Reaction of an alkynyllithium reagent with boron trifluoride: The structure of a likely intermediate in alkynyl anion chemistry

Article abstract of DOI:10.1039/a704185c

Lithium phenylacetylide reacts with 1 equiv. of BF₃·OEt₂ in thf-toluene to yield both a thf adduct of tris(phenylacetylido)borane 2, which is probably the true intermediate in RC≡CLi/BF₃-promoted alkynyl anion chemistry, and LiBF₄, products which are considered to form via sequential LiF elimination from previously proposed acetylidoborate intermediates; complex 2 is the first structurally characterised tris(alkynyl)borane and in the solid state it forms an unusual hydrogen-bonded dimer.

Full text of DOI:10.1039/a704185c

View Article Online / Journal Homepage / Table of Contents for this issue
Reaction of an alkynyllithium reagent with boron trifluoride: the structure of a
likely intermediate in alkynyl anion chemistry
John E. Davies, Paul R. Raithby, Ronald Snaith and Andrew E. H. Wheatley*
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge, UK CB2 1EW
Lithium phenylacetylide reacts with 1 equiv. of BF3·OEt2 in
thf–toluene to yield both a thf adduct of tris(phenylacetyli-
do)borane 2, which is probably the true intermediate in
RC·CLi/BF3-promoted alkynyl anion chemistry, and LiBF4,
products which are considered to form via sequential LiF
elimination from previously proposed acetylidoborate inter-
mediates; complex 2 is the first structurally characterised
tris(alkynyl)borane and in the solid state it forms an unusual
hydrogen-bonded dimer.
(i.e. mono and bis) acetylidoboranes. Dimerisation via hydro-
gen bonding is seemingly a new feature in the solid state for
acetylidoborane structures. The relevant interaction is between
the adduct O-centre and the m-aromatic proton on a phenyl-
acetylide ligand in an adjacent borane monomer. This results in
the formation of a 16-membered ring, incorporating an O···C
hydrogen bonding distance12 of 3.46(1) Å.
Notwithstanding the structure of 2, obtained from the 1:1
reaction of Ph–C·C–Li and BF3·OEt2 in thf, NMR studies show
that the bulk product is a co-crystalline mixture. While 1H and
13C NMR spectra of the product were consistent with the
observed structure of 2, 11B NMR spectroscopy in Me2SO
yielded only a first-order quintet at d 20.75. At 1.1 Hz, coupling
was identical to that demonstrated by the quartet observed in the
19F spectrum at d 2147.59 and accorded with previously
The value of organometallic nucleophilic addition to imines1
has been limited in the past by the poor reactivity of the latter as
acceptors,2 and by their propensity for enolisation in the
presence of, for example, organolithium species.3 While initial
attempts at activating imines towards nucleophilic addition of
organometallic reagents involved the use of organocopper–
boron trifluoride complexes4 it was subsequently reported that
supposed in situ generation of alkynylborane/borate species5
led to b-aminoacetylenes from aldimines.6 More recent work
has incorporated the use of similar in situ boron-containing
intermediates in the synthesis of allenes from N-aziridinyl-
imines.7 In each case, however, the existence of the alkynyl-
borane/borate intermediate was assumed, the presumption
being that lithium phenylacetylide8 reacted with 1 equiv. of
BF3·OEt2 to yield either the monophenylacetylidoborate, 1a, or,
following the elimination of lithium fluoride, the mono-
phenylacetylidoborane, 1b (Scheme 1). An unexplained anom-
aly was that the subsequent nucleophilic addition reactions gave
higher yields in the presence of 2 or, better, 3 equiv. of 1a/b.
1
observed J11 coupling for the tetrafluoroborate anion.13
BF
Confirmation that LiBF4, 3, was present in the isolated crystals
came from the observation of a very sharp singlet at d 20.11 in
the 7Li spectrum in Me2SO, suggesting the presence of
[Li(Me2SO)4]+.
At room temperature, 11B NMR spectroscopy of the bulk
product in Me2SO failed to yield any signal attributable to 2.
Studies were therefore undertaken in alternative media.
Whereas treatment of the co-crystalline product (2/3) with
Me2SO yielded a solution, 11B NMR spectroscopy demon-
strated that the necessity to filter the suspension afforded in
[2H6]benzene prior to spectral analysis reflected the complete
failure of the tetrafluoroborate component to dissolve. Accord-
ingly, the 11B NMR spectrum at room temp. now showed no
quintet, but rather a broad signal attributable to 2 (d ca. 0.03,
Dn1/2 = 450 Hz). Complete dissolution of 2/3 was afforded by
[2H8]thf, yielding both the tetrafluoroborate quintet (d 21.18)
i, BunLi
+
–
or
Ph
C
C
H
Ph
C
C
BF3Li
Ph
C C BF2
ii, BF3•OEt2
1a
1b
1
with a slightly expanded JBF coupling (1.8 Hz), and a broad,
Scheme 1
high-field signal (d ca. 27.32, Dn1/2 = 340 Hz), the relative
integrations suggesting a 1.25 :1 ratio of 2 to 3.
We have investigated and now report on the actual species
obtained from reacting a 1:1 mixture of a lithium acetylide with
BF3·OEt2. They are not the intermediates 1a/b suggested
previously (Scheme 1). They do, however, explain the prefer-
ence for excess intermediate to be present during subsequent
reaction with amide,5 aldimine,6 or N-aziridinylimine.7 Thus,
reaction of in situ generated lithium phenylacetylide with just 1
equiv. of BF3·OEt2 in thf–toluene affords, as one of the
products, an air-stable mono-thf adduct of tris(phenylacety-
lido)borane 2,† the X-ray crystal structure of which has been
determined.‡ A search of the Cambridge Crystallographic
Database shows that 2 is not only the first structurally
characterised tris(acetylido)borane, but that it is also the first
acetylidoborane structure to incorporate an O-donor molecule.
It exists as a hydrogen-bonded dimer in the solid state (Fig. 1),
the boron centres having distorted tetrahedral geometries, with
the fourth coordination site being occupied by the oxygen of the
thf [mean C–B–C 114.2(2)°, mean O–B–C 104.1(2)°]. Un-
surprisingly, at 1.633(3) Å, the observed B–O distance is rather
longer than any previously observed in three-coordinate
monoacetylidoborane structures containing direct B–O link-
ages.11 Both the B–C [mean 1.580(4) Å] and C·C [mean
1.209(3) Å] distances show close agreement with those in other
Spectroscopic evidence points, therefore, not only to the
elimination of LiF from 1a, yielding 1b in the first instance, but
also to the subsequent addition of LiF to an unreacted boron
C(11A)
C(19A)
C(15A)
C(12A)
H(15B)
C(20A)
C(27)
B(1A)
C(28A)
C(28)
B(1)
O(1A)
O(1)
C(20)
C(27A)
C(12)
C(15)
H(15A)
C(11)
C(19)
Fig. 1 Dimeric structure of 2; hydrogen bonding between thf and a
m-aromatic proton yields a 16-membered ring
Chem. Commun., 1997
1797

View Article Online
BF3•OEt2
+
278 °C. Warming to room temp. gave a white suspension, which afforded
–
Ph
C
C
Li
Ph
C
C
1a
BF3Li
a colourless solution on gentle heating. Storage at +5 °C for three days
afforded a colourless co-crystalline mixture of 2 and 3. Yield (based on 11
B
BF3•OEt2 –LiF
NMR spectroscopy and Scheme 2), 61% (2) and 16% (3), mp 163–164 °C.
1H NMR [250 MHz, 25 °C, (CD3)2SO], d 7.45–7.35 (m, 15 H, Ph), 3.59 (m,
4 H, thf), 1.76 (m, 4 H, thf). 13C NMR (400 MHz, 25 °C, Me2SO), d 131.1,
128.5 (o-,m-Ph), 127.5 (p-Ph), 124.7 (ipso-C-Ph), 95.3 (·C–Ph), 67.1, 25.2
+
BF2
Ph
C
C
LiBF4
3
1b
(thf). 19F NMR (235.361 MHz, 25 °C, Me2SO, CCl3F d = 0), d 2147.59
2Ph
C C Li/2BF3•OEt2 –2LiF
(q, 11BF4
,
1J11 1.1 Hz), 2147.53 (br m, 10BF42). 7Li NMR (155.508
2
BF
MHz, 25 °C, Me2SO, PhLi d = 0), d 20.11 [s, Li(Me2SO)4+]. 11B NMR
(128.379 MHz, 25 °C, Me2SO BF3·OEt2 d = 0), d 20.75 (qnt., BF42, 1JBF
1.1 Hz); (C6D6, BF3·OEt2 d = 0), d ca. 0.03 (br s, Dn1/2 450 Hz); ([2H8]thf,
BF3·OEt2 d = 0), d 21.18 (qut., 1B, 1JBF 1.8 Hz), ca. 27.32 (br s, 1.25B,
Dn1/2 = 340 Hz).
(Ph
C C)3B•thf
+
2LiBF4
3
2
Scheme 2
‡
Crystal data for 2: C28H23BO, monoclinic, space group P21/n,
trifluoride molecule, generating 3. Furthermore, it appears that
1b is attacked by 2 further equiv. of lithium phenylacetylide,
eliminating LiF each time and ultimately giving 2 (Scheme 2).
It is also apparent that an explanation of the favourability of
utilising 2 or 3 equiv. of the supposed acetylidoborate/borane
intermediate (1a,b) in subsequent nucleophilic addition chem-
istry is then at hand. Thus, the true alkynyl anion intermediate,
2, contains not one, but three phenylacetylido-ligands. The
implication is that one phenylacetylido-ligand is relatively
active with respect to subsequent nucleophilic behaviour while
the other two are largely inactive. This can be rationalised in
terms of the sterically congested intermediate species resulting
from the addition of 2 across just one carbonyl,5 imine6 or
N-aziridinylimine (Scheme 3).7
a = 10.188(4), b = 15.159(3), c = 14.5899(11) Å, b = 104.385(9)°,
U = 2182.6(9) Å3, Mr = 386.27, Z = 4, Dc = 1.176 Mg m23, m(Mo-
Ka) = 0.069 mm21, F(000) = 816. Data were collected by the w–2q scan
method on a Rigaku AKC5R four-circle diffractometer at 153(2) K using
graphite-monochromated Mo-KKa radiation (l = 0.71069 Å) in the range
5.16 < 2q < 45.00°, +h, +k, ±l; 3045 reflections of which 2857 were
independent (Rint = 0.0222) and used in all calculations. The structure was
solved using direct methods9 and subsequent Fourier difference syntheses
and refined10 by full-matrix least squares on F2 with anisotropic thermal
parameters for all non-hydrogen atoms. Hydrogen atoms were placed in
geometrically idealised positions and refined using a riding model. In the
final cycles of refinement a weighting scheme of the form w21 = [s2(Fo
)
2
+ (0.0745P)2 + 1.58P], where P = (Fo + 2Fc2)/3, was employed which
2
produced a flat analysis of variance. Final R(F)
= 0.0452 for 2084
reflections with [I > 2s(I)], wR(F2) = 0.2247 for all data; 271 parameters;
goodness of fit = 1.039. Maximum peak and hole in final Fourier difference
map 0.238 and 20.197 e Å23 respectively. CCDC 182/562.
O
1 R. A. Volkmann, in Comprehensive Organic Synthesis, ed. B. M. Trost
and I. Fleming, Pergamon, New York, 1991, vol. 1, ch. 1.12.
2 R. W. Layer, Chem. Rev., 1963, 63, 489; R. A. Volkmann, J. T. Davis
and C. N. Meltz, J. Am. Chem. Soc., 1983, 105, 5946.
3 A. I. Meyers, D. R. Williams and M. Druelinger, J. Am. Chem. Soc.,
1976, 98, 3032.
(Ph
C C)3B•thf
2
X
B
PhCC
PhCC
X
R1
R2
X = NR3, O
R1
R2
C
C
4 M. Wada, Y. Sakurai and K. Akiba, Tetrahedron Lett., 1984, 25,
1079.
Ph
5 M. Yamaguchi, T. Waseda and I. Hirao, Chem. Lett., 1983, 35.
6 M. Wada, Y. Sakurai and K. Akiba, Tetrahedron Lett., 1984, 25,
1083.
7 S. Kim, C. M. Cho and J.-Y. Yoon, J. Org. Chem., 1996, 61, 6018.
8 B. Schubert and E. Weiss, Angew. Chem., 1983, 95, 499; Angew. Chem.,
Int. Ed. Engl., 1983, 22, 496; B. Schubert and E. Weiss, Chem. Ber.,
1983, 116, 3212.
9 G. M. Sheldrick, Acta Crystallogr., Sect. C, 1990, A46, 467.
10 G. M. Sheldrick, SHELXL-93, Program for Crystal Structure Refine-
ment, University of Go¨ttingen, 1993.
Scheme 3
In conclusion, formation of 2, the first example of a
structurally characterised tris(acetylidoborane), is probably via
the sequential elimination of LiF from acetylidoborate pre-
cursors (e.g. 1a). Addition of the eliminated LiF to unreacted
boron trifluoride means that formation of 2 is concomitant with
that of LiBF4, 3, as evidenced by multinuclear NMR spectros-
copy. Finally, the fact that 2 is the true intermediate in RC·CLi/
BF3-promoted alkynyl anion chemistry, rather than 1a or 1b,
helps to explain the stoichiometries required for efficient
subsequent nucleophilic addition reactions.
11 A. Meller, W. Maringgele, G. Elter, D. Bromm, M. Noltemeyer and
G. M. Sheldrick, Chem. Ber., 1987, 120, 1437; W. Maringgele,
H. Knop, D. Bromm, A. Meller, S. Dielkus, R. Herbst-Irmer and
G. M. Sheldrick Chem. Ber., 1992, 125, 1807; H. Schulz, G. Gabbert,
H. Pritzkow and W. Siebert, Chem. Ber., 1993, 126, 1593.
12 Hydrogen-bonding interactions are generally regarded as having C···O
interactive distances of 3.0–4.0 Å. Analysis of 868 R3CH···O systems
gives a mean of 3.59 Å; R. Taylor and O. Kennard, J. Am. Chem. Soc.,
1982, 104, 5063.
We thank the UK EPSRC (A. E. H. W) and The Royal
Society (J. E. D., P. R. R.) for financial support.
Footnotes and References
* E-mail: cmc1006@cam.ac.uk
†
BunLi (1.88 ml, 1.6
m in hexanes, 3.0 mmol) was added to
phenylacetylene (0.33 ml, 3.0 mmol) in thf–toluene (5:1 ml) at 278 °C
under nitrogen. After stirring for 10 min, BF3·OEt2 (0.37 ml, 3.0 mmol) was
added and the resultant yellow solution was stirred for a further 10 min at
13 For NBun4BF4 in CH2Cl2, 1J11BF 1.0 Hz at 280 °C; J. S. Hartmann and
G. J. Schrobilgen, Inorg. Chem., 1972, 11, 940.
Received in Cambridge, UK, 16th June 1997; 7/04185C
1798
Chem. Commun., 1997