Exploiting boron-zinc transmetallation for the arylation of benzyl halides: What are the reactive species?

Article abstract of DOI:10.1002/anie.201202219

One step Beyond: The transmetallation reactions of ArB(OH)₂ and Ar₃B₃O₃ with Et₂Zn are far more complicated than previously supposed, with solvent-dependent equilibria between ArB(OY)₂ at one side and [RZn(solv)₃][BR₄] at the other (see picture). While the role of the highly reactive organozinc cation has not been implicated before, its importance for the activation of an otherwise sluggish class of electrophiles is shown. Copyright

Full text of DOI:10.1002/anie.201202219

Angewandte
Chemie
DOI: 10.1002/anie.201202219
Transmetallation Reactions
Exploiting Boron–Zinc Transmetallation for the Arylation of Benzyl
Halides: What are the Reactive Species?**
Robin B. Bedford,* Nicholas J. Gower, Mairi F. Haddow, Jeremy N. Harvey, Joshua Nunn,
Rukeme A. Okopie, and Rosalind F. Sankey
Following pioneering work by Bolm and co-workers,[1] B–Zn
transmetallation reactions between arylboronic acids and
diethylzinc have been exploited in a range of asymmetric
nucleophilic addition reactions to carbonyl and related
compounds in the presence of catalytic amounts of chiral
ligands and, in some cases, transition-metal catalysts.[2–7]
These reactions usually require a pre-activation step, in
which ArB(OH)2 and Et2Zn are heated in the absence of the
electrophile, and it is presumed that the active species thus
obtained are the mixed alkyl-aryl zinc reagents, EtZnAr, 1.
We have recently examined the catalytic formation of
diarylmethanes from benzyl halides[8,9] and wondered
whether species 1 would be sufficiently reactive to give the
arylation of these substrates in the absence of catalyst.
Gratifyingly, this proved to be the case and Scheme 1
highlights representative examples.[10]
highly surprised to find that under these conditions, very little
of the desired product 3a was produced, instead substantial
ethylation occurred, giving
5
(Scheme 1).[13] Diarylzinc
reagents are more reactive than the mixed species 1, yet
Ph2Zn also performed very poorly. If neither 1 nor Ar2Zn can
account for the reactivity alone, what other species are
present in the B–Zn mixture that might?
Figure 1a–d shows the 11B NMR spectra obtained upon
mixing Et2Zn with PhB(OH)2 in a range of solvents. Clearly,
several boron-containing products are formed in a solvent-
Figure 1. 11B NMR spectra (rt) of the reactions between PhB(OH)2 and
3–3.2 equivalents of Et2Zn at 608C after 30 min under various con-
ditions: a) in toluene, b) in THF, c) in glyme, d) in DMF, and e) in
toluene with PMDETA (1 equivalent per PhB(OH)2).
Scheme 1. Selected examples of the arylation of benzyl halides[10] in
dioxane at 608C for 18 h. [a] 1H NMR spectroscopic yields.
Reactions in dioxane gave the best results, but 1,2-
dimethoxyethane (glyme) could also be used, although the
reaction was slower, whereas the use of toluene gave
competing Friedel–Crafts benzylation of the solvent. Aryl
boroxines 4 could also be exploited as the boron reagent.[3c,11]
The reagent 1a (Ar= Ph) can also be generated in situ by
the reaction of Ph2Zn with Et2Zn.[12] Therefore we were
dependent manner. The spectrum of the reaction mixture
formed in toluene (Figure 1a) shows peaks at d = 85.7, 77.4,
and 71.9 ppm, assigned as BEt3,[14] BEt2Ph,[15] and BEtPh2,
respectively. Triphenylboron has been shown to act as a useful
aryl source in B–Zn transmetallation processes,[7a,16] thus the
observed arylboranes may well act as “reservoirs” for the aryl
nucleophile. A very similar spectrum was obtained when
triphenylboroxine, 4a, was used as the boron source,[10]
demonstrating that the species obtained in the equilibria are
independent of the “ArB(OY)2” precursor.
[*] Prof. Dr. R. B. Bedford, N. J. Gower, Dr. M. F. Haddow,
Prof. Dr. J. N. Harvey, J. Nunn, R. A. Okopie, Dr. R. F. Sankey
School of Chemistry, University of Bristol
The reaction in tetrahydrofuran (THF), another solvent
used in B–Zn transmetallations, gave the spectrum shown in
Figure 1b which shows a broad peak at 28 ppm, consistent
with an RB(OY)2 (R = Et or Ph) species and a smaller peak at
d = 52.3 ppm assigned as an “Et2B(OY)” intermediate, prob-
Cantock’s Close, Bristol, BS8 1TS (USA)
E-mail: r.bedford@bristol.ac.uk
[**] We thank the EPSRC and Syngenta for funding.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201202219.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
ably Et2B(OZnR).[17,18] A particularly intriguing peak was
observed at d = À10.8 ppm, consistent with a tetraorganobo-
rate anion. When the reaction was repeated in glyme, another
solvent exploited in B–Zn transmetallation reactions,[6] a sub-
stantial quantity of ZnO precipitated and the 11B NMR
spectrum of the supernatant (Figure 1c) showed five peaks in
this region, between d = À7.5 and À17.1 ppm, corresponding
to all five possible organoborates: [BPh4]À (trace); [BEtPh3]À;
[BEt2Ph2]À; [BEt3Ph]À (major species) and [BEt4]À, respec-
tively. This necessitates the concomitant formation of a coun-
terion, most likely [RZn(solv)x]+ (R = Et, Ph).[19] Presumably
the glyme acts as a good ligand for the organozinc cation,
shifting the equilibrium entirely in favor of the ion pairs
[RZn(solv)x][BR4], 6. Similarly the use of dimethylforma-
mide (DMF), a solvent known to stabilize alkyl zinc
cations,[20] gave substantial borate peaks (Figure 1d).
Figure 2. Crystal structure of [ZnEt(THF)3][BPh4], 6a. Thermal ellip-
soids set at 30% probability. Disorder of the ethyl and one of the THF
ligands omitted for clarity.
While no borates were detected in toluene solution
(Figure 1a), the ion pairs 6 may still be present at concen-
trations too low to be observed by 11B NMR spectroscopy. If
this is the case then, according to the Curtin–Hammett
principle,[21] they may still be expected to play a significant
role in subsequent reactions. Pentamethyldiethylenetriamine
(PMDETA) should react with the organozinc cation in 6 and
thus shift the equilibrium in favor of triamine coordinated
analogues 7 (Scheme 2).[22] The results in Figure 1e show that
this is indeed the case: on addition of PMDETA the only
boron-containing species in the reaction mixture were borate
anions.
Scheme 3. Proposed mechanism for the formation of 1.[3c]
suggest that this occurs through the equilibria shown in
Scheme 4. Our NMR results suggest that these steps can occur
with both boroxine 4a and PhB(OH)2 as the precursors.
B3LYP-D3 DFT calculations[25] predict that the DrG0 for
formation of PhBEt2 from ZnEt2(THF)2 and either 4a or
PhB(OH)2 is negative. Formation of one plausible intermedi-
Scheme 2. Trapping ion-pairs, 6, with PMDETA.
While 11B NMR spectroscopy can only reveal the pres-
ence of the borate anions of 6, evidence for the existence of
the [RZn(solv)x]+ counterion in solution was provided by
high-resolution electrospray ionization mass spectrometry of
the reaction mixture in DMF, which showed a peak at
239.0733, precisely that expected for [EtZn(DMF)2]+.[23]
Furthermore cooling a THF solution of a 1:2 mixture of
PhB(OH)2 and Et2Zn gave crystals of [EtZn(THF)3][BPh4],
6a, suitable for X-ray crystallography (Figure 2).[10] The
structure obtained is broadly similar to that of the closely
related complex [EtZn(OEt2)3][B(C6F5)4],[24] except that the
smaller anion in 6a allows considerably closer contact
between the ions (Zn···B separation of 6.55 vs. 7.63 ꢀ).
Maseras, Pericꢁs, and co-workers recently probed the
mechanism of the B–Zn transmetallation of phenylboroxine
(or PhB(OZnR)2, 8) with Et2Zn using a combined exper-
imental and computational approach. The results from their
study are summarized in Scheme 3.[3c] While this mechanism
elegantly describes the initial transmetallation step between
PhB(OY)2 and Et2Zn, it needs to be extended to account for
the presence of R2B(OY), R3B, and [RZn(solv)x][BR4]. We
Scheme 4. Extending the equilibria beyond the formation of 1.
ate, PhB(Et)(OZn(Et)(THF)2), is mildly uphill in free energy
terms.
Figure 3 shows the calculated free-energy profile for the
ensuing transmetallation between Me2Zn(OMe2)2 and
PhBMe2 (in these calculations, Me was used instead of Et
for convenience).[25] The transmetallation is modeled as
occurring through an ion pair species 9. The calculated
barriers for formation and reaction of 9 are low, consistent
with the rapid exchange processes that are observed exper-
imentally. Formation of 6b is predicted to be unfavorable in
free energy terms. This appears to be due to the use of
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
species, both of which are present in the complex reaction
mixture, it is not feasible at this stage to delineate the precise
role of the ion pairs 6 in aldehyde arylation.
In conclusion, we have exploited B–Zn transmetallation
as a method for arylating benzyl halides in the absence of
added catalyst. Probing the B–Zn mixtures showed that the
transmetallation reactions between ArB(OH)2 or Ar3B3O3
and Et2Zn give complex solvent- and condition-dependent
equilibria. These mixtures not only contain the boroxines (or
RB(OY)2) suggested previously but also R2B(OY) and BR3
species and the ion pairs [RZn(solv)x][BR4]. Given the
complex mixtures observed, it should not be assumed that
the simple organozinc species 1 are the sole nucleophiles
responsible for the reported activity in B–Zn arylation
reactions.
Figure 3. DFT relative free energies at 298 K for transmetallation in
a simplified model system.[25] Also shown the relative free energy for
forming the representative ion pair 6a.
Received: March 20, 2012
Published online: && &&, &&&&
Keywords: arylation · benzyl halides · boron · transmetallation ·
.
a truncated model: at the same level of theory,[25] the
calculated DrG0 for formation of 6a from BPh3 and EtZnPh-
(THF)2 is just + 0.5 kcalmolÀ1. The B–Zn distance in the
optimized full system 6a was found to be 6.36 ꢀ, and the sum
of the Mulliken charges for the Zn fragment was calculated as
+ 0.85.[26]
zinc
[1] a) C. Bolm, J. Rudolph, J. Am. Chem. Soc. 2002, 124, 14850; b) C.
Bolm, L. Zani, J. Rudolph, I. Schiffers, Synthesis 2004, 2173; c) J.
Rudolph, F. Schmidt, C. Bolm, Synthesis 2005, 840; d) S.
ꢂÅubukÅu, F. Schmidt, C. Bolm, Org. Lett. 2005, 7, 1407; e) F.
Schmidt, J. Rudolph, C. Bolm, Adv. Synth. Catal. 2007, 349, 703.
[2] Reviews: a) F. Schmidt, R. T. Stemmler, J. Rudolph, C. Bolm,
Chem. Soc. Rev. 2006, 35, 454; b) M. W. Paix¼o, A. L. Braga,
D. S. Lꢃdtke, J. Braz. Chem. Soc. 2008, 19, 813.
[3] Examples of additions to aldehydes and ketones: a) J. Ruan, G.
Lu, L. Xu, Y.-M. Li, A. S. C. Chan, Adv. Synth. Catal. 2008, 350,
76; b) M.-C. Wang, Q.-J. Zhang, W.-X. Zhao, X.-D. Wang, X.
Ding, T.-T. Jing, M.-P. Song, J. Org. Chem. 2008, 73, 168; c) C.
Jimeno, S. Sayalero, T. Fjermestad, G. Colet, F. Maseras, M. A.
Pericꢁs, Angew. Chem. 2008, 120, 1114; Angew. Chem. Int. Ed.
2008, 47, 1098; d) M.-J. Jin, S. M. Sarkar, D.-H. Lee, H. Qiu, Org.
Lett. 2008, 10, 1235; e) A. L. Braga, M. W. Paix¼o, B. West-
ermann, P. H. Schneider, L. A. Wessjohann, J. Org. Chem. 2008,
73, 2879; f) A. D. Wouters, G. H. G. Trossini, H. A. Stefani, D. S.
Lꢃdtke, Eur. J. Org. Chem. 2010, 2351; g) R. Infante, J. Nieto, C.
Andrꢄs, Org. Biomol. Chem. 2011, 9, 6691; h) M. Hatano, R.
Gouzu, T. Mizuno, H. Abe, T. Yamada, K. Ishihara, Catal. Sci.
Technol. 2011, 1, 1149.
Having observed both BR2Ph and [BR3Ph]À as potential
alternative nucleophiles in the B–Zn transmetallation pro-
cesses we next investigated whether isolated representative
examples of these nucleophiles would react directly with
benzyl halides and found that the reaction of 2a with BPh3,
Na[BPh]4, Li[BPhEt3], or Li[BPh3Et] again gave little or no
3a. The lack of arylation with either Zn- or B-based
nucleophiles implies that a halophilic species that can activate
the benzyl halide is also necessary and it seems highly likely
that the Lewis acid [RZn(solv)x]+ plays this role (Scheme 5).
Scheme 5. Suggested mechanism for the activation of benzyl halides.
[4] Ti-catalyzed addition to ketones: a) O. Prieto, D. J. Ramꢅn, M.
Yus, Tetrahedron: Asymmetry 2003, 14, 1955; b) V. J. Forrat, O.
Prieto, D. J. Ramꢅn, M. Yus, Chem. Eur. J. 2006, 12, 4431.
[5] Other additions: a) S. Wang, M. B. Onaran, C. T. Seto, Org. Lett.
2010, 12, 2690; b) K. Kobayashi, Y. Kondo, Org. Lett. 2009, 11,
2035.
[6] Ni-catalyzed cross-coupling with propargyl halides: S. W. Smith,
G. C. Fu, J. Am. Chem. Soc. 2008, 130, 12645.
[7] Conjugate addition to enones: a) D. PeÇa, F. Lꢅpez, S. R.
Harutyunyan, A. J. Minnaard, B. L. Feringa, Chem. Commun.
2004, 1836; b) L. Dong, Y.-J. Xu, L.-Z. Gong, A.-Q. Mi, Y.-Z.
Jiang, Synthesis 2004, 1057.
[8] Fe-catalyzed coupling with aryl zinc reagents: R. B. Bedford, M.
Huwe, M. C. Wilkinson, Chem. Commun. 2009, 600.
[9] Fe-catalyzed arylation using tetraorganoborates and co-catalytic
Zn: R. B. Bedford, M. A. Hall, G. R. Hodges, M. Huwe, M. C.
Wilkinson, Chem. Commun. 2009, 6430.
Finally we were interested to see whether the results
obtained had significance beyond the arylation of benzyl
halides. The asymmetric arylation of aldehydes, the arche-
typal B–Zn transmetallation reaction, usually exploits cata-
lytic amounts of chiral aminoalcohols, such as 1,1,2-triphenyl-
2-(piperidin-1-yl)ethanol 11,[12c] as ligands. When the reaction
described in Figure 1a was repeated in the presence of
8 mol% of (S)-11, the 11B NMR spectrum obtained was
essentially identical.[10] Adding PMDETA to this reaction
mixture gave a very similar spectrum to that shown in
Figure 1e.[10] This shows that both the ion pairs 6 and BR2Ph
species form even in the presence of aminoalcohols and
therefore that they must be present in most if not all of the B–
Zn transmetallation reactions reported. However, given the
propensity of aldehydes to react with arylzinc and ArBR2
[10] See supporting information for full details.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[11] N. A. Magnus, P. B. Anzeveno, D. S. Coffey, D. A. Hay, M. E.
Laurila, J. M. Schkeryantz, B. W. Shaw, M. A. Staszak, Org.
Process Res. Dev. 2007, 11, 560.
[12] a) C. Bolm, N. Herrmans, J. P. Hildebrand, K. MuÇiz, Angew.
Chem. 2000, 112, 3607; Angew. Chem. Int. Ed. 2000, 39, 3465;
b) J. Rudolph, T. Rasmussen, C. Bolm, P.-O. Norrby, Angew.
Chem. 2003, 115, 3110; Angew. Chem. Int. Ed. 2003, 42, 3002;
c) M. Fontes, X. Verdaguer, L. Solꢁ, Miquel A. Pericꢁs, A. Riera,
J. Org. Chem. 2004, 69, 2532.
[19] This contrasts with Mg– and Li–Zn transmetallation which can
yield Zn-based anions because of the lower electronegativities of
Li and Mg compared with Zn. For instance: a) E. Hevia, J. Z.
Chua, P. Garcꢆa-ꢇlvarez, A. R. Kennedy, M. D. McCall Hevia,
Proc. Natl. Acad. Sci. USA 2010, 107, 5294; b) M. Westerhausen,
Dalton Trans. 2006, 4755.
[20] L. Caggiano, R. F. W. Jackson, A. J. H. M. Meijer, B. T. Pickup,
K. A. Wilkinson, Chem. Eur. J. 2008, 14, 8798.
[21] J. I. Seeman, Chem. Rev. 1983, 83, 83.
[22] H. Tang, H. G. Richey Jr., Organometallics 1996, 15, 4891.
[23] Calculated m/z for C12H19N2O2Zn = 239.0733. See Figure S3 (in
the Supporting Information) for the isotope pattern.
[24] D. A. Walker, T. J. Woodman, D. L. Hughes, M. Bochmann,
Organometallics 2001, 20, 3772.
[13] The Et may be from Et2Zn rather than a mixed species 1a. Et2Zn
alone gave > 98% of 5. See also: K. A. Agrios, M. Srebnik, J.
Org. Chem. 1993, 58, 6908.
[14] d11B for BEt3 = 84 ppm, see: L. H. Toporcer, R. E. Dessy, S. I. E.
Green, Inorg. Chem. 1965, 4, 1649.
[25] The calculations at the B3LYP-D3 level include continuum
[15] J. J. Eisch, B. W. Kotowicz, Eur. J. Inorg. Chem. 1998, 761.
[16] For example: a) J. Rudolph, F. Schmidt, C. Bolm, Adv. Synth.
Catal. 2004, 346, 867; b) J. Rudolph, N. Hermanns, C. Bolm, J.
Org. Chem. 2004, 69, 3997.
solvation. See the Supporting Information for full details.
À
[26] 9: Zn-B = 2.37 ꢀ; Zn-CPh = 2.16 ꢀ; B CPh = 1.73 ꢀ (c.f. 1.57 ꢀ
À
À
À
in the free borane). 10: Zn B = 3.98 ꢀ; Zn CPh = 4.47 ꢀ; B
À
À
À
C
Ph = 1.64 ꢀ. 6b: Zn B = 6.00 ꢀ; Zn CPh = 5.73 ꢀ; B CPh bond
[17] Compare with d11B (THF) Et2B(OEt) = 52.4 ppm.
[18] Large peak at d = 76.3 ppm assigned as BEt3, (d11B pure BEt3 in
THF = 75.7).
length remained similar at 1.65 ꢀ. The sum of the Mulliken
charges calculated for the Zn fragments =+ 0.17, + 0.41, and
+ 0.87 for 9, 10, and 6b, respectively.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Transmetallation Reactions
One step Beyond: The transmetallation
reactions of ArB(OH)2 and Ar3B3O3 with
Et2Zn are far more complicated than
previously supposed, with solvent-de-
pendent equilibria between ArB(OY)2 at
one side and [RZn(solv)3][BR4] at the
other (see picture). While the role of the
highly reactive organozinc cation has not
been implicated before, its importance
for the activation of an otherwise sluggish
class of electrophiles is shown.
R. B. Bedford,* N. J. Gower,
M. F. Haddow, J. N. Harvey, J. Nunn,
R. A. Okopie,
R. F. Sankey
&&&& — &&&&
Exploiting Boron–Zinc Transmetallation
for the Arylation of Benzyl Halides: What
are the Reactive Species?
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!