Spot Reaction Experiments Part ZZZ:
Heterogeneous Reactions FRITZ FEIGL
Laboratoria Central da Produ~EoMineral, Ministerio da Agricultura, Rio d e Janeiro, Brazil
(Translated by Ralph E. Oesper, Uninersity of Cincinnati)
A.
NUMBER of experiments can be carried out to dlustrate heterogeneous reactions. 10. Condensation of Gases on the Surface of Finely Divided Platinum. Molecules on the free surface of a solid, that is on those parts that are in contact with gases or liquids, are in a different condition than those in the interior. The molecules in the interior are surrounded entirely by like molecules, either in disorder (amorphous materials), or in definite spatial configurations (crystalline bodies). In both cases the cohesion is due to the activity and saturation of the binding forces of the atoms. The cohesive forces of the molecules of the free surface are, however, satisfied only toward the interior. Forces directed toward the exterior are still available and can be satisfied by molecules of foreign materials, which are thus held more or less finnly. This adhesion is known as sorption. Gases, liquids, or dissolved materials can be held by sorption (absorption and adsorption). The molecules held by sorption are in a state of condensation as compared with their original distribution; sometimes they are also in a more reactive form. Both of these effects may be responsible for the fact that certain reactions of gases or dissolved materials, which normally proceed slowly, are more rapid if they occur on the surface of solids on which they are sorbed. If the products of these reactions do not remain a t the place of their origin, that is, on the surface, but break loose or are removed, the reaction continues by virtue of renewed
sorption. The free surface of a solid can thus accelerate a reaction without actually participating in it. This type of catalysis, which takes place on the interface of two phases, is known as heterogeneous catalysis. A very familiar instance of catalysis due to the condensation of gaseous reactants on a free surface is the behavior of h e l y divided platinum or palladium. The following experiment demonstrates that sufficient gas (illuminating gas-air, or hydrogen-air) condenses and reacts on the surface of platinum or palladium to bring the metal to glowing (because of the heat of reaction). Procedure: One drop of 0.3 per cent palladium or platinum chloride solution is placed on thin asbestos paper and then evaporated. A spot of finely divided platinum or palladium is left. While the paper is still warm, the spot is held in a stream of illuminating gas and soon begins to glow. Hydrogen will also condense on the metal and will react there with the oxygen of the air. The hydrogen can be drawn from a Kipp apparatus and directed through a glass tube against the spot of metal. Exceedingly small quantities of platinum or palladium (0.04 y Pt) can be detected by this catalytic action. 11. Reaction of Potassium Chromate and Manganese Diozide with Activated Hydrochloric Acid. In speaking of catalyzed reactions i t is common practice to state that a compound is "activated." This designation is not quite correct because in the majority of catalyzed
reactions both participants of a reaction, whose normal rate is slow, are activated by a catalyst, or by its insertion into the reaction mechanism. The term "activation" is just as inexact as calling a reaction an "oxidation" or "reduction." Actually it is both since no oxidation proceeds without concurrent reduction and vice versa. Therefore, if certain cases are spoken of as oxidations or reductions, the material whose oxidation or reduction seems more important is purposely emphasized. "Activation" is employed in this same way with respect to catalyzed reactions. The reactant whose heightened activity seems more important is thus purposely stressed. An example of this is presented in the following. An oxidizing action by potassium chromate or manganese dioxide is not surprising, since both these compounds are known to be strong oxidizing agents. On the other hand, a reduction of these by dilute hydrochloric acid is not normally expected. Therefore, if this reduction can be brought about with the aid of a catalyst there is justification for speaking of an "activation" of the hydrochloric acid, if the heightened activity of the acid seems worthy of emphasis. The following experiment demonstrates that this activation can be brought about by silver halides. This is an instance of heterogeneous catalysis. Concentrated hydrochloric acid reacts quickly with potassium chromate and manganese dioxide only if the temperature is raised. The equations for the reactions can be written:
+ +
2KaCrOd 16HC1= 2CrCb MnOz 4HC1 = MnCL
+ 4KClf 8Hx0 + 3Cb + 2H.O + CIS
Comparatively dilute hydrochloric acid reacts a t room temperature so slowly that even after several hours no perceptible reaction has occurred. However, if a suspension of MnOp in 3 N hydrochloric acid, or a solution of potassium chromate containing hydrochloric acid, is treated with a small quantity of a silver salt, a rapid readion sets in. This is evidenced by the discharge of the color in one case and by the change from orange to green in the other. Consequently, the silver salts function as catalysts.' Since chloride and silver ions produce slightly soluble silver chloride, it seems logical to regard silver chloride as the catalyst. In fact, the same acceleration can be brought about by adding previously precipitated silver chloride. It may be shown also that the still less soluble silver halides, such as AgBr, AgI, and AgCN, likewise are capable of this catalytic activity. Consequently, i t is justifiable to conclude that solid silver halides, in general, act as catalysts in these reactions, and that these are examples of heterogeneous catalysts. The mechanism of this effect obviously is that the silver halide adsorbs both hydrochloric acid and chromate (or chromic acid). Consequently the reactants are condensed or concentrated on the surface. LANG,2. anorg. Chem., 152,201 (1926); Bn.60,1359 (1927). BOBTELSKY, ET AL., Z. anorg. Chem., 189, 196 (1930); 205, 401 Bn.,65, 544 (1932); 209, 95 (1932). FEICL AND FRAENKEL, (1932).
The result is that the same conditions are then obtained as though chromic acid were brought into contact with concentrated hydrochloric acid.= It must be remembered that the condensation affects only the reactants that are adsorbed a t the particular time. After the reaction has occurred, the surface of the silver halide is available again for further adsorption-concentration and reaction. This sequence is repeated ,until all the chromate has been reduced. The action is quite analogous when a tine suspension of manganese dicxide is used in place of a solution of chromate. Procedure: A solution is prepared by mixing 1.5 ml. of 15 per cent potassium chromate solution with 3.5 ml. of concentrated hydrochloric acid. Three drops of this mixture are placed in each of two adjacent depressions of a spot plate. A drop of water is added to one and a drop of 0.01 per cent silver nitrate sohtion to the other. The silver-free solution remains unchanged, whereas the color of the other mixture rapidly turns brown and gradually becomes green. The experiment should be repeated, substituting minute quantities of washed, freshly precipitated AgC1, AgBr, and AgI for the silver nitrate. A colloidal solution of manganese dioxide may be used instead of the hydrochloric acid-chromate solution. I t is prepared by dissolving 0.6 g. of manganous sulfate in a mixture of 60 ml. water and 20 ml. concentrated hydrochloric acid. Twenty ml. of 0.1 N potassium permanganate solution are added and the mixture is shaken thoroughly. 12. Detection of the Beginning of Precipitation Reactions before the Precibitate Has Become Visible. If -insoluble compounds are formed in solutions as a result of chemical changes, the occurrence of the reaction is not perceived until the precipitate is actually seen. Precipitation reactions can be used for the detection or the determination of a material if a visible precipitate forms a t even low concentrations. Consequently, if the reaction A+ B- + AB, leads to the production of a slightly soluble material, AB, the solubility product of AB should determine the sensitivity of a test, or the accuracy of the gravimetric procedure or titrimetric precipitation method. Experience, however, has shown that the solubility product of the precipitate is by no means the sole determinant of the sensitivity of a precipitation reaction. Sometimes i t is easier to see the actual precipitation of a compound that has a larger solubility product and higher solubility than it is to detect the deposition of a compound that possesses a much smaller
+
The validity of the assumption that the reactants are concentrated on the surface of the silver halide is supported by the following experiment. Three drops of 10 per cent potassium dichromate solution are placed in a depression of a spot plate, four drops of concentrated sulfuric acid are added, and the mixture is stirred with a thin glass rod. A heap of silver iodide (the size of a pea) is added and the suspension allowed to stand without stirring. Almost instantly, the entire surface of the silver iodide becomes dark brown. Condensation of sulfuric acid and chromate has led to the deposition of CrOs. The color disappears if about 10 drops of water are stirred in; bright yellow silver iodide is left.
solubility product or a lower solubility. A case in point is provided by mercuric sulfide and lead sulfide. Although the solubility product of HgS = 2 X and of PbS = 2 X the precipitation of lead sulfide is more conspicuous and can be seen with smaller quantities of the metal. If, furthermore, the quantities of a material that can actually be detected in a given volume by means of a visible precipitate are compared with the quantities that might be expected to be visible on the basis of the insolubility of the compound, it will be found that these quantities by no means coincide. This is true of all precipitation reactions. The practical limits of identification represent quantities that are greater by several powers of ten than the theoretical quantities computed from the solubility data and the solubility product. The reason for this apparent incongruity is that the development of solid phases in a solution can be detected by unaided visual observation only after the quantity of precipitate exceeds a certain threshold value. The color of the solid phase also plays a role. But entirely aside from the physiological inadequacy of human sight, the visible formation of a solid phase is, of itself, only the end-product of chemical and physical processes that cannot be represented by the usual stoichiometric formulation of a chemical reaction. The general ionic reaction just given expresses only the fate of the ionic species A+ and B- up to the time of their union into the soluble single molecule AB. This however, is only the initial and shortest-lived stage of the total course of a precipitation reaction. I t is followed by a series of events, such as hydration or dehydration, polymerization of single molecules, aggregation of polymerates, passage through colloidal stages, supersaturation, and cxystallization. That which is observed as a precipitate is not even the last link of this chain. I t is well known that the properties of freshly precipitated products often change considerably when the precipitate ages. This alteration is due to further changes that occur even in the solid state. The nature and rate of the partial processes that follow the primary ionic reaction are influenced by the nature of the reactants, their concentration, the presence of other solutes, and the temperature. Consequently, it cannot be logically assumed that a precipitation reaction does not begin until the precipitate is actually seen. The visible deposition represents only a single stage of the aggregation that takes place relatively long after the beginning of the reaction that is stoichiometrically expressed by the chemical equation. I t can be proved that the visible production of a solid phase in a given reaction medium lags behind the primary chemical change. Reactions in which a gaseous product is formed along with a precipitate are used for this demonstration. If the gaseous reaction product is detected prior to the actual observation of the solid it follows that the reaction must have occurred before the precipitate becomes visible. A portion of the aggregation process that follows the
stoichiometrically formulated reaction occurs during the interval between the detection of the gas and the first sight of the solid. The action of acid on sodium thiosulfate, and on potassium nickel cyanide, are used here because they produce both a precipitate and a gas. The respective equations are: Na&08 2HC1= 2NaC1+ HnO+SO2 )S KzNi(CN)r
+ + 2HC1 = 2KC1+ 2HCN + Ni(CN)9
Under suitable conditions, the formation of SOz and HCN can be detected when there are as yet no signs of a turbidity due to the deposition of sulfur or nickel cyanide. This is possible despite the solubility of these gases in water, and even though the detection of the gases cannot be accomplished instantaneously, because the detection reaction necessarily requires a certain length of time for its execution. The detection of sulfur dioxide, following the acidification of a thiosulfate solution and before the visible separation of sulfur, deserves consideration for the following additional reason. It is usually assumed that when sodium thiosulfate is acidified the first product is free, unstable thiosulfuric acid, (I), which then gradually decomposes into sulfur and sulfur dioxide, (2) : Na&01
+ 2HC1 = 2NaC1+ H&Oa H&O, = HnO + SO* + S
(1) (2)
Accordingly, the interval during which an acidified solution of sodium thiosulfate remains clear has always been taken as the life span of the thiosulfuric acid. The present experiments show, however, that its lie, providing it has one, must, in any case, be shorter than this interval, because decomposition in the sense of equation (2) has already begun at the time of the detection of the sulfur dioxide, namely, before the visible separation of sulfur. The action of sulfur dioxide on a solution of ferric femcyanide is used to detect the gas. The redox reactions can be expressed: and 2Fe(CN).=
+ SO2 + 2H.O = 2Fe(CN)C + SO,- + 4H+
Consequently, Fe++ ions are formed and they, in turn, react with unchanged Fe(CN)b ions to produce Prussian blue. The other product, Fe(CN)B' ions, reacts with unchanged Fe+++ ions to form Turnbull's blue. The sensitivity of this test for sulfur dioxide is due, therefore, to the fact that both of the reduction products participate in the production of deep blue compounds. Some sodium bicarbonate is added to the thiosulfate solution in the present experiments; it evolves carbon dioxide when the acid is added and this aids in expelling the sulfur dioxide from the water. The hydrogen cyanide is detected through its reaction on a mixture of copper acetate and benzidine acetate; benzidine blue is formed. This oxidation
of benzidine is explained by assuming that prussic acid reacts with copper acetate: 2Cu++
+ 6CN-
=
+
CU~(CN)~-(CN)Z
Cyanogen has many of the characteristics of a free halogen, including oxidizing powers. Pro~edures: ( A ) Decomposition of Sodium Thiosulfate. The apparatus (Figure 1) used by the writer can be made from materials available in every laboratory. The bulb is filled to a height of about 2.5 cm.
with 0.01 N thiosulfate solution. Several granules of sodium bicarbonate are added and the mixture is acidified with a few drops of 6 N hydrochloric acid. The apparatus is closed immediately with the stopper, whose knob has been freed of grease and then dipped into ferric femcyanide solution. This reagent must be freshly prepared; it consists of 1 ml. of one per cent ferric chloride solution, 1 ml. of two per cent potassium ferricyanide solution, and 18 ml. of water. The closed apparatus is viewed against glossy black paper, which is held so that it extends up to the level of the liquid. After about 15 seconds the suspended drop of the yellow reagent solution becomes greenish and then gradually turns blue. The solution in the bulb is still perfectly clear and only after two minutes does a slight turbidity appear. I t gradually becomes more dense. Acco~d'ingly, ouly 15 seconds aftff acidifying the thiosulfate solution, sulfur dioxide is formed in sufficient quantities to be detected easily in the relatively large space above the liquid. The reaction mixture is still perfectly clear. This proves that the sulfur,
which is formed concurrently with the sulfur dioxide, passes through a stage where it is not visible. I t may be in true solution, or perhaps it forms a colloidal dispersion before it is transformed into a solid phase that can be seen. (b) Decomposition of Potassium Nickel Cyanide. A test tube, 5 cm. high and about 1 em. in diameter, is filled to within about 1.5 cm. from the top with four per cent solution of K2Ni(CN)+ The rest of the tube is then filled with 5 N acetic acid. A square of filter paper, on which a drop of copper-benzidine acetate solution has been placed, is immediately laid over the test tube. (The reagent, freshly prepared, is made by mixing equal parts of solutions I and 11. Solution I contains 0.29 g. of copper acetate in 100 ml. of water. Solution I1 consists of 47.5 ml. of a solution of benzidine acetate, saturated at room temperature, and then diluted to 100 ml.) By holding the tube properly, and observing from below, the formation of a blue color on the filter paper will be seen within several seconds. As soon as the color appears, the paper is removed and the clear solution is viewed against a dark background. A turbidity of nickel cyanide appears after about one and one-half minutes, and slowly increases in intensity. The further development of prussic acid during the precipitation of the nickel cyanide can be shown by means of a fresh piece of the reagent paper. Accordingly, prussic acid is formed while the reaction mixture is still perfectly clear, and despite its solnbility in water can be detected above the liquid. I t is thus proved that the nickel cyanide, formed simultaneously with the prussic acid, remains in true or colloidal solution for a considerable time before it can be detected visually. This experiment permits an interesting calculation of the threshold quantity of nickel cyanide that must be exceeded before a visible deposition occurs. The identification limit of prussic acid is known to be 0.25 y of HCN. Accepting this value, it can be calculated from the equation representing the decomposition of KaNi(CN)r by acid that two moles of HCN (54 g.) correspond to one mole of Ni(CN)n (111 g.). Consequently, at least 0.5 y of nickel cyanide must have been formed (although it was not visible) by the time 0.25 y of prussic acid was detected, that is, after three seconds. Nickel cyanide becomes visible ouly after 90 seconds. If, for the sake of simplicity, it is assumed that the rate at which the prussic acid forms remains constant, 15 y of nickel cyanide will be present when the threshold of visibility is reached after one and one-half minutes. If it is also considered that precipitated nickel cyanide is hydrated Ni(CN)%.7He0,twice this value is obtained for the weight of the precipitate actually involved.
