Investigations into the selective oxidation of vicinal diols to α-hydroxy ketones with the NaBrO3/NaHSO3 reagent: PH dependence, stoichiometry, substrates and origin of selectivity

Article abstract of DOI:10.1016/j.tet.2005.03.056

The NaBrO₃/NaHSO₃ reagent is one of the few oxidizing agents that chemoselectively oxidizes vicinal diols to α-hydroxy ketones with little overoxidation to the corresponding vicinal-dione or dicarboxylic acid. Oxidation reactions performed with this reagent showed strong pH dependence. cis-Vicinal diols reacted faster than trans-vicinal diols to the α-hydroxy ketone product. Hydroxy functional groups at axial ring positions were more readily oxidized than equatorial hydroxy groups. The application of the NaBrO₃/NaHSO₃ reagent for the chemoselective oxidation of vicinal diols was limited to simple systems and failed with more complex monosaccharide compounds probably due to acid catalyzed dehydrogenation reactions. Despite the simple reaction set-up and good selectivity towards the α-hydroxy ketone product, the actual oxidation reaction mechanism is highly complex and postulated to involve at least six different equilibria with a plethora of bromine containing species. A possible oxidation reaction mechanism is discussed.

Full text of DOI:10.1016/j.tet.2005.03.056

Tetrahedron 61 (2005) 4911–4917
Investigations into the selective oxidation of vicinal diols
to a-hydroxy ketones with the NaBrO /NaHSO reagent:
pH dependence, stoichiometry, substrates and origin of selectivity
3
3
Matthias Bierenstiel, Paul J. D’Hondt and Marcel Schlaf
*
Department of Chemistry, University of Guelph, Guelph, Ont., Canada N1G 2W1
Received 17 December 2004; revised 14 March 2005; accepted 15 March 2005
Available online 8 April 2005
3 3
Abstract—The NaBrO /NaHSO reagent is one of the few oxidizing agents that chemoselectively oxidizes vicinal diols to a-hydroxy
ketones with little overoxidation to the corresponding vicinal-dione or dicarboxylic acid. Oxidation reactions performed with this reagent
showed strong pH dependence. cis-Vicinal diols reacted faster than trans-vicinal diols to the a-hydroxy ketone product. Hydroxy functional
groups at axial ring positions were more readily oxidized than equatorial hydroxy groups. The application of the NaBrO /NaHSO reagent
3
3
for the chemoselective oxidation of vicinal diols was limited to simple systems and failed with more complex monosaccharide compounds
probably due to acid catalyzed dehydrogenation reactions. Despite the simple reaction set-up and good selectivity towards the a-hydroxy
ketone product, the actual oxidation reaction mechanism is highly complex and postulated to involve at least six different equilibria with a
plethora of bromine containing species. A possible oxidation reaction mechanism is discussed.
q 2005 Elsevier Ltd. All rights reserved.
1
ides or by a-oxidation of ketones
0
11,12
1. Introduction
and reduction of
a-diketones (Fig. 1). However, the synthetic route through
1
3
a-Hydroxy ketone (a-ketol) is a functional group entity of
selective oxidation of sec.-sec.-vicinal diols has only
14–17
1
–3
natural products with widely studied chemistry.
syntheses of a-hydroxy ketones have been achieved by
The
marginally been covered in the literature.
The secon-
dary vicinal diol functionality is a key structural feature in
sugars. Selective oxidation of one secondary hydroxyl to a
carbonyl function opens the synthetic pathway to pyrano-
or furanosiduloses also known as ‘oxo’-sugars. To our
knowledge, only two methods for this oxidation have been
reported in the literature: an enzymatic process involving a
4
,5
oxidation of alkenes,
by acyloin condensation of
6
7
8,9
diesters, from enol ethers, silylenol ethers and epox-
1
8,19
rare fungal enzyme
0,21
and brominolysis of tin-alkyl sugar
The difficulty in the selective oxidation of a
2
derivatives.
vicinal diol to the corresponding a-hydroxy ketone is the
prevention of overoxidation to a vicinal dione or—after
C–C bond cleavage—dicarboxylic acid (Scheme 1).
Assuming a first order reaction rate of the bimolecular
reaction between the OH-compound, i.e. vicinal-diol or
a-hydroxy ketone, and oxidizing species the relative ratio
of the reaction rate constants k (diol/a-OH-ketone)/k
1
2
(a-OH-ketone/dione) has to be S10 to support a
synthetically useful selectivity discrimination between the
Figure 1. Synthetic routes to a-hydroxy ketones.
Keywords: Chemoselective oxidation; Vicinal diols; a-Hydroxy ketone;
Hypobromous acid.
*
Corresponding author. Tel.: C1 519 824 4120x53002; fax: C1 519 766
499; e-mail: mschlaf@uoguelph.ca
1
Scheme 1. Oxidation of vicinal diols.
0040–4020/$ - see front matter q 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2005.03.056

4912
M. Bierenstiel et al. / Tetrahedron 61 (2005) 4911–4917
2
2,23
a-hydroxy ketone and the vicinal-dione compound.
Table 1. pH dependence of the effectiveness of the NaBrO
reagent in the oxidation of trans-1,2-cyclohexanediol
3
/NaHSO
3
a
This requirement is probably one of the reasons why the
chemoselective oxidation of sec.-sec.-vicinal diols is
limited to only few methods with Ishii’s NaBrO /NaHSO
b
Entry
pHinitial
pHfinal
a-Hydroxy cyclohexanone % (t)
3
3
c
2
4
1
2
0
1.7
0.5
2.3
95% (1 h)
74% (4 h)
reagent being one of them.
9
7% (24 h)
25% (4 h)
0% (24 h)
d
The simple oxidation reaction conditions for the NaBrO3/
NaHSO3 reagent are particularly useful for organic
syntheses as the chemicals are commercially available and
inexpensive off-the-shelf compounds. Yet, the actual
reaction mechanism is very intriguing and displays good
chemoselectivity. In a remarkable methodology, a mild
oxidizing agent is generated in situ by addition of both, an
oxidant, sodium bromate, and a reductant, sodium bisulfite.
Ishii et al. postulated that the actual oxidant was
hypobromous acid, HOBr, as bromohydrins are formed in
3
4
2.3
2.8
7
d
4.0
5
4.4
5
2.3
No reaction (48 h)
No reaction (48 h)
74% (4 h)
e
5
f
1
.7
97% (24 h)
a
Reaction conditions: 1.0 mmol diol, 1.2 mmol NaBrO
NaHSO in 2.2 mL H O and 2.0 mL CH
Determined by GC. Content of 1,2-cyclohexanedione was !5%.
SO added until desired pHinitial value reached.
Na HPO added until desired pHinitial value reached.
1.2 mmol Na SO
Addition of 0.6 mmol H
3
, 1.2 mmol
3
2
3
CN at 21 8CG1 8C.
b
c
d
e
f
H
2
4
2
4
2
4,25
2
3
.
reactions involving C]C bonds.
reagent has also been successfully employed in the
The NaBrO /NaHSO
3 3
2
4
SO .
2
6
g-lactone formation of o-alkyl substituted benzoic acids.
Other reactions that make use of bromate together with
reducing agents are the Belousov–Zhabotinskii class of
values !4 and that the oxidation reactions were faster at
lower pH. When the oxidation was completed the final pH
values of the reaction solutions were higher than the initial
pH values by approximately 0.5 units. In a control
experiment, Na SO —instead of NaHSO —was reacted
2
7,28
oscillating reactions.
Metsger et al. reported that the
oxidation of alkyl ethers occurred in a-position by Br2,
which is generated through the comproportionation reaction
of bromate and bromide to provide a low but steady
2
3
3
with sodium bromate and no reaction occurred within 48 h
Entry 5). With trans-1,2-cyclohexanediol as the substrate, a
reaction solution containing Na SO and NaBrO had an
(
2
9
concentration of bromine.
2
3
3
only slightly acidic pH of 5. When this reaction solution was
subsequently acidified with 0.5 equiv of conc. sulfuric acid
(
Here we report additional insights into the selective
oxidation of vicinal diols to the corresponding a-hydroxy
ketones using the NaBrO /NaHSO reagent, specifically its
scope and limitations with respect to pH, diol substrates, its
reaction mechanism and the origin of the chemoselectivity.
C
0.6 mmol, 1.2 equiv of H ) resulting in pH of 1.7, the
3
3
oxidation reaction immediately proceeded as with the
original NaHSO /NaBrO reaction condition (Entry 2).
3
3
2
.2. NaBrO /NaHSO reagent stoichiometry
3 3
2. Results
Various bromate and bisulfite amounts were added to trans-
1,2-cyclohexanediol substrate in order to investigate the
Oxidation reaction experiments were conducted on a small
scale (4.2 mL, 0.24 M diol substrate) and monitored by
quantitative gas chromatography calibrated against
dimethyl sulfone as the internal standard. Oxidation
products were identified by GC-MS and by comparison
with authentic commercial samples when available. The
oxidation experiments of the monosaccharide substrates
optimum oxidation reaction condition for the NaBrO
NaHSO reagent (Table 2). Ishii et al. reportedly applied
equal molar sodium bromate and sodium bisulfite quantities
/
3
3
2
4
in slight excess with respect to the diol substrate (Entry 2).
Throughout all reactions (Entries 1–12), the initial concen-
tration of trans-1,2-cyclohexanediol (0.24 M, 1.0 mmol)
was held constant and yields of a-hydroxy cyclohexanone
were determined by quantitative GC analysis. Generally,
it was found that even with more equivalents of oxidant, the
were conducted on an NMR scale in D O (0.7 mL, 0.24 M
2
1
13
substrate) and directly monitored by H and C NMR
spectroscopy.
1
,2-cyclohexanedione yield was not significantly increased
2
.1. pH dependence
always remaining in the range of 0–5%. Addition of an
equimolar ratio of NaBrO and NaHSO in half to three
3
3
The oxidation reaction of trans-1,2-cyclohexanediol exhibi-
ted a strong pH dependence (Table 1), an important
observation that has previously not been reported. The
reaction stoichiometry of diol substrate, sodium bromate
and sodium bisulfite was adopted from the established
equivalents with respect to the diol substrate generated
the a-hydroxy ketone quantitatively within 2–120 h (Entries
2–4). The 1,2-cyclohexanedione content was 15% at the
extended reaction time of 120 h (Entry 1). No reaction—
neither oxidation nor reduction—of trans-1,2-cyclohexane-
diol occurred when sodium bromate and sodium bisulfite
were added individually (Entries 5 and 6). The oxidation
reaction leveled off at 51% conversion after 48 h when the
relative ratio of bromate and bisulfite was 0.6–1.2 mmol
(Entry 7). However, addition of 1.2 mmol bromate and
2.4 mmol bisulfite to the diol substrate completed the
oxidation reaction within 24 h (Entry 9). An oxidation
reaction was not observed when the NaBrO : NaHSO ratio
2
4
oxidation procedure published by Ishii et al. Entry 2
corresponds to the reaction condition reported by Ishii et al.,
2
4
resulting in an initial pH of 1.7. Changes of the pH of the
reaction solution were effected by addition of sulfuric acid
or Na HPO . The desired a-hydroxy cyclohexanone
2
4
product was generated almost quantitatively within 1 h to
4 h while the overoxidation product, 1,2-cyclohexane-
2
dione, was only present in 0–5%. It was found that the
oxidation reaction proceeded only in acidic solution at pH
3
3
was less than 1:3 (Entries 10–12). Quantitative oxidation

M. Bierenstiel et al. / Tetrahedron 61 (2005) 4911–4917
4913
a
Table 2. trans-1,2-Cyclohexanediol oxidation with varying amounts of NaBrO
3
and NaHSO
3
reagent
b
Entry
1
NaBrO
3
(mmol)
NaHSO
3
(mmol)
pHinital
2.0
a-OH ketone % (time)
Colour
Yellow
0.6
0.6
20% (1 h)
5
7
8
5% (24 h)
0% (48 h)
5% (120 h)
c
2
3
4
1.2
2.4
3.6
1.2
2.4
3.6
1.7
1.5
1.3
74% (4 h)
7% (24 h)
60% (1 h)
6% (4 h)
73% (0.5 h)
00% (2 h)
Yellow
Yellow
Yellow
9
9
1
5
6
7
1.2
0
0.6
0
1.2
1.2
5.0
5.0
1.8
No reaction
No reaction
15% (1 h)
Colourless
Colourless
Yellow, colourless (48 h)
4
5
8% (24 h)
1% (48 h)
8
9
1.2
1.2
0.6
2.4
2.0
1.6
30% (1 h)
Light yellow
Orange
5
8
9
4% (6 h)
7% (24 h)
8% (48 h)
34% (1 h)
5
9
4% (4 h)
6% (24 h)
1
1
1
0
1
2
1.2
1.2
0.6
3.6
4.8
1.8
1.0
1.4
1.3
No reaction
No reaction
No reaction
Colourless
Colourless
Colourless
a
b
c
Reaction conditions: 1.0 mmol diol in 2.2 mL H O and 2.0 mL CH CN at 21 8CG1 8C.
Determined by GC based on a-hydroxy ketone. Content of dione was !5%.
Final 1,2-cyclohexanedione content was 15%.
2
3
was observed with 1.2 mmol of NaBrO and 0.6 mmol of
3
condition when the diol substrate and the NaBrO /NaHSO
3 3
NaHSO within 48 h (Entry 8).
3
reagent solution were combined immediately upon
preparation.
The colour of the NaBrO /NaHSO oxidation solution is an
3
3
excellent indicator for oxidation activity. The oxidation
reaction proceeded only when the solution was yellow-
orange coloured. No reaction occurred when the solution
was colourless. The initial yellow colour of the oxidation
reaction of entry 7 disappeared when the a-hydroxy-
cyclohexanone content leveled at 51%. In a control
experiment, it was found that the colour change was very
sensitive to the amount of bromate, bisulfite, sulfite and
sulfuric acid. Adjusting the pH of the reaction solution with
sulfuric acid (1.2 N) or NaOH (1.2 N) through a burette or
Pasteur pipette changed the colour of the solution with one
extra drop. The yellow colour/colourless sequence occurred
fully reversibly when base and acid were alternatively
added. When neutral or basic sulfite solutions were acidified
quickly by addition of conc. sulfuric acid brown vapours of
bromine were generated. The addition order of NaBrO3,
NaHSO or Na SO and H SO is irrelevant. Each different
2
.3. Solvent and temperature
The NaBrO /NaHSO reagent was active in various
3
3
solvents such as pure water, aqueous acetonitrile and
aqueous N,N-dimethylacetamide. The trans-1,2-cyclo-
hexanediol oxidation reactions proceeded in these three
solvent systems with similar reaction times, conversions and
selectivities comparable with Ishii’s published conditions
using aqueous-acetonitrile solution. An increase of
reaction rates without the formation of side products or
degradation of substrate or oxidant was found when the
temperature was increased from ambient temperature to 45
and 60 8C. The conversions of trans-1,2-cyclohexanediol to
a-hydroxy cyclohexanone were quantitative after 24, 6 and
2 h, respectively. The contents of the overoxidation product,
1,2-cyclohexane-dione, were not significantly different and
again remained within %5%.
2
4
3
2
3
2
4
order of addition produced the same yellow coloured
oxidizing reagent solution with the same activity towards
the trans-1,2-cyclohexanediol substrate.
2.4. Substrates
3
0
In order to investigate the reactivity and selectivity of the
oxidation of sugar substrates, we focused on relative
reactivity of cis and trans vicinal diols. Using Ishii’s
standard reaction conditions, the oxidation results for the
diol substrates showed that the cis compounds were slightly
more reactive than their corresponding trans isomers
(Table 3, entries 1–3; Table 2, entry 2). The cyclopentane-
diol compounds were oxidized marginally slower than their
cyclohexanediol homologues.
To explore the possibility of deactivation or formation of
other species originating from the oxidizing agent, four
reagent solutions containing sodium bromate (1.2 mmol)
and sodium bisulfite (1.2 mmol) were prepared and stirred
for 24 h and 7 days at room temperature and 60 8C,
respectively. These prepared reagent solutions were then
added to trans-1,2-cyclohexanediol (1.0 mmol) solutions at
room temperature. In each of the four experiments,
the oxidation reaction proceeded quantitatively to the
a-hydroxy cyclohexanone and was identical to the reaction
3
1
The influence of steric hindrance on the reactivity of

4914
M. Bierenstiel et al. / Tetrahedron 61 (2005) 4911–4917
Table 3. Oxidation of various diols and monoalcohols with NaBrO
NaHSO reagent
3
/
2,2,6,6-tetramethylcyclohexanol was oxidized within 2 h,
the oxidation of 2,2,6,6-tetrachlorocyclohexanol failed
a
3
b
(Entries 9 and 10).
Entry
Diol/monoalcohol substrate
Conversion % (t)
6
9
4% (1 h)
9% (6 h)
The intrinsic challenges associated with the quantitative and
qualitative detection of monosaccharide substrates and their
reaction products by GC or GC-MS lead us to conduct
1
2
3
3
2
6
9
4% (1 h)
8% (24 h)
1
13
H and C NMR experiments in D O instead. As demon-
2
strated above, the reactivity of the NaBrO /NaHSO
3
3
reagent in aqueous-acetonitrile solution was similar to
water solvent systems. Methyl a-D-glucopyranoside (1),
methyl a-D-galactopyranoside (2), methyl a-D-manno-
pyranoside (3) and 1,6-anhydro-D-glucopyranoside (4)
were examined under the NaBrO /NaHSO oxidation
5
8
0% (1 h)
9% (24 h)
8
1
5% (1 h)
00% (2 h)
3
3
conditions (Fig. 2). No reaction occurred after 24 h at
room temperature and thus the four samples were heated to
4
5
6
(K)-Menthol
6
0 8C. A slow oxidation process then occurred but after 5 d
3
1
4% (1 h)
00% (8 h)
the sugar substrates had reacted unselectively to complex
intractable product mixtures.
(
C)-Isomenthol
(C)-Neomenthol
100% (!15 min)
7
8
9
100% (1 h)
98% (1.5 h)
100% (2 h)
Figure 2. Investigated monosaccharide substrates.
2
.5. Other oxidizing conditions
Control experiments were conducted in order to determine
the oxidation mechanism and active oxidizing species. As
elemental bromine could be a possible oxidant, 1 equiv of
Br was added to an aqueous-acetonitrile solution contain-
2
ing trans-1,2-cyclohexanediol. The reaction solution was
yellow coloured and produced a complex product mixture,
which was analyzed by GC-MS and contained a-hydroxy
cyclohexanone, various brominated species and other
undetermined compounds. To simulate a low but steady
Br concentration in the reaction solution, a diluted Br
10
0% (24 h)
2
2
2
1 8CG1 8C.
solution (1 equiv) was added with a syringe pump over a 6 h
time period. Again, a complex product mixture was
obtained containing a-hydroxy cyclohexanone along with
a complex mixture of other products, albeit, less side
products were formed. In the NaBrO /NaHSO oxidation
a
Reaction conditions: 1.0 mmol diol, 1.2 mmol NaBrO
NaHSO in 2.2 mL H O and 2.0 mL CH CN at rt and pH 1.7.
3
, 1.2 mmol
3
2
3
b
Determined by GC based on a-hydroxy ketone or ketone. Content of
dione was !5%.
3
3
condition, no brominated side products were detected by
GC-MS. Hypobromous acid, HOBr, was produced by
mixing equal amounts of aqueous equimolar AgNO and
secondary alcohols was explored by the oxidation of a series
of monoalcohols (Table 3, entries 4–10). The steric
positioning of the hydroxyl function in cyclic systems
with respect to an alkyl substituent in axial or equatorial
position was investigated using menthol isomers and
cyclohexanol derivatives. In solution, the hydroxyl func-
tionality is predominantly axially oriented in (C)-neo-
menthol and equatorially in (C)-isomenthol. It has an
intermediate axial–equatorial position in (K)-menthol.
3
33
Br solutions. The trans-1,2-cyclohexanediol oxidation
2
with in situ generated HOBr yielded a complex mixture of
products after 6 h in stark contrast to chemoselective
oxidation with the NaBrO /NaHSO reagent.
3
3
(
K)-Menthol was oxidized in 2 h while its hydroxyl
3. Discussion
epimer, (C)-neomenthol, required less than 15 min to
produce (K)-menthone. (C)-Isomenthol was oxidized
to (C)-isomethone in 8 h.
In consideration of above results, the nature of the active
oxidant, the role of protons and the overall mechanism of
the diol oxidation by the NaBrO /NaHSO system remains
3
3
Additional methyl groups alpha to the secondary alcohol
functionality decreased the reactivity to the corresponding
ketone compound only marginally (Entries 7–9). While the
elusive, however, a few clear features of the apparently
complex system emerge and certain conclusion can be
made.

M. Bierenstiel et al. / Tetrahedron 61 (2005) 4911–4917
4915
K
respect of Br and a continuous low and steady regenera-
tion of HOBr concentration is therefore established.
3
.1. Equilibria and HOBr
Ishii suggested that hypobromous acid, HOBr, is the actual
oxidizing species as bromohydrins are synthesized when the
NaBrO /NaHSO reagent reacts with C]C double
0
0
K
HOBr þ RCHðOHÞR /RCðOÞR þ H2O þ Br þ H
þ
(7)
3
4,34
3
2
bonds.
It is, however, not immediately obvious that
3
.3. Control experiments
HOBr, is the actual oxidizing species, considering that a
:1 NaBrO :NaHSO reagent would generate 1 equiv of
1
3
3
The yellow colour—an indicator for oxidation activity—can
be associated with HOBr and Br but not with bromide,
bromous acid, HOBrO, as bisulfite is oxidized to sulfate
through a formal transfer of oxygen from bromate (Eq. 1).
Disproportionation of HOBrO could then generate HOBr,
but also reform bromate (Eq. 2), ultimately requiring an
excess of bisulfite, but as our stoichiometry study showed
the ratio of NaHSO to NaBrO does only influence the rate
2
bromate, sulfite and sulfate anions which are all colour-
less. Br , a potential active oxidant proposed by Metsger
3
5
2
2
9
et al., is probably not the active oxidant, as a very slow
addition of diluted Br solution to the diol substrate to
2
produce a low and steady Br concentration generated many
2
side products of which some were bromo-species, while
under our standard experimental NaBrO /NaHSO reaction
conditions no carbon–bromine species have been detected.
3
3
of reaction but not the selectivity of the oxidation.
Therefore, the actual oxidant is probably not directly
generated by NaHSO and NaBrO but through a cascade
3
3
3
3
of intricate equilibria involving at least all of the steps listed
in Equations 1–6. Thus the different bromo-species and
potential oxidants HOBrO , HOBrO, HOBr, Br , Br and
Ishii was able to isolate bromo-species only if a C]C
2
4
K
3
double bond was present in the substrate. This is an
indicator that the mechanism is probably ionic rather than
radical. In the control experiment when HOBr was
2
2
potentially the Br radical can be produced in the reaction
2
7,28,35–38
solution in various concentrations.
NaHSO reagent is not deactivated when kept at 60 8C for 7
The NaBrO3/
33
generated in situ from bromine and silver nitrate (Eq. 8),
3
the reaction products contained brominated species.
However, the HOBr concentration in the control experiment
may still have been relatively high so that HOBr can
undergo bromination side reactions in contrast to a low and
steady concentration generated by the NaBrO /NaHSO
days. Thus the actual oxidant must be fairly stable and does
not undergo side reactions once the equilibrium has been
established.
K
3
K
3
2K
4
BrO CHSO #HOBrO CSO
(1)
3
3
reagent.
K
C
2
HOBrO#HOBr CBrO CH
(2)
(3)
(4)
(5)
(6)
3
AgNO CBr CH O/AgBr CHNO CHOBr
(8)
3
2
2
ðsÞ
3
K
3
K
C
BrO C5Br C6H #3Br C3H O
2
2
3.4. Oxidation mechanism and selectivity
K
C
Br CH O#HOBr CBr CH
2
2
The actual oxidation can in principle either occur directly at
the hydroxyl function or at the alpha-position through an
enol type intermediate. An oxidation mechanism that
proceeds through an enol of the organic compound can,
however, be excluded, as the complete oxidation of 2,2,6,6-
tetramethylcyclohexanol, which contains no alpha-hydro-
gen, proceeded at similar reaction rates as the di- and
trimethylated cyclohexanol derivatives.
C
C
C
HOBr CH #½H O KBrꢀ #H O CBr
2
2
K
K
3
Br CBr #Br
2
3
.2. pH dependence
Protons play a key role in the oxidation mechanism. The
increased reaction rate at lower pH can be explained if
Br —or [Br–OH ] —were the actual oxidizing agent.
Br formation is favoured at higher H concentration
because the dehydration of HOBr is proton dependent
initial
Based on the observed pH dependence, we now postulate
that a small but persistent and continuously replenished
concentration of Br or possibly [H O–Br] rather than
neutral HOBr as proposed by Ishii is the actual oxidizing
agent. It is then logical to expect it to preferentially attack
the least acidic hydrogens geminal to the OH functional
group.
C
C
2
C
C
C
C
2
(
reactions (Eq. 2–4) are known to proceed at pH 1.5–2.5,
Eq. 5). The comproportionation and disproportionation
3
8
which asserts our experimental work as yellow colour
and oxidation activity was only detected when the solution
pH was !4.
Once the first oxidation to the a-hydroxy ketone has
occurred, the polarization of the remaining geminal
hydrogen is slightly less negative than that in a vicinal
diol as the carbonyl group withdraws electron density
(Fig. 3) resulting in the observed chemoselectivity of diol
oxidation.
Furthermore, during the oxidation process the pH of the
reaction solution increased, suggesting that protons were
consumed. This is on first sight contradictory as the
oxidation of a secondary alcohol with HOBr generates a
proton along with bromide and carbonyl compound (Eq. 7).
Another equilibrium must be responsible for the overall
consumption of protons. As bromide ions are continuously
produced, the comproportionation reaction of bromate and
bromide to hypobromous acid will require a total of three
protons (Eq. 3 and 4). The reaction is autocatalytic with
Figure 3. Geminal hydrogen polarization in a-hydroxy ketone and diol.

4916
M. Bierenstiel et al. / Tetrahedron 61 (2005) 4911–4917
5. Experimental
trans-1,2-Cyclohexanediol, cis-1,2-cyclohexanediol, trans-
,2-cyclopentanediol, cis-1,2-cyclopentanediol, a-hydroxy
1
cyclohexanone, cyclohexane-1,2-dione, cyclohexanone,
acetonitrile, 2,2,6,6-tetrachlorocyclohexanol, sodium bro-
mate, sodium bisulfite and sodium sulfite are commercially
available and were used as received. cis,cis-2,6-Dimethyl-
cyclohexanol [39170-84-8], cis-2,2,6-trimethylcyclohexa-
nol [20580-68-1] and 2,2,6,6-tetramethylcyclohexanol
C
Figure 4. Axial and equatorial attach of Br on cyclohexanol.
The preferential selectivity of axial vs. equatorial OH
functions studied in the menthol derivatives series, can then
C
also be rationalized by postulating Br as the actual oxidant
[6948-41-0] were obtained by methylation of cyclohexa-
none with MeI and subsequent reduction with LiAlH
which attacks the geminal hydrogen of the hydroxyl group
(
4
3
9
according to literature procedures. The identities of these
compounds as well as those of the corresponding oxidation
products cis,cis-2,6-dimethylcyclohexanone [471256-22-1],
cis-2,2,6-trimethylcyclohexanone [106036-02-6]/106033-
Fig. 4). An equatorial attack by the oxidant is more
preferred over an axial attack, which would experience
more shielding due to 1,3-diaxial interactions. In addition
moving the oxygen from the axial position towards a more
carbonyl-like transition state during the oxidation reaction
should also lead to a release of ring strain induced by the
5
[
6-1] (enantiomers) and 2,2,6,6-tetramethylcyclohexanol
1195-93-3] were established by H NMR and MS. The
1
w
pH values were measured with a Fisher Accumet pH meter
Model 620 and a calomel electrode. The NMR experiments
were conducted on a Varian Avance 400 MHz spectrometer.
GC and GC-MS experiments were conducted on a Varian
CP-3800 and Varian CP-3800-Saturn 2000. Environmental
note: residual solutions of the oxidation reactions should be
quenched with a saturated Na S O solution and properly
1
,3-interactions of the axial hydroxyl. The situation is
slightly different for cis-diols which were oxidized faster
than the trans-diol as there is always one geminal hydrogen
in axial position while in the trans-diol isomer both OH
functions occupy the equatorial position (Fig. 5). The
oxidation of 2,2,6,6-tetrachlorocyclohexanol failed prob-
ably due to the high steric shielding and/or the lowest
negative polarization of the geminal hydrogen due to the
four electron withdrawing chlorine atoms present.
2
2 3
discarded.
5
.1. General reaction conditions
trans-1,2-Cyclohexanediol (116 mg, 1.0 mmol) and sodium
bromate (181 mg, 1.2 mmol) were dissolved in 1 mL water
and 2 mL acetonitrile. A solution of sodium bisulfite
(124 mg, 1.2 mmol) in 1.2 mL water was then added via
syringe. The pH was measured with a pH meter, the reaction
stirred at room temperature (21 8CG1 8C) and monitored by
quantitative GC.
Figure 5. Relative conformations of trans- and cis-1,2-cyclohexanediols.
The chemoselective oxidation of polyhydroxy substrates,
e.g. monosaccharides failed or yielded only complex
intractable reaction mixtures, probably because these
substrates can undergo acid catalyzed condensation
reactions.
5
.2. NMR scale experiments
Methly a-D-glucopyranoside (33 mg, 0.17 mmol) was
reacted with NaBrO3 (30 mg, 0.20 mmol) and NaHSO3
(20 mg, 0.20 mmol) in 0.7 mL D O at room temperature for
2
24 h, then stirred at 60 8C for 5 days and the reaction
monitored by H and C NMR.
1
13
4
. Concluding remarks
The NaBrO /NaHSO reagent establishes an intricate
3
3
Acknowledgements
network of equilibria in which HOBr acts as the source of
but probably is not the actual oxidizing agent. The low and
constantly replenished concentration of the actual oxidant,
Funding for this research project was provided by
University of Guelph, NSERC and CFI. M. B. would like
to thank the Ernst-Schering Research Foundation, Berlin,
Germany for a doctoral stipend.
C
C
possibly [H O–Br] or Br , generates a constant redox
2
potential of the reagent solution. We propose that the redox
potential of the reaction solution reaches a specific value
which is sufficient enough for the first oxidation step of the
diol but does not allow—or only very little—of the second
2
2
oxidation step of a-hydroxy ketone to the dione to occur,
i.e. the NaBrO /NaHSO reagent combination effectively
acts as redox buffer.
3
3
Supplementary data
Supplementary material for this article detailing a generic
kinetic model for the oxidation of vic-diols as well as an
exemplary reaction protocol for larger scale applications of
the NaHSO /NaBrO method can be found, in the online
In order to further elucidate the reaction mechanism a full
quantitative analysis of each of the bromine containing
species in the reaction solution in real time would be
necessary, which constitutes a formidable challenge.
3
3
version, at doi:10.1016/j.tet.2005.03.056.

M. Bierenstiel et al. / Tetrahedron 61 (2005) 4911–4917
4917
References and notes
21. Tsuda, Y.; Hanajima, M.; Mastuhira, N.; Okuno, Y.;
Kanemitsu, K. Chem. Pharm. Bull. 1989, 39, 2344–2350.
2
2
2. See Supplementary Material.
1
2
3
. The Merck Index, 13th ed.; Merck: Whitehouse Station, NJ,
2
3. Atkins, P.; de Paula, J. Physical Chemistry; Freeman: New
York, 2002.
001.
. Larock, R. C. Comprehensive Organic Transformation, 2nd
ed.; Wiley-VCH: New York, 1999.
24. Sakaguchi, S.; Kikuchi, D.; Ishii, Y. Bull. Chem. Soc. Jpn.
1997, 70, 2561–2566.
. Hudlicky, M. Oxidations in Organic Chemistry; ACS Mono-
graph; American Chemical Society: Washington, DC, 1990.
. Sakata, Y.; Katayama, Y.; Ishii, Y. Chem. Lett. 1992, 671–674.
. Murahashi, S.-I.; Saito, T.; Hanaoka, H.; Murakami, Y.; Naota,
T.; Kumobayashi, H.; Akutagawa, S. J. Org. Chem. 1993, 58,
25. Kikuchi, D.; Sakaguchi, S.; Ishii, Y. J. Org. Chem. 1998, 63,
6023–6026.
4
5
26. Hayat, S.; Atta-ur-Rahman; Choudhary, I. M.; Khan, K. M.;
Bayer, E. Tetrahedron Lett. 2001, 42, 1647–1649.
27. Field, R. J.; Koros, E.; Noyes, R. M. J. Am. Chem. Soc. 1972,
2
929–2930.
. Finley, K. T. Chem. Rev. 1964, 64, 573–589.
6
7
9
4, 8649–8664.
. Hashiyama, T.; Morikawa, K.; Sharpless, K. B. J. Org. Chem.
992, 57, 5067–5068.
2
2
8. Noyes, R. M. J. Phys. Chem. 1990, 94, 4404–4412.
1
9. Metsger, L.; Bittner, S. Tetrahedron Lett. 2000, 56,
8
. Rubottom, G. M.; Vazquez, M. A.; Pelegrina, D. R.
Tetrahedron Lett. 1974, 4319–4322.
1
905–1910.
0. All possible reagent combinations of trans-1,2-cyclohexane-
diol, NaBrO , NaHSO , Na SO and H SO were permutated
3
9
. Takai, T.; Yamada, T.; Rhode, O.; Mukaiyama, T. Chem. Lett.
3
3
2
3
2
4
1
0. Tsuji, T. Bull. Chem. Soc. Jpn. 1989, 62, 645–647.
991, 281–284.
yielding quantitative a-hydroxy cyclohexanone within 24 h.
The 1,2-cyclohexanedione content was less than 5%.
1
1
1. Moriarty, R. M.; Hu, H.; Gupta, S. C. Tetrahedron Lett. 1981,
3
3
1. The solutions were allowed to cool to room temperature.
2. Derivation of these compounds, e.g. by exhaustive silylation
would be required to make them analyzable by GC. Given that
the reaction mixtures are aqueous and that the addition of
2
2, 1283–1286.
2. El-Qisairi, A. K.; Qaseer, H. A. J. Organometal. Chem. 2002,
59, 50–55.
1
1
6
3. Hekmatshoar, R.; Heravi, M. M.; Beheshtiha, Y. S. Monatsh. f.
Chem. 2002, 133, 195–197.
silylation agents such as Me Si-imidazolium chloride might
3
influence the reaction equilibria, this is not practical.
3. Derbyshire, D. H.; Waters, W. A. Nature 1949, 164, 446–447.
4. Takase, K.; Masuda, H.; Kai, O.; Nishiyama, Y.; Sakaguchi,
S.; Ishii, Y. Chem. Lett. 1995, 871–872.
1
4. Sakata, Y.; Ishii, Y. J. Org. Chem. 1991, 56, 6233–6235.
5. Curci, R.; D’Accolti, L.; Detomaso, A.; Fusco, C.; Takeuchi,
K.; Ohga, Y.; Eaton, P. E.; Yip, Y. C. Tetrahedron Lett. 1993,
3
3
1
3
4, 4559–4562.
3
3
3
5. Toth, Z.; Fabian, I. Inorg. Chem. 2004, 43, 2717–2723.
6. Toth, Z.; Fabian, I. Inorg. Chem. 2000, 39, 4608–4614.
7. Beckwith, R. C.; Wang, T. X.; Margerum, D. W. Inorg. Chem.
1
1
1
1
2
6. Bovicelli, P.; Lupattelli, P.; Sanetti, A.; Minicione, E.
Tetrahedron Lett. 1994, 35, 8477–8480.
7. Bovicelli, P.; Lupattelli, P.; Sanetti, A. Tetrahedron Lett. 1995,
1
996, 35, 995–1000.
3
6, 3031–3034.
3
8. Greenwood, N. N.; Earnshaw, A., 2nd ed. In Chemistry of the
Elements; Butterworth Heinemann: Oxford, 1997.
8. Giffhorn, F.; Koepper, S.; Huwig, A.; Freimund, S. Enzymol.
Micro. Tech. 2000, 27, 734–742.
9. Freimund, S.; Huwig, A.; Giffhorn, F.; Koepper, S. Chem. Eur.
J. 1998, 4, 2442–2455.
0. Liu, H.-M.; Sato, Y.; Tsuda, Y. Chem. Pharm. Bull. 1993, 41,
39. (a) Lissel, M.; Neumann, B.; Schmidt, S. Liebigs Ann. Chem.
1986, 263–264. (b) Becker, H. Organicum; Addison-Wesley:
Reading, MA, 1973; pp 521–524.
4
91–501.