Adsorption From Solution

U.S.A.

PREFACE

On the 31st July 1982, Douglas Hugh Everett retired from the position of Leverhulme Professor of Physical Chemistry in the University of Bristol, a post which he had held since 1954. This period of 28 years has seen marked changed in our University system and has also seen a great development of Colloid and Interface Science as a quantitative scientific discipline important both in industry and academe. A major role in this development has been played by Douglas Everett and this Symposium on Adsorption from Solution was held to honour Douglas Everett on the occasion of his formal retirement from the University.

Douglas Everett was born in Hampton, Middlesex, in December 1916, and after education at Hampton Grammar School, proceeded to the University of Reading to read chemistry. He graduated with first-class honours in 1938 and after a year as a research assistant at University College, Dundee, he went to Balliol College, Oxford, as a Ramsay Memorial Fellow, obtaining his D.Phil. in 1942. During this period he was much influenced by the late Lord Wynne-Jones and Professor R.P. Bell. After a period of wartime scientific service, for which he was awarded the M.B.E., he returned to Oxford as an I.C.I. Fellow in 1945 and then became a Fellow and Lecturer in Chemistry at Exeter College from 1947-1948. During this Oxford period the first series of fundamental papers on the thermodynamics of adsorption of gases on solids were written and later published in the Transactions of the Faraday Society. Following appointment to the position of Professor of Chemistry at University College, Dundee, in 1948, experimental work on this subject also began to develop. In 1954 Douglas Everett took up the position of Leverhulme Professor of Physical Chemistry at the University of Bristol in succession to Professor W.E. Garner who had held the chair since 1926; in fact Garner had succeeded J.W. McBain the first Leverhulme Professor who was appointed in 1919. Bristol provided the tradition and facilities for development of surface chemistry and there commenced in 1954 a long and fertile period of research into various aspects of interface and colloid chemistry. Amongst these topics came a succession of papers on Adsorption from Solution, firstly putting the thermodynamic analysis on to a sound footing and then developing experimental techniques for the precise determination of the extent of adsorption of small molecules on to various adsorbents. Later came an interest in the orientation of larger hydrocarbon molecules on to surfaces and then the development of operational methods for treating the adsorption of small polymer molecules and for considering the influence of adsorption on surface-surface interactions. A survey of this work is given in the Sir Eric Rideal Lecture of this Symposium. For his outstanding contributions to Colloid and Interface Science, Douglas Everett was appointed Tilden Lecturer of the Chemical Society in 1955, was the first recipient of the Chemical Society Award in Surface and Colloid Chemistry in 1971 and was elected a Fellow of the Royal Society in 1980. He was also President of the Faraday Division of the Royal Society of Chemistry from 1976-1978. Internationally his work has been recognised by Visiting Professorships in West Germany, in Canada, in the U.S.A. and Israel and many invitations to give Plenary Lectures. On the 1st August 1982, Douglas Everett became Emeritus Professor of Physical Chemistry in the University of Bristol.

This Symposium Volume is dedicated to Douglas Everett as a token of esteem from many friends, collaborators and scientific groups in many parts of the world. With it we express our thanks for the inspiration, clarification of thought, advice and scientific insight which we have received over the years. It is our hope that for Douglas Everett many more years of good health and research into the fascinating world of Colloid and Interface Science lie ahead.

R.H. Ottewill,     Bristol, January, 1983

THE SIR ERIC RIDEAL LECTURE

ADSORPTION FROM SOLUTION

Douglas H. Everett,     Department of Physical Chemistry, School of Chemistry, University of Bristol, U.K.

Publisher Summary

This chapter provides an overview of the study of adsorption from solution by solids. The adsorption of benzene from cyclohexane solution by Graphon is controlled almost entirely by enthalpy terms. The preferentially adsorbed component is that benzene exhibits more exothermic enthalpy of immersion in the pure state. The negative entropy of immersion in n–heptane compared with benzene presumably reflects the restriction imposed by the solid on the motions of an alkane chain. The phenomenon of adsorption from solution has important implications for other phenomena that arise from the relationship between adsorption and the free energy of surfaces. Adsorption from solution measurements can be of value in discussing wettability and liquid–liquid displacement problems. The chapter describes a study in which the observed contact angle for the heptane, water, and graphite system was 7° ± 2°, while the angle calculated from adsorption data was 8°. For benzene/water/graphite, no contact angle was observed, while the adsorption data gave a value of ‘cosθ’ > 1, indicating that heptane displaces water from a graphite surface. The influence of adsorption on the forces between colloidal particles is concerned with the way in which adsorption depends on the distance between the particle surfaces.

INTRODUCTION

This lecture surveys some of the more significant advances made during the last twenty years, in the study of adsorption from solution by solids.

John Kipling’s book (1), published in 1965, gives a very full account of progress up to the early 1960’s, and it was about that time that, largely as a result of discussions with him, I became interested in liquid/solid systems. His book shows that, despite the recognised importance of adsorption phenomena, and the extensive measurements made during the first half of this century largely on systems of practical interest, rather limited progress had been made towards a fundamental understanding of adsorption from solution. On the experimental side this was in part because the importance of working with well-characterised surfaces was insufficiently appreciated, and partly because the experimental techniques were of inadequate precision to enable any detailed theories to be tested. It is, however, surprising that theoretical studies of the solid/liquid interface were neglected since the thermodynamics of a simple model of the vapour/solution interface had been formulated in 1932 by Butler (2) and developed by several authors (3). Extension to the liquid/solid interface was not made until much later (4) even though this involves no major new concepts other than consideration of the added contributions to the energy of the system from interactions between the solid and molecules in the liquid phase. The last two decades have seen major progress in both the range and significance of experimental work and in its theoretical understanding, derived from work by many research groups worldwide (5,6,7,8). In this lecture, however, I want to make particular reference to the contributions made by workers in Bristol, whose names are listed in the Appendix.

The main topics dealt with are the following. First, the role of simple models and of elementary thermodynamic and statistical mechanical arguments is discussed in relation to the concept of a ‘surface phase’. Monolayer models of the adsorption region are often inadequate and thicker surfaces have to be assumed. A sharp boundary between surface and bulk phases is thermodynamically unacceptable and it is necessary to develop multilayer models and extend them to mixtures of molecules of different size. Although approaches of this kind form a useful basis for the interpretation of the behaviour of relatively simple systems, real systems exhibit more complex behaviour. It is not easy to perceive the modifications which need to be incorporated into the simpler models without guidance as to the source of the inadequacy of the theoretical models. It is therefore important to examine the usefulness of describing adsorption behaviour in purely thermodynamic terms without reference to any specific physical model. The results of such an analysis emphasize the importance of the interplay between enthalpic and entropic contributions to the free energy, and help to direct attention to the ways in which the simpler models need to be developed to provide more realistic theories of adsorption phenomena. Evidence is found for significant structural changes in the interfacial region, and a brief account is then given of some of the more striking examples. Finally, the role of adsorption from solution in controlling other phenomena such as wetting and liquid-liquid displacement, and the forces between colloidal particles is outlined.

SIMPLE MONOLAYER MODELS OF ADSORPTION

In developing a theoretical model for the solid/liquid interface it was natural initially to apply concepts and methods of approach analogous to those employed in discussing bulk phase equilibria. Thus the surface region was looked upon as a separate surface phase. The continuous distribution relating the average local concentrations x of different species to distance from the surface z was replaced by a step function (Figure 1). The so-called ‘surface phase’ occupies a definite region of space within which the concentration is uniform and different from the bulk. It was, however, appreciated (9) that the properties of such a phase are determined not only by forces between molecules in the surface, but also by interactions with the adjacent bulk phases: ‘surface phases’ are not truly autonomous.

Fig. 1 Approximation of continuous variation of concentration near a surface by a step function defining a ‘surface phase’, σ.

The condition of equilibrium between bulk and surface regions is that of constancy of chemical potential of each species on passing from locations adjacent to the solid into the bulk liquid phase. To apply this condition to the surface phase model it is necessary to express the surface chemical potential in terms of the local concentration in the surface phase, and to relate this local concentration to the observable adsorption parameters. One conveniently measured quantity (, or, if the specific surface area, as, of the solid is known, as the areal surface excess:

(1)

is related to the Gibbs relative adsorption of component 2 with respect to 1:

(2)

in the surface can be calculated from the equation:

(3)

are the areas subtended on the surface by molecules of types 1 and 2 respectively (molecular or molar surface areas). In the particular case of monolayer adsorption of molecules of the same size

(4)

Within these somewhat crude approximations one can envisage systems in which both bulk liquid and surface phases are ideal (can then be written

5(a)

5(b)

In ij are the same for molecules in the bulk and adsorbed phases and that the solid interacts appreciably only with molecules in the surface phase.

The resulting equilibrium conditions may be looked upon as defining an adsorption equilibrium constant, Ka, for the following exchange process taking place on a given area of solid (11):

(6)

which has the simple form

(7)

with

(8)

Taking equations (4) and (7) together leads to a relation between the observed specific surface excess and the bulk liquid composition:

(9)

as Ka increases. It is often found that experimental data for simple organic mixtures adsorbed by graphitised carbon or silica gel conform to the linear relation (9), from which values of Ka and nσ/m can be derived. If the size of the molecules is known, an estimate can be made of the specific surface area of the solid. In a surprisingly large number of examples the areas so obtained are in good agreement with those calculated from nitrogen gas adsorption by the BET method (12).

However, equation (9) is not a very sensitive test of the ideality of the system, and its use can be misleading. For example, there are cases in which linear relationships are observed experimentally even when the bulk solution is known to be non-ideal.

The model may be extended by introducing activity coefficients for both bulk and surface phases, (13), when the equilibrium constant adopts the more general form:

(10)

Conformity with .

The relationship between activity coefficients in the bulk and surface regions has been the subject of considerable speculation. Thus several authors (,

(11)

Here t is a proportionality factor* such that when t = 0 the surface is ideal, while when t = 1 the surface and bulk activity coefficients are the same.

The non-ideality of the surface phase can be assessed by combining (10) and (4) to give(13,16)

(12)

and compared with theoretical predictions. Alternatively, individual activity coefficients can be computed from the following thermodynamically derived equation: