Divergent reactivity of perfluoropentaphenylborole with alkynes

Article abstract of DOI:10.1021/om100334r

The reactions of perfluoropentaphenylborole (1) with the symmetrical internal alkynes RCCR (R = C₆F₅, C₆H ₅, CH₂CH₃) have been explored. For R = C ₆F₅, a slow reaction

Full text of DOI:10.1021/om100334r

5132 Organometallics 2010, 29, 5132–5139
DOI: 10.1021/om100334r
Divergent Reactivity of Perfluoropentaphenylborole with Alkynes§
Cheng Fan,† Warren E. Piers,*,† Masood Parvez,† and Robert McDonald‡
†Department of Chemistry, University of Calgary, 2500 University Drive N.W., Calgary, Alberta,
Canada, T2N 1N4, and ‡Department of Chemistry, University of Alberta, 11227 Saskatchewan Drive,
Edmonton, Alberta, Canada T6G 2G2
Received April 22, 2010
The reactions of perfluoropentaphenylborole (1) with the symmetrical internal alkynes RCCR
(R = C6F5, C6H5, CH2CH3) have been explored. For R = C6F5, a slow reaction occurs via a
Diels-Alder pathway established for the non-fluorinated system to give the new borane perfluoro-
heptaphenylborepin 2. The new perfluoroarylborane 2 and its pyridine adduct 2-py were char-
acterized crystallographically. In contrast, the reaction of 1 with diphenylacetylene is dominated by a
reaction path in which the borole electrophilically attacks the alkyne, triggering a 1,2 phenyl
migration and borole ring expansion to deliver the boracyclohexadiene 3 as the major product.
Minor products in this reaction were identified as diphenyl-substituted borepins resulting from the
Diels-Alder reaction channel. Finally, 3-hexyne reacts with 1 via the Diels-Alder pathway, but the
kinetic products are not the expected diethyl borepins but rather one of the initial Diels-Alder
adducts formed prior to ring-opening to the seven-membered borepin ring. This intermediate was
fully characterized and provides concrete evidence for the pericyclic reaction sequence proposed by
Eisch for the reaction of pentaphenylborole with diphenylacetylene.
Introduction
have been the subject of investigation for decades14-19 because
of their isoelectronic relationship to hydrocarbons of theore-
tical significance and their relevance to concepts of
aromaticity.20 Thus, five-membered borole rings I are antiaro-
matic 4π rings that are isoelectronic to the fleetingly stable
cyclopentadienyl [C5R5]þ cations, while the neutral, aromatic
6π borepin systems II are related to the tropylium cation.21,22
The aromaticity of II23 contributes to the stability of the
unsubstituted parent,24 but the unsubstituted C4H5B borole I
is not isolable. Typically, more substituted derivatives of both
heterocycles are required to bolster stability. In the case of I,
substituted14,16,25,26 or ring-annulated27-29 derivatives are
Materials containing π-conjugated boron heterocycles at
their core exhibit potentially exploitable photophysical, re-
dox, and anion-binding properties.1-13 However, although
interest in applications has recently arisen, these heterocycles
§ Part of the Dietmar Seyferth Festschrift. This article is dedicated with
much respect and gratitude to Prof. Dietmar Seyferth for his inspiring
scientific and leadership contributions to the organometallic chemistry
community.
*To whom correspondence should be addressed. E-mail:
wpiers@ucalgary.ca.
(1) Yamaguchi, S.; Wakamiya, A. Pure Appl. Chem. 2006, 78, 1413–
1423.
(2) Entwistle, C. D.; Marder, T. B. Chem. Mater. 2004, 16, 4574–
4585.
(16) Eisch, J. J.; Galle, J. E.; Kozima, S. J. Am. Chem. Soc. 1986, 108,
379–385.
(17) Fagan, P. J.; Burns, E. G.; Calabrese, J. C. J. Am. Chem. Soc.
€
(3) Jakle, F. Coord. Chem. Rev. 2006, 250, 1107–1121.
1988, 110, 2979–2981.
(4) Bosdet, M. J. D.; Piers, W. E. Can. J. Chem. 2009, 87, 8–29.
(5) Kim, S.; Song, K.-h.; Kang, S. O.; Ko, J. Chem. Commun. 2004,
68–69.
(18) Ashe, A. J.; Drone, F. J. J. Am. Chem. Soc. 1987, 109, 1879–1880.
(19) Sugihara, Y.; Yagi, T.; Murata, I.; Imamura, A. J. Am. Chem.
Soc. 1992, 114, 1479–1481.
(6) Wakamiya, A.; Mori, K.; Araki, T.; Yamaguchi, S. J. Am. Chem.
Soc. 2009, 131, 10850–10851.
(7) Agou, T.; Kobayashi, J.; Kawashima, T. Chem. Commun. 2007,
3204–3206.
(20) Schleyer, P. v. R. Chem. Rev. 2001, 101, 1115–1118.
(21) Wiberg, K. B. Chem. Rev. 2001, 101, 1317–1332.
(22) Schulman, J. M.; Disch, R. L.; Sabio, M. L. J. Am. Chem. Soc.
1982, 104, 3785–3788.
(8) Agou, T.; Kabayashi, J.; Kawashima, T. Chem. Eur. J. 2007, 13,
8051.
(9) Fukazawa, A.; Yamaguchi, S. Chem. Asian J. 2009, 4, 1386–1400.
€
(10) Hudnall, T. W.; Chiu, C.-W.; Gabbai, F. P. Acc. Chem. Res.
(23) Ashe, A. J.; Klein, W.; Rousseau, R. Organometallics 1993, 12,
3225–3231.
(24) Ashe, A. J.; Kampf, J. W.; Nakadaira, Y.; Pace, J. M. Angew.
Chem., Int. Ed. Engl. 1992, 31, 1255–1258.
2009, 42, 388–397.
(11) Mercier, L. G.; Piers, W. E.; Parvez, M. Angew. Chem., Int. Ed.
2009, 48, 6108–6111.
(25) Braunschweig, H.; Fernandez, I.; Frenking, G.; Kupfer, T.
Angew. Chem., Int. Ed. 2008, 47, 1951–1954.
(26) So, C.-W.; Watanabe, D.; Wakamiya, A.; Yamaguchi, S. Orga-
(12) Wood, T. K.; Piers, W. E.; Keay, B. A.; Parvez, M. Chem.
Commun. 2009, 5147–5149.
(13) Wood, T. K.; Piers, W. E.; Keay, B. A.; Parvez, M. Angew.
Chem., Int. Ed. 2009, 48, 4009–4012.
(14) Eisch, J. J.; Hota, N. K.; Kozima, S. J. Am. Chem. Soc. 1969, 91,
4575–4577.
nometallics 2008, 27, 3496–3501.
€
(27) Koster, R.; Benedikt, G. Angew. Chem., Int. Ed. Engl. 1963, 2,
323–324.
(28) Grigsby, W. J.; Power, P. P. J. Am. Chem. Soc. 1996, 118, 7981–
7988.
(29) Wehmschulte, R. J.; Diaz, A. A.; Khan, M. A. Organometallics
(15) Eisch, J. J.; Galle, J. E. J. Am. Chem. Soc. 1975, 97, 4436–4437.
2002, 22, 83–92.
r
pubs.acs.org/Organometallics
Published on Web 06/18/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 21, 2010 5133
required to prevent Diels-Alder dimerization pathways, while
for II, photochemical reactions that extrude “BR” and sub-
stituted benzenes are a principle reactivity mode30 disfavored
by substitution on the ring.
Chart 1
In terms of Lewis acidity, boroles are typically more Lewis
acidic than their borepin analogues due to their antiaroma-
ticity. In particular, the Lewis acid strength of perfluoroaryl
boranes has been augmented in this way. For example, the
quintessential perfluoroaryl borane B(C6F5)3, a strong, pro-
tically stable organometallic Lewis acid,31-34 is a slightly
weaker Lewis acid than the perfluorinated 9-phenyl-9-bora-
fluorene derivative III,35,36 which has at its core an antiaro-
matic five-membered borole ring. While steric effects and
relief of ring strain may also play a role in this Lewis acid
strength enhancement of III over B(C6F5)3, undoubtedly the
disruption of antiaromaticity is also a significant driver
behind the increased Lewis acidity of III. Thus, both anti-
aromaticity and perfluorination enhance the Lewis acid
strength of boron-containing compounds.37-39
In accordance with this assertion, the recently prepared
perfluorinated pentaphenylborole, 1,40 was found to be a
powerful Lewis acid in comparison to B(C6F5)3 and its
unfluorinated analogue.14,25 For example, in a competition
for the weak Lewis base CH3CN, the adduct of compound 1
was exclusively formed, leaving B(C6F5)3 completely free of
coordination despite its propensity to form a strong adduct
with CH3CN.41 We have thus begun to explore the reactivity
of 1, and given the propensity of the unfluorinated penta-
phenylborole to undergo Diels-Alder reactions with
alkynes,30 herein we report on the reactivity of 1 with three
symmetrically substituted alkynes, RCCR (R = C6F5, C6H5,
CH2CH3), with varying electronic properties. These studies
reveal competing pathways for the reactivity of 1 with alkyne
substrates involving conventional Diels-Alder reactivity or
electrophilic attack by the potently Lewis acidic boron
center.
characterize the reaction of the unfluorinated pentaphenyl-
borepin with diphenylacetylene to form heptaphenylbore-
pin; presumably the perfluorinated reactants engage in an
analogous reaction pathway to give 2, as depicted in
Scheme 1. Thus, a 1,3-suprafacial sigmatropic shift from
the initial Diels-Alder intermediate A yields the unstable,
homoantiaromatic43 7-borabicyclo[4.1.0]heptadiene inter-
mediate B,44,45 which undergoes electrocyclic ring-opening
to yield the observed heptaarylborepin product.
The course of the reaction was followed visually via the
loss of the intensely purple-colored 1 in favor of the white
suspension of 2, in a light yellow liquor, that results upon
complete conversion. The new perfluoroarylborane 2 was
isolated by filtration in 82% yield. Monitoring by NMR
spectroscopy was made difficult as a consequence of the poor
solubility of both 1 and 2 in common solvents even at higher
temperatures; indeed spectroscopic characterization of 2 was
to a large degree precluded by its insolubility. It was un-
ambiguously identified when crystals were harvested from a
1:1 reaction of 1 and C6F5CCC6F5 in CH2Cl2 at 50 °C that
took place over the period of about four weeks with no
stirring.
Spectroscopic confirmation of borepin 2 was carried out
indirectly via formation of its pyridine adduct 2-py. A
CH2Cl2 suspension of 2 went clear upon addition of one
equivalent of pyridine, forming the more soluble adduct,
which could be probed by the usual spectroscopic techni-
ques. The 1H NMR spectrum shows three downfield shifted
signals for the coordinated pyridine at 9.35 (broadened by
coupling to the 10B/11B nuclei), 8.42, and 7.93 ppm for the
ortho, para and meta protons, respectively. The 11B NMR
chemical shift observed for 2-py is at -2.3 ppm, consistent
with a neutral, four-coordinate boron center. The 19F NMR
spectrum is complex in the ortho and meta fluorine regions
due to rotational stasis in some of the rings, but shows four
para fluorine resonances in the expected 2:2:2:1 ratio at
-150.4, -151.5, -153.9, and -154.7 ppm.
The solid-state structures of both 2 and 2-py were deter-
mined by X-ray crystallography; thermal ellipsoid diagrams
of each are shown in Figure 1, along with selected metrical
parameters. Compound 2 crystallizes with two independent
molecules in the unit cell (only molecule A is shown in
Figure 1); the metrical parameters are similar for both
molecules of 2, the primary differences being the intraring
boron-carbon distances and the extent of puckering in the
Results and Discussion
Perfluoropentaphenylborole, 1, reacts slowly with elec-
42
tron-poor dienophile C6F5CCC6F5 at 110 °C in toluene
over the course of a week to form a new product, identified as
the perfluorinated heptaphenylborepin 2 (Scheme 1). Eisch
has detailed the “paradigm of pericyclic reactions”30 that
(30) Eisch, J. J.; Galle, J. E.; Shafii, B.; Rheingold, A. L. Organome-
tallics 1990, 9, 2342–2349.
(31) Piers, W. E.; Chivers, T. Chem. Soc. Rev. 1997, 26, 345–354.
(32) Beringhelli, T.; Donghi, D.; Maggioni, D.; D’Alfonso, G. Coord.
Chem. Rev. 2008, 252, 2292–2313.
(33) Erker, G. Dalton Trans. 2005, 1883–1890.
(34) Piers, W. E. Adv. Organomet. Chem. 2004, 52, 1–76.
(35) Chase, P. A.; Piers, W. E.; Patrick, B. O. J. Am. Chem. Soc. 2000,
122, 12911–12912.
(36) Chase, P. A.; Romero, P. E.; Piers, W. E.; Parvez, M.; Patrick,
B. O. Can. J. Chem. 2005, 83, 2098–2105.
(37) Williams, V. C.; Dai, C.; Li, Z.; Collins, S.; Piers, W. E.; Clegg,
W.; Elsegood, M. R. J.; Marder, T. B. Angew. Chem., Int. Ed. 1999, 38,
3695–3698.
(38) Metz, M. V.; Schwartz, D. J.; Stern, C. L.; Nickias, P. N.; Marks,
T. J. Angew. Chem., Int. Ed. 2000, 39, 1312–1316.
(39) Metz, M. V.; Schwartz, D. J.; Stern, C. L.; Marks, T. J.; Nickias,
P. N. Organometallics 2002, 21, 4159–4168.
(40) Fan, C.; Piers, W. E.; Parvez, M. Angew. Chem., Int. Ed. 2009, 48,
2955–2958.
(41) Jacobsen, H.; Berke, H.; Doring, S.; Kehr, G.; Erker, G.;
Frohlich, R.; Meyer, O. Organometallics 1999, 18, 1724–1735.
(42) Birchall, J. M.; Bowden, F. L.; Haszeldine, R. N.; Lever, A. B. P.
J. Chem. Soc. A 1967, 747–753.
(43) Williams, R. V. Chem. Rev. 2001, 101, 1185–1204.
(44) Wilkey, J. D.; Schuster, G. B. J. Org. Chem. 1987, 52, 2117–2122.
(45) Wilkey, J. D.; Schuster, G. B. J. Am. Chem. Soc. 1988, 110, 7569–
7571.

5134 Organometallics, Vol. 29, No. 21, 2010
Fan et al.
˚
Figure 1. (a) Thermal ellipsoid diagram (50%) of molecule A of compound 2; view from above ring. Selected bond distances (A):
B1-C1, 1.502(4); C1-C2, 1.392(3); C2-C3, 1.465(3); C3-C4, 1.369(3), C4-C5, 1.465(4); C5-C6, 1.370(4); B1-C6, 1.510(3);
B1-C71, 1.587(4). Selected bond angles (deg): C1-B1-C6, 122.4(2); C1-B1-C71, 118.7(2); C6-B1-C71, 118.3(2). Selected dihedral
planes (deg): θprow, 23.1(2) (20.2(2) for molecule B); θstern, 25.2(1) (23.2(1) for molecule B). (b) Thermal ellipsoid diagram (50%) of
˚
compound 2-py; view from side of ring. Selected bond distances (A): B1-C1, 1.604(7); C1-C2, 1.358(7); C2-C3, 1.492(6); C3-C4,
1.357(7), C4-C5, 1.480(7); C5-C6, 1.359(7); B1-C6, 1.640(7); B1-C12, 1.655(8); B1-N1, 1.604(7). Selected bond angles (deg):
N1-B1-C1, 111.2(4); N1-B1-C6, 108.9(4); C1-B1-C6, 104.4(4); N1-B1-C12, 107.4(4); C1-B1-C12, 113.2(4); C6-B1-C12,
111.8(4). Selected dihedral planes (deg): θprow, 45.8(4); θstern, 31.8(3). (c) Diagram illustrating the dihedral planes θprow and θstern
defining the deviation of the ring from planarity into a boat-like conformation.
Scheme 1
ring conformation. In molecule A, the B-C distances are
C3-C4, and C5-C6 exhibiting shorter distances than the
C2-C3 and C4-C5 lengths. The boron center is essentially
planar, with the sum of angles about B1 being 359.4(3)°; the
intraring C1-B1-C6 is larger than the extra-ring angles by
about 4°. The nonplanar nature of 2 contrasts with the
structure observed for the unsubstituted chloroborepin re-
ported by Ashe et al.,23 where the ring is highly planar. Thus,
the lack of planarity in 2 must be largely a consequence of the
need to accommodate the large C6F5 substituents around the
ring.
Coordination of pyridine to the boron center results in a
tetrahedral boron center; the side view of 2-py depicted in
Figure 1b illustrates more clearly the puckering of the
borepin ring. Indeed, the θprow and θstern angles are con-
siderably larger in 2-py than in 2 at 45.8(4)° and 31.8(3)°,
respectively. The occupation of the empty orbital on boron
by the pyridine base also serves to enhance the localization of
˚
1.502(4) and 1.510(3) A for B1-C1 and B1-C6, respectively,
while in molecule B the analogous values are somewhat
˚
shorter at 1.475(3) and 1.475(4) A. This may be due to the
somewhat more planar seven-membered ring in molecule B,
resulting in more π delocalization in the seven-membered
ring in molecule B. Nevertheless, in both molecules of 2 the
seven-membered ring is nonplanar,46 with both B1 and C3/
C4 rising out of the plane defined by C1-C2-C5-C6; the
angles between this plane and those defined by C1/B1/C6
and C2/C3/C4/C5 (θprow and θstern, see Figure 1c) are
23.1(2)° and 25.2(1)°, respectively, in molecule A, but slightly
smaller at 20.2(2)° and 23.2(1)° for molecule B. There is
distinct bond alternation within the ring, with C1-C2,
(46) Schulman, J. M.; Disch, R. L. Organometallics 2000, 19, 2932–
2936.

Article
Organometallics, Vol. 29, No. 21, 2010 5135
Scheme 2
ortho-fluorines on the C6F5 groups on C3 and C4, leaving the
pentafluorophenyl group on boron to assume an equatorial
conformation. The B1-N1 distance is comparable to those
of other perfluoroarylborane-pyridine adducts47 and slightly
shorter than various imine adducts of B(C6F5)3, which have
48
˚
B-N distances ranging from 1.627(3) to 1.658(6) A.
Clearly, despite the steric crowding about the boron center,
the perfluorination of the seven phenyl rings enhances the
Lewis acidity of the boron in this new member of the
perfluoroarylborane family of Lewis acids.34
Borole 1 is an electron-poor diene, and its slow reaction
with the perfluorodiphenylacetylene dienophile suggests that
a more electron-rich alkyne might react with 1 at a faster rate.
Indeed, the reaction of 1 with diphenylacetylene proceeds
essentially upon mixing at 25 °C in methylene chloride to
produce a complex mixture of products as determined by 19F
NMR spectroscopy. The chemistry that occurs is summar-
ized in Scheme 2. On the basis of an analysis of the products
that would be produced via the Diels-Alder/pericyclic
rearrangement pathway depicted in Scheme 1, four diphenyl-
substituted heptaarylborepin isomers are expected: two Cs
symmetric isomers (4,5-Ph2-2 and 2,7-Ph2-2) and two un-
symmetric isomers (2,3-Ph2-2 and 3,4-Ph2-2). (For a more
complete depiction as to how these isomers arise, see Scheme
S1 in the Supporting Information.) Since the 19F NMR
spectrum of the crude product mixture (see Figure S1 in
the Supporting Information) indicates that the major pro-
duct of the reaction has five inequivalent C6F5 groups, we
initially surmised that one of the two unsymmetric isomers of
Ph2-2 was the dominant product. However, an X-ray crystal-
lographic analysis of a single crystal of this major product
(which selectively crystallizes nicely from the crude product
mixture) reveals its structure to be the boracyclohexadiene
compound 3, in which the diphenylacetylene reagent man-
ifests itself as the exo diphenylmethylene unit shown in
blue and red in Scheme 2. From the 19F NMR spectrum
(Figures S1 and S2), it can be estimated that 3 comprises
approximately 75% of the product mixture; the rest is likely
made up of the diphenyl borepins Ph2-2, the major compo-
nent being the 3,4-Ph2-2 isomer. This was deduced through
Figure 2. Thermal ellipsoid diagram (50%) of compound 3,4-
˚
Ph2-2-py; view from above ring. Selected bond distances (A):
B1-C1, 1.612(5); C1-C2, 1.361(5); C2-C3, 1.480(5); C3-C4,
1.351(5), C4-C5, 1.474(5); C5-C6, 1.361(5); B1-C6, 1.646(5);
B1-C12, 1.653(5); B1-N1, 1.639(4). Selected bond angles
(deg): C1-B1-N1, 108.9(3); C1-B1-C6, 105.3(3); N1-B1-
C6, 109.3(3); C1-B1-C12, 113.0(3); N1-B1-C12, 107.9(3);
C6-B1-C12, 112.4(3). Selected dihedral planes (deg): θprow
47.3(2); θstern, 34.9(3).
,
double bonds in the seven-membered ring, as evidenced by
the elongation of the single bonds and contraction of the
double bonds of the ring relative to unligated 2. Particularly,
˚
the B1-C1 and B1-C6 distances in 2-py are ∼0.1 A longer
than those in 2, as conjugation between B1 and the alpha
carbons is disrupted. The pyridine occupies an axial site
on the prow of the boat and is slightly bent away from the
(47) Lesley, M. J. G.; Woodward, A.; Taylor, N. J.; Marder, T. B.;
Cazenobe, I.; Ledoux, I.; Zyss, J.; Thornton, A.; Bruce, D. W.; Kakkar,
A. K. Chem. Mater. 1998, 10, 1355–1365.
(48) Blackwell, J. M.; Piers, W. E.; Parvez, M.; McDonald, R.
Organometallics 2002, 21, 1400–1407.

5136 Organometallics, Vol. 29, No. 21, 2010
Fan et al.
Scheme 3
twisting of the C1-C6 double bond may also in part be
driven by the assumption of π-stacking interactions between
the C37 phenyl group with the C6F5 group on C2, and the
C43 phenyl moiety with the boron C6F5 substituent
(indicated by the double-headed arrows in Figure 3).
Given the expectation of Diels-Alder-dominated chem-
istry in the reaction of 1 with diphenylacetylene, the forma-
tion of 3 as the major product is surprising. We postulate that
it is a consequence of the very high Lewis acidity of 1, which
opens a new pathway for the reactivity of boroles with
alkynes that involves electrophilic attack at the carbon-
carbon triple bond, followed by rearrangement to a vinylidene-
like group via 1,2-migration of the phenyl group as shown in
Scheme 3. This alkyne-to-vinylidene sequence has ample pre-
cedence in transition metal chemistry;49,50 usually, it involves
terminal alkynes since the proton is an excellent migrating
group, but there is also precedence for this isomerization in
coordinated internal alkynes.51 Here, the coordination of the
alkyne to the electrophilic borole center is likely reversible, and
Diels-Alder chemistry to yield borepins Ph2-2 is competitive
with the rate-limiting migration of the phenyl group in zwitter-
ionic intermediate Ph2-C to irreversibly yield Ph2-D. The
reactivity of the highly electrophilic borole 1 with more
electron rich alkynes thus takes two divergent, competitive
pathways.
Figure 3. Thermal ellipsoid diagram (50%) of compound 3;
˚
view from above ring. Selected bond distances (A): B1-C1,
1.520(7); C1-C2, 1.469(6); C2-C3, 1.398(7); C3-C4, 1.438(6),
C4-C5, 1.390(7); B1-C5, 1.497(7); B1-C31, 1.566(7); C1-C6,
1.411(7). Selected bond angles (deg): C5-B1-C1, 117.0(4);
C5-B1-C31, 117.8(4); C1-B1-C31, 125.0(4); C1-C6-C43,
120.9(4); C1-C6-C37, 122.1(4); C43-C6-C37, 117.0(4). Se-
lected dihedral angles (deg): B1-C1-C6-C43, 42.0(6);
C2-C1-C6-C37, 37.2(7).
the isolation of its pyridine adduct via treatment of the
mother liquor of a crude reaction mixture with pyridine after
the bulk of compound 3 had been removed by precipitation.
Crystals of 3,4-Ph2-2-py suitable for X-ray analysis were
obtained from such an experiment, and a thermal ellipsoid
diagram is shown in Figure 2, along with selected metrical
data. Overall, the structure is very similar to that of 2-py,
except for the unfluorinated phenyl groups in the 3 and 4
positions of the ring. Here again, the pyridine occupies an
axial position on the boron, tucked into the pocket formed
by the prow and stern of the boat conformation of the
borepin ring.
Compound 3 was synthesized separately via the reaction
of the bromoborole 1-Br40 and diphenyl acetylene as shown
in Scheme 4. Interestingly, 1-Br reacts almost exclusively
via the electrophilic attack pathway to produce 3-Br as the
sole boron-containing product; conversion of this material
to 3 proceeds smoothly with prolonged heating at 100 °C
using the bis-pentafluorophenyl zinc reagent52 to install
the -C6F5 group on boron. In this way, analytically
pure samples of 3 (free of borepin side products) were
obtainable.
The molecular structure of boracycle 3 is shown in Figure 3,
along with selected metrical data. The intraring bond dis-
tances reflect the localization of carbon-carbon double
˚
bonds between C2/C3 and C4/C5 (1.398(7) and 1.390(7) A,
respectively) and single bonds between B1 and the sp2
˚
carbons C1 and C5 (1.520(7) and 1.497(7) A, respectively).
Indeed the atoms of the six-membered boracycle are essen-
tially coplanar, with deviations from the plane defined by the
˚
six atoms ranging from 0.010 to 0.033 A. The exo double
bond between C1 and C6 is somewhat elongated from typical
˚
values at 1.411(7) A, a consequence of its distortion from
Given these observations, it was of interest to attempt a
reaction of 1 with an electron-rich internal alkyne substituted
with groups that are less prone to migration, that is, those
coplanarity with the π system of the intra-ring diene. This
distortion is quite severe, as measured by the B1-C1-
C6-C43 and C2-C1-C6-C37 torsion angles of 42.0(6)°
and 37.2(7)° and the deviation of C6 from the plane defined
(49) Bruce, M. I. Chem. Rev. 1991, 91, 197–257.
(50) Bruneau, C.; Dixneuf, P. H. Acc. Chem. Res. 1998, 32, 311–323.
(51) Ikeda, Y.; Yamaguchi, T.; Kanao, K.; Kimura, K.; Kamimura,
S.; Mutoh, Y.; Tanabe, Y.; Ishii, Y. J. Am. Chem. Soc. 2008, 130, 16856–
16857.
(52) Sun, Y. M.; Piers, W. E.; Parvez, M. Can. J. Chem. 1998, 76, 513–
517.
˚
by B1 and C1-C5 of 0.170 A and likely arises due to steric
interactions between the phenyl groups on C6 with adjacent
C6F5 groups on B1 and C2. The C1-B1-C31 angle of
125.0(4)° is also impacted by this factor. However, the

Article
Organometallics, Vol. 29, No. 21, 2010 5137
Scheme 4
Essentially, compound 4 is one of the two possible isomers of
the 7-borabicyclo[2.2.1]heptadienes formed upon [4þ2]
Diels-Alder cycloaddition30 between 1 and 3-hexyne. Inter-
estingly, it is not the expected kinetic product with ethyl
groups in the 2,3 positions of the borabicyclo[2.2.1]heptadiene
ring, but rather the (apparently) more thermodynamically
stable 1,2-Et2-4 isomer as shown. While stable enough to
isolate (in 59% yield) and fully characterize, this product is
converted to a mixture of products upon heating in solution at
80 °C for 8 h, some of which are likely borepin isomers of Et2-2
on the basis of low-field boron signals in the 11B NMR
spectrum.
While the precise nature of all of the products in this
mixture was not determined, examination of the bond dis-
tances in the structure of 1,2-Et2-4 reveals some nuances in
the structure that suggest it is poised to undergo a suprafacial
1,3-sigmatropic shift that would lead to the 7-borabicyclo-
[4.1.0]heptadiene intermediate and borepin 3,4-Et2-2 shown
in Scheme 5. The compound crystallizes with two molecules
in the unit cell with similar metrical parameters; both are
given in the caption for Figure 4 for comparison. The ethyl
groups are on C1 and C2 in the view presented, and from this
perspective, it is clear that the B(C6F5) unit is tilted decidedly
toward C2 and C3. The angle subtended between the planes
defined by B1/C1/C4 and C1/C2/C3/C4 is acute (83.0(2)°),
while the one between the former plane and that defined by
C1/C6/C5/C4 is obtuse, at 152.7(2)°. Although the boron is
not four-coordinate, this tilt toward the C2-C3 double bond
may contribute to the unusual negative value of the chemical
shift for boron; another potential factor is the nonideal
C1-B1-C4 angle of 96.5(2)°. In any case, the double bond
Figure 4. Thermal ellipsoid diagram (50%) of molecule A of
˚
compound 1,2-Et2-4. Selected bond distances (A); bold para-
meters cited after are the corresponding values for molecule B:
B1-C1, 1.651(4), 1.651(4); B1-C4, 1.628(4), 1.634(4); C1-C2,
1.504(4), 1.506(4); C2-C3, 1.388(4), 1.398(4); C3-C4, 1.534(4),
1.530(4); C4-C5, 1.517(4), 1.518(4); C5-C6, 1.329(4), 1.336(4);
C1-C6, 1.513(4), 1.507(4); B1-C35, 1.574(4), 1.570(4). Se-
˚
lected nonbonding distances (A): B1-C2, 1.835(4), 1.825(5);
B1-C3, 1.812(4), 1.795(4). Selected bond angles (deg):
C1-B1-C4, 96.5(2), 96.6(2); C1-B1-C35, 127.6(3), 127.6(3);
C4-B1-C35, 135.0(3), 134.8(3). Selected dihedral planes (deg);
plane 1 contains C1/B1/C4; plane 2 contains C1/C2/C3/C4;
plane 3 contains C1/C6/C5/C4: angle between planes 1 and 2,
83.0(2), 82.0(2); angle between planes 1 and 3, 152.7(2), 152.5(2);
angle between planes 2 and 3, 124.8(2), 125.5(2).
˚
between C2 and C3 is 1.338(4) A, slightly longer than the
˚
linkage between C5 and C6, which is 1.329(4) A, by virtue of
this interaction. The other bonds in the C1-C6 rings are
clearly single bonds.
with sp3-hybridized carbons. Accordingly the reaction of 1
with 3-hexyne was explored. Not surprisingly, a rapid reaction
occurred upon mixing of these two reagents, as indicated by
the immediate loss of the deep purple color of 1 when
3-hexyne was added, forming a light orange-yellow solution.
Multinuclear NMR spectroscopy suggests the formation of a
single product, 4, with five inequivalent C6F5 groups and two
inequivalent (or diastereotopic) ethyl groups, consistent with
either of the reaction channels described above. However,
the 11B NMR shows one broad signal at -10.3 ppm, a
chemical shift inconsistent with either a borepin or a bor-
acyclohexadiene structure, whose boron nuclei resonate at
40-50 ppm. The observed -10.3 chemical shift is generally
associated with neutral (and perhaps anionic) four-coordinate
boron centers.
The B1-C distances in the molecule are interesting and
foreshadow the propensity for the molecule to convert to the
observed borepin, which involves a 1,3 shift that cleaves the
B1-C1 bond and forms a bond between B1 and C3. In
˚
accord with this, the B1-C1 distance of 1.651(4) A is slightly
˚
longer than the B1-C4 distance of 1.628(4) A, and B1 draws
˚
˚
slightly nearer to C3 (1.812(4) A) than it is to C2 (1.835(4) A);
the differences are more pronounced in molecule B. We
speculate that this isomer is the thermodynamic isomer
because the developing positive charge on C2 as this occurs
is more stabilized by an ethyl group than a pentafluorophe-
nyl group.
Regardless of this apparent structural bias toward decom-
position to yield borepin 3,4-Et2-2, the observation of other
products in the crude mixture formed upon thermal decom-
position of 1,2-Et2-4 suggests that conversion to other iso-
mers of Et2-2 and decomposition via other unidentified
modes (possibly via the intermediate Et2-C) are occurring
in this dynamic system.
The nature of 4 was determined conclusively via X-ray
crystallography; a thermal ellipsoid diagram is shown in
Figure 4, along with selected metrical parameters, and the
chemistry involved in this reaction is depicted in Scheme 5.

5138 Organometallics, Vol. 29, No. 21, 2010
Fan et al.
Scheme 5
Summary
over and distilled from CaH2 twice. Diphenylacetylene was pur-
chased from Aldrich and used as received. Perfluoropenta-
phenylborole,40 C6F5CtCC6F5,42 and Zn(C6F5)252 were prepared
The reactions of the perfluorinated pentaphenylborole 1
with three symmetrical internal alkynes have been studied.
These investigations show that, in addition to previously
observed Diels-Alder reactivity patterns for the reactions of
unfluorinated boroles with alkynes, the potent Lewis acidity
of 1 opens a new channel of reactivity for certain alkynes
involving electrophilic attack of the alkyne and the triggering
of an alkyne rearrangement reminiscent of those long known
for transition metal alkyne complexes. These two reaction
manifolds are competitive, with the nucleophilicity of the
alkyne and the ability of the alkyne substituent to undergo
1,2 migration being the main factors in determining the
dominant pathway. For example, the electron-poor alkyne
C6F5CCC6F5 exclusively takes part in Diels-Alder chemis-
try, yielding the new perfluoroaryl borane 2. Diphenylace-
tylene, on the other hand, undergoes both reactions, but the
electrophilic attack pathway dominates, because of its elec-
tron richness and the ability of the sp2-hybridized aryl ipso
carbon to lower the transition-state energy for 1,2 migration.
3-Hexyne, despite its electron-rich nature, does not suffer
electrophilic attack in the dominant pathway, although a
zwitterionic adduct akin to Et2-C may form and undergo
decomposition via mechanisms that do not involve ethyl group
migration at higher temperatures. Instead, Diels-Alder chem-
istry occurs, but the product at room temperature is the
diethyl-substituted 7-borabicyclo[4.1.0]heptadiene compound
that represents a previously proposed intermediate in the
reaction of boroles with alkynes. Characterization of 1,2-Et2-4
provides solid experimental evidence for the proposals of Eisch
regarding the mechanism of formation of borepins from boroles
and alkynes, representing one of the proposed intermediates
along this reaction path.
1
as previously reported. All H, 11B, and 19F NMR spectra were
recorded on a Bruker 400 MHz spectrometer (operating at 400
MHz (1H), 128 MHz (11B), and 376 MHz (19F)) at 25 °C or as
indicated. Chemical shifts reported were relative to SiMe4 (1H) or
BF3 (11B and 19F) standards. Due to poor solubility and 13C-19F
coupling, 13C spectra were not obtained. X-ray crystallography
analyses were performed on suitable crystals coated in Paratone oil
and mounted on a Nonius Kappa CCD diffractometer.
Synthesis of Perfluoroheptaphenylborepin, 2. Perfluoropenta-
phenylborole (1) (123 mg, 0.14 mmol) and C6F5CtCC6F5 (50
mg, 0.14 mmol) were mixed in toluene (3 mL). The suspended
mixture was sealed in a flask and heated to 110 °C for one week.
The white precipitate was filtered, washed by CH2Cl2 (3 ꢀ 3
mL), and dried in vacuo. Yield: 142 mg (82%). Anal. (%) Calcd
for C48BF35: C 46.04. Found: 45.71.
Synthesis of Compound 2-py. Perfluoroheptaphenylborepin (2)
(100 mg, 0.09 mmol) was suspended in CH2Cl2 (2 mL). Pyridine
(7 mg, 0.09 mmol) was added. The mixture was stirred and heated
to form a clear solution and maintained at room temperature
overnight to form colorless crystals. The crystals were filtered and
washed with hexane (3 ꢀ 3 mL) and dried in vacuo. Yield: 72 mg
(70%). 1H NMR (CD2Cl2): 9.35 (d, 2H), 8.42(t, 1H), 7.93 (t, 2H).
11B{1H} NMR (CD2Cl2): -2.3 (br). 19F{1H} NMR (CD2Cl2):
-121.5 (m, 2F, o-C6F5), -133.8 (m, 2F, o-C6F5), -135.0 (m, 2F,
o-C6F5), -137.6 (m, 2F, o-C6F5), -140.3 (m, 2F, o-C6F5), -141.6
(m, 2F, o-C6F5), -142.3 (m, 2F, o-C6F5), -150.4 (t, 2F, p-C6F5),
-151.5 (t, 2F, p-C6F5), -153.9 (t, 2F, p-C6F5), -154.7 (t, 1F,
p-C6F5), -160.2 (m, 2F, m-C6F5), -160.5 (m, 2F, m-C6F5),
-161.0 (m, 2F, m-C6F5), -161.4 (m, 2F, m-C6F5), -161.9 (m,
2F, m-C6F5), -163.5 (m, 2F, m-C6F5), -164.6 (m, 2F, m-C6F5).
Anal. (%) Calcd for C35H5BF35N CH2Cl2: C 45.79, H 0.50, N
3
1.00. Found: C 45.79, H 0.64, N 1.27.
Synthesis of Compound 3. 1-Br (1 g, 1.24 mmol) and diphe-
nylacetylene (0.22 g, 1.24 mmol) were mixed in toluene (20 mL)
and stirred overnight to form a dark blue solution. Zn(C6F5)2
(0.25 g, 0.62 mmol) was added, and the mixture was heated to
100 °C until complete conversion of 3-Br to 3 was observed by
19F NMR spectroscopy (3-5 days). The volatiles were removed
under reduced pressure, and the dark purple residue was then
dissolved in CH2Cl2 and filtered. After the solvent was removed,
the dark purple solid was recrystallized from hot toluene
(6 mL) and dried in vacuo. Yield: 0.85 g (64%). 1H NMR
Experimental Section
General Procedures. All operations were performed under a
purified argon atmosphere using glovebox or vacuum line
techniques. Toluene and hexane solvents were dried and purified
by passing through activated alumina and vacuum distilled from
Na/benzophenone. CH2Cl2, pyridine, and 3-hexyne were dried

Article
Organometallics, Vol. 29, No. 21, 2010 5139
(CD2Cl2): 7.62-7.57 (m, 2H, C6H5), 7.48-7.47 (m, 2H, C6H5),
7.41-7.33 (m, 4H, C6H5), 7.22-7.20 (m, 2H, C6H5). 11B{1H}
NMR (CD2Cl2): 41.3 (br). 19F{1H} NMR (CD2Cl2): -131.4 (m,
2F, o-C6F5), -136.6 (m, 2F, o-C6F5), -138.7 (m, 2F, o-C6F5),
-138.8 (m, 2F, o-C6F5), -140.8 (m, 2F, o-C6F5), -151.8 (t, 1F,
p-C6F5), -152.0 (t, 1F, p-C6F5), -153.6 (t, 1F, p-C6F5), -155.3
(t, 1F, p-C6F5), -155.6 (t, 1F, p-C6F5), -161.1 (m, 2F, m-C6F5),
-161.6 (m, 2F, m-C6F5), -162.4 (m, 2F, m-C6F5), -162.9 (m,
2F, m-C6F5), -163.3 (m, 2F, m-C6F5). Anal. (%) Calcd for
o-C6F5), -130.4 (br, 1F, o-C6F5), -135.4 (m, 1F, o-C6F5), -137.4
(m, 2F, o-C6F5), -137.5 (m, 1F, o-C6F5), -137.8 (m, 1F, o-C6F5),
-139.5 (m, 2F, o-C6F5), -139.9 (m, 1F, o-C6F5), -150.3 (t, 1F,
p-C6F5), -151.2 (t, 1F, p-C6F5), -152.0 (t, 1F, p-C6F5), -152.6 (t,
1F, p-C6F5), -153.5 (t, 1F, p-C6F5), -159.5 (m, 1F, m-C6F5),
-160.0 (m, 1F, m-C6F5), -160.6 (m, 1F, m-C6F5), -161.0 (m, 2F,
m-C6F5), -161.7 (m, 3F, m-C6F5), -162.5 (br, 2F, m-C6F5). Anal.
(%) Calcd for C40H10BF25 0.5C6H14: C 50.67, H 1.68. Found: C
3
50.75, H 1.82.
C48H10BF25 C7H8: C 56.73, H 1.56. Found: C 57.08, H 1.88.
3
Synthesis of Compound 4. Perfluoropentaphenylborole (1)
(500 mg, 0.56 mmol) and 3-hexyne (46 mg, 0.56 mmol) were
mixed in toluene (10 mL). After the mixture was stirred for 4 h,
all the volatiles were removed under vacuum. The orange solid
was redissoved in hexane (5 mL) to obtain a yellow solution.
Off-white crystals were allowed to grow at room temperature
from the yellow solution over the course of five days and isolated
by filtration. After washing with cold hexane (1 ꢀ 3 mL), the
product was dried in vacuo to obtain an off-white solid. Yield:
320 mg (59%). 1H NMR (CD2Cl2): 2.75 (m, 1H, CH2CH3), 2.47
(m, 1H, CH2CH3), 2.29(m, 1H, CH2CH3), 1.72(m, 1H, CH2CH3),
1.43 (m, 3H, CH2CH3), 0.87 (t, 3H, CH2CH3). 11B{1H} NMR
(CD2Cl2): -10.3 (br). 19F{1H} NMR (CD2Cl2): -122.6 (br, 1F,
Acknowledgment. Funding for this work was provided
by the National Science and Engineering Research Coun-
cil in the form of a Discovery Grant to W.E.P., the Xerox
Research Foundation, and the Institute for Sustainable
Energy, Economy and Environment (University of
Calgary). The authors thank Lauren G. Mercier for useful
discussions.
Supporting Information Available: Further details, including
NMR spectra, schemes, and crystallographic data (CIF) for com-
pounds 2, 2-py, 3,4-Ph2-2-py, 3, and 1,2-Et2-4. This information is
available free of charge via the Internet at http://pubs.acs.org.