Main-chain boron-containing oligophenylenes via ring-opening polymerization of 9-H-9-Borafluorene

Article abstract of DOI:10.1021/ja110947k

9-H-9-Borafluorene (H₈C₁₂BH; 5) can be generated in situ from 9-Br-9-borafluorene and Et₃SiH in benzene or hexane. Monitoring of the reaction by NMR spectroscopy at rt in C₆D ₆ reveals that 5 forms C₁-symmetric dimers (5)₂ under these conditions. DFT calculations on conceivable isomers of (5) ₂ and a comparison of calculated and experimentally determined ¹H, ¹³C, and ¹¹B NMR shift values lead to the conclusion that (5)₂ is not a classical dimer H₈C ₁₂B(μ-H)₂BC₁₂H₈, but contains one B-H-B three-center, two-electron bond together with a boron-bridging phenyl ring. Addition of 1 equiv of pyridine to (5)₂ leads to the clean formation of the pyridine adduct H₈C₁₂BH(py) (5?py). Likewise, (5)₂ can be employed in hydroboration reactions, as evidenced by its transformation with 0.5 equiv of tert-butylacetylene, which gives the hydroboration products tBuC(H)₂C(H)(BC₁₂H ₈)₂ (9) and tBuC(H)C(H)BC₁₂H₈ in almost quantitative yield. (5)₂ is not long-term stable in benzene solution. Addition of pyridine to aged reaction mixtures allowed the isolation of the adduct (py)H₂B-C₆H₄-C₆H ₄-(py)BC₁₂H₈ (10?py₂) of a ring-opened dimer of 5. Storage of a hexane solution of 9-Br-9-borafluorene and Et₃SiH for 1-2 weeks at rt leads to the formation of crystals of a ring-opened pentamer H[-(H)B-(C₆H₄)₂-] ₄BC₁₂H₈ (11) of 5 (preparative yields are obtained after 1-4 months). The polymer main chain of 11 is reinforced by four intrastrand B-H-B three-center, two-electron bonds. Apart from the main product 11, we have also isolated minor amounts of closely related oligomers carrying different chain ends, i.e., H₈C₁₂B-(C₆H ₄)₂[-(H)B-(C₆H₄)₂-] ₂BC₁₂H₈ (12) and H[-(H)B-(C₆H ₄)₂-]₅BH₂ (13). When the reaction between 9-Br-9-borafluorene and Et₃SiH is carried out in refluxing toluene, the cyclic dimer [-(μ-H)B-(C₆H₄) ₂-]₂ (14) can be obtained in a crystalline yield of 25%. The compounds 9, 10?py₂, 11, 12, 13, and 14 have been structurally characterized by X-ray crystallography. The entire reaction pathway leading from 5 to 10, 11, 12, 13, and 14 has been thoroughly elucidated by DFT calculations and we propose a general mechanistic scenario applicable for ring-opening polymerization reactions of 9-borafluorenes.

Full text of DOI:10.1021/ja110947k

ARTICLE
pubs.acs.org/JACS
Main-Chain Boron-Containing Oligophenylenes via Ring-Opening
Polymerization of 9-H-9-Borafluorene
Alexander H€ubner,† Zheng-Wang Qu,† Ulli Englert,‡ Michael Bolte,† Hans-Wolfram Lerner,†
Max C. Holthausen,*,† and Matthias Wagner*,†
†Institut f€ur Anorganische und Analytische Chemie, Johann Wolfgang Goethe-Universit€at Frankfurt,
Max-von-Laue-Strasse 7, D-60438 Frankfurt am Main, Germany.
‡Department of Inorganic Chemistry, RWTH Aachen, Landoltweg 1, D-52056 Aachen.
S Supporting Information
b
ABSTRACT: 9-H-9-Borafluorene (H8C12BH; 5) can be gen-
erated in situ from 9-Br-9-borafluorene and Et3SiH in benzene
or hexane. Monitoring of the reaction by NMR spectroscopy at
rt in C6D6 reveals that 5 forms C1-symmetric dimers (5)2 under
these conditions. DFT calculations on conceivable isomers of
(5)2 and a comparison of calculated and experimentally
determined 1H, 13C, and 11B NMR shift values lead to the conclusion that (5)2 is not a classical dimer H8C12B(μ-H)2BC12H8,
but contains one B-H-B three-center, two-electron bond together with a boron-bridging phenyl ring. Addition of 1 equiv of
pyridine to (5)2 leads to the clean formation of the pyridine adduct H8C12BH(py) (5 py). Likewise, (5)2 can be employed in
3
hydroboration reactions, as evidenced by its transformation with 0.5 equiv of tert-butylacetylene, which gives the hydroboration
products tBuC(H)2C(H)(BC12H8)2 (9) and tBuC(H)C(H)BC12H8 in almost quantitative yield. (5)2 is not long-term stable in
benzene solution. Addition of pyridine to aged reaction mixtures allowed the isolation of the adduct (py)H2B-C6H4-
C6H4-(py)BC12H8 (10 py2) of a ring-opened dimer of 5. Storage of a hexane solution of 9-Br-9-borafluorene and Et3SiH for
3
1-2 weeks at rt leads to the formation of crystals of a ring-opened pentamer H[-(H)B-(C6H4)2-]4BC12H8 (11) of 5 (preparative
yields are obtained after 1-4 months). The polymer main chain of 11 is reinforced by four intrastrand B-H-B three-center, two-
electron bonds. Apart from the main product 11, we have also isolated minor amounts of closely related oligomers carrying different
chain ends, i.e., H8C12B-(C6H4)2[-(H)B-(C6H4)2-]2BC12H8 (12) and H[-(H)B-(C6H4)2-]5BH2 (13). When the reaction between
9-Br-9-borafluorene and Et3SiH is carried out in refluxing toluene, the cyclic dimer [-(μ-H)B-(C6H4)2-]2 (14) can be obtained in a
crystalline yield of 25%. The compounds 9, 10 py2, 11, 12, 13, and 14 have been structurally characterized by X-ray crystallography.
3
The entire reaction pathway leading from 5 to 10, 11, 12, 13, and 14 has been thoroughly elucidated by DFT calculations and we
propose a general mechanistic scenario applicable for ring-opening polymerization reactions of 9-borafluorenes.
’ INTRODUCTION
boron-doped π-systems are relevant for a broad range of
applications as, e.g., luminophores and hole transporters in
organic light-emitting devices, as components of organic solar
cells, and as sensors for the detection of biologically or envir-
onmentally relevant anions.2,5-8
For the design of boron-containing π-conjugated polymers,
two cases have to be distinguished, i.e., the boron atoms can
either be introduced as lateral substituents or are incorporated as
integral parts of the polymer main chain.9,10 The dopant influ-
ence on the overall optoelectronic properties is maximized in the
latter class of compounds but only a few synthetic strategies exist
to prepare main-chain boron-containing π-conjugated systems.
The most prominent are polycondensation approaches11-18 and
the hydroboration polymerization.19-25
Incorporation of the main group elements B, Si, N, P, or S into
organic π-electron materials is a powerful approach to modifying
the optical and electronic properties of the parent systems.1 The
consequent notion that classical inorganic solid-state materials
can be replaced by lightweight and highly tunable organo-
element compounds sparked a vibrant interest in this field of
research.2 In this context, boron has lately moved to the center
stage of elements used as dopants for extended π-conjugated
organic frameworks. Boron atoms can efficiently interact with
organic π-systems through their vacant p-orbitals,3,4 and as a
result, the electronic structure of the material is modified with an
impact on its optical and charge transport properties. Moreover,
organoboranes possess a high Lewis acidity and can therefore
engage in Lewis acid-base reactions. As a result of adduct
formation, the local geometry about the boron center changes
from trigonal-planar to tetrahedral so that π-conjugation path-
ways across the boron p-orbital are disrupted, which can
be exploited for the design of molecular switches. Today,
Our group has recently reported on a novel polyconden-
sation protocol, which takes advantage of the fact that
Received: December 6, 2010
Published: March 02, 2011
r
2011 American Chemical Society
4596
dx.doi.org/10.1021/ja110947k J. Am. Chem. Soc. 2011, 133, 4596–4609
|

Journal of the American Chemical Society
ARTICLE
Scheme 1. π-Conjugated Polymers Synthesized by Polycon-
densation (2) and by Hydroboration Polymerization (4)
Figure 1. 9-H-9-Borafluorene (5) and the insertion products 6 and 7 of
BH3 and BH2Et into one B-C bond of 5 and 9-Et-9-borafluorene;
numbering scheme used for the assignment of NMR resonances.
syntheses. In the absence of a suitable trapping reagent (e.g., a
Lewis base or an unsaturated organic molecule), however, 5
undergoes ring-opening polymerization and the corresponding
pentamer has been characterized by X-ray crystallography. More-
over, we provide a thorough DFT study on the polymerization
mechanism and we elucidate the chain-termination processes
both by experimental and by theoretical means.
(organyl)(hydrido)boranes tend to undergo substituent scram-
bling reactions. In the particular case of 1,10-bis(dibro-
moboryl)ferrocene (1; Scheme 1), a partial Br/H exchange with
Et3SiH induces the formation of the boron-bridged poly-
(ferrocenylene) 2 with concomitant liberation of diborane(6)
(Scheme 1).12,14,17 In another approach, we have employed the
novel ditopic borane 9,10-dihydro-9,10-diboraanthracene [3]n
for the hydroboration polymerization of aromatic dialkynes,
which leads to highly photoluminescent materials.23-25
On the basis of these previous investigations, we have now
studied the reactivity of 9-H-9-borafluorene (5; Figure 1). 5
offers unique opportunities for the preparation of various new
classes of boron-doped π-electron materials, because the boron
atom is part of a formally antiaromatic central borole ring.26,27
Consequently, 9-borafluorenes are stronger Lewis acids than
acyclic analogs,28,29 and they are readily reducible to their
corresponding dianions.29-31 These pronounced σ- and
π-acceptor properties render 9-borafluorenes highly interesting
building blocks for anion sensors and electron-transporting
(n-type) materials.32
’ RESULTS AND DISCUSSION
In situ Generation and a Hydroboration Reaction of 9-H-
9-Borafluorene. The parent compound 9-H-9-borafluorene
(5; Figure 1, Scheme 2) has so far remained elusive and was
not available as a hydroboration reagent.35 However, its pyridine
adduct 5 py as well as lithium 9,9-dihydrido-9-boratafluorene
3
(which can be regarded as the hydride adduct Li[5 H]) are
3
literature-known: 5 py has been prepared by pyrolysis of pyr-
3
idine-2-biphenylylborane in liquid paraffin at 250 °C,36 and
Li[5 H] is accessible from 9-Cl-9-borafluorene and LiH in
3
THF/pentane.37
The major reason for the instability of free 5 is most likely a
consequence of the 4π antiaromatic nature of the central borole
ring, because the propensity to exclude the boron p-orbital from
the cyclic conjugation pathway increases the boron Lewis
acidity. In contrast to isolable derivatives of 5, like 9-Ph-9-
borafluorene38,39 or 9-OMe-9-borafluorene,39 the parent com-
pound is neither stabilized kinetically nor by virtue of a
π-donating exocyclic substituent.
9-H,10-H-9,10-dihydro-9,10-diboraanthracene (3; Scheme 1)
can be prepared in high yields using 9-Br,10-Br-9,10-dihydro-
9,10-diboraanthracene and Et3SiH.23 The compound is stable
under inert conditions even though it contains a formally
antiaromatic central ring with an electronic structure comparable
to that of the borole ring in 9-H-9-borafluorene (5). Given
this background, we investigated also the reaction between
9-Br-9-borafluorene (8; Scheme 2)40-42 and Et3SiH in greater
detail.
So far, all experimental and theoretical work on the incorpora-
tion of 9-borafluorene moieties into larger π-systems has focused
on the functionalization of their phenylene rings. A yet unex-
plored complementary concept would be to use the reactive
B-H bond in 5 (i) to attach lateral 9-borafluorenyl units to
conjugated polymer backbones via the hydroboration of pending
alkynyl substituents or (ii) to synthesize well-defined short-chain
4-type oligomers via the hydroboration of dialkynylarenes. Yet an
alternative option is the ring-opening polymerization of 5, which
should lead to 1,1-boranediyl-bridged biphenylenes [-B(H)-
C6H4-C6H4-]n. A driving force for such a polymerization se-
quence should be provided by the loss of antiaromaticity in the
monomer. One could envisage substituent scrambling as the
underlying reaction mechanism, similar to the reaction leading
from 1 to 2 (Scheme 1). In this context we note that the B-C
bonds in 9-borafluorene are known to insert BH3 or BH2Et
under formation of 633 or 734 (Figure 1), respectively, which is
precisely the kind of reactivity required for a ring-opening
polymerization of 5.
1H and 13C{1H} NMR spectroscopy on a mixture of 8 and
Et3SiH (C6D6, rt) reproducibly revealed a quantitative decrease
of the reactant signals and a concomitant appearance of a
complex pattern of well-resolved 1H and 13C resonances which
correspond to one single primary product. We note that the
conversion rate strongly depends on the absolute concentration
and the stoichiometric ratio of the starting materials. If 8 and
Et3SiH are employed in a 1:1 ratio, it takes several hours until the
starting material is no longer detectable by NMR spectroscopy;
used in a 1:2 ratio, full conversion is achievable within minutes.
Since the primary product is subject to a slow follow-up reaction,
we found it most convenient to prepare it by the use of 2 equiv of
Et3SiH, because in this case we can reach a stage at which 8 has
In this paper, we make use of both, the B-H bond and the
B-C bond reactivity of 5. We will show that this compound can
be generated in situ and used immediately for hydroboration
4597
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
completely vanished while almost no products other than
the primary product have formed. This critical point of clean
transformation is clearly indicated by a color change from yellow-
green to almost colorless.
Scheme 2. Synthesis of 5 and Trapping Reactions with Pyr-
idine (py) and tBuCCH at Different Points in Timea
The 11B{1H} NMR spectrum of the primary product shows
two broad resonances at 8.5 ppm and 10.3 ppm, testifying to the
presence of four-coordinate boron nuclei;43,44 no multiplets can
be resolved upon proton coupling, but the signal at 10.3 ppm
becomes significantly broader in the 11B NMR spectrum.
A convincing proposal for the molecular structure of the
primary product was possible only after a quantum chemical
assessment of the initial stages of the reaction cascade, which
1
allowed us to compute H, 11B, and 13C shift values of all
conceivable candidates. Comparison to the experimental data
led to the structural assignment presented in the corresponding
quantum chemical subsection below. At this point, it suffices to
state that (i) the observed NMR signature of Et3SiBr in the
reaction mixture indicates hydride transfer from silicon to boron,
(ii) the 1H and 13C signal pattern is in accord with the formation
of a C1-symmetric dimer (5)2, which does, however, certainly not
possess the classical structure H8C12B(μ-H)2BC12H8 (cf. [3]n;
Scheme 1).
To investigate the reactivity of (5)2, we repeated the transfor-
mation between 9-Br-9-borafluorene (8) and Et3SiH in an NMR
tube under the same conditions as before (stoichiometric ratio
=1:2) and quenched it at the stage of (5)2 with 1 equiv of pyridine
(experiment 1) or 0.5 equiv of tert-butylacetylene (tBuCCH;
experiment 2). In both cases, an almost quantitative conversion
was finished within minutes. Experiment 1 gave exclusively the
known pyridine adduct 5 py36 (Scheme 2; see the Supporting
3
Information (SI) for the full NMR data set). Experiment 2
resulted in the formation of the diborylmethane 9 (Scheme 2),
together with the monohydroboration product 9-tBuC(H)C-
(H)-9-borafluorene (product ratio =3:1; see the SI for the NMR
data and an X-ray crystal structure analysis of 9-tBuC(H)C(H)-
9-borafluorene).
The constitution of 9, which we have isolated in analytically
pure form in about 50% yield, has been confirmed by an X-ray
crystal structure analysis (see below), but is also evident from its
NMR spectra: (i) The 11B{1H} NMR spectrum shows one broad
signal at 71.0 ppm, which lies within the typical shift range of
triorganylboranes44 (cf. 9-Et-9-borafluorene: δ (11B) = 73.145).
(ii) The proton integral ratio between the four 9-borafluorene
resonances on the one hand and the tert-butyl singlet on the
other indicates a 2:1 stoichiometry. (iii) We observe a doublet
integrating 2H at 2.82 ppm and a triplet integrating 1H at 3.87
ppm; in the HSQC spectrum, the doublet shows a cross peak to a
sharp 13C signal at 46.6 ppm while the triplet is associated with a
very broad carbon resonance at 36.8 ppm. Since such line
broadening is typical for the signals of boron-bonded carbon
atoms, the HSQC spectrum leads to the conclusion that both
boryl substituents are attached to the same carbon center, which
necessarily has to be the terminal one.
a Reagents and conditions: (i) n = 2, C6H6, rt, 10-15 min. (ii) 1. n = 1,
hexane, rtf5 °C, 7d; 2. precipitate dissolved in C6H6, þ exc. pyridine, rt.
represents a rigid bifunctional organoborane, in which the two
boron atoms are placed in close proximity. Compounds of this
kind have found numerous applications,46 for example as hydride
sponges,47 as cocatalysts in ethylene polymerization,48 and as
catalysts for inverse electron-demand Diels-Alder reactions.49
In a more exotic context, the intramolecular interaction of the
boron pz-orbitals in 1,8-bis(diphenylboryl)naphthalene has been
exploited for the generation of a B-B one-electron σ-bond.50
Crystals of 9 suitable for X-ray analysis were grown from a
benzene solution (Figure 2; details of the X-ray crystal structure
analysis of 9 are compiled in the SI). As already deduced from the
NMR spectra, the compound contains two 9-borafluorenyl
moieties per alkyl fragment and both boryl groups are attached
to the same carbon atom. All key structural parameters of the
9-borafluorenyl substituents closely resemble those of 9-phenyl-
9-borafluorene.39
The formation of 5 py and 9 from (5)2 is revealing in several
3
respects: With regard to the general reaction mechanism it
becomes evident that (5)2 can readily convert back to 5 in the
presence of Lewis bases or CtC triple bonds. With regard to
applications it is pleasing to see that 9-H-9-borafluorene (5) can
be generated in situ and used efficiently in hydroboration
reactions.
The alkyl bridge of 9 is disordered over two positions (C and
C0), which impedes an accurate determination of the B-C(1)
bond lengths as well as the B(1)-C(1)-B(2) bond angle (cf.
Compound 9 not only constitutes important proof for the
existence of 5, but it is also interesting in its own right, because it
4598
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Figure 2. Molecular structure of 9 in the solid state; the minor sites (C0)
of the disordered alkyl bridge are not shown (30% probability ellipsoids;
H atoms omitted for clarity). Selected bond lengths (Å), atom atom
3 3 3
distance (Å), bond angles (deg), and dihedral angle (deg): B(1)-
C(1)/C(10) 1.586(3)/1.614(7), B(1)-C(11) 1.570(3), B(1)-C(21)
1.585(3), B(2)-C(1)/C(10) 1.598(2)/1.612(7), B(2)-C(31)
1.569(2), B(2)-C(41) 1.571(2); B(1) B(2) 2.534(2); C(11)-
3 3 3
Figure 3. Molecular structure of 10 py2 in the solid state (30%
B(1)-C(21) 103.2(1), C(31)-B(2)-C(41) 103.9(1), B(1)-C(1)/
C(10)-B(2) 105.5(2)/103.5(4); B(1)C(11)C(21)//B(2)C(31)C(41)
74.5.
3
probability ellipsoids; H atoms on carbon omitted for clarity). Selected
bond lengths (Å), bond angles (deg), and dihedral angle (deg): B(1)-
N(21) 1.619(4), B(1)-C(1) 1.637(4), B(1)-C(11) 1.632(4), B(1)-
C(31) 1.632(4), B(2)-N(51) 1.599(4), B(2)-C(42) 1.617(4);
N(21)-B(1)-C(31) 113.1(2), C(1)-B(1)-C(11) 98.9(2),
N(51)-B(2)-C(42) 109.5(2); Ar(C(31))//Ar(C(41)) 59.5.
Figure 2). The most important parameters are the B(1) B(2)
3 3 3
distance of 2.534(2) Å and the dihedral angle of 75° between the
two 9-borafluorenyl planes.
Trapping of a Ring-Opened Dimer of 9-H-9-Borafluorene.
The reaction mixture of 8 and Et3SiH in C6D6 was monitored by
NMR spectroscopy at rt over a period of several days. Thereby, it
became obvious that the initial intermediate (5)2 is not long-term
case of B(1) (cf. B(2)-N(51) = 1.599(4) Å, B(2)-C(42) =
1.617(4) Å).
The formation of 10 is reminiscent of the ligand redistribution
leading from 1,10-fc(B(H)Br)2 to the polyferrocenylene [-fc-
B(Br)-]n (2) and B2H6 (Scheme 1). At the same time, it is a ring-
opening process, which lends support to our initial supposition
that 9-H-9-borafluorene (5) may undergo a ring-opening po-
lymerization to boron-bridged oligophenylenes.
1
stable under these conditions, because the H and 13C NMR
spectra gradually changed with time. Noticeably, more and more
proton resonances overlapped in the chemical shift interval
6.6 ppm - 7.5 ppm to such an extent that an identification of
individual multiplets was no longer possible.
Ring-Opening Oligomerization of 9-H-9-Borafluorene. In
the absence of a trapping reagent, storage of a mixture of 8 and
Et3SiH (1:1) in hexane at rt results in the formation of thin
yellow crystal plates suitable for X-ray crystallography (details of
the X-ray crystal structure analysis are compiled in the SI). Under
the conditions specified in the Experimental Section, the first
crystals were typically observed after about 7 d, but the crop was
usually not harvested earlier than after 1 month. The crystals
consist of a remarkable main-chain boron-containing oligophe-
nylene 11 (Figure 4 top), which is a formal ring-opened
pentamer (Scheme 3) of 9-H-9-borafluorene (5). The molecule
bears one BH2 headgroup (i.e., B(1)) as part of a B2H3 terminus
and one 9-borafluorenyl end-group (i.e., B(5)). This structural
motif is reminiscent of the borane H2B-C6H4-C6H4-
In a series of subsequent experiments, we therefore tried again
to trap important constituents of the aged reaction mixtures. We
assumed that any higher molecular weight species should be less
soluble in aliphatic hydrocarbons than in benzene and therefore
switched to hexane as the reaction solvent. We also reduced the
reaction temperature to 5 °C to slow down the conversion rate of
(5)2. In this way, we obtained a pale yellow precipitate from a
hexane solution of 8 and Et3SiH, which had been stored at 5 °C
for 7 d. Since the material was not suitable for single crystal X-ray
analysis, we dissolved it at rt in C6H6 and treated the solution
with excess pyridine. Gas-phase diffusion of hexane into this
mixture led to the growth of a few single crystals of the double
pyridine adduct 10 py2 of a 2-biphenylylborane bearing a
3
9-borafluorenyl substituent (Scheme 2 and Figure 3; details of
BC12H8, which we have isolated as pyridine diadduct 10 py2.
the X-ray crystal structure analysis of 10 py2 are compiled in the
3
3
All four H-bearing boron atoms, B(1) to B(4), of 11 are engaged
in intramolecular B-H-B 3c-2e bonding, thereby establishing
two intrastrand cross-links. The tricyclic fragments B(1)-Ar(C-
(1))-Ar(C(11))-B(2) and B(3)-Ar(C(41))-Ar(C(51))-B(4)
closely resemble the literature-known compounds 6 and 7
(Figure 1), which form upon insertion of 0.5 (H2BR)2 into
one endocyclic B-C bond of 9-R-9-borafluorene (R = H,
Et).33,34
SI). It is important to note at this stage that the borane H2B-
C6H4-C6H4-BC12H8 (10), which has been trapped in the
form of 10 py2, is not identical with (5)2, because the latter
3
compound converts to 5 py when treated with pyridine.
3
The B(1)-N(21) = 1.619(4) Å bond length in 10 py2 agrees
3
with the B-N bond length in the pyridine adduct of 9-Cl-9-
borafluorene51 (average value =1.613(3) Å; two crystallographi-
cally independent molecules in the asymmetric unit). The
The molecular structure of 7 in the solid state has been
published.45 All key metrical parameters of 7 are very similar to
those of B(1)-Ar(C(1))-Ar(C(11))-B(2) and B(3)-Ar(C(41))-
Ar(C(51))-B(4): The endocyclic B-CAr bond lengths fall in the
interval between 1.548(6) Å (11) and 1.582(4) Å (11), the
exocyclic bond in 10 py2 (B(1)-C(31) = 1.632(4) Å) possesses
3
the same length as the two endocyclic bonds (B(1)-C(1) =
1.637(4) Å, B(1)-C(11) = 1.632(4) Å). Most likely as a result
of reduced steric congestion, the pyridine ligand as well as
the phenyl ring can approach B(2) more closely than in the
4599
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Figure 5. Molecular structure of 13 in the solid state (30% probability
ellipsoids; H atoms on carbon omitted for clarity; for a schematic view
illustrating the boron environments cf. Scheme 3). Selected bond
lengths (Å), atom atom distances (Å), and dihedral angles (deg):
3 3 3
B(1)-C(1) 1.531(8), B(2)-C(11) 1.556(7), B(2)-C(21) 1.578(7),
B(3)-C(31) 1.574(6), B(3)-C(41) 1.571(6), B(4)-C(51) 1.569(6),
B(4)-C(61) 1.582(6), B(5)-C(71) 1.583(6), B(5)-C(81) 1.548(7),
B(6)-C(91) 1.526(8), B(1) B(2) 1.847(9), B(3) B(4)
3 3 3
3 3 3
1.792(6), B(5) B(6) 1.849(9); Ar(C(1))//Ar(C(11)) 1.8, Ar(C-
3 3 3
(21))//Ar(C(31)) 57.3, Ar(C(41))//Ar(C(51)) 13.6, Ar(C(61))//
Ar(C(71)) 56.5, Ar(C(81))//Ar(C(91)) 1.0.
the X-ray crystal structure analysis are compiled in the SI). As to
be expected, 13 shows similar structural characteristics as 11,
both with regard to the B B distances and the overall
3 3 3
conformation of the oligomer chain.
Figure 4. (Top) Molecular structure of 11 in the solid state (30%
probability ellipsoids; H atoms on carbon omitted for clarity, one
terminal H atom is hidden behind B(1); for schematic views illustrating
the boron environments cf. Scheme 3). (Bottom) Superposition of the
molecular structures of 11 and 12, which cocrystallize in the same crystal
lattice (stoichiometric ratio: 11:12 ≈ 3:1). Selected bond lengths (Å),
Ring-Opening Dimerization of 9-H-9-Borafluorene. As
already outlined above, the H NMR spectrum of a mixture of
1
9-Br-9-borafluorene (8) and Et3SiH in C6D6 first shows the
signature of (5)2 and after some time develops an increasingly
complex pattern of multiple overlapping resonances, which
most likely correspond to higher oligomers. At an even later
stage, however, we observed the appearance of two very well-
resolved doublets of multiplets exhibiting an unusually strong
low-field shift (8.30 ppm, 8.63 ppm). Parallel to that, colorless
needle-shaped crystals grew in the NMR tube. Colorless
needles were also observed on the surface of the crystal cake
of 11-13 after prolonged storage of hexane solutions of 8 and
Et3SiH.
atom atom distances (Å), and dihedral angles (deg) of 11: B(1)-
3 3 3
C(1) 1.548(6), B(2)-C(11) 1.570(5), B(2)-C(21) 1.590(4), B(3)-
C(31) 1.572(4), B(3)-C(41) 1.582(4), B(4)-C(51) 1.561(4), B(4)-
C(61) 1.579(4), B(5)-C(71) 1.571(5), B(5)-C(81) 1.581(5), B(5)-
C(91) 1.560(5), B(1) B(2) 1.806(6), B(3) B(4) 1.800(4); Ar-
3 3 3
3 3 3
(C(1))//Ar(C(11)) 7.7, Ar(C(21))//Ar(C(31)) 53.5, Ar(C(41))//
Ar(C(51)) 7.0, Ar(C(61))//Ar(C(71)) 53.7, Ar(C(21))//Ar(C(61))
4.9, Ar(C(31))//Ar(C(71)) 5.1.
B
B distances cover the range between 1.800(4) Å (11) and
An X-ray crystal structure analysis of the needles revealed the
3 3 3
cyclic dimer 14 (Scheme 3; Figure 6).52
1.839(4) Å (7). Remarkably, 11 cocrystallizes with the closely
related oligomer 12 (Scheme 3), which bears two 9-borafluorene
end-groups (stoichiometric ratio of 11 and 12 in the crystal
≈3:1). A superposition of the two structures is shown in Figure 4
(bottom).
Compound 12 contains five biphenyl moieties, but only four
boron atoms. Formally, it can be generated from the pentamer 11
in a chain-termination step by the extrusion of BH3. We never
detected the 11B resonance of B2H6 during monitoring NMR
studies of the reaction between 8 and Et3SiH in sealed NMR
tubes. However, we take the isolation of compound 13, which
bears two B2H3-termini (Scheme 3), as evidence that the
extruded BH3 can be reinserted into the 9-borafluorene tail-
group of another molecule of pentamer 11.
13 crystallizes in very small amounts as yellow blocks that can
easily be overlooked in the crop of crystals of 11/12 (an X-ray
powder diffractogram of the sample containing 11/12 and 13
showed an excellent match with the simulated powder diffracto-
gram calculated from the single crystal data of 11/12). A plot of
the molecular structure of 13 is presented in Figure 5 (details of
All B-CAr bond lengths are very close to the corresponding
values in 7;45 however, the B B distance in 14 amounts to
3 3 3
1.780(3) Å, compared to a distance of 1.839(4) Å in 7.
Compound 14 adopts a twisted conformation. The dihedral
angles between C6H4 rings belonging to the same biphenyl unit
are 21.1° and 22.4°, while the dihedral angles Ar(C(1))//
Ar(C(21)) and Ar(C(11))//Ar(C(31)) are 30.4° and 31.3°,
respectively.
Compound 14 is already literature-known and has been
synthesized in 15% yield by the reaction of 7 with BF3 OEt2
3
(xylene, rt to 130 °C, 4 h).34 Since these conditions indicate that
14 is a thermodynamically favored molecule, we investigated the
reaction between 9-Br-9-borafluorene and Et3SiH also at ele-
vated temperature (toluene, reflux temperature, 3 h). Already
when the mixture was allowed to cool to rt again, crystals of 14
started to grow, and finally we obtained a crystalline yield of 25%.
Thus, our synthesis protocol not only provided an authentic
sample of 14, but it also gave access to a compound that
has promising potential as building block for two-dimensional
4600
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Scheme 3. Ring-Opened Pentamer 11 of 5, the Corre-
sponding Oligomers 12 and 13, Which Can Formally Be
Generated by Extrusion/Insertion of BH3 from/into 11, and
the Ring-Opened Dimer 14, Which Forms at Elevated
Temperatures
Figure 6. Molecular structure of 14 in the solid state (30% prob-
ability ellipsoids). Selected bond lengths (Å), atom
atom distance
3 3 3
(Å), bond angles (deg), and dihedral angles (deg): B(1)-C(1)
1.566(3), B(1)-C(21) 1.569(3), B(2)-C(11) 1.570(3), B(2)-
C(31) 1.568(3), B(1) B(2) 1.780(3); C(1)-B(1)-C(21) 132.2(2),
3 3 3
C(11)-B(2)-C(31) 131.6(2), B(1)-C(1)-C(2) 120.8(2), B(1)-
C(21)-C(22) 120.6(2), B(2)-C(11)-C(12) 120.5(2), B(2)-C(31)-
C(32) 120.2(2); Ar(C(1))//Ar(C(11)) 21.1, Ar(C(1))//Ar(C(21))
30.4, Ar(C(21))//Ar(C(31)) 22.4, Ar(C(11))//Ar(C(31)) 31.3.
boron-doped π-systems. The NMR spectroscopic investigation
of 14 confirmed that the signals at 8.30 ppm and 8.63 ppm, which
appeared last in the spectrum of the long-term monitoring study
(see above), indeed correspond to H-4 and H-1 of the ring-
opened cyclic dimer. We therefore conclude that oligomers like
11 (Scheme 3) are the kinetically favored products of the ring-
opening reaction of 5. At elevated temperatures, however, 14 is
formed preferentially.
Quantum Chemical Calculations. To provide further in-
sight into the details of the reaction mechanism underlying the
experimental observation and to propose a molecular structure
for the primary reaction intermediate (5)2, we now turn to the
discussion of quantum chemical results.
Dimerization Pathways. Conceptually, the formation of
covalently bonded dimers of 9-borafluorene (A) can be induced
by three independent elementary steps, i.e., B-H/B-H (the
B-H bond of the first monomer adds across the B-H bond of
the second monomer), and the corresponding B-H/B-C and
B-C/B-C addition reactions. We will show in the following
that the same three elementary steps are essential also for
chain initiation, propagation, and chain termination of higher
oligomers.
B-H/B-H reaction between two monomers results in the
slightly exoergic formation of the C2h-symmetric diborane B with
two B-H-B 3c-2e bonds (ΔGR = -3.3 kcal mol-1; Figure 7).
The structural motif of B is reminiscent of the unique B-H-B
bridged coordination polymer [3]n formed by 9,10-dihydro-
9,10-diboraanthracene in the crystal lattice.23 In the 9-borafluor-
ene case, however, we have so far not obtained experimental
evidence for the existence of a dimer B.
Figure 7. Dimer formation through B-H/B-H addition of 9-bora-
fluorene monomers (Gibbs free energies in kcal mol-1 relative to two
separate molecules of A).
B-H/B-C addition of two monomers leads to intermediate
C (Figure 8), which features a bridging phenyl ring. The
formation of C is exoergic but its kinetic stability is limited under
the experimental conditions applied: opening of the four-mem-
bered BHBC ring is feasible by simultaneous breaking of two
juxtaposed bonds, either leading back to A with a barrier of 8.7
kcal mol-1, or going forth via TSCD to D with a slightly higher
barrier of 11.3 kcal mol-1. D contains an intact 9-borafluorene
ring and a BH2 fragment linked by a biphenyl moiety. As we will
Figure 8. Dimer formation through B-H/B-C addition (Gibbs free
energies in kcal mol-1 relative to two independent molecules of A).
4601
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Figure 9. Dimer formation through endo-B-C/B-C addition (Gibbs free energies in kcal mol-1 relative to two independent molecules of A).
illustrate further below, our computed reaction paths indicate
that D represents a key structure for the build-up of higher
9-borafluorene oligomers for it provides two essential reaction
centers, i.e., a primary and a tertiary boron atom. Thus, the
experimental proof of the presence of D in the reaction mixture
(cf. compound 10 py2; Scheme 2) is an important validation of
3
this pivotal constituent of the complex overall reaction cascades.
Dimerization through B-C/B-C addition can proceed either
in an endo- or in an exo-approach of the monomers A. The initial
barrier of the endo-B-C/B-C addition (TSAE, ΔGq= 8.7 kcal
mol-1) is of comparable height to the processes investigated so
far. Dimer E features two boron-bridging phenyl rings (Figure 9).
A subsequent, almost barrierless ring-opening step through TSEF
leads to the essentially thermoneutral formation of intermediate
F. Only a small conformational adjustment is required for an
intramolecular B-H/B-C addition via TSFG to yield the
phenyl-bridged intermediate G with a moderate barrier. Opening
of the BHBC ring in G via TSGD provides a second viable path to
D. In addition, we located TSGH, a transition state which
connects G with H, a particularly interesting product because
not only D (as pyridine adduct 10 py2), but also H has been
3
characterized experimentally (cf. 14; Scheme 3). Yet, although H
is thermodynamically favored over D, the path GfTSGHfH
cannot compete kinetically (ΔGq = 29.6 kcal mol-1), and H
must form via an alternative route.
Figure 10. Dimer formation through exo-B-C/B-C addition (Gibbs
free energies in kcal mol-1 relative to two independent molecules of A).
H is also kinetically stable. In the sense of a dynamical least-
motion principle, all structures constituting this path are already
comfortably preorganized for the direct and exclusive formation
of the two stable B-H-B bonds present in H.
Thus, all reaction sequences along the B-H/B-H, the B-H/
B-C, and the endo-B-C/B-C paths involve low barriers and
no thermodynamic sinks occur. Only the exo-B-C/B-C path
yields a well-defined, thermodynamically and kinetically stable
product (H), provided that sufficient activation energy is avail-
able. Under our standard experimental conditions, however, the
reversible paths should be dominant and should lead to complex
dynamic equilibria. Accordingly, as shown in the NMR experi-
ments, only at very early stages the composition of the reaction
Indeed, the exo-B-C/B-C addition of two monomers A
provides an energetically favorable pathway to H (Figure 10).
Both, the first intermediate I and the corresponding transition
state TSAI, are higher in energy than those of their exo-counter-
parts E and TSAE due to the lack of stabilizing π-π interactions.
However, subsequent ring-opening and a further B-H/B-H
addition leads in a strongly exoergic final step to H. Conse-
quently, with an overall forward barrier of ΔGq = 17.4 kcal mol-1
and a reaction free energy of ΔGR = -9.0 kcal mol-1 (relative to
two monomers A) the exo-B-C/B-C addition path results in
the thermodynamically most favored product identified so far
(H). With a reverse reaction barrier of ΔGq = 22.8 kcal mol-1
,
4602
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
mixture is dominated by one single species (5)2. A comparison of
relative free energies of the intermediates along the reaction
pathways studied so far renders B, C, and G relevant candidates
to assign a structure to (5)2.
The experimental 1H NMR spectrum of the primary product
(5)2 is characterized by eight doublets and eight virtual triplets/
doublets of doublets, all possessing the same integral value
(neglecting 4JHH coupling in this description). The experimental
13C{1H} NMR spectrum contains 16 CH resonances together
with eight signals assignable to quaternary carbon nuclei, four of
which show the extreme signal broadening typical43 of boron-
bonded carbon atoms. Hence, at this point, we can already
exclude B because of its high symmetry.
The two remaining structures (C and G; Figure 11), however,
possess C1 symmetry and thus four magnetically inequivalent
C6H4 rings I-IV. In line with that, all experimentally observed
1H and 13C signals can be assigned to four different C6H4 rings
with the help of H,H COSY and HSQC experiments; a HMBC
experiment revealed that rings I and II as well as III and IV are
tied together to form two o,o0-diborylated biphenyl fragments.
We note that the 1H chemical shift values of I to IV are spread
over the wide range between 6.14 ppm and 8.40 ppm, but that the
proton signals of the biphenyl unit I-II have rather similar
chemical shifts. The 13C chemical shifts of corresponding CH
and C-5 (cf. Figure 1) nuclei in rings I to III fall within the narrow
range of (5 ppm. In contrast, ring IV is characterized by several
abnormal NMR features: (i) one CH atom resonates at 151.6
ppm, which is 15-30 ppm downfield from all other CH signals;
(ii) the chemical shift value of the C-5 nucleus is 169.6 ppm, as
opposed to 140.9, 142.0, and 145.0 ppm for rings I to III; (iii) the
chemical shift value of the C-6 (cf. Figure 1) nucleus is 114.2
ppm, which has to be compared to values of 137.4, 139.1,
and 149.7 ppm for rings I-III (see the SI for a plot of selected
NMR spectra and a detailed compilation of all chemical shift
values).
Figure 11. Selected computed 1H (red), 13C (black), and 11B (green)
NMR chemical shift values for C (top) and G (bottom; B97D/TZVP
GIAO calculations, deviations from experiment given in parentheses).
this compound as the phenyl-bridged species G. (ii) At room
temperature, a mixture of 9-Br-9-borafluorene (8) and Et3SiH in
C6D6 has to be kept for weeks before the NMR signature of H
(= 14) occurs; from refluxing toluene, however, 14 precipitates in
yields of 25% within a few hours. (iii) D obviously exists in the
aged reaction mixture and can be isolated in relevant amounts by
1
Computed H and 13C NMR shifts of C and G show a
significantly better match of experimental and computed data
for the latter (the SI contains a full presentation of data, including
a calibration for the NMR computations). Figure 11 shows the
most revealing features of C and G. For both species the unusual
13C resonances (see above) can be attributed to the boron-
bridging phenyl rings. However, we can discriminate between C
and G using the 11B shift values, as well as the 13C NMR data for
C-5 (cf. Figure 1) and the corresponding 1H signals of the ortho
protons of rings I and II. In line with investigations of
Wrackmeyer,45 fragments of type 5 show C-5 shift values around
150 ppm, whereas in fragments of type 6 the C-5 resonances
appear around 140 ppm. Also, the ortho 1H resonances appear
around 7.5 ppm in the former case, but around 8.5 ppm in the
latter. While (5)2 reacts as a direct source of 5 (as shown in the
trapping experiments), the NMR spectroscopic evidence indi-
cates that it is structurally more closely related to the ring-opened
compound 6 (Figure 1). The NMR calculations on G reproduce
all experimental chemical shift values of (5)2 with a mean signed
deviation of 0.14 ( 0.17 ppm (1H), -0.6 ( 1.9 ppm (13C),
or -1.1 ( 1.1 ppm (11B), and we feel safe to propose the identity
of (5)2 and G.53,54
trapping with pyridine (cf. 10 py2).
3
Against this background, we investigated the fate of D in more
detail, focusing first on its BH2 headgroup for its obvious
inherent reactivity.
Reactivity of the BH2 Headgroup in D. Relevant reaction
partners of D include added Lewis bases such as pyridine as well
as 9-H-9-borafluorene A. Dimerization of D is not considered
given the supposedly low stationary concentration of this highly
reactive species under the experimental conditions applied.
In line with the successful isolation of the pyridine adduct
10 py2, we computed the B-N binding energies of D py and
3
3
D py2 (which corresponds to 10 py2) and found a significant
3
3
thermodynamic driving force of ΔGR = -23.1 kcal mol-1 toward
the formation of D py2. In comparison, the formation of the
3
pyridine adduct A py (= 5 py; Scheme 2) is thermodynamically
3
3
less favored (ΔGR = -15.0 kcal mol-1; cf. SI).
As illustrated in Figure 12, K is readily formed from A and D
via B-H/B-H addition. Although this step represents a head-
group protection of D, we will demonstrate below that K still is
an important starting point of further reaction cascades.
Alternatively, D can also act as a hydride donor for a monomer
A through B-H/B-C addition. Figure 13 compiles the subse-
quent, surprisingly facile reaction cascade to N, which proceeds
So far, we conclude that our quantum chemical predictions are
fully in line with our most compelling experimental observations:
(i) neither 9-H-9-borafluorene A (= 5) nor its B-H-B-bridged
dimer B are detectable by in situ NMR spectroscopy; instead, we
observe clearly distinct signatures of a dimer (5)2, and we identify
4603
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
exclusively via the elementary steps already introduced above.
Trimer N is formed with a moderate overall barrier of ΔGq = 7.7
kcal mol-1 and constitutes (apart from the pyridine diadduct
encountered so far (ΔGR = -19.8 kcal mol-1 with respect to
D plus A; ΔGR = -21.1 kcal mol-1 with respect to three
independent units of A). We note that N already features all
relevant structural characteristics of compound 11 (Figure 4),
the major constituent of the crystalline precipitate, i.e., a 9-phen-
yl-9-borafluorene tail-group, a biphenylene linker, and a biphe-
nylene-bridged B2H3 moiety.
D py2) the thermodynamically most favorable species
3
Reactivity of the 9-Borafluorenyl Tail-Group in K. B-H/
B-C addition of a monomer A to K can occur across all three
bonds about the tertiary boron atom in the 9-borafluorene tail-
group. B-H addition across the exocyclic B-C bond constitutes
an unproductive identity reaction. B-H/B-C addition across
either of the two endocyclic B-C bonds leads to chain propaga-
tion; however, the inherent chirality of a biphenyl bridge renders
the two B-C bonds inequivalent. We explicitly studied both
paths using the unprotected dimer D as a model system (see the
SI for details). Since both reaction sequences do not exhibit any
appreciable kinetic or thermodynamic differences, we only
investigated the energetically (slightly) favorable chain propaga-
tion steps for the more realistic intermediate K (Figure 14). The
reaction cascade involves the usual elementary steps with an
overall barrier of ΔGq = 13.9 kcal mol-1 and first leads to P,
which contains a valence-unsaturated central diorganyl borane
group. Subsequent rotation of this group about two B-C bonds
results in a conformation from which two intramolecular B-H-
B bonds are established in a concerted process with concomitant
extrusion of the protective group A. Most notably, the final
species resulting from this sequence is identical to the product N
formed by the B-H/B-C attack of the BH2 headgroup of D on a
monomer A discussed above (Figure 13). N represents the
smallest oligomer of an entire series, comprising also the experi-
mentally characterized species 11 (Figure 4). We therefore focus
in the following section on chain propagation processes com-
mencing with N.
Figure 12. Protection reaction of the BH2 headgroup of D with A
(Gibbs free energies in kcal mol-1 relative to separated D plus A).
Chain Propagation. In principle, the transition from K to N
is just an isomerization reaction and its driving force is provided
by the loss of antiaromaticity resulting from the (formal) ring
expansion of the BH2-protecting 9-H-9-borafluorene moiety.
It is therefore not surprising that the reactivity of the 9-bora-
fluorenyl tail-groups toward an incoming monomer A is essen-
tially the same in K and N (cf. Figures 14 and 15). With R
(Figure 15), we encounter an intermediate that bears close
Figure 13. B-H/B-C addition of the BH2 headgroup of D to A
(Gibbs free energies in kcal mol-1 relative to separated D plus A).
Figure 14. B-H/B-C addition of A to the tail-group of K (Gibbs free energies in kcal mol-1 relative to K).
4604
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Figure 15. B-H/B-C addition of A to the tail-group of N (Gibbs free energies in kcal mol-1 relative to N plus A).
Figure 16. Oligomers N, N2, and N3 (Gibbs free energies in kcal mol-1
relative to the equivalent number of independent A).
resemblance to P (Figure 14), and it is again the unsaturated
nature of this species that governs the course of the subsequent
reactions.
0
As before, internal B-C rotation (TSRR ) orients the dangling
B-H bond such that the unsaturated active boron site is shifted
from its internal position (R/R0) to the periphery of the molecule
(S) in a concerted process. This shift generates a primary BH2
headgroup and thus substantially enhances the disposition of the
molecule for further intermolecular reactivity. S differs from D
only by the presence of an (innocent) internal spacer between
the two crucial functional groups, i.e., a BH2 head and a
9-borafluorenyl tail. The fate of S is readily deducible from the
reactivity patterns identified for D: both the headgroup reactivity
and the tail-group reactivity established before (Figure 13 and
Figure 14, respectively) will lead to the same N-type product
(N2; Figure 16), but now augmented by the first repeat unit of a
growing polymer chain. Because of the obvious structural and
energetic incremental behavior of the elementary steps involved,
we refrained from explicit calculations of the reaction pathway.
Our mechanistic proposal finds strong support by the success-
ful experimental characterization of N2 (= 11; Figure 4). How-
ever, the list of potential reaction partners is obviously more
complex than explicitly considered above, e.g., headgroup attack
of D to the tail-group of N represents a short cut to the higher
oligomer N2 (Figure 16). Yet, for a comprehensive description of
the general picture it suffices to focus on the few key elementary
Figure 17. Chain termination by B-H/B-C addition of A to the BH2
headgroup of S and subsequent elimination of BH3 (Gibbs free energies
in kcal mol-1 relative to S plus A).
steps involved, rather than on the wealth of possible reaction
partners.
It remains to be elucidated whether thermodynamics pose a
natural limit to chain propagation. Figure 16 compiles the
computed structures along the sequence of oligomers from N
to N3 and the relative free energies with respect to the corre-
sponding numbers of free monomers A. We conclude that
already for low-molecular-weight oligomers the two chain ends
behave essentially as independent fragments and every repeating
unit added contributes an increment of about ΔGinc = -23 kcal
mol-1 to the thermodynamic stability of the resulting polymer.
Chain Termination. In this section we exclusively consider
the two species that have been characterized experimentally, i.e.,
12 and 13 (Figures 4 and 5). Straightforward access would
formally involve BH3 elimination from 11 (= N2) and subse-
quent insertion into the 9-borafluorenyl tail-group of another
molecule of 11. While we did not identify any energetically
4605
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Figure 18. Chain termination by B-H/B-H addition of BH3 to the tail-group of N (Gibbs free energies in kcal mol-1 relative to N plus BH3).
feasible direct process for BH3 extrusion from N2, we found that
species like S represent bifurcation points either for chain
propagation (e.g., SþAfN2 = 11) or chain termination
(SþAfU = 12; Figure 17): B-H/B-C addition of A to the
BH2 headgroup of S yields intermediate T and subsequent BH3
elimination generates U.
shift values of various conceivable isomers of (5)2 gave an
excellent match for a structure featuring one B-H-B 3c-2e
bond and one boron-bridging phenyl ring. This species then ring-
opens to give the borane H2B-C6H4-C6H4-BC12H8 (10). 10
can react with a monomer 5 either via the BH2 headgroup or via
the 9-borafluorenyl tail-group. Inboth cases, the final products are
identical, i.e., the trimer H[-(H)B-(C6H4)2-]2BC12H8, which
stabilizes itself by two intramolecular B-H-B 3c-2e bonds and is
a shorter congener of oligomer 11. Further chain growth can
readily proceed via (i) B-H attack of a monomer 5 on an
endocyclic B-C bond of oligomers H[-(H)B-(C6H4)2-]n-
BC12H8 if n is an even number, or (ii) B-C attack of a monomer 5
on the free BH2 headgroup of H[-(H)B-(C6H4)2-]nBC12H8 if
n is an odd number. Chain termination occurs by BH3 extrusion
during BH2 headgroup attack of A and gives oligomers
H8C12B-(C6H4)2[-(H)B-(C6H4)2-]nBC12H8 bearing two
9-borafluorenyl termini (cf. the crystallographically characterized
compound 12; n = 2). Alternatively, BH3 can insert into the
9-borafluorenyl tail groups of growing polymer chains, thereby
generating a H[-(H)B-(C6H4)2-]nBH2 species (cf. the crystal-
lographically characterized compound 13; n = 5). At our present
stage of knowledge we have to assume that in the case of 11 the
chain length is mainly determined by the low solubility of the
compound in hexane, because theory has established that there is
no obvious thermodynamic limit to chain growth. We are pre-
sently evaluating possibilities to prepare higher oligomers or even
polymers by introduction of solubilizing side chains.
Our results indicate that it should be possible to transform
a mixture of, e.g., 11, 12, and 13 into pure 13 by treatment
with B2H6. Moreover, hydroboration of alkynes with such
oligomers should lead to corresponding oligophenylenes
[-(R)B-(C6H4)2-]n with three-coordinate boron bridges,
for which we expect highly interesting optoelectronic properties.
Finally, we note that intrastrand B-(μ-H)2-B cross-links as
in 11, 12, and 13 offer new ways to modulate the conformational
flexibility as well as the electronic structure of such macromole-
cules by, for example, temperature increase to thermally open the
B-H-B bonds.
The insertion of BH3 into an endocyclic B-C bond of the
9-borafluorene tail-group in N2 is indeed a viable process to form
oligomers 13, carrying two B2H3 termini. We studied the
corresponding reaction pathway for the smallest relevant model
system N (Figure 18). The highly exoergic reaction NþBH3fX
(ΔGR = -19.2 kcal mol-1) commences with an almost barrier-
less B-H/B-C BH3 addition step, followed by a rate-limiting
ring-opening via TSVW (ΔGq = 13.6 kcal mol-1), concerted
rotations about internal B-C and C-C bonds, and a subsequent
almost barrierless intramolecular B-H/B-H addition. Com-
pared to the series N to N3 (Figure 16), we find the same
incremental behavior (ΔGinc = -23 kcal mol-1) of the relative
free energies along the series of corresponding products of this chain
termination step, X to X3(cf. SI). The computed moderate barrier
for this reaction sequence (ΔGq = 13.6 kcal mol-1 with respect to
N) is fully in line with the fact that we did not observe any free B2H6
in our monitoring NMR studies described above.
’ CONCLUSION
9-H-9-Borafluorene (H8C12BH, 5), which can easily be gen-
erated by Br/H exchange between 9-Br-9-borafluorene and
Et3SiH, is a versatile reagent for the preparation of boron-doped
π-conjugated materials. We have shown that (i) it readily
hydroborates tert-butylacetylene but also that (ii) in the absence of
a trapping reagent it undergoes a ring-opening polymerization
reaction leading to B(H)-bridged biphenylenes H[-(H)B-
(C6H4)2-]nBC12H8. The corresponding pentamer 11 (n = 4)
has been characterized in the solid state by X-ray crystallography;
the compound shows two intrastrand B-(μ-H)2-B cross-links.
The reaction mechanism underlying polymer formation has
been elucidated in a joint experimental and theoretical study.
Three elementary steps, i.e., addition of a B-H bond across
another B-H bond, B-H/B-C addition, and B-C/B-C
addition, are sufficient to explain chain initiation, chain propaga-
tion as well as chain termination. In situ NMR spectroscopy on
solutions of 9-Br-9-borafluorene and Et3SiH (C6D6, rt) revealed
signals assignable to a C1-symmetric dimer (5)2 as primary
’ EXPERIMENTAL METHODS
General Remarks. All reactions were carried out under a nitrogen
atmosphere using Schlenk tube techniques or in an argon-filled glove-
box. Reaction solvents were either freshly distilled under argon from Na/
1
reaction intermediate. Computed H, 13C, and 11B chemical
4606
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
Pb alloy (alkanes) and Na/benzophenone (benzene, toluene, C6D6) or
dried over molecular sieves (4 Å; CDCl3) prior to use. NMR: Bruker
Avance 300, Avance 400, and Avance 600. Chemical shifts are referenced
needles (i.e., 14). An X-ray powder diffractogram of the crystal mixture
turned out to be a superposition of the powder diffractograms calculated
from the single-crystal data of 11/12 and 14; 11/12 were by far the
dominating species, and the signature of 13 could not be detected at all.
We note that crystallization of 11-14 is a continuous process. After
one month, an overall yield of about 20% (with respect to 8) can be
harvested; if the ampule is stored for 3-4 months, the oligomer yield
reproducibly exceeds 50%.
The colorless needles of 14 were manually removed from the sample
used for the elemental analysis. Elemental analysis calcd (%) for 0.75
C60H45B5 ꢀ 0.25 C60H42B4 ꢀ 0.5 C6H14 (859.64): C 88.02, H 6.01;
found: C 87.88, H 5.76.
Synthesis of 14. A stirred solution of 8 (336 mg, 1.38 mmol) in
toluene (20 mL) was heated to reflux temperature. A solution of Et3SiH
(350 μL, 255 mg, 2.19 mmol) in toluene (20 mL) was added dropwise
over a period of 1 h. After the addition was complete, heating was
continued for another 2 h. The clear yellow solution obtained was
allowed to cool to rt, whereupon thin colorless needles formed. The
mixture was stored at rt for 16 h, the mother liquor was removed using a
syringe, and the crystal crop was dried under vacuum. Yield: 56 mg (0.17
mmol; 25%). X-ray-quality crystals were obtained by recrystallization
from C6H6.
to (residual) solvent signals (1H, 13C{1H}) or external BF3 OEt2
3
(11B{1H}). Abbreviations: s = singlet, d = doublet, t = triplet, vt =
virtual triplet, m = multiplet, br = broad, n.o. = signal not observed. 8 was
synthesized following a literature procedure.42
Synthesis of 9. Neat Et3SiH (390 μL, 285 mg, 2.45 mmol) was
added via syringe at rt to a slowly stirred solution of 8 (311 mg, 1.28
mmol) in C6H6 (5.5 mL). After approximately 10 min, the color of the
mixture changed from yellow-green to almost colorless. This point
corresponds to the quantitative consumption of 8 with formation of (5)2
(NMR spectroscopic control). Neat tBuCCH (90 mL, 60 mg, 0.73
mmol) was added via syringe and stirring was continued at rt for 15 min,
whereupon the color of the solution changed to an intense yellow; a
slight heat development was noticeable. The mixture was stored at 5 °C
in a refrigerator for 1 h, whereupon a yellow microcrystalline solid
precipitated. The mother liquor was removed via syringe, and the solid
product, which turned out to be analytically pure 9, was dried under
dynamic vacuum. Yield: 124 mg (0.30 mmol; 47%). Single cystals of 9
were grown by recrystallization from C6H6.
All volatiles were removed from the mother liquor under dynamic
vacuum to yield a yellow oily residue. After approximately 12 h, part of
the oily residue had transformed into yellow crystals, which turned out to
be a second crop of 9 (NMR spectroscopic control). A close inspection
of the crystalline material revealed the presence of very few yellow-green
needles that consisted of the monohydroboration product 9-tBuC-
(H)C(H)-9-borafluorene. According to 1H NMR spectroscopy, the
ratio of 9 and 9-tBuC(H)C(H)-9-borafluorene in the oily residue was
1:1; we found no evidence for the presence of significant amounts of
other products (see the SI for NMR data and an X-ray crystal structure
analysis of 9-tBuC(H)C(H)-9-borafluorene).
1H NMR (400.1 MHz, C6D6): δ 3.12 (br, 2H; BH), 7.38 (vtd, 4JHH
1.2 Hz, 4H; H-2), 7.48 (vtd, 4JHH = 1.5 Hz, 4H; H-3), 8.30 (dm, 3JHH
=
=
8.2 Hz, 4H; H-4), 8.63 (dd, 3JHH = 7.5 Hz, 4JHH = 1.5 Hz, 4H; H-1). 1H
NMR (300.0 MHz, CDCl3): δ 7.59 (vtd, 4JHH = 1.2 Hz, 4H; H-2), 7.70
(ddd, 3JHH = 8.1, 8.2 Hz, 4JHH = 1.6 Hz, 4H; H-3), 8.52 (dm, 3JHH = 8.2
3
4
Hz, 4H; H-4), 8.76 (dd, JHH = 7.6 Hz, JHH = 1.6 Hz, 4H; H-1).
11B{1H} NMR (96.3 MHz, CDCl3): δ 11.0 (h1/2 = 300 Hz). 13C{1H}
NMR (75.4 MHz, CDCl3): δ 126.8 (C-2), 127.6 (C-4), 129.8 (C-3),
132.9 (C-1), 142.6 (C-5), n.o. (C-6).
Crystal Structure Analyses. Crystals of 9, 9-tBuC(H)C(H)-9-
1H NMR (400.1 MHz, C6D6): δ 0.93 (s, 9H; CH3), 2.82 (d, 3JHH
=
=
borafluorene, and 10 py2 were measured on a STOE IPDS-II diffract-
3
6.7 Hz, 2H; CH2), 3.87 (t, 3JHH = 6.7 Hz, 1H; CHCH2), 6.88 (vtd, 4JHH
ometer with graphite-monochromated Mo KR radiation. Crystals of 11/
12, 13, and 14 were measured on a Bruker CCD three-circle diffract-
ometer equipped with an INCOATEC microsource generator with
multilayer optics. The structures were solved by direct methods using
the program SHELXS55 and refined with full-matrix least-squares on F2
using the program SHELXL97.56 Hydrogen atoms bonded to C were
placed on ideal positions and refined with fixed isotropic displacement
4
1.3 Hz, 4H; H-2), 6.95 (vtd, JHH = 1.3 Hz, 4H; H-3), 6.99 (m, 4H;
H-4), 7.92 (dm, 3JHH = 6.9 Hz, 4H; H-1). 11B{1H} NMR (96.3 MHz,
C6D6): δ 71.0 (br). 13C{1H} NMR (75.4 MHz, C6D6): δ 29.7 (CH3),
32.6 (CCH3), 36.8 (br, CHCH2), 46.6 (CH2), 119.8 (C-4), 128.4 (C-2),
134.2 (C-3), 134.6 (C-1), 142.7 (br, C-6), 153.8 (C-5). Elemental
analysis calcd (%) for C30H28B2 (410.14): C 87.86, H 6.88; found: C
87.65, H 6.90.
parameters using a riding model. In 10 py2 the H atoms bonded to B
3
were found in a difference map and were freely refined. The H atoms
bonded to B in 11/12 were found in a difference map; the fully occupied
ones were freely refined, the partially occupied ones were refined using a
riding model. The H atoms bonded to B in 13 were found in a difference
map; the coordinates of six of the eight H atoms were refined, whereas
the remaining two were refined using a riding model. In 14, the H atoms
bonded to B were found in a difference map and were freely refined.
In 9, the neohexyl residue is disordered over two positions. The site
occupation factor of the major occupied sites is 0.748(4). The minor
occupied atoms were isotropically refined. Similarity restraints were
used for the two different sites.
Synthesis of 10 py2. A stirred solution of 8 (160 mg, 0.66 mmol)
3
in hexane (17 mL) was treated at rt with neat Et3SiH (130 μL, 95 mg,
0.82 mmol). The mixture was stored at 5 °C in a refrigerator for 7 d,
whereupon a pale-yellow microcrystalline precipitate formed. The
mother liquor was removed via syringe. The solid residue was dried in
vacuo, dissolved in C6H6 (5 mL), and treated with an excess of
neat pyridine (0.40 mL, 390 mg, 5.00 mmol). After 0.5 h stirring at rt,
the clear, pale-red solution was evaporated to dryness in vacuo in order
to remove unreacted pyridine. The resulting solid residue was redis-
solved in C6H6 (3 mL). Single crystals of 10 py2 suitable for X-ray
3
analysis were obtained by gas-phase diffusion of hexane into this
solution.
9-tBuC(H)C(H)-9-borafluorene formed nonmerohedrally twinned
crystals. The fractional contribution of the major domain was 0.655(5).
Hydrogen atoms were placed on ideal positions and refined with fixed
isotropic displacement parameters using a riding model.
Synthesis of 11, 12, and 13. Neat Et3SiH (200 μL, 146 mg, 1.26
mmol) was added via syringe at rt to an ampule containing a solution of 8
(271 mg, 1.12 mmol) in hexane (30 mL). The ampule was cooled to
liquid-nitrogen temperature, vacuum sealed, and allowed to warm to rt.
From the clear, yellow solution obtained, crystals grew over a period of
approximately 1 month. The ampule was opened in a glovebox, and the
mother liquor was removed from the crystalline material using a syringe.
The crystal crop was carefully examined with a microscope to determine
its homogeneity. The vast majority of crystals turned out to be thin
yellow platelets (i.e., 11/12), which were accompanied by very few
yellow crystals of slightly different habit (i.e., 13) and some colorless
10 py2 crystallizes with half a benzene molecule (located on a 2-fold
3
rotation axis) in the asymmetric unit.
Compounds 11 and 12 crystallize together in the same crystal. The
site occupation factor of 11 refined to 0.734(4) and was fixed to 0.75 in
order to avoid correlations between occupancy factors and displacement
parameters. The 1,2 and 1,3 distances between the minor occupied C
atoms were restrained to be equal, respectively, and these atoms were
refined with a common isotropic displacement parameter. 11/12 crystallizes
4607
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
with half a hexane molecule in the asymmetric unit, which is disordered
about a center of inversion over two equally occupied sites.
13 crystallizes with half a hexane molecule in the asymmetric unit,
located on a center of inversion. One of its CH2 groups is disordered
over two equally occupied sites.
This material is available free of charge via the Internet at http://
pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
Matthias.Wagner@chemie.uni-frankfurt.de; Max.Holthausen@
chemie.uni-frankfurt.de
CCDC reference numbers: 802380 (9), 802381 (9-tBuC(H)C(H)-
9-borafluorene), 802382 (10 py2), 802383 (11/12), 802384 (13),
3
802385 (14).
Computational Details. Geometry optimizations, analytic har-
monic frequency analyses, and intrinsic reaction coordinate (IRC)
calculations have been performed with the Gaussian09 program.57 Initial
geometry optimizations were performed employing the M06L
functional58 in combination with the small D95V basis.59 To speed up
the computations, we made use of the RI procedure60 employing the
DGA1 fitting basis61 as implemented in Gaussian09. The resulting,
rather moderate M06L/D95V/DGA1 level of density functional
theory allowed for efficient explorations of potential reaction pathways,
and it was found to provide very reasonable structures for minima
and transition states. This level was also used for IRC calculations, which
we performed to validate connectivities between minima and transition
structures implied above. The resulting structures were subsequently
reoptimized at the B3LYP(V)-D/TZVP level of DFT. In these calcula-
tions the B3LYP hybrid functional62 was employed incorporating the
VWN5 local correlation functional.63 Empirical dispersion corrections
devised by Grimme64 were employed, and the TZVP basis set65 was
used. The resulting B3LYP(V)-D/TZVP level of density functional
theory was recently identified as pleasingly accurate for related investiga-
tions.66 The nature of all stationary points reported in the following has
been characterized by the number of negative eigenvalues in the respective
diagonalized force constant matrices (0 for minima, 1 for transition states;
the wavenumbers of the imaginary modes of all transition states are given in
the figures below). Unscaled zero-point vibrational energies as well as
thermal and entropy corrections were obtained from computed Hessians
using the standard procedures in Gaussian09 and were used to obtain Gibbs
free energies, which relate to 298.15 K and ideal gas conditions (i.e, 0.04 mol
’ ACKNOWLEDGMENT
M.W. and M.C.H. gratefully acknowledge financial support
by the Beilstein-Institut, Frankfurt/Main, Germany, within the
research collaboration NanoBiC. A.H. is grateful to the Fonds
der Chemischen Industrie (FCI) for a Ph. D. grant.
’ REFERENCES
(1) (a) M€ullen, K; Wegner, G. Electronic Materials: The Oligomer
Approach, Wiley-VCH, Weinheim, 1998. (b) M€ullen, K; Scherf, U.
Organic Light-Emitting Devices: Synthesis Properties and Applications,
Wiley-VCH, Weinheim, 2005. (c) M€uller, T. J. J.; Bunz, U. H. F.
(Eds.) Functional Organic Materials: Syntheses, Strategies and Applica-
tions, Wiley-VCH, Weinheim, 2006.
(2) Entwistle, C. D.; Marder, T. B. Angew. Chem., Int. Ed. 2002, 41,
2927–2931.
(3) Entwistle, C. D.; Marder, T. B. Chem. Mater. 2004, 16, 4574–
4585.
(4) Yamaguchi, S.; Wakamiya, A. Pure Appl. Chem. 2006, 78, 1413–
1424.
(5) Doi, H.; Kinoshita, M.; Okumoto, K.; Shirota, Y. Chem. Mater.
2003, 15, 1080–1089.
(6) Jia, W. L.; Feng, X. D.; Bai, D. R.; Lu, Z. H.; Wang, S.; Vamvounis,
G. Chem. Mater. 2005, 17, 164–170.
(7) Hudnall, T. W.; Chiu, C.-W.; Gabbaï, F. P. Acc. Chem. Res. 2009,
42, 388–397.
(8) Wade, C. R.; Broomsgrove, A. E. J.; Aldridge, S.; Gabbaï, F. P.
Chem. Rev. 2010, 110, 3958–3984.
l
-1).67 All structures have been optimized within the highest possible
(9) J€akle, F. Coord. Chem. Rev. 2006, 250, 1107–1121.
(10) J€akle, F. Chem. Rev. 2010, 110, 3985–4022.
(11) Matsumi, N.; Naka, K.; Chujo, Y. J. Am. Chem. Soc. 1998, 120,
10776–10777.
(12) Scheibitz, M.; Bats, J. W.; Bolte, M.; Lerner, H.-W.; Wagner, M.
Organometallics 2004, 23, 940–942.
molecular symmetry. All resulting nontrivial point groups are given in the
figures below and all atoms involved in the chemical transformations have
been highlighted by a ball-and-stick representation; all other hydrogen
atoms have been omitted for clarity.
NMR isotropic shieldings σ were computed at the B97D/TZVP level
of theory64 within the GIAO68 framework on structures reoptimized at
this level of DFT. Tetramethylsilane (σ(1H) = 31.73, σ(13C) = 182.96)
(13) Sundararaman, A.; Victor, M.; Varughese, R.; J€akle, F. J. Am.
Chem. Soc. 2005, 127, 13748–13749.
1
was chosen as internal standard to calculate the H and 13C chemical
(14) Heilmann, J. B.; Scheibitz, M.; Qin, Y.; Sundararaman, A.; J€akle,
F.; Kretz, T.; Bolte, M.; Lerner, H.-W.; Holthausen, M. C.; Wagner, M.
Angew. Chem., Int. Ed. 2006, 45, 920–925.
(15) Niu, W.; Smith, M. D.; Lavigne, J. J. J. Am. Chem. Soc. 2006, 128,
16466–16467.
(16) Li, H.; J€akle, F. Angew. Chem., Int. Ed. 2009, 48, 2313–2316.
(17) Scheibitz, M.; Li, H.; Schnorr, J.; Sꢀanchez Perucha, A.; Bolte,
M.; Lerner, H.-W.; J€akle, F.; Wagner, M. J. Am. Chem. Soc. 2009, 131,
16319–16329.
shift values δ. Diborane(6) (σ(11B) = 85.78) was chosen as internal
standard to calculate δ(11B); the values obtained were corrected for the
experimentally used standard BF3 OEt2 (δ(B2H6) = 18.0 ppm vs
3
BF3 OEt2). We validated this method by comparison of computed and
3
measured 1H, 13C, and 11B chemical shifts for compounds 5 py, 6, 8, 9, and
3
14 and found very good agreement within 0.08 ( 0.32 ppm (1H), -0.7 (
1.6 ppm (13C), and -3.1 ( 3.4 ppm (11B, mean signed deviations).
(18) Cui, C.; Heilmann-Brohl, J.; Sꢀanchez Perucha, A.; Thomson,
M. D.; Roskos, H. G.; Wagner, M.; J€akle, F. Macromolecules 2010, 43,
5256–5261.
(19) Matsumi, N.; Naka, K.; Chujo, Y. J. Am. Chem. Soc. 1998, 120,
5112–5113.
(20) Matsumi, N.; Miyata, M.; Chujo, Y. Macromolecules 1999, 32,
4467–4469.
(21) Matsumi, N.; Chujo, Y.; Lavastre, O.; Dixneuf, P. H. Organo-
metallics 2001, 20, 2425–2427.
(22) Miyata, M.; Chujo, Y. Polym. Bull. 2003, 51, 9–16.
(23) Lorbach, A.; Bolte, M.; Li, H.; Lerner, H.-W.; Holthausen,
M. C.; J€akle, F.; Wagner, M. Angew. Chem., Int. Ed. 2009, 48
4584–4588.
’ ASSOCIATED CONTENT
S
Supporting Information. Synthesis and NMR spectra of
b
(5)2, 5 py, and 9-tBuC(H)C(H)-9-borafluorene. Crystallo-
graphic data of 9, 9-tBuC(H)C(H)-9-borafluorene, 10 py2,
3
3
11/12, 13, and 14. X-ray crystal structure analysis of 9-tBuC-
(H)C(H)-9-borafluorene. Quantum chemical results for pyri-
dine adducts of A and D, reaction pathways for the B-H/B-C
addition of A to D, free energies of oligomers X, X2, and
X3, computed NMR shifts for G, C, H, A py, ABr, Ac, and
3
A2tBE, tables with detailed computational results for all station-
ary points localized, as well as the full citation of reference 57.
4608
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609

Journal of the American Chemical Society
ARTICLE
(24) Lorbach, A.; Bolte, M.; Lerner, H.-W.; Wagner, M. Chem.
Commun. 2010, 46, 3592–3594.
(50) Hoefelmeyer, J. D.; Gabbaï, F. P. J. Am. Chem. Soc. 2000, 122,
9054–9055.
(51) Narula, C. K.; N€oth, H. Inorg. Chem. 1985, 24, 2532–2539.
(52) A structure plot of 14 is already shown in ref 34, however, the
crystal data have never been published. We have therefore performed a
full X-ray crystal structure analysis of the compound.
(25) Lorbach, A.; Bolte, M.; Lerner, H.-W.; Wagner, M. Organome-
tallics 2010, 29, 5762–5765.
(26) Eisch, J. J.; Galle, J. E.; Kozima, S. J. Am. Chem. Soc. 1986, 108,
379–385.
(53) We are aware of only one other compound featuring a
comparable B-(μ-aryl)-B structural motif: Pilz, M.; Allwohn, J.; Massa,
W.; Berndt, A. Angew. Chem., Int. Ed. Engl. 1990, 29, 399–401.
(54) Hofmann, M.; Scheschkewitz, D.; Ghaffari, A.; Geiseler, G.;
Massa, W.; Schaefer, H. F., III; Berndt, A. J. Mol. Mod. 2000, 6, 257–271.
(55) Sheldrick, G. M. Acta Crystallogr. 1990, A46, 467–473.
(56) Sheldrick, G. M. SHELXL-97. A Program for the Refinement of
Crystal Structures. Universit€at G€ottingen, G€ottingen, 1997.
(57) Frisch, M. J. et al. Gaussian 09, Revision A.02, Gaussian, Inc.:
Wallingfort CT, 2009.
(27) For selected recent references on the synthesis, structural
peculiarities, and applications of boroles, see: (a) Braunschweig, H.;
Fernꢀandez, I.; Frenking, G.; Kupfer, T. Angew. Chem., Int. Ed. 2008, 47,
1951–1954. (b) So, C.-W.; Watanabe, D.; Wakamiya, A.; Yamaguchi, S.
Organometallics 2008, 27, 3496–3501. (c) Fan, C.; Piers, W. E.; Parvez,
M. Angew. Chem., Int. Ed. 2009, 48, 2955–2958. (d) Huynh, K.; Vignolle,
J.; Tilley, T. D. Angew. Chem., Int. Ed. 2009, 48, 2835–2837.
(28) Chase, P. A.; Piers, W. E.; Patrick, B. O. J. Am. Chem. Soc. 2000,
122, 12911–12912.
(29) Romero, P. E.; Piers, W. E.; Decker, S. A.; Chau, D.; Woo, T. K.;
Parvez, M. Organometallics 2003, 22, 1266–1274.
(30) Grigsby, W. J.; Power, P. P. J. Am. Chem. Soc. 1996, 118, 7981–
7988.
(58) Zhao, Y.; Truhlar, D. G. J. Chem. Phys. 2006, 125 (194101),
1–18.
(59) Dunning Jr., T. H.; Hay, P. J. in Modern Theoretical Chemistry.
Ed. Schaefer III, H. F. Vol. 3, 1-28; Plenum: New York, 1976.
(31) Wehmschulte, R. J.; Khan, M. A.; Twamley, B.; Schiemenz, B.
Organometallics 2001, 20, 844–849.
€
(60) (a) Eichkorn, K.; Treutler, O.; Ohm, H.; H€aser, M.; Ahlrichs, R.
Chem. Phys. Lett. 1995, 240, 283–289. (b) Kendall, R. A.; Fr€uchtl, H. A.
Theor. Chem. Acc. 1997, 97, 158–163. (c) Dunlap, B. I. J. Mol. Struct.
(THEOCHEM) 2000, 529, 37–40. (d) Dunlap, B. I. J. Chem. Phys. 1983,
78, 3140–3142. (e) Eichkorn, K.; Weigend, F.; Treutler, O.; Ahlrichs, R.
Theor. Chem. Acc. 1997, 97, 119–124.
(32) For experimental as well as theoretical studies on 9-borafluor-
ene-containing materials, see: (a) Yamaguchi, S.; Shirasaka, T.; Akiyama,
S.; Tamao, K. J. Am. Chem. Soc. 2002, 124, 8816–8817. (b) Briꢁere, J.-F.;
C^otꢀe, M. J. Phys. Chem. B. 2004, 108, 3123–3129. (c) Thanthiriwatte,
K. S.; Gwaltney, S. R. J. Phys. Chem. A. 2006, 110, 2434–2439. (d)
Doskocz, J.; Doskocz, M.; Roszak, S.; Soloducho, J.; Leszczynski, J. J.
Phys. Chem. A. 2006, 110, 13989–13994. (e) Wakamiya, A.; Mishima, K.;
Ekawa, K.; Yamaguchi, S. Chem. Commun. 2008, 579–581. (f) Bonifꢀacio,
V. D. B.; Morgado, J.; Scherf, U. J. Polym. Sci., Part A: Polym. Chem. 2008,
46, 2878–2883. (g) Muhammad, S.; Janjua, M. R. S. A.; Su, Z. J. Phys.
Chem. C 2009, 113, 12551–12557.
(33) Hong, H.; Chung, T. C. J. Organomet. Chem. 2004, 689, 58–64.
(34) K€oster, R.; Willemsen, H.-G. Liebigs Ann. Chem. 1974, 1843–1850.
(35) 9-H-9-Borafluorene derivatives bearing bulky substituents at
their phenylene rings have been suggested (i) as reaction intermediate
during the reduction of 2,6-(2,4,6-iPr3C6H2)2C6H3BBr2 with KC8 (cf.
Ref 30) and (ii) as transient product of an intramolecular C-H activation
of the terphenylborane {2,6-(4-tBuC6H4)2C6H3BH(μ-H)}2 (cf.
Wehmschulte, R. J.; Diaz, A. A.; Khan, M. A. Organometallics 2003,
22, 83–92.
(61) Godbout, N.; Salahub, D. R.; Andzelm, J.; Wimmer, E. Can. J.
Chem. 1992, 70, 560–571.
(62) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098–3100. (b) Lee, C.;
Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789. (c) Stephens, P. J.;
Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem. 1994, 98,
11623–11627. (d) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(63) (a) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58,
1200–1211. (b) Hertwig, R. H.; Koch, W. Chem. Phys. Lett. 1997, 268,
345–351.
(64) Grimme, S. J. Comput. Chem. 2006, 27, 1787–1799.
(65) (a) Sch€afer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97,
2571–2577. (b) Sch€afer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994,
100, 5829–5835.
(66) Rakow, J. R.; T€ullmann, S.; Holthausen, M. C. J. Phys. Chem. A
2009, 113, 12035–12043.
(67) McQuarrie, D. A.; Simon, J. D. Physical Chemistry - A
Molecular Approach. University Science Books: Sausalito, CA, 1997.
(68) (a) Ditchfield, R. Mol. Phys. 1974, 27, 789–807. (b) Wolinski,
K.; Hinton, J. F.; Pulay, P. J. Am. Chem. Soc. 1990, 112, 8251–8260.
(36) van Veen, R.; Bickelhaupt, F. J. Organomet. Chem. 1973, 47,
33–38.
(37) Knizek, J.; N€oth, H. J. Organomet. Chem. 2000, 614-615, 168–
187.
(38) K€oster, R.; Benedikt, G. Angew. Chem. 1963, 75, 419.
(39) Kaufmann, L.; Vitze, H.; Bolte, M.; Lerner, H.-W.; Wagner, M.
Organometallics 2008, 27, 6215–6221.
(40) Gross, U.; Kaufmann, D. Chem. Ber. 1987, 120, 991–994.
(41) Wakamiya, A.; Mishima, K.; Ekawa, K.; Yamaguchi, S. Chem.
Commun. 2008, 579–581.
(42) H€ubner, A.; Lerner, H.-W.; Wagner, M.; Bolte, M. Acta Crystal-
logr. 2010, E66, o444.
(43) Wrackmeyer, B. Prog. Nucl. Magn. Reson. Spectrosc. 1979, 12,
227–259.
(44) N€oth, H.; Wrackmeyer, B. Nuclear Magnetic Resonance Spec-
troscopy of Boron Compounds. In NMR Basic Principles and Progress;
Diehl, P., Fluck, E., Kosfeld, R., Eds.; Springer: Berlin, Heidelberg,
New York, 1978.
(45) Wrackmeyer, B.; Thoma, P.; Kempe, R.; Glatz, G. Collect.
Czech. Chem. Commun. 2010, 75, 743–756.
(46) Piers, W. E.; Irvine, G. J.; Williams, V. C. Eur. J. Inorg. Chem.
2000, 2131–2142.
(47) Katz, H. E. J. Org. Chem. 1985, 50, 5027–5032.
(48) Spence, R. E. v. H.; Piers, W. E.; Sun, Y.; Parvez, M.;
MacGillivray, L. R.; Zaworotko, M. J. Organometallics 1998, 17, 2459–
2469.
(49) Kessler, S. N.; Wegner, H. A. Org. Lett. 2010, 12, 4062–4065.
4609
dx.doi.org/10.1021/ja110947k |J. Am. Chem. Soc. 2011, 133, 4596–4609