Role of catechol in the radical reduction of B-alkylcatecholboranes in presence of methanol

Article abstract of DOI:10.1039/b917004a

Mechanistic investigations on the previously reported reduction of B-alkylcatecholboranes in the presence of methanol led to the disclosure of a new mechanism involving catechol as a reducing agent. More than just revising the mechanism of this reaction, we disclose here the surprising role of catechol, a chain breaking antioxidant, which becomes a source of hydrogen atoms in an efficient radical chain process.

Full text of DOI:10.1039/b917004a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Role of catechol in the radical reduction of B-alkylcatecholboranes
in presence of methanolw
a
a
Guillaume Povie,z Giorgio Villa,z Leigh Ford,b Davide Pozzi,a
Carl H. Schiesserb and Philippe Renaud*a
Received (in Cambridge, UK) 18th August 2009, Accepted 10th November 2009
First published as an Advance Article on the web 25th November 2009
DOI: 10.1039/b917004a
Mechanistic investigations on the previously reported reduction
of B-alkylcatecholboranes in the presence of methanol led to the
disclosure of a new mechanism involving catechol as a reducing
agent. More than just revising the mechanism of this reaction,
we disclose here the surprising role of catechol, a chain breaking
antioxidant, which becomes a source of hydrogen atoms in an
efficient radical chain process.
reduction of radicals with trialkylborane–water mixtures8–10
and for the reduction with a titanium–aqua complex.11
So far, no direct experimental evidence for the presence of
complex A in solution has been reported. We report here a
detailed investigation of the reaction mechanism for the
reduction of B-alkylcatecholborane 3 in presence of methanol
based on NMR experiments and a new mechanism involving
free catechol as reducing agent is proposed.
In order to determine the rate constant of the reduction of
the radical by complex A, it was necessary to determine the
effective concentration of the active species A in solution
during the reduction process. In 1985, the study of
triphenylborane–methanol complexation was reported.12 The
rate of exchange between the complex and free species is fast
enough within NMR time scale to give only one weighted
average signal. Complexation was able to shift the triphenyl-
borane peak in 11B-NMR from 68 ppm in the absence
of methanol to 39 ppm in the presence of 3 equivalents of
methanol. Under these conditions, approximately 70% of the
triphenylborane was present as a complex.
Since the seminal work of H. C. Brown, organoboranes have
been widely used in organic synthesis.1 They are easily
obtained via hydroboration of alkenes and subsequent
oxidation of the resulting alkylborane leads efficiently to the
corresponding alcohol.2 However, reduction of alkylboranes
to saturated compounds is far less described. This transformation
is usually achieved by treatment under severe conditions with
propionic acid in refluxing diglyme,3 alkaline protonolysis,4 or
hydrogenolysis at high temperature (190–225 1C).5 Only the
selective radical reduction of one alkyl group of a trialkyl-
borane by an alkylthiol has been described to proceed
efficiently at room temperature.6
Similar behavior, i.e. a significant upfield shift of the
11B-NMR signal, was expected for the complex between
methanol and MeOBCat 4. Therefore, 1H-NMR and
11B-NMR spectra of MeOBCat 4 in C6D6 and in the presence
of increasing amounts of methanol (0.5 to 2 equiv.) were
recorded (Fig. 1). In the absence of methanol, a single 11B
NMR signal at 23.5 ppm was observed (Fig. 1, spectrum a).
A very small upfield shift (0.6 ppm) of the 11B-NMR signal
was observed when methanol was added to MeOBCat 4.
However, a second signal appeared at 18.8 ppm, increasing
in relative intensity with the amount of methanol (Fig. 1,
spectra b–d). The signal at 18.8 ppm matches the signal
observed for trimethylborate 5 (Fig. 1, spectrum f). This initial
We have recently reported a mild procedure for the
hydrogenation of alkenes via hydroboration with an excess
of catecholborane (2 equiv.) followed by treatment with
methanol (4 equiv.) in the presence of air as a radical initiator.
A typical example, the reduction of a-pinene 1a to pinane 2a,
is shown in Scheme 1, eqn (1), similar results were
obtained with a wide range of primary, secondary and tertiary
alkylcatecholboranes.7 Surprisingly, when the pure organo-
borane intermediate 3a was treated with methanol (4 equiv.)
under the same reaction conditions, only a very low yield of
pinane 2a (5%) was obtained [eqn (3)]. As a consequence, it
was proposed that complex A resulting from the complexation
of methoxycatecholborane 4 (MeOBCat, generated in situ
from the excess of catecholborane and methanol) and
methanol was the main source of hydrogen atoms in eqn (1).
It was hypothesized that the O–H bond of methanol
was activated by the complexation with the Lewis acidic
MeOBCat. Similar mechanisms were proposed for the
a University of Berne, Department of Chemisty and Biochemistry,
Freiestrasse 3, CH-3012 Berne, Switzerland.
E-mail: philippe.renaud@ioc.unibe.ch; Fax: (+41) 31-631-3426
b School of Chemistry, Bio 21 Molecular Science and Biotechnology
Institute, University of Melbourne, VIC 3010, Parkville, Australia
w Electronic supplementary information (ESI) available: Experimental
procedures and NMR data for all experiments. See DOI: 10.1039/b917004a
z Part of the PhD theses of G. P. and G. V.
Scheme 1 Alkene hydrogenation via hydroboration-radical hydrogen
transfer.
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 803–805 | 803

View Article Online
Scheme 2 Methanolysis of MeOBCat 4.
Fig. 1 11B-NMR of MeOBCat 4 in the presence of methanol.
Scheme 3 Methanolysis of PrBCat 3b.
result indicates that methanolysis of MeOBCat 4 according to
eqn (3) takes place rather than formation of complex A.
This was further demonstrated by the experiment where
(MeO)3B 5 and catechol were mixed together in benzene leading
to the same spectrum as the one obtained by mixing MeOBCat
4 and methanol (Fig. 1, spectrum e). However, a precise
interpretation of these results is difficult since both 1H and
13C-NMR spectroscopy show broad aromatic signals indicating
that a second fast equilibrium between the remaining MeOBCat
4 and the free catechol is probably taking place according to
eqn (4) (Scheme 2). This equilibrium generated the boric ester 6
known as the Meulenhoff’s free acid.13 This could also explain
the slight upfield shift observed for the MeOBCat 4 signal in
11B-NMR. Neglecting the formation of 6, the equilibrium
constant for the transesterification of MeOBCat 4 with MeOH
[eqn (3)] can be estimated from 11B-NMR integration to be
K3 = 14 Mꢁ1. To avoid complications resulting from the
reaction of catechol with MeOBCat 4 depicted in eqn (4)
(Scheme 2), a similar methanolysis study was performed with
B-n-propylcatecholborane 3b (PrBCat).14 The 11B-NMR
spectrum in benzene gives a single signal at 35.8 ppm. Addition
of methanol gives rise to a second signal at 31.9 ppm that was
All our experimental results (vide supra) indicate that
boronate complexes such as A cannot be the source of
hydrogen atoms for the reaction depicted in eqn (1). There-
fore, free catechol remains as the most plausible source of
hydrogen atoms for this reaction.16 In order to test this
hypothesis, pure B-isopinocampheylcatecholborane 3a was
treated with catechol (1 equiv.) using air as an initiator
(Scheme 4, eqn (6)). Under these conditions, the pinane 2a
was produced in 80% yield. As expected, no reaction takes
place in the absence of air and no reduction is observed in the
absence of catechol. The efficiency of the chain process is very
remarkable since a substoichiometric amount of oxygen
(0.3 equivalent) is sufficient to drive the reaction to comple-
tion. The reaction depicted in eqn (6) (Scheme 4) confirms that
free catechol is the reducing species.17 It is able to deliver a
hydrogen atom to the alkyl radical (Scheme 4, eqn (7)). This
process is followed by a rapid and efficient reaction of the
2-hydroxyphenoxyl radical with the organoborane 3 that
propagates the chain process by regenerating the initial alkyl
radical and the Meulenhoff’s free acid 6 (Scheme 4, eqn (8)).
The Meulenhoff’s free acid 6 is probably involved in the
hydrogen transfer step since reaction in the presence of
0.5 equivalent of catechol affords more than 50% yield of
reduced product. Since 6 has been described as a tricoordi-
nated planar boron species in the solid state,13 its hydrogen
donor ability is not expected to exceed that of catechol. It is
also reported that two molecules of 6 can liberate free catechol
upon formation of B2(C6H4O2)3.13 This mechanism is further
supported by the study of a 1 : 1 mixture of B-propylcatechol-
borane 3b and catechol by 11B-NMR. Under an argon
atmosphere, only one signal at 35.8 ppm corresponding to
3b is observed. As soon as a substoichiometric amount of
oxygen is allowed to enter the system, a second signal at 23.2 ppm
corresponding to the Meulenhoff’s free acid 6 appears.18
The reduction of B-alkylcatecholborane with methanol was
shown to be efficient when the boronate ester was prepared
in situ from an alkene such as a-pinene 1 by using an excess
(2 equiv.) of catecholborane and when 4 equivalents of
methanol were used. The use of a larger excess of methanol
led to a decreased yield. This is best explained by the fact that
under these conditions the excess of catecholborane was
converted first to MeOBCat and then very rapidly converted
to catechol according to eqn (3). The excess of one equivalent
of catecholborane used in the hydroboration process consumes
1
attributed to PrB(OMe)2 7b. H and 13C-NMR spectra gives
further evidence for the partial methanolysis of PrBCat 3b as
the spectrum of 3b/MeOH mixture represents the superposition
of those of 3b and 7b (see supporting information for copies of
spectraw). The absence of a significant modification of chemical
shifts in 11B-NMR (Dd r 0.07 ppm), 13C-NMR (Dd r
0.03 ppm), and 1H-NMR (Dd r 0.01 ppm) between pure
PrBCat 3b and 3b in the presence of variable amounts of
methanol indicates that no significant amount of a complex is
formed. This result is in accordance with the study of boronic
esters transesterification by Brown who reported that no tetra-
coordinated intermediate could be detected by 11B-NMR.15 The
equilibrium constant for methanolysis of PrBCat 3b according
to eqn (5) (Scheme 3) can estimated to be K5 = 0.3 Mꢁ1. This
value indicates clearly that B-alkylcatecholborane 3 does not
undergo methanolysis to the same extend than MeOBCat 4
(compare eqn (3) and (5)). When tert-butanol was used instead
of MeOH in the reduction process depicted in eqn (1)
(Scheme 1), the reduced product was obtained in very low
yield.7 Interestingly, 1H-NMR spectroscopy of a mixture of
PrBCat and tert-butanol shows no transesterification, probably
due to the bulkiness of the tert-butyl group.
ꢀc
This journal is The Royal Society of Chemistry 2010
804 | Chem. Commun., 2010, 46, 803–805

View Article Online
3 (a) H. C. Brown and K. Murray, J. Am. Chem. Soc., 1959, 81,
4108; (b) H. C. Brown and K. Murray, J. Org. Chem., 1961, 26,
631; (c) H. C. Brown and K. Murray, Tetrahedron, 1986, 42,
5497.
4 (a) P. R. Jones and T. F. O. Lim, J. Organomet. Chem., 1976, 120,
27; (b) L. Vasilev, V. V. Veselovskii and B. Mikhailov, Bull. Acad.
Sci. USSR, Div. Chem. Sci. (Engl. Transl.), 1977, 26, 1031.
5 (a) R. Koster, Angew. Chem., 1956, 68, 383; (b) F. L. Ramp,
¨
E. J. DeWitt and L. E. Trapasso, J. Org. Chem., 1962, 27,
4368.
6 (a) B. M. Mikhailov and Y. N. Bubnov, Zh. Obs. Khim., 1961, 31,
160; (b) A. G. Davies and B. P. Roberts, J. Chem. Soc. B, 1971,
1830.
7 (a) D. Pozzi, E. Scanlan and P. Renaud, J. Am. Chem. Soc., 2005,
127, 14204; (b) D. Pozzi and P. Renaud, Chimia, 2007, 61, 151.
8 D. A. Spiegel, K. B. Wiberg, L. N. Schacherer, M. R. Medeiros
and J. L. Wood, J. Am. Chem. Soc., 2005, 127, 12513.
9 Theoretical calculations: W. Tantawya and H. Zipse, Eur. J. Org.
Chem., 2007, 5817.
10 For kinetic and mechanistic investigations, see: (a) J. Jin and
M. Newcomb, J. Org. Chem., 2007, 72, 5098; (b) J. Jin
and M. Newcomb, J. Org. Chem., 2008, 73, 4740; (c) J. Boivin
and V. T. Nguyen, Beilstein J. Org. Chem., 2007, 3, 45;
(d) F. Allais, J. Boivin and V. T. Nguyen, Beilstein J. Org. Chem.,
2007, 3, 46.
11 J. M. Cuerva, A. C. Campana, J. Justicia, A. Rosales, J. L. Oller-
Lopez, R. Robles, D. G. Cardenas, E. Bunuel and J. E. Oltra,
Angew. Chem., Int. Ed., 2006, 45, 5522.
12 P. J. Domaille, J. D. Druliner, L. W. Gosser, J. M. Read,
E. R. Schmelzer and W. R. Stevens, J. Org. Chem., 1985, 50, 189.
13 A. Lang, J. Knizek, H. Noth, S. Schur and M. Thomann,
¨
Scheme 4 Reduction of B-alkylcatecholborane by catechol.
2.7 equivalents of methanol as 85% of the resulting MeOBcat is
methanolysed (estimated from 11B-NMR integration of the
reaction mixture). The remaining methanol (1.3 equiv.) is
involved in the methanolysis of the B-alkylcatecholborane 3
according to eqn (5). Since it is an equilibrated reaction, the
concentration of the B-alkylcatecholborane 3 remains sufficient
to be involved in a radical chain process as only 5% of the
PrBCat is methanolysed under these conditions. It is important
to note that the B-alkyldimethoxyborane 7 is not an efficient
source of radicals. The use of larger amount of methanol favors
the formation of the unreactive species 7 over 3 and therefore
makes the reduction process inefficient (the reduction process
gave a very low yield in pure methanol). When pure boronate 3
is used, the reaction is much less efficient (see Scheme 1, eqn (2)).
Indeed 11B-NMR analysis of the reaction mixture shows 65% of
methanolysis. Therefore both the concentration of catechol and
the concentration of the active organoborane 3 are reduced
relative to the conditions where the organoboranes was prepared
in situ with an excess of catecholborane.
Z. Anorg. Allg. Chem., 1997, 623, 901.
14 In the case of B-n-propylcatecholborane, the samples should be
prepared under strict exclusion of oxygen (glove box, degassed
solvent). Indeed, traces of oxygen were sufficient to initiate a
chain process leading within 2 h to a significant amount of
propane.
15 H. C. Brown and C. D. Roy, J. Organomet. Chem., 2007, 692, 784.
16 In 1H-NMR, the aromatic signals did not fit the expected signals of
catechol in deuterated benzene. Investigation of the 1H-NMR
spectra of catechol solutions in C6D6 with varying amounts of
methanol demonstrated cleanly that the changes of the aromatic
signals were caused by H-bonding interactions between catechol
and methanol (see supplementary information for detailsw).
B. G. Somers and H. S. Gutowsky, J. Am. Chem. Soc., 1963, 85,
3065.
17 Deuterium incorporationexperiments with catechol-d2 confirmed
that catechol was the source of hydrogen atoms. However, due to
well-documented exchange reactions between the O–D and the
aromatic C–H bonds, the deuteration was not quantitative.
(a) K. L. Tuck and H.-W. Tan, J. Labelled Compd. Radiopharm.,
2000, 43, 817; (b) R. O. Day, N. Hadipour and J. L. Ragle,
J. Magn. Res., 1984, 57, 369.
18 The signal at 23.2 ppm corresponds to 6 or to a mixture of 6 and
B2(C6H4O2)3: S. W. Hadebe and R. S. Robinson, Eur. J. Chem.,
2006, 4898.
19 X. Han, T. Shen and H. Lou, Int. J. Mol. Sci., 2007, 8, 950.
20 (a) C. Evans, B. P. Roberts and K. U. Ingold, J. Am. Chem. Soc.,
1992, 114, 4589; (b) P. Franchi, M. Lucarini, G. F. Pedulli,
L. Valgimigli and B. Lunelli, J. Am. Chem. Soc., 1999, 121, 507;
(c) G. Litwinienko and K. U. Ingold, Acc. Chem. Res., 2007, 40,
222; (d) I. Tejero, N. Gonzalez-Garcia, A. Gonzalez-Lafont and
J. M. Lluch, J. Am. Chem. Soc., 2007, 129, 5846; (e) R. Amorati,
S. Menichetti, E. Mileo, G. F. Pedulli and C. Viglianisi,
Chem.–Eur. J., 2009, 15, 4402.
21 (a) M. Kharasch, F. Kawahara and W. Nudenberg, J. Org. Chem.,
1954, 19, 1977; (b) M. Ye and R. H. Schuler, J. Phys. Chem., 1989,
93, 1898.
22 B-alkylcatecholboranes are known to undergo fast substitution
reactions with a wide range of oxygen centered radicals:
In conclusion, based on NMR studies and experimental results,
we have revised the mechanism for the reduction of B-alkyl-
catecholboranes by methanol. We disclose here that catechol, a
chain breaking antioxidant,19–21 is acting as a source of hydrogen
atom in an efficient radical chain process. This unexpected results
is made possible by the fact that aryloxyl radicals undergo a rapid
homolytic substitution (SH2) at the boron atom that prevents the
recombination reactions of the aryloxyl radicals.22,23 These results
uncover new opportunities in the quest for reagents to replace tin
hydride for radical reduction. Further application of this concept
for synthetic purpose is currently under investigation and will be
reported in due course.
We thank the Swiss National Science Foundation for
financial support and BASF Corporation for donation of
chemicals. L. F. and C. H. S. thank the Australian Research
Council through the Centers of Excellence Program for
financial support. We thank Prof. P. Bigler and his team for
help in NMR measurements.
Notes and references
1 (a) H. C. Brown, Organic Synthesis via Boranes, New York, 1975;
(b) H. C. Brown, in Nobel Lecture, 1979.
2 (a) U. Wietelmann, Janssen Chim. Acta, 1992, 10, 16;
(b) G. W. Kabalka, Org. Prep. Proced. Int., 1977, 9, 131;
(c) K. Burgess and M. J. Ohlmeyer, Chem. Rev., 1991, 91, 1179.
(a) A. G. Davies, Pure Appl. Chem., 1974, 39, 497;
(b) A. G. Davies and B. P. Roberts, Acc. Chem. Res., 1972, 5, 387.
23 (a) V. Darmency and P. Renaud, Top. Curr. Chem., 2006, 263, 71;
(b) A. Schaffner and P. Renaud, Eur. J. Org. Chem., 2004, 2291.
ꢀc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 803–805 | 805