Investigation of the Catalytic Properties of Non-Platinum Catalysts in the Oxidation of SO₂ to SO₃

Communication 1

Activity of Iron–Copper Catalysts

N. P. Kurin, S. A. Sigov, G. I. Semenova, and M. N. Vorob’eva.

As is known, at atmospheric pressure the reaction

\[ 2\mathrm{SO}_2 + \mathrm{O}_2 \rightleftharpoons 2\mathrm{SO}_3 \]

in the gas phase, in the temperature region in which the equilibrium is favorably shifted toward the formation of sulfuric anhydride, proceeds at an immeasurably low rate if the reacting molecules \(\mathrm{SO}_2\) and \(\mathrm{O}_2\) are not activated by any other forms of energy besides thermal energy (e.g., electrical discharges). The reasons for such reaction inertness lie in the fact that the atoms in the \(\mathrm{SO}_2\) and \(\mathrm{O}_2\) molecules are bound by relatively large affinity forces. The introduction of the necessary amounts of energy to overcome these forces by heating is in principle possible, but this would lead to that temperature region in which the occurrence of the reaction

\[ 2\mathrm{SO}_2 + \mathrm{O}_2 \rightleftharpoons 2\mathrm{SO}_3 \]

becomes thermodynamically impossible.

Activation of molecules by other forms of energy, for example electrical energy (silent and spark discharges), ultrasonic energy, may cause the reaction to proceed at low temperatures, but the practical results obtained in activating a number of other chemical reactions by these methods are still small, since the energy consumption per unit of product remains too high. The catalytic route proved to be the most fruitful. The use of homogeneous (nitrogen oxides) and heterogeneous catalysts, which reduce or split into parts the energy required in a noncatalytic process for activation, made it possible, at sufficiently high rates and on a vast technical scale, to carry out the process of obtaining sulfuric acid. Moreover, in the most recent period in the world sulfuric-acid industry there has been a shift toward the production of contact sulfuric acid, used, in particular, in the military industry. Studies by various authors (a review of works with indication of authors \(^{1}\)) established that many substances may serve as solid catalysts for accelerating the reaction

\[ 2\mathrm{SO}_2 + \mathrm{O}_2 = 2\mathrm{SO}_3, \]

however, among them platinum and various vanadium compounds are considered the most active. Platinum catalysts are expensive and very sensitive to contact poisons, which leads to large expenditures for preliminary purification of the gas; vanadium catalysts are less sensitive to poisons, but, like platinum, are relatively expensive. Therefore the problem of finding cheaper and sufficiently active catalysts other than platinum is of substantial interest both for theory and for practice.

On the basis of the foregoing, we undertook systematic studies of the activity of non-platinum catalysts with an iron base.

The first observation from which it was possible to draw a conclusion about the catalytic activity of iron oxide was made by Blondeau²). This investigator, passing a mixture of SO₂ with air through a porcelain tube filled with sand containing an admixture of clay and iron, observed the formation of small quantities of sulfuric acid; however, he carried out no further experiments in this direction. Wöhler and Mahla³) were the first to set up special experiments to reveal the catalytic ability of the oxides of iron, copper, and chromium. These authors found that pure oxides of iron, copper, and chromium, heated to a faint red heat, or co-precipitated oxides of the latter two metals, cause intensive formation of SO₃ when a mixture of sulfurous gas and air is passed through them. The works of Wöhler and Mahla aroused a certain interest in the oxides of iron and copper as catalysts for the contact process of obtaining sulfuric acid; a number of patents were filed (see, for example, 4). However, iron oxide in the form of pyrite cinders found industrial application only at the end of the last century in the method of the Society of Chemical Plants in Mannheim, published in 1898.

At the beginning of the present century, alongside patents, there appeared a whole series of research works aimed at clarifying the conditions of the catalytic oxidation of sulfurous gas in the presence of iron oxide. One of the first works devoted specifically to this question is the work of Lunge and Pollit⁵). In these experiments it was found that, when pure iron oxide is used, the maximum conversion lies at 620°C and amounts to 47.5%; below 600°C the yields are insignificant. When pyrite cinders were used and the gas-flow rate was only 25 cm³/min, with an SO₂ content of 5% in the initial gas, the maximum achieved was 46.9%. In a more recent work Lunge and Reinhardt,⁶) carrying out more careful temperature control while passing through pure iron oxide a gas mixture containing 2.4% SO₂ at a gas-flow rate of 25 ml/min (temperature 625°C), obtained a maximum conversion of 70.5%; moreover, beginning already at 400°C, a considerable increase in catalyst activity was observed. Knietsch⁷) also studied the action of iron oxide as a catalyst. According to the cited—

in Knietsch’s diagram cited in the work, the maximum conversion, amounting to only 47%, was achieved at a temperature of 550°C and a gas flow rate of 500 ml/min.

With the aim of elucidating the mechanism of the process of catalytic oxidation of sulfur dioxide by oxygen in the presence of iron oxide, Keppler 8), d’Ans 9), L. Wöhler and Pluddemann, and Wöhler 10) undertook special investigations. The principal leitmotif of these works was to test the applicability of the theory of intermediate compounds to the given catalytic process. They established the conditions under which the reacting molecules of sulfur dioxide and oxygen, entering into interaction with iron oxide, give sulfates of ferrous and ferric iron, which may be regarded as intermediate compounds in the catalytic process. On the basis of the temperature dependence of the dissociation pressure of these compounds, the optimal yields that could be obtained on iron oxide were determined. According to these investigators, the temperature optima lay, as a rule, above 600°C, and therefore the optimal yields of the reaction of oxidation of SO₂ to SO₃ should not exceed values of the order of ∼74%.

Among the works devoted to the study of iron catalysts carried out recently, the works of Neumann and Goebel 11), Neumann and Güttner 12), Luk’yanov with co-workers 13), and Kuz’minykh and Turkhan 14) should be noted.

Neumann and Goebel investigated the activity both of pure iron oxide and in the presence of various activators (oxides of bismuth, potassium, strontium, and tin) in the temperature interval 500–700°C and at gas-mixture flow rates of 150 and 300 ml/min. On pure iron oxide, with 7% SO₂ content in the gas, the optimum degree of conversion was achieved at 625°C and was 69.5% (gas flow rate 150 ml/min). The introduction into iron oxide of 4% bismuth oxide had a favorable effect on the activity of the iron oxide. The optimum degree of conversion, also observed at 625°C, was already 72.2%, i.e., 94.6% of the equilibrium value. Potassium oxide, introduced into iron oxide in an amount of 40%, strongly impaired the catalytic properties of iron (maximum yield at \(t\)-re 700°C—52.1%), whereas strontium oxide only slightly improved them (optimum temperature 600°C, maximum degree of conversion 71.2%). A catalyst obtained by mixing molecular quantities of the oxides of iron and tin proved to be considerably more active; the maximum degree of oxidation, observed at a temperature of 600°C, was equal to 76.2%. Arsenic oxides also proved capable of activating iron oxide.

The results of the work of Neumann and co-workers indicate the fundamental possibility of increasing the activity of iron catalysts by introducing promoters; however, the results obtained by them are not very large. It may be that the most active forms of the catalysts, owing to the small amount

of the compositions investigated for this pair of oxides (for example, iron oxide + activating oxide) escaped the attention of investigators. The number of activators studied is also small and, finally, the influence of the process of preparing iron-based catalysts on their final activity has not been the subject of detailed investigations. The catalytic properties of pure copper oxide were investigated by Neumann[^15]; it was found that this substance is a weak catalyst in the process of oxidizing \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) by oxygen. The maximum percentage of oxidation, observed at a temperature of \(700^\circ\mathrm{C}\), was \(58.73\%\). At a temperature of \(450^\circ\mathrm{C}\), however, the corresponding value was only \(0.14\%\). Despite the low activity of copper oxide, Lunpre and Pollitt[^5] had already indicated in their investigations that an admixture of copper oxide to iron oxide (for example, the use of pyrite cinders containing copper) increases the catalytic action of the cinders. Keppeler[^8] showed that copper oxide is reduced considerably more readily by sulfur dioxide and in this way explained the activating effect of \(\mathrm{CuO}\) on iron oxide.

From the works cited it is evident that, despite the interest shown in iron oxide as a catalyst, its catalytic properties, especially in the presence of activators, have been studied very little. Therefore, in order to fill the existing gap in this field of knowledge, we set ourselves the aim of carrying out systematic investigations of the activity of iron catalysts promoted by various activators. In the present communication we give the results that we obtained in studying catalysts consisting of mixtures of iron and copper oxides, mixed in various ratios.

Experimental Part

Most investigators have studied the catalytic process of oxidizing sulfur dioxide to sulfur trioxide by the dynamic method (the exceptions are the works of Bodenstein[^16], who, to establish the kinetics of the oxidation process on platinum, used the static method), with the catalyst obtained in one way or another in the form of pieces and charged into contact apparatuses, which usually consisted of wide tubes of quartz or glass. Thus, in the works of Neumann and co-workers the contact tubes had a diameter of \(2\ \mathrm{cm}\), and the length of the catalyst layer was \(20\ \mathrm{cm}\). Tubular electric furnaces were usually used as heating devices; in them the heat exchange between the contact apparatus placed in them and the walls of the electric furnace took place through air. Such an experimental technique was attractive because of its simplicity, but at the same time it has one very substantial drawback. The point is that the reaction of formation of sulfuric anhydride from oxygen and sulfur dioxide proceeds with the liberation of a considerable amount of

heat (21,550 cal per 1 mole of SO₃), and this leads to the fact that, owing to imperfect heat exchange, the temperature of the catalyst during the experiment differs both over the cross section and along the length of the contact apparatus. The temperature differences observed here may reach considerable magnitudes, and the use of furnaces with different densities of the electric-heater winding cannot fully eliminate these temperature distortions. The use, for measuring temperatures, of a thermometer that can be moved along the axis of the contact tube leads only to the ascertainment of the observed temperature difference in the various parts of the contact apparatus; therefore, in all these cases the observed activity of the catalyst is usually referred to some average temperature, which, taking into account the complexity of the temperature dependence of the catalyst activity, cannot be regarded as correct.

Proceeding from the foregoing, we set ourselves the goal of studying the activity of catalysts under conditions as close as possible to isothermal ones. For this purpose the contact apparatus was made of a thin-walled refractory glass tube 5 mm in diameter. This was to lead to minimal temperature differences over the transverse section of the apparatus; furthermore, for the most rapid removal of the heat of reaction, and thereby for creating the same temperature along the length of the contact apparatus, the latter was immersed in a liquid thermostat. As the heat-absorbing liquid, a melt of 55% KNO₃ and 45% NaNO₃ was used. A mixture of these two nitrates is liquid in the temperature range 230–635°C and is a very mobile liquid, ensuring intensive heat transfer. A schematic diagram of the apparatus is shown in Fig. 1.

Air from alternately operating glass gas holders (11), under a constant pressure created by a column of water from a Mariotte bottle (10), was passed through absorbers (12) and (13). In the first of these, the air was purified from carbon dioxide with a 40% solution of caustic soda, and in the second, filled with concentrated sulfuric acid (specific gravity 1.84), the air was freed from water vapor. From absorber (13) the dried air was directed into a two-capillary rheometer (14), by means of which the rate of the air stream was regulated and measured.

Sulfur dioxide gas from cylinder (1) was transferred into a glass gas holder (2), from which, under the pressure of water from the constant-level bottle (10), it was directed for drying through absorbers (3), filled with concentrated sulfuric acid. In order to prevent large losses of sulfur dioxide gas, owing to its dissolution, the surface of the water in gas holder (2) was protected by a 2 cm layer of wagon oil. Other grades of oil tested by us (for example, vaseline, transformer, machine oils) proved little suitable for this purpose, since in their presence sulfur dioxide gas was rather rapidly absorbed by the liquid phase. Pre-

Fig. 1. Diagram of the apparatus for studying the activity of iron–copper catalysts for the oxidation of sulfurous gas to sulfuric anhydride.

Preliminary transfer of sulfur dioxide from cylinder (1) into gas holder (2) was provided for purposes of economy, since the creation of a strictly constant pressure before the rheometers without gas holder (2), for example by means of a monostat, would inevitably have led to the discharge of considerable quantities of sulfur dioxide into the atmosphere.

From absorber (3) the dried sulfur dioxide passed through a two-capillary rheometer (4) and, after mixing with the air stream, was directed to the contact process. Sufficient accuracy in measuring the rate of the gas streams was achieved by using rheometers with a large difference in levels (700 mm) of sulfuric acid (specific gravity = 1.84) in the high- and low-pressure limbs. To avoid the influence of the normally fluctuating air temperature in the laboratory on the rheometer readings, the latter were enclosed in an air thermostat (15), in which, by means of a mercury–toluene thermoregulator with an extended surface and made in the form of a coil (16), a temperature of \(25 \pm 0.5^\circ\mathrm{C}\), equal to the rheometer calibration temperature, was maintained. Before entering the rheometers, the gases were preheated to the thermostat temperature in coils (17), (18). To strengthen heat transfer, thermostat (15) was provided with a propeller fan (19). After the rheometers, the mixture of sulfur dioxide with air passed through coil (20), the contact apparatus (21) filled with crushed catalyst, and, between experiments, was directed through vessel (7) into the atmosphere, while during experiments it was directed through two Reich \(SO_2\) absorbers (8) into aspirator (9). The height of the liquid in vessel (7) was selected so that, when the gas was switched to absorbers (8), no change in the rate of the gas stream occurred, which is easily noticed from a change in the rheometer readings. Thermostat (15), in which coil (20) and contact apparatus (21) were placed, consisted of a large crucible electric furnace (No. 5) and an iron crucible inserted into it, made of 2-millimeter iron. The crucible was filled with a melt of \(NaNO_3\) and \(KNO_3\). The furnace was suspended on blocks from the frame and its weight was balanced with lead weights, which made it possible to raise and lower the furnace with little effort. The movable construction of the thermostat was necessitated by the fact that continuous round-the-clock operation of the installation was impossible under laboratory working conditions. Therefore, to prevent the contact apparatus and coil from being crushed by the solidifying melt, it was necessary to remove them from the thermostat liquid. This was achieved by lowering the thermostat together with the melt. Somewhat greater difficulties arose when immersing the contact apparatus and coil in the molten saltpeter. In this case, if the furnace was raised rapidly, the glass of the contact apparatus and coil, not having had time to warm up, cracked; therefore the raising of the furnace and the immersion of the glass parts were carried out more slowly than its lowering.

To maintain a constant temperature in the thermostat, in the first trial experiments an air–mercury ther—

regulator; however, owing to the large thermal inertia of the furnace, the temperature was maintained with an accuracy of ±5°C.

Subsequently, temperature control was carried out by changing the strength of the electric current in the winding of the thermostat furnace, which was achieved by changing the resistance of a rheostat that, together with an ammeter, was connected in series into the furnace winding.

In this case, even under the most unfavorable operating conditions (sharp fluctuations of the current voltage in the network), the temperature was kept constant with an accuracy of ±1°C, and in most cases the temperature fluctuations were considerably smaller. The temperature of the thermostat liquid, and consequently also the temperature of the contact apparatus, were recorded by means of a platinum–platinum-rhodium thermocouple (6). To achieve good tightness of the setup, the separate apparatuses were connected with one another by soldering; the exception was the contact apparatus: it was attached on ground joints. To check the composition of the gas before contacting, a Reich apparatus (22) was connected in parallel to the gas line.

Preparation of Catalysts.

As is known, the activity of catalysts is determined not only by the nature of the individual components included in the catalyst, but to a very large extent it also depends on the method of preparing the catalyst. In this respect, the history of the development of active catalysts for the production of synthetic ammonia and methanol is most characteristic and instructive. Much money and effort were expended before highly active and stable catalysts were obtained, which in the initial experiments did not possess all the necessary qualities (see, e.g., 17). In preparing complex catalysts, the usual aim is to bring the individual components into as close contact with one another as possible, and also to obtain a fine structure of the catalyst characterized by minimal sizes of the elementary particles.

Iron–copper catalysts were prepared by mixing Fe₂O₃ and CuO in the appropriate quantitative ratios. Fe₂O₃ was obtained as follows: from a dilute aqueous solution (0.5 N) of chemically pure ferric nitrate, with heating, iron oxide hydrate was precipitated with ammonia. After filtration and thorough washing from \((\mathrm{NH}_4)\) and \((\mathrm{NO}_3)\) ions with hot distilled water, the precipitate was dried and, for the final removal of water, calcined for 2 hours in a muffle furnace at a temperature of 500°C; in this process \(\mathrm{Fe(OH)}_3\) was converted into \(\mathrm{Fe}_2\mathrm{O}_3\), colored black. The \(\mathrm{Fe}_2\mathrm{O}_3\) thus obtained was thoroughly ground in an agate mortar, whereby the powder acquired the reddish-brown coloration characteristic of \(\mathrm{Fe}_2\mathrm{O}_3\).

Chemically pure copper oxide was taken in ready-made form and, like Fe₂O₃, was subjected to thorough grinding. Weighed portions of the powdered oxides of iron and copper were introduced into a beaker with distilled water and thoroughly mixed and filtered.

The resulting mixture was slightly dried and pressed into tablets on a hydraulic press to a pressure of 5000 kg/cm². The duration of pressing at 5000 atm. was 5–7 minutes. To remove water, the tablets were again calcined at a temperature of 500°C and crushed. By means of sieves, pieces 1–1.5 mm in size were selected. Pressing of the catalyst was used to achieve still closer contact between the particles of Fe₂O₃ and CuO, and also to create good mechanical strength of the catalyst. The latter circumstance is of great importance in the industrial use of the catalyst.

Results of the experiments and their discussion. In the installation described, 11 catalysts of the following compositions were tested:

No.	Composition
1	Fe₂O₃
2	99% Fe₂O₃ + 1% CuO
3	96% Fe₂O₃ + 4% CuO
4	92% Fe₂O₃ + 8% CuO
5	90% Fe₂O₃ + 10% CuO
6	80% Fe₂O₃ + 20% CuO
7	70% Fe₂O₃ + 30% CuO
8	60% Fe₂O₃ + 40% CuO
9	50% Fe₂O₃ + 50% CuO
10	40% Fe₂O₃ + 60% CuO
11	30% Fe₂O₃ + 70% CuO

(The activity of catalysts with a CuO content from 70 to 100% is currently being studied by us, and the results will be given in a subsequent communication).

The activity of the catalysts was studied as a function of the space velocity (from 200 to 1000) and temperature (350 to 600°C), with a constant composition of the initial gas (7% SO₂, 19.5% O₂, and 73.43% N₂). By space velocity, in accordance with American investigators (see, for example, 18 and 19), who studied ammonia catalysts and were the first to introduce this concept into science and practice, we understand the number of volumes of gas mixture, under normal conditions, that is passed through one bulk volume of catalyst per hour. The value of the space velocity, thus determined, taking into account the correction for the change in gas volume with temperature and pressure, can be related to the residence time of the gas on the catalyst, i.e., the reaction time. This definition of space velocity is generally known and is invariably used by scientists and engineers working in the nitrogen industry. However, by a number of authors who have dealt

Percent oxidation of SO₂ to SO₃ in the presence of
Initial composition of the gas mixture in volume %%:

Tube temp., °C	Space velocity V	Fe₂O₃	99% Fe₂O₃ + 1% CuO	96% Fe₂O₃ + 4% CuO	92% Fe₂O₃ + 8% CuO	90% Fe₂O₃ + 10% CuO
350	200	—	—	20.3	—	—
350	400	—	—	10.4	—	—
350	600	—	—	4.9	—	—
350	800	—	—	1.59	—	—
350	1000	—	—	—	—	—
400	200	5.5	4.8	31.0	6.3	14.5
400	400	4.4	2.4	15.2	2.5	12.8
400	600	2.5	2.0	6.8	0.9	11.2
400	800	—	—	2.6	—	9.9
400	1000	—	—	—	—	5.5
450	200	11.8	9.0	7.9	10.4	18.0
450	400	9.5	3.6	—	4.8	15.7
450	600	5.8	2.2	—	3.4	12.6
450	800	3.6	1.59	—	—	11.4
450	1000	—	—	—	—	7.7
500	200	26.3	25.6	15.7	3.5	23.5
500	400	17.4	14.0	6.35	—	21.0
500	600	11.0	7.9	3.2	—	18.0
500	800	7.5	4.8	—	—	17.0
500	1000	6.9	2.6	—	—	15.0
550	200	40.0	42.6	73.4	50.0	35.0
550	400	31.0	31.0	60.5	34.0	26.0
550	600	24.0	23.0	49.0	23.0	22.2
550	800	18.0	17.4	39.8	16.7	17.0
550	1000	15.8	14.0	34.0	13.5	—
600	200	65.5	69.0	52.2	59.0	48.0
600	400	51.0	50.0	46.0	47.0	36.6
600	600	38.6	38.3	39.0	33.0	32.0
600	800	28.4	28.3	34.0	24.0	26.0
600	1000	26.0	21.7	30.4	20.0	20.0

iron-copper catalysts.
SO$_2$—7%; O$_2$=19,53%; N$_2$=73,47%. $P=1$ atm.

Table 1

SO$_2$ to SO$_3$, in %%; catalyst, in %%

80% Fe$_2$O$_3$ 20% CuO	70% Fe$_2$O$_3$ 30% CuO	60% Fe$_2$O$_3$ 40% CuO	50% Fe$_2$O$_3$ 50% CuO	40% Fe$_2$O$_3$ 60% CuO	30% Fe$_2$O$_3$ 70% CuO
—	1,2	7,1	26,4	21,7	17,0
—	0,3	3,7	23,7	11,1	13,5
—	—	0,8	18,7	8,7	3,95
—	—	—	14,9	7,1	2,4
—	—	—	9,5	4,8	1,1
35,0	16,5	50,0	28,7	33,2	22,5
11,0	7,6	46,6	27,0	26,1	21,7
4,7	4,8	41,5	26,5	21,9	17,3
3,1	1,5	37,1	20,7	18,1	13,9
1,5	—	32,8	19,2	14,3	11,8
24,0	19,6	32,5	18,4	22,3	19,9
21,8	11,0	27,1	7,9	18,8	12,8
18,6	7,3	24,1	2,4	16,0	9,5
15,3	2,4	20,6	1,6	10,9	6,68
11,0	—	15,7	—	7,3	5,6
36,0	42,5	38,3	20,7	25,9	29,9
27,0	12,7	31,0	18,7	23,9	25,7
20,0	8,9	27,8	16,0	21,1	22,6
17,0	6,4	23,6	17,5	14,3	18,1
16,0	3,2	18,9	7,1	12,0	14,8
52,8	64,0	41,3	29,4	27,9	31,8
34,0	49,0	37,9	27,5	27,0	28,6
27,8	36,9	35,2	18,9	26,1	27,2
23,4	29,6	34,0	15,8	16,5	25,0
20,3	—	21,8	12,6	12,9	17,5
59,8	70,0	62,0	57,1	71,5	73,6
39,0	53,4	58,5	40,5	46,6	52,4
30,9	44,1	52,5	39,9	38,4	39,1
27,9	37,0	43,0	29,4	32,9	35,0
23,3	—	35,2	23,1	31,2	31,8

research on catalysts for the contact production of sulfuric acid, as well as in textbooks on sulfuric acid¹, a new meaning has been introduced into the concept of space velocity; namely, space velocity is understood as the number of volumes of pure $\mathrm{SO_2}$, under normal conditions, which is passed through one bulk volume of catalyst per hour, independently of the concentration of $\mathrm{SO_2}$ in the gas mixture. Therefore this quantity cannot be directly compared with the reaction time, which is highly undesirable.

As already indicated, the investigations were carried out at temperatures of $350$—$600^\circ\mathrm{C}$. It was of interest to know how active iron–copper catalysts are in the region of relatively low temperatures, $350$—$450^\circ\mathrm{C}$, at which, according to thermodynamic conditions, almost complete conversion of $\mathrm{SO_2}$ into $\mathrm{SO_3}$ is possible. In the temperature interval $450$—$600^\circ\mathrm{C}$, complete conversion of $\mathrm{SO_2}$ into $\mathrm{SO_3}$ (at atmospheric pressure) is impossible; however, the equilibrium yields are still large, and this, in the case of an active catalyst, makes it possible to oxidize the main masses of $\mathrm{SO_2}$. Above $600^\circ\mathrm{C}$ there is no point in studying activity, since the equilibrium is strongly shifted toward the dissociation of $\mathrm{SO_3}$ into $\mathrm{SO_2}$ and $\mathrm{O_2}$.

All the investigations were conducted with one and the same initial composition of the gas mixture, which, in terms of $\mathrm{SO_2}$ concentration, corresponds to the gases used in the industrial production of contact sulfuric acid.

The process of studying the catalyst consisted of two periods: 1) the period of formation of the catalyst and establishment of constant activity, and 2) the period of direct testing of the activity with recording of all necessary data.

In the first period, a gas mixture of the composition indicated above was passed through the contact apparatus, filled with an oxide iron–copper catalyst, at $350^\circ\mathrm{C}$ and with a constant space velocity equal to 200. At the same time it was observed that the concentration of sulfurous gas at the outlet of the contact apparatus was lower than at the inlet, while at first sulfuric anhydride was simultaneously absent in the contacted gas mixture. After some time, $\mathrm{SO_3}$ appears in the exit gases, and then the percentage of $\mathrm{SO_2}$ in the contacted gases becomes constant; this characterized the established stationary state with the final catalyst activity characteristic of these conditions (temperature, $\mathrm{SO_2}$ concentration, space velocity). Thus, during this first period of the investigation, complete or partial absorption of sulfurous gas by the catalyst took place. Absorption of sulfurous gas by iron oxides was also observed by Lunge, Keppeler, Neumann, and Goebell. These authors explain these phenomena by sulfation of $\mathrm{Fe_2O_3}$. The duration of establishment of the stationary state associated with sulfation of $\mathrm{Fe_2O_3}$ and $\mathrm{CuO}$ ($t = 350^\circ\mathrm{C}$, concentration of $\mathrm{SO_2}$ in the gas 7%, space velocity 200) was:

Catalyst	Time
$\mathrm{Fe_2O_3}$	3 hours
90% $\mathrm{Fe_2O_3}$ + 10% $\mathrm{CuO}$	6 hours
80% $\mathrm{Fe_2O_3}$ + 20% $\mathrm{CuO}$	8 hours
70% $\mathrm{Fe_2O_3}$ + 30% $\mathrm{CuO}$	15 hours
60% $\mathrm{Fe_2O_3}$ + 40% $\mathrm{CuO}$	20 hours
50% $\mathrm{Fe_2O_3}$ + 50% $\mathrm{CuO}$	34 hours
40% $\mathrm{Fe_2O_3}$ + 60% $\mathrm{CuO}$	41 hours
30% $\mathrm{Fe_2O_3}$ + 70% $\mathrm{CuO}$	45 hours

The time of sulfation for pure iron oxide, according to the investigation of Neumann and Goebel, at temperatures below 625°C was a minimum of 3 hours, which agrees with the results obtained by us.

From the data presented, and also from Fig. 2, it is evident that the time required for the establishment of a steady state depends, first, on the composition of the catalyst and, second, increases with increasing percentage content of copper oxide in the catalyst. The latter circumstance may be explained either by the increasing density of the catalytic mass with increasing content of CuO in it, as a result of which the diffusion processes of $\mathrm{SO_2}$ and oxygen into the interior of the particle are impeded, or by the slower rate of sulfation of CuO in comparison with $\mathrm{Fe_2O_3}$. On the basis of our experiments we cannot give preference to one of the proposed explanations; it is possible that both of these causes act simultaneously.

For all catalysts, the degree of sulfation of $\mathrm{Fe_2O_3}$ and CuO, if judged by the amount of $\mathrm{SO_2}$ passed during the time required to establish the steady state, amounts to only a few percent. According to Keppler’s data, the technically used contact mass consisting of iron oxide contained 9.4% sulfuric acid, which agrees with what has just been set forth.

The results of experiments on the activity of iron–copper catalysts obtained in the second period of testing are given in Table 1 and shown in Figs. 3–13, 14–19, and 20–21. From the table presented and from Figs. 3–13 it is evident that, at constant temperature, for all the catalysts investigated, the degree of oxidation of sulfurous gas to sulfuric anhydride decreases with increasing space velocity; moreover, this decrease in the degree of oxidation, as a rule, occurs considerably more slowly than the space velocity increases. Thus, for example, in the presence of a catalyst whose initial form consisted of pure iron oxide, at a temperature of 600°C and at space velocities of 200 and 1000, the degree of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ proved to be 65.5 and 26.0%, respectively; i.e., the space velocity increased by a factor of 5, while the degree of oxidation decreased only by a factor of 2.5. An analogous dependence is also found for other temperatures and catalysts.

Fig. 2. Dependence of the time required to establish a stationary state on the content of copper oxide in the initial form of mixed iron–copper catalysts.

Fig. 3. Dependence of the degree of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) on the space velocity for iron–copper catalysts.

Fig. 4.

Fig. 4–5. Dependence of the degree of oxidation of SO₂ to SO₃ on the space velocity for iron–copper catalysts.

Fig. 6.

Figs. 6–7. Dependence of the degree of oxidation of SO₂ to SO₃ on space velocity for iron–copper catalysts.

Fig. 8.

Figs. 8–9. Dependence of the degree of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ on the space velocity for iron–copper catalysts.

Fig. 10.

Figs. 10–11. Dependence of the degree of oxidation of SO₂ to SO₃ on the space velocity for iron–copper catalysts.

Fig. 12.

Figs. 12–13. Dependence of the degree of oxidation of SO₂ to SO₃ on the space velocity for iron–copper catalysts.

The isotherms of the reaction rate under the conditions investigated, for all catalysts, with a decrease in temperature from 600 to 350°C, as a rule tend to straighten out, in some cases becoming straight lines (for example, the 450 and 400°C isotherms in the presence of a catalyst consisting of 90% \(\mathrm{Fe_2O_3} + 10\%\ \mathrm{CuO}\), Fig. 7, and others).

Turning to consideration of the influence of temperature on the catalytic properties of the substances we studied, one should first of all consider the catalyst whose initial form consisted of pure iron oxide. For this substance, at different but constant space velocities, with increasing temperature the degree of oxidation in the temperature interval investigated rises continuously; moreover, the activity, measured by the degree of oxidation, increases especially rapidly in the region of high temperatures, which is characterized by the steeper rise of the curves in Fig. 14a. Indeed, if the degree of oxidation at \(V = 200\) and \(t = 400^\circ\mathrm{C}\) was 5.5%, then at \(t = 600^\circ\mathrm{C}\) the analogous value increased to 65.5%. The same picture is also observed at other space velocities. The beginning of fully measurable contact for this catalyst at space velocities of 200–800 should be taken as a temperature of 400°C, although noticeable contact begins at somewhat lower temperatures (between 350 and 400°C). The maximum degrees of oxidation were observed by us at a temperature of 600°C; however, the optimum temperatures, which, as is known\(^{20}\), are a function of the space velocity, must lie above 600°C, and the higher, the greater the space velocity. Comparing the course of the curves in Fig. 14a with the equilibrium curve (the latter is not shown in the drawing), one can determine the region of optimum temperatures; apparently, for the catalyst obtained from pure iron oxide, with the gas composition indicated above and in the investigated interval of space velocities, the optimum temperatures lie in the range 615–630°C. According to the investigations of Neumann and Goebell, for the same catalyst, at a flow rate of 150 ml/min, which in terms of space velocity would be \(\sim 239\), the optimum temperature is 625°C. Comparing the degrees of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) observed by us in the presence of the catalyst obtained from pure iron oxide with the analogous values found by the authors just mentioned (see Table 2), one can see that the iron oxide used by us proved to be considerably more active than the same substance in the experiments of Neumann and Goebell and, all the more so, in the experiments of Lunge and Reinhardt.

This increased activity of the iron oxide obtained by us in comparison with the oxide studied by Neumann and Goebell is apparently explained by the fact that we precipitated iron hydroxide from a solution of ferric nitrate, and not from a solution of ferrous sulfate, as was the case in the experiments of the authors mentioned.

Fig. 14. Dependence of the degree of oxidation of \(SO_2\) to \(SO_3\) on temperature for iron–copper catalysts.

In the graph:

• \(b\) — degree of oxidation, %
• \(t^\circ C\)

(a) Catalyst \(Fe_2O_3\)

(b) Catalyst \(99\%\,Fe_2O_3 + 1\%\,CuO\)

Curves for space velocity:

Curve	Space velocity
I	200
II	400
III	600
IV	800
V	1000

Fig. 15. Dependence of the degree of oxidation of SO₂ to SO₃ on temperature for iron–copper catalysts.

In the figure:

• ordinate: degree of oxidation, \(\beta,\%\)
• abscissa: \(t^\circ\mathrm{C}\)

(a) catalyst \(96\%\ \mathrm{Fe_2O_3} + \mathrm{CuO}\ 4\%\)

(b) catalyst \(92\%\ \mathrm{Fe_2O_3} + 8\%\ \mathrm{CuO}\)

I — curve for space velocity 200
II — 400
III — 600
IV — 800
V — 1000

Fig. 16. Dependence of the degree of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) on temperature for iron–copper catalysts.

Fig. 17. Dependence of the degree of oxidation of \(SH_2\) to \(SO_3\) on temperature for iron–copper catalysts,

Labels in the figure:

• Degree of oxidation, %
• \(t^\circ C\)

(a) catalyst \(70\% Fe_2O_3 + CuO\)

curve I for a space velocity of 200
II — 400
III — 600
IV — 800
V — velocity 1000

(b) catalyst \(60\% Fe_2O_3 + 40\% CuO\)

\(v = 200\)
\(v = 400\)
\(v = 600\)
\(v = 800\)
\(v = 1000\)

Fig. 18.

Top plot:

(b)
catalyst \(40\%\mathrm{Fe_2O_3} + 60\%\mathrm{CuO}\)

curve I for space velocity 200
curve II — 400
curve III — 600
curve IV — 800
curve V — 1000

Middle plot:

(a)
catalyst \(50\%\mathrm{Fe_2O_3} + 50\%\mathrm{CuO}\)

curve I for space velocity 200
curve II — 400
curve III — 600
curve IV — 800
curve V — 1000

Vertical axis: degree of oxidation \(\delta,\%\)
Horizontal axis: \(t^\circ\mathrm{C}\)

catalyst \(30\%\mathrm{Fe_2O_3} + 70\%\mathrm{CuO}\)

curve I for space velocity 200
curve II — 400
curve III — 600
curve IV — 800
curve V — 1000

Vertical axis: degree of oxidation \(\delta,\%\)
Horizontal axis: \(t^\circ\mathrm{C}\)

Figs. 18–19. Dependence of the degree of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) on temperature for iron–copper catalysts.

Table 2

Catalysis temperature, °C	Neumann and Goebell, volumetric velocity 250¹)	Neumann and Goebell, volumetric velocity 500²)	Our studies, volumetric velocity 250¹)	Our studies, volumetric velocity 500²)
500	13.3	—	24	14
550	18.7	10	38	27.3
600	38.6	30	57	42.5

In addition, pressing iron oxide under such a high pressure as 5000 atm could also have been the cause of the increased activity of the catalyst.

Turning to consideration of the temperature dependence of the degree of oxidation for mixed catalysts, it should first be noted that the activity of the CuO catalyst, according to Neumann, is considerably lower than the activity of the purely iron catalyst. This is characterized by the following figures: at temperatures of 450, 500, 550, and 600°C and a gas flow rate of 150 ml/min, which corresponds to a volumetric velocity of 250, the degree of conversion was found to be, respectively, 0.14; 0.67; 1.90; 22.38%, i.e., with increasing temperature, just as for iron oxide, a continuous increase in catalyst activity is observed. Taking into account the properties of catalysts obtained from pure Fe₂O₃ and CuO, and also assuming the law of additivity to be valid, one could have expected that mixed iron–copper catalysts would, first, be less active than purely iron ones and, second, that the nature of the temperature dependence of the activity of the mixed catalysts should repeat the analogous dependences for pure iron and copper catalysts, i.e., the curves of percent oxidation = f(t°) at constant volumetric velocities should have a smooth character, without the presence of minima and maxima. Indeed, for 4 mixed catalysts with 1, 10, 20, and 30% CuO (Figs. 14 b, 16 a, 16 b, 17 a), the temperature dependence of the percentage oxidation of SO₂ to SO₃ is expressed, for different volumetric velocities, by smooth curves, along the entire length of which the reaction rate increases with temperature. Thus, for example, for a catalyst containing 1% CuO at a volumetric velocity of 200, the degree of oxidation of SO₂ to SO₃ for temperatures of 400, 500, and 600°C was found to be, respectively, 4.8; 25.6; 69.0%. Ana-

¹) In the experiments of Neumann and Goebell, the flow rate was expressed as ml/min for the entire amount of catalyst. The values of volumetric velocities given in these graphs were calculated by us on the basis of the dimensions of the contact apparatus, the height of the catalyst layer, and the diameter of the sleeve for the thermocouple placed in the contact mass. The diameter of the sleeve is not given in the work of Neumann and Goebell, and therefore we assigned it, taking it to be 6 mm.

²) The oxidation percentages at volumetric velocities of 250 and 500 were determined graphically from the curves shown in Fig. (14 a).

... a logical temperature dependence occurs for this catalyst at all the space velocities investigated, with the sole difference that the curves for percent oxidation \(= f(t^\circ)\), as the space velocity increases, are situated more gently sloping.

The curves of percent oxidation \(= f(t^\circ)\) (Fig. 17a) for the catalyst with 30% CuO, also indicating a continuous increase in the activity of this catalyst with increasing temperature, have a special form: namely, in the temperature regions 400–500°C and 550–600°C the curves are gently sloping, which indicates a slow increase in the activity of the catalyst with temperature, whereas in the temperature region 500–550°C, on the contrary, a sharp increase in the percent oxidation occurs.

The temperature dependence of the percent oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) in the presence of 6 mixed catalysts has an entirely different form; the initial form of these catalysts, in addition to \(\mathrm{Fe_2O_3}\), contained 4, 8, 40, 50, 60, and 70% CuO. Namely, in all these cases (Figs. 15a, 15b, 18a, 18b, and 19), at all the investigated space velocities, initially, when the temperature is raised from 350°C to 400°C, an increase occurs in the percent oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\); in the temperature region 400–450°C the activity falls, and with a further rise in temperature from 450 to 600°C an increase in the activity of these catalysts is again observed. Thus, the curves expressing the temperature dependence of the percent oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) for all catalysts of this group in the temperature region 350–600°C have one maximum and one minimum each. Moreover, the maximum, as a rule, is observed at 400°C and the minimum at 450°C. The exceptions are the catalysts containing 4% and 50% CuO: for the first of these a second activity maximum is also observed at a temperature of 550°; for the second, the first maximum and the minimum lie at higher temperatures, namely, respectively at 450 and 500°C.

A very interesting circumstance is that, in the presence of some catalysts of this group (with CuO contents of 4, 8, and 50%), the degree of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) at the temperature of the minimum, within the range of the space velocities used, proves to be close to or equal to zero, and thus these catalysts at lower temperatures (350°C) proved considerably more active than at 450°C. For all the other catalysts of this group, the absolute value of the percent oxidation at the temperature of the minimum is either almost equal to the percent oxidation at 350°C or considerably exceeds it (see, e.g., Fig. 17b). The increase in the activity of the catalysts when the temperature is raised above the temperature of minimum activity occurs with different intensity. For those catalysts whose initial form contains 4 and 8% CuO, the activity increases especially sharply in the temperature region 500–550°C (Figs. 15a and 15b). Conversely, for catalysts with 40, 50, 60, and 70% CuO, when the temperature is raised from 450 to 600°C, a uniform, comparatively slow increase in activ-

...which is characterized by a flatter arrangement of the curves in Figs. (17b), (18a), (18b), and (19).

For the mixed iron–copper catalysts that we studied, the highest percentages of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) at all space velocities were observed at a temperature of \(600^\circ\mathrm{C}\), i.e., at the highest temperature that we used in our experiments. The exception is the catalyst containing \(4\%\) \(\mathrm{CuO}\), which gave the highest percentages of oxidation at a temperature of \(550^\circ\mathrm{C}\) (Fig. 15a).

Thus, the highest activity of mixed iron–copper catalysts under the conditions of the study is observed at a temperature of \(600^\circ\mathrm{C}\); however, the maximum possible degrees of oxidation should lie in the region of temperatures higher than \(600^\circ\mathrm{C}\).

To elucidate the effect of additions of copper oxide on the activity of iron–copper catalysts, on the basis of the data in Table 1, Figs. 20a, 20b, 20c, 21a, and 21b were constructed; in these, the ordinates show the degrees of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\), and the abscissae show the percentage content of copper oxide in the iron–copper catalyst before its use; lines I, II, III, IV, and V in these figures are characterized by constancy of the space velocities. From these figures it is evident that no simple dependence is observed, of the type that might be predicted on the basis of the additivity rule, between the amount of copper oxide in the mixed catalyst and its catalytic activity, measured by the percentage of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\). On the contrary, at all the temperatures and space velocities studied, the curves \((\%\ \text{oxidation})_{\tau,V}=f_i(\%\ \mathrm{CuO})\) have a complex character, determined by the presence of several maxima and minima of activity; moreover, the occurrence of the latter on the curves is determined not only by the percentage of \(\mathrm{CuO}\) in the catalyst, but also by the temperature (the gas composition in our experiments was constant) at which the percentages of oxidation are recorded. Let us consider the effect of the addition of \(\mathrm{CuO}\) on the activity of the mixed iron–copper catalyst at a temperature of \(400^\circ\mathrm{C}\). At this temperature, even at the smallest space velocity, 200, the degree of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) in the presence of a catalyst obtained from “pure” iron oxide is only \(5.5\%\); introduction of \(1\%\) \(\mathrm{CuO}\) leads to a slight decrease in the activity of the catalyst (Fig. 20a), while increasing the \(\mathrm{CuO}\) content to \(4\%\) causes a considerable increase in activity relative to that of pure iron oxide; in this case the percentage of oxidation is \(31\%\). With an increase in the \(\mathrm{CuO}\) content from 4 to \(8\%\), the activity again falls, reaching, for the catalyst with \(8\%\) \(\mathrm{CuO}\), values close to those of pure iron oxide. After a small rise in the percentage of oxidation for the catalyst with \(10\%\) \(\mathrm{CuO}\), there follows a slow decline in activity up to the catalyst containing \(30\%\) \(\mathrm{CuO}\). The activity of catalysts having from 10 to \(30\%\) \(\mathrm{CuO}\) only slightly exceeds the activity of the pure iron catalyst.

Fig. 20.

Figs. 20–21. Dependence of the degree of oxidation of SO₂ to SO₃ on the CaO content in the initial form of iron-copper catalysts.

A further increase in the CuO content leads to a vigorous rise in activity; moreover, for a temperature of 400°C, among all those investigated the most active catalyst proved to be the one containing 40% CuO. Indeed, in the presence of this catalyst the observed percentage of oxidation of SO₂ to SO₃ at a space velocity of 200 is 50.0%, which is 9.1 times greater than the corresponding value characteristic of pure iron oxide; at other space velocities the increase in activity is still more vigorous. The introduction of still larger amounts of CuO than 40%, up to 70%, led to a constant decrease in activity; however, the absolute value of the percentage of oxidation for catalysts containing from 40 to 70% is considerably higher than the analogous value for pure iron oxide. This indicates the strong activating influence of CuO additions, even when they are introduced in large amounts (70%). From all that has been set forth it is evident that the curves (% oxidation)$_{T,V}=f(\% \mathrm{CuO})$, expressing the change in the catalytic properties of mixed iron–copper catalysts, have 3 minima and 3 maxima. The minimum values of activity under these conditions of investigation fall on catalysts containing 1, 8, and 30% CuO; activity maxima are observed for catalysts with 4, 10, and 40% CuO, and at a temperature of 400°C the highest activity is observed for the catalyst with 40% CuO.

The behavior of the catalytic system at a temperature of 450°C is somewhat different (Fig. 20c). Indeed, at this temperature an increase in the CuO content in the catalyst from 0 to 20% at first leads to a decrease in activity, and then to its increase. The first minimum of activity is observed at 4% CuO, whereas at 400°C this same catalyst gave an activity maximum. The first activity maximum occurs at 20% CuO; a further increase in the percentage of copper oxide to 30% causes the appearance of a second minimum, and at 40% a second maximum is observed, after which the drop in activity leads to the appearance (at 50% CuO) of a third minimum. In the range of CuO content in the catalyst from 50 to 70% there is one more weakly expressed third maximum of activity. It should be noted that at 450°C, at all space velocities, the activity maxima lie the higher, the lower the gas space velocity. From the diagram considered it is evident that copper oxide, mixed with iron oxide in definite amounts, gives catalysts that considerably surpass pure iron oxide in activity. At this temperature the most active catalyst proved to be (as also at 400°C) the one containing 40% CuO. In this case, at a space velocity of 200, the degree of oxidation of SO₂ to SO₃ is 32.5%, whereas for pure iron oxide this value is only 11.8%, i.e., the activity increases by 2.7 times. At a temperature of 450°C the maximum activity of the catalytic system, observed at 40% CuO, is considerably lower than that recorded at 400°C; thus, for example, at a space ...

at a space velocity of 200, the percent oxidation of SO₂ to SO₃ at 400 and 450°C is, respectively, 50.0 and 32.5%. However, at a temperature of 450°, catalysts containing from 10 to 30% CuO exhibit greater activity than at 400°C.

The change in the activity of this catalytic system at a temperature of 500°C and space velocities from 400 to 1000, as a function of the percentage content of copper oxide, is characterized by the presence of 2 maxima and 3 minima on the curves \([ \% \text{ oxidation} ]_{t,V} = f(\% \text{ CuO})\). The minimum activities were observed for catalysts whose initial form contained 8, 30, and 50% CuO.

The maximum activities fall, respectively, on catalysts containing 20 and 40% CuO. At a temperature of 500°C, for all space velocities (with the exception of \(V = 200\)), the maximum activity also falls on the catalyst containing 40% CuO. In this case copper oxide enhances the catalytic properties of iron oxide; however, the increase in activity, measured by the percent oxidation of SO₂ to SO₃, taking into account the increasing activity of the iron oxide itself, is small. Thus, for example, at a space velocity of 200, in the presence of this catalyst and of the catalyst made of Fe₂O₃, the degrees of oxidation of SO₂ to SO₃ are, respectively, 38.3% and 26.3%. Most of the catalysts studied at \(t = 500^\circ\text{C}\) proved more active than at 450°C; however, the most active catalyst with 40% CuO at a temperature of 500°C gave a lower percent oxidation than at 400°C.

Raising the temperature from 500 to 550°C leads to a sharp increase in the activity both of pure iron oxide and of its mixtures with copper oxide. The curves (Fig. 21a) \((\% \text{ oxidation})_{t,V} = f(\% \text{ CuO})\) have 2 maxima and 3 minima. The first, not very pronounced minimum occurs at 1% CuO; the other two minima correspond to catalysts with 10 and 50% CuO. The maxima of activity of this catalytic system (\(t = 550^\circ\text{C}\)) are observed at 4 and 30% CuO, and the maximum percent oxidation, equal to 73.4%, was given by the catalyst with 4% CuO (the percent oxidation in the presence of Fe₂O₃ is equal to 40%).

Catalysts with a content of from 20 to 40% CuO also proved more active than iron oxide. At a temperature of 600°C (Fig. 21b), the curves \((\% \text{ oxidation})_{t,V} = f(\% \text{ CuO})\) have 1 maximum (40% CuO) and 2 minima (10 and 50% CuO). It should be noted that at this temperature the introduction of various amounts of copper oxide into the iron catalyst does not cause sharp changes in the activity of the catalysts obtained. The maximum percent oxidation observed at space velocities from 400 to 1000 in the presence of the most active catalyst with 40% CuO only slightly exceeds the corresponding values for pure iron oxide. It is also interesting to note that at a temperature of 600°C, catalysts with 50 to 70% CuO proved considerably more active than at lower temperatures.

From all of the foregoing it is clearly seen that the introduction of additions of copper oxide into the iron-oxide catalyst may lead both to an increase and to a decrease in activity as compared with pure iron oxide. For the majority of the temperatures and space velocities investigated, the most active catalysts proved to be those containing about 40% copper oxide. However, the maximum percentage of oxidation, equal to 73.4 (space velocity 200), was observed at a temperature of 550°C on a catalyst containing only 4% CuO.

According to the investigations of Bodenstein and Fink ^16), the rate of the reaction \(2\mathrm{SO}_2 + \mathrm{O}_2 = 2\mathrm{SO}_3\) in the presence of a platinum catalyst in the temperature range 150–250°C, where the reaction is irreversibly shifted toward the formation of \(\mathrm{SO}_3\), is determined by the following kinetic equation:

\[ \frac{dx}{dt}=K\frac{(a-x)}{\sqrt{x}}, \tag{1} \]

where: \(a\) is the initial concentration of \(\mathrm{SO}_2\),

\((a-x)\) and \(X\) are, respectively, the concentrations of sulfurous gas and sulfuric anhydride at time \(t\).

According to this equation, valid with an excess of oxygen (relative to stoichiometric), the reaction rate is directly proportional to the concentration of sulfurous gas, does not depend on the concentration of oxygen, and is inhibited by the reaction product to the power \(1/2\). Thus, the apparent order of the reaction is equal to 1. Taylor and Lehner ^21) showed that at considerably higher temperatures, when the equilibrium is only partially shifted toward the formation of \(\mathrm{SO}_3\), in equation (1), instead of the quantity \(a\), one should substitute the difference between the initial concentration of \(\mathrm{SO}_2\) and the equilibrium concentration of this same gas.

If the degrees of oxidation of \(\mathrm{SO}_2\) to \(\mathrm{SO}_3\) at equilibrium and at time \(t\), respectively, are denoted by \(X_m\) and \(X\), in fractions of unity; and the initial concentration of \(\mathrm{SO}_2\) is still denoted by \(a\), then the differential kinetic equation of Taylor and Lehner takes the following form:

\[ \frac{dx}{dt}=K\frac{X_m-X}{\sqrt{ax}} \tag{1} \]

or, in integrated form:

\[ K=\frac{\sqrt{a}}{t} \left[ \sqrt{x_m}\,2.303\log \frac{\sqrt{x_m}+\sqrt{x}}{\sqrt{x_m}-\sqrt{x}} -2\sqrt{x} \right]. \tag{2} \]

The investigations of Boreskov and Pligunov ^22) showed that the Taylor and Lehner equation is applicable not only to the platinum catalyst, but also to the vanadium catalyst. Kuzminykh and Turkhan ^14), who investigated the catalytic activity of powdered colch—

of this cinder, also found that at \(650^\circ\mathrm{C}\) equation (2) is applicable for this catalyst. According to their data, in the presence of pyrite cinder the rate of the reaction of oxidation of sulfur dioxide to sulfur trioxide does not depend on the concentration of oxygen, which is confirmed by the satisfactory constancy of the rate constants calculated by equation (2) at various initial oxygen concentrations.

In the present work we also made an attempt to apply the Taylor and Lenher equation to the results obtained by us on iron–copper catalysts. The criterion for the applicability of this equation was to be the constancy of the rate constants calculated from the degrees of oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) observed at different space velocities and at constant temperature.

To calculate the rate constants by equation (2), it is necessary to know the residence time of the gas in the reaction space. However, from our experiments this quantity cannot be determined exactly, since the true volume occupied by the catalyst is unknown, as are also the diameter and the character of the arrangement of the pores. Therefore, instead of the reaction time, as was previously shown by one of us\(^{20}\), we substituted into equation (2) the values of the reciprocal space velocity, preliminarily corrected for the expansion of the gas with temperature, since \(t\) is proportional to

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

Thus, the final form of the kinetic equation was as follows:

\[ K_1 = V \frac{T}{273} \sqrt{a}\left[\sqrt{X_m}\, 2{,}303 \log \frac{\sqrt{X_m}+\sqrt{x}}{\sqrt{X_m}-\sqrt{x}} - 2\sqrt{x}\right] \tag{3} \]

The equilibrium degree of oxidation entering into equation (3) was calculated from the isochore equation for the Haber reaction.

The rate constants of the reaction \(2\mathrm{SO_2}+\mathrm{O_2}=2\mathrm{SO_3}\) in the presence of iron–copper catalysts, calculated by equation (3), are given in Table (3). From Table (3) it is seen that, for the majority of the catalysts studied by us, either at all temperatures or only at some of them, the rate constants calculated for different space velocities by equation (3) remain constant quantities. Some fluctuations of the rate constants for this group of catalysts may be explained by experimental errors and, in particular, apparently, by the slowness of the establishment of stationary conditions in passing from one space velocity to another, although we tried to eliminate these errors by prolonged passage of gas at the given vol-

...of the space velocity until constancy of the degree of conversion was established before the final recording of the experimental data. However, for some of the catalysts, either over the entire temperature range (400–600°C) or at some of the temperatures, the rate constants show a clearly expressed tendency to change; for these catalysts, in most cases, the rate constants steadily decrease with increasing space velocity. Thus, for example, for the catalyst whose initial form contained 8% CuO, at all temperatures from 400 to 600°C a decrease in the rate constants is observed with increasing space velocity; in this case the rate constants decrease especially strongly at low temperatures. For some catalysts, the decrease in rate constants with increasing space velocity occurs only at some temperatures. Thus, this phenomenon is observed at 400–450–500°C for the catalyst with 1% CuO; at higher temperatures (550 and 600°C) the constancy of the rate constants is maintained more or less satisfactorily. Finally, there are catalysts which, at some temperatures, give rise not to a decrease but to an increase in the rate constants with increasing space velocity.

Indeed, at a temperature of 400°C in the presence of the catalyst whose initial form contained 40% CuO, the rate constants also increase with increasing space velocity.

Thus, for the majority of the catalysts and temperatures investigated by us, the Taylor and Lehner equation correctly expresses the formal kinetic relationships occurring in the course of the reaction \(2\mathrm{SO}_2+\mathrm{O}_2=2\mathrm{SO}_3\) in the presence of iron–copper catalysts. However, for the catalyst with 8% CuO at all temperatures investigated by us, and also for some temperatures in the presence of other catalysts, this kinetic equation is inapplicable, since the steady change in the rate constants observed in these cases with increasing space velocity cannot be explained by experimental errors.

As is known, the activation energy of a chemical process is related to the rate constants through the Arrhenius equation \(^{23}\).

\[ \frac{d\ln K_1}{dT}=\frac{E}{RT^2} \tag{4} \]

where: \(E\), \(T\), \(K\), and \(R\) are, respectively, the activation energy in calories per mole of the substance formed, the temperature in °K, the reaction rate constant, and the gas constant, equal to 1.987.

Assuming \(E\) to be independent of temperature, after integrating equation (4) and replacing natural logarithms by decimal logarithms, we obtain:

\[ \log K_1=-\frac{E}{2.303R}\left(\frac{1}{T}\right)+\mathrm{Const} \tag{5} \]

If one plots \(\log K_1\) along the ordinate axis, and along the abscissa axis

\[ \left(\frac{1}{T}\right), \]

then, if the assumption made regarding the constancy of \(E\) is correct, a straight line should be obtained, the tangent of whose angle of inclination, multiplied by \(2.303R\), should give the value of the activation energy. Since the values of \(K_1\) were calculated from the concentrations of the reaction components in the gas volume, and not from those concentrations that actually exist in the adsorption layer on the solid surface of the catalyst, where the reaction takes place, the value of \(E\) determined from equation (5) will express not the true but the apparent activation energy.

The constructions described are shown in Figures (22–29).

From Figure 22 it is evident that, when \(\log K_1\) is plotted against

\[ \left(\frac{1}{T}\right) \]

for a catalyst of pure iron oxide, a straight line is obtained over the entire temperature range investigated by us (400–600°C).

However, a similar rectilinear dependence of \(\log K_1\) on

\[ \left(\frac{1}{T}\right) \]

for mixed catalysts is observed only in the temperature interval 450–600°C. In all these cases, the \(\log K_1\) corresponding to a temperature of 400°C clearly does not fall on the straight line, lying above it. Assuming that in these cases the catalysts, at temperatures from 400 to 450°C, pass from one catalytic modification into another, plots were made of \(\log K_1\) against

\[ \left(\frac{1}{T}\right) \]

and the apparent activation energies of the process of oxidation of \(SO_2\) to \(SO_3\) in the presence of the first modification (i.e., at 350–400°C) were calculated. The calculated values of the apparent activation energies are given in Table 4. From Table 4 it is evident that for pure iron oxide the activation energy proved to be equal to 26800 cal. The introduction of copper oxide into the catalyst can cause both an increase and a decrease in the activation energy. Indeed, copper oxide introduced into the catalyst in an amount of only 1% raises the activation energy to 34300 cal/mol, whereas increasing the CuO content to 20% in the temperature range (450–600°C) lowers the activation energy to 14300 cal. In the low-temperature region (350–400°C), for catalysts with a CuO content of 50–70%, the activation energies proved to be less than the analogous value for pure iron oxide.

It should be noted that the highest value of the apparent activation energy for the iron–copper catalysts investigated by us is equal to 34300 cal/mol. In the remaining cases the activation energy proved to be lower.

According to the investigations of the Kuz’mins and Turkhan, the activation energy of the process of oxidation of \(SO_2\) to \(SO_3\) in the presence of pyrite cinders in the temperature region 650–700°C varies from 69000

Values of the rate constants calculated by equation 3.
Initial composition of the gas mixture: SO₂—7%; O₂ = 19.53%; N₂ = 74.47%

Temperature in the furnace, °C	Volumetric velocity, V	100% Fe₂O₃	99% Fe₂O₃ + 1% CuO	96% Fe₂O₃ + 4% CuO	92% Fe₂O₃ + 8% CuO	90% Fe₂O₃ + 10% Cu
400°	200	1,18	0,965	18,86	1,46	5,34
400°	400	1,31	0,887	11,51	0,73	8,75
400°	600	1,095	0,779	4,90	0,25	10,64
400°	800	—	—	1,46	—	13,27
400°	1000	—	—	—	—	5,38
Average value of K		1,175	0,877	—	0,81	10,886
450°	200	4,134	2,70	—	4,44	7,99
450°	400	4,87	2,45	—	2,057	13,02
450°	600	4,099	0,929	—	1,85	13,74
450°	800	4,9	0,785	—	—	15,92
450°	1000	—	—	—	—	11,65
Average value of K		4,500	2,575	—	2,78	12,464
500°	200	16,87	16,14	179,56	—	13,919
500°	400	16,75	11,97	197,95	—	23,13
500°	600	12,26	7,33	189,26	—	26,93
500°	800	9,619	4,56	166,08	—	32,73
500°	1000	4,20	1,73	144,125	—	32,81
Average value of K		15,29	11,81	—	—	25,758
550°	200	37,854	45,55	113,95	65,87	32,54
550°	400	50,60	48,57	147,00	59,81	37,16
550°	600	50,112	46,46	155,63	46,61	39,04
550°	800	41,74	39,26	158,32	36,67	37,63
550°	1000	39,24	32,17	—	30,39	—
Average value of K		43,9	42,4	—	47,828	36,59
600°	200	192,28	244,36	113,95	143,90	80,92
600°	400	165,24	152,19	147,00	135,14	100,3
600°	600	152,29	165,54	155,63	112,18	95,64
600°	800	140,52	132,94	158,32	87,04	87,04
600°	1000	120,32	88,14	160,32	73,51	83,34
Average value of K		154,23	156,63	147,04	110,54	89,45

$P = 1$ atm.

Table 5.

Rate constants of the catalyst

80% Fe₂O₃ + 20% CuO	70% Fe₂O₃ 30% CuO	60% Fe₂O₃ 40% CuO	50% Fe₂O₃ 50% CuO	40% Fe₂O₃ 60% CuO	30% Fe₂O₃ 70% CuO
23,40 6,89 2,86 1,92 1,05	6,56 3,89 2,86 0,81 —	46,14 80,13 96,14 104,05 105,73	16,48 29,69 43,17 38,34 42,00	21,29 28,03 27,40 30,52 25,82	10,91 19,99 21,27 19,96 24,09
—	—	—	41,20	26,61	19,244
13,08 21,95 25,71 24,90 18,60	10,74 7,44 5,87 1,46 —	36,02 32,15 39,23 40,52 32,46	8,42 4,41 1,09 0,757 —	11,55 17,43 20,85 14,07 9,78	9,57 9,16 8,79 6,78 6,48
20.848	6,38	36,075	—	14,736	8,156
29,76 35,23 32,04 32,69 37,08	39,63 10,25 8,90 7,13 6,07	108,04 134,04 121,64 56,53 20,078	11,15 19,69 19,87 17,56 10,34	16,32 28,71 35,19 24,75 23,45	21,08 31,79 39,20 35,84 35,14
33,36	8,087	121,24	19,04	28,025	35,49
74,18 74,4 64,61 46,40 59,06	118,11 122,26 107,74 96,35 96,14	44,57 73,22 98,71 98,51 92,49	31,78 40,94 33,57 33,40 27,22	25,07 39,65 42,59 35,80 28,27	43,28 43,87 61,69 71,65 46,22
63,73	108,12	96,57	33,38	34,276	53,34
142,57 103,66 99,16 109,68 99,84	252,28 209,17 201,03 119,07 192,36	159,20 267,29 296,41 253,88 212,37	141,80 181,61 162,16 120,68 83,85	281,02 151,31 150,69 148,72 107,21	333,90 233,31 156,30 167,39 174,38
110,98	194,78	257,49	124,02	167,79	166,02

Catalyst: $\mathrm{Fe_2O_3}$

Fig. 22.

Catalyst: $99\%\,\mathrm{Fe_2O_3} + 1\%\,\mathrm{CuO}$

Figs. 22–23. Dependence of $\log K$ on the reciprocal of the absolute temperature for iron–copper catalysts.

Fig. 24.

Figs. 24–25. Dependence of \(\log K\) on the reciprocal of the absolute temperature for iron–copper catalysts.

Catalyst \(60\%\mathrm{Fe}_2\mathrm{O}_3 + 40\%\mathrm{CuO}\)

\[ \frac{1}{T}\times 10^{3} \]

Fig. 26.

Catalyst \(50\%\mathrm{Fe}_2\mathrm{O}_3 + 50\%\mathrm{CuO}\)

\[ \frac{1}{T}\times 10^{3} \]

Fig. 26–27. Dependence of \(\log K\) on the reciprocal of the absolute temperature for iron–copper catalysts.

Fig. 28.

Fig. 28–29. Dependence of \(\log K\) on the reciprocal of the absolute temperature for iron–copper catalysts.

to 48,000 cal/mol. At lower temperatures the cited authors did not carry out investigations.

It is also of interest to compare the activation energies for the oxidation of \(SO_2\) to \(SO_3\) in the presence of other catalysts. According to Taylor’s investigations, up to \(700^\circ C\) the apparent activation energy for a platinum catalyst is 16,000 cal/mol. As Maklakov and Arkhipova have shown \(^{24)}\), the activation energy in the presence of barium vanadate and calcium vanadate catalysts depends on temperature. Thus, in the temperature range \(400\)—\(450^\circ C\), for the first and second catalysts \(E\) proved to be equal, respectively, to 65,000 and 51,000 cal/mol; in the temperature range \(450\)—\(500^\circ C\) this quantity decreases sharply and, for both catalysts, as for platinum, proves to be only 16,000 cal/mol. Comparing the data cited with the results obtained by us, it should be noted that in the temperature range \(450\)—\(600^\circ C\) the activation energy for iron–copper catalysts is, as a rule, somewhat higher than the corresponding values for platinum and vanadium catalysts, but considerably lower than the values obtained by Kuzminykh and Turkhan for pyrite cinder.

Turning to an analysis of the mechanism of the process of oxidation of sulfur dioxide to sulfur trioxide in the presence of iron–copper catalysts, it should first of all be borne in mind that any catalytic process associated with the presence of a solid catalyst must consist of three consecutively occurring partial processes. The first of these is the process of supplying the initial gaseous components, in the present case \(SO_2\) and \(O_2\), to the surface of the catalyst and increasing their concentration in the adsorption layer. The degree of increase in concentration is determined both by the presence of simple physical adsorption, due to the existence of van der Waals attractive forces, and by activated adsorption, in which chemical affinity forces are manifested.

In the second partial process, the molecules adsorbed on the surface of the catalyst enter into interaction directly with one another or, combining with the substance of the catalyst, initiate a series of chemical processes in which the original bond of the atoms in the initial molecules is broken with the simultaneous appearance of the final reaction product or of intermediate compounds, which, reacting with one another or with the initial molecules, also give the final reaction product.

In the first case, when the reaction product is formed directly from the initial molecules, the acceleration of the reaction by the catalyst is explained by a simple increase in concentration and by a certain loosening of the reacting molecules due to the forces of physical adsorption; in the second case, when the reaction proceeds through intermediate compounds, the presence of sufficiently large forces of chemical affinity between the initial molecules and the substance of the catalyst is necessary.

And, finally, in the third period of formation the reaction products must be removed from the catalyst into the gas phase, so as to free the reactive surface for new portions of the initial substance.

All these processes, taking place in these three periods, proceed in time, and the total rate of the catalytic oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ will be determined by one of them, the one proceeding most slowly.

According to the investigations of Bodenstein $^{25}$), the oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ on a platinum catalyst proceeds exclusively as an adsorption process, the rate of which is determined by the rate of diffusion of $\mathrm{SO_2}$ or $\mathrm{O_2}$ through the film of $\mathrm{SO_3}$ adsorbed on the reaction surface.

Recently, in connection with the well-known objections against the diffusion theory of catalysis, Bodenstein rehabilitated his view by assuming that the final rate of the process is determined by the diffusion of $\mathrm{SO_2}$ or $\mathrm{O_2}$ (depending on the ratio $\mathrm{SO_2}$ to $\mathrm{O_2}$) in the layer of adsorbed two-dimensional gas from sites with low catalytic activity to active centers. Thus, in the present case there is no need to speak of the formation, from the catalyst and the reacting gaseous components, of intermediate, definite chemical compounds. Neumann $^{26}$), on the basis of the investigations of Wöhler, Foss, and Plüddemann $^{27}$), and Panzner and Goebel $^{28}$), assert that all noble metals (iridium, palladium, gold, silver), like platinum, bring about the oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ by adsorption catalysis.

As for oxide catalysts ($\mathrm{Fe_2O_3}$, $\mathrm{CuO}$, $\mathrm{Cr_2O_3}$, $\mathrm{V_2O_5}$), according to Neumann the process of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ in the presence of these substances proceeds chiefly through intermediate compounds (which can be detected by ordinary methods of chemical analysis), the formation of which is due to the manifestation of the ordinary forces of chemical affinity.

In the light of what has just been set forth, it is necessary first of all to determine whether ferric oxide and cupric oxide can not only bring about the physical adsorption of $\mathrm{SO_2}$ and $\mathrm{O_2}$, but also form with them more or less stable chemical compounds that could be regarded as intermediate compounds.

A number of investigators (Böther, Neumann and Goebel, and others) found that at $300^\circ$ and above, $\mathrm{SO_2}$ and $\mathrm{O_2}$ are completely or partially absorbed by $\mathrm{Fe_2O_3}$ without the simultaneous appearance of an equivalent quantity of $\mathrm{SO_3}$. Keppeler, as already indicated above, analyzing the composition of spent technical contact mass made of $\mathrm{Fe_2O_3}$, found the content of $\mathrm{H_2SO_4}$ to be 9.4%. It was further established that $\mathrm{Fe_2O_3}$, reacting with $\mathrm{SO_2}$ and oxygen, may, depending on the temperature, give both ferrous sulfate and ferric sulfate. In our experiments we likewise established that ferric oxide absorbs $\mathrm{SO_2}$ for a prolonged period of time. Cupric oxide also, reacting with $\mathrm{SO_2}$ and $\mathrm{O_2}$, gives the chemical compounds $\mathrm{Cu_2SO_4}$ and $\mathrm{CuSO_4}$.

Thus, between $\mathrm{Fe_2O_3}$, $\mathrm{CuO}$ and $\mathrm{SO_2}$ and $\mathrm{O_2}$ there exist sufficiently strong affinities, which lead to complete rupture of the bonds of the atoms in the molecules $\mathrm{SO_2}$ and $\mathrm{O_2}$ and to the formation of definite chemical compounds.

It remains to determine whether these sulfate compounds of iron and copper are the intermediate compounds that facilitate the process of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$, or whether their formation is due to side chemical reactions and the catalysis in this case proceeds either through other forms of compounds or is explained exclusively by adsorption phenomena. As is known, the successful course of a catalytic process through intermediate compounds can take place only when these compounds are not too stable, since otherwise, despite the ease of formation of intermediate compounds from the initial reacting molecules and the catalyst substance, their destruction will require considerable expenditure of energy, which will undoubtedly either stop or greatly slow down the overall process of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$. To establish whether the sulfates of iron and copper may be regarded as intermediate compounds, it is necessary, first of all, to consider the dissociation pressures and heats of formation of these compounds. Many investigators have already pointed to the connection between the dissociation pressure of sulfates and their catalytic activity. This was especially emphasized by Neumann, who asserted that the measure of the catalytic action of sulfates is the greater or lesser ease of their dissociation.

The dissociation pressure of ferric sulfate $[\mathrm{Fe_2(SO_4)_3}]$ has been studied by many investigators (d’Ans$^9$), Wöhler and Pluddemann$^{29}$), Bodenstein and Sudzuky$^{30}$), Wöhler and Grunzweig$^{31}$), but the results obtained by them agree little with one another. The same picture is also found for the data on the dissociation pressures of ferrous sulfate (the work of d’Ans and Greulich).

The works of the authors mentioned established that at a temperature of $600$—$610^\circ\mathrm{C}$ ferrous sulfate passes into ferric sulfate. In addition, the dissociation pressure of ferrous sulfate is greater than the corresponding value for ferric sulfate; thus, under the influence of temperature, $\mathrm{FeSO_4}$ will decompose considerably more vigorously.

For illustration, one may cite the data on dissociation pressures for $\mathrm{FeSO_4}$ and $\mathrm{Fe_2(SO_4)_3}$, measured respectively by Pluddemann and Greulich$^{32}$).

Dissociation pressures of $\mathrm{FeSO_4}$

Temp. in °C	235	251	315	376	386	433	482	522	614.5	631	634	645	654	680
$P_{\mathrm{mm}}\ \mathrm{Hg}$	1	2	10	20	28	32	73	84	254	295	328	587	548	760

Dissociation pressure of \(Fe_2(SO_4)_3\)
\(T\)-ra in °C	553	580	596	612	630	650	707
\(P_{\mathrm{mm}}\ Hg | 23 | 36 | 50 | 68 | 103 | 149 | 716 | If the partial pressure of \(SO_2\) in the gas used by us is taken into account \(\left(\dfrac{7.760}{100} = 53.2\ \mathrm{mm}\ Hg\right)\), then upon its complete oxidation the gas phase will contain \(\sim 53\ \mathrm{mm}\ Hg\) of \(SO_3\). Comparing this value with the tabulated data for the dissociation pressure of \(FeSO_4\), it can be seen that \(FeSO_4\) should decompose according to the equation \(2FeSO_4 \rightleftarrows Fe_2O_3 + SO_3 + SO_2\) at temperatures between \(\sim 450\) and \(500^\circ C\); the decomposition temperature for \(Fe_2(SO_4)_3\) will lie somewhat higher, namely around \(600^\circ C\). However, in fact, at the considerably lower degree of oxidation of \(SO_2\) to \(SO_3\) that was observed in the experiments, the decomposition of \(FeSO_4\) and \(Fe_2(SO_4)_3\) can occur at considerably lower temperatures. According to the investigations of Pluddemann, the dissociation pressure of \(CuSO_4\) at temperatures below \(640^\circ C\) is somewhat higher than the analogous value for \(Fe_2(SO_4)_3\); at higher temperatures this relation becomes the reverse. Thus, for the indicated gas composition as well, depending on the degree of oxidation of \(SO_2\) to \(SO_3\), copper sulfate can thermodynamically be formed and decomposed.

However, the thermodynamic possibility alone of the formation and decomposition of chemical compounds from the catalyst and the initial reacting gaseous components is not yet sufficient for the rapid course of the reaction. In addition, it is also necessary that the simultaneous energy expenditures not be too large, since the rapid communication of large portions of energy by the reacting gaseous components is a kinetically unlikely process. From this point of view it becomes understandable why substances such as copper sulfates and titanium oxides, which, in comparison with the sulfates of other elements, have greater dissociation pressures, turned out to be only slightly active catalysts for the oxidation of \(SO_2\) to \(SO_3\).

As was already indicated, we investigated the activity of iron-copper catalysts at temperatures up to \(600^\circ C\). For this temperature range, according to Neumann \(^{26}\), the kinetics of the process of oxidation of \(SO_2\) to \(SO_3\) is determined mainly by the rate of the reaction:

\[ 2FeSO_4 \longrightarrow Fe_2O_3 + SO_2 + SO_3. \]

According to our calculations, based on Hess’s law*), the heat effect of this reaction is endothermic and equal to \(-78500\) cal.

) Beskov.—Introduction to Technochemical Calculations* \(^{33}\).

at room temperature. If one takes into account that the activation energy of an endothermic reaction should be approximately equal to its heat effect, then, comparing the value of the activation energy found by us for a catalyst made of pure $\mathrm{Fe_2O_3}$ ($26800$ cal/mole $\mathrm{SO_3}$) with the heat effect of the decomposition reaction of ferrous sulfate ($-78500$ cal), one may conclude that the process of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ proceeds by another route. It is more probable to suppose that the kinetics of the given process in the presence of an iron-oxide catalyst is determined by the following reaction:

\[ 3\mathrm{Fe_2O_3}+\mathrm{SO_2}=2\mathrm{Fe_3O_4}+\mathrm{SO_3} \tag{1} \]

According to our calculations, the heat effect of this reaction at room temperature is $-31100$ cal/mole $\mathrm{SO_3}$, which is in satisfactory agreement with the magnitude of the apparent activation energy. The $\mathrm{Fe_3O_4}$ formed is rapidly oxidized again by oxygen according to the reaction:

\[ 4\mathrm{Fe_3O_4}+\mathrm{O_2}=6\mathrm{Fe_2O_3}+108090\ \text{cal} \tag{2} \]

The proposed mechanism also agrees with the form of the kinetic equation (1), according to which, in particular, the rate of oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ is directly proportional to the concentration of $\mathrm{SO_2}$ in the first power.

In addition to the indicated reactions 1 and 2, other reactions between $\mathrm{SO_2}$, $\mathrm{O_2}$, and $\mathrm{Fe_2O_3}$ may proceed simultaneously, leading, for example, to the formation of ferrous and ferric sulfates; however, they should be regarded not as the main reactions but as side reactions. Side reactions must influence the activity of the catalyst, since the sulfates formed can change the physicochemical structure and the structure of the catalyst surface.

The introduction of cupric oxide additives into an iron catalyst may increase, leave unchanged, or decrease the activation energy. Thus, an increase in the activation energy occurs for catalysts with 1 and 40% $\mathrm{CuO}$, for which this value is, respectively, $34300$ and $31600$ cal/mole.

The introduction into the catalyst of 50 and 70% $\mathrm{CuO}$ in the temperature range $450$–$600^\circ\mathrm{C}$ does not change the activation energy; or, finally, a substantial decrease in this value is observed for catalysts with 10 and 20% $\mathrm{CuO}$, for which the apparent activation energies are, respectively, $16300$ and $14300$ cal/mole.

It may be assumed that in all those cases where the activation energy for complex iron–copper catalysts (e.g., catalysts with 1; 40; 50; 70% $\mathrm{CuO}$) is approximately equal to the activation energy of a catalyst obtained from pure iron oxide, the mechanism of the oxidation of $\mathrm{SO_2}$ to $\mathrm{SO_3}$ on these catalysts remains the same as the mechanism for the pure iron ca-

catalyst. The CuO additives introduced in the indicated proportions only change the structure and activity of the reaction centers, which is reflected in the observed oxidation rate. For those complex catalysts we studied for which the activation energy is substantially lower than that for pure ferric oxide—for example, catalysts with 10 and 20% CuO—the mechanism of the oxidation of SO₂ to SO₃ must apparently be different.

In this case the following reactions may be assumed:

\[ 2\mathrm{CuO} + \mathrm{SO}_2 = \mathrm{Cu}_2\mathrm{O} + \mathrm{SO}_4 \tag{3} \]

\[ 2\mathrm{Cu}_2\mathrm{O} + \mathrm{O}_2 = 4\mathrm{CuO} \tag{4} \]

which is in agreement with the mechanism of the oxidation of SO₂ to SO₃ in the presence of a catalyst consisting of pure CuO, proposed by Neumann.

The thermal effects of reactions (3) and (4), according to our calculations, proved to be, respectively, equal to −11900 and +69600 cal. As can be seen, the activation energies for these catalysts exceed only slightly the negative heat of reaction (3). Thus, in the present case, apparently, the oxidation of SO₂ to SO₃ proceeds chiefly by the interaction of SO₂ with cupric oxide, which, as in the case of the catalyst made of pure ferric oxide, is also confirmed by the form of the kinetic equations (1 and 2′). The low values of the apparent activation energy for catalysts with 50–70% cupric oxide (13200, 15400 cal/mol) in the low-temperature region (350–400°C) apparently indicate that, under these conditions, the oxidation of SO₂ to SO₃ also proceeds by interaction of SO₂ with CuO. However, at higher temperatures the active forms of CuO are destroyed (for example, as a result of sulfation, etc.), and catalysis through the interaction of SO₂ with Fe₂O₃ begins to acquire the dominant importance; this apparently explains the presence of maxima on the curves of percent oxidation = f(t°). The activating and passivating influence of CuO at elevated temperatures (450–600°C) is apparently connected with a change in the physicochemical structure of the catalyst, both of CuO itself and of the products of its interaction with the initial gaseous components.

To clarify the causes responsible for the appearance of a maximum on the curves (percent oxidation) = f(t°) at some cupric oxide contents in the mixed iron–copper catalyst and their absence at other proportions between Fe₂O₃ and CuO, and also to answer the question of why cupric oxide, introduced into an iron catalyst in different amounts, can increase and, in some cases, decrease the activity of the mixed catalyst, it is advisable to carry out X-ray diffraction and electron-microscopic investigations of the structure of these catalysts.

Table 4

Values of the apparent activation energies of the process of oxidation of \(SO_2\) to \(SO_3\) in the presence of iron–copper catalysts

Composition of the initial form of the catalyst	Apparent activation energy \(E_{\text{app.}}\), cal/mol \(SO_3\)	Temperature range, °C
\(Fe_2O_3\)	26800	400–600
\(Fe_2O_3 + 1\%\ CuO\)	34300	450–600
\(Fe_2O_3 + 10\%\ CuO\)	16300	450–600
\(Fe_2O_3 + 20\%\ CuO\)	14300	450–600
\(Fe_2O_3 + 40\%\ CuO\)	31600	450–600
\(Fe_2O_3 + 50\%\ CuO\)	27000	500–600
\(Fe_2O_3 + 50\%\ CuO\)	13200	350–400
\(Fe_2O_3 + 60\%\ CuO\)	21000	450–600
\(Fe_2O_3 + 60\%\ CuO\)	22100	350–400
\(Fe_2O_3 + 70\%\ CuO\)	25100	450–600
\(Fe_2O_3 + 70\%\ CuO\)	15400	350–400

Conclusions

1. An experimental setup has been described for precise measurements of the activity of catalysts by the dynamic method under conditions close to isothermal.

The initial forms of the eleven catalysts investigated for the oxidation of \(SO_2\) to \(SO_3\) had the following compositions: \(Fe_2O_3\); \(Fe_2O_3 + 1\%\ CuO\); \(Fe_2O_3 + 4\%\ CuO\); \(Fe_2O_3 + 8\%\ CuO\); \(Fe_2O_3 + 10\%\ CuO\); \(Fe_2O_3 + 20\%\ CuO\); \(Fe_2O_3 + 30\%\ CuO\); \(Fe_2O_3 + 40\%\ CuO\); \(Fe_2O_3 + 50\%\ CuO\); \(Fe_2O_3 + 60\%\ CuO\); \(Fe_2O_3 + 70\%\) (by weight) \(CuO\).

The activity study was carried out by passing through the catalysts, dried over concentrated sulfuric acid, a gas mixture containing \(7\%\,SO_2\) (the remainder being air), in the temperature range \(350–600^\circ C\), at space velocities from 200 to 1000.

1. The time required to establish a steady state, characterized by constancy of the degree of contact at \(350^\circ C\) and a space velocity of 200, increases with increasing percentage content of \(CuO\) in the catalyst.

2. At constant temperature, for all the catalysts studied, with increasing gas space velocity the degree of oxidation of \(SO_2\) to \(SO_3\) decreases; moreover, this decrease, as a rule, occurs considerably more slowly than the space velocity increases.

3. At all space velocities, in the presence of a catalyst made from pure iron oxide, with increasing temperature the degree of oxidation in the temperature range studied continuously...

increases. An analogous temperature dependence occurs for catalysts with 1, 10, 20, and 30% CuO. The temperature dependence of the activity of catalysts with contents of 4, 8, 40, 50, 60, and 70% CuO has a different form. Namely, in the temperature region 350–400°C the activity increases with increasing temperature; in the temperature interval 400–450°C it decreases, and then, with a further increase in temperature, the activity again rises. For the catalyst with 4% CuO the first maximum of activity lies at a somewhat higher temperature—450°C.

1. No simple dependence of the type that could be predicted from the rule of additivity is observed between the amount of copper oxide in the mixed catalyst and its catalytic activity. The curves [% oxidation] \(tv = f\) (% CuO), at constant temperature and space velocity, have a complex character, determined by the presence of several maxima and minima of activity; moreover, the positions of the latter on the curves are determined not only by the % CuO in the catalyst, but also by the temperature.

2. The introduction into an iron-oxide catalyst of additions of copper oxide in certain proportions can both increase and, in some cases, decrease the activity of the mixed catalyst in comparison with the activity of the catalyst made of pure iron oxide.

For the majority of the temperatures and space velocities investigated, the most active catalysts proved to be those with a content of about 40% copper oxide. However, the maximum percentage oxidation of SO₂ to SO₃, equal to 73.4% (space velocity = 200), was observed at a temperature of 550°C in the presence of a catalyst containing only 4% CuO.

The particularly strong activating influence of copper oxide appears at low temperatures (350–400°C); under these conditions, for catalysts with a high copper-oxide content (from 40 to 70%), the increase in activity, as compared with the analogous value characteristic of pure iron oxide, amounts to from 6 to 20 times.

1. For the majority of the catalysts and temperatures investigated by us (including iron oxide), the kinetics of the process of oxidation of SO₂ to SO₃ is expressed by the Taylor and Lehner equation:

\[ \frac{dx}{dt}=K\frac{(X_m-x)}{\sqrt{ax}}, \]

which is confirmed by the satisfactory constancy of the rate constants calculated at different space velocities of the gas stream.

1. Using the Arrhenius equation, activation energies were calculated for those catalysts for which the Taylor and Lehner equation is applicable. For the catalyst obtained from pure iron oxide, \(E_{\text{app.}} = 26800\) cal/mole SO₃. The introduction of copper oxide into the catalyst in definite ratios

is capable of both somewhat increasing and decreasing the apparent activation energy relative to the analogous value characteristic of pure iron oxide. The maximum and minimum values of \(E_{\text{apparent}}\) in the temperature range \(450\text{—}600^\circ\mathrm{C}\) correspond, respectively, to catalysts with 1 and 20%.

1. An assumption is made regarding the mechanism of the catalytic oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\). In the presence of a purely iron catalyst, the following reactions apparently play the principal role:

\[ 3\,\mathrm{Fe_2O_3}+\mathrm{SO_2}=2\,\mathrm{Fe_3O_4}+\mathrm{SO_3}\ldots\ldots\ldots (1) \]

\[ 4\,\mathrm{Fe_3O_4}+\mathrm{O_2}=6\,\mathrm{Fe_2O_3},\ldots\ldots\ldots (2) \]

where the slowest reaction is reaction (1). In the presence of mixed iron–copper catalysts, especially those for which the activation energy is low, the oxidation of \(\mathrm{SO_2}\) to \(\mathrm{SO_3}\) may also proceed by the following reactions:

\[ 2\,\mathrm{CuO}+\mathrm{SO_2}=\mathrm{Cu_2O}+\mathrm{SO_3}\ldots\ldots\ldots (3) \]

\[ 2\,\mathrm{Cu_2O}+\mathrm{O_2}=4\,\mathrm{CuO}\ldots\ldots\ldots (4) \]

Here the slowest reaction is reaction (3). The \(\mathrm{SO_3}\) that is formed retards the process, since it screens part of the active surface of the catalyst.

REFERENCES

1. Waeser,—Handb. Schwefelsäurefabrikation, Braunschweig, 1930.
   Lukyanov,—Course of Chemical Technology of Mineral Substances, ONTI, 1935.
   Lukyanov,—Production of Sulfuric Acid by the Contact Oxidation Method, State Technical Publishing House, 192?
   Malin,—Technology of Sulfuric Acid and Sulfur, Part II, ONTI, 1937.
   Kuzminykh,—Production of Sulfuric Acid, ONTI, 1937.
2. Blondeau,—Compt. rend., 29, 405, 1849.
3. Wöhler and Mahla,—Ann. Chem. Pharm., 81, 255, 1852.
4. O. Kaush,—Catalysts, etc., ONTI, 1934.
5. Lunge and Pollit,—Z. angew. Chem., 15, 1105, 1902.
6. Lunge and Reinhardt,—Z. angew. Chem., 17, 1041, 1904.
7. Knietsch,—Ber. deutsch. Chem. Gesellsch., 34, 4069, 1901.
8. Keppeler,—Z. angew. Chem., 21, 552, 1908.
9. d’Ans,—Dissertation, Darmstadt, 48, 1905.
10. Wöhler,—Zt. Phys. Chem., 62, 64, 1903.
11. Neumann and Goebel,—Z. Elektrochem., 11, 734, 1928.
12. Neumann and Jüttner,—Z. Elektrochem., 2, 87, 1939.
13. Lukyanov with co-workers,—Zhurn. Khim. prom., 6, 5, 772, 1933.
14. Kuzminykh and Turkhan,—Zhurn. Khim. prom., 6, 22, 1931.
15. Neumann,—Z. Elektrochem. angew. Phys. Chem., 5, 304, 1932.
16. Bodenstein and Fink,—Z. Phys. Chem., 60, 1, 1907.
17. Kertis,—Fixed Nitrogen, ONTI, 1934.
18. Jarson and Brooks,—Jud. Eng. Chem., 18, 1305, 1926, et al.
19. Epstein,—Chemistry and Technology of Fixed Nitrogen, ONTI, 1935.
20. Kurin,—Zhurn. Khim. prom., 7, 475, 1937.
21. Taylor and Lehner,—Z. Phys. Chem., Festband, 1931.
22. Boreskov and Pligunov,—Zhurn. Prikladn. khimii, 5, 785, 1933.
23. Arrhenius,—Z. Phys. Chem., 4, 226, 1889.
24. Maklakov and Arzhanova,—Khimstroi, 1934, 6, 318.
25. Bodenstein,—Z. Phys. Chem., (B), 2, 345, 1929.
26. Neumann,—Z. Elektrochem., 1, 42, 1929.

1. Wöhler, Foss, and Pludderman.—Ber. Chem. Ges. 39, 3538, 1906.
2. Neumann, Panzner, and Goebl.—Z. Elektrochem. 34, 702, 1925.
3. L. Wöhler, Pluddermann, and P. Wöhler.—Z. Phys. Chem. 62, 641, 1908.
4. Bodenstein and Suzuki.—Z. Elektrochem. angew. Phys. Chem., 16, 912, 1910.
5. Wöhler and Grunzweig.—Berichte deutsch. Chem. Gesellsch., 46, 1857.
6. Greulich.—Z. anorg. Chem., 168, 197, 1928.
7. Beskov.—Introduction to Technochemical Calculations. GONTI, 1938. Handbook of Physico-Chemical Technical Quantities. Technical Encyclopedia, vol. 7, 1931.

1. visible footnote marker in the source; footnote text is not visible on this page. ↩