- Email: [email protected]
Effect of preparation conditions on the catalytic performance of copper manganese oxide catalysts for CO oxidation
APPLIED CATALYSIS A: GENERAL
ELSEVIER
Applied Catalysis
A: General
166 (1998) 143-152
Effect of preparation conditions on the catalytic performance of copper manganese oxide catalysts for CO oxidation G.J. Hutchings”,
A.A. Mirzaei”, R. W. Joynerb, M.R.H. Siddiqui”, S.H. Taylor”,*
aLeverhulme Centre for Innovative Catalysis, Deparhnent of Chemistry, Universiry of Liverpool, PO Box 147, Liverpool L69 SBX, UK b Catalysis Research Centre, Department
of Chemistry, Nottingham-Trent
University, Clifton Lane, Nottingham NGI I 8NS, UK
Received 5 May 1997; received in revised form 29 July 1997; accepted
29 July 1997
Abstract Copper manganese oxides are prepared using a coprecipitation procedure and studied for the oxidation of CO at ambient temperature. In particular, the effect of a range of preparation variables are investigated in detail. The variables investigated include the precipitate ageing time, pH and temperature of precipitation, the [Cu]/[Mn] ratio of the precipitation solution and the catalyst calcination temperature. The optimum preparation conditions are identified with respect to the catalyst activity for the oxidation of CO at ambient temperature. The results are interpreted in terms of the structure of the active catalyst. Generally it has been concluded that catalysts containing copper/manganese mixed phases are found to be the most active. c 1998 Elsevier Science B.V. Keywords:
Copper manganese
oxide catalyst; Preparation
condition;
1. Introduction The mixed
copper
manganese
oxide in the form of
CuMn204, is used as a catalyst
for the oxidation of CO at ambient temperature and is important in respiratory protection, particularly in the mining industry. Low temperature oxidation of CO has received renewed attention since Haruta et al. [l] demonstrated that supported Au catalysts could be active at sub-ambient temperatures. It is also recognised that low temperature CO oxidation catalysts are potentially important for new applications, such as CO2 laser technology. Hopcalite, however, is still the catalyst of choice for respiratory protection and stuhopcalite,
*Corresponding
author
0926-860x/98/$19.00 8 1998 Elsevier Science B.V. All rights reserved. PII SO926-860X(97)00248-2
CO oxidation
dies have been carried out on the mechanism of deactivation [2], the effect of surface enrichment of Cu and Mn [3], in addition to being used as a model system for the design of oxidation catalysts [4]. Mixed oxide catalyst precursors containing copper are typically prepared using a coprecipitation procedure in which suitable metal salts, typically the nitrates, are premixed and then a precursor is precipitated using sodium carbonate. This process has been well studied in the case of the CuO/ZnO/A120s precursor for the methanol synthesis catalyst [5]. In this case a complex mixture of hydroxy carbonates is formed as the precipitate which on subsequent calcination forms the mixed oxide precursor [6,7]. Recently there has been renewed interest in this precipitation process [8], in particular with respect to the
144
G.J. Hutchings et d/Applied
Catalysis A: General 166 (1998) 143-152
ageing time of the precipitate in the precipitation medium [7]. Although the CuO/ZnO/A1203 catalyst has been relatively well studied, there have been few detailed studies of other copper containing catalysts prepared by coprecipitation. Recently, we have shown [9] that the activity of copper manganese oxide catalysts for the oxidation of CO at 20°C was dramatically affected by variation in the precipitate ageing time. It was shown that this parameter is of crucial importance in controlling the catalytic performance and that catalysts which are aged for < 30min or > 300min gave the best performance. In this paper, we now extend these earlier studies to investigate the effect of a broad range of preparation parameters to enable the optimal catalyst preparation procedure to be identified.
2. Experimental 2.1. Catalyst preparation
and characterisation
All the catalysts tested in this study were prepared using a coprecipitation procedure. Aqueous solutions of Cu(N0&.3H20 (0.25 mol/l) and Mn(N0&.6H20 (0.25 mol/l) were pre-mixed and the resulting solution heated. Aqueous Na2C03 (0.25 mol/l) was added to the mixed nitrate solution which was continuously stirred at 150 rpm whilst the temperature was maintained isothermally in the range 25-80°C. The final pH achieved was varied between 7.5 and 10. This procedure took approximately 10 min to complete. The resulting precipitate was then left in this medium at the required pH and temperature used for the precipitation. For most preparations an ageing time of 12 h was used, this ageing time having been identified in the previous study as the optimum [9]. The precipitate was then filtered, washed several times with warm distilled water until no further Na+ was observed in the washings. The precipitate was dried at 120°C for 16 h to give a material denoted as the catalyst precursor which was subsequently calcined to give the final catalyst. Catalysts were characterised by powder X-ray diffraction using a Hiltonbrooks modified Philips 105OW diffractometer with a CuK, source. The surface area of the calcined catalysts was determined prior to testing by nitrogen adsorption in a Micromeritics
ASAP 2000 apparatus in accordance with the BET method. Thermal gravimetric analysis (TGA) was carried out on a Perkin Elmer Series 7 thermal analysis system under a flow of dry air. Copper and manganese elemental ratios were determined using atomic absorption spectroscopy. 2.2. Catalyst testing The catalysts were tested for CO oxidation using a fixed-bed laboratory microreactor. Typically CO (5% CO in He, 5 ml/min) and 02 (50 ml/min) were fed to the reactor at controlled rates using mass flow controllers and passed over the catalyst (100 mg) at 20°C; the products were analysed using on-line gas chromatography with a 3 m packed Carbosieve column. These conditions are equivalent to a total gas hourly space velocity of 33 000 hK’ and CO concentration of 0.45 mol%. Under these conditions the adiabatic temperature rise is < 7°C and consequently the reactor temperature could readily be maintained isothermally at 20°C.
3. Results 3.1. Effect
of
extended ageing time
In our previous study [9], we demonstrated the importance of ageing time with respect to catalyst activity. However, only ageing times of 12 h and shorter were studied, whilst in this study longer ageing times are also examined. A series of copper/manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, 80°C pH 8.3 f 0.1, 150 rpm stirring) with a range of ageing times for the precipitate under these conditions. The catalyst was prepared by calcination at 500°C for 17 h. Representative data for catalyst surface areas are shown in Table 1. The catalytic activity for the oxidation of CO was investigated for the range of materials and the activity with respect to time on line, for representative catalysts, is shown in Fig. 1. It is apparent that following a short initial period the catalyst activity is very stable with time on line. The effect of ageing time on catalyst performance is shown in Fig. 2. These results show that ageing the precipitate under these conditions for 12 h gives the
G.J. Hutchings et al/Applied
Es
-2
Catalysis A: General 166 (1998) 143-152
145
60 1
--=F’~~~==-------?=-?
50 -
‘1
20:: 10 . 0 0
CI’ 20
I”’ ” 40 60
.
w
.
100
“1 120
“I 80
-
.
-
-
,,
--
140
160
-
.
??
,, 180
200
220
240
260
Time on Line/min Fig. 1, Catalytic performance for CO oxidation at 20°C versus time on line (0.45% CO, GHSV = 33 000 hK’: Ageing time, 0 30 min: ?? 60 min; 0 150 min; + 180 min; v 300 min; A 12 h; 0 24 h.
Table 1 Effect of precipitate
3.2. Effect of precipitation ageing time on catalyst
pH
surface area
Ageing time
Surface area/m’g-’
30 min 150 min 300 min 12h
23 31 30 26
highest CO conversion. The BET surface areas of the catalyst precursors and final catalyst were all found to be similar and were in the range 23-31 m2gg’.
A series of copper manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, 80°C 12 h ageing time, 150 t-pm stirring) with a range of precipitation pH (7.5-10.0). Bulk copper/manganese ratios and surface area data are shown in Table 2. The catalytic activity for the oxidation of CO was investigated for the materials following calcination (500°C 17 h) are presented in Fig. 3. The bulk Cu/Mn ratio is very dependent on this preparation variable and at pH 7.5 a Cu rich precipitate is formed. This is not surprising as the pH for the onset of precipitation of
85 75 65 s .z 2 9 8 u
55
\I
45 35 25 15 5
AgeingTime Fig. 2. Steady state activity after 200 min on line versus ageing time.
146
G.J. Hutchings et &/Applied
Catalysis A: General 166 (1998) 143-152
Table 2 Effect of precipitation pH and temperature ratio on the catalyst bulk copper/manganese ratio and surface area
PH
Ageing
Ageing temperature/C
Surface area/m’gg’
Cu/Mn ratio a
7.5 8.3 9.0 10.0 9.0 9.0 9.0 9.0
80 SO 80 80 25 50 65 80
4.09 0.49 0.47 0.56 0.49 0.52 0.49 0.49
24 26 34 22 30 28 28 34
a Bulk Cu/Mn ratio determined by atomic absorption
molar
spectroscopy.
Cu2+ and Mn2+ from aqueous solution with CO:- are ca. 7 and 8 respectively, and markedly higher pH is required to achieve effective Mn2+ precipitation. The effect of precipitation pH on catalytic performance is shown in Fig. 3. It is apparent that the optimum pH is in the range 8.3-9.0 and this corresponds with a bulk Cu/Mn ratio of 0.47-0.49. The BET surface areas of the final catalysts are also given in Table 2 and it is apparent that although the sample prepared at pH 9.0 gives the highest activity per unit mass of catalyst, the sample prepared at pH 8.3 has a higher specific activity. 3.3. Effect of precipitation
temperature
Copper manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, pH 9.0 & 0.1,
12 h ageing time, 150 rpm stirring) at different temperatures in the range 25-80°C. Surface area and molar ratios of the prepared catalyst are shown in Table 2. The catalytic activity for the oxidation of CO was investigated for the materials following calcination (5OO”C, 17 h). The results are shown in Fig. 4. The temperature at which precipitation is carried out does not significantly affect the bulk Cu/Mn ratio of the precipitate, however, the activity of the final catalyst increases with increasing precipitation temperature. In this case, 80°C is considered to be a practical maximum operating temperature and in this study higher temperatures have not been investigated. 3.4. Effect of solution [Cu]/[Mn]
Copper manganese oxide catalysts were prepared by coprecipitation (SO’C, pH 9.0 ZL0.1, 150 rpm stirring) with a range of [Cu]/[Mn] solution ratios varying from 100% Cu to 100% Mn and the catalytic activity for the oxidation of CO was investigated for the materials following calcination (500°C 17 h). The results are shown in Fig. 5 and Table 3. Using these precipitation conditions the Cu/Mn ratio of the final catalyst is very similar to the Cu2+/Mn2+ ratio in the starting solution. Catalysts catalytic with Cu/Mn > 1 show very similar activity, however, Mn rich catalysts are much more active and the catalyst with Cu/Mn = 0.47 is the most active.
90
7.5
ratio
8.3
9 Ageing
10
pH
Fig. 3. Steady state activity after 200 min on line versus precipitation
pH.
G.J. Hutchings et al/Applied
50
25
147
Catalysis A: General 166 (1998) 143-152
65 Ageing
80
Temperature/C
Fig. 4. Steady state activity after 200 min on line versus precipitation
temperature.
90 80 70
l/O
411
211
l/l [Cu]/[Mn]
l/2 ratio
Fig. 5. Steady state activity after 200 min on line versus [Cu]/[Mn]
3.5. Effect Table 3 Effect of solution Cu/Mn ratio on the per/manganese ratio and surface area Solution
ICul/[Mnl l/O 4/l 211 1/t 112 t/4 O/l
Cu/Mn ratio a
catalyst
molar
cop-
Surface area/m’g-’ 22 19 22 13 34 21 44
3.92 1.92 0.94 0.47 0.28
ABulk Cu/Mn ratio determined
bulk
by atomic absorption
spectroscopy.
l/4
ratio.
of calcination conditions
A series of copper manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, 80°C precipitation pH 9.0, 12 h ageing time, 150 rpm stirring) following calcination at a range of temperatures 300-800°C for 17 h. Surface areas and molar ratios of the prepared catalysts are shown in Table 4 and the catalytic activity for the oxidation of CO was investigated for the materials. The variation of catalytic activity with time on stream and the effect of calcination temperature on catalytic activity are shown in Fig. 6. The optimum calcination temperature is ca. 500°C at higher temperatures the activity declines to a
148
G.J. Hutchings et al./Applied
Catalysis A: General 166 (1998) 143-152
80
60
’
50
I
300
400
500 Calcination
600
Calcination temperature/C
Cu/Mn ratio a
300 400 500 600 700 800
0.48 0.51 0.49 0.32 0.14 0.08
a Bulk Cu/Mn ratio determined
molar
bulk
cop-
Surface area/m2gg’ 110 66 34 14 12 3
by atomic absorption
spectroscopy.
steady value. Previous studies have also shown that high temperature calcination of hopcalite is deleterious to catalytic performance [ 1 I].
4. Discussion It is clear that the precipitation and calcination conditions have a significant effect on the performance of CuMnO, catalysts for the oxidation of CO at ambient temperature. These studies have shown that the optimum preparation conditions are: precipitation
: [Cu]/[Mn] = l/2,
pH = 9.0,
8O”C, 12 h ageing time; calcination
: 500°C
17 h.
800
TemperatureK
Fig. 6. Steady state activity after 200 min on line versus calcination
Table 4 Effect of calcination temperature on the catalyst per/manganese ratio and surface area
700
temperature.
These conditions resulted in the preparation of a catalyst with a bulk Cu/Mn ratio of 0.47 with a surface area of 34 m2 g-‘, which gives a stable catalytic performance of 5.9x lo-* mol gg’ h-’ CO conversion for the 500 min testing period. Changes in these preparation conditions lead to considerable variation in the final catalytic activity. The catalyst surface areas as determined by BET varied over a wide range, however, the majority of materials had surface areas in the range 20-40 m2g-‘. It is considered that the variation in catalyst performance cannot be solely attributed to surface area effects. 4.1. Ageing time Characterisation studies were carried out using powder X-ray diffraction for both the precursors and calcined catalysts. The unaged precursor was found to be well crystalline copper hydroxy nitrate together with manganese carbonate and hence immediately after the initial precipitation the copper and manganese components are present in separate phases. On ageing the copper hydroxy nitrate redissolves and only poorly crystalline manganese carbonate was identified by X-ray diffraction in all the aged precursors. The line shapes of the manganese carbonate are very broad and so it is not possible to determine if there are any changes in line spacing that are consistent with copper being in solid solution in the manganese carbonate.
149
G.J. Hutchings et al./Applied Catalysis A: General 166 (1998) 143-152
24h
12h
8OO(t
0-m
20
,, 25
30
35
40
45
50
55
80
65
70
20 Fig. 7. Powder X-ray diffraction
patterns for catalyst
The calcined catalysts show more variation, particularly with ageing time (Fig. 7) although all these materials were less crystalline than the Characterisation of the catalysts by precursors. powder X-ray diffraction after use showed that the phases initially present were unchanged. The effect of ageing on the catalyst structure has been discussed previously [9] and as the ageing time is increased the formation of mixed CuMnO, oxides becomes more predominant. In particular, the amount of Mn incorporation into the mixed oxide phase is enhanced. The best performance is obtained with the 12 h aged sample and the diffraction pattern of this material was significantly different. This material was less crystalline and the main phase was CuMn204 together with CuO. This composition was found to be the most active from the studies based on the variation in the Cu/Mn ration of the starting solution.
aged prepared
with varying ageing times.
4.2. Calcination
temperature
The calcination temperature has a marked effect on the catalytic performance and the X-ray diffraction patterns of the final catalysts are shown in Fig. 8. Catalysts calcined at 300 and 400°C showed similar diffraction patterns, the materials were poorly crystalline and comprised the phases, MnCOs and CuO. TGA confirmed the presence of undecomposed carbonate as further weight loss from the calcined catalysts was observed. However, the activity of the two catalysts were significantly different, the catalyst calcined at 300°C produced 59.9% conversion whilst the 400°C calcined material produced 76.6%. Although the Xray diffraction patterns were similar subtle differences were apparent, the diffraction peaks from CuO at 400°C were more narrow and the MnCOs peaks broader and considerably less intense than the 300°C calcined catalyst. The lower activity of the
G.J. Hutchings et al./Applied
Catalysis A: General 166 (1998) 143-152
0 1,;
25
30
35
11liilii"llll'Il~ 40 45
50
55
60
65
" ,,""1' '9 70
28
Fig. 8. Powder X-ray diffraction
patterns for catalysts
300°C catalyst can be attributed to the predominant MnCOs phase. TGA confirmed that the catalyst calcined at 300°C contained more MnC03 than the catalyst calcined at 400°C. The origin of the higher activity after 400°C calcination is unclear, however, it may be due to the formation of a mixed copper manganese oxide phase which has been identified after calcination at 500°C and above. A highly dispersed mixed oxide phase with a relatively small crystallite size or a poorly crystalline phase will not be detected by X-ray diffraction. Calcination at 500°C produced a relatively amorphous catalyst which showed the highest activity and was composed of the stoichiometric hopcalite phase CuMn204. After calcination at 600°C Cui.4Mn,.604 was the predominant phase with Mn203 a minor component. Calcination at higher temperatures produced similar catalysts although the Cu/Mn ratio of the mixed phase decreased as Cui.~MnL.s04 was the major phase. This was consistent with the incorporation of manganese from Mn203 in the mixed oxide
calcined
at different temperatures.
phase. The Tamman temperature of MnzOs is 403°C [IO] and, therefore, such solid state reactions can occur at this temperature and above. Striking differences in the catalyst diffraction patterns calcined above 600°C compared to those calcined below this temperature were apparent. The catalysts calcined above 600°C were considerably more crystalline than those calcined at lower temperature. It has previously been shown that crystalline copper manganese oxide phases are less active than amorphous ones [ 1l] and on the basis of weight normalised data the present study is in agreement with these findings. It is considered that the stoichiometry and the amorphous nature of the catalyst calcined at 500°C attributed to the highest activity. 4.3. Ageing pH The catalyst precipitated at pH 7.5 was the least active from the series, the only phase identified by Xray diffraction was Mn203, however, elemental ana-
G.J. Hutch&s
et al/Applied
Catalysis A: General
lysis of the catalyst indicated that it was rich in copper. The absence of a copper containing phase in the X-ray diffraction pattern indicates that the copper phase was either amorphous or comprised very small crystallites. Precipitation at pH 8.3 and above resulted in considerably more active catalysts and the solution pH did not influence the Cu/Mn bulk ratio of the prepared catalysts as they were close to those of the preparation media. The catalyst prepared at pH 8.3 showed Mn20a, CuMnzOh and CuO phases and was one of the most active synthesised. Increasing the pH to 9.0 resulted in the production of the most active catalyst, and along with the catalysts prepared at pH 10.0, the bulk phases C~i,~Mni.~0~ and Mn203 were identified. The catalyst prepared at pH 9.0 possessed a relatively high surface area and correcting for this the catalyst prepared at pH 8.3 showed the highest specific activity followed by pH 9.0 and 10.0. It was, therefore, concluded that although a more active catalyst was produced by precipitation at pH 9.0 the most active phase was stoichiometric CuMn204, prepared at pH 8.3. Whilst the formation of a copper manganese mixed oxide phase was more active than an unsubstituted Mnz03 phase precipitated at lower PH. 4.4. Ageing temperature Catalysts prepared by varying the temperature of the ageing solution all showed the same phase composition and Cu/Mn bulk ratio. The bulk phases present were Cu,,4Mn1.604 and MnzOa, however, the range of activity was broad and showed a direct relationship between increasing ageing temperature and CO oxidation activity. The diffraction patterns also showed general trends as the ageing temperature was increased, firstly the background diffraction increased indicating that more amorphous material was present after exposure to higher preparation temperatures. Secondly, the relative diffraction intensity from Mn*Os increased with temperature. Most significantly the increase in activity can be correlated with the increase in amorphous material. The surface areas were all of similar magnitude and, therefore, specific activity also increased in the same manner as total CO conversion. In order to investigate the nature of the amorphous phases temperature programmed reduction and transmission electron microscopy stu-
166 (1998)
151
143-152
dies are now in progress. The optimum practical ageing temperature was considered to be 80°C. although considering the trends observed the optimum ageing temperature may well be higher. 4.5. Copper/manganese
molar ratio
The catalyst prepared from copper nitrate solution with no manganese component was composed of NaCuO, this catalyst was active and showed a 45% CO conversion. Co-precipitation in the presence of a manganese component resulted in an increase in CO oxidation activity. The catalyst synthesised from the 4/ 1 Cu/Mn solution consisted mainly CuO, although a copper manganese mixed oxide phase was also present, but diffraction peaks were small and relatively broad making identification of the phase stoichiometry difficult. The activity increased steadily as the Cu/Mn ratio decreased and similar phases were identified for the catalysts with preparation ratios of 2/l and l/l, these phases were CuMn,O,, and CuO. The l/2 ratio catalyst was the most active and this preparation media corresponds to the ratio of the stoichiometric hopcalite phase CuMn,O, which has been identified as the phase with the highest specific activity for CO oxidation from the pH ageing studies. The actual phases identified in this catalyst under the specified preparation conditions were CU~.~M~,,~O~ and MnzOa. Clear differences in the diffraction pattern were evident when compared to the materials prepared from solutions of equivalent and excess molar concentrations of copper. These differences were the presence of Mn203 and the higher background diffraction indicative of amorphous phases. At this stage the exact nature of the amorphous phase is not clear, however, preparation under similar conditions (pH 8.3) has shown the formation of CuMn204 and it may be a highly dispersed phase of this type which imparts the high catalytic activity. The catalyst with a l/4 preparation ratio was also one of the most active, the major identified phase was Mn203 with Cu,.dMn,.604 present as a minor component. The catalysts prepared from a liquor containing only manganese nitrate was the least active and only produced a CO conversion in the region of 19%. The X-ray diffraction data from the series clearly shows that catalysts with bulk phases of copper manganese mixed oxide in conjunction with Mnz03 are
152
G.J. Hutch&s
et al./Applied
Catalysis A: General 166 (1998) 143-152
more active than mixed oxide phases in conjunction with CuO, it is also more beneficial to have excess manganese in the starting media. It is evident, as it is in the series with changing calcination temperature, that the amorphous phases which can be detected, but not identified by X-ray diffraction, are important for high catalytic activity.
5. Conclusions Many factors which can be varied during the catalyst preparative coprecipitation procedure and the subsequent calcination step are important in controlling the activity of copper manganese mixed oxide catalysts for ambient temperature CO oxidation. Preparation conditions for optimum catalytic activity are l/2 [Cu]/[Mn] ratio at pH 8.3 and 80°C for 12 h ageing time, followed by calcination at 500°C for 17 h. Relationships between bulk phases and catalytic activity were complex, although catalysts generally showed X-ray diffraction features which correspond to amorphous mixed copper manganese oxide phases. Catalysts which were calcined above 500°C are more highly crystalline than those calcined below this temperature and are the most active in terms of specific activity, although this was as a consequence of the very low surface areas not the high CO oxidation activity. The studies detailing the effect of ageing time can to some extent be correlated with the bulk phases of the catalysts system. These correlations cannot be made for other preparation factors such
as solution pH and temperature and we are now carrying out a detailed transmission electron microscopy study to further investigate these catalysts. However, from the results presented in this study it is clear that the precipitation and calcination conditions used in the preparation procedure are of crucial importance. In particular, the catalyst ageing time and calcination temperature have been found to be of most significance, and control of these parameters should be incorporated into the design of experimental programmes involving precipitation as the method of catalyst preparation.
References Ill M. Haruta, N. Yamada, T. Kobayashi, S. Iijima, .I. Catal. 115 (1989) 301.
121 S. Veprek, D.L. Cocke, S. Kehl, H.R. Oswald, J. Catal. 100 (1986) 250. r31 B.L. Yang, SF. Chan, W.S. Chang, Y.Z. Chen, J. Catal. 130 (1991) 52. [41 C. Yoon, D.L. Cocke, J. Catal. 113 (1988) 267. [51 J.C.J. Bart, R.P.A. Sneeden, Catal. Today 2 (1987) 1. [61 D. Waller, D. Stirling, ES. Stone, MS. Spencer, Faraday Disc., Chem. Sot. 87 (1989) 107. [71 R.W. Joyner, F. King, M.A. Thomas, G. Roberts. Catal. Today 10 (1995) 417. PI S. Fujita, A.M. Satriyo, G.C. Shen, Takezawa, Catal. Lett. 34 (1995) 85. [91 G.J. Hutchings, A.A. Mirzaei, R.W. Joyner, M.R.H. Siddiqui, S.H. Taylor, Catal. Lett. 42 (1996) 21. 1101 CRC Handbook of Chemistry and Physics, 74th Edition, (D.R. Lide Ed.), 1993-1994. El11 S.B. Kanungo, J. Catal. 58 (1979) 419.
ELSEVIER
Applied Catalysis
A: General
166 (1998) 143-152
Effect of preparation conditions on the catalytic performance of copper manganese oxide catalysts for CO oxidation G.J. Hutchings”,
A.A. Mirzaei”, R. W. Joynerb, M.R.H. Siddiqui”, S.H. Taylor”,*
aLeverhulme Centre for Innovative Catalysis, Deparhnent of Chemistry, Universiry of Liverpool, PO Box 147, Liverpool L69 SBX, UK b Catalysis Research Centre, Department
of Chemistry, Nottingham-Trent
University, Clifton Lane, Nottingham NGI I 8NS, UK
Received 5 May 1997; received in revised form 29 July 1997; accepted
29 July 1997
Abstract Copper manganese oxides are prepared using a coprecipitation procedure and studied for the oxidation of CO at ambient temperature. In particular, the effect of a range of preparation variables are investigated in detail. The variables investigated include the precipitate ageing time, pH and temperature of precipitation, the [Cu]/[Mn] ratio of the precipitation solution and the catalyst calcination temperature. The optimum preparation conditions are identified with respect to the catalyst activity for the oxidation of CO at ambient temperature. The results are interpreted in terms of the structure of the active catalyst. Generally it has been concluded that catalysts containing copper/manganese mixed phases are found to be the most active. c 1998 Elsevier Science B.V. Keywords:
Copper manganese
oxide catalyst; Preparation
condition;
1. Introduction The mixed
copper
manganese
oxide in the form of
CuMn204, is used as a catalyst
for the oxidation of CO at ambient temperature and is important in respiratory protection, particularly in the mining industry. Low temperature oxidation of CO has received renewed attention since Haruta et al. [l] demonstrated that supported Au catalysts could be active at sub-ambient temperatures. It is also recognised that low temperature CO oxidation catalysts are potentially important for new applications, such as CO2 laser technology. Hopcalite, however, is still the catalyst of choice for respiratory protection and stuhopcalite,
*Corresponding
author
0926-860x/98/$19.00 8 1998 Elsevier Science B.V. All rights reserved. PII SO926-860X(97)00248-2
CO oxidation
dies have been carried out on the mechanism of deactivation [2], the effect of surface enrichment of Cu and Mn [3], in addition to being used as a model system for the design of oxidation catalysts [4]. Mixed oxide catalyst precursors containing copper are typically prepared using a coprecipitation procedure in which suitable metal salts, typically the nitrates, are premixed and then a precursor is precipitated using sodium carbonate. This process has been well studied in the case of the CuO/ZnO/A120s precursor for the methanol synthesis catalyst [5]. In this case a complex mixture of hydroxy carbonates is formed as the precipitate which on subsequent calcination forms the mixed oxide precursor [6,7]. Recently there has been renewed interest in this precipitation process [8], in particular with respect to the
144
G.J. Hutchings et d/Applied
Catalysis A: General 166 (1998) 143-152
ageing time of the precipitate in the precipitation medium [7]. Although the CuO/ZnO/A1203 catalyst has been relatively well studied, there have been few detailed studies of other copper containing catalysts prepared by coprecipitation. Recently, we have shown [9] that the activity of copper manganese oxide catalysts for the oxidation of CO at 20°C was dramatically affected by variation in the precipitate ageing time. It was shown that this parameter is of crucial importance in controlling the catalytic performance and that catalysts which are aged for < 30min or > 300min gave the best performance. In this paper, we now extend these earlier studies to investigate the effect of a broad range of preparation parameters to enable the optimal catalyst preparation procedure to be identified.
2. Experimental 2.1. Catalyst preparation
and characterisation
All the catalysts tested in this study were prepared using a coprecipitation procedure. Aqueous solutions of Cu(N0&.3H20 (0.25 mol/l) and Mn(N0&.6H20 (0.25 mol/l) were pre-mixed and the resulting solution heated. Aqueous Na2C03 (0.25 mol/l) was added to the mixed nitrate solution which was continuously stirred at 150 rpm whilst the temperature was maintained isothermally in the range 25-80°C. The final pH achieved was varied between 7.5 and 10. This procedure took approximately 10 min to complete. The resulting precipitate was then left in this medium at the required pH and temperature used for the precipitation. For most preparations an ageing time of 12 h was used, this ageing time having been identified in the previous study as the optimum [9]. The precipitate was then filtered, washed several times with warm distilled water until no further Na+ was observed in the washings. The precipitate was dried at 120°C for 16 h to give a material denoted as the catalyst precursor which was subsequently calcined to give the final catalyst. Catalysts were characterised by powder X-ray diffraction using a Hiltonbrooks modified Philips 105OW diffractometer with a CuK, source. The surface area of the calcined catalysts was determined prior to testing by nitrogen adsorption in a Micromeritics
ASAP 2000 apparatus in accordance with the BET method. Thermal gravimetric analysis (TGA) was carried out on a Perkin Elmer Series 7 thermal analysis system under a flow of dry air. Copper and manganese elemental ratios were determined using atomic absorption spectroscopy. 2.2. Catalyst testing The catalysts were tested for CO oxidation using a fixed-bed laboratory microreactor. Typically CO (5% CO in He, 5 ml/min) and 02 (50 ml/min) were fed to the reactor at controlled rates using mass flow controllers and passed over the catalyst (100 mg) at 20°C; the products were analysed using on-line gas chromatography with a 3 m packed Carbosieve column. These conditions are equivalent to a total gas hourly space velocity of 33 000 hK’ and CO concentration of 0.45 mol%. Under these conditions the adiabatic temperature rise is < 7°C and consequently the reactor temperature could readily be maintained isothermally at 20°C.
3. Results 3.1. Effect
of
extended ageing time
In our previous study [9], we demonstrated the importance of ageing time with respect to catalyst activity. However, only ageing times of 12 h and shorter were studied, whilst in this study longer ageing times are also examined. A series of copper/manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, 80°C pH 8.3 f 0.1, 150 rpm stirring) with a range of ageing times for the precipitate under these conditions. The catalyst was prepared by calcination at 500°C for 17 h. Representative data for catalyst surface areas are shown in Table 1. The catalytic activity for the oxidation of CO was investigated for the range of materials and the activity with respect to time on line, for representative catalysts, is shown in Fig. 1. It is apparent that following a short initial period the catalyst activity is very stable with time on line. The effect of ageing time on catalyst performance is shown in Fig. 2. These results show that ageing the precipitate under these conditions for 12 h gives the
G.J. Hutchings et al/Applied
Es
-2
Catalysis A: General 166 (1998) 143-152
145
60 1
--=F’~~~==-------?=-?
50 -
‘1
20:: 10 . 0 0
CI’ 20
I”’ ” 40 60
.
w
.
100
“1 120
“I 80
-
.
-
-
,,
--
140
160
-
.
??
,, 180
200
220
240
260
Time on Line/min Fig. 1, Catalytic performance for CO oxidation at 20°C versus time on line (0.45% CO, GHSV = 33 000 hK’: Ageing time, 0 30 min: ?? 60 min; 0 150 min; + 180 min; v 300 min; A 12 h; 0 24 h.
Table 1 Effect of precipitate
3.2. Effect of precipitation ageing time on catalyst
pH
surface area
Ageing time
Surface area/m’g-’
30 min 150 min 300 min 12h
23 31 30 26
highest CO conversion. The BET surface areas of the catalyst precursors and final catalyst were all found to be similar and were in the range 23-31 m2gg’.
A series of copper manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, 80°C 12 h ageing time, 150 t-pm stirring) with a range of precipitation pH (7.5-10.0). Bulk copper/manganese ratios and surface area data are shown in Table 2. The catalytic activity for the oxidation of CO was investigated for the materials following calcination (500°C 17 h) are presented in Fig. 3. The bulk Cu/Mn ratio is very dependent on this preparation variable and at pH 7.5 a Cu rich precipitate is formed. This is not surprising as the pH for the onset of precipitation of
85 75 65 s .z 2 9 8 u
55
\I
45 35 25 15 5
AgeingTime Fig. 2. Steady state activity after 200 min on line versus ageing time.
146
G.J. Hutchings et &/Applied
Catalysis A: General 166 (1998) 143-152
Table 2 Effect of precipitation pH and temperature ratio on the catalyst bulk copper/manganese ratio and surface area
PH
Ageing
Ageing temperature/C
Surface area/m’gg’
Cu/Mn ratio a
7.5 8.3 9.0 10.0 9.0 9.0 9.0 9.0
80 SO 80 80 25 50 65 80
4.09 0.49 0.47 0.56 0.49 0.52 0.49 0.49
24 26 34 22 30 28 28 34
a Bulk Cu/Mn ratio determined by atomic absorption
molar
spectroscopy.
Cu2+ and Mn2+ from aqueous solution with CO:- are ca. 7 and 8 respectively, and markedly higher pH is required to achieve effective Mn2+ precipitation. The effect of precipitation pH on catalytic performance is shown in Fig. 3. It is apparent that the optimum pH is in the range 8.3-9.0 and this corresponds with a bulk Cu/Mn ratio of 0.47-0.49. The BET surface areas of the final catalysts are also given in Table 2 and it is apparent that although the sample prepared at pH 9.0 gives the highest activity per unit mass of catalyst, the sample prepared at pH 8.3 has a higher specific activity. 3.3. Effect of precipitation
temperature
Copper manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, pH 9.0 & 0.1,
12 h ageing time, 150 rpm stirring) at different temperatures in the range 25-80°C. Surface area and molar ratios of the prepared catalyst are shown in Table 2. The catalytic activity for the oxidation of CO was investigated for the materials following calcination (5OO”C, 17 h). The results are shown in Fig. 4. The temperature at which precipitation is carried out does not significantly affect the bulk Cu/Mn ratio of the precipitate, however, the activity of the final catalyst increases with increasing precipitation temperature. In this case, 80°C is considered to be a practical maximum operating temperature and in this study higher temperatures have not been investigated. 3.4. Effect of solution [Cu]/[Mn]
Copper manganese oxide catalysts were prepared by coprecipitation (SO’C, pH 9.0 ZL0.1, 150 rpm stirring) with a range of [Cu]/[Mn] solution ratios varying from 100% Cu to 100% Mn and the catalytic activity for the oxidation of CO was investigated for the materials following calcination (500°C 17 h). The results are shown in Fig. 5 and Table 3. Using these precipitation conditions the Cu/Mn ratio of the final catalyst is very similar to the Cu2+/Mn2+ ratio in the starting solution. Catalysts catalytic with Cu/Mn > 1 show very similar activity, however, Mn rich catalysts are much more active and the catalyst with Cu/Mn = 0.47 is the most active.
90
7.5
ratio
8.3
9 Ageing
10
pH
Fig. 3. Steady state activity after 200 min on line versus precipitation
pH.
G.J. Hutchings et al/Applied
50
25
147
Catalysis A: General 166 (1998) 143-152
65 Ageing
80
Temperature/C
Fig. 4. Steady state activity after 200 min on line versus precipitation
temperature.
90 80 70
l/O
411
211
l/l [Cu]/[Mn]
l/2 ratio
Fig. 5. Steady state activity after 200 min on line versus [Cu]/[Mn]
3.5. Effect Table 3 Effect of solution Cu/Mn ratio on the per/manganese ratio and surface area Solution
ICul/[Mnl l/O 4/l 211 1/t 112 t/4 O/l
Cu/Mn ratio a
catalyst
molar
cop-
Surface area/m’g-’ 22 19 22 13 34 21 44
3.92 1.92 0.94 0.47 0.28
ABulk Cu/Mn ratio determined
bulk
by atomic absorption
spectroscopy.
l/4
ratio.
of calcination conditions
A series of copper manganese oxide catalysts were prepared by coprecipitation ([Cu]/[Mn] = l/2, 80°C precipitation pH 9.0, 12 h ageing time, 150 rpm stirring) following calcination at a range of temperatures 300-800°C for 17 h. Surface areas and molar ratios of the prepared catalysts are shown in Table 4 and the catalytic activity for the oxidation of CO was investigated for the materials. The variation of catalytic activity with time on stream and the effect of calcination temperature on catalytic activity are shown in Fig. 6. The optimum calcination temperature is ca. 500°C at higher temperatures the activity declines to a
148
G.J. Hutchings et al./Applied
Catalysis A: General 166 (1998) 143-152
80
60
’
50
I
300
400
500 Calcination
600
Calcination temperature/C
Cu/Mn ratio a
300 400 500 600 700 800
0.48 0.51 0.49 0.32 0.14 0.08
a Bulk Cu/Mn ratio determined
molar
bulk
cop-
Surface area/m2gg’ 110 66 34 14 12 3
by atomic absorption
spectroscopy.
steady value. Previous studies have also shown that high temperature calcination of hopcalite is deleterious to catalytic performance [ 1 I].
4. Discussion It is clear that the precipitation and calcination conditions have a significant effect on the performance of CuMnO, catalysts for the oxidation of CO at ambient temperature. These studies have shown that the optimum preparation conditions are: precipitation
: [Cu]/[Mn] = l/2,
pH = 9.0,
8O”C, 12 h ageing time; calcination
: 500°C
17 h.
800
TemperatureK
Fig. 6. Steady state activity after 200 min on line versus calcination
Table 4 Effect of calcination temperature on the catalyst per/manganese ratio and surface area
700
temperature.
These conditions resulted in the preparation of a catalyst with a bulk Cu/Mn ratio of 0.47 with a surface area of 34 m2 g-‘, which gives a stable catalytic performance of 5.9x lo-* mol gg’ h-’ CO conversion for the 500 min testing period. Changes in these preparation conditions lead to considerable variation in the final catalytic activity. The catalyst surface areas as determined by BET varied over a wide range, however, the majority of materials had surface areas in the range 20-40 m2g-‘. It is considered that the variation in catalyst performance cannot be solely attributed to surface area effects. 4.1. Ageing time Characterisation studies were carried out using powder X-ray diffraction for both the precursors and calcined catalysts. The unaged precursor was found to be well crystalline copper hydroxy nitrate together with manganese carbonate and hence immediately after the initial precipitation the copper and manganese components are present in separate phases. On ageing the copper hydroxy nitrate redissolves and only poorly crystalline manganese carbonate was identified by X-ray diffraction in all the aged precursors. The line shapes of the manganese carbonate are very broad and so it is not possible to determine if there are any changes in line spacing that are consistent with copper being in solid solution in the manganese carbonate.
149
G.J. Hutchings et al./Applied Catalysis A: General 166 (1998) 143-152
24h
12h
8OO(t
0-m
20
,, 25
30
35
40
45
50
55
80
65
70
20 Fig. 7. Powder X-ray diffraction
patterns for catalyst
The calcined catalysts show more variation, particularly with ageing time (Fig. 7) although all these materials were less crystalline than the Characterisation of the catalysts by precursors. powder X-ray diffraction after use showed that the phases initially present were unchanged. The effect of ageing on the catalyst structure has been discussed previously [9] and as the ageing time is increased the formation of mixed CuMnO, oxides becomes more predominant. In particular, the amount of Mn incorporation into the mixed oxide phase is enhanced. The best performance is obtained with the 12 h aged sample and the diffraction pattern of this material was significantly different. This material was less crystalline and the main phase was CuMn204 together with CuO. This composition was found to be the most active from the studies based on the variation in the Cu/Mn ration of the starting solution.
aged prepared
with varying ageing times.
4.2. Calcination
temperature
The calcination temperature has a marked effect on the catalytic performance and the X-ray diffraction patterns of the final catalysts are shown in Fig. 8. Catalysts calcined at 300 and 400°C showed similar diffraction patterns, the materials were poorly crystalline and comprised the phases, MnCOs and CuO. TGA confirmed the presence of undecomposed carbonate as further weight loss from the calcined catalysts was observed. However, the activity of the two catalysts were significantly different, the catalyst calcined at 300°C produced 59.9% conversion whilst the 400°C calcined material produced 76.6%. Although the Xray diffraction patterns were similar subtle differences were apparent, the diffraction peaks from CuO at 400°C were more narrow and the MnCOs peaks broader and considerably less intense than the 300°C calcined catalyst. The lower activity of the
G.J. Hutchings et al./Applied
Catalysis A: General 166 (1998) 143-152
0 1,;
25
30
35
11liilii"llll'Il~ 40 45
50
55
60
65
" ,,""1' '9 70
28
Fig. 8. Powder X-ray diffraction
patterns for catalysts
300°C catalyst can be attributed to the predominant MnCOs phase. TGA confirmed that the catalyst calcined at 300°C contained more MnC03 than the catalyst calcined at 400°C. The origin of the higher activity after 400°C calcination is unclear, however, it may be due to the formation of a mixed copper manganese oxide phase which has been identified after calcination at 500°C and above. A highly dispersed mixed oxide phase with a relatively small crystallite size or a poorly crystalline phase will not be detected by X-ray diffraction. Calcination at 500°C produced a relatively amorphous catalyst which showed the highest activity and was composed of the stoichiometric hopcalite phase CuMn204. After calcination at 600°C Cui.4Mn,.604 was the predominant phase with Mn203 a minor component. Calcination at higher temperatures produced similar catalysts although the Cu/Mn ratio of the mixed phase decreased as Cui.~MnL.s04 was the major phase. This was consistent with the incorporation of manganese from Mn203 in the mixed oxide
calcined
at different temperatures.
phase. The Tamman temperature of MnzOs is 403°C [IO] and, therefore, such solid state reactions can occur at this temperature and above. Striking differences in the catalyst diffraction patterns calcined above 600°C compared to those calcined below this temperature were apparent. The catalysts calcined above 600°C were considerably more crystalline than those calcined at lower temperature. It has previously been shown that crystalline copper manganese oxide phases are less active than amorphous ones [ 1l] and on the basis of weight normalised data the present study is in agreement with these findings. It is considered that the stoichiometry and the amorphous nature of the catalyst calcined at 500°C attributed to the highest activity. 4.3. Ageing pH The catalyst precipitated at pH 7.5 was the least active from the series, the only phase identified by Xray diffraction was Mn203, however, elemental ana-
G.J. Hutch&s
et al/Applied
Catalysis A: General
lysis of the catalyst indicated that it was rich in copper. The absence of a copper containing phase in the X-ray diffraction pattern indicates that the copper phase was either amorphous or comprised very small crystallites. Precipitation at pH 8.3 and above resulted in considerably more active catalysts and the solution pH did not influence the Cu/Mn bulk ratio of the prepared catalysts as they were close to those of the preparation media. The catalyst prepared at pH 8.3 showed Mn20a, CuMnzOh and CuO phases and was one of the most active synthesised. Increasing the pH to 9.0 resulted in the production of the most active catalyst, and along with the catalysts prepared at pH 10.0, the bulk phases C~i,~Mni.~0~ and Mn203 were identified. The catalyst prepared at pH 9.0 possessed a relatively high surface area and correcting for this the catalyst prepared at pH 8.3 showed the highest specific activity followed by pH 9.0 and 10.0. It was, therefore, concluded that although a more active catalyst was produced by precipitation at pH 9.0 the most active phase was stoichiometric CuMn204, prepared at pH 8.3. Whilst the formation of a copper manganese mixed oxide phase was more active than an unsubstituted Mnz03 phase precipitated at lower PH. 4.4. Ageing temperature Catalysts prepared by varying the temperature of the ageing solution all showed the same phase composition and Cu/Mn bulk ratio. The bulk phases present were Cu,,4Mn1.604 and MnzOa, however, the range of activity was broad and showed a direct relationship between increasing ageing temperature and CO oxidation activity. The diffraction patterns also showed general trends as the ageing temperature was increased, firstly the background diffraction increased indicating that more amorphous material was present after exposure to higher preparation temperatures. Secondly, the relative diffraction intensity from Mn*Os increased with temperature. Most significantly the increase in activity can be correlated with the increase in amorphous material. The surface areas were all of similar magnitude and, therefore, specific activity also increased in the same manner as total CO conversion. In order to investigate the nature of the amorphous phases temperature programmed reduction and transmission electron microscopy stu-
166 (1998)
151
143-152
dies are now in progress. The optimum practical ageing temperature was considered to be 80°C. although considering the trends observed the optimum ageing temperature may well be higher. 4.5. Copper/manganese
molar ratio
The catalyst prepared from copper nitrate solution with no manganese component was composed of NaCuO, this catalyst was active and showed a 45% CO conversion. Co-precipitation in the presence of a manganese component resulted in an increase in CO oxidation activity. The catalyst synthesised from the 4/ 1 Cu/Mn solution consisted mainly CuO, although a copper manganese mixed oxide phase was also present, but diffraction peaks were small and relatively broad making identification of the phase stoichiometry difficult. The activity increased steadily as the Cu/Mn ratio decreased and similar phases were identified for the catalysts with preparation ratios of 2/l and l/l, these phases were CuMn,O,, and CuO. The l/2 ratio catalyst was the most active and this preparation media corresponds to the ratio of the stoichiometric hopcalite phase CuMn,O, which has been identified as the phase with the highest specific activity for CO oxidation from the pH ageing studies. The actual phases identified in this catalyst under the specified preparation conditions were CU~.~M~,,~O~ and MnzOa. Clear differences in the diffraction pattern were evident when compared to the materials prepared from solutions of equivalent and excess molar concentrations of copper. These differences were the presence of Mn203 and the higher background diffraction indicative of amorphous phases. At this stage the exact nature of the amorphous phase is not clear, however, preparation under similar conditions (pH 8.3) has shown the formation of CuMn204 and it may be a highly dispersed phase of this type which imparts the high catalytic activity. The catalyst with a l/4 preparation ratio was also one of the most active, the major identified phase was Mn203 with Cu,.dMn,.604 present as a minor component. The catalysts prepared from a liquor containing only manganese nitrate was the least active and only produced a CO conversion in the region of 19%. The X-ray diffraction data from the series clearly shows that catalysts with bulk phases of copper manganese mixed oxide in conjunction with Mnz03 are
152
G.J. Hutch&s
et al./Applied
Catalysis A: General 166 (1998) 143-152
more active than mixed oxide phases in conjunction with CuO, it is also more beneficial to have excess manganese in the starting media. It is evident, as it is in the series with changing calcination temperature, that the amorphous phases which can be detected, but not identified by X-ray diffraction, are important for high catalytic activity.
5. Conclusions Many factors which can be varied during the catalyst preparative coprecipitation procedure and the subsequent calcination step are important in controlling the activity of copper manganese mixed oxide catalysts for ambient temperature CO oxidation. Preparation conditions for optimum catalytic activity are l/2 [Cu]/[Mn] ratio at pH 8.3 and 80°C for 12 h ageing time, followed by calcination at 500°C for 17 h. Relationships between bulk phases and catalytic activity were complex, although catalysts generally showed X-ray diffraction features which correspond to amorphous mixed copper manganese oxide phases. Catalysts which were calcined above 500°C are more highly crystalline than those calcined below this temperature and are the most active in terms of specific activity, although this was as a consequence of the very low surface areas not the high CO oxidation activity. The studies detailing the effect of ageing time can to some extent be correlated with the bulk phases of the catalysts system. These correlations cannot be made for other preparation factors such
as solution pH and temperature and we are now carrying out a detailed transmission electron microscopy study to further investigate these catalysts. However, from the results presented in this study it is clear that the precipitation and calcination conditions used in the preparation procedure are of crucial importance. In particular, the catalyst ageing time and calcination temperature have been found to be of most significance, and control of these parameters should be incorporated into the design of experimental programmes involving precipitation as the method of catalyst preparation.
References Ill M. Haruta, N. Yamada, T. Kobayashi, S. Iijima, .I. Catal. 115 (1989) 301.
121 S. Veprek, D.L. Cocke, S. Kehl, H.R. Oswald, J. Catal. 100 (1986) 250. r31 B.L. Yang, SF. Chan, W.S. Chang, Y.Z. Chen, J. Catal. 130 (1991) 52. [41 C. Yoon, D.L. Cocke, J. Catal. 113 (1988) 267. [51 J.C.J. Bart, R.P.A. Sneeden, Catal. Today 2 (1987) 1. [61 D. Waller, D. Stirling, ES. Stone, MS. Spencer, Faraday Disc., Chem. Sot. 87 (1989) 107. [71 R.W. Joyner, F. King, M.A. Thomas, G. Roberts. Catal. Today 10 (1995) 417. PI S. Fujita, A.M. Satriyo, G.C. Shen, Takezawa, Catal. Lett. 34 (1995) 85. [91 G.J. Hutchings, A.A. Mirzaei, R.W. Joyner, M.R.H. Siddiqui, S.H. Taylor, Catal. Lett. 42 (1996) 21. 1101 CRC Handbook of Chemistry and Physics, 74th Edition, (D.R. Lide Ed.), 1993-1994. El11 S.B. Kanungo, J. Catal. 58 (1979) 419.