods and techniques for overbasing (U.S. patents 4,225,446; 4,148,740; 4,163,728; 4,347,147; 4,129,589; and 4,094,801). The process involves making a magnesium-organic complex either separately or in situ in the motor oil. The intermediate organic complex is commonly an alcoholate or phenate produced by the reaction of magnesium oxide with an alcohol, such as methanol or phenol. The type of magnesium oxide used in this process is normally a finely divided, medium reactivity, synthetic oxide. This organo-magnesium complex is then converted to a carbonate by the introduction of carbon dioxide gas in the presence of the protective colloids. It has also been found that the presence of a small quantity of water is necessary during the carbonation step. It has been postulated that the presence of water forms magnesium hydroxide dissolved in the aqueous phase of an oil – water emulsion. The reactions between the dissolved magnesium hydroxide and CO2 precipitates MgCO3, which then migrates to the polar sulfonate– oil/water interface where it is coated with magnesium sulfonate and transferred across the interface and into the oil phase (Dawson et al., 1959). The purpose of the protective colloids is to control the growth of the magnesium carbonate crystal size, thus preventing excessive growth and maintaining a colloidal size. This is important since larger particles may settle from the oil and clog filters.

14.8 MAGNESIUM PEROXIDE Magnesium peroxide is easily prepared by the reaction of a low-iron magnesium oxide or magnesium hydroxide with concentrated solutions of hydrogen peroxide (30 –70%) and can be represented by reaction (14.2) (Vannerberg, 1959; Pierron, 1950): MgO þ H2 O2 ! MgO2
H2 O

(14:2)

The necessity for a low-iron magnesium oxide stems from the fact that iron compounds present in the MgO have a tendency to catalyze premature decomposition of the peroxide. The reaction is fast and very exothermic and is carried out by rapidly mixing the reactants in a cooled vessel, the optimal processing temperature being around 30– 408C. Mixing time is kept to a minimum since the longer it is carried out the hotter the reaction mixture

220

MAGNESIUM SALTS AND MAGNESIUM METAL

becomes; this can result in a product with lower active oxygen content due to excessive decomposition of hydrogen peroxide. The slurry can then be dried using a spray drier (Doetsch et al., 1984; Doetsch and Casper, 2001). Magnesium peroxide with an active oxygen content of 10–18% can be produced using this process with a yield of about 35% MgO2. Active oxygen stabilizers, such as sodium silicate and soluble phosphate salts, can be added at the mixing stage. These stabilizers help to prevent the premature decomposition of the magnesium peroxide. Magnesium peroxide is widely used as a source of oxygen for aerobic microorganisms during the treatment and disposal of biological waste; see reaction (14.3): MgO2
H2 O ! Mg(OH)2 þ 12 O2

(14:3)

Odors that form in composting bins can be controlled using magnesium peroxide, as well as supplying oxygen to the aerobic bacteria, which can speed up the composting process. Magnesium peroxide can also be utilized in the remediation of soils contaminated with hydrocarbons. In this process, slurry of the peroxide is injected into a small bore hole that penetrates into the zone of contamination. The slow decomposition of the compound releases oxygen into the surrounding contaminated soil, which encourages the growth of aerobic hydrocarbon metabolizing bacteria (Konenigsberg, 1997).

14.9 MAGNESIUM METAL PRODUCTION Magnesium metal is produced primarily by either thermal or electrochemical processes. The thermal process operates at temperatures over 12008C and utilizes a metallothermic reduction in which magnesium metal volatilizes from MgO and is condensed to recover the metal. The electrochemical process is based on the electrolysis of fused anhydrous magnesium chloride. 14.9.1 Thermal Processes In the thermal reduction process, reaction (14.4), magnesium oxide (as a component of dolime) reacts with a metal such as silicon, which is present in ferrosilicon, to produce magnesium metal. The two thermal methods in operation today are the Pidgeon and Magnetherm processes. The Pidgeon process is a batch process in which dolime and silicon (usually ferrosilicon) are briquetted and fed into a gas-fired or electrical heat retorts. The retort is

BIBLIOGRAPHY

221

fitted with a condensing section that operates at about 5008C and 0.1 mmHg pressure, where the magnesium vapor is condensed: Si(Fe)(s) þ 2CaO
MgO(s) ! 2Mg(g) þ (CaO)2 SiO2(s) þ Fe(Si)(s)

(14:4)

In the Magnetherm process, alumina is added to melt the dicalcium silicate slag that forms at about 15008C; see reaction (14.5): Si(Fe)(s) þ 2CaO
MgO(2) þ 0:3Al2 O3(s)

! 2Mg(g) þ (CaO)2 SiO2(l) þ 0:3Al2 O3 þ Fe(Si)(l)

(14:5)

14.9.2 Electrolytic Production Electrolytic reduction methods make use of reaction (14.6): MgCl2(l) ! Mg(g) þ Cl2(g)

(14:6)

Anhydrous magnesium chloride in the molten state has been produced extensively from MgO obtained either from calcined magnesium hydroxide from seawater or from calcined magnesite, as shown in reaction (14.7): MgO þ C þ Cl2 ! MgCl2 þ CO

(14:7)

The reduction process is carried out in refractory-lined electrolytic cells, the chlorine gas being collected and reused in the process, while molten magnesium metal is collected from the cathodes.

BIBLIOGRAPHY Allain, R. J., and Fong, D. W. (1982), U.S. Patent 4,347,147. Arnold, J. D., Fair, H. J., Fair, L. V., and Berry, J. D. (1980). U.S. Patent 4,225,446. Cease, V. J., and Kirk, G. R. (1979). U.S. Patent 4,148,740. Cheng, W. J., and Guthrie, D. B. (1979). U.S. Patent 4,163,728. Copp, A. N., and Wardle, R. W. (1981). Magnesium Compounds. In Kirk-Othmer Encyclopedia of Chemical Technology, Vol. 14, 3rd ed. Wiley: New York. Dawson, C. R., Ratner, H., and Roberts, L. R. (1959). Presentation before the Division of Petroleum Chemistry of the American Chemical Society, Atlantic City Meeting, September 13 – 19. Doetsch, W., and Casper, O. (2001). U.S. Patent 6,193,776.

222

MAGNESIUM SALTS AND MAGNESIUM METAL

Doetsch, W., Dillenburg, H., Fuchs, P. W., and Honig, H. (1984). U.S. Patent 4,427,644. Eliades, T. I., and Horner, J. D. (1978). U.S. Patent 4,129,589. Forsberg, J. W. (1978). U.S. Patent 4,094,801. Konenigsberg, S. (1997). Environ. Protection, 19– 22, February. Pierron, P. (1950). Bull. Soc. Chim., France, 291. Rivett, A. C. D., and Packer, J. (1927). J. Chem. Soc. 130, 1342– 1349. Russo, V. J. (1968). Ind. Eng. Chem. 7(1), 69. Spearot, J. (1974). National Meeting of the American Chemical Society, Los Angeles, March 31 –April 5, 19(3), 598. Vannerberg, N.-G. (1959). Akiv. Kemi. 14, 99.

15 PULP APPLICATIONS

15.1 SULFITE PULPING Sulfite pulping is a full chemical pulping process that uses mixtures of 2þ þ 2þ sulfurous acid and/or its alkali salts (Naþ, NHþ 4 , Mg , K , or Ca ) to remove lignin from the cellulose matrix. The lignin present in wood is solubilized through the formation of sulfonate groups and cleavage of the lignin bonds. Once the dominant pulping process, it has been surpassed by Kraft pulping and now only accounts for about 10% of pulp production. The advantages of sulfite pulping are that it produces a brighter more easily bleached pulp and also has a higher pulp yield than the Kraft process. Its disadvantages are that not all species of wood can be pulped and the pulp is also weaker in strength than Kraft pulp.

15.2 MAGNEFITE PULPING PROCESS The Magnefite pulping process utilizes acid magnesium bisulfite solution at a pH of about 4.5 as the pulping liquor. The process produces pulp that is used to make reinforcing pulps for newsprint; see Figure 15.1, which displays a schematic of the Magnefite process. The recovery of MgO is necessary for

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 223

224

Figure 15.1

Schematic diagram of the Magnefite pulping process. (Courtesy of The Babcock & Wilcox Company.)

15.2 MAGNEFITE PULPING PROCESS

225

economic viability as magnesium oxide is the most expensive alkali compared to other alkalis used in sulfite pulping. 15.2.1 Pulping Liquor Preparation Sulfur dioxide is prepared by the burning of sulfur to form sulfur dioxide, which is dissolved in water to form sulfurous acid (H2SO3); see reactions (15.1) and (15.2). The sulfurous acid is then reacted with slaked MgO to form magnesium bisulfite, reaction (15.3). Further reaction of magnesium bisulfite with excess magnesium hydroxide produces magnesium sulfite, reaction (15.4). S þ O2 ! SO2 SO2 þ H2 O

$

(15:1) H2 SO3

2H2 SO3 þ Mg(OH)2

$

Mg(HSO3 )2 þ Mg(OH)2

(15:2) Mg(HSO3 )2 þ 2H2 O

$

2MgSO3 þ 2H2 O

(15:3) (15:4)

The pH of the pulping liquor has to be maintained at a pH below 5.0 due to the limited solubility of magnesium sulfite, which is about 5.5 g/L at 188C, while that of magnesium bisulfite is 43 g per 100 g of solution at 258C and 1 atm sulfur dioxide (The Institute of Paper Chemistry, 1960). Figure 15.2 displays the molar fractions of sulfur dioxide, bisulfite, and sulfite present in solution as a function of pH.

Figure 15.2

22 Fractional speciation of aqueous SO2, HSO2 3 and SO3 as function of pH.

226

PULP APPLICATIONS

15.2.2 Pulping Process Wood species such as spruce, hemlock, aspen, poplar, birch, beach, and maple are suitable for sulfite pulping; however, resinous species such as southern pine and Douglas fir are not suitable since the resin contains a compound that inhibits pulping. The wood chips are cooked with the liquor for about 4.5 h at 130–1358C at a pressure of about 620 kPa (90 psi). After the cooking time has elapsed, the digester pressure is lowered to about 275 kPa (40 psi) prior to blowing the chips. This sudden decompression has the effect of separating the wood fibers. The sulfite process produces a medium-strength pulp with soft flexible fibers and low lignin content. It is also a more readily bleachable pulp, using about one-half the bleaching chemicals of that required in the Kraft process. The pulping liquor is then washed from pulp prior to the bleaching phase. 15.2.3 Chemical Recovery Spent sulfite liquor is concentrated to about 55% solids and burned at 13508C in order to recover both the MgO and SO2. The magnesium oxide ash from the combustion of the liquor is recovered through the use of mechanical cyclone dust collectors. The ash is conveyed to a retention tank to dissolve soluble calcium, sodium, and potassium salts contaminating the MgO and then filtered. The washed MgO cake is then slaked at 1958F. A makeup of dry MgO or Mg(OH)2 slurry at 25% solids is introduced at the slaking stage. Magnesium hydroxide slurry containing 10% solids is then pumped to venturi absorbers. Cooled flue gas from the recovery furnace flows through two Venturi scrubbers where magnesium hydroxide slurry is added to the recycled liquor and absorbs SO2 to form magnesium bisulfite. By splitting the absorption of sulfur dioxide into two stages, the pH is separately controlled, which optimizes the formation of magnesium bisulfite in the first absorber and, by controlling to a higher pH in the second, minimizes the loss of SO2. Control of pH is by regulation of magnesium hydroxide feed to the two Venturi scrubbers and high recirculation rates of slurry favors rapid conversion to the bisulfite.

15.3 PULP BLEACHING Bleaching is the treatment of wood pulps with chemical agents to increase their optical brightness. Mechanical and chemical pulps undergo different

15.3 PULP BLEACHING

227

treatments for the bleaching process. The bleaching of chemical pulps, such as that produced by the Magnefite or Kraft process, is achieved by lignin removal. However, mechanical pulps are brightened by chemically altering portions of the lignin molecule that absorb light, that is, has color, thus decolorizing them. Most of the lignin remains intact. Brightening of mechanical pulps can also be accomplished with reducing agents, such as dithionite, or oxidants, such as hydrogen peroxide. Bleaching of chemical pulps can be achieved using a number of compounds, such as chlorine (C stage), chlorine dioxide (D stage), hypochlorite (H stage), oxygen (O stage), and hydrogen peroxide (P stage). 15.3.1 Oxygen Bleaching and Delignification Oxygen delignification of pulp uses oxygen at 551– 690 kPa (80 –100 psi) of pressure and NaOH at a level of 3– 4% on pulp. Delignification is carried out at 90 –1308C for 20 –60 min. Although oxygen delignification is a more environmentally friendly and less costly bleaching process, it does suffer from the disadvantage of being the least specific for lignin removal and does cause depolymerization of the cellulose, which results in a marked decrease in pulp viscosity and strength. The key to the success of this process was the introduction of small amounts of magnesium ion into the bleaching solution, which was found to mitigate the degradation of cellulose during the delignification process (Robert et al., 1966, 1964; French Patent 1,387,853). Mg2þ is normally applied at levels between 0.05 and 0.1 wt % on oven-dried pulp basis and can be applied in the form of Epsom salt or magnesium oxide or hydroxide. The selectivity of oxygen delignification toward lignin, rather than cellulose, is affected by the presence of pulp contaminants, such as transitionmetal ions, for example, iron, copper, and manganese. These ions catalytically generate free-radical species that attack both lignin and cellulose. One approach to dealing with the presence of transition-metal ions is to remove them from the pulp before the oxygen delignification stage. This can be achieved either by acid washing the pulp or by treating with chelating agents such as ethylenediaminetetraacetic acid (EDTA). Another approach is to add a carbohydrate protector, such as magnesium ion, which helps prevent degradation of the cellulose polymer. Magnesium additives are thought to function by forming magnesium hydroxide in situ, which either adsorbs the transition-metal ions or forms complexes, thus deactivating their catalytic activity (Robert and Viallet, 1971; Gilbert et al., 1973). Although sodium hydroxide has been the primary alkali used in oxygen delignification of sulfite pulps, its replacement by magnesium hydroxide

228

PULP APPLICATIONS

and oxide has also been evaluated (Luo and Christensen, 1992; Yonghao et al., 1998). Results from this evaluation have shown that magnesium hydroxide and oxide are suitable replacements, producing a pulp with higher yields but a lower average degree of polymerization than pulp O2 delignified with caustic soda.

15.3.2 Hydrogen Peroxide Bleaching The use of hydrogen peroxide in pulp bleaching is now in widespread use as the industry strives to eliminate chlorine from the bleaching process. This has arisen over environmental concerns about organochlorine compounds being discharged into the environment, which are produced as by-products in chlorine bleaching processes. Hydrogen peroxide under alkaline conditions reacts with the hydroxyl ion (OH2) to yield the perhydroxyl anion (HOO2); see reaction (15.5): H2 O2 þ OH

! HOO

þ H2 O

(15:5)

The perhydroxyl anion is a strong nucleophile and is primarily responsible for the bleaching by hydrogen peroxide. There are two competing reactions occurring in hydrogen peroxide bleaching; the first leads to a brightness increase, while the second parallel reaction results in the decomposition of hydrogen peroxide into oxygen and water. The decomposition of hydrogen peroxide is catalyzed by transition-metal ions (Tatsumi et al., 1987), which is shown in reactions (15.6)– (15.8):




Mnþ þ H2 O2 ! HO þ OH þ Mðnþ1Þ


!O
þ H O
þ H O ! O þ HO
þ OH

HO þ HOO O2

(15:6)

2

2

2

(15:7)

2

(15:8)

2




The presence of low concentrations of hydroxyl radical (HO ) and superoxide anion radical (O2 2 ) may be necessary for hydrogen peroxide bleaching to proceed. However, high concentrations of these species lead to poor selectivity and decreased bleaching efficiency. Again, controlling the metal profile in the pulp before bleaching by chelation and acid washing may be necessary. The addition of magnesium compounds, such as magnesium sulfate, along with sodium silicate, has been found to hinder hydrogen peroxide decomposition (Colodette et al., 1989). The proposed mechanism involves the magnesium ion disrupting the free-radical chain reaction by the formation of a metastable magnesium superoxide complex, reaction




BIBLIOGRAPHY

229

(15.9), which subsequently decomposes to produce oxygen and magnesium peroxide [reaction (15.10)]:




2O2 þ Mg2þ ! Mg(O2 )2 Mg(O2 )2 ! MgO2 þ O2

(15:9) (15:10)

Processes have also been proposed whereby the sole alkaline source in the peroxide bleaching (which is normally caustic soda) process is replaced by magnesium oxide or hydroxide (Vincent and McLean, 2000, 2003). Other processes involve the peroxide bleaching of mechanical pulp, utilizing magnesium ion derived from magnesium hydroxide. The addition of Mg(OH)2 into the refiners has also been proposed (Haynes et al., 2002; Parrish et al., 2004; Gibson and Wajer, 2003). Magnesium oxide and hydroxide provide the benefits of being not only alkalis but also pulp protectors and peroxide stabilizers.

15.4 DEINKING The use of recycled fiber in the paper industry has increased significantly in recent years. Before reuse, it has to be thoroughly deinked. Most commercial operations use either floatation or washing as the main methods of ink removal. In the deinking process, the ink is dislodged by swelling the fiber structure and pulping under high alkalinity, which assists this process. Caustic soda is the most common source of alkalinity; however, some research has been conducted into replacing sodium hydroxide with other alkalis, such as magnesium oxide (Mahagaonkar et al., 1996). Initial studies showed that despite the fact that magnesium oxide displayed the same ink removal ability as caustic soda when deinking newspaper and magazines, the resulting pulp did have lower brightness, tear, tensile, and burst indices. The addition of a wetting agent was discovered to improve the deinking efficiency of the MgO system to a level similar to that of NaOH and has undergone successful plant trials (Stack et al., 2000).

BIBLIOGRAPHY Biermann, C. J. (1996). Handbook of Pulping and Papermaking, 2nd ed. Academic: San Diego, California. Darmstadt, W. J., and Tomlinson, G. H. (1960). Magnesia Base Pulping and Recovery, Presented before the Acid Pulp Session of TAPPI, February, 23. The Babcock & Wilcox Company Bulletin BR-751.

230

PULP APPLICATIONS

REFERENCES Colodette, J. L., Rothenberg, S., and Dence, C. W. (1989). J. Pulp Pap. Sci. 15(2), J45. Gibson, A. R., and Wajer, M. T. (2003). U.S. Patent Application No. 2003/ 0024664A1. Gilbert, A. F., Pavlovova, E., and Rapson, W. H. (1973). TAPPI 56(6), 95. Haynes, K. K., Campbell, R. O., Brooks, Z. L., Parrish, A., and Hamilton, R. T. (2002). U.S. Patent Applications 2002/0189021A1. Luo, J., and Christensen, P. K. (1992). TAPPI J. June, 183. Mahagaonkar, M., Banham, P., and Stack, K. (1996). APPITA J. 49, 403. Parrish, A., Campbell, R. O., Harrison, R. E., Mobley, P. B., and McCarthy, G. (2004). U.S. Patent Application No. 2004/0112557A1. Robert, A., and Viallet, A. (1971). ATIP Rev. 25, 237. Robert, A., Rerolle, P., Viallet, A., and Martin-Borret, O. (1964). ATIP Rev. 18(4), 151. Robert, A., Traynard, P., and Martin-Borret, O., French Patent No. 1,387,853. Robert, A., Viallet, A., Rerolle, P., and, Andreolety, J. P. (1966). ATIP Rev. 20(5), 207. Stack, K., Clow, M., Kirk, M., and Maughan, S. (2000). APPITA, 459. Tatsumi, K., Murayama, K., and Terashima, N. (1987). TAPPI International Oxygen Delignification Conference Proceedings, TAPPI Press, Atlanta, p. 99. The Institute of Paper Chemistry (1960). Project 2184, Progress Report 2, Use of Magox in Magnesium-Base Sulfite Pulping. Vincent A. H., and McLean, I. A. (2000). U.S. Patent No. 6,056,853. Vincent A. H., and McLean, I. A. (2003). U.S. Patent No. 6,524,437 B1. Yonghao, N. I., Van Heiningen, A. R. P., and Kang, G. J. (1998). International Patent Publication WO 98/31872.

16 MAGNESIA CEMENTS

16.1 INTRODUCTION In 1867 Sorel announced the discovery of excellent cement formed from the combination of magnesium oxide and magnesium chloride solution. This cement type is known by many different names, such as Sorel, magnesite, and magnesium oxychloride (MOC) cement. There are two other known magnesia cements; the first is magnesium oxysulfate (MOS), which is the sulfate analog of magnesium oxychloride and is formed by the reaction between magnesium oxide and magnesium sulfate solution. The second is magnesium phosphate cement (MAP), formed by the reaction between magnesium oxide and a soluble phosphate, such as dibasic ammonium phosphate (NH4H2PO4).

16.2 MAGNESIUM OXYCHLORIDE CEMENT Magnesium oxychloride has many properties that are superior to that of Portland cement. It does not need wet curing, has high fire resistance, low thermal conductivity, and good resistance to abrasion. It also has high transverse and crushing strengths; 48– 69 MPa is not uncommon. Magnesium oxychloride also bonds very well to a wide variety of inorganic and The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 231

232

MAGNESIA CEMENTS

organic aggregates, such as saw dust, wood flour, marble flour, sand, and gravel, giving a cement that has high early strength, insecticidal properties, resiliency, electrically conducting, and is unaffected by oil, grease, and paints. The major commercial applications of magnesium oxychloride cement are industrial and residential flooring and decking, fire protection coatings and paneling, grinding wheel binder, and, because of its resemblance to marble, has been used for rendering wall insulation panels and for stuccos. 16.2.1 Phase Formation The main bonding phases found in hardened cement pastes are Mg(OH)2, 5Mg(OH)2.MgCl2.8H2O (5-form), and 3Mg(OH)2.MgCl2.8H2O (3-form) (Sorrell and Armstrong, 1976; Urwongse and Sorrell, 1980a). Five-form is the phase with superior mechanical properties and is formed using a molar ratio of MgO : MgCl2 : H2O ¼ 5 : 1 : 13. A slight excess of MgO and an amount of water as close as possible to theoretical required for formation of the 5-form and hydration of the excess MgO to form Mg(OH)2 is also necessary. The 5-form phase appears about 2 h after the cement is mixed with water and results in the formation of needlelike crystals, see Figure 16.1, which interlock in a rapid growth. At the stage when crystal growth becomes crowded due to lack of space, the crystals then begin to

Figure 16.1 Electron micrograph showing interlocking needlelike growths of magnesium oxychloride.

16.2 MAGNESIUM OXYCHLORIDE CEMENT

233

intergrow into a denser structure (Matkovic and Young, 1973). The strength of the bond therefore depends on the formation of 5-form, which has good space-filling properties and forms a dense microstructure with minimum porosity. Dehua and Chuanmei (1999) consider that the formation of the magnesium oxychloride phases are produced in the following reaction sequence: the MgO first dissolves in the magnesium chloride solution followed by the formation of the polynuclear complexes [Mgx(OH)y(H2O)z]2x2y. These complexes then further react with the chloride ion to form a continuous phase or hydrogel, which subsequently converts to the crystalline phases. Requirements of Magnesium Oxide The reactivity of the MgO greatly influences reaction rates and products thus affecting the development of strength (Harper, 1967). The magnesium oxide should conform to certain requirements of chemical and physical properties. Conditions of calcination, particle size, and active lime content must be carefully controlled. The “active lime” content is defined as that portion of the total lime that will react with the magnesium chloride and includes calcium oxide, calcium hydroxide, and some forms of calcium silicate. This reaction results in an increase in volume in the cement during the setting process and will result in decreased strength and durability. When the active lime content does not exceed 2.5%, the effect is minimized and can also be compensated for by the addition of magnesium sulfate to the magnesium chloride gauging solution where the sulfate reacts with the active lime to form calcium sulfate. Since the reaction to form MOC cement requires the dissolution of the magnesium oxide in the magnesium chloride solution, surface area is important, which in turn is related to mean crystallite size, size distribution, and the state of particle aggregation. Magnesium chloride is generally applied as a 218 Be´ gauging solution (19.5 wt % MgCl2) and should have a maximum of 0.5% calcium chloride and 1.0% total alkali chlorides. The minimum amount of gauging solution to yield a plastic mix of satisfactory workability should be used. 16.2.2 Water Resistance of MOC Cement The main reason why magnesium oxychloride cement has not remained popular in the building industry is that the magnesium oxychloride phase is not stable in prolonged contact with water, which will result in the leaching of magnesium chloride from the cement phase, leaving magnesium hydroxide behind as the binder. Since magnesium hydroxide does not have adequate strength, the cement bond as a whole is severely weakened. Various

234

MAGNESIA CEMENTS

additives have been added to MOC cements to try and combat this problem of water resistance with varying degrees of success, but on the whole no additive has produced an adequate long-term waterproofing effect. Over a period of time, atmospheric carbon dioxide will react with magnesium oxychloride to form a surface layer of magnesium chlorocarbonate, Mg(OH)2.2MgCO3.MgCl2.6H2O (Cole and Demediuk, 1955). This layer serves to slow the leaching of magnesium chloride from the cement. Eventually, additional leaching results in the formation of hydromagnesite, 5MgO.4CO2.5H2O, which is insoluble and enables the cement to maintain structural integrity. Improvement in the water resistance of MOC cement by admixtures such as ZnHPO4.H2O and AlH3(PO4)2.3H2O (Matkovic et al., 1976) as well as calcium phosphates (Matkovic, 1981, FHA Report) and other phosphate compounds has been documented (Deng, 2003). It has also been reported that the addition of silicates, such as serpentinite and diopside, can also improve the water resistance of MOC cement (Vereshchagin et al., 1997).

16.3 MAGNESIUM OXYSULFATE (MOS) CEMENT Magnesium oxysulfate cement is formulated by the reaction between magnesium oxide and magnesium sulfate solution, and like that of magnesium oxychloride it has very good binding properties. The strength of MOS cement is on par with that of MOC cement provided similar porosities are considered (Beaudoin and Ramachandran, 1978). Four oxysulfate phases are formed at temperatures between 30 and 1208C: 5Mg(OH)2.MgSO4.3H2O (5-form), 3Mg(OH)2.MgSO4.8H2O (3-form), Mg(OH)2.MgSO4.5H2O, and Mg(OH)2.2MgSO4.3H2O. Only the 3-form and 5-form are stable at 258C (Cole and Demediuk, 1957). The phase relationship in MOS cement has also been studied by Urwongse and Sorrell (1980b). The major use of MOS cement is in the manufacture of lightweight insulating panels. MOS cement suffers from the same lack of water resistance as does MOC cement.

16.4 THERMAL INSULATIVE AND FIRE RESISTANCE PROPERTIES OF SOREL CEMENT 16.4.1 Thermal Insulation The ASTM-tested (American Society for Testing and Materials) foamed MOC cement is used as thermal insulation in the cavity of house walls

16.5 MAGNESIUM PHOSPHATE CEMENT

235

(Rossiter and Brown, 1990). The foamed cement was formulated as a calcium magnesium oxychloride silicate composition and was pumped into the cavity walls of the test house. The density of cement was 46.5 kg m23 and had a measured thermal conductivity of 0.0430 W/m8C at 248C, which is comparable to other commercially available insulation materials. 16.4.2 Fire Resistance The water of hydration and hydroxyl water associated with the MOC 5-form and 3-form are 44 and 49%, respectively. When heated to 2978C, the chemically bound water will be converted to steam with an energy requirement of about 1000 Btu per pound of water released. The MOC cement beneath the surface exposed to the heat will not be heated above this temperature until all of the water has been released and driven from the cement. Because of the high energy requirement for this process to occur, the insulative effect from the water of hydration is considerable and constitutes the principle means of insulation (Montle and Mayhan, 1973). Thermally decomposed MOC cement is primarily MgO and as such has a high reflectivity, which is also a significant factor in the overall insulative capability of magnesium oxychloride cement. It has been calculated that a 5-cm thickness of typical MOC cement with a density of 960 kg m23, containing approximately 35% bound water and no fillers, requires over 6 h for the nonheated face to reach a temperature of 10008F (5388C).

16.5 MAGNESIUM PHOSPHATE CEMENT Cements based on the reaction between magnesium oxide and soluble phosphate salts have long been used as investment cements for the casting of alloys. Magnesium phosphate cement has a rapid set and very high early strength and has found utility as a rapid patching mortar for roads and aircraft runways, which can typically be reopened after about 45 min (El-Jazari, 1982). Compressive strengths of 13.8 MPa can be developed in under 2 h. Practical cements, however, are only formed by the reaction of magnesium oxide with the water-soluble acid phosphate salts of ammonium, aluminum, and potassium; the reaction between MgO and phosphoric acid results in a water-soluble cementitious mass that has no practical use. 16.5.1 Reaction Mechanism Cement formation between MgO and various acid phosphate salts involves both acid–base and hydration reactions, and the reaction products can be

236

MAGNESIA CEMENTS

either crystalline or amorphous. The most important characteristic of the magnesium oxide powder used to form the cement is its reactivity. Hardburned or dead-burned magnesium oxides are typically used, which have a specific surface area less than 1.0 m2 g21. Soude´e and Pe´ra (2002) have examined the influence of calcining and grinding of the MgO on its reactivity in MAP cement. This low reactivity is necessary due to the rapidity of the reaction with acid phosphate salts; lighter burned magnesium oxides react almost instantaneously such that there is virtually no working time available. The rapidity of the set of MAP cement can be ameliorated by the use of set retarders. Sodium tetraborate decahydrate (borax) can the used to decrease the rate of reaction between MgO and the acid phosphate salt. The normal set time for nonretarded cement is about 3 min, however, the addition of 20% borax can extend the set time to .20 min. It is thought that borax ions, which then act as an Mg2þ cation acceptor. This yields B4O22 7 complex precipitates on the surface of the MgO grains, forming a barrier that slows the release of further Mg2þ ions, retarding the set of the cement (Sugama and Kukacka, 1983a). 16.5.2 Magnesium Phosphate Cement Derived from Ammonium Dihydrogen Phosphate The reaction between magnesium oxide and ammonium dihydrogen phosphate (ADP) in water yields struvite, MgNH4PO4.6H2O, and schertelite, Mg(NH4)2(HPO4)2.4H2O; see reaction (16.1) (Abdelrazig et al., 1988, 1989): MgO þ 2NHþ 4 þ 2H2 PO4 þ 3H2 O ! Mg(NH4 )2 (HPO4 )2
4H2 O (16:1) If there is sufficient water present, then the reaction can go to completion with the subsequent formation of struvite: Mg(NH4 )2 (HPO4 )2
4H2 O þ MgO þ 7H2 O

! 2MgNH4 PO4
6H2 O

(16:2)

The reaction is exothermic, which occurs with the loss of some ammonia, and can be retarded by the addition of borax or sodium tripolyphosphate (STPP). The reactions described by Abdelrazig et al. (1988 and 1989). show that the solution of ADP in water forms ammonium and phosphate ions immediately, which react with the magnesium oxide particles in suspension. At the initial stages of the reaction, when phosphate concentration is at its highest, the intermediate tetrahydrate schertelite is formed. As the

16.5 MAGNESIUM PHOSPHATE CEMENT

237

reaction proceeds, and the concentration of phosphate decreases, the formation of struvite is favored. Abdelrazig et al. formed mortars containing MgO : ADP : silica : water ¼17.1 : 12.9 : 70.0 : 12.5 water/solid ratio 1 : 8 that set within 7 min of mixing. Crystal growth of struvite was appreciable within 5 min and the rodlike crystallites of struvite appeared to grow between the silica grains. The compressive strength of the mortar reached 14.9 MPa (2161 psi) in 24 h. The presence of STPP in the mortar improved compressive strength and reached 19.5 MPa in 24 h along with a reduction of the total pore volume. In the presence of STPP there is also a change in crystal morphology from acicular to hexagonal plates. A water deficient mortar (water : solid ¼ 1 : 16) still had ADP present after 3 weeks and was slower to set. Schertelite was the major reaction product initially; however, after a while struvite became the major product with schertelite still being a significant constituent. The incompletely hydrated cement is stronger than the fully hydrated version having a compressive strength of 27.4 MPa (3794 psi). This is in keeping with the general rule that the strength of cement increases with decreasing water content. The commercial magnesium phosphate cement SET-45, produced by Master Builders, sets in 15 min and hardens in 30–60 min. The compressive strength reached 24 MPa in the first hour and 47 MPa after 24 h. After 1 month the compressive strength reached 56 MPa (8122 psi). 16.5.3 Magnesium Phosphate Cement Derived from Diammonium Phosphate Reaction of magnesium oxide with aqueous solutions of diammonium phosphate [(NH4)2HPO4)] have been investigated by Sugama and Kukacka (1983a). This yields cement with a compressive strength of 6.2 MPa (900 psi) after 1 h. X-ray diffraction studies of the cured cement revealed that the major reaction product was magnesium orthophosphate tetrahydrate [Mg3(PO4)2.4H2O]. Struvite and magnesium hydroxide were also detected. However, the work of Abdelrazig and Sharp (Abdelrazig and Sharp, 1985) disagreed with the work of Sugama and Kukacka (1983a) and concluded that the most likely reaction products are the hydrates of the magnesium ammonium phosphates: struvite, dittmarite, and schertelite. 16.5.4 Magnesium Phosphate Cement Derived from Ammonium Polyphosphate Ammonium polyphosphate solution (APP) contains about 44% water and is known commercially as Poly-N. The major constituents of APP are pyrophosphate, along with small amounts of tri- and tetrapolyphosphate. These

238

MAGNESIA CEMENTS

polyphosphates all have linear chain structures, and the total amount of phosphate as polyphosphate ranges from 56 to 58%. Poly-N is commonly used as an agricultural fertilizer. The reaction between APP and hard-burned MgO is strongly exothermic, with the cement setting within 3 min. The early strength of the cement was 13.8 MPa (2000 psi) after 1 h and over 20 MPa (2900 psi) after 5 h (Sugama and Kukacka, 1983b). Struvite was identified as the major cementitious phase by X-ray diffraction. 16.5.5 Magnesium Phosphate Cement Derived from Potassium Dihydrogen Phosphate The reaction between hard-burned MgO, water, and potassium dihydrogen phosphate produces a quick setting cementitious mass, KMgPO4.6H2O. This cement system was originally developed at Argonne National Laboratory for the stabilization and encapsulation of hazardous and radioactive wastes (Wagh et al., 1998). BIBLIOGRAPHY Wilson, A. D., and Nicholson, J. (1993). Acid-Base Cements and Their Biomedical and Industrial Applications, Cambridge University Press, Cambridge, United Kingdom.

REFERENCES Abdelrazig, B. E. I., and Sharp, J. H. (1985). Cement Concrete Res. 15, 921– 922. Abdelrazig, B. E. I., Sharp J. H., and El-Jazairi, B. (1988). Cement Concrete Res. 18, 415– 425. Abdelrazig, B. E. I., Sharp J. H., and El-Jazairi, B. (1989). Cement Concrete Res. 19, 247– 258. Beaudoin, J. J., and Ramachandran, V. S. (1978). Cement Concrete Res. 8(1), 103– 112. Cole, W. F., and Demediuk, T. (1955). Aust. J. Chem. 8(2), 234– 251. Cole, W. F., and Demediuk, T. (1957). Aust. J. Chem. 10, 287 –294. Dehua, D., and Chuanmei, Z. (1999). Cement Concrete Res. 29, 1365 – 1371. Deng, D. (2003). Cement Concrete Res. 33, 1311 –1317. El-Jazari, B. (1982). Rapid Repair of Concrete Paving, Concrete, pp. 12 –15. Harper, F. C. (1967). J. Appl. Chem. 17, 5 –10. Matkovic, B. (1981). Federal Highway Aministration Report, No. FHWA/ RD-80/128.

REFERENCES

239

Matkovic, B., and Young, J. F. (1973). Nature (London), Phys. Sci. 246(153), 79 – 80. Matkovic, B., Rogic, V., and Kaluza, F. (1976). Kem. Ind. (Zagreb) 25(11), 610 – 624. Montle, J. F., and Mayhan, K. G. (1973). The Role of Magnesium Oxychloride as a Fireproofing Material, AICHE Meeting. Rossiter, W. J., and Brown, P. W. (1990). Standard Technical Publication of the ASTM, STP 1030, pp. 38 –51. West Conshohocken, PA. Sorrell, C. A., and Armstrong, C. R. (1976). J. Am. Ceram. Soc. 59(1 –2), 51– 54. Soude´e, E., and Pe´ra, J. (2002). Cement Concrete Res. 32, 153– 157. Sugama, T., and Kukacka, L. E. (1983a). Cement Concrete Res. 13, 407 –416. Sugama, T., and Kukacka, L. E. (1983b). Cement Concrete Res. 13, 499 –506. Urwongse, L., and Sorrell, C. A. (1980a). J. Am. Ceram. Soc. 63(9 –10), 501– 504. Urwongse, L., and Sorrell, C. A. (1980b). J. Am. Ceram. Soc. 63(9– 10), 523– 526. Vereshchagin, V. I., Smirenskaya, V. N., and E´rdman, S. V. (1997). Glass Ceramics 54(11 –12), 368 –372. Wagh, A. S., Singh, D., and Jeong, S. -Y. (1998). U.S. Patent 5,830,815.

17 MISCELLANEOUS MAGNESIA APPLICATIONS

17.1 SUGAR MANUFACTURE The process of raw sugar manufacture is essentially that of separating by chemical and mechanical means sucrose from the sugar cane. The process can be broadly divided into the following steps; see Figure 17.1 for a flow diagram of the process: .

.

.

Juice Extraction Juice is extracted from the cane either by milling between heavy rolls or by leaching with water. Juice Screening The bagasse (spent cane) is then screened from the sugar cane juice. Clarification The juice from the mill contains many impurities, such as field soil, silica, bacteria, yeast, molds, and spores along with waxes, starches, and gums. Owing to the presence of natural enzymes in the juice, at this stage in the process, sugar inversion (conversion of sucrose to glucose and fructose) is occurring. To stop the enzymatic action, the juice is raised to a pH of ca. 7.5 and heated to near 1008C. At the same time, a large portion of the suspended matter is removed by settling in clarification tanks. Traditionally, lime is added at this stage (also known as defecation) to aid in the clarification process by the

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 241

242

Figure 17.1

Schematic diagram of cane sugar manufacturing process.

17.1 SUGAR MANUFACTURE

.

243

formation of insoluble precipitates, such as calcium phosphate from phosphate present in the juice. The mud produced in the defecation process is then filtered off from the clarified juice. Evaporation The clarified juice is now concentrated by the removal of water to the point of crystallization of sugar. Since clarified cane juice contains significant quantities of inorganic ions, including calcium (introduced by the defecation process) and sulfate, the heating surfaces of the evaporators are quick to scale and require frequent cleaning.

The use of magnesium oxide as a partial replacement for lime has been found to have the following benefits. 17.1.1 Clarification and Filtration The use of MgO in the defecation process results in a higher density and faster settling mud, along with an increase in clarifier capacity. Filtration capacity is also increased since a thicker, denser mud is obtained. Filtering efficiency is also improved by virtue of the reduced volume of more easily filtered mud. 17.1.2 Reduction in Scaling Magnesium oxide reduces the amount of scaling from calcium deposits in the juice heaters, evaporators, and vacuum pans. The neutralization of cane juice using MgO, instead of lime, produces soluble magnesium salts, such as magnesium sulfate. Lime produces calcium sulfate that precipitates from solution and forms a hard scale on exposed metal surfaces. In practice, it is necessary to only add sufficient magnesium oxide to keep the calcium content of the limed juice below a critical level. Although this level depends to a large extent on the factory producing the sugar and on the season of production, in general, the critical level is about 260 ppm CaO in the clarified juice. If the CaO is kept below 260 ppm by addition of MgO to the lime, then scale formation is significantly reduced. This can be achieved by substituting approximately 30–40% of the lime by MgO. 17.1.3 Additional Benefits The ash content of raw sugar has been found to be lowered by the use of magnesium oxide, along with a stabilization of pH of the juice in the clarifiers during periods of downtime, thus reducing the loss of sucrose due to inversion, which normally occurs when there is a decrease in juice pH. The use of

244

MISCELLANEOUS MAGNESIA APPLICATIONS

magnesium oxide has also been found to result in a reduction in the quantity of molasses produced as a percentage of net cane with little effect on its purity. 17.2 CHROME TANNING OF LEATHER The primary function of a tanning agent is to stabilize the collagen fibers of the animal hide so that they are no longer subject to biodegradation. Over 95% of the leather produced in the United States is chrome tanned. The initial preparation of the raw hide involves the following steps: . . .

.

Hair removal using a solution of lime and sodium sulfide. Removal of excess fat using “flesh scraping” machines. Deliming—This involves washing the pelt using aluminum sulfate solution to remove any remaining lime from the hide. Enzymatic bating—A pancreatic enzymatic agent is used to open up the fiber structure by degrading proteins.

Once the pelt has undergone the above preparation, it is now ready for tanning. The hides are placed in a solution of sodium chloride and sodium formate buffer; the presence of the salt prevents the acid added in the next step from damaging and swelling the hide. Sulfuric acid is then pumped into the moving drum and rotated for 1 – 3 h where the pH is depressed to less than 3.0. Chrome tanning salt [chromium(III) sulfate] is then dissolved in the acid solution and allowed to fully penetrate the hide. Once this has occurred, the pH of the solution is slowly raised using magnesium oxide. The MgO used in this step is generally a hard-burned high-purity grade; the lower reactivity is necessary so that it does not raise the pH too swiftly and cause spotting of the pelt. Once the pH has reached between 3.4 and 3.6, the chrome has reacted with the collagen to produce a fully preserved, tanned hide, at which point it said to be in the blue. Chromium(III) sulfate forms many different types of complexes in solution (Slabbert, 1975), and the reaction of these basic chromium sulfate tanning solutions with collagen have been studied (Shuttleworth, 1958). It has been firmly established that cross-linking is accomplished by various chromium species with free carboxyl groups in the collagen side chains. The purpose of the magnesium oxide is to provide alkalinity for the olation reaction, which involves elimination of water and formation of a linkage between two collagen chrome complexes. After chrome tanning the leather then undergoes further tanning with synthetic resins and natural tannins and is then finally dyed and “fat liquored”

17.3 MAGNESIA AS A CATALYST SUPPORT

245

with natural and synthetic oils to replace the natural greases removed in the preceding steps.

17.3 MAGNESIA AS A CATALYST SUPPORT The purpose of the catalyst support is to provide a solid surface onto which the catalytic component can be coated. This should result in a stable structure that can withstand the rigors of the environment under which the catalyst will perform, which will typically be high temperatures and gas or liquid flow. Another purpose of the support material is to act as a thermal stabilizer and prevent the agglomeration of lower melting point catalytic materials such as copper, platinum, and the oxides of zinc and molybdenum. Since the catalytic metals typically have a higher density than the support and are also more expensive, the support acts to decrease the overall density of the catalytic system as well as diluting the more expensive component with a less expensive one. Magnesium oxide has all of the above desired properties; it has a high melting point (28008C) as well as being a basic material that will enhance reactions that are benefited by an alkaline environment. Furthermore, magnesium hydroxide has a tendency to form a hydrous gel, which upon drying, coalesces and densifies forming a compact structure with a small pore size but high surface area. One method of preparing the catalyst is to co-precipitate magnesium hydroxide from magnesium nitrate solution along with the catalytic components, which provides the desired intimate contact between catalyst and support. An added benefit of the hydroxide is that it lends itself to extrusion as well as prilling. Since magnesium hydroxide, and typically the catalytic metals, form insoluble carbonates, hydrated magnesium carbonate can be co-precipitated along with the carbonate of the catalytic material. However, magnesium carbonate supported catalyst can only be used at temperatures well below its decomposition point, otherwise weakening and disintegration of the support may occur. 17.3.1 Example of Magnesium Oxide as Catalyst Krylov et al. (1965) have investigated the basic properties of MgO by studying the adsorption and isotopic exchange of carbon dioxide on both MgO and partially dehydrated magnesium hydroxide. Three types of basic centers were identified to be present at the surface of partially dehydrated magnesium hydroxide: (a) strongly basic O22 centers that are transformed into carbonate anions upon adsorption of CO2, (b) strongly basic centers that were derived from O22 ions that were adjacent to surface OH groups, and (c) weakly basic surface OH groups.

246

MISCELLANEOUS MAGNESIA APPLICATIONS

The electron-donor centers on metal oxides can be measured by adsorbing certain organic molecules on the surface of the oxide. The transfer of an electron from the donor site of the oxide to the adsorbed molecule creates a paramagnetic ion detectable using electron paramagnetic resonance (EPR) spectroscopy. Che et al. (1972) studied the adsorption of tetracyanoethylene (TCNE) on MgO that had been pretreated between 100 and 8008C. Using EPR methods they identified the presence of adsorbed TCNE radical anion. As the pretreatment temperature increased from 100 to 8008C, the concentration of the radical anion passed through two maxima, one at 2008C and the other at 7008C. The electron-donor centers were found to be associated with OH2 and O22 ions with a low coordination number. The isomerization of 1-butene (CH255CH22CH222CH3) over magnesium oxide has been studied by Baird and Lunsford (1972) as a function of pretreatment temperature, which was varied from 300 to 9008C. The results showed that the isomerization reaction was dependent upon pretreatment temperature, as was the initial cis/trans ratio of the product 2-butene (CH322CH55CH22CH3). The reaction rate was also found to pass through a maximum at 7008C. The proposed mechanism is that 1butene adsorbs onto the MgO surface where O22 ions on corner sites aid in removing an allylic proton and transferring it to the terminal methylene group.

17.4 FUEL ADDITIVES The combustion of sulfur-containing fossil fuels such as coal and oil generates not only sulfur dioxide but also a small, but significant, quantity of sulfur trioxide (SO3). Since power generation utilities consume enormous quantities of both coal and oil, the quantities of sulfur dioxide and trioxide produced is also large. Some of the SO3 is generated in the combustion flame, while the remainder is produced by the catalytic oxidation of sulfur dioxide by contact of the gas with solid surfaces such as vanadium oxide (which is a common constituent of oil ash), iron, and other metals present throughout the boiler. The conversion rate of SO2 to SO3 is typically around 0.5% (Cullis and Mulcahy, 1972). Typical SO3 levels are 5–20 ppmv if excess oxygen in the gas after combustion is 0.5–1.0% (Davies et al., 1981). Sulfur trioxide then reacts with moisture present in the flue gases to form sulfuric acid. However, sulfur trioxide present in the combustion gases as well as vanadium compounds present in fuel oil can also cause significant damage to the boiler components in the form of acid corrosion and slag deposit corrosion.

17.4 FUEL ADDITIVES

247

17.4.1 High-Temperature Corrosion High-temperature corrosion problems are experienced mainly by boilers firing residual fuel oils. The corrosion is due primarily to the presence of vanadium, sodium, and sulfur compounds in the fuel oil (vanadium can be as high as 500 ppm, Na 300 ppm, and sulfur 40,000 ppm). During combustion the presence of these compounds react and give rise to complex lowmelting-point materials that deposit on heat-transfer surfaces and supporting structures; see reactions (17.1)–(17.3) (Niles and Sanders, 1962): Na2 SO4 þ V2 O5 ! 2NaVO3 þ SO3 Na2 SO4 þ 3V2 O5 ! Na2 O
3V2 O5 þ SO3

(17:1) (17:2)

Na2 SO4 þ 6V2 O5 ! Na2 O
6V2 O5 þ SO3

(17:3)

The sodium vanadates (Na2O.xV2O5, where x ¼ 1, 3, or 6) formed during combustion, which are present in the fuel oil ash, are low-melting-point compounds (typically in the range 580– 6008C), which deposit and stick tenaciously to boiler surfaces. These deposits also trap large amounts of ash, which impedes heat transfer and affects boiler efficiencies. The mechanism of corrosion by vanadates has not been elucidated; proposed mechanisms include: (1) vanadates serve as oxygen carriers, (2) they distort the normal crystal lattice of the protective iron oxide film on the metal surfaces, and (3) molten vanadates dissolve the protective iron oxide layer. It is generally thought that a liquid film speeds up corrosion; so metal wastage will be greatest when the vanadates have a melting point lower than the temperature of the metal substrate. The other mechanism involves the alkali metals present in the fuel. The iron oxide boiler surface is covered with an alkali sulfate layer that is formed from alkalis present in the fuel reacting with SO2. Particles of ash adhere to this alkali layer and accumulate. As the ash layer becomes thicker, a thermal gradient through it increases and the exposed outer layer melts. The increased temperature within the deposit leads to dissociation of sulfates with subsequent release of SO3, which then migrates to the cooler oxide-covered surface. The SO3 reacts with the alkali sulfates and the iron oxide surface, which then dissolves the protective iron oxide film and forms alkali iron trisulfates; see reactions (17.4)–(17.7). When deslagging occurs, the layer spalls off leaving a fresh iron surface to be oxidized and start the cycle over again (Reid, 1974): Na2 O þ SO2 þ 12 O2 ! Na2 SO4

(17:4)

Na2 SO4 ! Na2 O þ SO3

(17:5)

248

MISCELLANEOUS MAGNESIA APPLICATIONS

3Na2 SO4 þ 3SO3 þ Fe2 O3

! 2Na3 Fe(SO4 )3

4Fe þ 3O2 ! 2Fe2 O3

(17:6) (17:7)

High-temperature corrosion can be combated by the introduction of MgO into the boiler where it reacts with vanadium pentoxide to form magnesium orthovanadate; see reaction (17.8): 3MgO þ V2 O5 ! 3MgO
V2 O5

(17:8)

Magnesium orthovanadate has a melting point of around 11908C, which is considerably higher than that of sodium –vanadium compounds. This means that magnesium orthovanadate has a lesser tendency to deposit; and, when it does, it is loose and powdery and easily removed using the soot blowers (Salooja, 1972). 17.4.2 Low-Temperature Corrosion The formation of sulfur dioxide during combustion gives rise to sulfuric acid when the flue gas temperature drops below the acid dewpoint, which itself is dependent upon SO3 concentration (Reese et al.). This can occur in the economizer and air heater sections and leads to corrosion of the heat exchangers and supporting structures as well as forming a visible plume of acid aerosols. The acid deposit, which is 75–80% H2SO4, also creates a sticky surface that collects soot particles. These soot agglomerates, the so-called acid smuts, can eventually reentrain in the flue gas and are emitted from the stack into the surrounding environment causing significant damage. Another, more recent event affecting the formation of sulfur trioxide in power generation boiler systems is the advent of NOx reduction technology in the form of selective catalytic reduction (SCR). Although this technology reduces the NOx content of flue gas, it has also been found to be effective in catalyzing the oxidation of sulfur dioxide to sulfur trioxide, thus exacerbating an already challenging problem. Magnesium compounds, such as MgO and Mg(OH)2, and magnesium carbonate can again be used to ameliorate the problem of low-temperature corrosion (Alexander et al., 1961). Injection of these compounds into either the furnace or into the economizer outlet effectively neutralizes sulfuric acid forming magnesium sulfate (Lewis and Goldberg, 1978). MgO has also been found to coat the superheater surfaces, thus preventing catalytic formation of a large amount of the SO3 (Barrett, 1967). 17.4.3 Types of Additives for High-Temperature Corrosion Additives for high-temperature oil and coal-ash fouling and corrosion can be applied as powders or slurries. The form of the additive determines the

17.4 FUEL ADDITIVES

249

delivery system and point of addition into the combustion train. The additive can react with the problem-causing components of the oil ash in the gas stream before they reach the tube section, or they can be applied directly to the deposits on the tube. Because of the ease of application, most additives are applied as dispersions. The dispersions can be either water or oil based; oil-based dispersions are generally metered into the fuel while water-based dispersions are applied through the soot blowers. Magnesium oxide particles in the 2- to 7-mm median particle size range deposit in the convection pass preferentially, while finer sized particles are not deposited and tend to get carried into the back end of the boiler. When an MgO of 2-mm median particle size is fed into the fuel at treatment rates of 0.5–1.0 parts by weight Mg to 1.0 part V in the oil, the quantity of MgO deposited on the superheater tubes is 2–3 times that of the V and Na. This treatment rate results in a high Mg : V ratio, which is needed to raise the ash-melting point and reduce tube deposits. Application of aqueous MgO-based slurry via the soot blowers should theoretically result in lower treatment rates than other modes of application. However, data suggests that dosage rates are in the same range as oil-based MgO dispersions. However, with this mode of application, inadequate dosage or poor distribution can result in an increased ash burden, thus exacerbating the problem. 17.4.4 Types of Additives for Low-Temperature Corrosion Magnesia-based additives for treating low-temperature corrosion and fouling in oil-fired and coal-fired boilers can again be applied in either fine powder (325 mesh) or slurry form. The magnesia-based additives are used to neutralize sulfuric acid after it has been formed in the combustion train and again coats iron oxide surfaces thus interfering with catalysis of SO2 to SO3. While MgO is a good acid neutralizer, for it to be effective for low-temperature corrosion, it must be transported to the cold end of the boiler by flue gas or added at the economizer outlet. When the additive is added to the fuel or added into the fire side of the boiler, the transport to the cold end is, among other factors, dependent upon the particle size distribution of the particulates; the finer the size, the further they travel to the cold end. Powders are best added at the economizer outlet where it is available to neutralize acid condensing in the cold end of the boiler. The feed rate for powders injected into the economizer outlet for boilers firing high-sulfur oil (2% or greater S) and operating with low excess air is 165–621 lb MgO equivalent per 1000 barrels of oil. The feed rate for oilbased magnesia dispersions is 127 lb MgO equivalent per 1000 bbl of oil. Coal-fired boilers, using a high-sulfur bituminous coal, typically require

250

MISCELLANEOUS MAGNESIA APPLICATIONS

0.27 lb Mg(OH)2 per ton of coal when dosed into the upper furnace as an aqueous slurry (Radway and Rohrbach, 1976, and Radway and Boyce, 1977). Oil-based dispersions fed into the upper furnace require approximately 0.4–0.6 lb MgO equivalent per ton of coal (Dainoff and Schenck, 1984).

17.5 WELL-DRILLING FLUIDS Well-drilling fluids perform a variety of functions such as: . . . . .

. .

Cooling and lubrication of the drill bit. Removal of cuttings from the well bottom. Prevention of circulation loss. Seal formation fluids from well bore. Control subsurface pressures. As the well becomes deeper, it is necessary to increase the drilling mud density to hold back subsurface formation pressure. Transmit hydraulic power to the bit. Control of annulus and pipe corrosion by maintaining a high pH.

Drilling muds are a sophisticated mixture of many additives that alter the fluid density, rheology, and pH, which is beyond the scope of this book to cover this topic in detail. Magnesia is used in brine drilling mud to act as scouring agent in order to keep the annulus clean as well as to neutralize acid gases released from the formation during the drilling operation. MgO is also used to control fluid alkalinity, particularly in clay-free polymerbased fluids (Jackson, 1976). Attapulgite clay is also used as a rheology modifier in drilling mud, and the addition of small quantities (approximately 2% by weight) of lightburn MgO helps to increase fluid viscosity. Dead-burned magnesia has been employed in expansive cement used to seal a steel casing to the bore-hole walls (Cheung, 1999). The dead-burned magnesia slowly hydrates and expands the cement to compensate for any shrinkage that the cement should undergo during curing. Sorel cement has been reported to be useful as a bridging agent in drilling and servicing fluids. A bridging agent deposits as a filter cake on the walls of the well bores within the producing formations. This prevents the drilling or servicing fluid from being lost into the formation and also prevents solids from entering the porosity of the formation and causing blockage. The Sorel cement bridging agent can be removed by dissolving with an organic acid, once drilling is complete, to allow the production of hydrocarbons (Bradley, 2002).

17.6 NANOPARTICULATE MAGNESIA

251

17.6 NANOPARTICULATE MAGNESIA The blossoming science of nanotechnology holds promise of development in many areas from pharmaceuticals to material science. Nanoscience deals with the realm of particles sizes between that of molecules and that of condensed matter, in the range of 1– 100 nm. Many physical and chemicals properties vary with the size of the particle, such as magnetic and optical properties, melting points, specific heats, and surface reactivity. The hot pressing of nanoparticulate ceramic powders has yielded ceramic bodies with higher strength and flexibility than conventionally prepared articles. The surface activity of nanoparticles is dramatically different from that of macroparticles. Macroparticles are loosely those in excess of 100 nm, although this borderline is somewhat arbitrary. Nanocrystalline powders can have enormous surface areas. When one considers the fact that a 3-nm diameter particle has about 50% of its atoms on the surface, these types of particles can react as nearly stoichiometric reagents. The properties of nanoparticulate magnesia has been studied quite extensively as have the surface properties of microparticulate magnesium oxide. It has been found that nanoscale MgO particles display unusual surface morphologies and have a more reactive surface due to the presence of more surface defects at edge/corner sites. It has been possible to prepare MgO with surface areas greater than 500 m2 g21 and crystallite size ,4 nm by the use of alcogel-aerogel techniques (Utampanya et al., 1991; Itoh et al., 1993). This high surface area suggests that about 40% of the atoms are on the surface of the particle, which gives this material a tremendous gas adsorption/reaction capacity. 17.6.1 Synthesis of Nanoparticulate Magnesia Conventional methods of preparation of magnesium oxide yield products that have large and varied grain sizes and fairly low surface areas. The most popular method of nanoparticle synthesis has been via sol –gel processing. Other liquid-phase methods involve the use of hydrothermal synthesis, which has yielded rod, tube, and needle-shaped morphologies (Ding et al., 2001). Klabunde (2001) has reviewed the various synthetic methods. 17.6.2 Chemical and Catalytic Properties of Nanocrystals Surface chemistry plays a vital role in many processes, including corrosion, oxidation, reduction, catalysis, and adsorption. As mentioned above, as a particle decreases in size, a greater percentage of its constituent atoms are featured on the surface compared with the atoms present in the bulk of the

252

MISCELLANEOUS MAGNESIA APPLICATIONS

particle. Whereas a 3-nm particle has 50% of its atoms exposed on the surface, a 20-nm particle has fewer than 10%. Also, as a particle decreases in size, crystal shape may also change as well. In the case of MgO, the crystal shape tends to change from cubic to polyhedral as the size decreases. Nanocrystalline MgO possessing superbasic sites has been prepared by the doping of potassium atoms using vapor diffusion (Sun and Klabunde, 1999a). This creates Kþ and e2 sites on the crystal surface where the free electron associates with surface sites to produce superbasic reactive zones. Exposure of alkenes to this material result in the formation of allyl anions, formed by proton abstraction, which can alkylate ethane; see reaction (17.9): CH3 CH55CH2 þ CH255CH2 ! CH3 CH2 CH2 CH55CH2

(17:9)

Nanocrystalline MgO has demonstrated some unusual catalytic properties when chlorine gas is absorbed into the nanocrystals (Sun and Klabunde, 1999b). Contacts of this adduct with alkanes results in their catalytic chlorination; see reaction (17.10). It appears that the chlorine is behaving in a manner more consistent with chlorine atoms being formed on the surface of the MgO by dissociative chemisorption. MgO22MgO þ Cl2 ! MgO
Cl22MgO
Cl þ RH ! MgO
R22MgO
Cl þ HCl ! RCl þ MgO22MgO

(17:10)

Nanocrystalline MgO also has a large capacity for adsorbing acid gases such as SO2, CO2, HCl, HBr, and SO3 at near stoichiometric proportions. Work by Klabunde has demonstrated that nanocrystalline magnesium oxide has an enhanced surface reactivity over that anticipated from surface area alone. Nanocrystalline MgO chemisorbed 6 molecules of SO2 per nm2, whereas larger microcrystalline MgO only chemisorbed 1.8 molecules per square nanometer (Stark and Klabunde, 1996). The proposed mechanism for the difference in adsorption capacities is due to a monodentate-type adsorption mechanism of SO2 via the sulfur atom in the instance of nanocrystalline MgO, whereas microcrystalline MgO favors a bidentate adsorption mode through sulfur and an oxygen atom. The surface of nanocrystalline MgO has been found to interact strongly with polar organic molecules, such as aldehydes, ketones, and alcohols, by a dissociative chemisorption process, which results in the destruction of the organic molecule. This is in contrast to high-surface-area activatedcarbon absorbents, which merely absorb the moiety with no resultant reaction. The chemisorption of acetaldehyde on MgO nanocrystals results in

17.6 NANOPARTICULATE MAGNESIA

253

an exothermic reaction and the development of an orange color. The proposed reaction mechanism is the loss of the aldehydic proton by initial coordination of the carbonyl oxygen to Mg2þ, followed by Hþ transfer to the surface O22 with the probable formation of polymeric materials (Khaleel et al., 1999). This chemisorption process has proven useful for the development of a new class of materials that could find utility for protection against chemical and biological warfare agents. These agents are either organophosphonates or sulfur halides, which have been found to interact strongly with nanocrystalline MgO, at room temperature, with the production of nontoxic residues (Wagner et al., 1999a,b). The nerve agent Soman [CH322(F)P(55O) 22OCH(CH3)C(CH3)3] reacts with MgO nanoparticles with the elimination of HF by utilization of residual 22OH groups. The HF reacts with basic MgO to produce MgF2 and water. The nerve agent VX [o-ethyl-S-(2-diisopropylamino)]ethylmethylphosphonothiolate and mustard gas [Cl(CH2)2S(CH2)2S] were also found to be destructively chemisorbed onto nanocrystalline MgO. All of these reactions appear to require the presence of trace quantities of water, which suggests that surface hydroxyl groups play a role in the reaction mechanism. Virus, bacteria, bacterial spores, as well as toxins have been found to be degraded by MgO chlorine adducts. The bacterial outer membrane appears to be ruptured by the oxidizing action of the chlorine as well as the abrasive action of the MgO crystals. It also appears that the polar bacterial DNA (deoxyribonucleic acid) is strongly adsorbed by the MgO. Magnesium oxide nanocrystals containing a monolayer of a transitionmetal oxide on the surface have a greatly enhanced efficiency for the destructive adsorption of chlorocarbons, hydrogen sulfide, and sulfur dioxide (Klabunde et al., 1995; Decker and Klabunde, 1996; Jaing et al., 1998). A layer of vanadium pentoxide on MgO nanocrystals catalyzes the decomposition of carbon tetrachloride at 4508C to carbon dioxide and magnesium chloride; see reaction (17.11). The initial reaction results in the reaction of CCl4 with V2O5 to form a vanadium chloride moiety and carbon dioxide. The exchange of chloride ion with a crystal lattice oxide anion, (O22) forms magnesium chloride and the regeneration of vanadium oxide: MgO22V2 O5 þ CCl4 ! MgO22VClx ! MgCl222V2 Ox

(17:11)

The exchange of lattice oxide anions in nanocrystals of MgO has been observed when H2O18 is passed over MgO particles at various temperatures (Li and Klabunbe, 1992). At 3008C, the surface hydroxyls underwent exchange; however, at higher temperatures surface lattice O22 exchanged, while at 7008C, inner lattice oxide anions exchanged.

254

MISCELLANEOUS MAGNESIA APPLICATIONS

17.7 TRANSFORMER STEEL COATING Magnesium oxide is used to produce the insulating glassy layer on the surface of grain-oriented silicon steel for use in electrical transformers. This is achieved by coating the steel surface with aqueous slurry of magnesium oxide, which during the high-temperature annealing process reacts with silicon at the steel surface to form a glassy insulating layer of magnesium silicate. The magnesium oxide has to be of high purity and contain very low levels of elements such as chlorine, bromine, calcium, fluoride, and sodium, which can produce an unacceptably high electrical conductivity in the insulting layer. The MgO also has to be very finely sized to produce slurry that is stable and has the desired coating characteristics. BIBLIOGRAPHY Betz Handbook of Industrial Water Conditioning, 6th ed., 1962, Trevose, PA. Hamilton, R. K., and Payne, J. H. (1965). 12th Congress of the International Society of Sugar Technologists. Paper No. 52, San Juan, Puerto Rico. Radway, J. E., and Hoffman, M. S. (1987). Operations Guide for the Use of Combustion Additives in Utility Boilers, CS-5527, Research Project 1839-3, EPRI. The Use of Magnesium Oxide to Prevent Evaporator Scaling in Sugar Factories. Premier Chemicals, LLC. King of Prussia, PA.

REFERENCES Alexander, P. A., Fielder, R. S., Jackson, P. J., Raask, E., and Williams, T. B. (1961). Inst. Fuel 34, 53. Baird, M. J., and Lunsford, J. H. (1972). J. Catal. 26, 440 –450. Barrett, R. E. (1967). Trans. ASME J. Eng. Power Ser. A 89, 288. Bradley, T. L. (2002). Europ. Patent 1 223 207 A1. Che, M., Naccache, C., and Imelik, B. (1972). J. Catal. 24, 328. Cheung, P. S. (1999). U.S. Patent 5,942,031. Cullis, C. F., and Mulcahy, M. F. R. (1972). Combustion Flame 18(2), 225. Dainoff, A. S., and Schenck, H. N. (1984). Mechanisms for the MgO Treatment of Coal-Fired Utility Boilers. Presented at the Engineering Foundation Conference on Fouling and Slagging from Impurities in Combustions Gases at Copper Mountain, Colorado, July 29– August 3. Davies, I., Laxton, J. W., and Owers, M. J. (1981). J. Inst. Energy March, 21– 30. Decker, S., and Klabunde, K. J. (1996). J. Am. Chem. Soc. 118, 12465. Ding, Y., Zhang, G., Wu, H., Hai, B., Wang, L., and Qian, Y. (2001). Chem. Mater. 13(2), 435– 440.

REFERENCES

255

Itoh, H., Utampanya, S., Stark, J. V., Klabunde, K. J., and Schlup, J. R. (1993). Chem. Mater. 5, 71. Jackson, J. M. (1976). U.S. Patent 3,953,335. Jaing, Y., Decker, S., Mohs, C., and Klabunde, K. J. (1998). J. Catal. 180, 24– 35. Khaleel, A., Kapoor, P. N., and Klabunde, K. J. (1999). Nanostruct. Mater. 11, 459. Klabunde, K. J., Ed. (2001). Nanoscale Materials in Chemistry. Wiley-Interscience: New York. Klabunde, K. J., Khalee, A., and Park, D. (1995). High Temp. Mater. Sci. 33, 99. Krylov, O. V., Markova, Z. A., Tretiakov, I. I., and Fokina, E. A. (1965). Kinet. Katal. 6, 128. Lewis, D. C., and Goldberg, J. H. (1978). Combustion, December, p. 32. Li, Y. X., and Klabunde, K. J. (1992). Chem. Mater. 4, 611. Niles, W. D., and Sanders, H. R. (1962). Trans. ASME, J. Eng. Power April. Radway, J. E., and Boyce, T. (1977). Reduction of Coal Ash Deposits with Magnesia Treatment. In R. W. Bryers, Ed. Ash Deposits and Corrosion Due to Impurities in Combustion Gases. Hemisphere: Washington, DC, pp. 401 –416. Radway, J. E., and Rohrbach, R. R. (1976). ASME Paper 76-WA/APC-9. Reese, J. T., Jonakin, J., and Caracristi, V. Z. (1964). ASME Paper No. 64-PWR-3. Reid, W. T. (1974). In Combustion Technology, H. B. Palmer and J. M. Beer Eds. Academic: New York. Salooja, K. C. (1972). J. Inst. Fuel, pp. 37– 42, January. Shuttleworth, S. G. (1958). The Chemistry and Technology of Leather, Vol. 2. Reinhold: New York, pp. 281– 322. Slabbert, N. P. (1975). Proc. XIV Congr. Int. Union Leather Chem. Technol. Socs. I, 240. Stark, J. V., and Klabunde, K. J. (1996). Chem. Mater. 8, 1904. Sun, N., and Klabunde, K. J. (1999a). J. Catal. 185, 506. Sun, N., and Klabunde, K. J. (1999b). J. Am. Chem. Soc. 121, 5587. Utampanya, S., Klabunde, K. J., and Schlup, J. R. (1991). Chem. Mater. 3, 175. Wagner, G. W., Bartram, P. W., Koper, O., and Klabunde, K. J. (1999a). J. Phys. Chem. B 103, 3225. Wagner, G. W., Koper, O., Lucas, E., Decker, S., and Klabunde, K. J. (1999b). J. Phys. Chem. B 104, 5118.

APPENDIX

TEST PROCEDURE FOR DETERMINATION OF CO2 IN ACIDIFIED SEAWATER 1

Apparatus Graduated cylinder, 100 mL Erlenmeyer flask, 250 mL Burette, 50 mL, 0.1-mL graduation

2

Reagents Phenolphthalein indicator Methyl orange indicator Sodium hydroxide, C.P., 0.2 N, standardized Sulfuric acid, C.P., 0.02 N, standardized

3 Method 3.1. Transfer 100 mL of sample to an Erlenmeyer flask and add 3 or 4 drops of phenolphthalein indicator. 3.2. If sample turns red, then proceed to 3.3, if not, add 0.2 normal NaOH dropwise until red color persists. The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 257

258

APPENDIX

3.3. Titrate with 0.02 N H2SO4 until pink color just disappears. 3.4. Add 4 drops of methyl orange (solution will turn blue) and titrate with 0.02 N H2SO4 to a steel gray endpoint. Record mL required for this step only. 4 Calculations 4.1. Titer from step 3.4  8.8 ¼ ppm CO2. TEST PROCEDURE FOR DETERMINING MgO CONTENT OF REACTOR OVERFLOW This method is used for measuring the solids content MgO per gallon or pounds MgO per thousand gallons in magnesium hydroxide slurry of varying concentrations. 1 Apparatus Pipette, 5 and 10 mL Erlenmeyer flask, 250 mL Burette, 50 mL, 0.1-mL graduations 2 Reagents Methyl orange indicator Nitric acid, C.P., 2.0 N and 0.0336 N Sodium hydroxide, C.P., 0.2 N and 0.0336 N 3 Method 3.1. Reactor underflow and thick slurry (.20% solids). 3.1.1. From a thoroughly mixed sample pipette 5 mL into a 250-mL Erlenmeyer flask washing down the inside if the pipette with distilled or deionized water. 3.1.2. Add a few drops of methyl orange indicator. 3.1.3. Add 2.0 N nitric acid from a 50-mL burette until all solids are in solution, and the solution has turned a permanent orange color. Record mL of HNO3 used. 3.1.4. Back titrate with 0.2 N NaOH until a permanent blue color is obtained. Record mL of NaOH used. 3.2. Reactor overflow and medium slurry (5–15% solids). 3.2.1. Pipette a 50-mL sample of slurry and repeat steps 3.1.2–3.1.4. 3.3. Reactor overflow and thin slurry (trace solids).

TEST PROCEDURE FOR DETERMINING EXCESS MgO

259

3.3.1. Pipette 50 mL of sample slurry into a 250-mL Erlenmeyer flask and add 2 drops methyl orange indicator. 3.3.2. Add 0.0336 N HNO3 in 10-mL increments until a permanent orange color is obtained. Record mL HNO3. 3.3.3. Back titrate with 0.0336 N NaOH until a permanent blue color is obtained. Record mL NaOH titrated. 3.3.4. Repeat 3.3.1–3.3.3 with a 50-mL sample of filtered slurry (use filtrate) and use as sample blank. 4 Calculations 4.1. Tank underflow and thick slurry, 5-mL sample. (mL HNO3  normality

mL NaOH  normality)  0:0336

¼ MgO, lb=gal 4.2. Reactor overflow and medium slurry, 50-mL sample. (mL HNO3  normality

mL NaOH  normality)  0:00336

¼ MgO, lb=gal 4.3. Tank overflow and thin slurry, 50-mL sample. (mL HNO3

mL NaOH)

(mL HNO3

mL NaOH)

10 ¼ MgO lb=103 gal TEST PROCEDURE FOR DETERMINING EXCESS MgO This method is intended for determining the amount of soluble Mg (as MgO lb per 1000 gal) in reacted seawater. 1

Reagents 0.0595 N sodium hydroxide, standardized Thymolphthalein indicator 0.05 N alcoholic calcium chloride, not standardized

2

Apparatus Erlenmeyer flask, 250 mL with rubber stopper Burette, 50 mL, 0.1-mL graduations Graduate cylinder, 100 mL

260

APPENDIX

3 Method 3.1. Measure 100 mL of 0.05 N alcoholic calcium chloride into Erlenmeyer flask and add several drops of indicator. 3.2. Titrate to a permanent blue endpoint with 0.0595 N NaOH, stopper flask. 3.3. Pipette 10 mL of sample into flask (if color turns a deeper blue, stop there and report as lime side) and titrate to the same permanent blue with 0.0595 N NaOH. 4 Calculations mL of 0:0595 N NaOH required (step 3:3) ¼ excess MgO lb=1000 gal

SLURRY COMPACTION TEST PROCEDURE This method is intended for measuring the compaction or the ability to settle of Mg(OH)2 slurry. 1 Apparatus Compaction tube, 2 in. in diameter, 48 in. long with 1.0-in. graduations 2 Method 2.1. Determine density, lb MgO/gal, on a mixed sample. 2.2. Calculate the amount of Mg(OH)2 slurry required to give 2400 mL of solution containing 0.30 lb MgO/gal and add to a 3000-mL graduate. 2.3. Dilute to 2400 mL with distilled water or deionized water, mix well and determine density, lb MgO/gal. 2.4. Pour the well-mixed slurry into the compaction tube and record the original level of the liquid. 2.5. At the end of 24 h record the level of the sludge in the compaction tube. 3 Calculations 3.1. To calculate the mL of slurry required in 2.2: mL undiluted slurry ¼

desired density (0:30)  quantity (2400 mL) undiluted slurry density

¼ 720=undiluted density

CAUSTIC MAGNESIA ACTIVITY (ACETIC ACID)

261

3.2. To calculate compaction: Compaction ¼ diluted density  original level=sludge level at 24 h

CAUSTIC MAGNESIA ACTIVITY (ACETIC ACID) This method of test is intended for determining the reactivity of MgO by acid neutralization. 1

Reagents Acetic acid, 1.0 N Phenolphthalein indicator

2

Apparatus One-quart Waring blender jar and jar holder for a Waring commercial blender (model 31BL92) mounted on a Dayton 12 hp AC–DC series motor (model 2Ml45) and a Dayton electronic speed control (model 4X797) whose input voltage is controlled by a Staco variable transformer (type 291, 3 amp, 0–120 Vac) Digital tachometer (Shimpo DT-4LB) with magnetic pickup speed sensor (Shimpo 3030 AM) and 5-ft shielded cable (all from Clarence J. Marx Company, Cleveland, Ohio) A 30-tooth 20 DP steel gear, 38-in. bore, required for the magnetic pickup Steel box for mounting tachometer and speed control equipment Analytical balance, 0.002 g readability Stop watch Thermometer, 0–2308F Graduates: 100, 500 mL

3 Method 3.1. Accurately weigh a 5.00-g sample on an analytical balance. 3.2. Measure 300 mL of distilled or deionized water at 778F + 8F into the blender container. 3.3. Add the 5.00-g sample to the water in the blender and add 8–10 drops phenolphthalein indicator. 3.4. Mix for approximately 10 s at 2500 rpm. 3.5. Quickly add by graduate 100 mL of 1.0 N acetic acid to the sample slurry while continuing to mix at 2500 + 50 rpm. 3.6. Start the stop watch as the acid is being added.

262

APPENDIX

3.7. Record the time in seconds for the development of the red phenolphthalein color. If the red color has not developed in the first 45 s, add 4 more drops of phenolphthalein indicator and repeat every 30 s until endpoint. This will ensure a concise color change. 4 Report The caustic magnesia activity time is the elapsed time, in seconds, from the acid addition to the development of a red color.

INDEX

Acid mine drainage, 172 – 173 Adatom, 128 Aerobic digestion, 164 Air classification, 75 Alkalinity, 158 Aluminum trihydrate, 179 Aman process, 59 AmphoMag, 197 Amphoteric, 196 Anaerobic digestion, 165 ANFO, 68 Animal nutrition, 203 – 204 Antibacterial activity, 197 Artinite, 5 Ball-mill, 117 – 119 Bilecik, 27 Bioflocculation and solids settling, 166 –167 Biological wastewater treatment, 163 – 166 Black, Joseph, 2 Bone magnesite, 20 Boron, 49– 50, 135

Breitenau, 28 Brine extraction, 42 Brine precipitation, 54– 55 Briquetting press, 144 Brucite, 4, 5, 12, 15, 16, 23, 198 Worldwide occurrence, 12– 30 Brucitic marble, 16 Brumado, 17 Brunnerite, 6 Calcium sulfate, 161 Calcite, 15 Calcining magnesite, 83– 88 Effect of time and temperature, 85 Energy requirements, 85 Kinetics of calcination, 85– 88 Calcining magnesium hydroxide, 88– 93 Decomposition mechanism, 90– 92 Effect of time and temperature, 94– 96 Energy requirements, 89 Kinetics of decomposition, 93 Calcium carbonate, 7, 8, 49 Carbon dioxide sequestration, 197 – 198 Carnallite, 5

The Chemistry and Technology of Magnesia, by Mark A. Shand Copyright # 2006 John Wiley & Sons, Inc. 263

264

INDEX

Catalyst support, 245 Chalkidiki peninsula, 26 Chemical contour, 68, 69 Chrome tanning of leather, 244 Compaction, 51 Cone crushers, 71–72 Crown spraying, 172 Crystal growth, 94 Crystal structure, 30, 34, 122 Dairy ration buffers, 208 Davy Humphrey, 2 Dead burned magnesia, 133 – 144 Decarbonation, 48 De-inking, 229 Delignification, 227 Densification, 94 Density, 34, 122 Dewatering, 159 – 161 Diffusion, 139 Dispersants, 79, 147 Dolime, 6, 47, 56 Dolomite, 4, 6, 11 Dravo thiosorbic process, 194 Drilling, 63– 68 Bench Height, 64 Blast timing, 67 Blast-hole pattern, 66 Blasting agents, 68 Burden and Spacing, 64– 65 Hole-Diameter, 64 Hole-Stemming, 66 Secondary blasting, 68 Sub-Drilling, 65 Dye adsorption, 163 Electrical resistance, 123 Environmental Impact, 163 Epsomite, 5 Eskisehir, 27 Evaporite salt, 44 Ferrofluids, 76 Fertilizers, 203 Filtration, 52– 54 Fire resistance, 235 Flame retardant, 180 – 184 Floatation machines, 80 Floatation regents, 79

Float-sink separation, 76 Flue additives, 246 Flue gas desulphurization, 189 – 194 Fosterite, 15, 137 Froth flotation, 77– 81 Fused magnesia, 144 –146 Gabbs, 13– 15 Gas burner, 105 Grain growth, 139 –144 Grass Tetany, 207 Gravity concentration, 76– 77 Grinding, 55, 115 – 119 Gypsum, 161 Gyratory crusher, 71– 72 Handling requirements, 161 Hard burned magnesia, 133 Hardness, 30, 34 Hazardous spill clean-up, 196 –197 Heat balance, 98 Heat capacity, 124 Heat of formation, 31, 34 High temperature corrosion, 247 – 248 History of magnesia, 1 Horizontal rotary kilns, 103 – 105 Human nutrition, 209 Hydration kinetics and mechanisms, 150 – 151 Hydroclones, 76 Hydrogen peroxide bleaching, 228 Hydrogen sulfide, 169 – 172 Hydromagnesite, 5, 9, 15 Impact crusher, 72 Index of refraction, 30, 34, 122 Industrial Wastewater Treatment, 155 – 157 Jaw crusher, 72 Kawasaki process, 192 Kieserite, 5 Kink, 128 Kunwarara, 17 Ku¨tahya, 27 Lansfordite, 5 Liaoning province, 22 Limpopo province, 24

INDEX

Low temperature corrosion, 248 Luster, 30, 33 Magnefite pulping process, 223 – 226 Magnesia alba, 1 – 3 Magnesia cements, 231 – 235 Magnesite Biogenic, 8 Cryptocrystalline, 10– 11, 20 Crystalline, 11 Dissociation pressure, 84 Mineralogical name, 3 Secondary nodular, 8 Sedimentary, 7 Serpentine alteration, 10 Solubility constant, 7 Worldwide occurrence, 12– 30 Magnesium Acetate, 215 Alkyls, 215 Bisulfite, 225 Chloride, 216 Nitrate, 217 Overbase sulfonates, 218 Oxychloride, 231 Oxysulfate, 234 Peroxide, 219 Soaps, 218 Sulfate, 217 Sulfite, 225 Magnesium bicarbonate, 152 – 153 Magnesium bioavailability, 206 Magnesium hydroxide slurry quality control, 151 Magnesium hydroxide slurry, 146 Magnesium hydroxide slurry, synthetic, 146 – 148 Magnesium metal production, 220 – 221 Magnesium oxide properties, 121 – 125 Chemical properties, 125 – 126 Dissolution, 126 – 127 Hydration, 148 – 150 Molecular adsorption, 129 – 130 Surface structures, 127 – 129 Magnesium phosphate cement, 235 – 238 Magnetic susceptibility, 125 Media mills, 147 – 149 Melting point, 124 Metal hydroxide solubility, 156

265

Metals removal, 158 – 159 Mining operations, 63– 69 Moonmilk, 9 Mount Brussilof, 17 Mpumalanga province, 24 Muck maps, 68– 69 Multiple hearth furnaces, 98– 103 Back Rabble (Four Arms per Hearth), 102 Full Progressive Rabble (Fours Arms per Hearth), 102 Full Progressive Rabble (Two Arms per Hearth), 102 Refractory Linings, 102 – 103 Single Progressive Rabble (Four Arms per Hearth), 101 Nanoparticulate magnesia, 251 Neck formation, 140 Neoprene, 185 Nesquehonite, 5 Neutralization curve, 158 Neutralization reactions, 161 – 162 Nickel hydroxide, 159 Nitrification, 164 Nuclear waste disposal, 196 Nutrition, 201, 203 Odor control, 168 – 172 Oral supplementation, 207 Overburden Removal, 63 Overliming, 50 Oxygen bleaching, 227 Packaging, 56 Pattinson process, 151 – 153 Pelletizing, 134, 137 Periclase, 15, 122 pH control, 157 Phase diagram, 136, 139 Phosphorus removal, 167 – 168 Plant nutrition, 201 –203 Porosity, 142 Poultry, 208 Precipitation curve, 163 Precipitation process, 50 Precipitation reaction, 45 Preheater, 105 Primary Crushing, 70– 71 Pulp bleaching, 226 – 229

266

INDEX

Pulp liquor preparation, 225 Pyrohydrolysis of magnesium chloride hexahydrate, 59– 60 Quality Control, 56– 58 Rabble, 98– 102 Radenthein, 28 Radium hot springs, 16 Refractory lining, 102 – 103 Refractory, 133 Remediation applications, 194 – 196 Ring-roller mill, 116 Roll crusher, 72 Rumen fluid, 206 Ruminant animals, 204 – 205 Safety, 157 Satka, 29 Scaling reduction, 243 Screening methods, 74 Seawater chemistry, 41 Seawater pretreatment, 48 Sedimentation, 9 Serpentization, 11 Settling, 51 Sewer corrosion, 168 – 172 Shaanxi province, 23 Shaft Kilns, 107 – 113 Annular Shaft Kiln, 111 – 112 Discharge, 107 Double-Inclined Kiln, 109 Modern Shaft Kiln, 109 Multi-Chamber Kiln, 109, 111 Ore Charging, 107 Parallel-Flow Regenerative Kiln, 113 Sheet molding compound, 184 –185 Siderite, 6

Silica Removal, 173 – 176 Sinter aids, 142 – 144 Sintering, 139 – 142 Size separation, 73– 81 Soil, 201 – 203 Solubility constants, 7 Sorbtech process, 194 Sorel cement, 231 – 235 Specific gravity, 30 Specific heat capacity, 3, 34 Specific surface area, 94, 95 Stone size, 88 Struvite formation, 167 – 168 Subsurface brines, 45 Sugar manufacture, 241 Sulfite pulping, 223 Sump leaching, 42 Surface treatment, 182 – 184 Swine, 207 – 208 Synthetic magnesia manufacturers, 40 Tamil Nadu, 25 Terrace, 128 Tertiary crushing, 81 Thermal conductivity, 31, 34, 122 Thermal expansion, 31, 34, 124 Thermal insulation, 234 Thuddungra, 21– 22 Toxicity characteristic leaching procedure, 195 Transformer steel coating, 254 Valley, 128 Vanadates, 247 Washing, 52 Wastewater treatment, 155 – 168 Water Coolers, 105 – 107 Well-Drilling Fluids, 250