METHOD FOR THE HYDROTHERMAL PREPARATION OF MOLYBDENUM-BISMUTH-COBALT-IRON-BASED MIXED OXIDE CATALYSTS

Abstract

The present invention relates to a process for preparing molybdenum-bismuth-iron-cobalt-based multielement oxide catalysts by means of hydrothermal synthesis, wherein the hydrothermal synthesis is conducted with an aqueous solution and/or an aqueous suspension of precursor compounds of the elements present in the multielement oxide catalyst to be prepared, the pH of which has been adjusted to a value between about 6 and about 8. The present invention also further relates to the multielement oxide catalysts obtainable by this process and to the use thereof in the partial gas phase oxidation of olefins and tert-butanol.

Claims

1. A process for preparing a multielement oxide catalyst, the process comprising: providing a mixture of an aqueous solution and/or an aqueous suspension of precursor compounds of elements present in the multielement oxide catalyst in an amount to achieve stoichiometry thereof, setting a pH of the mixture obtained from the providing to a value between 5.5 and 8.5, reacting the mixture comprising the precursor compounds obtained from the setting under solvothermal reaction conditions in an autoclave at a temperature of from 100 C. to 600 C. to form the multielement oxide catalyst, and separating the multielement oxide catalyst from the aqueous solution and/or suspension, wherein the multielement oxide catalyst is of formula (I):
Mo.sub.aBi.sub.bCo.sub.cFe.sub.dNi.sub.gX.sub.fX.sub.gX.sub.hX.sub.iX.sub.jO.sub.x (I), wherein X is W or P, X is Li, K, Na, Rb, Cs, Mg, Ca, Ba or Sr, X is Ce, Mn, Cr or V, X is Nb, Se, Te, Sm, Gd, La, Y, Pd, Pt, Ru, Ag or Au, X is Si, Al, Ti or Zr, a is 12, b is 1 to 4, c is 4 to 10, d is 1 to 4, e is 0 to 4, f is 0 to 5, g is 0 to 2, h is 0 to 5, i is 0 to 2, j is 0 to 800, and x is a number which is determined by a valency and frequency of the elements other than oxygen.

2. The process according to claim 1, wherein the precursor compounds in step a) are salts.

3. The process according to claim 1, wherein the setting is setting a pH of of the mixture to a value between 6.5 and 7.5.

4. The process according to claim 1, wherein the reacting is reacting at a temperature of from 100 C. to 400 C.

5. The process according to claim 1, wherein the reacting is conducted for a period of from 5.5 to 48.5 hours.

6. The process according to claim 1, further comprising: washing the multielement oxide catalyst obtained from the separating at least once, and drying and/or calcining the multielement oxide catalyst.

7. The process according to claim 1, further comprising: applying the multielement oxide catalyst obtained from the separating, the washing, or the drying and/or calcining to a support to obtain a supported catalyst, or forming the multielement oxide catalyst obtained from the separating, the washing, or the drying and/or calcining to obtain an unsupported catalyst, and additionally drying and/or calcining the multielement oxide catalyst obtained from the applying or the forming.

8. A multielement oxide catalyst of formula:
Mo.sub.aBi.sub.bCo.sub.cFe.sub.dNi.sub.eX.sub.fX.sub.gX.sub.hX.sub.iX.sub.iO.sub.x, wherein X is W or P, X is Li, K, Na, Rb, Cs, Mg, Ca, Ba or Sr, X is Ce, Mn, Cr or V, X is Nb, Se, Te, Sm, Gd, La, Y, Pd, Pt, Ru, Ag or Au, X is Si, Al, Ti or Zr, and a is 12, b is 1 to 4, c is 4 to 10, d is 1 to 4, e is 0 to 4, f is 0 to 5, g is 0 to 2, h is 0 to 5, i is 0 to 2, j is 0 to 800, and x is a number which is determined by a valency and frequency of the elements other than oxygen, and wherein the multielement oxide catalyst comprises areas, in which molybdenum, bismuth and iron are simultaneously present, and the areas have a diameter of from 10 nm to 25 m.

9. A multielement oxide catalyst obtained by the process according to claim 1, wherein a pH of between 6.5 and 7.5is set in the setting.

10. The multielement oxide catalyst according to claim 8, wherein e to i are each 0.

11. The multielement oxide catalyst according to claim 8, wherein a is 12, b is 1 to 2, c is 5 to 8 and d is 2 to 3.

12. The multielement oxide catalyst according to claim 8, wherein the multielement oxide catalyst does not contain a MoO.sub.3-phase.

13. The multielement oxide catalyst according to claim 8, wherein the multielement oxide catalyst has an alpha-bismuth molybdate phase and a beta-cobalt molybdate phase.

14. A process, comprising: performing a partial gas phase oxidation and/or an ammoxidation of olefins or tert-butanol in presence of the multielement oxide catalyst according to claim 8.

15. The process according to claim 14, wherein the olefin is propene and/or isobutene.

Description

DESCRIPTION OF THE FIGURES

[0184] FIG. 1: Correlation between the pH and the composition of the catalysts of experiments 1 to 3 and the comparative examples C1 to C3.

[0185] FIG. 2: Correlation between the pH and the composition of the catalysts of experiments 4 to 6.

[0186] FIG. 3: BET surface area of the Mo.sub.12Bi.sub.1Co.sub.8Fe.sub.3O.sub.x catalyst prepared hydrothermally as a function of the pH in the hydrothermal synthesis and the drying temperature prior to the BET measurement (.circle-solid.=drying at 320; .square-solid.=drying at 150 C., wherein the temperatures are the drying temperatures prior to the BET measurements).

[0187] FIG. 4: SEM-EDX image of the multielement catalyst from experiment 1 (the elements bismuth Bi, molybdenum Mo and iron Fe are coloured as stated in the image).

[0188] FIG. 5: SEM-EDX image of the multielement catalyst from experiment 2 (the elements bismuth Bi, molybdenum Mo and iron Fe are coloured as stated in the image).

[0189] FIG. 6: SEM-EDX image of the multielement catalyst from experiment 3 (the elements bismuth Bi, molybdenum Mo and iron Fe are coloured as stated in the image).

[0190] FIG. 7: Phase transformations of the catalysts prepared by hydrothermal synthesis at pH=6, 7 or 8 of experiments 1 to 3 (indicated as pH=6, pH=7 and pH=8 in FIG. 7) and the catalyst prepared by coprecipitation according to WO 2007/042369 A1 (indicated as CP in FIG. 7) (Raman spectra before the catalysed reaction (dark line) and thereafter (light line)).

[0191] FIG. 8: Phase compositions of the catalyst Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x synthesized hydrothermally at pH values of from 6 to 10.

[0192] FIG. 9: Time-dependent phase transformation of the Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.1.8O.sub.x catalyst obtained by coprecipitation.

[0193] FIG. 10: Time-dependent phase transformation of the Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x catalyst synthesized hydrothermally at pH=7.

EXAMPLES

Example 1

Preparation of MoBiCoFe Oxides

[0194] A first solution (referred to hereinafter as solution I) was prepared by first dissolving amounts of bismuth(III) nitrate pentahydrate, cobalt(II) nitrate hexahydrate and iron(III) nitrate nonahydrate as defined by the figures in the table in 20 ml of nitric acid (concentration 2 M) and stirring thoroughly for 15 minutes. In parallel, a second solution (referred to hereinafter as solution II) was prepared by dissolving stoichiometric amounts of ammonium heptamolybdate according to the figures in Table 1 in 20 ml of demineralized water and stirring for 15 minutes. In each case, the molar amounts of the precursor compounds of bismuth, molybdenum, cobalt and iron were chosen such that they added up to 20 mmol in total. Solution I and solution II were combined in an autoclave insert made of Teflon. Thereafter, the pH was adjusted by dropwise addition of a 25% ammonia solution by means of a titrator (Schott Instruments), and the solution obtained was stirred for a further 15 minutes. Thereafter, the autoclave insert together with the solution present therein was transferred into a steel autoclave having a volume of 250 ml (from Bergerhof). Then the autoclave was closed and heated to a temperature of 180 C. for 24 hours. After this period of time, the autoclave together with its contents was allowed to cool down to room temperature over a period of a further 24 hours. Thereafter, the product obtained was filtered through a G4 glass frit (nominal pore size 10 to 16 micrometres). The solid product was washed three times with 10 ml of water and three times with 10 ml of acetone. Finally, the solid product was dried at room temperature for 48 hours and calcined at 320 C. for 5 hours.

[0195] The compositions of the catalysts from experiments 1 to 6 and comparative experiments C1 to C3 were determined in Example 3.

TABLE-US-00001 TABLE 1 Overview of the hydrothermally synthesized multielement oxide catalysts Molar proportions/% Experiment Bi Mo Co Fe pH n (HNO.sub.3)/mmol V(NH.sub.3)/ml Bi(NO.sub.3).sub.3/g (NH.sub.4).sub.6Mo.sub.7O.sub.24/g Co(NO.sub.3).sub.2/g Fe(NO.sub.3).sub.3/g C1 7.2 59.2 24.8 8.8 5 0.4 4.3 0.695 2.092 1.445 0.709 1 7.2 59.2 24.8 8.8 6 0.4 4.5 0.695 2.092 1.445 0.709 2 7.2 59.2 24.8 8.8 7 0.4 5.0 0.695 2.092 1.445 0.709 3 7.2 59.2 24.8 8.8 8 0.4 5.9 0.695 2.092 1.445 0.709 C2 7.2 59.2 24.8 8.8 9 0.4 9.4 0.695 2.092 1.445 0.709 C3 7.2 59.2 24.8 8.8 10 0.4 10.5 0.695 2.092 1.445 0.709 4 4.2 50.0 33.3 12.5 6 0.4 4.4 0.404 1.766 1.940 1.010 5 4.2 50.0 33.3 12.5 7 0.4 4.8 0.404 1.766 1.940 1.010 6 4.2 50.0 33.3 12.5 8 0.4 5.9 0.404 1.766 1.940 1.010

Example 2

Determination of Surface Area of the Catalysts According to the Invention

[0196] By means of the BET analysis, the specific surface area of the Mo.sub.12Bi.sub.1Co.sub.8Fe.sub.3O.sub.x catalyst prepared by hydrothermal means at different pH values was determined. For this purpose, 100 to 500 mg of the catalyst were dried at 150 C. and at 340 C. under reduced pressure for 5 hours, wherein the temperatures are the drying temperatures prior to the BET measurements. The BET analyses were conducted with a BELSORP-Mini II from Rubotherm GmbH. The purge gas used was helium. After drying, the catalyst sample was evacuated and cooled to a temperature of 77 K with liquid nitrogen. The partial pressure of the nitrogen adsorption gas was increased stepwise, and a standard adsorption isotherm in the range of p/p.sub.0=0 to 1 was recorded. Finally, the desorption isotherm was measured. The BET surface area was determined with the aid of BET theory in the range of p/p.sub.0=0.05 to 0.3.

[0197] The BET surface areas determined were correlated both with the pH in the synthesis of the particular catalyst and with the drying temperatures. FIG. 3 shows these correlations.

[0198] The specific surface area of the catalysts rises with the pH established in the synthesis. Drying conducted at higher temperatures likewise leads to a greater specific surface area.

Example 3

Elemental Analysis of the Catalysts Prepared

[0199] The composition of the catalysts of experiments 1 to 6 and comparative examples C1 to C3 was conducted by means of optical emission spectroscopy with inductively coupled plasma (ICP-OES, Agilent 720/725-ES). The plasma was generated by a 40 MHz high-frequency generator, and argon was used as plasma gas. The digestion of about 40 mg of sample for the optical emission spectroscopy was effected by suspending the sample in a mixture of 6 ml of concentrated hydrochloric acid, 2 ml of concentrated nitric acid and 1 ml of hydrogen peroxide, and subsequent treatment of the mixture obtained in a microwave at 600 watts for 45 minutes.

[0200] The metal fractions per mole determined for the respective catalysts were correlated with pH values established for the hydrothermal synthesis. These correlations are shown in FIG. 1 for the catalysts of experiments 1 to 3 and comparative experiments C1 to C3 and in FIG. 2 for the catalysts of experiments 4 to 6.

[0201] The multielement oxide catalyst prepared at pH=6 from experiment 1 had a composition of roughly Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.1.8O.sub.x. Proceeding from pH=5, the molar proportion of the catalytic molybdenum component decreases continuously with rising pH from about 62 mol % down to about 37 mol % at pH=9. At the same time, with rising pH, the molar proportion of cobalt increases from 15 mol % at pH=5 up to 35 mol % at pH=9. The proportion of iron and bismuth increases only by a few mol % with rising pH.

[0202] The multielement oxide catalyst prepared at pH=7 from example 5 had a composition of roughly Mo.sub.12Bi.sub.1Co.sub.8Fe.sub.3O.sub.x. Proceeding from pH=6, the molar proportion of the catalytic molybdenum component decreases continuously with rising pH from 55 mol % down to about 40 mol % at pH=8. At the same time, with rising pH, the molar proportion of cobalt increases from 25 mol % up to about 40 mol % at pH=8. The proportions of iron and bismuth remain essentially constant irrespective of the pH at about 5 mol % and about 15 mol % respectively.

[0203] The correlation between the pH values in the hydrothermal synthesis and the particular composition of the catalyst shows that the pH has a great influence on the composition of the catalyst. Particularly the molar proportion of the catalytically active molybdenum component is greatly affected by the pH.

Example 4

SEM-EDX Analyses

[0204] The surface of the multielement oxide catalysts from experiments 1 to 3 was examined in a scanning electron microscope with energy-dispersive x-ray spectroscopy (SEM-EDX).

[0205] The images of the surfaces of the multielement oxide catalysts of experiments 1 to 3 which have been made by means of SEM-EDX are shown in FIGS. 4 to 6. In these images, the elements molybdenum Mo, bismuth Bi and iron Fe have been coloured. Areas of the catalysts in which all three elements are present togethercorresponding to the overlapping of the colours for the individual elementsare shown in white in the figures. Only the multielement oxide catalyst from experiment 2, prepared at pH=7, has white areas, i.e. areas in which all three elements molybdenum, bismuth and iron are present together (see FIG. 5). The diameter of these areas was determined to be in the range of from 10 nm to 25 m.

Example 5

Phase Determinations

[0206] The phases of the catalysts Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x, synthesized hydrothermally at a pH of 6, 7 or 8 and synthesized by coprecipitation (FIG. 7), the catalysts Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x of experiments 1 to 3 and comparative experiments C2 to C3, synthesized hydrothermally at a pH of from 6 to 10 (FIG. 8), the catalyst Mo.sub.12Bi.sub.1 5Co.sub.5Fe.sub.1.8O.sub.x, synthesized by coprecipitation (FIG. 9), and the catalyst Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x, synthesized hydrothermally at pH=7 (FIG. 10), were determined by means of Raman spectroscopy. For this purpose, a Renishaw inVia reflex spectrometer system from Renishaw equipped with a research-grade Leica DM2500 microscope was used. The samples were analysed with a 17 mW helium-neon laser at 623 nm. For this purpose, typically the spectral range of 50 to 1300 cm.sup.1 was recorded.

[0207] For the recording of FIG. 8 (catalysts Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x of experiments 1 to 3 and comparative experiments C2 to C3, synthesized hydrothermally at a pH of from 6 to 10), the laser beam was focused onto the sample at room temperature with a Leica 50 objective lens. In each case, more than 2000 spectra were recorded over areas of about 300*300 m and averaged over this area. The Raman spectra obtained were assigned with the aid of literature values to the corresponding phases of Bi, Mo, Co and Fe.

[0208] For the spectra shown in FIG. 7 (catalysts Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x of experiments 1 to 3, synthesized hydrothermally at a pH of 6, 7 or 8), FIG. 9 (catalyst Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.1.8O.sub.x, synthesized by coprecipitation), and FIG. 10, (catalyst Mo.sub.12Bi.sub.15Co.sub.5Fe.sub.3O.sub.x, synthesized hydrothermally at pH=7), 30 mg of catalyst having a particle size of 150 to 300 m were introduced in each case into a quartz glass capillary of diameter 2 mm. A gas stream of 10 ml per minute having the composition of N.sub.2/O.sub.2/C.sub.3H.sub.6/H.sub.2 in a ratio of 72/18/8/8 was passed through this quartz glass capillary. The quartz glass capillary was heated with an Oxford gas blower to a temperature of up to 445 C. The laser beam was focused onto this capillary with the aid of a video fibre-optic from Renishaw GmbH connected to the Raman spectrometer. The spectra in FIG. 7 were each recorded at 80 C. before (dark line) and after the catalysed reaction (light line). The spectra in FIGS. 9 and 10 were recorded during the catalysed reaction, with the following temperature profile defined by means of the gas blower: 180 C. (0 min).fwdarw.4 K min.sup.1.fwdarw.320 C. (20 min).fwdarw.4 K min.sup.1.fwdarw.345 C. (46 min).fwdarw.4 K min.sup.1.fwdarw.370 C. (46 min).fwdarw.4 K min.sup.1.fwdarw.395 C. (46 min).fwdarw.4 K min.sup.1.fwdarw.420 C. (23 min).fwdarw.4 K min.sup.1.fwdarw.445 C. (23 min).fwdarw.10 K min.sup.1.fwdarw.180 C. (18 min).fwdarw.10 K min.sup.1.fwdarw.60 C.

Example 6

Testing of the Catalysts According to the Invention

[0209] The catalysts according to the invention of experiments 1 to 3 (HS-Bi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH6, HS-Bi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH7, and HS-Bi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH8), the catalyst Mo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x prepared by means of coprecipitation according to WO 2007/042369 A and the binary catalyst prepared Bi.sub.1Mo.sub.1O.sub.x by means of hydrothermal synthesis (HS-Bi.sub.1Mo.sub.1O.sub.x-pH6) were tested in the partial gas phase oxidation of propene. For each test, a reactor having an internal diameter of 6 mm was charged with 800 mg in each case of the catalyst to be tested of a sieve fraction from 300 to 450 m, such that its interior was filled essentially completely by the catalyst. The resulting fixed catalyst bed had a height of 2.7 cm. There was one thermocouple each at the upper and lower ends of the fixed catalyst bed, which was in contact with the fixed catalyst bed. Via an external heat supply, the temperature of the fixed catalyst bed was adjusted to a value of about 380 C. Each reactor charged with a specific catalyst was supplied in three experiments with a reactant gas mixture of the composition N.sub.2/O.sub.2/C.sub.3H.sub.6/H.sub.2 in a ratio of 70/14/8/8 with a flow rate of 100, 150 and 200 ml (STP) per minute; based on the mass of the respective catalyst used, the modified residence time (as a quotient of mass of catalyst used divided by the flow rate) for the different flow rates was thus 0.48 g s ml.sup.1, 0.32 g s ml.sup.1 and 0.24 g s ml.sup.1 respectively. During the test, the temperature of the fixed catalyst bed was kept essentially constant at 380 C. by, in the event of temperature fluctuations in the fixed catalyst bed, correspondingly adjusting the temperature of the external heat source. The reaction mixture leaving the reactor was passed via a heated conduit to an Agilent 7890B gas chromatograph.

[0210] The gas chromatograph was equipped with two sample loops for permanent gases, especially nitrogen, oxygen, carbon monoxide and carbon dioxide, and for hydrocarbons, especially propene, acrolein and acrylic acid. The first sample loop consisted of two series-connected Agilent Hayesep Q separation columns and a final micro-packed column of the Agilent Molsieve 5A type. The first Hayesep Q separation column served to remove permanent gases from the hydrocarbons; the second Hayesep Q separation column served to separate the permanent gases: carbon dioxide was the first gas eluted from the second Hayesep Q separation column, and then the other permanent gases were eluted together from the second Hayesep Q separation column and passed through a valve to a final separation column of the Molsieve 5A type which served for separation of the permanent gases N.sub.2, O.sub.2, CO, CO.sub.2 and the permanent gases eluted from the Molsieve 5A separation column were detected with a thermal conductivity detector. The second sample loop consisted of an Agilent HP-FFAP capillary column. The compounds eluted from this separation column, propene, acrolein and acrylic acid, were detected with a flame ionization detector. The gas chromatograph was calibrated with gas mixtures of known concentration of N.sub.2, CO, CO.sub.2, propene and propane and a gas mixture of N.sub.2 and O.sub.2, and with standard solutions of propene, acrolein and acrylic acid in methanol. This calibration was followed by the separation and determination of the individual components of the reaction mixtures in the individual catalyst tests. The amounts of propene used and the amounts of propene, acrolein, acrylic acid and carbon oxides CO and CO.sub.2 detected were used to ascertain the propene conversion and the selectivities for the formation of acrolein, acrylic acid and carbon oxides CO and CO.sub.2. These values are compiled in Tables 2 to 4 below, wherein HS indicates obtained by hydrothermal synthesis and CP indicates obtained by coprecipitation.

TABLE-US-00002 TABLE 2 Overview of the catalytic properties of the catalysts tested at w/F = 0.48 g s ml.sup.1. T(cat.) T(cat.) Propylene Acrolein CO + CO.sub.2 Acrylic acid Oven at at Catalyst conversion/% selectivity/% selectivity/% selectivity/% temperature/ C. inlet/ C. outlet/ C. T(cat)/K CPMo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x 47.0 63.5 27.7 8.3 373 395 366 29 HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 6 29.0 50.5 44.6 4.4 348 363 393 30 (experiment 1) HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 7 63.4 75.9 20.9 2.8 354 376 384 8 (experiment 2) HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 8 52.6 59.7 25.7 14.2 370 378 383 5 (experiment 3) HSBi.sub.1Mo.sub.1O.sub.x-pH 6 18.1 79.7 19.7 0.0 380 377 383 6

TABLE-US-00003 TABLE 3 Overview of the catalytic properties of the catalysts tested at w/F = 0.32 g s ml.sup.1. T(cat.) T(cat.) Propylene Acrolein CO + CO.sub.2 Acrylic acid Oven at at Catalyst conversion/% selectivity/% selectivity/% selectivity/% temperature/ C. inlet/ C. outlet/ C. T(cat)/K CPMo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x 38.0 65.7 20.2 13.8 371 390 370 20 HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 6 19.6 63.3 36.3 0.0 350 361 397 36 (experiment 1) HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 7 56.4 77.2 18.0 4.6 353 378 383 5 (experiment 2) HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 8 41.3 69.1 21.5 9.2 369 373 387 14 (experiment 3) HSBi.sub.1Mo.sub.1O.sub.x-pH 6 12.5 94.0 5.6 0.0 381 377 384 7

TABLE-US-00004 TABLE 4 Overview of the catalytic properties of the catalysts tested at w/F = 0.24 g s ml.sup.1. T(cat.) T(cat.) Propylene Acrolein CO + CO.sub.2 Acrylic acid Oven at at Catalyst conversion/% selectivity/% selectivity/% selectivity/% temperature/ C. inlet/ C. outlet/ C. T(cat)/K CPMo.sub.12Bi.sub.1.5Co.sub.5Fe.sub.3O.sub.x 32.0 70.5 19.0 10.1 370 388 368 20 HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 6 16.2 62.8 33.0 3.4 353 360 400 40 (experiment 1) HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 7 47.5 79.3 14.9 5.6 350 375 385 10 (experiment 2) HSBi.sub.1.5Mo.sub.12Co.sub.5Fe.sub.3O.sub.x-pH 8 33.8 72.5 19.0 8.3 369 370 390 20 (experiment 3) HSBi.sub.1Mo.sub.1O.sub.x-pH 6 10.9 96.0 2.7 0.0 381 377 384 7