CATALYSIS. The term catalytic agent was introduced by J. J. Berzelius into chemistry to include those substances which by their presence accelerate the rate of reactions proceed ing with a decrease in free energy towards equilibrium. In gen eral, the catalyst provides an alternative path for the reaction; thus the chemical union of two components to form a product, Ad-B AB, may proceed directly or via a series of reactions, such as A+X–sAX, AX+B–>AXB, AXB--AB+X, in which the substance X is a catalyst. In reversible systems catalysts must necessarily accelerate both forward and reverse reac tions equally, since the equilibrium point cannot be affected by a material which acts solely as a catalytic agent. Catalytically accelerated reactions may be either homogeneous or heterogeneous in character, the former taking place in gases and liquids, the latter at interfaces, more especially at gas–solid and liquid–solid interfaces.
Catalysts both in homogeneous and heterogeneous systems the oretically possess an indefinite life, but in practice loss generally occurs, (a) through side reactions unconnected with the main reaction, as in the gradual reduction of sulphuric acid in the etherification processes; (b) through the formation of inactive compounds with impurities present in the reactants, as in the formation of nickel sulphide from sulphur compounds present as impurity in hydrogen employed in hydrogenation (q.v.) proc esses; (c) through removal from the reacting system either by vaporization, as in the case of catalysis of platinum in the for mation of sulphur trioxide, or by coagulation, such as occurs in colloidal metal catalysts employed for hydrogenation; (d) through sintering and fusion due to overheating of the active surfaces.
Catalysis in Gas Reactions.—Simple collision between the molecules of two reacting gases is not sufficient to ensure reaction. In general but a small number of such collisions are effective in promoting chemical reaction, the number of these effective col lisions rising, in the case of bimolecular reactions, exponentially with the temperature. It was first suggested by Arrhenius that effective collisions were those which took place between excited or active molecules, whilst normal molecules did not so react. In the majority of cases it is sufficient to assume that excited or active molecules are produced not prior to but in the act of col lision itself. It is yet unknown whether the excited molecule resulting in collision possesses a transitory existence before reac tion, or whether excitation and reaction are simultaneous proc esses. An activating collision results when the energy available for activation exceeds a critical value, the "critical energy incre ment." The source of this energy is, in simple molecules, the kinetic energy available on collision; in complex molecules it appears that part or all of the potential energy in the molecule can supplement the kinetic. The fraction of the total number of collisions effective will be dependent both on the complexity of the molecule and the magnitude of the critical energy increment. In troduction of a catalyst may augment the velocity of a reaction by effecting a two-stage reaction with two small critical energy increments, as alternative to a one-stage reaction, with a large value for the critical increment. Although the catalytic influence of small quantities of water on the rate of many chemical reac tions (e.g., certain combinations of sulphur dioxide; or the thermal combination of hydrogen and chlorine) has long been established, the mechanism of any one of these reactions is not sufficiently well known to establish beyond doubt their homogeneous nature.
We have noted that only certain high-speed collisions between normal molecules effect chemical reaction, but similar results may be obtained by slow-speed collision provided that one of the colliding molecules already possesses sufficient potential energy available on collision to effect the required chemical change. These potentially active or excited molecules may be generated from the normal species by photochemical means, and thus they serve as catalysts or photosensitizing agents in what are primarily photochemical reactions. The case of the catalysis of the union of hydrogen and oxygen by light of wave length X = may be cited. These gases do not combine with appreciable speed at ordinary temperature when exposed to radiation of this fre quency. On the addition of a small quantity of mercury vapour, however, rapid combination is effected. The energy of activation for the hydrogen-oxygen combination is supplied, as shown by Franck and Cario, by collision with mercury atoms which by the absorption of radiation have been converted into excited mercury atoms. Numerous similar gas reactions may be carried out with similar photosensitizing catalysts which include, not only mer cury in effecting hydrogenation of oxygen, carbon monoxide and ethylene or the decomposition of ammonia, but also bromine and chlorine in the decomposition of ozone. In considering the equilibrium point attained when such photocatalysts are em ployed, not only the active masses of the reactants, but likewise the radiation density must be taken into consideration.
In the recombination of atomic hydrogen to the molecular state it is evident that the resultant molecule is not stable unless the energy of dissociation can be liberated on association. If the re sultant molecule possesses an electric moment, this energy which must be liberated can escape in the form of radiation, but for molecules possessing small or no electric moments such radiation cannot take place, and thus atomic recombination does not occur. If, however, molecules are present, intermolecular collisions be tween two atoms and a molecule may result in the transfer of part or all of the superfluous energy to the molecule, and the atoms can combine the molecules thus acting in a catalytic man ner. The recombination of bromine atoms is apparently facili tated by the presence of bromine molecules.
Finally mention may be made of the catalytic influence of ions in chemical reactions. In the case of polymerization of acetylene, the energy of activation may be supplied by the recombination of ions of other and indifferent gases. Around each ion both positive and negative clusters of acetylene molecules are formed, and on reaction of these ions with each other the acetylene clusters absorb the energy set free on recombination and are converted into the polymeride.
Type i. The catalytic agent is a simple ion causing reaction in a neutral molecule in the presence of a large excess of the other reactant. This type includes most of the catalytic reactions in volving oxonium and hydroxyl ions, such as the mutarotation of glucose, ester hydrolysis, the decomposition of diazoacetic ester and nitrosotri-acetoneamine, as well as the decomposition of nitramide by salts and acids. If in ester hydrolysis the con centration of the ester be represented by [E] and that of the cata lytic oxonium ion by we obtain for the reaction velocity the expression where fo, fl, f2, etc., are the activity coefficients for neutral molecules, univalent and divalent ions, respectively.
In some cases the rate of reaction is dependent on the rate of diffusion of the reactants to, or of the products from, the surface; this is more frequently the case in reactions which involve either the interaction of a gas and a liquid at a solid surface or combus tive reactions taking place at relatively high temperatures; but in the majority of cases of low temperature catalysis the reaction proceeds sufficiently slowly to ensure attainment of equilibrium between the adsorbed gas and the gas in' close proximity to the catalyst. For such reactions several different types of surface action can readily be identified. In a number of cases, such as the decomposition of alcohols at a copper surface or the hydrol ysis of sugars by certain enzymes, the reaction rate is indepen dent of the concentration of the reactant; we may regard the catalyst surface as being saturated with reactant over the whole range of concentrations, and the rate of reaction measured is consequently merely the rate of liberation of the products of reaction from the surface. In some cases, e.g., the decomposition of ozone at a silver surface, every molecule striking the surface appears to undergo decomposition, and the rate of reaction is consequently strictly proportional to the partial pressure of the reactant. It is evident that for low concentrations of reactant or an inactive catalyst, the first case, viz., a zero-order reaction may be transformed into one of this second or unimolecular type.
For bimolecular reactions, e.g., the hydrogenation of ethylene at metal surfaces or the oxidation of oxalic acid in solution at carbon surfaces, the reaction velocity is a function of the con centrations of both of the reactants and attains a maximum at some suitable ratio of the reactant concentrations. At this point the reactants are adsorbed in the optimum ratio for interaction. Occasionally the reaction products inhibit the reaction ; thus the sulphur trioxide formed in the catalytic union of sulphur dioxide and oxygen, and the oxygen set free in the catalytic decomposition of nitrous oxide hinder the reactions by which they are formed. In these cases the reaction products do not volatilize immediately on formation from the surface but are relatively strongly adsorbed, so that, in effect, the area of surface available for reaction is reduced.
The quantitative study of poisoning action has revealed the fact that, for low temperature catalysis at least, only a small quantity of poison is necessary to effect complete destruction of the catalytic activity. This result has led to the view that only a small fraction of the surface is catalytically active, or that the seats of catalytic activity are active parts small in area and distributed irregularly over the surface of the catalyst. These active patches likewise appear to be differentiated from the remainder of the catalyst surface in other ways ; thus, they are much more susceptible to the influence of temperature which effects sintering of the material, and on adsorption of many gases the heat evolution is markedly different from that occurring on adsorption on the remainder of the material. At high temperatures the differentiation between the active patches and the rest of the surface is not so marked, and the whole surface appears to be nearly uniform in catalytic activity.
The efficiency of a given weight of catalyst is dependent on the specific surface of the active patches at low temperatures and on the total surface at high temperatures. Such a desired increase in surface may be effected by utilization of catalyst supports such as magnesium sulphate or asbestos (see SULPHURIC ACID) for the distribution on their surfaces of catalysts such as platinum or vanadium pentoxide. Nickel oxide can be employed as a support for nickel employed in processes of hydrogenation.
It has been noted that an increase in catalytic activity for a certain weight of material may be obtained, not only by effecting an increase in the specific surface, but also by incorporating into the material other substances termed promoters. The catalytic efficiency is found to be greater for such promoted catalysts than for either constituent singly. In industrial practice the use of promoted catalysts is now general, e.g., iron-potash-alumina mix tures for the ammonia synthesis, iron and chromium oxides for obtaining hydrogen from water gas, and zinc and chromium for the synthesis of methanol. (See METHYL ALCOHOL.) Two views have been advanced as to the function of promoters. By some, it is considered that the reaction accelerated by the promoted reaction is in reality a two-stage reaction in which the velocity of one stage are greatest on one constituent and that of the second stage greatest on the other constituent. In the hydrogenation of carbon monoxide one stage may be regarded as an association of carbon monoxide and hydrogen to form a methanolic complex, and the second stage might be considered as a dehydration of this complex. Nickel promoted with dehydrating catalysts is particularly effective for this reaction. The alternative view is based upon early observations of Faraday that perfect crystals of a hydrated salt do not effloresce in a dry atmosphere, but on scratching, efflorescence spreads from the patch throughout the crystal. The molecules at phase interfaces appear to exist in a particularly labile condition. This observation can be readily confirmed in the auto-accelerated decomposition of many solids, and in the reduction of copper oxide by hydrogen, which only takes place at the boundary between the oxide and the metal. Promoted catalysts are characterized by the presence of numerous discontinuities in the composition of the surface phase, and it is to be anticipated that, if surface actions take place more readily at these points, it is possible that catalytic actions on these dis continuities may likewise proceed more easily. The mechanism by which such reactions take place on surfaces is unknown ; there is ample evidence that in complex organic molecules adherence to the surface is effected by certain portions of the molecule, thus primary fatty alcohols adhere to a copper surface by means of their — CH_,OH group and it is in this group that reaction takes place.
Heterogeneous catalysts include the colloidal catalysts, both organic (the enzymes) and inorganic (colloidal platinum and palladium) . The former are remarkable in that some of them are almost specific in their action, whilst both are characterized by possessing a specific reacting surface which is liable to great modi fications with small changes in environment. We find also, as is the case in the decomposition of hydrogen peroxide or in the oxi dation of benzaldehyde, that the walls of the containing vessel, as well as particles of dust in the liquids, may readily provide surfaces which possess catalytically active portions.