Activated Carbon Surfaces in Environmental Remediation [1 ed.] 0-12-370536-3, 978-0-12-370536-5

Citation preview

Activated Carbon Surfaces in Environmental Remediation

INTERFACE SCIENCE AND TECHNOLOGY Series Editor: ARTHUR HUBBARD In this series: Vol. 1: Clay Surfaces: Fundamentals and Applications Edited by F. Wypych and K.G. Satyanarayana Satyanarayana Microfluidics Vol. 2: Electrokinetics in Microfluidics By Dongqing Li Interfaces Vol. 3: Radiotracer Studies of Interfaces Edited by G. Horányi Horanyi

Vol. 4: Emulsions: Structure Stability and Interactions Edited by D.N. Petsev Vol. 5: Inhaled Particles By Chiu-sen Wang Vol. 6: Heavy Metals in the Environment Edited by H.B. Bradl Vol. 7: Activated Carbon Surfaces in Environmental Remediation Edited by T.J. Bandosz

INTERFACE SCIENCE AND TECHNOLOGY -– VOLUME 7

Activated Carbon Surfaces Surfaces in Environmental Remediation

Edited by

Teresa J. Bandosz The City College of New York New York, USA

, 1

t y

ELSEVIER

Amsterdam • Boston • Heidelberg •• London New York •• Oxford • Paris •• San Diego San Francisco • Singapore • Sydney • Tokyo Academic Academic Press Press isis an an imprint imprint of of Elsevier Elsevier

ELSEVIERB.V. ELSEVIER B.V. Radarweg 29 211, 1000 1000 AE Amsterdam P.O. Box 211, The Netherlands

ELSEVIER Inc. 1900 525 B Street, Suite 1900 San Diego, CA 92101-4495 USA

ELSEVIER Ltd The Boulevard, Langford Lane Kidlington, Oxford OX5 1GB UK

ELSEVIER Ltd 84 Theobalds Road WC1X8RR London WC1X 8RR UK

© 2006 Elsevier Ltd. All rights reserved. following terms and conditions apply to its use: This work is protected under copyright by Elsevier Ltd., and the following

Photocopying Single photocopies of single chapters may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profit non-profit educational classroom use. Elsevier*s Rights Department in Oxford, UK: phone ((+44) 1865 843830, fax ((+44) Permissions may be sought directly from from Elsevier’s +44) 1865 +44) 1865 853333, e-mail: [email protected]. Requests may also be completed on-line via the Elsevier homepage 1865 (http://www.elsevier.com/locate/peiTnissions). (http://www.elsevier.com/locate/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: ((+1) +1) (978) 7508400, fax: ((+1) +1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: ((+44) +44) 20 7631 5555; fax: ((+44) +44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments.

Derivative Works Tables of contents may be reproduced for internal circulation, but permission of the Publisher is required for external resale or distribution of such material. Permission of the Publisher is required for all other derivative works, including compilations and translations. Electronic Storage or Usage Permission of the Publisher is required to store or use electronically any material contained in this work, including any chapter or part of a chapter. Except as outlined above, no part of this work may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Elsevier's Rights Department, at the fax and e-mail addresses noted above. Address permissions requests to: Elsevier’s Notice pei'sons or property as aamatter No responsibility is assumed by the Publisher for any injury and/or damage to persons matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material verification of diagnoses and drug dosages herein. Because of rapid advances in the medical sciences, in particular, independent verification should be made.

First edition 2006 B r i t i of sh Congress Library Cataloging Cataloguing in Publication Data Library catalogue record is available from B r i t i of sh Congress. Library. A catalog from the Library

British Library Cataloguing in Publication Data from the British Library. A catalogue record is available from

ISBN-13: 978-0-12-370536-5 ISBN-10: 0-12-370536-3 @ The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper). Printed in The Netherlands.

Working together to grow libraries in developing countries •ww.elsevier.com | www.bookaid.org | www.sabre.org

ELSEVIER

BOOK AID International

Sabre Foundation

v

PREFACE

One may ask: "what's special in chars? They are cheap, abundant, unsophisticated and dirty. Why bother to study them and their applications? They will always work efficiently provided that enough quantity is applied". How much truth is in this statement and how much common misunderstanding? Does adding the "buzz" word of the last decade and converting the old "chars" into "nanocarbons" change anything? Are they really the same chars, which were used 2000 years B.C.? I hope that the answers to these questions can be easily found in the pages of this book. The idea of editing a book addressing the environmental applications of carbonaceous materials from fundamental and, application stand points is a result of current needs, an increase in environmental awareness and last but not least, a recent boom in development of carbon science. The latter not only brought to the market completely new materials as fullerenes, nanotubes or nanohorns but also open nonexisting before, new fields for their applications as, for instance, microelectronic devices, nanosensors, drug carriers, or energy storage media. All of these promoted extended research on carbon surfaces. Many of us started to realize how important are the pore sizes and shapes and how big is a difference between one Angstrom and one nanometer; the presence of functional groups and their chemical nature became crucial for sophisticated reactions leading to nanostructure formation in confined pore space or biocompatibility of the final product. And thus the myth of cheap carbon started to dissipate An ancient art of application of carbon surfaces in environmental remediation is the one, which benefits a lot from recent boom in carbon science. The statement from the first paragraph about the superiority of quantity over quality of adsorbent is not true anymore. Stricter and stricter environmental regulations demand the high purity for water and gas phases. The "zero emission" level very often can be reached with a sophisticated adsorbent whose surface, besides having pores of specific sizes and shapes is decorated with various chemical functional groups to maximize the efficiency of separation. This efficiency is very often enhanced by catalysis and surface reactions. Summarizing, very often the nano- or microreactors are constantly working within the pores spaces of carbonaceous adsorbents This book intends to provide a comprehensive summary of environmental applications of activated carbons. I hope it will serve as a handbook or reference book. To understand the removal of contaminants and pollutants on activated carbons, the theoretical basis of adsorption

vi

Preface

phenomena are addressed. The effeets of pore structure and surface chemistry are discussed from both scientific and engineering points of view. The book intends to be written for graduate students and beginning professionals in environmental science and engineering. Those interested in surface science will find in this book how surface features are engaged in environmental remediation. The practical aspects addressed in the book cover the broad spectrum from air and water cleaning, through warfare gases removal to biomedical applications. At the end of this short preface I would like to address my gratitude to my colleagues and friends who agreed to contribute to this book. As you can see, they represent different fields of carbon science and engineering but certainly their common feature is the deep passion to understand carbon and to make it better and more efficient adsorbent and catalyst. I am grateful for their enthusiasm, involvement and patience. Special thanks are given to those who helped me in providing scientific and editorial reviews. The list includes but is certainly not limited to Prof. Michel Evans (Royal Millitary College, Canada), Prof. Ljubisa Radovic (Penn State University), Prof. Emeritus Amos Turk (CCNY), Prof. James (Chip) Kilduff (Rensselaer Polytechnic Institute), Drs. Fred Baker (ORNL), Dr. James Graham (US Filters), Dr. Teus Wigmans (Kisuma), Mr. Christopher Karwacki (USArrny, ECBC), and Drs. Seyed A. Dastgheib, Yanping Guo, Cindy Lee, Mark A. Schlautman, and Hocheol Song of Clemson University. Besides them I am sure they are others who deserve my thanks for their involvement in this project. Last but not least I would like to express my gratitude to Professor Arthur Hubbard, the Editor of INTERFACE SCIENCE AND TECHNOLOGY series for encouragement to take this project and to the Elsevier Editor, Mr. Derek Coleman, for his guidance, useful information in styling of the manuscript and for constant support in numerous E-mails.

New York, September 2005

vii CONTENTS TABLE OF CONTENTS Preface v Preface............................................................................................................................................... Volume xi Contributors to this Volume.............................................................................................................xi Chapter 1. Types of carbon adsorbents and their production production....................................................11 J. J. Ángel Angel Menéndez-Díaz Menendez-Diaz and I. Martín-Gullón Martin-Gullon 4. Types of carbons carbon and a d their structure sorbents ............................................................................................. 1 1. 5. Historical background of the activated carbons carbons. ............................................................................43 2. carbons 5 3. Physical and chemical properties of activated carbons................................................................. adsorbents 4. Production of carbon adsorbents................................................................................................. 11 5. Regeneration of exhausted activated carbons ............................................................................. 43 References ....................................................................................................................................... 45 References Chapter 2. Pore formation and control in carbon materials .................................................... 49 Tascon M. Inagaki and J.M.D. Tascón 1. Pores in carbon materials ............................................................................................................ 49 1. characterization 2. Pore characterization................................................................................................................... 51 formation in carbon materials ............................................................................................. 68 3. Pore formation 4. Some important types of porous carbons .................................................................................... 76 5. Novel techniques to control pore structure ................................................................................. 82 References ..................................................................................................................................... 100 References Chapter 3. Characterization of nanoporous carbons by using gas adsorption Isotherms ....................................................................................................................107 107 M. Jaroniec and J. Choma 1. Introduction ............................................................................................................................... 107 1. adsorbents 2. Carbon adsorbents..................................................................................................................... 109 6. Gas adsorption Adsorption isotherms .......................................................................................................... 113 3. 7. Basic Basicsfor adsorption isotherms ................................................................................................ 118 135 4. parameters ..................................................................................................... 8. Comparative adsorption analysis .............................................................................................. 121 5. 9. Adsorption Conclusion spotential a n d f u distribution rtherperspe .............................................................................................. 6. 130 ctives........................................................................................130 isotherms.................................................................................................... .......153 7. Basics for adsorption isotherms ................................................................................................ 135 8. Pore size analysis ...................................................................................................................... 144 9. Conclusions and further perspectives........................................................................................ perspectives 152 References ..................................................................................................................................... 153 References characterization Chapter 4. Surface chemistry of activated carbons and its characterization........................ 159 CO. Ania T.J. Bandosz and C.O. 1. Introduction ............................................................................................................................... 159 1. functionalities 2. Surface functionalities............................................................................................................... 160 3. Electrochemical aspects of carbon surface ............................................................................... 169 functionalities 4. Techniques for the characterization of the surface functionalities............................................ 179 5. Effect Effect of mineral matter ............................................................................................................ 213 carbons 6. Role of surface chemistry in the environmental applications of activated carbons.................. 215 References ..................................................................................................................................... 218 References

viii

Table of Contents

Desulfurization on activated carbons..................................................................... carbons Chapter 5. Desulfurization 231 T.J. Bandosz 1. Introduction ............................................................................................................................... 231 1. environmentally detrimental sulfur sulfur compounds......................................... compounds 2. Basic chemistry of environmentally 233 sulfur dioxide from from flue gas .................................................................................. 237 3. Removal of sulfur 7. Removal of hydrogen sulfide sulfide .................................................................................................... 249 4. carbons 5. Removal of mercaptans on activated carbons........................................................................... 271 sulfur compounds from from liquid fuel ...................................................... 282 6. Removal of thiophenic sulfur 7. Concluding remarks .................................................................................................................. 287 References ..................................................................................................................................... 288 References Chapter 6. Applications of nanotextured carbons for supercapacitors and I n t r o d u c storage ...........................................................................................................................293 hydrogen 293 Raymundo-Pifiero and F. Béguin Beguin E. Raymundo-Piñero 1. Introduction ............................................................................................................................... 293 1. 2. Energy storage in electrochemical capacitors based on carbon materials ................................ 294 3. Hydrogen storage ...................................................................................................................... 324 4. Conclusion and perspectives ..................................................................................................... 337 References ..................................................................................................................................... 339 References treatment Chapter 7. Activated carbon adsorption in drinking water treatment.................................. 345 Tanju Karanfil Karanfil Tanju 1. Introduction ............................................................................................................................... 345 1. sources 2. Typical water quality characteristics of drinking water sources............................................... 347 Adsorption interactions ................................................................................................................369 3. Adsorption 349 4. Conclusions ............................................................................................................................... 367 Acknowledgment .......................................................................................................................... 369 References ..................................................................................................................................... 369 References Chapter 8. Adsorption of organic compounds onto activated carbon –....375 applications in water and air treatments .................................................................................. 375 Pierre Le Cloirec and C. Faur 1. Introduction ............................................................................................................................... 375 1. carbon 2. Fundamentals of adsorption onto activated carbon................................................................... 376 treatment 3. Organics adsorption onto activated carbon in waste water treatment....................................... 387 4. Air treatment-voc and odor removal ......................................................................................... 399 5. Conclusion and trends ............................................................................................................... 416 References ..................................................................................................................................... 416 References Chapter 9. Activated carbon filters and their industrial applications ................................... 421 Przepiorski J. Przepiórski 1. Introduction ............................................................................................................................... 421 1. filtering media ................................................ 423 2. Applications of activated carbons as industrial filtering applications 3. Liquid phase applications.......................................................................................................... 427 4. Gas phase applications .............................................................................................................. 447

Table TableofofContents Contents

ixix

medium 5. Activated carbon fiber as a filtering medium............................................................................ 461 6. Regeneration of spent activated carbon .................................................................................... 462 References ..................................................................................................................................... 466 References Chapter 10. Adsorption of chemical warfare agents ............................................................... 475 P. Lodewyckx 1. Introduction ............................................................................................................................... 475 1. adsorp agents tion.. ........................................................................................................... 475 2. Chemical warfare CWAs 3. The use of activated carbons in the protection against CWAs.................................................. 479 7. Adsorption mechanisms ............................................................................................................ 489 4. adsorption.................................................................................................................511 5. Modelling adsorption ................................................................................................................ 511 influencing the adsorption of CWAs............................................................................ CWAs 6. Factors influencing 516 7. Conclusions ............................................................................................................................... 523 References ..................................................................................................................................... 523 References Chapter 11. Activated carbons as medical adsorbents ............................................................ 529 S.V. Mikhalovsky and V.G. Nikolaev 1. Introduction ............................................................................................................................... 529 1. AC 2. Modes of medical use of AC..................................................................................................... 530 3. Activated carbon in poisoning treatment .................................................................................. 532 4. Conclusions ............................................................................................................................... 555 Acknowledgments 555 Acknowledgments......................................................................................................................... References ..................................................................................................................................... 555 References INDEX .......................................................................................................................................... 563

This Page is Intentionally Left Blank

xi

CONTRIBUTORS TO THIS VOLUME C O . Ania Department of Chemistry City College of New York 138* Street and Convent Ave New York, NY 10031 J. Angel Menendez Diaz INCAR-CSIC Apartado 73 33080 Oviedo - Spain [email protected] TJ. Bandosz Department of Chemistry City College of New York 138* Street and Convent Ave New York, NY 10031 [email protected] F. Beguin Centre National de la Recherche Scientifique Universite D* Orleans Centre De Recherche sur la Matiere Divisee IB, rue de la Ferrolerie 45071 Orleans, Cedex 2 France [email protected] J. Choma Institute of Chemistry Swietokrzyska Academy Checinska 5 25-020 Kielce Poland M. Ingaki Aichi Institute of Technology Yakusa Toyota 470-0392 Japan

xii

list of Contributors List

M. Jaroniec Sorption and Surface Science Center Department of Chemistry Kent State University Kent, OH 44242 j aroniee @ kent.edu C. Faur Responsable du Dept. Systemes Energetiques et Environnement Ecole des Mines de Nantes GEPEA, UMR CNRS 6144 4 rue Alfred Kastler, BP 20722 44307 Nantes cedex 03, France Tanju Karanfil Clemson University Department of Environmental Engineering and Science 342 Computer Court Anderson, SC 29625 tkaranf @ clemson.edu Pierre Le Cloirec Responsable du Dept. Systemes Energetiques et Environnement Ecole des Mines de Nantes GEPEA, UMR CNRS 6144 4 rue Alfred Kastler, BP 20722 44307 Nantes cedex 03, France Pierre.Le-Cloirec @ emn.fr P. Lodewyckx Belgian Army Service of Technological Applications NBC Division Martelarenstraat 181 B-1800 Vilvoorde (Peutie) Belgium [email protected] I. Martin-Gullon Chemical Engineering Department Universidad the Alicante, PO Box 99 03080 Alicante - Spain

List of Contributors

S. V. Mikhalovsky School of Phannacy and Biomolecular Sciences University of Brighton Brighton BN2 4GJ United Kingdom [email protected] V. G. Nikolaev R.E. Kavetsky Institute of Experimental Pathology Oncology and Radiobiology Nat. Acad.Sci. Vasilkovskaya Str. 03022 Kiev Ukraine J, PrzepiorsM Institute of Chemical and Environment Engineering Technical University of Szczecin ul. Pulaskiego 10 70-322 Szczecin, Poland j aeek.przepiorski Ops.pl E. Rayrnundo-Pinero Centre National de la Recherche Scientifique Universite D'Orleans Centre De Recherche sur la Matiere Divisee IB, rue de la Ferrolerie 45071 Orleans, Cedex 2 France J.M.D. Tascon INCAR-CSIC Apartado 73 33080 Oviedo - Spain [email protected]

xiii

This Page is Intentionally Left Blank

Activated Carbon Surfaces in Environmental Remediation T.J. Bandosz (editor) © 2006 Elsevier Ltd. All rights reserved.

1

Types of carbon adsorbents and their production J. A. Menendez-Diaz" and I. Martln-Gullonb "Institute National de Carbon, CSIC, PO box 73,33080 Oviedo, Spain b

Chemical Eng. Dept, Universidad de Alicante, PO box 99, 03080 Alicante, Spain

1, TYPES OF CARBONS AND THEIR STRUCTURE Carbon materials are mainly composed of the element carbon. Because of its atomic structure (Is 2 , 2s2,2p2) this element has unique bonding possibilities, both with other elements and with itself. Depending on the hybridization type of the carbon atoms, these are able to bond with other carbon atoms, giving rise to three major* allotropic forms of carbon [1]: •

•

•

Diamond is a rigid and isotropic 3D-structure in which each carbon atom shares four o-bonds with four other neighbouring carbon atoms, that form a cubic structure (sp3-based structure). It has the highest atomic density of any solid and is also the hardest material with the highest thermal conductivity and melting point. Graphite is a layered structure in which the graphene layers (single graphite plane) are formed by atoms of carbon bonded by c - and 7i-bonds to another three neighbouring carbon atoms (sp2-based structure). The graphite planes tend to exhibit a parallel alignment which is maintained by dispersive and Van der Wals forces. This leads to a high degree of anisotropy. Graphite (in plane) has an even higher thermal conductivity than diamond and exhibits a good electrical conductivity. Fullerenes are three-dimensional carbon structures where the bonds between the carbon atoms are bent to form an empty cage of sixty (Cgo) or more carbon atoms. This is possible due to the fact that re-hybridisation occurs, resulting in a sp 2+s form, which is intermediate between sp2 and sp3 [2].

On an atomic scale, the majority of carbons exhibit the allotropic form of graphite, i.e. a sp2-based structure, while diamond-like carbons, fullerenes and their derivatives, such as nanotubes, represent a shorter variety of carbon forms. However, depending on the degree of * There are also other less abundant allotropic forms of carbon like carbynes (sp-based hybridization) or rhombohedral graphite [1].

2

J.A. Menéndez-Díaz Menendez-Diaz and and I. Martín-Gullón Martin-Gullon J.A.

crystallographic order in the third direction (c-direction), carbons based on the allotropic form of graphite can be classified

into graphitic carbons

(which have a measurable

crystallographic order in this direction irrespective of the presence of structural defects) and non-graphitic carbons (without any measurable crystallographic order in the c-direction direction apart from a more or less parallel stacking) [3],

••. ••••: ••. •: ••

•

allotropic forms

hexagonal graphite

fullerene (C60)

cubic diamond

non-graphitic carbons non-graphitizable carbons (chars)

diamond-like carbons

graphitizable carbons (cokes, pitch) carbon nanotubes

graphitic carbons

activated carbon mi croe structure

carbon nanofibres

pyrolytic graphite (planar symmetry)

carbon fibre (concentric texture) carbon black (concentric texture)

glassy carbon (ribbon model]

i so tropic graphite

carbon fibre (radial texture)

Fig. 1. Major allotropic forms of carbon and a schematic representation of some of the carbon structures derived from these forms [4, 5]

Types of Carbon Adsorbents and their Production

3

Non-graphitic carbons can in turn be divided into graphitizable and non-graphitizable carbons. Thus, a graphitizable carbon is "a non-graphitic carbon which upon graphitization (heat treatment) is converted into graphitic carbon", while a non-graphitizable carbon is "a non-graphitic carbon which cannot be transformed into graphitic carbon by high-temperature treatment up to 3300 K under atmospheric pressure or a lower pressure" [2]. Moving up from nano-scale to micro-scale, carbons exhibit very different structures. Some of these microstructures are arranged in preferential directions, like synthetic graphite or graphitized carbon fibres, while disordered microstructures are characteristic of chars or activated carbons. Such a wide variety of possible structures gives rise to a large amount of different types of carbons. Fig. 1 shows a schematic representation of some of these carbon structures. Powder (particle size lower than 100 • 10"6 m) and granular (including extruded and pelletized) activated carbons are typical carbon adsorbents. These are non-graphitic, non-graphitizable carbons with a highly disordered microstructure. Other forms of carbons are also used as adsorbents such as activated carbon fibres, fabrics and felts prepared from a wide variety of precursors including coal, petroleum pitch, viscose or rayon. Moreover, exfoliated graphite can also be used to adsorb heavy oils [6]. The use of carbon nanotubes for hydrogen storage is also a current subject of study. 2. HISTORICAL BACKGROUND OF THE ACTIVATED CARBONS The use of carbon extends so far back in time that its origin is impossible to determine exactly. Prior to the use of what we call today activated carbon (which has a highly developed porous structure), either wood char, or coal char or simply a partially devolatilized carbonaceous material was employed as an adsorbent. The first recorded case dates back to 3750 BC, when both the Egyptians and Sumerians used wood char for the reduction of copper, zinc and tin ores in the manufacture of bronze, and also as a smokeless fuel [7]. In 2650 BC, the Egyptians used bonechar to wallpaint Perneb's grave. The first proof of the medicinal use of carbon was found in Thebes (Greece), in a papyrus document from 1550 BC [8]. Later on, Hippocrates (around 400 BC) recommended that water should be filtered with woodchar prior to consumption, in order to eliminate bad taste and odor and to prevent several diseases, including epilepsy, chlorosis and anthrax. In relation to the treatment of ancient drinking water, recent studies indicate that on Phoenician ships drinking water was stored in charred wooden barrels from 450 BC, a practice which continued until the 18th century as a means of prolonging the supply of drinking water on transatlantic voyages. Nevertheless, the first reported application of activated carbons as a gas phase adsorbent did not take place until as late as 1793 AD, when Dr. D.M. Kehl applied woodchar in order to mitigate the odours emanating from gangrene. The above gentleman also recommended filtering water with woodchar. The first application of activated carbon in the industrial sector took place in England in 1794, when it was used as a decolorizing agent in the sugar production industry, this event

4

Menendez-Diaz and I. Martín-Gullón Martin-Gullon J.A. Menéndez-Díaz

marking the beginning of research with activated carbons in liquid phase. This application remained a secret until 1812 (18 years), when the first patent appeared in Great Britain, although several sugar refineries were already using woodchar as a decolorizing agent before 1808. In 1811, it was proved that bone char was more effective than woodchar and its regeneration became the established objective. The first patent for the thermal regeneration of activated carbon dates from 1817, and a continuous process for manufacturing and regenerating bone char was developed in Germany in 1841. The first large scale gas-phase application took place in the mid 19th century. In 1854, the Mayor of London ordered the installation of wood char filters in all the sewer ventilation systems to remove nasty odours, while in 1872 gas masks with carbon filters were used in chemical industries to prevent mercury vapours from being inhaled. The term 'adsorption' was first used in 1881 by Kayser to describe the uptake of gases by chars. At about the same time, activated carbon material, as it is known nowadays, was discovered by R. von Ostrejko, who is considered the father and/or inventor of activated carbon [9], In 1901, he patented two different methods of producing activated carbons: • The carbonization of lignocellulosic materials with metal chlorides (the basis of chemical activation). • The mild gasification of chars with steam or carbon dioxide at red temperatures (thermal or physical activation). At around this time, von Ostrejko also patented specific equipment and utilities for producing activated carbons. In 1910, Wijnberg and Sauer acquired the patent rights, and were the first to apply activated carbons to the sugar industry (Norit White Sugar Company), referring to these carbons as 'noir epure'', 'eponit' or 'norit', although the carbons were purchased from Stettiner Spritt Werke and Erste Osterreichisehe Ceserin Werke AG in Stockerau [10]. Later, the company decided to produce its own carbons from peat during the First World War in Zaandam, under the name of NV Nederlandse NORIT Maatsehappij (Dutch NORIT Company). On the other hand, the first commercial chemically activated carbon (Carboraffin) was produced in Aussig (the Czech Republic) in 1914, using sawdust as raw material and ZnClz as activating agent. The First World War stimulated the development of both the production and application of activated carbons. The use of poisonous gases by the German army against the French, British and Russians on different fronts, posed a severe problem for the allies, and as a consequence, there was an urgent need for gas mask development. Nikolai Zelinski, a professor at Moscow University was the first to suggest packing activated carbons inside a canister fitted to a gas mask [11]. A bit later, an intensive research program was carried out in the USA with the development of coconut-shell based granular activated carbons that could be packed inside a canister, with the two fold objective of removing poisonous gas and of offering a low pressure drop. Without doubt, the First World War was the starting point of activated carbon development, when it was used not only in the white sugar industry, but also as an adsorbent

Types of Carbon Adsorbents and their Production

5

for water treatment and the adsorption and removal of vapours in gas phase. The rapid development of society over the 20th century, promoted by the medical and scientific improvements as a consequence of the industrial revolution of the previous century, has also affected the use of activated carbon. Indeed the production and utilization of activated carbon has increased with every decade, specially in the second half of the last century due to the stricter environmental regulations regarding both water resources, clean gas application and economic recovery of valued chemicals. In the last thirty years, the use of activated carbon as a metal catalyst support instead of carbon blacks has also been widespread [12]. The study of activated carbons, or carbon adsorbents, is included in what is commonly called Science and Technology of Carbon Materials, where a variety of carbon based materials are employed (such as the above mentioned carbon blacks, nanotubes, etc). Although the development and use of carbon adsorbents have brought enormous benefits to mankind, the words of Professor Harry Marsh (1997) [13] claiming that "Activated carbon is the Cinderella of the Carbon family, (...) and after purifying tonnes of sugar, cleaning up oceans of water, and enormous amounts of food, (...), they were never taken to the BalP cannot be ignored. Nevertheless, in recent years a growing interest has been shown in Activated Carbon Fibres (ACF), derived from the carbon fibers of viscose rayon (1966), Saran (1970), phenol-formaldehide resin (1980) or coal pitch (1985) [14]. 3. PHYSICAL AND CHEMICAL PROPERTIES OF ACTIVATED CARBONS Carbon adsorbents have a porous carbon structure, which contains small amounts of different heteroatoms such as oxygen and hydrogen. Some activated carbons also contain variable amounts of mineral matter (ash content) depending on the nature of the raw material used as precursor. The porous structure is perhaps the main physical property that characterizes activated carbons. This is formed by pores of different sizes which according to IUPAC recommendations [15] can be classified into three major groups (see Fig. 2): Fxternal surface

Fig. 2. Schematic representation of the pore network of a carbon adsorbent

6

Menendez-Diaz and I. Martín-Gullón Martin-Gullon J.A. Menéndez-Díaz

• • •

Micropores with a pore width of less than 2-10" HI. Mesopores with widths from 2.0 to 50 • 10" m. Macropores with a pore width larger than 50 • 10"9 m.

Furthermore, the presence or absence of surface groups, formed by heteroatoms (atoms different to the carbon atom), that may bond themselves to the carbon atoms at the edges of the basal planes gives rise to carbons with different chemical properties. The physical and chemical properties are of great importance for the behaviour of carbon adsorbents and they will be discussed in depth in other chapters of this book. We would like, however, to outline here some basic concepts and provide an overall view of the importance that these properties have for the adsorption processes. 3.1. The importance of the porous texture The structure of the carbon skeleton can be considered as a mixture of graphite-like crystallites and non-organized phase composed of complex aromatic-aliphatic forms. The crystallites are composed of a few (about three) parallel plane layers of graphite, the diameter of which are estimated to be about 2 - 1 0 " m, or about nine times the width of one carbon hexagon [16]. The regular array of carbon bonds on the surface of the crystallites is disrupted during the activation process, yielding relative free valences. The crystallites are randomly oriented and extensively interconnected. The structure of an activated carbon may be visualized as a stack of poorly developed aromatic sheets (crystallites), distributed and crosslinked in a random manner, separated by disorganized carbonaceous matter and inorganic matter (ash) derived from the raw material. The anisotropic crystallite alignment is associated with the presence of voids. During the activation the spaces between the crystallites become cleared of less organized carbonaceous matter and, at the same time, some of the carbon is removed from the crystallites. The resulting channels through the graphitic regions and the interstices between the crystallites of the activated carbon, together with fissures inside and parallel to the graphitic planes, make up the porous structure which usually has a large surface area. These range over several sizes and are therefore classified into the three groups mentioned above. In the different adsorption processes, both in gas and liquid phase, the molecules or atoms (adsorbable) are fixed (adsorbed) on the carbon (adsorbent) surface by physical interactions (electrostatic and dispersive forces) and/or chemical bonds. Therefore, a relatively large specific surface area is one of the most important properties that characterize carbon adsorbents. The surface of the activated carbons consists mainly of basal planes and the edges of the planes that form the edges of microcrystallites. Adsorption capacity related parameters are usually determined from gas adsorption measurements. The specific surface area is calculated by applying the Brunauer-EmmettTeller (BET) equation [17] to the isotherms generated during the adsorption process. The adsorption of N2 at 77 K or CO2 at 273 K are the most commonly used to produce these isotherms. The BET theory is based upon the assumption that the monolayer is located on

of Carbon Adsorbents and their Production Types of

7

surface sites of uniform adsorption energy and multilayer build-up via a process analogous to the condensation of the liquid adsorbate. For convenience, the BET equation is normally expressed in the form:

n{po-p)

J L+ ^ w± j h nC nmCp°

e

r

cc = where e

e

\

RT

which requires a linear relationship between (p/p°)l{n(p°-p) and p/p°, from which the monolayer capacity, nm (mmol g"1), can be calculated. In activated carbons the range of linearity of the BET plot is severely restricted to the p/p° range of 0.05-0.20. The alternative form of linearization of the BET equation employed by Parra and coworkers [18] appears to be more convenient for a microporous solid since the choice of the appropriate experimental interval is free of ambiguity. The BET equation, however, is subject to various limitations when applied to microporous carbons. Thus, constrictions in the microporous network may cause molecular sieve effects and molecular shape selectivity. Diffusion effects may also occur when using N2 at 77 K as the adsorbate since at such low temperatures the kinetic energy may be insufficient to penetrate all the micropores. For this reason adsorption of CO2 at higher temperatures (273 K) is also used. CO2 and N2 isotherms are complementary. Thus, whereas from the CO2 isotherm micropores of up to approximately 10"9 m width can be measured, the N2 can be used to test larger pores. Despite these limitations the BET surface area is the parameter most commonly used to characterize the specific surface area of carbon adsorbents. On the basis of volume-filling mechanism and thermodynamic considerations, Dubinin and Radushkevich [19] found empirically that the characteristic curves obtained using the Potential Theory for adsorption on many microporous carbons could be linearized using the Dubinin-Radushkevich (DR) equation,

where Vg is the micropore volume, E is an adsorption energy, flh an adsorption characteristic constant dependent on the adsorbate, and e= RTln (p°/p) is the adsorption potential at the temperature T (K). For some microporous carbons the DR equation is linear over many orders of magnitude of pressure. For others, however, deviations from the DR equation are found. For such cases the Dubinin-Astakhov equation has been proposed in which the exponent 2 of the DR equation is replaced by a third adjustable parameter, n, where 1 < n < 3. Both the BET and the Dubinin models are widely thought to adequately describe the physical adsorption of gases on solid carbons. BET surface areas from many microporous carbons range from 500 to 1500 m2 g"1. However, values of up to 4000 m2 g"1 are found for some super-activated carbons and these are unrealistically high. The relatively high values of the surface areas of activated carbons are mainly duetothe contribution of the micropores and most of the adsorption takes place in these pores. At least

8

Menendez-Diaz and I. Martín-Gullón Martin-Gullon J.A. Menéndez-Díaz

90-95% of the total surface area of an activated carbon may correspond to micropores. However, meso- and macropores also play a very important role in any adsorption process since they serve as the passage through which the adsorbate reaches the micropores. Thus, the mesopores, which branch off from the macropores, serve as passages for the adsorptive to reach the micropores. In such mesopores capillary condensation may occur with the formation of a meniscus in the adsorbate. Although the surface area of the mesopores is relatively low in most activated carbons, some may have a well developed mesoporosity (200 m g" or even more). In addition, depending on the size of the adsorbate molecules, especially in the case of some organic molecules of a large size, molecular sieve effects may occur either because the pore width is narrower than the molecules of the adsorbate or because the shape of the pores does not allow the molecules of the adsorbate to penetrate into the micropores. Thus, slit-shaped micropores formed by the spaces between the carbon layer planes are not accessible to molecules of a spherical geometry, which have a diameter larger than the pore width. This means that the specific surface area of a carbon is not necessarily proportional to the adsorption capacity of the activated carbon. Pore size distribution, therefore, is a factor that cannot be ignored. The suitability of a given activated carbon for a given application depends on the proportion of pores of a particular size. In general highly microporous carbons are preferred for the adsorption of gases and vapours —and for the separation of gas molecules of different dimensions if the carbon possesses a suitable distribution of narrow size pores (molecular sieves)— while well developed meso- and macroporosity is necessary for the adsorption of solutes from solutions. 3.2, The importance of the surface chemistry It has already been pointed out that a high surface area and an adequate pore size distribution are necessary conditions for a carbon adsorbent to perform well in a particular application. However, there are many examples of carbons with similar textural characteristics, which show a very different adsorption capacity with the same adsorbate [12]. The reason for these different behaviours is that an adequate porous texture is a necessary but not a sufficient condition for the optimization of the adsorption capacity of activated carbons. The nature and amount of surface groups that may be present on the carbon surfaces must also be taken into account. Carbon atoms located at the edges of the basal planes are unsaturated carbon atoms, which possess unpaired electrons. These sites are usually bonded to heteroatoms giving rise to surface groups. Among these groups, oxygen-containing surface groups are by far the most common in carbons. In particular, activated carbons (non-graphitizable carbons) have a relatively large edge area, which results in a strong propensity for oxygen chemisorption. Thus, molecular oxygen can dissociate into atoms that react chemically with atoms of carbon to form oxygen surface compounds. This oxidation process is particularly significant as the temperature of the reaction increases, but also at room temperature for carbons previously

9

Types of Carbon Adsorbents and their Production

treated at high temperatures, which present a highly reactive surface. Oxygen-containing surface groups are not only formed by reaction with oxygen but can also result from reaction with many other oxidizing gases such as ozone, nitrous oxide, carbon dioxide, etc. and with oxidizing solutions like nitric acid, hydrogen peroxide, etc. Thus, the surface chemistry of activated carbons can be tailored by oxidation with different agents in order to create oxygen functionalities or by heat treatment in order to remove them either selectively or completely depending on the temperatures used [20]. Fig. 3 summarizes the most important types of surface groups that may be present on carbon surfaces. Carboxyl O x / O H Hidroxy1 H """• C OH Ether

Carbonyl O o Lactone ||

quinone Q O

O

Chromene-like

Pyrone-like

Fig. 3. The most important types of surface groups that may be found on a carbon surface As mentioned above, oxygen-containing groups are the most common and abundant, but hydrogen is also found combined with edge carbon atoms. In addition, treatments with ammonia, melamine or urea can be used to introduce nitrogenated functionalities, which confer special characteristics on the activated carbons. Of course not all of these functionalities are present at the same time on a carbon. The nature and amount of the different surface oxygen-containing carbon groups on a carbon surface may vary depending on the oxidation conditions. The surface sites associated with functional groups represent a small proportion of the total surface area. However, small variations in the chemical nature of an activated carbon may produce important changes in its adsorption capacity. The importance of the surface groups lies in the fact that their presence or absence can have an important effect on the interaction of carbons with different adsorbates. Two principal effects must be considered. One is the modification of the hydrophobic/hydrophilic character of the carbon. Carbons are, in general, of a hydrophobic nature. However, the presence of polar oxygen-containing

10 10

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

surface groups causes an increase in hydrophilicity since water molecules can form hydrogen bonds with the oxygen atoms of the carbon surface. These molecules may in turn form new hydrogen bonds with new water molecules. The mechanism is schematized in Fig. 4. This effect is of particular importance in the preparation of carbon-supported catalysts from aqueous solutions, where the wettability of the carbon will determine the degree of impregnation of the carbon surface by the solution that contains the catalyst to be supported [12]. In the case of the adsorption of the compounds in a gas stream, the presence of oxygen functionalities can be detrimental since the adsorption of moisture (water molecules) may block the access of the adsorbate to a part of the mieropores, thereby impairing the adsorption capacity of the activated carbon.

& II

Fig. 4. Increase in the hydrophilic character of a carbon surface as a consequence of the presence of oxygen-containing surface groups The other important effect of the surface groups is the influence that they have on the acidic or basic character of the carbons. Activated carbons are amphoteric by nature. That is, they have acidic and basic sites that coexist on their surface. Thus, the carboxyl, hydroxyl and lactone groups behave as acidic groups, if the pH of the medium is higher than the pKa of these groups (basic medium). However, the nature of the basic sites of the activated carbons is more controversial. It is generally admitted that pyrone-like and chromene-like groups [21], among the oxygen functionalities, and the delocalized Tt-electrons of the basal planes [22] are the main basic sites that occur on a carbon surface, although some nitrogen-containing functionalities can also behave as basic sites (these less common functionalities are not discussed in this introductory chapter). However, whether basic oxygen-containing basic groups have a greater impact on carbon basicity than the delocalized it-electrons of the basal plane is an issue that can only be resolved after an evaluation of the relative concentration of the different types of groups and their strengths as bases (pKa) on the carbon surface [20]. Unfortunately these two values cannot be easily determined. By the same token, the overall acidity or basicity of a given carbon depends on the concentration and the strength as acids or bases of the acidic and/or basic sites on the carbon surface. Thus, if the acidic groups are more numerous or their overall acidic strength is higher than that of the basic groups, the

11 11

Types of Carbon Adsorbents and their Production

carbon will be of a basic nature and vice versa. Another factor to be considered is the pH of the medium in relation to the point of zero charge (pHpzc) of the activated carbon. Thus, if the pH > pHpzc, acidic functionalities will dissociate, releasing protons into the medium and leaving a negatively charged surface on the carbon. On the other hand, if the pH < pHpzc, basic sites combine with protons from the medium to leave a positively charged surface. This behaviour is schematically represented in Fig. 5. Basic medium

Acidic medium

o

Fig. 5. Schematic representation of the acidic and basic behaviour of the oxygen-containing surface groups and delocalized ji-electrons of the basal plane

In the light of the above considerations, the interactions and the adsorption capacity of the activated carbons can be optimised by modifying the surface chemistry of the carbon (and/or the pH of the medium, when this is possible). Thus, in a simplistic approach, basic carbons are preferable for adsorbing acidic molecules while acidic carbons will perform better for the adsorption of basic compounds. Moreover, the adsorption of cations will be favoured (by electrostatic forces) if the carbon surface is negatively charged, while the adsorption of anions will be enhanced on a positively charged surface.

4. PRODUCTION OF CARBON ADSORBENTS Activated carbons were the first adsorbents to be developed. As stated in previous sections, activated carbons are produced from a solid carbonaceous based material, which is non-graphitic and non-graphitizable, and has an initial isotropic structure. The precursor is transformed or 'activated' by means of medium to high temperature treatments, which remove solid mass, and at the same time, create pores where the removed mass was previously located. The common properties of activated carbons and other kinds of carbon adsorbents is their well developed pore network, and the similar ways in which they are

12 12

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

produced. In this section details of activated carbon production (raw materials, chemical process, industrial reactors, etc.) will be presented. 4.1. Raw materials Since an activated carbon is structurally a non-graphitic carbonaceous material, almost any carbonaceous solid material can be converted into activated carbon. There are, therefore, plenty of possible raw materials such as wood, ngnocellulosic biomass, peat, lignite and coals that can be used to make activated carbon. However, there are some limitations. Since activated carbon is disordered and isotropic, the raw material must not be allowed to pass through a fluid or pseudofluid state due to the fact that the solid structure tends to transform into an ordered structure. Consequently, coking coals are not an appropriate raw material, unless they are first treated with anti-coking processes (e.g. low-temperature preoxidation). Similarly, thermoplastic wastes are not a suitable raw material because they melt with increasing temperature, losing their initial shape and particle size. On the other hand, a stabilized solid char from a thermoset material does have the necessary characteristics of an appropiate raw material. In practice, wood, coconut shells, fruit stones, coals, lignites, petroleum coke, etc. are all inexpensive materials with a high carbon content and a low inorganic content, and consequently, are suitable for use as an activated carbon precursor. The resulting properties of the product are dependent on the precursor, and consequently, the carbons can be tailored for selected applications. Furthermore, the resulting activated carbon properties are also influenced to a great extent by the activation treatment. Selection of the appropriate raw material is based on the following criteria [23]: •

Possibility of yielding a good activated carbon in terms of adsorption capacity, high density and hardness. • Low in inorganic matter. The adsorption capacity is measured per mass unit, and since inorganic materials are non-porous, their presence reduces the adsorption capacity [24]. • Availability and cost. As with any other product, the price of the raw material affects the final cost, so a high availability is important to ensure stable prices. It should also be taken into account that there is a considerable mass loss in all the activation treatments, and the lower the product yield, the higher the cost. The product yields may vary considerably, and can be as low as 5-10% for wood-based carbons. Moreover, raw material availability obviously depends on the part of the world in which the plant is located.

Fig. 6 shows the estimated world production of activated carbon per region and per precursor in 1993 [25]. Total production exceeded 350000 tonnes/year: coconut shells (34%), coal (31%) and wood (24%) these being the most important precursors. However, the raw materials used per world region shows a very different picture. In Europe, the raw material

13 13

Types of Carbon Adsorbents and their Production

most used was peat (36000 ton/year, 36% of total European production), whereas the production of coconut shell based carbon occupied fourth place with only 12% of the share. On the other hand, coconut based carbons led in Asia with over 60% of the share, North America occupying a position between Asia and Europe, with nearly the same production of coal, coconut and wood based activated carbons.

Rest of Asia 39 -, 11%

North America ——^ 140 39% j)

Japan Japan 76.6 76.6 22% 22%

Wood 85 24%

Peat 36 10%

Other 3.6 1% 1%

^

Coconut shells 124 124 34%

Coal 110 31%

Europe 100 28%

Fig. 6. 1993 world production of activated carbon per region and per precursor [25]. Production is given in kton/year

In 2002 the total production of activated carbon was estimated to be 750000 ton/year [26], which represents a rise of two times the level of production in less than 10 years. This change comes at a time when the production capacity is shifting from western industrialised countries to China and South-East Asia, where raw material, energy and labour costs are lower. By world regions, Asia is leading world production with 54% of the total share, followed by America with 32% and Europe with only 14%. By countries, Fig. 7 shows the current world producers, where it can be seen that China manufactures 23% of the world production. UK 3% France 4%

Others 12%

USA 29%

India 6% Philippines 5% Netherlands 6%

Japan 12%

China 23%

Fig. 7. World production capacity per leading country [26]

14 14

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

As stated above, the precursor is very important for both the activation process and the final properties of a given carbon adsorbent. As will be commented in section 4.2, the production methods can be grouped into i) physical, or more properly thermal activation, which consists in thermal devolatihzation followed by char gasification with an oxidating agent, and ii) chemical activation, which consists basically in the inert carbonization of mixtures of the raw material with a chemical agent. When the raw material is subjected to thermal activation, the higher the volatile content of the raw material, the lower the yield. This type of treatment, however, gives rise to a more accessible micropore structure, thereby ensuring a higher and more homogeneous pore development. On the other hand, low volatile raw materials (such as anthracites or cokes) lead to a greater density and hardness. Most activated carbon materials are produced in either granular or powdered forms, depending on the final application. The type of raw material used also determines whether the activated carbon produced from it will be in powdered or granular form. 4.2. Thermal activation There are many ways of producing activated carbon, and thousands of patents have been registered worldwide. Nevertheless, as mentioned in the previous section, all these production methods can be classified into two clearly defined groups: •

•

Thermal Activation for physical activation). Generally (although not necessarily), this consists of two consecutive steps. The first step is thermal carbonization of the raw material, where devolatilization takes place, carried out at medium or high temperatures, to produce a char rich in carbon. The second step is activation, where the remaining char is partially gasified with an oxidizing agent (mostly steam) in direct fired furnaces. If both steps are carried out simultaneously, the process is called direct activation. Chemical Activation. This is carried out in a single carbonization stage. The raw material is first impregnated with appreciable amounts of a chemical agent (the usual chemical reagent vs. raw material ratios being around 1-4), and then heated up. The product must be washed to eliminate any excess chemical agent after carbonization. The thermal treatment temperature depends on what chemical agent is used. The most common are phosphoric acid, ZnClj and alkaline hydroxides.

Thermal activation is generally divided into two thermally dependant steps, carbonization followed by activation. However, additional treatments are often applied. These are common industrial operations and are commented on below. The main scheme for activation production is shown in Fig. 8.

Briquettes Briquettes \

Raw material

Powder, Powder, granular granular?~i .

\ '*, • • • • • - ''

preoxidation preoxidation

\

J

Wash

""-A

••••••, . ; ; - •- , .

Carbonization

Fig. 8. General flowsheet for the production of thermal activated carbons

Mill Mill Briquetting Briquetting

Crushing, sieving

Pretreatments Pretreatments

Activation

1

f

Types of Carbon Adsorbents and their Production

15

16 16

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

4.2.1. Pretreatments The raw material will need some pretreatments prior to its actiYation. In the case of regular powdered or granular broken carbons, the raw material must first be prepared to the required particle size by crushing and sieving. Sometimes, a washing step is applied with water or acid in order to remove any dirt coming with the raw material and to reduce any mineral matter. If the raw material is a coking coal, preoxidation treatment is necessary: a long period of air-preoxidation at moderate temperatures (say 453-573 K), stabilizes the coal by creating crosslinkage within the coal structure, which prevents the fluid phase from occurring during carbonization [27]. Besides the powdered or granular carbons, the initial raw material is sometimes pulverized, briquetted and again crushed to an appropiate particle size. This treatment is normally applied to low volatile precursors (medium-high rank coals), where the meso and macropore network is not well developed in the carbonization step and the activating agent can diffuse poorly in the subsequent activation step. If a granular material with a well-developed internal porosity is required from these precursors, the coal must be finely pulverized and then briquetted. When the briquettes have been crushed, the material resembles the granules which have only been milled, but they have a well developed network of transport pores, so that the activated agent can diffuse easily to all the individual coal particles in the briquette. As an example, Fig. 9 shows the meso and macropore size distribution of two bituminous coal based activated carbons, one of which has not been briquetted, whereas the other has. The pore size distribution was determined by mercury porosimetry. Fig. 9 shows that for the briquetted carbon, there is a huge peak at pore sizes around 10"6 m, whereas the amount of pore volume distribution for the non-briquetted carbon is similar regardless of the pore size range. The peak is often called briquetted peak (or extrusion peak, if the carbon granules are produced by extrusion) and it is derived from the interparticle void volume among the pulverized particles of the briquettes. 4.2.2. Carbonization As Fig. 8 shows, pre-treatment is followed by the carbonization step. This step is usually critical in the overall production of activated carbons, because it is during this step when the micropore structure starts to form. The terms pyrolysis and carbonization are often used without distinction, and both processes are almost identical. However, there are differences in the objectives of each process. Both pyrolysis and carbonization refer to the thermal degradation of a carbonaceous material (coal, lignite, wood, polymer, etc) in an inert atmosphere, where the total or partial devolatilization of the raw material takes place. The difference is that, in pyrolysis, it is the gaseous and volatiles compounds evolved from the solid that are the objective of the process, and all process conditions (temperature, residence time, etc.) are focused on producing gaseous compounds (e.g. pyrolysis conditions to favour the light olefin fraction), regardless of the final char residue. On the other hand, in a carbonization process, thermal treatment focuses on the final properties of the char (e.g. porous texture, hardness, density, etc), regardless of the volatile compounds evolved.

17 17

Types of Carbon Adsorbents and their Production 0.06 -non-briqueted non-briqueted

diff. pore volume (cm3/g)

•briquetted briquetted

E 0.04 0.04 o 4>

I

0.02 -

0 1

10

100 100

1000 1000

10000

pore diameter (nm)

Fig. 9. Mercury porosimetry of a briquetted and a non-briquetted based granular activated carbon A scheme of the carbonization (or pyrolysis) of a carbonaceous material is shown in Fig, 10. The starting material, based on an organic macromolecular structure, decomposes during the thermal treatment to yield: • •

A gaseous fraction, rich in hydrogen, light hydrocarbons and tar. A solid fraction, rich in carbon, called char.

The gases and vapours that first evolve from the solid are primary products. These vapours come directly from fragments of the carbonaceous structure, so there are many radical species. Once this fraction is in gas phase, these species may react among themselves, depending on the treatment temperature and residence time, to yield secondary (cracking) products. At medium temperatures (say 773 K), there are few secondary reactions, and the gas phase consists mainly of stabilized primary products. Consequently, the gas composition is highly dependent on the precursor used under mild conditions. As the temperature is increased, the secondary reactions become more important, yielding at extreme conditions (above 1273 K) methane, hydrogen and soot If the temperature is 1073-1173 K, intermediates of soot formation (stabilized polyaromatics) can appear, together with oleflns, methane, hydrogen , water and carbon oxides. If the heat treatment is carried out at high heating rates, cracking reactions are so quick that soot may be deposited over the char particles. A knowledge of what is happening, in terms of chemical reactions, during the carbonization is important for an understanding of the final properties of the final char.

18 18

J.A. Martin-Gullon J.A. Menendez-Diaz Menéndez-Díaz and I. Martín-Gullón

Volatiles

H22O H

CH 4

CH44 OH

H2

H2

CO

Devolatilization

Secondary cracking

H 2O

CO

O O

O

O O O

O O

O

O

Condensation Soot deposition

O

Raw material

o

o

Char

Fig. 10. Typical carbonization reaction scheme of a carbonaceous material With respect to the solid char formed, there is an enrichment in both relative carbon content and aromaticity, compared to the precursor. Because a lot of mass is evolved during carbonization, this increase in aromaticity is accompanied by an increase in incipient microporosity, since a lot of functional groups and bridge chains are evolved, yielding void spaces. The char is composed of a group of disordered short graphitic crystals. Between the crystals there is little voidspace (the micropores), which are often not accessible from the external surface because the meso/macropore network is blocked by soot deposition (especially at high temperatures and high heating rate treatments). Thus, the carbonization heating rate is very important in establishing the final textural properties of the char. A high heating rate produces a very quick devolatilization, giving rise to a solid with a well developed meso and macropore network, a low density and a low abrasion and hardness index. On the other hand, low heating rate chars experience a slow release of volatiles during the devolatilization stage. This situation does not favour the formation of a large meso/macro network, and consequently, both density and hardness values are higher than in the case of high heating rates. However, the incipient microporosity developed is as high as that created at the high heating rates. In consequence, a low heating rate treatment, at not very high

19 19

Types of Carbon Adsorbents and their Production

temperatures and long soaking times, seems to offer the appropriate conditions for producing chars that upon activation give rise to dense and hard activated carbons. In industrial processes, it is common to use direct heated kilns in an oxidant atmosphere (steam with exhausted air from burners), at very low stoichiometric ratios. Final temperatures of 873 K and heating rates of 100-300 Kh"1 are reported [28]. 4.2.3. Activation The product from the carbonization step has still only an incipient porous structure and cannot be used as an adsorbent unless this porous structure is enhanced, upgraded, or "activated". The thermal activation consist in partially oxidizing the char with steam, carbon dioxide or air. These gases react with the carbon atoms and remove some of the mass of the internal surface of the solid, in the incipient micropores, creating a well developed mieroporous material. In addition, some internally blocked micropores may also become accessible due to tap burn-out. The activation rate is conditioned by the characteristics of the precursor and the activating agent. The most reactive agent is oxygen, whereas the lowest reactive agent is carbon dioxide. 4.2.3.1. Activation with oxygen Chemical reaction of carbon with oxygen yields simultaneously carbon monoxide and dioxide: C + O2 2C + O2

> CO2 > 2CO

AH=-387 kJ mol"1 AH=-226 kJ mol"1

(3) (4)

Both reactions are highly exothermic. Although the combustion reaction is one of the most important for mankind, the reaction mechanisms are still uncertain. It seems that both carbon monoxide and carbon dioxide are primary products, and the ratio CO/CO2 increases with temperature. Due to the high enthalpy, the temperature of the reaction is extremely difficult to control, and often a reaction runaway (self ignition or uncontrolled temperature) takes place so that the reaction is governed by diffusion control burning the carbon only on the surface of the particle and not in the inner surface. Consequently oxygen activation is scarcely applied. However, there is a typical process for producing a very specific developed mieroporous material using oxygen activation at laboratory scale. It consists of a cyclic process [29]: first oxygen is chemisorbed at low temperatures, below the ignition point (say 473-573 K); next the atmosphere is changed to inert, while at the same time the char is rapidly heated up to high temperatures. The initial chemisorbed oxygen burns out the corresponding carbonoxygen complexes. After a short time, the sample is cooled down in an inert atmosphere. This method of producing activated carbon gives rise to extraordinary high capacity molecular sieves, and adsorbent materials with a narrow and defined pore size distribution. However, this is a very expensive process for carrying out at industrial scale.

20

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

4.2.3.2. Activation with steam The steam gasification reaction also "burns out" carbon (like the combustion reaction) but follows an endothermic pattern. C + H2O H 2 +CO

AH=132kJmol"1

(5)

This reaction has been extensively studied, not only for activating carbons, but also because it is the main reaction of water shift. Because this reaction is endothermic, the reaction runaway does not take place. In addition to the heterogeneous reaction, gaseous products may react among themselves in gas phase, yielding: CO + H2O o H 2 +CO 2 C + 2H2 CH4

AH = -41.5kJmol"1 AH = -87.5kJmol4

(6) (7)

Nevertheless, the kinetics of the two latter reactions is very low and they hardly take place. From the kinetic point of view, the steam-carbon gasification reaction shows a typical reaction rate for most precursors, the conversion degree being expressed as: d

^ = K{\-aY at

(8)

where a is the carbon conversion degree (mass of reacted carbon vs. mass of initial carbon), K is the apparent kinetic constant (which depends on temperature and the partial pressures of the gas compounds) and n is the reaction order, which can have the following values [30]: •

n = 0. This is very unusual, although sometimes the reaction rate is independent of the carbon conversion degree in some catalyzed gasifications. • n = 1. In this case, the reaction follows the uniform conversion model [31], which indicates that it proceeds under chemical reaction control, and the reactive agent first diffuses through the pore network and later reacts at any internal active site. The particle will retain its initial size during the conversion • n = 0.67. This reaction order corresponds to the shrink core reaction model. The reaction is under diffusion control, and the oxidizing agent reacts only on the surface because it has no time to diffuse internally, so the particle size becomes smaller as the conversion degree increases. In this case, the apparent kinetic constant will depend on the particle size and parameters derived from internal diffusion, in addition to the parameters mentioned before. In order to produce activated carbons, it is necessary that the steam gasification progresses under chemical control to allow internal porosity to develop inside the char particles.

Types of Carbon Adsorbents and their Production

21

Otherwise the reaction would only take place on the outside of the particle. Due to the specific nature of steam gasification, chemical control is achieved at relatively high temperatures. Consequently, steam can be considered as a good activating agent. In most situations, chemical control can be maintained at temperatures up to 1273 K. Internal porosity (in terms of the BET specific surface area or the micropore volume) increases with the conversion degree, although the micropore sizes also increase, which indicates that activation is taking place throughout the porosity. The way in which the internal porosity is developed (development of existing micropores by widening, which may even become mesopores at high conversion degrees) depends on what the active sites for the previously complex C(H20) are. The way in which porosity develops is similar for all chars but it is not exactly the same for several reasons. First, each char has a different origin and chemical structure, so the active sites are not the same. Second, most of the precursors have a mineral content, which is able to catalyse the steam gasification. This catalysis may affect slightly the way in which porosity develops. One example is the coconut shell char. Compared with other raw materials, coconut shells are excellent due to their extraordinary hardness and high and narrow micropore development. The reason for this behaviour is that the mineral matter of coconut shells contains appreciable amounts of potassium. With regards to the apparent kinetic constant, there is experimental evidence that hydrogen inhibits the reaction kinetics significantly, whereas carbon monoxide does not. From this evidence, the apparent kinetic constant is an expression of the LangmuirHinshelwood type [32]:

»l + k2

±l

\RT where h are Arrhenius type preexponential factors, Ea the activation energy, Ez and E3 adsorption enthalpies and P,- the partial pressures. Steam activation is the most widely used method for producing activated carbons in the world. In terms of adsorption capacity, it easily reaches specific surface areas of 1000 m2 g"1 at 50% of activation conversion degree (regardless of carbonization yield), when an acceptable raw material with an initial ash content below 10% is used. Basically, steam activated carbons are microporous materials, with a micropore size that increases with the activation degree, but with no mesopore development. As an example, Fig. 11 shows the nitrogen adsorption isotherms of activated carbon produced by the steam activation of an anthracite [33]. Each adsorption isotherm corresponds to an activated carbon with activation degrees of 20, 35, 50, 70 and 80% burn-off (samples: AC-720, AC-735, AC-750, AC-770 and AC-780 respectively), the sample AC-700 being the non-activated char. These isotherms not only provide specific numerical parameters correlated with adsorption capacity, but also provide certain qualitative information derived from their shape.

22 22

Menendez-Diaz and I. Martín-Gullón Martin-Gullon J.A. Menéndez-Díaz

Thus, the char (non-activated) is at the base of the plot, with nearly no adsorption at all. For the rest of the samples, the sharp increase in adsorption at nearly nil relative pressures and the knee close to a horizontal plateau indicate a narrow microporosity (pores below 1 run). As the knee becomes more open, the micropore structure changes from the narrow micropores to wide micropores. The occurrence of a positive slope of the plateau (in the relative pressure range of 0.2 to 0.7) is an indication that mesopore development has started. Thus, from Fig. 11 it can be inferred that at low burn-off, the carbon has only narrow micropores. As the burnoff increases, the adsorption capacity clearly increases (the isotherm plateau has higher values) and the isotherm knee becomes more open, which is indicative of micropore development and an enlargement of the micropores. It should be noted that the knees start from the same point (close to the Y axis) for the last three samples, which is indicative of this enlargement. Only at very high burn-offs (80%) does this micropore widening lead to the partial formation of mesopores (positive slope). This tendency is valid for nearly any precursor with steam activation, with some differences regarding the starting material and kinetic conditions (e.g. the mesopore development could be higher at high temperatures).

• AC-700

400

• AC-720 •• AC-733 - AC-750

350

" AC-770 - AC-780 - ' " /

d 300

j | . 200 *• • 150

*~

"

100 50 0

0.0

0.4

0.6

0.8

1.0

p/pO

Fig. 11. Nitrogen adsorption isotherms for steam anthracite based carbons with different burn-off degrees. Reprinted from [33], with permission ftom Elsevier

4.2.3.3. Carbon dioxide activation The heterogeneous reaction between carbon and carbon dioxide yields carbon monoxide: C + CO2 o

2CO

AH=159kJmor

(10)

23

Types of Carbon Adsorbents and their Production

following an endothermic pattern, which is positive for the activation process. The kinetic expression for the process is:

RT ) P c m

dt

{l-af

(11)

exp —*- \PC

RT

where the meanings of all the parameters have been previously defined. The value of the reaction order n, as in the case of the steam gasification, may be: 0, 0.67 and 1 depending on the reaction control. From the kinetic point of view, carbon dioxide gasification is slower than steam gasification, for the same temperature and partial pressure. Depending on the partial pressures of the reagents and products, the temperature where the transition takes place is around 1123 K (or even lower), when the reaction rate is not very high. Under chemical control, the porosity development attained is as high as that obtained with steam [34]. On the other hand, under diffusion control, porosity development is poor and unsatisfactory, giving rise to external particle burning [35,36], Consequently, in order to work with carbon dioxide it is necessary to find out whether the process proceeds under chemical control or not. Usually, laboratory reactors with a low carbon dioxide mass flow vs. mass of carbon, operate under chemical control, whereas high carbon dioxide mass flow vs. mass of carbon (fluidized beds, rotary kirns, etc.) fall under diffusion control [37]. n (mmol/g)

0

0.2

0.4

0.6

0.8

P/P° Fig. 12. Nitrogen adsorption isotherms of COj activated carbons from a bituminous coal char, with activation burn-offs of 25, 48 and 62% [38] An example of the variation of porosity development with the activation degree can be observed in Fig, 12. This corresponds to the activation with CO2 of a bituminous coal char at 25, 48 and 62% bum-off. Once again, it can be seen that a low activation degree gives rise to a narrow microporosity (close isotherm knee) and that higher burn-offs also produce

24

J.A. Martin-Gullon J.A. Menendez-Diaz Menéndez-Díaz and I. Martín-Gullón

microporous materials. It must be noted that contrary to what occurred in the case of steam activation (Fig. 11), the isotherm knees are less open (for all the samples) than those shown in Fig. 11, and they separate from the Y axis at different heights. 4.2.3,4. Carbon dioxide vs. steam activation Although industries have always used steam instead of carbon dioxide to produce activated carbons by thermal activation, there has been constant disagreement about which oxidizing agent performs better. It is clear that steam reacts under chemical control over a wide temperature range, thereby developing porosity. COi, on the other hand, may react under undesirable diffusion control, which represents an obstacle for its use as an oxidizing agent. The porosity developed (assuming chemical control for both agents) from these reactive gases is compared in Fig. 13. The Fig. shows two activated carbons, with similar burn-offs, from the same starting material, one activated with steam and the other with CO2 [39], It is clear that both samples attained similar adsorption capacities (similar isotherm heights). However, the porosity developed by CO2 is a little narrower than that of the carbon activated with steam (slightly more open knee for the latter, with some positive slope in the plateau). The choice of which carbon to use will depend on the final application. Thus, carbons with narrow or normal micropore sizes are usually more appropriate for gas adsorption, whereas carbons with an open micropore structure are preferred for liquid phase adsorption, since the wide micropores facilitate diffusion of the solute into the particles. 16

steam 14

N (mmol/g)

12

CO2

10 8 6 4 2 0 0

0.2

0.4

0.6

0.8

1

P/Po

Fig. 13. Nitrogen adsorption isotherms of steam and CO2 activated carbons at similar burn-offs. Reprinted with permission from [37], Copyright 2000 American Chemical Society

25

Types of Carbon Adsorbents and their Production

Fig. 14 shows the variation in the pore diameter with the pore size distribution at different activation degrees for these two activating agents, for the activation of a bituminous coal char in a fixed bed reactor [39]. The carbons activated with CO2 present a micropore structure of the same pore size regardless of bum-off. Only the adsorption capacity is improved when the bum-off increases. This means that new micropores, of similar sizes to the existing ones, were developed. On the other hand, the carbons activated with steam present a pore widening which increases with the burn-off. Thus, the mean micropore size increases with burn-off enhancing, its adsorption capacity due to the enlargement of the existing pores and the destruction of the initial narrow micropores. In a comparison of the micropore size distribution at 1 nm, it can be observed that the carbon with the higher degree of activation has practically no pores of this size with respect to the less activated carbons. This does not mean that CO2 activated carbons are better for adsorption applications than the steam activated carbons. Micropores are necessary since they are the sites responsible for adsorption. However, some applications will require only narrow micropores while others (specially liquid applications) will require wide micropores. 1 0.9

dW/dx

CUA12-29

CO2

0.8

CUA12-42

0.7

CUA12-57

0.6

CUA12-66

0.5 0.4 0.3 0.2 0.1 0 0

0.5 0.5

11

1.5 1.5

22

2.5 2.5

3

pore pore diameter (nm) (nm)

dW/dx

1 0.9

steam

0.8 0.7 0.6

CUA12-s22 CUA12-s42 CUA12-s69

0.5 0.4 0.3 0.2 0.1 0 0

0.5

11

1.5 1.5

22

2.5

3

pore pore diameter diameter (nm) (nm)

Fig. 14. Micropore size distribution of CO2 and steam carbons at different burn-offs. Reprinted from 13*->|- wuli permission iruiii Flscxicr

26

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

Although the tendency explained is valid in relative terms for any raw material, each specific raw material must be studied individually, as the quantitative parameters cannot be predicted due to the influence of the organic structure of the precursor and the different behaviours of the mineral contents of each precursor. As mentioned above, steam activation is preferred by industry to carbon dioxide activation. It has the advantage of having a higher gasification kinetics and, at the same time, it operates under chemical reaction control up to very high temperatures. This ensures the development of porosity while the reaction is taking place. At industrial scale, direct fire-heated rotary kilns and multiple hearths are commonly used in steam activation. The type of reactor is also an important factor, although to a lesser extent than the activation agent, in determining the final properties of the activated carbon, due to the different characteristics of the gas-solid contact. The diffusion of gas through the carbon particles in a rotary kiln is poor, resulting in a low gasification kinetics due to the low reactive partial pressures. This effect is positive for narrow micropore development (as chemical control is ensured), although higher temperatures and higher residence times (and consequently, larger kilns) are required. On the other hand, in multiple hearth reactors the gas solid contact is more intense, resulting in a high reaction kinetics (lower temperatures and residence time are then required), although the microporosity may be a little wider due to the burning of the particles externally and/or the burning of existing pores instead of the development of new narrow micropores (which would require a total chemical control). Steam gasification is a endothermic process, and external heating is obviously necessary for maintaining the activation temperatures. At industrial scale, the heating process will add to the cost of the final product. Direct fire-heated systems are the most economic, and they are used in steam activation. This means that, at industrial scale in a continuous operation, both steam and air are fed into the kiln. As steam activation is endothermic, the heat required is provided by the exothermic gas phase reaction of CO and H2 (steam gasification products) with the appropiate amount of air. Air must be fed into the reactor at different reactor points in order to prevent carbon-oxygen gasification and so that they react only in gas phase. In this way, heat supplied by gas phase oxidation can balance the steam gasification requirements, and no fuel is needed by the fire burners except for the non-stationary starting operations. 4.3. Chemical Activation A second commercial route for producing activated carbons involves the reaction of a carbon precursor with a chemical reagent, and is known as chemical activation. Porosity develops at the same time as the transformations that occur during heat treatment in an inert atmosphere at temperatures in the range of 623 to 1173 K. Compared to the two kilns that are normally employed for carbonization and activation in industrial thermal activation, chemical activation only requires a single kiln. The reagents that have been most used by industry are zinc chloride, phosphoric acid and potassium hydroxide. Each reagent produces a very different pore development in the carbon precursor. The feasibility of chemical activation processes is strongly dependent upon efficiently recovering the reagent for recycling. This involves

27

Types of Carbon Adsorbents and their Production

subsequent leaching stage, followed by an additional operation consisting of drying the washed carbon. This section will describe the chemical activation process with the most commonly used reagents: zinc chloride, phosphoric acid and potassium hydroxide. Other low-volume processes utilize various Lewis acids such as aluminium chloride and ferric chloride with fibrous materials such as rayon (a product of cellulose) to produce activated cloths or fibres, 4.3.1. Zinc Chloride Activation The zinc chloride process was the principal method of chemical activation up to 1970, and it is still used in China. The preferred precursors are those with a high amount of volatiles (and oxygen content) such as lignocellulosic materials. Wood, therefore, is the most common precursor, although other materials, such as ground olive stones, are also used, Low recovery efficiencies, corrosion problems, the presence of residual zinc in the carbon, and the need for stringent control of plant emissions have contributed to their displacement by phosphoric acid. Initially, the precursor is mixed with the chemical in an aqueous solution to form a paste, which is later evaporated. During evaporation, there is a weakening of the lignocellulosic structure due to hydrolysis reactions (with loss of volatile matter), leading to an increase in elasticity and a swelling of the precursor particles [40]. The higher the amount of zinc chloride, the stronger the changes during impregnation and evaporation. After evaporation, the impregnated carbon is heat treated in an inert atmosphere at final temperatures between 773 and 1073 K. The ZnClz restricts tar formation, preventing the contraction of the particle and giving rise to a wide and open mieroporosity (border micro-mesopores). This results in a slightly higher adsorption capacity and a wider porosity than in the case of thermal activation. Fig. 15 shows the nitrogen adsorption isotherms of activated carbons obtained using different ZnCb/precursor ratios.

V (c m 3 /g )

1.21.2

D

0.9

C

0.6

B

A

0.3 0 0

0.2

0.4

0.6

0.8

1

P/Po

Fig. 15. Butane adsorption isotherms of ZaCh based activated carbons with heat treatment temperature at 773 K and Zn/precursors ratios of (A) 0.24, (B) 0.48 and (D) 0.96. (C) was made at HTT of 1073 K and Zn/C of 0.96. Reprinted from [40], with permission from Elsevier

28

J.A. Menendez-Diaz Martin-Gullon Menéndez-Díaz and I. Martín-Gullón

It can be seen that all the Isotherm knees leave the Y axis at the same level, although each one reaches a different plateau level. However, the knees are so open that the plateau is not reached above relative pressures of 0.4, which means that the porosity developed is wider than that obtained with thermal activation. The carbon yield is much higher than that achieved with thermal activation but the particle density and abrasion/hardness values are lower. 4.3.2. Phosphoric acid activation. The ability to produce high surface area carbons with phosphoric acid in an inert atmosphere depends primarily on the structure of the starting material, the acid-to-precursor ratio and the heat treatment conditions (time-temperature profile and maximum heat treatment temperature). As with ZnCl2 activation, wood and some agricultural by-products (such as almond shells and olive stones) as well as carbonaceous materials with a high volatile content are the precursors most commonly used in this method of activation. The steps involved in the production of activated carbons by chemical activation with H3PO4 are: 1) grinding and classification of the starting material; 2) mixing with phosphoric acid (recycled acid plus make-up); 3) heat treatment initially at a temperature between 373 to 473 K in an inert atmosphere. 4) continuing the heat treatment at a final temperature between 673 to 773 K; 5) washing the product to recover acid for recycling, followed by drying and classification of the washed carbon. Both steps 3 and 4, can be carried out in the same kiln/reactor and no cooling down is necessary between steps. Soaking times at these two temperatures are around 1 hour for each. Compared with carbons produced by thermal activation, the wood-based carbons activated with H3PO4 have a lower density, lower abrasion resistance and a more developed mesoporosity. These properties are related to the hollow fibrous structure of wood, which gives rise to an important macropore volume in the activated carbons. Frank Derbyshire and coworkers at the University of Kentucky [7,41] who carried out an extensive study of the H3PO4 activation of different raw materials, proposed an activation mechanism for the activation of lignocellulosic materials, the only materials which display interesting adsorption characteristics with this chemical reagent. Wood or lignocellulosic materials (hard fruit stones and shells) are composed of cellulose (42-50%), hemicellulose (19-25%) and lignin (16-25%). Wood is often compared to a composite material, where most of the cellulose forms microfibrils with a crystallite structure, whereas hemicellulose and lignin form the matrix, an amorphous paste that surrounds the microfibrils. A HsPCVpreeursar ratio of 1.5 is the most frequently used proportion in this method of activation. After the chemical reagent is mixed with the precursor, the mixture is slowly heated up to 373-473 K (low temperature treatment). During this stage, the acid attacks the hemicellulose and lignin hydrolyzing the glycosidic linkage to produce polysaccharides (hemicellulose) and cleaving the aril ether bonds of the lignin [42]. These reactions may be

Types of Carbon Adsorbents and their Production

29

accompanied by secondary reactions such as degradation and condensation. During this low temperature treatment, there is an extensive evolution of CO/CO2 and methane (which would not occur under this heat treatment without H3PO4). This process also leads to a reduction in the volume of the particles. When the temperature is increased, during the second heat treatment, towards the final temperature, the rate of weight loss diminishes appreciably and the structure begins to dilate, developing porosity. Under this regime, crosslinking reactions promoted by phosphate esters among the cellulose fibres are predominant. These produce the effect of dilation, which is maximum at around 473-523 K. Fig. 16 shows the porosity developed in terms of surface area vs. the heat treatment temperature and Fig. 17 shows the particle dimensions vs. the heat treatment temperature for an oak wood precursor, together with the variation undergone when only thermal pyrolysis is applied. It can be observed (Fig. 16) that micropore development begins a little above 473 K, increasing sharply and attaining a maximum at around 573 K, with a BET surface area of 1500 m2 g"1. This micropore increase runs parallel to particle dilation (Fig. 17). The particles undergo an initial contraction, which is then followed by a dilation above 473 K. Mesopore surface area also reaches a maximum at 723 K. As a consequence, the micropores widen into mesopores when the material is heated from 573 to 723 K. After 723 K, the micro and mesopore areas experience a decrease, which is accompanied by a contraction of the particles. These changes are accompanied by a shrinkage of the pore size distribution. The reason for this contraction is the breakage of the (previously formed) phosphate crosslmkages between the cellulose fibres. This leads to an increase in the aromaticity of the solid resulting from a thorough reorganization of the structure of the solid. 2000

'

• C 1500 -

1 1 1 1

"P


S2+ 2H2O

(5)

From the standpoint of organic chemistry hydrogen sulfide can add to double bonds, form hydroxythiols with epoxides and olthiols with aldehydes. 2.4. Mercaptans Mercaptans (thiols) are organic compounds where SH group is bonded to the carbon atom in the molecule. Methyl mercaptan (CH3SH) is a colorless, flammable gas with an extremely strong and repulsive smell. Mercaptans are the products of anaerobic digestion and thus they are present in fossil fuel or digester gas in various concentrations reaching 500 ppm. At very high concentrations methyl mercaptan is highly toxic and affects the central nervous system. Its penetrating odor provides warning at dangerous concentrations. Mercaptans have two primary uses: as chain transfer agents in polymerization and as chemical intermediates in agrochemical, pharmaceutical, and a variety of industrial applications such as gas odorants. They have very high reactivity towards metals. Methyl mercaptan (MM) in the presence of radicals is oxidized to dimethyldisulfide: 2CH3SH + 1/2 O; -* (CH3)2S2 + H2O

(6)

2.5. Thiophenic compounds

R

R S S

Benzothiophenes Benzothiophenes

XSJ

S SX Dibenzothiophenes Dibenzothiophenes X

R S H33C

H3C

CH3

CH 3

Dimethyldibenzothiophenes 4,6 Dimethyldibenzothiophenes

Fig. 4. The most common thiophenic compounds present in diesel fuel

236

T.J. Bandosz

Rcaetant

DBT MW=184

4.6DMDBT MW=212

Identified reaction products

DBT 5-oxide MW=200

4.6DMDBT 5-oxide MW=228

DBT 5,5-dioxide MW=216

4.6DMDBT 5,5-dioride MW=244

HOC fi-methyldibenzothiphene-4-carbaldeMde MW=226 Fig.5. Products of catalytic photooxidation of DBT and 4,6-DMDBT [13], Reprinted with permission form: S. Matsuzawa, J. Tanaka, S. Sato and T. Ibusuki. J. Photochem. Photobio. A, 149 (2002) 83

Thiophenic compounds are naturally present in crude oil. Although hydrodesulfurization targets them and converts toward hydrocarbon and hydrogen sulfide [7] still significant quantities are present after this treatment. The reactivities of the 1- to 3-ring compounds decreases in the order thiophenes > benzothiophenes > dibenzothiophenes. In gasoline benzothiophenes are present, in jet fuel - benzothiophenes and dibenzothiophenes whereas in diesel fuel dibenzothiophenes and 4, 6-dimethyldibenzothiophene (4,6-DMDBT) are found. The latter is considered as the most refractory sulfur compound towards HDS. The most common refractory sulfur compounds in liquid fuels are presented in Fig. 4. Although hydrodesulfurization is difficult to achieve on thiophenic compounds, these species are relatively easily oxidized as a result of photochemical process or in the presence of oxidants providing OH radicals [13,14], As a result, the products listed in Fig. 5 were identified. Besides oxidation, it is believed that at certain conditions, on the surface of undisclosed reactive adsorbent such as that in the S-Zorb process of ConocoPhilips [15, 16], the sorbent surface attracts sulfur, removes it from the molecule and hydrocarbon molecule is released back to the system (Fig. 6). The sorbent likely contains reduced metal, which reacts with sulfur forming sulfldes [7,15].

237

Desulfurization on Activated Carbons

S S

+ H2

ET +

SORBENT

Fig. 6. Example of reactive adsorption in the case of thiophenic compounds [15]

3. REMOVAL OF SULFUR DIOXIDE FROM FLUE GAS Complaints about SOa pollution were known at least back to thirteenth century [5]. The effects caused an increase in the acidity of natural waters, fast rate of abrasion of buildings and monuments, and associated with these health problems. To remedy these problems, the desulfurization of fossil fuels along with removal of SO2 from stock gases are the technologies which have been developing rapidly during the last thirty years. Although conventional methods for SO2 abatement utilize basic scrubbers where acidity of sulfur dioxide is neutralized and salts are formed [9], for removal of low concentration of SO2, activated carbons [17-39] and activated carbon fibers [22, 24] were shown as feasible removal media. Numerous studies indicate good efficiency of SO2 removal on these materials either at low [22, 28, 32] or high temperatures [21, 24, 31, 32]. Process of SO2 adsorption has been studied extensively and such parameters as porosity [18-27, 34-35], surface chemistry [18, 21, 23, 25, 27-29, 32] and constituents of ash [36-39] were taken into consideration. The products of surface reactions were analyzed from the point of view of removal efficiency and the feasibility of adsorbent regeneration [24, 34]. Due to the higher oxidation state of sulfur in SO2 than in H2S, the chemistry involved in immobilization is expected to be much less complex than that for oxidation of hydrogen sulfide. Since usually the process is carried out in the presence of moisture and oxygen, it is generally accepted that sulfur dioxide is oxidized to sulfuric acid as a final product of the reaction. That acid is strongly retained in the pore system of activated carbons. Higher extent of oxidation usually results in more SO2 adsorbed [26]. Adsorption/oxidation of SO2 in oxygen atmosphere and in the presence of water occurs as follows: SOjgas^SOzsds.

(7)

-»2Omj.

(8)

SO 2ads + O a d s ^SO 3 a < t e

(9)

H2Ogas -* H2Oads

(10)

SO3 ads + HaCU, -* H2SO4MS ,

(11)

O2m

where subscripts "gas" and "ads" refer to the presence of reactants in the gas phase and the adsorbed state, respectively. It was also found that three forms of adsorbed sulfur oxides could be present in such a situation. They are: weakly adsorbed SO2, physically adsorbed SO3 (after oxidation of SO2), and strongly adsorbed H2SO4 [24-26].

238

T.J. Bandosz

3.1. Role of carbon porosity It is well known that small pores, similar in size to the adsorbate molecule, cause overlapping of the adsorption potential resulting in enhancement of adsorption forces [40]. Since the diameter of sulfur dioxide molecule is estimated to be about 0.43 nm [35] the most favorable pores for its adsorption are those with widths smaller than 0.8 nm. The effect of pore width on micropore filling mechanism of SOj in carbon micropores was studied by Wang and Kaneko [35], Based on the comparison of the adsorption study on four carbon samples with significant differences in porosity (pore sizes equal to 0.79, 1.01, 1.13 and 1.45 nm for P10, P15, P2Q, and P25, respectively) and taking into account cases either with or without dipole-induced dipole interactions they found that dipole-induced dipole interaction contributes significantly for stabilization of SO2 molecules at the potential minimum and the contribution becomes larger with an increase in the pore width. This is due to a decrease in the dispersive forces. These interactions affect the pore filling. In small pores the dipoles should be oriented with each other but in larger pores the overlapping effect of the molecule-pore wall interactions in the central position of the micropores is not enough for SO2 to fill the pores and the molecules are adsorbed only on the pore walls. The interaction potential profiles calculated for SO2 adsorption in micropores of carbons with various pore diameters (pore size increase from P10 to P25) are presented in Fig.7. The induced image potential method refers here to the polar interactions of SO2 molecule with graphitic slit pore [35], The nonpolar interactions were calculated using 10-4-3 potential proposed by Steele [41]. Table 1 Comparison of calculated and experimental adsorption energies of SO2M1 a micropore. Reprinted with permission form Ref. [35], Copyright (1998) American Chemical Society sample P10 P15 P20 P25 (P5) Wm [nm] (0.75) 0.79 1.01 1.13 1.45 (37.8) 35.1 29.4 28.6 27.5 qd [kJ mol"1] (37 ±2 ) 33 ±2 36 ±2 29 ± 2 30 ± 2 qa, ^=0.4 [kJ mol 1 ] The differential adsorption energies calculated from the total interaction energy, qa, in comparison with the experimental energies at the fractional filling § = 0.4 are listed in Table 1. The 40 kJ mol"1) is linked to the interactions of SO2 with surface functional groups. These interactions are addressed in section 2.3 of this chapter. Since the molecular size of SO2 is around 0.43 ran (LJ parameter a s= 0.429 ran) the presence of pores smaller than 0.8 nm in the structure of carbon should be crucial for physical adsorption of this molecule. Indeed the importance of such pores in the process of SO2 removal was pointed out in the literature. Raymundo-Pinero and coworkers studied the dependence of the amount adsorbed on various carbons on the porosity measured using Dubinin-Radushkevich method, CO2 adsorption and the total pore volume calculated from nitrogen adsorption [26]. The results obtained show the relatively good correlation for the volume of micropores calculated form carbon dioxide adsorption. It has to be pointed out here that it is believed that CO2 at experimental conditions chosen in that research adsorbs only in pores smaller than 0.7 nm. Such correlation is found only when oxygen is present in the system. Lack of oxygen decreases the amount adsorbed by a factor of two to six depending on the type of carbon. The evidence on adsorption of sulfur dioxide in micropores in the absence of oxygen was found by Molina-Sabio and coworkers [27]. While calculating the micropore volumes of various carbons using CO2, N2, and SO2, a relatively good agreement in the values was obtained. A small discrepancy found in the case of SO2 was explained by the polarity effect. The strong adsorptive -adsorptive interaction in the gas phase caused weaker adsorbent-adsorbate interaction than in the case of N2 and CO2. SGQ

oWelfAir) 400

-

300 *" • 9

200 B

a

100

•

. , - •-

•

. . - - - •••" *

L

0 o

0.0$

0,1

0.15

•

A

0.2

Selected! incremental irtcropore volume,

Fig. 9. Dependence of SO2 adsorption capacity on the pore volume between 0.679 nm and 0.858 nm (calculated from DFT). Reprinted with permission from Ref. [34]. Copyright (2005) American Chemical Society

Desulfurization Desulfurization on onActivated Activated Carbons Carbons

241 241

A significant effect of very small micropores on SO2 adsorption was also noticed by Bagreev and coworkers [34]. Fig, 9 shows the dependence of SO2 adsorption capacity on the volume of pores with widths between 0.679 and 0.858 nm. Two slopes distinguished in this figure suggest two different steps/mechanisms of adsorption. It was further concluded that the adsorption capacity is governed by two surface features: porosity and surface chemistry. Their contributions are impossible to separate and the combined effect is addressed in section 3.3. A well pronounced effect of the volume of pores on the capacity for SO2 removal was noticed for carbonaceous adsorbents derived from sewage sludge [37]. In spite of the low surface area, the capacity of such adsorbents is significant and when experiments were run at dry conditions where catalytic function of the adsorbent was not fully activated, good correlation between the capacity and total pore volume was found.

- • ] •

•

— • — ACF-10 — * — ACF-15 — • — ACF-35

.... I

i .... . ..

c ram)

swo 3000 Concentration {ppmvj

Fig. 10. SOa adsorption from flue gas onto ACFs [28]. M.A. Daley, C.L. Mangun, J.A. DeBarr, S. Riha, A,A, Lizzio, G.L. Donnals, and J. Economy, Carbon, 35 (1997) Daley and coworkers found that the SO2 adsorption capacity on the activated carbon fibers was inversely proportional to pore size, pore volume and pore size distribution [28], The kinetics of adsorption on three different activated carbon fibers with increasing pore sizes are presented in Fig. 10. ACF10 which has the smallest average pore size adsorbs more than the ACF15 and ACF25 at shorter times. The equilibrium adsorption isotherms on those activated carbon fibers indicated that only pore size affected the amount adsorbed and the pore volume effects were little.

242

T.J. Bandosz

3.2. Effect of oxygen and water The uptake of SO2 on activated carbon is related to the degree of its conversion to SO3, which requires the presence of oxygen. It was found that oxidation of sulfur dioxide to sulfur trioxide occurs mainly in the 0.7 nm pores [26, 27]. With an increase in the size of pores in the carbon adsorbents less SO2 is converted, which results in smaller uptake of sulfur dioxide. The controversial results are reported regarding the effect of pore volume and SO2 uptake in the presence of oxygen. While Bagreev and coworkers reported the importance of porosity [34], Raymundo-Pinero and coworkers could not find any correlation between volume of micropores and the amount of SO2 adsorbed in the presence of oxygen [26].

200

AB&

£00

eoo

1000

Temperature; fG]

Fig.ll. Changes in the amount of thermally desorbed products for the surface of SCN-4 samples run at various conditions (D-dry in the presence of oxygen; N-no moisture, no oxygen; conditions for SCN-4: 3,000 ppm SO2, 80% humidity, air, room temperature). Reprinted with permission from Ref. [34]. Copyright (2002) American Chemical Society Due to the fact that sulfur dioxide is oxidized in carbon pores, there is a general agreement that it is adsorbed on the surface with two adsorption energies [18-26]. The low energy, about 50 kJ/mol, corresponds to weak physical adsorption, and the second, about 80 kJ/mol, to chemisorption [18, 34] of SO3. The latter is not formed when oxygen is absent. Those two adsorption energies are revealed as two desorption peaks on DTG [18, 34] and TPD curves [26]. The first one at about 373 K represent removal of weakly adsorbed SO2 and the second at about 523 K is related to desorption of strongly adsorbed species. Those strongly adsorbed species are likely physically adsorbed SO3 or SO3 from decomposition of sulfuric acid (Fig. 11)

Desulfurization on Activated Carbons

243

Formation of sulfuric acid during SOa adsorption and its extractability by water-washing was investigated by Lisovskii and coworkers [20]. It was found that adding water to the gas mixture, even in the absence of oxygen, results in a two fold increase in the amount of SO2 adsorbed. This was linked it to the formation of sulfuric acid as a result of SO2 oxidation by catalytic action of oxygen containing groups. The effect of water on the formation of strongly bonded species on the surface of carbon is also seen in Fig. 11 where in the presence of oxygen and water the amount of SO2 retained increased 3 ^ times compared to the process run in anaerobic and dry conditions [34]. The negative role of oxygen, if any, in the amount of SO2 adsorbed is linked to its ability to react with carbonaceous matrix and the formation of surface groups, which decrease the surface area of adsorbent. 3.3. Role of surface chemistry Although effects of porosity are crucial for physical adsorption, the importance of the catalytic effects of surface chemistry increases when weak adsorption forces exist. In the case of acidic gases such as SO2, the positive effect on adsorption should be observed when the basicity of the surface increases. Numerous researches found that heat treatment of activated carbons or activated carbon fibers at temperatures about 1300 K results in an increase in the amount of sulfur dioxide adsorbed [21, 23, 28, 30]. Such treatment, besides removal of oxygen-containing acidic groups, should increase carbon basicity [11]. It was specifically found that when basic groups containing oxygen are present on the carbon surface the adsorption of SO2 is significantly enhanced [34]. In such a case basic groups (pyronic and pyronic-like type) are responsible for strong physical adsorption of sulfur dioxide. These acid/base interactions do not introduce any catalytic effect leading to the formation of H2SO4 and its chemisorption on the surface. A strong adsorption of sulfur dioxide is enhanced by the presence of oxygen [23,24, 26]. These oxygen-containing sites are proposed to act as catalytic centers for oxidation of SO2 to SO3 [21, 23]. According to Davini, oxygen present in the system plays an important role in the variations of SO2 adsorbed (Fig. 12) [24]. Carbon oxidized at low temperature with high content of oxygen groups tent to adsorb less SOa since their surface contains less basic groups. On the other hand, carbons oxidized at high temperature, in spite of the fact that their oxygen content is smaller, adsorb more SO2 as a result of the basic character of their surfaces. A decrease in the SO2 uptake upon the presence of oxygen-containing acidic groups was noticed by Daley and coworkers [28]. They found a correlation between an increased SO2 capacity and the amount of CO/CO2 evolved during heat treatment of carbon fiber surfaces, which led to the formation of new active centers. Similar correlation was investigated by Mochida and coworkers [22]. On the other hand, Daley and coworkers found that when dry SO2 was adsorbed, the presence of oxygen containing functional groups significantly enhanced the performance bellow temperatures 348 K [28]. That enhancement was explained by surface reactions of quinines with SO2 and water forming the diol and sulfuric acid. The effect of surface chemistry on the SO2 oxidation step was also discussed in detail by Raymundo-Pinero and coworkers [26], However, contrary to Daley and coworkers, they did not find any correlation between the amount of groups decomposed during heat treatment and an increase in the SO2

244

T.J. Bandosz

adsorbed, they confirmed that removal of oxygen form the surface forms new high energy adsorption/oxidation centers.

Fig. 12. SO2 (a) and on SO2 (b) adsorbed on carbon samples (group I and 11) versus % oxygen of each carbon sample (fixed bed data). Groups I low temp oxidation; groups II- high temperature oxidation; aphysically bound, b- chemically bound [23]. Reprinted with permission from P. Davini. Carbon, 28 (1990) 565

An increase in the uptake of SO2 upon oxidation of carbon was found by Lisovskii and coworkers [24]. As mentioned above, they concluded that these surface acidic groups are catalysts for SO2 oxidation. Moreover, the presence of strong basic functionality was suggested as not beneficial for the process of sulfur dioxide removal due to an increase in the retention of sulfuric acid, which is undesirable from the point of few of adsorbent regeneration [44]. High adsorption on chars containing high surface acidity and basicity was also noticed by Rubio and Izquierdo [30]. Based on the performance of their materials, they concluded that not only the amount of surface groups but also their accessibility have an effect on SO2 adsorption/oxidation. Basic groups were also found as a factor enhancing the capacity for sulfur dioxide removal on activated carbon fibers, especially those PAN-based [22, 32, 42]. Kawabuchi and coworkers noticed a significant increase in the sorption capacity when activated carbon fibers were modified with pyridine and basic nitrogen functionalities were introduced to the surface [42]. Pyridine provided basic functionality, which increased catalytic removal of SO*. Since introduction of nitrogen functionality to activated carbon surfaces is known as a way to improve its basicity, the effect of such treatment on the performance of carbons as SO2 removal media was investigated in detail [34, 46]. It was found that nitrogen containing pyridinic species, which are placed at the edges of graphene layers, noticeably increase the amount of SO2 adsorbed and its catalytic conversion to sulfuric acid (Fig. 13). The effect is even more pronounced when those groups are present in small pores [34]. The only negative part related to

245

Desulfurization on Activated Carbons

the application of these materials is strong adsorption of sulfuric acid leading to the difficulty in adsorbent regeneration [44]. Fig. 14 shows that the presence of nitrogen-containing groups in BX-U and Centaur® carbon results in significant amount of SO2 adsorbed as H2SO4, which is shown on DTG curves as a weight loss peak between 473 and 673 K.

• A20 0.1

• KLA1

- ^

* Aox

I

s^^

0.1S ^ * ^

A.

0.1

0.0$

0

(1.5

3.5

Fig.13. SO2 to SO3 conversion percentage by surface area versus pyridine-like nitrogen percentage [46], Reprinted with permission from E. Raymundo-Pinero, D. Cazorla-Amoros, A. Linares-Solano, Carbon, 41 (2003)1925

Taking into account that complete separation of the roles of porosity and chemistry seems to be impossible, the dependence of the specific capacity on the incremental amount of basic groups was investigated. This incremental amount was calculated as the product of the average density of basic groups and the incremental surface area in pores between 0.679 and 0.858 nm. As seen from Fig. 15, a relatively good linear trend exists for the majority of the samples studied. This implies that either volume of pores having sizes around 0.7-0.8 nm in diameter or basicity of the samples leads to the similar final effect, which is the enhanced SO2 adsorption capacity. This happens for the samples that may have a combination of both these factors provided, that sufficient pore volume exists for the "storage" of sulfuric acid (two slopes in Fig. 10). As in the case of hydrogen sulfide, the presence of ash and its composition should have an effect on the amount of SO2 retained on the surface. This effect was observed by Lu and Do [36] studying the SO2 adsorption on activated coal rejected char. Its high content of inorganic matter/ inorganic oxides was expected to affect the amount of SO2 oxidized to sulfuric acid. Titanium oxide was identified as the most active ingredient. The enhancement in the oxidation of SO2 due to the presence of active inorganic matter was also found by Bashkova and coworkers on carbonaceous adsorbents derived from sewage sludge [37]. In those materials, a high content of CaO was identified as a favorable factor. The effect of calcium was also studied when fly ash mixtures with calcium hydroxide were tested as SO2 adsorbents [38]. It was found that Ca(OH)2 enhances the dispersion of calcium reagent and thus improves the efficiency of the adsorbent.

246

T.J. Bandosz

I

0.25 •

BUC BMMruiia! errtaur

n

A

;

a-is

1 •\

a i • SI' • V ''

f M

0.G5-

iv 5 \

3Bgj h i r

0"

0

ZDD

4QQ

600

SQO

1000

Temperatiiie, fCJ Fig. 14. DTG curves in nitrogen for BAX and Centaur®. Reprinted with permission from Ref. [34]. Copyright (2002) American Chemical Society S.0

5.0

0

[ »

* 9 10

1.0

a

d'

c

9(L00

O.10

0.20

Q.30

O.ACt

Inoementstl amount Of baste groups, [mmtf/gl Fig. 15. Dependence of SO2 adsorption on the incremental basicity in pores between 0.679 and 0.858 nm. Dotted line shows the trend in the data. Reprinted with permission from Ref. [34]. Copyright (2002) American Chemical Society

247

Desulfurization on Activated Carbons

3.4. Mechanism of SO2 interactions with the carbon surface

o H

H

H

o

H

A

/

Chcmisorpiion ^

ads

40.5 kcal.'mol

x$&: xwx H

Ehyd=-14.4kcaltaol

H

H

H

A2

Al

\

Hydiatioo Ha

H

H

H

>3H

H

A4 E o x = -49.9 kcal/md

^ -

H

H

H

A5 Fig. 16. Reaction pathways for carbon-catalyzed SO2 oxidation by Q2/H2O to form sulfuric acid [45]. Reprinted with permission from F.H. Yang and R.T. Yang, Carbon, 41 (2003) 2149

All the empirical results point out the importance of surface basicity and pore sizes in the process of sulfur dioxide removal with oxidation to SO3 followed by formation of sulfuric acid as a favorable factor. Based on these, ab ibido molecular orbital calculations were performed on the possible pathways of SO2 oxidation on activated carbon in the presence of water and oxygen. Yang and Yang found that when surface oxides are present on the zigzag edge sites sulfuric acid is formed with sulfurous acid as an intermediate (Fig. 16) [45]. On the other hand, chemisorption was found unfavorable on the edge sites containing twin oxides. Those findings can help to

248

T.J. Bandosz

explain some discrepances described above regarding the role of surface oxygen groups in SO2 adsorption/oxidation on activated carbons. When oxygen is present in the form of surface functional groups, the most favorable mechanism is chemisorption of SO2 with formation of SO3 as intermediate followed by continuation along two possible reaction pathways to form the sulfuric acid intermediate (Fig.17).

Chcinisorpliun

E o x = -«f>.0 kcal/mo!

Hydration \ E hj .j = -9.8 Kcat/mol

Fig.17. Reaction pathways for carbon-catalyzed SO2 oxidation by O2/H2O to form sulfuric acid on graphite sites with oxide [45]. Reprinted with permission from F.H. Yang and R.T. Yang, Carbon, 41 (2003) 2149

249

Desulfurization on Activated Carbons

4. REMOVAL OF HYDROGEN SULFIDE 4.1. Air odor control 4.1.1. Activated carbon based odor control systems Odor problems associated with waste water treatment plants are common even around well operated facilities [47]. The odors are generated usually by some forms of anaerobic bacterial activity which produces and releases gases into the ambient environment. The main odor components are hydrogen sulfide, ammonia, and volatile organic compounds containing sulfur or nitrogen such as mereaptans, indoles, scatoles, etc. In spite of the variety of species present, a leading malodorant arising from sewage treatment facilities and geothermal vents is hydrogen sulfide. To remove hydrogen sulfide odor various measures are available. Among them are addition of chemicals to waste water, adjustment and alteration of the sewage flow regime and enclosure and subsequent treatment of the odorous headspace air either by catalytic oxidation, combustion, wet scrubbing, dilution with clean air or activated carbon adsorption. The H2S concentration in influent sewage air varies from 0.1 ppm to 28 ppm with typical value of about 5.3 ppm. The odor concentration varies from 34 SOUm"3 to 250 SOU'm"3 with an average of 120 SOUm"3 [47]. Traditionally, activated carbons used for removal of high concentrations of H2S in sewage treatment plants are those impregnated with caustic materials such as NaOH or KOH [48-56]. Air currents around odor generating facilities are initially washed in scrubbers, during which they intake high levels of humidity, and are then blown through the activated carbon vessels [54, 55]. The residual H2S quickly reacts with the strong base and is immobilized. The presence of humidity in air facilitates the reaction [53, 56, 57]. The schematic view of the typical activated vessel installed in water treatment plant is presented in Fig. 18. New York City Department of Environmental Protection own over 100 such vessels in its 11 plants. Each vessel contains about 11 tons of activated carbon. Clean air

Activated ' carbon (dual bed).

/ * •

Fan

\

Fig. 18. Typical activated carbon vessel used to polish air in water treatment plants

250

T.J. Bandosz

4.1.2. Caustic activated carbons as hydrogen sulfide odor removal media Field studies indicate caustic impregnated carbons as efficient media to remove hydrogen sulfide odor in water treatment plants [47,49, 50, 52]. The carbon bed is mostly used as a support for the caustic material and storage of the oxidation products. The removal capacity of such carbon estimated using accelerated ASTM D6646-01 test [58] exceeds 0.14 g'cm"3 of carbon bed. Recent study of the adsorption/oxidation mechanism on sodium hydroxide impregnated activated carbons showed [57] that at least 3 moles of H2S are adsorbed per one mole of NaOH. This indicates the catalytic effect of sodium hydroxide. NaOH shifts the dissociation of hydrogen sulfide to the right increasing the content of HS" ions, which can be further oxidized either on adsorbed sulfur [59] or on the activated carbon surface [60]. The reaction proceeds until all NaOH is consumed in the surface reaction and deposited in the form of salts, either sulfites or carbonates, and the regeneration of basic environment does no longer occur. The catalytic action of NaOH impregnated carbon can be summarized by the following reactions [57]: NaOH + H2S —• NaHS + H2O 2 NaOH + H2S -*Na2S + H2O NaHS+ l/2 0 2 ^ S + Na0H Na2S + I/2O2 + H2O -* S + 2 NaOH HS" + H2O -» H2S + OH" S2" + H2O -» HS" + OH" 2 NaOH + H2SO4 -*Na2SO4 + 2H2O

(12) (13) (14) (15) (16) (17) (18)

The shortcoming in the applications of caustic impregnated activated carbon is the fact that impregnation decreases the ignition temperature of the carbon and poses a hazard of selfignition [55, 61]. Another disadvantage is the oxidation of hydrogen sulfide to elemental sulfur [55, 57, 61], which cannot be removed from carbons by washing with water [52], Moreover, the activity of caustic carbons toward H2S oxidation is exhausted when the caustic is consumed and the carbon pores are blocked by sulfur and sodium or potassium salts. Both advantages and disadvantages of caustic impregnated carbons directed the attention of researchers toward other impregnants, which can sustain basic properties with less exothermic reaction in the system. An example is potassium carbonate, which was studied in details by Przepiorski and coworkers [62-64]. According to them, hydrogen sulfide dissolves more favorably in aqueous solution of K2CO3 than in water. H2S, due to its small size, is able to access the small micropores as KHS (also KHCO3 is formed), which rapidly converts to H 2 S. The H2S located in small pores reacts with oxygen forming elemental sulfur. Important for surface catalysis is neutralization of KHCO3 to K2CO3. Since oxidation of hydrogen sulfide to sulfur either in direct reaction or via dissociation to HS" and its oxidation releases significant heat, the risk of bed self-ignition still exists.

Desulfurization on Activated Carbons

251

4.2. Carbon surface as a catalyst of hydrogen sulfide oxidation This risk of self-ignition of the carbon bed along with hazardous conditions of working with high pH carbons caused virgin (unimpregnated) activated carbons [47, 59, 60, 65-98] or carbon with specific surface modifications, such as nitrogen functionality [98-102], to be investigated as H2S removal media. However, considerable removal capacities for hydrogen sulfide have been reported in the literature for carbons serving at temperatures around 473 K, the use of unmodified activated carbon for H2S removal at the ambient temperatures [60, 86-91], is not yet common. This might be related to a relatively low capacity of virgin carbon compared to caustic impregnated one, which for the best materials, coconut based carbons, is seven times smaller than that on the impregnated counterparts [61]. Moreover, the mechanism on unimpregnated carbons seems to be complex and very detailed features of carbon surfaces play a role in adsorption and catalytic oxidation. For this reason that most of the results reported so far have been based on an empirical analysis of specific types of carbon, which are sometimes difficult to reproduce [49104]. A simple mechanism of adsorption/oxidation of hydrogen sulfide was first proposed by Hedden and coworkers [77], According to them, dissociation of hydrogen sulfide occurs in the film of adsorbed water at the virgin carbon surface and then hydrogen sulfide ions, HS", are oxidized by oxygen radicals to elemental sulfur. Since then many studies have been carried out to account for such factors as a role of water [73, 78, 80, 82, 83, 86, 92, 95] role of oxygen [59, 65-73, 84], autocatalysis by sulfur [59, 74], influence of pore sizes [66, 75, 79, 81, 84], role of carbon surface chemistry [87-90], the effects of inorganic matter [95, 96, 108—111] and last but not least, speciation of surface oxidation products [87-92]. 4.2.1. Role of porosity In all studies of hydrogen sulfide adsorption the presence of micropores is indicated as an important factor. Although opinions about the first location of adsorbed sulfur vary [65, 66], the filling of micropores by elemental sulfur or sulfates seems to be the limiting factor for the activated carbon capacity [87, 91, 94]. Steijn and Mars found that the strong sulfur adsorption is in carbons having pores between 0.5—1 nm [66], which is expected based on the size of sulfur chains and the overlapping of adsorption potential in pores similar in size to the adsorbate molecule. Moreover, when sulfur is adsorbed in such small pores the presence of large polymers is unlikely, and isolated adsorbed sulfur radicals are further oxidized to SO2 and then SO3. On such carbons, sulfuric acid is the important product of surface reaction [87], It was also found that when the H2S capacity of carbons is normalized to their pore volume, the similar values are obtained [87,92]. Although direct relationship between the porosity of carbons and their hydrogen sulfide adsorption capacity was not established, the pore sizes should play a role in energetics of physical adsorption. Analysis of the isosteric heats of adsorption, Qst, obtained in dry anaerobic conditions indicates that they do not depend on surface chemistry in the studied temperature range [93]. This finding excludes the oxygen from surface functional groups as the contributor to oxidation of hydrogen sulfide. The heat of H2S adsorption in the case of all unmodified carbons is slightly higher than 40 kJ mol"1. The value reported by Doleva and coworkers [112] on carbon black is

252

T.J. Bandosz

20.6 kJ mol"1. Taking into account the enhancement in the energy of adsorption owing to the interaction of the H2S molecules with the two pore walls, the highest value of the heat should be about twice of that obtained for a nonporous surface [40]. Indeed, in the Qst found by Bagreev and coworkers [93] is 1.9 to 2.3 times the value reported by Doleva and coworkers [112], probably the result of pores different in shape than slits. Since physical adsorption in very small pores is the most likely mechanism of the adsorption process in dry conditions, the dependence of Qst on the relative microporosity expressed as the ratio of the volume in micropores to total pore volume (DFT) was analyzed. Although the data scatter, the correlation was found with R equal to 0.93 indicating that micropores are very important for H2S adsorption [93]. They are the high-energy centers, and oxidation of hydrogen sulfide should be enhanced when the structure of carbons is homogeneous and the volume of micropores is high. Hydrogen sulfide, when immobilized in small pores, is likely oxidized to sulfur radicals and than to SO2 and SO3 resulting in high yields of sulfuric acid, which is important from the point of view of carbon regeneration by water washing [44,91, 92]. In the case of nitrogen-enriched carbons, the heat of hydrogen sulfide adsorption is significantly higher than that on their unmodified counterparts [104]. Reaching 50 kJ mol 4 indicate the contribution of nitrogen containing centers to the adsorption process. They catalyze oxidation of H2S to SO2 and their effect is discussed in section 4.2.3. 4.2.2.Role of water and oxygen As indicated in numerous experimental works, water is a very important factor in the process of hydrogen sulfide removal [78, 83, 86]. In the mechanism of the adsorption/oxidation process proposed by Hedden and coworkers, H2S to be oxidized by active oxygen has to be first dissociated to hydrogen sulfide ions [77]. It is well known that the activated carbon surface is hydrophobic in its nature [10, 11, 113]. Its low affinity to water is the result of a high degree of aromatization and the presence of graphene-like sheets. Adsorption of water can be enhanced when functional groups containing oxygen exist at the edges of graphene-like sheets [11,114]. To evaluate the effect of water the H2S breakthrough capacity was measured at both wet (80% humidity) and dry conditions [82, 84, 88, 89]. The capacity at dry conditions is usually very small and it represents physical adsorption in the small pores of carbons. It is interesting that to achieve high capacity on some carbons, not only moisture is necessary in the challenging air but also water should be adsorbed on the carbon surface in sufficient quantity. The breakthrough experiments (accelerated tests) were carried out on as-received and prehumidified carbons of various origins and from various suppliers [95]. The results showed that the H2S adsorption capacities of prehumidified carbons are about two to six times higher than those of the asreceived carbons. Since analyzing the dependence between the amount of sulfur deposited on the surface and the amount of water adsorbed the increasing trend was revealed until 60 % humidity [82, 84] (Fig. 19), the dependence of the normalized capacity (per unit pore volume of carbon) on the amount of water adsorbed was studied [87]. The relationship obtained suggests that the affinity for water adsorption should not be greater than 5% to reach the maximum capacity. It is likely that, when the affinity of carbon to adsorb water is very high, the small pores are filled by condensed adsorbate and the direct contact of HS" with carbon surface in the smallest pores is

253

Desulfurization on Activated Carbons

limited. The other factor that plays a role is the degree of carbon oxidation. When more oxygen groups are present the surface becomes more acidic suppressing dissociation of hydrogen sulfide [89]. ,

.

.

450

.

?

cut

...3,:...;.^.....:

450

• — * - C

|fT.,

^^pa

~]

.__._.

I \ '—j—-'•'•

\

:

C ^Sg/te 4 * m = 2Qni4__ M

: - .

... _

.

.

i

20

.

,

.

i

,

,

,

i

.

.

• i

40 60 80 Relative Humidily (%}

.

.

.

? aoo S

: 250 200

100

Fig. 19. H2S adsorption on ACC as a function of relative humidity level (qe represents the amount of loaded gas). Reprinted with permission from Ref. [84]. Copyright (2003) American Chemical Society The majority of research results confirm the importance of water for H2S adsorption/oxidation. Only Coskun and Tollefson [65] found that the presence of water at temperatures close to ambient decreases the catalytic activity of carbon surfaces. 4.2.3. Role of surface chemistry In a general application of activated carbons for adsorption processes surface chemistry is rarely taken into account and it is even not considered as one of the required specifications. When adsorption of H2S on activated carbons is studied the role of species present on the carbon surface is seldom addressed. Discussing this subject seems to be natural especially in combination with the role of water as a medium for dissociation of hydrogen sulfide. It is well known that the degree of acid dissociation depends on the pH of the system, and dissociation is feasible when pH is greater than pKa of an acid under study. Hydrogen sulfide is a weak diprotic acid with first and second pKa equal to 7.2 and 13.9, respectively. These values can be different in the small pores of activated carbon where the effect of the enhanced adsorption potential is very strong [115]. Analysis of the performance of carbons shows the dependence of the capacity on the acidity of carbon [87, 94]. This was demonstrated using number of carboxylic groups, basic groups, CO/CO2 evolved during TPD experiments, amount of water adsorbed at low relative pressure and surface pH as variables describing the degree of surface acidity [87]. The analyses of the plots suggest the existence of a threshold in the surface acidity. When the number of carboxylic groups crosses the range between 0.15 and 0.3 mmol g"1 a significant difference in the normalized capacity is observed. Based on the above findings and the detailed characteristics of the samples, the surface chemistry seems to be the factor, which plays a dominant role in the H2S

254

T.J. Bandosz

breakthrough performance of carbons [87-95]. The local pH in the pore system has a significant effect on the efficiency of hydrogen sulfide dissociation and its oxidation to various sulfur species. A moderately low average pH of the carbon surface is expected to suppress the dissociation of H2S and the creation of hydrogen sulfide ions. Those ions, when present in low concentration in small pores, are oxidized to sulfur oxides from which sulfuric acid is formed. On the other hand, a pH in the basic range promotes the dissociation of H2S. This results in a high concentration of HS" ions, which are then oxidized to sulfur polymers having chain or ring-like shapes. When the pH value is very low only physical adsorption can occur. This raises a question about the practical limits of acidity of the carbon adsorbent surface since this can lead to different products of surface reaction. The dependence of the normalized capacity on the pH of the carbon surface is presented in Fig. 20. The threshold value derived from the analysis of the data occurs at a pH around 4.5. When the pH is lower than 4.5 only physical adsorption can occur, and the concentration of dissociated hydrogen sulfide ions is negligible. The justification for the threshold in surface pH is based on the steps of hydrogen sulfide adsorption/oxidation on unmodified carbons [87, 94]. They are as follows: (1) H2S adsorption on the carbon surface, (2) its dissolution in a water film, (3) dissociation of H2S in an adsorbed state in the water film, (4) and surface reaction with adsorbed oxygen.

H2Sgas

-

H2Sads

(19) (20)

Ka TT

o

W o jjK/jf

,

j»

i O

^^

TJG"

,

- I - TJ"1"

* Si ads ' v l x l

HS a*& +3 O*a^ ~3>SO2 ads ~*~ OH H + + OH" -* H 2 O

f™i 1 ^

(^zZSlJ

(22b) (23)

where H2Sgas, H2S a 250 250

E

O I

48

450

C 300

g

to

100

250

.

I-

f

200 150

100 100

0

6

12

18

24

30

36

f

42

0 0

48

•

{ J

0

6

•

—

•

-D i

50

Time, months

—

H2S VOC VOC H2O H2O Total

\A

50

0

.

12

18

24

,.4-4-

30

36

42

48

Time, months

Fig. 28. Dependence of H2S (A), VOC (B) and H2O adsorption (C) on time and the dependence of the average amount adsorbed (total) and that of each component on time (D). Experimental data is represented as points and calculated as lines. Reprinted with permission from Ref. [126] In general, Eq. (41) may be used to predict the breakthrough time for any inlet concentration if a kinetics coefficient and a capacity of adsorbent are known. Neglecting the first term in Eq. (41) due to its small value in comparison with the second one, and taking logarithm of both side of Eq. (41) one gets: log (tb) = log ((xL - In (Co/Cb - 1)) / k) - log (Co)

(43)

Stabilizing value C0/Cb at some certain level, for example Q/Cb = 100, and taking into account that for one component adsorption XL does not depend on Co and time, the linear equation can be obtained: = Ki-K2log(C0), where Ki = log ((xL - In (CyC b - 1)) / k) and K2 = 1

(44)

111 271

Desulfurization on Activated Carbons

Plotting the experimental data for breakthrough time versus the initial concentration (Fig. 29) and fitting these data using Eq. (44) coefficients Ki and Ka can be found. Then the breakthrough time for a low concentration can be easily predicted by extrapolation of experimental data to "ppb" level or from a direct calculation of tb using Eq. (43) for the particular concentration. In the case of multicomponent adsorption, XL value depends on the kinetics of VOCs adsorption. If it is taken into account, the nonlinear dependence of the breakthrough time versus concentration is observed in the range of H2S concentration where competitive adsorption takes place. The results of these calculations are collected in Fig. 29 in comparison with experimental data and "working" line determined based on the micropore volume only. This volume limits the amount adsorbed. The best result is obtained for the model with multicomponent adsorption [126]. 10000000

N

• DM DM (H2S) (H2S) -DM(H2S+VOC) DM(H2S+VOC) Vmic —Vmic o Lab Lab A NR NR

Breakthrough time, h

1000000

•=

e

100000 100000

10000

A

s s.

A

N

\ 1000

V

N

100

N

10

1 0.01

0.1

11

10

100

1000 1000

10000

ppm H 2S concentration, ppm

Fig 29. Dependence of the breakthrough time on inlet concentration of H2S. Thin solid line and dashed line correspond to breakthrough time predicted by dynamics models (DM) for multicomponent (H2S + VOC) and one component (H2S) adsorption. Thick solid line predicts the amount adsorbed based on the micropore volume. Points represent the experimental data from the dynamics tests at high H2S concentrations (Lab) and field conditions (NR). Reprinted with permission from Ref. [126]

A similar approach was used for prediction of the hydrogen sulfide removal capacity of activated carbons used as desulfurization media from digester gas [125]. The breakthrough time at 50 and 100 ppm was predicted based on the measurements at 1000, 2000 and 5000 ppm. 5. REMOVAL OF MERCAPTANS ON ACTIVATED CARBONS Although mercaptans present in air or gaseous fuel can cause similar environmental and catalyst poisoning effects as hydrogen sulfide, their adsorption on activated carbons has been addressed less frequently in the scientific literature [48, 78, 132-139]. Very often their adsorption is studied

272

T.J. Bandosz

along with that of hydrogen sulflde. Mercaptans, besides their natural occurrence in the gaseous fossil fuel are also added to the gas on purpose in order to increase the leak detection. This is linked to their specific rotten cabbage odor and its low threshold. The main difference between hydrogen sulfide and methyl mercaptan (MM) is the presence of hydrocarbon moiety in the case of the latter compound. It causes that MM molecule is much stronger adsorbed by physical forces on the surface of activated carbons than hydrogen sulfide. The likely surface oxidation product is dimethyldisulfide not elemental sulfur or sulfur dioxide. 5.1. Role of carbon surface features An extensive study of the effects of the carbon surface features on adsorption-oxidation of methyl mercaptan at ambient temperature was performed by Bashkova and coworkers [135-139]. To evaluate the effects of porosity, surface area and surface chemistry, activated carbons with a broad range of pore sizes were chosen (from very microporous to mesoporous) and with various surface chemical heterogeneity (oxygen or nitrogen containing surface groups). Moreover, the effect of inorganic impurities such as iron content in ash was also evaluated [139]. Table 5 pH of the surface, amount of water preadsorbed, breakthrough time and CH3SH breakthrough capacities for initial and oxidized commercial carbons (bed of 1 cm x 6 cm, 1-2 mm granules carbons, relative humidity 80 % at 298 K, flow rate 0.5 L/min; capacity arbitrary calculated at 50 ppm). Reprinted with permission from Ref. [136]. Copyright (2002) American Chemical Society Water pHE Brth. Time CH3SH pH CH3SH Sample capacity preadsorbed desorbed 1 [min] [mg g"1] [mg g"1] [mg g" ] 6.78 13 163.4 7.20 0.48 BAX 28.2 14 4.07 3.30 168.8 3.56 BAX-0 23.0 3.82 7.41 BPL 155 89.9 0.03 216.8 79 3.92 99.6 6.60 0.05 BPL-0 104.3 5.94 162.2 143 7.47 92.6 0.04 S208 203.2 184 5.75 98.1 0.04 S208-O 8.13 59 5.34 7.57 78.2 PCB 68.2 0.11 65 4.57 94.5 6.35 0.23 PCB-0 72.9 Analysis of the surface oxidation products using either TA or GC/MS indicated that the predominant product of surface oxidation is DMDS. In the case of the coconut shell based carbon with very small pores present, S208, the traces of methyl methane thiosulfonate (C2H6O2S2) were detected. It can be the product of the reaction of DMD with oxygen: C2H6S2 + 2 O * ^ C2H6O2S2

(45)

273

Desulfurization on Activated Carbons

or the disproportionation of sulfonic acid: 3CH3SO2H - • C2H6O2S2 + CH3SO3H + H2O

(46)

Support for the presence of these oxidation products was only less than 3 pH unit decrease in the pH, which excluded the formation of sulfonic acid [136]. The overall effect of oxidation on the capacity was mixed and it was believed that the observed changes were the results of the combined influence of the surface chemistry and porosity. More acidic was the surface, less strong was the sorption of MM, which was reflected in the amount of desorbed MM by air purging after the adsorption process. 5.1.1. Nitrogen enriched carbons; the effect of surface basicity New light on the importance of the specific surface features on the feasibility of removal of methyl mereaptan on activation carbon was thrown after study of nitrogen-enriched carbons as media for methyl mereaptan removal [138,140]. The capacity on carbons modified with urea was much higher than that on the as received adsorbents (Table 6) and even ten times enhancement can be obtained using this route of surface modification. The enhanced performance is attributed to the effect of basic nitrogen incorporated within the carbon matrix. Table 6 pH of the surface, amount of water preadsorbed, breakthrough time and CH3SH breakthrough capacities for the initial and nitrogen modified (at two different temperature expressed by the number in Celsius degree) commercial carbons. Reprinted with permission from Ref. [138]. Copyright (2003) American Chemical Society pH pHE Amount of water Bth.time CH3SH CH3SH Sample desorbed capacity [min] [mg g"1] [mg g'1] [mg g"1] BAX 163.4 28.2 0.48 7.20 6.78 13 1.82 BAXU-450 6.18 5.45 161.1 30.0 15 BAXU-950 7.43 4.88 0.03 145.4 299.0 180 BPL 0.03 7.41 3.82 89.9 216.8 155 0.04 7.84 5.98 129.4 321.1 232 BPLU-450 BPLU-950 0.02 8.46 3.67 102.0 440.6 303 0.04 S208 7.47 5.94 92.6 162.2 143 S208U-450 0.05 8.49 6.56 93.9 221.9 199 0.03 254 S208U-950 9.41 6.77 68.9 272.9 78.2 68.2 PCB 0.11 7.57 5.34 59 162 PCBU-450 0.09 8.81 7.26 81.7 203.5 PCBU-950 0.05 9.07 7.35 75.7 192.3 156 BAX 163.4 28.2 0.48 7.20 6.78 13

274

T.J. T.J.Bandosz Bandosz

In order to see the effect of nitrogen, the normalized capacity (with respect to the surface area of materials) was plotted versus the ratio of carbon to nitrogen content determined from elemental analysis. For this purpose only initial carbons and those modified with urea at 1223 K were taken into consideration to ensure that nitrogen is in similar chemical form as a result of high temperature heat treatment [104]. The results indicated the maximum in the capacity at C/N content about 0.02. It confirms the results obtained by Strelko and coworkers [141] who found using the quantum chemical calculation of model nitrogen-containing carbon clusters that at an atomic concentration of quaternary nitrogen within the carbon matrix between 2 and 3 % the minimum of the band gap occurs. The lowest width of the band gap, AE, indicates the highest catalytic activity in electron transfer reactions. Moreover, the results of quantum chemical calculations showed that carbons with quaternary nitrogen-containing groups at the edges of grapheme layers have the highest charge mobility in the carbon matrix and the best donoracceptor properties. The importance of basicity was also shown when the density of basic groups on the surface determined using Boehm titration method was linked to the MM removal capacity. The majority of carbons used for that research followed the linear trend. The discrepancy in some case was caused by the differences in the content of catalytically active phase, iron oxide, which will be addressed later in this section.

CH

3SHe

gas

*hf CH 3 S\

CHaS\

fin O\

liquid CHaSSCH3

Fig. 30. Schematic representation of catalytic adsorption/oxidation of MM on nitrogen-containing carbon, a) surface reactions; b) final products. Reprinted with permission from Ref. [138]. Copyright (2003) American Chemical Society On the basis of the systematic study [135—138] and the ability of CH3SH to dissociate, the following scheme of CH3SH adsorption/oxidation on the nitrogen containing carbons was proposed (Fig. 30): 1) methyl mercaptan is first adsorbed from the gas phase on the carbon surface, where due to the presence of water film, it is dissolved and, depending on the pH, it can dissociate with the formation of thiolate ions and protons; 2) since the positively charged quaternary nitrogen enhances the ion exchange properties of activated carbons, thiolate ions is adsorbed in the vicinity of the nitrogen center; 3) then nitrogen accept an electron from sulfur and transfer it to the oxygen adsorbed on the surface; 4) as a result, thiolate radicals and superoxide

275 275

Desulfurization Desulfurization on on Activated Carbons Carbons

ions Oa" can be formed with the latter triggering the formation of hydroxyl radicals; 5) the final step of the oxidation process is formation of DMDS and water. The reaction proceeds until all the pores with positively charged nitrogen centers, and other active centers of the carbon surface are filled with the reaction products, and then only physical adsorption of MM takes place. 5.1.2. Dual role of water Since after adsorption of MM the displacement of the preadsorbed water was observed as change in the intensity of DTG desorption peaks for water about 353 K and DMDS 473 K (Fig. 31), and following the assumption that either H2O or DMDS are adsorbed only in pores smaller than 5 nm, the data was normalized based on the volume of these pores. Fig. 32 shows the relationship between the normalized amount of DMDS and water. The correlation coefficient and slope are equal to 0.89 and 0.99, respectively. The slope represents the density of DMDS (1.06 g cm"3). The small discrepancy is likely related to the fact that not all pores are filled by oxidation products owing to the existence of some physical hindrances (blocked pore entrances). The thin line represents theoretical limit of adsorption assuming real density of DMDS and H2O. The fact that almost all points are located below this line validates hypothesis about the "active" pore volume [7-10]. All points used for this correlation represent equilibrium data. If equilibrium conditions, for instance for adsorption of water, are not fulfilled the amount of DMDS is usually small and the point "moves" from the established dependence line.

out pHS (HT

04

0.3

•H

pH10

0.1

100

200

300

4C0

SCO

Temperature [°C ]

Fig. 31. DTG curves after MM adsorption fro S208 carbons with various surface pH. Reprinted with permission from Ref. [137]. Copyright (2002) American Chemical Society The data presented in Fig. 31 suggests dual role of water in the process of methyl inercaptan adsorption on activated carbons. DMDS, which is the main product of surface reactions, has to compete with water for adsorption sites. However, the competition exists, DMDS is always a "winner" owing to its strong adsorption on carbons [137]. On the other hand, the formation of significant amount of DMDS would not be possible without the presence of

276 276

T.J. Bandosz T.J.

water in the system. Water facilitates dissociation of methyl mercaptan leading to its oxidation by oxygen, mainly from air. —

1.0 0.8

s

So

0,6

c y= 0.9942X-H Ml

o -CH3 > -Cl > -NO2 Very important effect when the adsorbing species is ionized (weak acid and basic species) Inorganic salts (CaCla, NaCl) can enhance adsorption of ionized species [52] or humic acids [53] Due to the exothermic character of adsorption, an increase in temperature generally results in a decrease in adsorption

Adsorption of Organic Compounds onto Activated Carbon -– Applications in Water and Air Treatments

391

1200 1200

Qm phenol (µmol g -1)

1000 1000 "Q o

800 800

Ig aoo 600 I % a

400 400 200

200 0

0

0

1

2

3

4

5

concentration of acidic groups (meq gg"-11)

Fig. 7, Influence of the concentration of acidic surface groups on the maximum adsorption capacity of phenol (reprinted with permission from [48], Copyright 1968 American Chemical Society). The specific surface area of the adsorbent is one of the principal characteristics affecting the adsorptive capacity of an activated carbon. This parameter ranges from 500 to 2500 m 2 g"1 for commercial activated carbons [44]. However, the specific surface area alone is generally inadequate to explain adsorption differences and must be correlated with pore size distributions. Micropores contribute to the major part of the specific surface area and many of them possess molecular dimensions that make them suitable for the adsorption of organic micropollutants [45] by the overlapping of adsorption forces [46]. Nevertheless, higher molecular weight compounds will be excluded from pores smaller than their size. For dyes, a relationship dc = L 74 Lc was assessed between the critical pore diameter of the activated carbon de and the critical minor diameter Lc of the adsorbate molecule [47]. In addition to pore structure, the surface chemistry of the adsorbent may have a strong influence on the adsorption of aromatic compounds. Fig. 7 (from [48]) shows that the Langmuir adsorption capacity of phenol is dependent on the concentration of acidic surface oxides, as defined by Boehm [49]. Furthermore, for acidic and basic species, the effect of the adsorbent's surface charge combined with that of the pH of the solution is extremely important because it determines the nature of the forces (attractive/repulsive) between the adsorbate and the adsorbent surface. For cationic dyes, adsorption is promoted when the surface charge is negative: adsorption capacities are 1.7 to 2.3 times higher at basic pH and adsorption rates are doubled [50]. 3.2. In a dynamic reactor Although batch laboratory studies give useful information on the application of adsorption to the removal of specific waste constituents, continuous carbon filters provide the most practical application of this process in waste treatment. Indeed, the continuous process enables the use of high capacities in equilibrium with influent concentration rather than effluent concentration and allows biological activity, which may affect adsorption, to be taken into account.

392

Four P. Le Cloirec and C. Faur

Table 4 presents some adsorption data obtained in continuous activated carbon systems. Like in batch systems, the extent of adsorption is dependent on the operating conditions, adsorbate and adsorbent properties. For the adsorption of micro-organics (p-nitrophenol, benzoic acid) in fixed-bed columns, the half breakthrough time increases proportionally with increasing bed depth but decreases inversely proportionally with increasing water flow rate [54,55]. By studying the adsorption of chloroethylenes on activated carbon fibers, Sakoda et al. [56] determined a linear relationship between the overall mass transfer coefficient (Kja) and the flow rate UQ. The influence of temperature on the dynamic adsorption of phenol on fibrous activated carbon has also been demonstrated [57]. The influence of the pore characteristics of activated carbon on their dynamic properties has been extensively studied. Breakthrough curves obtained with p-nitrophenol and various activated carbons exhibit different shapes due to differences in pore size distribution: microporous activated carbons induce a slow intraparticular diffusion resulting in flattened curves whereas more meso- and macroporous adsorbents possess a sharper curve because of an enhancement of mass transfer [58]. The adsorption of trihalomethanes on granular and fibrous activated carbon also shows adsorption capacities proportional to the micropore volume [59]. Table 4 Some data related to dynamic adsorption of organic species. GAC: granular activated carbon (AC); ACF: AC fibers; Co: influent concentration; Z: bed depth; M: AC weight; Uo: influent velocity; Qo = influent volumetric flow rate; 1*,: breakthrough time; CUR: carbon usage rate; Q s : saturation adsorption capacity Activated carbon

Organic species

GAC

Operating conditions

Breakthrough parameters

Ref.

p-Nitrophenol

Co = 4.4 mmol L"'

Half time t w =117-315 h

[54]

Benzoic acid

Z = 3-6 cm

[55]

U o = 3.6-14.3 cm h"1 Several

p-Nitrophenol

GAC

Co = 150 and 300 mgL" 1

tb = 8-157 h

Z = 20-60 cm

CUR = 0.62-2.25 g I/ 1

[58]

Qo = 60-180 Ld"1 Co THM = 113-315 W L"1 Z (CHC13, CHBrClj, = 15 cm CHBr2Cl, CHBr3) Uo = 77 m h"1

GAC, ACF THM

ACF

Phenol

C0 = l g L -

1

M = 4g Uo — 5 nxL min"1

Per individual THM:

[59]

Q5 = 0.33 - 0.50 mg g"1 for GAC Q5 = 2.04 - 5.00 mg g 1 for ACF Q5 = 365 mg g"1

[60]

Adsorption of Organic Compounds onto Activated Carbon -– Applications in Water and Air Treatments

393

0.1 r—•-

0.08

0.06 o

y o

0.04

0.02

1000

2000

3000

4000

5000

Temps (rrin) * Experimental — RNstatique

RN dynamique Deterministe

Fig, 8. Comparison of knowledge model (linear driving force model) and statistical tools (static and dynamic perceptrons) to predict breakthrough times of pesticides onto activated carbons. In order to include different properties of activated carbons as influent variables, an original modeling approach using neural networks was developed in [61]. The authors compared the ability of a common knowledge model, the linear driving force model (LDFM), with two neural networks, a static and a dynamic three-layer perception, to predict breakthrough curves of pesticides onto five different activated carbons. Whereas LDFM is based on mass balance, diffusion and equilibrium equations, neural networks are non-linear statistical tools, which enable some factors related to operating conditions, adsorbate or adsorbent properties to be included as explicative variables. Specific surface area or microporosity, which were demonstrated to influence strongly pesticide adsorption, were considered as input neurons. As presented in Fig. 8, the dynamic neural network is more effective than both other models at predicting the breakthrough times of pesticides. The static perception, and especially the LDFM, underestimate the breakthrough time. In the case of the mathematical LDFM, this poor agreement may be due to the restrictive hypothesis of the model. Concerning neural networks, it seems that the dynamic one is more suitable for describing a dynamic process like the exhaustion of an activated carbon bed. 3.3. Some industrial applications 3,3.1. Types ofadsorber Depending on the nature of the wastewater, one of several types of granular carbon column design may be employed, as presented in Fig. 9: 1) Water may be applied to either upflow (packed or expanded) or downflow beds. 2) Single-stage contactors may be used, or columns may be arranged in series or in parallel.

394

Four P. Le Cloirec and and C. Faur

In

1

Granular aciivaied carbon

-A / \

uu u •

Ou:

Downfiow in

In

Granula activated carbon

Oul

A _/\ >

. fJ

Out Downfiow in parallel

rrr Ou

In

Upflcrw expanded m series

Fig. 9. Types of GAC column design (reprinted with permission from [33], copyright 2000 McGraw Hill eds.). Table 5 presents guidelines for the application of the different configurations of adsorbers (from [33,62]). Table 5 Guide to selection of a GAC adsorber (reprinted with permission from [33], copyright 2000 McGraw Hill eds. And from [62], copyright 1998 McGraw Hill eds.) Adsorber configuration Single adsorber Fixed beds in series Fixed beds in parallel Upflow moving beds Upflow expanded beds

Application Low carbon adsorption kinetics High adsorption kinetics and a high effluent quality must be ensured To minimize pressure drops and ensure a high flow rate High adsorption kinetics - suspended solids are present in the influent or biological action occurs in the bed Suspended solids are present in the influent and some of them may be tolerated in the effluent

3.3.2. Design parameters In wastewater treatment, carbon adsorption systems often use granular activated carbon. The design of the activated carbon system consists essentially in determining: - the contact time; - the carbon breakthrough and exhaustion; the hydraulic loading rate; - the carbon usage rate. The empty bed contact time (EBCT) is calculated as follows:

Adsorption of Organic Compounds onto Activated Carbon –- Applications in Water and Air Treatments

V L'bed EBCT = — =

Q

395 395

(31)

u,

where V is the volume of the empty bed (occupied by the activated carbon), Q is the volumetric flow rate of influent, L^d is the depth of activated carbon and Ui is the linear velocity of water. The EBCT has a significant impact on the performance of the activated carbon column. As shown in Fig. 10 (from [63]), as EBCT increases, the bed life (expressed in bed volumes of product water to breakthrough) will increase. Current EBCT ranges from a few minutes to more than 4 h for the removal of some specific contaminants present at high concentrations. The breakthrough can be defined as the point where the organic concentration in the effluent is higher than that required. This given time, called breakthrough time ft, may be determined using the Bed Depth-Service Time (BDST) approach proposed by Hutchins [64] from the Bohart-Adams equation [65]:

h

(32)

=•

BVCn

where No is the adsorption capacity of the bed per unit bed volume, othe fluid velocity, s the bed porosity, Co the influent concentration and (Z - Zo) the amount of bed depth in excess of the critical depth. The values of the parameters No and Zo can be found by plotting the data on h vs. Z.

Bed Dtplli. m

Fig. 10. Service time vs. EBCT for the removal of chloroform from Miami, Florida water by GAC (reprinted with permission from [63], copyright 1999 Lewis Publishers).

396

Four P. Le Cloirec and C. Faur

The hydraulic loading rate is defined as the rate of a volume of water passing through a given area of the activated carbon bed, usually expressed in m3 m~2 h"1. This parameter, correlated with head loss, is important when the mass transfer is controlled by external transfer (the ease of highly adsorbable compounds). When the intraparticular rate is the controlling step, the hydraulic loading rate is not important. Values of hydraulic loading rates range between 1 to 30 m h"1, although 7 to 10 m h"1 is more commonly used as design criteria [43]. The Carbon Usage Rate (CUR), calculated as the mass of activated carbon required per unit volume of treated water, characterizes the rate at which carbon will be exhausted and how often it must be replaced. 3.3.3. Industrial examples The data from 1987 on the use of activated carbons for liquid-phase applications in the US shows that 15,900 tons of activated carbon was used for wastewater treatment, 7,300 in the granular form and 8,600 in the powdered form [66]. Of 15,900 tons, 13,000 tons was used industrially and 2,900 tons municipally. Industrial applications of activated carbon are related to, in decreasing order, decolorization, treatment of chemical effluents, Pharmaceuticals or mining groundwater. The organic contaminants removed include BOD, TOC, phenol, color, cresol, polyethers, toluene, xylene, nitro- and chlorophenols, insecticides, refinery wastes and acetic acid. Generally, the flow being treated is less than 80 m3 d"1 and thermal reactivation of the carbon is used [67]. Table 6 Industrial applications of wastewater treatment by activated carbon (reprinted with permission from [67], copyright 1978 Ann Arbor Science Publishers) Industry

Flow

Organic pollutants

Textile mill, Virginia Oil refinery, California Chemicals, New York Chemicals, Texas

190

Dyes

Residence Type of Regeneration time adsorber (min) Moving bed 57 None

15,960

COD

60

57

Phenol, COD

200

5,700

Nitrated aromatics phenol

40

Pharmaceuticals, Switzerland Herbicide, Oregon

95 570

90

Chlorophenols 105 Cresol

Gravity beds in parallel Downflow beds in series Moving beds

Multiple hearth furnace None Rotary kiln

Downflow beds None in series Upflow beds in Multiple hearth furnace series

Adsorption of Organic Compounds onto Activated Carbon -– Applications in Water and Air Treatments

397

Some specific examples are presented in Table 6, including parameters like wastewater flow, organic contaminants, contact time, and type of contactor or regeneration process where available. For a particular application, activated carbon may be coupled with another treatment. This is the case for the PACT Process, which consists in adding powdered activated carbon to the aerator of a submerged culture biological process in order to combine adsorption and biodegradation processes [68]. This process was applied to the removal of toxic organics, aniline and phenol [69]. Starting with initial concentrations of 500 mg L"1 for both organics, low residuals of 0.15 mg L"1 and less than 0.04 mg L"1 were reached for phenol and aniline respectively. A percentage of non-biodegraded phenol ( 4 - 9 %) and aniline (15 - 32 %) remained adsorbed on the powdered activated carbon surface. In the case of biodegradable compounds like dyes, PAC addition does not improve the biological process efficiency [70]. Powdered activated carbon may also be coupled to ultrafiltration (UF). This hybrid system, referred to in the literature as the PAC/UF process, incorporates the adsorption capabilities of activated carbon and the microorganism and particle removal ability of the UF membranes [71]. Combined with coagulation as a pre-treatment, the coupling has been successfully applied to the treatment of domestic wastewater [72]. Metivier-Pignon et al. [73] have proposed a combination of ultrafiltration with activated carbon cloth to treat complex aqueous solutions with suspended solids, colloids and dissolved organic molecules, like highly colored wastewaters from ink manufacture. 3,3.4. Regeneration For powdered activated carbon, inexpensive disposal methods and a low initial cost have led to the predominant use of fresh carbon. For granular activated carbon, the high initial cost makes regeneration essential and three common methods may be used: thermal, steam and chemical [62]. Thermal regeneration is the most used, the two other methods being more dedicated to adsorbate recovery. Thermal regeneration is divided into 4 steps [43]: - drying up to 200 °C, to remove high volatile adsorbates; - vaporization of volatile adsorbates between 200 and 500 °C; - pyrolyzing non-volatile adsorbates at 500 - 700 DC; activating carbon at a temperature above 700 °C. The factors affecting regeneration efficiency are primarily those that affect the oxidation step of activation, namely the type of activating agent (CO2 or steam), the time and temperature of activation, the amount and type of adsorbate, the type and quantity of inorganic substances accumulated on the carbon and the type of activated carbon. Due to its high temperature, thermal regeneration is usually not conducted in situ but requires shipment of the spent activated carbon to special regeneration units such as multiple hearth furnaces or rotary kilns. The design of the reactivation system is dependent on the carbon loading, in mg g"1, and on the carbon usage in the adsorber i.e. the mass of carbon exhausted per unit time, in kg d"1 [43].

398

Four P. Le Cloirec and C. Faur

t - 20 (min) -20

4• 0

4

s S 10 Reactivation Cycle

b)

Fig, 11. Influence of a) regeneration temperature and b) regeneration time on recovery {reprinted with permission from [63], copyright 1999 Lewis Publishers and from [67], copyright 1978 Ann Arbor Science Publishers).

In general, thermal regeneration results in mass losses of between 9 and 19 %, due primarily to the evaporation of moisture (~ 4 %), to the volatilization and combustion of adsorbed organic components and to the combustion of a relatively small portion of the activated carbon [67]. This weight loss indicates a weakening of the carbon structure and a clogging of the smaller pores into mesopores [74]. Furthermore, a decrease in particle diameter occurs, leading to higher kinetic rates but also an increase in pressure drops in fixedbed adsorbers. Mass losses may be related to the activation temperature [67] and to the activation time [63], as presented in Fig. 11. Typical design data for thermal regeneration are presented in Table 7 [75]. Chemical regeneration is an alternative to thermal treatment that enables in situ regeneration and avoids the loss of carbon. Organic (acetone, methanol, acetic acid) and inorganic regenerants (sodium hydroxide) may be used. Thermal and chemical regeneration of activated carbons exhausted with various aromatic compounds have been compared [76]. Whereas thermal treatment is independent on the characteristics of the adsorbate, the efficiency of the chemical treatment, which varies from 15.2 to 96.8 %, is a function of the properties of adsorbate and the choice of chemical

Adsorption of Organic Compounds onto Activated Carbon -– Applications in Water and Air Treatments

399

regenerant. Organic solvents possess a relatively higher regeneration potential than other chemical regenerants. Furthermore, a high solubility of the adsorbate in the regenerant solution is required to reach high regeneration efficiencies [77]. Steam regeneration is suitable for those adsorbates that are volatile enough. Table 7 Design data for thermal regeneration of carbon (reprinted with permission from [75], copyright 1990 McGraw Hill eds.) Parameter Multiple-hearth furnace loading Residence time Natural gas fuel Steam Heat loss

Typical value 55 - 75 kg m"3 h"1 30min 0.19 - 0.34 m3 kg"1 C l-3kgkg'C 15%

4. AIR TREATMENT - VOC AND ODOR REMOVAL Activated carbon is a universal adsorbent for organic pollutants present in industrial emissions. Thus, a large number of adsorbers used in air pollution treatments are inventoried. This paragraph presents a general approach to solid - pollutant interactions in the gas phase and a description of the adsorption - desorption processes carried out for emission control. 4.1. Air pollutant adsorption; general approach 4.1.1. Factors affecting adsorption Specific surface area and pore diameter First, the adsorption capacity of a molecule onto a porous medium is proportional to the adsorbent surface. A microporous material, with a pore diameter less than 2 run, is generally used for air treatment. Mechanically, the pore diameter has to be much bigger than the molecule size to achieve a rapid diffusion inside the porous volume in order to reach the adsorption sites. Pollutant structure The transfer and the adsorption are strongly influenced by the size of the compound (surface and volume) and by the functional groups present in the molecule (alcohols, aldehydes, ketones, carboxylic acids, amines, mercaptans, halogenated molecules) inducing some polarization effects. Some studies have reported a quantitative structure activity relationship (QSAR) between molecular structure and adsorption parameters (adsorption capacities, energies) [32,37,78]. The reactivity of some compounds leads to oxidation at the adsorbent surface, which plays a catalytic role. Moreover, a mixture of molecules in air

400

Four P. Le Cloirec and C. Faur

results in an adsorption competition between the different molecules and thus reduces the adsorption capacity of the molecule alone. Humidity and temperature The influence of air moisture is due to competition between water molecules and adsorbate. Activated carbon is generally useful at room temperature until a relative humidity of about 70 ~ 80 % for pollutant concentrations ranging from 1 to 1,000 mg m"3. However, the most important parameter is the ratio of molecule and water concentration. That is to say, a high humidity will inhibit the removal of solvent traces. In contrast, a high humidity will have no effect on high solvent concentration removal. Due to the exothermic reactions, temperature influences adsorption; it is less efficient at high temperature. Thus, adsorption at a temperature less than 40 °C is recommended. 4.1.2. Adsorption — desorption energies The molecule-solid interaction energy is an important parameter not only for adsorption but also in terms of desorption and regeneration of the saturated porous material. These values are determined by different calorimetric analyses [78-82]. Some data are presented in Table 8. These values are useful to calculate the adsorption exothermic heat or the energy to be introduced into the system in order to regenerate the porous media [83,84]. An example of a model presenting the heat and mass transfer in a VOC adsorber will be given in section 4.1.5. 4.1.3. Some specific cases Ketones Akubuiro and Wagner [85,86] proposed ketone oxidation mechanisms at the surface of an activated carbon. Ketones are oxidized to produce peroxides, which are very unstable and decompose with strong exothermic reactions. These by-products give earboxylie acids, aldehydes and/or diketones. An example of the reaction pathways is shown in Fig. 12. Hydrogen sulfide and ammonia Hydrogen sulfide, always found in odorous gaseous emissions, is removed by physisorption onto activated carbons with an adsorption capacity of less than 3 % [67]. An oxidation reaction generally occurs with the oxygen in air at room temperature, catalyzed by the surface of the medium: H2S + 1/2 O2 -» 1/8 S8° + H2O

(33)

Adsorption of Organic Compounds onto Activated Carbon -– Applications in Water and Air Treatments

401

Table 8 Adsorption - desorption energies of VOC in a microporous AC obtained by differential scanning calorimetry (reprinted with permission from [78], copyright 1998 Elsevier). Desorption Energy Volatile organic compound Adsorption Energy I (kJ.maP) AH^ (kJ.mor1) -AHJJ 52.6 48.5 Aeetaldehyde 55.2 Acetone 50.6 68.7 2-Bromopropane 54.2 Butylamine 85.6 74.9 Butyraldehyde 65.5 53.2 Chloroform 56.3 50.5 68.5 49.4 2-Chloropropane Cyclohexane 65.1 55.4 73.2 63.5 Cyclohexene 1,2-Dichloroethane 60.1 51.2 53.2 Dichloromethane 48.6 67.3 Diehloroether 56.9 93.4 Diisopropylamine 74.4 Dimethoxymethane 71.4 64.2 Ethanol 54.7 48.6 70.4 61.2 Ethyl acetate 83.2 76.2 Ethyl acrylate 61.7 51.4 Ethyl formate 49.3 Formaldehyde 42.3 70.4 Heptane 72.9 Hexane 78.7 61.4 74.1 1 -Hexene 63.0 73.9 1-Hexyne 56.9 Isopropyl acetate 78.7 68.5 Isopropylic ether 70.5 62.2 Isobutylvinylether 71.1 64.7 Methanol 85.6 40.8 65.5 Methylethylketone 56.2 56.3 Methylethyldioxolane 70.9 Methylisobutylketone 68.5 67.6 62.5 2-Methyl-1 -propanol 56.6 73.2 Pentane 57.9 60.1 Propanol 50.0 Propionaldehyde 58.7 49.4 Tetrachloroethylene 67.3 70.2 Tetrahydrofuran 54.1 49.1 Toluene 71.4 63.1 54.7 Trichloroethylene 65.6 Triethylamine 70,5 81.9 2,2,4-Trimethyl pentane 87.9 75.3

402

LeCloirec and C. Faur Four P. Le HO

HjC—C—C—CH 3 H

HO

H, O

"* > H 3 C—C—c—CHj

HO

*• H a C — C ^ C — C H 3

O—OH

*- H 3 C—C

O^OH

O

II

III HO

H

o

O

- I — - I - II

— HjC—C—C—CH 3

H»C

o

III 0

C—CB, OH

IV

o

I II

-«- O — C—C—CHj

+

+ H

2°

V

OH

IV 0

H3C-C-C-CH5

0 - j ^ ~ 2C-CH, OH

V IV Fig, 12. Probable oxidation mechanisms of methyl ethyl ketone (reprinted with permission from [85,86], copyright 1992,1993 American Chemical Society). Hydrogen sulfide is transformed into sulfur, which stays on the porous structure. Mechanisms of oxidation, the influence of water and the autocatalysis of sulfur produced on the surface are discussed in [87], In the case of wastewater treatment plant emissions, the amount of sulfur removed by adsorption and oxidation is about 900 mg g"1. Activated carbon does not seem to be a good adsorbent of ammonia present in air. Thus, in order to increase the adsorption capacities of such a material, an impregnation with metal or metal oxides is performed. Iron, copper or other metal-impregnated activated carbons are commonly used. 4.1.4. Organic pollutant adsorption capacities From the breakthrough curves, data of adsorption capacities have been determined for a large set of volatile organic compounds or odorous molecules adsorbed onto activated carbons. Values are presented in Table 9. 4.1.5. Modeling mass and heat transfers [82-84] The adsorption of VOC vapor with activated carbon is a mass transfer from fluid to solid and an exothermic reaction with an increase in the local temperature. Fig. 13 shows the temperature evolution, obtained by an infrared camera, inside the column during the removal of acetone hi air. The polluted air flows from the top to the bottom of the adsorber.

Adsorption of of Organic Compounds onto Activated Carbon -– Applications in in Water and and Air Treatments

403

Table 9 Vapor adsorption of different solvents onto activated carbon in a dynamic system, T = 20 °C, P = 1 atm (reprinted with permission from [67], copyright 1978 Ann Arbor Science Publishers) Compound Acetaldehyde Acetic acid Acetone Acetylene Acrylic acid Amyl acetate Butyl acetate Ethyl acetate Methyl acetate Isopropyl acetate Bromhydrie acid Butyric acid Aery aldehyde Amylic alcohol Ammonia Benzene Butane Butanol Butylene Butyl chloride Butyne Butaraldehyde Camphor Caprylie acid Carbon disulfide Carbon tetrachloride ChloThydric acid Chlorine Chloroform Cresol Crotonaldehyde Decane Diethyl ketone Ethanol Ethyl chloride Butyl ether Ethyl ether Ethylene Ethyl mercaptan

Maximal adsorption capacity (%) 7 30 15 2 20 34 28 19 16 23 12 35 15 35 negligible 24 8 30 8 25 8 21 20 35 15 45 12 15 40 30 30 25 30 21 12 20 15 3 23

404

Le Cloirec and Four P. Le and C. Faur

Table 9 (continued) Vapor adsorption of different solvents onto activated carbon in a dynamic system, T = 20 °C, P = 1 atm (reprinted with permission from [67], copyright 1978 Ann Arbor Science Publishers). Eucalyptol Fluorhydric acid Formaldehyde Formic acid Heptane Hexane Hydrogen sulfide Indole Iodhydric acid Iodine Iodoform Isopropanol Methanol Methylene chloride Methyl ethyl ketone Methyl isobutyl ketone Methyl mercaptan Naphthalene Nicotine Nitric acid Nitrobenzene Nitrogen dioxide Nonane Octane Ozone Palmitic acid Pentane Phenol Propane Propionic acid Propylene Propyl mercaptan Propyne Pyridine Skatole Sulfur dioxide Sulfuric acid Sulfur trioxide, oleum Toluene Valeric acid Xylene

20 10 negligible 7 23 16 3* 25 15 40 30 26 10 25 25* 30* 20 30 25 20 20 10* 25 25 Decomposition* 35 12 30 5 30 5 25 5 25 25 10 30 10 29 35 34

Adsorption of Organic Compounds onto Activated Carbon -– Applications in Water and Air Treatments

405

AT ∆T (°C)

0,7 6,1 1

11,4 11,4 16,8 16,8 t= 2 22,2

27,6 t = 2 min

t ==116 6 mmin in

t = 30 min 30min

t = 38 min 38min

t = 46min 46 min

t = 58 min 58min

Fig, 13. Local temperature evolution (IR camera) in an activated carbon column fed by air loaded with acetone (Co = 50 g m"3 and Uo = 500 m h"1). The system is modeled by a set of 3 partial differential equations (PDEs): the energy balance and the adsorbate mass balances within the gas and solid phases. The following assumptions are made to develop the model: - the pressure drop through the bed is negligible; - the plug flow model describes the flow pattern. Temperature, adsorbate concentration, gas flow and porosity are uniform at any cross section of the bed; - the physical properties of the adsorbent are considered constant and the accumulation of energy in the gas phase is negligible; - heat loss is negligible in large industrial columns, which operate under near adiabatic conditions contrary to laboratory scale columns. An overall heat transfer coefficient is used to express heat loss; - heats of adsorption are usually considered as constant in mathematical models. However, values of adsorption energies are dependent on the surface loading so this phenomenon must be introduced into the energy balance. Variations of adsorption energies with temperature are small and are not considered in the model; - the linear driving force (LDF) model represents the mass transfer rate. Because of the high VOC concentration range, granular external surface diffusion is generally considered to be the main contribution to the mass transfer. However, the mass transfer coefficient is an effective lumped-resistance coefficient. Based on the above assumptions, the mathematical model could be written according to the following equations: - mass balance for the gas phase:

{mass flux}+ < \ +\ [ in the gas phase J [_ in the solid phase Un 0

b

da

a

„

(34)

406

P. Le Cloirec and C. Faur

- mass balance for the solid phase: -^ = kp(99.9 %) of mercury (II) removal over a wide pH range (2.5 to 11). The other activated carbons used showed the maximum total mercury (II) uptake at pH 4 to 5 with the percentage of mercury (II) removal reduced distinctly at other pH values. The adsorption of zinc was also found to depend on the treated solution pH [92]. The pH effect was recognized as related to the interactions between complexes formed on the carbon surface and Zn(II) ions in a treated solution. For the enhanced removal of heavy metals (Hg, Cd, Pb) some authors suggest application of activated carbon containing sulfur in its structure [93-94].

439

Activated Carbon Filters and their Industrial Applications

Table 7 Total organic carbon reduction by granular activated carbon in some industrial waste waters, adapted from [90] Industry

Food Textiles Paper Printing Chemicals Rubber Stone, glass, clay

Waste water Total Organic Carbon, [mg C dm"3] 25-5300 9-4670 100-3500 34-170 36-4400 120-8350 12-8300

Carbon usage,

Average removal,

fern3]

[%] 90 93 90 98 92 95 87

0.1-41 0.12-29 0.4-19 0.52-0.55 0.13-17 0.62-20 0.34-36

Activated carbon is one of the most commonly used meda for dye removal [95]. It very effectively adsorbs cationic, mordant, and acid dyes while the adsorption of dispersed, direct, pigment, and reactive dyes is lower. Because the effectiveness of activated carbon in adsorption of dyes depends on the characteristics of the waste water, in certain cases carbon adsorption can be ineffective [5], Other examples of carbon filtration in waste water treatments include an effective controlling of perfluorocarbons (PFCs) emissions from the semiconductor manufacturing processes, purification of waste waters in the paper industry, etc [96-97]. In general, those organics, which are adsorbed poorly on carbon, undergo biodegradation easily. On the other hand, contaminants which are hard to biodegrade can be easily adsorbed onto activated carbon. Adsorption onto activated carbon is considered as the best available technology for removal of non-biodegradable species from aqueous streams [98]. Treatment with activated carbon is usually used as the last stage (after biological treatment and primary filtration) for treating of waste waters and dumpsite leachates. In common practice, GAC or extruded activated carbon can be used to adsorb residual contaminants in tertiary treatment systems [99]. In this mode, carbon filter units are installed after the primary filtration and secondary biological treatment. The tertiary treatment involves the adsorption of both organic and inorganic compounds. Typically over 99% reduction of biological oxygen demand can be achieved [6]. However, the concentration of pollutants in industrial waste waters can be much higher than in drinking water. This, in turn, has a strong impact on the adsorption lifetime of the activated carbon beds. Although high concentrations are accommodated in a plant design, the carbons serve effectively only from several months to a few days. For this reason, the use of activated carbons in purification of waste waters is economically feasible if appropriate regeneration of spent carbon is possible [100]. Treatment of waste waters can be also carried out using powdered carbons but only the treatment with PAC can be used independently only in specific cases. In the most common technology, PAC is used together with activated sludge [101]. The powdered activated carbon

440

J. Przepiórski Przepiorski

treatment (PACT) process [102-103] based on a direct addition of PAC to the aerated activated sludge has been developed by DuPont Co, and USFilter's Zimpro process in the early seventies. Processes taking place in aerated tanks holding PAC and the activated sludge result in both the adsorption of non-biodegradable organics and improvement of the entire biological process through fixing bacteria on the carbon surface. Some authors suggested a synergistic effect [104-105], while others reported an additive effect including adsorption onto activated carbon and biodegradation by the activated sludge [106-107]. The process can effectively treat various complex municipal and industrial waste waters (e.g. kraft mill [108] or textile dyeing [109] waste waters), contaminated groundwater, and landfill leachates in a single step combining adsorption and biological treatment. Data reported by Orshansky and coworkers [110] showed very efficient removal of phenol and aniline (from 500 mg dm"3 to 0.15 mg dm"3 and 0.04 mg dm"3, respectively) from solutions subjected to the PACT process. The PACT process is designed to improve the performance of the secondary biological treatment step [111] (Fig. 10.), through the stabilization of biomass against shock loadings or upsets. Thus, no deterioration in the biological activity is associated with the PACT technology. The spent activated carbon is removed along with the microorganisms and in most cases incinerated. The application of the PACT process results in the following advantages over the processes, which conventionally use activated sludge [105, 108]: - higher removal of biological and chemical oxygen demands, - improved sludge settling, - increased stability of the system to shock loads. Wastewater

Pretreatment

PAC

•

Biological filtration Biological

Clarification

Sand filtration

Treated water

Fig. 10. Utilization of activated carbon in PACT process

Activated Carbon Filters and their Industrial Applications

441

3.3,2, Biological filters Filtration by biological activated carbon (BAC) can be applied to the treatment of waste waters and to the production of drinking water. The process is based on bacteria immobilization on granular activated carbon [112]. Principally, BAC is used for biodegradation of dissolved organic carbon (DOC) [113-115] and for oxidation of ammonia [116-118]. According to Koch and coworkers [119], activated carbon with a growing microbial colony removed ca. 40% of the initial DOC from solution. Three independent mechanisms (surface degradation, film degradation, and pore degradation) were proposed for the removal. BAC is also a known medium for of removal TOC from waste waters. As reported [120], BAC employed for treatment of waste waters from wood treatment plants, after 64 bed volumes was able to reduce TOC by at least 80% or 50% at TOC content in influent 320 mg dm"3 or 900 mg dm"3, respectively. Improved uptaking of perchlorates and metals by biological carbon filters was also reported [121-122] and the efficiency of BAC filters was recognized as dependent on the inlet contaminants concentrations [123]. Removal of dyes in the BAC process was recognized [112] as more effective than a combination of individual carbon adsorption and biological processes (Table 8). In order to decompose some high molecular weight organic compounds to easily biodegradable species, ozone is often introduced into the water to be purified [42, 124]. As reported [125], the combination of pre-ozonation and biofiltration effectively removes color and greatly reduces THM formation potential and chlorine demand of treated waters. However, the total removal of humic compounds is lower, normally below 30%. Taking into consideration the fact that pre-ozonation deteriorates adsorption of some organics [126], a growing interest is observed in the use of alternative adsorbents to GAC (activated alumina, bone char, activated bauxite) [127]. Table 8 Removal of acid dyes in a stirred tank reactor by free bacterial (P. putida) cells and bacteria immobilized on GAC F400 at 298 K, adapted from [112]

Acid Black 26A Acid Blue 277 Acid Red 361 Acid Orange 156

Color removal after 24 h, [%] Free cells BAC 80 95 90 30 80 5 97 5

Another process carried out by means of BAC is biological nitrification. It is a two-step process involving a sequential oxidation of ammonia into nitrite and nitrite into nitrate by Nitrosomonas and Nitrobacter bacteria, respectively [128]. Ammonia contained in water is nitrified by bacteria immobilized on the external surface of activated carbon [129] according to following equations:

442

Przepiorski J. Przepiórski

NH 4 + + 3/2O2 + H2O - ^ NO2" +2H3O+

(1)

NO2" + 14O2^NO3"

(2)

The nitrification process has been recognized as increasing the chemical and biological stability of water in the distribution system and decreasing the formation of undesirable chlorinated by-products. Properly optimized nitrification can significantly reduce the presence of nitrite, taste, and odors as well as of bacterial regrowth in water distribution systems. One of the main factors affecting the nitrification process is temperature [130-132]. It is generally believed that the rate drops sharply at temperatures below 288 K [117, 133]. As confirmed for a variety of contaminants, the important advantage of the filtering systems employing BAC units is an increased effluent output until the breakthrough of the filter [134]. Some researchers have reported that the bacterial activity in GAC filters contributes to an increased service life of GAC [135-136]. From a practical point of view this benefits in a less frequent regeneration and thus in a reduction of down-time. On the other hand, microbial growth entails clogging and thus more frequent backwashing is required owing to an increased pressure drop [137]. 3.3.3. Ex situ ground water remediation technology Concern over contaminated groundwater sources increased drastically in the 1980s. Therefore, consumption of activated carbon for groundwater treatment increased in the late eighties followed by a further growth over the next years. Liquid-phase GAC can be potentially used for remediation of waters contaminated with various classes of compounds (Table 9) and is generally used in conjunction with other remediation technologies such as soil flushing, biodegradation, and pump-and-treat [138]. The major target contaminants for carbon adsorption are semi volatile organic compounds (SVOCs) and ordnance compounds [139]. Adsorption on activated carbon filters is also applicable for remediation of waters from polychlorinated biphenyls (PCBs) as well as other halogenated volatile organic compounds (perchloroethane, trichloroethylene) [140]. Table 9 Examples of organic compounds in hazardous material spills and cleaned up with the use of activated carbon treatment, adapted from [2] Contaminant

Waste, [m3]

PCB Chlordane Kepone Toluene Dimethylphenol

2270 11 852 946 946

Concentration in influent, [ppb] 400 1430 4000 120 1220

Concentration in effluent, [ppb] < 0.075 0.43 423 K)

Dopants (B, P, As, Sb)

- corrosion of silver contacts (especially H2S) - corrosion of copper and aluminum by H2SO42" - formation of crystalline compounds in reaction of SO42" orNO 3 "withNH 4 + - defects in the chemical amplified resist processes - defects in pattern generation processes - formation of crystals (by neutralization with acids dissolved into moisture on the wafer surfaces after cleaning with deionized water) - lowered illuminance resulted from lens contamination - problems in deep ultra-violet photolithography - defects in voltage resistance and current leakage (by carbides in oxidation gate processes) - defects in forming oil membrane on wafer surfaces (by high-molecular weight organic compounds) - formation of SiO2 (in reaction of siloxane with ozone) - inhibition thin film growth and deposition - evolution of secondary contaminants during high temperature processes - problems caused by accidental doping of semiconductors - defects in electric properties in the LSI and LCD devices - problems in performing diffusion processes

Table 19 Examples of molecular contaminants to be removed and carbon impregnating chemical used for purification of air for semiconductors production processes Example of target gas

Semiconductor manufacturing process

Reaction with the additive

NH 3 HF H 2 S, SO2 TOC, Toluene

Photolithography Etching Outside air handling Cleaning/diffusion

NH 3 + H3PO4 -> NH 4 H 2 PO 4 2HF + K2CO3 ->• 2KF + CO 2 + H2O H2S + K2CO3 -> K2S + H2O + CO2 Physical adsorption

Activated Carbon Filters and their Industrial Applications

459

Fig, 16, Example of high-flow separator type filter used for clean-room circulation

Honeycomb matrix

Fig. 17. Scheme of filter unit with a honeycomb structure Owing to the possibility of carbon settling, an appropriate distribution of activated carbon during filling of the honeycomb filter is of primary concern [87]. If settling does occur, the filtered air may bypass activated carbon media, thus making the filter less effective. The adsorption capacities and efficiencies of honeycomb carbon filters offered by industry can be easily adjusted by controlling the wall thickness [88]. Chemical filters containing folded impregnated activated carbon fiber exhibit an outstanding efficiency for the removal of target gases. This kind of sorbent is able to efficiently capture the molecular contaminants even from the initial concentration of few tens of ppb to a level below 1.0 ppb and the final effect depends on airflow. Apart from the high removal efficiency, the use of the corrugated ACF sheets offers a low filtered air pressure drop. In order to achieve an ideal clean environment, the carbon-based chemical filters often work in a combination with low-outgassing filters or PTFE membranes.

460

Przepiorski J. Przepiórski

4.8. Oil vapor removal from compressed air Gas compressors are often directly lubricated by oil in order to work properly. Thus, to some extent, the traces of oil are transferred into the compressed gas. Additionally, since heat is generated during compression, the compressor oil undergoes both degradation and oxidation, usually with the formation of acidic products [220]. The presence of the oil and its degradation products in the compressed gas often should be avoided due to a potential negative influence on the subsequent processes. Since the oil appears in compressed gases mainly as small droplets or as a mist, a mechanical way of its removal seems to be the most feasible. However, filtering units (made of cotton, shaped sintered metals or porous ceramics) conventionally employed for that purpose were found to be ineffective to capture those parts of oil present as vapor (5 to 20 vppm), gaseous products of oil degradation, as well as the airborne contaminations normally present in the air supply, including water and odors [221]. Because the capacity of activated carbon for adsorption of oil in vapor state and other organic contaminants is very high, its application for cleaning oil-containing air is common. Activated carbon filters are often used for purification of compressed air in critical applications (mainly food production, electronic, pharmaceutical, and fermentation industry), where only compressed air of the highest quality is acceptable [222]. 4.9. Other gas phase applications Gas phase activated carbon filters find a limited use in several other applications. There is a choice of commercially available carbon filters designed especially for purification of byproduct CO2 from contaminants in food industry or in soft drink bottling plants. In order to meet the odor control requirements, carbon filtration is sometimes employed in the catering industry. Activated carbon (particularly brominated) finds application for the removal of ethylene from fruit storage facilities [223]. The filtration systems are also installed for protection of museum exhibits from the ravages caused by airborne pollutants or for odor control at landfill sites, sewage, and waste water plants. Activated carbon filters are sometimes used for the control of discharged or displaced air arising from tank emptying or filling. A very specific is the use of carbon adsorbent as adsorbing medium in off-gas delay beds in the nuclear industry. Here, radioactive noble gas isotopes are adsorbed in activated carbon with very fine pores [224]. Consequently, after time long enough for the radioactive isotopes to decay to acceptable levels of radioactivity, the gases are released into the atmosphere. A relatively new application field for activated carbon is air filtration in vehicle cabins [225] and aircrafts [226]. In this specific application, activated carbon granules are usually used together with a high efficiency particulate filter medium [227]. In recent years activated carbon filters have found the application for continuous gas cleaning in the cement clinkering process. A particular application of porous carbon as filter is the gas separation [228] by carbon molecular sieves (CMS). This highly microporous (predominant pore size

- - Mean Mean inhaled inhaled Flow Flow rate

-inn 100

c

Volume inhaled aled (litres/min) (Ii nin)

— Real inhaled flow rate

° "I

...

...

...

-

•

1

)

000c

10 10

20

3 30

Time (s)

Fig.8. Real vs mean inhaled flow rate (individual filter)

3.3. Collective filtration Collective filters differ from the individual ones by the fact that they are not connected to ones gas mask. They can provide clean air to either a closed and sealed room such as a shelter, vehicle or helicopter, or to a set of individual gas masks linked to a central filter in an Armed Infantry Fighting Vehicle. In the first case, the most common one, the room will normally be placed in overpressure, to ensure all air entering the room will pass through the filtration system. Many of the problems encountered with individual canisters do not exist with collective filters: • The airflow is not regulated by the breathing rhythm of the wearer. It is possible to have a steady, constant flow through the filter, eliminating any flow pattern related problems. • Neither mass nor available space are major limitations for collective filters. Even though in some circumstances they will have to be taken into account, for instance on vehicles or

Adsorption of Chemical Warfare Agents

•

•

•

483

helicopters, these constraints will be less stringent than for individual protection. It is possible to calculate the minimum amount of carbon necessary for a given protection time against different, well-defined threats, and to increase this amount by a considerable safety margin. As the collective filters are overdimensioned, the risk of a premature breakthrough is much lower than in the case of individual filters. It is also much more feasible to install an end-of-service indicator. A popular set-up is two filters, placed in series, with a dedicated detection device placed in between them. In this way the first filter can be replaced as soon as it is saturated and the agent is detected, while the second filter assures the continued supply of clean air. As there is more than enough carbon present in the filters, any irregularities in the packing of the bed still result in a loss of protection efficiency, but this loss will not be as catastrophic as in the case of individual filters. Given the bigger dimensions of the filter, it is also easier to obtain a good packing density. New military technology for collective filtration is focusing on PTSA-systems (Pressure and Temperature Swing Adsorption). Basically these use two filters alternatively, one providing protection while the other one is being "cleaned", i.e. the toxic compounds desorbed by a combined action of reduced pressure and temperature. Due to the nature of the adsorption processes, this is very difficult to achieve with activated carbon filters as changing temperature and pressure will reverse physical adsorption, but will have a negligible effect on chemisorption (see section 4).

3.4. Aerosol filters Aerosol filters are not activated carbon filters, but they are an integrated part of all military individual canisters (see Fig.6) and collective filtration systems. They usually comprise two parts: a coarse filter (a metallic "sieve") and the actual aerosol filter, a folded cellulose, glassfibre or other type of sheet. The first will stop larger physical objects, such as dust particles, to enter the filter. In this way it prevents these particles damaging the second filter, and lessens the possibility of clogging of the filter inlet. It also keeps out any radioactive dust in the event of a radiological or nuclear incident. The second filter will retain smaller airborne particles (i.e. aerosols) typically in the diameter range of around 0.5-1 urn. These can be liquid droplets from chemical agents, or biological agents such as spores and bacteria. Military aerosol filters are very efficient and are usually tested with aerosols of G.3um diameter. Whereas industrial HEPA (High Efficiency Particulate) filters typically present a protection factor (quantity of aerosol retained) [13] of 99.95 to 99.97%, military filters go quite commonly up to 99.997% [12]. There is no real interaction between the aerosol filter and the activated carbon, except as a barrier to prevent solid and liquid particles penetrating into and being deposited on the carbon bed. Similarly, the carbon bed will adsorb any toxic vapours that might evaporate from the liquid droplets retained by the aerosol filter.

484

P. Lodewyckx

3.5. Protective clothing The role of activated carbon in modern protective garments is rather similar to its interaction with the aerosol filter, as it is shown in Fig.9 [14]. Military protective clothing is usually of the (semi-)permeable type. This is different from the chemical industry, where most protective suits are of the impermeable type. This is due to their apparent advantage: as long as impermeable clothes are not degraded by the chemicals (e.g. strong acids), nothing will come through, be it liquid or vapour. For military applications there is another constraint, the minimisation of heat stress (see section 3.6.1), as military personnel can be compelled to wear these garments for several hours, even days. This represents a situation that is quite different from the one in industry where workers put on their protective suits immediately before entering a contaminated area, leaving it rather quickly, constrained by occupational health laws.

Toxic droplets & aerosols

Liquid water (rain,...) (rain,…) Outer layer

Activated carbon Binder

Toxic vapours

Water vapour (evaporated from skin)

Inner layer

Fig.9. The role of activated carbon in protective garments Permeable garments usually consist of a minimum of three layers [15]: • An inner layer, that does not contribute to the protection but only acts as a support and increases the comfort of the wearer, avoiding contact with the activated carbon layer. • An outer layer that provides the necessary protection against liquids and aerosols. It is usually a plastic membrane with excellent liquid repellant capabilities, but permeable to vapours (hence the term "semi-permeable"), so as to allow the evaporation of transpiration (i.e. water vapour) in order to limit heat stress [16]. • The activated carbon layer that adsorbs toxic vapours. These can be present in the atmospheric environment, or they can be liberated by the droplets and aerosol particles retained by the outer layer. There are several different types of activated carbon layers [17,18], including activated carbon (nano)particles, activated carbon impregnated foams,

Adsorption of Chemical Warfare Agents

485

activated carbon microbeads and activated carbon cloth (woven or non-woven activated carbon fibres). Activated carbon particles have also been used to limit the permeability of the outer layer, while still having good water vapour exchange capacity. This has been realised by embedding carbon particles in an impermeable plastic film. In this way, the open spaces in the membrane are replaced by activated carbon "windows", retaining toxic vapours but allowing water vapour to pass [19]. Recently much research has been focussed on increasing the reactivity of the carbon. Because the nature of clothing, and the constraints on weight and heat exchange, it is only possible to include a very thin carbon layer. So, even with the low concentrations passing the membrane, it is difficult to obtain complete adsorption / destruction of the toxic compounds. One way to overcome this problem is the introduction of so-called reactive carbon nano-particles. These are small carbon particles, impregnated with reactive substances (mostly metal oxides) that will react with the most common toxic compounds [20,21]. Usually the destruction process will involve hydrolysis, followed by adsorption. Interestingly, the impregnation products, such as Y1O2, AI2O3 and MgO, are quite different from the ones used in filters (i.e. Whetlerites, see section 4.2.2.1). 3.6. Operational effects of using chemical protection 3,6.1. Heat stress The main problem related to wearing a chemical protective garment is the heat stress [19,2225]. As the protective layers are preventing the toxic chemicals from reaching the skin, they also prevent transpired water evaporating. Even with special materials, specifically designed to form a barrier against the permeation of toxics but to let the sweat out, the rate of transpiration will go down. Chemical protection suits will also increase the body temperature by isolation as they present a heat resistance (expressed as K.m2.W"1) against any heat exchange between the skin and the environment. As a result, skin and body temperatures can experience a significant increase over a short period of time. This will also have a direct, negative effect on the heart rate (cardiovascular strain) and the blood pressure. As a rule of thumb, continuing operations wearing protective clothing becomes dangerous if at any one of the following conditions is met: • • •

Body temperature (calculated as 0.8?reci«r+ 0.27>*») becomes higher than 311.6K(38.5°C) Heart rate exceeds 80% of the maximum rate for that person Blood pressure gets higher than 22/11

A supplementary indication is the difference between the rectal and the skin temperature, which should not exceed IK, as it is a direct indication of a heat loss problem.

486

P. Lodewyckx

3.6.2. Breathing resistance Theoretically, the deeper the filter bed, the better the respiratory protection. However, the size and depth of an individual canister are limited. Size (and mass) are obviously limited by the fact that the filter has to be worn in connection to a gas mask. Too big dimensions will hamper head movement, will cause neck fatigue and even neck pain, and can limit the protection factor of the gas mask by deforming the rubber during head movements, creating leakages between the face and the mask, thus jeopardising the facial seal. This problem can be avoided by not connecting the canister directly to the mask. In some systems the filter unit is located elsewhere, e.g. in a pouch on the persons waist or hip, and connected to the mask by means of a flexible hose. It is clear this solution is not well suited for heavy duty in the field as it involves extra equipment, limitation of movement, and possible entanglement of the tube. In other cases, e.g. aircraft crews, it constitutes a viable solution. The height of the carbon layer is not only limited by the dimensions, but also by the pressure drop over the bed. This pressure drop results in a breathing resistance, making it more difficult for the wearer to breathe. As this resistance increases with the flow rate, it will hinder most when the wearer is performing heavy duty work and really needs high quantities of clean air. One way of reducing flow resistance is the use of bigger activated carbon particles, as flow resistance around a particle is known to be directly proportional to its specific surface (Darcy's law), which in turn is inversely proportional to its diameter. 120 -i

0.06-1 E 0.05 -

y = 0,04x,-D,B3 R2 = 0,999

a §• 0.04

*• a

?

I 1 0-03 K

f 100 - • 80',;

60

g 0.02-

40-

0.01 -

20 •

E

u

• •

o 0

1

2

3

4

Mean particle diameter (mm)

Fig. 10. Flow resistance vs mean particle diameter

0 1000 2000 3000 Vapour pressure at 293K (mmHg)

Fig.l 1. Breakthrough time vs vapour pressure for monochlorinated hydrocarbons

This is illustrated in Fig. 10 [26] where the flow resistance per unit of bed depth and per unit of flow velocity is plotted against the mean diameter of the particles. Unfortunately, the kinetics of adsorption are inversely proportional to the mean diameter of the activated carbon particles. As a result, smaller particles exhibit better adsorption characteristics but cause a higher breathing resistance. This can be overcome by using assisted filtration in which

Adsorption of Chemical Warfare Agents

487

filters are equipped with a powered fan (usually carried on the hip and connected to the mask via a hose) to overcome the breathing resistance. But as this device needs a power supply, the total breathing "package" consisting of mask, filter, power supply and fan becomes rather voluminous and heavy, making it unsuitable for most military applications. Consequently, the size of the particles in an individual military canister is the result of a compromise, resulting in the use of a mean particle diameter of around 1 mm (usually expressed as fractions passing/not passing the ASTM sieves = 12x30 Mesh size). 3.6.3. Psychological strain Wearing a chemical protective suit, including gas mask and filter, leads to a significant amount of psychological strain [23,24]. Not only is the wearer confronted with heat stress (3.6.1), leading to a general discomfort of the body, and increased breathing resistance (3.6.2), but there is also the limitation in the field of vision as many military gas masks have a pair of goggles. Even the ones with a panoramic visor induce limitations on the peripherical field of vision, leading to a diminished sense of equilibrium, loss of dexterity and even claustrophobia. Dexterity is also hampered by the gloves and the rigidity of the protective suit as a whole. The gas mask will also have negative effects on communications. Even with the help of special membranes, voice recognition and speech intelligibility are reduced. Communication is also more difficult as it is harder to recognise people and eye contact is nearly impossible. These diminished capabilities frustrate the wearer and increase his stress levels which are usually already rather high given the particular circumstances that made it necessary to put on the chemical protective suit in the first place... 3.7. End-of-service indicators Both in military and in industrial applications it is very important to know when a filter has reached the end of its service life, i.e. when a toxic compound is on the verge of breaking through the filter. In an industrial environment it is a question of economics. As the employer must provide adequate protection to the employees, filters must be changed regularly. This can be done either by a one-use policy, which is, of course, a very expensive way, or by relying on predictive models for estimating service life. These models always present an inherent uncertainty and they are developed for one continuous use, not for the intermittent use that is customarily in industrial applications. As a result, one has to take into account rather large safety margins, increasing again the total annual cost of safety equipment. For military applications the situation is quite different as there is normally no intermittent use, the filters being disposed of after use. Military personnel will usually have to stay in the contaminated environment for longer periods and if they cannot leave this area, changing the filters risks exposing themselves. In short, End of Service Life Indicators (ESLIs) are primarily of economic interest in industry but can be life-saving in military applications. There are several types of ESLIs. The bigger ones are real chemical detectors based on infrared, photo-ionisation, acoustic wave, etc., that use (micro)probes to measure gas

488

P. Lodewyckx

concentrations inside the activated carbon bed. As such, they can only be used in large filters, i.e. filters for collective protection [27]. In individual filters there exist three promising techniques, chemiresistor microsensors, signal compounds and colorimetric indicator films [28]. Microwatt chemiresistor microsensors [29] give a signal (audio or LED) triggered by a chemically induced change in a local resistance when the adsorption front reaches a certain point in the filter, e.g. 80% of the total depth. The advantage of this technique is its general response as "all" gases will provoke a change in resistance, giving raise to a signal. However, there are some major problems, specifically the sensitivity and response time, the fact the sensor has to be embedded in the carbon bed and the fact it needs a power supply source however small it may be. Signal compounds are pre-adsorbed chemical compounds [30]. They are situated in the original carbon, or in a supplementary layer of a different carbon. When the toxic compound reaches this depth of the filter, the pre-adsorbed compound will be desorbed (expelled) by the toxic material and can be detected at the filter outlet, either by the human nose or by an instrumental detector inside the gas mask. This is a very clean and straightforward method, suffering however from one major drawback. In order to give a warning against a wide range of different compounds the pre-adsorbed compound must be very weakly bound by the carbon. As a result, as will be described further on (see section 4.1.1), this signal compound will be easily desorbed from the carbon by the simple action of (clean) air flowing through the filter. While this is not really a problem in an industrial environment, where gas masks and filters are usually only worn when there is a toxic concentration present, it poses severe problems in a military context. Prior to an attack, the mask can be worn for quite a long time, flowing air through the filter and causing premature alerts from desorbing the signal compound. Finally, the most promising route seems to be the use of colorimetric indicator films [31,32]. The compounds with which these films are impregnated react with certain gases, giving differently coloured products. When part of the filter body is transparent, these films can be placed along the perimeter of the carbon bed, and the saturation of the bed can be followed, in real time, by the colour change along the film. The downside of this method is its specificity. While it can be extremely useful in an industrial environment, where one usually knows exactly which toxic compounds to expect, it becomes rather tedious in military applications. Either the film should react with all chemical warfare agents and toxic industrial chemicals, which is not feasible, or one needs to place a significant number of these films in parallel in the filter body. While in theory this is a viable option, it would almost certainly lead to confusion in operational circumstances.

489

Adsorption of Chemical Warfare Agents

4. ADSORPTION MECHANISMS 4.1. Physisorption 4,1.1, General Physisorption of gases and vapours by Van der Waals forces is the basis of the protection afforded by activated carbons. Even for weakly adsorbed gases, where one needs special impregnants on the carbon in order to stimulate neutralisation reactions, it plays two important roles. First of all, it is nearly impossible to get chemisorption (see section 4,2) without prior physisorption. If the different components of the reaction, especially the chemical warfare agent, are not physisorbed on the carbon even for a very small period of time, this reaction will not take place. There is a need to the temporally immobilisation of the molecules, to give them enough contact time with the impregnates. The same is true for the hydrolysis reactions by which many compounds are destroyed. These reactions take place between the physically adsorbed vapour and the physically adsorbed ambient water vapour. Added to that, it has been shown that for several reactions the carbon surface itself is not inert, but plays a significant role in the actual reaction, for instance as an electron donor or receptor [33]. Secondly, the by-products of the neutralisation reaction can be harmful themselves. Fortunately they are usually retained by the activated carbon as they are more strongly physically adsorbed than the initial CWA. How well a CWA, and a vapour in general, is retained by physisorption on a bed of activated carbon is determined by two parameters, the affinity of the carbon for this vapour, and the volatility (i.e. the saturated vapour pressure) of the CWA. This can be illustrated by the use of the Dubinin-Radushkevich equation [34]:

(1)

Where We is the equilibrium adsorption capacity (g.gcarbon"1), Wo the mieropore volume (enrlgau-boiT1), di the liquid density of the organic vapour (g.cnr3), B a structural constant of the carbon (K"2), Fthe air temperature (K), /?the affinity coefficient of the organic vapour, Cs the concentration of the contaminant corresponding to the saturation vapour pressure of the contaminant P$ (g.cnr3) and Co the contaminant concentration in air (g.cnr3) This equation gives the equilibrium adsorption capacity of a given carbon (described by the parameters Wo and B) for a given concentration, Cn of a vapour (determined by /Jand C,). The affinity coefficient /? can be calculated through the use of the molar polarisation P, and a reference vapour, usually benzene, with a value of Pe of 26.2 cm^mol"1 [35]. The molar polarisation can be calculated from the tabulated values of the refractive index at the sodium-D line, nD, the liquid density HOCN + HCN

(13)

whereas further hydrolysis of HOCN will yield CO2 and NH3. This hydrolysis could be catalysed by the presence of the Cu-Cr combination [61-63]. Another possible retention mechanism for cyanogen is a result of direct reaction with Cu2+ (Eq. 14) [40]:

CuO + (CN) 2 -»CuO.C2N 2

or

OCN / Cu

(14)

\ CN It has also been observed experimentally that there is an increase in HCN breakthrough time when using ASC carbon instead of A or AS. This suggests that chromium has a beneficial effect on HCN retention. Moreover, it has been shown that this is primarily an effect on (CN)j retention, rather than on HCN itself [40,64,65]. The overall reaction seems to

499

Adsorption of Chemical Warfare Agents

be the formation of oxamide. This reaction is catalysed by Cr1^ according to Eq.15 (formation of cyanoformamide and oxamide):

(CN)2 + H2O _ ^ I CONH 2 _ C ^ ^

'

I CN

? ? H2N-C-C-NH2

So, very generally, the complete chemisorption mechanism for HCN on an ASC whetlerite can be described by Eq.16: HCN

Cu2+ • (CN)2

Ot Oxamid • e

(16)

In this process copper is actively involved, and will be reduced from Cu2+ to Cu1+ and subsequently to Cu°. Chromium on the other hand will act as a pure catalyst and will remain in its oxidised Crs+ state. This is confirmed by experimental data from ESCA (Electron Spectroscopy for Chemical Analysis) before and after HCN adsorption [45]. Other sources report a reduction of chromium from the hexavalent to the trivalent state suggesting an oxidation reaction rather than a catalytic process, but no possible reaction schemes were given. However, this reduction to Cr + could explain the fact that carbons loose their capacity against HCN before being saturated with oxamide. As the breakthrough is usually by cyanogen rather than HCN, the limiting factor seems to be the availability of hexavalent chromium. Of course, this could also be explained by a gradual shielding of the chromium salts by deposition of the reaction products, i.e. the oxamide [45]. Clearly it seems that the overall reaction is HCN -> Cyanogen -> Oxamide

(17)

but the individual reactions are less obvious, especially in the presence of water vapour which can have an additional reducing effect on the hexavalent chromium. Probably all cited reactions will take place in real filters, their individual relevance in the overall process depending on many different factors such as HCN concentration, the moisture content of the air stream, temperature, carbon porosity and carbon surface chemistry. For the next generation ASZM whetlerites, in which chromium is replaced by zinc and molybdenum, the overall reaction scheme seems to hold true. The first step (reaction with copper) is identical, and the Zn^/Mo64" combination seems to play a role equivalent to the Cr6" of ASC [53]. Apparently, Zn2+ can also react directly with HCN, however without the formation of (CN)2 that is responsible for early breakthrough in copper-containing whetlerites [52]. Molybdenum would then just play a stabilising role, preventing the Cu2+ from reacting with the HCN rather than the Zn2+. Even though there are experimental data to support this mechanism, it does not explain the presence of oxamide on ASZM carbons after exposure to HCN [53]. It seems that, once again, the exact reaction scheme, and the individual contributions to the general adsorption process, depend highly on the experimental conditions. Finally, a supplementary impregnation with triethylenediamine (TEDA) or other organic amines does not seem to have

500

P. Lodewyckx

any beneficial effect on HCN retention. On the contrary, most experimental results seem to indicate a slight loss in capacity [48]. This can be explained by the loss in physisorption capacity due to pore blocking, and in chemisorption capacity due to a shielding of the copper and chromium ions by the TEDA deposition. 4.3.4.2 Cyanogen chloride Cyanogen chloride is one of the other chemical warfare agents that are only weakly physisorbed by unimpregnated carbons. Comparable to HCN, this is the result of its high volatility. As is the case for HCN, the reactions involved in the retention of CNC1 on whetlerite carbons are still subject to much debate. There is one thing all sources agree upon, which is that the chemisorption of CNC1 is the result of a combined effect of the different impregnated metals. Impregnation with only copper or chromium gives the carbon some capacity, but by far not enough to be considered as a protection against a CNC1 threat. One possible pathway is given by Eq.18 through 20 [66]: CNC1 + whetlerite.H2O -» HOCN + HC1

(18)

2 HC1 + (NH4)2CO3 -> 2 NH4C1 + H 2 CO 3

(19) (20)

This accounts for the experimentally determined liberation of CO2 and the presence of NH4CI on the carbon. The (NH4)2CO3 was present in the Cu/Cr impregnation solution used to obtain the ASC whetlerite (see Table 3). If this reaction scheme is correct, there should be no liberation of CO2 after altering the impregnation procedure, using CuCNOsJi.SHiO as the impregnating solution rather than the carbonate (Eq.21-23) [67]: CNC1 + whetlerite.HiO -> HOCN + HC1

(21)

2 HC1 + Cu(NO3)2.3H2O -> 2 HNO 3 + CuCl2 + 3 H2O

(22)

HNO3 -» ? (Not CO 2 !)

(23)

Experimentally, there was no significant difference between the amount of CO2 released in both cases, suggesting some other reaction scheme rather than the one represented by equations 18 through 20. Another possibility would be a direct catalytic oxidation according to Eq.24: Cu^/Cr^ 2CNC1 + O 2

• CI2+CO2 + N2

(24)

Chlorine could be hydrolysed and follow the normal pathway for CI2 adsorption (see section 4.3.3.2). This would also effectively account for the presence of CO2 but not for the experimentally determined presence of NH3 [61]. It seems highly unlikely that nitrogen would

501

Adsorption of Chemical Warfare Agents

convert into NH3 on the carbon surface. The advantage of this reaction is that it can take place in the absence of water vapour. As it has been experimentally determined that there is indeed a limited adsorption of CNC1 from a dry air stream on a dried activated carbon [50], this reaction cannot be excluded completely. It will probably take place, but only to a limited extend under normal circumstances of humidity and pre-wetting of the carbon bed. In the end, the most probable reaction path is given by the equations 25-27 [40,44,50,61,64,67,68]: CNC1 + H2O HOCN + H 2 O

u

/ur

> HOCN + HC1 (with some formation of HOC1+HCN)

(25)

• NH3+CO2

(26)

2HCl + CuO -> CuCl2 + H 2 O

(27)

When the carbon runs out of available Cu2+, the HC1 could be destroyed according to [50,61]: 6HCl + 2 C r 6 + - » 3 C l 2 + 2 C r ! + + 6 H +

(28)

This would account for the observed change in the ratio Cr +/Cr +. It has been demonstrated by XPS [61,66] that when adsorbing CNC1 this ratio decreases, i.e. Cr + is reduced to Cr 3 ^ but the same studies show there is still a significant amount of Cr6+ left after breakthrough. This is explained by the shielding of the Cr64 by reaction products such as CUCI2 and local polymerisation products (e.g. the formation of C3N3CI3). It seems strange however that chromium would only start to act as a reagent instead of a catalyst after all copper has been "used". Not the least because this would mean that the initial hydrolysis of the CNC1 which, according to the majority of the authors is to be attributed to the Cu^/Cr^ combination, would then stop with the immediate consequence of halting any HC1 production. Neither is there any evidence of a decrease of the Crfr+/Cr3+ ratio when directly adsorbing HC1 or Cl2. Apparently this has never been investigated, although it could shed some light on the issue of chromium reduction due to CNC1 adsorption. Nor does this destruction of HC1 by Cr6+ explain the huge gain in breakthrough time between ASC and AS carbon. Normally there should be sufficient copper present on the AS carbon to give a reasonable retention time, which is not the case. One really needs the chromium to obtain decent CNC1 breakthrough times. On the other hand there is evidence to show that the ageing of ASC whetlerite (see section 6.6.2), which is essentially a reduction of the Crfrt/Cr3+ ratio, is the result of the interaction of the copper-chromium salts with water. So one could suggest that, in the adsorption of CNC1, Cr6+ acts as a pure catalyst, working together with Cuz+ in the hydrolysis of CNC1 to HOCN and subsequently to NH3 and CO2. The reduction of Cr6+ to Cr3+ could then be attributed to either secondary reactions, or to the direct influence of water, exactly as in the case in HCN adsorption (see section 4.3.4.1). As for chromium-free whetlerites, e.g. ASZM-TEDA, there is very little known about the actual processes involved in the CNC1 retention. As there seems to be very little difference in the capacity to remove CNC1 between ASZ (Cu2+, Zn2+) and ASZM (Cu2+, Zn2+, Mo6+) carbons [52], it is safe to suggest that the Zn2+ takes over the tasks of Cr6+, as catalyst,

502

P. Lodewyckx

oxidising agent, or co-catalyst with copper, its most probable role. Molybdenum is only introduced into the impregnation to increase HCN retention times under humid/aged conditions (see section 4.3.4.1) and, apparently, does not interfere with CNC1 adsorption. All whetlerites in use today contain TEDA (triethylenediamine), essentially to enhance the CNC1 capacity under humid conditions. Apparently TEDA works in two different ways, preventing, at least partially, the ageing of the carbon (see section 6.6) and providing a supplementary ehemisorption capacity. As opposed to the Cu2+/Cr6+ combination, this role of TEDA seems to be only marginally affected by the presence of water vapour. The exact interactions between TEDA and CNC1 are still not completely clear, but a reaction analogous to the one proposed for methyl radio iodide (see section 4.3.7) seems highly likely (Eq.29). s

,N+2 CNC1-*C1- [cN + N ^ N + C N ] Cl"

(29)

4.3.4.3 Arsine Arsine (ASH3) is a gas so, like other highly volatile materials, it is hardly retained by unimpregnated carbon. Fortunately it is quite easily oxidised by the catalytic action of copper. According to some sources [40], however, this catalytic activity decreases with increasing air stream humidity. At high humidities silver takes over as the main catalyst. Different oxidation states of arsenic are possible, so the arsenic is mostly converted to a solid mixture of oxides and elemental As which tend to shield the catalysts and reduces their effectiveness. 4.3.5. Neurotoxic agents (see section 2.3.6) Nerve agents are large organic molecules, with high molar masses and very low (VX, Soman, Tabun) to low (Sarin) volatility. Their low values of Cs lead to a very strong physisorption. Therefore, from a purely respiratory protection point of view, nerve agents do not pose a severe threat. As long as one is wearing a well-fitted gas mask, most commercial activated carbon filters will provide plenty of protection. However, the same problem arises as with the vesicants in that the physically adsorbed compounds are hydrolysed under the influence of water vapour (see section 6.4). Except for YX, the decomposition products can be harmful and volatile. For example the hydrolysis of Soman (Eq.30) yielding organophosphorous hydroxide, which is large enough to be well retained by physisorption, and HF which, on the contrary, is very volatile and toxic, and could pass through the activated carbon bed rather rapidly. Fortunately, the standard impregnation of the military carbons (see section 4.2.2) includes Cu-ions and organic amines that will react with HF gas in order to retain it on the carbon. Direct chemical binding in the form of CuF2 will result in the retention of this compound by the carbon filter. (30) O II

CH, CH 3 I I

H3C — p — O — C

— C— CH3+H5O

O II •

H3C — p — O — C

I I F

CHj CHj I I — C —

I I CHi

OH

CH 3

C H 3 + H F

Adsorption of Chemical Warfare Agents

503

In the case of VX, none of the hydrolysis products are volatile so that even under humid conditions physisorption will suffice for protection against VX. 4.3.6. Industrial gases and vapours (see section 2.3.7) 4.3.6.1 Organic vapours Most industrial organic vapours are purely physisorbed. It is however possible that in some cases hydrolysis in the adsorbed state will occur, and there may also be interactions of the vapour or its hydrolysis products with the metal impregnants. These will only enhance the retention on the carbon. One should remember that, in general, carbons for military use (whetlerites) do have a smaller micropore volume than industrial carbons dedicated to physisorption. Compared to "general purpose" industrial activated carbons designed for use against both organic vapours and acid gases, military carbons perform quite well, with a physisorption capacity that is more than equivalent, due to a larger amount of activated carbon per filter. There has been some concern about some highly volatile and toxic organic molecules, so-called "mask breakers", of which the most well known examples are PFIB (2trifluoromethyl-l,l,3,3,3-pentafluoropropene or perfruoroisobutene) and CPFP (or 3-chloropenta-fluoropropene). These vapours are only very weakly retained on unimpregnated carbons, but it has been shown that by a catalytic effect of Cu and/or TEDA, and, probably, some hydrolysis, these compounds are transformed on whetlerites into less volatile chemicals that are sufficiently retained by physisorption [69-74]. 4.3.6.2 Chlorine As in the past chlorine was considered to be a CWA, it has already been treated with the other choking agents (see section 4.3.3.2). 4.3.6.3 Ammonia Ammonia is a major concern for people working with military carbons. Due to its volatility, it is only weakly bound by physisorption, and there is very little indication of any chemisorption or catalytic activity of the copper, chromium or silver salts towards this gas. Some sources [40] suggest a slightly higher uptake on ASC-whetlerite, compared to type AS. This would suggest a limited activity of the chromium. There is some retention of NH3 on military filters, probably as a result of dissolution of the gas into adsorbed water. This is shown in Fig. 12, where the breakthrough time of ASCT carbons is plotted against the pre-wetting of the filter. This pre-wetting was obtained by equilibrating the filter with flowing air at a certain relative humidity prior to the exposure to ammonia. The filter was then challenged with 1600 ppm NH3 in an air stream at the same relative humidity to avoid any exchange of water between the carbon and the air stream. It is clear that the humidity of the filter causes an increase in ammonia adsorption capacity. The drop for very high humidities (> 80%), is probably the result of kinetic effects. At these

504

P. Lodewyckx

values water is filling the pore system, slowing down the physisorption of the ammonia which is a first, and necessary, step towards its dissolution in the adsorbed water [75].

E 40 -

./ •

v

D30-

|

20 H

B 10 m

i

0

25

50

75

100

% Relative Humidity

Fig. 12. Breakthrough time of NH3 on ASC-T carbon as a function of filter pre-wetting. 4.3.6.4 Sulphur dioxide Sulphur dioxide is adsorbed to some extent on non-metal impregnated carbons by physisorption [76-78]. A chemisorption capacity can also exist which depends essentially on the basicity of the surface. A possible reaction scheme is given by a combination of hydrolysis and oxidation (Eq.31 & 32) of the SO2 in the physisorbed state [79,80]. This effect will be even more pronounced when a high number of nitrogen atoms is present on the surface, i.e. when the surface basicity of the carbon increases [81]. Given the impregnation process of whetlerites (see section 4.2.2.1), it is clear that the residual ammonia from the impregnation solution will enhance the capacity of the carbon for SO2 adsorption. The presence of the basic TEDA molecules will reinforce this effect. SO2 + C-[basic group] -)• C-SO2 C-SO2 + H2O + V2 O 2 -> C-H2SO4

(31) (32)

Apparently a direct reaction between the metal salts and SO2 has, as yet, not been documented. However, a reaction similar to the one with other acid gases (see section 4.3.6.5) cannot be ruled out completely. The H2SO4, formed by a combination of hydrolysis and oxidation, is likely to react with the copper according to Eq.33: H z SOt + CuO -» CuSO 4 + H2O

(33)

4.3.6.5 Other acid gases Acid gases such as HC1, H2S and HF will react with the metal salts, especially copper, after an initial weak physisorption on the carbon. The reaction will proceed according to the general scheme of Eq.34 [40,82]:

505

Adsorption of Chemical Warfare Agents

CuO-» CuX2 +H2O

(34)

This reaction is very efficient, as illustrated by the protection time afforded by military filters compared to industrial, even specifically impregnated, carbons. For example, when testing with a concentration of 5000 ppm H2S, at a relative humidity of 70%, European law prescribes a minimum breakthrough time of an industrial filter of 40 minutes (EN 141 - Class B2 [83]). The Belgian military filter, as an example, yields breakthrough times of more than 120 minutes, which is more than double the protection of the industrial filter. Even when taking into account differences in amount of carbon, breathing resistance, mass, etc., it illustrates the efficiency of whetlerites in retaining acid gases. 4.3.7. Radioactive gases (see section 2.3.8) Non-impregnated activated carbons exhibit a low to medium retention against volatile radioactive gases such as CH3I and CHjBr [84,85]. Most industrial filters eliminate radioactive iodine (primarily 131I, but also 132I and 123I) and bromine by ion exchange with non-radioactive I or Br. Usually the activated carbon has been impregnated with potassium iodide (KI, see Table 1). Another possibility is an impregnation with triethylenediamine (TEDA) [40,86-88]. There are several possible reaction schemes, but the most probable one is the formation of a quaternary ammonium salt (Eq.35) [88].

""*" +2CH 3 I^r[ + CH 3 N^^NCH 3 + ]r

(35)

As TEDA is present on all modern types of whetlerite, military filters exhibit a rather good retention of radioactive gases by direct chemisorption. However, one has to bear in mind this is only a retention, not a destruction. This means that the filter itself will become a radioactive source, much like when retaining radioactive aerosol particles (see section 2.3.8), and has to be replaced, and disposed of, as soon as possible. 4.4. Adsorption of water Of course, water is not a chemical warfare agent. However, it plays a major role in the various adsorption mechanisms, either directly, in a benevolent role (e.g. hydrolysis), or indirectly, usually influencing in a negative way the adsorption behaviour of the carbon (see sections 6.4 and 6.5). Therefore it is quite important to have a good understanding of the water adsorption behaviour of carbons in general, and military carbons in particular. This behaviour is characterised by two parameters, the water uptake and the time to reach this uptake. The water uptake depends on the structure and surface chemistry of the carbon. This is illustrated by the water adsorption isotherm (see Fig. 13), that presents a very distinctive shape, especially when compared to the isotherms of other vapours such as chloropicrin (see Fig. 14). Even when

506

P. Lodewyckx

Water uptake (g/gcarbon)

these vapours have a saturation vapour pressure very close to that of water (Ps (chloropicrin) : 27 hPa, Ps (H2O) = 28 hPa). • 0.4 n

Chloropicrin uptake (g/gcarbon)

S

0.5 JS

0.3 H

I£ 0.2-

•

c £ 0.3 0.3

0.2

'5..o> 0.2 •

cc

3.

= 0.1 0.1 •

I.

••» » ••

0.4 0.4-

0.1

O

0

0 0

0.2

0.4

0.6

0.8

1

p/po

Fig, 13. Example of a water isotherm (Type V) on activated carbon

0

0.2

0.4

0.6

0.8

1

p/po

Fig,14. Example of organic vapour isotherm (Type I) on activated carbon

This is the result of the very weak interaction (physisorption) between water vapour and activated carbon [89]. The value of the Dubinin-Radushkevich affinity parameter /?for water is approximately 0.06, much lower than the values for organic vapours which are roughly between 0.5 and 1.5. Basically, the water isotherm of military carbons such as whetlerites can be divided in three parts. The first one is the chemisorption of water vapour by surface groups. These can be oxygen groups (involuntarily introduced during the activation process) or other species, such as impregnated metal salts (copper, chromium, etc.) or organic molecules (TEDA). When sufficient molecules are adsorbed in this way, pore filling of the micropores will start. It is still not very clear how this transition takes place, but pore filling initiated by the formation of water clusters around the primary adsorption sites (hydrogen bonding), the clusters behaving as water in a semi-liquid state, seems a viable pathway [90,91]. Just as for organic vapours, the micropore filling results in a steep rise of the isotherm. Subsequently, after filling the total micropore volume, adsorption will start in the mesopores due to capillary condensation [92]. Another typical feature of water isotherms is the large hysteresis when comparing adsorption and desorption. This is, probably, related to the very weak adsorption forces between water and carbon. But at this time, no satisfactory general explanation has been found for this phenomenon. The second parameter is the time to reach an equilibrium uptake for a given moisture level in the environment. Whereas there is still much debate regarding the exact nature of water adsorption and, consequently, the interpretation of the isotherm, there is even less known about the kinetics of water adsorption. Up to now the only existing models are purely based on experimental data and, as such, they are completely empirical [93-97].

507

Adsorption of Chemical Warfare Agents

4.5. Novel forms of carbon 4.5.1. Industrial-military carbons From the previous sections it is clear that military filters filled with whetlerite carbon retain most industrial vapours and gases at least as effectively as dedicated industrial filters, especially when the supplementary impregnation with TEDA is added. This is also illustrated in Fig. 15, where the breakthrough times for a number of gases of some typical military filters (filled with whetlerites) are compared with the minimum requirements of an industrial filter ABEK-2, tested conform the European Standard EN141 [83]. This norm describes the testing conditions and requirements of the most common types of industrial filters. The data presented are for tests with a flow rate of 30 litres/min, an inlet concentration of 5000 ppmv and a relative humidity of the air stream of 70%. - D ABEK-2 80ASC-T(175g) ASC (100g) -A • S CASC-T - T ( 1 2(120g) 0g ) -A • S CASC-T - T ( 1 7(175g) 5g ) -A • BE K-NB C ABEK-NBC

Breakthrou gh (min)

140 120 100 80 60 40 20 0 Cl2 C12

H2S

HCN

CCl4 CC14

SO2

NH3

C2H4O

Fig.l 5. Comparison between an industrial filter type ABEK-2 and some typical military filters regarding the protection against certain industrial compounds

Fig. 15 shows clearly the weak and strong sides of the whetlerite carbon. ASC carbon will protect rather well against acid gases, SOi and organic compounds, CCU. The supplementary TEDA impregnation enhances this, but only very slightly. The weak points of the whetlerite are ammonia and highly volatile organic compounds such as ethylene oxide. The small capacity against highly volatile organic compounds is logical as these compounds are very weakly physisorbed (see section 4.1.1) and do not react in any way with the impregnants. On the other hand, when taking proper precautions, the risk from these vapours is minimal, the high volatility making the build up of a toxic concentration in free air very unlikely. Proper procedures for entering industrial buildings can therefore eliminate almost completely the risk from these compounds. The main problem is the very low capacity of whetlerites against NH3. Ammonia is, together with chlorine, the most commonly employed toxic chemical throughout the world [98]. Consequently, a low protection against NH3 puts a severe burden on planning the deployment of troops in an industrialized environment, with a non-negligible risk to the health of these troops. Therefore several countries have sought to improve the capacity of the military carbons against ammonia. This way the conventional military filter, offering already

508

P. Lodewyckx

good protection against most industrial gases, will effectively become a combined CWA Industrial Gases filter (also known as ABEK-NBC). Some countries (e.g. France) have indeed developed this kind of filter. There is, however, a downside to this enhanced spectrum of protection. In order to retain a good physisorption capacity, and a good distribution of the impregnants throughout the filter, the amount of activated carbon has to be increased considerably. This will lead to a higher filter mass and an increased breathing resistance through the carbon bed. Both these will have a negative influence on the already high physiological burden of wearing respiratory protection. It has also been shown that in some cases an increase in the mass of the filter will reduce the overall protection factor of the mask - filter ensemble by creating deformations of the rubber mask during use, inducing small leaks around the facial seal. However, many countries and manufacturers pursue this line of research, as the advantages of one single filter being able to protect against both "all" chemical warfare agents and as many industrial gases as possible are evident. Another approach is to incorporate two types of carbon in one filter, either by mixing them, or by putting the two different beds in series in the canister. The obvious advantage of this technique is the possibility to optimise the 'tailoring" of each adsorbent for a given set of gases. In this way it is possible to introduce e.g. both basic sites for SO2 adsorption and acidic sides for NH3-adsorption, without negative interactions. The result are filters that are effective against a very wide range of dangerous substances, from very volatile organic compounds (e.g. ethylene oxide), and different industrial gases (SO2, H2S and even NH3), to Chemical Warfare Agents (CNC1, HCN, AsH3) [99]. Of course, as it is the case for filters filled with "good for everything"-carbons, the main task of obtaining maximum protection for a minimum breathing resistance and a minimum total mass still remains a dilemma. 4.5.2 Carbon monoliths Monoliths are usually cylinders, with a high number of interior channels (see Fig. 16) with a circular or square cross-section. They are either completely made of carbon, or from a skeleton material (such as ceramics), coated with carbon. The surface of this carbon is activated, either by physical or by chemical activation (Activated Carbon Monolith or ACM), The main advantages of this form of activated carbon are the high micropore volume, the very fast adsorption kinetics and the low breathing resistance. The first two are a result of the almost total absence of meso- and macroporosity. All micropores are directly connected to the inner surface of the channels. Indeed, the particular form of the monolith makes it easier to "tailor" the pore size distribution and favour the creation of micropores during the activation process. There is no need to create a complete pore network, as one has to inside a carbon particle. The channels are effectively playing the role of macropores and give an unobstructed access to the activation agents, and, later, to the contaminated air stream. This also results in fast adsorption kinetics as, compared to activated carbon granules, several steps in the diffusion process are absent (e.g. Knudsen diffusion). The low breathing resistance is due to the fact that, contrary to granular activated carbon, the path of the air flow is free of any tortuosity, resulting in an almost perfect laminar flow pattern. Indeed, the flow will essentially

509

Adsorption of Chemical Warfare Agents

Concentration

pass through the large channels and will not have to "find its way" through a bed of granules. A final advantage of activated carbon monoliths is the possibility to eliminate the use of a canister. The monolith is usually cylindrical in shape, but could, theoretically, be manufactured in any given form. This way it could, for example, be integrated in a helmet or another rigid structure. But activated carbon monoliths do also present some major disadvantages which are a direct result of the specific geometry of the channel and pore system. The first of these is leakage. It has been reported on several occasions that, in spite of the fast adsorption kinetics and the high adsorption capacity, there is an immediate breakthrough of the activated carbon monolith filter, but only at a very low concentration of the chemical compound. After the amount of time that one could normally expect from breakthrough models, this is fallowed by a real breakthrough curve. This is illustrated in Fig. 17. The effect is directly related to the flow pattern, i.e. to the effect of the channels. As the flow is (pseudo-)laminar, the carbon monolith can be compared to a cylindrical plug flow reactor. For these reactors it is well known that the residence time of a reagent molecule inside the reactor is a mean value. In reality, there is a spread of residence times, some molecules reacting immediately, others passing through the reactor without reacting at all.

- - Inlet concentration concentration — Outlet concentration • Breakthrough Breakthrough criterion

Time

Fig. 16. General appearance of a carbon monolith

Fig. 17. Breakthrough and leaking of an ACM

Here one can observe a similar effect. Some of the molecules of the chemical compound will get through the monolith, staying in the centre of a channel, without being adsorbed in the micropores on the channel walls. As in military applications one is generally confronted with very toxic chemicals, this is a highly undesirable effect. Usually the leakage concentration, i.e. the amount of the chemical that gets through the monolith without being adsorbed, is higher than the concentration that is deemed to present a health hazard. A possible solution can be to use several, shorter, cylinders placed in series with intervening void spaces in order to make the laminar flow become turbulent in the void spaces, enhancing mixing and reducing chances of molecules passing through the filter unadsorbed. This is shown in Fig. 18. Experiments show that increasing the number of cylinders for a given filter length by increasing the number of void spaces decreases filter leakage. As a result, this method yields better performance of monolith filters without, however, eliminating completely the problem of premature breakthrough [100]. There seems to be an exponential decay relation between the number of voids and leakage, leading to a zero chance of leakage

510

P.Lodewyckx P. Lodewyckx

for an infinite number of voids. This last solution, however, represents nothing else than a granular carbon bed, the granules being very small monoliths. It is clear this solution is not viable, as, apart from the technical feasibility, it eliminates most of the other advantages of monoliths. Another major problem lies in the impregnation procedures for monoliths. From the previous sections it is clear that the adsorption of chemical substances in general, and warfare agents in particular, is the result of a joint process of physisorption and chemisorption. Consequently, in the case of monoliths, the impregnants must be situated inside the channels. The pore entrances of the micropore system are also situated directly in the channels, as there is no meso- and macropore system between these micropores and the surface as it is the case in granular activated carbon. This leads to a disproportionate part of the micropore system being blocked by the impregnants, severely reducing the physisorption capacity of the monoliths, which, in turn, not only leads to a diminished capacity for organic molecules, but also has a negative effect on the chemisorption capacity itself [101]. At this time no satisfactory solution has been found to this problem.

Fig.l 8. Effect of void spaces in a carbon monolith filter 4.5.3. Carbon fibres Another recent development is the use of activated carbon fibres (ACF). These look very promising for use in military applications since they couple a large micropore area to very fast kinetics, nearly all pores being micropores situated directly beneath the surface of the fibre. One would also expect a lower breathing resistance, as the fibres will present more void space for the same physisorption capacity when compared to conventional granular activated carbons. Also, the fibres could fulfill the role of an aerosol filter, eliminating another part of the canister, thereby reducing mass, volume and breathing resistance. However, at this point the future of ACF in military respiratory protection does not seem to be very promising. There are several reasons for this, the first one being the same as that encountered with monoliths. "Whetlerising" commercial ACF gives very poor results, both for the chemical and physical adsorption capacity. Apparently the metal salts do not behave in the same way as on granular carbon, resulting in much shorter breakthrough times for HCN and CNC1 than would be expected on the basis of micropore volume and impregnation level. The micropore capacity, which is initially very high, is dramatically reduced by the blocking of the pore entrances [102, 103]. However, the main problem lies in the void volume between the fibres. Unless the fibres are really pressed together, there is a problem of leakage as some of the gas molecules will find a pathway through the filter with a maximum of void spaces separated by thin, easily saturated volumes occupied by activated carbon fibre. This can be overcome by

511

Adsorption of Chemical Warfare Agents

compressing the fibres, reducing dramatically the voids inside the filter. One way of doing this is pressing the carbon fibres into a felt and placing several layers of felt on top of each other. This way, from a leakage point of view, an ACF filter becomes as good as a granular filter, and the necessary high military protection capacities can be obtained. However, the pressure drop across the filter increases dramatically, to a point where it becomes impossible to breath through it (see Fig.19) [104]. This means that, at the current state of technology, activated carbon fibres are only efficient in respiratory protection when short protection times are required, and merely against organic vapours or very specifically known inorganic gases. One example of this is escape hoods for chemical incidents or fires in which case the fibres will also effectively block the solid and liquid aerosol particles, but for military applications the presently available activated carbon fibres do not provide adequate respiratory protection. An area in which fibres could prove very useful is in protective garments, i.e. body protection. As stated earlier (see section 3.5), protective clothing contains, amongst other things, a layer of activated carbon. Usually this layer consists of spherical particles of granular carbon embedded in a resin or another retaining substance. The use of woven or non-woven activated carbon fibres adds more flexibility to the garment and will reduce the chance of leakage. Indeed, when mechanically disturbed the carbon beads can be pushed apart, giving rise to a carbon-free "channel" towards the inner layers and, eventually, the skin. As the air flow through the garment is less important, and certainly a lot slower, than through a carbon filter leakage and pressure drop are much less of an issue than in the case of respiratory protection. 100

- • - Pressure drop -• • - Breakthrough time

80 80

60

8 e l 4 040- !

60

Breakthrough time 1% hexane (min)

Presure drop at 80l/min (mmH2O)

80

2

"

E

40 40 I S 0

20 20-

20

0

0

0.02

0.06

0.1

Density fibres (g/cm3)

Fig.19. Effect of carbon fibre density on breathing resistance (pressure drop) and organic vapour breakthrough time.

5. MODELLING ADSORPTION 5,1. Chemical engineering models The first group of models that is used to describe adsorption onto activated carbon filters are the so-called chemical engineering models. These consider the filter as a chemical reactor, using the conventional equations for mass and heat transfer to describe the transport processes

512

P. Lodewyckx

through the filter. Even with modern computational techniques, one has to make some assumptions and simplifications to the real phenomena [105-109]: • There is no radial dispersion, i.e. one uses a unidimensional model of a plug flow. This is generally true, as the (usually) cylindrical shape of the filter, in combination with a high flow rate, will result in a negligible radial dispersion. Axial dispersion, however, can not be neglected. • There is no pressure drop over the filter. Of course this is not true (see section 3.6.2), but the influence of pressure drop on the different equations (e.g. by changing the partial pressure values) is very small. • The heat capacities (cp and cv) remain constant. In the range of the measured differences in pressure and temperature this assumption is correct. • The contaminant is passive. This means the contaminant will not change the macroscopic dynamic parameters of the carrier gas (usually air). The limited amount of the toxic compound will not change gas parameters such as viscosity and density, hence one can use the known parameters of dry or humid air to calculate the dynamics of the gas flow. • The filter is considered to be adiabatic, i.e. there is no heat exchange between the activated carbon bed and the environment. This is certainly not true, and some models do take the heat exchange (loss) with the environment into account. However, this is not a feature of the activated carbon bed, but is directly related to the filter housing or canister. • The temperature of the gas and the activated carbon, at a certain spot inside the filter, are always equal. This presumes an infinitely fast heat exchange between the gas (heated by the exothermic reactions) and the carbon. This would also mean that all the carbon downstream from the adsorption front would have the same temperature as the gas in the adsorption front. Experimental evidence shows clearly this is not the case [58]. However, as this heat exchange is very case sensitive, it is in most cases impossible to calculate it correctly. With these provisos, one can rely on the following set of equations: • A mass balance over the filter (mterparticle diffusion) • An energy (heat) balance over the filter • A mass balance over one carbon particle (intraparticle diffusion) • An energy (heat) balance over one carbon particle As one can see, this is a system of coupled differential equations. Some of these balances can even involve more than one equation, e.g. intraparticle diffusion that can be a combination of Knudsen diffusion and surface diffusion. A supplementary equation is needed to describe the adsorption itself. In the case of pure physisorption, this is simply the mathematical description of the adsorption isotherm, e.g. the Dubinin-Radushkevieh equation (see section 4.1.1). In the case of chemisorption, one has to find an equation to describe the interaction between the carbon (and the impregnants) and the specific gas. In most cases equilibrium conditions can be fitted with a Langmuir-type isotherm. A combination of these equations, including a correct set of boundary conditions, will yield a time-concentration profile through the filter. These equations have been verified by measuring the breakthrough

Adsorption of Chemical Warfare Agents

513

curve (concentration at the filter outlet as a function of time). This curve is nothing more than the mirror image of the adsorption front (see Fig.20). Real time in situ measurements with computer tomography [110] have also been used to verify these equations. These models have two main advantages: they simulate the complete breakthrough curve and they are, within the limits of the assumptions and simplifications, physically correct. The first one is not really needed in the case of protection of military personnel since the most important parameter that has to be determined is the breakthrough time (see section 5.2). The complete breakthrough curve is of particular interest in gas separation applications and more complex systems such as PTSA (Pressure and Temperature Swing Adsorption). The main advantage, the physical correctness, of these models comes at a price, as one has to describe, in detail, every physical and chemical phenomenon inside the carbon, and one has to know all the interaction parameters. Some of these, such as the tortuosity, the different diffusion coefficients and any chemical interactions, are difficult to calculate or even to evaluate from a limited number of experiments. As a result, these models that are supposed to be very general can only be used for one very specific adsorbent-adsorbate couple. If either changes, one has to go back to the drawing board, estimating, guessing or determining experimentally the missing parameters. Even though this limits the practical use of these models, their importance should not be underestimated. A detailed comparison between the results of the model and experimental breakthrough curves can yield very important information about the adsorption and transport processes, especially if one is interested in the values and effects of adsorption heats and heat transport through the bed. 5,2. Breakthrough models 5,2,1, Different breakthrough equations For most military applications, and for protective purposes in general, there is no need to simulate the complete breakthrough curve. The point of interest is when the breakthrough curve reaches a certain concentration, the so-called breakthrough concentration. This is a predetermined value, different for each toxic compound. Logically, it is the concentration at which the person wearing the protective equipment starts to experience adverse effects. For industrial chemicals these are normally tabulated values, such as TLV (Threshold Limit Value) or MAK (Maksimale Arbeitsplatz Konzentration), that take into account long term effects on safety and health. For the chemical warfare agents these are typically IDLH-values (Immediate Danger to Life and Health). This is purely a practical approach, as, sadly enough, for many CWA these are the only thresholds that have been established experimentally. Once this value is reached at the filter outlet, the protection is considered compromised and the filter has to be exchanged for a new one. Nearly all these models are based on some sort of a mass balance over the filter. In other words, all other mass and energy balances (see section 5.1) are either ignored, or are not treated separately. Usually these equations, especially the ones related to intraparticle diffusion, only influence adsorption kinetics. The capacity of the activated carbon bed is then

514

P. Lodewyckx

determined by a combination of the mass balance over the filter, combined with an expression defining the maximum uptake per unit of carbon, i.e. the adsorption isotherm. All kinetic terms are usually combined into one or two parameters that "correct" the capacity of the filter in order to give the proper breakthrough time. A very good, and comprehensive, overview of these models and equations can be found in Ref. [111,112].

Concentration

Adsorption front (C) Breakthrough curve (C)

C = C inlet

C=0 Saturated carbon

Virgin carbon

Time & Space

Fig.20. Adsorption front and breakthrough curve for an activated carbon bed 5.2.2. The Wheeler-Jonas equation One of the most commonly used models is the one proposed by Wheeler and Jonas in the early 70s [113,114]. This equation, also known as the Reaction Kinetic equation, can be expressed in several forms. The most explicit one is given in Eq.36:

(36)

Ma, The breakthrough time fe (min) is expressed as a function of the sorption capacity We (ggas per gem-ban), the total mass of carbon in the filter JF(gcart»n), the volumetric flow rate Q (cm3.min"1), the inlet concentration Cm (ggas.cnr3), the bulk density of the carbon in the filter/jj (gcarbon.cm"3), the breakthrough concentration cmt (ggas.cm"3) and the overall mass transfer coefficient kv (min"1). There are some limitations to this equation: • The flow pattern has to be a perfect plug flow with axial, but no radial diffusion. This is normally satisfied when the diameter of the bed is not too small compared to the bed length. For filter systems it is commonly accepted that diameters have to exceed 2 cm to be considered plug flow. • The original equation was based on physisorption into micropores (see further). • The rate constant k, has to be of a first order with respect to the number of gas molecules (= C&,). For pure physisorption this is only true in the first, convex, part of the sigmoidal breakthrough curve, i.e. for values of comlcin < 4 %.

515

Adsorption of Chemical Warfare Agents

The specific arrangement of the different terms clearly shows the rationale of the model: the first term is the total capacity of the carbon (W. Wt) for a given vapour divided by the total amount of vapour entering the filter per unit of time (Q.Cm)- In other words, this is the time the filter would resist if the adsorption was instantaneous and the adsorption front (see Fig.20) would have zero depth. After t = % the concentration at the outlet of the filter would jump from zero to c»,. In reality, the adsorption front has a certain width, and there are vapour molecules in front of the saturated part of the filter (see Fig.20). Consequently, the breakthrough time will be shortened by the second term. This term is of course a function of the reduction factor R (cm.coull\ i.e. the point chosen on the breakthrough curve to define the breakthrough time. But the most important parameter is the overall mass transfer coefficient h,. This factor accounts for all possible resistances against mass transport during the adsorption process. In this way it covers both the interparticle and intraparticle diffusion. The Wheeler-Jonas equation can be used as such to extrapolate experimental data to other circumstances, varying airflow, concentration, bed depth, etc. To do this, We and kv can be treated as fitting parameters and derived from a number of breakthrough experiments. Then, the obtained values can be used to make the necessary extrapolations. However, if one wants to predict breakthrough times for a given filter-toxic vapour system, there has to be a way to calculate We and £„ without any prior breakthrough experiment. For We this problem is not too difficult to solve [115] as the capacity, in the case of physisorption, can be approximated by the static capacity as given by the adsorption isotherm. Accordingly, We can be calculated, rather straightforwardly, from the Dubinin-Radushkevieh equation with c0 = cin (see section 4.1.1):

The only unknown (not tabulated) parameters are Wo and B. These can be derived from any known isotherm of the activated carbon (e.g. N2 at 77K). The affinity coefficient/? can be found in the literature or calculated [35]. Estimating the overall mass transfer coefficient k, is more difficult. It is, a priori, impossible to differentiate between the different diffusion steps in the adsorption process. Therefore h? is estimated on the basis of semi-empirical equations. Up to now, the most complete one is given by Eq.37 [116]:

k, =800# J \rf p L 5 .vf s .p^l M wJ

(37)

In this equation, kv is the overall mass transfer coefficient (min"1), fl is the affinity coefficient of the Dubinin-Radushkevieh equation , dp the mean diameter of the carbon particles (cm), VL the linear velocity of the air stream through the bed (em.s 4 ), We the equilibrium adsorption capacity (gvapour.gcarbon"1) and Aft? the molar mass of the toxic organic

516

P. Lodewyckx

vapour (g.moF1). As such, this equation does not account explicitly for the influence of temperature on the adsorption rate, unlike the equation to calculate We, This is not that big a problem, as the influence of temperature on transport phenomena is known to be far less important than its influence on the adsorption capacity. 5.2.3. Extensions of the Wheeler-Jonas equation The Wheeler-Jonas equation has been used extensively in the case of pure physisorption on granular carbons. But recently, it has been demonstrated it can be equally well applied in a number of very divergent cases. The first extension is on the type of adsorbent, especially new types. The validity of the Wheeler-Jonas equation has been demonstrated [117,118] for both activated carbon fibres (ACF) and activated carbon monoliths (ACM). The equation itself and the calculation of the capacity We remain unchanged. As for the estimation of the overall mass transfer coefficient, the normal equation stays valid, providing a correct interpretation of the "equivalent diameter of the particles". For ACFs, dp has been calculated from the total external surface. Given their small diameter, dp is essentially related to the length of the fibres [119]. For ACMs, dp seems to be related to the internal diameter of the channels. A second effort has been directed towards chemisorbed gases [117]. Here the situation is reversed as in many cases ky can be calculated in the same way as for physisorbed vapours, especially at intermediate to high values of the inlet concentration c(H. This is due to the physisorption, which always precedes chemisorption (see section 4.2), being the rate controlling step in the overall adsorption process. Apparently, in many cases both the hydrolysis and the chemical reactions with surface complexes proceed at a higher rate than the diffusion through the bed and/or through the carbon particles. But it is clear the adsorption capacity We is no longer exclusively related to the available micropore volume of the carbon. In the case of chemisorption, Wt has to be experimentally determined for a given adsorbentadsorbate couple. As such, the Wheeler-Jonas equation can not be used to estimate breakthrough times without any prior experiments, but it can still be used to extrapolate experimental data towards other conditions of airflow, concentration, etc. [120,121], and to evaluate the effect of changing environmental conditions on the adsorption capacity and kinetics. 6. FACTORS INFLUENCING THE ADSORPTION OF CWAs 6.1. General Many factors influence adsorption of chemical warfare agents and toxic industrial chemicals by activated carbon. Some of them have already been mentioned in the previous sections. In this section the nature of this influence will be described and explained, where possible. In most cases this can be illustrated by looking at the effect of changing a particular parameter on breakthrough time predictions. Therefore, where possible, the Wheeler-Jonas equation (Eq.36) will be used to explain the effect of changing a given parameter.

Adsorption of Chemical Warfare Agents

517

6.2. Temperature The influence of temperature has been explained, partially, in section 4: an increase in temperature normally has a negative effect on the adsorption capacity of the carbon [122]. This is true in all cases when dealing with pure physisorption (see Dubinin-Radushkevich equation in 4.1), and in most cases involving chemisorption. However, according to diffusion theories, the overall mass transfer coefficient k¥ should increase with increase in temperature, thus having a positive effect on the breakthrough time, h, should be approximately linearly proportional to the temperature T. Up to now, no predictive equation for kv takes this into account, even though it has been suggested [116]. On the contrary, Eq.40 (see section 5.2.2) shows a proportional dependency of ky on We, and thus an inverse proportionality with T. However, as We is exponentially proportional to T, and k, only linearly, the effect of higher temperatures on breakthrough time will almost always be negative. 6.3. Flow rate and flow pattern 6.3.1. Flow rate Flow rate will influence both the total capacity and the mass transfer ky,, but to a different extent. An increased flow rate will, just as an increase in inlet concentration [123,124], result in a higher quantity of toxic vapour per unit of time being adsorbed by the activated carbon bed. As a result, an increase in flow rate will lower the first term of the Wheeler-Jonas equation. On the other hand, for a same filter geometry, a higher flow rate means a higher linear velocity Vi through the bed [125], which, in turn, leads to a higher k,, faster kinetics, and an increased breakthrough time. Here the effect on both terms of the Wheeler-Jonas equation is more equal than in the case of temperature, the first term being proportional to vi1 and the second term to v£0'75. Overall, as the capacity term usually accounts for more than 75% of fe, an increase in flow rate will reduce breakthrough time. 6.3.2. Flow pattern As explained in section 3.2, real flow patterns in filters differ from the theoretical constant flow. They can be simulated by a half-sine function (see Fig.8). It has been experimentally determined that this flow pattern exerts a negative influence on both the capacity Wt and the mass transfer coefficient ky, resulting in shorter breakthrough times than in the case of a constant flow [126,127]. The first term, We, is only marginally reduced, but the second, ky, is reduced to a much larger extent, which accounts for the apparent absence of this phenomenon in filters with larger bed depths [128-131]. In shallow beds the breakthrough time is influenced by both terms of the Wheeler-Jonas equation, whereas for larger bed depths the high value of the capacity term, We, becomes predominant. The small loss of capacity can be attributed to the fact that, in this specific case, the dynamic capacity does not reach the theoretical equilibrium capacity. As yet, this influence on We can not be calculated or estimated theoretically.

518

P.Lodewyckx P. Lodewyckx

The diminished mass transfer rate can be related directly to the flow pattern via Eq.37, more specifically via the linear velocity vi. When taking into account the real flow pattern, calculated values of h, are typically 20% lower than those calculated for a constant mean flow. Physically, this can be interpreted as a loss in driving force each time no air is flowing over the carbon bed, slowing down the adsorption process and broadening the adsorption front. If the exhaled air passes through the filter, as is the case in some half masks or escape hoods, breakthrough times for organic vapours actually increase [132]. This is a result of (weakly) adsorbed vapours being partly desorbed by the air stream into the ambient air during exhalation. 6.3.3. Intermittent use It has been shown that intermittent use will shorten breakthrough times. Intermittent use is defined as use of the filter for a given time interval, storing it, using it again, storing, etc. Apparently, in some cases, it has been demonstrated that vapours can desorb and re-adsorb inside the bed during storage, broadening the adsorption front [84]. Theoretically, this could lead to a uniform level of contamination throughout the carbon bed, resulting in an immediate breakthrough once the filter is challenged again. This phenomenon is primarily observed with highly volatile (i.e. weakly bound) compounds and very rarely with chemisorbed species. It is also of lesser importance in military applications as in most armies filters are treated as oneuse items and are changed, and disposed of, after any type of real use. 6.3.4. Bed geometry Influence of bed geometry on breakthrough times has also been reported [133]. However, when taking care to keep all other parameters, particularly the amount of activated carbon, constant, the observed differences are very small. They can probably be attributed to local and/or gradual differences in the flow pattern throughout the filter, for example in a conical bed where there will be a depth-dependent linear velocity, resulting in a non-constant k,value. 6.4. Grain size The influence of grain size has already been discussed in points 3.2 (packing of the bed), 3.6.2 (breathing resistance) and 5.2.2 (kinetics of adsorption). In general, appropriate grain size will depend on filter dimensions. Adsorption kinetics are optimized by using the smallest particles as possible, without raising the breathing resistance of the activated carbon bed above an acceptable level. 6.5. Humidity 6.5.1. Effect of humidity Humidity has a marked effect on breakthrough times. This is not in the least due to the abundance of moisture in the ambient air. A relative humidity of 60% at 296K, which is not at

519

Adsorption of Chemical Warfare Agents

all exaggerated and quite normal in most parts of the world, is equivalent to a concentration of approximately 16500 ppmv of water molecules. For a filter, this concentration comes into contact with the activated carbon through the inhaled air stream. For clothing, the transpiration going out will be an additional source of water "contamination" of the activated carbon. Thus even in the absence of any toxic compound, the activated carbon in filters and clothing will be exposed to water vapour. This is often referred to as pre-wetting, or prehumidification, of the carbon bed. When a chemical incident occurs, the toxic compound will be in competition with the water vapour in the air stream, and with the pre-adsorbed water, if any, for available adsorption space. This can be "real" space, i.e. micropore volume in the case of physisorption, or reactive/catalytic species on the surface of the carbon (e.g. Cu ions on whetlerite). On the other hand, one should not forget that in a number of cases the presence of water can reinforce adsorption and, sometimes, is even necessary as a prerequisite for hydrolysis or for certain types of chemisorption (see section 4.3, e.g. Fig. 12). But even then, an abundance of water can have a negative overall effect as it will slow down adsorption kinetics and limit the adsorption space available to other adsorptives. 6,5.2, Influence on adsorption capacity In the case of physisorption, the loss of adsorption capacity can be modelled. The most successful way of doing this is by volume exclusion [134-136]. Where there is water, there cannot be any vapour, and if the vapour is to replace the water, it will have to displace it from the micropore volume it occupies. As can be seen from the theory of physisorption (see section 4.1), adsorption forces depend primarily on the volatility of the compounds. Hence the more volatile the toxic compound, the more it will be influenced by water adsorption as it is unable to replace the more strongly adsorbed water molecules. This can be expressed by equations 38 and 39 [136]: W\ = WB- W^ - AWair + AWS

(38) C

o

)

{

P

W ) \

(39)

AWS < 0 => AWS = 0 The available micropore volume WB' can be calculated from the dry micropore volume, Wo, the amount of pre-adsorbed water, WprB, the amount of water ad- or desorbed by the contaminated air stream, AWm,, and the amount of water replaced by the vapour AWs. The latter being function of the total amount of water on the carbon (Wpm + AWair), the ratio of the amounts of water and vapour in the air stream ([Cw+Co]/Co) and the ratio of their saturation vapour pressures {[Ps&PwVPw}. The value of Wo' thus obtained can be used in the DubininRadushkevich equation (see section 5.2.2) to calculate the capacity (We') under humid conditions.

520

P. Lodewyckx

6.5.3. Influence on adsorption kinetics The water adsorbed on the carbon will also influence adsorption kinetics, Various authors [137,138] have demonstrated this. In contradiction to the capacity which can be influenced positively or negatively (see section 6.5.1), mass transfer will always slow down. The overall mass transfer coefficient, £w of the Wheeler-Jonas equation does not differentiate between the different types of diffusion. Consequently, every impact of water on the adsorption kinetics will be translated into a drop afk? values.

k,:=t, 1-

(40)

TPV

This is illustrated by equation 40 [137], which expresses the mass transfer coefficient under humid conditions (£„*) as a function of ky'm the absence of water, the total pore volume {TPV) of the carbon and the total amount of water adsorbed. 250 -

• Carbon Carbon pre-humidifation pre-humidifation = 0%RH 0%RH • Carbon Carbon pre-humidifation pre-humidifation = 90%RH 90%RH O Model MnHpl \J IVIULJCI

D Model Model

200 )

o

i

•

Breakthrough time(min) Bre ;thr

-£•

E "of

O

!! 150 -

•

°

^

8

8

O) 3

O

100 -

(S

• 50 -

0

—i

0

1

1

u

B-

20 40 60 80 Air stream humidity (%RH)

100

Fig.21. Theoretical vs Experimental breakthrough times for CC14 on BPL-type carbon under varying humidity conditions. Clearly, it is not only the presence of water in the micropore system that plays a role (as it is the case for the physisorption capacity), but also in the rest of the pore system [137139]. It has been verified that even in cases where water has a beneficial effect on capacity (e.g. CNC1 [140]), k, will decrease for high water loading. Probably, the adsorbed water

Adsorption of Chemical Warfare Agents

521

hampers transport of the toxic compound to the adsorption sites (micropore volume or reactive sites) and, when there is replacement of water by the vapour, it will also impede this water from exiting the carbon bed. A model based on the loss of physisorption capacity and diminished mass transport can adequately describe the breakthrough of organic vapours, even under severe conditions of carbon pre-wetting and moist air. This is illustrated in Fig.21 [141]. 6.6. Ageing 6.6.1. Influence on physisorption capacity The prolonged exposure of activated carbon to moist air is known as ageing. It was first witnessed for whetlerite carbons as it can have a very marked effect on the retention of chemical warfare agents, especially CNCl (see section 6.6.2). Evidence of changes in the physisorption capacity, i.e. in the pore structure of the carbons, is inconclusive. Some groups have found evidence of an increase in micropore volume after ageing, essentially due to a widening of the smallest pores, whereas others did not observe any pore structure related changes [142,143]. In any case these changes, if any, are small and of the same order of magnitude as the experimental error in measuring micropore volumes and pore size distributions. 6.6.2. Influence on chemisorption capacity Contact with water vapour, especially over long periods of time, will change the composition of the surface of the carbon: it will oxidise this surface, introducing additional oxygencontaining complexes, such as carboxylic groups. Most of these groups are hydrophilic and will enhance water vapour adsorption [144]. This effect is permanent, even when the carbon is subsequently dried [142,143]. The result is an increased sensitivity of the carbon towards water vapour, i.e. a higher uptake for a given partial water pressure (i.e. air stream humidity) and faster water adsorption kinetics, which, in turn, will lead to a lower physisorption capacity under humid conditions (see section 6.5). Not only is the carbon surface affected, but also other chemical structures can be oxidised by ageing, specifically the copper and chromium ions on whetlerites. In this case, ageing reduces the individual metals from Cu2+ to Cu1+ and from Cr6*" to Cr3* [49,145-152]. The exact reactions involved are still not completely understood, but the effects are reinforced by higher temperatures [147,153]. The reduction of copper makes the whetlerite carbons less effective in almost all chemisorption reactions (see section 4.3), and in conjunction with the reduction of chromium, it has a disastrous effect on both HCN and CNCl adsorption mechanisms. In particular, breakthrough times for the latter component (CNCl) are drastically reduced, usually to values well below any acceptable threshold. New whetlerites such as ASZM also suffer from ageing effects, even though the exact mechanisms are even less well known than in the case of the conventional ASC type whetlerites [52], It is interesting to note the particular behaviour of TEDA on whetlerites. The presence of TEDA seems to reduce considerably the effects of ageing. It is commonly

522

P. Lodewyckx

assumed that TEDA, in some yet unknown way, shields the copper-chromium complexes from the degradation by water vapour. More importantly, the TEDA itself seems not to be affected at all by ageing. It keeps its full capacity to chemisorb CNCl (see section 4.3.4.2) and takes over part of the task that is usually performed by the Cu-Cr complexes. 6.7. Multi-component adsorption Multi-component adsorption can be described by some of the more sophisticated simulation programs (see section 5.1) [154], but it is still a difficult task especially when more than two different vapours or gases are involved. In general, the effects of multi-component organic vapour adsorption are very similar to the case of vapour-water co-adsorption explained in section 6.5. More strongly adsorbed, i.e. less volatile, vapours will win the competition for the available sorption space from the more volatile ones. Taking the co-adsorption of two vapours as an example, experiments have shown the breakthrough time for the less volatile compound to remain more or less unchanged, usually a bit shorter than for the single vapour due to the slower kinetics. For the more volatile vapour, breakthrough will be much shorter than in the case of this vapour alone. There will also be the effect of "rollup", as the more strongly adsorbed vapour will replace the already adsorbed, weakly retained vapour, increasing the outlet concentration of the latter to well above its concentration in the contaminated air. This is illustrated in Fig.22. Of course, things get even more complicated when some of the compounds are physisorbed and others are primarily chemisorbed. 25

Vapour A = more volatile

20

Vapour B = less volatile Concentration

I

16

15

Inlet concentration 1 AA B-

O 10 10 •

5

0 0

10

20

30

40 A 40

50

60

Time (min)

B A Single vapour breakthrough time Fig.22. Multi-component organic vapour adsorption: rollup effect

Whereas this phenomenon is quite common in industrial applications, it is not really relevant to the military use of activated carbons, except, of course, the case of water as explained in the previous sections. It is indeed unlikely that troops in operations will be subjected to mixtures of chemical warfare agents, and even more unlikely that in such

Adsorption of Chemical Warfare Agents

523

instances the CWAs will have very different volatilities. Sometimes toxic chemicals are mixed with tear or vomit agents, but as most of these are distributed as aerosols there is little or no interaction on the activated carbon. Only when confronted by industrial threats could this become a real problem. Even in this case, since most military canisters exhibit a high to very high physisorption capacity and the troops will normally be in the open, multicomponent adsorption and rollup phenomena are not really an issue. 7. CONCLUSIONS The protection against Chemical Warfare Agents is a very small, but highly specialised field amongst activated carbon applications. Even after several decades of intensive use and elaborate research, many aspects of the military carbons, most of which are known as whetlerites, remain unclear. Much as is the case for other applications, activated carbon provides an adequate answer to a wide variety of problems, but raises a lot of questions as to the how and why of its excellent performance. This will, however, not prevent it from continuing to play its role as the first line of defence against the threat of CWAs hi the near and even in the not so near future.

REFERENCES [1 ]

[2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

Convention on the prohibition of the development, production, stockpiling and use of chemical weapons and on their destruction. Organisation for the prohibition of chemical weapons (OPCW), 1993 (Updated 2004). Sadayoshi Ohbu, Akira Yamashina, Nobukatsu Takasu, et al., South. Med. J., 90 (1997) 587. H. Morita, N. Yanagisawa, T. Nakajima, et al., Lancet 346 (1995) 290. Journal "L'illustration", n° 3801 (1916) 44. Journal "J'ai vu ", n° 56 (1915) 783. S. Franke et al. Lehrbuch der Militarehemie [Textbook of military chemistry] Vbl 1. Militlrverlag der Deutschen Demokratischen Republik, East Berlin, 1967. IUPAC, Pure Appl. Chem. 74 (2002) 187. Boris Plotnikoff. Private Collection of gas masks. Mask ARFA(GIAT)- publicity folder. History of the Army's Protective Mask, US Army Soldier and Biological Chemical Command, Aberdeen Proving Ground, MD (1996). Journal "J'ai vu ", n° 86 (1916) 435. BI-CK-3000B: Technische "Notitie ~ Filterdozen voor gasmaskers [Technical note ~ Filter canisters for gas masks]. Belgian Defence (1997). European Standard EN143: Respiratory protective devices - Particle filters - Requirements, testing, marking, European Committee for Standardization (February 2000). C. Willis, S. Brewer, C. Stone, M. Dennis, and B. Beadle, Proc. of the 8* Int. Symp. on Protection against CBWA, Goteborg (Sweden) (2004). L. Pleister, Proc. of the 8th Int. Symp. on Protection against CBWA, Goteborg (Sweden) (2004).

524

P. Lodewyckx

[16] E. Wilusz and W.X. Zukas, Proc. of the 8* Int. Symp. on Protection against CBWA, Goteborg (Sweden) (2004). [17] U. Danielsson and J. Keinanen, Proc. of the 5th Int. Symp. on Protection against Chemical and Biological Warfare Agents, Stockholm (Sweden)(1995) 157. [18] S. Etienne, B. Melin, J.Y. Pelicand, P. Dziedzinl, A. Charpenet, N. Zeinou and B. WarmeJanville, Proc. of the 5 Int. Symp. on Protection Against Chemical and Biological Warfare Agents, Stockholm (Sweden) (1995) 163. [19] M.G Katz, Proc. of the 3rd Int. Symp. on Protection Against Chemical Warfare Agents, Umea (Sweden) (1989) 25. [20] H. Axtell and T. Pease, Proc. of the 8th Int. Symp. on Protection against CBWA, GSteborg (Sweden) (2004). [21] H.L Schreuder-Gibson, J.E Walker, W. Yeomans and D. Ball, Proc. of the 8* Int. Symp. on Protection against CBWA, Gateborg (Sweden) (2004). [22] C. Eon and M. Degranges, Proc. of the 1st Singapore Int. Symp. on Protection against Toxic Chemicals, Singapore (Singapore) (1998) 51. [23] J. Davey, Proc. of the 4* Int. Symp. on Protection against Chemical Warfare Agents, Stockholm (Sweden) (1992). [24] D. Kelm, UK J. Defence Science 1(1996) 406. [25] C, Carton, E, Degrave and J. Willems, Proc. of the 6* Int. Symp, on Protection against Chemical Warfare Agents, Stockholm (Sweden) (1998). [26] A. Bailey, In: Patrick J.W. editor. Porosity in Carbons, Edward Arnold, London (1995). [27] G,P. Smith, Edgewood Arsenal Technical Report EM-TR-76093, Aberdeen Proving Ground MD, USA (1977). [28] J.E. Roehl, T.W. Caraher, K.A. Kalmes and E.A. Isley, Proc. of the 1st Joint Conf. on Point Detection for CB Defense, Williamsburg VA (USA) (2000). [29] E.S. Mayer, M.W. Findlay, G. Maclay and J. Stetter, Am. Ind. Hyg. Assoc. J. 54 (1993) 417. [30] M.J.G. Linders, E.A. Bal, P J. Baak, J.H.M. Hoefs and J.J.G.M. van Bokhoven, Proc. of the Int. Carbon Conf. 2001, Lexington KY (USA) (2001). [31] P. Gardner, D. Tevault and R.A. McGill, Proc. of the 8th Int. Symp. on Protection against CBWA, Goteborg (Sweden) (2004). [32] P. Gardner, Short Abstract, 24* Army Science Conf, Orlando FL (USA) (2004). [33] V.R. Deitz, J.N. Robinson and E.J. Poziomek, Carbon 13 (1975) 181. [34] M.M. Dubinin, Chem. Rev. 60 (1960) 235. [35] G.O. Wood, Carbon 39 (2001) 343. [36] G.O. Nelson and C.A. Harder, Am. Ind. Hyg. Assoc. J. 35 (1974) 391. [37] H.F. Stoeckli, P. Rebstein and L. Ballerini, Carbon 28 (1990) 907. [38] R.H. Bradley, Proc. of the Int. Carbon Conf. 1990, Paris (France) (1990) 660. [39] K-D. Henning and S. Schafer, Gas Sep. Purif. 7 (1993) 235. [40] D.T. Doughty, J.E. Groose and S.L. Liller, Literature Review n°20, Calgon Carbon Corporation (1986). [41 ] F. Rodriguez-Reinoso, M. Molina-Sabio and V. Perez, Proc. of the Int. Carbon Conf. 1994, Granada(Spain) (1994) 258. [42] M. Molina-Sabio, V. Perez and F. Rodriguez-Reinoso, Carbon 32 (1994) 1259.

Adsorption Chemical Warfare Agents Adsorption ofofChemical Warfare Agents

525525

[43] P. Marchand, Internal Report R88/02, Belgian Defence, TDLM/CT-NBC, Vilvoorde-Peutie, Belgium (1988). [44] P.N. Krishnan, S.A. Kate, A. Birenzvige and H. Salem, Carbon 26 (1988) 914. [45] P.N. Brown, G.G. Jayson, G. Thompson and M.C. Wilkinson, Proc. of the Int. Carbon Conf. 1990, Paris (France) (1990) 42. [46] P.N. Brown, G.G. Jayson, G. Thompson and M.C. Wilkinson, J. Colloid Interface Sci. 116 (1987)211. [47] J. Kloubek, Collect. Czech. Chem. Commun., 47 (1982) 2368. [48] B.R. Alves and M.P. White, Laboratory Report n" M875/82, CDE Porton Down, UK (1982). [49] M.M. Ross, R.J. Colton and V.R. Deitz, Carbon 27 (1989) 492. [50] E.J. Poziomek, R.A. Mackay and R.P. Barrett, Carbon 13 (1975) 259. [51] Y. Suzin, L.C. Buettner andCA. LeDuc, Carbon 37 (1999) 335. [52] R.W. Morrison and D.T. Doughty, NATO Defence Research Group 25th anniversary celebration lectures on the Future of Defence Research, Imperial College of Science & Technology, London (UK) (1992). [53] V.R. Deitz and CJ. Karwacki, Carbon 32 (1994) 703. [54] V.R. Deitz and CJ. Karwacki, Proc. of the Int. Carbon Conf. 1993, Buffalo NY (USA) (1993) 464. [55] V.R. Deitz and CJ. Karwacki, Carbon 31 (1993) 237. [56] W.A. Waters and J.H. Williams, J. Chem. Soc. (1950) 18. [57] C.T. Chiou and P. J. Reueroft, Carbon 15 (1977) 49. [58] L. Verhoeven, D. Van Rompaey and P. Lodewyckx, Proc. of the Int. Carbon Conf. 2001, Lexington KY (USA) (2001). [59] P. Lodewyckx and L. Verhoeven, Carbon 41 (2003) 1215. [60] P. Lodewyckx, L. Verhoeven, D. Van Rompaey, E.F. Vansant, A. Lavanchy, A. Wittwer and M. Gurtner, Proc. of the Int. Carbon Conf. 2000, Berlin (Germany) (2000). [61] E. Biron and R. Stavisky, Carbon 33 (1995) 1413. [62] A. Birenzvige, G.K. Knoss and P.N. Krishnan, Carbon 28 (1990) 249. [63] V.R. Deitz, L.C. Cullen and J.S. Murday, CROC, US Chemical Weapons Conf. (1983). [64] B.R. Alves and AJ. Clark, Carbon 24 (1986) 287. [65] V.R. Deitz and J.A. Rehrmann, Carbon 28 (1990) 387. [66] V.R. Deitz and J.A. Rehrmann, Carbon 28 (1990) 363. [67] R. Willaert, Internal Report, Belgian Defence, TDLM/CT-NBC, Vilvoorde-Peutie, Belgium (1991). [68] E. Biron and R. Stavisky, Proc. of the Int. Carbon Conf. 1995, San Diego, CA (USA) (1995) 490. [69] CJ. Karwacki and G.A. Stickel, Carbon 29 (1991) 1292. [70] M.C. Biesinger, N.S. Mclntyre, S J. Splinter and S. Liang, Carbon 35 (1997) 463. [71] MJ. Chinn, I.W. Lawston and M.W. Smith, Proc. of the Int. Carbon Conf. 1994, Granada (Spain) (1994) 282. [72] C.R. Hall, I.W. Lawston and A.M. Tinsley, Chem. Ind. 6 (1989) 145. [73] P. Genin, R. Fangeat, P. Dziedzinl and M.V. Mathieu, Proc. of the Int. Carbon Conf. 1990, Paris (France) (1990) 58.

526

P. Lodewyckx

[74] M.J. Chinn, I.W. Lawston, M.W. Smith and D. Stanton, Proc. of the Int. Carbon Conf. 1994, Granada (Spain) (1994) 218. [75] L. Verhoeven and P. Lodewyctac, Proc. of the Int. Carbon Conf. 2001, Lexington KY (USA) (2001). [76] Z.W. Wang and K. Kaneko, J. Phys. Chem. 99 (1995) 16714. [77] M. Molina-Sabio, M.A. Mufieeas, F. Rodriguez-Reinoso and B. McEnaney, Carbon 33 (1995) 1777. [78] P.G. Gray, Gas Sep. Purif. 7 (1993) 213. [79] I. Mochida, S. Kisamori, Y, Kawabushi, K, Kuroda and S. Kawano, Proc, of the Int. Carbon Conf. 1994, Granada (Spain) (1994) 270. [80] J.A. DeBarr and A.A. Lizzio, Proc. of the Int. Carbon Conf. 1994, Granada (Spain) (1994) 268. [81] E. Raymundo-Pifiero, D. Cazorla-Amoros and A. Linares-Solano, Proc. of the Int. Carbon Conf. 2001, Lexington KY (USA) (2001). [82] S.R. Choi, S.K. Ryu, S.J. Park and D.D. Edie, Proc. of the Int. Carbon Conf. 2004, Providence RI (USA) (2004). [83] European Standard EN141; Respiratory protective devices - Gas filters and combined filters Requirements, testing, marking, European Committee for Standardization (February 2000). [84] F.A.P. Maggs and M.E. Smith, Ann. Occup. Hyg. 18(1974) 111. [85] F. Billard, G. Chevalier, J. Caron and J. Van Der Meersoh, Rapport CEA-R 2908, Centre d'Etudes Nucleates, Fontenay-aux-roses, France (1966). [86] G.O. Wood, Am. Ind. Hyg. Assoc. J, 42 (1981) 570. [87] G.O. Wood, Am. Ind. Hyg. Assoc. J. 46 (1985) 251. [88] S. Kitani, T. Noro and T. Kohara, J. Nucl. Sci. Technol. 9 (1972) 197. [89] P. Lodewyckx and E.F. Vansant, Carbon 37 (1999) 1647. [90] T. Iiyama, M. Ruike and K. Kaneko, Chem. Phys. Lett. 331 (2000) 359. [91] J.K. Brennan, T.J. Bandosz, K.T. Thomson and K.E. Gubbins, Colloids Surf., A 187/188 (2001) 539. [92] P. Lodewyckx, D. Van Rompaey, L. Verhoeven and E.F. Vansant, Carbon 39 (2001) 309. [93] V.R. Deitz, Proc. of the Int. Carbon Conf. 1991, Santa Barbara CA (USA) (1991) 68. [94] P. Lodewyckx and E.F. Vansant, Carbon 35 (1997) 310. [95] P. Lodewyckx and E.F. Vansant, Carbon 36 (1998) 304. [96] P. Lodewyckx, E.R. Vansant, P.R. Norman and L. Pears, Proc. of the Int. Carbon Conf. 1997, Perm State PA (USA) (1997). [97] N.J. Foley, K.M. Thomas, P. Forshaw, D. Stanton and P. Norman, Langmuir 13 (1997) 2083. [98] NATO Document AC/225(LG7)D(2004)0005(INV) [NATO Confidential] (2004). [99] C.J. Karwacki, Proc. of the 8th Int Symp on Protection against CBWA, Goteborg (Sweden) (2004). [100] M.J. Chinn, P.T. Green, H.E. Keel and M.W. Smith, Proc. of the Int. Carbon Conf. 2001, Lexington KY (USA) (2001). [101] M.J. Chinn, CJ. Hindmarsh and L.A. Pears, Proc. of the 2nd Symposium on Carbon Adsorbents: Porosity, Surface Chemistry & Use, Aberdeen (UK) (2004). [102] Lj.V. Rajakovic, M.R. Ilic, P.B. Jovanic and P.B. Radosevic, Carbon 33 (1995) 1433. [103] H. Jin, S.Y. Jeong, J.K. Suh, O.Y. Kwon, S.J. Park, J.M. Lee and S.K. Ryu, Proc. of the Int. Carbon Conf. 1995, San Diego CA (USA), (1995) 488.

Adsorption of Chemical Warfare Agents

527

[104] C.J. Hindmarsh and P.L. Phillips, Proc. of the Int. Carbon Conf, 2004, Providence RI (USA) (2004). [105] M.J.G. Linders, TU Delft (The Netherlands), PhD thesis (1999). [106] P. Lodewyckx, Katholieke Universiteit Leirven (Belgium) - FTW-CIT, MSc thesis (1991). [107] A. Lavanchy, R. Touzani and M. Stoekli, Proc. of the Int. Carbon Conf. 1991, Santa Barbara CA (USA) (1991) 38. [108] U. Huber, Bericht TA-8-SIG Wo/Le nr 8143, GRD AC-Labor, Spiez, Switserland (1981). [109] R.H. Van Dongen and P.C. Stamperius, Report 1974-4, Chemisch Laboratorium, TNO, Rijswijk (The Netherlands) (1974). [110] A. Wittwer and A. Lavanchy,, Proc. of the Int. Carbon Conf. 1990, Paris (France) (1990) 102. [111] E. Balieu, Proc. of the Int. Symp. on Gas Separation Technology, Antwerpen (Belgium) (1989) 91. [112] M. Van Zelm and J. Medemal. In: Protection against toxic compounds - chemical and technological aspects. Chemical Laboratory TNO, Rijswijk, The Netherlands (1973). [113] L.A. Jonas and J.A. Rehrmann, Carbon 10 (1972) 657. [114] L.A. Jonas, Soc. Chem. Ind. (1974) 181. [115] G.O. Wood, Carbon 30 (1992) 593. [116] G.O. Wood and P. Lodewyckx, Am. Ind. Hyg. Assoc. J. 64 (2003) 646. [117] P. Lodewyckx, G.O. Wood and S.K. Ryu, Carbon 42 (2004) 1345. [118] C. Martin, J-P. Joly and A. Perrard, Proc. of the Int. Carbon Conf. 2003, Oviedo (Spain) (2003). [119] P. Lodewyckx and S.K. Ryu, Proc. of the Int. Carbon Conf. 2002, Beijin (PRC) (2002). [120] N. Koenig, Internal Report WWDBw fuer ABC-Schutz, Munster, Germany (1979). [121] B. Staginnus, Proc. of the Int. Carbon Conf. 1982, London (UK) (1982) 228. [122] G.O. Nelson andN.A. Correia, Am. Ind. Hyg. Assoc. J. 37 (1976) 514. [123] G.O. Nelson and CA. Harder, Am. Ind. Hyg. Assoc. J. 37 (1976) 205. [124] CJ. Littleton and J.R. Feeney, J. Int. Soc. Resp. Prot. Winter (1991-1992) 6. [125] J.A. Rehrmann and L.A. Jonas,. Carbon 16 (1978) 47. [126] Y. Suzin, I. Nir and D. Kaplan, Carbon 38 (2000) 1129. [127] I. Nir, Y. Suzin and D. Kaplan, Carbon 40 (2002) 2437. [128] M.J.G. Linders, E.P.J. Mallens, J.J.G.M. van Bokhoven, F. Kapteijn and J.A. Moulijn, Am. Ind. Hyg. Assoc. J. 64 (2003) 173. [129] J.J.G.M. van Bokhoven and H. Jager, Proc. of the 4th Int. Symp. of Protection against Chemical Warfare Agents, Stockholm (Sweden) (1992) 253. [130] R. Fangeat, Journee d'etudes "Materiaux adsorbants earbones: Caracterisation et Applications [Carbonised adsorbing materials: characterisation and applications]", Centre d'Etudes du Bouehet, Vert-le-Petit (France) (1991). [131] G.O. Nelson and CA. Harder, Am. Ind. Hyg. Assoc. J. 33 (1972) 115. [132] M. Jovasevic, B. Stojanovic, M. Polovina and B. Kaludjerovic, Proc. of the Int. Carbon Conf. 1995, San Diego CA (USA) (1995) 492, [133] J.J.G.M. van Bokhoven. In: Groothuizen ThM, editor. TNO Prins Maurits Laboratory - Diverse and Dynamic (1988). [134] M. Manes, Ext. Abstracts Engineering Foundation Conf. on Fundamentals of Adsorption, Schloss Elmau (Germany) (1983) 335,

528

P. Lodewyckx

[135] RJ, Grant, R.S. Joyce and J.E. Urbanic, Ext. Abstracts Engineering Foundation Conf. on Fundamentals of Adsorption, Schloss Elmau (Germany) (1983) 219. [136] P. Lodewyckx and E.F. Vansant, Am. Ind. Hyg. Assoc. J. 60 (1999) 612. [137] P. Lodewyckx and E.F. Vansant, Am. Ind. Hyg. Assoc. J. 61 (2000) 461. [138] G.O. Wood and P. Lodewyckx, J. Int. Soc. Resp. Prot. 19 (2002) 58. [139] R.C. Hall and RJ. Holmes, Proc. of the Int. Symposium on Gas Separation Technology, Antwerpen (Belgium) (1991) 231. [140] P. Lodewyckx and H. Wullens, Proc. of the Int. Carbon Conf. 2004, Providence RI (USA) (2004). [141] P. Lodewyckx, University of Antwerpen (Belgium), PhD thesis (1998). [142] A. Wittwer and A. Lavanchy, Proc. of the 7* Int Symp on Protection against Chemical and Biological Warfare Agents, Stockholm (Sweden) (2001). [143] A. Wittwer, A. Lavanchy and M. StOckli, Labornotiz SI 2002.01, Labor Spiez, Fachsektion ABC-Schutz, Spiez, Switserland (2002). [144] L.B. Adams, C.R. Hall, RJ. Holmes and R.A. Newton, Carbon 26 (1988) 451. [145] P. Ehrburger, J. Lahaye, P. Dziedzinl and R. Fangeat, Carbon 29 (1991) 297. [146] Ch. Leclercq and M.V. Mathieu, C.R. Acad Sc Paris, Tome 304, Serie II, n°14 (19S7) 793. [147] Nurcan Bac, J.L. Hammarstrom and A. Sacco Jr, Carbon 25 (1987) 545. [148] J. Rossin, E. Petersen, D. Tevault, R. Lamontagne andL. Isaacson, Carbon 29 (1991) 197. [149] P. Ehrburger, J. Lahaye, P. Dziedzinl and R. Fangeat, Proc. of the Int. Carbon Conf. 1989, State College PA (USA) (1989) 36. [150] J.L. Hammarstrom and A. Sacco Jr, J. Catal. 112(1988) 267. [151] N.S. Mclntyre, G.R. Mount, T.C. Lipson, R. Humphrey, B. Harrison, S. Liang and J. Pagotto, Carbon 29 (1991) 1071. [152] P.N. Brown, G.G. Jayson, G. Thompson and M.C. Wilkinson, Carbon 27 (1989) 821. [153] P. Lodewyckx, Royal Military Academy, Brussels (Belgium), MSc thesis (1985). [154] Xijun Hu, G.N. Rao and D.D. Do, Gas Sep. Purif. 7 (1993) 39.

Activated Carbon Surfaces in Environmental Remediation T.J. Bandosz (editor) © 2006 Elsevier Ltd. All rights reserved.

529

Activated carbons as medical adsorbents S. V. Mikhalavsky" and V. G. Nikolaevb a

University of Brighton, Lewes Road, Brighton BN2 4GJ, UK

R.E. Kavetsky Institute of Experimental Pathology, Oncology and Radiobiology, Nat. Acad. Sri., 45, Vasilkovskaya Str., 03022, Kiev, Ukraine 1. INTRODUCTION Activated carbons (AC) have been used in medicine since ancient times. An excellent description of findings on oral use of AC from ancient Egypt up to early 1990s is given in the book 'Activated Charcoal in Medical Applications' by D. Cooney [1]. To avoid repetition, in this chapter we would mainly refer to this source for works published before 1994 with one notable exception, hi Ch. 21 of his book entitled 'Reports from the Soviet Union' Cooney mentioned works carried out in the then Soviet Union as those that require treatment '...with extreme caution' because '...the results are so striking that independent confirmations ... are needed' [1]. One of the paragraphs in that very short chapter (less than five full pages) had a title 'Enterosorption1 studies described by Nikolaev', referring to the review [2], one of the very few papers on the subject of medical use of oral activated carbons published at the time by the Soviet scientists in English. Since one of the authors of this chapter is the same Nikolaev, we have decided to describe results of research on medical AC carried out in the USSR in more detail. Much of that work is relevant to the main theme of this book and particularly work aimed at protection of the population affected by the Chernobyl accident and the personnel (known in literature as 'liquidators', in the sense of liquidation of the consequences of this infamous nuclear disaster) who worked on the affected territory. hi a broad sense, the medical applications of activated carbons are based on their powerful non-specific adsorption capacity unrivalled by any other material. They are used to remove undesirable and harmful substances - toxins - from the human body. These substances either enter the human body from the external environment via skin, eyes, breathing airways or with food and drinks, or they can be produced internally by the body itself due to organ malfunction, autoimmune diseases, infection or trauma. Considering ourselves as an inseparable part of the environment, both externally and internally generated toxins cause nothing else but pollution of our body, similar to the environmental pollution, which therefore has to be

1 'Enterosorption' is the term often used in Russian literature instead of oral adsorption, and 'haemosorption* is used instead of 'hEemoperfusion'.

530

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

decontaminated2. We limit our review to medical applications of activated carbon in the treatment of intoxication and poisoning caused by external (exogenous) events. These include AC for oral administration, which is classified as a medicine and an extracorporeal AC-filled haemoperfusion column for blood purification which is classified as a medical device. A gas mask containing AC is an example of a technical device which also protects human body from exogenous poisoning, but it is not a medical device and therefore is not covered in this chapter. 2. MODES OF MEDICAL USE OF AC AC is used internally as an oral adsorbent or externally - in extracorporeal devices such as haemoperfusion columns and materials such as wound bandages. The whole extracorporeal unit is a complex engineering system which comprises equipment controlling the blood flow, preventing air bubble formation and additional lines feeding supplementary nutrients and drugs into the blood circulation. A haemoperfusion column is a relatively simple device in which a plastic column is filled with AC. The word 'haemoperfusion' indicates that such a column is used for direct blood purification; the blood passes through a layer of granulated AC and then returns to the patient (Fig. 1). In some more sophisticated extracorporeal systems AC is used in suspension [3]. 2.1. What does the word 'medical" mean? In simple terms 'a medical grade' AC should at least obey the dictum of 'First, do no harm' 3 [4]. To satisfy this statement, AC has to be biocompatible and in the case of direct contact with blood, haemocompatible. The phenomenon of bio/haernocompatibility is very complex and still poorly understood, thus making a biomaterial still relies more on the trial and error approach rather than firm knowledge. This state of our knowledge is reflected in the definition of biocompatibility and formulating criteria, which the biomaterial should satisfy. One of the latest definitions of biocompatibility formulates it as '...the ability of a material, device, or a system to perform without a clinically significant host response in a specific application' [5]. Respectively, the definition of a biomaterial given by an authoritative group of experts, states what the biomaterial should NOT DO rather than what it should do: - absence of thrombogenic, toxic, allergic or inflammatory reactions; - no damage of blood cells; - no changes in plasma proteins and enzymes; - no immunological reaction; - no carcinogenic effect; - no deterioration of adjacent tissue [6]. Following this definition a medical carbon should not release uncontrolled impurities or harmful substances into the body and it should not remove useful metabolites and nutrients from In Russian scientific literature a special word - 'endoecology', i.e. internal ecology, has been coined to reflect the concept of human body being part of environment; in the Western literature this word however has not been used. Although it is often attributed to Hippocrates and his oath, the author of this dictum translated from Latin 'Primum non nocere' was probably Galen.

3

Activated Carbons as Medical Adsorbents

531

it. The first part of this requirement is relatively easy to satisfy by restricting methods of carbon activation to 'physical' treatment with steam or carbon dioxide rather than using chemical activation by metal salts or hydroxides. Proper selection of a raw material for carbonisation also reduces risk of uncontrolled contamination of AC. Some natural precursors such as cellulose and certain types of wood have very low mineral content, but AC of the highest purity are produced from synthetic polymers such as polystyrene or phenol-formaldehyde resins, as the polymerisation process can be strictly controlled at every step. Even when all these conditions are met, there is still a possibility of AC contamination during physical activation, caused by iron oxides falling from the furnace walls that can be incorporated in carbon. Iron and other mineral impurities are usually removed by AC treatment with hydrochloric acid followed by washing with water and saline [7-9]. It is much more difficult to prevent AC from removing useful substances from the body simply because it is against the nature of this material, which is the most powerful non-selective adsorbent ever created. This however hardly causes any problem for oral use of AC, which has to achieve the main aim of removing the toxic substance. No reports of negative effects of orally administered AC associated with depletion of nutrients or other useful substances have been reported; apparently this means that temporary loss of these substances due to non-selective adsorption by AC is either negligible or can be easily compensated by a normal diet. It is however recommended to avoid taking antidotes or other drugs and AC at the same time because carbon can adsorb the drug and neutralise its action [1]. 2.2, Ilaemocompatibiliry of activated carbon Requirements for AC in a haemoperfusion column are the strictest. Blood is much more sensitive to the properties of a biomaterial than the gastrointestinal tract where the orally taken AC is located. In early publications on haemoperfusion, changes in the mineral composition of blood after contact with AC were reported; those included reduction of Ca2+, K+ and phosphate concentration. These results varied from carbon to carbon and in general were not considered to be a problem [10]. More importantly, significant blood cell losses were found after haemoperfusion over AC [11]. There is however another problem with the use of AC for blood purification, and it is potentially more dangerous and serious than all the others. Blood serum contains 6.6-8.3% of proteins [12], which are powerful surfactants. Whilst exposed to a surfactant solution, an activated carbon with poor mechanical strength releases a large number of microparticles [13]. These particles are loosely attached to carbon granules and they are not removed by washing with a saline. Once released into the bloodstream they may end up in small diameter blood vessels blocking them and causing mieroemboli. These observations in 1960s-1970s led to the conclusion that AC is not sufficiently haemocompatible (i.e., compatible with blood) and the solution to this problem was found in coating activated carbon granules with a more biocompatible material such as cellulose or albumin (Fig. 2a) [9,10,14]. Although AC coating resolved the problem of its poor biocompatibility, it created another one, which restricted use of adsorption therapy in comparison with alternative blood purification techniques such as dialysis or filtration. The reason for such a restriction is rather obvious: a 0.5-5 um thick semi-permeable coating (i) slows down the

532

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

external diffusion to the adsorbent surface and (ii) prevents larger molecules from contact with the adsorbent completely [7]. It means that removing so-called 'middle molecules' (substances with a molecular weight MW between 1,500 and 20,000 Da), larger proteins and protein complexes that are associated with many pathological states, via adsorption by coated AC or other adsorbents becomes inefficient [15]. As a result, haemoperfusion over coated AC has been mostly used for acute poisoning treatment only, which is usually caused by low MW substances and the duration of the haemoperfusion session is as long as that of dialysis or filtration, in the range of 4-6 hours. Even so, haemoperfusion is not always efficient in removing small molecules because they may be bound with serum albumin in large molecular weight complexes, as the human serum albumin molecule has MW 67,000 Da [16]. An example of such a small drug molecule is ibuprofen, which has molecular weight of 352 but it is 99% albumin-bound [17]. 2.3. Uncoated activated carbon for haemoperfusion In the then USSR the problem of poor AC haemocompatibility was solved by producing it from synthetic polymers in strictly controlled conditions and pre-conditioning of AC prior to contact with blood by keeping the adsorbent in a solution with mineral composition mimicking that of blood. In earlier works AC was pre-conditioned in a saline. The most successful medical AC was produced from polystyrene (SUGS and SCS) and poly(vinylpyridine) (SCN) [2,18-20]. AC produced from synthetic polymers is mechanically strong; its spherical shape granules withstand attrition upon transport and storage and do not release microparticles in contact with blood (Fig. 2b). A simple pre-washing of the AC haemoperfusion column reduces the number of microparticles with the size in the range of 2-10 um in the solution to within the limit allowed by the British Pharmacopoeia for solutions used for intravenous injections (1000 particles per ml for particles within the range of 2-5 urn and 100 particles per ml for particles larger than 5 um). Although cell losses occur upon direct blood-AC contact, they are minimised by adding heparin — an anti-clotting agent, which is always added in an extracorporeal treatment by dialysis, filtration or haemoperfusion over coated AC [21 ]. 3. ACTIVATED CARBON IN POISONING TREATMENT The most likely cause of poisoning is by intentional or accidental swallowing or drinking of a toxic substance or its solution. Although acute poisoning is a common cause of admission to emergency treatment, its mortality rate is low and special treatment is considered to be unnecessary unless the patient is in a life-threatening state [21]. 3.1. Oral use of activated carbon (charcoal) To account for all the publications in which activated carbon or as most medical publications prefer to call it, charcoal, has been used as an oral adsorbent is a task next to impossible. This reflects a firm belief of a large number of clinicians in the ability of AC to adsorb toxic substances in the gut thus decontaminating the patient's body and reducing the risk of serious complications. Interestingly, this belief goes back perhaps to the XlXth century, when in 1831 a pharmacist Pierre Touery swallowed a mixture of one grain of strychnine (10 times the

Activated Carbons as Medical Adsorbents

533

lethal dose) mixed with 15 g of charcoal apparently without any ill effect! [1]. He demonstrated this experiment for the French Academy of Medicine, which was so impressed that nobody has questioned the therapeutic ability of AC ever since. The case of AC is unique as no systematic clinical trials were undertaken to prove its effect. This argument was used in a Civil Court Case: US State of Hawaii vs. Eli Lilly in 1997 by an expert in toxicology Dr Healy, who examined the history of the use of AC. He testified that activated charcoal appears to be the only substance, for which the current registration and licence for the use in drug overdose treatment has been issued on the basis of that single case report [22], Despite a massive number of publications reporting the use of AC in poisoning treatment, there is a lack of systematic comprehensive clinical studies which should have involved a randomised statistically significant number of patients in several clinical centres. The usefulness of AC in poisoning treatment has never been proven or questioned ever since, and the attitude of clinicians to this subject seems to be ambivalent. The general perception is that administration of AC is beneficial to the patient as it adsorbs the toxic substance and reduces its concentration in the body. Some authors, however, argue that despite this reduction, it has never been proven that administration of AC actually improves the clinical outcome [24]. 3.1.1. Treatment of poisoning with organic substances Activated carbon is the best-known adsorbent for organic molecules, and the list of substances, which AC has been used to eliminate from poisoned patients, is impressive (Table 1). Activated carbon was administered either orally or via a nasogastric tube if the patient was unconscious or refused to swallow the adsorbent, or in most severe cases, haemoperfusion or plasma perfusion was used. (Publications 9 and 31 refer mainly to the extracorporeal treatment). From the analysis of the data some general conclusions and recommendations can be drawn [36,41]. Activated carbon used as an oral adsorbent accelerates and increases the elimination of most low MW organic substances from the body confirming that it is a truly universal antidote. There are no satisfactorily designed in vitro experiments that can be related to the animal and clinical studies. Although such experiments try to mimic the conditions of the gastro-intestinal (GI) tract, they fail to model the crucial step of drug absorption from the small intestines and its distribution in the rest of the body. According to [41], there are no satisfactorily designed clinical studies assessing efficiency of administration of AC either. The adsorption efficiency of AC determined in vitro is almost invariably higher than that measured in vivo. This result is understandable taking into account the presence of substances competing for adsorption in vivo and distribution of the target substance in other body compartments inaccessible to AC. In all cases early administration of AC is more efficient than a delayed administration. Position statements published by the American Academy of Clinical Toxicology (ACCT) and the European Association of Poisons Centres and Clinical Toxicologists (EAPCCT) on the medical use of activated carbon recommend the administration of AC up to one hour following ingestion of a life-threatening amount of poison [36,41]. Multiple-dose administration of AC even after the first hour may be recommended for treatment of patients who ingested drugs with a

u

Air trap

Flow meter

Fig. 1. Schematic of an extracorporeal circuit for blood purification. Detoxifying unit is a haemoperfusion column or a dialyser, or a filtration unit. The blood is pumped from an artery or vein and is returned to the same or a different vein.

To vein

From artery/vein

Pump

Flow meter

Anticoagulant solution

g

534 S.V. Mikhalovsky and V.G. Nikolaev

x2,500

x10,000 xl0,000

Fig. 2 (adapted from [23]). Scanning electron micrographs of (a) cellulose-coated commercial AC used in Adsorba® 300C haemoperfusion column (Gambro, Sweden) and (b) uncoated AC produced from phenol-formaldehyde resin (MAST Carbon Ltd., Guildford, Surrey, UK). Note that the coated surface of AC in (a) is smooth, whereas the exposed carbon surface in the place where the coating is damaged is ruudi. In cnnirasi, I IK- SUI-UKC ol'ihc uncoi'iiul pol\ IIKT pwolvscd At in (b) is

x25

1 mm

b

a

r

I

Activated Carbons as Medical Adsorbents

535

536

5. V. Mikhalovsky and V.G. Nikolaev S.V.

Table 1 Toxic organic substances and drugs adsorbed by activated carbo^ Name Class of substance In Animal Patient vitro studies treatment Y Y Y Strychnine Alkaloid Y Y Nicotine Y Y Atropine Y Y Aconitine Y Y Veratrine Y# Y Y Aspirin Salieylate Y Y Y Sodium salieylate Y Y Y Acetaminophen Analgesic (paracetamol) Y Y Y Propoxyphene Y Y Nefopam Y Morphine Y Y Tilidine Y Hypnotic or sedative Diethylbarbituric acid Y Y Y Allylpropynal Y Y Pentobarbital Y Y Glutethimide Y Y Methaqualone Y Y Secobarbital Y Y Y Phenobarbital Y Y Amobarbital Y Y Barbital (Veronal) Y Methyprylon Y Ethchlorvynol Y Zolpidem Y Y Sulfanilamide Antimicrobial Y Carbamazepine Anticonvulsant Y NSAID Mefenamic acid Y Piroxicam Y Phenylbutazone Y Indomethacin Y* Y Y Tricyclic depressant Imipramine Y Y Desipramine Y Amitryptiline Y Nortriptyline Y Doxepin Y Clozapine Y Furosemide Diuretic

Ref. 1,25 1 1 1 1 1,26-28 1,9,26 1,29 1 1 1 1 1 1 1 1 1 1 1,26,27,30,31 26 1,31 1 1 32 1 1 1 1 26 1 1,26 1 1,31 1 1 33,34 1

537

Activated Carbons as Medical Adsorbents

Class of substance Cardiac glycoside

Organic solvent

Antihistamine

Antibiotic

Anti-tuberculosis Anti-malarial

Sulphone Anti-arrhythmic

Beta-blocker

Ca-channel blocker Anticonvulsant

In Animal Patient vitro studies treatment Y Y Y Digoxin Y Y Y Digitoxin Y Methyl proscillaridin Y Y Y p-Msthyl digoxin Y Bufadienolides Y Kerosene Y Benzene Y Diethylaniline Y Tetrachloroethane Y Y Carbon tetrachloride Y Aniline Y Pyridine Y Carbon disulphide Y 1,2-Dichloroethane Y# Y Y* Ethanol Y** Y Y* Ethylene glycol Y Y Chlorpheniramine Y Diphenhydramine Astemizole Y# Y ** Gentamicin Y#* Ciprofloxaein Name

Vancomycin Aminosalicylic acid Isoniazid Chloroquine Quinine Sulphoxidine Dapsone Amiodarone Disopyramide Flecaininde Quinidine Nadolol Sotalol Propranolol Diltiazem Phenytoin Carbamazepine Valproic acid

Y Y Y Y Y

Y Y Y

Y* Y

Y* Y Y* Y Y Y Y Y* Y Y

Y Y Y Y Y** Y Y Y

Ref 1,26 1 1 1 1 1 1 1 1 1,9,26,31 1 1 1 1,26 1,31 1,31 1 1 35,36 1 1 36,37 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1

538 Class of substance Tranquiliser

CNS agent GIdrug

Antidiabetic Bronchodilator Preservative, anaesthetic Flame retardant

5. V. Mikhalovsky and V.G. Nikolaev S.V.

Name Diazepam Chlorpromazine Meprobamate Methamphetamine Propantheline Diphenoxylate Nizatidine Sulphonylureas Theophylline Camphor

In Animal Patient vitro studies treatment Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y** Y#

Ref 1,31 1 1 1 1 1 1 1 1,26,38 1,39

Y** 1 Polybrominated biphenyl Y 1 Y* Y Herbicide Paraquat 1 Y Insecticide Y Malathion Y 1 Y Diazinon Y Y 1,26 Carbophos 31 Trichlorfon Y (Chlorophos) Y Pesticide Neguvon Immunosuppressant Cyclosporin Y Y* Chemotherapeutie Methotrexate Y Anti-infective Trimethoprim Y Drug of abuse Y Cocaine Y Y 40 Phencyclidine There is also a very large number of clinically relevant substances whose adsorption was studied only in vitro. These results are not included. *Not effective if administered with a delay. **Not effective; low adsorption level. Controversial results. NSAID - Non-steroidal anti-inflammatory drug, CNS - central nervous system, GI gastrointestinal. prolonged elimination half-life and slow diffusion through the membrane of the GI tract walls [36,42]. An interesting exception is related to the controversy in the data regarding adsorption of ethanol. Ethanol is poorly adsorbed by AC even in vitro experiments and one may expect that it would be inefficient in vivo [1,41]. The results are however equivocal; although most of them confirm inefficiency of AC, apparently there is some evidence that administration of AC before ethanol intake or very soon thereafter reduce the hangover symptoms following alcohol consumption [1]. The authors of this chapter also received similar (verbal) conformation from a

Activated Carbons as Medical Adsorbents

539

number of persons. The most likely explanation to this mystery is that AC adsorbs products of oxidation of ethanol and higher alcohols that may be present in the beverage. Controversies in the results for other substances may be due to the fact that the clinical studies are often carried out in very different conditions, which include AC dosage, frequency of use and use of supplementing methods [42]. 3.1.2. Treatment of poisoning with inorganic substances and radionuclides Unlike organic molecules, inorganic species are not efficiently bound by AC [1,21]. The list of inorganic substances than are accidentally taken orally is also much shorter than that of organic substances as the former are not so easily available. It includes metallic mercury, iodine (tincture), metal salts (mostly potassium, thallium, lithium, mercury, lead, copper and iron), alkalis, hypochlorite, acids, cyanide and fluoride. Among these substances, cyanide is known to be adsorbed by AC, however its biological effect is very rapid making the use of activated carbon problematic. In animal experiments the survival rate of rats that were given AC orally immediately after oral administration of KCN, was significantly higher than in the control group and onset of poisoning symptoms was delayed [1]. Among other inorganic species those which tend to stay in the uncharged molecular form, are adsorbed much better than ionic species. Thus intoxication with some compounds of heavy elements such as HgO [43], HgCb [44], TI2SO4 [1, 45] and AS2O3 [46] has been successfully treated using oral administration of AC. Only a few results were reported and usually AC was administered along with a chelating agent, so it is difficult to say to which extent AC was necessary. In one case though, the patient had ingested HgCb and having received lavage and activated carbon, she refused to take any chelating therapy and eventually was discharged without serious consequences [44]. Oral adsorbents that have ion exchange properties, such as polymer resins and polysiloxanes, may be more efficient than AC in binding ionic inorganic species. This idea led to the development of oxidised AC as an oral adsorbent [47]. Oxidised carbon has a range of acidic functional groups and therefore a higher adsorption capacity towards metal cations [48]. It has been suggested that impregnation of the oxidised AC with physiologically relevant cations such as K+ and Mg2+ may be used as an additional (ion exchange) function of the orally administered AC [47]. It has to be said that bioinorganic chemistry of metals is a complex subject and the metal ion behaviour in living body is unlikely to obey relatively simple relationships established in laboratory experiments. Important data on the efficiency of oral AC in heavy metal removal have been obtained after the Chernobyl nuclear power plant accident on April 26* , 1986. During the first 7 days after the accident significant amounts of radionuclides were released from the nuclear reactor to the environment, which required extensive measures and man-power to prevent further spreading of the radioactive contamination and clean-up of the contaminated territories [49,50]. The military personnel who worked in the Chernobyl exclusion zone, known as 'liquidators' because they liquidated the consequences of this disaster, were exposed to elevated levels of radioactivity. Although these levels could not cause the radiation sickness disease, they were sufficiently high to affect health of 'liquidators' particularly if the radionuclides became incorporated in the body

540

S. V.Mikhalovsky Mikhalovskyand andV.G. V. Nikolaev G. Nikolaev S.V.

[51,52]. Health of the 'liquidators' who worked in the Chernobyl zone in early days after the accident was at highest risk The estimated radiation dose received by the personnel was 30-50 ber (biological equivalent of roentgen) in the first week after the accident, 15-20 ber in the first half of May 1986 and 5-10 ber in the second half of May 1986 [52]. These data do not take into account accumulation of radioiodine in thyroid. After Chernobyl a large proportion of the population has become continuously exposed to low-level radioactivity, consequences of which are unknown. Most of the work done before focused on exposure of the living species to highlevel radioactivity that could be expected in the event of a nuclear war. Extrapolation of the previous data to the post-Chernobyl situation suggests a complex effect of external ionising radiation and incorporated radionuclides on the human body. In addition to direct molecular and cellular damage it comprises secondary processes such as free radical generation, oxidative damage, formation of radiotoxins, alteration of enzymatic transformations, immune system deficiency, carcinogenesis and ageing [53-56]. What substances fall within the definition of radiotoxin is not clear, but this concept reflects a well established fact that the blood of animals irradiated with ionising radiation accumulates certain toxic substances, and, if injected, it causes pathological changes in healthy animals [57,58]. Products of metabolite oxidation such as quinones (from phenols), lipoperoxides and hydroperoxides (from unsaturated lipids), crosslinked fragments of DNA, free radicals and a number of other types of substances contribute to the toxicity of blood of the irradiated animals. At low level of radioactivity the secondary processes are likely to produce a more serious impact on health than direct ionising radiation. Formulations based on activated carbon SCN (Ukrainian Academy of Sciences) [59] were given to several groups of 'liquidators' in order to reduce accumulation of radionuclides in the organism. This study was carried out in early months after the Chernobyl accident and AC was administered to the personnel of group 1 - only during work in the zone (number of persons n = 50); group 2 - only after leaving the zone (n = 25) and control group (n = 25) which did not receive AC [52]. Group 2 also served as the control group during its staying in the zone. AC was administered in a 5-7 g dose, three times a day for two weeks, at least 1-1.5 h before meal. All participants in the study were male, 25-50 years old. All three groups worked in the zone simultaneously. By the time of leaving the zone, the average concentration of radionuclides in the blood and urine of group 1 was 6.1 and 2.2 times lower than in the control group, whereas in the faeces it was 1.5 times higher. Two weeks later the level of radionuclides in the blood of group 2 was 3 times lower than in the control group, whereas it was 2 times higher in the faeces. Similar results were obtained in other studies carried out in April ~ August 1986. An example of radionuclide distribution in the blood of 'liquidators' is given in Table 2. The conditions and AC dosage were the same as described above. These data confirm that oral administration of AC accelerated removal of incorporated radionuclides from the human body. The exact mechanism of this action is not clear. Activated carbon usually bears some ion-exchange functionality due to the presence of both basic sites and oxygen-containing acidic groups [60]. Oxidation of AC creates higher surface density of acidic functional groups and enhances adsorption of metal cations from simple solutions of their salts [48,61]. Metal ion adsorption on AC is only partly reversible suggesting that there are at least two different processes contribute to the adsorption mechanism, one being physical adsorption

Activated Carbons as Medical Adsorbents

541

(reversible) and another - chemisorption (irreversible). Adsorption affinity of metal ions with the oxidised AC resembles the stability series of metal complexes with carboxylic acids implying that the surface complex formation rather than ion exchange could be the main process responsible for metal ion retention by AC [48]. Being produced from polyvinylpyridine, activated carbon SCN has also ca. 1% of nitrogen heteroatoms in its structure [59]. According to the flow microcalorirnetry and Mossbauer spectroscopy data nitrogen-containing surface functionalities form complexes with transition metal ions which are stronger than complexes with oxygencontaining functionalities [62,63]. Affinity of polyvalent transition metal ions with the carbon surface is higher than affinity of cations of groups I and II of the periodic table. The effect of oral AC administration on the level of incorporated radionuclides was the biggest in the 'liquidators' who worked in the zone in the early days after the accident. They had the highest level and the widest variety of incorporated radionuclides. Later on the spectrum of radionuclides became dominated by radiocaesium and radiostrontium and the efficacy of oral AC diminished. It is likely that the mechanism of the detoxifying action of AC comprises surface binding of radionuclide ions via ion exchange and complex formation and physical adsorption of radionuclide complexes with biomolecules. Table 2 Effect of the prophylactic administration of the activated carbon SCN on the concentration of radionuclides, nCi I"1, in the blood of'liquidators' (May-June 1986) (adapted from [51]). Note that with the exception of iodine and caesium, all the radionuclides are polyvalent transition metals. Isotope half-life Prophylactic use Control Control/ (n = 30) Prophylactic of AC (n = 57) 51.6 ±4.1 7.3 8.04 days 1-131 7.11 ±0.67 28.6 ±1.5 Ru-103 + Ru-106 39.3 + 368.2 days 6.61 ± 0.80 4.3 17.2 518 ±47 40.2 hours La-140 30.1 ± 3.8 Cs-134 + Cs-137 2.06+ 31 years 2.0 23.4 ±1.9 11.9 ±8.0 33.2 ± 3.2 2.0 35 days Nb-95 16.8 ±1.1 21.5 ±3.4 2.6 64 days Zr-95 8.21 ± 0.80 Ce-141+Ce-144 32.5 + 284.3 days 23.2 ± 2.5 2.5 56.9 ± 6.7 7.1 733.2 ±211 Total 103.9 ±23.8 Isotope

3.1.3. Treatment of poisoning with biological toxins Biological toxins are produced by bacteria, viruses, protozoa and fungi (mycotoxins). They can also be present in some fish, plants, mushrooms and other species. Biological pathogens can enter the human body with food, water, air, through skin and other routes. Some of them can also be used as biological warfare [64,65]. Biological toxins are usually substances of polypeptide or proteinaceous nature which are much more deadly than chemical poisons. Food infections, acute poisoning and other pathological conditions caused by biotoxins have been treated with oral AC or haemoperfusion as shown in Table 3.

542

5. V. Mikhalovsky and V.G. Nikolaev S.V.

AC is included in the recommended treatment of ricin, saxitoxin, T2 mycotoxins and botulinum toxin poisoning although its efficacy remains to be proven [65,71]. It is interesting that rather large objects such as bacteria (0,1-10 jjm) can be retained by AC at all; these results suggest that the nature of AC surface — bacteria interaction cannot be attributed to physical adsorption only; for smaller objects such as viruses (10-100 nm) and molecules of biotoxins which have molecular weight up to 1000 kDa contribution of adsorption in mesopores may be important. Table 3 Treatment of biotoxin caused poisoning and pathological conditions with AC. Origin In Name Animal Patient Ref vitro* studies treatment 1 Fungus Aspergillus flavus Y Aflatoxin Bi Y Toxin T-2 Y Fungal mould, Fusarium 1,65 sporotrichioides (Trichothecene) 1 Fungal mould Aspergillus y Ochratoxin A Y and Penicillium spp. 1 Y Blue-green algae Microcystin-LR Y Microcystis aeruginom Y Paralytic shellfish Saxitoxin and 1,65 poisoning (caused by Tetrodotoxin algal species) Puffer fish Tetrodotoxin Y Red algae Brevetoxin Y Clostridium botulinum Toxin type A Y Y Mycotoxin, fungus Sporidesmin Pithomyces chartarum Snake Cobra venom Y Y Virus Foot-and-mouth Y Y Virus Sheep pox Y Thermobacterium Bacteria helveticum 1 Y Bacteria Streptococcus lactis 1 Bacteria Saccharomyces Y ellipsoideus 1 Bacteria Lactobacillus Y acidophilis 1 Y Bacteria Pseudomonas chlororaphis 1 Y Bacteria Bacillus megatherium

543

Activated Carbons as Medical Adsorbents

Name

Origin

Patulin Diphtheria toxin

Penicillium spp. fungae Corynebacterium diphtheriae Tetanus toxin Clostridium tetani Tuberculin toxin Mycobacterium tuberculosis Amatoxins (mostly Death cap mushroom Amanita phalloides a-amanitine), phallotoxins Baeteraemia, Pseudomonas Aeruginosa gastrointestinal infections Escherichia coli Endotoxin Vibrio cholerae Enterotoxin Escherichia coli Sepsis Bacillary Shigella flexneri dysentery, or shigellosis Infection, sepsis, Salmonella typhimurium salmonellesis Staphylococcus aureus Sepsis Klebsiella spp. Sepsis Enterobacter spp. Sepsis Sepsis, septicaemia Neisseria menigitidis (Meningococcus) Citrobacter spp. Sepsis Proteus mirabilis Infection, sepsis Hemolysin Proteus vulgaris Ricin Castor bean, Ricinus communis * Adsorption of bacteria was studied. Controversial results.

Animal Patient In vitro* studies treatment Y Y Y

Ref 1 1 1 1

Y

66,67

Y

Y Y Y* Y

Y Y Y

Y Y Y# Y

Y

1,19,31

Y

1 1 1,47,66-68 66,67

Y

66,67, 70

Y Y# Y Y*

47, 66,67, 69,70 47,66-68 26,47, 68 47

Y Y Y Y

47 47 68,70 65

3.1.4. Comparison of oral AC treatment with alternative methods It seems obvious that the most efficient poisoning treatment would be to administer a specific antidote/antagonist to the particular poison. There are however surprisingly few antidotes available and their indiscriminate use may even be harmful [21,72]. Quite frequently a poisoned patient is unconscious, which makes identification of the toxic substance more problematic and delayed. Hence non-specific treatment methods prevail in the management of

544

5. V. Mikhalovsky and V.G. V. G. Nikolaev S.V. Nikolaev

poisoned patients. These include symptomatic procedures, supportive care and gut decontamination methods such as gastric lavage, ipecac-induced emesis, activated carbon and whole bowel irrigation (WBI) [21]. Use of cathartics, such as sorbitol or saline cathartics, according to the most recent position paper published by AACT/EAPCCT, has no role in the management of the poisoned patient, nor their use in combination with AC is beneficial to the patient [72,73]. For the most severely poisoned patients other methods which accelerate elimination of poisons may be used. These include multiple-dose AC, urinary alkalinisation and extracorporeal methods such as haemoperftision, dialysis and ultrafiltration [1,21,36,41]. Gastric lavage, or stomach washout, has been used for over 180 years [72]. It is an uncomfortable and time-consuming procedure [1]. Clinical studies carried out in 1970s-1990s found that gastric lavage has very few if any benefits compared to using AC and is associated with high risk [72,74,75]. Ipecac syrup is made from the rhizome and roots of ipecacuanha plant; it contains potent and toxic alkaloids [1,72]. Ipecac has been used in clinical practice to induce emesis and over three million patients received it during 1983-96, but by the time of the AACT/EAPCCT Position Statement in 1997 clinical studies have not confirmed any benefit from its use in the management of poisoned patients [76]. Moreover, a prospective evaluation of the outcomes of poisoned patients which received ipecac and AC vs. patients who received AC only, revealed a significantly greater complication rate and greater length of stay in the first group. The administration of ipecac may delay the use of AC by 1-2 hours, so the benefits of using ipecac are at least doubtful [1,76,77]. It may however be used before AC administration to increase the efficiency of the latter [21], Whole bowel irrigation, which is the oral administration of large volumes of lavage fluids, substantially decreases the bioavailability of slowly released or slowly absorbed ingested drugs. Usually a solution of polyethylene glycol with electrolytes is used [21]. It should be considered for potentially toxic ingestions of sustained-release or enteric-coated drugs, e.g., theophylline, ingested packets of illicit drugs and inorganic substances that cannot be adsorbed by AC such as lithium, lead, zinc, arsenic, mercury and iron compounds [1,78]. The concurrent administration of WBI and AC may decrease the effectiveness of AC. In general, oral AC administration is more effective than WBI [1]. Urinary alkalinisation increases the urine elimination of certain poisons by shifting the urine pH above 7.5 [1,21,79]. It is achieved by intravenous injection of a sodium bicarbonate solution. Raised pH of urine facilitates elimination of acidic drugs such as salicylates and barbiturates as they become ionised. The rate of reabsorption of an ionised drug into the blood is significantly lower than that of a non-ionised drug [1, 79]. In all cases however AC administration is more efficient than urinary alkalinisation [1,80]. It should be also kept in mind that WBI and urinary alkalinisation are relatively slow procedures which may cause delay with AC administration. Extracorporeal methods are used when removal of the toxic substance by oral AC is impossible or inefficient, for example if the patient is in coma or if the poison has already been absorbed into the blood [1]. There is limited data on the use of haemofiltration in poisoned patients and at present it is not recommended for the poisoning treatment [21].

Activated Carbons as Medical Adsorbents

545

Because of the complexity of metabolic processes in which the toxic substance is involved in the body, it is unlikely that a simple relationship between the chemical structure of the substance and clinical efficiency of the activated carbon can be found, but the clinical evidence that removal of such a large number of substances is accelerated by AC suggests that it can be used as an almost universal antidote, particularly if the nature of the poison is unknown. And even if it is known, contrary to the common belief there are very few specific antidotes that are efficient in neutralising particular substances and the use of AC will be still beneficial in reducing bioavailability of the toxic species [21 ]. J.i.J. Efficiency of oral AC treatment There have been few comparative studies of different commercially available activated carbons and most of them were carried out in simple bench experiments [1,81]. In these experiments efficiency of adsorption of low molecular weight substances was determined. The main conclusion is that the performance of AC usually correlates with its surface area, although it is not clear to which extent these results could be extrapolated to clinical cases. Activated carbon is used in a single-dose or multiple-dose treatment of acute poisoning [1,21]. It can be given orally or via nasogastric tube if the patient is unconscious or refuses to take AC orally. The main direct mechanism of the detoxifying action of AC is adsorption of the toxic substance from the gastrointestinal tract and thus diminishing its concentration and bioavailability. Hence the common perception is that AC should be administered as early as possible after the poison ingestion. The recommendation to take oral AC within an hour of ingestion given in many publications is evidence based rather than supported by proper clinical trials [38,72,75]. There is also evidence of circulation of some drugs between the gut and other body compartments, the socalled enterohepatic or enteroenteric circulation [82], secretion of poisons such as digoxin directly into the gut ('gastrointestinal dialysis') [83] and formation of a toxic substance in the gut. It means that the orally administered AC can interrupt this circulation and prevent absorption of the toxic substance by the body reducing its bioavailability and concentration in the blood. It is worth looking at the data on delayed administration of AC. This information is important because in reality a large percentage of the poisoned patients would not receive treatment within the first hour. For example, a prospective study carried out at the Accident and Emergency Department of Hull Royal Infirmary (UK) showed that only 17% of the overdosed patients presented within 1 h and almost 11% presented at 12 h and later after the ingestion [84]. Although undoubtedly less effective than within the first hour, administration of AC at 6 h reduced absorption of theophylline by 57.3% (vs. 91.2% at 1 h) in experiments with simulated theophylline overdose in healthy volunteers [38] the conclusions were made by comparing concentration of theophylline in the blood plasma of volunteers and the control group. In similar experiments with paracetamol only the group of volunteers which received oral AC at 1 h had statistically significant lowering of the drug concentration in the blood and no difference was recorded at 2 and 3 h [85]. These results contradict the findings of the group that used 'superactivated' charcoal [86]; they found that the paracetamol level in the blood of the volunteers who received 3 g of the drug and took 75 g of AC at 3 h, was significantly lower than

546

S. V.Mikhalovsky Mikhalovskyand andV.G. V. Nikolaev G. Nikolaev S.V.

in the control group (the data were corrected for the weight of the volunteers and the drug dose) at 4 and 7 h after the paracetamol injection. The results in the other group of volunteers, who took 2 g of the drug, did not differ from the control group. These conflicting results confirm the AACT/EAPCCT statement that there are insufficient data to support or exclude the delayed use of AC after 1 h of ingestion [41]. It seems that the higher dose of AC administered to the patient with severe drug overdose has a lowering effect on the drug level in the blood at times much longer than 1 h. It has been suggested that AC has the highest adsorption capacity by low molecular weight substances in the range of 100-1000 Da [72]. This is probably true for the AC formulations currently used (Table 4) because these carbons are predominantly mieroporous. Although large clinical trials have never been carried out, a number of animal, volunteer and clinical studies have shown that multiple-dose AC (MDAC) enhances clearance of phenobarbital, earbamazepine, theophylline, quinine, digoxin, paraquat, vancomycine, dapsone and the fungus Amanttaphalloides and shortens the time spent in hospital [21,27,36,37,87,88]. Studies on volunteers have demonstrated benefits of using MDAC also in the treatment of other poisonings [36]. These studies involved a limited number of patients and, not surprisingly, sometimes produced contradicting results, particularly regarding clinical benefits of using MDAC such as the time to complete recovery [89,90]. Multiple doses of AC are usually administered every 4-6 h in the amount of 25-50 g [72]. MDAC has been reported to significantly reduce mortality and life-threatening cardiac arrhythmias after yellow oleander poisoning [91], in this single-blind, randomised placebo-controlled trial 201 patients received 50 g of AC every 6 h for 3 days vs. placebo (sterile water); repeated doses of AC reduce the half-life of digoxin from 37 h to 22 h and increased the body clearance from 12 to 18 L h"1. It is possible to formulate the criteria for the drug to have enhanced clearance by using multiple-dose AC: slow release preparations, low clearance (long half-life), small volume of distribution, delayed action [42,88]. Apparently these data confirm that some drugs diffuse out of the bloodstream back into the intestinal lumen and multiple-dose AC treatment of poisoning with such substances would be beneficial. The common sense suggests that reducing bioavailability and concentration of the poison in the organism should be beneficial to the patient, and AC undoubtedly achieves this aim via adsorption. The main objection raised against using single- or multiple-dose AC is that there is insufficient evidence that it actually changes the outcome of the poisoning treatment. Lack of large controlled trials - and they are unlikely to ever take place! - restricts oral use of AC to lifethreatening conditions but so far there is no better and universal alternative to this treatment. The common sense also advises not to administer AC along with other drugs as their action may be neutralised by adsorption on the former [1], There are few actual data which confirm or reject this statement [92,93], but it seems logical and to avoid possible interference between AC and drugs they are administered with a reasonable time interval. Such an interaction may become beneficial in preparing delayed drug release AC formulations. The preparation comprising the antibiotic mitomycin C pre-adsorbed on activated carbon with particle size within the range of 40-150 um was dispersed in the peritoneal cavity of patients undergoing gastrectomy for gastric carcinoma. Due to its slow release (desorption) from AC, high

Activated Carbons as Medical Adsorbents

547

mitomycin concentration was maintained in peritoneal exudate for 24 h as compared to the control group of patients. Three years after treatment the mortality rate in the group of ACmitomycin treated patients was 2.5 times lower than in the control group [94], A number of other materials have been used as oral adsorbents. It has been shown that natural mineral adsorbents such as zeolites and clays (bentonites, kaolins, montmorillonite or Fuller's earth, and palygorskite), natural organic adsorbents (cellulose, chitin and lignin) and synthetic adsorbents such as polymeric ion-exchangers, non-ionic resins, silicas and siloxanes can remove toxic substances from the human organism [1,2,26,52,95-98]. They all have one important psychological advantage over AC — they are not black. Their palatability is no better than the palatability of AC. Some of these adsorbents have ion exchange properties and, unlike activated carbon, can be used for removing ions. Combinations of AC and mineral adsorbents carbon-mineral adsorbents have been suggested for removing inorganic ions [52,99,100]. Different mineral additives have been tested. They include titanium and zirconium dioxides, zirconium and titanium phosphates, alumina, natural zeolites and Prussian blue (for radiocaesium removal) [52]. Nevertheless, AC is a much more powerful non-specific adsorbent in comparison with other oral adsorbents and there seems to be very little doubt that it is more effective in poisoning treatment [1,26]. 3.1.6. Complications associated with oral administration of activated carbon Taking into account that in the management of acute poisoning the recommended dose of AC is huge (25-100 g for adults and 0.5-1 g kg" for children up to 12 years, orally or by nasogastric tube) [41], it can be considered as a well tolerated substance and few adverse effects have been reported indeed [1,21]. It is a very rare example of a medicine that is safe to take by pregnant and breast-feeding women. AC is associated with fewer serious complications than the alternative methods of gut decontamination such as gastric lavage or ipecac induced emesis [1,84]. In the most recently published report, 555 patients with suspected acute intoxication received aqueous Carbomix suspension in an out-of-hospital treatment provided by emergency care services and no adverse effects were reported [101]. Those complications that have been reported are rare and include severe pulmonary complications and death caused by AC aspiration [1,21,87], constipation or even intestinal obstruction, regurgitation with subsequent aspiration into the lungs, gastrointestinal tract perforation, nausea, emesis/vomiting and fluid and electrolyte abnormalities [1,21,24, 36,41,102]. Aspiration of AC into the lungs followed by pulmonary complications or lung injury was caused by inadequate airway management, misplacement of the nasogastric tube or vomiting. Single dose AC does not cause GI obstruction or constipation, but multiple doses of AC can result in constipation and formation of charcoal 'briquettes'. It has been suggested to give sorbitol or lactulose together with multiple-dose AC to reduce the risk of bowel obstruction although the need for this remains unproven [27]. There has been a report suggesting that sorbitol inhibits the adsorption efficiency of AC in acute acetaminophen poisoning [103]. On the other hand, addition of sorbitol was shown to enhance the antidotal effectiveness of some charcoals in sodium pentobarbital removal from rats [104]. Apparently AC particles can cause corneal abrasion as reported in one case [24]. Fluid and electrolyte abnormalities are associated with excessive cathartic administration [102]. In one

548

S. V.Mikhalovsky Mikhalovskyand andV.G. V. Nikolaev G. Nikolaev S.V.

paper multiple-dose AC has been reported as the cause of acute appendicitis [105]. In the study based on 878 medical records of patients who received multiple-dose activated charcoal, it was found that five (0.6%) patients had clinically significant pulmonary aspiration (none of them died), one patient had a corneal abrasion (0.1%), 53 (6.0%) had hypernatraemia, 27 (3.1%) had hypermagnesaemia and none had gastrointestinal obstruction [106]. It is debatable whether even those few cases can be defined as adverse effects of AC itself rather than poor treatment management. The risk of aspiration pneumonia associated with AC administration is significantly higher in unconscious patients and can be reduced by protecting them with intubation by the properly trained personnel [107]. None out of 115 paediatric patients who were administered AC in the home (following instructions from Kentucky Regional Poison Center) had any aspirations or complications [108]. The mean dose in this study was 12.1 g and 95% of the children received AC within first hour after ingestion of a toxic substance. Findings of this study have been subjected to a good deal of criticism because the average AC dose was subtherapeutic and the condition which required AC administration was uncommon with less than two children per week meeting the criteria in the whole state [109,110]. Similar results were obtained in a prospective randomised trial which involved 34 overdose adult patients. No aspirations were reported in the group of patients who were administered AC before hospital treatment Average time for pre-hospital administration of AC was significantly shorter than in the hospital group (28 min. vs. 118 min.) [111]. After the publication of AACT/EAPCCT position statements on activated carbon [36,41 ] new animal data have been produced which seem to give evidence of direct pulmonary effects caused by aspiration of AC rather than gastric contents probably via activation of alveolar macrophages, 'volutrauma' and obstruction of small distal airways [24,112,113]. Alveolar macrophages trigger potent inflammatory response, whereas the presence of charcoal particles in the lungs leads to excessive pressure in airways, pulmonary oedema and damage - 'volutrauma'. AC does not increase lung neutrophil activity excluding neutrophil-mediated lung injury [112]. Contrary to the generally accepted opinion of low incidence of side-effects and complications associated with oral use of AC, a recent study carried out on healthy adult volunteers who received an overdose of acetaminophen reported a significant number of adverse effects caused by consumption of AC [114]. This study involved a limited number of volunteers (48 including 24 in the control group) and was restricted to one particular type of AC ~ 'superactivated' charcoal. Nausea and vomiting after AC administration are not rare; incidence of vomiting varies between 6.5 and 23%. hi a randomised-controlled trial of 1479 patients higher incidence of vomiting was reported in AC group (23% vs. 13% in control group) but no difference in incidence of aspiration was found [115]. Vomiting can be controlled by intravenous injection of conventional anti-emetics such as metoclopramide or ondansetron [21]. A fear of alleged carcinogenic properties of AC stands alone as a complicating factor in administering activated carbon. Although it is obvious to a specialist in AC that its manufacturing process completely eliminates all the organic matter including carcinogenic molecules, the words 'activated charcoal' commonly used in medical literature are frequently mistaken for charcoal used in barbeques, whose products of incomplete burning can be

Activated Carbons as Medical Adsorbents

549

carcinogenic indeed! Unfortunately, scientific literature simply ignores this almost anecdotal misunderstanding and the clarification can be usually found in 'alternative medicine' publications which do not have the same level of credibility [116]. 3.1.7. Palatability of AC This is by far the biggest problem with oral use of AC. Its palatability is poor, its slurry is unpleasant to drink and the black colour does not make it more attractive either. A prospective study conducted on poisoned patients attending the Accident and Emergency Department of Hull Royal Infirmary (Yorkshire, UK) revealed that out of 102 patients who were prescribed AC, 40 (39%) refused it and only 15 (24%) patients took all the prescribed amount [84]. The prospective follow-up study of 656 poisoned patients who were prescribed AC out-of-hospital by the emergency medical services in Helsinki, Finland, showed that in total 101 patients (15.4%) were not given activated carbon because of patient refusal (n = 72 patients, 11%), inability to ingest a charcoal mixture (n = 23, 3.5%) or other reasons (n = 6, 0.9%) [101]. Rejection of AC is particularly common among children who are the most likely cohort of population prone to accidental poisoning [1,117]. Encapsulation of AC is not a good option taking into account large quantity of activated carbon that should be administered in the case of poisoning - 1-2 g kg"1 body weight in children. To overcome this problem it has been suggested to use AC together with sorbitol [118]. Sorbitol as a 70% solution with sweet taste masks the AC presence to some extent and, due to its cathartic function, facilitates passage of the toxin through the GI tract [117]. Other attempts to improve AC palatability, such as adding flavours [1], mixing with soda pop [119], diet cola or cola [120,121], ice cream [122], milk chocolate, fruit juice [117,123] or sherbet, appear to make it more tolerable but decrease the adsorption capacity of activated carbon competing for the adsorption sites with the target substance. Although reports confirming decrease of the adsorptive capacity have been published [1,124], there is no conclusive evidence supporting or rejecting this approach as no systematic research in this area has been undertaken [117]. In infants, administration of AC as a suspension in warm water through a nasogastric tube may be the best option. For older children nasogastric administration should be avoided as they easily get frightened by the procedure and recommendations such as giving AC in a nontransparent container, presence of parents and involving children in a game which requires them to swallow AC may be considered [117]. Age-dependent psychological factors, education, and fantasy elements should not be underestimated in administering AC particularly to paediatric patients. It has been also noticed that the results obtained with volunteers are not necessarily reflective of real patients" behaviour and patients who had taken a potentially toxic overdose were 5.4 times more likely to take charcoal than patients who had taken mild or moderate (nontoxic) overdose [84]. Similarly comparison of different groups of poisoned patients has to be done cautiously, as the rejection of AC is higher in volunteer groups rather than in actual poisoning cases when the safety of the patient is at real risk [108,125,126]. 3.1.8. Availability of oral AC Poor palatability of activated carbon and its rejection by a large percent of patients, particularly children, has led to a paradoxical situation. On the one hand, AC is the only

550

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

substance that can claim the title of the universal antidote; its use in poisoned patients reduces bioavailability of most ingested substances [127]. It has been officially recommended in the position statements published by the American Academy of Clinical Toxicology and the European Association of Poison Centers and Clinical Toxicologists [36,41]. On the other hand, many physicians are reluctant to prescribe or fail to administer AC because of the palatability problem. According to an anonymous telephone survey which involved all 75 poison centres in the US and Canada and seven leading toxicologists, regarding recommendations for the treatment of a patient with acute aspirin poisoning (hypothetical), the toxicologists preferred use of AC for GI decontamination whereas answers from poison centres were widely variable [28]. Nevertheless, five most common recommendations from poison centres included use of AC. The authors of this survey concluded that the recommendations were based on habit and intuition indicating that more research is needed. A recent prospective trial did not produce evidence that AC improved outcome in the mild and moderately poisoned patients compared to conventional supportive care [115]. It has been suggested therefore to re-assess the risks and benefits of using AC in such patients and perhaps restrict it to life-threatening conditions [24]. On the one hand, availability of AC for home/pre-hospital administration clearly reduces the mean time after drug injection which is potentially beneficial to the poisoned patient [101], but on the other hand there is scepticism among some pharmacists and medics regarding risks and failures of home administration of AC [109,121]. Some poison centres in the United States recommend that AC be available at home to provide pre-hospital treatment, however it has not become common practice to stock AC in pharmacy stores. Patient education materials explaining the use of AC are nonexistent and many pharmacists are not familiar with this information [128,129]. A survey in Georgia, USA, has shown that the state hospitals have improved their antidote stocking practices (including AC) since the Olympic Games in Atlanta, 1996, when several bioterrorism hoaxes took place, however the authors have concluded that no individual hospital would be able to cope with a large number of simultaneously poisoned people [130]. A survey among paramedics regarding their general knowledge of using AC in the treatment of poisoned patients demonstrated that up to 40% of answers to multiple-choice questions were incorrect, including the preferred time frame for AC administration [131]. Concerns were expressed regarding the increasing number of patients who refused to take AC [84]. According to the prospective study conducted in 1998 - 1999 this figure increased to 39% vs. 9.9% between 1984 and 1994 [132]. The trend should be reversed by both educating medical personnel and convincing patients; it has been shown that distributing information leaflets, newsletters and posters among customers, medical societies and services and pharmacy stores, producing press releases to the local TV and radio stations and newspapers helps to achieve this goal [133]. 3.1.9. Activated carbon currently available/or oral administration The source of activated carbon is usually not disclosed by the manufacturers of oral AC. The manufacturers and suppliers given in Table 4 are those companies which prepare and package AC formulations rather than produce activated carbon. According to [1], most of the carbons originate from Norit Company (Amersfoort, the Netherlands). Carbomix is an AC from

Activated Carbons as Medical Adsorbents

551

the Norit A or Norit C range and CharcoAid and Liqui-Char are also based on a Norit charcoal NoritUSPXXn[l]. Table 4 Formulations Brand name ActidoseAqua Advance (formerly ActidoseAqua)

of AC available in the UK and the US for poisoning treatment. Excipients Manufacturer/Supplier Formulation Sucrose, citric Aqueous or aqueous/ Cambridge Laboratories, Wallsend, Tyne & Wear, UK; acid, sorbitol, sorbitol oral propylene Dales Pharmaceuticals Ltd., suspension, AC 25-50 g / 120-240 ml Skipton, North Yorkshire, UK; glycol, glycerol, in a bottle or squeeze Paddock Laboratories Inc., methylparabe Minneapolis MN tube

Meadow Laboratories, Romford, Citric acid, Dry granulated or glycerol, powdered AC, 25-50 gEssex, UK; acacia in 350-500 ml bottle Penn Pharmaceuticals Ltd., or 5 g in a sachet Tredegar, Gwent UK Charcodote Oral suspension, AC Pliva Pharma Ltd., Petersfield, Water 50 g in 250 ml Hants, UK; Dominion Pharma Ltd., Haslemere, Surrey, UK; Dales Pharmaceuticals Ltd., Skipton, North Yorkshire, UK; Liqui-Char* Sorbitol/aqueous Sorbitol Oxford Pharmaceuticals, suspension, AC Harrow, UK; 25- 50 g (1:1 with Jones Pharma Inc., Bristol, TN sorbitol) in a bottle or squeeze tube Medicoal* Granulated AC, 5 g in Concord Pharmaceuticals Ltd., Sodium bicarbonate, a 10 g sachet Dunmow, Essex, UK citric acid (effervescent) Requa Inc., Bridgeport, CT CharcoAid G 'Superactivated' granulated charcoal Poison 15 g Actidose-Aqua Alliance Medical Inc., Bowman Ipecac syrup Antidote Kit charcoal suspension Pharmaceuticals Inc., Canton, and 1 oz. ipecac syrup OH EZ-Char* * Pellets, 25 g per bottle Paddock Laboratories Inc., 20% bentonite Minneapolis, MN and 80% AC * Discontinued in the UK ** Only pelleted AC that has been recommended for poisoning treatment is included in this table. Most pelleted AC are used to treat indigestion problems rather than poisoning. Carbomix

552

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

The surface area of AC used in these formulations is in the range from 950 to 1600 m g" and they are microporous carbons [1 and our own data]. 'Superactivated' powdered charcoal developed in mid-1970s, reportedly has surface area up to 2800-3500 m2 g"1 [1], although the commercial charcoal powder is more likely to have the surface area of ca. 2000 m2 g"1 [86]. A number of oral AC were developed in the former USSR and are used in the New Independent States; these include Carbolen (powdered or pelleted, Norit type), Vaulen and Belosorb (fibrous, from cellulose), SCN or SKN (microbeads, from polyvinylpyrrolidone), SKT6A (granulated, from peat) and KAU (granulated, from crushed fruit stones) [26,134]. Among all the AC formulations an aqueous slurry or suspension remains the preferred option because tablets, capsules and pastes have other ingredients which reduce the effect of the formulation compared to the pure AC per mass unit [41,135]. 3.2. Haemoperfusion over activated carbon We have already pointed out that the clinical applications of haemoperfusion (HP) in the West have been very different from the former USSR. The main explanation to these differences is in the type of AC used. All the commercial HP columns produced in the West utilise coated activated carbon, whereas in the former USSR uncoated AC has been used. Use of coated AC dramatically narrowed application of haemoperfusion almost exclusively to acute poisoning, and even in these applications clinicists are reluctant to use it. Partly this is due to the doubts regarding efficiency of this procedure in the management of acute poisoning and partly to the low rate of mortality from poisoning and thus lack of necessity although it has been acknowledged that charcoal HP is simple and safe for use in any intensive care unit that has experience of extracorporeal techniques [136]. Currently in the UK it is recommended for consideration in the treatment of patients poisoned with theophylline, paraquat, barbiturates and carbamazepine [21,136], Haemodialysis also seems to have limited use and haemofiltration is not receommended at all [21]. As a result, charcoal haemoperfusion columns are not commonly available in the poisons centres although they can be obtained from the supplier on the same day. At present the main haemoperfusion products on the market are DHP column with poly-HEMA coated charcoal (Kuraray, Osaka, Japan), Hemosorba with uncoated charcoal (Asahi Kasei Medical, formerly Asahi Medical, Japan), Adsorba C with cellulose-coated Norit RBX charcoal (Gamhro, Sweden) and Clark column with heparin-hydrogel coated charcoal (Clark R&D, Inc., Folsom, La). No characteristics of charcoals used in these columns are provided except for the total surface area (in Adsorba C column), which corresponds to the specific surface area of ca. 1000 m2 g"1 [137]. In the New Independent States of the former USSR in all the HP columns only uncoated AC has been used. Following the recommendations of the Sklifosovskii First Aid Research Institute (Moscow), haemoperfusion units were installed in the ambulance vehicles and successfully used for pre-hospital treatment of patients poisoned with organophosphates, tricyclic antidepressants, barbiturates, dichloroethane, salicylates and alkaloids [9]. Outcome of the poisoning treatment which includes haemoperfusion is determined by the level of the poison in the blood, age of the patient and time of the treatment from the poison ingestion [31]. These factors are the same as in the oral AC treatment although usually the delay of the first

Activated Carbons as Medical Adsorbents

553

haemoperfiision session is significantly longer than that of oral AC administration. Uncoated granulated AC used in the HP columns are of the same origin as those used for oral administration except for the larger size of the granules in the former [9,26,47,138-140]. The most popular uncoated AC haemoadsorbent SCN produced from polyvinylpyridine cross-linked with divinylbenzene was used in columns containing 100-350 ml of AC [9,47,139]. Duration of HP was usually 2-4 h with the blood flow rate 80-200 ml min"1 [9,31,47,141]. Most severely poisoned patients had up to four HP sessions, with 6-12 h interval between sessions and total volume of perfused blood from 4 to 95 1 [31]. As with the oral administration of AC, the earlier HP was performed the more effective it was. It has been claimed that efferent therapy, mostly haemoperfusion, reduced the mortality in acutely poisoned patients by 2.5 times [142]. It is difficult to directly compare results of haemoperfusion undertaken in the former USSR and the West, as the statistics of poisoning causes is very different and different AC are used is but in general specialists seem to be in agreement on the conditions in which HP should be given. In salicylate poisoning it is recommended to consider using HP when the level of salicylate in the blood is above 700 mg I"1 [143], and in the former USSR haemoperfusion was used in patients with salicylate level in the range of 1.14-1.5 g I"1 [9]. Mushroom poisoning particularly with Death cap mushroom Amanita phalloides has been treated with HP along with other measures [31]. All patients (n = 47) with moderate and high degree of poisoning recovered if the HP was done within 48 h after the ingestion. This number includes elderly patients. In total, in the reported study 34 patients with high degree of poisoning were treated, of whom 23 survived and 11 died. Among the survived patients were persons who received HP only on the fourth or fifth day after poisoning [31]. Similar conclusions about the efficiency of HP 30-48 h after mushroom poisoning were made in [144]. In this paper 13 patients out of 19 were reported to have survived even being treated with HP 72 hours after ingestion. In another study all eight patients poisoned with Amanita phalloides received haemoperfusion and survived [145]. The number of mushroom poisoning cases reported in the Western literature is much lower as mushroom hunting is more common in Eastern Europe [146]. In a more recent report both patients who consumed amatoxin-producing mushrooms and received oral AC followed by haemoperfusion survived [146]. In a different study four patients poisoned with mushrooms received multiple doses of AC orally and two of them died. No haemoperfusion was performed [147]. Overall, the number of cases and variety of poisonings treated by HP in the former USSR was larger than in the West [148]. For most poisons there is a consensus among the clinicians regarding the usefulness of haemoperfusion but in some cases the views are quite different. For example, oral AC and therefore HP are considered to be inefficient in the treatment of patients poisoned with organic solvents and alcohols [21]. Publications in the former USSR suggest the opposite [9]. Positive results were achieved using HP for the treatment of severe poisoning with carbon tetrachloride and dichloroethane particularly when the treatment was first applied at the pre-hospital stage [9,31]. In combination with haemodialysis, HP gave positive results in the treatment of ethylene glycol and methanol poisoning although the number of patients was too small to make definitive conclusion.

554

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

An interesting application of AC haemoperfusion in oncotherapy was suggested in [47,139]. Treatment of tumours often requires use of highly toxic cytostatic agents (chemotherapy) or radiotherapy. Consequently, patients suffer from poisoning and develop cytostatic hepatitis and toxicoses. It has been suggested to use HP over activated carbon to control local and systemic pharmacokinetics of the anitumour agents. Results of 205 haemoperfusion sessions applied to 162 oncological patients were very encouraging as HP improved clinical symptoms and reduced the severity and number of complications associated with chemo- and radiotherapy [47,139]. Administration of oral AC also has a positive effect on such patients [149,150]. In the experiments on animals HP was shown to be an efficient treatment modality of the radiation sickness disease of the III and FV grade at the early stages [58,141,151]. In these studies, 69 mongrel dogs developed radiation sickness after exposure to 525 rad (5.25 Gy) dose irradiation. The survival rate in the dogs that received haemoperfusion was 62.4% (n = 19) vs. 3.2% in the control group (n = 31) at 24 h after irradiation. 3.3. Mechanism of haemoperfusion A single session of HP usually removes 0.35-2.5% of the total amount of the toxin in the body [47,142]. Therefore although the main role of AC in haemoperfusion is direct adsorption of the target substance but the real mechanism is more complex. Apparently, HP has an effect on enzymatic and immune components of the blood and the AC surface interacts with the blood formed elements passing through the column. Adsorption of the secondary products of poisoning, bacterial toxins and radiotoxins also should be taken into account. An original idea which reveals another mechanism of therapeutic action of haemoperfusion over AC was suggested in [152]. The authors have shown that human serum albumin which is usually loaded with low MW ligands is purified ('deligandised') in the course of perfusion. Thus its own detoxifying and transport function is restored providing amplification to the detoxifying action of the HP adsorbent. These conclusions have been confirmed by the work on ibuprofen adsorption from albumin solutions on activated carbon [17]. Normally ibuprofen is strongly bound with albumin in a complex which has dissociation constant K4 = 3.0 x 10"s; nevertheless it is broken down by the carbon surface into ibuprofen molecule which is adsorbed and albumin molecule (purified) which returns into the medium. In the extracorporeal system MARS (Teraklin, Germany) currently available on the market, two circuits are used for blood purification; in the main circuit serum albumin is used as a liquid sorbent that binds toxins and in the secondary circuit albumin loaded with the ligand is regenerated by passing through a column with AC [153]. In the single-column HP system these two processes are combined in one. 3.4. Complications of haemoperfusion Complications which were common in the early days of haemoperfusion, seldom happen in the modern extracorporeal treatment. Most frequently patients undergoing HP develop mild fever, hypotension, bleeding and perturbation of the blood clotting system. Sometimes hypotension could be significant with arterial pressure dropping by 35-40%. Patients with severe poisonings followed by liver failure are particularly prone to hypotension. This reaction to

Activated Carbons as Medical Adsorbents

555

haemoperfusion is probably a body response to the external circuit connection which has a substantial volume up to 500 ml. Bleeding is most likely due to the incorrect dosage of heparin or its antagonists, whereas the nature of the increased fibrinolytic activity of blood is not clear. Sometimes it happens immediately after HP but more often within the first 24 h and continues for a few days [9,31]- These complications are usually easy to control and they rarely have any serious consequences. A very rare complication known as re-coma, or secondary coma, occurs in the patients severely poisoned with barbiturates or other psychotropie agents. It sometimes happens after successful HP when the concentration of the poison in the blood dropped below toxic level. It may be a result of the brain hypoxia caused by a rapid removal of the drug from the body. As an operation, charcoal haemoperfusion is considered simple and safe for use by experienced personnel [136]. 4. CONCLUSIONS If administered orally, activated carbon has its disadvantages mostly associated with its poor palatability. But according to the data published by the American Association of Poison Control Centers, 64 participating centres reported 2,395,582 cases of poisoning in 2003. Single use of AC has been the second most frequent mode of treatment accounting for 134,619 cases (compared to the most frequently used dilution/irrigation without using any drug at all — 1,108,359 cases and the third — cathartic — 48,839 cases); multi-dose AC was used in 5,793 cases [154]. In total, AC was administered in 5.9% of exposures. These most recent data confirm an important and yet underestimated role that activated carbon plays in the management of severe and acute poisoning. Oral use of AC is considered one of the main modalities in the treatment of persons exposed to some of the deadliest biological and chemical warfare [64,65,71]. AC was successfully used for the protection of personnel subjected to radioactive contamination in the clean-up of the Chernobyl zone. Due to its almost universal ability to adsorb most substances, activated carbon should become available to every household particularly facing the new reality of the XXI century — a threat of terrorist actions that may use biological, chemical and radiological warfare. A major technogenic (man-made) environmental disaster could also affect large territories with a big number of casualties. Ability to protect population at the earliest stages following contamination, when the nature of the contaminating agent or agents is not known yet, is absolutely crucial to reduce the potential impact of such an event. This can be done only by the people themselves and an immediate access to activated carbon is the key issue. ACKNOWLEDGMENTS Authors acknowledge financial support of the NATO Collaborative Linkage Grant LST.CLG.978860 and EPSRC (UK) Grant GR/R05154. V.G. Nikolaev acknowledges support of the Royal Society of his visit to Brighton, UK. REFERENCES [1]

D.O. Cooney, Activated Charcoal in Medical Applications, Marcel Dekker, New York, 1995 and references therein.

556 [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]

[21] [22] [23] [24] [25] [26] [27] [28] [29] [30]

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G. Nikolaev S.V. Nikolaev

V.G, Nikolaev, Biomater., Artif. Cells, Artif, Org., 18 (1990) 555. J. Steczko, S.R. Ash, D.B. Blake, D.J. Catr and R.H. Bosley, Artif. Organs, 23 (1999) 310. T.L. Beauehamp and J.F. Childress, Principles of Biomedical Ethics, 3rd ed., Oxford University Press, New York, 1989. HJ. Gurland and A.M. Davison, Nephrol., Dial, Transpl, 9 (1994) 3-3 S2. H. Klinkmann, U. Baurmeister and A.M. Davison, Nephrol., Dial, Transpl., 9 (1994) 1-2 S2. E. Denti and J.M. Walker, in: Sorbents and Their Clinical Applications, C. Giordano (ed.), Academic Press, New York- London, 1980,101. G. Dunea and WJ. Kolff, Trans. Amer. Soc. Artif. Int. Organs, 11 (1965) 178. Iu.M. Lopukhin, Hemosorption, Mosby, New York, 1979. H. Yatzidis and D. Oreopoulos, Kidney Int., 10 (1976) S215. J,D. Andrade, K. Kumitomo, R. van Wagenen, B. Kastigir, D. Gough and WJ. Kolff, Trans. Amer. Soc. Artif. Int. Organs, 17 (1971) 222. R.J. Henry, D.C. Cannon and J.W. Winkelman (eds.), Clinical Chemistry: Principles and Technics, 2nd ed., Harper & Row, New York, 1974. K.E. Hagstam, L.E. Larsson and H. Thysell, Acta Med. Scand., 180 (1966) 593. T.M.S. Chang, Kidney Int., N7 (1975) S347. J.F. Winchester, C. Ronco, J.A. Brady, J. Clemmer, T.E. Muller, V. Davankov, M. Tsyurupa, L. Pavlova, M. Pavlov andN.W. Levin, Contrib. Nephrol., 133 (2001) 131. P.A. Routledge, Br. J. Pharmacol., 22 (1986) 99. M. Melillo, GJ. Phillips, J.G. Davies, A.W. Lloyd, S.R. Tennison, O.P. Kozynchenko and S.V. Mikhalovsky, Carbon, 42 (2004) 565. V.V. Strelko, Yu.F.Korovin and V.G. Nikolaev, in Hemoperfusion and Artificial Organs, E. Piskin and T.M.S. Chang (eds.), Meteksan Publ. Co., Ankara, Turkey 1982,144. V.G. Nikolaev and K.S. Ternovoy, in Hemoperfusion and Artificial Organs, E. Piskin and T.M.S. Chang (eds.), Meteksan Publ. Co., Ankara, Turkey 1982, 200. V.G. Nikolaev, L.B. Pinchuk, M.A. Umansky, V.G. Pinchuk, T.N. Burushkina, S.V. Petrenko, N.K. Rodionova, E.A. Snezhkova and N.P. Bichkova, Artif. Organs, 17 (1993) 362. A.L. Jones and P.I. Dargan, Toxicology, Harcourt Publ., Edinburgh, UK, 2001. http://www.pssg.org/forsyth.htm S.V. Mikhalovsky, Perfusion, 18 (2003) S47. D. Seger, J. Toxicol., Clin. Toxicol., 42 (2004) 101. D.M. Wood, E. Webster, D. Martinez, P.I. Dargan and A.L. Jones, Crit. Care, 6 (2002) 455. N.A. Belyakov (ed.), Enterosorption (in Russian), Centre of Sorption Technologies, StPetersburg, Russia, 1991. S.M. Bradberry and J.A. Vale, J. ToxicoL, Clin, Toxicol., 33 (1995) 407, D.N. Juurlink and M.A. McGuigan, J. Toxicol., Clin. Toxicol., 38 (2000) 465. R.D. Hofbauer and J.S. Holger, J. Toxicol., Clin. Toxicol., 42 (2004) 61. M.L. Frenia, J.L. Schauben, R.L. Wears, J.L. Karlix, C.A. Tucker and T.A. Kunisaki, J. Toxicol., Clin. Toxicol., 34 (1996) 169.

Activated Carbons as Medical Adsorbents

[31] [32] [33] [34] [35] [36]

[37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52]

[53] [54]

557

L.V. Usenko (ed.), Hemosorption in the Treatment of Acute Exotoxicoses (in Russian), Zdorov'ya Publ., Kiev, Ukraine, 1986. D.L. Kurta, L.B. Myers and E.P. Krenzelok, J. ToxicoL, Clin. ToxicoL, 35 (1997) 453. M.R. Welber and S. Nevins, J. Emerg. Med., 13 (1995) 199. D. Reith, J.P, Monteleone, I.M. Whyte, W, Ebelling, N.H. Holford and G.L. Carter, Ther, Drug Monit, 20 (1998) 92. K. Hoppu, T. Tikanoja, P. Tapanainen, M. Remes, O. Saarenpaa-Heikkila and K. Kouvalainen, Lancet, 338 (1991) 538. American Academy of Clinical Toxicology (ACCT) and the European Association of Poisons Centres and Clinical Toxicologists (EAPCCT) Position Statement and Practice Guidelines on the Use of Multi-Dose Activated Charcoal in the Treatment of Acute Poisoning, J. ToxicoL, Clin. ToxicoL, 37 (1999) 731. S. Kucukguclu, Y. Tuncok, H. Ozkan, H. Guven, A. Uguz and F. Maltepe, J. ToxicoL, Clin. ToxicoL, 34 (1996) 83. N.A. Minton and J.A. Henry, J. ToxicoL, Clin. ToxicoL, 33 (1995) 43. J.N. Love, M. Sammon and J. Smereck, J. Emerg. Med., 27 (2004) 49. H.T. Milhorn, Amer Family Phys., 43 (1991) 1293. AACT/EAPCCT Position Statements: Single-Dose Activated Charcoal, J. ToxicoL, Clin. ToxicoL, 35 (1997) 721. P.A. Chyka, J. ToxicoL, Clin. ToxicoL, 33 (1995) 399. B.T. Ly, C, Rangan, M. Ingels, S.R. Williams and R.F. Clark, J. ToxicoL, Clin. ToxicoL, 38 (2000) 550. C. Greethong, W. Tai and A.A. Salvucci, J. ToxicoL, Clin. ToxicoL, 37 (1999) 624. P.A. Lehmann and L. Favari, L. Arch. ToxicoL, 57(1985) 56, B. Vantroyen, J.-F. Heilier, A. Meulemans, A. Michels, J.-P. Buchet, S. Vanderschueren, V. Haufroid and M. Sabbe, J. ToxicoL, Clin. ToxicoL, 42 (2004) 889. V.G. Nikolaev, Haemocarboperfusion in Experiment and Clinic (in Russian), Naukova Dumka Publ., Kiev, Ukraine, 1983 and references therein. LA. Tarkovsaya, Oxidised Carbon (in Russian), Naukova Dumka Publ., Kiev, Ukraine, 1981. V.K. Savchenko, The Ecology of the Chernobyl Catastrophe, UNESCO and Parthenon Publishing Group, Paris - Canforth, Lanes. - New York, 1995. V.G. Baryakhtar (ed.), Chernobyl Catastrophe, Naukova Dumka, Kiev, Ukraine, 1996. I.K. Dedenko, A.V. Starikov and V.V. Strelko, Efferent Methods of Treatment of Radiation Related Diseases (in Russian), ISPE, Kiev, 1996. I.K. Dedenko, A.V. Starikov, V.F. Torbin and V.M. Postrelko, Efferent Methods of Treatment of Radiation and Toxic Encephalopathy (in Russian), Nora-Print, Kiev, Ukraine, 1998. K.N. Prasad, Human Radiation Biology, Lippincott Williams and Wilkins, Philadelphia, PA, 1974. J.J, Mangano, Low Level Radiation and Immune System Damage: An Atomic Era Legacy, CRC Press, London, 1998.

558 [55] [56]

[57] [58] [59] [60] [61] [62] [63]

[64]

[65]

[66] [67]

[68] [69] [70] [71] [72] [73] [74] [75] [76] [77]

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

A.M. Kuzin, Foundations of Radiation Biology (in Russian), Atomizdat Publ., Moscow, 1964. E.B. Burlakova and V.I. Naidich, The Effects of Low Dose Radiation: New Aspects of Radiobiologioal Research Prompted by the Chernobyl Nuclear Disaster, VSP International Science Publishers, 2005. R.L. Hugg, R. Trautman and D.C. Van Dyke, Amer. J. Physiol., 161 (1950) 56. K.S. Ternovoy (ed.), Hemosorption in the Treatment of Acute Radiation Sickness (in Russian), Naukova Dumka Publ., Kiev, Ukraine, 1983. J. Lahaye, G. Nanse, A. Bagreev and V. Strelko, Carbon, 37 (1999) 585. L.R. Radovic, C. Moreno-Castilla and J. Rivera-Utrilla, in: Chemistry and Physics of Carbon, Vol. 27 (L. R. Radovic, ed.), Marcel Dekker, New York, 2001, 227. Y.F. Jia and K.M. Thomas, Langmuir, 16 (2000) 1114. Y.F. Jia, B. Xiao and K.M. Thomas, Langmuir, 18 (2002) 470. A.E. Pol'shina, A.M. Puziy, E.V. Pol'shin, R.G.A. Faragher and S.V. Mikhalovsky, in Fundamentals of Adsorption, M.D. LeVan (ed.), Kluwer Academic Publ., Boston, MA, 1996, 741. E.M. Eitzen, Jr., in: Textbook of Military Medicine: Medical Aspects of Chemical and Biological Warfare, Office of the Surgeon General, Department of the Army, Bethesda, MD, USA, 1997,208. Treatment of Biological Warfare Agent Casualties, Field Manual, Headquarters, Departments of the Army, the Navy, and the Air Force, and Commandant, Marine Corps, Washington, DC, 2000. A.V. Grigoryev, V.A. Znamenskiy and V.I. Bugaev, Microbiol. J. (in Russian), 63 (1991) 98. A.V. Grigoryev, N.T. Kartel, V.M. Bondarenko, V.A. Znamensky, L.G. Kupchinsky, M.E. Shor-Chudnovsky, S.L. Medvedev, E.V. Sudzhian, L.N. Rachkovskaya and I.M. Samodumova, J. Microbiol., Epidem., Immunol, (in Russian), 7 (1991) 11. M.E. Shor-Chudnovsky, N.T.Kartel and A.V. Grigoryev, Clin. Surg. (in Russian), 9 (1988) 21. V.B. Fenelonov, Y.L. Chernov, L.N. Rachkovskaya, G.N. Kryukova, O.A. Mashkov, V.Y. Gavrilov and G.V. Plaksin, React. Kinet. Catal. L., 25 (1984) 225. I.K. Dedenko, V.A. Litvinyuk and V.F. Torbin, Efferent Methods of Treatment of Food Toxicoinfections (in Russian), Nora-Print, Kiev, Ukraine, 1998. Working Group on Civilian Biodefense, Consensus Statement: Botulinum Toxin as a Biological Weapon. Medical and Public Health Management, JAMA, 285 (2001) 1068. E.P. Krenzelok, Toxicol. Letters, 127 (2002) 299. AACT/EAPCCT Position Paper: Cathartics, J. Toxicol., Clin. Toxicol., 42 (2004) 243. AACT/EAPCCT Position Paper: Gastric Lavage, J. Toxicol., Clin. Toxicol., 42 (2004) 933. O. Lapatto-Reiniluoto, K.T. Kivisto and P.J. Neuvonen, Br. J. Clin. Pharmacol., 48 (1999) 148. AACT/EAPCCT Position Paper: Ipecac Syrup, J. ToxicoL, Clin. Toxicol., 42 (2004) 133. T.E. Albertson, R.W. Derlet, G.E. Foulke, M.C. Minguillon and S.R. Tharratt, Ann. Emerg. Med., 18 (1989) 56.

Activated Carbons as Medical Adsorbents

[78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91 ]

[92] [93] [94] [95] [96] [97]

[98] [99] [100]

[101] [102] [103]

559

AACT/EAPCCT Position Paper: Whole Bowel Irrigation, J. Toxicol., Clin. Toxicol., 42 (2004) 843. A.T. Proudfoot, E.P. Krenzelok and J.A. Vale, Position Paper on Urine Alkalinization, J, Toxicol., Clin. Toxicol., 42 (2004) 1. M.L. Frenia, J.L. Schauben, R.L. Wears, J.L. Karlix, C.A. Tucker and T.A. Kunisaki, J. Toxicol., Clin. Toxicol., 34 (1996) 169. D.O. Cooney, J. Toxicol., Clin. Toxicol., 33 (1995) 213. Y. Wakabayashi, S. Mamyama, K. Hachimura and T. Ohwada, J. Toxicol., Clin. Toxicol., 32 (1994) 419. G. Levy, New England J. Med., 307 (1982) 676. R.M. Lynch and R. Robertson, Accident Emerg. Nursing, 11 (2003) 63. M. Tenenbein, R. Green, R. Grierson and D.S. Sitar, J. Toxicol., Clin. Toxicol., 37 (1999) 610. R.L. Sato, JJ. Wong, S.M. Sumida, R.Y. Marn, N.R. Enoki and L.G. Yamamoto, Am. J. Emerg. Med,, 21 (2003) 189, B. Mokhlesi, J.B. Leiken, P. Murray and T.C, Corbridge, Chest, 123 (2003) 577. A.J. Jones and G. Volans, Br. Med. J., 319 (1999) 1414. S. Wason, R.C. Baker, P. Cardan, R. Seigel and R.W. Druckenbrod, J. Toxicol., Clin. Toxicol., 30 (1992) 39. S.M. Pond, K.R. Olson, J.D. Osterloh and T.G. Tong, J. Am. Med. Assoc, 251 (1984) 3104. H.A. de Silva, M.M.D. Fonseka, A. Pathmeswaran, D.G.S. Alakahone, G.A. Ratnatilake, S.B. Gunatilake, CD. Ranasinha, D.G. Lalloo, J.K. Aronson and H.J. de Silva, Lancet, 361 (2003) 1935. P.E. McKinney, R. (Milan and W.A. Watson, J. Toxicol., Clin. Toxicol., 30 (1992) 549. R.D. Hofbauer and J.S. Holger, J. Toxicol., Clin. Toxicol., 42 (2004) 61. A. Hagiwara, T. Takahashi, O. Kojima, K. Sawai, T. Yamaguchi, T. Yamane, H. Taniguchi, K. Ritamura, A. Noguchi, K. Seiki and C. Sakahura, Lancet, 339 (1992) 629. A.F. Vozianov, N.A. Kolesnik and I.M. Samodumova, Vrachebnoe Delo (in Russian), 2 (1992) 90. D.I. Shvets, T.I. Denisova and V.V. Strelko, Central European J. Chem., 1 (2003) 178. A.A. Chuiko, O.O. Pentyuk and V.K.. Pogorelyi, in; Chemistry, Physics and Technology of Surfaces, Issues 4-6, A.A. Chuiko (Ed.), Institute of Surface Chemistry, Kiev, Ukraine (2001) 290. J.L. Sandier, T.F. Nealon, Jr. and J.H. Gibbon, J. Am. Med. Assoc, 179 (1962) 201. V.V. Strelko and A. Clearfield, Abstr. Pap. Am. Chem. Soc., 217 (1999) 252 -IEC Part 1. L.N. Rachkovskaya, I.I, Frolova, A,K. Rovina, Y.L. Chernov, G.V. Plaksin, V.D. Sokolovskii, N.N. Prutovykh, A.P. Shevchenko, O.A. Mashkov and V.N. Kokhno, Biomater., Artif. Cell., 15(1988)701. A.O. Alaspaa, MJ. Kuisma, K. Hoppu and PJ. Neuvonen, Ann. Emerg. Med., 45 (2005) 207. L.S. Mauro, J J. Nawarskas and V.F. Mauro, Ann, Pharmacother., 28 (1994) 915. T. Nakamura, Y. Oida, K. Matsumoto, N. Kawasaki and S. Tanada, J. Environ. Sci. Health. A - Tox. Hazard. Subst. Environ. Eng., 37 (2002) 905.

560 [104] [105] [106] [107] [108] [109] [110] [Ill] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132]

S. V.Mikhalovsky Mikhalovskyand andV.G. V.G.Nikolaev Nikolaev S.V.

CD. Curd-Sneed, J.G. Bordelon, K.S. Parks and J.J. Stewart, J. Toxicol. din. Toxicol, 25 (1987) 555. A. Eroglu, U. Kucuktulu, N. Erciyes and H. Turgutalp, J. Toxicol., din. Toxicol., 41 (2003) 71. C.L. Dorrington, D.W. Johnson and R. Brant, Ann. Emerg. Med., 41 (2003) 370. J. Liisanantti, P. Kakoranta, M. Martikainen and T. Ala-Kokko, Resuscitation, 56 (2003) 49. H.A. Spiller and G.C. Rodgers Jr., Pediatrics, 108 (2001) E100. American Academy of Pediatrics Policy Statement: Poison Treatment in the Home, Pediatrics, 112 (2003) 1182. G.R. Bond, Pediatrics, 109 (2002) 145. C. Keyes and L. DeTamble, J. Toxicol. Clin. Toxicol., 37 (1999) 610. T.C. Arnold, B.H. Willis, F. Xiao, S.A. Conrad and D.L. Garden, J. Toxicol. Clin. Toxicol., 37(1999)1. C. Tomaszewski, J. Toxicol. Clin. Toxicol., 37 (1999) 17. R.L. Sato, J.J. Wong, S.M. Sumida and L.G. Yamamoto, Hawaii Med. J., 61 (2002) 251. K.S. Merigian and K.E. Blaho, Am. J. Ther., 9 (2002) 301. A.M. Thrash and C. Thrash, Home Remedies, Yuchi Pines, Seale, Alabama, 1981; R.C Kaufman, J. MegaHealth Soc, 5 (1989) N3 (issue 23). L. West, Pediatric Nursing, 23 (1997) 616. M.A.Q. Curley, J.B. Smith and P.A. Moloney-Harmon (eds.), Critical Care Nursing of Infants and Children, W.B. Saunders Co., Philadelphia, PA, 1996. S. Smith, D. Cobaugh, V. Dragalin, P. Wax and R. Lawrence, J. Toxicol. Clin. Toxicol., 37 (1999)611. C. Rangan, S.P. Nordt, R. Hamilton, M. Ingels, and R.F. Clark, Ann. Emerg. Med., 37 (2001) 55. E.J. Scharman, H.A. Cloonan and L.F. Durback-Morris, J. Emerg. Med., 21 (2001) 357. L.C. Hoegberg, H.R. Angelo, A.B. Christophersen and H.R. Christensen, Pharmacol. Toxicol., 93 (2003) 233. T.F. Eisen, P.A. Grbcich, P.G. Lacouture, M.W. Shannon and A. Woolf, Ann. Emerg. Med., 20 (1991) 143. G.K. McEvoy (ed.), American Hospital Formulary Service Drug Information, Am. Soc. of Health-System Pharmacists, Bethesda, MD, 2004. H.A. Spiller, J. Emerg. Med., 22 (2002) 106. A. Lamminpaa, J. Vilska and K. Hoppu, Hum. Exp. Toxicol., 12 (1993) 29. M.A. McGuigan, Clin. Pediatr. Emerg. Med., 1 (2000) 191. R. Rose and E. Waring, J. Toxicol, Clin. Toxicol,, 38 (2000) 561. L. McGoodwin and S. Sehaeffer, J. Toxicol. Clin. Toxicol., 38 (2000) 562. F. Manney, RJ. Geller, M.L. Tarantino and C.S. Parramore, J. Toxicol. Clin. Toxicol., 38 (2000) 563. A. Hoifiiung, P. Wax, S. Schneider and R. Lawrence, J. ToxicoL Clin, Toxicol., 37 (1999) 598. A.F. MacNamara, M.S. Riyat and D.N. Quinton, J. Roy. Soc. Med., 89 (1996) 608.

Activated Carbons as Medical Adsorbents

[133] [134] [135] [136] [137] [138] [139] [140]

[141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151]

[152] [153]

[154]

561

G. Banach and H, Kuhl, J. Toxicol. Clin. Toxicol, 37 (1999) 599. V.V. Strelko, Y.F, Korovin, N.T. Kartel and A.B. Shcherbitskii, J, Applied Chem., (in Russian), 57(1984)1134, H.P, Remmert, M. Oiling, W. Slob, W.F. van der Giesen, A, van Dijk and A.G. Rauws, Eur, J. Clin. Pharmacol., 39 (1990) 501. D. Webb, Br. J. Hospital Med., 49 (1993) 493. Specification for Adsorba C hemoperfusion cartridge, Gambro Renal Products, Lund, Sweden, 2002. S.P. Valueva, N.M, Kaznacheeva, S.I. Surinova, B.S. Eltsefon, E.M. Kopylova, R.I. Pertsov and G.A. Trikhanova, Chem. & Pharm. J. (in Russian), 17 (1983) 105. V.G. Nikolaev and V.V. Strelko, Hemosorption on Activated Carbons, Naukova Dumka Publ, Kiev, Ukraine, 1979. S.V. Mikhalovsky, in: Microspheres, Mierocapsules and Liposomes, The MML Series, Vol. 2: Medical & Biotechnology Applications, R. Arshady (ed.), Citus Books, London, 1999, 133. V.G. Nikolaev (ed.), Modem Methods of Sorption Therapy in Clinical Practice (in Russian), Ukrainian Academy of Sciences, Kiev, 1998. Yu.M. Lopukhin, A.S. Parfenov and D.V. Kulaev, Efferent Therapy (in Russian), 1 (1995) 14. D.G. Richlie and RJ. Anderson, Seminars in Dialysis, 9 (1996) 257. O, Bartels, Acta Hepato-GastroenteroL, 25 (1978) 324. J.P. Wauters, C. Rossel and J.J. Farquet, Br. J. Med., 2 (1978) 1465. D.A. Feinfield, H.C Mofenson, T. Caraccio and M. Kee, J. Toxicol., Clin. Toxicol., 32 (1994) 715. S. Zevin, D. Dempsey and K. Olson, Morbidity & Mortality Weekly Report, 46 (1997) 489. V.G. Nikolaev, Int. J. Artif. Organs, 7 (1984) 289. L.V. Bonatskaya, V.M. Plotnikov and V.G. Nikolaev, Experimental Oncology (in Russian), 11(1989)71. G,V. Muravskaya, V.G, Nikolaev, V.P, Sergeev, N.I. Rrutilina, L.V. Bonatskaya, V.N. Klevtsov, V.V. Surovikina and V.V. Sinajko, Biomater., Artif. Cell, 19 (1991) 167. V.G. Nikolaev, L.B. Pinchuk, M.A. Umansky, V.G. Pinchuk, T.N. Burushkina, S.V. Petrenko, N.K.. Rodionova, E.A. Snezhkova and N.P. Biehkova, Artif. Organs, 17 (1993) 362. V.V.Sarnatskaya, W.E. Lindup, A.I. Ivanov, L.A. Yushko, J. Tjia, V.N. Maslenny, N.M. Gurina and V.G. Nikolaev, Nephron Physiol., 95 (2003) 10. S.R. Mitzner, J. Stange, S. Klammt, T. Risler, CM. Erley, B.D. Bader, E.D. Berger, W. Lauchart, P. Peszynski, J. Freytag, H. Hickstein, J. Loock, J.M. Lohr, S. Liebe, J. Emmrich, G. Korten and R. Schmidt, Liver Transplant., 6 (2000) 277. W.A. Watson, T.L. Litovite, W. Klein-Schwartz, G.C. Rodgers, J. Youniss, N. Reid, W.G. Rouse, R.S. Rembert and D. Borys, Am. J. Emerg. Med., 22 (2004) 335.

This Page is Intentionally Left Blank

563

INDEX

abrasion index, 35, 36, 37, 39 accessibility of pores, 286, 312, 3449, 354, 357 acidic character, 162, 206, 207, 210, 356, 357, 359, 366, 368 acidic groups, 162,173, 179,181, 182, 185-187, 209, 211, 243, 244, 285, 391, 539, 540 acidity constants, 183 activated carbon based electrodes, 301 activated carbon cloth, 486 activated carbon fibers, 52, 53, 55, 57, 61, 63, 77, 78, 81,237, 241,243, 244,281, 425,457, 459,461 activated carbon filters, 415, 480 activated carbon microbeads, 486 activation, 14,15, 20, 21, 27, 28, 32,109111,151,397,398,406 activation energy, 22 active sites, 160,162,165,174,208, 212 Adams-Bohart-Thomas method, 378 Adamsite, 477 adsorber, 381, 393, 394, 396, 397, 400, 402, 408-411, 416, 426, 449 adsorption, 183, 215, 220, 226, 529, 531, 533, 538-542, 545-549, 554 adsorption ealorimetry, 61, 62 adsorption capacity, 8, 9,10,11, 12,13, 22,23, 26,28, 31, 32, 33, 35, 39, 40, 41, 44, 46,109,128,166,183,212, 213,240, 241,245, 251,257, 258,278, 279,283, 376, 379, 383, 384, 390, 391, 395, 399, 400,403, 406 adsorption characteristics, 387 adsorption energy, 111,119,131,491 adsorption exothermic heat, 400 adsorption forces, 215, 492, 509, 523 adsorption isotherms, 22-25, 28, 32, 34, 109, 111, 13,116-118,121-124,130136,140-148,151, 429, 443, 462 adsorption mechanisms, 498

adsorption/oxidation 244, 248, 250, 252, 252, 254,257, 259,260, 264,274, 277, 278, 279,281 adsorption potential, 109,130-43,147,148, 152,153 adsorption sites, 434,461 aerosol filters, 484 air pollution, 231, 232, 262 alcohols, 193, 207, 216 algal toxins, 432 alkali metal, 75, 76, 91 alpha plot, 57, 59 amines, 163,164, 182,216, 256, 399 ammonia, 493, 495, 506, 510 anionic dopant, 319 anthracite, 22, 23, 34 antibacterial activated carbons, 437 aqueous solution adsorption, 376 arsine, 478,498 asymmetric capacitors, 322, 342 atomic force microscopy, 50, 51 atrazine adsorption, 362,362, 366,389 autocatalysis, 251, 259 autocatalysis of sulfur, 402 B backwashing, 427, 434,442 bacteriostatic action, 461 Barrett-Joyner-Halenda, 57,144,146 basic groups, 181, 182,216, 243,245, 253, 255, 274,280, 390, 391, 454 basic oxides, 162 basicity, 11,243, 244-247,273, 274,281 basicity of carbons, 162,213 bed depth -service time, 395 bed geometry, 522 BET, 52, 55, 57, 60, 76, 82, 83, 86, 93, 112,118-120,124, 128,132,303,304, 309,310,311,313,314,334 binder, 37,298 binding energy, 195 biocompatibility, 530, 531 biodegradation, 359 biofiltration, 441 biological regeneration, 466

564

Index

bioterrorism, 550 biotoxins, 541, 542 bituminous coal, 17,24, 26 blood agents, 478, 500 blood purification, 530, 531, 534, 554 Boehmtitration, 179-183, 185, 202, 209, 274 boron-doped carbon ,169 boundary conditions, 267 breakthrough 484,491, 500, 501, 503, 506, 507, 509, 512-526 breakthrough concentration, 426 breakthrough curves, 382,383,426, 512, 516,518 breakthrough plot, 44 breakthrough time, 382, 392,393, 395 breathing resistance, 487, 488, 507, 510, 511,513,514,522 Broekhoff and de Boer equation, 144 Brunauer-Emmett-Teller (BET) 7, burn-off, 22, 23, 24,26, 33, 35,40, 42

calcium hydroxide, 243 calorimetric methods, 61, 204,205, 208 capacitance, 295-319, 320, 322, 324, 331, 337, 338

capillary condensation, 116,165 car canisters, 80 carbon basal planes, 162 carbon-based monolithic structures 42 carbon-based recovery systems, 107 carbon blacks, 110 carbon dioxide, 425,445, 446 carbon dioxide activation, 23 carbon fabrics, 41 carbon fibres, 3,35,37,39,40,42,107108,110,121,326 carbon films, 42 carbon filters, 4, 426,446, 492,494, 497, 505, 514 carbon foams, 98 carbon functionalization, 177 carbon-halogen complex, 167 carbon modified with urea, 256, 273 carbon monolith, 511 carbon nanopipes, 88 carbon nanotubes, 43

carbon-oxygen functional groups, 160,194 carbon precursors, 109 carbon surface, 161-163,166, 168,170, 171,174,176-181,185-191,197, 200217, 437, 438, 440, 449, 451, 454, 457, 490, 492, 495, 496, 501, 503, 525 carbonization, 4,14,15, 17,18, 19, 20, 22, 27, 31, 33, 34, 36,42, 43,109-112,151 carboxylic acids 164,180, 182,184, 194, 210,211,216,399,400 catalysis, 22,36 catalyst impregnated activated carbons, 466 catalyst support, 217 catalysts, 492, 499, 501, 503, 504 catalytic activity, 253, 257, 259,274, 496, 504, 506 catalytic oxidation, 233, 249,251, 257, 264, 277 catalytic properties, 112,163,188,214, 216 catalytic reaction, 216,217,423,431 cation exchange reactions, 216 caustic impregnated carbons, 250,255 char gasification, 14 characteristic adsorption curve, 130-134, 147 charge density, 170 charge storage capacity, 309 chelating therapy, 539 chemical activation, 4,14, 27,28, 29, 33, 36,38,41,109 chemical engineering models, 514 chemical oxidation, 181 chemical protection suite, 486 chemical warfare, 475,476, 478, 479,489, 490, 494, 500, 502, 508, 511, 516, 520, 524, 526 chemisorbed oxygen, 186, 202,265 chemisarption,113, 115,164, 165,171, 217,242,243, 247,248, 423,45, 484, 490,493,494, 499, 501, 502, 504, 506, 508, 509, 513, 515, 519, 520, 522, 525, 542 chlorination, 197 chlorine, 499 chloroethylenes adsorption, 392 chlorophenols, 396

565

Index

chromium, 466 choking agents, 478,498 classification of pores, 50 Clean Air Act, 232 CMK-1, 86, 87, 89 CMS, 426,460 CNC1,475, 478, 494-497, 502-504, 511, 513, 524, 525 CNTs, 308, 318, 322, 324 coagulation, 347, 348, 352, 356 colloid imprinting method, 90,112 competitive adsorption, 270, 359 composite materials, 98 contact time, 361,362, 368 conversion rate, 262,281 copper, 421,455, 458,466 covalent bonding, 326 Craston and Inkley method, 144 CS, 477, 478, 491 CS2,282 CWAs, 475-478, 480, 494, 498, 520, 526, 527 cyanogen chloride, 478,494, 497 cyclic voltammetry, 299, 300,312, 322 cylindrical pores, 115,145

dipole-dipole interactions, 454 disinfection, 425,432, 436,437, 445 dispersive interaction, 4, 7,186,189,203, 215,238 disposal methods, 397 dissociation of H2S, 254,256 dissociatively adsorbed, 254, 256 dissolution, 254 dissolved organic carbon, 347 dissolved oxygen, 356, 357, 358, 367, 368 Dollimore and Heal method, 144 DOM, 345, 347, 351-368 double layer, 295315, 322, 329, 331, 333, 334, 337,338 double layer capacitors, 80,171 drinking water, 424,431,432,433,435, 436, 437, 439, 441, 464 Dubinin-Astakhov method, 135, 152 Dubinin-Radushkevich (DR) equation, 7, 57,135, 140,142,240, 328, 490,491, 508,515,518-520,523 dye removal, 439 dynamic adsorption, 383, 3865, 392 dynamics of water adsorption, 268 dynamic systems, 281

D

E

decolorization, 396,424, 428,44,445, 463 degree of activation, 305 demineralisation, 213 density, 4,12,14,17,19,29, 36, 37, 39 density of basic groups, 245 desorption of gas, 413 desulfurization, 214,217, 218,232,233, 237, 261,262, 271,282-287, 446, 451 detoxifying action, 541, 545, 554 devolatilization, 14,17, 19 dibenzothiophene, 283,285,286 differential adsorption energies, 238,240 differential enthalpy, 109, 133,134, 152 differential entropy, 109, 133,134,152 differential scanning calorimetry, 62 diffusion, 20, 21, 24, 25, 27, 37,41, 45,46 diffusion control, 20, 21, 24,25 diffusion kinetics, 380 dimethyldisulfide (DMDS), 235, 272, 275282 dioxins, 425, 447,450, 451, 452

EDLCs, 80, 81, 85, 91, 295, 297, 302, 309, 310 effects of porosity, 243, 272,277,285 effects of porous texture, 302 electrical conductivity, 294,298, 308, 320, 238 electrical double layer, 159,170,171,297, 304, 329 electroactive groups, 178 electrochemical behavior, 169,170,173175,186 electrochmeical capacitors, 292,295, 298, 312

electrochemical polarization, 316 electrochemical techniques, 177 electrodeposition of water, 322, 326, 337, 339 electrolyte concentration, 297,298 electromagnetic energy, 198 electron donor or receptor, 490 electron transfer processes, 174

566

Index

electron transfer reactions. 171-176,274 electronically conducting polymers, 318 electrostatic interaction, 170,186 elemental analysis, 179 elemental sulfur, 233,234,250,251,256262, 266,268,272, 281,451, 454 emission control, 399,409,410 emission of mercury, 168 end-of-service indicators, 488 energetic heterogeneity, 109,151 152 energy density, 293 energy storage, 293, 294, 295, 338 enthalpies of immersion, 205-207 ESLIs, 489 ESR spectroscopy, 199 exfoliated graphite, 50, 63, 64, 65, 66, 67, 68, 71, 72, 73, 99 exfoliation, 64, 68, 71, 72, 73, 74, 99 exhausted carbons, 44,46 external mass transfer, 383 extracorporeal devices, 530

Fe2O3, 258, 281 filtration, 426-439,443-461 filtration cake, 429 fixed-bed adsorbents, 389,398 flow rate, 483, 487, 509, 515, 517, 520 fractal dimension, 52, 55, 60, 61, 65, 66 free energy of adsorption, 208 Freundlieh equation, 378, 384 fuel, gaseous, 261 Fullerenes, 4 functional groups 159-217, 347, 355-359, 433 G GAC, 422,424,-428,431-443, 446,453, 457, 461,463,464, 465 GAC filters, 348, 359, 361 GAC sandwich filter, 437 galvanostatic charge, 299, 332 gas adsorption, 110,113 gas-adsorption calorimetry, 205 gas phase adsorption, 305, 4222,424,425, 447, 460,461 gas masks, 480, 481, 483, 488, 489, 527 gas purification, 447,448, 454

gas separation, 79 gasification, 4, 21,24,27, 35,40, 41, 93, 99,311,312,464 gas-phase application, 4,475,492 grand canonical Monte Carlo simulations, 149 granular activated carbons, 35 graphene layers, 160-164,1689, 174,175, 199,205,315,316,331 graphite flakes, 71,99 graphite planes, 4,40 graphitic carbons, 2, 3 graphitization, 3 grinding and classification of the starting material, 29 ground water treatment, 431, 442 H H2S 165, 186,193, 216, 251-254, 265-269, 271-280,493,507,511 H2S breakthrough capacity, 252, 255,265, 269 halogenated species, 197,432 halogenated volatile organic compounds, 442 hard and soft acids and bases, 163 hardness, 12, 14,17, 19, 22, 29, 36, 37, 39, 347 hazardous waste, 349,444 HC1,493, 498, 499, 502, 503, 507 HCN, 475,478, 493,496, 500-504, 511, 513, 52 heat of adsorption, 405 heat capacity of sorbent, 406 heat of immersion, 62,109,152 heat treatment 29-33,164, 167,192, 194, 202 heavy metals, 425,438, 446, 451 hemoperfusion, 529-535, 541, 544, 552, 553, 554 heteroatoms, 5, 6, 9,159,160,165,174, 175,179,181, 182,197,199, 202, 355, 389 heterogeneity, 107,18, 111, 119,134,135, 140,141,153 HF, 494, 505, 507 high-energy adsorption sites, 215,216 high-pressure adsorption, 152

567

Index

Horvath-Kawazoe, 57,144,146 honeycomb carbon filters, 459 honeycomb internal structure. 457 humic substances, 347, 356, 461 humidity, 503, 504,506, 507, 509, 522, 524, 525 HVAC, 452 hybridization, 4 hydrodesulfurization, 233, 236, 283 hydrogen adsorbed, 326, 327, 331, 325 hydrogen adsorption capacity 326, 327 hydrogen bonding, 208,215,348,355, 454, 509 hydrogen cyanide, 478, 500 hydrogen storage, 293,294, 325, 326, 328, 329, 334, 335, 337, 338, 339 hydrogen sulfide, 163,186,214, 216, 231280

hydrogen sulfide ion, 251,251,254,260 hydrolysis, 486,490, 492, 498,499, 500, 503, 505-508, 519, 522 hydrophilicity, 10,110, 205, 252, 260, 283, 357, 358, 366, 387 hydrophilic surface, 454 hydrophobic carbons, 215 hydrophobic components, 355, 367 hydrophobic functional groups, 306, 307 hydroxyl radicals, 275 hysteresis, 116,117

impendance spectrscopy, 299 impregnants, 250,281,490,494,495, 4907,499,505,510,513,515 impregnation, 10, 28, 33, 36,41,437, 455, 456 impregnation with copper, 257 industrial applications, 425 industrial gas filters, 510 influence of temperature, 519, 520 infrared spectroscopy, 179, 189, 208 inorganic contaminants, 432 inorganic matter, 13,203, 213, 245,251, 257,258, 277, 286 intraparticle pores, 49 intraparticular diffusion, 392 intraparticular mass transfer, 384

intraparticular tortuosity, 406 interactions, 160,169, 170,174, 187, 207, 209,210,213,215,217 intercalation compounds, 50, 71, 75, 76 intermolecular interactions, 240 internal surface, 109 interparticle pores, 49 intramolecular charge repulsion, 357 inverse gas chromatography, 208 ion exchange, 539,541,547 ion-exchangers, 547 ionic bonding of hydrogen, 326 ionic diffusion, 319 ionic effective dimensions, 305 iron oxide, 257,274, 466 irreversible capacity, 174,177 isoelectric point(IEP), 186-1188 isosteric heats of adsorption, 251

Jaroniec-Choma equation, 141,153 K K2CO3, 250, 281 Kelvin equation, 144, 146-149 kinetic curve, 376, 378 kinetics of adsorption, 241,262, 283, 345, 366,368,511,517,522,523 KIO3, 281 KOH, 33, 34, 35, 36, 39, 249, 256, 281, 3003, 303, 305, 313, 314, 327, 329, 330,331,333,336,336

Langmuir equation, 378, 406, 515 layer-by-layer mechanism.,121 LDFT, ISO Lenard-Jones potential, 150 lethal agents, 477,478 linear driving force model, 384, 393, 405 linear-sweep voltammery, 177 liquid phase applications, 35, 36, 108, 425, 427 liquid repellant capabilities, 485 lithium insertion, 174 lithium storage capacity, 177

568

LSD25, 478 M macroporous carbons,6, 8, 96, 108,109, 392 magnesium oxides, 257 mass balance equation, 376, 379, 384, 515, 517 maximal adsorption capacity, 377 MBI, 362,369 mean adsorption enthalpy, 406 mechanical properties, 294, 318, 319, 338, 447, 531 mechanisms of adsorption, 241, 278 mercaptans, 233, 235, 249, 271, 272, 399 mercury porosimetry, 17, 50,68,99 mercury vapor adsorption, 455 mesophase pitch, 39,40, 89, 90, 98,112 mesopores, 6,22, 30, 35, 41 50, 51, 52, 68, 69, 77, 82, 86, 88, 89, 90, 91, 93S 94, 116,121, 132-136,144, 146,147,151, 285, 309-312, 335, 338, 352, 542 mesoporous activated carbon cloth, 378 mesoporous carbon aerogels, 93 metal hydrides, 325, 326 metal impregnated activated carbon, 402, 496 metal ion adsorption, 540 metal oxides, 297, 315, 318, 320 methyl mercaptan, 214, 216, 235,272,277,281 microorganisms, 426,432,437, 438, 440, 461 micropollutants, 428,432 micropores, 7, 8, 10,19, 20, 22, 23, 25, 26, 27, 30, 34, 35, 41, 52, 61, 93,238-242, 250-258, 266, 268, 278, 285, 349, 352, 362, 366, 368 micropore filling, 55,133, 153, 238, 509 micropore size, 309, 310 micropore volume, 11,121,122-124,128, 129, 132, 133,140, 147-149, 271, 349, 389, 392, 490-496, 505, 509, 511, 513, 519, 522-524 microporous carbon membranes, 83 microreactors, 276, 277 migration of ions, 311 military carbon, 494

Index

military filters, 481, 482,484,497, 506510 mineral impurities, 531 mineral matter, 5, 17, 22, 33, 213 mobility of water molecules, 308 molecular sieving carbons, 79,, 107,108, 110, 142 molecular sieving effect, 304, 305 molecule-solid interaction energy, 400 molybdenum, 496, 501 monolayer capacity, 116, 126,132,133 monoliths, 41,42 moving bed filters, 448 multicomponent adsorption, 271, 379, 525 multilayer adsorption, 116,144,147 multiple columns, 427 municipal waste water, 431 MWNTs, 313, 314, 319, 320, 321, 322 N NaOH, 249,250, 256, 281, 466 nanoporous carbons, 107-112,123, 125, 135,144, 152,153 nanotubes. 294, 309-322, 326, 334-338 NDFT, 144,149 nitro groups, 193 nitrogen dioxide, 231 nitrogen containing carbons, 163,174, 182 nitrogen containing precursor, 163, 316 nitrogen incorporation, 163,182 nitrogenated functionalities, 10, 163,182, 197,210,211,256 NMR.200 NO2,231 NOx, 217, 450, 451,458,461 NOM, 432, 433 non-specific dispersion forces, 355 non-graphitic carbons, 2 normalized capacity, 252-254, 274 O oligomerization, 356, 357 oncotherapy, 5545 optical microscopy, 51, 99 oral adsorbents, 530, 532, 533, 539 ordered mesoporous carbon, 84, 85, 86 organic micropollutants, 391

569

Index

organic impregnations, 497 organometallic complexes, 282 organo-sulfur compounds, 165 oxidation, 161, 162, 166,169, 171-173, 177,178,181,184,187,188,191,193, 194,196,197, 199, 200, 202, 205, 206, 211,212, 214, 216,217. 279,285, 431, 438,441, 451, 452,454, 458,460, 494, 496, 501, 502, 504, 506, 507, 541 oxidation yield, 69, 70 oxidizing agents, 162, 211, 345, 347, 348, 358, 368, 432, 454 oxygen-containing surface groups, 9,161, 163,172, 306, 316, 509 oxygen content, 178, 180,187, 188,197, 202, 206 oxygen-sulfur interactions, 285

p-nitriphenol, 392 PAC, 397,422, 424,425,428, 430,435, 436,439, 443-445,452,461, 465 partial pressures, 21,22,24, 27 particle size, 428,432, 434, 447,448, 455 particle size distribution, 428 pH effect, 438 pH of carbon, 186, 255,280 pharmaceutical production, 396,415,421, 445,449, 454, 460 phosgene, 478, 499 phosphoric acid, 15, 27, 28, 29, 31, 32, 39, 166 photocatalysis, 445 physical activation, 4,14, 531 physical adsorption, 113,135,233, 240243,251-254, 275, 331, 332, 334, 337, 490, 491, 492, 494, 495, 497, 498, 500, 502, 504-508, 5109, 513, 515, 518, 519, 520, 522, 524-526 physical and chemical properties, 5 plasma treatment, 316 poisoning treatment, 532 polar interactions, 238 pokrography, 177,178 polychlorinated biphenyls, 432,442 polydisperse polyelectrolytes, 348 polysulfides, 260

polyvalent cation interactions, 348 pore blocking, 456 pore connectivity, 116,117 pore size distribution, 8, 52, 55, 56, 57, 59, 77, 107, 109, 110,116, 149-151, 304, 308-310,335,349,362 pore structure, 3, 5, 6, 20, 43,45, 50, 55, 63, 64, 65, 72, 73, 77, 82, 90, 93, 95, 99,166, 173 pore volume, 110-112,118,128,130,146148, 150-153 pore wall thickness, 305, 306 porosity of carbon bed, 267 porous carbons, 77, 80 potentiometric titration, 183-186, 179, 185, 213 powdered activated carbons, 35 prehumidified carbons, 252 preoxidation, 12,17,40 pre-ozonation, 441 pressure drop, 4, 35, 37, 41 pressure swing adsorption, 426,450,462 production of activated carbon, 13,37 products of surface reactions, 237 PSD, 52, 55, 59, 60, 76, 77, 88, 95, 98, 99 pseudocapacitance, 171,172,174,297, 306,315-318,320,322 PTSA-systems, 484 purification of indoor air, 453 pyrolysis, 316, 338,464 pyrolysis conditions, 17 PZC, 181,186,188 R radicals, 234-236, 251, 252, 256, 257, 260, 274,277,281 radioactive gases, 479, 507 radioactive isotopes, 460 radionuclides, 539, 540, 541 raw materials, 12,13,14, 22, 29, 33 reaction pathways, 493 redox activity, 175,176 redox pseudocapacity reactions, 314 reference adsorbent, 111, 121-123,128, 136 reference solid, 121-123,133 reference vapor, 136

570

Index

regeneration, 4, 44, 45, 46, 376, 397, 398, 399, 400, 408,410, 412, 413,415, 416, 421, 422, 427,438, 439, 442,443, 448, 449,451,463,464-466 relative humidity, 506 remediation technologies, 442 residence times, 27 resonance structures, 174,175 retention time, 503 role of water, 251,253,275 S sand filter, 436 Sarin, 477, 478,499, 504 saturated vapor pressure, 490 saturation adsorption capcity, 382,392 SAXS, 63 scanning electron microscopy, 63 scanning tunneling microscopy, 50 secondary biological treatment, 439, 440 sedimentation, 428, 435 self-ignition temperature, 497 SO2,163,165, 208,212-214, 216, 450, 451, 455, 457, 458, 561, 493, 506, 507, 510,511 SO2 emissions, 232 SEM, 51, 63, 64, 69, 70, 72, 99, 281 selective oxidation, 233 selectivities of adsorbent, 286 self-ignition, 250, 251 sewage sludge derived materials, 241,245, 257, 258 side-effects, 548 silica, 76, 82, 84, 85, 86, 88, 89, 90 similarity coefficient, 135 slit-like pores, 110,117 SNU-1, 84, 85 SOC, 345, 348, 349, 351, 355-362, 365368 sol-gel condensation, 95 solvation energy, 387 sorption capacity, 244 spz-based structure, 4,2 specific interactions, 180, 204-207, 209, 216 specific surface area, 297, 302-314,317, 322, 334, 338 spectroscopic methods, 189

spinel-like structures, 257 statistical firm thickness, 145,146 steam, 4,14,20, 21,22, 23,24, 25,26, 27, 31,35,37,39,40,42,45 STM, 50, 51, 52, 53, 54, 88 structural heterogeneity, 110,135,142, 143,146,153,204 sulfur adsorbed, 237,250, 251, 259,268, 269 sulfur compounds, 232, 233,236, 282, 283,284 sulfur containing activated carbons, 438 sulfur containing functionalities, 212 sulfur dioxide, 231,-234, 237-240, 242244, 247, 272 sulfuric acid, 234, 237, 242-248, 251-256, 259,260,285 sulfurous acid, 234, 247, 260 supereapaeitors, 293-299,302, 314, 318320, 323, 324, 334, 335, 338, 339 supercritical conditions, 45 supermicropores, 50, 52, 69,109, 138 superoxide ions, 256, 275 surface acidity, 244, 253 surface area, 54, 57,59-62,69,76-86, 9095,107-136 surface binding, 541 surface charge, 347,349, 354,359 surface chemistry, 8, 9-11, 159-161,182, 183, 191-193, 202-217, 233, 237, 241, 243, 251, 253, 255, 257, 260, 272, 273, 277,285,287, 345, 349, 355, 356, 361, 366,391 surface complexes, 161,165,186,192, 197,206,212 surface functionalities, 161-166,171-189, 191-195,208-213,294, 297, 306, 326, 334,338, 509 surface hydroxyls, 177 surface oxidation, 177 surface oxygen functional groups, 91,93. 177,192,193 surface pH, 253, 254,260, 275,278,280 surface polarity, 351, 358,358,359 surface reactions, 203,211,216,250,251, 254,259,279,281,282 surface reactivity, 204 survival rate, 539, 554

Index

system heterogeneity, 345

t-curve, 1440146 TEDA, 494, 496, 497, 499, 502, 504, 505, 507-510,525 TEM, 51, 55, 56, 82, 83, 86, 87, 88, 93 template derived carbons, 43, 82 templating procedure, 310 tertiary treatment systems, 439 thermal activation, 14,20,25,27,28,29, 31,33,35,36,37,39 thermal analysis, 179,212 thermal regeneration, 4, 45 thermal stability, 192 thermodynamic equilibrium, 149 thiolate ions, 279 thiophenic compounds, 235,236,237,283, 286 total organic carbons (TOC), 347 toxic agents, 477,478 toxic compounds, 475, 476, 484, 486, 489, 531 TPD, 172, 179,209 transition metals, 171, 203, 216, 218 treatment temperature, 11 tricholoroethylene, 349, 356 triethylenediamine, 497, 501, 504, 507 trihalomethanes, 432 U ultrafiltration, 352,367, 397, 544 ultramicropores, 50, 52, 57, 59,60, 63,69, 109, 310, 328, 335,336, 339 unimpregnated carbons, 251

VOCs adsorption, 231, 249, 268, 271, 412, 415, 432, 442, 443,449, 453, 454, 461, 510,511 volumetric adsorption analyzers, 113 W

washing carbon,, 29 waste water, 249

571

waste water treatment plants, 249 water adsorption, 208, 252, 267, 278, 358, 366, 508, 509, 523, 525 water clusters, 278,279, 358 water film, 254, 256, 260, 274 water layers, 296 water purification, 108 water remediations, 442,443,444 water treatment process, 355,356,357, 402 water washing, 243,252,260 water vapor contents, 264 water vapor exchange capacity, 486 wettability, 294,306, 307,308 Wheeler-Jonas equation, 517-521, 523 Whetlerites, 486,494

X-ray difiractograms, 89 X-ray fluorescence, 202 X-ray photoelectron spectroscopy, 195 X-ray spectroscopy, 179

zeolite, 50, 76, 82, 83, 110,112,126, 547 zeta potential, 359 zinc chloride, 27,28 zinc impregnated carbons, 265 zinc oxides, 257

This Page is Intentionally Left Blank