Studies on BCl and on the exchange reaction of BCl3 with polyboron flourides

Article abstract of DOI:10.1016/S0277-5387(01)01032-4

Gaseous BCl has been prepared at low pressure by reaction of BCl₃ with boron at 2000 °C or by decomposition of B₂Cl₄ using either an electric discharge or using flash thermolysis at 1150 °C, and then condensed at - 196 °C. Reasons are discussed why each method gives BCl capable of generating (Cl₂B)₃BCO when the condensate at - 196 °C is treated with CO, but B₄Cl₄ is only formed from BCl prepared from B₂Cl₄. The stepwise conversion of (F₂B)₃BCO to (Cl₂B)₃BCO by treatment with BCl₃ has been followed by ¹¹B NMR spectroscopy. Evidence from mass spectrometry and ¹¹B NMR spectroscopy suggests that reaction of polyboron fluorides, particularly B₈F₁₂ and B₁₀F₁₂, with BCl₃ yields a new form of B₉Cl₉ of, as yet, unknown structure which converts slowly to closo-B₉Cl₉ or closo-B₈Cl₈ depending on the solvent.

Full text of DOI:10.1016/S0277-5387(01)01032-4

Polyhedron 21 (2002) 543–548
www.elsevier.com/locate/poly
Studies on BCl and on the exchange reaction of BCl3 with
polyboron fluorides
Jennifer A.J. Pardoe, Nicholas C. Norman, Peter L. Timms *
School of Chemistry, Uni6ersity of Bristol, Bristol BS8 1TS, UK
Received 16 August 2001; accepted 7 December 2001
Abstract
Gaseous BCl has been prepared at low pressure by reaction of BCl3 with boron at 2000 °C or by decomposition of B2Cl4 using
either an electric discharge or using flash thermolysis at 1150 °C, and then condensed at −196 °C. Reasons are discussed why
each method gives BCl capable of generating (Cl2B)3BCO when the condensate at −196 °C is treated with CO, but B4Cl4 is only
formed from BCl prepared from B2Cl4. The stepwise conversion of (F2B)3BCO to (Cl2B)3BCO by treatment with BCl3 has been
followed by 11B NMR spectroscopy. Evidence from mass spectrometry and 11B NMR spectroscopy suggests that reaction of
polyboron fluorides, particularly B8F12 and B10F12, with BCl3 yields a new form of B9Cl9 of, as yet, unknown structure which
converts slowly to closo-B9Cl9 or closo-B8Cl8 depending on the solvent. © 2002 Elsevier Science Ltd. All rights reserved.
Keywords: Boron monochloride; Halogen exchange; Polyboron halides
1. Introduction
pumped, glass vacuum lines equipped with Teflon stop-
cocks, with Viton O-ring joints (J. Young) and with
efficient means of separating volatiles on a low temper-
ature vacuum distillation column [5]. Pre-dried solvents
Knowledge of the chemistry of the polyboron halides
has been growing for about 50 years since the original
discovery of B4Cl4 by Schlesinger [1]. The chemistry of
polyboron chlorides and bromides was reviewed in
1991 [2] and there have been some developments since
[3]. An earlier review also embraced the polyboron
fluorides [4]. The work reported here concentrates on
two aspects of this field, on the chemistry of BCl and
on halogen exchange reactions by which polyboron
fluorides are converted into chlorides.
,
were further dried over vacuum treated 3 A molecular
sieves just before use; dry CO was obtained by collect-
ing liquid CO from a cylinder in a liquid nitrogen
cooled Schlenk tube and allowing the liquid CO to
evaporate only at very low temperatures; BCl3 (Aldrich)
was pumped on to remove HCl before use; B2Cl4 was
prepared either by reaction of excess BCl3 with B2F4 in
the gas phase at room temperature or by reduction of
BCl3 with copper atoms [6]; (F2B)3BCO was prepared
by reaction of CO with B8F12 which had been prepared
from low temperature condensation of BF [7].
2. Experimental
A 10–1000 amu Micromass quadrupole mass spec-
trometer integral to the vacuum lines permitted imme-
diate analysis of products of sufficient volatility. Less
volatile samples were put in capped vials under strictly
oxygen free conditions and analysed using a VG Ana-
lytical Autospec mass spectrometer. All samples for 11B
NMR spectroscopy were fusion sealed in 5 mm Pyrex
tubes; the spectra were run at 96 MHz on a JEOL 300
MHz Eclipse spectrometer.
All boron compounds described in this work were
handled under high vacuum conditions on diffusion
* Corresponding author. Tel.: +44-117-9287654; fax: +44-117-
9251295.
E-mail address: peter.timms@bristol.ac.uk (P.L. Timms).
0277-5387/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0277-5387(01)01032-4

544
J.A.J. Pardoe et al. / Polyhedron 21 (2002) 543–548
3. Results and discussion
pound of this type, (F2B)3BCO, was prepared by reac-
tion of CO with B8F12 [7]. However, (F2B)3BCO can
also be made by reacting CO at low temperatures with
(F2B)2BF, which disproportionates above −50 °C to
B8F12 and B2F4. The formation of (Cl2B)3BCO strongly
suggests that some (Cl2B)2BCl was formed by insertion
of BCl into the BꢀB bond or a BꢀCl bond of B2Cl4 and
that this reacted with CO as shown in Eq. (3):
3.1. Studies on BCl prepared by different methods
Boron monochloride is a well studied molecule be-
cause of its interest to spectroscopists [8,9], its impor-
tance in plasma chemistry of BCl3 [10] and its
involvement in the formation of B2Cl4 [11]. We have
recently employed three routes to making BCl to use it
as a chemical reagent. Using two procedures originally
described 25 years ago [12,13], BCl has been prepared
by passing the vapour of B2Cl4 at the rate of about 1
mmol min−1 and a pressure of a few mbar, either
through a 2 kV, 50 Hz electrical discharge in a copper
tube or through a short length of 2 mm bore quartz
tubing heated to 1150 °C to allow flash thermolysis to
occur. The off-gases from the discharge or flash ther-
molysis units were expanded into a vacuum at a pres-
sure B10−3 mbar and were condensed rapidly on the
liquid nitrogen cooled walls of the vacuum chamber.
The discharge preparation is somewhat easier experi-
mentally as in the flash thermolysis system there are
two competing reactions shown in Eqs. (1) and (2):
2(Cl2B)2BCl+CO“(Cl2B)3BCO+B2Cl4
(3)
We have also recently prepared BCl by a new route,
the reduction of BCl3 by passing the gas over boron at
approximately 2000 °C at low pressure. The apparatus
used was similar to that previously used to prepare BF
[7]. The BCl3 was passed at a rate of 0.8 mmol min−1
and at pressure of B2 mbar through a column of
granular boron contained in an inductively heated
graphite tube. The off-gases from the tube emerged into
a vacuum at B10−4 mbar and passed directly to the
walls of a stainless steel vessel, part immersed in liquid
nitrogen. By measuring the loss of weight of boron and
calculating using Eq. (4) below, about 12% of the BCl3
seemed to be converted to BCl.
2B(s)+BCl3(g)“3BCl(g)
(4)
B2Cl4(g)“BCl(g)+BCl3(g)
3B2Cl4“2B(s)+4BCl3(g)
(1)
(2)
On allowing the stainless steel reaction vessel to
warm up, BCl3 and B2Cl4 were pumped out below
room temperature. With the vessel warmed to room
temperature (22 °C), a mixture of green and yellow–
brown solids could be pumped out slowly. Release of
these solids was speeded up by warming the stainless
steel vessel to approximately 40 °C. Studies of these
solids by 11B NMR and mass spectrometry showed the
most volatile compound to be B8Cl8 while the less
volatile solids were mainly B9Cl9 with some B10Cl10,
B11Cl11 and B12Cl12. No B4Cl4 was detected. An insolu-
ble, involatile but very air-sensitive solid was left on the
walls of the reactor after exhaustive pumping to pres-
sures B10−4 mbar. Repeating the experiment with
addition of CO to the cold reactor after the condensate
had collected, resulted in the formation of (Cl2B)3BCO,
suggesting that some (Cl2B)2BCl was among the prod-
ucts formed on the cold surface. The formation of
(Cl2B)2BCl and its decomposition at low temperatures
may explain the formation of the observed poly-boron
chlorides via the mechanism proposed by Davan and
Morrison [15] in their studies of the thermal decomposi-
tion of B2Cl4.
The reaction of Eq. (1) is less thermodynamically
favoured than the reaction of Eq. (2) but the reaction
of Eq. (1) can be kinetically favoured because there is a
high activation energy for nucleation of solid boron in
the homogeneous gas phase reaction of Eq. (2). Unfor-
tunately, it is hard to suppress the formation of boron
in a heterogeneous reaction with the tube walls and the
flash thermolysis tube often blocked after only a few
mmol of B2Cl4 had passed through.
The only molecular product of condensation of the
off-gases from the discharge or flash thermolysis treat-
ment of B2Cl4 was B4Cl4, characterised by 11B NMR
spectroscopy and by mass spectrometry [2]. This was
formed together with a larger amount of a colourless,
intractable solid. Maddren [13] was able to suppress
B4Cl4 formation by condensation of the off-gases from
the discharge with a large excess of ethyne (which gave
C4H4B2Cl2, 1,4-diboracyclohexa-2,5-diene), showing
that the B4Cl4 formed on the cold surface not in the
discharge (or flash thermolysis) region. However, we
have found that if the condensate from the discharge or
flash thermolysis system is allowed to warm from
−196 °C in the presence of approximately 0.5 atm
pressure of CO, a new compound not found in the
earlier work can be isolated in addition to B4Cl4. This
solid was identified as (Cl2B)3BCO from its 11B NMR
spectrum (peaks at 68.2 and −20.7 ppm), from its IR
The complete absence of B4Cl4 formation from BCl
in the above experiments was a surprise but the result
may give a clue as to the true mode of formation of
B4Cl4. Earlier work has shown that B4Cl4 is not one of
the products of decomposition of B2Cl4 even at temper-
atures up to 300 °C [15]. Apart from our work on flash
thermolysis of B2Cl4, B4Cl4 has only been formed by
the use of electrical discharges in BCl3 or in B2Cl4. In
spectrum (showing a strong y CO stretch at 2176 cm−1
and from its crystal structure [14]. The original com-
)

J.A.J. Pardoe et al. / Polyhedron 21 (2002) 543–548
545
an electrical discharge, there is energy available to form
exchange of PꢀF for PꢀCl was observed within the time
scale needed to exchange BꢀF for BꢀCl.
1
BCl in its singlet ground state, X S+, and also in its
3
triplet excited state, a P1, which lies 242 kJ mol−1
The stepwise nature of the above fluorine-for-chlo-
rine exchange process was revealed by 11B NMR spec-
troscopy. Liquid BCl3 was allowed to interact with
(F2B)3BCO for 50 min at 0 °C, followed by cooling to
−70 °C, pumping boron trihalides away and then
dissolving the remaining solid in CD2Cl2. The 11B–11B
COSY spectrum of the product is shown in Fig. 1.
Progressive replacement of F by Cl as (F2B)3BCO
converts to (Cl2B)3BCO can give a total of ten different
compounds containing combinations of BF2, BFCl and
BCl2 groups. The spectrum of Fig. 1 gives evidence for
at least eight of these ten possible structures. Consider-
ing first the four-coordinate boron environments rang-
ing in l values from −50.6 ppm for (F2B)3BCO to
−20.7 ppm for (Cl2B)3BCO. There are seven different
peaks corresponding to successive replacements of F by
Cl in compounds of type B4F6−xClxCO. The peaks at
−45.0 and −24.4 ppm are directly assignable to
(F2B)2(FClB)BCO and to (Cl2B)2(FClB)BCO, respec-
tively. Of the remaining three peaks, that at −29.0
ppm has an asymmetry indicating that both
(F2B)(Cl2B)2BCO and (Cl2B)(FClB)2BCO are present
although the precise positions of their respective reso-
nances cannot be resolved from the COSY spectrum.
The COSY spectrum also suggests that both
(F2B)(FClB)(Cl2B)BCO and (FClB)3BCO and both
(F2B)(Cl2B)2CO and (FClB)2(Cl2B)BCO contribute to
the peaks at −34.3 and −39.8 ppm, respectively. In
addition, the spectrum shows three broad resonances
corresponding to BF2, BFCl and BCl2 groups with
chemical shifts at +32.4, +50.8 and +68.9 ppm,
respectively. These values are similar to those previ-
ously reported for species of the type B2F4−xClx [19]; in
pure (BF2)3BCO, the chemical shift for the BF2 groups
is +31.3. The COSY spectrum shows how the 11B
chemical shifts for these BF2, BFCl and BCl2 groups
move slightly to higher field as the fluorine content of
the molecules increases.
above the ground state [9]. As suggested earlier [16], it
could be that B2Cl4 arises from insertion of singlet BCl
into the BꢀCl bonds of BCl3 while B4Cl4 arises from
aggregation of triplet BCl. If both singlet and triplet
BCl were condensed on a cold surface, polymerisation
of the latter may be the source of the observed B4Cl4.
The idea of forming a mixture of singlet and triplet
forms of BCl seems less obvious in relation to the work
on flash thermolysis. There would not appear to be
sufficient energy available at 1150 °C to form BCl in its
excited triplet state yet B4Cl4 is formed in amounts
similar to that observed from the discharge route. The
answer may lie in the mode of decomposition of B2Cl4
by the gas phase reaction of Eq. (1). As shown by
McKee for the flash thermolysis of 1,2-P2B4Cl4 [17],
energy barriers to apparently simple modes of decom-
position can be high, providing kinetic enhancement of
thermodynamically less favourable pathways. So flash
thermolysis of B2Cl4 could yield a mixture of the ex-
pected singlet BCl with some of the higher energy
triplet form.
Preparing BCl at 2000 °C in a heterogeneous reac-
tion from boron and BCl3 is likely to be a route to the
most thermodynamically favoured form, i.e. singlet,
ground state BCl, and this on condensation gives no
B4Cl4. All routes to BCl give some singlet BCl and it is
believed that this is the form most suited to react
efficiently with BCl3 to give B2Cl4 or to react with B2Cl4
to give (Cl2B)2BCl which is the precursor to
(Cl2B)3BCO.
It is clear from this research that more theoretical
work is needed to give a proper understanding of the
mode of formation of B4Cl4 from BCl.
3.2. Con6ersion of (F2B)3BCO to (Cl2B)3BCO by
reaction with BCl3
The conversions of B2F4 to B2Cl4 and of B2Cl4 to
B2Br4 by reaction with BCl3 and BBr3, respectively,
have been known for many years [4]. We have found
that halogen exchange also works well starting with
(F2B)3BCO and treating it with BCl3.
Reaction of (F2B)3BCO with BCl3 in the vapour
phase at room temperature gives a virtually quantita-
tive conversion to (Cl2B)3BCO in less than an hour with
liberation of BF3 [14]. This is by far the most conve-
nient preparation of (Cl2B)3BCO in our laboratories
because (F2B)3BCO can be made in gram quantities by
allowing the products formed by condensing BF at
−196 °C to warm to room temperature in the presence
3.3. Reaction of polyboron fluorides with BCl3
Boron trichloride has also been reacted with some of
the polyboron fluorides originating from condensation
of BF at −196 °C which contain eight or more boron
atoms [4]. The reactions have yielded a new, unstable
polyboron chloride, the detailed structure, of which can
not yet be assigned from the available NMR data.
Some explanation of the nature of the starting boron
fluoride materials is needed as single crystal X-ray
structures have only recently been obtained for two of
the polyboron fluorides, B8F12 and B10F12 [20]. The
structure of B8F12 differs from the diborane-like struc-
ture B2(BF2)6 proposed on the basis of its reactivity [7]
mainly through the presence of a short, central BꢀB
of CO gas.
A similar reaction occurs between
(F2B)3BPF3 [18] and BCl3 yielding (Cl2B)3BPF3 (show-
ing 11B resonances at l +65.9 and −17.1 ppm); no

546
J.A.J. Pardoe et al. / Polyhedron 21 (2002) 543–548
formed by reaction with the BCl3 in the vapour. If the
temperature is then raised to 60 °C in the presence of
BCl3 vapour before all the highly volatile boron tri-
halides are pumped away, the residue is a dark brown–
purple solid. Direct extraction of this solid with CD2Cl2
at room temperature gives a brown solution which
initially exhibits a an 11B NMR spectrum showing six
main resonances, the chemical shifts and relative inten-
sities of which are given in Fig. 2. Essentially the same
spectrum can be obtained by heating the brown–purple
solid to 40 °C and pumping out a solid by short-path
distillation on to a cold finger, and dissolving the
material on the cold finger in CD2Cl2 or C6D6. The
colour of the sublimate can vary from run to run from
almost colourless to yellow–brown. It seems likely that
bond, supported by bridging BF2 groups and a folded
rather than planar central B2(m-BF2)2 unit. The exis-
tence of B10F12 was not apparent from previous studies
of higher boron fluorides based largely on mass spec-
trometry [4], but it is now clear that this molecule with
a central tetrahedral core of four boron atoms with
four terminal and two bridging BF2 groups having S4
symmetry, is the most abundant component of the
polyboron fluorides of lower volatility than B8F12.
The reaction of either B8F12 or of higher polyboron
fluorides rich in B10F12 with excess liquid BCl3 follows
the same pattern. Reaction begins below −30 °C with
a darkening in colour, which intensifies slowly. Holding
the temperature at 0 °C for several hours gives a
red–brown solution and BF3, BF2Cl and BFCl2 are
Fig. 1. 11B–11B COSY NMR spectrum of compounds of type B4F6−xClxCO prepared by reacting (BF2)3BCO with BCl3 (CD2Cl2 as solvent).

J.A.J. Pardoe et al. / Polyhedron 21 (2002) 543–548
547
resonances in the 11B NMR. However, our results do
strongly suggest the existence of a type of polyboron
chloride never previously described. Morrison’s pro-
posals on the way in which B2Cl4 molecules react
together to form closo-BnCln structures with elimina-
tion of BCl3 [2] hint at the existence of open structured
intermediates but he has not reported an intermediate
of this type stable enough to be sublimed. The relatively
high stability of the polyboron chloride observed in our
work may be due to the way it is formed from B10F12,
i.e. it may not be possible to form such a species by
decomposition of B2Cl4. Further support for the idea
that the polyboron chloride giving the six 11B NMR
resonances may be accessible only from the reaction of
BCl3 with polyboron fluorides comes from the follow-
ing observation. The product was not detected among
the closo-BnCln compounds generated under very mild
conditions from the products obtained made by con-
densing at −196 °C, BCl made from boron and BCl3
at 2000 °C.
The solutions obtained either by direct extraction or
by sublimation followed by dissolution of the subli-
mate, which give the peaks at positions shown in Fig. 2,
are not stable. When held at room temperature, the
colour intensity increases and the 11B NMR spectrum
changes in a complex but fairly reproducible way. All
six of the peaks of Fig. 2, decrease in intensity and
become broader, but there is massive growth of one
sharp resonance. For samples dissolved in C6D6, this
occurs at l +58.5 ppm corresponding to closo-B9Cl9.
For one sample decomposing just in CD2Cl2, the single
peak, which grew was at 65.2 ppm corresponding to
closo-B8Cl8 [2]. Morrison has reported that the forma-
tion of closo-BnCln species from B2Cl4 is solvent sensi-
tive with closo-B8Cl8 favoured over B9Cl9 in CCl4 as
solvent [2]. In the 11B NMR spectra of samples under-
going decomposition, there was some increase in the
intensity of the resonance due to BCl3, but not enough
to indicate that BCl3 was eliminated to a major extent
by the transformation. If the product of reaction of
BCl3 with polyboron fluorides was left for a few days at
room temperature under vacuum in the absence of a
solvent and was then pumped on to remove very
volatile products, the resulting intensely coloured
product showed from its 11B NMR and mass spectrum
the presence of a mixture of the known closo-boranes
from B8Cl8 to B12Cl12, with B9Cl9 as by far the major
product.
Fig. 2. 11B NMR data for the dominant product formed by reacting
polyboron fluorides rich in B10F12 with BCl3 at 0 °C until chlorine
for fluorine exchange was complete (CD2Cl2 as solvent).
the main component of the sublimate is colourless but
that it is unstable and becomes contaminated to a
varying extent by intensely coloured decomposition
products.
The mass spectrum of the sublimate, run within a few
minutes of dissolving the solid in C6D6, showed BCl2+
as the base peak, with B8Cl6+ (m/z 293–307, centred at
299) as the dominant high mass peak, approximately 40
times more intense than the highest mass peak group at
m/z 410–426, centred at 416, corresponding to B9Cl9+
.
The mass spectrum of closo-B9Cl9 run under the same
conditions shows a much lower ratio of the intensities
of B8Cl6+: B9Cl9+ in line with that reported by Mor-
rison [2]. The mass spectrum suggests that the main
volatile component of the sample may have been B9Cl9
but an isomer of different structure to the known
closo-B9Cl9, with a much greater tendency to eliminate
BCl3 to form B8Cl6+
.
In the 11B NMR spectrum, the six peaks identified in
Fig. 2 seem to belong to a single molecular species as
they are always found in the same ratio and decline in
intensity together as the sample decomposes over time.
The exact position and relative intensity of the peak at
+65 ppm is least certain as this region in the spectrum
overlays a group of peaks associated with the closo-bo-
ranes and with BCl2 groups from a variety of sources.
It has been clearly demonstrated by 11B and 19F NMR
spectra, that none of the six peaks is caused by any
residual BꢀF containing species. The way the peaks are
inter-related is shown in Fig. 2 from the 11B–11B COSY
spectrum couplings. The peak at −18.2 ppm is coupled
to peaks at +103.7, +65.0 and +50.7 ppm while the
peak at +0.8 ppm is coupled to the peaks at +50.7
and +48.7 ppm. The wide range of 11B chemical shifts
within a single molecule points to a variety of chemical
environments. The peaks at +0.8 and at −18.2 ppm
may be from boron atoms without direct chlorine sub-
stituents {cf. the central boron of (Cl2B)3BCO}. The
peaks at +65.0, +50.7 and +48.7 ppm are likely due
to boron atoms with one or two chlorine substituents.
The peak at +103.7 ppm seems to suggest a boron
atom of a BCl group in a more highly strained environ-
ment than the boron atoms in B4Cl4 which resonate at
+85.5 ppm [2].
The 11B NMR spectra of the samples after treatment
with BCl3 always contain other small peaks than the six
noted in Fig. 2, including a very sharp peak at +47.7
ppm due to BCl3 which occurs in all the samples and is
used as internal reference. However, the intensity ratios
and chemical shifts vary appreciably from sample to
sample unlike the six peaks discussed above. Their
presence simply reinforces the idea that the reaction of
The NMR data alone by no means proves the struc-
ture or composition of the compound giving the six

548
J.A.J. Pardoe et al. / Polyhedron 21 (2002) 543–548
BCl3 with the polyboron fluorides is an exceedingly
complex process, generating a wide range of boron
containing products.
tion of B2Cl4 and on the mechanism by which B4Cl4
forms from BCl.
The fact that the product which showed in its 11B
NMR spectrum the dominant six peaks of Fig. 2 could
be obtained by reaction of excess BCl3 with either B8F12
or B10F12 at 0 °C can be explained by the following
observation. When B8F12 was treated with BCl3 at
−60 °C for a few minutes, mass spectrometry on the
product showed that the B8F12 had been largely con-
verted to B10F12 and BF3 with no evidence for chlorine-
for-fluorine exchange in the polyboron fluorides. Thus,
it can be assumed that conversion of B8F12 to B10F12
was a rapid first step in the reaction with BCl3 at 0 °C.
Investigations into the reaction of BCl3 with poly-
boron fluorides are on-going. The primary objective
will be to try to get a crystal structure of the key
intermediate giving the six 11B NMR resonances dis-
cussed. It may not be possible to crystallise this com-
pound from solution because of its instability and low
temperature laser techniques on the solid may also
promote the transformation to a stable closo-BnCln
structure, making this a difficult problem.
Acknowledgements
We thank Dr. Colin Western for helpful discussions
about the electronic state of BCl and we are grateful to
the Engineering and Physical Sciences Research Coun-
cil for their support of this research and for a stu-
dentship to Jennifer A.J. Pardoe.
References
[1] G. Urry, T. Wartik, H.I. Schlesinger, J. Am. Chem. Soc. 74
(1952) 5809.
[2] J.A. Morrison, Chem. Rev. 91 (1991) 35.
[3] W. Ho¨nle, Y. Grin, A. Burkhardt, U. Wedig, M. Schultheiss,
H.G. von Schnering, R. Kellner, H. Binder, J. Solid State Chem.
133 (1997) 59.
[4] P.L. Timms, in: A.L. Rheingold (Ed.), Homoatomic Rings,
Chains and Macromolecules of Main-Group Elements, Elsevier,
Amsterdam, 1977, p. 151.
[5] D.F. Shriver, The Manipulation of Air-sensitive Compounds,
McGraw-Hill, New York, 1969, p. 91.
[6] P.L. Timms, J. Chem. Soc., Dalton Trans. (1972) 830.
[7] P.L. Timms, J. Am. Chem. Soc. 89 (1967) 1629.
[8] K.K. Baeck, Y. Joo, Chem. Phys. Lett. 337 (2001) 190.
[9] K.K. Irikura, R.D. Johnson, J.W. Hudgens, J. Phys. Chem. A
104 (2000) 3800.
Related experiments on the reaction of BBr3 with
higher boron fluorides have also been carried out. The
end result is a series of known closo-BnBrn compounds
but the intermediates are at least as complicated as
those from the BCl3 reaction.
[10] G.A. Hebner, M.G. Blain, T.W. Hamilton, J. Vac. Sci. Technol.
A 17 (1999) 3218.
[11] A.G. Briggs, A.G. Massey, M.S. Reason, P.J. Portal, Polyhe-
dron 3 (1984) 369.
4. Conclusions
[12] P.L. Timms, in: M. Moskovits, G.A. Ozin (Eds.), Cryochem-
istry, Wiley, New York, 1976, p. 108.
[13] P.S. Maddren, Ph.D. Thesis, University of Bristol 1975.
[14] J.C. Jeffrey, N.C. Norman, J.A.J. Pardoe, P.L. Timms, Chem.
Commun. (2000) 2367.
[15] T. Davan, J.A. Morrison, Inorg. Chem. 25 (1986) 2366.
[16] T. Davan, J.A. Morrison, Inorg. Chem. 18 (1979) 3194.
[17] M.L. McKee, J. Phys. Chem. 95 (1991) 9273.
[18] R.W. Kirk, D.L. Smith, W. Airey, P.L. Timms, J. Chem. Soc.,
Dalton Trans. (1972) 1392.
The research has provided clear results relating to the
synthesis of (Cl2B)3BCO and the stepwise conversion of
(F2B)3BCO to (Cl2B)3BCO by reaction with BCl3. It
has also provided some evidence for the existence of a
new type of open-structured polyboron chloride, which
can readily isomerise to the much more stable closo-
BnCln compounds. Full characterisation of this com-
pound will be experimentally challenging. It has been
shown that the formation of B4Cl4 from BCl is sensitive
to the way the BCl is synthesised and this suggests that
there is a need for theoretical work on the decomposi-
[19] D.A. Saulys, J. Castillo, J.A. Morrison, Inorg. Chem. 28 (1989)
1619.
[20] J.A.J. Pardoe, N.C. Norman, P.L. Timms, S. Parsons, I. Mackie,
C.R. Pulham, D.W.H. Rankin, in preparation.