LEWIS STUDIES IN CATALYSIS. PART XEI. CSXV1I.-Studies in Catahpis. Part XII. Catalytic Cderia and the Radiation Hypothesis. By WILLIAM CUDMORE MCCULLAGH LEWIS. TEE criteria which have been suggested from time to time as apply-ing to the phenomenon of catalysis are as follows (compare Rideal a.nd Taylor “Catalysis in Theory and Practice,” Chap. 2): (I) The chemical composi$ion of the catalytic agents is un-changed on completion of the reaction process. (2) Minimal amounts of a catalytic agent are adequate for the transf armation of large quantities of the reacting substances. (3) A catalyst does not affect the final state of equilibrium. (4) A catalyst modifies the velocity of two inverse reactions to the same degree LEWIS STUDIES IN CATALYSIS. PART XII. 1361 (5) A catalytic agent is incapable of starting a reaction; it c&~l only modify the velocity of the reacbion.Criteria (I) and (2) are closely related (2) in fact being the corollary of (1). Both would be accepted a t once provided secondary effech are excluded. To this extent they state a fact of experience and indicate that catalysis is simply a special case of ordinary chemical reactivity. Criteria (3) to (5) which form a group by themselves are in a different category as representing generalisations which may or may not be true. Criterion (4) is the corollary of (3) so that the group contains two distinct criteria. Considerable difference of opinion exissts a t the present time regarding the validity of these conclusions according to the point of view adopted as the basis of criticism.It is of some interest therefore to examine criterion (3) or ( 4 ) and criterion (5) from the point of view of the radiation hypothesis of chemical reactivity. On the radia-tion hypothesis the possibility of a reaction occurring depends on the existence of radiation of a type or frequency absorbable by the reacting substance the quantum of which radiation is suffici-ently large t o communicate the necessary critical increment to the molecule. I n the case of thermal radiation which is the kind of radiation envisaged in the quantum theory theoretically all possible wavelengths or frequencies are represented a t any temperature. Consequently the type of radiation necessary for any reaction is present in the space occupied by the matter and therefore every reaction is correspondingly possible.This mu* include the so-called catalytic reactions as well as those to which this name is not applied. From this point of view therefore c a ~ we conclude that a catalyst does not initiate but simply accelerates a process which would occur although under certain conditions infinitely slowly ? This cannot be affirmed withont qualification. In the form in which criterion ( 5 ) is stated it is evidently assumed that the same process may occur whether the catalyst be'present or not but this assumption is not necessarily true. It seems necessary to ascribe in certain cases if not in all a definite stoickeiometric molecular mechanism to a catalyst just as one would to any other reactant and consequently by adding such 8 catalyst a new process commences (the origin of which is the field of radiation) that happens to give rise to certain end-produh, which we believe might be attained in the absence of the catalyst.The fact appears to be that criterion (5)) as ordinarily stated, involves a false antithesis. From the point of view of the radis-We shall consider criterion (5) in the first place 1362 LEWIS STUDIES IN CATALYSIS. PART XII. tion hypothesis a catalyst may be said either ta render a reaction possible by supplying the necessary matter or it may merely accelerate according to circumstances. In no case is it the funda-mental initiator of a process. The r81e played by the catalyst may be conveniently illustrated by the catalytic effect of an acid in the inversion of sucrose or the hydrolysis of an ester.Prior to the addition of the acid the reaction is possible involving reaction between a molecule of sucrose or ester and either a molecule of water or its ions probably the undissociated molecule. On addi-tion of the catalysing acid the hydrogen ion accelerates the process already begun by the hydrogen ions already present. The undis-sociated molecule of the acid may also accelerate the reaction but in doing so it is almost certain that it produces an intermediate substance which was not formed in its absence. In so far as the intermediate stage is concerned the molecule of the acid has rendered a new intermediate process possible although the final produds are independent of the nature of this intermediate stage.The real source or origin of initiation of any reaction on the radiation hypothesis is the radiation itself. The material catalyst, if it acts simply as a transformer hastens a reaction which radia-tion has already initiated. The catalyst may also act as a mole-cular reactant giving rise under the stimulus of radiation to new intermediate products. The validity of conclusion (5) depends, therefore on the particular view adopted regarding the mechanism of the process. It seems that two distinct modes of mechanism are possible and are apparently realised in the well-known acid catalysia. On one mode the catalyst simply accelerates; on the other it renders a new mechanism possible from the material point of view. Turning now to criterion (3) or (4) which possesses much greater practical significance the radiation hypothesis leads t o the con-clusion that as a general principle criterion (3) or (4) is certainly not true.Let us take the simplest possible case of reversible reaction, represented by A B. The substance A is characterised by being capable of absorbing radiation of frequency v, as a result of which it is transformed into B. The substance B is likewise capable of absorbing radiation of frequency vE as a result of which the process is reversed. The heat evolved & on passing from A to B is then given by the expression where N is the Avogadro number and & is referred to one gram-molecule of .4 transformed. Q = Nh(v - V A ) LEWIS STUDIES IN CATALYSIS. PART XII. 1363 Let us consider the special case in which vA=vB or approxim-ately so.A catalyst acting as a transformer will in this case be unable to distinguish between the two types of molecules A and B since each is capable of absorbing the same type or approximately the same tspe of radiation. It will therefore catalyse both the direct and the reverse reaction equally. That is the opposing velocity constants will be equally increased and the equilibrium constant will remain unaffected by the presence of the catalyst. This result is in harmony with the criterion. If on the other hand the heat of the process is considerable, that is v A differs considerably from vB then it no longer follows that a positive catalyst will equally accelerate both reactions. In general it would not be expected to do so and consequently in general the equilibrium point will be affected by the catalyst.It is a significant fact in view of the conclusion just drawn that those reactions such as esterification or hydrolysis in which the equilibrium point is not sensibly affected by the catalyst are precisely those in which the heat effect is small. In the above case we have been considering mainly homogeneous catalysis by means of ions. Let us now take the case of catalysis by the undissociated molecule such as the molecule of hydrochloric acid which is generally regarded as functioning through the form-ation of an intermediate ternary compound. Thus Falk and Nelson ( J . Amer. Chem. SOC. 1915 37 1732) represent the inter-mediate oxonium complex in the case of hydrolysis of esters aa (ester,HCI,H,O).I n the reverse process the corresponding com-pound is (carboxylic acid,HCl,alcohol). These two compounds are tautomeric and may be identical. I f they are identical as Falk and Nelson assume then the hydrochloric acid molecule will equally affect the direct and the reverse process and thus leave the equilibrium point unchanged. This explanation of the mechanism of the effect produced by the undissociated molecule of the cata-lysing acid has certainly the advantage of simplicity. It has this implication however. Such additive compounds are generally formed rapidly compared with the rate of any further decomposi-tion which they may undergo If this is so and if the same inter-mediate compound is formed in the hydrolysis as in the esterifica-tion it would follow that the velocity constants should be the same, and the equilibrium constant should therefore be unity.This is not in agreement with experiment although it is significant that the value of K is not greatly removed from unity. Thus experi-ment has shown that In this case Q is zero or approximately so. [methyl acetate] x [water1 [methyl alcoholl x [acetic acid] = 4.6 R 1364 LEWIS STUDIES IN CATALYSIS PART XII. (compare Part V of this series of papers T. 1916 189, 71). This ratio means that the velocity constant of es'teriflcation is 4-6 times the velocity constant of hydrolysis. On the radiation view this ratio is mainly determined by the relative value of the exponential terms that is by where Nhvl is the critical increment of esterification and Nhv is the critical increment of hydrolysis that is, ~ M ~ Y z - v1 )IB T = 4.6.A t 300° absolute we find therefore that vZ-vl=1 x 1013. Both v1 and v do not however lie very far from the value 2 x 1014 so that the difference in respect of position of absorption of infra-red radiation is extremely small being of the order of one-twentieth of the absolute value of either frequency. The same idea is con-veyed by saying that the heat effect does not exceed 1000 calories. The fact therefore that the equilibrium constant possesses a value not unity but not far removed therefrom means on the radiation basis that the intermediate compounds are not identical but tauto-meric and further that both kinds of molecule absorb almost the same frequency so that any change in the equilibrium constant introduced by altering the concentration of the catR1yst is insensible.( I t may be noted that the relatively large change in R observed by Lapworth has its origin as Lapworth has shown, in what is virtually a distinct reaction not directly connected with the actual esterification-hydrolysis process itself .) I n the case con-sidered we conclude therefore that as a practical guide the criterion (3) or (4) is true that is in those cases in which the equilibrium constant is not far removed from unity. The con-clusion is obviously comparable with that drawn in connexion with ion catalysis. In addition to honiogeneous catalysis by dissolved substances in a given solvent it is well known that different solvents themselves exert their own catalytic effect.From the point of view of radia-tion we conclude that in this type of catalysis criterion (3) or ( 4 ) cannot in general be even approximately true since each solvent is characterised by its own electromagnetic properties that is by its power of absorption at different wave-lengths which differs from solvent to solvent and consequently entails a different dis-tribution of radiation density. The particular type of radiation required by the reactant is therefore present to a different extent, according to the nature of the solvent and consequently th LEWIS STUDIES IN CATALYSIS. PdRT XII. 1365 velocity constant and equilibrium constant is a function of the solvent. Specific differences are further introduced by the mutual interaction of solvent and solute whereby the effective frequency itself is altered to a slight extent.Finally as regards heterogeneous catalysis evidence has been collected and presented by Bancroft ( J . Physical Chem. 1917 21, 573; 1918 22 433) to show that the catalyst affeda the equil-ibrium point of the process. Heterogeneous catalysis has been examined in a preliminary manner from the point of view of radiation.(T. 1919 115 182) and it is concluded that the equil-ibrium point must be a function of the nature and extent of the catalytic material owing t.0 the alteration in the values of the critical incremenh which is introduced by the presence of the catalyst. The catalytic layer here considered is only one molecule, or possibly two molecules in thickness that is it is of the order 10-8 cm.In this layer the final amounts of reactants and resultants will differ in general from the true equilibrium amounts characteristic of the homogeneous gas phase for in the adsorption layer the relative amounts are determined by the relative adsorp-tion capacities. If this is a complete statement of the phenomenon, i t is evident that criterion (3) or (4) is inapplicable. There is, however a further possibility to be considered. The adsorbed reactants and resultants are in an activated con-dition as long as they are actually in the adsorption layer. I n time they necessarily pass out into the homogeneous gas phase, owing to desorption. If these molecules a t the moment of leav-ing the adsorption layer lose the extra energy which they possess and become immediately transformed into molecules of normal energy content it follows that their relative concentration in the homogeneous gas phase becomes identical with that in the adsorp-tion layer for in general the reaction in the homogeneous phase is extremely slow.There is the possibility however that the activated reactants and resultants on leaving the true adsorption layer do not immediately revert to the inactive state but may retain their activity for a short space corresponding perhaps with a layer 10-6 cm. in thickness. If this is the case there will be a rapid chemical change in this extra-adsorption layer which will tend t;o bring the concentrations of the reactants and resultants into the ratio required by the law of mass action for the homo-geneous phase.Whether the true equilibrium point would be attained or not would depend on the average life of the activated molecules. A t moderately high temperatures the average mole-cular velocity may be taken to be 105 cm. per second; hence it would require 10-11 second for a molecule to traverse a distanc 1366 LUMSDEN CRITERIA OF THE DEGREE OF of 10-6 cm. It is possible that the activate’d state may be main-tained for a longer period of time than this and therefore the more likely is the process to attain the true equilibrium position. The more selective the nature of the adsorption material &he further in general will the “ equilibrium ” of the adsorption layer depart from the true mass action equilibrium. Hence even if such a compensating effect as that suggested above actually operates, criterion (3) or (4) cannot be regarded as valid. It is concluded therefore that criterion (3) or (4) is in general, not true; in homogeneous systems it approximates more closely to experiment the smaller the heat effect accompanying the reaction ; in heterogeneous systems it is not certain whether even this approximation to validity holds good. M u s ~ u r r LABORATORY OF PHYSICAL DD ELECWRO-UHEMISTRY, UNIVERSITY OF LIVERPOOL. [Receiued October 31at 1919.