Selective and environmentally benign aerobic catalytic oxidation of alcohols by a molybdenum-copper system

Article abstract of DOI:10.1002/(sici)1099-0682(200004)2000:4<655::aid-ejic655>3.0.co;2-r

The oxidation of activated primary and secondary alcohols to the corresponding aldehydes and ketones can be carried out with molecular oxygen, in the presence of the bimetallic molybdenum-copper system MoO₂(acac)₂- Cu(NO₃)₂ as catalyst.

Full text of DOI:10.1002/(sici)1099-0682(200004)2000:4<655::aid-ejic655>3.0.co;2-r

FULL PAPER
Selective and Environmentally Benign Aerobic Catalytic Oxidation of Alcohols
by a Molybdenum-Copper System
Christian Y. Lorber,[a][°] Sebastian P. Smidt,[a] and John A. Osborn*[a]
Keywords: Molybdenum / Copper / Oxidation / Alcohols / Catalysis
The oxidation of activated primary and secondary alcohols to
the corresponding aldehydes and ketones can be carried out
with molecular oxygen, in the presence of the bimetallic mol-
ybdenum–copper system MoO2(acac)2–Cu(NO3)2 as catalyst.
Introduction
tion of benzylic alcohol. Knowing the capability of cop-
per(I) for the fixation and/or activation of molecular oxy-
gen,[6] we introduced a catalytic amount of copper(I/II) salt
in place of the sulfoxide. As we expected, the oxidation of
PhCH2OH was effective with this new heterobimetallic
Mo–Cu–O2 catalytic system in toluene at 100 °C
(Scheme 2). Table 1 shows blank experiment results proving
that both molybdenum and copper catalysts are essential to
oxidize PhCH2OH to PhCHO under a molecular oxygen at-
mosphere.
The oxidation of primary and secondary alcohols to al-
dehydes and ketones, respectively, is a fundamental trans-
formation in organic synthesis, that can be now accomp-
lished by several oxidizing reagents. However, in most cases
these reagents are very toxic and must be used in stoichi-
ometric amounts.[1] Considering cost and environmental
factors, the development of ‘‘green’’ catalytic processes is
an important goal.[2]
During the last few years we have been interested in using
high oxidation state oxo complexes of transition metals as
catalysts for the oxidation and/or isomerisation of alco-
hols,[3,4] and we recently discovered the catalytic oxidation
of primary and secondary alcohols with cis-dioxomolyb-
Scheme 2
denum(VI) complexes with sulfoxides or N-oxides as co- Table 1. Comparative blank and catalytic experiment results on
oxidants (Scheme 1).[5]
PhCH2OH oxidation
Catalyst[a]
t[b] (h)
PhCHO (%)[c]
MoVI
5
3
3
3
5
3
3
5
5
MoVI/O2
CuII
6
0
Ͻ 2
0
6
98
75
Scheme 1
CuII/O2
FeII/O2
MoVI/CuII
MoVI/CuII/O2
Although DMSO is a cheap oxidizing agent, DMS would
[d]
[d]
be an unpleasant side-product in large-scale applications.
Here we wish to report the selective catalytic oxidation of
alcohols by a new heterobimetallic molybdenum-copper
system that uses molecular oxygen as the oxidizing source.
MoVI/FeII/O2
[a] MoVI ϭ 1 equiv. MoO2(acac)2, CuII ϭ 1 equiv. Cu(NO3)2, FeII
ϭ
[b]
1 equiv. FeBr2. –
20 equiv. PhCH2OH, 5 equiv. 4-toluic acid,
molecular sieves (4 A), toluene, 100 °C. – [c]GC determination. –
˚
[d]
Catalytic run.
Results and Discussion
Typically, a great variety of Mo complexes can be used
(e.g. alkoxides, chlorides, dithiocarbamates...), but further
studies showed that the nature of the ligands around the
molybdenum center is critical to the success of the reaction,
with acetylacetonate (acac–) proving the most effective.[7]
On the other hand, varying the nature of the copper(II)
counterion had little effect on the rate/yield of the reaction,
for example CuSO4, Cu(OAc)2, CuCl2, Cu(NO3)2, Cu-
(acac)2 and copper(I) salts such as CuCl can also be used.
Other transition metals salts known to activate O2, such as
CoII, FeII and MnII have also been tested. With Co(acac)2,
MnBr2, or FeBr2 instead of a CuII salt we observed a slower
oxidation of PhCH2OH with selectivities ranging from 75–
94%. The Mo–Fe system, although less active, behaved sim-
In our previous catalytic system, R2SϭO was used to re-
oxidize the molybdenum-dioxo catalyst.[5] Initial investi-
gations showed that substituting R2SϭO by dioxygen gave
only stoichiometric amount of benzaldehyde for the oxida-
[a]
´
´
Laboratoire de Chimie des Metaux de Transition et de Catalyse,
UMR 7513 CNRS, Universite Louis Pasteur,
4 rue Blaise Pascal, 67000 Strasbourg, France
Fax: (internat.) ϩ 33-3/88416171
E-mail: osborn@chimie.u-strasbg.fr
Present address: Laboratoire de Chimie de Coordination, UPR
8241 CNRS,
[°]
205 route de Narbonne, 31077 Toulouse Cedex, France
E-mail: lorber@lcc-toulouse.fr.
Fax: (internat.) ϩ 33-5/615530023
Eur. J. Inorg. Chem. 2000, 655Ϫ658
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1434Ϫ1948/00/0404Ϫ0655 $ 17.50ϩ.50/0
655

C. Y. Lorber, S. P. Smidt, J. A. Osborn
FULL PAPER
ilarly to the Mo–Cu system, i.e. no oxidation is observed more than 20 equivalents of substrate lead to lower rates
without the Mo catalyst (cf. blank experiments in Table 1); due to deactivation of the catalyst by the water formed.
however, we verified that CoII or MnII alone (i.e. without Therefore, the activity of the Mo–Cu system appears to be
Mo catalyst) is responsible for some of the alcohol oxida- lower than that of Ru–Cu systems,[2b,2c,4] and is limited to
tion.[9]
The addition of 5 equivalents of a carboxylic acid (e.g. 4-
the oxidation of benzylic alcohols.
Studies on the oxidation of a series of para-substituted
toluic acid) to the molybdenum-copper system improved benzylic alcohols (R ϭ Me, H, Cl, CF3, NO2) show no cor-
the rate and yields of the reaction; stronger acids such as p- relation with Hammett σ parameters and little variation in
TsOH have the opposite effect.[10] For similar reasons, we rate on the para substituent. Hence, benzyl radicals would
˚
used powdered molecular sieves (4 A) to trap the water appear not to be involved in the rate determining step of
formed during the alcohol oxidation, which improved the the oxidation process,[12,13] a conclusion which is further
catalyst lifetime.[2a,2c,5] Therefore, our typical catalytic sys- supported by the observation that radical traps such as
tem consists of MoO2(acac)2 (1 equiv.), Cu(NO3)2 (1 2,4,6-tri-tert-butylphenol or galvinoxyl have no effect on
equiv.), 4-toluic acid (5 equiv.), substrate (alcohol, 20 the rate of the oxidation. The catalytic oxidation of
˚
equiv.), powdered activated molecular sieves (4 A), 1 atm of PhCH(D)OH with the Mo–Cu system at 100 °C gives a
O2 in toluene[11] at 100 °C.
product ratio of PhCDO/PhCHO ϭ 3.0. In parallel experi-
The catalytic oxidation of various alcohols takes place ments, the overall rate of the catalytic oxidation of
under these conditions, and a few results are presented in PhCH2OH and PhCD2OH at 100 °C gives a value of kH/
Table 2. Oxidation of various alcohols with the Mo–Cu–O2 catalytic system
[a]
[b]
˚
20 equiv. alcohol, 1 equiv. MoO2(acac)2, 1 equiv. Cu(NO3)2, 5 equiv. 4-toluic acid, MS 4 A, toluene, 100 °C, 1 atm O2. –
GC
[c]
[d]
determination. – Solvent: 1,2-dichlorobenzene, side-reaction: isomerisation into 2-methyl-3-buten-2-ol (32%). – Side-reaction: ether-
ification into ROR (R ϭ cyclohexene) (30%). – At 120 °C. – Selectivity in aldehyde ϩ acid: 66%.
[e]
[f]
Table 2. Primary benzylic alcohols are oxidized with nearly kD ϭ 1.8. Overall, these results support the direct involve-
total selectivity to aldehydes with no further oxidation to ment of the benzylic C–H bond in the rate-determining
the corresponding acid (Entries 1–5). Secondary benzylic step.
alcohols are oxidized to ketones with good selectivity but
Although the complete mechanism of oxidation with this
at lower rates (Entry 6). Allylic alcohols are oxidized by this new catalytic system is not yet fully understood, we propose
method with poor selectivity (Entries 7–8), a common side- the catalytic cycle depicted in Figure 1 based on the above
reaction being isomerisation[5] to an allylic alcohol or de- observations for the oxidation of PhCH2OH. An initiation
hydration leading to formation of an ether. Longer periods would first result in the formation of an alkoxide complex
of reaction are needed for the oxidation of aliphatic alco- (a) by displacement of the acac ligand, as already seen in
hols, with only moderate selectivity (Entries 9–12), and in the oxidation of alcohols with other molybdenum cata-
the case of tBuCH2OH further oxidation into tBuCO2H lysts[5] and by previous experiments carried out in our
was observed at 120 °C (Entry 11). Attempts to oxidize laboratory, which show that such exchange processes occur
656
Eur. J. Inorg. Chem. 2000, 655Ϫ658

Selective and Environmentally Benign Aerobic Catalytic Oxidation of Alcohols
FULL PAPER
˚
distilled and stored under nitrogen (over molecular sieves 4 A for
liquids) in a glove box. 4-Toluic acid (Aldrich) was dried over P2O5
˚
under vacuum. Molecular sieves (4 A) were activated at 150 °C
under vacuum, powdered and stored in a glove box. Toluene was
dried over sodium, distilled and stored over molecular sieves in a
glove box. MoO2(acac)2 was purchased from Aldrich or prepared
following a literature procedure.[22] Cu(NO3)2 was obtained by dry-
ing Cu(NO3)2 2.5H2O under vacuum at 100 °C for 48 hours and
stored under nitrogen. PhCD2OH was prepared by treating ethyl
benzoate with LiAlD4 in diethyl ether, followed by hydrolysis with
sodium hydroxide (10%), and distillation from CaH2. PhCH(D)OH
was prepared by treating PhCHO with LiAlD4 in THF, followed
by hydrolysis with Na2SO4 10H2O, dried with anhydrous MgSO4,
and distilled from CaH2. The analyses of the catalytic reactions
were carried out with a Hewlett–Packard 5890 Series II Gas Chro-
matograph equipped with a flame ionisation detector (FID) and a
Hewlett–Packard HP-1 methyl silicon gum column of length 10 m,
diameter 0.53 mm and 2.65 µm film thickness.
Figure 1. Proposed mechanism for the Mo–Cu–O2-catalysed oxida-
tion of PhCH2OH; apparently low coordinate species b represented
in this figure will have additional neutral ligands, such as alcohol,
which have been omitted for simplicity
readily at electrophilic MoVI centers.[3,14] This point is fur-
ther consistent with the higher rate of oxidation of
PhCH2OH vs. PhCH(Et)OH which may result from the fact
that prior coordination of the alcohols to the metal would
General Procedure for the Oxidation of PhCH2OH: In a typical
oxidation experiment, under nitrogen (glove box), PhCH2OH
(257 mg, 20 equiv.), MoO2(acac)2 (39 mg, 1 equiv.), Cu(NO3)2
be disfavored for the latter. A concerted process involving (23 mg, 1 equiv.), 4-toluic acid (81 mg, 5 equiv.), powdered molecu-
˚
lar sieves (4 A) (400 mg), and 2 g of toluene were placed in a 5- or
10-mL flask equipped with a stir bar and a reflux condenser. The
reaction flask was connected to a molecular oxygen vacuum
Schlenk line and heated at 100 °C for 3 h under 1 atm of O2. Yield
of PhCHO: 98% (determined by GC).
the transfer as a hydride of a β-hydrogen atom of the alkoxo
ligand to one of the molybdenum-oxo entities via a five-
membered cyclic transition state follows in a rate-determin-
ing step, with formation of PhCHO and a MoIV species (b).
This transfer would result in the observed moderate prim-
ary kinetic isotope effect and is similar to the hydride trans-
fer in other Mo-catalysed systems for the oxidation of alco-
hols.[5] The reoxidation of (b) into (a) by the copper salt in
the presence of molecular oxygen is unclear. We can exclude
a mechanism in which a copper(II) species may serve as the
sole oxidant for the molybdenum by a pure redox Wacker-
type process,[15] because the oxidation of PhCH2OH in the
absence of molecular oxygen but with a twofold excess of
CuII salt relative to the substrate (i.e. Mo/Cu/substrate ϭ
1:40:20) proceeds with different characteristics (slow rate
and very low selectivity in benzaldehyde) than the normal
catalytic system under dioxygen atmosphere (Mo/Cu/sub-
strate ϭ 1:1:20).[16] In consequence, the regeneration of the
MoVI active species (a) most certainly involves the activa-
tion of dioxygen by the copper co-catalyst.[17]
Acknowledgments
We would like to thank Dr. B. Meunier, Dr. P. Cassoux, Dr. J.
Kress and Dr. J.-P. Le Ny for helpful discussions. We gratefully
acknowledge the support of this research by the Centre National
de la Recherche Scientifique.
[1] [1a]
S. V. Ley, J. Norman, W. P. Griffith, S. P. Marsden, Syn-
[1b]
thesis 1994, 639–666. –
R. A. Sheldon, J. K. Kochi, in
Metal-Catalyzed Oxidations of Organic Compounds, Academic
Press, New York, 1981.
For catalytic oxidations see, for example, ref.[1a,4,5] and:
W.
[2]
[2a]
P. Griffith, S. V. Ley, G. P. Whitcomb, A. D. White, J. Chem.
[2b]
Soc., Chem. Commun. 1987, 1625–1627. –
R. Lenz, S. V.
Ley, J. Chem. Soc., Perkin Trans. 1 1997, 3291–3292. – [2c] I. E.
´
´
Marko, P. R. Giles, M. Tsukazaki, I. Chelle-Regnaut, C. J.
Urch, S. M. Brown, J. Am. Chem. Soc. 1997, 119, 12661–
[2d]
´
I. E. Marko, P. R. Gilles, M. Tsukazaki, S. M.
12662. –
Further studies are needed to clarify the molybdenum re-
oxidation pathway and the exact function of the copper and
carboxylic acid co-catalysts.
Brown, C. J. Urch, Science 1996, 274, 2044–2046. – [2e] A. Dijk-
sman, I. W. C. E. Arends, R. A. Sheldon, Chem. Commun.
[2f]
1999, 1591–1592. –
S. I. Murahashi, T. Naota, Y. Oda, N.
Hirai, Synlett 1995, 733–734.
[3]
[4]
[5]
[6]
C. Y. Lorber, J. A. Osborn, Tetrahedron Lett. 1996, 37, 853–
Conclusion
856.
K. S. Coleman, C. Y. Lorber, J. A. Osborn, Eur. J. Inorg. Chem.
1998, 1673–1675.
C. Y. Lorber, I. Pauls, J. A. Osborn, Bull. Soc. Chim. Fr. 1996,
133, 755–758.
In conclusion, we describe a new, ecologically benign,
Mo–Cu catalytic system for the aerobic oxidation of alco-
hols. Although so far limited to the oxidation of benzylic
alcohols and with only moderate activities, this catalytic
system remains very interesting from a mechanistic point of
view and has been extended to the oxidation of alcohols by
other oxo-metal complexes with higher activities.[4,21]
[6a]
See for example:
N. Kitajima, Adv. Inorg. Chem. 1992, 39,
[6b]
1–77. –
N. Kitajima,Y. Moro-oka, Chem. Rev. 1994, 94,
B. J. Hathaway, in Comprehensive Coordination
[6c]
737–757. –
Chemistry (Eds.: G. Wilkinson, R. D. Gillard, J. A. McClev-
[6d]
erty), Pergamon Press, 1987, vol. 5, pp. 534–774. –
K. D.
Karlin, Z. Tyeklar, A. D. Zuberbühler, in Bioinorganic Catalysis
(Ed.: J. Reedijk), Marcel Decker, New York, 1993, pp. 261–315.
[7]
Compare, for example, the oxidation of PhCH2OH into
PhCHO at 100 °C: MoO2(acac)2 3 h/98% PhCHO, MoCl5 5 h/
55% PhCHO, [PPh4][MoO2(NCS)4] 6 h/63% PhCHO,
MoO2(Et2dtc)2 12 h/82% PhCHO. We have also used with suc-
cess immobilized Mo catalyst such as PBI Mo (PBI ϭ polybenz-
Experimental Section
General: All the substrates were purchased from Aldrich or Lancas-
ter and were carefully dried over a suitable drying agent before use,
Eur. J. Inorg. Chem. 2000, 655Ϫ658
657

C. Y. Lorber, S. P. Smidt, J. A. Osborn
FULL PAPER
imidazol)[8] which is less sensitive to degradation by water
‘‘normal’’ Mo–Cu–O2 system). In the presence of carboxylic
acid, we observed the formation of 30% of PhCHO, but with
a low selectivity (selectivity was only 60%, instead of 98% for
the Mo–Cu–O2 system). We believe that most of the PhCHO
formed in anaerobic conditions in presence of 4-toluic acid re-
sults from a Cu-catalysed oxidation of PhCH2OH because in a
blank experiment (40 equivalents CuII, 5 equivalents 4-toluic
acid, but no Mo catalyst) we formed ca. 18% of PhCHO. We
also noticed the same behavior with our Mo–Fe system when
performing the same kind of experiments with an excess of
FeIII salt under argon [MoO2(acac)2 1 equivalent, FeCl3 40
equivalents, 4-toluic acid 5 equivalents, PhCH2OH 20 equiva-
lents 100 °C, 5 hours] with formation of only 3% PhCHO (with
a selectivity in PhCHO of only 4%) which is in contrast to 75%
PhCHO (selectivity in PhCHO: 75%) with the Mo–Fe system
under O2 (Mo/Fe 1:1).
(yield of PhCHO after 10 h: 98%).
[8]
M. M. Miller, D. C. Sherrington, J. Chem. Soc., Chem. Com-
mun. 1994, 55–56.
[9]
H. Mimoun, in Comprehensive Coordination Chemistry (Eds.:
G. Wilkinson, R. D. Gillard, J. A. McCleverty), Pergamon
Press, 1987, vol. 6, pp.318–410.
[10]
Strong acids like p-TsOH have a detrimental effect on the cata-
lysis, therefore the beneficial effect observed with a carboxylic
acid may not come from its acidic proton character (as previ-
ously seen with other alcohols oxidation catalysts[5]) but from
the coordinating properties of its carboxylate anion.
[11]
1,2-Dichlorobenzene can also be used, but we observed in some
cases unwanted side-products. With donor solvents such as
DMF and MeCN, rates are strongly decreasing, probably by
inhibition of the copper function.
S. L. Scott, A. Bakac, J. H. Espenson, J. Am. Chem. Soc. 1992,
114, 4205–4213.
[17]
Copper(I) and iron(II) complexes are well known to coordin-
ate, activate and transfer molecular oxygen.[6,18–20] We believe
the MoIV complex is first oxidized to MoV by a redox process
induced by the CuII salt which would generate a CuI species.
The CuI species is essential for the next step which would in-
volve the activation of O2 and would result in the oxidation of
the MoV complex with regeneration of the MoVI catalytic spe-
cies a, possibly through a heterobimetallic µ-peroxo- or µ-oxo-
MoVI–CuII complex.
[12]
[13]
Non linear plots can also be due to other causes such as com-
plications arising from side-reactions, steric effects from large
para substituents, or cyclic transition states: [13a] H. Maskill, in
The Physical Basis of Organic Chemistry, Oxford University
[13b]
Press, 1985, chapter 10, pp. 405–473. –
J. March, in Ad-
vanced Organic Chemistry, 3rd ed., Wiley Interscience, 1985,
chapter 9, pp. 237–250.
See for example: [18a] G. Davies, M. A. El Sayed, R. E. Fasano,
[18]
[19]
[14] [14a]
J. Belgacem, J. Kress, J. A. Osborn, J. Am. Chem. Soc.
1992, 114, 1501–1502. – [14b] J. Belgacem, J. Kress, J. A. Osborn,
J. Mol. Catal. 1993, 86, 267–285.
[18b]
Inorg. Chim. Acta 1983, 71, 95–99. –
S. Mahapatra, J. A.
Halfen, E. C. Wilkinson, G. Pan, C. J. Cramer, L. Que, Jr., W.
[18c]
B. Tolman, J. Am. Chem. Soc. 1995, 117, 8865–8866. –
[15]
[16]
S. D. Ittel, G. W. Parshall, in Homogeneous Catalysis, 2nd ed.,
Wiley, New York, 1992.
`
C. Jallabert, C. Lapinte, H. Riviere, J. Mol. Catal. 1980, 7,
127–136.
In a Wacker-type process, to convert one equivalent of
PhCH2OH to PhCHO, the catalytic system would need two
equivalents of Cu catalyst in order to reoxidize the Mo center.
To check the possibility of a pure redox Wacker-type process
we performed the oxidation of PhCH2OH (20 equiv.) in the
absence of O2 with 1 equivalent of MoO2(acac)2 and 40 equiv.
of copper(II) salt [Cu(NO3)2 or Cu(OAc)2], at 100 °C for 3
hours. The reaction was carried out both in the absence of 4-
toluic acid and with 5 equivalents of 4-toluic acid. In the reac-
tion performed in the absence of carboxylic acid, we observed
only the formation of 2% of PhCHO (instead of 29% with the
E. Ochiai, Inorg. Nucl. Chem. Lett. 1973, 9, 987–990.
A. L. Feig, S. J. Lippard, Chem. Rev. 1994, 94, 759. –
J. B. Vincent, G. L. Olivier-Lilley, B. Averill, Chem. Rev. 1990,
[20] [20a]
[20b]
90, 1447–1467.
[21]
[22]
K. S. Coleman, M. Coppe, C. Thomas, J. A. Osborn, Tetrahed-
ron Lett. 1999, 40, 3723–3726.
M. C. Chakravorti, D. Bandyopadhyay, Inorg. Synth. 1992, 29,
129–134.
Received October 13, 1999
[I99361]
658
Eur. J. Inorg. Chem. 2000, 655Ϫ658