Liquid phase oxidation of toluene to benzaldehyde with molecular oxygen over copper-based heterogeneous catalysts

Article abstract of DOI:10.1002/adsc.200505107

We conducted the liquid phase oxidation of toluene with molecular oxygen over heterogeneous catalysts of copper-based binary metal oxides. Among the copper-based binary metal oxides, iron-copper binary oxide (Fe/Cu = 0.3 atomic ratio) was found to be the

Full text of DOI:10.1002/adsc.200505107

FULL PAPERS
DOI: 10.1002/adsc.200505107
Liquid Phase Oxidation of Toluene to Benzaldehyde with
Molecular Oxygen over Copper-Based Heterogeneous Catalysts
Feng Wang, Jie Xu,* Xiaoqiang Li, Jin Gao, Lipeng Zhou, Ryuichiro Ohnishi
State Key Laboratory of Catalysis, Dalian Institute of Chemical Physics, Graduate School of the Chinese Academy of
Sciences, Chinese Academy of Science, Zhongshan Road 457, P. O. Box 110, Dalian, 116023, Peopleꢀs Republic of China
Fax: (þ86)-411-8437-9245, e-mail: xujie@dicp.ac.cn
Received: March 12, 2005; Accepted: September 8, 2005
Abstract: We conducted the liquid phase oxidation of tive adsorption measurements that pyridine could re-
toluene with molecular oxygen over heterogeneous duce the adsorption of benzaldehyde. At a long reac-
catalysts of copper-based binary metal oxides. Among tion time of 4 h, the conversion increased to 25% and
the copper-based binary metaloxides, iron-copper bi-
benzoic acid became the predominant reaction prod-
nary oxide (Fe/Cu¼0.3 atomic ratio) was found to be uct (72% selectivity) in the absence of pyridine. The
the best catalyst. In the presence of pyridine, overox- yield of benzoic acid was higher than that in the
idation of benzaldehyde to benzoic acid was partially Snia-Viscosa process, which requires corrosive halo-
prevented. As a result, highly selective formation of gen ions and acidic solvents in the homogeneous reac-
benzaldehyde (86% selectivity) was observed after tion media. The catalyst was easily recycled by simple
2
h of reaction (7% conversion of toluene) at 463 K filtration and reusable after washing and drying.
and 1.0 MPa of oxygen atmosphere in the presence
of pyridine. These catalytic performances were similar
or better than those in the gas phase oxidation of tol- Keywords: benzaldehyde; CÀH bond activation; het-
uene at reaction temperatures higher than 473 K and erogeneous catalysis; liquid phase; molecular oxygen;
under 0.5–2.5 MPa. It was suggested from competi- toluene
Introduction
for benzaldehyde. Generally, the conversion has to be
kept at less than 4% to attain 70% selectivity of benzal-
3
Selective oxidation of sp hybridized carbon of inactive dehyde and to avoid the formation of carboxylic acids,
hydrocarbons to industrially important intermediates phenols and decomposition to carbon oxides (CO and
2
[
1–5]
[10–12]
still remains a major challenge.
can be converted into oxidation products such as benzyl
For example, toluene CO) and tar.
The liquid phase oxidation of toluene with homoge-
alcohol, benzaldehyde and benzoic acid. Among these neous metal salt catalysts was industrially realized in
products, benzaldehyde is the most desirable and val- the Rhodia, Dow and Snia-Viscosa processes using
[
13–16]
ue-added product. However, benzaldehyde is easily oxygen or peroxides as oxidants.
For example,
overoxidized to benzoic acid. Due to rather poor per- the Snia-Viscosa process operates at 1658C and under
formance of direct oxidation, various substituted aro- 10 atm of air in the presence of a homogeneous cobalt
matic aldehydes are still manufactured in the traditional catalyst in acetic acid. At the optimal conditions, ben-
way of organic chemistry. Traditionally, benzaldehyde is zoic acid as the target product is produced with 90% se-
produced by side-chain chlorination of toluene and sap- lectivity, and benzaldehyde as a minor by-product is
onification of the resulting dichloromethyl group to obtained in 3% selectivity at 15% conversion of tol-
form the aldehyde group. The product, still containing uene. In these processes, however, halogen ions and
the chlorinated impurities, does not meet food and acidic solvents are indispensable, and they easily cause
drug grade specifications. Besides, these processes pro- erosion of the facility. The formation of benzoic acid via
duce much waste causing environmentalpro be lm s
and, additionally, the efficiencies of these routes are CO and ionic liquids as solvents.
toluene oxidation was also reported in supercritical
[
17,18]
In the latter re-
2
not so high. There are some reports on vapor phase ox- port, the maximum conversion of toluene was only
idation of toluene with oxygen. However, the common 4.7% after 48 h. Toluene oxidation by the use of cobalt
reaction conditions (T>473 K and under 0.5–2.5 MPa tetraphenylporphyrin as catalyst has also reported re-
[6–9]
[19]
pressure
seem too harsh for improving the selectivity cently.
Adv. Synth. Catal. 2005, 347, 1987 – 1992
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1987

FULL PAPERS
Feng Wang et al.
sions in the oxidation of toluene. Only the copper-iron
binary metal oxide catalyst (Table 2, entry 6) gave a
promising activity of 7.4% and high selectivity (96.5%)
of oxidized products at the benzylic position. At the
long reaction time of 4 h, the conversion increased to
Scheme 1. Liquid phase oxidation of toluene with molecular 25.4% and benzoic acid became the predominant reac-
oxygen over Cu-based heterogeneous catalysts.
tion product (71.6% selectivity). The yield of benzoic
acid (25Â72%¼20%) was higher than that of the
Snia-Viscosa process (90Â15%¼13.5%). Considering
We report here the liquid phase oxidation of toluene the indispensable usage of corrosive halogen ions and
with molecular oxygen on copper-based binary metal acidic solvents in the Snia-Viscosa process, this work
oxide catalysts (see Scheme 1). To the best our knowl- could be judged to be the better system. The main prod-
edge, no report is available on the heterogeneous cata- uct can be shifted from benzaldehyde to benzoic acid by
lytic oxidation of toluene to benzaldehyde with mo- prolonging the reaction time, as was expected.
lecular oxygen in the liquid phase.
The molar ratio of Fe to Cu was optimized to be 0.3:1,
as presented in Figure 1. Yields of benzaldehyde and
benzyla cl oholbecame maximalat Fe/Cu
the increases in benzaldehyde and benzyl alcohol selec-
tivities were compensated with the decrease in the con-
¼0.3 since
Results and Discussion
It can be seen from Table 1 that the blank run without version, as the iron content increased. Magnesium and
catalyst gave no products. The support g-Al O itself titanium(IV) oxide were employed as supports for
2
3
gave minor conversion of 2%. Among the single compo- iron-copper oxide (Fe/Cu¼0.3). Low conversions of be-
nent catalysts tested, only copper supported on g-Al O low 2% were obtained on these catalysts, demonstrating
gave the best conversion of 2.5%, although it is still very that g-Al O is a suitable support for the oxidation of tol-
2
3
2
3
low. GC-MS analyses of the post-reaction mixture uene. Although we do not know how the iron-copper
showed that the main by-products were of the phenolic metaloxide promotes the reaction, transition metasl
series and esters [(o-, p- and m-)cresols, benzoic acid salts are known to pull hydrogen out from organic com-
methylester and formic acid benzylester)], the products
of ring-opening reactions (maleic anhydride), decarbox- position of peroxide species.
pounds forming radicals and/or to promote the decom-
[23,24]
ylation reactions (benzene and phenol) and coupling re-
A dramatic change in conversion and selectivity was
actions (biphenyl, methyl-substituted biphenyl and 1,2- observed when a small amount (1% of toluene) of polar
biphenylethane).
After screening severalmetaloxides, we focused our
attention on copper-based binary metaloxide cata ly sts.
organic additives was added to the reaction media. As
shown in Table 3, NBS and 1,1,2,2-tetrobromoethane
promoted the reaction while phenol retarded it. These
The copper and cobalt binary oxide catalyst was tested results are typical for the case of a radical auto-oxidation
first and gave a relatively low conversion of 0.7% (Ta- reaction, where the former two additives act as initiators
ble 2, entry 1). Among these catalysts, those of entries of the radicalreaction by supp ly ing bromine radicaland
2
and 4 have been reported to be highly selective (96% phenolacts as a radicalscavenger. A part of acetic acid
selectivity) and active (97% conversion) in the oxida- was converted to benzylacetate, which may be formed
tion of cresols with oxygen to corresponding aldehydes from a reaction of benzyla cl oholwith acetic acid. The
at 348 K.
[20–22]
However, both catalysts gave low conver- two bases, 3-methylpyridine and pyridine, were charac-
[a]
Table 1. Catalytic oxidation of toluene with single components catalysts.
[
b]
Entry
Catalyst
Conversion
mol%)
Product selectivities (%)
(
[c]
[c]
[c]
CHO
CH OH
COOH
Others
2
1
2
3
4
5
6
7
No catalyst
0
–
–
49.9
4.3
12.5
11.5
20.6
13.0
–
0
0
0
0
0
0
–
0
0
0
0
1.6
1.3
g-Al O
2.0
1.1
1.4
1.8
2.0
2.5
50.1
95.7
87.5
88.5
77.8
85.7
2
3
Fe/Al O
2
3
Mn/Al O
2
3
Zn/Al O
2
3
Co/Al O
2
3
Cu/Al O
2
3
[
[
[
a]
b]
c]
Toluene (50 mL, 0.47 mol) was oxidized in a 500-mL stirred autoclave at 463 K and 1.0 MPa O for 2 h on 1.0 g catalyst
Conversion was calculated from GC measurement using 1,2,4,5-tetramethylbenzene (durene) as an internal standard.
2
CHO, CH OH and COOH refer to benzaldehyde, benzyl alcohol and benzoic acid, respectively, unless otherwise noted.
2
1988
asc.wiley-vch.de
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1987 – 1992

Liquid Phase Oxidation of Toluene to Benzaldehyde
FULL PAPERS
[a]
Table 2. Catalytic oxidation of toluene with binary oxides catalysts.
[
b]
Entry
Catalyst
Conversion
Product selectivities (%)
(%)
CHO
CH OH
COOH
Others
2
1
2
3
4
5
6
7
8
9
CuCo [1: 0.3] (10)
CoCu/C [1: 0.8] (10)
CuZn [1: 0.3] (10)
CuMn/C [1: 0.2] (7)
CuMn [1: 0.3] (10)
CuFe [1: 0.3] (10)
CuFe [1: 0.3] (10)
CuFe[1: 0.3]/MgO
CuFe[1: 0.3]/TiO2
0.7
1.1
1.1
1.4
1.9
7.4
25.4
1.2
1.8
12.4
57.0
86.6
71.7
74.9
45.6
27.4
41.5
27.2
36.7
28.0
13.4
28.3
19.1
23.8
1.0
29.9
0
0
21.0
15.0
0
0
0
3.2
27.1
71.6
0
2.8
3.5
0
14.1
29.7
[c]
44.4
37.9
5.2
[
a]
Toluene (50 mL, 0.47 mol) was oxidized in a 500-mL stirred autoclave at 463 K and 1.0 MPa O for 2 h on 1.0 g catalyst,
2
which was supported on g-Al O unless otherwise noted.
Numbers in square brackets and parentheses are the molar ratios of two metals and the weight % of metal to support.
Reaction time 4 h.
2
3
[
[
b]
c]
Figure 1. The effect of the Fe/Cu molar ratio. Reaction condi-
tions: toluene 50 mL; catalyst 1.0 g; reaction temperature 463
Figure 2. The competition tests.
K; reaction pressure 1.0 MPa O ; reaction time 2 h.
2
was preferentially adsorbed on the catalyst over benzal-
dehyde. The amount of pyridine adsorbed did not vary
while that of benzaldehyde decreased to 30% of that
terized by high selectivity for benzaldehydeþbenzylal-
cohol( >92%) and low selectivity for benzoic acid (< in the single component experiment. It can be concluded
8%). In particular, pyridine promoted the reaction from the above experiments that the weakly bonded
with 86% selectivity for benzaldehyde after 2 h of reac- benzaldehyde on the catalyst surface is expelled by pyr-
tion at 7.3% conversion of toluene, which is the same as idine, which is more strongly held on the catalyst surface.
the originalactivity (7.4%). These cata ly tic performan-
3-Methylpyridine (pK ¼5.80) is more basic than pyri-
a
ces were similar to or better than those obtained in the dine (pK ¼5.67) and thus expected to adsorb more
a
gas phase oxidation of toluene.
strongly.
The preferentialdesorption of adsorbed benza dl e-
In Figure 3 are shown the time courses of the reaction
hyde by pyridine could be responsible for the selectivity catalyzed by a 10 wt % CuFe (1:0.3)/g-Al O (lower),
2
3
increase for benzaldehyde. Accordingly, we conducted and its homogeneous counterpart of a copper and iron
competitive adsorption experiments with benzalde- nitrate mixture (1:0.3 in mole) (upper). Benzyl hydro-
hyde, pyridine, and a mixture of both on the catalyst sur- peroxide was not detected by iodometry in the heteroge-
face in toluene solution. As shown in Figure 2, the adsor- neous reaction media at the early stages of reaction al-
bed amount of pyridine was almost two times more than though it is reported in the literature on the homogene-
[
25]
that of benzaldehyde when they were adsorbed sepa- ous catalysis reaction. Because of the preferentialfor-
rately. On competitive adsorption, however, pyridine mation of unidentified products as detected by GC, the
Adv. Synth. Catal. 2005, 347, 1987 – 1992
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1989

FULL PAPERS
Feng Wang et al.
[a]
Table 3. The effect of different additives.
[
b]
Additives
Conversion
Product distribution (%)
(
%)
CHO
45.6
CH OH
COOH
27.1
Others
3.5
2
None
7.4
23.8
Phenol0.34
-Methylpyridine
Acetic Acid
Pyridine
83.5
0
0
16.5
3
2.0
5.6
7.3
9.2
63.7
25.8
85.9
18.8
18.7
36.3
33.1 (27.6)
6.1
14.5
9.1
0
41.0
8.1
60.9
61.2
0
0
0
5.8
11.0
[
c]
1
,1,2,2-Tetrabromoethane
[d]
NBS
31.0
[
[
[
[
a]
b]
c]
d]
Reaction conditions are the same as those of Table 1.
Molar ratio of toluene to additive is 100.
The number in the parentheses is the selectivity of benzyl acetate.
N-Bromosuccinimide.
Figure 3. Time-course of reactions over CuFe/g-Al O (lower) and a homogeneous catalyst (upper). Reaction conditions: tol-
2
3
uene 50 mL; CuFe (1:0.3)/g-Al O 1.0 g; reaction temperature 463 K; reaction pressure 1.0 MPa O ; reaction time 2 h.
2
3
2
sum of selectivities was almost zero at the early stage of that the rate of reaction (slope of conversion vs. time) in-
the reaction. As the reaction proceeded, the amount of creased with time in the case of the heterogeneous reac-
unidentified products decreased and oxidation products tion although it was constant in the homogeneous case.
at the benzylic position were produced. The product dis- This result suggests that the reaction intermediates are
tribution among the oxidation products in the heteroge- accumulating on the heterogeneous catalyst. The accu-
neous reaction was quite different from that in the ho- mulating intermediates will surely speed up the reaction
mogeneous reaction. Adsorption of the reaction prod- rate through enhancing the extent of desorption of ben-
ucts on the heterogeneous catalyst may account for zaldehyde.
this difference. Compounds having an acidic proton
In Figure 4 are shown the reusability tests of catalyst.
such as benzyl alcohol and benzoic acid should be In the reusability test, a used catalyst was washed with
strongly held on the heterogeneous catalyst, as com- ethanoland deionized water, and then emp lo yed for an-
pared with benzaldehyde. Consequently, benzaldehyde other reaction. After the catalyst was used for four times,
was observed at the early stage of reaction while benzyl the conversion of toluene and the selectivity for benzal-
alcohol and benzoic acid were not. It should be noted dehyde (column 4 in Figure 4) still remained compara-
1990
asc.wiley-vch.de
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1987 – 1992

Liquid Phase Oxidation of Toluene to Benzaldehyde
FULL PAPERS
and 1.0 MPa of oxygen pressure for 2 h. At the
long reaction time of 4 h, the conversion increased
to 25.4% and benzoic acid became the predomi-
nant reaction product at 71.6%.
b) A small amount of polar additives caused a dramat-
ic change in conversion and selectivity. Pyridine im-
proved the selectivity for benzaldehyde to 86%
while conversion retained at 7.3% for 2 h. It was
found that pyridine was more strongly held on
the catalyst surface than benzaldehyde and thus
prevented further oxidation of benzaldehyde.
c) The catalyst was reusable after washing and drying.
After the catalyst had been used four times, about
95% conversion of toluene and 97% selectivity for
benzaldehyde are still maintained with respect to
their originalva ul es.
Although further researches will be required for appli-
cations of the copper-oxide based catalyst system in syn-
thesizing benzaldehyde, such an impressive activity and
selectivity, as we have showed here, opens a new possi-
Figure 4. Catalyst reusability tests. Reaction conditions: tol-
uene 50 mL; CuFe (1:0.3)/g-Al O 1.0 g; reaction tempera-
2
3
ture 463 K; reaction pressure 1.0 MPa O ; reaction time 2 h.
2
3
bility for activating sp hybridized carbon atoms of inac-
tive hydrocarbons to synthesize the corresponding alde-
hydes through heterogeneous catalysis with molecular
oxygen in the liquid phase.
Experimental Section
Catalyst Preparation and Characterization
The g-alumina support was impregnated with an aqueous sol-
ution of metalnitrate(s) according to the meta l- to-metalratio
and metal weight %, as listed in Table 1, using the incipient
wetness method. After impregnation, the samples were dried
at 393 K and then calcined at 823 K for 3 h in a quartz tube
for transformation to the oxides.
X-ray powder diffraction patterns of the catalysts were re-
corded on a Rigaku Miniflex with CuKa radiation (a¼
0
.1542 nm) at 30 kV and 15 mA. The pore diameter and the
Figure 5. The XRD spectra of fresh catalyst and four times
used catalyst.
BET surface area were characterized using a Micromeritics
ASAP 2000 Physi/Chemisorption apparatus. The g-Al O -sup-
2
3
ported catalysts have a mean pore diameter of 66 Š and a spe-
2
cific surface area of 233 m /g (BET).
ble to the values of the first run (column 1 in Figure 4).
Figure 5 shows the recorded XRD spectrum, which indi-
cates that the main characteristics of the catalyst were
preserved during recycling and reuse.
Catalytic Reaction
The oxidation reactions were carried out in a 500-mL stirred
autoclave at 463 K and 1.0 MPa of oxygen atmosphere. In a
typical reaction, toluene (50 mL, 0.47 mol) and catalyst
Conclusion
(
1.0 g) were charged into the reactor. The autoclave was heated
up to a set temperature in a nitrogen atmosphere. At the set
temperature, nitrogen was replaced with oxygen and the reac-
tion started. After a certain reaction time interval, products
were analyzed by an HP-4890 gas chromatograph and quanti-
fied using the internalstandard 1,2,4,5-tetramethy bl enzene
In summary, following conclusions are drawn from these
studies:
a) An activity of 7.4% and more than 90% selectivity
of oxidized products at the benzylic position were
obtained in liquid phase oxidation of toluene on a (durene). The internalstandard was added to samp el s before
copper-iron binary metaloxide cata yl st at 463 K
injection into GC.
Adv. Synth. Catal. 2005, 347, 1987 – 1992
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1991

FULL PAPERS
Feng Wang et al.
The data of saturated vapor pressure and explosive limits for
toluene in air and oxygen atmosphere were carefully calculat-
ed before we launched the reaction. The explosive limit for tol-
uene is 1.27–7.75% (volume ratio). Equation (1) used for this
purpose is as follows with y standing for the saturated vapor
pressure and x for the temperature (8C):
[3] S. S. Stahl, Angew. Chem. Int. Ed. 2004, 43, 3400–3420.
[4] C. Limberg, Angew. Chem. Int. Ed. 2003, 42, 5932–5954.
[5] A. D. Sadow, T. D. Tilley, Angew. Chem. Int. Ed. 2003,
4
2, 803–805.
[
6] C. Subrahmanyam, B. Louis, F. Rainone, B. Viswana-
than, A. Renken, T. K. Varadarajan, Catal. Commun.
2
002, 3, 45–50.
[
[
7] A. Martin, U. Bentrup, A. Brückner, B. Lücke, Catal.
Lett. 1999, 59, 61–65.
8] F. Konietzni, H. W. Zanthoff, W. F. Maier, J. Catal. 1999,
ð1Þ
1
88, 154–164.
Based on the calculated data, the toluene oxidation reaction
under 10 atm oxygen or air pressure at 1908C will fall into
the lower explosive limit zone and thus is a safe reaction in
the laboratory.
[9] A. Martin, U. Bentrup, G.-U. Wolf, Appl. Catal. A 2002,
227, 131–142.
10] F. Bruhne, E. Wright, Ullmannꢀs Encyclopedia of Techni-
cal Chemistry, 6th edn., 1998, Electronic Release (ben-
zaldehyde entry).
[
[
11] F. Konietzni, U. Kolb, U. Dingerdissen, W. F. Maier, J.
Catal. 1998, 176, 527–535.
Competitive Absorption
[
[
12] S. Lars, T. Andersson, J. Catal. 1988, 98, 138–149.
13] Y. Ishii, S. Sakaguchi, T. Iwahama, Adv. Synth. Catal.
Amount of absorption was determined using three stock solu-
tions of benzaldehyde (4.8 mM), pyridine (6.2 mM), and a mix-
ture of benzaldehyde and pyridine(4.8 and 6.2 mM, respective-
ly), in toluene solution with an internal standard of durene
2
001, 343, 393–427.
[
14] K. Nomiya, K. Hashino, Y. Nemoto, M. Watanabe, J.
Mol. Catal. A 2001, 176, 79–86.
15] T. Garrell, S. Cohen, G. C. Dismukes, J. Mol. Catal. A
2002, 187, 3–15.
(
10.7 mM). Catalyst (0.5 g) was added to 10 mL of the stock
[
solutions, and the concentration of each component was moni-
tored by GC at a time intervals of 1, 2 and 3 days.
[
[
16] Snia-Viscosa, Hydrocarbon Proc. 1977, 134, 210–212.
17] A. R. Zhu, S. C. Tsang, Chem. Commun. 2002, 2044–
2
045.
Acknowledgements
[
[
18] K. R. Seddon, A. Stark, Green Chem. 2002, 2, 119–123.
19] C.-C. Guo, Q. Liu, X.-T. Wang, H.-Y. Hu, Appl. Catal. A
2005, 282, 55–59.
[20] F. Wang, J. Xu, S. Liao, Chem. Commun. 2002, 626–627.
[21] F. Wang, G. Yang, W. Zhang, W. Wu, J. Xu, Chem. Com-
mun. 2003, 1072–1073.
This workreceived support from the National Natural Science
Foundation of China (Project 20233040) and from the National
Hi-Tech Research and Development Program of China (Project
2002AA321020).
[
22] F. Wang, G. Yang, W. Zhang, W. Wu, J. Xu, Adv. Synth.
Catal. 2004, 346, 633–638.
References
[
23] G. B. Shulꢀpin, J. Mol. Catal. A 2002, 189, 39–66.
[
1] R. A. Sheldon, H. van Bekkum, Fine Chemicals through
Heterogeneous Catalysis, Wiley-VCH, Weinheim, 2001,
pp 1–10.
[24] T. V. Bukharkina, N. G. Digurov, Org. Process Res. Dev.
2004, 8, 320–329.
[25] G. A. Russell, J. Am. Chem. Soc. 1957, 79, 3871–3877.
[2] J. M. Thomas, R. Raja, G. Sankar, R. G. Bell, Nature
1
999, 398, 227–230.
1992
asc.wiley-vch.de
ꢁ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1987 – 1992