Redirection of oxidation reactions by a polyoxomolybdate: Oxydehydrogenation instead of oxygenation of alkanes with tert-butylhydroperoxide in acetic acid

Full text of DOI:10.1021/ja005620e

J. Am. Chem. Soc. 2001, 123, 6437-6438
Redirection of Oxidation Reactions by a
6437
Polyoxomolybdate: Oxydehydrogenation Instead of
Oxygenation of Alkanes with tert-Butylhydroperoxide
in Acetic Acid
Alexander M. Khenkin and Ronny Neumann*
Department of Organic Chemistry
Weizmann Institute of Science, RehoVot, Israel 76100
ReceiVed September 19, 2000
Selective oxidation of alkanes remains a very difficult but very
important research goal. In this context many catalysts have been
described that use hydroperoxides such as tert-butylhydroperoxide
(TBHP) and hydrogen peroxide as oxidants for oxygenation of
alkanes to alcohols, ketones, and other or further oxygenated
products. Among the many catalytic systems looked at in this
area during recent years, there may be pointed out the somewhat
mechanistically controversial iron-based biomimetic systems1 and
the metal-substituted zeolite catalysts.2 Also in the area of our
present interest, various transition metal-substituted polyoxo-
metalates have been shown to effectively catalyze oxygenation
of alkanes.3 For oxidation of alkanes with TBHP and H2O2 it is
now well-accepted by many researchers in the field that the
reaction mechanism is of a radical nature and a Haber-Weiss
type mechanistic scheme may be invoked (Scheme 1, reactions
1-4).
Figure 1. Type of product distribution in the oxidation of alkanes with
TBHP catalyzed by H3PMo12O40. Reaction conditions: 1.8 M alkane,
0.6 M TBHP, 0.03 M H3PMo12O40 in acetic acid, 80 °C in air. The
reaction time was 20-60 min (until all TBHP was used up). The type of
product is given as mol % specific product type out of all product types.
Control experiments (a) without H3PMo12O40 the reactions were very slow,
72 h, 9% conversion, and only oxygenated products were observed. (b)
In oxidation of 1.8 M cyclooctane with 0.03 M mononuclear MoO2(acac)2
the reaction was slower, 5 h, the conversion was lower, 14%, and
selectively to cyclooctene was only 14%. (c) In other solvents such as
acetonitrile, nitromethane and tert-butyl alcohol, the selectivity in
cyclooctane oxidation was only ∼40% to dehydrogenation, and a
significant amount of epoxide was formed. Acetic acid was unique in
that no epoxidation was observed.
We have now found that in the presence of a Keggin-type
polyoxomolybdate, H3PMo12O40, as catalyst, the usual pathway
to oxygenated products such as ketones, alcohols, and peroxides
is strongly inhibited, and instead dehydrogenated products,
alkenes, are formed with high selectivity.
Scheme 1. Mechanism for Formation of Alkenes and Acetates
A reaction was typically carried out by heating a solution of
1.8 M alkane, 0.6 M TBHP (70%, 30% H2O) and 0.03 M H3-
PMo12O40‚24H2O in acetic acid at 80 °C in air. Reactions under
argon gave practically the same results. Depending on the alkane
used, the reaction time was about 20-60 min, that is until all of
the TBHP is consumed. The catalyst was unchanged under the
reaction conditions as measured by 31P NMR. The completion of
the reaction was easily observable. In the presence of TBHP, the
polyoxometalate was in the yellow oxidized form; once the TBHP
was used up the green reduced polyoxometalate appeared. The
results showing the selectivity to the product types and the various
dehydrogenated products are summarized in Figure 1 and Table
1, respectively. From Figure 1 one may observe that for a series
of cyclic alkanes, the formation of the dehydrogenated product(s)
dominates with typically 90 ( 5% selectivity. The minor products
were (a) oxygenates such as alcohols, ketones, and peroxides,
(CH3)3COO-R (RH is the cyclic alkane) and (b) esters of acetic
acid which, however, were probably not formed from intermediate
alcohols (see below). For compounds with cyclohexyl moieties
there was some tendency for aromatization. Thus, cyclohexane,
methylcyclohexane, and 1,3-dimethylcyclohexane yielded some
benzene, toluene, and m-xylene, respectively. cis-Decalin gave a
broader product distribution with formation mainly of octahy-
dronaphthalene and tetralin. In general, dienes were found only
in small (decalin) or trace (other examples) amounts. Most
interestingly, cycloalkenes were preferably formed at the less
substituted positions. Thus, decalin was not dehydrogenated at
the bridging position, and in the oxidation of methylcyclohexane,
1-methylcyclohexene was a minor product compared to 3- and
4-methylcyclohexene.
Acylic alkanes as substrates also showed that the dehydroge-
nated products were dominant. The relative percentage of
dehydrogenation was a function of the linearity of the alkane.
Thus, as the branching of the alkane is increased, the selectivity
toward dehydrogenation increased from 60% for n-octane to 100%
for 2,2,4-trimethylpentane. It is notable, however, that this
increased selectivity is at the expense of reduced yield on TBHP
(Table 1). As was observed in the oxidation of cyclic alkanes,
the formation of the double bond at the less substituted carbon
centers was preferred. Thus, in the oxidation of 2,2,4-trimethyl-
pentane, the terminal alkene, 2,2,4-triMe-4-pentene, was formed
in 4-fold excess relative to the internal alkene, 2,2,4-triMe-3-
pentene. An exemplary alkyl aromatic compound, ethylbenzene,
showed only the formation of styrene and R- and â-phenylethyl
acetate. The acetates were obtained in a ∼5:1 ratio.
(1) (a) MacFaul, P. A.; Wayner, D. D. M.; Ingold, K. U. Acc. Chem. Res.
1998, 31, 159. (b) Walling, C. Acc. Chem. Res. 1998, 31, 155.
(2) (a) Arends, I. W. C. E.; Sheldon, R. A.; Wallau, M.; Schuchardt, U.
Angew. Chem., Int. Ed. Engl. 1997, 36, 1144. (b) Huybrechts, D. R. C.; De
Bruycker, L.; Jacobs, P. A. Nature 1994, 345, 240.
(3) (a) Faraj, M.; Hill, C. L. J. Chem. Soc., Chem. Commun. 1987, 1487.
(b) Neumann, R.; Khenkin, A. M. Inorg. Chem. 1995, 34, 5753. (c)
Cramarossa, M. R.; Forti, L.; Fedotov, M. A.; Detusheva, L. G.; Likholobov,
V. A.; Kuznetsova, L. I.; Semin, G. L.; Cavani, F.; Trifiro´, F. J. Mol. Catal.
1997, 127, 85. (d) Matsumoto, Y.; Asami, M.; Hashimoto, M.; Misono, M. J.
Mol. Catal. 1996, 114, 161.
As noted in the literature, reactions with TBHP invariably
proceed with initial formation of tert-butylperoxy and tert-
10.1021/ja005620e CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/07/2001

6438 J. Am. Chem. Soc., Vol. 123, No. 26, 2001
Communications to the Editor
Table 1. Distribution of Oxidative Dehydrogenation Products in
the Oxidation of Alkanes with TBHP Catalyzed by H3PMo12O40
by the polyoxomolybdate must be Very fast because alkyl radicals
react with molecular oxygen at diffusion-controlled rates. Such
electron-transfer oxidation of alkyl radicals is rather unique; it
has mostly been described in the past using Cu(II) salts, preferably
CuII(OAc)2.5 Also, in related research Cu salts have been shown
to decompose alkyl hydroperoxides in macrocycles with intramo-
lecular formation of a double bond,6 and TBHP in the presence
of copper chelates and iron compounds have been reported to
dehydrogenate cyclooctane with low selectivity in air.7 It is also
very worthwhile mentioning that polyoxotungstates have been
shown to partially dehydrogenate photochemically formed alkyl
radicals to alkenes under anaerobic conditions.8 In air oxygenated
products dominated.
A more complete discussion of the reaction mechanism is
beyond the scope of this paper, but some initial comments that
would explain the product selectivity toward the formation of less
substituted alkenes is worthwhile. A competitive oxidation at 75
°C of C6H12/C6D12 showed an isotope effect of kH/kD ) 3.5 (
0.3. This would seem to indicate that the rate-determining step
of the reaction is hydrogen abstraction of the alkane by the tert-
butyloxy radical. In a substrate with 3°, 2°, or 1° H atoms, radical
formation at the more substituted carbon is preferred, but perhaps
not exclusive. The formation of less substituted alkenes can be
explained using 2,2,4-triMe-pentane as an example.9 A carbo-
cation at the 4-position would lead to a 3/1 mixture of 2,2,4-
triMe-4-pentene/2,2,4-triMe-3-pentene from strictly statistical
considerations. However, results are skewed in consideration of
the likelihood of the formation of other carbocations and their
relative stability. Also, it is likely that since 3° carbocations are
more stable to further reaction (alkene formation), the yield based
on TBHP consumption is reduced in 2,2,4-triMe-pentane com-
pared to that based on n-octane since TBHP is decomposed in a
parallel reaction.
distribution of dehydrogenation
substrate
cyclooctane
cycloheptane
cyclohexane
yieldb
products, mol %c
30
22
15
12
cyclooctene (>99)
cycloheptene (>99)
cyclohexene (94), benzene (6)
3- & 4-Me-cyclohexened (83),
1-Me-cyclohexene (11), toluene (6)
1,3-diMe-5-cyclohexene (91),
1,3-diMe-3 & 4-cyclohexenec (8),
3-xylene (2)
Me-cyclohexane
1,3-diMecyclohexane
12
cis-decalin
43
octahydronaphthalenee (32),
hexahydronaphthalene (4),
tetralin (53),
dehydronaphthalene (8),
naphthalene (3)
octane
27
14
1-octene (23), 2-octene (27),
3-octene (27), 4-octene (23)
2-Me-4-pentene (43),
2-Me-pentane
2-Me-3-pentene (45),
2-Me-2-pentene (5),
2-Me-1-pentene (7)
2,2,3-triMe-pentane
2,2,4-triMe-pentane
ethylbenzene
7
4
6
2,2,3-triMe-3-pentene (50),
2,2,3-triMe-4-pentene (50)
2,2,4-triMe-4-pentene (79),
2,2,4-triMe-3-pentene (21)
styrene (100)
a Reaction conditions: 1.8 M alkane, 0.6 M TBHP, 0.03 M
H3PMo12O40 in acetic acid, 80 °C in air. The reaction time was 20-60
min (until all TBHP was used up). b Yield based on TBHP assuming
1 equiv needed per formation of one double bond and oxygenated
product. c The amount of product is given as mol % specific product
out of all dehydrogenated products. d Mixtures, the components of
which were not separated by GC. e Dehydrogenation at the 1,2 and
3,4 positions not at the 9,10 position.
A polyoxometalate compound has been shown to catalyze an
alkane to alkene oxydehydrogenation with TBHP as oxidant by
a mechanism proposed to be a combination of tert-butoxy radical
formation and electron transfer oxidation of an intermediate alkyl
radical.
butyloxy radicals. The importance of the formation of these
radicals in this case could also be demonstrated using two accepted
probe reactions.1a Addition of an effective radical scavenger, tert-
butylcatechol, in the oxidation of cyclooctane totally inhibited
formation of any product. Also when using a probe hydroperoxide,
PhCH2C(CH3)2OOH instead of TBHP no cyclooctane conversion
was noticeable, since the corresponding tertiary alkoxy radical
undergoes a very fast â-scission prior to hydrogen abstraction
from cyclooctane. The initially formed tert-butylperoxy and tert-
butyloxy radicals were also trapped and observed by ESR
(Supporting Information) using two spin traps, 5,5-dimethyl-1-
pyrroline-N-oxide (DMPO) and N-tert-butyl-R-phenylnitrone
(PBN). The use of DMPO allows one to differentiate between
carbon- and oxygen-centered radicals. The spectrum shows that
only oxygen-centered radicals were trapped. The kinetic profile
of the reaction at 55 °C (Supporting Information) indicates that
there was no induction period, autocatalysis, or other complex
behavior.
Acknowledgment. This research was supported by the Basic Research
Foundation administered by the Israeli Academy of Science and Humani-
ties. Dr. Lev Weiner is thanked for the ESR measurements.
Supporting Information Available: The ESR spectrum of the trapped
radicals and the reaction profile (PDF). This material is available free of
charge via the Internet at http://pubs.acs.org.
JA005620E
(4) Papaconstantinou, E. J. Chem. Soc., Faraday Trans. 1 1982, 78, 2769.
(5) Jenkins, C. L.; Kochi, J. K. J. Am. Chem. Soc. 1972, 94, 843.
(6) (a) Cekovic, Z.; Green, M. M. J. Am. Chem. Soc. 1974, 96, 3000. (b)
Schreiber, S. L. J. Am. Chem. Soc. 1980, 102, 6163.
(7) (a) Barton, D. H. R.; Be´vie`re, S. D.; Hill, D. R. Tetrahedron 1994, 50,
2665. (b) Barton, D. H. R.; Be´vie`re, S. D.; Chavasiri, W.; Doller, D.; Hu, B.
Tetrahedron Lett. 1993, 34, 567. (c) Launay, F.; Patin, H. New J. Chem. 1997,
21, 247.
(8) Renneke, R. F.; Hill, C. L. Angew. Chem., Int. Ed. Engl. 1988, 27,
1526. (b) Renneke, R. F.; Hill, C. L. New J. Chem. 1987, 11, 763. (c) Renneke,
R. F.; Pasquali, M.; Hill, C. L. J. Am. Chem. Soc. 1990, 112, 6585.
(9) Disproportionation as an explanation for formation of less substituted
alkenes is not likely under aerobic conditions, because relatively long-lived
radicals would have to be postulated. In ref 8 such a mechanism has been
proposed for anaerobic conditions, but in air there is autooxdiation.
Our explanation of these results, Scheme 1, is that first the
alkyl radical is formed by reaction with the tert-butyloxy radical.
The alkyl radical is then oxidized by electron transfer to the
polyoxomolybdate to yield a carbocation.4 The latter may then
yield the alkene by loss of a proton or an acetate after nucleophilic
attack by the solvent. It is clear that the capture of the alkyl radical