J. Chem. SOC., Faraday Trans. I, 1982, 78, 3281-3286 Production of CO from CO, by Reduced Indium Oxide BY KIYOSHI OTSUKA,* TAKAO YASUI AND AKIRA MORIKAWA Department of Chemical Engineering, Tokyo Institute of Technology, Ookayama, Meguro-ku, Tokyo 152, Japan Received 4th January, 1982 The forward and backward reactions of the process In,O,(s) + xCO(g) $ In20a-z(s) + xCO,(g) proceed rapidly and repeatedly at 673 K when the degree of reduction of In,O, is low. However, the yield of the backward reaction decreases as the degree of reduction rises above ca. 40%. This is ascribed to a decrease in the number of surface-active sites for the decomposition of CO,. The free-energy change of the reaction evaluated from the ratio of partial pressures P(CO,)/P(CO) at equilibrium agreed well with results obtained by other workers, assuming indium metal as the reduced species.Since carbon dioxide is the most stable and most highly oxidized state of carbon, temperatures > 2000 K are needed to decompose it thermally into carbon monoxide and oxygen. Recently we have reported that the decomposition of water proceeds smoothly at or above 573 K on reduced indium oxide.' The preliminary reduction of indium oxide has been carried out by various reductants such as carbon, biomass,l hydrogen or carbon monoxide.2T We may expect that the decomposition of carbon dioxide should also occur on reduced indium oxide according to the following (1) reaction : In,O,-, + xC0, -+ In,03 + xC0 where In203-, represents crystalline In,O or a mixture of indium metal and In,03. In this report we examine the conditions for the decomposition of CO, on a reduced sample of In,O,.The existence of crystalline In20 has been suggested by several w o r k e r ~ . ~ - ~ However, there is still controversy concerning its exi~tence.''~ Hence the reduced state of In,03 will be described on the basis of the free-energy change for eqn EXPERIMENTAL (1). The In,03 used was a reagent-grade powder (purity > 99.9%) obtained from the Wako Pure Chemical Co. The surface area of the fresh In,O, measured by nitrogen adsorption (using the B.E.T. method) was 13.9 m2 g-l. Carbon dioxide was purified three times by trap-to-trap distillation. Carbon monoxide was purified by passing it through a silica-gel adsorbent cooled at 195 K. The experiments were carried out using a conventional mercury-free gas-circulation apparatus of cu.320 cm3 volume. After the partial reduction of In,O, by CO or H,, the gases in the system were pumped out and the decomposition of CO, was started by adding CO, (21.3 x lo3 Pa). The progress of the reaction was followed by measuring the amounts of CO and CO, present by gas-chromatographic analysis. RESULTS AND DISCUSSION Fig. 1 shows repeated forward and backward reactions of step (1) at 673 K. Before the experiments the In,O, sample was reduced by hydrogen at 673 K to a degree of 328 13282 PRODUCTION OF CO USING In,O, reduction of 27.8%, as determined on the basis of the oxygen atoms removed. The right-hand vertical axis indicates the percentage recovery of the reduced oxide to In,O, evaluated from the quantity of CO produced.Decomposition of CO, proceeded rapidly at 673 K and the reaction was almost complete within 50 min. The condensation of CO, at 77 K (at A) caused the backward reaction; i.e. the reduction of the oxide was initiated by CO which had been formed before point A. Evaporation of the condensed CO, at point B reinitiated the decomposition of CO,. The results in fig. 1 show that both the forward and backward reactions of equilibrium (1) pro- ceed smoothly and repeatedly at 673 K. 6.0 - 0 E -3 4.0 2 z P I 0 - 3 E .- 2 2.0 cc 0 4- t: m F 0 t B I 1 I 50 100 150 ti me / m in 100 m 0 N e 0 x 0 I u 50 & 0) w * E & 0) a 1 FIG. 1 .-Repeated forward and backward reactions of equilibrium (1) at 673 K. The weight of In,O, used was 0.20 g.Fig. 2 shows the percentage recovery of the reduced oxide to In,O, by CO, at 673 K as a function of time for various samples with different initial degrees of reduction. The positions of the downward and upward arrows indicate the length of time for which the temperature of the reactor increases and the time at which the temperature reaches 773 K, respectively. Although the rate of CO production for the sample having an 85% degree of reduction was very slow at 673 K, the reaction was completed rapidly when the temperature was raised to 773 K. Fig. 3 shows the effect of CO, pressure on the rate of formation of CO at 673 K for samples of 1 1.4 and 8.1 % degree of reduction. The rate depends on the pressure of CO,. This suggests that the rate-determining step is not the solid-state diffusion process of a reduced species of the oxide but the reaction of CO, with this species on the surface.The large difference observed between the apparent activation energies of the decomposition of water (50.1 kJ rn~l-')~ and of CO, (ca. 110 kJ mol-I, see fig. 5 ) also supports the above considerations. The quantity of CO produced and the percentage recovery of the reduced sample obtained from fig. 2 are plotted in fig. 4 as functions of the degree of reduction of In,O,. The maximum in the amount of CO produced in 20 min lies at 2540% reduction. For the sample having a low degree of reduction, the rate of CO productionK . OTSUKA, T. YASUI AND A. MORIKAWA 3283 FIG. 2.-Percentage recovery of the reduced oxide to In,O, as a function of time for samples with different initial degrees of reduction.Initial degree of reduction and weight of the sample as follows: V, 4.9% (0.50 g); 0 , 9 . 6 % (0.50 g); D, 17.7% (0.20 g); V, 19.1% (0.20 g); A, 27.8% (0.20 g); 0,42.6% (0.20 g); 0, 55.3% (0.50 g); 0, 65.1 % (0.20 g); A, 79.6% (0.20 8); a, 84.8% (0.10 g). 0 5 10 15 20 p(CO,)/ lo3 Pa FIG. 3.-Effect of CO, pressure on the rate of CO production for the following degrees of reduction of In,O,: 0, 11.4%; A, 8.1 %. is fast enough to complete the reaction in 50 min, as can be seen in fig. 2. Hence, the quantity of CO produced for samples of low degrees of reduction is restricted by the quantity of reduced species able to react with CO,. On the other hand, the quantity of CO produced is governed by the rate of reaction of CO, with the reduced species for the sample having a high degree of reduction.Both the percentage recovery and the quantity of CO produced in 20 min decrease sharply when the degree of reduction is increased above ca. 40%. The rates of decomposition of CO, have been measured3284 PRODUCTION OF CO USING In,O, at two different temperatures (598 and 648 K) for samples with the same degree of reduction. Sets of rates were obtained for samples of different degrees of reduction (1 0.9-80.2%). The apparent activation energies and pre-exponential factors, obtained from an Arrhenius plot of the rates at the two different temperatures, are plotted as functions of the degree of reduction in fig. 5 . The results in the figure indicate that the activation energies do not change with the degree of reduction but that the pre-exponential factors drop sharply above ca.30% reduction. This suggests that the nature of the reduced In,O, species active in CO, decomposition does not change with the degree of reduction up to ca. 80%. The drop in the values of the pre-exponential factor strongly suggests that the sharp decrease in the yield of CO above ca. 40% reduction level (fig. 4) can be ascribed to the decrease in the number of surface-active sites to react with CO,, probably owing to the decrease in the surface area of the reduced particles caused by their sintering. 8 8 6, 0 2 x Y 0 50 100 degree of reduction (%) FIG. 4.-Effect of the degree of reduction on the quantity of CO produced and the percentage recovery to In,O,.Quantity of CO produced: A, in 5 min; A, in 20 min. Percentage recovery to In,O,: 0, in 20 min; 0, final value at 773 K. The final percentage recovery to In,O, at 773 K (closed circles in fig. 4) becomes lower as the degree of reduction decreases, which can be ascribed to the presence of highly reducible surface oxygen atoms on fresh In,O,: such oxygen atoms cannot be reformed by the oxidation with CO,., The dashed curve in fig. 4 is the least-squares plot for the closed circles assuming the following equation: Y = 100(1-;) (2) where a is a constant, Y is the percentage recovery to In,O, and X is the degree of reduction of the sample. The number of active surface oxygen atoms evaluated from the value of a is 2.0 x mol (g In203)-l. The crosses in fig.4 are the percentages of final recovery to In20, observed in the case of water decomp~sition.~ The number of active surface oxygen atoms obtained from water-decomposition data is 1.8 x mol g-l. Both values are in fair agreement with each other.K. OTSUKA, T. YASUI A N D A. MORIKAWA 3285 150 - I - 0 E 21 --. x $100 S a, s 0 m .A Y .- + U 50 2 a a 0 4,000 100 50 10 50 70 0 degree of reduction (%) FIG. 5.-Apparent activation energy of CO production and pre-exponential factor as functions of the degree of reduction of In,O,. TABLE 1 .-EXPERIMENTAL VALUES OF P(CO,)/P(CO) AND THE FREE-ENERGY CHANGE OF EQN (3) AGO of eqn (3)/kJ temp./K P(CO,)/P(CO) this work ref. (10) ref. (1 1 ) ref. (12) 550 0.053 _+ 0.008 - 40.2 -41.2 - 38.9 - 44.4 600 0.088 f 0.003 - 36.3 - 38.7 - 35.6 -41.0 646 0.1 13 f 0.003 - 35.2 - 36.3 - 32.6 - 38.0 673 0.134 0.008 - 33.7 - 34.9 - 30.8 - 36.2 700 0.158 f 0.006 - 32.9 - 33.5 - 28.9 - 34.4 773 0.245 &0.001 - 27.2 - 29.7 - 23.8 - 29.5 Broch and Christensen8 could not prove the existence of crystalline In,O after detailed X-ray analysis of reduced In,O,.Our X-ray spectroscopic studies on the partially reduced In,O, could only indicate the existence of In,O, and indium metal. Hence, the reduced state of the oxide has been studied by measuring the equilibrium pressures of CO, and CO over the partially reduced indium oxide (< 51 % degree of reduction) at different temperatures. The ratios of the observed pressures of CO, and CO are shown in the second column of table 1 and yield the free-energy change of reaction (3) as shown in the third column of table 1.Stubbs et aLIO and Hochgeschwender and Ingrahaml' determined the free-energy change for reaction (4) 2 In(1) + 3 CO,(g) In,O,(s) + 3 CO(g) (3) using gas-equilibration techniques. Combining their data with literature data for the free energies of formation of water vapour and of CO, gas,12 the free-energy changes3286 PRODUCTION OF co USING h203 of eqn (3) were calculated at different temperatures. They are listed in the fourth and fifth columns of table 1. The values listed in the last column were obtained from the free energies of formation of CO, and those of In203 determined by Newns and Pelmore13 using e.m.f. measurements with the galvanic cell In(l), In,03(s)~0.85Zr02 + O . 1 5CaO(Ni(s), NiO(s). The values obtained in this work agree well with those of other workers and support the previous suggestion3 that the In203-,(s) in equilibrium (1) is a mixture of indium metal and In203. K. Otsuka, Y. Takizawa, S. Shibuya and A. Morikawa, Chem. Lett., 1981, 347. K. Otsuka, T. Yasui and A. Morikawa, J. Catal., 1981, 72, 389. K. Otsuka, T. Yasui and A. Morikawa, Bull. Chem. Soc. Jpn, 1982, 55, 1768. W. Klemm and H. U. V. Vogel, Z. Allg. Chem., 1934, 219, 45. K. A. Klinedinst and D. A. Stevenson, J. Chem. Thermodyn., 1973, 5, 21. T. J. Anderson and L. F. Donaghey, J. Chem. Thermodyn., 1977, 9, 617. L. Brewer, Chem. Rev., 1953, 52, 38. N. C. Broch and A. N. Christensen, Acta Chem. Scand., 1966, 20, 1996. A. J. Van Dillen, J. W. Geus and J. H. W. de Wit, J . Chem. Thermodyn., 1978, 10, 895. K. Hochgeschwender and T. R. Ingraham, Can. Metall. Q., 1967, 6, 293. I. Barin and 0. Knacke, Thermochemical Properties of Inorganic Substances (Springer-Verlag, Berlin, 1973 and 1977). lo M. F. Stubbs, J. A. Schufle and A. J. Thompson, J. Am. Chem. Soc., 1952, 74, 6201. l3 G. R. Newns and J. M. Pelmore, J. Chem. Soc. A, 1968, 360. (PAPER 2/007)