118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure.This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point.These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order. The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility.The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No. 13.118 ELECTRICAL THEORY OF ADBORPTTON The writer considers the double layer as consisting of a swface of rigidly fixed atoms under continuous bombardment of positively and negatively charged ions, any particular point on the rigid surface becoming in turn negative, neutral and positive, these conditions arisdg in any order.The observed contact difference is the average effect of these conditions. Where several kinds of atoms are present in the solution the average number of any one of them at the surface will depend on their concentbration, valency and mobility. The variation of contact Werence from negative to neutral and positive was observed with cotton and aluminium sulphate near the neutral point. These variations occurred during the same experiment, the readings being direct measurements of E.1I.F.s developed by filtration under pressure. This point would be covered by putting n2 = 1 and = 2 or 3 in Mukherjee’s equation No.13. ON THE ADSORPTION OF GASES. SECTION I. EXPERIMENTAL METHODS. INTRODUCTORY PAPER TO SECTION I. BY ERIC K. RIDEAL. Received I 9th December, I93 I . In an introductory paper it is impossible even to give a sketch of the enormous field embraced under the discussion. I have accordingly attempted to give in briefest outline only a short account of a few of the problems connected with the experimental investigation. We may note as of primary importance, the determination of the specific surface of the adsorbent. In logical sequence we must examine the various methods which have been employed for the examination of the final distribution of the gas which has been sorbed, i e . , between the surface and the bulk phase respectively.Examination of the rates of attainment of equilibrium involve the measurement both of the critical energy increments and the reaction velocities of the various processes taking place when a solid is exposed to a gas. Finally, we must mention the experimental methods which have been devised to examine in more detail the process of formation of surface phases. The Specific Surface. Of fundamental importance is the nature and extent of the surface phase of the solid. With the realisation of the somewhat porous structure of the surface layers of most solids, the evaluation of the accessible area rather than the true specific area has been more important. Some five methods have been employed for this purpose. These may be briefly enumerated as follows :- ( I ) Comparison of the rate of solution in some reagent with that of a uniform surface of similar material, which is a method originally employed by Schmidt,l W ~ l f f , ~ Durau and more recently by Schwab and Rudolph,* for evaluation of the specific surface of reduced nickel.Lack of uniformity in the rate of solution is the chief objection to this method. ( 2 ) By determination of the Newtonian " temper " colours developed by metals when attacked by reactive gases. From a knowledge of both the film thickness determined optically and either the increase in weight of the metal or the volume of gas which has been taken up by the ISchmidt, 2. PhysiR. Chcm., I I ~ A , 236, 1925. 'Wolff, 2. angew. Chem., 35, 138, 1922. 8 Durau, 2. Physik, 38, 419, 1926. 4 Schwab and Rudolph, Z.physik. Chem., I ~ B , 427, 1931. I39I 40 INTRODUCTORY PAPER TO SECTION I metal, the specific surface can be evaluated. This method has been employed both by Dunn5 and by Constable.6 Uniformity of film thick- ness and uniformity of composition of the film is, as observed by not readily obtainable, and all irregularities in the surface texture must be assumed to possess a substrate thickness a t least equal to that of the resulting film. (3) The change in electrode potential caused by the passage of a definite electric flux 6.010~ coulombs per sq. cm. for a 100 mv. change in interface potential across a liquid metal interface, permits of the evaluation of the specific surface. This method was developed by Bowden and Rideal,8 and has been used by Bowdens and by Volmer and Erdly Griz.lo A theoretical treatment of the method on the basis of the wave mechanics has recently been given by Gurney.ll (4) The fact that the conditions of reversible adsorptive equilibrium are quickly established between the surface of an insoluble salt such as lead sulphate and a solution containing isotopic ions such as the radio- active thorium B, has been utilised by Paneth and his co-workersl* to evaluate the specific surface of such salts, and by Hahn.13 ( 5 ) A number of investigators have attempted to evaluate specific surfaces from experimental data on the adsorption of solutes which are readily quantitatively determined in small quantities.It is assumed in all cases that the saturation maximum corresponds with a close packed unimolecular layer.Dyes estimatable by colorimetric methods have been employed by Paneth and Radu,l* by Bancroft and Barnett,16 and by Bancelin.ls Metallic ions such as silver determined by electropotentio- metric methods have been utilised by Euler.17 The inapplicability of the former, even when not colloidal, to fine grained or porous materials, and of the latter to interfaces which in general possess a potential difference, e.g., mercury in a Lippmann electrometer to the exact deter- mination of the specific surface, is clearly evident ; the method although simple is very restricted in scope and the accuracy attainable very problematical. After the evaluation of the specific surface of the solid the next prob- lem which arises in the experimental investigation is the examination of the detailed molecular structure of the surface, this is necessary since the characteristics of the adsorption isotherms as well as the thermal changc-; appear to be dependent on the structure.It cannot be said that much progress has been made in this direction. We may note the at- tempts which have been made for the evaluation of the mean grain size when crystalline, utilising Debye's method of X-ray examination by Astbury and Clark,l* and by FeichtknechtlQ and the more recent ex- 5Dunn, P.R.S., IOIA, 203, 1926. Constable, P.R.S., 117A, 376, 192s ; 119, 196, 1928. Evans and Bannister, P.R.S., 125A, 370, 1929. Bowden and Rideal, P.R.S., 120A, 63, 1928. Bowden, P.R.S., 125A, 446, 1929. lo Volmer and Erdly Griz, 2. physik. Chem., 150A, 203, 1930.IlGurney, P.R.S., I ~ A , 137, 1931. 12 Paneth, 2. physik. Chem., I O I , 445, 480, 1922 ; Ber., 578, 1215, 1924. l3 Hahn, 2. $&ysik. Chem., 144A, 161, 1929. l4 Paneth and Radu, Ber., 57B, 1221, 1924. l6 Bancroft and Barnett, COX Symp., 6, 73, 1928. l6 Bancelin, J . Ckim. Physique, 22, 522, 1925. l7 Euler, 2. Elektrochem., 28, 2, 1922. l8 Astbury and Clark, J.A.C.S., 47, 2261, 1925. Is Feichtknecht, 2. Elektrochem., 35, 142, 1929.E. K. RIDEAL periments of G. P. Thomson,20 in interpreting the diffraction patterns obtained with high speed electrons. The existence of a network postulated by Smekal2I and Lennard Jones and Dent,22 has been demonstrated by optical methods by Z ~ i c k y . ~ ~ The problem of the surface structure is evidently capable of being attacked in an alternative manner not by analysis but by synthesis.By laying down a fresh surface in vacuo either as small aggregates in cathodic spluttering or molecularly by evaporisation, its properties and the change of these properties on ageing and with time have been the subject of several investigations. We may mention the work of Estermann 24 on the growth of crystals of silver and cadmium deposited by molecular rays of these metals. Optical methods for the examination of the deposits have been developed by Reinders and Hamburger,25 whilst the numerous observations on the change in electrical conductivity with age have been the subject of theoretical discussion by Frenke126 and Ehrenberg and The adsorptive powers of deposited films for various gases has been the subject of detailed investigations by Frankenburger and his co- workers.% The experiments of Bastow29 on the sorption of nitrogen and hydrogen by deposited platinum are of interest in this connection.The Surface Phase. If a metal surface be exposed to a gas and sorption takes place, it is important to consider what experimental methods can be devised to dis- cover whether the gas is retained on the surface of the metal or no. A number of such methods have in fact been employed and a t least one of them has been found to be capable of yielding quantitative, as compared with qualitative, results as to the existence, extent and nature of the surface phase. We may briefly review these methods :- (a) At high temperatures the thermionic emission from a clean surface is given by Richardson's equation i = AT2e-+IKT.The work function + is very considerably modified by the presence of foreign atoms in the surface phase. For dilute films this modification may be expressed in the form 4' = + + a0 where 9 is the fraction of the surface covered with foreign atoms, a being either positive or negative. For surface concentrations of less than 10 per cent. in the case of caesium and for closely packed films, this simple expression is no longer valid. The method has been developed as a weapon for the examination of sur- face film structure, more especially by Langmuir, Becker, and Kingd~n.~O Whilst the method of examination by thermionic emission is naturally restricted to relatively high temperatures, the change in potential a t a gas liquid interface caused by the insertion of a film can readily be measured at ordinary temperature^,^^ and it has been found possible zoG.P. Thomson. P.R.S., 133, I, 1931. 21 Smekal, Physik. Z., 26, 700. 1925 ; 45, 869, 1927. 28 Lennard Jones and Dent, P.R.S., I ~ I A , 247, 192s. a3 Zwicky, Proc. Nut. Acad. Scz., 15, 253, rg2g. 24 Estermann, 2. physik. Chem., 106, 403, 1923 ; 2. Physzk., 33, 320, 1925. 2SReinders and Hamburger, Rec. Trav. Chzm. Pays Bas, 50, 357, 475, 1931. a7 Ehrenberg and Honl, 2. Physik, 68, 289, 1931. 28 Frankenburger, 2. Elektrochem., 35, 920, 1929, et seq. 2Q Bastow, J.C.S., 193~. 30 Langmuir, Becker, and Kingdon, see Schottky, Handbuch expt. Physik. ; 31 Schulman and Rideal, P.R.S., 130A, 259. 1931. Frenkel, Physical Rev., 36, 1604.1930. Dushman, Reviews of Modem Physics. 2, 381, 1930.142 INTRODUCTORY PAPER TO SECTION I (unpublished work by H. Whalley) to extend this method to metal gas interfaces. ( b ) The presence or absence of a surface phase can be demonstrated by the nature of the diffraction pattern obtained on reflection of electrons of relatively low velocities from metallic surfaces, a method due to Davisson and Germer,32 and extended very considerably by R ~ p p . ~ ~ The necessity for high vacua precludes the methods from becoming a general one for the determination of the conditions of equilibrium. It may be noted in passing that the difficulties of obtaining a metal sur- face really free from gas is clearly demonstrated by this method of examination. (c) Many attempts have been made to obtain information to the existence or absence of a surface film from an examination of the long wave threshold value of the photo-electric emission.Those of Ives 34 may be cited as an example of the method in its application to films of the alkali metals. The difficulties inherent in the method are great, and although as yet its general applicability must be considered restricted, it is a method worth further serious attention. (d) The most suitable optical method at present available for the detection of a unimolecular film is that based upon the observations of Jamin35 that plane polarised light when reflected from the surface of water was slightly elliptically polarised. The theoretical treatment by Drude (Theory of Optics) revealed on reflection of plane polarised light both the existence of a change in amplitude and in phase of the two component beams when there existed a transition layer or film of refractive index different from either medium.By suitable measure- ments it should be possible to obtain not only the film thickness and the refractive index but also the dispersion of the film forming material. I t is possible that with the development of the theory of molecular scatter- ing, the method may be capable of giving information on the electrical properties of the adsorbed phase. The most exhaustive examination of a surface phase by this method has been carried out by Tronstad on oxide films on metals. In all cases his films were relatively thick. Unimolecular films have been detected by numerous investigators and making assumptions which in some cases a t least are somewhat un- warranted, molecular dimensions have been calculated from the observations.The work of Sissingh and Haak36 on mercury surfaces, of Ives and Johnson 37 on thin films of rubidium, of Frazer 38 on alcohol on glass, and of Bouchet,S9 may be mentioned as representative of the experimental methods employed. (e) There are two thermal methods of investigation which appear to be capable of more precise development. The amount of energy carried away from a wire at a temperature T, by the impingement of molecules of a pure gas at a temperature TI on its surface is given by Knudsen’s (T2 - T1) where a is the accommodation co- equation in the form Q = Ka dc 39 Davisson and Germer, Physical Relr., 2, 35, 705, 1927.83 Rupp, Ann. Physik, 5 , 453, 1930. Ives, Physical Rev., 34, 117, 1929. 86 Jamin, JOUY. de Chem., 31, 165, 1851. *a Sissingh and Haak, P.R. Acad. Sci. Amst., 21, 678, 1919. ST Ives and Johnson, Proc. Opt. SOC., 15, 374, 1927. 88 Frazer, Physical Rev., 33, 47, 1929. 89 Bouchet, C.R., 185, 200, 1927 ; and J . Physique, 1931.E. K. RIDEAL '43 efficient. The accommodation coefficient a is dependent on the nature of the surface. In the presence of an adsorbed layer of a foreign gas the value of a rises. The experimental work of Hughes and Bevan,m Chapman and Ha11,O F a r k a ~ , ~ ~ and especially the investigations of Roberts 43 may be cited as applications of the method. Both the monochromatic as well as the total thermal emissivity of a surface is dependent on the nature of the surface.We may cite the values of the latter for (bright) platinum and oxidised platinum (black ?) from the I.C.T. Temperature. Pt 400' 600' 4.66 7-50 PtO. 8.6 11'0 Van Praagh and Rideal 44 found that the temperature of a tungsten wire maintained in vacuo varied for identical energy consumptions from 44" C. to 80" C., the latter in presence of a film of tungstic iodide on the surface computed to be ten molecules in thickness. A film this thick- ness still possessed the characteristic dissociation pressure of a bulk phase. The total emissivity is accordingly less for this film coated sur- face than for the bare metal. It is possible that the change in emissivity caused by the presence of a unimolecular film is and that the applicability of the method is restricted to thicker films.The Bulk Phase. The methods of examination available apart from those involving the measurement of velocities. for the presence of the sorbed gas in the bulk phase of the adsorbent are somewhat less general than those described above for the detection in the surface phase, Whilst analy- tical methods present no difficulty, the distribution and state of the gas or vapour are not so readily discerned. We may divide the possible forms of distribution into the following : compound formation, lattice distribution, distribution in the slip planes and between the crystallites, distribution in micro-capillaries or channels. The increase in specific volume suffered by substances as varied as charcoal, metals such as palladium, platinum and iron, zeolites and clays, when sorption occurs is readily m e a ~ u r a b l e , ~ ~ but on account of the slow rate of attainment of true equilibrium the exact reversibility of the dimensional increase is difficult to demonstrate.Further expansion on sorption of gases has been demonstrated in the case of both platinum and palladium by means of the X-rays,47 but as is well known the phase diagrams for these systems are still in an unsatisfactory state. A few measurements have likewise been made of the change in electric resistance of wires on sorption and desorption of a g a ~ . ~ 8 Hughes and Bevan, P.R.S., I 17A, 102, 1928. 4 1 Chapman and Hall, i&id., 126, 478, 1929. 42 Farkas, 2. physih. Chem., 1931.49 Roberts, P.R.S., 129A, 146, 19.30. 44 Van Praagh and Rideal, P.R.S., 134, 400, 1931. 45 L O G . cit. See also Langmuir, J.A.C.S., 38, 2271, 1916. 46 See McBain, The Sorption of'Gases by Solids. 47 Handwalt, Physical Rev., 33, 444, 1929 ; Osawa, Tohuku Imp. Univ., I, 14, UI Sieverts, 2. Metallurg. 3, 37, 1913, 45, 1925.144 INTRODUCTORY PAPER TO SECTION I Rates of Sorption. Measurements of the rate of establishment of sorption equilibrium in themselves give but little information, unless the rates of the various processes taking place can be separately evaluated. It would appear that surface adsorption equilibrium is almost instantaneously attained. The adsorbed gas may now undergo further operations involving thermal changes; it may undergo reaction with the substrate to form a chemi-adsorptive compound, such a process requires an energy of activation and involves a thermal change. This phenomenon appears similar to, if not identical with, Taylor’s hypothesis of acti- vated adsorption.The adsorbed gas may penetrate into the sub- strate either along the intergranular boundaries or slip planes or possibly in the finer dimensions postulated by Smekal, or actually penetrate through the space lattice of the crystalline solid. It has been fairly definitely established that both these latter processes mimic ordinary chemical reactions, in that energies of activation are required for the diffusion processes and that these differ for the intergranular or slip plane and lattice diffusim processes respectively. Furthermore, the rate a t which both these processes occur follows the ordinary Fick diffusion law in the sense that the rate of passage from the surface to the interior and in the opposite sense on desorption is dependent on the concentration gradient from the interior to the surface, the concentration a t the surface being proportional to the amount of gas actually superficially adsorbed which is by no means necessarily proportional to the pressure of the gas.We may cite the experimental work of Langmuir49 and Cla~sing,~o on the migration of thorium through tungsten ; of Dunn,51 F e i ~ h t n e c h t , ~ ~ and Wilkins and Ridea153 on the diffusion of oxygen through copper oxide, and of Ward 64 on the diffusion of hydrogen through copper, as examples of direct measurement of rates of such diffusive processes and the determination of the critical energy increments.Measurements of the rates and critical energy increments of the other possible reaction of the adsorbed gas, namely, conversion into a chemi- adsorptive compound, have generally been under conditions when the possibly important processes of intergranular and lattice diffusion have been ignored, nevertheless the existence of this transformation has been clearly demonstrated although the values given for the critical energy increments based upon the supposed rates may be somewhat inexact. We may cite the work of Garner and his co-workers on the adsorption of oxygen by charcoal,55 and of Taylor and his colleagues on the ad- sorption of hydrogen by oxides. We may note in passing that since the rate of the intergranular dif- fusion into a fully evacuated solid is governed by the amount of gas adsorbed on the surface and the texture of the surface, ( i t ? ., the number and size of the fissures), this rate may be very considerably affected by the presence of any substance sufficiently strongly adsorbed, which not only reduces the amount of gas adsorbed on the surface but may also reduce the number and size of the fissure entrances, this phenomenon is 51Dunn, P.R.S., III, 210, 1926. 49 Langmuir, Physic. Rev., 22, 357, 1923. 50 Clausing, Physica, 7, 193, 1927. 62 Feichtnecht, 2. Elektrochem, 35, 142, 1929. 53 Wilkins and Rideal, P.R.S., 128, 394, 1930. “Ward, P.R.S., 133, 523, 1931. 66 Garner, J.C.S., 125, 1288, 1924 ; 2451, 1927 ; 2870, 1928 : Nature, 124, 4’9, Ig29.E.I(. RIDEAL I45 exemplified in the work of Maxted on the influence of hydrogen sulphide on the rate of sorption of hydrogen by platinum.56 Rates of Surface Migration. We have already referred to the measurement of the rates of inter- granular and lattice diffusion processes involved in the phenomenon of gaseous sorption and to the fact that adsorbed molecules may undergo reaction with the substrate to form a chemi-adsorptive compound. The chemi-adsorptive compound is presumably rigidly held on to the sub- strate surface and can only undergo lattice diffusion a t elevated tem- peratures. We must esamine the experimental methods for determining whether the adsorbed molecules are held in the sense that they perform oscillations within a small compass on the surface or whether they can migrate over the surface. Direct evidence for the lateral mobility of molecules moving over a substrate of the same material and of molecules moving over glass have been provided by actual visual and micro-balance observations by Volmer and his c o - ~ o r k e r s .~ ~ The movement of metallic atoms deposited on glass and on metal surfaces from beams of molecular rays has been directly observed by Estermann, by Stern,5s and by C o c k ~ r o f t . ~ ~ Whilst the numerous ex- amples of mobility a t relatively low temperatures provided by the phenomenon of sintering are well known, Becker’s observations on the migration of the alkali metals and of the alkaline earths over platinum by observation of the change in the thermionic emission, likewise provide us with direct evidence on this point.The method of measurement of the change in the gas solid interfacial potential men- tioned above is likewise capable of providing similar information as well as on the rates of formation and evaporation of surface phases at low temperatures. A study of kinetics of the oxidation of copper61 provides us with indirect evidence for the lateral mobility of oxygen over copper oxide. The hypothesis of lateral mobility of adsorbed gases provides us with a very useful mechanism for the interpretation of surface actions,62 al- though if the activation process for a surface catalytic chemical reaction involves the activation of a chemi-adsorptive compound which by some is supposed to be the only species capable of entering into a subsequent chemical reaction, the hypothesis of lateral mobility of a t least one of any two surface reacting gases must be given up. A very brief consideration of the observed effect of temperature on these lateral velocities show that in general the rate of movement is not only far less than in the free gas phase a t corresponding temperatures but, furthermore, possesses a t least in the cases carefully examined on relatively clean and dry surfaces (e.g., the experiments of Becker) a definite exponential temperature coefficient indicative of an energy of activation being requisite for the process of lateral diffusion, i.e., in passing from point to point in a non-uniform atomic field.I t is possible that this energy of activation for lateral diffusion may prove to be identical with the ct seq.68 Stern, 2. fihysik. Chem., 106, 397, 1923 ; and 39, 774, 1926. sB Cockcroft, P.R.S., I 19A, 1928. 6o Becker, Trans. Amer. Electrochem. Soc., 1929. 61 Wilkins, P.R.S., 178A, 407, 1930. 68 See especially Schwab, Katalyse. I 1146 INTRODUCTORY PAPER TO SECTION I value obtained for the process of diffusion in the intergrannlar boundaries and fissures of the solid. We may make an approximate calculation for this energy of activation of surface diffusion from Wilkins' experi- ments (Zoc. cit.) on the oxidation of copper. The rate of oxidation as determined by the rate of change of pressure at 500' K. at 300 mm. pressure was found to be 1'74 mm./sec. The area of the foil was 2 sq.cm. and the volume of the system about 100 C.C. Hence the rate of loss of oxygen into the interior was 0.115 C.C. (at N.T.P.) per sq. cm. per sec. At this pressure by the Herz Knudsen equation a t least 3'103 C.C. would strike the surface per second. If free lateral mobility were possible for all atoms striking and the accommodation coefficient were unity on the oxide coated surface, i t is clear that the limiting pressure observed, viz., 300 mm. would be much less. If only a fraction e - of the condensing moiec:iles acquire sufficient lateral mobility to migrate to a fissure we obtain = 0*115/3-10~ whence E = 12,400 cals. per gm. mol. for the energy of activation for surface migration, a value not widely different to 9500 cals. for the energy of activation of the intergranular diffusion process.In Davisson and Germer's experiments 32 on adsorbed hydrogen films on nickeI, the apparent " melting-point" of the film was found to be 400° K. If a t this temperature the number of hydrogen atoms moving from point to point becomes so large that the electron diffraction pattern becomes too weak for observation we obtain the small value of E = Ca 1000 calories. The occlusion rate of hydrogen through nickel has been measured by Sieverts 63 who noted a rapid sorption above 200' C. It is an interesting speculation how far an atom can travel when once liberated in this manner, but the evidence provided by molecular ray experiments appears to indicate that the range of action when in the liberated state is small and that the molecule suffers rapid readhesion.On Energy Exchange between Gas and Solid on Adsorption. We have mentioned the experimental methods which have been devised to test the hypothesis that molecules impinging on a surface may condense and then undergo chemi-adsorptive reaction or pass into the interior via the intergranular channels or through the lattice. When condensation does not occur the impinging molecule is reflected. The life of a molecule on the surface may thus be veqf short or very long; we may summarise the various experimental methods which have been devised to examine the result of collision of a molecule with a surface and the effect of increasing the lifetime of the molecule on the surface, ie., in altering the relative magnitudes of 4 and kT. For extremely short lifetimes comparable with the lattice frequency (10 - l3 secs.) the interaction must be regarded as governed by the laws of wave mechanics. If the atoms are arranged in a regular lattice a certain proportion then are reflected or diffracted and for them there is no interchange of energy a t all.We are indebted to Stern and K n a ~ e r , ~ ~ Estermann and Stern,= and Johnson for confirmation of the ap- plicability of de Broglie's wave equation by the method of molecular rays.67 In this case, as shown by Roberts,68 the accommodation co- 6* Sieverts, 2. Elektrochsm., 16, 707, 1910. 62 Stem and Knauer, 2. Physik, 53, 779, 1929. 6s Estermann and Stem, 2. Physik, 61, 114, 1930. 66 Johnson, J . Franklin Inst., a%, 308, 1928 ; 207, 635,1929 ; 210, 145, 1930. *' Molecular Rays, 1931, Fraser, Cambridge Univ.Press. Roberts, P.R.S., 1a9A, r46, 1930.E. K. RIDEAL I 4 7 efficients are small and the extent of this energy interchange is thus very small; an approximate theory of this case has been developed by Jackson. When the lifetime increases in this manner and the energy interchange becomes more pronounced the reflection becomes more and more diffuse as shown by the experiments of L a n g r n ~ i r , ~ ~ Knudsen, '* and Wood.71 With molecular rays the conditions for condensation can readily be examined and Frenkel's 72 analysis tested. At the critical stream densities and temperatures the lives are naturally much greater than those obtained for specular reflection. W e r t e n ~ t e i n , ~ ~ by observation of the rate of condensation of mercury vapour on glass, obtained ca.10 - 13 seconds a t 300" K. Cockcroft 74 obtained I O - ~ seconds a t 250' K. for cadmium on copper. These lifetimes are all maximum values. When condensation occurs not only has thermal equilibrium with the surface taken place but the condensed atoms have reacted with one another t o form a condensate. We may regard this phenomenon as the formation of a chemi-adsorptive compound with the substrate followed by activation and reaction with a neighbour. The importance of the lifetime on the surface as a factor in facilitating the transfer of energy is thus evident. The question of energy transfer in relationship to time of contact is clearly important in chemi-adsorptive reactions as the rate of transfer of kinetic energy of translation, of rotational energy and of internal vibrational energy may not be identical.That differences exist in the rate of transfer for the different kinds of energy in simple molecular collisions is apparent from the recent work of Kneser on carbon dioxide where a relaxation period of as long as I O - ~ seconds was found for the transfer of the vibrational energy, whilst Herzfeld and Rice,75 observed as a mean time for the transfer of rotational energy I O - ~ seconds. We may note in this connection that Rice and Byck 76 obtained no signs of decomposition a t a platinum target a t 1600~ C. of impinging beams of acetone or dimethyl mercury. In the latter case, if equipartition had been attained, with an energy of activation of 35,000 cals./gm. mol. over 6 per cent. decomposition should have been attained. According to Taylor's values the energy of activation for chemi-adsorption of hydrogen on a metal is about 12,000 cal./gm. mol. Taking hydrogen a t roo mm. pressure 103 C.C. strike I sq. cm. of the surface per second according to the Herz Knudsen equation. If a fraction e - E/RT of these underwent the chemi-adsorptive reaction, reaction would be complete in about I O - ~ seconds. If, on the other hand, the reaction be supposed to take place after adsorption and we consider N molecules adsorbed per sq. cm. a t any time, the rate at which these enter into reaction will be Nve - E/RT where I/V = T is this period of interchange. If we insert the ordinary vibration frequency T = 10 - l3 in the above equation we obtain as rate of reaction LV . 4-10s, i.e., an extremely rapid rate of reaction A value of v = I O ~ or T = I O - ~ seconds gives us a rate equal to 0.42, which if the surface is b u t sparsely covered is a measurable rate of reaction. The same point may be demonstrated in another way. Langmuir, Proc. Nat. Acad. Sci., 141, 1917. 70 Knudsen, Ann. Physik, 48, 1115, 1915. Wood, Phil. Mag.. 30, 300, 1915, 72 Frenkel, 2. PhysiR, 26, 117, 1924. 73 Wertenstein, J . Physique, 4, 281, 1923. 74 Cockcroft, P.R.S., I r9A, 1928. 75 Herzfeld and Rice, Physical Rev., 31, 691, 1928. 76 Rice and Byck..P.R.S., 50, 132, 1931.