STUDY OF THE ACTIVITY OF SULFIDE ZINC–COPPER METHANOL CATALYSTS

N. P. Kurin and E. N. Streletskaya

Communication III

In previous communications I¹) and II²), data were presented on the catalytic activity of a system consisting, before reduction, of zinc and copper sulfides. The activity, characterized by the degree of decomposition of methanol vapors, was studied as a function of the CuS content in the catalyst before reduction (from 0 to 70% by weight), the space velocity (from 420 to 2100), and the temperature (250–425°C). As a result, it was established that, under the conditions studied, the highest activity of this catalytic system is exhibited by catalysts containing, before reduction, from 1 to 4% copper sulfide. In addition, it was shown that the change in activity does not obey the rule of additivity and has a complex character. However, when the CuS content is increased from 50 to 70%, the activity of the system steadily decreases. In view of the foregoing, there naturally arises the need to investigate this catalytic system at still higher CuS concentrations (from 70 to 100%). For this purpose, we prepared and tested catalysts with 90 and 95% CuS, according to the procedure and on the apparatus described in Communications I and II.

In addition, the influence of the method of preparation on the activity of sulfide zinc–copper methanol catalysts was also studied. The point is that all our previous studies were carried out with catalysts prepared by the method of mechanical mixing of separately precipitated zinc and copper sulfides. It is known³) that, for a number of processes, catalysts obtained by coprecipitation of the components, with the same chemical composition, proved to be more active in comparison with catalysts prepared by mechanical mixing.

Furthermore, many authors have shown⁴) that the activity of precipitated oxide catalysts depends to a considerable degree on the concentration of the initial solutions.

In connection with the above, three catalysts of identical chemical composition \((\mathrm{ZnS} + 4\%\, \mathrm{CuS})\), corresponding to the most active catalyst obtained by mechanical mixing of the components, were prepared by the method of coprecipitation and tested according to the previously established procedure. Coprecipitation of zinc and copper sulfides was carried out from 0.5, 1.0, and 2.0 normal solutions with respect to zinc nitrate. For this purpose, ammonium acetate was added to an aqueous solution having the specified normality with respect to zinc nitrate (e.g., 0.5, etc.) and containing the necessary amount of copper nitrate, and then a stream of hydrogen sulfide was passed through until the precipitation of zinc and copper sulfides ceased.

It was observed that, when hydrogen sulfide was passed through the solution, copper sulfide was precipitated first, which, with insufficient stirring,

with stirring, settles to the bottom, and only then does zinc sulfide separate out. To prevent the copper sulfide from settling to the bottom and, thereby, to ensure uniform distribution of the latter in the zinc sulfide, an intense stream of hydrogen sulfide was passed through the solution, which also stirred the solution. The temperature during precipitation, as in the preceding experiments, was maintained at room temperature. The precipitates obtained by coprecipitation settled and filtered more rapidly than was the case with separate precipitation, which indicates their lower dispersion. Further treatment of the precipitates was carried out as described in Communication I.

RESULTS OBTAINED AND THEIR DISCUSSION.

Let us first consider the results relating to catalysts obtained by mixing zinc and copper sulfides.

The reduction time for catalysts with 90 and 95% CuS was, respectively, 31 and 25 hours, with the final reduction temperature being raised to 450°C. The data presented in Table 1 indicate

Table 1.

Percent decomposition of methanol as a function of temperature and space velocity in the presence of a catalyst whose initial form consisted of:

10% ZnS + 90% CuS

the very low activity of the catalyst which, before reduction, contained 90% CuS. At a temperature of 450°C and the lowest space velocity used by us, 420, the percent decomposition of methanol is only 3.11%. At lower temperatures, decomposition of methanol vapor with formation of noncondensable gaseous products was not observed. The catalyst which, before reduction, contained 95% CuS proved inactive, since in its presence methanol did not decompose even at 450°C. The composition of the gaseous products of methanol decomposition (Table 2),

Table 2.

Composition of the gaseous products of methanol decomposition in the presence of a catalyst which, before reduction, contained 10% ZnS + 90% CuS.

obtained in the presence of a catalyst with 90% CuS, is characterized by a reduced content of carbon monoxide (from 17.6 to 22.3%), a high content of methane, and considerable amounts of carbon dioxide and organic products absorbed by sulfuric acid. The ratio $\frac{H_2}{CO}$ is consid-

differs appreciably from the stoichiometric ratio (2.93—4.0 instead of 2). All this indicates that, in addition to the main reactions determining the course of decomposition according to the overall stoichiometric equation
$\mathrm{CH_3OH \rightarrow CO + 2H_2}$, side reactions also proceed simultaneously, leading to the formation of $\mathrm{CO_2}$, $\mathrm{CH_4}$, etc.

Summarizing the results of the study of the catalytic activity of the system which, before reduction, consists of zinc and copper sulfides and was obtained by the method of mechanical mixing of separately precipitated ZnS and CuS, it should be pointed out that, among all the catalysts studied and described in Communications I, II, and III, the most active proved to be those which, before reduction, contained from 1 to 4% CuS (see Figs. 1, 2, and 3); these should, first of all, be regarded as the most promising for tests under pressure and for industrial use.

The results of the activity study of jointly precipitated sulfide catalysts are given in Table 3 and presented in Figs. 4, 5, 6, 7, 8, and 9.

Table 3.

Percentage decomposition of methanol in the presence of jointly precipitated zinc-copper sulfide catalysts. Catalyst composition before reduction: 96% ZnS + 4% CuS.

From Table 3 and Figs. 4—9 it is evident that, in the presence of jointly precipitated zinc-copper sulfide catalysts, the dependence of the percentage decomposition of methanol vapor on the space velocity and temperature has the same form as in the presence of catalysts obtained by mixing separately precipitated sulfides; i.e., with increasing space velocity the percentage decreases, while with increasing temperature it continuously increases. An exception is the catalyst precipitated from a 2-normal solution. In this case, an increase in the percentage decomposition at space velocities of 1200—2100 with increasing temperature occurs only up to 375°C; with a further increase in temperature to 400°C the percentage decomposition decreases, which is apparently associated with destruction of the active modifications of the catalyst as a result of sintering.

Fig. 1

Fig. 2

Fig. 3

Fig. 4. Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained by the method of coprecipitation of zinc and copper sulfides from a 0.5 normal solution.

Fig. 5. Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained by the method of coprecipitation of zinc and copper sulfides from a 0.5 normal solution.

Fig. 6. Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained by the method of coprecipitation of zinc and copper sulfides from a 1.0 normal solution.

Fig. 7. Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained by the method of coprecipitation of zinc and copper sulfides from a 1.0 normal solution.

Fig. 8. Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained by the method of coprecipitation of zinc and copper sulfides from a 2.0-normal solution.

Fig. 9. Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained by the method of coprecipitation of zinc and copper sulfides from a 2.0-normal solution.

Comparing the percentage decomposition of methanol vapor in the presence of catalysts jointly precipitated from 0.5 and 1.0 normal solutions, it may be noted that the absolute activity values of these catalysts differ little from one another.

Thus, increasing the concentration of the zinc nitrate solution from 0.5 to 1.0 normality does not change the activity of the jointly precipitated catalyst of composition 96% ZnS + 4% CuS, and a certain deterioration of the catalytic properties occurs only when the solution concentration is increased to 2.0 normal.

Comparing the activity of jointly precipitated sulfide zinc-copper catalysts (Table 3) with catalysts of the same initial chemical composition (96% ZnS + 4% CuS), but prepared by mixing separately precipitated zinc and copper sulfides (Communication I, Table 1), it can be seen that the catalysts prepared by the sulfide-mixing method proved to be more active.

It was of interest to trace the influence of joint precipitation and of the concentration of the initial solution on the composition of the gaseous products of methanol vapor decomposition. The results obtained in this direction are given in Table 4. From this table it is evident that the composition of the gaseous

Table 4.

Composition of the gaseous products of methanol decomposition as a function of temperature and space velocity in the presence of jointly precipitated catalysts containing, before reduction, 96% ZnS + 4% CuS.

products of the decomposition of methanol vapors depends little on the concentration of the initial zinc nitrate solution. The compositions of the gases obtained in the presence of catalysts precipitated from 0.5 and 1.0 normal solutions are practically identical and are characterized by a low content of organic substances and carbon dioxide, and also by a ratio \(\frac{H_2}{CO}\) that differs little from the stoichiometric value determined by the reaction \(\mathrm{CH_3OH \to 2H_2 + CO}\).

The composition of the gas in the presence of the catalyst precipitated from a 2-normal solution proved to be of somewhat poorer quality than was the case for the first two catalysts. In particular, for most conditions (temperatures and space velocities) an increased content of carbon dioxide was observed, as well as a greater deviation from the stoichiometric value—

Table 5.

Rate constants for the reaction of decomposition of methanol vapors in the presence of coprecipitated catalysts consisting, before reduction, of \(96\%\,\mathrm{ZnS} + 4\%\,\mathrm{CuS}\), calculated by equation (1).

the value of $\dfrac{\mathrm{H}_2}{\mathrm{CO}}$. The low methane content or its complete absence is noteworthy, and from this point of view the catalysts obtained by this method possess increased selectivity of action.

In view of the low activity of catalysts with a high percentage of CuS (90 and 95%), the data necessary for calculating the temperature dependence of the rate constants and the activation energies were not obtained for them. As for the jointly precipitated catalysts, as Table 5 shows, in their presence the decomposition of methanol vapors proceeds according to zero order and is satisfactorily expressed by the kinetic equation of N. P. Kurin (see Communications I and II):

\[ K = V(1+\alpha)-\frac{273}{T}\alpha . \tag{1} \]

The values of the activation energies calculated by the Arrhenius equation, by the method described in Communication I, are given in Table 6.

Table 6.

Values of the activation energy for catalysts obtained by joint precipitation
(initial composition ZnS + 4% CuS).

From Fig. 10 it is seen that the straight lines $\log K = f\left(\dfrac{1}{T}\right)$ have a sharp break; moreover, the activation energies for catalysts jointly precipitated from 0.5 and 1.0 normal solutions, in accordance with Fig. 10, for the regions of lower and higher temperatures proved to be respectively equal to 56700 and 27000 cal/mol, i.e., changing the concentration of the solution from 0.5 to 1.0 normality does not change the activation energy. Increasing the concentration of the solution to 2-normal leads to some increase in the activation energy at lower temperatures (60000 cal/mol). In general, the activation energy of the jointly precipitated catalysts proved to be somewhat higher in comparison with analogous values for the majority of catalysts obtained by mixing separately precipitated ZnS and CuS. The zero order of decomposition and the presence of an upward-directed break in the straight line

\[ \log K = f\left(\frac{1}{T}\right) \]

show that in the presence of jointly precipitated catalysts, as in the case of catalysts obtained by mixing, the mechanism of methanol decomposition consists of two

Fig. 10.

consecutive reactions, in one of which methanol adsorbed on the surface of the catalyst decomposes with the liberation of formaldehyde and hydrogen, and in the second reaction formaldehyde decomposes into hydrogen and carbon monoxide (see Communications I and II).

As was already noted in Communications I and II, there is no simple relationship between the initial composition of the zinc–copper sulfide catalysts and their catalytic activity, for example of the type determined by the additivity rule. An analogous picture is also observed when comparing the activity, measured by the percent decomposition, with the activation energy. Indeed, comparing the percentage decomposition of methanol (Figs. 1, 2, and 3) with the corresponding activation energies (Table 7) for some cases, for example, an increase in the activation energy

Table 7*).

Values of the activation energies of the process of decomposition of methanol vapors in the presence of zinc–copper catalysts.

leads to a decrease in activity; however, in other cases a change in the value of \(E\) is accompanied by results opposite to those expected, i.e., instead of an increase or decrease in the rate of decomposition, respectively, its slowing or acceleration is observed.

As a probable explanation of this phenomenon, the following may be given. As is known, the formal relationship of the rate constant of a chemical reaction to the activation energy is expressed by equation (2)

\[ K = \mathrm{Const.}\, e^{-\frac{E}{RT}} . \tag{2} \]

For monomolecular homogeneous gas reactions, Const. is a temperature-independent factor which, according to Polanyi and Wigner\(^5\), in most cases corresponds in order of magnitude to an intramolecular frequency. It may be assumed that the laws governing the rate of chemical reactions in three-dimensional gases are also valid for two-dimensional gases, which are films adsorbed on the solid surface of a catalyst, in which molecules can move only in two directions. As was already indicated above, the decomposition of methanol vapors in the presence of the catalysts investigated by us is a zero-order process, characterized—

* The table is borrowed from Communications I and II.

...undergo complete coverage by the reacting substance of the active contact surface; thus, only actively adsorbed molecules can decompose. The true process of decomposition of methanol vapors in the adsorption layer consists of two successive monomolecular reactions:

\[ \mathrm{CH_3OH \to CH_2O + H_2,} \]

\[ \mathrm{CH_2O \to CO + H_2,} \]

where the first of them is the slower one and therefore determines the overall rate of the process at elevated temperatures, while the second does so at lower temperatures.

The value of Const in equation (2) for the indicated reactions, proceeding in a two-dimensional gas on the surface of catalysts, may be represented as the product \(A \cdot \mathrm{const}\), where \(\mathrm{const}\) is analogous to the value Const in equation (2), but referred to a unit of active surface; the quantity \(A\) is nothing other than the total active surface possessed by one bulk volume of the crushed catalyst. Thus,

\[ K = A \cdot \mathrm{const}\ e^{-\frac{E}{RT}} . \tag{3} \]

In this connection, the discrepancy noted above between the rate of decomposition of methanol vapors in the presence of zinc-copper catalysts and the activation energy can be explained by the fact that the value \(A \cdot \mathrm{const}\) in equation (3) may vary so greatly that an increase in the rate of the decomposition process occurring, for example, as a result of a decrease in the activation energy will be completely offset or even surpassed. If, by analogy with a monomolecular reaction in the gas phase, the value of \(\mathrm{const}\) in equation (3) is taken as independent of temperature and equal for all the zinc-copper catalysts we studied, despite the different energetic value (measured by the activation energy) of their active surface, then, knowing for each of them the activation energy and the rate constant \(K\), one can calculate the degree of development of the active surface of one catalyst relative to another. For this purpose, the expression for the temperature dependence of the rate constant for the catalyst taken as the standard will be written with indices (1):

\[ K_1 = A_1 \mathrm{const}\ e^{-\frac{E_1}{RT}}, \tag{4} \]

and the analogous function for any other catalyst compared with the standard with indices (2),

\[ K_2 = A_2 \mathrm{const}\ e^{-\frac{E_2}{RT}} . \tag{5} \]

Dividing equation (5) by equation (4), we obtain

\[ \frac{K_2}{K_1} = \frac{A_2}{A_1}\ e^{\frac{E_1-E_2}{RT}}, \]

whence the degree of development of the active surface of catalyst II relative to catalyst I will be equal to

\[ \frac{A_2}{A_1} = \frac{K_2}{K_1}\ e^{\frac{E_2-E_1}{RT}} . \tag{6} \]

The values of \(\frac{A_2}{A_1}\) calculated from equation (6) are given in Table 8. In the calculations for the region of reduced temperatures (250–325°C), the active surface of the catalyst consisting of ...

Table 8.

Values of the quantity $\dfrac{A_2}{A_1}$

reduction from pure ZnS; for the region of elevated temperatures, the catalyst with the initial composition ZnS + 25% CuS was adopted as the standard. As Table 8 shows, if the assumption made concerning the constancy of the quantity const in Eq. (3) is correct, the introduction into the catalyst of various amounts of copper sulfide at temperatures of 250–325°C in most cases increases the active surface of the mixed catalyst in comparison with the active surface of the catalyst obtained from pure ZnS. For the most active catalysts, with a content in their initial form of from 1 to 4% CuS, the increase in active surface in comparison with the standard catalyst is from 150 to 600 times; however, for some less active catalysts (with 35 and 30% CuS) this increase reaches 1,29·10⁶–3,25·10⁸ times.

As regards elevated temperatures (350–400°C), at which the limiting reaction is $\mathrm{CH_3OH \rightarrow CH_2O + H_2}$, an increase in the amount of copper sulfide above 25% in the initial form of the catalyst leads first to a slow, and then to a rapid, decrease in the active surface. Thus, from all that has been set forth above, it is evident that the reaction surface of reduced zinc–copper catalysts differs, depending on the initial composition, both in quality (measured by the activation energy) and in the quantity of active sites (centers), or active surface. The reasons for this phenomenon lie in the fact that the initial, amorphous, finely dispersed sulfide form itself is not a catalyst, and during the period of reduction, under the influence of methanol vapor, carbon monoxide, and hydrogen, it is destroyed with the removal of sulfur and the formation of a new catalytically active crystalline system consisting of zinc oxide, metallic zinc and copper, and possibly also solid solutions.

Since, depending on the CuS content in the mixed catalyst, the conditions of formation and shaping of the catalyst and of its active surface change, differences should therefore also be expected in the quality and quantity of active centers, and this must be connected primarily with the degree of dispersion. However, one should not expect a direct dependence between the degree of dispersion and activity. As a number of investigators have shown⁶), an increase in the degree of dispersion leads to an increase in activity only up to a certain limit, beyond which the activity begins to decrease.

In this connection, the presence of minima and maxima in the activity of the catalytic system we studied upon changing its initial composition can be explained by the possibility of the formation of different degrees of dispersion, lying both before and after the optimum regions, for the catalytically active components of which the reduced sulfide zinc–copper catalysts consist, and by the unavoidable superposition of their activities that occurs in this case.

It should be noted that, from the formal standpoint, the absence of parallelism between changes in the rate of methanol decomposition and in the activation energy can also be explained in another way. Namely, it is sufficient for this purpose to assume that in equation (3) \(A\) remains constant, while the value of \(\mathrm{const}\) may change upon transition from one catalyst to another and thereby compensate for or outweigh the changes occurring in the decomposition rate as a result of the simultaneous change in activation energy. In addition to the two limiting cases indicated, an intermediate case is also possible, when in equation (3), upon transition from one catalyst to another, both \(A\) and \(\mathrm{const}\) change simultaneously. In the latter two cases the data presented in Table 8 will correspond to the values

\[ \frac{\mathrm{const}_2}{\mathrm{const}_1} \quad \text{and} \quad \frac{A_2 \mathrm{const}_2}{A_1 \mathrm{const}_1}. \]

In this connection, carrying out X-ray investigations in order to determine the structure and degree of dispersion of the reduced zinc–copper catalysts would make it possible to verify the correctness of the considerations given above concerning the relation of the rate constant to the degree of dispersion and the activation energy for the catalytic system studied by us.

CONCLUSIONS

1. According to the procedure developed earlier (Communication I), the activity of sulfide zinc–copper methanol catalysts obtained by the method of mixing separately precipitated ZnS and CuS of the following initial compositions was studied: \(\mathrm{ZnS} + 90\%\, \mathrm{CuS}\) and \(\mathrm{ZnS} + 95\%\, \mathrm{CuS}\).

In addition, catalysts prepared by the method of coprecipitation of sulfides from aqueous solutions of zinc and copper nitrates of various concentrations were also tested.

Conditions of the investigation: temperature \(250—450^\circ\mathrm{C}\), space velocity \(420—2100\).

1. A catalyst prepared by the mixing method and containing, before reduction, \(90\%\, \mathrm{CuS}\) only at \(450^\circ\mathrm{C}\) began to decompose methanol with the formation of noncondensable gaseous products. The catalyst with initial composition \(\mathrm{ZnS} + 95\%\, \mathrm{CuS}\) proved to be completely inactive and did not decompose methanol even at \(450^\circ\mathrm{C}\).

2. In the presence of coprecipitated sulfide zinc–copper catalysts (composition before reduction \(96\%\, \mathrm{ZnS} + 4\%\, \mathrm{CuS}\)), as well as for catalysts obtained by mechanical mixing, with increasing space velocity the degree of decomposition decreases, while with increasing temperature it increases.

3. Raising the concentration of the solutions from \(0.5\) to \(1.0\) normality practically does not change the activity of the coprecipitated catalysts, and some deterioration of the catalytic properties occurs only when the concentration of the solution is increased to 2-normal.

4. The composition of the gaseous products of methanol-vapor decomposition depends little on the concentration of the initial solution and is characterized by a low content of organic substances and carbon dioxide, and by a low conten-

with, or complete absence of, methane and an \(H_2:CO\) ratio close to the stoichiometric one.

The decomposition of methanol vapor in the presence of jointly precipitated catalysts proceeds with an apparent zero order and is satisfactorily expressed by the kinetic equation of N. P. Kurin:

\[ K = V(1+\alpha)\frac{273}{T}\,\alpha . \]

The activation energies for catalysts prepared by the method of joint precipitation from 0.5- and 1-normal solutions, for the regions of lower (275–325°C) and higher (325–350°C) temperatures, proved to be respectively equal to 56,700 and 27,000 cal/mole; for a catalyst precipitated from a 2-normal solution, in the temperature range 300–350°C, \(E\) is equal to 60,000, and at 350–400°C to 20,800 cal/mole. The mechanism of methanol decomposition is the same as in the presence of catalysts prepared by the method of mechanical mixing.

1. The absence of a normal relationship between activation energy and activity for catalysts prepared by mixing separately precipitated zinc and copper sulfides is explained by differences in the degree of development of the reactive surface, which in turn is associated with the dispersity of the reduced catalysts. The appearance of maxima and minima on the curves of decomposition percentage \(= f(\%\,\mathrm{CuS})\) is also associated with this latter factor.

2. Summing up the study of the catalytic system consisting, before reduction, of mechanical mixtures of separately precipitated zinc and copper sulfides, it should be noted that, of all the catalysts described in Communications I, II, and III (from 100% ZnS to 100% CuS), the most active proved to be those which, before reduction, contained from 1 to 4% CuS; they should be regarded first of all as the most promising for testing under pressure and for industrial use.

LITERATURE

1. Kurin and Tantsov—Proceedings of the Tomsk Industrial Institute named after S. M. Kirov, vol. 59, III, 1941.

2. Kurin and Prorvin—Proceedings of the Tomsk Industrial Institute named after S. M. Kirov, vol. 59, III, 1941.

3. Frolich and Fenske, Ind. Eng. Chem. 20, 694, 1928.

4. Ipatiev and Dolgov—Journal of Chemical Industry 6, 825, 1931.
   Dolgov—Methods of Chemical Utilization of Carbon Monoxide, ONTI, 1936.

5. Polanyi and Wigner, Z. physik. Chem. (A) 139, 439, 1928.

6. Levi and Haardt, Atti R. Accad. dei Lincei Roma, 6, 3, 91 215; Gazz. chim. Ital., 56, 424, 1926.
   Dankov, Journal of Chemical Industry 4, 326, 1933.
   Dankov and Kochetkov—Reports of the Academy of Sciences of the USSR, 359, 1934.
   Rubinshtein, Acta Physicochimica URSS, 7, 101, 1937; Proceedings of the Academy of Sciences of the USSR, Chemical Series 4, 46, 1938.