Hypervalent λ3-bromane strategy for Baeyer-Villiger oxidation: Selective transformation of primary aliphatic and aromatic aldehydes to formates, which is missing in the classical Baeyer-Villiger oxidation

Article abstract of DOI:10.1021/ja104330g

A conceptually distinct, modern strategy for Baeyer-Villiger oxidation (BVO) was developed. Our novel method involves initial hydration of water to carbonyl compounds, followed by ligand exchange of hypervalent aryl-λ³-bromane on bromane(III) with the resulting hydrate, yielding a new type of activated Criegee intermediate. The intermediate undergoes BV rearrangement and produces an ester via facile reductive elimination of an aryl-λ³-bromanyl group, because of the hypernucleofugality. The novel strategy makes it possible to induce selectively the BV rearrangement of straight chain primary aliphatic as well as aromatic aldehydes, which is missing in the classical BVO: for instance, octanal and benzaldehyde afforded rearranged formate esters with high selectivity (>95%) under our conditions, while the attempted classical BVO produced only carboxylic acids. This firmly establishes the powerful nature of new methodology for BVO.

Full text of DOI:10.1021/ja104330g

Published on Web 06/22/2010
Hypervalent λ3-Bromane Strategy for Baeyer-Villiger Oxidation: Selective
Transformation of Primary Aliphatic and Aromatic Aldehydes to Formates,
Which is Missing in the Classical Baeyer-Villiger Oxidation
Masahito Ochiai,*,† Akira Yoshimura,† Kazunori Miyamoto,† Satoko Hayashi,‡ and Waro Nakanishi*,‡
Graduate School of Pharmaceutical Sciences, UniVersity of Tokushima, 1-78 Shomachi, Tokushima 770-8505,
Japan and Department of Material Science and Chemistry, Faculty of Systems Engineering, Wakayama UniVersity,
930 Sakaedani, Wakayama 640-8510, Japan
Received May 19, 2010; E-mail: mochiai@ph.tokushima-u.ac.jp
Scheme 1. Classical and Modern Strategies for BVO
Abstract: Aconceptuallydistinct,modernstrategyforBaeyer-Villiger
oxidation (BVO) was developed. Our novel method involves initial
hydration of water to carbonyl compounds, followed by ligand
exchange of hypervalent aryl-λ3-bromane on bromane(III) with
the resulting hydrate, yielding a new type of activated Criegee
intermediate. The intermediate undergoes BV rearrangement and
produces an ester via facile reductive elimination of an aryl-λ3-
bromanyl group, because of the hypernucleofugality. The novel
strategy makes it possible to induce selectively the BV rearrange-
ment of straight chain primary aliphatic as well as aromatic
aldehydes, which is missing in the classical BVO: for instance,
octanal and benzaldehyde afforded rearranged formate esters
with high selectivity (>95%) under our conditions, while the
5 instead of BVO products (formate esters 4), owing to the facile
attempted classical BVO produced only carboxylic acids. This
ꢀ-elimination of acid ArCO2H (or preferential 1,2-shift of hydrogen
firmly establishes the powerful nature of new methodology for
over the carbon moiety) in Criegee intermediates 1 (Scheme 2):2,8
BVO.
thus, primary aliphatic aldehydes as well as aromatic aldehydes
generally produce the corresponding carboxylic acids 5 upon peracid
treatment. Oxidation of some R-branched and R-nitrogen-substituted
The Baeyer-Villiger oxidation (BVO) of carbonyl compounds
aliphatic aldehydes2,9 as well as electron-rich aromatic aldehydes
with peracids, discovered as early as 1899,1 directly affords valuable
(Dakin oxidation)10 is an exception to the general rule and affords
esters or lactones through a 1,2-shift of an alkyl group from a carbon
formates 4. We report herein difluoro-λ3-bromane-induced BVO
to an oxygen atom in the initially formed Criegee intermediates 1,
of aliphatic and aromatic aldehydes in the presence of water, based
with concomitant release of acid (Scheme 1a).2 The migration step
on the unique ligand exchange strategy depicted in Scheme 1b.
is concerted with retention of stereochemistry of the migrating group
Most importantly, our method makes it possible to selectively
and rate-limiting in the vast majority of cases.3 In contrast to this
induce the normal BV rearrangement of straight chain aliphatic and
classical BVO, we envisioned a conceptually distinct, novel strategy
aromatic aldehydes, yielding formates 4, which is missing in the
for the BV reaction: our method involves initial hydration of water
classical BVO. This firmly establishes the powerful nature of the
to carbonyl compounds, followed by a ligand exchange of hyper-
new methodology for BVO.
valent aryl-λ3-bromane 2 with the resulting hydrate (gem-diol),
producing a new kind of Criegee intermediate R-hydroxyalkoxy-
λ3-bromane 3 (Scheme 1b). An aryl-λ3-bromanyl group exhibits
vastly enhanced leaving group ability:4 nucleofugality of PhBr(BF4)
is greater than that of hyperleaving PhI(BF4), the latter being ∼106
times greater than a superleaving group triflate.5 Therefore, alkoxy-
λ3-bromane 3 would function as a highly activated Criegee
intermediate, yielding a rearranged ester via facile reductive
elimination of an aryl-λ3-bromanyl group. Furthermore, the ligand
exchange step of aryl-λ3-bromane 2 with gem-diol, yielding Criegee
adduct 3, seems to be a low energy process, based on the chemistry
of hypervalent organo-λ3-iodanes.6 The new methodology will be
particularly suitable for the BVO of aldehydes, because the initial
hydration equilibrium is shifted more efficiently toward gem-diols
in aldehydes rather than in ketones.7
Scheme 2
In the presence of 1 equiv of water, exposure of linear decanal
to p-trifluoromethylphenyl(difluoro)-λ3-bromane (2) (1.5 equiv) in
dichloromethane at 0 °C for 3 h under argon resulted in the selective
formation of nonyl formate (4a) in 70% yield (Table 1, entry 2).11
A small amount (8%) of decanoic acid (5a) was produced in the
reaction. This is the first example of highly selective BVO of
unsubstituted linear primary aldehydes. Higher yields (76-88%)
of BV product 4a were obtained by the use of 2-10 equiv of water,
while in the absence of water no formation of 4a was detected.
These results strongly suggest that the initial rapid hydration
equilibrium of decanal plays a critical role in our BVO, which is
in good agreement with our strategy, depicted in Scheme 1b. Use
of acetonitrile as a solvent dramatically changed the reaction course
One of the well established events in the oxidation of aldehydes
with peracids is the selective formation of carboxylic acids RCO2H
† University of Tokushima.
‡ Wakayama University.
9
9236 J. AM. CHEM. SOC. 2010, 132, 9236–9239
10.1021/ja104330g  2010 American Chemical Society

C O M M U N I C A T I O N S
Table 2. λ3-Bromane-Induced BVO of Aldehydesa
and afforded acid 5a (69%) as a major product at the expense of
the formate route (entry 8). Acetonitrile with a larger ꢀ value (0.40)
of solvent hydrogen-bond acceptor basicities than that of dichlo-
romethane (0.10) probably enhances the rate of ꢀ-elimination of
Criegee adduct 3 (R′ ) H) to yield acid 5a.12
Table 1. BVO of Decanal with Difluorobromane 2a
yield (%)b
entry
solvent
H2O (equiv)
time (h)
4a
5a
1
2
3
4
5
6
7
8
9c
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CHCl3
Cl(CH2)2Cl
MeCN
CH2Cl2
0
1
2
3
10
2
2
2
2
3
3
1
1
1
1
1
1
1
0
70
88 (80)
82
76
66
80
27
0
0
8
6
5
5
4
5
69
0d
a Conditions: decanal (0.05 M)/bromane 2 (1.5 equiv)/0 °C/Ar. b GC
yields. Parentheses are isolated yields. c Difluoroiodane p-CF3C6H4IF2
was used instead of 2. d Decanal (80%) was recovered.
Frohn and Giesen reported that difluorobromane 2 is highly
susceptible toward hydrolysis and undergoes a spontaneous redox
reaction with generation of molecular dioxygen and p-CF3C6H4Br
in aqueous acetonitrile solution at room temperature;11a however,
the results of entry 5 indicate that ligand exchange of bromane 2
with gem-diol derived from decanal, yielding intermediate 3, is
faster than the redox reaction of 2 by a large excess (10 equiv) of
water in dichloromethane at 0 °C. In contrast to bromane 2, difluoro-
λ3-iodane p-CF3C6H4IF2 did not undergo BVO of decanal under
our conditions (entry 9). Because the initial carbonyl hydration is
a common step and because the subsequent ligand exchange of
aryl-λ3-iodane with gem-diol will be a rapid process with a low-
energy barrier,6 the differences in nucleofugality between fluo-
ro(phenyl)-λ3-iodanyl and -bromanyl groups during the 1,2-shift
with reductive elimination of halobenzenes in the halogan(III)
Criegee intermediate 3 probably reflect the observed differences
in reactivity.4
To our delight, the new strategy has found general use in the
BVO of aldehydes (Table 2): straight chain unsubstituted primary
aldehydes with Et, Pr, Bu, n-C5H11, n-C7H15, and n-C9H19 groups
undergo λ3-bromane-induced BVO in good to high yields in the
presence of water (2 equiv) under mild conditions and afforded
formates 4 with more than 80% selectivity. In marked contrast,
the attempted classical BVO of pentanal and octanal with m-
chloroperbenzoic acid (m-CPBA) exclusively afforded carboxylic
acids 5 (entries 5, 8).8b The only exception is the reaction with
acetaldehyde, which resulted in preferential formation of acetic acid
with 95% selectivity over the formate route. This is probably due
to the reported extremely low migratory aptitude of the methyl
group compared to the other alkyl groups.13
R-Branching in the carbon chain of aldehydes enhanced a
tendency for the 1,2-shift of alkyl moieties in Criegee intermediates
3, which was reflected in the higher ratios (>98%) of formate
production (Table 2, entries 14-17). On the other hand, introduction
of electron-withdrawing chlorine and bromine atoms to the terminal
methyl group of hexanal decreased the ratios of the formate route
from 94% to approximately 80% (entries 6, 12, 13). The percent
carbon chain migration seems to depend primarily upon the
electron-donor ability of migrating alkyl groups: in fact, experi-
mental relative rate factors kR/kH for formation of 4 relative to that
of 5 (Scheme 2), calculated from the product ratios, correlate very
well with Taft’s polar substituent constants σ* for alkyl groups (r
a Conditions: an aldehyde (0.05 M)/bromane 2 (1.5 equiv)/H2O (2
equiv)/CH2Cl2/0 °C/Ar. b GC yields. Parentheses are isolated yields.
c Ratios of 4:5, determined by GC. d Classical BVO with m-CPBA or
peroxomonophosphoric acid at room temperature. Data taken from refs
2a and 8b.
) 0.97, Figure 1). Taft’s σ* is an electronic parameter and a
measure of the inductively electron-donating power of an alkyl
group.14 The excellent linear σ* plot clearly indicates that the rates
of the 1,2-shift of an alkyl moiety to the electron-deficient oxygen
atom attached to bromane(III) in Criegee intermediate 3 increase
with an increase in the electron-donating power of an alkyl group.
Similar electronic effects of alkyl substituents were reported in the
classical BVO of methyl ketones by permaleic acid and trifluo-
roperacetic acid (see Figure S1).13 The evaluated large negative
F* value of -15 strongly suggests a rate-determining migration of
alkyl groups in the λ3-bromane-induced BVO.8a
Aromatic aldehydes exclusively undergo the λ3-bromane-induced
BVO through a ligand exchange pathway (Table 2). The attempted
classical BVO of benzaldehydes p-XC6H4CHO (X ) H, Cl, Br,
Me, etc.) with PhCO3H, SeO2/H2O2, or [Me(n-C8H17)3N]HSO4/H2O2
produced only benzoic acids 5 (for instance, entries 24, 27),2 while
exclusive formation of aryl formate esters 4 was detected in our
BVO. The only exception is p-CF3C6H4CHO with a powerful
electron-withdrawing substituent (CF3: σp ) 0.54), which deceler-
ates the 1,2-migration of the aryl group.15
Relative rates of BVO for a series of substituted benzaldehydes
were measured by competitive reactions, in which a mixture of
each 5-fold excess of two competing substrates was used. Electron-
withdrawing p-Cl and p-Br groups decreased the relative rates krel
of the reaction: 1.9 (p-MeC6H4CHO), 1.0 (PhCHO), 0.69 (p-
FC6H4CHO), 0.57 (p-ClC6H4CHO), 0.60 (p-BrC6H4CHO). A Ham-
mett plot showed a satisfactory correlation of the relative rate factors
with σp constants and gave the reaction constant F ) -1.2 (r )
0.96).15 A similar moderately large and negative F value of -1.1
9
J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010 9237

C O M M U N I C A T I O N S
Figure 1. Plot of logarithm of 4:5 ratios log (kR/kH) vs Taft’s polar
substituent constants σ* for aliphatic groups.
was reported for classical BVO of substituted acetophenones using
trifluoroperacetic acid in dichloroethane, in which decomposition
of the Criegee adduct constitutes the rate-limiting step.16 Further-
more, these two relative rate factors for the λ3-bromane-induced
and the classical peracid-induced BV reactions correlate very well
with each other (see Figure S2). These results indicate that our
BVO of benzaldehydes probably involves a rate-limiting migration
of aryl groups and that a similar kind of electronic effect of
substituents operates in the transition states of both classical and
modern BVO.
Figure 2. Energy profile for the Me and H migration in BVO of
acetaldehyde based on ab initio calculations. Br (large dark purple), O (red),
C (gray), and F (blue purple).
calculated for the 1,2-shift of the methyl group in λ3-iodane-derived
Criegee adducts (see Figure S4) agrees with our experimental
findings (Table 1, entry 9). Thus, calculations strongly suggest that
the reaction proceeds easily and reasonably to give the products
via the protonated species such as 11 and 12, although the solvent
effect other than the protonation is not considered in the calculations.
Further calculations for BVO of aldehydes RCHO (R ) Et, i-Pr,
t-Bu) revealed that the activation energy for the concerted 1,2-shift
of an alkyl group in λ3-bromane-derived Criegee adducts decreases
considerably with an increasing degree of alkyl substitution in the
order (kJ mol-1) Me (48.7), Et (24.5), i-Pr (16.2), and t-Bu migration
(9.4). Differences in these estimated activation barriers are correlated
well (r ) 0.95) with Taft’s σ* values of alkyl groups (see Figure
S5), again suggesting that the electronic nature of alkyl groups is
of paramount importance in the BV rearrangement.18
Scheme 3
Scheme 3 clearly demonstrates that the essential feature of the
classical BVO holds for our unique BVO technology: oxidation of
chiral (S)-aldehyde 6 with difluorobromane 2 takes place with
extensive retention of stereochemistry at the migrating center.
Scheme 3 also depicts BVO of cyclohexanones, producing seven-
membered lactones in high yields. Preferential migration of the more
substituted methine group of 2-methylcyclohexanone yielding
lactone 8b as a major product was observed, as in the case of the
classical BVO.2
Thus, we have developed a conceptually distinct, new strategy
for the BVO. The method involves a hydration-ligand exchange
sequence to Criegee intermediates using difluorobromane(III). Our
method makes it possible to selectively induce the normal BVO of
straight chain aliphatic as well as aromatic aldehydes, which is
missing in the classical BVO.
Acknowledgment. This work was supported by a Grant-in-Aid
for Scientific Research (B) (JSPS). We thank Central Glass Co.
Ltd., Japan for a generous gift of BrF3.
Calculated structures and the energy profile for the methyl and
hydrogen migration of the activated (protonated) intermediate,
R-hydroxyethoxy(phenyl)-λ3-bromane 11, in the bromane-induced
BVO of acetaldehyde at the MP2/6-311G(d) method of the Gaussian
03 program are illustrated in Figure 2.17 The leaving phenyl-λ3-
bromanyl and the migrating methyl groups in transition state 13
occupy an anti periplanar arrangement with a dihedral angle of
-169.9° about the O-C bond, along which the migration takes
place. A similar anti arrangement (-176.8°) in 10 was evaluated
for the hydrogen migration (see also Figure S3). Transition state
10 for hydrogen shift is predicted to lie 25.3 kJ mol-1 higher in
energy than 13 for the methyl shift, which is not compatible with
the experimental observations of the preferred formation of acetic
acid over methyl formate production (Table 2, entry 1). Therefore,
it seems reasonable to assume that the acetic acid formation
probably involves a ꢀ-elimination process, but not a 1,2-H shift of
Supporting Information Available: Experimental procedures,
supplementary figures, and computational results. This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Baeyer, A.; Villiger, V. Ber. Dtsch. Chem. Ges. 1899, 32, 3625.
(2) For reviews, see: (a) Krow, G. C. Org. React. 1993, 43, 251. (b) Strukul,
G. Angew. Chem., Int. Ed. 1998, 37, 1198. (c) Renz, M.; Meunier, B. Eur.
J. Org. Chem. 1999, 737. (d) Brink, G.-J.; Arends, I. W. C. E.; Sheldon,
R. A. Chem. ReV. 2004, 104, 4105.
(3) For a catalytic version of BVO, see: Corma, A.; Nemeth, L. T.; Renz, M.;
Valencia, S. Nature 2001, 412, 423.
(4) Ochiai, M.; Tada, N.; Okada, T.; Sota, A.; Miyamoto, K. J. Am. Chem.
Soc. 2008, 130, 2118.
(5) (a) Okuyama, T.; Takino, T.; Sueda, T.; Ochiai, M. J. Am. Chem. Soc.
1995, 117, 3360. (b) Stang, P. J.; Rappoport, Z.; Hanack, M.; Subramanian,
L. R. Vinyl Cations; Academic Press: New York, 1979.
the hemiacetal group. A large activation barrier (91 kJ mol-1
)
9
9238 J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010

C O M M U N I C A T I O N S
(6) Ochiai, M. In Topics in Current Chemistry; Wirth, T., Ed.; Springer: Berlin,
2003; Vol. 224, p 5.
(14) Taft, R. W. In Steric Effects in Organic Chemistry; Newman, M. S., Ed.;
Wiley: New York, 1956.
(7) Carey, F. A.; Sundberg, R. J. AdVanced Organic Chemistry; Kluwer
(15) Hansch, C.; Leo, A.; Taft, R. W. Chem. ReV. 1991, 91, 165.
(16) Hawthorne, F.; Emmons, W. D. J. Am. Chem. Soc. 1958, 80, 6398.
Academic: New York, 2000.
(8) (a) Ogata, Y.; Sawaki, Y. J. Am. Chem. Soc. 1972, 94, 4189. (b) Lehtinen,
C.; Nevalainen, V.; Brunow, G. Tetrahedron 2000, 56, 9375. (c) Sato, K.;
Hyodo, M.; Takagi, J.; Aoki, M.; Noyori, R. Tetrahedron Lett. 2000, 41,
1439. (d) Karrer, P.; Haab, O. HelV. Chim. Acta 1949, 32, 973.
(9) (a) Alcaide, B.; Aly, M. F.; Sierra, M. A. J. Org. Chem. 1996, 61, 8819.
(b) DeBoer, A.; Ellwanger, R. E. J. Org. Chem. 1974, 39, 77.
(10) Hassall, C. H. Org. React. 1957, 9, 73.
(17) Structures and energies evaluated by quantum chemical (QC) calculations
essentially correspond to those in the gas phase. We must be careful when
they are discussed on the basis of QC calculations, since other factors,
such as the solvent effect in solutions, would be stronger than those
predicted by QC calculations. However, factors to stabilize the structures
in the gas phase must also operate in solutions, which enables us to clarify
the mechanism of reactions by comparing suitably the calculated and
observed results.
(11) (a) Frohn, H. J.; Giesen, M. J. Fluorine Chem. 1998, 89, 59. (b) Ochiai,
M.; Nishi, Y.; Goto, S.; Shiro, M.; Frohn, H.-J. J. Am. Chem. Soc. 2003,
125, 15304.
(18) For transition state structures of bromane-induced BVO of aldehydes RCHO
(12) Reichardt, C. SolVents and SolVent Effects in Organic Chemistry; Wiley-
(R ) Et, i-Pr, t-Bu), see Figure S6.
VCH: Weinheim, 2003.
(13) Winnik, M. A.; Stoute, V. Can. J. Chem. 1973, 51, 2788.
JA104330G
9
J. AM. CHEM. SOC. VOL. 132, NO. 27, 2010 9239