J . Chem. Soc., Faraday Trans. I, 1986,82, 2665-2672 Catalytic Decomposition of Isopropanol over Chromite Spinels MCr,O, (M = Ni, Mn and Mg) Krithivasa Balasubramanian Applied Science Division, Madras Institute of Technology, Anna University, Madras 600 044, India Vengaimuthu Krishnasamy Department of Chemical Engineering, A .C. College of Technology, Anna University, Madras 600 025, India The decomposition of isopropanol over nickel, manganese and magnesium chromite spinel catalysts has been investigated in the vapour phase in an integral reactor. Its decomposition follows first-order kinetics. Kinetic and thermodynamic parameters have been calculated using the Arrhenius and Eyring equations. The activity pattern is found to be NiCr,O, > MnCr,O, > MgCr,O,. The chromite spinels have been characterised by X-ray studies, i.r.spectral analysis, conductivity and thermoelectric potential measurements. All three chromites were found to be p-type semiconductors in the temperature range 150-400 "C. Exclusive dehydrogenation is shown by NiCr,O, and MnCr,O,, whereas MgCr,O, functions as a dehydro- genation and dehydration catalyst. A linear correlation exists between the entropy of activation and the activation energy for electrical conduction for the chromite spinels studied. Spinels are a group of inorganic compounds represented by a general formula AB,0,,1-3 where A and B are bivalent (Mg, Mn, Co, Ni, Cu and Zn) and trivalent (Al, Cr, Fe) metals, respectively. Spinels are thermally stable and they maintain enhanced and sustained activities for a variety of industrially important reactions such as decomposition of nitrous ~ x i d e , ~ hydrodesulphurisation of crude petrole~m,~ oxidation and dehydro- genation of hydrocarbons6? and oxidation and methanation of carbon This study involves the preparation, characterisation and comparison of the catalytic activity of nickel, manganese and magnesium chromite spinels.Experiment a1 Catalyst Preparation A mixture of a 10% solution of the metal nitrates was taken in the ratio of Cr: M = 2: 1 and the mixture was heated to 60-80 "C. To this hot mixture a 5% ammonia solution was added dropwise with constant and uniform stirring. The solution was maintained at pH 9 during the precipitation. The mixture was digested for another 2 h at this temperature to complete precipitation.The precipitate was filtered and dried at 105 "C for 24 h. Calcination of magnesium and nickel chromites was done at 700 "C in a muffle furnace for 8 h in a current of pure dry air and manganese chromite in hydrogen. The purity of isopropanol (AR, BDH) was tested by gas chromatography. 26652666 Catalytic Decomposition of Isopropanol Table 1. X-Ray d spacing data (A) on NiCr,O,, MnCr,O, and MgCr,O, ~ NiCr,O, MnCr,O, MgCr204 plane dlit.b dobs.a dcalc. dobs.a dlit. 111 4.78 w 4.79 (20) - - 4.80 m 4.813 (65) 220 2.90 m 2.93 (30) 3.145 m 2.98 2.94 w 2.95 (14) 31 1 2.49 vs 2.50 (100) 2.542 vs 2.54 2.51 vs 2.51 (100) 400 2.05 m 2.07 (35) 2.11 m 2.1 1 2.10 s 2.08 (55) 422 1.68 w 1.70 (15) 1.72 w 1.72 - - 511, 333 1.58 s 1.60 (60) 1.625 s 1.624 1.587 m 1.603 (40) 440 1.47 s 1.47 (80) 1.496 s 1.492 1.468 s 1.473 (55) a vs = Very strong; s = strong; m = medium; w = weak.Peak intensity in parentheses. Powder diffraction file lattice parameter: NiCr,O, = 8.331 A, MnCr,O, = 8.437 A, MgCr,O, = 8.329 A. Table 2. I.r., conductivity and Seeback potential data on NiCr,O,, MnCr,O, and MgCr,O, ~ metal chromites Cr-0 Cr-0 M-0 M-0 Ua and colour obtained reported obtained reported EJeV /pV OC-l NiCr,O, 575-585 510 630-640 630 1.066 380 MnCr,O, 490-5 10 490 610-630 620 1.054 92 MgCr204 525 525 600-620 650 0.652 6.8 (green) (grey ) (grey-green) a Positive Seeback potential indicates that all three chromite spinels are p-type semiconductors in the temperature range studied. Catalyst Characterization X-Ray Study The spinels were characterised by X-ray powder pattern studies1° using a Philips X-ray diffractometer (PW lOSO), and 'd' spacings and lattice parameter values are given in table 1, along with the reported data for comparison.ll I.R.Study The i.r. spectra of spinels were obtained in a double-beam spectrophotometer (Perkin- Elmer 983). The maxima of the absorption bands for Cr-0 (580-490 cm-l) and M-0 (650-620 cm-l) are given in table 2, along with the reported data for comparison.ll Conductivity and Thermoelectric Potential Measurements Electrical conductivity measurements12 were carried out using a two-probe conductivity cell in the temperature range 150-400 "C. The activation energies for electrical conduction (E,) were obtained from the slope of the plot of log 0 vs.T1/2 and the values are included in table 2. Thermoelectric (Seeback) potential, a meas~rements~~ for all three chromites were also carried out in this temperature range. The nature of charge carriers (holes orK. Balasubramanian and V . Krishnasamy 2667 electrons) was determined by checking the sign of the Seeback potential. For this purpose the following modification in the conductivity measuring apparatus was made. An auxiliary heating element was wound over a silica tube and kept at one end of the conductivity cell containing the catalyst pellet. After equilibrating the sample at each temperature for 15 min, the auxiliary heater was switched on and a temperature difference of ca. 10 K was maintained between the sample ends. After a steady state had been attained the Seeback potential was measured using a d.c.microvoltmeter. In this way, the Seeback potential has been measured at various temperatures. The cold end of the conductivity cell was positive, indicating p-type semiconduction (hole conduction) and the a values are given in table 2. Conductivity data of many spinels have been related to their magnetic properties, both of which in turn depend on the crystal structure.14~ l5 Goodenough,16 using a one-electron energy diagram, has found out that it is the octahedral (B site) cations which are responsible for electrical conduction by virtue of site symmetry in an AB,O, spinel. Neel17 has explained the magnetic behaviour of ferrite spinels on the basis of exchange interactions that occur between the transition-metal ions at A and B sites in the spinel structure.The three possible interactions are A-A, A-B and B-B. A-B interactions can occur through 125" A-0-B super-exchange type, while B-B interactions can be either 90" B-0-B super-exchange type or a direct B-B one. The A-A interaction is usually very weak owing to the large distance between the A-site cations. Thus the relative magnitude of interaction in spinels is in the following order: AB > BB > AA. All three exchange interactions are normally antiferromagnetic and the spin vector would tend to align antiparallel. 18* l9 In NiCr,O, and MnCr,O, all the three exchange interactions are present and the distance between the Cr3+-Cr3+ ions is shortened owing to the higher ionic radii of Ni2+ (0.69 A) and Mn2+ (0.80 A).The resulting antiferromagnetic order thus enhances the activation energy for electrical conduction (table 2). MgCr,O, has a much lower activation energy for electrical conduction owing to the very weak B-B interaction (Mg2+ non-magnetic, radius 0.66 A). Apparatus and Procedure The reactions were carried out in a fixed-bed flow-type integral reactor, 50 cm long with 1.5 cm internal diameter in the temperature range 290-380 "C and at different contact times, W/F (W is the weight of the catalyst and I; is the weight rate of the reactant per hour). Pyrex glass beads (4 mm diameter) were placed above the catalyst bed to a height of 5 cm. The reactor tube was inserted into a cylindrical furnace and heated electrically to the requisite temperature. Using a thermocouple the temperature was monitored along the length of the catalyst bed and the temperature required for a particular run was maintained constant.The liquid products, containing acetone and unreacted isopropanol, were identified and estimated by a CIC gas chromatograph (FFAP column, f.i.d. detector, column temperature 90 "C, injection port temperature 150 "C, detector temperature 90 "C, carrier gas argon, 1.75 kg cm-2, fuel H,, 1.2 kg cm-2, sample size lop3 cm3). The gaseous products consisted mainly of H, with all three catalysts, with a measurable amount of propene over MgCr,O,. Products were estimated using Orsat's gas analyser. At higher contact times, traces of carbon dioxide were noticed and measured. Results and Discussion A plot of logl,[lOO/(lOO-x)] us.contact time (fig. l), where x is the percentage isopropanol converted to acetone, results in straight lines passing through the origin, indicating that the decomposition of isopropanol follows first-order kinetics. The rate2668 Catalytic Decomposition of Isopropanol 0.15 0.0 5 0.5 0.1 5 0.2 contact time/h Fig. 1. First-order plot for the formation of acetone at 290 "C. @, NiCr204; A, MnCr,04; 0, MgCr204* Table 3. Rate constants for dehydrogenation rate constant/h-l catalyst T/"C first-order plot initial rate Guggenheim NiCr,04 290 340 380 MnCr,O, 290 340 380 MgCr204 290 340 380 1.49 2.4 1 4.14 0.69 1.26 2.07 0.46 1.12 2.16 1.33 2.87 4.23 0.77 1.50 2.54 0.46 1.11 1.92 1.11 2.70 4.1 0.63 1.14 1.97 0.37 1.25 2.71 constants calculated from the slopes of these plots are presented in table 3, along with the values obtained by the method of initial rate.The first-order rate constants obtained by Guggenheim's20 finite contxt time method are also included in table 3 for comparison. The formation of acetone as a function of contact time over NiCr,O,, MnCr,O, and MgCr,O, at 340 "C is plotted in fig. 2. The rate of formation of acetone decreases with increasing contact time. Increase of contact time may facilitate the adsorbed acetone to react with lattice oxygen to form CO,. This is evident from the liberation of CO, (table 4). Oxidation of acetone to CO, was confirmed by adding acetone and identifying the liberated CO,. The decrease of acetone may also be due to the occurrence of the reverse reaction, as reported by Daniel and Kuriakose.Over magnesium chromites, dehydration also occurs and the extent of dehydrogenation and dehydration are shown in fig. 3. The effect of temperature on the formation of acetone for the above three catalysts is illustrated in fig. 4. The activation energy (E,) obtained from the Arrhenius plots and thermodynamic parameters Am, AS$ and AG$ evaluated for the activated state2, of the system are given in table 5.K. Balasubramanian and V. Krishnasamy 2669 LO 30 4 2 2 20 ---. E a, Y 20 0.1 0.2 0.3 0.4 contact time/h Fig. 2. Effect of contact time on the formation of acetone at 340 "C. 0 , NiCr20,; A, MnCr,O,; 0, MgCr20,. Table 4. CO, (mol% ) liberated during the decomposition of isopropranol over chromite spinels CO, liberated/mol NiCr,O, MnCr,O, MgCr204 T/"C 0.2a 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 290 0.56 1.1 2.1 0.21 0.64 0.9 0.16 0.4 0.7 340 0.69 1.6 2.7 0.32 0.85 1.4 0.28 0.65 0.9 380 0.9 2.25 3.2 0.42 1.2 1.8 0.4 1 .o 1.3 a Contact time/h.The order of activity, based on the values of activation energy for dehydrogenation, is: NiCr,O, > MnCr,O, > MgCr,O,. The negative ASS values indicate the formation of an ordered activated complex by absorption of isopropanol on the spinel, with resultant loss in the internal degrees of freedom. The higher entropy of activation (- 210 J mol-1 K-l) and lower energy of activation (38.37 kJ mol-l) for dehydration of isopropanol over MgCr,O, compared to the respective values of dehydrogenation shows that activated complex for the former process is relatively stable and easily attainable.This is clear evidence for the prevalence of a different active site over MgCr,O, for isopropanol decomposition. In situ electrical conductivity measurements in ambient atmospheres have been carried out on the three catalysts. The initial conductivities at 340 "C for NiCr,O,, MnCr,O,2670 90 80 70 * E .+ 50 4 0 30 20 z 60 . c 0 Catalytic Decomposition of Isopropanol 0 01 0 -2 0.3 0.4 contact time/h Fig. 3. Effect of contact time on the product distribution over MgCr,04 at 340 "C. A, Unchanged u 20 10 0 isopropanol ; 0, acetone ; 0 , propylene. 1 1 I 290 34 0 38 0 TIo C Fig. 4. Effect of temperature on the formation of acetone. 0, NiCr,O,; A, MnCr,O,; 0, MgCr204.K . Balasubramanian and V. Krishnasamy 267 1 % d --. 0.8 0.7 0.6 Table 5.Activation and thermodynamic parameters at 563 K - - 0 I I Ea A S AGI A S catalyst /kJ mol-1 /kJ mol-l /kJ mol-1 /J mol-l K-I dehydrogenation NiCr,O, 34.46 29.78 130.56 - 175.16 MnCr,O, 37.44 32.76 141.95 - 193.95 MgCr20'4 5 1.06 46.38 149.83 - 183.75 dehydration MgCr!204 38.37 33.69 158.22 -210.2 Fig. 5. Relationship between entropy of activation and activation energy for electrical conduction. 0, CoCr,O,; H, CuCr,O,; 0, ZnCr,O,; A, MnCr,O,; 0, NiCr,O,; 0, MgCr,O,. and MgCr,O, were 6.13 x 2.49 x lop4 and 1.36 x lop3 R-l cm-l, respectively. When isopropanol vapour was introduced the conductivity decreased slowly and attained a constant value of 4.41 x and 0.79 x lW3 R-l crn-l, respec- tively, after 10 min. The decrease in conductivity is attributed to the transfer of electrons from the oxygen of isopropanol to the catalyst during the process of adsorption. This lends further support to the p-type nature of the chromite spinels.The constant value of conductivity may be due to the attainment of adsorption-desorption equilibrium. An exactly similar trend is observed in presence of acetone vapour, except that the conductivity decrease is drastic, while there is no change in conductivity in the atmosphere of hydrogen. The existence of a linear correlation between activation energy for electrical conduction (E,) and entropy of activation (AS$) for Mn, Co, Ni, Cu and Zn chromites is shown 1.83 x2672 Catalytic Decomposition of Isopropanol in fig. 5 (the values for Co, Cu and Zn are taken from our earlier which indicates the exclusive dehydrogenating nature of these spinels, while the dual-function MgCr,O, catalyst does not fit into the linear plot (fig.5). The authors thank the Director of A. C . College of Technology, Dr B. Jaganadhasamy, and the Director of M.I.T., Dr S. Sathikh, for the facilities provided by them to carry out this investigation. References 1 E. W. Corter, Philips Res. Rep., 1954, 9, 295. 2 G. Blasse, Philips Res. Rep., Suppl., 1964, 3, 1. 3 A. B. Van Groeou, P. F. Bongers and A. L. Stuyts, Mater. Sci. Eng., 1968, 3, 317. 4 Reinercker, Gunther, Chem. Tech. (Berlin), 1959, 11, 246. 5 G. K. Boreskov, V. V. Proporskii and V. A. Sezonov, Proc. 4th Int. Congr. Catal., Moscow, 1968. 6 Gillot, Bernard, Bull. SOC. Chim., 1968, 6, 2382. 7 Petro Tex Chemical Corp., Neth. (Pat), Chem. Abs., 1966, 64, 15739d. 8 F. G. Dwyer, Catal. Rev., 1972, 6, 261. 9 P. Schoubye, J. Catal., 1969, 14, 238. 10 R. J. Rennard and W. L. Kehl, J. Catal., 1971, 21, 282. 11 T. M. Yur'eva, G. K. Boreskov, V. I. Zharkov, L. G. Karakchiev, V. V. Poporskii and V. A. Chigrina, 12 V. Krishnasamy, Indian J. Chem., 1979, 17A, 437. 13 R. M. Rose, L. A. Shepard and J. Wulf, The Structure and Properties of Materials (Wiley Eastern Ltd, 14 J. Smit and H. P. J. Wijn, Ferrites (Wiley, New York, 1959), 157. 15 E. J. W. Verwey, P. W. Heayman and F. C. Romejin, J. Chem. Phys., 1947,15, 181. 16 J. B. Goodenough, Magnetism and the Chemical Bond (Interscience, New York, 1963). 17 L. Neel, Ann. Phys., 1948, 3, 137. 18 P. W. Anderson, Phys. Rev., 1959, 115, 2. 19 P. I. Slick, Ferromagnetic Materials, ed. E. P. Wohlfarth (North Holland, Amsterdam, 1980), vol. 2, 20 E. A. Guggenheim, Philos. Mag., 1926, 2, 538. 21 C. Daniel and J. C. Kuriakose, Indian J. Chem., 1969, 6, 648. 22 K. J. Laidler, Chemical Kinetics (Tata McGraw-Hill, Bombay, 1965), p. 89. 23 K. Balasubramanian and V. Krishnasamy, Indian J. Chem., 1982, 21A, 813. Kinet. Catal., 1968, 9, 1063. New Delhi, 1971), vol. IV, p. 169. chap. 3. Paper 511788; Received 15th October, 1985