INVESTIGATION OF THE ACTIVITY OF SULFIDE ZINC–COPPER METHANOL CATALYSTS

N. P. Kurin and N. N. Tantsov.

Communication I.

As is known, the catalysts that have been subjected to the most comprehensive investigation are those consisting, in their initial form, of oxides of the elements; among them, zinc–chromium and copper catalysts have proved to be the most active and suitable for the purposes of industrial methanol synthesis. The former possess good selectivity of action and good activity only at high temperatures, while the latter are very active at reduced temperatures but are very sensitive to contact poisons (sulfur compounds) and to overheating. Thus, these two groups of oxide catalysts only partially satisfy the requirements imposed by industry, and therefore the problem of finding better methanol catalysts is highly topical.

The aim of the present work is to investigate the activity of zinc–copper sulfide catalysts. These catalysts may be expected to possess increased activity even at reduced temperatures, since ZnS and CuS, when precipitated in the cold, separate in the form of a highly dispersed precipitate which, upon subsequent reduction by methanol vapors or by $\mathrm{H_2}$ and $\mathrm{CO}$, should retain its dispersity, and this will determine the increased activity of the catalyst.

Furthermore, catalysts obtained from zinc and copper sulfides apparently should not be poisoned so strongly by sulfur compounds as is observed for oxide catalysts; and if poisoning does occur, it should to a considerable extent be reversible in character.

In the literature, apart from brief patent indications[^1], there is on this question only one work, by Dolgov, Karpinskii, and Silina[^2], who investigated the rate of decomposition of methanol in the presence of zinc sulfides and its mixture with 1% and 5% sulfides of copper, cadmium, molybdenum, and antimony at temperatures from 225 to 400°C and at a constant rate of passage of methanol vapors. The cited authors, in agreement with the premises set forth above, observed increased activity of catalysts whose initial form consisted of mixtures of sulfides of the enumerated metals.

In the present work the activities of catalysts were studied whose initial form consisted of zinc sulfide, copper sulfide, and mixtures of these substances in the following ratios: $\mathrm{ZnS} + 0.5\%\,\mathrm{CuS}$; $\mathrm{ZnS} + 1\%\,\mathrm{CuS}$; $\mathrm{ZnS} + 2\%\,\mathrm{CuS}$; $\mathrm{ZnS} + 4\%\,\mathrm{CuS}$; $\mathrm{ZnS} + 8\%\,\mathrm{CuS}$; and $\mathrm{ZnS} + 16\%\,\mathrm{CuS}$, at temperatures from 250 to 400°C and space velocities from 420 to 2100*). Stud—

*) By space velocity is meant the number of volumes of vaporous methanol (at 0°C and 760 mm Hg) that is passed per hour through one bulk volume of catalyst.

was conducted by the method of reverse decomposition of methanol; it was assumed that the observed activities of the contact substances in the reverse reaction would, to a considerable extent, correspond to their activities in the direct reaction. The method of methanol decomposition has already been used repeatedly by other investigators³).

EXPERIMENTAL PART.

The testing of the catalysts consisted of two periods: the 1st period, during which the catalyst was treated, on heating, with methanol vapors until a constant activity was established, and the 2nd period, the direct investigation of the catalyst activity at various temperatures and space velocities.

For the investigation we adopted the dynamic method, the main advantages of which are:

1. Sufficient accuracy in determining the degree of decomposition of methanol.
2. The possibility of establishing a stationary state in which, throughout the entire experiment, constant concentrations are established of the initial, final, and also intermediate substances formed on the surface of the catalyst.

The impossibility of establishing a stationary state in the static method leads to difficulties in processing the results obtained⁴).

The investigation of the catalyst was carried out in the apparatus shown in Fig. 1. All the apparatus was glass, and in order to achieve good tightness the connection of the individual pieces of apparatus was made by sealing.

To obtain methanol vapors, evaporator 1 was used; pure methanol was poured into its inner flask, while the interstitial space between the outer and inner flasks, communicating with the atmosphere through a reflux condenser, was filled with carbon tetrachloride. The evaporator was placed in a water bath; when the latter was heated, the carbon tetrachloride, regardless of the degree of heating of the water bath (above 76°C), boiled at a constant temperature equal to its boiling point at approximately atmospheric pressure.

The evaporated carbon tetrachloride condensed in the reflux condenser and flowed back into the flask. Owing to the constancy of the temperature, the excess pressure of methanol vapors in the evaporator, necessary for overcoming the resistance of all the subsequent apparatus, was maintained very constant, which was especially important for creating a strictly constant rate of gas flow. From the evaporator the methanol vapors were directed into rheometer (2), where their rate was measured.

To prevent condensation of the methanol vapors, the rheometer was placed in a wooden thermostat (3), the front wall of which, for observation, was made of glass. The thermostat was heated by means of an electric heater (4). The temperature of the thermostat (80°C) was maintained constant (±0.2°C) by means of a mercury–toluene thermoregulator (5), made in the form of a coil with a highly developed surface and large volume. To create good heat transfer to the coils of the thermoregulator and the gas heater (6), and to establish the same temperature in all parts of the thermostat, the air in the latter was intensively mixed by a fan (7). Bromonaphthalene was chosen as the liquid for the rheometer. This liquid does not react and does not mix with methanol and, at the temperature of the thermostat, has an insignificant vapor pressure⁵). The height of the arms of the rheometer was 75 cm, which ensured sufficient accuracy in readings of the rate of methanol vapors, which was regulated by a three-way stopcock (8). From the rheometer the methanol vapors entered coil (9), where they were heated to a tempe-

Fig. 1.

Visible labels in the figure:

• “To the current source and thermoregulator”
• “Methyl alcohol”

Numbered callouts visible in the diagram: 1–19.

the contact temperature and then directed into a U-shaped contact apparatus (10) filled with catalyst. The size of the glass coil was selected so that, over the range of space velocities studied, complete preheating of the methanol to the contact temperature occurred. The contact apparatus was made of a thin-walled glass tube 5 mm in diameter, with a catalyst volume of 5 cm³. This design of the contact apparatus was adopted so that the activity of the catalyst in the decomposition of methanol vapor could be studied under truly isothermal conditions, in which phenomena of local overheating were reduced to a minimum. Such overheating, owing to imperfect heat transfer, is an entirely unavoidable phenomenon in contact apparatuses of large diameter. To achieve the greatest isothermicity of the process over the diameter and length of the contact apparatus, the latter, together with the gas preheater, was immersed in an electric furnace (11) filled with a molten mixture of KNO₃ and NaNO₃ (in the ratio, respectively, of 45:55 parts by weight). The furnace liquid was vigorously stirred with an agitator. The temperature of the thermostat liquid was maintained constant to an accuracy of ±2°C by regulating the current with rheostats connected in series, together with an ammeter, into the furnace winding. The temperature of the contact apparatus, equal to the temperature of the thermostat liquid, was measured with a mercury thermometer.

After the nitrate had melted, the furnace was slowly raised by means of a system of lead counterweights and pulleys, and the contact apparatus was immersed in the heat-transfer liquid. This design of a suspended movable electric furnace was used in order to make it possible, after completion of the experiments, to remove the contact apparatus and the preheater from the molten nitrate, since otherwise, upon solidification, the latter would crush these apparatuses.

From the contact apparatus the gases were directed into a condenser cooler (12), where condensation took place of the undecomposed methanol vapors, which flowed as a liquid into a measuring burette (13), where they were also recorded.

The condenser cooler was cooled to a temperature of −20°C with a mixture of snow and common salt. The gases then passed through scrubbers (14) and (15), filled respectively with methanol and solid calcium chloride, where absorption took place of the products of side and intermediate reactions that formed in part during the contact process (for example, CH₂O, HCOOCH₃, and others), and then they were directed into an aspirator (16). From the volume of water (in some cases mercury) that flowed out of the aspirator, and from the temperature and pressure, the total quantity of gases formed per unit time was determined. In the intervals between measurements of the rate of gas formation, part of the gases was directed into gas receivers (17), and from them into a modified Orsat apparatus for analysis of the content of organic impurities (absorbed by concentrated sulfuric acid), CO₂, CO, H₂, and CH₄. Determination of the last two gases was carried out by combustion either in a Drehschmidt capillary or in an explosion pipette.

To prevent condensation of methanol vapors, the glass tubes from the methanol evaporator to the rheometer, from the rheometer to the contact apparatus, and from the latter to the condenser cooler were equipped with electric heating.

The catalysts were investigated as follows. First, the contacting ability of the walls of the contact apparatus was determined. For this purpose, methanol vapors were passed through the apparatus. However, it was not possible to observe any appreciable decomposition of methanol up to a temperature of 480°C within the range of the space velocities at which all further experiments with catalysts were carried out.

After the catalyst had acquired a constant activity, the test itself was carried out. Methanol vapors, at the given temperature and space velocity, were passed for 30–40 minutes through the contact apparatus so that all processes associated with the contact surface would become stationary; then all the data necessary for determining the activity of the catalyst were recorded, i.e., the rate of passage of methanol vapor was measured with a rheometer, the amount of undecomposed liquid methanol condensed in the cooler-condenser was measured, the rate of gas formation was measured from the aspirator readings, and, finally, a portion of the gas was sent for analysis. On the basis of these data the percentage decomposition of methanol was calculated.

PREPARATION AND REDUCTION OF THE CATALYSTS

The catalysts were prepared by mixing zinc and copper sulfides. The sulfides were precipitated with hydrogen sulfide from dilute (0.5 normal) aqueous solutions of the nitrates with additions of ammonium acetate²). The precipitates obtained were filtered off and washed with distilled water, dried for 3–4 hours in a thermostat at 130°, and mixed in the dry state in the specified ratio.

To ensure uniform and close contact of the components, the sulfide mixture was thoroughly ground in a mortar with the addition of distilled water to form a paste, which was then dried for 4 hours at a temperature of 130°C, crushed, and sieved. Pieces of catalyst 1.5–2 mm in diameter were charged into the contact apparatus.

At present it may be regarded as firmly established that the activity of methanol catalysts, both zinc-based and, especially, copper-based, depends not only on the physicochemical properties of the unreduced catalyst, but also very strongly on the temperature regime of reduction. Namely, as a rule, catalysts reduced at low temperatures possess considerably higher activity than catalysts reduced at higher temperatures. The reason for this phenomenon lies in the fact that the finest particles of contact substance formed as a result of reduction, under the influence of high temperature, sinter into larger aggregates having lower activity. This point was demonstrated by X-ray studies by a number of authors who investigated oxide and hydrate forms of zinc and copper catalysts⁶). It may be assumed that an analogous picture should also be observed for highly dispersed sulfide zinc-copper catalysts. Unfortunately, X-ray studies of these catalysts, with the exception of a brief indication in the work of Dolgov, Karpinskii, and Silina²), are absent from the literature.

In the present work, in order to create a highly developed and active surface, and also to obtain a fine structure, sulfide zinc-copper catalysts were reduced with methanol vapors for 15–20 hours at a gradually increasing temperature up to 325°C. The space velocity used in reducing the catalysts was 600.

During the reduction process, especially in its initial period, just as Dolgov, Karpinskii, and Silina did, we observed intensive formation of mercaptans with their unpleasant and characteristic odor*).

*) The reduction process of sulfide zinc-copper catalysts should be regarded as a new method for the organic synthesis of mercaptans.

Table 1

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

1. ZnS

2. ZnS + 0.5% CuS

3. ZnS + 1% CuS

4. ZnS + 2% CuS

5. ZnS + 4% CuS

6. ZnS + 8% CuS

7. ZnS + 16% CuS

RESULTS OF THE EXPERIMENTS

After reduction, the activity of the catalysts was investigated at temperatures from 250 to 400°C and methanol vapor space velocities from 420 to 2100. The results obtained for the catalysts ZnS; ZnS + 0.5% CuS; ZnS + 1.0% CuS; ZnS + 2% CuS; ZnS + 4% CuS; ZnS + 8% CuS; ZnS + 16% CuS (by weight) are summarized respectively in Table 1, and are also presented in Figures 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, and 14. The catalyst prepared from CuS proved to be completely inactive, since in its presence no appreciable decomposition of methanol occurred at temperatures up to 480° in the investigated range of space velocities.

From the table and figures presented it is evident that the percentage decomposition of methanol for catalysts of all investigated compositions decreases regularly with increasing space velocity. However, this decrease in the percentage decomposition of methanol proceeds considerably more slowly than the space velocity increases. Thus, for example, for the catalyst obtained by reduction of pure ZnS at a temperature of 300°C, increasing the space velocity from 420 to 2100, i.e., by a factor of 5, leads to a decrease in the percentage decomposition only from 53.2 to 18.2%, i.e., by a factor of 2.92. An analogous dependence is observed at all temperatures for this catalyst, as well as for the other catalysts investigated by us.

At constant space velocity, the percentage decomposition of methanol increases with increasing temperature for all catalysts investigated by us. Thus, for example, the ZnS catalyst at a space velocity of 1200, when the temperature is raised from 250 to 350°C, causes an increase in the percentage decomposition from 1.11 to 100%. It should be noted here that, for catalysts whose initial forms contained no more than 4% CuS, this increase in activity, expressed by an increase in the percentage decomposition, occurs most sharply in the temperature range 275–325°C. For catalysts containing a larger percentage of copper, however, the temperature interval in which the most vigorous increase in activity is observed shifts toward higher temperatures, namely 300–350°C.

The temperature at which appreciable decomposition of methanol vapors begins for sulfide catalysts containing, before reduction, no more than 4% CuS is 250°C; catalysts that contained, before reduction, 8% and 16% CuS had a somewhat higher decomposition-onset temperature—275°C. An analogous picture is also observed for the temperatures of 100% decomposition. Specifically, for catalysts prepared from zinc–copper sulfides in which the CuS content does not exceed 4%, this temperature at space velocities of 420–600 is 325°C (with the exception of the catalyst containing only 0.5% CuS), whereas catalysts with a higher CuS content gave 100% decomposition at higher temperatures, 350–375°C. The foregoing is characterized by Table 2.

TABLE 2.

Temperature of 100% decomposition of methanol vapors as a function of the composition of catalysts whose initial form consisted of zinc and copper sulfides (space velocity 420–600).

Fig. 2.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS.

Fig. 3.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS.

Fig. 4.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS + 0.5% CuS.

Fig. 5.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS + 0.5% CuS.

Fig. 6.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS + 1% CuS.

Fig. 7.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS + 1% CuS.

Fig. 8.

Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst prepared from ZnS + 2% CuS.

Fig. 9.

Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst prepared from ZnS + 2% CuS.

Fig. 10

Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst prepared from ZnS + 4% CuS.

Fig. 11.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst prepared from ZnS + 4% CuS.

Fig. 12.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst prepared from ZnS + 8% CuS.

Fig. 13.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst prepared from ZnS + 8% CuS.

Fig. 14.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS + 16% CuS.

Fig. 15.
Dependence of the percentage decomposition of methanol on space velocity and temperature in the presence of a catalyst obtained from ZnS + 16% CuS.

It is also evident from Table 2 that the temperature of 100% decomposition of methanol, characteristic of the catalyst obtained from pure ZnS, does not decrease when copper sulfide is added to zinc sulfide, as was observed by Dolgov, Karpinskii, and Silina, but at a certain ratio between ZnS and CuS it even increases, although in absolute magnitude it remains below the temperatures of 100% decomposition for zinc-oxide catalysts.

If activity is judged by the temperatures of 100% decomposition, then the introduction of various amounts of copper sulfide into zinc sulfide does not very substantially change the activity of the resulting complex catalyst. This may be explained by the leveling effect of the elevated temperatures (corresponding to the temperatures of 100% decomposition) on the activity of different reaction centers.

However, the influence of the composition of zinc–copper sulfide catalysts on their activity is sharply manifested at temperatures below the temperatures of 100% decomposition.

As Figures (16, 17, 18, 19, 20) show, the activity of the mixed catalyst, with increasing copper sulfide content, changes according to a complex curve which, over the investigated range of content in the ca-

Fig. 16.

Fig. 17.

Fig. 18.

in the CuS catalyst, 2 minima (0.5 and 8% CuS) and 1 maximum (4% CuS). When 0.5% CuS is introduced into the catalyst, its activity falls and the first minimum is observed; then, with a further increase in the CuS content from 0.5 to 4%, the activity increases and becomes considerably greater than the activity of the catalyst prepared from pure ZnS. The increase in catalyst activity when the CuS content is raised from 1 to 4% proceeds, as a rule, more slowly, and at a CuS content of ≃4% the catalyst has maximum activity. Introduction of still larger amounts of CuS into the catalyst again leads to a sharp drop in activity, and at a content of 8% a second minimum is observed. With an increase

Fig. 19.

Fig. 20.

in the CuS content to 16%, the activity increases, while remaining, in absolute magnitude, considerably lower than that characteristic of catalysts obtained from pure ZnS and zinc–copper catalysts containing, before reduction, from 1 to 4% CuS.

The dependence of activity on catalyst composition described here is preserved over the investigated temperature range at all gas space velocities.

The absolute value of the minima and maxima on the activity curves depends on temperature and space velocity. Specifically, with increasing temperature at constant space velocity, the absolute values of the minimum and optimum activities increase, which is explained by the general increase in catalyst activity with increasing temperature. With increasing space velocity, at constant temperature, the absolute values of the minimum and optimum percentages of decomposition naturally decrease, since the residence time of methanol vapors on the contact decreases.

On the basis of the foregoing, for methanol synthesis, zinc–copper catalysts whose initial sulfide form contains from 1 to 4% CuS may be preliminarily recommended as catalysts first of all, as the catalysts with the highest activity and also decomposing methanol closest to the stoichiometric composition (see below).

The generally increased activity of the catalysts investigated by us in comparison with zinc–copper catalysts²) is apparently explained by their highly developed reaction surface. The latter, in turn,

is determined by the degree of dispersion of the zinc and copper sulfides, which does not change appreciably in the course of reduction of the latter.

The reasons determining the change in activity as a function of the content, in the initial form of the catalyst, of catalytically inactive copper sulfide cannot be established on the basis of the results of the present investigation.

As regards the chemistry of the decomposition of methanol vapor and the influence on it of changes in the composition of the catalyst, then, in agreement with the observations of Dolgov, Karpinskii, and Silina\(^{2}\), a catalyst prepared from pure ZnS decomposes methanol mainly according to the overall stoichiometric equation

\[ \mathrm{CH_3OH \rightarrow 2H_2 + CO}. \]

This is confirmed by the results of analysis of the gas phase given in Table 3. It is seen from Table 3 that the ratio between hydrogen

Table 3

Changes in the composition of gaseous decomposition products as a function of temperature in the presence of a catalyst prepared from ZnS

and carbon monoxide at temperatures of 200 and 350°C is close to 2. Here it is also necessary to note the elevated methane content, which in the presence of this catalyst reaches 7–8%.

Table 4 gives data characterizing the change in the composition of the gases as a function of the CuS content. At 300°C and a space velocity of 600, additions of CuS stimulate side reactions. This is characterized—

Table 4

Changes in the composition of gaseous products of methanol decomposition as a function of catalyst composition (temperature 300° C, space velocity 600)

The data in the last column are at 350° C.

is due to the fact that the ratio $\dfrac{\mathrm{H_2}}{\mathrm{CO}}$ steadily increases with an increase in the CuS content up to 4%. Thus, for example, in the presence of catalysts obtained from ZnS and ZnS + 4% CuS, the ratio $\dfrac{\mathrm{H_2}}{\mathrm{CO}}$ is respectively equal to 1.84 and 3.51. However, when the catalyst contains 8% CuS, $\dfrac{\mathrm{H_2}}{\mathrm{CO}}$ is very close (1.99) to the theoretical ratio corresponding to the equation $\mathrm{CH_3OH} \rightarrow 2\mathrm{H_2} + \mathrm{CO}$, and with a further increase in the CuS content an increase in the ratio $\dfrac{\mathrm{H_2}}{\mathrm{CO}}$ is again observed (at 16% CuS it is equal to 3.03).

The introduction of copper sulfide into a zinc catalyst leads, at \(t = 300^\circ\mathrm{C}\) and \(V = 600\), to a considerable decrease in the methane content in the decomposition products (7 and 3–4%, respectively). Under these conditions the contents of hydrogen, carbon monoxide, and carbon dioxide are respectively 58.8–73.3; 32.0–24.1; 1.3–4.14%.

The influence of temperature and space velocity on the composition of the methanol decomposition products in the presence of catalysts of initial composition ZnS + 8% CuS and ZnS + 16% CuS is presented in Table 5. From this table

Table 5

Change in the composition of the gaseous products of methanol decomposition as a function of temperature and the space velocity of the catalyst, whose initial form contained ZnS + 8% CuS

ZnS + 16% CuS

it is evident that at a space velocity of 600 and with a catalyst containing 8% CuS, the ratio $\dfrac{\mathrm{H_2}}{\mathrm{CO}}$ initially falls and then rises as the temperature is increased from 300 to 400. Increasing the space velocity (\(t = 350^\circ\mathrm{C}\)) in the presence of this same catalyst does not substantially change the gas composition; however, the decomposition deviates strongly from the stoichiometric ratio, which is characterized by large values of $\dfrac{\mathrm{H_2}}{\mathrm{CO}}$ (5.65–7.42) and a high content of CO₂. The composition of the products of methanol decomposition in the presence of the catalyst made from ZnS + 16% CuS depends little on temperature and space velocity.

At the end of the investigation of the activity of the catalyst obtained from ZnS, it was poisoned with 100% hydrogen sulfide. For this purpose a stream of hydrogen sulfide is passed at a rate of 50 cm³/min for 15 minutes, at a temperature of 325°C. The subsequent decomposition of methanol on the catalyst poisoned in this way showed that the phenomenon of poisoning is also characteristic of the catalyst obtained from zinc sulfide. However, this catalyst, when pure methanol vapor was passed at a temperature of 325°C and a flow rate of 50 cm³/min, completely regenerated its activity within 10 minutes, whereas this phenomenon is not observed for catalysts based on copper oxide\(^7\). However, in order to establish an exact picture of hydrogen-sulfide poisoning of catalysts obtained from zinc and copper sulfides, special experiments are required.

According to the investigations of Dohse\(^8\), the decomposition of methanol when it is passed through a contact substance consisting of zinc oxide proceeds according to the law of reactions of zero order. Assuming that, in the presence of the catalysts studied by us, whose initial form before reduction consisted of zinc and copper sulfides, the methanol decomposition reaction likewise has zero order, one may, from the point of view of formal kinetics, write the following equation for the rate of decomposition:

\[ \frac{dx}{dt}=K_1. \tag{1} \]

After integration we obtain

\[ x=K_1t+\mathrm{Const}. \]

In the initial period, at time \(t=0\), \(x=0\), and \(\mathrm{Const}\) is also equal to 0.

Hence

\[ K_1=\frac{1}{t}x. \tag{2} \]

Introducing instead of \(x\) the degree of decomposition of methanol vapor \(\alpha=\frac{x}{a}\) in fractions of unity, where \(a\) is the initial constant concentration of methanol vapor, we obtain

\[ K_2=\frac{1}{t}\alpha. \tag{3} \]

To calculate the rate constants by formula (3), it is necessary to know the residence time of the gas on the catalyst, during which a known amount of methanol vapor decomposed. In our experiments, carried out by the dynamic method with a catalyst having a granular form, the reaction time cannot be determined, since the true volume occupied by the catalyst is unknown. Therefore, in the case when the gas volume as a result of the reaction remains constant or changes only very slightly, instead of the residence time \(t\) one may substitute into the kinetic equation the reciprocal of the space velocity, previously corrected for gas expansion with temperature\(^9\), since

\[ t \text{ is proportional to } \frac{1}{V \frac{T}{273}}, \]

where \(V\) is the space velocity,
\(T\) is the temperature in °K.

For the case of methanol decomposition, instead of \(t\) the quantity

\[ \frac{1}{V \frac{T}{273}} \]

may be substituted only for those conditions under which the degree of methanol decomposition

is small; at high degrees of decomposition such a substitution will not correspond to reality, since during the decomposition of methanol to CO and H₂ there is a large increase in volume. Therefore, in addition to the temperature correction, it is necessary to introduce another correction for the increase in gas volume as a result of the course of the reaction itself.

The analyses of the gaseous products given above indicate that the decomposition of methanol proceeds mainly according to the overall equation

\[ \mathrm{CH_3OH} \rightarrow 2\mathrm{H_2} + \mathrm{CO}, \]

i.e., the increase in volume upon complete decomposition of methanol is approximately threefold. Therefore the volumetric velocity in the contact apparatus will also increase. We assume that the mean volumetric velocity of the gas will be equal to the arithmetic mean of the initial and final volumetric velocities, i.e.

\[ V_{\text{mean}} = \frac{V + V_1}{2}, \]

where \(V\) is the initial volumetric velocity of methanol vapors;
\(V_1\) is the final volumetric velocity of the gaseous decomposition products.

But \(V_1 = V(1 + 2\alpha)\), and hence

\[ V_{\text{mean}} = V(1 + \alpha), \]

where \(\alpha\) is the degree of methanol decomposition in fractions of unity, equal to the percentage decomposition divided by 100.

After substituting into equation (3), instead of \(t\), the quantity proportional to it and equal, for the case of methanol decomposition, to

\[ \frac{1}{V(1+\alpha)\dfrac{T}{273}}, \]

the final kinetic equation (4) was obtained.

\[ K = V(1+\alpha)\frac{T}{273}\alpha . \tag{4} \]

The rate constants of decomposition in the presence of all the catalysts investigated by us, calculated from equation (4), are presented in Table 6. It is evident from this table that, within the limits of experimental error, the rate constants, at constant temperatures but at different volumetric velocities, remain constant quantities. The exception is the catalyst whose initial form contained \(\mathrm{ZnS} + 8\%\,\mathrm{CuS}\). In this case, with increasing volumetric velocity at temperatures of 275–350°C, a systematic decrease in the rate constants is observed.

Thus, in the presence of this catalyst, the course of the methanol decomposition reaction apparently deviates from the given law.

From the calculated rate constants, the activation energies were found for all the catalysts investigated by us.

As is known, the relation between the activation energies and the rate constants of a reaction is expressed by the Arrhenius equation \(^{10)}\)

\[ \frac{d \ln K}{dT} = \frac{E}{RT^2}, \tag{5} \]

where \(K\) is the reaction rate constant,
\(T\) is the absolute temperature,
\(R\) is the gas constant in calories,
\(E\) is the activation energy in calories per mole of decomposed substance.

Table 6

Rate constants of the methanol decomposition reaction in the presence of a catalyst prepared from ZnS

Continuation of Table 6

Continuation of Table 6

Assuming $E$ to be independent of temperature, after integrating equation (5) and replacing natural logarithms by common logarithms, we obtain equation (6).

\[ \log K=-\frac{E}{2.303\,R}\left(\frac{1}{T}\right)+\text{Const.} \tag{6} \]

Equation (6) is the equation of a straight line, the tangent of whose angle of inclination, multiplied by $2.303\,R$, is equal to the activation energy. Since the gas reaction

\[ \mathrm{CH_3OH}\to 2\mathrm{H_2}+\mathrm{CO} \]

proceeds in the presence of a solid catalyst, i.e., it is heterogeneous, calculations by equation (6) give not the true value, but the apparent activation energy.¹¹)

Plotting \(\left(\dfrac{1}{T}\right)\) along the abscissa axis, and \(\log K\) along the ordinate axis, we obtained, in the interval of temperatures investigated, straight lines for the catalysts tested by us, which are presented in Figs. (21) and (22).

According to these figures, the tangents of the angles of inclination were determined graphically; multiplying them by

\[ 2.303 \cdot R = 2.303 \cdot 1.987 \]

gave the sought values of the apparent activation energies. The results of these calculations are presented in Table 7.

Fig. 21

On the basis of the satisfactory constancy of the rate constants found at different space velocities and constant temperatures, one may conclude that the assumptions made by us earlier concerning the zero order of the reaction of decomposition of methanol vapors in the presence of catalysts, the initial form of which consisted of zinc and copper sulfides, are correct. In this connection, the mechanism of methanol decomposition in our case may be represented as follows: methanol vapors, reaching the catalyst as a result of intensive adsorption, almost completely cover the reaction-capable active surface of the contact; as a result of this, the concentration of the starting product in the adsorption layer will not change substantially when the concentration of methanol vapors in the gas phase decreases, and thus the recorded rate of decomposition under isothermal conditions will be a constant quantity, determined by the kinetic equation characteristic of zero-order reactions:

\[ \frac{dx}{dt} = K. \]

Fig. 22.

Our studies, which establish the zero order of decomposition of methanol, do not give direct indications concerning the decomposition processes taking place in the adsorption layer itself, since in reactions of apparent zero order the true order of the reaction may be very differ-

However, it may be assumed that the mechanism of methanol decomposition in the adsorption layer on the catalysts investigated by us is the same as that established by Dohse\(^8\) for zinc oxide. Namely, the decomposition of adsorbed methanol vapors is a consecutive reaction consisting of two successively occurring reactions, according to the following equations:

\[ \mathrm{CH_3OH \rightarrow CH_2O + H_2,} \tag{1} \]

\[ \mathrm{CH_2O \rightarrow CO + H_2.} \tag{2} \]

For the first stage of decomposition, characterized by the formation of formaldehyde, according to Dohse’s investigations the activation energy in the presence of ZnO is 22,500 cal. The activation energy of the second, most difficult process, according to the investigations of the same author, is 35,000 cal/mol.

According to our investigations, as Table 7 shows, the activation energy of methanol decomposition varies depending on the composition of the catalyst. Thus, for a catalyst obtained from pure ZnS, the apparent activation energy is 42,400 cal/mol; upon introducing up to 1% CuS into the catalyst, the activation energy increases to 49,500 cal/mol. A further increase in the CuS content to 4% does not lead to a substantial change in the activation energy, and only increasing the CuS content to 8% causes an increase in the activation energy to 54,500 cal/mol. The activation energy measured at higher temperatures for a catalyst containing 16% CuS decreases to 32,700 cal/mol.

As is known\(^12\), the usual indication of the occurrence of a consecutive reaction is the presence of a break in the curve obtained by plotting

\[ \log K \text{ versus } \left( \frac{1}{T} \right) \]

and a flatter position of the upper straight line in comparison with the lower one. However, in order to decide the question of precisely which partial reaction (the 1st or the 2nd) a given straight line corresponds to, and consequently also the activation energy, additional experiments are required.

Dohse established that the upper, flatter straight line on the diagram

\[ \log K \text{ versus } \left( \frac{1}{T} \right) \]

with an activation energy of 22,500 cal corresponds to reaction (1), and the steeper straight line with an activation energy of 35,000 cal/mol to partial reaction (2).

Table 7

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

Apparently, for the catalysts we investigated, the calculated activation energies also correspond to the reaction of decomposition of formaldehyde (i.e., the reaction of the 2nd stage), since in the plots of $\log K$ versus $\left(\frac{1}{T}\right)$ for catalysts containing up to 4% CuS, the logarithms of the rate constant at a temperature of 325°C systematically deviate from the straight line toward a smaller slope (see diagrams 21 and 22), and, when the temperature interval of the investigation is extended toward higher temperatures, it will apparently be possible to observe a flatter part of the straight line corresponding to the reaction of decomposition of methanol into formaldehyde and hydrogen, which is probably also observed for the catalyst containing 16% CuS; in this case the activation energy of reaction (1) is 32,700 cal/mol.

Thus, the process of decomposition of methanol in the adsorption layer on the surface of the catalysts we investigated is apparently a consecutive reaction consisting of two stages: in the first of them formaldehyde and hydrogen are formed from methanol, and in the second formaldehyde decomposes into carbon monoxide and hydrogen. The fact that the composition of the gas after decomposition, as a rule, deviates from the stoichiometric one is explained by the presence of side reactions that lead to the formation of CH$_4$, CO$_2$, etc. The change in activity and activation energy with the composition of the catalyst is apparently connected with changes in the physicochemical structure of the catalysts and, in particular, with the size and structure of the finest particles of which the catalyst is built, and with the development of its surface. However, the answer to this question must be provided by X-ray diffraction and electron-microscopic investigations.

CONCLUSIONS

1. An apparatus for studying the rate of methanol decomposition has been described, and catalysts were studied which, before reduction, consisted of ZnS, CuS, and their mixtures of the following compositions:

ZnS + 0.5% CuS; ZnS + 1% CuS; ZnS + 2% CuS;

ZnS + 4% CuS; ZnS + 8% CuS; ZnS + 16% CuS (by weight),

at space velocities $V = 420$–2100 and temperatures of 250–400°C.

1. The catalyst obtained from CuS proved to be completely inactive, since within the range of the space velocities investigated it did not give noticeable decomposition up to temperatures of 480°C.

2. Catalysts whose initial forms consisted of ZnS and its mixture with CuS, at $V = 420$–600, completely decomposed methanol at temperatures of 325–350°C (with the exception of the catalyst obtained from ZnS + 8% CuS, for which the temperature of 100% decomposition was 375°C), i.e., they proved to be considerably more active than zinc oxide catalysts.

3. In the presence of all the catalysts investigated and at constant temperature, the percentage decomposition of methanol decreases with increasing space velocity; however, this decrease in the percentage decomposition occurs considerably more slowly than the space velocity increases.

4. With increasing temperature at $V = \mathrm{Const.}$, for all the catalysts investigated the percentage decomposition of methanol increases.

5. The activity of the complex catalyst, as the content of copper sulfide in it increases, changes along a complex curve having, in the interval investigated, 2 minima (at 0.5 and 8% CuS) and 1 maximum (at 4% CuS).

6. The decomposition of methanol vapors proceeds according to a law characteristic of a reaction of apparent zero order. The rate constants have been calculated.

1. From the temperature variations of the rate constants, the apparent activation energies were calculated; for catalysts obtained from ZnS, ZnS + 0.5% CuS, ZnS + 1% CuS, ZnS + 2% CuS, ZnS + 4% CuS, ZnS + 8% CuS, and ZnS + 16% CuS, they proved to be, respectively, 42,400, 43,200, 49,500, 47,800, 49,300, 54,500, and 32,700 cal/mole.

2. It is assumed that the process of methanol decomposition in the adsorption layer is a consecutive reaction consisting of two stages: in the first of them methanol decomposes into formaldehyde and hydrogen, and in the second formaldehyde decomposes into carbon monoxide and hydrogen. The activation energies calculated by us (with the exception of \(E = 32700\) cal.) apparently refer to the second stage of methanol decomposition. The presence in the gaseous decomposition products, in addition to CO and H\(_2\), of other gases (CO\(_2\), CH\(_4\), etc.) is explained by the presence of side reactions.

3. For the synthesis of methanol, zinc–copper catalysts, whose initial sulfide form contains from 1 to 4% CuS, may be positively recommended as catalysts.

LITERATURE

1. Bader and Thomas.—Celanese Corp. of America USP 1942530 (of 10.II—30; 9.I.1934).

2. Dolgov, Karpinskii, and Silina.—“Chemistry of Solid Fuel,” V. 5, 470, 1934.

3. Arkius and Perkins,—J. Phys. Chem., 32, 221, 1928.
   Smith and Hawk.—J. Phys. Chem., 32, 415, 1928.
   Storch.—J. Phys. Chem., 32, 1743, 1928.
   Lewis and Fröhlich.—Ind. Eng. Chem., 20, 285, 1928.
   Audibert and Raineau.—Ind. Eng. Chem., 20, 1105, 1928.
   Hoffman and Dodge.—Ind. Eng. Chem., 21, 1056, 1929.
   Plotnikov and Ivanov.—Journal of Chemical Industry, 6, 941, 1929.

4. On this question see Emmett’s article in Curtis’s book Bound Nitrogen, GKhTI, 1933, in which, in particular, the process of ammonia decomposition on iron catalysts is considered.

5. Kahlbaum.—Z. Phys. Chem., 26, 577, 1898.

6. Natta.—G. chim. ind. ed appl., 12, 13, 1930.
   Hüttig and Feher.—Z. anorg. Chem., 197, 129, 1931.
   Hüttig, Kostelitz and Feher.—Z. anorg. Chem., 198, 206, 1931.
   Hüttig and Steiner.—Z. anorg. Chem., 199, 149, 1931.
   Ushakov, Zimakov, and Zinov’ev.—Journal of Chemical Industry, 2, 590, 1931.
   Ivannikov, Frost, and Shapiro.—“Reports of the Academy of Sciences,” 1933.
   Dolgov.—Methods for the Chemical Utilization of Carbon Oxides. ONTI, 1936.

7. On the process of poisoning of zinc–chromium catalysts, see the works of Velistova and Karpov, Bulletin of GIIVD, 7, 65, 1933.

8. Dohse.—Z. Phys. Chem., 8, 3, 159, 1930.

9. Kurin.—Journal of Chemical Industry, 7, 475, 1937.
   Kurin and Blokh.—Journal of Applied Chemistry, 5, 734, 1938.

10. Arrhenius.—Z. Phys. Chem., 4, 226, 1889.

11. Hinshelwood.—Kinetics of Gas Reactions. GGTI, 1933.

12. Eucken and Rakovskii.—Chemical Kinetics and Catalysis, GKhTI, 1932.