Catalytic Oxidation of Carbon Monoxide with Oxygen in the Presence of Hydrogen

N. P. Kurin and G. A. Orman

Catalytic Activity of Hopcalite

As is known[^1], carbon monoxide is a contact poison for ammonia catalysts. Therefore, at present, in the production of synthetic ammonia it is necessary to resort to careful purification of the nitrogen–hydrogen mixture from carbon monoxide. The amount of carbon monoxide in the nitrogen–hydrogen mixture or in hydrogen depends on the method of production. Thus, a nitrogen–hydrogen mixture obtained from coke-oven gas by deep cooling according to the Linde method contains hundredths of a percent of carbon monoxide; when hydrogen produced by the iron–steam method is used, the concentration of CO in the nitrogen–hydrogen mixture reaches 0.4–0.8%; and, finally, the gas mixture after conversion of water gas contains 3–4% carbon monoxide.

At low CO concentrations, the nitrogen–hydrogen mixture is purified by the methanation method[^2]. At higher concentrations (3–4%), carbon monoxide is removed from the gas mixture by washing it in ammoniacal copper solutions. However, this purification method must be classed among the expensive ones, since, in order to prevent the adverse effect of elevated temperature (as a result of the heat of absorption being released) on the process of carbon monoxide absorption, it is necessary to use a special ammonia refrigeration unit to cool the solution. In addition, special equipment is required for regeneration of the spent absorbent.

The low profitability of such wet purification of a nitrogen–hydrogen mixture containing 0.4–0.8% carbon monoxide forces the unpurified nitrogen–hydrogen mixture to be sent for processing into ammonia, which inevitably leads to very rapid poisoning of the catalyst and accumulation of methane in the circulating gas mixture, i.e., ultimately to a decrease in the productivity of industrial installations and to an increase in the cost of synthetic ammonia. In light of the foregoing, the search for new methods of more rational and inexpensive purification of gases from carbon monoxide should be considered expedient; in this connection, very attractive prospects are opened up by the catalytic selective oxidation of CO to CO₂ in the presence of hydrogen, since removal from the gas mixture of the carbon dioxide formed—for example, with the aid of an alkaline solution—would not present substantial difficulties.

The essence of selective catalytic oxidation consists in the fact that carbon monoxide present in a nitrogen–hydrogen mixture or in hydrogen, in the presence of certain catalysts, burns to carbon dioxide at the expense of previously introduced appropriate amounts of oxygen, while the hydrogen either does not react with oxygen at all, or reacts only slightly; thus, under certain conditions carbon monoxide can be oxidized faster than hydrogen[^3].

A number of substances have been tested as catalysts for the selective oxidation of carbon monoxide in the presence of hydrogen. Thus, Rideal and Taylor\(^{4}\) showed that copper oxide at a temperature of 110°C and the oxides of iron and chromium at 250–350°C bring about preferential oxidation of carbon monoxide as compared with hydrogen. The rate of selective oxidation can be substantially increased by introducing into the catalysts such additives as the oxides of cerium and thorium.

Further, Rideal\(^{5}\) investigated the activity of Fe\(_2\)O\(_3\), Fe\(_2\)O\(_3\) with additions of 2.5% Cr\(_2\)O\(_3\) and 0.5% CeO\(_2\), CuO, NiO, and Pd in the form of palladized asbestos in the temperature range 150–380°C, with an initial composition of the dry gas of CO from 2.5 to 16.6, O\(_2\) from 5.5 to 39.0, and H\(_2\) from 48.5 to 90.3% (by volume), and at various rates of passage through the catalyst (according to our calculations, the space velocities of the gas in Rideal’s experiments varied from 42.4 to 400 volumes of gas at 0° and 760 mm Hg per 1 bulk volume of catalyst per hour). It turned out that at temperatures of about 150°C all the listed catalysts, with the exception of palladium, are only slightly active; at higher temperatures, simultaneous oxidation of both carbon monoxide and hydrogen takes place, and an increase in temperature lowers the selectivity of CO oxidation.

Lamb, Scalione, and Edgar\(^{6}\) used hopcalite\(^{7}\), invented in America during the World War, for the selective oxidation of carbon monoxide in the presence of hydrogen. This substance, used for filling gas masks, even at room temperature brings about very intense oxidation of carbon monoxide present in air. The authors named studied the behavior of hopcalite when dry and moist mixtures of hydrogen and air, carbon monoxide and air, and hydrogen, carbon monoxide, and air were passed through it. It was found that in a dry gas mixture containing 90% H\(_2\) and 10% air, oxidation of hydrogen begins only at temperatures not lower than 60–70°C (space velocity = 6000); if the gas mixture is humidified (\(P_{\mathrm{H_2O}} = 50\) mm Hg), the temperature at which hydrogen oxidation begins rises to 125°C.

On the other hand, when a dry gas mixture containing 0.5% CO, 89.5 H\(_2\), and 10% air is passed through this catalyst, complete oxidation of CO to CO\(_2\) occurs at room temperature (space velocity 6000), while the hydrogen does not burn; however, if the CO concentration is raised to 1%, then under the same conditions, owing to the evolution of the heat of combustion of CO, intense heating of the catalyst occurs, causing simultaneous intense combustion of hydrogen.

Further, in the case of a moist mixture of carbon monoxide with a content of CO—0.5–1, air 2.5–5, and N\(_2\)—94–97%, at a space velocity of 6000, complete oxidation of CO to CO\(_2\) occurs at temperatures of 90–100°C, i.e., approximately 20–25°C below the temperature at which hydrogen oxidation begins.

To determine the constancy of the catalyst activity, a moist nitrogen–hydrogen mixture containing 0.5% CO and 0.5% oxygen was passed for a long time through hopcalite placed in a copper contact apparatus, at a temperature of 101.8°C and a space velocity of 6000. It was found that under these conditions, over the course of 200 hours, the degree of oxidation becomes constant and equal to 93%. In conclusion, the same authors studied the activity of hopcalite when a gas mixture containing nitrogen, carbon monoxide, and oxygen was passed through it, as a function of temperature, space velocity, oxygen concentration, etc.; these factors were related to the thermal phenomena occurring on the catalyst.

Kuentzel\(^{8}\), with the aim of finding conditions under which it would be possible selectively to oxidize carbon monoxide to carbon dioxide, the carbon monoxide being present in the gas mixture at a considerably higher concentration (1–2 and more percent), studied the following substances as catalysts:

properties. Two catalysts (types 1 and 2) used for the conversion of water gas of the composition:

1. \(94\%\, \mathrm{Fe_2O_3} + 5\%\, \mathrm{Al_2O_3} + 1\%\, \mathrm{KOH};\)
2. \(95\%\, \mathrm{Co_2O_3} + 5\%\, \mathrm{Al_2O_3};\)
3. \(100\, \mathrm{CuO};\)
4. \(100\, \mathrm{MnO_2};\)
5. two-component hopcalite—\(60\%\, \mathrm{MnO_2} + 40\%\, \mathrm{CuO};\)
6. five-component hopcalite—\(39\%\, \mathrm{MnO_2} + 30\%\, \mathrm{CuO} + 17\%\, \mathrm{Co_2O_3} + 8\%\, \mathrm{Fe_2O_3} + 6\%\, \mathrm{Ag_2O_3}.\)

Tests of catalyst activity were carried out at a space velocity of 720 and with a gas-mixture composition of \(\mathrm{CO_2}\)—27.5; \(\mathrm{CO}\)—1.1; \(\mathrm{O_2}\)—1.1; \(\mathrm{CH_4}\)—0.3; \(\mathrm{H_2}\)—47.6 and \(\mathrm{N_2}\)—22.6% at various temperatures and with a large amount of water vapor (steam/gas ratio \(=3\)).

The first two catalysts do not possess selectivity of action and are typical conversion catalysts. Copper oxide in the temperature range 185–225°C has high selective properties with respect to the oxidation of CO to \(\mathrm{CO_2}\) in the presence of hydrogen; the degree of oxidation of CO to \(\mathrm{CO_2}\) reached 72–95% under these conditions. However, on going from high temperatures to lower ones, the activity of the catalyst drops sharply, which is explained by reduction of the catalyst by hydrogen. \(\mathrm{MnO_2}\) behaves similarly as a catalyst.

Of all the catalysts investigated by Kuentzel, the most active proved to be two-component hopcalite. The attainable degree of oxidation of CO to \(\mathrm{CO_2}\) in the temperature range 150–163°C was respectively 90–95%, although after several hours of operation it fell to 81.3–90.7%, which the author relates to the slight reduction of the catalyst observed in the first period of operation, which then ceases. Raising the temperature to 165–190°C makes it possible to obtain a degree of oxidation above 95%. At temperatures of about 220°C the catalyst loses its activity. The five-component catalyst proved less active than the two-component one.

Subsequently Kuentzel\(^9\) showed that, using commercial two-component CWS hopcalite at 160–195°C, with gas composition \(\mathrm{CO}\)—1.3; \(\mathrm{O_2}\)—1.5; \(\mathrm{CO_2}\)—28.0; \(\mathrm{H_2}\)—49.5 and \(\mathrm{N_2}\)—19.7%, a space velocity of 100 (calculated on dry gas), and a steam/gas ratio \(=3\), it is possible to oxidize from 97 to 99% of the carbon monoxide present in the gas. Steam promotes rapid dissipation of the heat of reaction and, like oxygen, prevents reduction of the catalyst. Increasing the oxygen excess in the gas from 100 to 145% only slightly increased the degree of oxidation of CO to \(\mathrm{CO_2}\).

Kazarnovskii, Borshchevskii, and Kastin\(^ {10}\) investigated the following catalysts as catalysts for the selective oxidation of CO to \(\mathrm{CO_2}\) in the presence of hydrogen: 1 and 2—\(100\%\, \mathrm{Fe_2O_3}\); 3—\(\mathrm{Fe_2O_3} + 2.5\%\, \mathrm{Cr_2O_3} + 0.5\%\, \mathrm{Ce_2O_3}\), this catalyst being deposited on a silica-gel support in an amount of 30%; 4—\(\mathrm{Fe_2O_3} + 2.5\, \mathrm{Cr_2O_3} + \mathrm{Ce_2O_3}\); 5—hopcalite.

The activity study was carried out at a space velocity of 200 and with a dry nitrogen–hydrogen mixture containing 0.5–1.0% CO and 0.5–1.0% \(\mathrm{O_2}\). As a result of the study it was found that the activity of iron catalysts, and especially of unactivated ones, is considerably lower than the activity of hopcalite. Hopcalite gives good results at a temperature of about 100°C. The degree of oxidation of CO to \(\mathrm{CO_2}\) in this case is 93–95%. At higher temperatures, 130–150–200°C, hopcalite changes color and loses activity, which the authors explain by reduction of the catalyst. In the second part of the work, the influence of various factors on the activity of hopcalite was studied. It turned out that operation on moist gas (saturation with water vapor at 25–30°C) proceeds satisfactorily. Increasing the excess of oxygen relative to the amount required for complete

increasing the oxidation of CO to CO₂ above 100% does not give any substantial increase in the degree of oxidation. When the space velocity is increased, the degree of oxidation decreases, which is partly connected with easier overheating of the catalyst and its destruction.

Finally, the change in the activity of hopcalite as a function of operating time under various conditions was investigated. It turned out that over 96 hours, at a gas space velocity of 5000 and with the use of a dry and pure nitrogen–hydrogen mixture, the activity does not change. However, when using a technical nitrogen–hydrogen mixture containing 5–7 mg of organic sulfur compounds per 1 m³ of gas, during operation for 27–38 hours the activity falls from 90–93 to 67.5–64.2%, which is explained by poisoning of the catalyst. During operation on a moist nitrogen–hydrogen mixture (7–8% H₂O) and with the use of a glass contact apparatus, the activity of hopcalite decreases somewhat with time. A similar picture is also observed when using a copper contact apparatus (over 100 hours of operation at a space velocity of 5000 and \(t = 101^\circ\mathrm{C}\), the degree of oxidation fell from 95.6 to 65%).

Thus, from the works cited it is evident that, in principle, it is possible to carry out selective heterogeneous catalytic oxidation of carbon monoxide to carbon dioxide in the presence of hydrogen. Hopcalite proved to be the best catalyst. In the presence of this substance, with the aid of small amounts of oxygen, CO can be oxidized to CO₂ even if carbon monoxide is present in the nitrogen–hydrogen mixture in amounts of 1–2 percent or more. In the latter case, to prevent damage to the catalyst due to overheating and reduction, larger amounts of water vapor should be admixed with the nitrogen–hydrogen mixture, which is precisely what takes place in the conversion method for obtaining \( \mathrm{N_2 + 3H_2} \).

The purpose of the present work is to investigate the activity of hopcalite as a function of the temperature and the space velocity of the gas. As is known, temperature and space velocity, in the practical implementation of catalytic processes, are subject to the most frequent regulation; therefore, determining the degree of catalytic oxidation of CO to CO₂ in the presence of H₂ as a function of these two factors is highly desirable. In addition, studying the rate of this process as a function of temperature and space velocity will make it possible to draw certain conclusions regarding the mechanism of the catalytic oxidation of CO to CO₂ in the presence of hydrogen.

EXPERIMENTAL PART.

To investigate the activity of hopcalite, we chose a dynamic method, which makes it possible to study this quantity under established stationary states and to analyze, with sufficient accuracy, the composition of the gases after the catalyst at low initial concentrations of carbon monoxide. Taking into account the negative effect of overheating on the activity of hopcalite noted by previous investigators, our study of the latter was carried out under conditions close to isothermal. In addition, apparatus was provided for careful regulation and measurement of the rates of the gas flow.

The scheme of the experimental setup is shown in Fig. 1. The gas mixture, consisting of a nitrogen–hydrogen mixture, carbon monoxide, and air from a gas holder (1) (for continuous operation, two such gas holders were installed), under the pressure of a water column from a water-pressure bottle (with a constant level) (2), was passed through absorption columns (3) and (4), filled respectively with 40% NaOH solution and concentrated sulfuric acid. In column (3), purification of the gas mixture took place

Figure 1

Visible labels in the figure: $\mathrm{CO_2}$; $\mathrm{Ba(OH)_2}$.

from carbon dioxide and in the column (4) its drying. Next, the mixture was passed through a flowmeter (5), by means of which the rate of the gas stream was regulated and measured, and was sent for contacting into the thermostat (6). The flowmeter (5) had two capillaries and a large difference in levels (750 mm) of concentrated sulfuric acid in the low- and high-pressure arms, which made it possible to measure the rate of gas streams over wide ranges and with great accuracy. After passing through the preheater (7) and the contact apparatus (8) filled with catalyst, the gas mixture was sent for analysis. The contact apparatus and the preheater were located in a metal reservoir (6) filled with liquid and insulated on the outside with felt in order to reduce heat losses. The use of a liquid, rather than a gas, as the heat-transfer medium was dictated by the desire to minimize, as far as possible, the temperature difference both over the cross section of the contact apparatus and along its length, arising as a result of the release of heat in the reaction of combustion of carbon monoxide to CO₂. In experiments on the activity of hopcalite at temperatures up to 100°C, the thermostat was filled with water, and at higher temperatures with castor oil. The temperature of the thermostat liquid was maintained constant to an accuracy of ±0.1°C by means of an electric heater (9) and a mercury-toluene thermoregulator (10). To increase the intensity of heat transfer, the thermostat liquid was vigorously stirred by means of a stirrer (11). The size of the coil preheater (7) was chosen with a large margin, so that the gas mixture entering the contact apparatus, even at the highest volumetric velocities, had time to be heated to the temperature of the thermostat liquid. To exclude the catalytic influence of the walls on the activity of hopcalite in the selective oxidation of CO to CO₂ in the presence of hydrogen, the contact apparatus (8) was made of a glass tube and had a U-shaped form.

However, taking into account the statements of previous investigators concerning the harmful effect of overheating on the catalyst, as well as our desire to study the activity of hopcalite under conditions close to isothermal ones, the contact apparatus was made of a thin-walled and narrow tube. The wall thickness of the contact apparatus was 0.5 mm, and the tube diameter was 4 mm.

The nitrogen-hydrogen mixture was obtained by decomposing ammonia over iron turnings at high temperature (~700°C). For this purpose, ammonia from the cylinder (12) was passed through a cracker (13), filled with iron turnings and heated by an electric furnace. In the columns (14), filled with sulfuric acid, the nitrogen-hydrogen mixture was freed from residues of undecomposed ammonia and was then sent into one of the gas holders (1).

Carbon monoxide was obtained by decomposing formic acid with sulfuric acid. For this purpose, concentrated sulfuric acid was poured from vessel (17) into flask (15). There it was heated to 150–160°C, and then formic acid was added dropwise from vessel (16). The carbon monoxide formed was passed through a condenser (18), where condensation of the vapors of undecomposed formic acid occurred, and, after purification from CO₂ and H₂O, was collected in the gas holder (19). For continuous operation, the receivers (16) and (17) were connected to the tube conducting gas away from the flask. With this arrangement, the gas pressure in the flask and above the liquid in the receivers (16) and (17) is the same, and therefore the formic acid can flow in continuously.

When preparing a gas mixture of the required composition, carbon monoxide from the gas holder (19), with the stopcock (20) closed and the stopcock (21) open, was passed through the absorption columns (22) and (23), filled, respectively, with a solution of caustic potash and concentrated sulfuric acid.

(for removal of the CO₂ evolved from the water in the gasometer (19) and of water vapor) and entered the measuring burette (24), in which mercury was used as the confining liquid. After this the required volume of CO from burette (24) was transferred into one of the gasometers (1), previously completely filled with water. Into the same gasometer, from another similar burette (25), the required quantity of air was transferred, after prior removal of CO₂ and H₂O in the wash bottles (26) and (27). After this, a nitrogen–hydrogen mixture from the ammonia-decomposition apparatus was introduced into gasometer (1) up to a specified mark.

The degree of oxidation of carbon monoxide was determined from the amount of CO contained in the gas before the catalyst and the amount of carbon dioxide trapped from the same amount of gas after the catalyst, when it was passed through a solution of caustic baryta. The concentration of the caustic baryta solution before and after passage of the gas under analysis was determined by titration with a solution of oxalic acid.

To prevent breakthrough of carbon dioxide, two absorption bottles with caustic baryta were connected in series. U-shaped tubes filled with P₂O₅ made it possible to determine gravimetrically the amount of water formed as a result of hydrogen combustion.

PREPARATION OF THE CATALYST

In the present communication, results are given concerning the activity of hopcalite obtained by mixing pure precipitated cupric oxide and manganese dioxide.

Cupric oxide was prepared¹²) by precipitation from a solution of copper sulfate with a solution of caustic soda. The copper oxide hydrate, initially obtained as a bluish precipitate, upon heating of the solution rapidly changed into cupric oxide, which was thoroughly washed with distilled water until the reaction for the ion \((\mathrm{SO}_4)''\) disappeared, and was dried in a drying oven at a temperature of 60–80°C.

Active MnO₂ was obtained¹²) by the interaction of potassium permanganate and manganous sulfate in aqueous solution while boiling. The precipitate of MnO₂ obtained was thoroughly washed with distilled water until the reaction for the sulfuric acid ion disappeared and was likewise dried at a temperature of 60–80°C. The dried oxides, weighed out in the appropriate amounts, were mixed with a small quantity of distilled water. For more complete mixing, the paste obtained was ground for a long time in an agate mortar, then spread on glass, dried, and heated at a temperature of 200°C for one hour. The hopcalite obtained was crushed, and for the activity studies a fraction with a particle diameter from 1.5 to 2 mm was used. The volume of the contact space in which the catalyst was placed was 3.85 ml; the weight of the catalyst was 2.101 g.

The activity of the hopcalite was studied as follows: initially, at each given temperature and space velocity, the gas mixture was passed through the catalyst for 0.5–1 hour, and only after this were the contacted gases sent for analysis. This was necessary in order to establish a stationary state between the gas phase and the solid catalyst.

RESULTS OBTAINED AND THEIR DISCUSSION

In the present communication, data are given concerning the activity, measured by the percentage oxidation of CO to CO₂, of a two-component hopcalite (60% MnO₂ and 40% CuO) at space velocities \((V)\) from 500 to 9300

and at temperatures of 25–135°C when a dry nitrogen–hydrogen mixture containing 0.5% carbon monoxide and 0.5 and 0.3% oxygen was passed through the catalyst.

The results obtained are given in Table 1 and presented in Figs. (2), (3), (4), and (5).

Table 1

Dependence of the degree of oxidation of CO to CO₂ in the presence of a nitrogen–hydrogen mixture on hopcalite on temperature, space velocity, and initial O₂ concentration. Initial CO concentration—0.5% (vol.)

From Table 1 and Figs. (2) and (3) it is seen that the degree of oxidation of carbon monoxide to carbon dioxide at all the temperatures and oxygen concentrations studied decreases with increasing space velocity of the gas stream. As a rule, the degree of oxidation falls considerably more slowly than the space velocity increases; this is especially the case at higher temperatures. Indeed, for example, at an initial oxygen content of 0.3% and a temperature of 55°C, an increase in space velocity from 521 to 9300, i.e., by a factor of 17.9, lowers the degree of oxidation correspondingly from 91 to 13.9%, i.e., only by a factor of 6.55. With increasing temperature, the decrease in the degree of oxidation as the space velocity increases proceeds still more slowly, which is characterized by a flatter arrangement of the isotherms (Figs. 2 and 3) at high temperatures.

In the temperature interval studied by us (from 25 to 135°C), the degree of oxidation of CO to CO₂ increases with increasing temperature at all space velocities, both at an initial O₂ content of 0.3 and of 0.5% (vol.) (Figs. 4 and 5). Thus, if at an initial oxygen concentration of 0.3% and a space velocity of 3000 (Fig. 4) the degree of oxidation at a temperature of 50°C is 28.4%, then at a temperature of 115°C the analogous value increases to 87.9%. It should be noted that the most intensive increase in the activity of hopcalite occurs in the temperature region 75–95°C. At these temperatures the curves (degree of oxidation) \(v = f(t^\circ C)\) rise especially steeply. A further increase in temperature above 95°C when a gas containing 0.3% O₂ is passed through hopcalite leads only to a slight increase in the degree of oxidation, which, apparently, is explained by partial poisoning of the catalyst by water vapor formed as a result of hydrogen oxidation.

According to studies by previous investigators, an increase in the initial O₂ concentration above the stoichiometric value only slightly increases the degree of oxidation. Thus, Kazarnovskii, Borshchevskii, and Kostina found that increasing the excess oxygen, relative to the stoichiometric amount necessary for complete oxidation of CO to CO₂, from 0 to 100% at a space velocity of 100 raises the degree of oxidation only from 90 to 96% (determined by us from Fig. 3 of the cited work of the indicated ...

authors). Comparing with one another the results obtained by us at identical temperatures and space velocities but different initial oxygen concentrations (Table 2), one may note that, indeed, in those cases where the degree of oxidation of CO to CO₂ at 0.3% O₂ is high (80–90%), increasing the oxygen concentration to 0.5% (i.e., a 100% excess) increases the degree of oxidation only slightly; however, under all other conditions (low temperatures or high space velocities), when the degree of oxidation at O₂—0.3% is low, increasing the initial oxygen concentration to 0.5% very substantially increases the degree of oxidation. Thus, for example, if at \(V = 520\) and \(t = 50^\circ\mathrm{C}\) increasing the initial oxygen concentration from 0.3 to 0.5% raises the degree of oxidation of CO to CO₂ from 91.7 to 96.6, then at the same temperature but at a space velocity of 7500 the degree of oxidation at initial oxygen concentrations of 0.3 and 0.5% proves to be, respectively, 15.7 and 34.4%, i.e., an increase by a factor of 2.

Fig. 2. Dependence of the degree of oxidation of carbon monoxide to carbon dioxide on space velocity. Initial concentrations: CO—0.5 and O₂—0.3% (by volume).

The data presented in this communication were obtained on the same hopcalite over 2.5 months of periodic operation (at night the investigation was discontinued because of the working conditions in the laboratory), and during this time no substantial change in its activity occurred.

The process of catalytic oxidation of carbon monoxide by oxygen may be represented by the following stoichiometric equation:

\[ 2\mathrm{CO} + \mathrm{O}_2 = 2\mathrm{CO}_2. \]

Therefore, it is of primary interest to know whether this equation is justified kinetically under the conditions studied. To answer this question, the reaction-rate constants were calculated from the experimental data using the integral form of the following differential equation:

\[ -\frac{dC_{\mathrm{co}}}{dt} = K C_{\mathrm{co}}^{2} C_{\mathrm{O}_2}, \tag{1} \]

and it turned out that, for a constant temperature, the rate constants \((K)\), calculated from the experimental data, do not remain constant quantities, but show a steady tendency toward strong growth as the space velocity increases. Thus, the kinetic relations in the catalytic selective oxidation of carbon monoxide in the presence of hydrogen cannot be expressed by equation (1). On the other hand, the results obtained by us do not fit the kinetic equation of zero order

\[ -\frac{dC_{\mathrm{co}}}{dt} = K, \tag{2} \]

the validity of which for the oxidation of carbon monoxide on hopcalites (composition 70% MnO₂ and 30% CuO, and the American grade MSA) in the absence of hydrogen was demonstrated by Shurmovskaya and Bruns¹³).

Fig. 3. Dependence of the degree of oxidation of carbon monoxide to carbon dioxide on the space velocity. Initial concentrations: CO—0.5 and O₂—0.5% (by volume).

Fig. 4. Dependence of the degree of oxidation of carbon monoxide to carbon dioxide on temperature. Initial concentrations: CO—0.5 and O₂—0.3%.

Fig. 5. Dependence of the degree of oxidation of carbon monoxide to carbon dioxide on temperature. Initial concentrations: CO—0.5 and O₂—0.5% (by volume).

We assumed that, in the presence of the hopcalite studied by us, the rate of oxidation of carbon monoxide to carbon dioxide is directly proportional to the concentrations, or partial pressures, of carbon monoxide and oxygen, and that the reaction is not inhibited by the carbon dioxide formed, i.e., the following kinetic equation is valid:

\[ -\frac{dC_{\mathrm{CO}}}{dt}=K C_{\mathrm{CO}}\cdot C_{\mathrm{O_2}}. \tag{3} \]

The latter assumption concerning the absence of an inhibiting effect of \(\mathrm{CO_2}\) on the rate of the catalytic oxidation of CO is in agreement with the work of Whitesell and Frazer\(^{14}\) and of Elovich and Kondorf\(^{15}\).

Denoting by \(A\) and \(B\) the initial amounts, respectively, of carbon monoxide and oxygen in the original gas mixture, and by \(x\) the amount of CO that has reacted by reaction time \(t\), one may write

\[ \frac{dx}{dt}=K(A-x)\left(B-\frac{1}{2}x\right). \tag{4} \]

For integration, we rewrite equation (4) in the following form:

\[ \frac{2\,dx}{(A-x)(2B-x)}=K\,dt, \]

after integration we obtain

\[ Kt=\frac{2}{2B-A}\ln\frac{2B-x}{A-x}+\mathrm{Const}. \]

At the initial moment of the reaction \(t=0\), and consequently \(x\) is also equal to zero, whence

\[ \mathrm{Const}=-\frac{2}{2B-A}\ln\frac{2B}{A}; \]

substituting the value of \(\mathrm{Const}\) and replacing natural logarithms by common logarithms, we shall have

\[ Kt=\frac{2\cdot 2.303}{2B-A}\log\frac{(2B-x)A}{(A-x)2B}. \tag{5} \]

Equation (5) can be expressed through the degree of oxidation of carbon monoxide \((\alpha)\); \(\alpha=\frac{x}{A}\), where \(A\) and \(x\) have the meanings given above.

After substituting the value of \(x\), equation (5) assumes the following form:

\[ K=\frac{4.606}{t(2B-A)}\log\left(\frac{1-\frac{A}{2B}\alpha}{1-\alpha}\right). \tag{6} \]

To calculate the rate constant \(K\) from equation (6), it is necessary to know the reaction time \(t\). However, under the conditions of the dynamic method of investigation used by us and with the catalyst in lump form, this quantity cannot be determined exactly; therefore, as was already shown earlier by one of us (N. P. Kurin\(^{16}\)), instead of the reaction time \(t\) in equation (6) one may substitute the reciprocal of the volumetric velocity, first divided by the temperature correction that accounts for the expansion of the gas with temperature, since \(t\) is proportional to

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

Ultimately, the equation, by

to which the calculations were performed after the indicated substitution, had the following form

\[ K = V \frac{T}{273} \cdot \frac{4.606}{(2B - A)} \log \left( \frac{1 - \frac{A}{2B}\alpha}{1 - \alpha} \right). \tag{7} \]

In calculating the rate constants, the degree of oxidation of carbon monoxide \((\alpha)\), in fractions of unity, and the initial concentrations of CO and \(O_2\), in volume percent, were substituted into equation (7).

The rate constants calculated from equation (7) at various temperatures and space velocities are given in Table 2.

It is apparent from Table 2 that the constancy, required by equation (7), of the rate constants calculated at constant temperatures but different space velocities is, as a rule, observed quite satisfactorily, if one takes into account the difficulty of accurately measuring the rates of chemical reactions.

In addition, the correctness of kinetic equations (3) and (7), which imply first apparent order with respect to oxygen, apart from the constancy of the rate constants at different space velocities, is also confirmed by the satisfactory agreement of the constants calculated at constant temperatures but different initial oxygen concentrations.

Thus, from the foregoing it is evident that the overall apparent order of the reaction of oxidation of carbon monoxide in the presence of hydrogen on the hopcalite studied by us is equal to 2, and the rate of oxidation is directly proportional to the partial pressures or concentrations of carbon monoxide and oxygen. However, in order to resolve unambiguously the question of the reaction order with respect to carbon monoxide, it is necessary to investigate the dependence of the half-time of oxidation of CO to \(CO_2\) on the initial pressure of CO.

As is known, reliable methods for calculating the activation energies of chemical reactions from the most general properties of the reacting molecules are still lacking. Therefore, to determine this quantity one usually uses the temperature dependence of the rate constants found experimentally, expressed by the Arrhenius equation, which in integral form has the following form:

\[ \log K = -\frac{E}{2.303 \cdot R}\left(\frac{1}{T}\right) + \mathrm{Const}. \tag{8} \]

Equation (8), at constant \(E\), is the equation of a straight line; the tangent of the angle of inclination of this line to the \(\left(\frac{1}{T}\right)\) axis, multiplied by \(2.303 \cdot R\), should give the desired value of the activation energy. The indicated plots, according to the data of Table 2, are presented in Fig. 6.

In Fig. (6), the lines Ia—Ib and IIa—IIb express the relation between \(\log K\) and \(\left(\frac{1}{T}\right)\) for the reaction of oxidation of CO to \(CO_2\) in the presence of hydrogen on hopcalite, respectively at 0.3 and 0.5% oxygen in the initial gas mixture. From this figure it is seen that, when plotting \(\log K\) against \(\left(\frac{1}{T}\right)\), straight lines are indeed obtained; however, they have a sharp break at a temperature of \(75^\circ C\), and the portions of the straight lines (Ib and IIb) corresponding to elevated temperatures \((75—115^\circ C)\) are situated at a smaller angle of inclination to the \(\left(\frac{1}{T}\right)\) axis than is the case at lower tempera-

types (Ia and IIa). From the tangents of the angles of inclination of the straight lines (found graphically from Fig. 6), Ia, Ib, IIa, and IIb, the apparent activation energies were calculated; they proved to be equal for the low-temperature region (25–75°C) to 6550 and 6860, and for the elevated-temperature region (75–115°C) to 16500 and 17300 cal/mole.

Table 2

Rate constants of the reaction of oxidation of carbon monoxide by oxygen in the presence of a nitrogen–hydrogen mixture on hopcalite, calculated according to equation (7).

| Temperature, °C | Reciprocal of absolute temperature | Volumetric gas velocity \(V\) | \multicolumn{3}{c}{Rate constants at initial concentration CO—1.5% (vol.) and O\(_2\); 0.3% (vol.)} | \multicolumn{3}{c}{Rate constants at initial concentration CO—1.5% (vol.) and O\(_2\); 0.5% (vol.)} |
|---:|---:|---:|---:|---:|---:|---:|---:|---:|
| | | | \(K\) | \(K_{\text{mean}}\) | \(\log K_{\text{mean}}\) | \(K\) | \(K_{\text{mean}}\) | \(\log K_{\text{mean}}\) |
| 25 | 0.003356 | 520 | 1780 | 1860 | 3.270 | 7060 | 1900 | 3.279 |
| 25 | 0.003356 | 720 | 1920 | 1860 | 3.270 | 1990 | 1900 | 3.279 |
| 25 | 0.003356 | 1000 | 1680 | 1860 | 3.270 | 2040 | 1900 | 3.279 |
| 25 | 0.003356 | 2000 | 2040 | 1860 | 3.270 | 1670 | 1900 | 3.279 |
| 50 | 0.003096 | 520 | 17900 | 4570 | 3.660 | 6720 | 5840 | 3.766 |
| 50 | 0.003096 | 720 | 2645 | 4570 | 3.660 | 4480 | 5840 | 3.766 |
| 50 | 0.003096 | 1000 | 5020 | 4570 | 3.660 | 4110 | 5840 | 3.766 |
| 50 | 0.003096 | 2000 | 4060 | 4570 | 3.660 | 5250 | 5840 | 3.766 |
| 50 | 0.003096 | 3000 | 4130 | 4570 | 3.660 | 6550 | 5840 | 3.766 |
| 50 | 0.003096 | 5000 | 4830 | 4570 | 3.660 | 7960 | 5840 | 3.766 |
| 50 | 0.003096 | 7500 | 5200 | 4570 | 3.660 | 5810 | 5840 | 3.766 |
| 50 | 0.003096 | 9300 | 4140 | 4570 | 3.660 | | 5840 | 3.766 |
| 75 | 0.002873 | 520 | — | 8680 | 3.939 | 9940 | 9940 | 3.997 |
| 75 | 0.002873 | 720 | 11500 | 8680 | 3.939 | 5690 | 9940 | 3.997 |
| 75 | 0.002873 | 1000 | 6470 | 8680 | 3.939 | 5390 | 9940 | 3.997 |
| 75 | 0.002873 | 2000 | 5760 | 8680 | 3.939 | 9020 | 9940 | 3.997 |
| 75 | 0.002873 | 3000 | 5870 | 8680 | 3.939 | 11070 | 9940 | 3.997 |
| 75 | 0.002873 | 5000 | 8590 | 8680 | 3.939 | 15850 | 9940 | 3.997 |
| 75 | 0.002873 | 7500 | 11100 | 8680 | 3.939 | 16750 | 9940 | 3.997 |
| 75 | 0.002873 | 9300 | 11500 | 8680 | 3.939 | 5840 | 9940 | 3.997 |
| 95 | 0.002717 | 2000 | 41700 | 38700 | 4.588 | — | 47700 | 4.679 |
| 95 | 0.002717 | 3000 | 33500 | 38700 | 4.588 | 59190 | 47700 | 4.679 |
| 95 | 0.002717 | 5000 | 31100 | 38700 | 4.588 | 74100 | 47700 | 4.679 |
| 95 | 0.002717 | 7500 | 41600 | 38700 | 4.588 | 46310 | 47700 | 4.679 |
| 95 | 0.002717 | 9300 | 45600 | 38700 | 4.588 | 37480 | 47700 | 4.679 |
| 115 | 0.002577 | 2000 | 82100 | 70500 | 4.848 | | | |
| 115 | 0.002577 | 3000 | 68800 | 70500 | 4.848 | | | |
| 115 | 0.002577 | 5000 | 56700 | 70500 | 4.848 | | | |
| 115 | 0.002577 | 7500 | 79500 | 70500 | 4.848 | | | |
| 115 | 0.002577 | 9300 | 65400 | 70500 | 4.848 | | | |
| 135 | 0.002451 | 2000 | 77300 | 82800 | 4.915 | | | |
| 135 | 0.002451 | 3000 | 89500 | 82800 | 4.915 | | | |
| 135 | 0.002451 | 5000 | 97200 | 82800 | 4.915 | | | |
| 135 | 0.002451 | 7500 | 79800 | 82800 | 4.915 | | | |
| 135 | 0.002451 | 9300 | 68400 | 82800 | 4.915 | | | |

Turning to the discussion of the question of the mechanism of the catalytic oxidation of carbon monoxide on hopcalite in the presence of hydrogen, one may first of all imagine that the oxidation of CO to CO₂ occurs as a result of: 1) collision of carbon monoxide and oxygen molecules adsorbed on the surface of the catalyst; 2) impact, from the gas phase, of an oxygen molecule upon an adsorbed carbon monoxide molecule; 3) impact of a carbon monoxide molecule upon an adsorbed oxygen molecule; and, finally, 4) the reaction may proceed through the formation and decomposition of more or less stable surface compounds formed from the initial gaseous components and the substance of the catalyst. All these mechanisms presuppose, above all, the presence of adsorption by hopcalite of one or both participants in the reaction. Benton¹⁷) measured the adsorption of CO₂, CO, N₂, O₂, and H₂ on such oxide catalysts as Co₂O₃, hopcalite, CuO, MnO₂, Fe₂O₃, V₂O₅, and SiO₂, and found that there are two types of adsorption, “primary” and “secondary,” with the arrangement of the indicated substances in descending order of their catalytic activity agreeing with a decrease in the magnitude of the primary adsorption. On this basis he gives the following scheme for the catalytic oxidation of CO to CO₂ in the presence of metal oxides:

Fig. 6. Effect of temperature on the rate constant of the reaction of oxidation of CO to CO₂ on hopcalite in the presence of a nitrogen–hydrogen mixture.

\[ \begin{aligned} 1.\quad &>MO + CO \rightleftarrows >MO\!\cdot\!CO \rightleftarrows >MOCO \rightleftarrows >M\!\cdot\!OCO \rightleftarrows M + CO_2,\\ 2.\quad &2 > M + O_2 \rightleftarrows 2 > MO, \end{aligned} \]

where \(>MO\) and \(>M\) are, respectively, active portions of the surface of the metal oxide and of the metal; i.e., according to Benton, the surface of the metal oxide is alternately reduced by carbon monoxide and oxidized by oxygen, which agrees with the views of Jones and Taylor¹⁸) regarding the oxidation of CO to CO₂ in the presence of copper oxide.

The concept of alternate reduction and oxidation of the catalyst substance in the catalytic oxidation of carbon monoxide on Fe₂O₃ is also supported by Eckell¹⁹). In the latter’s opinion, the oxidation of CO to CO₂ in the presence of Fe₂O₃ proceeds through an intermediate compound obtained as a result of the reduction of Fe₂O₃ by carbon monoxide (apparently Fe₃O₄), followed by oxidation of this complex by oxygen to Fe₂O₃. In the opinion of Kobozev and Ivanov²⁰), the catalytic oxidation of carbon monoxide in the presence of Fe₂O₃ proceeds through oxidation by oxygen of a phase intermediate compound of the iron oxycarbonyl type \([(Fe_xO_y)(CO)_z]\), the determining stages of the process being the formation and oxidation of this compound. Frazer²¹), Neumann, Kroger, and Iwanowski²²), likewise explaining the process of oxidation of CO to CO₂ by alternate reduction and oxidation of oxide catalysts, assert that the catalytic activity of oxides is the greater, the higher the dissociation elasticity of the oxide with respect to oxygen. However, there are a number of authors who consider it unlikely that the catalytic oxidation of CO to CO₂ proceeds through intermediate oxides and carbonyls. Thus, for example, Bone and Andrew²³) assert that on the surface of gold, as well as on the surface of nickel, copper, and their oxides, an activated layer of oxygen is formed, which reacts with likewise activated molecules of carbon monoxide.

A systematic study of the process of oxidation of carbon monoxide by atmospheric oxygen in the presence of \(\mathrm{MnO_2}\) was carried out by Roginskii and Zel’dovich\(^{24}\), Elovich and Kharakhorina\(^{25}\), Elovich and Roginskii\(^{26}\), Elovich and Korndorf\(^{27}\), and Elovich and Kachur\(^{28}\). On the basis of experimental study of the sorption of CO and \(\mathrm{O_2}\) on \(\mathrm{MnO_2}\), as well as of the kinetics of oxidation of CO to \(\mathrm{CO_2}\), the indicated group of investigators proposed a scheme for the catalytic oxidation of CO to \(\mathrm{CO_2}\) in the presence of \(\mathrm{MnO_2}\), according to which the limiting stage of the process is the stage of activated adsorption of carbon monoxide; moreover, this is confirmed mainly by the good agreement between the activation energies of the oxidation reaction and of the activated adsorption. Thus, for example, according to the calculations of Kharakhorina and Elovich, the activation energy of activated adsorption of CO on \(\mathrm{MnO_2}\) is 6225 cal; the analogous value for activated adsorption and the oxidation reaction of CO to \(\mathrm{CO_2}\), according to Elovich and Kachur, is 6860 cal.

In considering the mechanism of the catalytic oxidation of carbon monoxide in the presence of hydrogen on the two-component hopcalite studied by us, attention should first of all be paid to the character of the arrangement of the straight lines in Fig. 6. The straight lines \(\log K = f\left(\frac{1}{T}\right)\) at a temperature of \(75^\circ\mathrm{C}\) have a sharp break, the latter being directed downward, which, as is known\(^{29}\), indicates the parallel course of two processes. In this connection, at low temperatures \((25—75^\circ\mathrm{C})\), the catalytic oxidation of CO to \(\mathrm{CO_2}\) in the presence of the hopcalite studied by us apparently proceeds according to the following scheme:

\[ \begin{aligned} 1.\;& \mathrm{MnO_2 + CO \rightleftharpoons MnO_2CO},\\ 2.\;& \mathrm{MnO_2CO + O_2 \rightleftharpoons MnO_2O + CO_2},\\ 3.\;& \mathrm{MnO_2O + CO \rightleftharpoons MnO_2 + CO_2}. \end{aligned} \]

This scheme, which assumes, in accordance with the views of Roginskii and Zel’dovich, the presence of activated adsorption of carbon monoxide on \(\mathrm{MnO_2}\) (reaction 1), is confirmed by the fact that the activation energies obtained by us, 6550 and 6860 cal, are in good agreement with the analogous values cited above for activated adsorption and the oxidation reaction of CO on pure \(\mathrm{MnO_2}\). It should be pointed out that the value of the activation energy found by us also coincides with the activation energy calculated by Roginskii from the data of Lamb and Veil (6700 cal), and also with the activation energy obtained for two-component hopcalite \((70\%\,\mathrm{MnO_2} + 30\%\,\mathrm{CuO})\) and MSA-type hopcalite by Shurmovskaya and Bruns (5210 and 6320 cal, respectively).

As regards the oxidation of CO to \(\mathrm{CO_2}\) on the hopcalite studied by us in the region of higher temperatures \((75—135^\circ\mathrm{C})\), under these conditions the mechanism of oxidation is somewhat different than at low temperatures. Apparently, at elevated temperatures the catalytic oxidation of CO to \(\mathrm{CO_2}\) proceeds according to the following reactions:

\[ \begin{aligned} 1.\;& \mathrm{MnO_2 + CO \rightleftharpoons MnO_2CO},\\ 2.\;& \mathrm{MnO_2CO \rightleftharpoons MnCO_3},\\ 3.\;& \mathrm{MnCO_3 \rightleftharpoons MnO + CO_2},\\ 4.\;& \mathrm{2MnO + O_2 \rightleftharpoons 2MnO_2}. \end{aligned} \]

According to this scheme, a molecule of carbon monoxide, striking the surface of \(\mathrm{MnO_2}\), as a result of activated adsorption gives the adsorption complex \(\mathrm{MnO_2CO}\), in which the bonds of the atoms in the initial components \(\mathrm{MnO_2}\) and CO are not destroyed. Then, in this complex, the original bonds between the atoms are broken and a new chemical compound, manganese carbonate, is formed, which then dissociates with liberation of the final product

CO₂ and MnO. MnO is again oxidized by adsorbed atmospheric oxygen to MnO₂. The most difficult stage in this mechanism is apparently the dissociation of MnCO₃ into MnO and CO₂. This reaction is endothermic, and according to our calculations, based on Hess’s law, its heat effect is equal to 23400 cal.

If one takes into account that the activation energy of an endothermic reaction should be approximately equal to its heat effect, then the activation energy we found for this temperature region (16500–17300 cal) is somewhat at variance with the heat effect of the reaction. However, in reality the agreement between the heat effect and the activation energy is apparently better, since the carbon dioxide formed as a result of the dissociation of MnCO₃ is initially in the adsorbed state; therefore, for the decomposition of MnCO₃ it is necessary to expend not 23400 cal, but an amount smaller by the heat of adsorption of CO₂.

If the heat of adsorption is taken to be 5000 cal, then only 18400 cal will have to be expended for the decomposition of MnCO₃, which is in satisfactory agreement with the value of E found by us and, consequently, confirms the correctness of the assumption made regarding the mechanism of oxidation of CO to CO₂ in the presence of hopcalite.

The proposed schemes of the mechanism also agree with kinetic equation (3), which indicates that the rate of catalytic oxidation of CO to CO₂ is directly proportional to the concentrations of carbon monoxide and oxygen in the first degree. This can be explained by the fact that the adsorption of CO and O₂ on the active surface of hopcalite is small, and therefore the concentrations of these substances in the adsorption layer will be directly proportional to their partial pressures in the gas phase.

As for the second component of hopcalite—CuO—according to the study of Bessalov and Kobozev, up to 120°C carbon monoxide in the presence of pure powdered copper oxide, and also on the considerably more active precipitated aerosol of CuO, is practically not oxidized. According to the study of the authors mentioned, the activation energy of the process of oxidation of CO by atmospheric oxygen on powdered CuO in the temperature range 150–300°C is equal to 3200 cal, and at 350–390°C—40000 cal; for the precipitated aerosol the analogous value at 125–170°C is equal to 32200 cal. As can be seen, these activation energies differ considerably from the values found by us, and therefore it may be assumed that the role of CuO in hopcalite is reduced to activation of MnO₂ and stabilization of the active surface. However, if one takes into account the possibility of mutual activation of the components constituting the catalyst, then it is quite possible, especially at elevated temperatures, that the oxidation of CO to CO₂ also partly occurs on CuO by reactions analogous to those for MnO₂.

As a number of investigators have shown³), ⁴), ⁵), ⁶), when a mixture containing CO, H₂, and O₂ is passed over the catalyst, along with the oxidation of carbon monoxide, hydrogen may also be oxidized with the formation of water; the latter poisons hopcalite. In our experiments at low temperatures no formation of water was observed. However, at higher temperatures, 95–115°C, appreciable oxidation of hydrogen occurred, and the water vapor formed, being adsorbed on the active surface of hopcalite, partially poisoned the catalyst; this apparently explains the deviation of the value of log K from the straight line

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

(Fig. 6) at temperatures of 115°C and 135°C.

CONCLUSIONS

1. An experimental apparatus has been described for investigating the selective catalytic oxidation of carbon monoxide to CO₂ in the presence of hydrogen by the dynamic method under conditions close to isothermal.

1. As a catalyst, the activity of a two-component hopcalite of composition 60% \( \mathrm{MnO_2} + 40\% \ \mathrm{CuO} \) was studied by passing through the catalyst a dry nitrogen–hydrogen mixture \((\mathrm{N_2} + 3\mathrm{H_2})\), containing 0.5% CO and 0.3 and 0.5% (by volume) \( \mathrm{O_2} \), at temperatures of 25–135°C and space velocities from 520 to 9300 volumes of gas mixture \((0^\circ\mathrm{C}\) and 760 mm Hg) per 1 bulk volume of catalyst per hour.

2. It has been established that, with increasing gas space velocity \((t^\circ = \mathrm{Const})\), the degree of oxidation decreases; however, this decrease occurs more slowly than the space velocity increases.

3. With increasing temperature and constant gas velocity, the degree of oxidation of CO to \( \mathrm{CO_2} \) increases; a particularly intense increase in the degree of oxidation is observed in the temperature range 75–95°C.

4. Increasing the oxygen concentration in the initial gas mixture increases the degree of oxidation of CO to \( \mathrm{CO_2} \). At a temperature of 95°C, with a 100% excess of oxygen relative to the amount required for complete oxidation of CO and a space velocity of 5000, the degree of oxidation is 97%.

5. The rate of the reaction of oxidation of CO to \( \mathrm{CO_2} \) in the presence of a nitrogen–hydrogen mixture is directly proportional to the concentrations of CO and \( \mathrm{O_2} \) in the first degree and is expressed by the following kinetic equation:

\[ -\frac{d C_{\mathrm{co}}}{dt} = K C_{\mathrm{co}} \cdot C_{\mathrm{o2}} . \]

In the temperature range 25–75°C, for initial \( \mathrm{O_2} \) concentrations of 0.3 and 0.5%, the apparent activation energies are close to one another and are, respectively, 6500 and 6860 cal. However, in the temperature range 75–115°C the activation energy rises sharply to 16500 and 17300 cal.

1. On the basis of the data obtained, a scheme is given for the mechanism of the catalytic oxidation of CO to \( \mathrm{CO_2} \) in the presence of hopcalite.

REFERENCES

1. Epstein.—Chemistry and Technology of Fixed Nitrogen, ONTI, 1935.
2. Curtis.—Fixed Nitrogen, ONTI, 1934.
3. Tarry.—English patent, 127609, 1917.
4. Rideal and Taylor.—Analyst, 44, 89, 1919.
5. Rideal.—J. Chem. Soc. L., 993, 1919.
6. Lamb, Scalione and Edgar.—J. Am. Chem. Soc., 44, 738, 1922.
7. Lamb, Bray and Frazer.—J. Ind. Eng. Chem., 12, 213, 1920.
8. Kuentzel, J. Am. Chem. Soc., 52, 437, 1930.
9. Kuentzel, J. Am. Chem. Soc., 52, 445, 1930.
10. Kazarnovskii, Borshchevskii and Kostin.—Zhurn. Khim. Prom., 4, 41, 1938.
11. Lidov.—Gas Analysis, 143, ed. 1928.
12. Vanino.—“Handbuch der Präparativen Chemie,” pp. 505 and 695. Stuttgart, 1925.
13. Shurmovskaya and Bruns.—Zhurn. Fiz. Khim., 9, 3, 301, 1937.
14. Whitesell and Frazer, J. Am. Soc., 45, 2347, 1928.
15. Elovich and Korndorf.—Zhurn. Obshch. Khim., 9, 8, 673, 1939.
16. Kurin.—Zhurn. Khim. Prom., 7, 475, 1937.
17. Benton.—J. Am. Chem. Soc., 45, 887, 900, 1923.
18. Jones and Taylor.—J. Phys. Chem., 27, 623, 1923.
19. Eckell.—Z. Elektrochem., 39, 807, 1933.
20. Kobozev and Ivanov.—Zhurn. Fiz. Khim., 10, 1, 1, 1937.
21. Frazer.—J. Phys. Chem., 35, 405, 1931.
22. Neumann, Kroger u. Iwanowski.—Z. Elektrochem., 37, 121, 1931.
23. Böne and Andrew.—Proc. Roy. L., A, 110, 16, 1926; A, 109.
24. Roginsky and Zeidowitsh.—Acta Phys. Chim., 1, 554; 1, 595, 1934.
25. Charachorin and Elowitz.—Acta Phys. Chim., 5, 325, 1936.
26. Elowitz and Roginsky.—Acta Phys. Chim., 7, 296, 1937.
27. Elovich and Korndorf.—Zhurn. Obshch. Khim., 9, 8, 673, 1939.
28. Elovich and Kachur.—Zhurn. Obshch. Khim., 9, 8, 714, 1939.
29. Hinshelwood.—Kinetics of Gas Reactions, ONTI, 1932.
30. Lamb and Veil, J. Am. Chem. Soc., 47, 123, 1925.