Deoxygenation of alcohols employing water as the hydrogen atom source

Article abstract of DOI:10.1021/ja052185l

Trialkylboranes (BMe₃, BEt₃, and BBu₃) have been shown to mediate reductive deoxygenation reactions of O-alkyl-S-methyl dithiocarbonates (methyl xanthates) in which water or deuterium oxide functions as the source of hydrogen or deuterium. This method has proven versatile with regard to substrate scope and is capable of providing protio- or deuterioalkane products in high yields with excellent levels of D-incorporation. Ab initio calculations suggest that the trialkylborane-water complex possesses an unusually low O-H bond dissociation energy (73 kcal/mol) and support a radical chain mechanism for this process. Taken together, this report provides evidence for fundamentally novel and previously overlooked modes of reactivity for water and trialkylboranes of wide ranging importance in both theoretical and applied investigations. Copyright

Full text of DOI:10.1021/ja052185l

Published on Web 08/18/2005
Deoxygenation of Alcohols Employing Water as the Hydrogen Atom Source
David A. Spiegel, Kenneth B. Wiberg, Laura N. Schacherer, Matthew R. Medeiros, and
John L. Wood*
Sterling Chemistry Laboratory, Department of Chemistry, Yale UniVersity, New HaVen, Connecticut 06520-8107
Received April 5, 2005; E-mail: john.wood@yale.edu
Since its initial disclosure in 1975,1 the Barton-McCombie
reaction has proven to be widely useful in the deoxygenation of
alcohols.2 This process has been shown to involve the radical
decomposition of an O-alkyl thiocarbonate (xanthate) ester to the
corresponding alkyl radical, followed by hydrogen atom transfer
from an organotin hydride to provide an alkane product.3-8
A
substantial drawback associated with this method is that organotin
hydrides are costly, often difficult to separate from desired reaction
products, and highly toxic.9-11 Therefore, a variety of methods have
been reported for replacing stoichiometric tin in deoxygenation
processes.12 In the course of an ongoing effort directed toward a
synthesis of the phomoidrides (1-4, Figure 1), we discovered that
water could replace toxic metal hydrides as the source of hydrogen
in trialkylborane-mediated variants of the Barton-McCombie
process. Herein, we describe the events surrounding this discovery
and provide insight into the factors underlying this remarkable and
previously overlooked mode of reactivity for water.
Figure 1. The phomoidrides.
Scheme 1
During the course of our synthetic studies, we prepared complex
isotwistane 5 (Scheme 1) in anticipation of accessing olefin 7
through the intermediacy of radical 6. Initial attempts to effect
fragmentation employing traditional Barton-McCombie conditions
resulted predominantly in the formation of 8,13 presumably owing
to the rapid rate of H-transfer from the metal hydride species.1 In
an effort to avoid rapid reduction, we sought alternative conditions
for the generation of 6 and became intrigued by an isolated example
from Barton and co-workers wherein the treatment of a secondary
xanthate ester with triethylborane/air appeared to furnish products
arising from the slow reduction of an intermediate alkyl radical.3,14
Although this study did not address the origin of the H-atom in
reduction products, we reasoned that BEt3 might have been involved
either through a disproportionation pathway involving ethyl radi-
cals15 or via the homolysis of the R-boryl C-H bond.16 Since our
goal was to impede radical reduction altogether and produce olefin
7, we reasoned that using BMe3 in place of BEt3 might accomplish
the desired result. Thus, we explored the ability of BMe3 to promote
the fragmentation of 5.
Table 1. Deuterium-Labeling Studies18
Astonishingly, under these conditions, 5 undergoes conversion
to 8 in quantitative yield and high purity following simple solvent
evaporation. Intrigued by this discovery, we decided to explore its
generality by applying these tin-free, workup-free conditions to a
series of simple xanthate test substrates. In these initial studies,
the corresponding deoxygenated products were formed in irrepro-
ducible yields that were uniformly lower than those reported for
the metal-hydride-mediated protocols (data not shown).17 Clearly,
optimization of the process was required; it therefore became critical
to establish the hydrogen atom source.
a For all entries, xanthate 1 (3 mg, 0.006 mmol, 1.0 equiv) was dissolved
in benzene (0.2 mL) with or without additive (20 equiv), and trialkylborane
(20-50 equiv) was bubbled through reaction mixtures until TLC analysis
indicated completion. b Values in entries 1-3 were obtained from integration
of 1H NMR spectra, and those in entries 4-7 were calculated from isotopic
ratios in electrospray ionization mass spectra.
replaced with their deuterated variants (entry 4), only a minimally
deuterated sample of 8 was isolated.
Since the above reductions were conducted on relatively small
scales, we reasoned that adventitious water could have been present
as a contaminant. However, since the BDE of an O-H bond in
water is greater even than that of a C-H bond in the reaction solvent
benzene (117.6 versus 113.5 kcal/mol),20 we considered water an
unlikely H-atom source. Nevertheless, we addressed the issue by
exploring the deoxygenation of 3 using BMe3-d9 in the presence
To this end, the deuterium-labeling studies listed in Table 1 were
initiated. Initially, each hydrogen-containing reagent was substituted
1
with its deuterated counterpart (entries 1-3). However, H NMR
analysis of 8 revealed no detectable levels of D-incorporation.19
Unexpectedly, when all H-containing reagents were simultaneously
9
10.1021/ja052185l CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 12513-12515
12513

C O M M U N I C A T I O N S
Table 2. Trialkylborane/H2O/D2O-Mediated Xanthate Reductions
Table 3. Generality of Reduction Reactionsa
a Reactions conducted as in Table 2, except for reaction scale (30-145
mg, 0.06-0.53 mmol in substrate), and the indicated trialkylborane was
added in place of BMe3. b See Supporting Information for experimental
details. c With 5.0 equiv of D2O added. d Yields determined by GC analysis.
Scheme 2. Results of Ab Initio Calculations22
genation products were also obtained in good yields with high levels
of deuterium incorporation. In general, however, these reductions
were not found to be as efficient as those with BMe3. These results
indicated that the observed patterns of reactivity were general, at
least for simple primary trialkylboranes. To rationalize the unex-
pected behavior of water in these reactions,23 we considered the
known Lewis acidity of trialkylboranes24 and postulated that
complexation of water to the trialkylborane might furnish a species
activated for O-H bond homolysis. We therefore turned to the
Gaussian 3 (G3) model chemistry developed by Pople et al.22 to
determine the enthalpies of formation for several relevant inter-
mediates. Calculation of reaction enthalpies using these data
indicated a dramatic decrease in the O-H BDE for the Me3B•OH2
complex (endothermic by 86 kcal/mol compared to 116 kcal/mol
for uncomplexed water derived through the same method).19
Importantly, these results indicated that the derived Cs symmetric
radical (17) was a transition structure that upon geometry optimiza-
tion underwent dissociation into methyl radical (19) and Me2BOH
(18). The latter process was determined to be exothermic by 13
kcal/mol. Overall, the computations indicated that homolysis of the
O-H bond in 16 to provide 18 and 19 was endothermic by 73
kcal/molsa value 43 kcal/mol lower than the BDE predicted for
the O-H bond in uncomplexed watersand slightly lower than the
Sn-H bond in tributyltin hydride.25 Interestingly, analogous
calculations performed on the Me3Al•OH2 complex (not shown)19
yielded an O-H BDE of 116 kcal/mol, implying that the observed
trend may be unique to trialkylborane-water systems.
On the basis of these calculations, along with experimental
observations,26 the mechanism in Scheme 3 is proposed. Here, the
process begins with air oxidation of the trialkylborane (20) to
liberate an alkyl radical (21). This reacts with the xanthate (22) to
provide intermediate 23, which rapidly decomposes to S-alkyl-S-
methyl dithiocarbonate (24) and substrate-derived alkyl radical
25.2-8 This species is then reduced by R3B•OH2 (26) to provide
the observed product (28), dialkyl borinic acid 27, and an equivalent
of alkyl radical 21 capable of propagating the radical chain. While
this mechanism is consistent with the results of experiments and
calculations, alternative pathways, including the intermediacy of
the hydroperoxy radical (HOO•)27 and/or thiol-mediated proton
transfer, could not be ruled out. However, with regard to the latter,
the observation of deuterated butane production upon exposure of
tributylborane/water/benzene mixtures to air suggests that thiol-
a Unless otherwise indicated, substrates (90-106 mg, 0.19-0.39 mmol,
1.0 equiv) were dissolved in benzene (0.03 M) in the presence of either
H2O (5.0 equiv) or D2O (20 equiv) as indicated. Air (0.8 equiv of O2) was
then added at a rate of 1.3 mL/h. b See Supporting Information, Scheme 1,
and/or Table 1 for experimental details. c Yield determined by gas
chromatographic (GC) analysis.
of D2O (entry 5) and were surprised to observe very high deuterium
incorporation in the product (6). Taken together with control
experiments conducted in parallel (entries 6 and 7), these data
clearly indicated that H2O or D2O was serving as the source of
hydrogen or deuterium.
This observation enabled the development of optimized reaction
conditions, which were surveyed on the substrates listed in Table
2. As shown, introduction of trimethylborane gas into a solution
of water or deuterium oxide in benzene, followed by the addition
of air, converted the methyl xanthate esters of tertiary (entries 1
and 2), secondary (entries 3-12), and primary (entries 13 and 14)
alcohols into the corresponding protio- and deuterioalkanes in good
to excellent yields. These compared favorably with metal-hydride-
based protocols, and in some cases, proved better than existing
methods (e.g., entries 1 and 9).13,21 Conveniently, the formation of
C-H versus C-D bonds depended solely upon whether H2O or
D2O was added to reaction mixtures. Furthermore, as with the
examples depicted in Table 1, reactions employing BMe3 did not
require aqueous workup; products were obtained in high purity
following simple low-pressure removal of volatile byproducts.19
When BMe3 was replaced by BEt3 or BBu3 (Table 3), deoxy-
9
12514 J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005

C O M M U N I C A T I O N S
(7) Crich, D. Tetrahedron Lett. 1988, 29, 5805-5806.
Scheme 3. Proposed Mechanism for Xanthate Reductions
(8) Bachi, M. D.; Bosch, E. J. Chem. Soc., Perkin Trans. 1 1988, 6, 1517-
1519.
(9) Appel, K. E. Drug. Metab. ReV. 2004, 36, 763-786.
(10) Boyer, I. J. Toxicology 1989, 55, 253-298.
(11) Dopp, E.; Hartmann, L. M.; Florea, A. M.; Rettenmeier, A. W.; Hirner,
A. V. Crit. ReV. Toxicol. 2004, 34, 301-333.
(12) For reviews, see: (a) Baguley, P. A.; Walton, J. C. Angew. Chem., Int.
Ed. 1998, 37, 3073-3082. (b) Studer, A.; Amrein, S. Synthesis 2002, 7,
835-849. (c) Gilbert, B. C.; Parsons, A. F. J. Chem. Soc., Perkin Trans.
2 2002, 2, 367-387. For lead references on stoichiometric tin hydride
substitutes, see: for silanes (a) Chatgilialoglu, C. Acc. Chem. Res. 1992,
25, 188. For phosphites: (b) Barton, D. H. R.; Jang, D. O.; Jaszberenyi,
J. C. J. Org. Chem. 1993, 58, 6838-6842. For functionalized cyclohexa-
dienes: (c) Studer, A.; Amrein, S.; Schleth, F.; Schulte, T.; Walton, J. C.
J. Am. Chem. Soc. 2003, 125, 5726-5733 and references contained therein.
For phosphine-boranes: (d) Barton, D. H. R.; Jacob, M. Tetrahedron
Lett. 1998, 39, 1331-1334. For cyclohexane: (e) Quiclet-Sire, B.; Zard,
S. Z. J. Am. Chem. Soc. 1996, 118, 9190-9191. For catalytic tin: (f)
Lopez, R. M.; Hays, D. S.; Fu, G. C. J. Am. Chem. Soc. 1997, 119, 6949-
6950.
(13) Reduction of 5, as described in ref 1, provided 8 in 78% yield.
(14) The reduction product was isolated in 62% yield versus 90% in reactions
containing tributyltin hydride (ref 3). Further evidence for slow reduction
under tin-free conditions was the isolation of products from radical
recombination (9%) and disproportionation (12%).
(15) Kochi, J. K. Free Radicals; Wiley: New York, 1973.
(16) Walton, J. C.; McCarrroll, A. J.; Chen, Q.; Carboni, B.; Nziengui, R. J.
Am. Chem. Soc. 2000, 122, 5455.
mediated hydrogen transfer is not a necessary step in the reaction
mechanism.17
In summary, this report provides evidence that water-trialkyl-
borane solvent mixtures, in the absence of metal hydrides, are
capable of mediating the deoxygenation of xanthate esters, perhaps
through the intermediacy of trialkylborane-water complexes. These
results not only represent a potentially useful alternative to available
methods for replacing alcohols with hydrogen, deuterium, or
potentially tritium but also suggest more broadly the existence of
a novel mode of reactivity for water. In this regard, the analogy of
trialkylborane-water complexes to metal-water complexes seen
in biological systems (e.g., photosystem II)28-31 is particularly
intriguing. Further investigations into the mechanistic details of this
deoxygenation and other uses of triakylborane aqua complexes are
currently in progress.
(17) The full details of these studies will be disclosed at a later date.
(18) Since compound 8 obtained upon BMe3 treatment followed by solvent
1
evaporation was pure by H and 13C NMR analysis, and neither workup
or flask transfer was necessary for product purification, the possibility
that a C-H bond in 5 or 8 was providing the H-atom in these reactions
was discounted.
(19) For full details of these studies, please refer to the Supporting Information.
(20) For the BDE of water (H-OH), see: (a) Ruscic, B.; Wagner, A. F.;
Harding, L. B.; Asher, R. L.; Feller, D.; Dixon, D. A.; Peterson, K. A.;
Song, Y.; Qian, X.; Ng, C.-Y.; Liu, J.; Chen, W.; Schwenke, D. W. J.
Phys. Chem. A 2002, 106, 2727. For the BDE of benzene (H-C6H5),
see: (b) Davico, G. E.; Bierbaum, V. M.; DePuy, C. H.; Ellison, G. B.;
Squires, R. R. J. Am. Chem. Soc. 1995, 117, 2590.
(21) To the best of our knowledge, the highest yield that has been reported for
the deoxygenation of 12 is 86% (Togo, H.; Matsubayashi, S.; Yamazaki,
O.; Yokoyama, M., J. Org. Chem. 2000, 65, 2816-2819).
(22) Curtiss, L. A.; Raghavachari, K.; Reforn, P. C.; Rassolov, V.; Pople, J.
A. J. Chem. Phys. 1998, 109, 7764.
Acknowledgment. The authors wish to thank Professors Glenn
C. Micalizio, Jerome A. Berson, David A. Evans, and John Hartwig
for helpful discussions, as well as Dr. Joshua Lawrence for helpful
discussions and use of reagents. We also acknowledge Dr. TuKiet
Lam (W. M. Keck Foundation, Biotechnology Resource Laboratory,
New Haven, CT) for assistance in obtaining FTICR mass spectra.
D.A.S. thanks the Yale Medical Scientist Training Program and
the NIH Cancer Education Program for funding. J.L.W. acknowl-
edges the NIH (Grant No. 1 RO1 CA/GM 93591-01A), Bristol-
Myers-Squibb, Eli Lilly, GlaxoSmithKline, Yamanouchi, and
AstraZeneca for financial support through their Faculty Awards
Programs, and the Henry Dreyfus Foundation for a Teacher-
Scholar Award.
(23) Indeed, many reactions reported to involve free-radical intermediates have
been conducted in water-BEt3-air mixtures. Substantial kinetic and/or
mechanistic differences between these methods and the reductions reported
herein are likely operating. For more detail on aqueous radical reactions,
see: Yorimitsu, H.; Shinokubo, H.; Oshima, K. Synlett 2002, 5, 674-
686.
(24) Larson, J. W.; McMahon, T. B. J. Am. Chem. Soc. 1985, 107, 766.
(25) The BDE for Bu3Sn-H bond has been reported at 78 kcal/mol (Laarhoven,
L. J. J.; Mulder, P.; Wayner, D. D. M. Acc. Chem. Res. 1999, 32, 342-
349).
(26) Many of the examples in Table 2 required only substoichiometric oxygen,
implying that a chain mechanism is operating. However, the uncontrolled
entry of additional oxygen through rubber septa cannot be excluded at
this time. Furthermore, the slow introduction of air was critical to achieving
high yields in these reactions, as observed in other trialkylborane-mediated
free-radical processes. See, for example: (a) Brown, H. C.; Kabalka, G.
W. J. Am. Chem. Soc. 1970, 92, 714-716. (b) Yoshimitsu, T.; Arano,
Y.; Nagaoka, H. J. Org. Chem. 2003, 68, 625-627. (c) Yoshimitsu, T.;
Tsunoda, M.; Nagaoka, H. Chem. Commun. 1999, 17, 1745-1746. The
further observation that the reaction of xanthate 11 was completely
inhibited by galvinoxyl suggests that the initial generation of alkyl radicals
from trialkylboranes (Onak, T. Organoborane Chemistry; Academic
Press: New York, 1975; p 360) is critical to the mechanism. The isolation
of 24 (R′ ) Bu) in 97% yield also supports the proposed mechanism.
(27) If formed upon interaction of water with the dialkylborane peroxy radical
(R2BOO•), this species, based upon its BDE (47.6 kcal/mol), would be
capable of delivering hydrogen atom. See: Flowers, B. A.; Szalay, P. G.;
Stanton, J. F.; Ka´llay; Gausss, J.; Csa´sza´r, A. G. J. Phys. Chem. A 2004,
108, 3195.
Supporting Information Available: Experimental details and
compound characterization (13 pages, print/PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
References
(1) Barton, D. H. R.; McCombie, S. W., J. Chem. Soc., Perkin Trans. 1 1975,
16, 1574-1585.
(2) For reviews, see: (a) Crich, D.; Quintero, L. Chem. ReV. 1989, 89, 1413-
1432. (b) Hartwig, W. Tetrahedron 1983, 39, 2609-2645.
(3) Barton, D. H. R.; Jang, D. O.; Jaszberenyi, J. C. Tetrahedron Lett. 1990,
31, 3991-3994.
(28) For a review, see: Tomos, C.; Babcock, G. T. Acc. Chem. Res. 1998, 31,
18-25.
(29) Hoganson, C. W.; Babcock, G. T. Science 1997, 277, 1953-1956.
(30) McEvoy, J. P.; Brudvig, G. W. Phys. Chem. Chem. Phys. 2004, 6, 4754-
4763.
(4) Barton, D. H. R.; Crich, D.; Lobberding, A.; Zard, S. Z. Tetrahedron
1986, 42, 2329-2338.
(5) Barton, D. H. R.; Crich, D.; Lobberding, A.; Zard, S. Z. Chem. Commun.
1985, 10, 646-647.
(31) For a review, see: Cukier, R. I. Biochem. Biophys. Acta 2004, 1655,
37-44.
(6) Bachi, M. D.; Bosch, E.; Denenmark, D.; Girsh, D. J. Org. Chem. 1992,
57, 6803-6810.
JA052185L
9
J. AM. CHEM. SOC. VOL. 127, NO. 36, 2005 12515