Reference values for the B-X bond lengths of BI3 and BBr 3

Article abstract of DOI:10.1002/zaac.200700510

High-resolution X-ray analysis of BI₃ and BBr₃ single crystals as well as high-level theoretical calculations agree on accurate but long B-X bond distances that should be used as future reference to account for their structure, bondi

Full text of DOI:10.1002/zaac.200700510

DOI: 10.1002/zaac.200700510
Reference Values for the B-X Bond Lengths of BI3 and BBr3
˜
G. Santiso-Quinones and I. Krossing*
Freiburg, Germany, Institut für Anorganische Chemie der Albert-Ludwigs-Universität
Received September 14th, 2007; accepted November 13rd, 2007.
Abstract. High-resolution X-ray analysis of BI3 and BBr3 single
crystals as well as high-level theoretical calculations agree on accu-
rate but long B-X bond distances that should be used as future
reference to account for their structure, bonding and reactivity.
Keywords: boron trihalides; high resolution X-ray crystallography;
ab initio calculation; DFT calculation; Lewis acid
Introduction
Crystal growth out of solution
BI3: Sublimed BI3 (10 g) was put into a previously dried Schlenk
vessel equipped with a Young valve with glass stem. Hexane was
condensed into the flask (ca. 20 ml). The flask was put inside a
dewar vessel (5 l) containing ethanol. The dewar vessel was placed
inside a 0 °C refrigerator. After 2 days at this temperature crystals
suitable for X-ray diffraction were obtained.
It is well known that the BX3 molecules do not obey the
expected trend in Lewis acidity and, surprisingly, BF3 is a
much weaker Lewis acid than BI3 [1]. The unexpected be-
havior of the boron trihalides may be explained in terms of
σ- and π-bond donation [2], even though not everybody [3,
4] necessarily agrees with these findings. Experimentally, the
extent of halogen-boron π-bonding has been investigated
on the basis of core binding energies [5a], mass spectro-
metric analyses of trimethylamine adducts [5b] as well as
IR and Raman studies of phosphine adducts [5c]. So far,
many theoretical [2Ϫ4, 6] but very few experimental reports
address the problem and there is a clear need for further
experimental data on BX3 (X ϭ F-I) to finally resolve this
long standing problem. One input for this discussion is an
accurate determination of the B-X bond length. While these
are known from gas phase studies on BF3 [7a-c] and BCl3
[7c], the available data for BBr3 are only reasonable but
those of BI3 are less accurate and in part even lack the
assignment of a standard deviation [8].
BBr3: Distilled BBr3 (5 g) was condensed into a previously dried
Schlenk vessel equipped with a J. Young valve with glass stem.
Hexane was condensed into the flask (ca. 10 ml). The flask was put
inside a dewar vessel containing 5 L ethanol. The dewar vessel was
placed inside a Ϫ78 °C refrigerator. After 4 days at this temperature
crystals suitable for X-ray diffraction were obtained.
Crystal growth inside a capillary
BBr3: Distilled BBr3 was condensed into a previously dried Linde-
mann glass capillary (outer diameter 0.2 mm) which was flamed
sealed under vacuum. Using a procedure similar to the one re-
ported in the literature [10] a single crystal was grown in situ on
the diffractometer by slowly cycling the temperature of the cryo-
stream between 227 K and 203 K. The crystallinity of the sample
was checked every 3 seconds during the cycling procedure by the
quality of the frames obtained. Close to the point where the sample
started to give single crystal reflections, heating the sample with a
green Laser (532 nm, 0.1 mA) for some seconds produced the de-
sired homogeneity of the sample. After this point the sample was
slowly (ϳ 30 K/h) cooled down to 100 K, the temperature at which
the measurement was performed. With this procedure reproducibly
single crystals of BBr3 could be prepared.
Experimental Section
All manipulations were carried out inside a glove box with inert
gas atmosphere. Hexane was dried by distillation from Na and
stored in a greaseless Schlenk vessel over molecular sieves (4 nm).
Commercially available BBr3 was distilled and stored in a greaseless
Schlenk vessel containing some milligrams of Hg. BI3 was prepared
from NaBH4 and I2 in heptane according to reported procedures
[9].
Data collection
For both structures, data collection was performed at 100 K on a
Nonius Kappa goniometer equipped with a Bruker APPEX II
CCD detector and using Mo Kα (0.70173 A) radiation.
Prof. Dr. Ingo Krossing, Dr. Gustavo Santiso-Quin˜ones
Institut für Anorganische Chemie
Albert-Ludwigs-Universität Freiburg
Albertstr. 21
D-79104Freiburg / Germany
Tel: ϩ49(0)7612036122
˚
A BI3 single crystal was mounted at Ϫ70 °C on a specially designed
device using a stream of cold gaseous N2. Data collection was per-
˚
formed up to a limit of 80° in 2θmax. a ϭ b ϭ 6.9966 A, c ϭ
˚
7.2615 A, α ϭ β ϭ 90°, γ ϭ 120°. Z ϭ 2, space group P63/m,
Fax:ϩ49(0)7612036001
E-mail: krossing@uni-freiburg.de, gustavo.santiso@ac.uni-freiburg.de
2θmax ϭ 80°, 24375 reflections collected (659 unique), R1 ϭ 0.0156
704
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2008, 634, 704Ϫ707

Reference Values for the B-X Bond Lengths of BI3 and BBr3
for 582 Fo > 4sig(Fo) and 0.0221 for all 659 data and 10 param-
minimize absorption effects. The refinement has higher
accuracy than previously reported values and should be
used as a basis for further discussions of the bonding in BI3
(σ- and π-contributions, Coulomb interaction).
eters, wR2 ϭ 0.0273, GooF ϭ S ϭ 1.367. Crystal size ϭ 0.02 x 0.02
3
˚
x 0.02 mm. max/min larg. res. peak, 1.01/Ϫ0.95 e/A . Absorption
correction type: Spherical (r ϭ 0.01). Tmin ϭ 0.7431, Tmax
ϭ
0.7472.
The intramolecular I-I distances (409 pm) from mol-
ecules within the same plane are longer than the sum of the
van der Waals radii (rB ϭ 200 pm, rI ϭ 198 pm) [15] see
figure 1. All I-I and B-I intramolecular contacts to mol-
ecules in the upper and lower planes are above 409 pm.
From this, it is evident that the B-I bond is not affected by
weak intermolecular interactions in the unit cell and its
length relies almost exclusively on the nature of the interac-
tion between the boron and iodine atoms composing the
molecule.
BBr3 inside the capillary was measured up to 72° in 2θmax. a ϭ b ϭ
˚
˚
6.4304 A, c ϭ 6.8466 A, α ϭ β ϭ 90°, γ ϭ 120°. Z ϭ 2, space
group P63/m, 2θmax ϭ 72°, 15634 reflections collected (426 unique),
R1 ϭ 0.0261 for 372 Fo > 4sig(Fo) and 0.0339 for all 426 data,
wR2 ϭ 0.0628, GooF ϭ S ϭ 1.386. Crystal size (cylindrical): r ϭ
3
˚
0.099 mm, l ϭ 0.3 mm. max/min larg. res. peak, 1.11/Ϫ0.69 e/A .
Cylindrical absorption correction (r ϭ 0.08 mm). Tmin ϭ 0.0472,
Tmax ϭ 0.0853.
Both structures were solved (SHELX 6.14) by the Patterson heavy
atom method and successive interpretation of the difference
Fourier maps, followed by least-square refinement. All atoms were
refined anisotropically.
Calculations
DFT (BP86), ab initio (MP2) and hybrid (PBE0) computations
were done with the program TURBOMOLE [11a]. The geometries
of all species were optimized with the spilt valence polarization
SVP, the TZVP, or the TZVPP basis sets respectively, as im-
plemented in the program [11b]. Frequency calculations were
performed at the BP86/SVP level, and all structures represent true
minima without imaginary frequencies on the respective hyper sur-
face. Hybrid HF-DFT (PBE1PBE ϭ PBE0 from Turbomole) and
CCSD(T) calculations using (SDB)-Aug-cc-pVTZ and (SDB)-
Aug-cc-pVQZ as basis set were performed with the program
GAUSSIAN 03 Revision C.02 [12]. In all cases the core electrons
of Br (10 e) and I (28 e) were replaced by a scalar relativistic
ECP. The basis sets and ECP’s were downloaded from
http://emsl.pnl.gov/forms/basisform.html, references to these basis
sets are quoted there.
Fig. 1 In plane packing of the BX3 molecules showing the inter-
molecular B-X and the shortest intramolecular X-X distances. a)
BI3, b) BBr3. Thermal ellipsoids are drawn at 50 % probability
level.
Results and Discussion
As a cross check we performed a series of quantum
chemical calculations. As a quantum chemical reference the
BI3 distance was assessed at the highly correlated CCSD(T)
level using the very flexible (SDB)-Aug-cc-pVQZ basis set.
This calculation essentially reproduced the experimental
value within 0.1 pm, i.e. B-ICCSD(T) ϭ 212.6 pm (see table
1). By contrast, standard DFT calculations (BP86) over-
estimate the B-I bond length, while ab initio calculations
(MP2) and hybrid HF-DFT methods with very flexible QZ
basis sets predict shorter B-I bonds (211.3 to 211.8 pm). It
is noteworthy that the B-I bonds shrink for MP2/HF-DFT
from values larger than the experiment to a distance shorter
than the experimental bond length if going from a DZ to a
TZ and QZ basis. Only the highly correlated CCSD(T)
method with a scalar relativistic effective core potential and
a flexible valence basis for I reproduced the experimental
B-I distance. This shows the necessity to employ highly ac-
curate experimental and theoretical methods to understand
the true nature of the B-X bond in the simple but delicate
boron trihalides.
We report here high resolution single crystal X-ray data for
BI3 and BBr3. BI3 single crystals were grown either by subli-
mation or from a hexane solution; the later provided the
best crystallographic data sets. Single crystals were grown
by slowly cooling a saturated hexane solution from room
temperature to 0 °C. As in previous powder experiments
[13]. BI3 crystallizes in the space group P63/m and the mol-
ecules are trigonal planar (see figure 1a). Our best data set
(2θmax ϭ 80°, R1 ϭ 1.56 %, wR2 ϭ 2.73 %) gave a B-I bond
length of 212.51(3) pm, which is longer than the spectro-
scopic value provided by Binbrek et al. (212 pm, no stand-
ard deviation given) [8] or the x-ray powder data [13] report
which concluded a value of 211.2(8) pm. Correction for li-
bration by a riding model [14] did not change our B-I dis-
tance. We note here that according to our experience the B-I
bond length depends critically on the quality of the crystal,
apparently with lower crystal quality leading to shorter B-
I distances. The new data set presented here emerged as the
last of a series of continuous improvements and was meas-
ured from a very small crystal of 0.02x0.02x0.02 mm to
Z. Anorg. Allg. Chem. 2008, 704Ϫ707
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
705

G. Santiso-Quin˜ones, I. Krossing
For BBr3, a liquid with a melting point of Ϫ46 °C, the
crystallization process in hexane at Ϫ78 °C is as straightfor-
ward as for BI3. However, the mounting of single crystals
has been problematic due to the reactivity of the material.
We have been able to mount and measure single crystals of
BBr3 grown from a hexane solution, but the quality of these
data sets was only average. Nevertheless, we reproducibly
have been able to obtain good data sets (2θmax ϭ 72°, R1 ϭ
0.0261, wR2 ϭ 0.0628) of BBr3 single crystals grown in situ
on the cryo-stream of the diffractometer (see figure 1b),
using a procedure similar to the one reported in the litera-
ture [10]. BBr3 crystallizes in the same space group as BI3,
namely P63/m. To the best of our knowledge, there is no
literature value available for the B-Br bond length in the
condensed phase [16]. Gas electron diffraction of BBr3 sug-
gested d(B-Br)GED ϭ 190.0(4) pm [17]. This is in very good
agreement with the value that we have obtained for the con-
densed phase, namely d(B-Br)x-ray ϭ 189.85(5) pm. It
should be noted that the standard deviation in our structure
is by about one order of magnitude smaller. However, cor-
rection for libration [14] according to a riding model gives
a corrected d(B-Br)x-ray,libr of 190.3 pm indicating that some
residual libration may be present at the temperature of the
measurement (100 K).
Agreements to experimental values from previously re-
ported calculations (e.g Ref’s [2Ϫ4, 6]) using any other
method with lower quality basis sets as the one we have
used were in part fortuitous, due to the inferior quality of
the experimental B-X distance to which they were com-
pared. Future analysis of structure and bonding of the
heavier BX3 molecules should be done with reference to the
values reported here.
In conclusion we note that according to high quality x-
ray crystal structures as well as highly correlated ab initio
calculations with very flexible basis sets the B-X bond
lengths of the heavier BX3 molecules are rather long. Of
course that does not answer the question about the bonding
in BI3 and BBr3, but gives an insight that πϪback donation
might be of minor importance. With this work we have
shown that it is possible to obtain very good experimental
crystallographic data sets for both BI3 and BBr3 which are
in very good agreement with high level theoretical calcu-
lations. However, we failed to obtain data sets sufficient to
be used for a multipole refinement [18] in order to deter-
mine as accurate as possible the electron distribution in
such delicate systems. This would indeed give more insight
on the real bonding in BI3 and BBr3.
In BBr3, the shortest intermolecular B-Br and Br-Br dis-
tances are as long as the sum of the van der Waals radi
(rB ϭ 200 pm, rBr ϭ 185 pm) [15], again providing evidence
that the crystal structure provides direct information on the
nature of the undistorted B-Br bond.
As in the case of BI3, the DFT calculations overestimate
the B-Br bond length. The MP2 and the hybrid HF-DFT
calculations gave shorter bond lengths even with a very flex-
ible QZ basis set (189.3 to 189.6 pm). Again, only a highly
correlated method like CCSD(T) gives a converged distance
close to the experimental x-ray value corrected for libration,
i.e. d(B-Br)CCSD(T) ϭ 190.2 pm, see table 1.
Notes and References
[1] A short (incomplete) selection: a) F. Bessac, G. Frenkin, Inorg.
Chem. 2003, 42, 7990Ϫ7994. b) C. Aubauer, E. Irran, T. M.
Klapötke, W. Schnick, A. Schulz, J. Senker, Inorg. Chem. 2001,
40, 4956Ϫ4965. c) V. J. Apel, J. Grobe, Z. Anorg. Allg. Chem.
1979, 453, 28Ϫ36. d) J. R. McDivitt, G. L. Humphrey, Spec-
trochim. Acta 1974, 30A, 1021Ϫ1033. e) T. B. Eames, B. M
Hoffman, J. Am. Chem. Soc. 1971, 93, 3141Ϫ3146. f) G. F.
Lanthier, J. M. Miller, J. Chem. Soc. (A). 1971, 346Ϫ350.
[2] a) V. Branchadell, A. Oliva, J. Mol. Struct. 1991, 236, 75Ϫ84.
b) G. Frenking, S. Fau, C. M. Marchand, H Grützmacher, J.
Am. Chem. Soc. 1997 119, 6648Ϫ6655.
[3] a) H. P. A. Mercier, M. D. Moran, G. J. Schrobilgen, R. J.
Suontamo, J. Fluorine. Chem. 2004, 125, 1563Ϫ1578. b) I.
Krossing, I. Raabe, J. Am. Chem. Soc. 2004, 126, 7571Ϫ7577.
[4] a) H. Hirao, K. Omoto, H. Fujimoto, J. Phys. Chem A. 1999,
103, 5807Ϫ5811. b) E. A. Robinson, S. A. Johnson, T. H.
Tang, R. J. Gillespie, Inorg. Chem. 1997, 36, 3022Ϫ3030.
[5] a) D. B. Beach, W. L. Jolly, J. Phys. Chem. 1984, 88,
4647Ϫ4649. b) G. F. Lanthier, J. M. Miller, J. Chem. Soc. (A)
1971, 346Ϫ350. c) J. R. Durig, S. Riethmiller, Y. S. Li, Inorg.
Chem. 1974, 13, 2729Ϫ2735.
Table 1 Comparison of the B-X bond lengths in BX3 (X ϭ I, Br)
molecules based on experimental work and theoretical calculations.
Values in [ ] were taken from the literature. All values in pm.
DZ
TZ(d) TZ(2d,f) QZ(3d,2f,1g)
b)
BP86
BBr3 PBE0
MP2
191.5 191.6 191.2
189.9 189.9 189.7
190.5 190.4 189.8
191.2
189.6
189.3
190.2
Experimental
[190.0(4)]
b)
b)
a)
189.85(5)
[6] D. Yu, Z-D Chen, F. Wang, S-Z. Li, Gaodeng, Xuexiao
Huvaxue Xuebao. 2001, 22, 1193Ϫ1196.
c)
d)
CCSD(T)
Ϫ
Ϫ
190.2
190.3
DZ
TZ(d) TZ(2d,f) QZ(3d,2f,1g)
[7] a) S. Yamamoto, R. Kuwabara, M. Takami, K. Kuchitsu, J.
Mol. Spectrosc. 1986, 115, 333Ϫ352. b) K. Kuchitsu, S.
b)
BP86
PBE0
MP2
214.6 214.0 213.4
213.0 212.5 212.0
213.4 212.6 211.4
213.3
211.8
211.3
212.6
´
Konaka, J. Chem. Phys. 1966, 45, 4342Ϫ4347. c) H. A. Levy,
b)
BI3
Experimental
[212]
212.51(3)
L. O. Brockway, J. Am. Chem. Soc. 1937, 59, 2085Ϫ2092.
[8] O. S. Binbrek, N. Krishnaurthy, A. Anderson, J. Chem. Phys.
1974, 60, 4400.
[9] A. G. Briggs, Naturwissenschaften 1990, 77, 595Ϫ597.
[10] a) V. R. Thalladi, R. Boese, H. C. Weiss, J. Am. Chem. Soc.
2000, 122, 1186Ϫ1190. b) V. R. Thalladi, H. C. Weiss, D.
Bläser, R. Boese, A Nangia, G. R. Desiraju, J. Am. Chem. Soc.
1998, 120, 8702Ϫ8710.
b)
e)
c)
CCSD(T)
Ϫ
Ϫ
212.7
a) See ref.[17]. b) DZ ϭ def-SV(P), TZ(d) ϭ def-TZVP, TZ(2d,f) ϭ
def-TZVPP, QZ(3d,2f,1g) ϭ (SDB)-Aug-cc-pVQZ. c) TZ(2d,f) ϭ
(SDB)-Aug-cc-pVTZ, QZ(3d,2f,1g)
ϭ
(SDB)-Aug-cc-pVQZ.
d) Corrected for libration by a riding model, see ref. [14]. e) See
ref. [8].
706
www.zaac.wiley-vch.de
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Z. Anorg. Allg. Chem. 2008, 704Ϫ707

Reference Values for the B-X Bond Lengths of BI3 and BBr3
[11] a) TURBOMOLE Version 6: K. Eichkorn, O. Treutler, H.
Oehm, M. Haeser, R. Ahlrichs, Chem. Phys. Lett. 1995, 242,
652Ϫ660. b) For the corresponding literature for the basis sets
see ftp://ftp.chemie.uni-karlsruhe.de/pub/basen.
M. W. Wong, C. Gonzalez, J. A. Pople, Gaussian, Inc., Wall-
ingford CT, 2004.
[13] B. Albert, K. Schmitt, Z. Anorg. Allg. Chem. 2001, 627,
809Ϫ810.
[12] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M.
A. Robb, J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven,
K. N. Kudin, J. C. Burant, J. M. Millam, S. S. Iyengar, J. Tom-
asi, V. Barone, B. Mennucci, M. Cossi, G. Scalmani, N. Rega,
G. A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toy-
ota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y.
Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E. Knox, H.
P. Hratchian, J. B. Cross, C. Adamo, J. Jaramillo, R. Gom-
perts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C.
Pomelli, J. W. Ochterski, P. Y. Ayala, K. Morokuma, G. A.
Voth, P. Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dap-
prich, A. D. Daniels, M. C. Strain, O. Farkas, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. V. Ortiz,
Q. Cui, A. G. Baboul, S. Clifford, J. Cioslowski, B. B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. L. Martin,
D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayak-
kara, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen,
[14] C. K. Johnson, Crystallogr. Computing 1970, 220Ϫ226.
[15] The van der Waals radius for boron was assumed to be 200 pm
but most probably is even shorter. See notes and references in
the CSC web page: http://www.ccdc.cam.ac.uk/products/csd/radii.
[16] We have checked the ICSD and the SciFinder data base for a
possible crystal structure of BBr3. The only reported x-ray
data of this compound is based on powder diffraction from
1947 (Rolier M. A., Riva, A. Gaz. Chim. Ital. 1947, 77,
361Ϫ366.), there is no B-Br bond length given in this article.
[17] K.-G. Martinsen, N. Vogt, H. Vidar Volden, V. S. Lyutsarev,
J. Vogt, J. Mol. Struct. 1996, 385, 159Ϫ165.
[18] As part of a collaboration with Prof. Dr. Niels Hansen,
´
´
Universite H. Poincare, Nancy, France. a) N. Hansen, P.
Coppens, Acta Crystallogr. 1978, A34, 909Ϫ921. b) M. Tafi-
polsky, W. Scherer, K. Oefele, G. Artus, B. Pedersen, W. A.
Hermann, G. S. McGrady, J. Am. Chem. Soc. 2002, 124,
5865Ϫ5880.
Z. Anorg. Allg. Chem. 2008, 704Ϫ707
© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
www.zaac.wiley-vch.de
707