524 



NA rUKIi 



[April 3, 1902 



When, however, we attempt to derive from it the slightest 

 tjuiclance towards experimental and exploratory work, or to 

 deduce an idea of the possible laws of catalytic actions— and 

 this is, indeed, the only purpose of such hypotheses — we are 

 convinced of its entire fruitlessness. That the hypothesis of 

 molecular vibrations drove the whole matter into a blind 

 alley is obvious from the fact that steady scientific inves- 

 tigation of the problem, previously prosecuted with much 

 zeal, did not follow. For quite a long lime only isolated 

 investigators concerned themselves with describing catalytic 

 phenomena. .Schonbein himself, to whose investigations we 

 are so much indebted for what we know of the facts, took no 

 part in the theoretical discussions as to the causes ; on the 

 contrary, it obviously gave him pleasure to investigate pheno- 

 mena for which contemporary chemistry, to which he paid 

 little heed, could find no explanation or place of refuge. 



It is possible to speak much more favourably of another view 

 which had long before been advocated, but meanwhile had been 

 neglected. This is the idea of intcnn(diate reactions. This 

 had its origin in the first scientific investigation of the chemical 

 processes that take place in the lead chamber in the manu- 

 fiiclure of sulphuric acid. Clement and Desormes, in the year 

 1806, in a classical research gave the explanation, still generally 

 accepted, of the action of oxides of nitrogen in the oxidation of 

 sulphurous acid by atmospheric oxygen. As everyone knows, this 

 rests upon the assumption that the sulphurous acid is oxidised by 

 higher oxides of nitrogen, which are thereby reduced to nitric 

 oxide. This then unites again with atmospheric oxygen, and 

 the process can begin anew. Thus a very small quantity of 

 oxides of nitrogen serves to oxidise a large quantity of sul- 

 phurous acid. It is remarkable that at the time of the discus- 

 sion between Berzelius and Liebig this case was not brought 

 forward, and only later do we find applications of this old view to 

 other cases where similar chemical processes are brought about. 

 Then, however, this mode of viewing the phenomenon became 

 more and more general, and to-day we must regard it as the 

 oldest and most important attempt to explain certain, though 

 not all, catalytic processes. As a rule, however, there certainly 

 exists a weakness in this view. In confronting a catalytic phe- 

 nomenon, we seek for the possible intermediate product in the for- 

 mation of which the catalyst could take part, and we consider the 

 problem essentially solved when we can fix upon such. If we 

 succeed, indeed, in getting some of the assumed intermediate 

 products from the materials, the view is considered to be proved. 

 Whether such a .substance is truly an intermediate product, and 

 not merely some b)e- product, is a question which is hardly raised 

 and still less answered. 



If we lest the idea from our present standpoint, we find at 

 first something contradictory in it. In order that a process can 

 occur at all, it must be associated with the diminution of free 

 energy. This loss depends only on the initial and final stage of 

 the reaction, not upon its course. On the other hand, the 

 velocity of the reaction in strictly comparable cases is propor- 

 tional to this loss. Hence we should be inclined to conclude 

 that the velocity of reaction in a given system must have the 

 .same value, whether the process is direct or indirect — whether 

 it takes place in one leap or in several steps. 



Such a conclusion would be incorrect, for besides the loss of 

 free energy, the velocity of reaction depends upon many other 

 factors, which are by no means all known. A well-known 

 example is the very great influence which temperature has, an 

 influence which counts much more than the corresponding 

 increase of free energy. Chemical energetics also teach 

 us that while definite generalisations may be made con- 

 cerning the equilibrium of a given system, this is not the case 

 with the actual velocity with which equilibrium is attained. 

 There is thus no contradiction of general laws, if we as.sume 

 that a certain course of reactii>n takes place more quickly through 

 an intermediate substance than it does directly. Nothing can 

 be said for, but something against, the view that this is generally 

 the case. 



Coming back now lo our classical example, which indeed in 

 other respects will be historical, we can assume that sulphurous 

 acid is oxidised more slowly by the oxygen of the air than by 

 the two reactions. Oxidation of sulphurous acid by nitrogen 

 peroxide and oxidation of nitric oxide by atmospheric oxygen 

 run concurrently, although the concentration of the intermediate 

 products must necessarily be smaller than the concentrations 

 which act in a direct reaction. Whilst, however, we may regard 

 this view as scientifically founded, there is still a chief thing 



NO. 1692, VOL. 65] 



wanted. The velocity of the reaction concerned must be 

 actually measured, and until this is done one can only speak of 

 a conjecture and not of an explanation, and what is said here is 

 true in general for the assumption of an intermediate reaction. 

 No catalytic acceleration is explained, unless it is also shown 

 that the intermediate reactions actually take place more rapidly 

 than the direct reactions, under the given conditions. Up to 

 now, no case of this kind has been satisfactorily investigated, 

 and no such explanation actually proved for a given case. I 

 certainly hope that this gap will not long remain, as investi- 

 gations directed to this point are approaching their conclusion. 



If we assume that in certain cases the correctness of the 

 theory of intermediate products is proved (which to all 

 appearances is the case), we have the further question, whether 

 all catalyses will find an explanation in this way. I believe this 

 question may certainly be answered in the negative. I believe 

 that there are a number of cases in which such an explanation 

 is not applicable. In particular, I see no possibility of explaining 

 the retarding catalytic influences by the assumption of inter- 

 mediate products, for if a reaction goes more slowly via the 

 intermediate products than in the direct path it will take the 

 latter, and the possibilityof intermediate products has no influence 

 on the process. 



Another theory of catalysis has been put forward by Euler. 

 Starting from the assumption that all chemical reactions are re- 

 actions of ions and that the velocity depends on the concentration 

 of active ions, he supposes that the catalytic substance has the 

 property of altering the concentration of the ions. In accordance 

 with this alteration in the concentration, the velocity of the re- 

 action must also alter. 



So far as I can see, such a theory is admissible, that is to say, 

 it would usually be possible to make the assumption without 

 contradicting the laws of general chemistry Whether or not 

 ultimately contradictions would appear, cannot be foreseen at 

 present. There appears to be onedifticulty in the fact, often ob- 

 served, that two catalysts acting together efl'ecl a much greater 

 acceleration than would be calculated from the sum of their 

 separate effects. It is not obvious how, by the .simultaneous 

 action of the two catalysts (for example, Cupriion atnl Ferri- 

 ion), so much larger quantities of reactive ions could be formed 

 than these can form acting apart. 



One may say of this theory, then, that it might explain some 

 catalyses, but by no means all. 



A more complicated case of catalytic phenomena is found in 

 processes where one of the substances taking part in the reaction 

 acts also in the capacity of catalyst. Among the possibilities 

 of this auto-catalysis, I will only mention the case of a reaction 

 producing an accelerator. This occurs, for example, in one of 

 the best-known reactions, the solution of metals in nitric acid. 

 The nitrous acid here formed increases in a high degree the 

 velocity of reaction of the nitric acid, whence arise the following 

 phenomena : — If the metal is brought into pure acid, the re- 

 .action begins with extreme slowness. In the same degree as it 

 progresses, it becomes more rapid and in the end tumultuous. 

 When this period is passed, the process slows down and ends 

 with a velocity converging on zero. 



This stands in striking contradiction to the usual course of 

 reactions, which begin with a maximum of velocity and, owing 

 to the gradual consumption of active substances, become con- 

 stantly slower. 



Physiological analogies present themselves irresistibly at this 

 point. We have here a typical fever phenomenon. Another 

 important physiological fact can be illustrated in the same way, 

 hahit and niemcry. I have here two specimens of the same 

 nitric acid, differing only in this, that in one a small piece of 

 copper has been already dissolved. I now bring two similar 

 pieces of copper foil into the two acids, which are at the same 

 temperature. Vou see that the acid which has already once 

 dissolved the copper has become "habituated" to the work 

 and begins instantly and vigorously to carry it out, while the 

 other unpractised acid does not know how to begin with the 

 copper, and sets about its work so clumsily and so slowly that 

 we will not wait for it. 



That we are here dealing with catalysis by means of nitrous 

 acid will be obvious when I add some sodium nitrite to the 

 dilatory acid ; the copper is now at once atl.acked and dissolved. 



(3) Ileterogeneous Catalysis. 



The best-known case of heterogeneous catalysis is the action 

 of platinum on combustible mixtures of gases. Whilst previously 

 the chief interest centred round mixtures of hydrogen and 



