Synthesis and study of new binuclear compounds containing bridging (μ-CN)B(C6F5)3 and (μ-NC)B(C 6F5)3 systems

Article abstract of DOI:10.1039/b302163g

The new compounds [K(18-crown-6)][SCNB(C₆F₅) ₃] (1), [K(18-crown-6)][(C₆F₅) ₃B(μ-NC)B(C₆F₅)₃] (2), [K(18-crown-6)][NCB(C₆F₅)₃] (3), [Me ₃Si(μ-NC)B(C₆F₅)₃] (4) and [Hg{(μ-CN)B(C₆F₅)₃}₂] (5) have been synthesised. Reaction between [IrCl(PPh₃)₂(CO)] and one equivalent of 4 gives [Ir{(μ-NC)B(C₆F₅) ₃}(PPh₃)₂CO] (6). Similarly, the reaction between [Fe(η-C₅H₅)(CO)₂Cl] and 4 gives [Fe{(μ-NC)B(C₆F₅)₃}(η-C ₅H₅)(CO)₂] (7a). The NMR-scale reaction between [Fe(η-C₅H₅)(CO)₂CN] and B(C ₆F₅)₃ gives the isomer [Fe{(μ-CN)B(C ₆F₅)₃}(η-C₅H ₅)(CO)₂] (7b). In the compounds 3, 4, 6 and 7a the anion [NCB(C₆F₅)₃]^- is coordinated to the metal through the nitrogen. Complexes 1, 2 and 7b incorporate a (μ-CN)B unit. The compounds containing the isomeric bridging (μ-CN)B or (μ-NC)B systems have been investigated by ¹¹B{¹H} NMR spectroscopy and by DFT calculations. The ¹¹B{¹H} NMR spectra enable distinction between the isomers. The single-crystal X-ray structures of 1, 3 and 7a have been determined. The Royal Society of Chemistry 2003.

Full text of DOI:10.1039/b302163g

View Article Online / Journal Homepage / Table of Contents for this issue
Synthesis and study of new binuclear compounds containing
bridging (ꢀ-CN)B(C6F5)3 and (ꢀ-NC)B(C6F5)3 systems
Ino C. Vei, Sofia I. Pascu, Malcolm L. H. Green,* Jennifer C. Green, Richard E. Schilling,
Grant D. W. Anderson and Leigh H. Rees
Inorganic Chemistry Laboratory, South Parks Road, Oxford, UK OX1 3QR
Received 24th February 2003, Accepted 2nd May 2003
First published as an Advance Article on the web 23rd May 2003
The new compounds [K(18-crown-6)][SCNB(C6F5)3] (1), [K(18-crown-6)][(C6F5)3B(µ-NC)B(C6F5)3] (2), [K(18-crown-
6)][NCB(C6F5)3] (3), [Me3Si(µ-NC)B(C6F5)3] (4) and [Hg{(µ-CN)B(C6F5)3}2] (5) have been synthesised. Reaction
between [IrCl(PPh3)2(CO)] and one equivalent of 4 gives [Ir{(µ-NC)B(C6F5)3}(PPh3)2CO] (6). Similarly, the reaction
between [Fe(η-C5H5)(CO)2Cl] and 4 gives [Fe{(µ-NC)B(C6F5)3}(η-C5H5)(CO)2] (7a). The NMR-scale reaction
between [Fe(η-C5H5)(CO)2CN] and B(C6F5)3 gives the isomer [Fe{(µ-CN)B(C6F5)3}(η-C5H5)(CO)2] (7b). In the
compounds 3, 4, 6 and 7a the anion [NCB(C6F5)3]Ϫ is coordinated to the metal through the nitrogen. Complexes 1, 2
and 7b incorporate a (µ-CN)B unit. The compounds containing the isomeric bridging (µ-CN)B or (µ-NC)B systems
have been investigated by 11B{1H} NMR spectroscopy and by DFT calculations. The 11B{1H} NMR spectra enable
distinction between the isomers. The single-crystal X-ray structures of 1, 3 and 7a have been determined.
(C6F5)3] (1), and the reaction between 18-crown-6 ether, KCN
and B(C6F3)3 (in a 1 : 1 : 1 ratio or a 1 : 1 : 2 ratio) in toluene
gave [K(18-crown-6)][(C6F5)3B(µ-NC)B(C6F5)3] (2) instead of
Introduction
Compounds incorporating the weakly coordinating anion
[(C6F5)3B(µ-CN)B(C6F5)3]Ϫ together with cationic metallocenes
the expected [K(18-crown-6)][NCB(C6F5)3] (3). The anion
have recently been reported by Bochmann and co-workers to
[(C6F5)3B(µ-CN)B(C6F5)3]Ϫ has been reported earlier as the
give active homogenous catalysts for olefin polymerisation.1,2
[Ph3C]ϩ salt.1,2 Alternative routes for the synthesis of 3 via
The compound [Zr{(µ-NC)B(C6F5)3}Me(CpЉ)2], where CpЉ =
treatment of [K(18-crown-6)][CN] with B(C6F5)3 were explored.
1,3-C5H3(SiMe3)2, containing the anionic isonitrile ligand
The reaction was performed in toluene, CH2Cl2, THF and
[NCB(C6F5)3]Ϫ has also been reported.2 The complexes
[NHMe2Ph]2[Ni{(µ-CN)B(C6F5)3}4] and [Ph3C]2[Ni{(µ-CN)-
CH3CN, but gave only [K(18-crown-6)][(C6F5)3B(µ-CN)-
B(C6F5)3] (2) and not 3. An NMR scale reaction between
Me3SiCN and B(C6F5)3 in C6D6 or CD2Cl2 gave a clean
spectrum consistent with the formation of [(Me3Si)(µ-NC)-
B(C6F5)3] (4). This was surprising, since the product expected
was [(Me3Si)(µ-CN)B(C6F5)3]. However, we note that rapid
B(C6F5)3}4] are active as co-catalysts towards olefin polymeris-
ation.2
Earlier we have described some coordination chemistry of
B(C6F5)3 and of the anionic ligand [HOB(C6F5)3]Ϫ. 3–8 Here we
describe the potassium salts of the new anions [SCNB(C6F5)3]Ϫ
equilibrium between R3SiCN and R3SiNC is well estab-
and [NCB(C6F5)3]Ϫ, the adducts of B(C6F5)3 with Me3SiCN and
lished.12–15 A larger scale synthesis in CD2Cl2 led to the formation
Hg(CN)2, and the organometallic compounds [Ir{(µ-NC)-
B(C6F5)3}(PPh3)2(CO)], [Fe{NCB(C6F5)3}(η-C5H5)(CO)2] and
[Fe{(µ-CN)B(C6F5)3}(η-C5H5)(CO)2].
The neutral isonitrile adducts [RNCBC6F5)3] and nitrile
of 4 in 77% yield. Since a facile cleavage of the anion
[NCB(C6F5)3]Ϫ from the adduct 4 was expected, the reaction of
this adduct with KCl (in presence of 18-crown-6 ether) was
9
carried out in toluene and the desired compound [K(18-crown-
9–11
adducts [RCNB(C6F5)3]
(R = alkyl, aryl)) have been
6)][NCB(C6F5)3]Ϫ (3) was isolated in 73% yield.
reported previously. They contain the units E1NCB or E1CNB,
respectively, where E1 = alkyl, aryl, hydrocarbyl. The organic
compounds E1NC and E1CN are commonly described as iso-
nitriles (isocyano) or nitriles (cyano), respectively. However
when we consider the bridging systems E1(µ-CN)E2 or
E1(µ-NC)E2 where E1 and E2 are not carbon bonding groups the
use of the terms “nitrile” or “isonitrile” can lead to confusion.
Of course, for E1(µ-CN)E2, if E1 is a hydrocarbyl group and E2
is, for example, a metal, then description as “an organo nitrile
ligand to the metal” would be appropriate. There are in fact
only a few two-atom bridging (µ-AB) ligands which can adopt
both E1(µ-AB)E2 and E1(µ-BA)E2 structures with a similar facil-
ity, and we are not aware of a systematic notation for isomers
in the class E1(µ-CN)E2 and E1(µ-NC)E2. Therefore, we will
describe the compounds containing the units E1(µ-CN)E2 or
E1((µ-NC)E2 simply by defining the elements E1 and E2 in each
case. In this work the E2 atom is boron.
Another attempt to synthesise the isonitrile anion [CNB-
(C6F5)3]Ϫ was made by cleavage, using potassium metal, of the
adduct [Hg{(µ-CN)B(C6F5)3}2] (5). The new compound 5 was
synthesised by addition of [Hg(CN)2] to B(C6F5)3 in CH2Cl2.
The adduct 5 is decomposed by water to give, inter alia,
the known adduct [H2OB(C6F5)3].3 Attempts to form either
H[CNB(C6F5)3] or H[NCB(C6F5)3] by treatment of 5 with H2S
were unsuccessful. The reduction of the Hg() adduct 5 with
potassium metal in the presence of 18-crown-6 ether was
carried out in toluene at 90 ЊC. The nitrile compound [K(18-
crown-6)][NCB(C6F5)3] (3) was isolated in moderate yield, as
white crystals. There was no evidence for the isomer [K(18-
crown-6)][CNB(C6F5)3].
Reaction between [IrCl(PPh3)2(CO)] and [Me3Si(µ-NC)-
B(C6F5)3] (4) in CH2Cl2 gave an initial pale-yellow mixture
which changed to a bright yellow after 2 h. Bright-yellow
crystals of the complex [Ir{(µ-NC)B(C6F5)3}(PPh3)2(CO)] (6)
were obtained by crystallisation from CH2Cl2. Treatment of
[Fe(η-C5H5)(CO)2Cl] with one equivalent of 4 in pentane
as room temperature formed red crystals of [Fe{(µ-NC)-
B(C6F5)3}(η-C5H5)(CO)2] (7a). The isomer [Fe{(µ-CN)-
B(C6F5)3}(η-C5H5)(CO)2] (7b) could be prepared in solution by
the treatment of [Fe(η-C5H5)(CO)2(CN)] with B(C6F5)3 in C6D6
and has been characterised only by NMR data in solution.
Results and discussion
Syntheses
The reaction between KSCN and B(C6F3)3 in the presence
of 18-crown-6 ether, in toluene, gave [K(18-crown-6)][SCNB-
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
2550

View Article Online
Scheme 1 Formation of compounds 1–6, 7a and 7b.
Scheme 1 shows the syntheses of compounds 1–6, 7a and 7b,
spectrum of the complex [Me3Si(µ-NC)B(C6F5)3] (4) shows a
which are highly air- and moisture-sensitive. The new com-
pounds 1–7a, 7b have been characterised by microanalysis and
by infrared and NMR spectroscopy and the data are given in
the experimental sections or Tables 1 and 2 and are discussed
below. The crystal structures of 1, 3 and 7a have been
determined.
band at 2255 cmϪ1 compared to 2190 cmϪ1 for Me3SiCN. The
corresponding wavenumber for [Fe(η-C5H5)(CO)2CN] is 2114
cmϪ1 and for [Fe{(µ-NCB(C6F5)3}(η-C5H5)(CO)2] (7a) it is
2252 cmϪ1
.
X-Ray crystallography
The crystal structures of compounds 1, 3 and 7a have been
determined. Selected molecular parameters are given in Table 3
and Table 4 presents the X-ray crystallographic data. Single
crystals of 1 suitable for the X-ray structure analysis were
Spectroscopic studies
The 11B NMR spectra are particularly diagnostic. The 11B
resonances differ markedly between three-coordinate borane
B(C6F5)3 and the four-coordinate boron found in the Lewis
acid–base adducts [(Donor)B(C6F5)3] for which the 11B reson-
ance is shifted to lower frequency with respect to B(C6F5)3 by
more than 50 ppm.3–7, 16
grown by diffusion of pentane into
a dichloromethane
solution. An ORTEP of 1 is presented in Fig. 1. The X-ray
crystal structure analysis of 1 confirms that only one equivalent
of 18-crown-6 ether is present in the asymmetric unit and
The 11B data for a number of known compounds and for
compounds described in this work are listed in Table 2. The
data show that when boron is directly bonded to nitrogen, as in
E1(µ-CN)B, the 11B{1H} resonance is broad and occurs in the
range δ Ϫ9.0 to Ϫ13.0 ppm. When boron is directly attached to
carbon, as in E1(µ-NC)B, the 11B{1H} resonance is sharp and
occurs in the range δ Ϫ20.0 to Ϫ23.0 ppm. On this basis the
structures of the isomeric compounds 7a and 7b can be
assigned as shown in Table 2.
The 19F NMR spectra of all the new compounds show reson-
ances broadly similar to those of B(C6F5)3 but shifted to lower
frequency. The para – fluorine resonances were found to be the
most sensitive of the aromatic fluorine shifts. The 19F NMR
data for derivatives of nitrile- or isonitrile-borates, (containing
the RCNB and RNCB units) have been found to be distinct
from the free borane, however they do not permit a reliable
distinction between two isomers such as 7a and 7b.
The IR spectra of compounds 1–6 and 7a in Table 1 show
sharp bands in the region 2146–2284 cmϪ1 which can be
assigned to the νCN or νNC stretching frequencies. As previously
shown,9 the coordination of B(C6F5)3 to a E1CNϪ group causes
a significant band shift to higher wavenumbers. Thus the IR
Fig. 1 ORTEP of [K(18-crown-6)][SCNB(C6F5)3] (1).
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2551

View Article Online
Table 1 IR (KBr disks, ν in cmϪ1) and NMR data (C6D6, δ in ppm) at room temperature
19F-ortho
19F-para
(br, 1F)
Compound
IR
1H
11B{1H}
(d, 2F, 2JF–F ≈20 Hz)
19F-meta (br, 2F)
B(C6F5)3
[K(18-crown-6)K][SCNB(C6F5)3] 1
[K(18-crown-6)][(C6F5)3B(µ-NC)B(C6F5)3] 2
58
Ϫ128.8
Ϫ134.7
Ϫ134.6; Ϫ135.7
Ϫ144.6
Ϫ160.5
Ϫ161.3,
Ϫ161.4
Ϫ160.5
Ϫ155.2
Ϫ156.7
Ϫ159.8
Ϫ161.5
Ϫ165.7
Ϫ168.0, Ϫ168.4
2146 (νCN
2294.8;
2295.3 (νCN
2212 (νCN
2256 (νCN
2284 (νCN
2232 (νCN
)
3.70 (s)
3.63 (s)
Ϫ12.4
Ϫ12.1;
Ϫ21.9
Ϫ22.5
Ϫ21.7
Ϫ10.7
Ϫ22.8
)
[K(18-crown-6)][NCB(C6F5)3] 3
[Me3Si(µ-NC)B(C6F5)3] 4
)
)
)
)
3.54 (s)
Ϫ0.30 (s)
Ϫ132.6
Ϫ133.7
Ϫ134.4
Ϫ132.16
Ϫ165.8
Ϫ163.2
Ϫ164.3
Ϫ164.8
[Hg{(µ-CN)B(C6F5)3}2] 5
[Ir{(µ-NC)B(C6F5)3}(PPh3)2(CO)] 6a
7.38
(m, 6H)
7.56
2012 (νCO-sym
)
(m, 4H)
1962 (νCO-asym
2252 (νCN
2076 (νCO-sym
)
[Fe{(µ-NC)B(C6F5)3}(η-C5H5)(CO)2] 7a
)
5.15 (s)
Ϫ22.4
Ϫ12.0
Ϫ134.0
Ϫ135.6
Ϫ158.6
Ϫ159.3
Ϫ164.7
Ϫ169.5
)
2032 (νCO-asym
–
)
[Fe{(µ-CN)B(C6F5)3}(η-C5H5)(CO)2] 7b
a 31P{1H} NMR (C6D6) at 25 ЊC: δ 22.9 (s).
5.28 (s)
Table 2 Comparative 11B{1H} NMR data. 11B{1H}NMR line shape is described here as sharp if ν1/2 is <15 1 Hz, and as broad if ν1/2 is >300
10 Hz
Compound
δ
11B{1H}
Line shape
Assignment
Ref.
a[Ph3C][Pd{(µ-CN)B(C6F5)3}4]
b(MeCN)B(C6F5)3
Ϫ12.8
Ϫ10.3
Ϫ9.6
–
N–B
N–B
N–B
N–B
N–B
N–C
N–B
C–B
N–B
C–B
C–B
C–B
1
3
3
3
Broad
Broad
Broad
Broad
Broad
Broad
Sharp
–
b[(p-NO2C6H4CN)B(C6F5)3]
b[(p-MeC6H4CN)B(C6F5)3]
Ϫ10.2
Ϫ10.7
Ϫ12.0
Ϫ12.1
Ϫ21.9
Ϫ11.9
Ϫ21.6
Ϫ21.8
Ϫ21.7
Ϫ21.0
Ϫ19.6
Ϫ22.5
Ϫ21.7
Ϫ22.8
Ϫ22.4
b[Hg{(µ-CN)B(C6F5)3}2] 5
This work
This work
This work
b[Fe{(µ-CN)B(C6F5)3}(η-C5H5)(CO)2] 7b
b[K(18-crown-6)][(C6F5)3B(µ-NC)B(C6F5)3] 2
c[Ph3C][(C6F5)3B(µ-NC)B(C6F5)3]
2
b[(tBuNC)B(C6F5)3]
Sharp
Sharp
Sharp
–
Sharp
Sharp
Sharp
Sharp
3
3
b[(Me3C–CH2–CMe2–NC)B(C6F5)3]
b[(2,6-Me2C6H3)NCB(C6F5)3]
a[Zr{(µ-NC)B(C6F5)3}Me(CpЉ)2]d
b[K(18-crown-6)][NCB(C6F5)3] 3
b[Me3SiB(µ-NC)B(C6F5)3] 4
C–B
C–B
C–B
C–B
C–B
2
This work
This work
This work
This work
b[Ir{(µ-NC)B(C6F5)3}(PPh3)2(CO)] 6
b[Fe{(µ-NC)B(C6F5)3}(η-C5H5)(CO)2] 7a
NMR data in a CD2Cl2, b C6D6, c CDCl3. d CpЉ = 1,3-C5H3(SiMe3)2.
the thiocyanate group is connected by a B(1)–N(1) bond
(1.531(3) Å). Noteworthy, this bond is significantly shorter than
the corresponding boron–nitrogen distances found in [Ph3C]-
[(C6F5)3B(µ-CN)B(C6F5)3]Ϫ (1.593(2) Å), [Ph3C]2[Ni{(µ-CN)-
B(C6F5)3}4] (1.574(2) Å), [HNMe2Ph]2[Ni{(µ-CN)B(C6F5)3}4]
(1.582(3) and 1.578(3) Å) and [Ph3C]2[Pd{(µ-CN)B(C6F5)3}4]
(1.577(3) and 1.571(3) Å).1,2 In 1, the C(1)–N(1) bond is
1.180(3) Å, somewhat longer that those found for several
neutral RCNB(C6F5)3 adducts (1.124(3) for R = Me and
X-Ray diffraction methods alone cannot reliably distinguish
between carbon and nitrogen in the systems E1(µ-CN)E2 and
E1(µ-NC)E2. However, the refinement of crystals structures of 3
and 7a considering a NCB system produced better solution to
the data set (with smaller R and wR values) than for the CNB
isomer. Furthermore, on inspection of the parameters for the
thermal ellipsoids, it has been observed that the sizes of the
ellipsoids are consistent for the NCB systems, without any large
discrepancies, unlike the case of the CNB systems. The X-ray
diffraction data, correlated with the observations from 11B{1H}
NMR spectroscopy, strongly indicate that 3 and 7a contain
BCN units.
The asymmetric unit of 3 consists of one 18-crown-6 macro-
cycle containing potassium, one [NCB(C6F5)3]Ϫ anion unit and
one molecule of toluene in a general position. Also present in
the asymmetric unit is half a molecule of toluene disordered
about an inversion centre. The asymmetric unit of 7a contains
only [Fe{(µ-NC)B(C6F5)3}(η-C5H5)(CO)2]. Figs. 3 and 4 show
the ORTEP diagrams of the compounds [K(18-crown-
6)][NCB(C6F5)3] (2) and [Fe{(µ-NC)B(C6F5)3}(η-C5H5)(CO)2]
(7a), respectively. The geometry around the boron centre is dis-
torted tetrahedral in 3 and 7a. For 3, the C(1)–N(1) distance is
1.113(4) Å, whereas in 7a, upon coordination to the Fe()
centre, the corresponding distance increases to 1.135(3) Å. Both
distances are slightly shorter than that found for [Zr{(µ-NC)-
1.135(3) for R = p-NO2C6H4),9 or for the anion [(C6F5)3-
1,2
B(µ-CN)B(C6F5)3]Ϫ (1.144(2) Å).
The C(1)–S(1) distance is
1.598(2) Å and the N(1)–C(1)–S(2) angle is close to linear
(178.3(2)Њ). The coordination geometry about the boron centre
is distorted tetrahedral, with N–B–C angles 108.03(18),
106.35(16)Њ and 106.89(17)Њ and the C–B–C angles measuring
112.02(16)Њ (C(11)-B(1)–C(31)), 112.12(16)Њ (C(11)–B(1)–
C(21)) and 112.86(19)Њ (C(21)–B(1)–C(31)), and the mean
B–C(aromatic) distance is ca. 1.650 Å. No solvent molecules
are present and the closest intermolecular contact is 2.892(13)
Å between K(2) and F(25) in an adjacent molecule in which the
C(25)–F(25)–K(2) angle is 136.35(16)Њ. The cation is bound to
B(C6F5)3 via fluorine bridges, to form a chain-like supra-
molecular structure, as shown in Fig. 2. This structure accounts
for the exceptional stability of 1 towards the attack of nucleo-
philes such as H2O or PMe3.
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2552

View Article Online
Fig. 2 Extended chain-like structure of [K(18-crown-6)][SCNB(C6F5)3] (1).
Table 3 Selected bond lengths (Å) and angles (Њ) for compounds 1, 3
and 7a
1
C(1)–N(1)
C(1)–S(1)
1.180(3)
1.598(2)
1.650(3)
1.654(3)
1.647(3)
1.531(3)
2.8902(13)
2.8024(13)
2.8039(13)
2.7855(13)
N(1)–C(1)–S(1)
C(1)–N(1)–B(1)
C(11)–B(1)–C(21)
C(11)–B(1)–C(31)
C(21)–B(1)–C(31)
C(11)–B(1)–N(1)
C(21)–B(1)–N(1)
C(31)–B(1)–N(1)
178.3(2)
173.1(2)
C(11)–B(1)
C(21)–B(1)
C(31)–B(1)
N(1)–B(1)
K(2)–F(25)
K(2)–O(53)
K(2)–O(56)
K(2)–O(59)
112.12(16)
112.02(16)
112.86(19)
106.03(18)
106.35(16)
106.89(17)
3
N(1)–C(1)
C(1)–B(1)
1.113(4)
1.609(4)
1.637(4)
1.655(4)
N(1)–C(1) –B(1)
C(1)–B(1)–C(11)
C(1)–B(1)–C(21)
C(11)–B(1)–C(21)
C(1)–B(1)–C(31)
C(11)–B(1)–C(31)
C(21)–B(1)–C(31)
K(1)–N(1)–C(1)
173.7(3)
111.3(2)
104.2(2)
108.3(2)
105.0(2)
113.0(2)
114.7(2)
121.60(11)
C(11)–B(1)
C(21)–B(1)
C(31)–B(1)
K(1)–N(1)
K(1)–O(53)
K(1)–O(56)
K(1)–O(41)
K(1)–O(44)
K(1)–O(47)
K(1)–O(50)
1.649(4)
2.9189(14)
2.9064(14)
2.8387(14)
2.8094(14)
2.8709(14)
2.7734(14)
2.8431(14)
7a
Fe(1)–N(2)
Fe(1)–C(38)
Fe(1)–C(40)
Fe(1)–C(42)
Fe(1)–C(43)
Fe(1)–C(44)
Fe(1)–C(45)
Fe(1)–C(46)
N(2)–C(3)
B(4)–C(3)
B(4)–C(5)
B(4)–C(16)
B(4)–C(27)
C(38)–O(39)
C(40)–O(41)
1.9198(18)
1.809(2)
1.793(2)
2.097(2)
2.078(2)
2.064(2)
2.103(2)
2.117(2)
1.135(3)
1.614(3)
1.636(3)
1.653(3)
1.646(3)
1.127(3)
1.122(3)
Fe(1)–N(2)–C(3)
N(2)–C(3)–B(4)
C(3)–B(4)–C(5)
C(3)–B(4)–C(16)
C(5)–B(4)–C(16)
C(3)–B(4)–C(27)
C(5)–B(4)–C(27)
C(16)–B(4)–C(27)
Fe(1)–C(38)–O(39)
Fe(1)–C(40)–O(41)
178.11(17)
174.9(2)
Fig. 3 ORTEP of [K(18-crown-6)][NCB(C6F5)3] (3).
111.55(16)
105.08(15)
107.28(15)
103.97(15)
113.94(15)
114.71(15)
179.7(2)
177.2(3)
B(C6F5)3}Me(CpЉ)2], where CpЉ = η-1,3-C5H3(SiMe3)2 (1.149(3)
Å).2 The C–B distances for 3 and 7a were found to be 1.609(4)
and 1.614(3) Å, respectively, close to the distance found for
C–B in [Zr{(µ-NC)B(C6F5)3}Me(CpЉ)2] (1.614(4) Å).2 For both
compounds, the B–C–N angles are approaching linearity
(174.9(2)Њ in 7a and 173.7(3)Њ in 3) The C(1)–N(1)–K(1) angle
found for 3 was 121.60(11)Њ, whereas the C–N–Fe unit of 7a
is almost linear (178.11(17)Њ). The latter is close to the angle
C–N–Zr in [Zr{(µ-NC)B(C6F5)3}Me(CpЉ)2], CpЉ
= η-1,3-
C5H3(SiMe3)2, which is 175.1(2)Њ.2
Fig. 4 ORTEP of [Fe{(µ-NC)B(C6F5)3}(η-C5H5)(CO)2] (7a).
DFT calculations
time. Geometry optimisations were carried out for the neutral
species: [Me3Si(µ-NC)BF3] (I), [Me3Si(µ-CN)BF3] (IЈ), [Hg-
{(µ-NC)BF3}2] (II), [Hg{(µ-CN)BF3}2] (IIЈ), [Fe{(µ-NC)-
BF3}(η-C5H5)(CO)2] (III), [Fe{(µ-CN)BF3}(η-C5H5)(CO)2]
(IIIЈ) and for the anionic species [NCBF3]Ϫ (IV), [CNBF3]Ϫ
Density functional calculations have been carried out to
investigate the relative stability of the system E1(µ-NC)B vs.
E1(µ-CN)B for several of the new compounds. The C6F5 groups
were substituted by fluorine atoms to reduce computational
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2553

View Article Online
Table 4 Crystallographic data for compounds 1, 3 and 7a
1
3
7a
Chemical formula
Formula weight
C31H24NBF15KO6S
873.47
C31H24NBF15KO6ؒC10.5H12
979.62
C26H5NBF15FeO2
714.97
Crystal system
Space group
Monoclinic
P21/c
Triclinic
P1
Triclinic
P1
¯
¯
Unit cell dimensions:
a/Å
b/Å
c/Å
α/Њ
β/Њ
8.9130(5)
21.0570(15)
18.7380(14)
90
97.339(5)
90
3488.0
4
1.66
12.8240(5)
12.9440(6)
13.6630(6)
108.484(2)
98.684(2)
90.807(2)
2121.7
2
9.9080(4)
10.4020(4)
12.6890(6)
97.594(2)
92.383(2)
99.853(2)
1274.4
2
γ/Њ
V/Å3
Z
Dc/g cmϪ3
1.533
1.86
T /K
150
0.34
150
0.240
150
0.73
µ/mmϪ1
Reflections collected
Independent reflections
35534
6629
17785
8022
7026
4866
Rint
0.04
0.018
0.03
Final R indices:
R
wR
0.0884
0.0152
0.0828
0.0712
0.0443
0.0409
Note: criterion for observation, I > 3σ(I ).
(IVЈ) and [F3BNCBF3]Ϫ (V). Optimisations have also been
performed on the model precursors Me3SiCN, Me3SiNC,
Hg(NC)2, Hg(CN)2, [Fe(NC)(η5-C5H5)(CO)2] and [Fe(CN)-
(η-C5H5)(CO)2]. The results of all geometry optimisations are
given in Table 5 together with the experimental values for com-
pounds [(C6F5)3B(µ-NC)B(C6F5)3],1,2 7a and 3. All E1–N–C–B
and E1–C–N–B systems are very close to linear. Reasonable
agreement between calculated and experimental distances
confirms that, in general, the model works well despite the
replacement of the C6F5 groups with fluorine atoms. In general
the N–C(B) distances are shorter than the C–N(B) ones. In
both systems the C–N distances decrease very slightly on co-
ordination by BF3. The N–B distances for the E1(µ-CN)B unit
and the C–B distances for the E1(µ-NC)B unit show significant
variation in the order E1 = Hg > Me3Si > Fp > BF3 > anion for
both series. The B–F distances correlate inversely with the B–N
and B–C distances.
Energies for the model compounds were calculated using
both the local density approximation (LDA) and using gradient
corrections (GGA). Table 6 gives the calculated isomerisation
energy for both the parent nitriles and isonitriles and the
E1(µ-CN)B vs. E1(µ-NC)B systems. Both methods give a similar
pattern. The almost universal preference is for the isomer
(E1(µ-CN)B.
Isomerisation between simple nitriles and isonitriles has been
investigated both experimentally and computationally.15,17,18
The general conclusions are that σ-accepting substituents
favour nitriles (E1CN) over isonitriles (E1NC) and that
π-accepting substituents favour isonitriles (E1NC) over nitriles
(E1CN).17 Our results are in agreement with this. In particular
they agree with the low energy preference found for silylnitriles
Si–CN where both σ- and π-accepting effects are important.
For example values ranging between 18 and 31 kJ molϪ1 have
been calculated for the isomerisation energy of H3Si–CN to
H3Si–NC using DFT.18b
the C and the N but is more localised on C. Thus the C is
generally the preferred basic atom in coordination to a Lewis
acid. As it is primarily non-bonding in character, little change
in CN distance results from coordination to a Lewis acid. Table
5 lists the Mulliken charges of C in the E1NC systems and of N
in the E1CN systems. In the parent E1–NC and E1–CN mole-
cules they are reduced compared with free cyanide, and further
reduced on coordination to BF3. Within a series the charge
varies Hg < SiMe3 < Fp < anion (the anion having the most
negative charge). This correlates well with the C–B and N–B
distances noted above, the least charged atoms forming the
longer bonds to BF3. Thus the better E1 as a σ acceptor, the
weaker the further bond for E2–BF3. From Table 6 it can be
seen that the isomerisation energies are less favourable after BF3
coordination. Just as nitriles E1CN are generally found to be
more stable than isonitriles E1NC because C is the more basic
site, the E1(µ-NC)B systems are relatively more stabilised on
BF3 coordination as in this case it is the C atom that binds to
the BF3 group. The isomerisation energy difference for E ᎐
᎐
1
SiMe3 is just sufficient to reverse the preference so that
[Me3Si(µ-NC)BF3] is found to be marginally more stable than
[Me3Si(µ-CN)BF3]. Since Me3SiNC and Me3SiCN are known
to be in equilibrium at room temperature, it is not surprising
that treatment of Me3SiCN with B(C6F5)3 leads to formation of
[Me3Si(µ-NC)B(C6F5)3] (4). In all other cases the calculations
indicate the complexes E1(µ-CN)B to be the thermodynamically
favoured isomer.
Conclusions
New main group and transition element derivatives of the
anions [SCNB(C6F5)3]Ϫ, [CNB(C6F5)3]Ϫ and [NCB(C6F5)3]Ϫ
have been prepared and characterised. New routes towards
formation of complexes containing the bulky anionic ligand
[NCB(C6F5)3]Ϫ are described. We have observed formation of
the E1(µ-NC)B(C6F5)3 linkages in the complexes 6 and 7a.
Calculations indicate that in general energetic preference is for
the E1(µ-CN)B.
The orbital responsible for the basic properties of the cyanide
ion is the HOMO, 5σ (Fig. 5). This has electron density on both
Experimental
All reactions were carried out under dinitrogen, either in a
glove-box or using conventional Schlenk-line techniques. The
compounds [FeCl(η-C5H5)(CO)2],19 [Fe(CN)(η-C5H5)(CO)2],19
Fig. 5 The 5σ orbital of CNϪ.
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2554

View Article Online
Table 5 Selected DFT calculated distances (Å) and angles (Њ) for I–V, IЈ-IVЈ, Me3SiCN, Me3SiNC, FpCN, FpNC, Hg(NC)2 and Hg(CN)2. Where
available experimental distances of the related B(C6F3)3 derivatives are given for comparison (Fp = Fe(η-C5H5)(CO)2)
Type E1NC
E1–N
N–C
C–B
B–F
Charge on C
I [Me3Si(µ-NC)BF3]a
Me3SiNC
II [Hg{(µ-NC)BF3}2]
Hg(NC)2
1.793
1.760
1.935
1.94
1.860
1.9198(18)
1.882
–
1.160
1.178
1.162
1.177
1.158
1.135(3)
1.175
1.161
1.113(4)
1.174
1.683
–
1.712
–
1.655
1.614(3)
–
1.621
1.609(4)
1.365
–
1.361
–
1.372
–
–
0.08
Ϫ0.11
0.03
Ϫ0.09
Ϫ0.07
III [Fp(µ-NC)BF3]
7ab
FpNC
Ϫ0.19
Ϫ0.07
IV [NCBF3]Ϫ
3c
1.391
–
–
[CN]Ϫ
Ϫ0.5
0.07
V [BF3(µ-NC)BF3]Ϫ
[(µ-NC){B(C6F5)3}2]Ϫ d
1.578
–
1.152
1.144(2)
1.644
1.583(2)
1.379
–
Type E1CN
E1–C
C–N
N–B
B–F
Charge on N
IЈ [Me3Si(µ-CN)BF3]a
Me3SiCN
IIЈ [Hg{(µ-CN)BF3}2]
Hg(CN)2
1.885
1.885
1.986
1.989
1.828
1.872
–
1.152
1.161
1.150
1.160
1.158
1.165
1.168
1.174
1.152
1.144(2)
1.643
–
1.752
–
1.596
–
1.537
1.353
–
1.350
–
1.367
–
1.386
Ϫ0.08
Ϫ0.22
0.03
Ϫ0.19
Ϫ0.13
Ϫ0.30
Ϫ0.20
Ϫ0.5
IIIЈ [Fp(µ-CN)BF3]
FpCN
IVЈ [CNBF3]Ϫ
[CN]Ϫ
V [BF3(µ-CN)BF3]Ϫ
1.644
–
1.578
1.593(2)
1.379
–
Ϫ0.2
[(µ-CN){B(C6F5)3}2]Ϫ d
Note: a Average distances (Å) for Si–C(Me): 1.845 (I) and 1.850 (IЈ). b Average distances (Å): Fe–C(Cp) 2.070 (IIЈ), 2.074 (II); Fe–C(O): 1.743 (II),
1.739 (IIЈ), C–O 1.148 (II) and 1.147 (IIЈ). c X-Ray values, this work. d X-Ray values, literature data.1,2
Table 6 Calculated energies ∆Eisom (kJ molϪ1) for the isonitrile (E1–NC and E1(µ-NC)B)
at both the LDA and the GGA level
nitrile (E1–CN and E1(µ-CN)B) isomerisation energies
Isomerisation
∆Eisom/kJ molϪ1 (GGA) ∆Eisom/kJ molϪ1 (LDA)
[Me3Si(µ-NC)BF3]
[Me3Si(µ-CN)BF3]
[Hg{(µ-CN)BF3}2]
[Fp(µ-CN)BF3]
I
IЈ
ϩ3
Ϫ20
Ϫ96
Ϫ149
Ϫ60
Ϫ79
ϩ24
ϩ2
Ϫ18
Ϫ102
Ϫ153
Ϫ73
Ϫ87
ϩ20
Me3SiNC
Me3SiCN
[Hg{(µ-NC)BF3}2]
II IIЈ
[Hg(NC)2]
[Hg(CN)2]
[Fp(µ-NC)BF3]
III IIIЈ
IV IVЈ
[FpNC]
[FpCN]
[NCBF3]Ϫ
[CNBF3]Ϫ
and B(C6F5)3,20 were prepared according to the literature
methods. [IrCl(PPh3)2(CO)] was used as supplied by Aldrich.
The compound B(C6F5)3 was freshly sublimed before use. NMR
spectra were recorded on either a 300 MHz Varian Mercury or
a 500 MHz Varian Unity spectrometer, referenced internally
using the residual protio-solvent, and chemical shifts were
74.5%) was recrystallised from a mixture of CH2Cl2 and penta-
ne to give white crystals of 1. The reaction was subsequently
repeated in benzene, THF or CH2Cl2 and similar results were
obtained.
Elemental analysis (%). Calc. (found) for C31H24NBF15KO6S:
C 42.6 (42.3), H 2.8 (2.9), N 1.6 (1.5). FABϩ m/z (intensity (%)):
873.2 (100) [SCNB(C6F5)3]Ϫ.
.
reported with respect to SiMe4 (1H and 13C, δ = 0), BF3 Et2-
O (11B, δ = 0), H3PO4 (31P, δ = 0) and CFCl3 (19F, δ = 0). All
chemical shifts are quoted in δ (ppm) and coupling constants in
Hz. Infrared spectra were recorded on either a Perkin-Elmer
1710 IR FT spectrometer or a Perkin-Elmer 1600 FTIR. Ele-
mental analyses and mass spectra were provided by the Micro-
analytical Services of the Inorganic Chemistry Laboratory,
Oxford.
[K(18-crown-6)][(C6F5)3B(ꢀ-CN)B(C6F5)3] (2). Method 1. In
a glove-box, a Schlenk vessel was charged with KCN (25 mg,
0.38 mmol), B(C6F5)3 (200 mg, 0.39 mmol). A separate Schlenk
vessel was charged with 103 mg (0.391 mmol) 18-crown-6 ether
and toluene (50 mL) was added and the resulting solution was
added slowly to the initial mixture. This was then stirred at
room temperature for 24 h and the solution turned orange. The
white solid formed was isolated by filtration and dried under
Preparations
1
reduced pressure. H- and 11B{1H}-NMR spectra showed that
[K(18-crown-6)][SCNB(C6F5)3] (1). In a glove-box, a Schlenk
vessel was charged with KSCN (38 mg, 0.39 mmol), B(C6F5)3
(200 mg, 0.39 mmol). A separate Schlenk vessel was charged
with 18-crown-6 ether (103 mg, 0.39 mmol) and 50 mL of tolu-
ene was added. This solution was added slowly under stirring
to the Schlenk vessel containing a mixture of KSCN and
B(C6F5)3. The mixture was stirred at room temperature for
5 days. Volatiles were removed under reduced pressure to
yield an off-white oily material, which was washed with 20 mL
cold pentane. The resulting powder (crude yield: 254 mg,
this solid contains mainly starting materials. From the filtrate,
volatiles were removed under reduced pressure to yield a brown
oil. Recrystallisation from CH2Cl2 gave a small amount of
light-yellow crystals (yield <10%) characterised by 11B{1H}
NMR as [K(18-crown-6)][(C6F5)3B(µ-CN)B(C6F5)3] (2). The
reaction was subsequently repeated in a number of different
solvents (THF, toluene, CH2Cl2 or CH3CN) using 2 : 2 : 1 or
1 : 1 : 2 ratios of 18-crown-6 ether, KCN and B(C6F5)3. In each
case, only the compound [K(18-crown-6)][(C6F5)3B(µ-CN)-
B(C6F5)3] (2) was obtained in low yield.
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2555

View Article Online
Method 2. In a glove-box, a Schlenk vessel was charged with
18-crown-6 ether (1 g, 3.8 mmol) and KCN (247 mg, 3.8 mmol).
Dry methanol was then added, and the mixture was stirred at
40 ЊC overnight. The solution was then cooled to room temper-
ature and the MeOH removed under reduced pressure. The
residual white solid, was soluble in CH2Cl2 and insoluble in
toluene and was washed with pentane and dried under reduced
pressure to yield 1.16 g (93% yield) of [K(18-crown-6)CN].
Then B(C6F5)3 in CH2Cl2 (50 mL) was added slowly to a solu-
tion of [K(18-crown-6)CN] (300 mg, 0.91 mmol) in 20 mL
CH2Cl2. The mixture was stirred at room temperature under
N2 for 30 min. The volatiles were removed from the light-
yellow solution under reduced pressure to give a creamy-white
solid. This was extracted with toluene (10 mL). The toluene was
removed under reduced pressure and the resulting brown, oily
solid was recrystallised from a mixture of CH2Cl2 and pentane
to give light-yellow crystals of [K(18-crown-6)][(C6F5)3B(µ-CN)-
B(C6F5)3] (2). Yield: 540 mg, 36%.
were mixed in a Schlenk vessel and 50 mL CH2Cl2 added. The
mixture was stirred at room temperature under N2 for 15 h.
From the clear solution, the volatiles were removed under
reduced pressure to give a white solid, which was washed with
pentane (2 × 10 mL) and dried under reduced pressure to give
1.1 g (87% yield) 5 as a white solid.
Elemental analysis (%). Calc. (found) for C38N2F30B2Hg: C
35.75 (36.0), N 2.2 (2.3). FABϩ m/z (intensity (%)): 1277.6 (85)
[Hg{(µ-CN)B(C6F5)3}2]ϩ.
[Ir{(ꢀ-NC)B(C6F5)3}(PPh3)2(CO)]
(6).
[Me3Si(µ-NC)-
B(C6F5)3] (4) was prepared in situ by addition of Me3SiCN
(0.033 mL, 0.25 mmol) to a solution of B(C6F5)3 (128 mg,
0.25 mmol) in CH2Cl2 (50 mL). The mixture was then stirred for
5–10 min before being added to a solution of [IrCl(PPh3)2(CO)]
(195 mg, 0.25 mmol) in CH2Cl2 (50 mL). The solution was left
stirring for 2 h, and the initially pale yellow colour changed to a
bright-yellow. The solution was subsequently filtered, volatiles
removed under reduced pressure and the resulting residue was
washed with pentane (2 × 20 mL) to yield a yellow powder. The
solid was then recrystallised from CH2Cl2 at Ϫ80 ЊC to give
compound 5 (120 mg, 37% yield) as a microcrystalline yellow
powder.
Elemental analysis (%). Calc. (found) for C13H24NKO6: C
47.4 (47.7), H 7.3 (7.2), N 4.25 (4.3). Calc. (found) for
C49H24NB2F30KO6: C 43.5 (43.27), H 1.8 (1.9), N 1.0 (1.0). ESϪ
m/z (intensity (%)): 1046.28 (30) [(C6F5)3B(µ-CN)B(C6F5)3]Ϫ.
[K(18-crown-6)][NCB(C6F5)3] (3). Method 1. In a glove-box, a
two-neck, round bottom flask was charged with potassium
metal (230 mg, 5.88 mmol), [Hg{(µ-CN)B(C6F5)3}2] (5) (1 g,
0.783 mmol) (prepared as described below) and 18-crown-6
ether (415 mg, 1.57 mmol). Toluene (200 mL) was added and
the mixture was stirred at the room temperature for 1 h. The
mixture became cloudy, and was then heated at 90 ЊC for 7 h.
The resulting black solid was removed by filtration. From the
light-yellow filtrate, volatiles were removed under reduced pres-
sure to yield an off-white solid. Recrystallisation from toluene
at Ϫ80 ЊC gave the compound [K(18-crown-6)][NC{B(C6F5)3}2]
(3) (450 mg, 29% yield) as colourless needle-like crystals.
Method 2. A solution of 18-crown-6 ether (140 mg, 0.53
mmol) in toluene (50 mL) was added to KCl (140 mg, 1.88
mmol) and the resulting mixture was stirred for 2 h. at room
temperature. A solution of [Me3Si(µ-NC)B(C6F5)3] (4) (0.31 g,
0.53 mmol) (prepared as described below) in toluene (50 mL)
was subsequently added and the mixture stirred for 2 h. The
solids were removed by filtration and, from the filtrate, volatiles
were removed under reduced pressure to give a white solid.
After recrystallisation from toluene at Ϫ80 ЊC, 323 mg (73%
yield) of compound 3 were obtained as fine colourless crystals,
suitable for study by X-ray diffraction.
Elemental analysis (%). Calc. (found) for C56H30NOBF15IrP:
C 52.4 (52.2), H 1.1 (1.2), N 2.4 (2.65). FABϩ m/z (intensity
(%)): 1282.5 (8.5) [Ir{(µ-NC)B(C6F5)3}(PPh3)2(CO)]ϩ.
[Fe{(ꢀ-NC)B(C6F5)3}(ꢁ-C5H5)(CO)2] (7a). [Me3Si(µ-NC)-
B(C6F5)3] (4) was prepared in situ by addition of Me3SiCN
(0.13 mL, 1.00 mmol) to a solution of B(C6F5)3 (512 mg, 1.00
mmol) in CH2Cl2 (50 mL). The mixture was then stirred for 30
min before being added to a solution of [Fe(η-C5H5)(CO)2Cl]
(213 mg, 1.00 mmol) in CH2Cl2 (50 mL). The solution was then
left stirring overnight, after which time the red colour had dark-
ened to red–brown. The volatiles were subsequently removed
under reduced pressure to a red solid. This was washed with
light petroleum (bp 40–60 ЊC) (2 × 15 mL) and recrystallised
from a 1 : 1 mixture of pentane and CH2Cl2 to give 486 mg
(68% yield) of 7a as a red, microcrystalline solid. Crystals of
7a, suitable or for X-ray analysis were obtained by diffusion of
pentane into a concentrated CH2Cl2 solution.
Elemental analysis (%). Calc.(found) for C26H5NBF15FeNO2:
C 43.7 (43.2), H 0.7 (0.3) N 2.00 (1.8) Fe 7.8 (7.1). FABϩ
m/z (intensity (%)): 715.0(5) [Fe{NCB(C6F5)3}(C5H5)(CO)2]ϩ,
658.9(13.5) [Fe{(µ-NC)B(C6F5)3}(η-C5H5)]ϩ.
Elemental analysis (%). Calc. (found) for C31H24NBF15KO6:
C 44.25 (44.0), H 2.8 (2.85), N 1.7 (1.7). ESϪ m/z (intensity (%)):
536.09 (100) [CN{B(C6F5)3}2]Ϫ.
[Fe(ꢀ-CN)B(C6F5)3(ꢁ-C5H5)(CO)2} (7b). The solids B(C6F5)3
(20 mg, 0.039 mmol) and [Fe(CN)(η5-C5H5)(CO)2] (5 mg,
0.025 mmol) were placed in a Young’s tap NMR tube and
1
C6D6 (or CD2Cl2) added. H-, 19F- and 11B{1H}-NMR spectra
[Me3Si(ꢀ-NC)B(C6F5)3] (4). NMR-scale reaction. The com-
pounds Me3SiCN (0.013 mL, 0.1 mmol) and B(C6F5)3 (0.05 g,
0.1 mmol) were mixed at room temperature in a glove-box in a
Young’s tap NMR tube and C6D6 was added. The NMR
spectra, recorded within 15 min, showed quantitative formation
of [Me3Si(µ-NC)B(C6F5)3] (4). Synthesis. Me3SiCN (0.15 mL,
1.12 mmol) in 40 mL CH2Cl2 was added to B(C6F5)3 (0.6 g, 1.17
mmol) in 20 mL CH2Cl2 and the resulting mixture was sub-
sequently stirred for 1 h. The solvents were removed under
reduced pressure, the crude yellow–white powder was washed
with pentane (2 × 15 mL), and the product was recrystal-
lised from a saturated CH2Cl2 solution layered with pentane.
Compound 4 (530 mg, 77% yield) was isolated as a white,
micro-crystalline material.
recorded within 15 min showed the quantitative formation of
[Fe{(µ-CN)B(C6F5)3}(η-C5H5)(CO)2] (7b).
X-Ray crystallography
In each case, a single crystal was selected under an inert atmos-
phere, encased in perfluoro-polyether oil, and mounted on the
end of a glass fibre. The fibre, secured in a goniometer head was
then placed under a stream of cold nitrogen maintained at 150
K and data collected on an Enraf-Nonius DIP2000 image plate
diffractometer with graphite monochromated Mo-Kα radiation
(λ = 0.71069 Å) The images were processed with the DENZO
and SCALEPACK programs21 and corrections for Lorentz
and polarisation effects were performed. All solution, refine-
ment, and graphical calculations were performed using the
CRYSTALS22 and CAMERON23 software packages. The
structures were solved by direct methods using the SIR9224
program and refined by full-matrix least squares procedure on
F. All non-hydrogen atoms were refined with anisotropic dis-
placement parameters. All hydrogen atoms were generated and
Elemental analysis (%). Calc. (found) for C22H9NBF15Si: C
43.2 (42.9), H 1.5 (1.1), N 2.3 (2.0). FABϩ m/z (intensity (%)):
610.2 (40) [Me3SiNCB(C6F5)3]ϩ.
[Hg{(ꢀ-CN)B(C6F5)3}2] (5). In a glove-box, the compounds
[Hg(CN)2] (0.25 g, 0.97 mmol) and B(C6F5)3 (1 g, 1.95 mmol)
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2556

View Article Online
2 J. Zhou, S. J. Lancaster, D. A. Walker, S. Beck, M. Thornton-Pett
allowed to ride on their corresponding carbon atoms with fixed
thermal parameters. An empirical absorption correction was
applied.25 The crystallographic data are summarised in Table 3.
In addition, refinement using full-matrix least squares on all
F 2 data26 was also performed for 3 and 7a. Both refinements
methods (either on F or on F 2) gave better results (lower R and
wR) for the model with the CN group N-bound to the metal
(K in case of 3, or Fe in case of 7a) than for the model where
this group is C-bond to the metal. The inspection of the param-
eters for the thermal ellipsoids also indicated that the model
containing NCB units are better solutions to the data sets than
those incorporating the CNB units. These results are consistent
with the observations from 11B{1H} NMR spectroscopy and the
previously reported data.2
and M. Bochmann, J. Am. Chem. Soc., 2001, 123, 223.
3 A. A. Danopoulos, J. R. Galsworthy, M. L. H. Green, S. Cafferkey,
L. H. Doerrer and M. B. Hursthouse, Chem. Commun., 1998,
2529.
4 L. H. Doerrer, J. R. Galsworthy, M. L. H. Green, M. A. Leech and
M. Mueller, J. Chem. Soc., Dalton Trans., 1998, 3191.
5 L. H. Doerrer, J. R. Galsworthy, M. L. H. Green and M. A. Leech,
J. Chem. Soc., Dalton Trans., 1998, 2483.
6 L. H. Doerrer, A. J. Graham and M. L. H. Green, J. Chem. Soc.,
Dalton Trans., 1998, 3941.
7 L. H. Doerrer and M. L. H. Green, J. Chem. Soc., Dalton Trans.,
1999, 4325.
8 J. L. Priego, L. H. Doerrer, L. H. Rees and M. L. H. Green,
Chem. Commun., 2000, 779.
9 H. Jacobsen, H. Berke, S. Doering, G. Kehr, G. Erker, R. Froehlich
and O. Meyer, Organometallics, 1999, 18, 1724.
10 L. Li and T. J. Marks, Organometallics, 1998, 17, 3996.
11 C. Bergquist, B. M. Bridgewater, C. J. Harlan, J. R. Norton,
R. A. Friesner and G. Parkin, J. Am. Chem. Soc., 2000, 122,
10581.
CCDC reference numbers 204648–204650.
See http://www.rsc.org/suppdata/dt/b3/b302163g/ for crystal-
lographic data in CIF or other electronic format.
12 D. E. J. Arnold, S. Cradock, E. A. V. Ebsworth, J. D. Murdoch,
D. W. H. Rankin, D. C. J. Skea, R. K. Harris and B. J. Kimber,
J. Chem. Soc., Dalton Trans., 1981, 1349.
13 W. R. Hertler, D. A. Dixon, E. W. Matthews, F. Davidson and
F. G. Kitson, J. Am. Chem. Soc., 1987, 109, 6532.
14 K. Georgiou and A. C. Legon, J. Mol. Struct., 1982, 78, 257.
15 I. Ugi, Isocyanide Chemistry, Academic Press, New York,
1971.
16 G. Barrado, L. H. Doerrer, M. L. H. Green and M. A. Leech,
J. Chem. Soc., Dalton Trans., 1999, 1061.
17 H. Vahrenkamp, A. Geiss and G. N. Richardson, J. Chem. Soc.,
Dalton Trans., 1997, 3643.
Computational details
All calculations were performed using density functional
methods of the Amsterdam Density Functional Package (ver-
sion 2000.2).27–30 The electronic configurations of the molecular
systems were described using type IV basis sets with triple ζ
accuracy sets of Slater type orbitals, with a single polarisation
function added to the main group atoms. The cores of the
atoms were frozen up to 2p for Fe, 4d for Hg, 1s for C, O and N,
2p for Si. First-order relativistic corrections were made to the
cores of all atoms using the ZORA formalism. The generalised
gradient approximation (GGA non-local) method was used,
using Vosko, Wilk and Nusair’s local exchange correlation31
with non-local exchange corrections by Becke32 and non-local
correlation corrections by Perdew.33 The non-local correction
terms were not utilised in calculating gradients during geometry
optimisations as better agreement with experimental structure
is normally achieved with a local density approximation.
18 (a) K. Sung, J. Org. Chem., 1999, 64, 8984; (b) C. Zanchini, Chem.
Phys., 2000, 254, 187.
19 T. S. Piper, F. A. Cotton and G. Wilkinson, J. Inorg. Nucl. Chem.,
1955, 1, 165.
20 J. L. W. Pohlmann and F. E. Brinckmann, Z. Naturforsch., Teil B,
1965, 20, 5.
21 Z. Otwinowski and W. Minor, Methods Enzymol., 1996, 276.
22 D. J. Watkin, C. K. Prout, J. R. Carruthers and P. W. Betteridge, in
CRYSTALS, Issue 10, Chemical Crystallography Laboratory,
Oxford, 1996.
23 D. J. Watkin, C. K. Prout and L. J. Pearce, in CAMERON, Chemical
Crystallography Laboratory, Oxford, 1996.
Acknowledgements
24 A. Altomare, G. Carascano, C. Giacovazzo and A. Guagliardi,
J. Appl. Crystallogr., 1993, 26, 343.
25 N. Walker and D. Stuart, Acta Crystallogr., Sect. A, 1983, 39,
158.
26 G. M. Sheldrick, in SHELXL93 - Program for Crystal Structure
Refinement, Göttingen, 1993.
27 E. J. Baerends, E. G. Ellis and P. Ros, Chem. Phys., 1973, 2, 41.
28 C. Fonseca Guerra, J. G. Snijider, B. te Velde and E. J. Baerends,
Theor. Chem. Acc., 1998, 99, 391.
29 B. te Velde and E. J. Baerends, J. Comput. Phys., 1992, 99, 84.
30 L. Verluis and T. Ziegler, J. Chem. Phys., 1988, 88, 322.
31 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1990, 58,
1200.
We thank Dr Neale G. Jones for helpful discussions, Drs David
J Watkin and Andrew R. Cowley for assistance with crystallo-
graphy, Balliol College Oxford for a Dervorguilla Scholarship
(S. I. P), the AG Leventis Foundation for a grant (to I.C.V) and
EPSRC for support (to G. D. W. A.). Part of this work was
carried out using the resources of the Oxford Super Computing
Centre and part using the EPSRC Columbus cluster at the
Rutherford Laboratory.
References
1 S. J. Lancaster, D. A. Walker, M. Thornton-Pett and M. Bochmann,
Chem. Commun., 1999, 1533.
32 A. D. Becke, Phys. Rev. A, 1988, 38, 2398.
33 J. Perdew, Phys. Rev. B, 1986, 33, 8822.
D a l t o n T r a n s . , 2 0 0 3 , 2 5 5 0 – 2 5 5 7
2557