Is there B-N bond-length alternation in 1,2:3,4:5,6-Tris(biphenylylene) borazines?

Article abstract of DOI:10.1002/cplu.201300110

The alternation of the B-N bonds in the central borazine ring of the overcrowded 1,2:3,4:5,6-tris(biphenylylene)borazine (2 a) and its tribromo derivative (2 g) is investigated by computational methods and compared with their experimentally obtained crystal structures. The calculations are performed with a meta-generalized-gradient-approximation (GGA) density functional (Tao-Perdew-Staroverov-Scuseria (TPSS)) without and with dispersion corrections, including Becke-Johnson damping, in conjunction with a polarized triple-ζ basis set. These data show a small bond-length alternation (BLA) of around 0.01A in 2 a and 2 g. This outcome is in good agreement with X-ray diffraction data for 2 g, but at variance with earlier X-ray diffraction measurements that gave a BLA of 0.06A for 2 a. A re-investigation of the crystal structure of 2 a reveals a positional disorder that precludes a discussion of the B-N bond lengths. The synthesis of 2 g is the first example of an electrophilic aromatic substitution of an aryl borazine with elemental bromine. Successful bromination was also demonstrated for hexaphenylborazine.

Full text of DOI:10.1002/cplu.201300110

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300110
Is There BÀN Bond-Length Alternation in 1,2:3,4:5,6-
Tris(biphenylylene)borazines?
Matthias Mꢀller,[a] Cꢁcilia Maichle-Mçssmer,[b] Peter Sirsch,[b] and Holger F. Bettinger*[a]
In memory of Detlef Schrçder
The alternation of the BÀN bonds in the central borazine ring
of the overcrowded 1,2:3,4:5,6-tris(biphenylylene)borazine (2a)
and its tribromo derivative (2g) is investigated by computa-
tional methods and compared with their experimentally ob-
tained crystal structures. The calculations are performed with
around 0.01 ꢂ in 2a and 2g. This outcome is in good agree-
ment with X-ray diffraction data for 2g, but at variance with
earlier X-ray diffraction measurements that gave a BLA of
0.06 ꢂ for 2a. A re-investigation of the crystal structure of 2a
reveals a positional disorder that precludes a discussion of the
BÀN bond lengths. The synthesis of 2g is the first example of
an electrophilic aromatic substitution of an aryl borazine with
elemental bromine. Successful bromination was also demon-
strated for hexaphenylborazine.
a
meta-generalized-gradient-approximation (GGA) density
functional (Tao–Perdew–Staroverov–Scuseria (TPSS)) without
and with dispersion corrections, including Becke–Johnson
damping, in conjunction with a polarized triple-z basis set.
These data show a small bond-length alternation (BLA) of
Introduction
As bond-length equalization is one criterion of aromaticity,[1–4]
bond-length alternation (BLA) in annulated benzenoid mole-
cules has been investigated extensively.[5–18] Extreme examples
are “starphenylene”[5] and tris(bicyclo[2.1.1]hexeno)benzene,[11]
the central six-membered rings of which exhibit cyclohexa-
triene-like geometries with BLA of 0.14 and 0.089 ꢂ, respective-
ly. But even in a relatively simple polycyclic aromatic hydrocar-
bon such as triphenylene, a BLA of 0.059 ꢂ in the central ring
is observed.[19,20] This has been explained by cyclic p effects.[21]
The aromatic character of “inorganic” benzene,[22–27] borazine
B3N3H6 (1), is still debated (Scheme 1).[28–49] The BÀN bond
lengths in 1 alternate to a very small extent and their average
(electron diffraction: 1.4355(21) ꢂ;[50] X-ray: 1.429(1) ꢂ[51]) lies
between the values for a typical BÀN single bond (for example,
in ammonia borane: 1.6576(16) ꢂ[52] (microwave spectroscopy)
and 1.564(6) ꢂ[53] (X-ray)) and a B=N double bond (e.g., in ami-
noborane: 1.391(2) ꢂ[54] (microwave spectroscopy)). A survey of
the geometries of borazine molecules deposited in the Cam-
bridge Structural Database (CSD) revealed that BLA in borazine
derivatives is usually small (see the Supporting Information for
details).
Scheme 1. Borazine (1) and 1,2:3,4:5,6-tris(biphenylylene)borazine (2a,
X=H).
An unusual borazine derivative in this regard is 1,2:3,4:5,6-
tris(biphenylylene)borazine (2a), the borazine analogue of hex-
abenzotriphenylene,[21,55–59] which was synthesized first by
Kçster et al. in 1965 (Scheme 1).[60,61] The crystal and molecular
structure of 2a was reported by Roberts et al. in 1974.[62] The
steric overcrowding due to the biphenyl rings precludes
a planar geometry for 2a, and a propeller-like structure with
the borazine ring distorted from planarity is preferred.[62] The
BÀN bond lengths were reported to alternate: the shorter BÀN
bonds (average 1.430 ꢂ vs. 1.488 ꢂ) were found between pairs
of atoms bridged by the biphenylylene substituent.[62] The re-
sulting BLA of 0.06 ꢂ is larger than that found in hexabenzotri-
phenylene (0.04 ꢂ),[21] but similar to that of triphenylene
(0.06 ꢂ).[20]
[a] M. Mꢀller, H. F. Bettinger
Institut fꢀr Organische Chemie
Universitꢁt Tꢀbingen
Auf der Morgenstelle 18, 72076 Tꢀbingen (Germany)
E-mail: Holger.Bettinger@uni-tuebingen.de
[b] C. Maichle-Mçssmer, P. Sirsch
Institut fꢀr Anorganische Chemie
Universitꢁt Tꢀbingen
We have recently reported an alternative synthesis for 2a
and found that the molecule has interesting photophysical
properties.[63] Although the geometry of 2a computed at the
B3LYP/6-31G* level of theory was overall in good agreement
Auf der Morgenstelle 18, 72076 Tꢀbingen (Germany)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201300110.
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 988 – 994 988

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
with experiment, the bond-
length alternation (0.01 ꢂ) was
significantly smaller than the
one derived experimentally
(0.06 ꢂ).
One may argue that the
B3LYP/6-31G* level of theory is
not of sufficient quality to pro-
duce highly accurate geometric
parameters for the overcrowded
2a. However, the B3LYP func-
tional in conjunction with simi-
lar-quality basis sets reproduced
the experimental structure and
BLA in hexabenzotriphenylene[21]
and triphenylene[64] well. More-
over, more sophisticated com-
putational methods do not
bring experiment and theory in
accord, as demonstrated below.
In view of the importance of an
accurate computational descrip-
tion of molecular structures, the
discrepancy in bond-length al-
Scheme 2. Synthetic routes to 2.
ternation between experiment and theory prompted the pres-
ent study. Herein, we report for the first time the synthesis and
structural characterization of a derivative of 2a that was ob-
tained by electrophilic aromatic substitution, a reaction that
has not been reported previously for aryl-substituted bora-
zines. In addition, the geometries of 2a and its derivative were
studied using more sophisticated computational treatments of
2a. Our results indicate that the bond-length alternation in 2a
is significantly smaller than concluded previously.
Another synthetic route to 2b uses base-induced dehydro-
halogenation[63] of 10-chloro-3-isopropyl-9-aza-10-boraphenan-
threne (5). Although the synthesis of 5 is straightforward, the
dehydrohalogenation step failed.
As 2a is sensitive to water and Lewis acids in solution, we
focused on its mild direct functionalization (Scheme 3). Transi-
tion-metal-mediated borylation of 2a to yield 2d using bis(pi-
nacolato)diboron/IrI[66] led to the recovery of the starting mate-
rial. Some of the classical electrophilic aromatic substitution re-
actions (acid-free nitration[67,68] and Vilsmeier–Haack formyla-
tion[69]) gave rise to complex mixtures instead of the expected
derivatives 2e and 2 f. Only the bromination of 2a yielded tri-
bromo derivative 2g in reasonable quantities. Treatment of 2a
at room temperature with an excess amount of bromine with-
out catalyst in dichloromethane produces derivative 2g in
35% yield according to mass spectrometry and heteronuclear
two-dimensional NMR spectroscopy. The proton spectrum dis-
plays seven signal groups and thus points to a structure with
Results and Discussion
Syntheses
High-temperature dehydrogenation of 3a at 4058C was de-
scribed by Kçster et al.[60,61] to produce 2a in a yield of 16%
(Scheme 2). We were able to increase the yield to 25% by
strict control of the argon flow that blows off the released hy-
drogen. During the course of the reaction the crystalline by-
product 4 also formed in 6% yield, as we reported previous-
ly.[63]
To reveal a motif-inherent bond-length alternation, it is nec-
essary to synthesize derivatives of 2a for structural characteri-
zation. However, the high temperature required for the dehy-
drogenation step is expected to severely limit the scope of
substituents that would be tolerated. Indeed, the high-temper-
ature synthesis of isopropyl derivative 2b failed. The synthesis
of the tert-butyl derivative was impossible as the precursor
borazine 3b could not be obtained. Attempts to lower the re-
action temperature required for dehydrogenation using Hart-
wig’s[65] iridium(I) complex resulted in mixtures that could not
be characterized any further.
Scheme 3. Attempted functionalization of 2a.
ChemPlusChem 2013, 78, 988 – 994 989
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
threefold symmetry. Two sets consisting of four and three hy-
drogen atoms can be identified by 1H,1H-COSY and TOCSY,
which indicates that only one of the rings in the biphenylyl
system carries a bromine atom. Owing to efficient relaxation
induced by boron nuclei, adjacent carbon atoms are not seen
in carbon NMR spectra.
Table 1. Computed distances (in ꢂ) in 2a and its derivative 2g using the
TPSS meta-GGA functional without and with D3 dispersion corrections,
including Becke–Johnson damping D3 (BJ) in conjunction with the def2-
TZVP basis set. See Scheme 1 for a definition of distances a–c.
Level
Distance a
Distance b
Distance c
X=H (2a)
The preferred bromination of the nitrogen-bound phenyl
ring is expected as it is more electron rich owing to the meso-
meric effect of the borazine nitrogen atoms. To the best of our
knowledge, this is the first example of an electrophilic aromatic
bromination of an aryl borazine. This reaction is not limited to
2a: the three nitrogen-bound phenyl rings of hexaphenylbora-
zine (6) can be brominated to give the corresponding tribromo
derivative 6a (Scheme 4). As the structure cannot be con-
TPSS
TPSS-D3
TPSS-D3(BJ)
1.453
1.451
1.449
1.464
1.462
1.461
3.150
3.131
3.097
X=Br (2g)
TPSS
1.453
1.451
1.449
1.445
1.464
1.462
1.460
1.453
3.142
3.122
3.088
3.097
TPSS-D3
TPSS-D3(BJ)
exptl[a]
1
firmed by X-ray crystallography, a H,15N HMBC NMR spectrum
[a] Averaged.
We hence used these techniques also for the computation of
the structures of 2a and its derivative 2g (Table 1).
The results obtained without dispersion correction parallel
the earlier B3LYP/6-31G* data: again, the bond-length alterna-
tion is around 0.01 ꢂ and therefore much smaller than that in
experiment. Switching on the dispersion correction (D3) slight-
ly shortens the BN distances, but does not change the differ-
ence in bond lengths: it remains at 0.01 ꢂ. Likewise, additional
BJ damping very slightly shortens the BN distances further, but
the degree of alternation does not change. Only distances be-
tween neighboring aromatic systems shrink by 0.05 ꢂ. Hence,
the calculations show that the bond-length alternation in 2a is
smaller than that derived earlier from X-ray diffraction and is
not influenced markedly by the close contact of the organic ar-
omatic p systems present. The same also holds true for deriva-
tive 2g that has not been structurally characterized previously.
Scheme 4. Bromination of aryl borazines 6 and 7.
was recorded. The data clearly show a correlation between the
ortho protons of the brominated aryl ring and the borazine ni-
trogen. A good indicator for the reaction success is the strong
BN stretching band detected by IR spectroscopy around
1360 cmÀ1. Upon bromination of N,N’,N’’-triphenylborazine (7)
this band disappears, which indicates decomposition of the
borazine core during the reaction.
Single-crystal X-ray crystallography
In light of the theoretical results above, we have carried out an
X-ray diffraction study on 2g for the first time, and also re-in-
vestigated the crystal structure of 2a at low temperature and
with Mo instead of Cu radiation. Suitable crystals for 2a were
obtained from the melt of 3a. The same space group and unit
cell as in the previous report were obtained for 2a.[62] In that
earlier study, the authors noted that both potential models for
the assignment of the B and N positions in the borazine ring
had been tested; the one with the lower R value (0.084) was fi-
nally selected as the best model. In our refinements, we in ad-
dition employed a model in which the B and N atoms were
treated as being disordered over all six ring positions. With this
approach, the R value dropped significantly (to 0.048) and the
occupancy factors for the ring sites were refined to a B/N ratio
of about 1:1 (i.e., each position is shared by boron and nitro-
gen by equal amounts). This strongly suggests the presence of
positional disorder in the borazine ring, which is also facilitated
by the molecular symmetry (close to C3) of the compound: The
overall geometry of the molecule (i.e., the relative positions of
the biphenylyl groups), and hence the intermolecular interac-
Computational treatment
The repulsion between hydrogen atoms of the biphenylyl sys-
tems results in the propeller-like structure of 2a. The distances
between the adjacent rings are 3.1–3.2 ꢂ, a range that is just
below the sum of the van der Waals radii of carbon (3.4 ꢂ).[70]
It may be argued that the bond-length alternation in 2a as de-
termined by X-ray diffraction[62] arises from steric overcrowding
of the biphenylyl substituents at the borazine ring. The accu-
rate treatment of such close contacts between aromatic units
can be achieved nowadays by density functional methods
using the corrections for dispersion interactions without (D3)
and with Becke–Johnson damping (D3(BJ)) as introduced by
Grimme and co-workers.[71,72] Grimme and Mꢀck-Lichtenfeld
have shown recently for cyclophanes with p–p distances short-
er than the sum of the van der Waals radii that the combina-
tion of common density functionals, such as the Tao–Perdew–
Staroverov–Scuseria (TPSS)[73] meta-generalized-gradient-ap-
proximation, and the D3 and D3(BJ) corrections provides geo-
metries in very good agreement with X-ray crystallography.[74]
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 988 – 994 990

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
tions within the crystalline framework, are not affected by the
occupation scheme of the borazine ring. Consequently, bond
lengths and in particular their alternation cannot be reliably
discussed using the X-ray data owing to disorder.
Crystals of 2g suitable for X-ray crystallography were ob-
tained by slow evaporation of the solvent dichloromethane.
Compound 2g crystallizes in the space group P21/c along with
one molecule of dichloromethane in the asymmetric unit (see
Table 2 and Figure 1).
Table 2. Crystal data and structure refinement for 2g.
empirical formula
formula mass
T [K]
C37H23B3Br3Cl2N3
852.64
100(2)
l [ꢂ]
1.54178
crystal system
space group
a [ꢂ]
monoclinic
P21/c
14.2604(2)
b [ꢂ]
c [ꢂ]
8.21000(10)
28.3417(4)
Figure 1. Structure of 2g in the solid state. A dichloromethane molecule
was omitted for clarity. Ellipsoids are drawn at the 50% probability level.
a, b, g [8]
V [ꢂ3]
90, 94.821(1), 90
3306.45(8)
4, 1.713
6.274
1680
Z, 1calcd [MgmÀ3
]
m [mmÀ1
F(000)
]
1) The previously determined pronounced BLA of 0.06 ꢂ in
the overcrowded 1,2:3,4:5,6-tris(biphenylylene)borazine
(2a) results from a misinterpretation of the diffraction data.
A re-investigation of the crystal structure of 2a arrives at
disorder that precludes discussion of the boron–nitrogen
bond lengths.
crystal size [mm3]
0.50ꢄ0.06ꢄ0.04
3.13 to 66.65
À16ꢀhꢀ16
À7ꢀkꢀ9
q range for data collection [8]
index ranges
À33ꢀlꢀ32
29199/5726 (Rint =0.0337)
98.0
reflections collected/unique
completeness to 2q=66.658 [%]
absorption correction
max./min. transmission
structure solution
refinement method
data/restraints/parameters
GOF on F2
2) The sterically overcrowded 2a can be functionalized further
with high chemoselectivity using Br2 (to give 2g), whereas
a number of other methods used for functionalization of ar-
omatic rings fail.
numerical
0.7937/0.1428
direct methods
full-matrix least-squares on F2
5726/0/433
3) X-ray crystallography of 2g demonstrates that the BLA in
the 1,2:3,4:5,6-tris(biphenylylene)borazine system is small.
4) Computations using modern density functional methods
with appropriate corrections for dispersion interactions,
TPSS-D3 and TPSS-D3(BJ), calculate a structure for 2g that
is in very good agreement with experiment. Based on the
good performance for 2g, the computed BLA of the closely
related parent 2a can safely be assumed to be similarly
small.
1.042
final R indices (I>2s(I))
R indices (all data)
R1 =0.0293, wR2 =0.0714
R1 =0.0351, wR2 =0.0743
0.860 and À0.861
largest diff. peak and hole [eAÀ3
]
In contrast to 2a, the boron- and nitrogen-atom positions
could now be clearly located within the borazine ring. It seems
that it is the presence of the three bromine atoms in the pe-
riphery of the molecule that prevents the occurrence of disor-
der in 2g: The bromine atoms participate in the intermolecular
interactions in the crystal and therefore also imply a distinct
occupation scheme in the borazine ring. The average BÀN
bond lengths in 2g are 1.445 and 1.453 ꢂ, respectively, and
thus the BLA is now less than 0.01 ꢂ. This is in good agree-
ment with the theoretical treatments that arrive at a BLA
slightly above 0.01 ꢂ.
Experimental Section
General
All experiments were performed under anhydrous conditions using
argon as the protective gas. Dry dichloromethane was collected
from a MBraun solvent purification system (SPS-800) and xylene
was distilled from P2O5 prior to use. All NMR spectra were recorded
on Bruker DRX 250 and Avance II 400 spectrometers. The spectra
were referenced to residual solvent signals (1H, 13C: SiMe4) and ex-
ternally (11B: BF3·OEt2). The NMR spectra were measured at room
temperature in CD2Cl2 and [D8]THF that were purchased from Deu-
tero GmbH. IR spectra were recorded on a Bruker Tensor 27, ele-
mental analysis was performed using a EuroEA system, LDI mass
spectra were performed on a Bruker AutoFlex with a steel plate,
Conclusion
The experimental and computational results reported here
allow the following conclusions to be drawn:
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 988 – 994 991

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
and APCI mass spectra were acquired on a Bruker Esquire3000plus.
High-resolution mass spectrometry was performed on a Finnigan
MAT MAT95 (HRMS EI) or Bruker maxis4G (HRMS APCI) instrument.
Melting points were determined with a Bꢀchi B-540 heat block.
The borane–triethylamine complex (Sigma–Aldrich), bromine, and
2-aminobiphenyl (Acros) were purchased and used as received.
FWHM is the full width at half-maximum, which was determined
by a Lorentzian fit with TopSpin 2.1 (Bruker).
10-Chloro-6-iPr-9-aza-10-boraphenanthrene (5)
5-iPr-2-aminobiphenyl (4.311 g, 20.4 mmol) that was synthesized as
described previously[76,77] was dissolved in dry xylene (125 mL) and
BCl3 (34 mL, 34 mmol, 1m in hexanes) was added at 08C over
7 min. After stirring at 08C for 30 min and for 30 min at room tem-
perature a 11B{1H} NMR spectrum (d=6.8 ppm (FWHM=13 Hz)) of
the reaction mixture indicated the formation of the amine–boron
adduct. AlCl3 (545 mg, 4.1 mmol) was added to this mixture and
was heated under reflux for 15 h. After the evaporation of the sol-
vent, the residue was distilled at 908C under vacuum (10À3 mbar).
The colorless oil solidifies upon standing (1.353 g, 26%). 1H NMR
(400 MHz, C6D6): d=8.49 (dd, 3J(H,H)=8 Hz, 4J(H,H)=1 Hz, 1H),
8.09 (d, 4J(H,H)=2 Hz, 1H), 7.06 (dd, 3J(H,H)=8 Hz, 4J(H,H)=2 Hz,
1,2:3,4:5,6-Tris(biphenylylene)borazine (2a)
This compound was synthesized according to Kçster et al.[61] with
minor modifications. In a Schlenk tube, N-tri(biphenylyl)borazine
(6; 5.773 g, 10.75 mmol) prepared as described previously[75] was
heated to 4058C with an argon flow of 2.3 sccmminÀ1. During
heat-up the compound melted and a viscous, colorless oil was re-
leased. At 4058C the melt became brownish and after 1 h a color-
less precipitate was formed. After 6.5 h the almost-solid melt was
brought to room temperature. The resulting brown tar was sus-
pended 4 times with dichloromethane (40 mL) and the slurry was
dissolved in dichloromethane (total volume: 800 mL). After filtra-
tion the filter was filled with off-white crystals (432 mg) that were
identified as 4.[63] The solvent of the filtrate was evaporated and
the resulting orange foam was suspended in acetone (30 mL). This
suspension was centrifuged for 5 min at 13.4 krpm. The superna-
tant was discarded and fresh acetone was added. After the three-
fold repetition of this centrifugation procedure the sediment was
air-dried and finally isolated as an off-white, fine powder (1.275 g,
22%). 1H NMR (400 MHz, [D8]THF): d=8.28 (m, 6H), 7.64 (m, 3H),
7.50–7.56 (m, 6H), 7.37 (m, 3H), 7.06–7.11 ppm (m, 6H); MS (LDI-
TOF): m/z (%): 531 (100) [M+].
3
3
1H), 6.91 (brs, 1H), 6.41 (d, J(H,H)=8 Hz, 1H), 2.84 (sept, J(H,H)=
3
7 Hz, 1H), 1.23 ppm (d, J(H,H)=7 Hz, 6H); 13C{1H} NMR (101 MHz,
C6D6): d=142.5, 140.0, 136.6, 134.8, 132.1, 126.72, 126.68, 123.3,
122.5, 121.9, 119.2, 34.4, 24.5 ppm; 11B{1H} NMR (80 MHz, C6D6): d=
34.5 ppm (FWHM=168 Hz); MS (EI quadrupole, 70 eV): m/z (%):
255 (44) [M+], 240 (100) [M+ÀMe], 213 (28) [M+ÀC3H6]; HRMS (EI
sector field, 70 eV): m/z: calcd for C15H15BClN [M+]: 255.09861;
found: 255.09824.
N-Tris(5-iPr-biphen-2-yl)borazine (3b)
5-iPr-2-aminobiphenyl (9.800 g, 46.38 mmol)[76,77] was mixed with
triethylaminoborane (6.81 mL, 46.38 mmol) and heated for 6.5 h at
2058C. The product was recrystallized from hexanes (60 mL) and
isolated as an off-white powder (6.267 g, 61%). M.p. 123–1258C;
1H NMR (250 MHz, CD2Cl2): d=6.05–7.41 (m, 24H), 4.13 (brs, 3H),
2.94 (sept, 3J(H,H)=7 Hz, 3H), 1.28 ppm (d, 3J(H,H)=7 Hz, 18H);
1H{11B} NMR (250 MHz, CD2Cl2): d=6.07–7.44 (m, 24H), 4.29 (brs,
3H), 2.96 (m, 3H), 1.30 ppm (d, 3J(H,H)=6 Hz, 18H); 13C{1H} NMR
(101 MHz, CD2Cl2): d=146.3, 143.5, 141.0, 137.8, 130.7, 129.0, 126.8,
126.3, 34.0, 24.2 ppm; 11B{1H} NMR (80 MHz, CD2Cl2): d=33.5 ppm
(FWHM=953 Hz); IR (KBr): 1401 (B<C-N), 2555 cmÀ1 (BÀH); MS
(APCI-TOF): m/z: 664 [M++H]; HRMS (APCI-TOF): m/z: calcd for
C45H49B3N3 [M+] (monoisotopic): 664.42208; found: 664.42204.
3,11,19-Tribromohexabenzo[a,b,g,i,m,o]-1a,5a,9a-triaza-
4a,8a,12a-triboratriphenylene (2g)
Borazine 2a (115.3 mg, 0.22 mmol) was dissolved in dichlorome-
thane (125 mL) and elemental bromine (1.01 mL, 19.8 mmol) was
added at room temperature over 2.5 min. The red solution was
clipped from the argon and stirred overnight. Then the solution
was cooled and decolorized within 5 min using saturated Na2SO3
solution (50 mL). The solvent was removed and the yellow residue
was suspended in acetone (4 mL) and centrifuged (5 min at
13.4 krpm). The supernatant was removed and replaced by fresh
acetone. After treatment in an ultrasonic bath the suspension was
centrifuged once more (5 min, 13.4 krpm). The sediment was then
dried under vacuum and obtained as a colorless solid (58.2 mg,
35%). For elemental and X-ray analysis the sample was purified on
a column chromatography system using hexane and dichlorome-
thane as eluent and silica gel as the stationary phase. M.p. 312–
3158C; H NMR (400 MHz, [D8]THF): d=8.46 (d, J(H,H)=2 Hz, 3H),
8.30 (d, 3J(H,H)=8 Hz, 3H), 7.58–7.61 (m, 6H), 7.50 (m, 3H), 7.26
(dd, 3J(H,H)=9 Hz, 4J(H,H)=2 Hz, 3H), 7.16 ppm (t, 3J(H,H)=8 Hz,
3H);13C{1H} NMR (101 MHz, [D8]THF): d=140.3, 140.1, 135.0, 132.3,
130.2, 130.0, 128.3, 128.2, 127.1, 124.2, 116.9 ppm; 11B{1H} NMR
(80 MHz, [D8]THF): d=36.1 ppm (FWHM=1000 Hz); IR (KBr):
1363 cmÀ1 (BÀN); UV/Vis (CH2Cl2): lmax =325, 234 nm; fluorescence
(CH2Cl2): lex =246, 324 nm; lem =388 nm; MS (APCI ion trap): m/z:
809 [M++CH3CN+H], 768 [M++H]; elemental analysis calcd (%) for
C37H23B3Cl2N3: C 52.12, H 2.72, N 4.93; found: C 52.10, H 2.71,
N 4.68.
N-Tris(4-bromophenyl)-B-triphenylborazine (6a)
Hexaphenylborazine (170 mg, 0.32 mmol) that was prepared from
N-triphenyl-B-trichloroborazine and phenylmagnesium bromide[78]
was dissolved in dichloromethane (125 mL). At room temperature,
bromine (1.48 mL, 28.80 mmol) was added to the colorless solution
over 3.5 min and was stirred for 18 h. The mixture was washed
twice with saturated Na2SO3 solution (50 mL) and dried over
Na2SO4. After the evaporation of the solvent the crude solid was
subjected to column chromatography on a column chromatogra-
phy system using hexane and dichloromethane as eluent and silica
gel as the stationary phase (Rf(CH2Cl2)=0.81). The product was iso-
lated as an off-white powder (87 mg, 35%). M.p. 330–3338C;
1H NMR (400 MHz, [D8]THF): d=6.94 (m, 6H), 6.86 (m, 15H),
6.67 ppm (m, 6H); 13C{1H} NMR (101 MHz, [D8]THF): d=164.8,
133.0, 132.1, 131.1, 127.3, 127.2, 118.1 ppm; 11B{1H} NMR (80 MHz,
[D8]THF): d=34.8 ppm (FWHM=714 Hz); MS (APCI-iontrap): m/z:
1
4
774 [M++H]; IR (KBr): 1369 cmÀ1 (BÀN); UV/Vis (CH2Cl2): lmax
=
231 nm; fluorescence (CH2Cl2): lex =257 nm; lem =282 nm; HRMS
(APCI-TOF): m/z: calcd for C36H28B3Br3N3 [M++H] (monoisotopic):
772.01093; found: 772.00676.
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 988 – 994 992

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[11] H.-B. Bꢀrgi, K. K. Baldridge, K. Hardcastle, N. L. Frank, P. Gantzel, J. S.
Siegel, J. Ziller, Angew. Chem. 1995, 107, 1575; Angew. Chem. Int. Ed.
Engl. 1995, 34, 1454.
[12] T. M. Krygowski, M. Wisiorowski, S. T. Howard, L. Z. Stolarczyk, Tetrahe-
dron 1997, 53, 13027.
[13] S. T. Howard, T. M. Krygowski, A. Ciesielski, M. Wisiorowski, Tetrahedron
1998, 54, 3533.
[14] A. Stanger, J. Am. Chem. Soc. 1998, 120, 12034.
X-ray diffraction
Suitable single crystals were coated with paratone oil and mount-
ed using a nylon loop. Data collection was performed on a Smart
APEX II diffractometer (2a) using MoKa radiation (l=0.71073 ꢂ) at
153 K, and a Smart APEX II Duo diffractometer (2g) equipped with
a Cu microfocus tube (l=1.54178 ꢂ) at 100 K, respectively. Raw
data were integrated with the program SAINT, and absorption ef-
fects were corrected numerically using SADABS.[79] The structures
were solved by a combination of direct methods (SHELXS) and iter-
ative difference Fourier syntheses (SHELXL).[80] All non-hydrogen
atoms were refined anisotropically, while the hydrogen atoms were
placed at calculated positions and refined using a riding model. As
mentioned before, the atoms within the borazine ring in 2a were
refined using a model of positional disorder for the boron and ni-
trogen atoms. Further details of the refinements and crystallo-
graphic data are listed in Table 2 (2g) and in the Supporting Infor-
mation (2a). Plots were generated using the program ORTEP-3.[81]
CCDC-930452 (2a) and CCDC-930451 (2g) contain the supplemen-
tary crystallographic data for this paper. These data can be ob-
tained free of charge from The Cambridge Crystallographic Data
Centre via www.ccdc.cam.ac.uk/data_request/cif.
´
[15] T. M. Krygowski, M. K. Cyranski, Synlett 2003, 7, 0922.
[16] N. L. Frank, K. K. Baldridge, P. Gantzel, J. S. Siegel, Tetrahedron Lett. 1995,
36, 4389.
[17] N. L. Frank, K. K. Baldridge, J. S. Siegel, J. Am. Chem. Soc. 1995, 117,
2102.
[18] F. Cardullo, D. Giuffrida, F. H. Kohnke, F. M. Raymo, J. F. Stoddart, D. J.
Williams, Angew. Chem. 1996, 108, 347; Angew. Chem. Int. Ed. Engl.
1996, 35, 339.
[19] F. R. Ahmed, J. Trotter, Acta Crystallogr. 1963, 16, 503.
[20] G. Filippini, J. Mol. Struct. 1985, 130, 117.
[21] L. Barnett, D. M. Ho, K. K. Baldridge, R. A. Pascal, Jr., J. Am. Chem. Soc.
1999, 121, 727.
[22] A. Stock, E. Pohland, Ber. Dtsch. Chem. Ges. 1926, 59, 2215.
[23] E. Wiberg, Naturwissenschaften 1948, 35, 182.
[24] E. Wiberg, A. Bolz, Ber. Dtsch. Chem. Ges. 1940, 73, 209.
[25] E. Wiberg, K. Hertwig, Z. Anorg. Chem. 1947, 255, 141.
[26] E. Wiberg, K. Hertwig, Z. Anorg. Chem. 1948, 257, 138.
[27] E. Wiberg, K. Hertwig, A. Bolz, Z. Anorg. Chem. 1948, 256, 177.
[28] C. W. Rector, G. W. Schaeffer, J. R. Platt, J. Chem. Phys. 1949, 17, 460.
[29] B. A. Kozlov, T. K. Rebane, Zh. Fiz. Khim. 1961, 36, 143.
[30] C. B. Chowdhury, R. Basu, J. Indian Chem. Soc. 1968, 45, 469.
[31] D. L. Cooper, S. C. Wright, J. Gerratt, P. A. Hyams, M. Raimondi, J. Chem.
Soc. Perkin Trans. 2 1989, 719.
[32] P. W. Fowler, E. Steiner, J. Phys. Chem. A 1997, 101, 1409.
[33] P. v. R. Schleyer, H. Jiao, N. J. R. van Eikema Hommes, V. G. Malkin, O.
Malkina, J. Am. Chem. Soc. 1997, 119, 12669.
[34] P. v. R. Schleyer, G. Subramanian, H. Jiao, K. Najafian, M. Hofmann in Ad-
vances in Boron Chemistry (Ed.: W. Siebert), Royal Society of Chemistry,
London, 1997, pp. 3.
[35] E. D. Jemmis, B. Kiran, Inorg. Chem. 1998, 37, 2110.
[36] I. D. Madura, T. M. Krygowski, M. K. Cyranski, Tetrahedron 1998, 54,
14913.
Computations
The non-empirical meta-GGA functional introduced by Tao,
Perdew, Staroverov, and Scuseria (TPSS)[73] was employed for all ge-
ometry optimizations using the def2-TZVP basis set in conjunction
with the resolution-of-the-identity (RI) approximation and the rec-
ommended fitting basis sets.[82,83] To take into account the disper-
sion interaction that is not included in the TPSS functional,
Grimme’s dispersion correction with and without Becke–Johnson
damping (D3 and D3(BJ)) was employed.[71,72] All calculations were
performed with the Turbomole program using tightened SCF
(10À8 au) and geometry convergence criteria (10À4).[84]
[37] B. Kiran, A. K. Phukan, E. D. Jemmis, Inorg. Chem. 2001, 40, 3615.
[38] E. Steiner, P. W. Fowler, R. W. A. Havenith, J. Phys. Chem. A 2002, 106,
7048.
[39] A. Y. Timoshkin, G. Frenking, Inorg. Chem. 2003, 42, 60.
[40] D. Benker, T. M. Klapçtke, G. Kuhn, J. Li, C. Miller, Heteroatom. Chem.
2005, 16, 311.
Acknowledgements
This study was supported by the Deutsche Forschungsgemein-
schaft and the Fonds der chemischen Industrie through a fellow-
ship to M.M.
[41] J. J. Engelberts, R. W. A. Havenith, J. H. Van Lenthe, L. W. Jenneskens,
P. W. Fowler, Inorg. Chem. 2005, 44, 5266.
[42] R. Miao, G. Yang, C. Zhao, J. Hong, L. Zhu, Theochem 2005, 715, 91.
[43] A. Soncini, C. Domene, J. J. Engelberts, P. W. Fowler, A. Rassat, J. H. Van
Lenthe, R. W. A. Havenith, L. W. Jenneskens, Chem. Eur. J. 2005, 11, 1257.
[44] E. Steiner, A. Soncini, P. W. Fowler, J. Phys. Chem. A 2006, 110, 12882.
[45] P. K. Chattaraj, D. R. Roy, J. Phys. Chem. A 2007, 111, 4684.
[46] I. Fernꢆndez, G. Frenking, Faraday Discuss. 2007, 135, 403.
[47] R. Islas, E. Chamorro, J. Robles, T. Heine, J. C. Santos, G. Merino, Struct.
Chem. 2007, 18, 833.
Keywords: borazine
·
density functional calculations
·
electrophilic substitution · isoelectronic analogues · structure
elucidation
[48] P. Seal, S. Chakrabarti, J. Phys. Chem. A 2007, 111, 9988.
[49] W. Shen, M. Li, Y. Li, S. Wang, Inorg. Chim. Acta 2007, 360, 619.
[50] W. Harshbarger, G. Lee, R. F. Porter, S. H. Bauer, Inorg. Chem. 1969, 8,
1683.
[1] P. von Raguꢅ Schleyer, H. Jiao, Pure Appl. Chem. 1996, 68, 209.
[2] A. Stanger, Chem. Commun. 2009, 0, 1939.
´
[3] T. M. Krygowski, M. K. Cyranski, Z. Czarnocki, G. Hꢁfelinger, A. R. Katritz-
ky, Tetrahedron 2000, 56, 1783.
´
[51] R. Boese, A. H. Maulitz, P. Stellberg, Chem. Ber. 1994, 127, 1887.
[52] L. R. Thorne, R. D. Suenram, F. J. Lovas, J. Chem. Phys. 1983, 78, 167.
[53] M. Bꢀhl, T. Steinke, P. von Raguꢅ Schleyer, R. Boese, Angew. Chem. 1991,
103, 1179; Angew. Chem. Int. Ed. Engl. 1991, 30, 1160.
[54] M. Sugie, H. Takeo, C. Matsumura, J. Mol. Spectrosc. 1987, 123, 286.
[55] J. W. Barton, A. R. Grinham, J. Chem. Soc. Perkin Trans. 1 1972, 634.
[56] N. P. Hacker, J. F. W. McOmie, J. Meunier-Piret, M. Van Meerssche, J.
Chem. Soc. Perkin Trans. 1 1982, 19.
[4] T. M. Krygowski, M. K. Cyranski, Chem. Rev. 2001, 101, 1385.
[5] R. Diercks, K. P. C. Vollhardt, J. Am. Chem. Soc. 1986, 108, 3150.
[6] R. Faust, E. D. Glendening, A. Streitwieser, K. P. C. Vollhardt, J. Am. Chem.
Soc. 1992, 114, 8263.
[7] A. Stanger, J. Am. Chem. Soc. 1991, 113, 8277.
[8] K. K. Baldridge, J. S. Siegel, J. Am. Chem. Soc. 1992, 114, 9583.
[9] J. S. Siegel, Angew. Chem. 1994, 106, 1808; Angew. Chem. Int. Ed. Engl.
1994, 33, 1721.
[10] R. Boese, D. Blꢁser, W. E. Billups, M. M. Haley, A. H. Maulitz, D. L. Mohler,
K. P. C. Vollhardt, Angew. Chem. 1994, 106, 321; Angew. Chem. Int. Ed.
Engl. 1994, 33, 313.
[57] D. PeÇa, A. Cobas, D. Pꢅrez, E. Guitiꢆn, L. Castedo, Org. Lett. 2000, 2,
1629.
[58] D. PeÇa, D. Pꢅrez, E. Guitiꢆn, L. Castedo, Org. Lett. 1999, 1, 1555.
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 988 – 994 993

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[59] D. PeÇa, D. Pꢅrez, E. Guitiꢆn, Chem. Rec. 2007, 7, 326.
[60] R. Kçster, S. Hattori, Y. Morita, Angew. Chem. 1965, 77, 719.
[61] R. Kçster, K. Iwasaki, S. Hattori, Y. Morita, Liebigs Ann. Chem. 1968, 720,
23.
[62] P. J. Roberts, D. J. Brauer, Y.-H. Tsay, C. Krꢀger, Acta Crystallogr. Sect. B
1974, 30, 2673.
[63] S. Biswas, M. Mueller, C. Toenshoff, K. Eichele, C. Maichle-Moessmer, A.
Ruff, B. Speiser, H. F. Bettinger, Eur. J. Org. Chem. 2012, 4634.
[64] T. Keszthelyi, G. Balakrishnan, R. Wilbrandt, W. A. Yee, F. Negri, J. Phys.
Chem. A 2000, 104, 9121.
[65] T. Ishiyama, J. Takagi, K. Ishida, N. Miyaura, N. R. Anastasi, J. F. Hartwig, J.
Am. Chem. Soc. 2002, 124, 390.
[72] S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456.
[73] J. Tao, J. P. Perdew, V. N. Staroverov, G. E. Scuseria, Phys. Rev. Lett. 2003,
91, 146401.
[74] S. Grimme, C. Mꢀck-Lichtenfeld, Isr. J. Chem. 2012, 52, 180.
[75] R. Kçster, H. Bellut, S. Hattori, Liebigs Ann. Chem. 1968, 720, 1.
[76] H.-I. Chang, H.-T. Huang, C.-H. Huang, M.-Y. Kuo, Y.-T. Wu, Chem.
Commun. 2010, 46, 7241.
[77] K. Venkateswarlu, K. Suneel, B. Das, K. N. Reddy, T. S. Reddy, Synth.
Commun. 2008, 39, 215.
[78] S. J. Groszos, S. F. Stafiej, J. Am. Chem. Soc. 1958, 80, 1357.
[79] Bruker, APEX, SAINT, SADABS and XPREP. Area detector control and
data integration and reduction software control and data integration
and reduction software, Bruker Analytical X-ray Instruments Inc., Madi-
son, WI, USA, 2008.
[66] T. Ishiyama, J. Takagi, J. F. Hartwig, N. Miyaura, Angew. Chem. 2002, 114,
3182; Angew. Chem. Int. Ed. 2002, 41, 3056.
[67] N. Iranpoor, H. Firouzabadi, N. Nowrouzi, D. Firouzabadi, Tetrahedron
Lett. 2006, 47, 6879.
[80] G. M. Sheldrick, Acta Crystallogr. Sect. A 2008, 64, 112.
[81] L. J. Farrugia, J. Appl. Crystallogr. 1997, 30, 565.
[68] J. J. P. Selvam, V. Suresh, K. Rajesh, S. R. Reddy, Y. Venkateswarlu, Tetrahe-
dron Lett. 2006, 47, 2507.
[82] F. Weigend, Phys. Chem. Chem. Phys. 2006, 8, 1057.
[83] F. Weigend, R. Ahlrichs, Phys. Chem. Chem. Phys. 2005, 7, 3297.
[84] TURBOMOLE V6.4 2012, a development of University of Karlsruhe and
Forschungszentrum Karlsruhe GmbH, 1989–2007, TURBOMOLE GmbH,
since 2007; available from http://www.turbomole.com.
[69] H. G. O. Becker, R. Beckert, W. Berger, G. Domschke, E. Fanghꢁnel, J.
Faust, M. Fischer, F. Gentz, K. Gewald, R. Gluch, W. D. Habicher, H.-J.
Knçlker, R. Mayer, P. Metz, K. Mꢀller, D. Pavel, H. Schmidt, K. Schollberg,
K. Schwetlick, E. Seiler, G. Zeppenfeld, Organikum, Wiley-VCH, Wein-
heim, 2009.
[70] A. Bondi, J. Phys. Chem. 1964, 68, 441.
Received: March 25, 2013
[71] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104.
Published online on May 17, 2013
This paper is part of a Special Issue dedicated to the memory of Detlef Schrçder. To view the complete issue, visit:
http://onlinelibrary.wiley.com/doi/10.1002/cplu.v78.9/issuetoc
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2013, 78, 988 – 994 994