Activation of nitrous oxide and selective oxidation of alcohols and alkylarenes catalyzed by the [PV2MO10O40]5- polyoxometalate ion

Article abstract of DOI:10.1002/anie.200390062

A vanadium-substituted polyoxomolybdate ion ([PV₂Mo₁₀O₄₀]^5-, top right) can activate nitrous oxide to promote the highly selective catalytic oxidation of primary and secondary alcohols to aldehydes and ketones, and the oxidation of various alkylaromatic substrates (see scheme).

Full text of DOI:10.1002/anie.200390062

Communications
yield phenol and methanol, respectively.[6] In homogeneous
media, noncatalytic oxygen transfer from nitrous oxide to
alkyl, alkyne, and imido ligands of transition-metal complexes
3
0, 6051 ± 6054; c) K. Shibata, K. Miyazawa, Y. Goto, Chem.
Commun. 1997, 1309 ± 1310.
[
8] Although we have examined other a,b-unsaturated carbonyl
compounds bearing substituents at the a and/or b positions, no
desired addition/elimination reaction product was obtained.
9] H. Nakagawa, S. Sakaguchi, Y. Ishii, 81st Annual Meeting of the
Chemical Society of Japan, 1G549 (2002). In our present study,
the reaction of (4-MeC H )EtSi(OH) (0.3 mmol) with butyl
[
7]
has been observed.
cleaved nitrogen±nitrogen bonds and N O has been decom-
Molybdenum(iii) complexes have
[
8]
2
[
posed at subambient temperatures to N and O2 using a
2
[
9]
ruthenium complex. More relevant to the work reported
herein is the reported oxygen transfer from nitrous oxide to
6
4
2
acrylate 2a (0.9 mmol) in toluene/H O at 12 08C in the presence
of [{Ir(OMe)(cod)} ] (3 mol%) afforded the Mizoroki±Heck-
type product(46%), in which 36% of 2a was converted into 9.
2
[10]
[11]
2
form metal±oxo species.
In one particular case,
a
ruthenium porphyrin has been used to prepare a high-valent
ruthenium±dioxo species capable of oxygen-transfer to nu-
cleophiles such as alkenes and sulfides. Under more extreme
[
10] The rhodium-catalyzed reaction under similar conditions
(
3 mol%
of
[{Rh(OH)(cod)} ],
4-MeC H SiEt(OH)
2
6 4 2
(
0.6 mmol), 2a (0.3 mmol), and 3b (0.6 mmol) in THF (3 mL)
at 708C for 24 h) afforded 27% of the Mizoroki±Heck-type
product 3a and 3% of 4b with 75% recovery of 3b.
conditions (1408C, 10 atm N O), the ruthenium porphyrin has
2
recently been shown to catalyze oxygen transfer to trisub-
[
12]
[13]
stituted alkenes, and the oxidation of alcohols to ketones.
With regard to the ongoing interest in the catalytic activity
[
14]
of polyoxometalates in oxidation reactions, especially with
benign oxidants such as O and H O , we have very recently
2
2
2
found that a manganese-substituted ™sandwich∫-type poly-
oxometalate can also activate N O and lead to the highly
2
[
15]
selective epoxidation of alkenes.
In this paper we will
describe a different set of N O-activation reactions involving
2
the vanadium-containing Keggin-type polyoxomolybdate,
Polyoxometalate Catalysts
[{nBu N} {PV Mo O }]. In this case, selective oxidation by
4
5
2
10 40
CꢀH bond activation was possible leading to the oxidation of
Activation of Nitrous Oxide and Selective
a) primary and secondary alcohols to aldehydes and ketones,
Oxidation of Alcohols and Alkylarenes Catalyzed respectively (Equation (1)) and b) alkyl aromatics to ketones
5
ꢀ
by the [PV Mo O ] Polyoxometalate Ion**
(Equation (2)), or alternatively leading to their oxydehydro-
genation (Equation (3)).
2
10 40
Revital Ben-Daniel and Ronny Neumann*
Nitrous oxide is a potentially interesting oxidant for selective
oxidation of organic substrates since it contains 36 wt%
oxygen, and the by-product of an oxidation reaction would be
environmentally benign molecular nitrogen. These advan-
tages are difficult to realize since N O is generally considered
2
[
1]
[2]
to be both inert and a poor ligand. Thus, very few catalysts
have been reported for the activation of N O and its use as a
For the oxidation of alcohols (Table 1) it was shown that
secondary and primary alcohols were oxidized selectively
(greater than 99.9%) to ketones and aldehydes, respectively,
2
selective oxidant in hydrocarbon oxidation. In the area of
heterogeneous catalysis, various metal oxides have been
[
3]
investigated. The most effective catalysts appear to be iron-
containing acidic zeolites[4] which at elevated temperatures
are thought to yield surface-activated iron±oxo species (a-
Table 1: Oxydehydrogenation of alcohols with N O catalyzed by
2
a]
[
5]
[{nBu
Substrate
-octanol
2-octanol
4
N}
5
{PV
2
Mo10O
40}].[
oxygen). These species are capable of oxygen transfer to
rather inert hydrocarbons, such as benzene and methane, to
Product
Conversion [mol%]
1
1-octanal
42.5
48.1
68.2
89
50.5
50.3
56.8
[
*] R. Neumann, R. Ben-Daniel
Department of Organic Chemistry
Weizmann Institute of Science
2-octanone
cyclohexanone
cyclooctanone
2-hexenal[
2-hexenal[
1-octen-3-one
benzaldehyde
cyclohexanol
cyclooctanol
cis-2-hexen-1-ol
trans-2-hexen-1-ol
1-octene-3-ol
b]
76100 Rehovot (Israel)
b]
Fax: (þ972)8-934-4142
E-mail: ronny.neumann@weizmann.ac.il
benzyl alcohol
>99
[
**] This research was supported by the Basic Research Foundation
administered by the Israeli Academy of Science and Humanities and
the Israeli Ministry of Science. Dr. Lev Weiner is thanked for the ESR
measurements. R.N. is the Rebecca and Israel Sieff Professor of
Organic Chemistry.
[
a] Reaction conditions: alcohol (1 mmol), [{nBu N} {PV Mo O }]
4 5 2 10 40
(
0.02 mmol), N O (1 atm), benzonitrile (1 mL), 1508C, 15 h. Selectivity
2
to the aldehyde or ketone was > 99.9%. [b] The configuration of the
unsaturated aldehyde was not definitively determined. However, since
the catalytic system is inert to alkenes it is likely, but not certain, that the
configuration was retained.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
9
2
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4201-0092 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 1

Angewandte
Chemie
with no further oxidation of aldehydes to carboxylic acids.[16]
Acyclic primary, secondary, and allylic alcohols showed
similar reactivity; yields between approximately 40±60%
were possible. Higher yields were obtained for cyclic secon-
dary alcohols (approximately 60±80%), while benzylic alco-
hols gave quantitative yields of products. Attempts at
increasing the conversion in the cases of 1-octanol and 2-
octanol by lengthening the reaction time, coupled with
Another notable observation is that the quantification of
N O, N , and O by gas chromatography showed no non-
2
2
2
productive decomposition of N O to N and O at 150±1808C
2
2
2
in the absence or presence of a reaction substrate such as
benzyl alcohol or xanthene. In the presence of the afore-
mentioned substrates, N formation was consistent with the
2
degree of reaction of the organic substrate within a 10%
margin of error.
addition of more N O or increasing the reaction temperature
Some preliminary insight into the mode of N O activation
2
2
were not successful. Similar reaction conversions and selec-
tivity were obtained. An increase in the amount of catalyst to
and the reaction pathway is most warranted. First, heating
[{nBu N} {PV Mo O }] (0.01m) in benzonitrile or decalin at
4
5
2
10 40
5
mol% unfortunately led to a decrease in reaction selectivity.
1508C in a N O atmosphere (1 atm) yielded a green solution
2
Importantly, when using O in place of N O there was no
(UV/Vis shows an absorption at 750 nm) indicative of the
2
2
reaction. One should note that in the past we have reported
oxidation of N O by the polyoxometalate, which is reduced
accordingly (Figure 1).
2
5
ꢀ
on the use of the [PV Mo O ] polyoxometalate ion in the
2
10 40
aerobic oxidation of alcohols; however in those cases the
The degree of reduction, calculated by calibration with the
reduction of pure [nBu N] [PV Mo O ] by Zn, is approx-
[
17]
catalyst is only active when supported on active carbon;
{nBu N} {PV Mo O }] is inactive.
4
5
2
10 40
[
imately 55% after 6 h. In the absence of N O, that is, by
4
5
2
10 40
2
The results of the oxidation of alkyl aromatic compounds
heating [{nBu N} {PV Mo O }] in benzonitrile or decalin
4
5
2
10 40
are summarized in Table 2. For substrates without a hydrogen
atom at the a-position, with respect to the position of the
benzylic carbon atom, there was oxygenation at the benzylic
position and selective formation (greater than 99.9%) of the
diaryl ketone or triaryl alcohol as sole products. Lengthening
the reaction time coupled with repressurization of the
reaction vessel could increase the yields. For other substrates,
such as ethylbenzene and i-propylbenzene, the presence of an
a-hydrogen atom led to oxydehydrogenation and the forma-
tion of vinyl benzene products. Control reactions with molecular
oxygen gave much lower yields and reaction mixtures typical
for autooxidation pathways. [{nBu N} {PV Mo O }] did not
only, slight polyoxometalate reduction (approximately 5%)
was observed within 6 h. Importantly, the apparent reduction
of the [{nBu N} {PV Mo O }] polyoxometalate is reversible,
4
5
2
10 40
that is, upon removal of N O, the [{nBu N} {PV Mo O }] is
2
4
5
2
10 40
reoxidized. Additionally, an ESR spectrum of this solution
yielded an anisotropic spectrum which can be assigned to a
reduced [PV Mo O ] species (see Supporting Information).
2
10 40
1
5
Under the same reaction conditions the N NMR peaks of
1
5
N O at d ¼ 150 and 210 ppm disappeared, apparently
2
because of a paramagnetic peak-broadening effect caused
by the reduction of the polyoxometalate by N O. These lines
2
of evidence seem to suggest an unusual and unexpected redox
4
5
2
10 40
catalyze the alkene epoxidation of 1-octene, cyclooctene, or
cyclohexene. In the latter case, some oxydehydrogenation,
or donor±acceptor (D±A) interaction between N O and
2
[{nBu N} {PV Mo O }] (Equation (4)).
4
5
2
10 40
3
mol% conversion to 1,3-cyclohexadiene and then to ben-
zene was observed. It should also be noted that although the
examples given in Table 1 and Table 2describe the use of
quite polar benzonitrile as solvent, similar reactions with
alcohol and alkylaromatic substrates such 1-octanol, 2-
octanol, and diphenylmethane with apolar decalin (decahy-
dronaphthalene) as solvent gave almost identical results. In
contrast fluorobenzene, which was previously used in epox-
idation reactions with nitrous oxide,[ did not lead to
selective oxidation, especially in the case of the oxidation of
alkylaromatic substrates.
Such a reaction is rather surprising considering the high
[
18]
ionization potential of N O (approximately 12.8 eV), but
2
may conceivably be explained by an outer-sphere electron
transfer, in which the polyanionic nature of polyoxometalates
energetically increases the favorability of such D±A inter-
15]
Table 2: Oxidation of alkylaromatics with N O catalyzed by
2
[
a]
[
{nBu N} {PV Mo O }].
4 5 2 10 40
Substrate
Product
Conversion [mol%][a]
triphenylmethane
diphenylmethane
toluene
triphenylmethanol
benzophenone
no product
52.4 (89)
6.2 (32)
0
fluorene
xanthene
ethylbenzene
i-propylbenzene
fluoren-9-one
xanthen-9-one
styrene
11.1 (50)
14.0 (54)
6.8
a-methylstyrene
27.6
[
a] Reaction conditions: substrate (1 mmol), [{nBu N} {PV Mo O }]
4 5 2 10 40
(
0.01 mmol), N O (1 atm), benzonitrile (1 mL), 1508C, 8 h. Selectivity to
Figure 1. The reduction of [{nBu N} {PV Mo O }] by N O. Reaction
2
4
5
2
10 40
2
the product was > 99.9%. [b] Yields in parentheses are those after 48 h,
conditions: [{nBu N} {PV Mo O }] (0.01m) in benzonitrile, 1508C,
4 5 2 10 40
with reintroduction of N O after 24 h.
N O (1 atm).
2
2
Angew. Chem. Int. Ed. 2003, 42, No. 1
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4201-0093 $ 20.00+.50/0
93

Communications
rate-determining step. Thus, for both alkyl
arenes and alcohols, there is initial evidence
indicating homolytic CꢀH bond cleavage in the
rate-determining oxidation step. This is support-
ed by kinetic isotope effects (KIE) which gave
k /k ¼ 11.0 and 15.9 for benzyl alcohol/
H
D
[
D ]benzyl alcohol and xanthene/[9-D ]xan-
2 2
thene, respectively.
Although it is premature to present a full
picture of the reaction scheme, the initial results
on the novel selective oxidation reactions pre-
sented above indicate that initially there is
activation by the electron transfer of N O by
2
the polyoxometalate species, as suggested by
UV/Vis, ESR, and N NMR spectroscopy. The
Figure 2. Reactivity for oxygenation of alkyl aromatics as a function of benzylic CꢀH
bond energy (from gas phase IR CꢀH stretch vibration). Reaction conditions: alkylar-
ene (1 mmol), [{nBu N} {PV Mo O }] (0.01 mmol), N O (1 atm), benzonitrile (1 mL),
15
activated N O then reacts with (or activates) the
2
4
5
2
10 40
2
organic substrate leading to H abstraction from
the latter, as supported by the correlation
between the rate of alkyl arene oxidation, the
CꢀH homolytic bond strength, and the low
1
508C.
actions and the formation of D±A complexes.[19] A similar
redox interaction has been recently observed between a
1 value obtained for benzylic alcohol oxidation; the measured
KIEs support such a scheme. In the case of alcohol oxidation,
initial H abstraction is presumably followed by additional
dehydrogenation to yield the aldehyde or ketone. In the case
of alkyl arene oxidation, the intermediate radical species is
probably oxidized through oxygen transfer from nitrous
oxide.
[
15]
manganese(iii)-substituted polyoxometalate and N O.
2
For further insight into the mode of activation of the alkyl
arene and alcohol substrate, a correlation between reactivity
and a relevant property of the substrates was sought. In the
oxidation of alkyl arenes, the conversion initially increased
linearly as a function of time indicating a zero-order reaction
in the organic substrate. Using these data, one may observe
that the relative reactivity towards the oxidation of alkyl Experimental Section
arenes was in the order: triphenylmethane > xanthene > flu-
Reactions were carried out in 15-mL glass pressure tubes. Typically,
appropriate amounts (see Table and Figure captions) of the substrate
(all highest purity commercial products) and the [{nBu N}
Mo O }] catalyst were dissolved in the reaction solvent. N O
orene > diphenylmethane > toluene, and was well correlated
2
4
{PV2-
5
(
r ¼ 0.96) to CꢀH homolytic bond strength at the benzylic
[
20]
10 40
2
position (Figure 2). Notably, the use of approximately 70%
(greater than 99.95% purity) was introduced to the pressure tube
1
8
O-labeled [{nBu N} {PV Mo O }] (xanthene (1 mmol),
4
5
2
10 40
by four consecutive pump±thaw cycles using a Schlenk line. Con-
versions of the organic substrate were measured by GLC (HP 6890)
using a 5% phenylmethylsilicone (30 m, 0.32mm internal diameter,
0.25 mm coating) column. The identity of the products was verified by
GC-MS.
[
{nBu N} {PV Mo O }] (1 mmol), N O (1 atm), benzonitrile
4 5 2 10 40 2
(
1 mL), 1508C) yielded only the unlabeled product indicating
[
21]
that the source of oxygen in the product was N O.
2
Similarly, a Hammett plot of the oxidation of a series of
2
benzylic alcohols yielded 1 ¼ 0.59 (r ¼ 0.97, Figure 3). The
Received: July 10, 2002
Revised: October 8, 2002 [Z19706]
low 1 value is consistent with the formation of a radicaloid
species at the benzylic carbon in the transition state of the
[
[
[
1] R. G. S. Banks, R. J. Henderson, J. M. Pratt, J. Chem. Soc. A
968, 2886 ± 2889.
2] F. Bottomley, I. J. B. Lin, M. Mukaida, J. Am. Chem. Soc. 1980,
02, 5238 ± 5342.
3] K. J. Zhen, M. M. Khan, C. H. Mak, K. B. Lewis, G. A. Somorjai,
J. Catal. 1985, 94, 501 ± 507; H. F. Liu, R. S. Liu, K. Y. Liew, R. E.
Johnson, J. H. Lunsford, J. Am. Chem. Soc. 1984, 106, 4117 ±
1
1
4121.
[
4] G. I. Panov, CATTECH 2000, 4, 18 ± 32; G. I. Panov, A. Uriate,
M. A. Rodkin, V. I. Sobolev, Catal. Today 1998, 41, 365 ± 385.
5] N. P. Notte, Top. Catal. 2000, 13, 387 ± 394.
[
[
6] M. A. Rhodkin, V. I. Sobolev, K. A. Dubkov, N. H. Watkins, G. I.
Panov, Stud. Surf. Sci. Catal. 2000, 130A, 881 ± 886.
[
7] P. Mansunaga, G. L. Hillhouse, A. L. Rheingold, J. Am. Chem.
Soc. 1993, 115, 2075 ± 2077; G. A. Vaughan, D. Chadwick, G. L.
Hillhouse, A. L. Rheingold, J. Am. Chem. Soc. 1989, 111, 5491 ±
5493; A. M. Baranger, T. A. Hanna, R. G. Bergman, J. Am.
Chem. Soc. 1995, 117, 10041 ± 10046.
Figure 3. Hammett plot for the oxidation of benzylic alcohols. Reaction
conditions: benzylic alcohol (1 mmol), [{nBu N} {PV Mo O }]
4
5
2
10 40
(
0.01 mmol), N O (1 atm), benzonitrile (1 mL), 1508C.
2
9
4
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4201-0094 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 1

Angewandte
Chemie
[
8] A. R. Johnson, W. M. Davis, C. C. Cummins, S. Serron, S. P.
Carbohydrate±Carbohydrate Interactions
Nolan, D. J. Masaev, K. Morokuma, J. Am. Chem. Soc. 1998, 120,
2
071 ± 2085; C. E. Laplaza, A. L. Odom, W. M. Davis, C. C.
Cummins, J. D. Protasiewicz, J. Am. Chem. Soc. 1995, 117, 4999 ±
000.
Probing Specificity in Carbohydrate±
Carbohydrate Interactions with Micelles and
Langmuir Monolayers**
5
[
9] C. B. Pamplin, E. S. F. Ma, N. Safari, S. J. Rettig, B. R. James, J.
Am. Chem. Soc. 2001, 123, 8596 ± 8597.
[
[
[
[
[
10] R. R. Conry, J. M. Mayer, Inorg. Chem. 1990, 29, 4862± 4867;
W. A. Howard, G. Parkin, J. Am. Chem. Soc. 1994, 116, 606 ± 615.
11] J. T. Groves, J. S. Roman, J. Am. Chem. Soc. 1995, 117, 5594 ±
Paul V. Santacroce and Amit Basu*
5
595.
Carbohydrate±carbohydrate interactions (CCIs) between cell
surface glycolipids are important mediators of cell adhesion
during development, metastasis, and signal transduction.[
CCIs are also important for the higher-order structure of
polysaccharides, compaction of the myelin sheath, sperm±egg
cell adhesion, and proteoglycan-mediated sponge cell aggre-
12] T. Yamada, K. Hashimoto, Y. Kitaichi, K. Suzuki, T. Ikeno,
Chem. Lett. 2001, 268 ± 269.
13] K. Hashimoto, Y. Kitaichi, H. Tanaka, T. Ikeno, T. Yamada,
Chem. Lett. 2001, 922 ± 923.
14] N. Mizuno, M. Misono, Chem. Rev. 1998, 98, 171 ± 199; C. L. Hill,
C. M. Prosser-McCartha, Coord. Chem. Rev. 1995, 143, 407 ±
1]
4
55; I. V. Kozhevnikov, Chem. Rev. 1998, 98, 171 ± 198; R.
[2]
gation. The interaction has been studied by a variety of
Neumann, Prog. Inorg. Chem. 1998, 47, 317 ± 370.
[
3a,b]
techniques including vesicle adhesion,
surface plasmon
[
15] R. Ben-Daniel, L. Weiner, R. Neumann, J. Am. Chem. Soc. 2002,
[
3c,d]
resonance (SPR) spectroscopy,
atomic force microscopy
nuclear magnetic resonance (NMR) spectros-
1
24, 8788 ± 8789.
[
3e]
(
AFM),
[
16] It should be noted that alcohols can also be oxidized with benign
oxidants such as O and H O , but the reaction is rarely selective
[
3f,g]
3j]
[3h,i]
copy,
mass spectrometry (MS),
infrared (IR) spectros-
2
2
2
[
[,3k]
for aliphatic primary alcohols, see: R. A. Sheldon, Catal. Today
000, 57, 157 ± 166.
17] A. M. Khenkin, I. Vigdergauz, R. Neumann, Chem. Eur. J. 2000,
copy, surface force measurements,
quartz crystal micro-
2
[3l]
balance (QCM) measurements, and Langmuir monolayer
[
[
[
[
[
[3m]
compression isotherms.
These interactions, which are
6, 875 ± 882; R. Neumann, M. Levin, J. Org. Chem. 1991, 56,
generally calcium ion dependent, require multivalent carbo-
hydrate±-carbohydrate contacts, and the oligomeric nature of
the association has hampered efforts to structurally character-
ize the carbohydrate aggregates at high resolution. In this
communication we report a new method for detecting CCIs
by monitoring the interactions of glycolipid micelles with a
glycolipid monolayer. We show that small changes in carbo-
hydrate structure significantly affect glycolipid association.
The adhesion of a melanoma cell to an endothelial cell is
mediated by carbohydrate±carbohydrate recognition be-
tween the melanoma cell surface ganglioside sialosyllactosyl-
5707 ± 5710.
18] There is no exact correlation between the gas-phase ionization
potential and the solution oxidation potential which is the more
relevant, but unknown, parameter.
19] I. A. Weinstock, Chem. Rev. 1998, 98, 113 ± 170; A. M. Khenkin,
L. Weiner, Y. Wang, R. Neumann, J. Am. Chem. Soc. 2001, 123,
8531 ± 8542.
20] Benzylic CꢀH bond strengths are proportional to the benzylic
CꢀH stretch vibration, as taken from the gas-phase IR spectra.
IR data from NIST at http://webbook.nist.gov/chemistry.
21] The inverse experiment using N218O is also desirable; N218O was
not commercially available and is prohibitively expensive to
synthesize.
ceramide (GM , 1) and the glycosphingolipid lactosylcera-
3
OH
CO2H
O
OH
OH
O
OH
HO
OH
O
HN
C17H35
HO
O
AcNH
O
O
O
C13H27
HO
OH
OH
OH
GM3,
1
O
OH OH
O
OH
O
HN
C17H35
HO
O
O
C13H27
HO
OH
OH
OH
LacCer, 2
[
*] Prof. A. Basu, P. V. Santacroce
Department of Chemistry
Box H, Brown University
Providence, RI02912 (USA)
Fax: (þ1)253-540-0698
E-mail: amit_basu@brown.edu
[**] We gratefully acknowledge financial support from Brown University
(startup funds; Salomon Faculty Research Award), Rhode Island
Foundation, Research Corporation (Research Innovation Award),
and the Petroleum Research Fund administered by the American
Chemical Society. A.B. is an NSF Faculty Early CAREER awardee.
P.V.S. is a Department of Education GAANN Fellow. We thank
Professor Zimmt for the use of his spectrofluorometer.
Angew. Chem. Int. Ed. 2003, 42, No. 1
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4201-0095 $ 20.00+.50/0
95