New insight into the mechanism of the reaction between α,β-unsaturated carbonyl compounds and triethylborane (Brown's reaction)

Article abstract of DOI:10.1016/S0040-4039(99)02279-0

A study of the reaction of α,β-unsaturated carbonyl compounds with triethylborane under free radical conditions (Brown's reaction) including spectroscopic analyses (¹¹B NMR, IR, EPR) of products and intermediates indicated that these reactions involve the prior formation of an 'α,β- unsaturated carbonyl compound-organoborane' complex. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(99)02279-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 1195–1198
New insight into the mechanism of the reaction between
α,β-unsaturated carbonyl compounds and triethylborane
(Brown’s reaction)
Valérie Beraud,a Yves Gnanou,b John C. Walton c and Bernard Maillard a,∗
aLaboratoire de Chimie Organique et Organométallique, associé au CNRS UMR 5802, Université Bordeaux 1,
F-33405 Talence Cedex, France
bLaboratoire de Chimie des Polymères Organiques, UMR 5629, ENSCPB-CNRS, Université Bordeaux I, BP 108,
33402 Talence Cedex, France
cSchool of Chemistry, University of St Andrews, St Andrews, Fife KY16 9ST, UK
Received 8 October 1999; accepted 6 December 1999
Abstract
A study of the reaction of α,β-unsaturated carbonyl compounds with triethylborane under free radical conditions
(Brown’s reaction) including spectroscopic analyses (11B NMR, IR, EPR) of products and intermediates indicated
that these reactions involve the prior formation of an ‘α,β-unsaturated carbonyl compound–organoborane’ com-
plex. © 2000 Elsevier Science Ltd. All rights reserved.
Keywords: boron and compounds; complexation; carbonyl compounds; enones; radicals and radical reactions.
Organoboranes are known to undergo exceptionally fast 1,4-addition to many α,β-unsaturated carbo-
nyl compounds such as methyl vinyl ketone (MVK)1,2 and acrolein.2,3 Subsequent addition of water to
the reaction mixture results in rapid hydrolysis of the intermediate vinyloxyborane (1), with liberation of
the corresponding aldehyde or ketone (2) (Scheme 1). Brown and Kabalka suggested that this addition
occurs via a free radical chain mechanism.4 Recently, Bieber and co-workers reported that alkylation
of 2-methoxy-1,4-quinone, via free radical reaction with trialkylboranes, occurred regioselectively in
the 5-position.5 They invoked the existence of a quinone–borane complex to explain this result. More
recently, Malacria and co-workers demonstrated the efficiency of triethylborane mediated free radical
cyclizations onto the carbonyl of a ketone.6 They also suggested the formation of a complex between the
borane and the ketone to account for the high reactivity of C_O and the observed chemoselectivity in the
cyclization of enal precursors. These recent papers and the review of Renaud and Gerster7 on the use of
Lewis acids in free radical chemistry prompted us to present our results on Brown’s reaction, including
∗
Corresponding author. Tel: 33 5 56 84 64 45; fax: 33 5 56 84 69 94; e-mail: b.maillard@lcoo.u-bordeaux.fr (B. Maillard)
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(99)02279-0
tetl 16207

1196
our spectroscopic observations of the complexes formed between α,β-unsaturated carbonyl compounds
and triethylborane.
Scheme 1. Free-radical alkylation of enones (Brown’s reaction)
In connection with the possible free radical homopolymerization of MVK,8 the fact that Brown did not
detect the presence of any oligomer in the reaction medium is surprising. However, we confirmed that
when freshly distilled MVK9 was stirred for 1 h in dry benzene, at 60°C, under an argon atmosphere, in
the presence of catalytic amounts of AIBN (0.2 mol% of the ketone), a poly(MVK) sample of 104 000
g. mol−1 molar mass was isolated. By contrast, when the same reaction was performed in the presence
of one equivalent of triethylborane,10 with respect to MVK, the conditions being otherwise similar, the
only product formed was hexan-2-one (84%). No polymer or oligomer could be detected by GC after
evaporation of the solvent. Brown’s explanation for this singular result was that the intermediate MVK
adduct-radical would be oxygen centred and hence would react very rapidly with triethylborane.
Applying this hypothesis in connection with the analogous fast addition of alkoxyl radicals to
phosphorous compounds (e.g. the rate constant for t-butoxyl radical addition to triethylphosphite is ca.
2×109 M−1 s−1 11
)
implies that the AIBN initiated reaction of MVK with P(OEt)3 should give some
hexan-2-one. However, irrespective of the conditions used, only poly(MVK) was isolated after removal
of the solvent. This lack of reactivity of the growing polymeric radicals with phosphite seems to indicate
that the specific reactivity of radicals arising from the MVK/BEt3 reaction cannot be merely attributed to
a high electron density on the carbonyl oxygen atom of the propagating species.
Boron has a high affinity for oxygen, as evidenced by the B–O bond dissociation energy in several
molecules (808.8 kJ mol−1 12
and hence has a penchant to react with a variety of Lewis bases forming
)
corresponding complexes.13 It occurred to us therefore that Brown’s reaction might involve a transient
1
‘carbonyl compound–BR3’ complex. After confirming by H NMR that no addition reaction occurred
when MVK and BEt3 were mixed, the formation of a complex was demonstrated by 11B NMR. Recorded
in oxygen free conditions14 the 11B NMR spectrum indicated that the chemical shift of the boron atom
in BEt3 was significantly affected by the presence of MVK (Table 1, entries 1 and 2). Indeed, the single
peak due to the boron atom now appeared at −53.39 ppm as against −86.45 ppm in the absence of ketone
(relative to BF3 etherate). Likewise, IR spectra of MVK–triethylborane equimolar mixtures, performed
in oxygen free conditions15 showed that the νC_O band of the ketone was shifted to higher frequency by
the presence of BEt3 (Table 1).
We also examined whether this complexation is a general phenomenon or is peculiar to MVK.
Acrolein2,3 and 3-methylbut-3-en-2-one16 were also successfully used in the Brown reaction, so their
equimolar mixtures with triethylborane were analysed by IR and 11B NMR. In each case, the νC_O band
was found to be shifted by the presence of BEt3 (Table 1, entries 3 and 4) and the 11B chemical shifts
were in agreement with complex formation (Table 1, entries 1, 3 and 4).
The question that then arose was whether MVK formed similar complexes with alkylborinates. In
fact, Et2BOMe failed to participate in the AIBN-induced free radical reaction with MVK and only
poly(MVK) (Mn of about 2090) was ultimately obtained. Since the chemical shift of the boron atom
in the ‘BEt3–MVK’ complex is of the same order of magnitude as that observed for alkylborinates, the

1197
Table 1
11B NMR and IR data for carbonyl compound–organoborane equimolar mixtures
presumed 11B chemical shift for a ‘Et2BOMe–MVK’ complex should appear around −30 ppm, which is
the chemical shift observed for a two oxygen bonded boron atom.17 It turns out that the 11B NMR of an
equimolar mixture of this alkylborinate and MVK exhibits a single peak at −53.86 ppm (Table 1, entry
6), indicating that Et2BOMe and MVK do not form any complex when mixed together.
The mechanism implied that the complexed propagating radical 3 underwent a dissociative rearran-
gement to afford vinyloxyborane and release an alkyl radical, in preference to intermolecular addition
to another enone molecule (Scheme 2). Evidence for the existence of complexed radicals was sought by
examining the 9 GHz EPR spectra obtained on photolysis of a solution of BEt3 and MVK in di-t-butyl
peroxide. t-Butoxyl radicals produced by photolysis were expected to displace ethyl radicals from BEt3.
These ethyl radicals should then add to MVK to produce complexed adduct radicals 3 (Y=Me). The
EPR spectrum at 270 K displayed a quartet structure with a(3H)=18.4 G which corresponds to radical 3
(Y=Me) with the α-H and two β-H having coincidentally equal hyperfine splitting (hfs). For comparison
the uncomplexed adduct radical was generated by hydrogen abstraction from hexan-2-one also using t-
BuO· radicals. In this case the t-BuO· radicals abstracted unselectively and the spectrum consisted of at
least two radicals one having a quartet structure [a(3H)=19.2 G] which we attribute to the uncomplexed
adduct radical (hexan-2-one-3-yl) and another having a double quintet corresponding to the hexan-2-
one-5-yl radical. To confirm our identification of the uncomplexed hexan-2-one-3-yl radical we used
Roberts’ polarity reversal catalysis method (with Et3NBH3) to selectively abstract hydrogen adjacent to
the carbonyl group:18
Photolysis of hexan-2-one and t-BuOOBu-t with added amine–borane furnished solely the quartet EPR
spectrum of hexan-2-one-3-yl with a(3H)=19.3 G at 270 K. The significantly smaller hfs of the radical
in the presence of BEt3 is entirely consistent with complexation which results in measurable withdrawal
of spin density from the enone moiety.
In conclusion, it appears that Brown’s reaction indeed occurs via an ‘α,β-carbonyl com-
pound–organoborane’ complex. Considering that MVK easily copolymerizes with several monomers, it
appears to be of great interest to take advantage of the ‘MVK–BEt3’ complex as a chain transfer agent in
free radical polymerizations. This possibility is currently under active investigation in our laboratories.

1198
Scheme 2. Mechanism of alkylation via complexed enones
Acknowledgements
We gratefully acknowledge the award of a post-doctoral scholarship by CNRS and AKZO NOBEL
to V.B. and are thankful to Dr. J. Mascetti (University Bordeaux 1) for fruitful discussions about IR
analyses. J.C.W. thanks the EPSRC for financial support (Grant GR/L49185).
References
1. Suzuki, A.; Matsumoto, H.; Itoh, M.; Brown, H. C.; Rogic, M. M.; Rathke, M. W. J. Am. Chem. Soc. 1967, 89, 5708–5709.
2. For a recent application, see: Ollivier, C.; Renaud, P. Chem. Eur. J. 1999, 5, 1468–1473.
3. Brown, H. C.; Rogic, M. M.; Rathke M. W.; Kabalka, G. W. J. Am. Chem. Soc. 1967, 89, 5709–5710.
4. Brown H. C.; Kabalka, G. W. J. Am. Chem. Soc. 1970, 92, 714–716.
5. Bieber, L. W.; Rolim Neto P. J.; Generino, R. M. Tetrahedron Lett. 1999, 40, 4473–4476.
6. Devin, P.; Fensterbank, L.; Malacria, M. Tetrahedron Lett. 1999, 40, 5511–5514.
7. Renaud, P.; Gerster, M. Angew.Chem., Int. Ed. 1998, 37, 2562–2579.
8. Young Min, K.; Chang Sik, H.; Seong Jin, C.; Dong Kyu, P.; Won Jei, C. Pollimo 1992, 16, 563–565.
9. The commercially available methyl vinyl ketone was purified under classical conditions to remove the stabilizers: Perrin, D.
D.; Armarego, W. L. F. In Purification of Laboratory Chemicals, 3rd. edn; Pergamon Press: Oxford, 1988, pp. 232.
10. 0.1 M solution in hexanes.
11. Roberts, B. P.; Scaiano, J. C. J. Chem. Soc., Perkin Trans. 2 1981, 905–911.
12. Lide, D. R. In: CRC Handbook of Chemistry and Physics, 78th edn; CRC Press: New York, 1997.
13. Santelli, M.; Pons, J.-M. In: Lewis Acids Selectivity in Organic Synthesis; CRC Press: Boca Raton, 1995.
14. Solutions of carbonyl compounds in anhydrous benzene were degassed and placed under an argon atmosphere before adding
BEt3.
15. Solutions of carbonyl compounds in anhydrous CCl4 were degassed and placed under an argon atmosphere before adding
BEt3. The IR cells were filled and closed in a glove box.
16. Fenzl, W.; Koester, R.; Zimmermann, H. J. Justus Liebigs Ann. Chem. 1975, 12, 2201–2210.
17. Goetze, R.; Nöth, H.; Pommerening, H.; Sedlack, D.; Wrackmeyer, B. Chem. Ber. 1981, 114, 1884–1893.
18. Paul, V.; Roberts, B. P.; Willis, C. R. J. Chem. Soc., Perkin Trans. 2 1989, 1953–1961.