Crystal structure of the molecular addition compound diphenylchloroborane tetrahydrofuran

Article abstract of DOI:10.1021/ic991148q

Crystalline Ph₂BCl·O(CH₂)₄ is an unexpectedly stable complex containing an intact THF ligand in the presence of a boron halide bond. The short B-O and long B-Cl bonds of this ether cleavage precursor complex are rationalized using Gillespie's ligand close-packing arguments.

Full text of DOI:10.1021/ic991148q

2690
Inorg. Chem. 2000, 39, 2690-2692
Notes
Crystal Structure of the Molecular Addition
Compound
Diphenylchloroborane‚Tetrahydrofuran
Wendy I. Cross, Matthew P. Lightfoot,
Francis S. Mair,* and Robin G. Pritchard
Department of Chemistry, UMIST, P.O. Box 88, Manchester
M60 1QD, U.K.
ReceiVed September 27, 1999
Introduction
Figure 1. Molecular structure of 2. See Table 2 for significant bond
lengths and angles.
A very common usage of haloboranes in organic synthesis
is in the cleavage of ethers.1 Diorganohaloboranes in particular
are valued in this capacity.2 Ether cleavage is particularly
facilitated in the case of cyclic ethers, such as tetrahydrofuran
(THF), so much so that it is common practice to employ low
temperatures when utilizing haloboranes in such solvents.
Indeed, cleavage of tetrahydrofuran can even occur with short
reaction times at 0 °C, such as was observed in the only
crystallographic confirmation of the course of a boron-mediated
ether cleavage reaction in the literature,3 that of tetraphe-
nylphosphonium 2-(4-chlorobut-1-oxy)-1-carba-closo-undecahy-
drododecaborate(1-) (1; Chart 1). This was obtained from the
insertion reaction of BCl3‚SMe2 with Li3[CB10H11] in THF.
It was particularly surprising for us to discover, therefore,
that from the reaction of diphenylchloroborane with a THF
solution of tetralithiated tetra-tert-butyl calix(4)arene, the well-
formed colorless crystals isolated were in fact of the pseudot-
etrahedral complex Ph2BCl‚THF (2, Figure 1).
Chart 1
Table 1. X-ray Data Collection and Processing Parameters for 1
molecular formula
fw
C16H18BClO
272.56
temp/°C
wavelength/Å
space group
-70(2)
0.710 73
P21/c
a/Å
b/Å
c/Å
9.087(2)
14.019(4)
11.680(2)
108.324(10)
1412.4(6)
4
1.282
0.258
0.0301, 0.0821
Experimental Section
â/deg
V/Å3
Ph2BCl and tetra-tert-butyl calix[4]arene were prepared according
to literature methods.4,5 To a THF solution (7 mL) of tetra-tert-butyl
calix[4]arene (0.34 g, 0.52 mmol) was added dropwise nBuLi (1.35 M
in hexanes, 1.56 mL, 2.1 mmol). The solution was stirred for 2 h, and
an orange solution was formed. The solution was cooled to -78 °C,
and BPh2Cl (0.58 g, 2.89 mmol) in hexane (2 mL) was added. Gradual
warming produced a white, cloudy suspension. As the solution reached
room temperature, a clear, yellow solution was formed that was left
stirring for 1 h. Some of the solvent (6 mL) was removed under vacuum,
and the remaining solution was placed at -50 °C to form colorless
Z
D(calcd)/g cm-3
abs coeff/mm-1
final R indices: R1,a wR2b
a R1 ) ∑||Fo| - |Fc|/∑|Fo| (observed reflections). b wR2 ) [∑w(Fo
2
- Fc2)2/∑(wFo )2]1/2 (all data).
2
Results and Discussion
1
rhomboidal crystals of 2, mp ) 124-130 °C. H NMR (200 MHz,
The successful structural characterization of 2 constitutes the
first case of THF found intact in the presence of a boron-
halogen bond. We therefore have confirmed structural data on
both the precursor and postcursor3 of the ether cleavage reaction.
C6D6): δ 8.08 (dd, 4H), 7.34 (m, 6H), 3.61 (m, 4H, THF), 1.12 (m,
4H, THF). Found: C, 70.28; H, 6.91; Cl, 12.89. Calcd for C16H18-
BClO(%): C, 70.50; H, 6.66; Cl, 13.00. Details of the structure analysis
are summarized in Table 1.
In 2, the C-O lengths (see Table 2) of 1.4835(18) and
1.4899(18) Å are longer than those in free THF (1.429 Å at
-175 °C, 1.435 Å at -125 °C) and significantly longer than
the mean value of all crystallographically determined uncoor-
dinated THF C-O bond lengths (1.403 Å).6 This lengthening
can be ascribed to the Lewis acid nature of the boron. The
lengthening is significant, the THF C-O bond lengths in 2 being
among the longest known for accurately determined6 structures.
(1) March, J. AdVanced Organic Chemistry, 4th ed.; Wiley-Interscience,
New York, 1992; pp 433-434.
(2) Guindon, Y.; Yoakim, C.; Morton, H. C. Tetrahedron Lett. 1983, 24,
2969. Guindon, Y.; Therien, M.; Girard, Y.; Yoakim, C. J. Org. Chem.
1987, 52, 1680 and references therein.
(3) Mair, F. S.; Morris, J. H.; Gaines, D. F.; Powell, D. J. Chem. Soc.,
Dalton Trans. 1993, 135.
(4) Niedenzu, K.; Bayer, H.; Dawson, J. W. Inorg. Chem. 1962, 1, 738.
(5) Gutsche, C. D.; Iqbal, M. Org. Synth. 1985, 68, 234.
10.1021/ic991148q CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/11/2000

Notes
Inorganic Chemistry, Vol. 39, No. 12, 2000 2691
Table 2. Bond Lengths [Å] and Angles [deg] for 2
and hence maximizes its electrostatic attraction for chloride.
Cl(1)-B(1)
O(1)-C(13)
O(1)-C(16)
1.8932(17)
1.4835(18)
1.4899(18)
O(1)-B(1)
B(1)-C(1)
B(1)-C(7)
1.5691(19)
1.599(2)
1.602(2)
The longest, 1.799 Å in 1-chloroborepin (C6H6BCl), is presum-
ably ascribable to the involvement of the carbocyclic π system
in reducing the positive charge on boron.12 The donation of π
density from a carbon to boron is much more plausible on the
grounds of electronegativity than from a halide to boron. The
full range of B-Cl lengths in these three-coordinate analogues
lies outside the observed B-Cl length of 1.8932(17) Å in 2.
This value is in fact very close to the mean B-Cl value (1.905
Å) found in four-coordinate neutral Lewis base adducts of
diorganochloroboranes.13 The systematic lengthening on in-
creasing coordination number is in full accord with Gillespie’s
“ligand-close-packing” model of molecular structure. There is
no necessity to invoke lost back-bonding as a further reason
for lengthening in the case of group 13 halides. The most
pertinent comparisons to draw, then, are between similar four-
coordinate ether solvates. Experimental data are available for
Ph3B‚THF14 but, of course, not for the reactive Cl3B‚THF.
However, a model of Cl3B‚OMe2 has been recently computed
at the B3LYP density functional level.9
The B-O bond is marginally shorter in 2 than in the
(computed) Cl3B‚OMe2 (1.633 Å)9 because the carbon atoms
of the phenyl groups take up less room around the boron atom
than chlorine ligands, allowing closer approach of the ether
oxygen in 2. The B-O bond in 2 is shorter than in Ph3B‚THF
(1.660(4) Å),14 despite the presence of the chlorine, because
the positive charge on boron in 2 is greater than in Ph3B‚THF.
This short B-O bond in 2 is at the expense of a long B-Cl
bond; the presence of three first-row elements in the coordination
sphere of boron forces the remaining, more weakly held chlorine
ligand out further than in comparable trichloroborane complexes
(Cl3B‚OMe2, 1.831 and 1.844 Å;9 Cl3B‚NMe3, 1.839(9) and
1.827(9) Å).15 The charge on boron in monochlorinated 2 is
presumably less than in these two trichloroborane complexes,
thus re-enforcing the ligand close-packing argument. Thus, the
ionic model, in concert with ligand close-packing, deals
satisfactorily with the apparent paradox of an isolable compound
with a shorter B-O and a longer B-Cl bond than those
computed for a reactive and as yet unisolated Cl3B analogue.
C(13)-O(1)-C(16) 108.80(11)
C(13)-O(1)-B(1) 118.51(11)
C(16)-O(1)-B(1) 118.00(11)
O(1)-B(1)-C(1)
O(1)-B(1)-C(7)
C(1)-B(1)-C(7) 114.75(12)
O(1)-B(1)-Cl(1) 104.14(10)
C(1)-B(1)-Cl(1) 111.17(10)
C(7)-B(1)-Cl(1) 111.65(11)
C(6)-C(1)-C(2) 116.66(14)
106.77(11)
107.61(12)
While the isolation and characterization of 2 does not directly
proVe that it is the precursor to the ether cleavage reaction, this
observation strengthens the view of 2 as a model for the
immediate precursor to B-X insertion into the C-O bond.
However, the presence of the two phenyl groups appears to
moderate the reactivity of the B-X bond. Gillespie has recently
championed a novel way of rationalizing the structures and
behaviors of many first-row compounds, including haloboranes,
using the results of the atoms-in-molecules approach.7 This view
emphasizes the ionic component of bonding8 and asserts that
arguments based on pπ-pπ interaction are likely to be spurious
because it is not chemically sensible to consider a halogen atom
as donating π-electron density to a boron atom on grounds of
electronegativity.8 The conventional argument may run thus.
The phenyl groups, unlike halides, do not donate π density.
Therefore, they free the boron’s unutilized p orbital to accept
π density from the single remaining chlorine; indeed, through
their electron-withdrawing capacity, they increase the demand
of the boron for π density from the chlorine. A less reactive
B-Cl bond than in BCl3 results. However, the B-Cl bond
length in 2 is longer than that in the recently computed Cl3B‚
OMe2 molecule (mean, 1.835 Å).9 Conversely, an assumption
of predominantly ionic bonding8 means that the principal
determinant of bond strength/length is the closest packing of
anionic ligands. Reexamination of experimental and computa-
tional structural data in a wide variety of molecular oxides and
fluorides supports this view.8 Applied to 2, this view explains
the observations rather better (vide infra), without the need to
employ arguments requiring pπ-pπ back-bonding.
Unfortunately, no data are available for a direct comparison
of uncomplexed Ph2BCl. However, from those uncomplexed
diorganochloroboranes that have been structurally characterized,
the expected trends are discernible. The mean B-Cl length in
this class is 1.768 Å, similar to the distance in BCl3 (1.75(2)
Å).10 While the carbon ligands are smaller than chlorine, thereby
allowing a closer approach for the remaining chloride, they are
also much less electronegative so that the positive charge on
boron is much less well developed, leading to cancellation of
the two effects. The shortest in this group is 1.746(5) Å for
(C6F5)2BCl;11 the electron-withdrawing character of the per-
fluorophenyl substituents increases the positive charge on boron
The reduced reactivity of B-Cl bonds in diorganochlorobo-
ranes with respect to BCl3, and therefore the lack of THF
cleavage in 2, is in stark contrast to 1. While solutions of 2
require extended reflux to promote reaction, conversion of BCl3‚
SMe2 and Li3[CB10H11] in THF to 1 is quantitative after 30
min of stirring at 0 °C.3 It is an open question as to whether the
insertion reaction itself resulted in facilitating the ether cleavage.
Aside from the issue of C-O cleavage reactivity, the sluggish-
ness of Ph2BCl‚THF in reacting with the lithiated calix-4-arene
is also quite unexpected. In contrast, Ph2BCl reacts smoothly
at low temperatures with LiNC(tBu)2 in hexane.16 In the realm
of lithium chemistry, movement from hexane to THF as reaction
solvent normally results in an increase in reactivity, through
the breakdown of aggregated species. Here the opposite is true.
We ascribe this result to a combination of protection of the boron
(6) Luger, P.; Buschmann, J. Angew. Chem., Int. Ed. Engl. 1983, 22, 410.
A search of the Cambridge Crystallographic Database for uncoordi-
nated THF molecules free of disorder yielded 342 hits. The mean C-O
length of these was 1.403 Å, significantly shorter than the distances
in 2. However, the range was 0.931-1.721 Å! Clearly, many of the
entries, even those allegedly free of disorder, were subject to
crystallographic error. While there was significant clustering of values
around the mean (lower quartile ) 1.326 Å, higher quartile ) 1.471
Å), we consider the safest comparison to draw is with the accurate
low-temperature data from pure, crystalline THF.
(11) Piers, W. E.; Spence, R. E. v. H. Acta Crystallogr., Sect. C 1995, 51,
1688.
(12) Ashe, A. J., III; Klein, W.; Rousseau, R. Organometallics 1993, 12,
3225.
(7) Bader, R. F. W. Atoms in Molecules: A Quantum Theory; Clarendon
Press: Oxford, U.K., 1991.
(8) Robinson, E. A.; Johnson, S. A.; Tang, T.-H.; Gillespie, R. J. Inorg.
Chem. 1997, 36, 3022. Gillespie, R. J. J. Chem. Educ. 1998, 75, 923.
(9) Roswell, B. D.; Gillespie, R. J.; Heard, G. L. Inorg. Chem. 1999, 38,
4659.
(13) Cambridge Crystallographic database search produced Refcodes
DECDAQ, DONXIN, and HESTAA.
(14) Evans, W. J.; Shreeve, J. L.; Ziller, J. W. Acta Crystallogr., Sect. C
1996, 52, 2571.
(15) Hess, V. H. Acta Crystallogr., Sect. B 1969, 25, 2338.
(16) Collier, M. R.; Lappert, M. F.; Snaith, R.; Wade, K. J. Chem. Soc.,
Dalton Trans. 1972, 370.
(10) Atoji, M.; Lipsomb, W. N. J. Chem. Phys. 1957, 48, 1571.

2692 Inorganic Chemistry, Vol. 39, No. 12, 2000
Notes
from alkoxide attack by the saturation of its coordination sphere
by THF and the reluctance of the extensively chelated calixarene
to break down its aggregated form17 even in THF. After
extended reflux, some conversion of parent calixarene is
apparent, but we have so far failed to isolate pure products.
Prof. R. J. Gillespie is thanked for useful discussions and for
providing a manuscript prior to publication. This work made
use of the Crystallographic Database maintained by the Dares-
bury Laboratory, U.K., funded by EPSRC.
Supporting Information Available: X-ray crystallographic files,
in CIF format. This material is available free of charge via the Internet
at http://pubs.acs.org.
Acknowledgment. The financial support of EPSRC (stu-
dentships to W.I.C. and M.P.L.) is gratefully acknowledged.
(17) Davidson, M. G.; Howard, J. A. K.; Lamb, S.; Lehmann, C. W. Chem.
Commun. 1997, 1607. Bock, H.; John, A.; Na¨ther, C.; Havlas, Z. J.
Am. Chem. Soc. 1995, 117, 9367.
IC991148Q