Structural evidence for antiaromaticity in free boroles

Article abstract of DOI:10.1002/anie.200704771

(Chemical Equation Presented) Pentaphenylborole and a ferrocenylborole were synthesized and structurally characterized. Both experimental and theoretical data reveal that rather weak intermolecular interactions are able to significantly alter the bond lengths in the borole ring of pentaphenylborole (see picture). Moreover, the high Lewis acidity of boroles is demonstrated by a significant Fe...B interaction in the ferrocenylborole in the solid state.

Full text of DOI:10.1002/anie.200704771

Angewandte
Chemie
DOI: 10.1002/anie.200704771
Boron Heterocycles
Structural Evidence for Antiaromaticity in Free Boroles**
Holger Braunschweig,* Israel Fernµndez, Gernot Frenking,* and Thomas Kupfer
Unsaturated boron-containing heterocycles such as borirenes
characterize a simple cyclopentadienyl cation have failed so
far, as these species are highly reactive.
[
1]
[2]
[3]
[7g–h,8]
(I), boroles (II), and borepins (III) have attracted
fundamental interest owing to their electronic structure.The
According to ab initio calculations, the isoelectronic
borole (II) is predicted to have an antiaromatic singlet
ground state featuring strongly alternating bond lengths
within the BC4 backbone, which is destabilized by the
interaction of the empty p orbital at boron in these systems
z
with the unsaturated carbon backbone might result in a
stabilization (I and III) or destabilization (II) of the entire
p system, depending on the number of available p electrons.
In particular, boroles (II) are of interest because of their close
relationship to the cyclopentadienyl cation (IV), which
[9]
delocalization of the four p electrons. As a consequence of
the antiaromaticity, boroles are regarded as highly reactive
species, and the parent molecule II has only been generated in
[10]
the coordination sphere of transition metals. To date, the
number of stable and monomeric boroles that have been
isolated and spectroscopically characterized is restricted to
[2c]
the pentaphenyl-substituted derivative PhBC Ph (1),
4
4
whereby the antiaromatic character was demonstrated both
[2c,e,3a]
by UV/Vis spectroscopy and by reactivity studies.
However, no X-ray diffraction data of any monomeric, non-
annulated borole derivative have been reported to date, even
though structural data are of interest for determining the
consequences of p-electron delocalization in this four-elec-
tron, formally antiaromatic system.
represents one of the prototypical molecules in the theory
of aromaticity and antiaromaticity.Both ESR spectroscopic
[
4]
[5]
data as well as photoelectron spectroscopic studies have
led to the conclusion that the electronic ground state of the
cyclopentadienyl cation is a triplet state with almost ideal D5h
symmetry, but that several singlet states lie very close in
Herein, we report on the synthesis and full character-
ization of 1 and the related ferrocenylborole FcBC Ph [2;
4
4
5
5
Fc = (h -C H )Fe(h -C H )], as well as on the elucidation of
5
5
5
4
their electronic structure by quantum chemical methods.The
crystal structure analyses of 1 and 2 provide a link between
experimental results and the electronic configuration of these
borole species.
[
6]
energy. This result has been authenticated by numerous
quantum chemical calculations, which predict perfect D5h
symmetry for the triplet ground state and a Jahn–Teller
[7]
distortion for the excited singlet states. In addition, Schleyer
and co-workers have demonstrated that the delocalization of
the four p electrons in the triplet species results in an
aromatic stabilization of the entire molecule, whereas the
electronic singlet states are destabilized owing to antiaroma-
Compounds 1 and 2 were prepared by boron–tin exchange
reactions of the stannole precursor 3 and stoichiometric
[
2c,11]
amounts of the appropriate boron dihalides (Scheme 1).
[7f]
ticity.
However, all attempts to isolate or structurally
[*] Prof. Dr. H. Braunschweig, Dr. T. Kupfer
Institut für Anorganische Chemie
Julius-Maximilians-Universität Würzburg
Am Hubland, 97074 Würzburg (Germany)
Fax: (+49)931-888-4623
Scheme 1. Syntheses of 1 and 2.
E-mail: h.braunschweig@mail.uni-wuerzburg.de
Dr. I. Fernµndez, Prof. Dr. G. Frenking
Fachbereich Chemie
Philipps-Universität Marburg
Hans-Meerwein-Strasse, 35043 Marburg (Germany)
Fax: (+49)6421-282-5566
E-mail: frenking@chemie.uni-marburg.de
Crystallization of 1 from a saturated CH Cl2 solution at
2
ꢀ358C yielded deep blue needles that were suitable for X-ray
[
12,13]
diffraction analysis (Figure 1).
The solid-state structure of
1
is characterized by a planar BC five-membered ring (root-
4
mean-square deviation: 0.0175 Š) with internal ring dihedral
angles of 4.1(1), ꢀ3.7(2), 1.7(2), 1.02(2), and ꢀ3.1(1)8, as well
as a propeller-like arrangement of the five phenyl substituents
[
**] T.K. thanks the FCI for a PhD fellowship. I.F. thanks the MEC (Spain)
for a postdoctoral grant. We are grateful to Prof. Martin Bröring and
Dr. Olaf Burghaus (Marburg) for performing the ESR and SQUID
measurements.
[
for example, C1-B1-C51-C52: ꢀ32.9(2)8 (ꢀ35.18); B1-C4-
C41-C46: ꢀ51.3(2)8 (ꢀ51.38), C2-C3-C31-C32: ꢀ52.1(2)8
ꢀ49.98)]. The bond lengths within the central ring
moiety are alternating, but the alternations are much less
[
14]
(
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
Angew. Chem. Int. Ed. 2008, 47, 1951 –1954
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1951

Communications
Figure 1. Molecular structure of 1. Selected bond lengths [Š] and
angles [8] (calculated values at the BP86/def2-SVP level are given in
italics): B1-C1 1.526(2) (1.595), B1-C4 1.539(2) (1.595), C1-C2 1.428(2)
Figure 2. Molecular structure of 2. Only one molecule of the asymmet-
ric unit is shown for clarity. Selected bond lengths [Š] and angles [8]
(
1.380), C2-C3 1.470(2) (1.533), C3-C4 1.426(2) (1.380); C1-B1-C4
05.4(1) (105.2), B1-C1-C2 107.5(1) (106.1), C1-C2-C3 109.9(1)
111.3), C2-C3-C4 109.8(1) (111.3), C3-C4-B1 107.3(1) (106.1).
(calculated values at the BP86/def2-SVP level are given in italics): B1-
1
(
C1 1.597(3) (1.600), B1-C4 1.582(3) (1.599), C1-C2 1.358(3) (1.379),
C2-C3 1.518(3) (1.525), C3-C4 1.353(3) (1.381), Fe1-CCp 2.021(2)–
2.068(2) (2.040–2.072), Fe1-B1 2.664 (2.825); C1-B1-C4 103.6(2)
(
104.5), B1-C1-C2 106.4(2) (106.5), C1-C2-C3 111.8(2) (111.2), C2-C3-
pronounced than predicted by theoretical calculations on the
C4 110.5(2) (111.4), C3-C4-B1 107.8(2) (106.3).
[
9]
parent compound and on 1.The B ꢀC bond lengths [between
1
.516(2) and 1.539(2) (1.595) Š] are significantly shortened
with respect to the typical value of a boron–carbon single
typical boron–carbon single bond (1.61 Š). Additionally, the
carbon–carbon bond lengths [C1ꢀC2: 1.358(3) (1.379) Š; C3ꢀ
C4: 1.353(3) (1.381) Š; C2ꢀC3: 1.518(3) (1.525) Š] deviate
bond (1.61 Š). However, the lengths for the C1ꢀC2 [1.428(2)
(
1.380) Š] and C3ꢀC4 bonds [1.426(2) (1.380) Š] are signifi-
cantly elongated compared to the expected value of a carbon–
carbon double bond (1.32 Š), but strongly resemble those
found in delocalized ring systems (1.40 Š). Additionally, the
formal carbon–carbon single bond [C2ꢀC3 1.470(2)
substantially from the corresponding values for 1.Whereas
the first two bonds are only slightly longer than a typical CꢀC
double bond (1.32 Š), the values for the latter bond lie
3
between those of a single bond between sp -hybridized
(
1.533) Š] is somewhat shorter than an isolated single bond
carbon atoms (1.54 Š) and a single bond between uncon-
2
2
(
1.54 Š) or a single bond between unconjugated sp -hybri-
jugated sp -hybridized carbon atoms (1.49 Š). Similiar bond
dized carbon atoms (1.49 Š). The experimental values for the
lengths have been observed in cis,cis-1,2,3,4-tetraphenylbuta-
[
19]
bond lengths suggest a significant p –p* conjugation between
1,3-diene (1.356 and 1.484 Š); thus, the bonding situation is
best described as an isolated diene system that is bridged by a
boron center.
p
2
the unsaturated sp -hybridized boron center and the carbon
backbone, which is surprising because the borole is an
antiaromatic species with four p electrons.Investigations on
the spin multiplicity of 1 by variable-temperature NMR and
ESR spectroscopy, as well as magnetic SQUID measure-
ments, revealed no evidence for paramagnetic contributions
and, consequently, a significant population of a triplet state
can be excluded.
We calculated the geometries of 1 and 2 in the singlet state
at the BP86/def2-SVP level to analyze the structures and
[
20]
bonding situation of the molecules. Relevant geometrical
data are given in the captions to Figures 1 and 2.The most
striking aspect of the data for 1 is the large difference between
the theoretical and experimental values for the bond lengths
in the borole ring.The calculations give much longer C ꢀB
The antiaromaticity in borole systems is manifested by the
high Lewis acidity of the boron center that has enabled, for
instance, the application of borole derivatives as potent Lewis
bonds (1.595 Š) than the experiment (1.526–1.539 Š), and the
theoretical CꢀC bond alternations are much more pro-
[
15]
acids in the polymerization of ethylene. Here, the Lewis
acidity of the boron center is convincingly demonstrated by
nounced (1.380 Š for the short bonds and 1.533 Š for the
long bonds) than experimentally determined (1.426–1.428 Š
for the short bonds and 1.470 Š for the long bonds). We
optimized the geometry of 1 using B3LYP/def2-SVP and RI-
MP2/def2-SVP to check whether the discrepancy comes from
a failure of the BP86 method.However, all three methods
[
12,16]
the solid-state structure of ferrocenylborole (2, Figure 2),
which allows for the direct observation of a strong FeꢀB
interaction.The pronounced bending of the boryl ligand
towards the iron center with a dip angle a* = 29.4 (23.2)8 is
significantly increased in comparison to other ferrocenylbor-
[
21]
gave very similar bond lengths for the borole ring. We also
optimized 1 in the triplet state.The calculated geometry at the
BP86/def2-SVP level indicates stronger conjugation in the
borole ring, and the calculated bond lengths (B1ꢀC1/C4:
1.595 Š; C1ꢀC2: 1.459 Š; C2ꢀC3: 1.431 Š) appear to be in
[17]
anes.
As a consequence of this interaction, the p –p*
p
conjugation in the borole subunit is at least partially
interrupted in comparison to 1.Hence, the antiaromatic
character of 2 seems to be less pronounced, which is further
[
11,18]
confirmed by UV/Vis spectroscopy.
These findings are
better agreement with experiment than the theoretical values
for the singlet state.However, the calculations predict the
opposite order for the CꢀC bond lengths, that is, C1ꢀC2 >
C2ꢀC3 while the experimental data shown in Figure 1 give
C1ꢀC2 < C2ꢀC3.Moreover, the triplet state of 1 at the BP86/
also supported by the bond lengths within the BC4 five-
membered ring.The B ꢀC bond lengths [B1ꢀC4: 1.582(3)
(
1.599) Š; B1ꢀC1: 1.597(3) (1.600) Š] are notably elongated
compared to those found in 1 and equal to the value of a
1
952
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1951 –1954

Angewandte
Chemie
ꢀ
1
def2-SVP level is 14.1 kcalmol higher in energy than the
singlet state.The theoretical data thus support the exper-
imental results, which suggest that 1 has a singlet state.
Reexamination of the crystal structure helped to solve the
puzzling discrepancy between theory and experiment.Inspec-
tion of the X-ray data revealed that there are dimeric subunits
of 1 with noticeably short BꢀC and BꢀH separations between
measured in the solid state (1.638 Š) is more than 0.8 Š
shorter than in the gas phase (2.473 Š).
[
24]
Unlike for compound 1, the theoretical and experimental
data for the borole ring of 2 are in very good agreement
(Figure 2).The calculations suggest that the B ꢀC and CꢀC
bond lengths in 2 are nearly the same as in 1, while the
experimental data indicate a significant difference.The other
calculated data of 2 also agree quite well with experiment
except for the FeꢀB bond length (2.825 Š), which is clearly
the dimers.The boron phenyl substituent of each borole lies
above or below the boron atom of the other borole, which
allows for intermolecular phenyl!boron p donation.The
shortest intermolecular BꢀC separation in the dimer is only
longer than the measured value (2.664 Š). The iron–boron
donor–acceptor bond should not be very strong, and thus, it
may become significantly shorter in the solid state, which is in
agreement with a previous study about the length of donor–
acceptor bonds in the gas phase and in the condensed
3
.635 Š, but even the intermolecular BꢀC separations
between the boron atom and a C(phenyl) atom of adjacent
dimers of the borole are only 3.855 Š (see the Supporting
Information).This finding indicates that the electron-defi-
cient boron atom in free 1 receives electronic charge through
intermolecular p donation.
[23]
phase.
In this contribution we have reported on the determina-
tion of the solid-state structure of pentaphenylborole (1).The
structural parameters of this antiaromatic heterocycle with
four p electrons differ significantly from those obtained by
To estimate the effect of the intermolecular interactions
on the bond lengths of 1, we tried to optimize a dimer (1) at
2
the MP2 level, which is necessary for an adequate treatment
of the intermolecular attraction.A partial optimization
showed that the bond lengths in the borole ring change
from the values in free 1 to values that approach the X-ray
data.We then optimized the geometry of the complex of 1
with the weak electron donor CO to mimic the effect of
charge donation to the borole.The optimized complex 1·CO
has a BꢀCO bond length of 1.550 Š and a bond dissociation
quantum chemical calculations and suggest an unexpectedly
2
strong p –p* conjugation between the unsaturated sp -
p
hybridized boron center and the carbon backbone.Reexami-
nation of the crystal packing revealed distinct borole dimers
with short BꢀC intermolecular contacts.According to quan-
tum chemical studies the discrepancies between the exper-
imentally and theoretically determined bond lengths can be
attributed to the increase of electron density at the electron-
deficient boron atom in free 1 through intermolecular
p donation from adjacent phenyl groups.
ꢀ1
energy of D = 17.3 kcalmol
(BP86/def2-SVP).These
o
values suggest a slightly longer and weaker bond than in
H B·CO (experimental values: BꢀCO: 1.534 Š, D =
3
o
ꢀ
1
[22]
2
4.6 kcalmol ).
The impact of the BꢀCO bond on the
Received: October 15, 2007
Revised: November 22, 2007
Published online: January 28, 2001
bond lengths in the borole ring is very large.The calculated
bond lengths of 1·CO are 1.644/1.649 Š for B1ꢀC1/C4,
1
.389 Š for C1ꢀC2, and 1.490 Š for C2ꢀC3.The latter
Keywords: antiaromaticity · boroles · boron ·
bond, which is rather distant from the B-CO moiety, is
predicted to be much shorter than the calculated value for the
free borole (1.533 Š), while its length is very similar to the
experimental value of 1 (1.470 Š).
.
density functional calculations · intermolecular interactions
We also calculated 1 at the BP86/def2-SVP level with
fixed CꢀC and CꢀB bond lengths for the borole ring (taken
[1] a) K.Krogh-Jespersen, D.Cremer, JD. .Dill, JA. .Pople,
P.von R. Schleyer, J. Am. Chem. Soc. 1981, 103, 2589 – 2594;
b) S.M. van der Kerk, P.H.M. Budzelaar, A. van der Kerk-
from the X-ray data), while the rest of the molecule was
completely optimized.The energy difference between the
latter species and the completely optimized compound is only
van Hoof, G.J.M. van der Kerk, P.von R. Schleyer,
Angew.
Chem. 1983, 95, 61; Angew. Chem. Int. Ed. Engl. 1983, 22, 48;
c) C.Pues, A.Berndt, Angew. Chem. 1984, 96, 306 – 307; Angew.
Chem. Int. Ed. Engl. 1984, 23, 313 – 314; d) J.J.Eisch, B.Shafii,
A.L. Rheingold, J. Am. Chem. Soc. 1987, 109, 2526 – 2528;
ꢀ1
ꢀ1
5
.2 kcalmol at the BP86/def2-SVP level (5.0 kcalmol at
RI-MP2/def2-SVP).This energy may easily be provided by
intermolecular interactions between the phenyl substituents
of the borole, which serve as p donors, and the electron-
deficient boron atoms.Thus, the discrepancy between theory
and experiment is not a sign for a shortcoming of either
method.It is rather an interesting finding which shows that
already rather weak intermolecular interactions may signifi-
cantly alter the bond lengths in the borole ring.We predict
that the geometry of 1 in the gas phase is quite different from
the geometry in the solid state, and that the gas-phase values
will exhibit a much larger bond alternation in the borole ring.
Large differences between bond lengths in the gas phase and
in the solid state have been noted before, but they were
e) J.J. Eisch, B. Shafii, J.D. Odom, A.L. Rheingold,
J. Am.
Chem. Soc. 1990, 112, 1847 – 1853, and references therein; f) H.
Braunschweig, I.Fernµndez, G.Frenking, K.Radacki, F.Seeler,
Angew. Chem. 2007, 119, 5307 – 5310; Angew. Chem. Int. Ed.
2007, 46, 5215 – 5218; g) H.Braunschweig, T.Herbst, D.Rais, F.
Seeler, Angew. Chem. 2005, 117, 7627 – 7629; Angew. Chem. Int.
Ed. 2005, 44, 7461 – 7463.
[2] a) J.J.Eisch, N.K.Hota, S.J.Kozima,
J. Am. Chem. Soc. 1969,
91, 4575 – 4576; b) G.E. Herberich, B. Buller, B. Hessner, W.
Oschmann, J. Organomet. Chem. 1980, 195, 253 – 259; c) J.J.
Eisch, J.E.Galle, S.Kozima, J. Am. Chem. Soc. 1986, 108, 379 –
385; d) P .J .Fagan, E .G .Burns, J .C .Calabrese,
J. Am. Chem.
Soc. 1988, 110, 2979 – 2981; e) J .J .Eisch, J .E .Galle, B.Shafil,
A.L. Rheingold, Organometallics 1990, 9, 2342 – 2349; f) P.J.
Fagan, W.A.Nugent, J.C.Calabrese, J. Am. Chem. Soc. 1994,
[
23]
confined to bond lengths of donor–acceptor complexes. For
example, the nitrogen–boron bond length in HCN·BF3
Angew. Chem. Int. Ed. 2008, 47, 1951 –1954
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
1953

Communications
1
16, 1880 – 1889; g) S.Kim, K.Song, S.Ook, J.Ko,
Chem.
[13] The crystallographic data set was of excellent quality and did not
show any evidence for a disorder of 1, regardless of the high
symmetry.Considering the difficulties associated with the crystal
structure of trimesitylborirene MesBC Mes (Mes = 2,4,6-
Commun. 2004, 68 – 69.
[
3] a) J.J.Eisch, J.E.Galle, J. Am. Chem. Soc. 1975, 97, 4436 – 4437;
b) J.J.Eisch, J.E.Galle, J. Organomet. Chem. 1976, 127, C9 –
C13; c) A.J.Ashe III, F.J.Drone, J. Am. Chem. Soc. 1987, 109,
2
2
[
1d]
Me C H ), in our hands the assignment of boron and carbon
3
6
2
1
879 – 1880; d) Y.Sugihara, T.Yagi, I.Murata, A.Imamura,
Am. Chem. Soc. 1992, 114, 1479 – 1481; e) A.J.Ashe III, J.W.
Kampf, W.Klein, R.Rousseau, Angew. Chem. 1993, 105, 1112 –
113; Angew. Chem. Int. Ed. Engl. 1993, 32, 1065 – 1066; f) J.
J.
atoms of the central ring moiety in the structure of 1 was
unambiguously derived from the observed bond lengths and
angles, as well as from the thermal ellipsoids, which were found
to be almost isotropic.
1
Schulman, R.L. Disch, Organometallics 2000, 19, 2932 – 2936.
4] M.Saunders, R.Berger, A.Jaffe, J.M.McBride, J.OꢀNeill, R.
Breslow, J.M.Hoffman Jr., C.Perchonock, E.Wasserman, R.S.
Hutton, V.J.Kuck, J. Am. Chem. Soc. 1973, 95, 3017 – 3018.
5] a) H.J. Wörner, F. Merkt, Angew. Chem. 2006, 118, 299 – 302;
Angew. Chem. Int. Ed. 2006, 45, 293 – 296; b) H.J. Wörner, F.
Merkt, J. Chem. Phys. 2007, 127, 034303.
[14] The values in italics are derived from theoretical calculations,
which are discussed below.
[
[15] P.A.Chase, W.E.Piers, B.O.Patrick,
J. Am. Chem. Soc. 2000,
122, 12911 – 12912.
[
[16] Compound 2 crystallizes in the monoclinic space group P2 with
1
two independent molecules in the asymmetric unit; one moiety
exhibits a weak disorder within a carbon-bound phenyl group.
Hence, for simplicity reasons, only the structure of the non-
disordered molecule is discussed.
[
6] E.Wasserman, R.S.Hutton, Acc. Chem. Res. 1977, 10, 27 – 32.
[
7] a) M .J .S .Dewar, R .C .Haddon,
J. Am. Chem. Soc. 1973, 95,
5
836 – 5837; b) J.Feng, J.Leszczynski, B.Weiner, M.C.Zerner,
[17] So far, the most noticeable interactions have been reported by
Jäkle et al.for an oxidized diboradiferrocene ( a* = 22.68: K.
J. Am. Chem. Soc. 1989, 111, 4648 – 4655; c) M.N.Glukhovtsev,
B.Reindl, P.von R.Schleyer, Mendeleev Commun. 1993, 3, 100 –
Venkatasubbaiah, I.Nowik, R H. .Herber, F.Jäkle,
Chem.
1
12; d) P.von R. Schleyer, C. Märker, A. Dransfeld, H. Jiao,
N.J.R.Eikema Hommes, J. Am. Chem. Soc. 1996, 118, 6317 –
318; e) H.Jiao, P .v on R.Schleyer, Y.Mo, M .A .McAllister,
T.T. Tidwell, J. Am. Chem. Soc. 1997, 119, 7075 – 7083; f) V.
Gogonea, P.von R. Schleyer, P.R. Schreiner, Angew. Chem.
998, 110, 2045 – 2049; Angew. Chem. Int. Ed. 1998, 37, 1945 –
948; g) K.B. Wiberg, Chem. Rev. 2001, 101, 1317 – 1331;
Commun. 2007, 2154 – 2156) and by M.Wagner et al.for FcBBr 2
(a* = 18.98 and 17.78: M.Scheibitz, M.Bolte, J.W.Bats, H -. W.
Lerner, I.Nowik, R H. .Herber, A.Krapp, M.Lein, M C. .
6
Holthausen, M.Wagner, Chem. Eur. J. 2005, 11, 584 – 603).
[18] Whereas the delocalization of the p electrons in 1 is supported
ꢀ1
ꢀ1
1
1
by a characteristic visible band at 561 nm (e = 361 Lmol cm ),
the corresponding absorption in 2 is found significantly blue-
ꢀ
1
ꢀ1
h) A.D.Allen, T.T.Tidwell, Chem. Rev. 2001, 101, 1333 – 1348.
8] Examples: a) R.Breslow, H.W.Chang, R.Hill, E.Wasserman, J.
Am. Chem. Soc. 1967, 89, 1112 – 1119; b) J.B.Lambert, L.Lin, V.
Rassolov, Angew. Chem. 2002, 114, 1487 – 1489; Angew. Chem.
Int. Ed. 2002, 41, 1429 – 1431; c) M.Otto, D.Scheschkewitz, T.
Kato, M.M.Midland, J.B.Lambert, G.Bertrand, Angew. Chem.
shifted at 390 nm (e = 2175 Lmol cm ), a region that has been
[
1c]
[
described for the related base adducts.
[19] J.W.Bats, B.Urschel, Acta Crystallogr. Sect. E 2006, 62, 748 –
750.
[20] All calculations were carried out at the BP86 level [a) A.D.
Becke, Phys. Rev. A 1988, 38, 3098 – 3100; b) J.P.Perdew, Phys.
Rev. B 1986, 33, 8822 – 8824] using the def2-SVP basis set (F.
Weigend, R.Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297 –
3305) with the Gaussian03 (rev.D 0. 1) suite of programs (M.J.
Frisch et al.; full reference is given in the Supporting
Information).
[21] Selected bond lengths [Š]: B3LYP/def2-SVP level: B1-C1/C4
1.588, C1-C2 1.363, C2-C3 1.533; RI-MP2/def2-SVP level: B1-
C1/C4 1.587, C1-C2 1.374, C2-C3 1.516. The latter calculation
was carried out using the TURBOMOLE (v5.90) program: R.
Ahlrichs, M.Baer, M.Haeser, H.Horn, C.Koelmel, Chem. Phys.
Lett. 1989, 162, 165 – 169.
2
2
002, 114, 2379 – 2380; Angew. Chem. Int. Ed. 2002, 41, 2275 –
276.
[
9] a) E.J.P. Malar, K. Jug,
Tetrahedron 1986, 42, 417 – 426;
b) P.von R. Schleyer, P.K. Freeman, H. Jiao, B. Goldfuss,
Angew. Chem. 1995, 107, 332 – 335; Angew. Chem. Int. Ed.
Engl. 1995, 34, 337 – 340; c) M.K. Cyranski, T.M. Krygowski,
A.R. Katritzky, P.von R. Schleyer,
333 – 1338.
10] G.E.Herberich, U.Englert, M.Hostalek, R.Laven, Chem. Ber.
991, 124, 17 – 23.
J. Org. Chem. 2002, 67,
1
[
[
1
11] The experimental section including the syntheses, full character-
ization, and spectroscopic data of all compounds, as well as the
UV/Vis spectra can be found in the Supporting Information.
12] Experimental details of all X-ray crystal structure determina-
tions can be found in the Supporting Information.CCDC-650933
[22] a) A.C. Venkatachar, R.C. Taylor, R.L. Kuczkowski,
J. Mol.
Struct. 1977, 38, 17 – 23; b) S.G. Lias, J.E. Bartmess, J.F.
[
Liebman, J.L. Holmes, R.D. Levin, W.G. Mallard,
Chem. Ref. Data 1988, 17, Suppl.1.
J. Phys.
(
1) and 650934 (2) contain the supplementary crystallographic
[23] V.Jonas, G.Frenking, M.T.Reetz, J. Am. Chem. Soc. 1994, 116,
8741 – 8753.
data for this paper.These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
[24] W.A. Burns, K.R. Leopold,
11622 – 11623.
J. Am. Chem. Soc. 1993, 115,
1
954
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 1951 –1954