Springer

Published in cooperation with NATO Public Diplomacy Division

Proceedings of the NATO Advanced Research Workshop on

Recent Advances in Adsorption Processes for Environmental Protection and Security

Kyiv, Ukraine

9-12 September 2006

AC.I.P. Catalogue record for this book is available from the Library of Congress.

ISBN 978-1-4020-6804-1 (PB) ISBN 978-1-4020-6803-4 (HB) ISBN 978-1-4020-6805- 8 (e-book)

Published by Springer,

P.O. Box 17, 3300 AA Dordrecht, The Netherlands. www.springer.com

Printed on acid-free paper

All Rights Reserved

© 2008 Springer Science+Business Media B.V.

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, microfilming, recording or otherwise, without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work.

ORGANIZATION

Workshop Committee

José Paulo Mota (Universidade Nova de Lisboa, Portugal)

Svetlana Lyubchik (Institute of Physical Organic and Coal Chemistry, Ukraine)

Sandor Barany (University of Miskolc, Hungary) Francois Beguin (CNRS-University, France)

Eiliv Steinnes (Norwegian University of Science and Technology, Norway) Isabel Fonseca (Universidade Nova de Lisboa, Portugal)

Advisory Committee

Sandor Barany (University of Miskolc, Hungary) Francois Beguin (CNRS-University, France)

Tatiana Burdejnaya (Institute of Physical Organic and Coal Chemistry, Ukraine)

Isabel Fonseca (Universidade Nova de Lisboa, Portugal)

Peter Lodewyckx (Royal Military Academy, Belgium)

Francisco Rodriguez-Reinoso (Universidad de Alicante, Spain)

Eiliv Steinnes (Norwegian University of Science and Technology, Norway)

Fritz Stoeckli (Université de Neuchatel, Switzerland)

Local Organizing Committee

Joao Fraga Anna Kladova Olena Lygina Natalie Lygina Andriy Lyubchik Sergiy Lyubchik Yury Neylo Patricia Oliveira Tatiana Shendrik Liliya Tikhonova

PREFACE

The NATO Advanced Research Workshop on Recent Advances in Adsorption Processes for Environmental Protection and Security was held during, September 9-12, 2006, at Hotel Salyut, Kiev, Ukraine. There were 45 participants from 12 NATO countries, 8 Eligible Partner and Mediterranean dialog home and one key speaker from Switzerland. We hope that everyone went home with excellent memories and new ideas to enhance environmental protection and security through adsorption science and technology.

The purpose of the Workshop was to bring together researchers and engineers working in adsorption-related fields, to share knowledge on the latest advances on adsorption processes for environmental security and protection, as well as to cross-link and disseminate to the scientific community the main results and achievements of recent NATO SfP projects on environmental security and protection.

Topics covered by the Workshop include recent theoretical and experimental developments on environmental adsorption, adsorption processes, as well as synthesis and tailoring of novel adsorbents, including the assessment of materials and processes.

The invited lectures provided a comprehensive report on adsorption and colloids, carbon materials and adsorbents for various industrial applications, and ecological safety and antiterrorism. Because rapidly developing areas in nanotechnology for fine chemistry, air quality, and environmental protection, are based on the synthesis and modification of the adsorbents, special attention was given to synthesis and chemical tailoring of porous materials to achieve desired properties as adsorbents and separation media.

We hope that the Workshop helped to intensify the cooperation between scientists from NATO countries, USA, Central Europe and FSU countries, and to shorten the gap between Partner/Mediterranean Dialogue countries and NATO countries with respect to environmental protection and security standards. Finally, we also hope that the Workshop opened a forum for discussion on possibilities for further improvements and areas still lacking critical knowledge and expertise.

We would like to express our gratitude to the members of the local organizing committee for all their hard work to make this Workshop a success. The Workshop would not have been possible without the generous financial support from the NATO Science for Peace and Security Programme.

Jose Paulo Mota Svetlana Lyubchik

CONTENTS

Organization v

Preface vii

List of participants xi

Extension of Dubinin's Theory to Adsorption From Aqueous

Solutions 1

F. Stoeckli, D.M. Nevskaia, E. Castillejos-Lopez, T.A. Centeno

Applications of Immersion Calorimetry in Dubinin's Theory and in Electrochemistry 9

T.A. Centeno, F. Stoeckli

Adsorption on Activated Carbon: One Underlying Mechanism? 19

P. Lodewyckx

Adsorption of Organic Vapour Pollutants on Activated Carbon 29

A.J. Fletcher, M.J. Kennedy, X.B. Zhao, J.B. Bell, K.M. Thomas

Adsorption Behaviour of Lignosulphonates on the Interfaces Water-Inorganic/Organic Solids, Used for Paper Production 55

G. Telysheva, T. Dizhbite, A. Andersone, A. Volperts

Adsorption Properties of Polymer Adsorbents 65

J. Hradil

SAXS Characterization of Solid/Vapor Interfaces in Polymer-Based Microporous Carbons With Different Surface Chemistry 75

CONTENTS

Co-Adsorption of Cesium Chloride Molecules and Thulium Atoms on 2DGF on Re

N.M. Nasrullayev

Controlling Porosity to Improve Activated Carbon Applications

A. Linares-Solano, D. Lozano-Castelló, M.A. Lillo-Ródenas, D. Cazorla-Amorós

Liquid-Phase Adsorption/Oxidation of Sulfur-Containing Species by Activated Carbon 107

R.V.R.A. Rios, J. Silvestre-Albero, A. Sepulveda-Escribano, F. Rodriguez-Reinoso

Adsorption Properties of Functional Silicas Towards Some

Toxic Metal Ions in Water Solutions 119

V. Tertykh, L. Polishchuk, V. Yanishpolskii, E. Yanovska, A. Dadashev, V. Karmanov, O. Kichkiruk

Waste Conversion into Activated Carbon for Heavy Metal Removal From Waste Water 133

S. Lyubchik, M. Khodorkovskij, T. Makarova, L. Tikhonova, J.P.B Mota, I. Fonseca

Modeling Carbon Mask Adsorptive Filters 147

J.M. Loureiro, A.M. Ribeiro

Adsorption Processes in Gas Mask Filter Canisters: Practical Aspects, New Materials and Modeling 155

M.J.G. Linders

Hydrogen Adsorption on Carbon Materials at High Pressures and Different Temperatures 165

F. Suárez-García, M. Jordá-Beneyto, D. Lozano-Castelló, D. Cazorla-Amorós, A. Linares-Solano

Adsorbed Natural Gas Technology J.P.B. Mota

LIST OF PARTICIPANTS

Andriy Lyubchik Donetsk National University, Department of Chemistry, Donetsk, Ukraine

Angel Linares-Solano Department of Inorganic Chemistry, University of Alicante, Alicante, Spain

Anna Kladova Chemistry Department, Faculty of Sciences and Technology, Universidade Nova de Lisboa, Caparica, Portugal

Boris Voevoda Donetsk National Technical University, Department of Geology, Donetsk, Ukraine

Christina Hakopian Yerevan State University, International Scientific Research Center of Water, Climatic and Recreational Resources, Yerevan, Armenia

Dilek Gulbayir Yildiz Technical University, Chemical Engineering Department, Davutpasa Campus, Esenler, Istanbul, Turkey

Dmitriy Shvets Institute for Sorption and Endoecology, NAS of Ukraine, Kiev, Ukraine

Eiliv Steinnes Department of Chemistry, Norwegian University of Science and Technology, Trondheim, Norway

Elena Lygina Donetsk State University of Economics and Trade, Department of Chemistry, Donetsk, Ukraine

Encarnación Raymundo-Pinero Research Center on Divided Matter, Orléans University, Orlans, France

Eugene Katz Ben-Gurion University of the Negev, Department of Solar Energy and Environmental Physics, Israel

Fabian Suarez Garcia Department of Inorganic Chemistry, University of Alicante, Alicante, Spain

Francisco Rodriguez-Reinoso Department of Inorganic Chemistry, University of Alicante, Alicante, Spain

Francois Beguin Research Center on Divided Matter, Orlans University, Orlans, France

Fritz Stoeckli IMT-Chimie des Surfaces, University of Neuchatel, Neuchatel, Switzerland

Georgios Gallios Aristotle University of Thessaloniki, Chemical Technology Division, School of Chemistry, Thessaloniki, Greece

Ilknur Kucuk Yildiz Technical University, Chemical Engineering Department, Davutpasa Campus, Esenler, Istanbul, Turkey

Janos Lakatos University of Miskolc, Department of Chemistry, Miskolc-Egyetemvaros, Hungary

Jelena Chirkova Latvian State Institute of Wood Chemistry, Riga, Latvia

Jose Paulo Mota Chemistry Department, Faculty of Sciences and Technology, Universidade Nova de Lisboa, Caparica, Portugal

Juriy Hradil Institute of Macromolecular Chemistry AS CR, Prague, Czech Republic

K. Mark Thomas University of Newcastle upon Tyne, School of Natural Sciences, Newcastle upon Tyne, UK

Kristina Laszlo Department of Physical Chemistry, Budapest University of Technology and Economics, Budapest, Hungary

Marco Linders TNO Defense, Security and Safety, Rijswijk, The Netherlands

Marina Sosina Institute for Sorption and Endoecology, NAS of Ukraine, Kiev, Ukraine

Massoud Rostam-Abadi Illinois State Geological Survey, University of Illinois, Champaign, Illinois, USA

Miguel Loureiro LSRE/DEQ, Faculty of Engineering, University of Porto, Porto, Portugal

Miroslava Vaclavikova Institude of Geotechnics of the Slovak Academy of Sciences, Kosice, Slovac Republic

Nataliya Mishchuk Institute of Colloid and Water Chemistry, Vernadskogo, Kiev, Ukraine

Nazim Narsullayev Baku State University, Department of Physics, Baku, Azerbaijan

Nik Kanellopoulos NCSR "Demokritos", Institute of Physical Chemistry, Agia Paraskevi Attikis, Greece

Nikolai Kartel Institute of Sorption and Problems of Endoecology, NAS of Ukraine, Kiev, Ukraine

Nina Tchanishvili The Eliava Institute of Bacteriophage, Microbiology and Virology, Tbilisi, Georgia

Patricia Oliveira Chemistry Department, Faculty of Sciences and Technology, Universidade Nova de Lisboa, Caparica, Portugal

Peter Lodewyckx Royal Military Academy, Belgium

Sandor Barany University of Miskolc, Department of Chemistry, Miskolc-Egyetemvaros, Hungary

Sergei Lyubchik Donetsk National University, Department of Chemistry, Donetsk, Ukraine

Sergey Mikhalovsky University of Brighton, Brighton, UK

Svetlana Lyubchik Institute of Physical Organic and Coal Chemistry, National Academy of Science of Ukraine, Donetsk, Ukraine

Tata Burburatshvili The Eliava Institute of Bacteriophage, Microbiology and Virology, Tbilisi, Georgia

Tatiana Shendrik Institute of Physical-Organic and Coal Chemistry, Department of Coal Chemistry, Donetsk, Ukraine

Tatjana Dizbite Latvian State Institute of Wood Chemistry, Riga, Latvia

Teresa Centeno Instituto Nacional del Carbon-CSIC, Oviedo, Spain

Valentin Tertykh Institute of Surface Chemistry of Academy of Sciences of Ukraine, Kiev, Ukraine

Valentin Tret'yakov Lomonosov Moscov State Academy of the Fine Chemical Technology, Moscow, Russia

EXTENSION OF DUBININ'S THEORY TO ADSORPTION FROM FROM AQUEOUS SOLUTIONS

FRITZ STOECKLI*

IMT-Chimie des Surfaces, Université de Neuchâtel. Rue Emile Argand 11, CH-2009 Neuchâtel, Switzerland

DASHA M. NEVSKAIA, EVA CASTILLEJOS-LOPEZ

Departamento de Quimica Inorganica, Facultad de Ciencias UNED, Paseo Senda del Rey 9, 28040 Madrid, Spain

TERESA A. CENTENO

Instituto Nacional del Carbón-CSIC. Apartado 73, 33080 Oviedo, Spain

Abstract. Adsorption of sparingly soluble organics from aqueous solutions, by activated carbons, can be described within the framework of Dubinin's theory by using a modified Dubinin-Radushkevich-Kaganer (DRK) equ-equation, where relative pressures are replaced by relative concentrations. With respect to the descriptions based on the Langmuir model and similar expressions, this approach has the advantage that it allows predictions on the basis of simple physico-chemical properties of the solid and of the adsorbate. Preliminary experiments indicate that in the case of dilute binary mixtures, the model of independent coadsorption, based on the DRK equation, applies. However, more experimental evidence is needed to confirm this potentially very useful approach in filtration technology.

Keywords: Dubinin's theory; adsorption; aqueous solutions; binary mixtures

1. Introduction

The removal of sparingly soluble organics from aqueous solutions is a relatively important topic, in particular for the purification of drinking

* To whom correspondence should be addressed. Institut de Microtechnique-LCPS, rue Emile Argand 11, CH-2009 Neuchatel, Switzerland; E-mail: [email protected]

J.P. Mota and S. Lyubchik (eds.), Recent Advances in Adsorption Processes for Environmental Protection and Security, 1-8. © 2008 Springer.

water by activated carbon. So far, a variety of theoretical models have been proposed, which can reproduce accurately the experimental data.1'2

However, a major shortcoming in the use of Langmuir- or Freundlich-based approaches is the fact that predictions are difficult, and even impossible in many cases, due to the complexity of the parameters used in these expressions. It is therefore necessary to find alternative descriptions, which provide a tool for quantitative or at least semi-quantitative predictions in filtration technology.

Active carbons are very efficient adsorbents and the adsorption of vapours is well described by Dubinin's theory,3'4 which uses relatively simple parameters and allows predictions over relatively wide pressure and temperature ranges. As shown in detail elsewhere,5-7 for relative concentrations ceq/cs < 0.01-0.05 (submonolayer conditions) adsorption of sparingly soluble species such as phenol and its derivatives can be described by the modified Dubinin-Radushkevich-Kaganer (DRK) equation.

where

In the case of sparingly soluble species, Nam corresponds to the monolayer capacity of the walls of the micropores and not to the filling of these pores, as opposed to adsorption from the vapour phase. This means that the sorptive capacity of a microporous carbon is often smaller in the case of adsorption from dilute aqueous solutions, than from the vapour phase. This point is clearly illustrated by the adsorption of phenol from both phases.8

It appears that the principle of temperature invariance observed by Dubinin for micropore filling also applies here, as parameters n and Es are practically constant over a range of 20°-30°.

One can also introduce specific scaling factors (surface affinity coefficients ps) for Es, to correlate the adsorption of different species relative to a reference solute (benzene or phenol). These properties are reflected in the logarithmic plot (Figure 1) for the adsorption of various phenolic compounds from aqueous solutions, at different temperatures and on three different activated carbons, with n = 4. As shown in the case of the DRK equation9 (adsorption of vapours by non-porous solids), exponent n is related to the heterogeneity of the surface and it appears that for certain systems n = 3 (or even 2) also provides a good fit.

As observed for adsorption from the vapour phase, the characteristic adsorption energy Es, depends on both the adsorbent and the adsorbate. It appears that in the case of aqueous solutions of phenol,5-7 Es is practically equal to the characteristic energy of adsorption of benzene vapours Eo, which is a reference in adsorption by microporous carbons.3 For other species (i) removed from aqueous solutions, one finds specific scaling factors ps relative to phenol and defined by

Typical values are shown in Table 1, with the values observed for adsorption from the vapour phase or calculated as described elsewhere.

300000

Figure 1. Logarithmic plot of Eq. (2) for the adsorption of various organics from aqueous solutions onto activated carbons: ◊ carbon CP (phenol, 4-cresol, 3-chlorophenol, 3-aminophenol, 4-nitrophenol); ▲ carbon PC (phenol, ra-chlorophenol); A carbon AC (phenol, aniline).

TABLE 1. Typical affinity coefficients ps (average values) for the adsorption of sparingly soluble organics from aqueous carbons onto activated carbons. For comparison purposes, the affinity coefficients for adsorption from the vapour phase9,10 are also given (in the case of compounds with low vapour pressures, p is estimated10 from parachors or molar volumes).

TABLE 1. Typical affinity coefficients ps (average values) for the adsorption of sparingly soluble organics from aqueous carbons onto activated carbons. For comparison purposes, the affinity coefficients for adsorption from the vapour phase9,10 are also given (in the case of compounds with low vapour pressures, p is estimated10 from parachors or molar volumes).

Compound

ßs = Es/Eo

P (vapour phase)

ßs/ß

Phenol

1.03

1.09

0.95

3-chlorophenol

1.02

1.24

0.82

4-nitrophenol

0.90

1.20

0.75

Aniline

0.80

1.05

0.74

Benzene

0.54

1.00

0.54

Benzoic acid

0.80

1.04

0.77

Caffeine

1.28

1.85

0.69

These results show that predictions can be made for the removal of sparingly soluble organics by activated carbons at ceq/cs < 0.05, on the basis of relatively simple physico-chemical and structural properties. The latter are the characteristic energy Eo of the carbon, which varies typically between 30 kJ/mol (micropore width around 0.5 nm) and 16-17 kJ/mol

(supermicropores of ~2 nm) and the surface area of the micropore walls Smi. This area can be assessed by different techniques, but in view of the good correlation observed with the adsorption of phenol from aqueous solutions based either on the solution isotherm or immersion calorimetry, the latter technique is now used to determine surface areas (see the contribution of Centeno and Stoeckli in this volume). For example, on the basis of a molecular surface area of 45 x 10-20 m2 (molecule lying flat on the carbon surface), the specific enthalpy hi(phenol) = -(0.109 ± 0.03) J/m2. Moreover, and in spite of the fact that phenol is excluded from the oxygen-containing surface sites, it appears that, due to a compensation effect, the enthalpy of immersion AiH(J/g) of a carbon into an aqueous solution of phenol (usually 0.4 M) leads to a good approximation of the total surface area by6

This area corresponds to Smi and the external surface Se area found in larger pores and on the outside. Experiments with graphitized carbon blacks suggest a similar relation for caffeine solutions with -(0.112 ± 0.015 J/m2), but in view of the structure of this molecule, the technique is limited to micropores larger than 0.55-0.6 nm.

In the low concentration range, Eq. (1) corresponds to adsorption on the micropore walls, where the adsorption energy is higher than on the external surface area. Typically, Eo is around 11 kJ/mol for non-porous carbons, and the adsorption of phenol from an aqueous solution onto graphitized carbon black N234-G leads to Es(phenol) = 13 kJ/mol. This experiment confirms that adsorption on the open surface, also limited to a monolayer, becomes effective only at relatively high relative concentrations. In Figure 1, this corresponds to an additional section, not shown, at the upper end of the graph (ceq/cs > 0.05-0.1). On the other hand, in calorimetric experiments, the concentration is high enough (0.4 M) to form a monolayer on Stot.

2. Binary Adsorption from Aqueous Solutions

Following the success obtained with the adsorption of single species, it is tempting to extend the Dubinin-based approach to binary and, hopefully, to multiple adsorption. The advantage is obvious, as it should allow predictions for the simultaneous removal of organics from water on the basis of simple parameters and over a range of temperatures. Obviously, within the framework of the modified DRK theory, the situation should correspond to the submonolayer region and describe the final stages of purification of water.

A convenient working hypothesis is the model of independent coad-sorption, which assumes that each species is adsorbed according to its DRK

equation (1), but only on the surface area left free by the other component. In the case of binary adsorption, this corresponds to the set of coupled equations

NaA = [NamA - Nb (NamA /NamB)] eXp{-[RTln(cs / Ce^A /Esa]"} (5) NaB = [NamB -NaA(NamB /NamA)] eXp{-[RTln(cs / Ce^B / EsbT} (6)

The subsidiary condition, leading to the pre-exponential terms, is

where AmA and AmB are the molecular surface areas of species A and B and NAv is Avogadro's number.

In the case of a single species, exponent n = 4 often provides a good fit.5-7 However, it is not a prerequisite, as n probably reflects the heterogeneity of the surface. As discussed elsewhere,9 this is clearly the case with the original DRK equation, where the fixed value n = 2 is responsible for the discrepancy often observed between the monolayer capacities Nam(DRK) and Nam(BET). Exponent n may therefore vary from carbon to carbon and it is possible to obtain a better overall fit with n = 3 or even 2.

The set of Eqs. (5) and (6) is potentially interesting, provided that two requirements are fulfilled:

1. The principle of temperature invariance of Es and n applies, which allows predictions over a range of concentrations and temperatures

2. The parameters EsA and EsB should be the same as those obtained for the adsorption of the individual species.

At this stage, the analysis of the phenol + aniline mixture in water, adsorbed by activated carbon AC at 298 K1 and 313 K (new data) provides a first and interesting illustration of the potentiality of the present approach.

The analysis of the nitrogen adsorption isotherm at 77 K, using the Dubinin-Radushkevich equation, leads to a micropore volume Wo = 0.418 cm3/g, and a characteristic energy Eo = 23.0 kJ/mol. With the help of the correlations3,4

one obtains respectively Lo = 0.93 nm and Smi = 898 m2/g.

The adsorption isotherms of the single species at 298 and 313 K provide the best fits for Eq. (1) with n = 3, rather than n = 4, where a slight upward curvature is observed in the logarithmic plots. This leads to the values of Nam and Es given in Table 2. They correspond to the domain of relative concentrations ceq/cs between 5 x 10-5 and 0.01.

TABLE 2. Parameters of Eq. (1) with n = 3 for the single adsorption from unbuffered aqueous solutions by carbon AC (Eo = 23.0 kJ/mol and Sml = 898 m2/g) at 298 and 313 K.

Adsorbate (single)_Phenol_Aniline

This data is coherent and suggests average surface affinity coefficients Ps(phenol) = 1.01 ± 0.04 and ps(aniline) = 0.78 ± 0.04, in agreement with other determinations.7 Moreover, the molecular surface area of 45 x 10-20 m2 obtained for phenol adsorbed on graphitized carbon blacks, leads to a micropore surface area of 848 m2/g, which is in reasonable agreement with the value derived from Eq. (9). The difference may be ascribed to the presence of surface oxygen, on which water is preferentially adsorbed in unbuffered solutions. The data also suggests that the molecular surface area of aniline is close to the value observed for phenol, as expected.

The analysis of the data for binary adsorption based on Eqs. (5) and (6) using again n = 3 and the values for Nam(phenol) and Nam(aniline) of Table 2, leads to the best fit values of Es given in Table 3.

TABLE 3. Values of Es(phenol) and Es(aniline) for binary adsorption at 298 and 313 K.

Adsorbate (in the mixture)

Phenol

Aniline

Es(298 K) (kJ/mol)

22.5 ± 0.5

18.1 ± 0.5

Es(313 K) (kJ/mol)

22.3 ± 0.5

18.1 ± 0.5

These values are in good agreement with those obtained for the single adsorption (Table 2) and suggest that for the present binary system the adsorption equilibrium between 298 and 313 K could have been predicted with a reasonable accuracy by using the structural parameters of the carbon (Eo = 23.0 kJ/mol and Smi = 898 m2/g), with the affinity coefficients ps of Table 1 and the saturation concentrations of phenol and aniline found in standard tables. The correlation between the experimental and amounts adsorbed in the submonolayer region and the values calculated using the best fits for Es and Nam is shown in Figure 2. The data corresponds to adsorption in the submonolayer region, with Na(phenol) + Na(aniline) below 2-2.5 mmol/g.

Obviously, the uncertainty on ps introduces a corresponding uncertainty in the amounts adsorbed. However, the approach outline here and based on the extended DRK equation provides a good estimate, as opposed to the Langmuir-based models, where the parameters cannot be determined "a priori" or extrapolated to other temperatures.

Figure 2. Correlation between the calculated and experimental amounts of phenol (■) and aniline (▲) adsorbed from binary mixtures in water, by active carbon AC, at 298 and 313 K.

3. Conclusions

The present study shows that Dubinin's theory can be extended to the adsorption, by activated carbons, of sparingly soluble organics from aqueous solutions. The basic equation is a modified Dubinin-Raduskkevich-Kaganer equation, where relative pressures are replaced by relative concentrations and the exponent n around 3-4. The principle of temperature invariance of parameters Es and n has been established for single adsorption of a variety of organics and predictions can be made for a range of concentrations and temperatures.

Preliminary studies on binary systems in water, for example, phenol + aniline adsorbed by a typical industrial active carbon, suggest a similar pattern in the submonolayer region. In this case, the model of independent coadsorption can be applied and, with a good approximation, the individual isotherms are those obtained for single adsorption. This extension must be confirmed by a systematic study of more systems and special attention should be given to the value of exponent n, probably on the basis of a refined analysis of the carbon itself.

The approach outlined here is of great relevance to filtration technology, where semi-quantitative and quantitative predictions can be made on the basis of relatively simple physico-chemical properties of the adsorbates and the structural parameters of the carbon.

References

1. D.M. Nevskaia, E. Castillejos-López, A. Guerrero-Ruiz, and V. Muñoz, Effect of the Surface chemistry of carbon materials on the adsorption of phenol-aniline mixtures from water, Carbon 42(3), 653-665 (2004).

2. R. Leyva-Ramos, J. Zoto-Zuñiga, J. Mendoza Barron, and R.M. Guerro Coronado, Adsorption of phenol from aqueous solutions onto activated carbon. Effect of solvent, temperature and particle size, Adsorpt. Sci. Technol. 17(7), 533-543 (1999).

3. F. Stoeckli, in: Porosity in Carbons, edited by J. Patrick (E. Arnold, London, 1995), pp. 67-97.

4. F. Stoeckli and T.A. Centeno, On the determination of surface areas in activated carbons, Carbon 43(6), 1184-1190 (2005).

5. F. Stoeckli, M.V. López-Ramón, and C. Moreno-Castilla, Adsorption of phenolic compounds from aqueous solutions, by activated carbons, described by the Dubinin-Astakhov equation, Langmuir 17(11), 3301-3306 (2000).

6. E. Fernández, D. Hugi-Cleary, M.V. López-Ramón, and F. Stoeckli, Adsorption of phenol from dilute and concentrated aqueous solutions by activated carbons, Langmuir 19(23), 9719-9723 (2003).

7. D. Hugi-Cleary, A. Slasli, and F. Stoeckli, Helv. Chim. Acta 88, 470-477 (2005).

8. F. Stoeckli and D. Hugi-Cleary, Russ. Chem. Int. Ed. 50(11), 2060-2063 (2001).

9. D. Hugi-Cleary, S. Wermeille, and F. Stoeckli, Chimia 57(10), 611-615 (2003).

10. G.O. Wood, Affinity coefficients of the Polanyi/Dubinin adsorption isotherm equations, Carbon 39(3), 343-356 (2001).

APPLICATIONS OF IMMERSION CALORIMETRY IN DUBININ'S THEORY AND IN ELECTROCHEMISTRY

TERESA A. CENTENO*

Instituto Nacional del Carbón-CSIC. Apartado 73, 33080 Oviedo, Spain

FRITZ STOECKLI

IMT-Chimie des Surfaces, Université de Neuchatel. Rue Emile Argand 11, CH-2009 Neuchatel, Switzerland

Abstract. This study shows that immersion calorimetry is a useful technique which simplifies considerably the analysis of porosity and chemical nature of activated carbons.

The characterization of activated carbons in the general theoretical framework of Dubinin's theory with its extensions to calorimetry and adsorption from solutions allows the identification of some key parameters for the performance of these materials in electrochemical capacitors.

Keywords: activated carbon; immersion calorimetry; porosity; electrochemical capacitor

1. Introduction

Activated carbons can be characterized within the framework of Dubinin's theory. The basic relation is the Dubinin-Radushkevitch (DR) equation1

v^Eoy

where W represents the volume adsorbed at temperature T and relative pressure p/ps, Wo is the limiting volume adsorbed in the micropores and

*To whom correspondence should be addressed. E-mail: [email protected]

J.P. Mota and S. Lyubchik (eds.), Recent Advances in Adsorption Processes for Environmental Protection and Security, 9-18. © 2008 Springer.

A = RTln(ps/p); p and Eo are specific parameters depending on the adsorptive and on the adsorbent. Wo and Eo are usually obtained from the adsorption of small molecular probe (typically CO2, CH2Cl2, and C6H6) with critical dimensions around 0.33-0.40 nm, and unhindered by constrictions in the structure.

In view of its thermodynamic basis, the DR equation can easily be used in the context of immersion calorimetry.12 Starting from the definition of the isosteric heat of adsorption of a vapour qst (>0)

The inversed DR equation leads to qst as a decreasing function of the micropore filling qst(T;0) = PE0{ [ln(1/0)f2 + (aT/2)[ ln(1/0)] -1/2 ]} + AHmp (3)

where a is the thermal expansion coefficient of the adsorbate. On the other hand, the net heat of adsorption defined as qnet (T; d)= qst - Hp (4)

is related to the enthalpy of immersion of the solid into the corresponding liquid by

Taking into account that there is practically no liquid-vapour interface in micropores, it follows that

Since active carbons often present an external (non-microporous) surface area, Se, the experimental enthalpy of immersion will include an extra contribution, corresponding to the wetting of this surface and

It appears, as a thermodynamic consequence of Dubinin's theory, that the enthalpy of immersion of a microporous carbon into organic liquids is related to the parameters of the adsorption isotherm (Wo, Eo and Se). This formal link between both approaches leads to a detailed picture of the porous structure and of the chemical nature of activated carbons.

The experimental set-up required for immersion calorimetry is relatively simple.1 The sample is outgassed at 10-5 Torr and around 523 K in a glass bulb with a brittle end. The bulb is sealed, introduced into a cell containing around 5 cm3 of the wetting liquid and placed inside the calorimeter. Once thermal equilibrium is achieved, the brittle end of the bulb is gently broken and the liquid wets the sample. The corresponding heat evolution through 180 thermocouples is monitored as a function of time and the integration of this signal leads, with the help of an electrical calibration, to the enthalpy of immersion AiH. A typical experiment lasts approximately 30-45 min, which allows an important gain in time, if compared with classical adsorption experiments.

2. Application of Immersion Calorimetry for Structural and Chemical Characterization of Activated Carbons

2.1. STRUCTURAL ASPECTS OF IMMERSION CALORIMETRY

2.1.1. Calorimetric studies of selective adsorption from aqueous solutions: determination of specific surface areas

This section shows how the enthalpy of immersion of carbons into aqueous solution of caffeine (0.1 M) or phenol (0.4 M) becomes another source of information for the determination of specific surface area of carbons.1'3-5

Some years ago, a calorimetric approach3 showed that the enthalpy of immersion of carbon blacks into aqueous solutions of caffeine is a linear function of the mass of carbon. The comparison of AiH (J g-1) with the SBET (m2 g-1) of the materials led to an average specific enthalpy hi of -(0.112 ± 0.015) J m-2. This correlation, confirmed by the corresponding adsorption isotherm from the solution, implies that adsorption is limited to a single layer. This is also true for microporous carbons, where no volume filling process takes place, as opposed to adsorption from the vapour phase. It follows that for both porous and non-porous carbons the total surface area can be determined with a good accuracy from the enthalpy of immersion into an aqueous solution of caffeine with the help of the simple relation13

Since the caffeine molecule cannot penetrate into micropores of less than approximately 0.6 nm, this procedure provides an estimate of the total surface area in carbons with pores wider than 0.6-0.7 nm.1'3'4

Subsequently, this technique has been used successfully with diluted aqueous solutions of phenol (0.4 M) and the study of carbon blacks and active carbons5'6 suggested a process similar to that observed for caffeine. Phenol appears to form a monolayer, which provides information on the surface area of pores above approximately 0.45 nm. In this case, the specific enthalpies hi corresponded to -(0.109 ± 0.003) J m-2 and

O ( 2 -1 \ A iH phenol (Jg 1 ) (9) S total (m 2 g 1 )=-p-2--(9)

It should be pointed out that the enthalpy of immersion into aqueous solution of phenol is affected by the surface acidity of the carbon since water is adsorbed preferentially by the oxygen-containing surface groups.56 The reduction in specific surface area has been estimated to be around 70 m2 per mmol of surface oxygen, but due to a compensation effect, Eq. (9) still provides a good assessment of Stotal in the case of oxidized carbons.

On the other hand, it appears that the enthalpies of immersion of the microporous carbons into concentrated solutions are generally larger than observed for the dilute solution. This increase confirms that the adsorption of phenol from concentrated aqueous solutions corresponds to a process of micropore filling and that it is not limited to the coating of the micropore walls, as observed for dilute solutions.6

2.1.2. The assessment of micropore size distribution of carbons

Equation (7) can be used to evaluate, on the basis of the experimental enthalpies of immersion, the volumes W(Lc) filled by a molecular probe of critical dimension Lc,1

Eo is the characteristic energy obtained from the reference isotherm of a small adsorbate (CH2Cl2, C6H6, CO2, etc.) and Lc is the smallest micropore width accessible to the molecules of the liquid.

The use of a series of liquids of known molecular dimensions (i.e. dichloromethane [0.33 nm], benzene [0.41 nm], cyclohexane [0.54 nm], carbon tetrachloride [0.63 nm], cyclododeca-1,5,9-triene [0.76 nm], tri-2,4-xylyl phosphate [1.5 nm]) leads to the micropore size distribution.17-9

However, this technique reflects the true pore size distribution only as long as the entry into wider pores is not limited by constrictions smaller than their actual size. If such "gate" effects are present, one obtains an apparent pore size distribution.8'9 The difference between the apparent and the real pore size distributions can been illustrated if one uses the approach based on the modelling of CO2 adsorption, a molecule which is small enough to bypass gate effects.9 The two techniques are therefore complementary and provide a refined picture of the porous structure.

2.2. CHEMICAL ASPECTS OF IMMERSION CALORIMETRY

2.2.1. Detection of surface oxygen from immersion into water

As opposed to organic liquids and vapours, where the volume filling of micropores and adsorption on the external surface area Se are the fundamental processes, water interacts strongly with functional groups of carbon surface. An interesting consequence is the possibility to estimate the number of surface groups from the experimental enthalpies of immersion of active carbons into water and into benzene at 293 K and taking into account the chemistry of the surface through an excess enthalpy of immersion, AiH (H2O)excess. As reported earlier,1011 water interacts in a similar fashion with surface functionalities containing mainly oxygen (-12.1 J mmol-1) and with basic groups (-10.3 J mmol-1). The latter are characterized by their equivalents of HCl and it appears that most of them do not contain oxygen. An overall assessment for a large variety of activated carbons led to10

AiH (HO) (Jg-1) = 0.210 AiH (CH) - 9.9 (Jmmol-1) [O + HCl] (11)

with average specific interactions around -(9.9 ± 0.7) J mmol-1 of oxygen or HCl mequivalents of basic sites. As the concentration of the basic sites of typical carbons is not high, the bulk of the specific interactions is related to oxygen atoms.

2.2.2. Determination of acidic and basic groups on carbon surface

The correlation between the net enthalpy of neutralization into 2N aqueous solution of NaOH

and the number of equivalents of acid obtained from direct titration leads to a net enthalpy of neutralization of -(41.1 ± 1.8) kJ eq-1 for acidic sites. A similar approach for the net enthalpy of neutralization into 2N HCl

leads to a neutralization energy of -(52.3 ± 2.0) kJ eq-1 for basic sites.12

Further calorimetric experiments with 1N aqueous solutions of NaHCO3 allow a clear distinction between carboxylic sites the other acidic sites. Selective neutralization of carboxylic groups reported a value of -(39.7 ± 1.7) kJ eq-1 which is close to the result obtained with NaOH for the bulk of the acidic groups.12

3. Application of Immersion Calorimetry to the Characterization of Carbons Used in Electric Double Layer Capacitors

Electrochemical capacitors have generated wide interest in recent years for high power applications where high efficiency and long cycle life are required. At the present, most commercial devices correspond to those referred to as electrical double layer capacitors (EDLCs) which perform by an electrostatic attraction between electric charges accumulated on the electrode surface and ions of opposite charge in the electrolyte side.13

Activated carbons are widely used as electrodes in EDLCs systems, as far as their high surface areas provide large interfaces for the formation of the electric double layer.14 The reliable characterization of activated carbons in the general theoretical framework of Dubinin's theory, with its extensions to calorimetry and adsorption from solutions, leads to the identification of key parameters for high performance in electrochemical capacitors.15-17 In this context, the determination of the real surface of the carbons is of prime importance and immersion calorimetry plays a major role, as illustrated below. Useful information can only be gained from several independent techniques, including the technique based on aqueous phenol solutions (see Section 2.1). As discussed in detail elsewhere, SBET, is often too large18 and leads to erroneous surface related properties.15'17 As a consequence, it is difficult to establish reliable correlations.

In the case of 2M H2SO4 aqueous electrolyte, the study of a large variety of activated carbons showed that the specific capacitance at low current density (1 mA cm-2), Co, depends on standard contributions from the total surface area and from the surface groups which desorb as CO in Thermally Programmed Desorption (TPD)15:

Co[H2SO4](Fg-1) = 0.081 (Fm-2) Stot + 63 (Fmmol-1)[CO](mmolg-1) (14)

This approach reflects the important role played by both the surface area Stot and its chemical nature in the overall capacitance of activated carbons. It also establishes the origin of a pseudocapacitive process (via redox reactions involving CO-desorbing groups) which should be added to the purely EDLC mechanism. Furthermore, this approach explains the important scatter for the values of surface-related capacitance in (F m-2)

quoted in the literature with no clear linear correlation between the specific capacitance and the specific surface area of activated carbons.

On the other hand, surface oxide related pseudocapacitance contributions in the aprotic electrolyte 1M (C2H5)4NBF4 in acetonitrile are much weaker (9 F mmol-1 of CO generated in TPD, for the best performing carbons) than in the H2SO4 electrolyte (63 F mmol-1 of CO).16 As illustrated by Figure 1, the specific capacitance at low current density, Co, increases linearly with the total specific surface area, following the approximate correlation

Co[(C2H5)4NBF4] (Fg-1) = 0.09 (Fm-2) Stot (m2g-1) (15)

The deviation observed for some carbons is attributed to exclusion effects in the aprotic electrolyte. The fact that these materials display average micropore sizes or "gate" effects around 0.5-0.6 nm16 limits the internal surface area accessible to ions (C2H5)4N+ with critical dimension around 0.69 nm.

Total surface area (m2g-1)

Figure 1. Variation of the specific capacitance (1 mA cm-2) of activated carbons with the total surface area in 1M (C2H5)4NBF4/acetonitrile. The deviation for some carbons (□) reflects exclusion effects in the aprotic electrolyte.

Total surface area (m2g-1)

Figure 1. Variation of the specific capacitance (1 mA cm-2) of activated carbons with the total surface area in 1M (C2H5)4NBF4/acetonitrile. The deviation for some carbons (□) reflects exclusion effects in the aprotic electrolyte.

Finally, the present work confirms the possibilities of immersion calorimetry used alone for the prediction of the specific capacitance at low current density of unknown activated carbons in the aprotic electrolyte 1M (C2H5)4NBF4 in CH3CN (Figure 2). As far as the specific capacitance and the enthalpy of immersion into benzene, -AjH[C6H6], are surface-related properties of the materials, one obtains a relatively good correlation between both parameters

Co[(C2H5)4NBF4] (Fg-1) = -0.57 (F J-1) AH[C6H6] (Jg-1) (16)

As suggested by recent calorimetric tests,17 the micropore system of typical activated carbons where the average micropore size is above 0.8 nm displays a similar accessibility to (C2H5)4NBF4/acetonitrile and to benzene. The deviations observed in Figure 2 (□) confirm the limited accessibility of the aprotic electrolyte into micropores of less than 0.7 nm.

Figure 2. Correlation between the specific capacitance (1 mA cm-2) of different activated carbons in the aprotic electrolyte, Co [(C2H5)4NBF4], and the enthalpy of immersion of carbons into benzene, -A;H(C6H6). (□) Carbons with micropore widths below 0.7 nm.

Figure 2. Correlation between the specific capacitance (1 mA cm-2) of different activated carbons in the aprotic electrolyte, Co [(C2H5)4NBF4], and the enthalpy of immersion of carbons into benzene, -A;H(C6H6). (□) Carbons with micropore widths below 0.7 nm.

0 50 100 150 200 250 300

Figure 3. Evolution of the specific capacitance (1 mA cm-2) of different activated carbons in H2SO4 aqueous electrolyte, Co [H2SO4], and the enthalpy of immersion of carbons into benzene, -A;H(C6H6).

On the other hand, Figure 3 shows the limitations of this technique to assess, with a good accuracy, the suitability of a carbon to be used in H2SO4 aqueous capacitors. It is not surprising, in view of the significant influence of the CO-generating surface groups on the capacitance in H2SO4 aqueous medium (see Eq. (14)) whereas the enthalpy of immersion into a nonspecific liquid such as benzene does not depend on the presence of oxygen.

4. Conclusions

Immersion calorimetry provides complementary information to the adsorption isotherms and simplifies considerably the assessment of porosity and chemical nature of activated carbons.

The assessment of activated carbons in the framework of Dubinin's theory with its extensions to calorimetry and adsorption from solutions led to the identification of key parameters for the performance of these materials in electrochemical capacitors. For 2M H2SO4 aqueous electrolyte, the limiting capacitance of activated carbons at low current densities depends essentially on the total surface area and on the surface groups which generate CO in TPD. In the case of the aprotic electrolyte 1M (C2H5)4NBF4 in acetonitrile, the specific capacitance increases linearly with the total specific surface area, the contribution from surface oxygen groups being less significant.

Micropore widths or "gate" effects of less than 0.7 nm notably reduce the internal surface area accessible to ions (C2H5)4N+ of the aprotic electrolyte, as opposed to the SO42- ion of the aqueous medium.

This work confirms the possibilities of immersion calorimetry used alone for the prediction of the specific capacitance of carbons in 1M (C2H5)4NBF4 in CH3CN, but it also shows the limitations of this technique to assess, with a good accuracy, the suitability of a carbon to be used as electrodes in H2SO4 aqueous capacitors.

References

1. F. Stoeckli, in: Porosity in Carbons, edited by J. Patrick (E. Arnold, London, 1995), pp. 67-97.

2. H.F. Stoeckli and F. Kraehenbuehl, The enthalpies of immersion of active carbons, in relation to the Dubinin theory for the volume filling of micropores, Carbon 19(5), 353356 (1981).

3. L. Ballerini, D. Huguenin, P. Rebstein and F. Stoeckli, Determination of the total surface area in carbonaceous adsorbents by the selective adsorption of caffeine from water, J. Chim. Phys. 87, 1709-1714 (1990).

4. F. Stoeckli, T.A. Centeno, J.B. Donnet, N. Pusset and E. Papirer, Characterization of industrial activated carbons by adsorption and immersion techniques and by STM, Fuel 74(11), 1582-1588 (1995).

5. F. Stoeckli, M.V. López-Ramón and C. Moreno-Castilla, Adsorption of phenolic compounds from aqueous solutions, by activated carbons, described by the Dubinin-Astakhov equation, Langmuir 17(11), 3301-3306 (2000).

6. E. Fernández, D. Hugi-Cleary, M.V. López-Ramón and F. Stoeckli, Adsorption of phenol from dilute and concentrated aqueous solutions by activated carbons, LanSmuir 19(23), 9719-9723 (2003).

7. T.A. Centeno and F. Stoeckli, The oxidation of an asturian bituminous coal in air and its influence on subsequent activation by steam, Carbon 33(5), 581-586 (1995).

8. F. Stoeckli and T.A. Centeno, On the characterization of microporous carbons by immersion calorimetry alone, Carbon 35(8), 1097-1100 (1997).

9. F. Stoeckli, A. Slasli, D. Hugi-Cleary and A. Guillot, The characterization of microporosity in carbons with molecular sieve effects, Micropor. Mesopor. Mat. 51, 197-202 (2002).

10. F. Stoeckli and A. Lavanchy, The adsorption of water by active carbons, Carbon 38(3), 475-494 (2000).

11. M.V. López-Ramón, F. Stoeckli, C. Moreno-Castilla and F. Carrasco-Marín, Specific and non-specific interactions of water molecules with carbon surfaces from immersion calorimetry, Carbon 38(6), 825-829 (2000).

12. M.V. López-Ramón, F. Stoeckli, C. Moreno-Castilla and F. Carrasco-Marín, On the characterization of acidic and basic surface sites on carbons by various techniques, Carbon 37(8), 1215-1221 (1999).

13. R. Kotz and M. Carlen M, Principles and applications of electrochemical capacitors, Electrochim Acta 45(15-16), 2483-2498 (2000).

14. A.G. Pandolfo and A.F. Hollenkamp, Carbon properties and their role in supercapacitors, J. Power Sources 157(1), 11-27 (2006).

15. T.A. Centeno and F. Stoeckli, The role of textural characteristics and oxygen-containing surface groups in the supercapacitor performances of activated carbons, Electrochim. Acta 52(2), 560-566 (2006).

16. T.A. Centeno, M. Hahn, J.A. Fernández, R. Kotz and F. Stoeckli, Correlation between capacitances of porous carbons in acidic and aprotic EDLC electrolytes, Electrochem. Comm. 9, 1242-1246 (2007).

17. T. A. Centeno and F. Stoeckli, in: Recent Advances in Supercapacitors, edited by V. Gupta (Transworld Research Network, Kerala, 2006), pp. 57-77.

18. F. Stoeckli and T.A. Centeno, On the determination of surface areas in activated carbons, Carbon 43(6), 1184-1190 (2005).

ADSORPTION ON ACTIVATED CARBON: ONE UNDERLYING MECHANISM?

PETER LODEWYCKX*

Department of Chemistry, Royal Military Academy, Renaissancelaan 30, B-1000 Brussels, Belgium

Abstract. This paper deals with the differences and similarities that are observed for isotherms of different adsorbates on activated carbon. In stead of the commonly used classification of isotherms in six "basic shapes", a more general approach is suggested. Each isotherm can be subdivided in different, distinct, parts, each of which is the result of a different adsorption mechanism. The relative importance of these parts, even the fact if they exist or not, depends on a number of parameters. The most important are the pore size distribution (PSD) and active surface groups of the adsorbent, the strength of the adsorbate-adsorbent interactions and the temperature at which the isotherm is measured. As all these parameters are either constant (e.g. PSD) or can be estimated (e.g. interaction energies) it is possible to calculate a complete isotherm for one adsorbate on the basis of that of another one. But in order to achieve this, one must fully comprehend the meaning of each and every part of the isotherm: isotherm equations must reflect the physical processes involved rather than just fit the (complete) isotherm.

Keywords: adsorption; activated carbon; mathematical modelling

1. Introduction

Activated carbons are widely used, and the number of possible applications is still growing. In many, if not all, of these applications adsorption comes into play. Therefore, most researchers are used to characterize their activated

*To whom correspondence should be addressed. E-mail: [email protected]

J.P. Mota and S. Lyubchik (eds.), Recent Advances in Adsorption Processes for Environmental Protection and Security, 19-28. © 2008 Springer.

carbons by some kind of gas phase adsorption measurements, leading to adsorption isotherms. These isotherms contain a wealth of information about the carbon but, sadly, there remains a lot of uncertainty about their exact interpretation. Not only do the applied models differ from the ones used for other adsorbents (e.g. zeolites), but in many cases different models are used for different adsorbates. This is the direct result of (apparent) dissimilarities between the isotherms. In this work we will try to demonstrate that there is one underlying setcvcx of adsorption mechanisms that is applicable to all adsorbates (perhaps even to all adsorbents), the differences in the shape of the isotherms being only the result of interaction energy levels between the activated carbon and the vapour.

2. Type I Isotherms

2.1. GENERAL

According to the BDDT-classification isotherms can be subdivided into six major categories.1 Only four of them will be treated. Type III is in fact identical to the first part of type V, and type VI is a step-wise isotherm that is very seldom seen in real adsorption measurements on activated carbon. The type I isotherm is the typical shape attributed to adsorption into micropores. However, it can also be found in other adsorption processes: the first part of both type II and type IV isotherms is clearly of this same shape. This is also valid for the very first part of the type V isotherm: if ones carefully studies the isotherm under p/po = 0.2 it becomes apparent that the very start of the convex type V isotherm is in fact concave towards the relative pressure axis (see Sections 2.3 and 5.1).

2.2. MICROPORE FILLING

The typical knee of the "real" type I isotherm is attributed to micropore filling. This can be described by a number of equations, the most commonly used being the Dubinin-Radushkevich or DR equation1 (Eq. (1)) that gives the amount adsorbed We as a function of partial pressure p/po or concentration c/cS. The parameters depending on the adsorbent-adsorbate combinations are the micropore volume of the adsorbent Wo, the liquid density of the adsorbate dL, the scaling factor p (dependent on the energy of interaction) and B, related to the interaction energy and proportional to the mean micropore half-width.

It has been observed that in some cases, where there is clearly volume filling of micropores, the DR equation does not provide a good fit. In these cases one should use the Dubinin-Astakhov (DA) equation in which the factor 2 is replaced by a factor n.

2.3. NON-MICROPORE FILLING: THE WATER ISOTHERM

The start of the water adsorption isotherms on activated carbons seems to be of type I. It has also been demonstrated that it can be fitted with a DR or DA equation.2 However, it is clear this first part is not related to micropore filling as this phenomenon does not start before p/po « 0.3 (see Section 5.2). Hence fitting by DR or DA is just a mathematical tool but has no physical meaning. It is generally agreed upon3 that water adsorption on activated carbon starts with specific adsorption on active surface sites. These sites contain mostly oxygen (and sometimes nitrogen) and can be of an acidic or basic nature. Then water-water interactions (hydrogen bonding) promote the formation of water clusters around these active sites. Subsequently, these clusters are adsorbed in the micropore volume (see Section 5.2). Therefore, this part of the water isotherm can be considered as an adsorption on a non-porous surface with active sites, thus being described by a Langmuir type equation1:

In which n represents the amount adsorbed, nm the monolayer, b is related to the interaction energy and po/p is the partial water vapour pressure. Equation 2 gives a very good fit of the experimental isotherms for a wide variety of activated carbons. An example can be found in Figure 1.

p/po

Figure 1. Langmuir fit of the initial part of the water isotherm.

The fitted parameters differ from one activated carbon to another. But there is a clear positive correlation between nm and the number of oxygenated surface sites. The coefficient b is very small, in the order of magnitude of 5. This is rather surprising, b being a measure for the interaction energy. Therefore it seems logical to assume that b is not only related to the specific interaction between water and the active sites, but also comprises the mutual bridging between water molecules leading to the formation of water clusters. As these clusters will be formed around the initially (specifically) adsorbed water molecules, the number of clusters will still be related to the number of active sites. Therefore, nm is still related to the number of active sites, even though the value of b is the result of two separate mechanisms.

2.4. NON-MICROPORE FILLING: TYPE II ISOTHERMS

The initial part of a type II isotherm, typical for adsorption onto non-porous surfaces, also has the characteristic type I form. It is clear that in this case it does not represent micropore filling, but only the formation of a monolayer on the surface of the adsorbent. How this monolayer is exactly forming on an activated carbon is, however, not really understood. It is clear it is not an instant coverage of the total surface. Most likely, as it is the case for water vapour adsorption, it will start at those spots that present the highest interaction energy with the adsorbate, i.e. oxygen surface groups. Then adsorption will spread out to form a monolayer. Even though it is clear that in some places multilayer adsorption will precede complete

0 0

Post a comment