Synthesis of nickel-monocarbollide complexes by oxidative insertion

Article abstract of DOI:10.1021/ic051115c

Treatment of the 11-vertex carborane anion [closo-2-CB₁₀H ₁₁]^- with Ni(0) reagents in tetrahydrofuran (THF) affords - via oxidative insertion reactions - 12-vertex Ni(II) complexes, isolated as the salts [N(PPh₃)₂][2,2-L₂-closo-2,1-NiCB ₁₀H₁₁] (L = CO (1a), CNBu^t (1b), and CNXyl (1c; Xyl = C₆H₃Me₂-2,6); L₂ = cod (1d; cod = 1,2:5,6-η-cyclo-octa-1,5-diene)). One CO ligand in 1a is readily replaced by donors L′ in the presence of Me₃NO to give the species [N(PPh₃)₂][2-CO-2-L′-closo-2,1-NiCB ₁₀H₁₁] (L′ = PEt₃ (1e), PPh₃ (1f), CNBu^t (1g), and CNXyl (1h)). The anionic complexes themselves readily react with hydride abstracting reagents in the presence of donor ligands to yield zwitterionic complexes in which boron vertexes bear substituents that are bound through C, N, or O atoms. Thus, for example, 1c with H⁺ and CNXyl gives [2,2,7-(CNXyl)₃-closo-2,1-NiCB₁₀H ₁₀] (2b), while 1f with Me⁺ in the presence of OEt ₂ affords [2-CO-2,11-{μ-PPh₂(C₆H ₄-o)}-7-OEt₂-closo-2,1-NiCB₁₀H₉] (4), in which an additional cycloboronation of one phosphine phenyl ring has occurred. In contrast, 1f with Me⁺ in the presence of NCMe gives a mixture of the isomers [2-CO-2-PPh₃-7-{(X)-N(Me)=C(H)Me}-closo-2,1- NiCB₁₀H₁₀] (X ≡ E (5c) and Z (5d)). X-ray diffraction analyses of compounds 1a, 2b, 4, and 5c confirmed their important structural features.

Full text of DOI:10.1021/ic051115c

Inorg. Chem. 2005, 44, 8135−8144
Synthesis of Nickel−Monocarbollide Complexes by Oxidative Insertion
Thomas D. McGrath, Andreas Franken, Jason A. Kautz, and F. Gordon A. Stone*
Department of Chemistry and Biochemistry, Baylor UniVersity, Waco, Texas 76798-7348
Received July 5, 2005
Treatment of the 11-vertex carborane anion [closo-2-CB10H11]- with Ni(0) reagents in tetrahydrofuran (THF) affords
via oxidative insertion reactions 12-vertex Ni(II) complexes, isolated as the salts [N(PPh3)2][2,2-L2-closo-2,1-NiCB10H11]
(L C6H3Me2-2,6); L2 cod (1d; cod 1,2:5,6- -cyclo-octa-1,5-
CO (1a), CNBut (1b), and CNXyl (1c; Xyl
diene)). One CO ligand in 1a is readily replaced by donors L in the presence of Me3NO to give the species
[N(PPh3)2][2-CO-2-L
-closo-2,1-NiCB10H11] (L′ ) PEt3 (1e), PPh3 (1f), CNBut (1g), and CNXyl (1h)). The anionic
s
s
)
)
)
)
η
′
′
complexes themselves readily react with hydride abstracting reagents in the presence of donor ligands to yield
zwitterionic complexes in which boron vertexes bear substituents that are bound through C, N, or O atoms. Thus,
for example, 1c with H+ and CNXyl gives [2,2,7-(CNXyl)3-closo-2,1-NiCB10H10] (2b), while 1f with Me+ in the presence
of OEt2 affords [2-CO-2,11-{µ-PPh2(C6H4-o)
of one phosphine phenyl ring has occurred. In contrast, 1f with Me+ in the presence of NCMe gives a mixture of
the isomers [2-CO-2-PPh3-7- (X)-N(Me) C(H)Me -closo-2,1-NiCB10H10] (X E (5c) and Z (5d)). X-ray diffraction
analyses of compounds 1a, 2b, 4, and 5c confirmed their important structural features.
}-7-OEt2-closo-2,1-NiCB10H9] (4), in which an additional cycloboronation
{
d
}
≡
Introduction
monocarbollide derivatives. In an attempt to constructively
address this imbalance,4 we continue to prepare and study
the reactivity of “piano stool” complexes in which a metal
center is bonded on one side by a monocarbollide ligand
and on the other by CO groups, the latter often in combina-
tion with isocyanides or phosphines.6-10
The area of carbametallaborane chemistry was initiated
almost 40 years ago by the discovery that the dicarbollide
ligand [nido-7,8-C2B9H11]2- could form π complexes with
transition metals.1 It was subsequently shown that the
monocarbollide ligand [nido-7-CB10H11]3- had similar ligat-
ing properties.2 Like the isolobal cyclopentadienide ligand
[C5H5]-, these di- and monocarbollide ligands behave as
pentahapto 6π-electron donors to transition element centers,
but they have, in addition, a higher negative charge that leads
to often pronounced differences in behavior.3,4 However,
whereas cyclopentadienide-metal complexes have played
a central role in the development of modern organometallic
chemistry,5 there has been far less study of the corresponding
carbollide species:3,4 this is particularly so in the case of the
Thus far, the 12-vertex parent species [2,2,2,2-(CO)4-closo-
2,1-MCB10H11]- (M ) Mo8 and W10) and [2,2,2-(CO)3-closo-
2,1-MCB10H11]n- (n ) 2, M ) Re;7 n ) 1, M ) Fe,8 Ru,9
and Os9) have all been reported and their chemistry has been
explored.7-14 In seeking to extend this series to encompass
(5) See, for example: (a) ComprehensiVe Organometallic Chemistry;
Wilkinson, G., Abel, E. W., Stone, F. G. A., Eds.; Pergamon Press:
Oxford, U.K., 1982. (b) ComprehensiVe Organometallic Chemistry
II; Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.; Pergamon
Press: Oxford, U.K., 1995.
(6) Batten, S. A.; Jeffery, J. C.; Jones, P. L.; Mullica, D. F.; Rudd, M.
D.; Sappenfield, E. L.; Stone, F. G. A.; Wolf, A. Inorg. Chem. 1997,
36, 2570.
(7) Blandford, I.; Jeffery, J. C.; Jelliss, P. A.; Stone, F. G. A. Organo-
metallics 1998, 17, 1402.
* To whom correspondence should be addressed. E-mail: gordon_stone@
baylor.edu.
(1) Hawthorne, M. F.; Young, D. C.; Wegner, P. A. J. Am. Chem. Soc.
1965, 87, 1818.
(2) (a) Hyatt, D. E.; Little, J. L.; Moran, J. T.; Scholer, F. R.; Todd, L. J.
J. Am. Chem. Soc. 1967, 89, 3342. (b) Knoth, W. H. J. Am. Chem.
Soc. 1967, 89, 3342.
(3) Grimes, R. N. In ComprehensiVe Organometallic Chemistry; Wilkin-
son, G., Abel, E. W., Stone, F. G. A., Eds.; Pergamon Press: Oxford,
U.K., 1982; Vol. 1, Section 5.5. Grimes, R. N. In ComprehensiVe
Organometallic Chemistry II; Abel, E. W., Stone, F. G. A., Wilkinson,
G., Eds.; Pergamon Press: Oxford, U.K., 1995; Vol. 1, Chapter 9.
Grimes, R. N. Coord. Chem. ReV. 2000, 200-202, 773.
(4) McGrath, T. D.; Stone, F. G. A. AdV. Organomet. Chem., in press.
(8) Ellis, D. D.; Franken, A.; Jelliss, P. A.; Stone, F. G. A.; Yu, P.-Y.
Organometallics 2000, 19, 1993.
(9) Ellis, D. D.; Franken, A.; McGrath, T. D.; Stone, F. G. A. J.
Organomet. Chem. 2000, 614-615, 208.
(10) Du, S.; Franken, A.; Jelliss, P. A.; Kautz, J. A.; Stone, F. G. A.; Yu,
P.-Y. J. Chem. Soc., Dalton Trans. 2001, 1846.
(11) Jeffery, J. C.; Jelliss, P. A.; Rees, L. H.; Stone, F. G. A. Organome-
tallics 1998, 17, 2258.
(12) Ellis, D. D.; Franken, A.; Jelliss, P. A.; Kautz, J. A.; Stone, F. G. A.;
Yu, P.-Y. J. Chem. Soc., Dalton Trans. 2000, 2509.
10.1021/ic051115c CCC: $30.25
Published on Web 09/22/2005
© 2005 American Chemical Society
Inorganic Chemistry, Vol. 44, No. 22, 2005 8135

McGrath et al.
Chart 1
elements further to the right of the periodic table, we have
described the cobalt-nitrosyl complex [2-CO-2-NO-closo-
2,1-CoCB10H11]-,15 an analogue of the hitherto elusive
dicarbonyl species [2,2-(CO)2-closo-2,1-CoCB10H11]2-. We
now report upon the synthesis and reactivity of the corre-
sponding dicarbonyl nickel anion [2,2-(CO)2-closo-2,1-
NiCB10H11]-, along with several analogues, which extends
the series of carbonyl-metal-monocarbollide complexes
into group 10.
Results and Discussion
Previously, the monocarbollide-metal-carbonyl com-
plexes [2-(CO)x-closo-2,1-MCB10H11]n- have been prepared
via two major routes: namely, treatment of a suitable metal
carbonyl reagent with the sodium salt of the carborane
trianion [nido-7-CB10H11]3-, followed by oxidation where
necessary,7,8,10 or interaction of a metal carbonyl with [nido-
7-CB10H13]- at elevated temperatures, in which case the
metal center is oxidized by the concomitant reduction of two
face-bridging protons from the carborane and their elimina-
tion as H2.8,9 For our target nickel-monocarbollide complex,
[2,2-(CO)2-closo-2,1-NiCB10H11]- the former route appeared
somewhat unattractive, as it would either involve the
presumed formal intermediate [2-CO-closo-2,1-NiCB10H11]3-
or require a reagent such as [NiCl2(CO)2]. However, an
alternative route also presents itself. Some three decades ago,
it was discovered that zerovalent d10 metal centers could
oxidatively insert into closo-carboranes, with the resulting
net two-electron transfer opening the cluster to afford a nido-
carborane ligand which complexed with the divalent metal
fragment so formed.16 Indeed, the potential of this “oxidative
insertion” approach has already successfully been demon-
strated for the reaction of M0 reagents (M ) Ni, Pd, and Pt)
with {closo-2-CB10} fragments to give species with a {closo-
2,1-MCB10} architecture.17
It was recognized that extending this methodology to use
[Ni(CO)4] as the Ni0 reagent could provide an entry into
carbonyl-nickel-monocarbollide chemistry, provided that
the nickel reagent is sufficiently nucleophilic. Thus, treatment
of [NMe4][closo-2-CB10H11] with [Ni(CO)4] (conveniently
prepared in situ18 from [Ni(cod)2] and CO) in THF, followed
by addition of [N(PPh3)2]Cl and chromatographic workup,
afforded the desired parent species [N(PPh3)2][2,2-(CO)2-
closo-2,1-NiCB10H11] (1a) in good yields (see Chart 1 for
the structure). Note that the aforementioned species [N(PPh3)2]-
[2-CO-2-NO-closo-2,1-CoCB10H11], formally isoelectronic
with 1a, was similarly prepared using [Co(CO)3(NO)] as the
metal reagent.15 Two bis(isocyanide) analogues of 1a, namely
[N(PPh3)2][2,2-(CNR)2-closo-2,1-NiCB10H11] (R ) But (1b)
and Xyl (1c)), could likewise be obtained from [NMe4][closo-
2-CB10H11] and [Ni(CNR)2(cod)], with the latter being
prepared in situ from [Ni(cod)2] and CNR, followed by
addition of [N(PPh3)2]Cl. Moreover, the NiII-diene complex
[N(PPh3)2][2,2-(cod)-closo-2,1-NiCB10H11] (1d) was syn-
thesized analogously using [Ni(cod)2] itself as the zerovalent
nickel source. Although the anion of 1b, isolated as the
[NMe4]+ salt, was prepared in a similar fashion in the original
study upon this system,17 its NMR spectroscopic properties
are reported fully here for the first time.
Compound 1a is surprisingly stable. Analogous dicarbol-
lide species [1,2-Me2-3,3-(CO)2-closo-3,1,2-NiC2B9H9]19 and
[2,2-(CO)2-closo-2,1,7-NiC2B9H11]20 have been prepared and
found to be somewhat unstable, particularly with respect to
carbonyl ligand loss. In contrast, 1a is essentially unchanged
after many months at room temperature in the solid state;
compounds 1b-d are similarly robust. The anion of 1d is
itself also noteworthy as a rare example of an anionic
Ni(II)-olefin complex.21
Data characterizing all of the compounds 1a-d are
presented in Tables 1-3. The CO and CN stretching
frequencies of anionic 1a and b may be compared with those
of their neutral dicarbollide analogues. For 1a, the IR
spectrum shows two strong bands in the CO region at 2035
and 2084 cm-1, whereas for the corresponding {closo-
NiC2B9} derivatives the comparable frequencies are 2081
and 2093 cm-1 (3,1,2-isomer)19 and 2098 and 2130 cm-1
(2,1,7-isomer).20 Similarly, compound 1b shows CN stretches
at 2132 and 2163 cm-1, compared with 2178 and 2197 cm-1
(both the 3,1,2- and 2,1,7-isomers).19,20 These differences are
readily understood in terms of the greater electron density
at nickel in the anionic species than that in the neutral ones
and, hence, the stronger metal-to-ligand back-bonding. In
(13) Franken, A.; Du, S.; Jelliss, P. A.; Kautz, J. A.; Stone, F. G. A.
Organometallics 2001, 20, 1597.
(14) Du, S.; Kautz, J. A.; McGrath, T. D.; Stone, F. G. A. J. Chem. Soc.,
Dalton Trans. 2001, 2791.
(15) Kautz, J. A.; McGrath, T. D.; Stone, F. G. A. Polyhedron 2003, 22,
109.
(16) Stone, F. G. A. J. Organomet. Chem. 1975, 100, 257.
(17) Carroll, W. E.; Green, M.; Stone, F. G. A.; Welch, A. J. J. Chem.
Soc., Dalton Trans. 1975, 2263.
(18) Jolly, P. W. In ComprehensiVe Organometallic Chemistry; Wilkinson,
G., Abel, E. W., Stone, F. G. A., Eds.; Pergamon Press: Oxford, U.K.,
1982; Vol. 6, Section 37.2.
(19) Carr, N.; Mullica, D. F.; Sappenfield, E. L.; Stone, F. G. A. Inorg.
Chem. 1994, 33, 1666.
(20) Hodson, B. E.; McGrath, T. D.; Stone, F. G. A. Inorg. Chem. 2004,
43, 3090.
(21) Jolly, P. W. In ComprehensiVe Organometallic Chemistry; Wilkinson,
G., Abel, E. W., Stone, F. G. A., Eds.; Pergamon Press: Oxford, U.K.,
1982; Vol. 6, Section 37.5. Chetcuti, M. J. In ComprehensiVe
Organometallic Chemistry II; Abel, E. W., Stone, F. G. A., Wilkinson,
G., Eds.; Pergamon Press: Oxford, U.K., 1995; Vol. 9, Chapter 3.
8136 Inorganic Chemistry, Vol. 44, No. 22, 2005

Synthesis of Nickel-Monocarbollide Complexes
Table 1. Analytical and Physical Data
analb
H
a
yield
(%)
IR νmax
(cm-1
compd
color
orange
orange
orange
brick red
yellow-orange
orange-red
orange
orange-red
orange-red
yellow
)
C
N
[N(PPh3)2][2,2-(CO)2-closo-2,1-NiCB10H11] (1a)
[N(PPh3)2][2,2-(CNBut)2-closo-2,1-NiCB10H11] (1b)
[N(PPh3)2][2,2-(CNXyl)2-closo-2,1-NiCB10H11] (1c)
[N(PPh3)2][2,2-(cod)-closo-2,1-NiCB10H11] (1d)
[N(PPh3)2][2-CO-2-PEt3-closo-2,1-NiCB10H11] (1e)
[N(PPh3)2][2-CO-2-PPh3-closo-2,1-NiCB10H11] (1f)
[N(PPh3)2][2-CO-2-CNXyl-closo-2,1-NiCB10H11] (1h)
[2,2,7-(CNBut)3-closo-2,1-NiCB10H10] (2a)
[2,2,7-(CNXyl)3-closo-2,1-NiCB10H10] (2b)
[2-CO-2-PEt3-7-OEt2-closo-2,1-NiCB10H10] (3a)
[2-CO-2,11-{µ-PPh2(C6H4-o)}-7-OEt2-closo-2,1-
NiCB10H9] (4)
68
40
82
51
64
79
36
11
42
36
50
2084 s, 2035 s
60.3 (60.2)c
63.2 (63.1)
62.8 (62.5)d
64.7 (64.6)
60.0 (60.4)
62.0 (62.0)d
61.0 (61.3)d
44.0 (43.9)
52.6 (52.2)d
35.0 (35.2)
49.6 (49.6)e
5.3 (5.5)
6.4 (6.6)
5.6 (5.7)
6.4 (6.4)
6.5 (6.5)
5.3 (5.3)
5.6 (5.5)
8.4 (8.5)
6.0 (5.9)
8.5 (8.6)
5.9 (5.8)
1.9 (1.8)
4.2 (4.7)
4.0 (3.9)
1.8 (1.7)
1.6 (1.6)
1.3 (1.3)
3.0 (3.0)
9.6 (9.6)
6.4 (6.3)
2163* m, 2132* m
2142* m, 2107* m
2004 s
2016 s
2147* m, 2034 m
2250* m, 2181* s, 2160* s
2225* m, 2159* s, 2134* s
2022 s
orange
2032 s
[2-CO-2-PPh3-7-{N(Me)dC(H)Me}-closo-2,1-
NiCB10H10] (5c, 5d)
red-orange
39
2032 s
48.7 (48.8)
5.7 (5.8)
2.4 (2.4)
a Measured in CH2Cl2; CO or *CN stretching frequencies; in addition, the spectra of all compounds show a broad, medium-intensity band at ca. 2500-
2550 cm-1 due to B-H absorptions. b Calculated values are given in parentheses. c Crystallizes with 0.25 equiv of C5H12. d Crystallizes with 1.0 equiv of
CH2Cl2. e Crystallizes with 0.5 equiv of CH2Cl2.
Table 2. 1H and 13C NMR Dataa
compd
1Hb (δ)
13Cc (δ)
1a
1b
7.69-7.46 (m, 30H, Ph), 2.34 (br s, 1H, cage CH)
7.69-7.46 (m, 30H, Ph), 2.01 (br s, 1H, cage CH),
1.43 (s, 18H, But)
194.2 (CO), 134.0-126.7 (Ph), 47.5 (br, cage C)
150.1 (CtN), 134.0-126.7 (Ph), 56.6 (CMe3),
42.4 (br, cage C), 30.5 (Me)
1c
1d
1e
7.69-7.46 (m, 30H, Ph), 7.12 (m, 2H, C6H3),
7.06 (m, 4H, C6H3), 2.43 (s, 6H, Me),
2.31 (br s, 1H, cage CH)
7.69-7.46 (m, 30H, Ph), 5.13 (br m, 4H, dCH),
2.75 (br s, 1H, cage CH), 2.58 (m, 4H, CH2),
2.23 (m, 4H, CH2)
7.69-7.46 (m, 30H, Ph), 2.04 (br s, 1H, cage CH),
1.82 (dq, J(HH) ) J(PH) ) 8, 6H, CH2),
1.13 (dt, J(PH) ) 16, 9H, CH3)
7.75-7.38 (m, 45H, Ph), 2.33 (br s, 1H, cage CH)
7.69-7.46 (m, 30H, Ph), 2.18 (br s, 1H, cage CH),
1.48 (s, 9H, But)
163.7 (CtN), 135.0, 127.9, 127.8 (C6H3),
134.0-126.7 (Ph), 43.8 (br, cage C), 18.8 (Me)
134.0-126.7 (Ph), 100.9 (dCH),
46.4 (br, cage C), 30.7 (CH2)
198.5 (d, J(PC) ) 21, CO), 134.0-126.7 (Ph),
45.1 (br, cage C), 17.8 (d, J(PC) ) 27, CH2), 8.0 (br, CH3)
1f
1g
197.3 (d, J(PC) ) 18, CO), 134.9-126.7 (Ph), 47.2 (br, cage C)
196.5 (CO), 145.9 (t, J(NC) ) 17, CtN), 134.0-126.7 (Ph),
66.2 (CMe3), 45.0 (br, cage C), 30.3 (Me)
1h
2a
7.69-7.46 (m, 30H, Ph), 7.16-7.07 (m, 3H, C6H3),
2.42 (s, 6H, Me), 2.35 (br s, 1H, cage CH)
2.46 (br s, 1H, cage CH), 1.58 (s, 9H, B-CNBut),
1.50 (s, 18H, Ni-CNBut)
160.3 (CtN), 135.4, 128.6, 128.1 (C6H3), 134.0-126.7 (Ph),
45.7 (br, cage C), 19.0 (Me)
144.1 (t, J(NC) ) 17, Ni-CN), 128.1 (br q, J(BC) ≈ 55, B-CN),
60.0 (B-CNCMe3), 57.8 (Ni-CNCMe3), 48.3 (br, cage C),
30.5 (Ni-CNCMe3), 29.8 (B-CNCMe3)
157.2 (Ni-CN), ca. 141.0 (br, B-CN), 136.7, 136.5, 131.4,
129.3, 128.8, 128.4 (C6H3), 50.6 (br, cage C),
18.9 (Ni-CNC6H3Me2), 18.4 (B-CNC6H3Me2)
195.6 (d, J(PC) ) 22, CO), 78.8 (OCH2), 43.5 (br, cage C),
18.1 (d, J(PC) ) 28, PCH2), 13.3 (OCH2CH3),
8.2 (br, PCH2CH3)
2b
3a
7.32-7.11 (m, 9H, C6H3), 2.78 (br s, 1H, cage CH),
2.42 (s, 12H, Ni-CNC6H3Me2), 2.37 (s, 6H, B-CNC6H3Me2)
4.62, 4.50 (br m × 2, 2H × 2, OCH2), 2.10 (br, 1H, cage CH),
1.90 (dq, J(PH) ) J(HH) ) 8, 6H, PCH2CH3),
1.46 (vt, J(HH) ) 7, 6H, OCH2CH3),
1.14 (dt, J(PH) ) 15, 9H, PCH2CH3)
7.87-7.23 (m, 14H, Ph and C6H4),
4
193.4 (d, J(PC) ) 21, CO),d 134.1-126.3 (Ph and C6H4),
77.7 (CH2), 42.7 (br, cage C), 13.1 (CH3)
4.41 (q × 2, J(HH) ) 7, 2H × 2, OCH2 × 2),
2.26 (br s, 1H, cage CH), 1.27 (t, J(HH) ) 7, 6H, CH3)
7.82 (br, 1H, dCH), 3.70 (d, 4J(HH) ) 1, 1H, NMe),
2.54 (d, 3J(HH) ) 6, 3H, dCMe),
5a
195.2 (d, J(PC) ≈ 20, CO), 173.8 (NdC), 53.7 (NMe),
47.2 (br, cage C), 20.8 (dCMe), 18.2 (d, J(PC) ) 29, PCH2),
8.3 (br, PCH2CH3)
2.33 (br s, 1H, cage CH), 1.89 (dq, 2J(PH) ) 9, 3J(HH) ) 8, PCH2),
1.12 (dt, 3J(PH) ) 16, PCH2CH3)
5b
5c
5d
2.63 (s, 3H, NCMe), 2.28 (br, 1H, cage CH),
1.91 (dq, 2J(PH) ) 9, 3J(HH) ) 8, PCH2),
1.15 (dt, 3J(PH) ) 16, PCH2CH3)
7.76 (br, 1H, dCH), 7.53-7.16 (m, 15H, Ph),
3.70 (br, 3H, NMe), ca. 2.52 (sh, 1H, cage CH),
2.39 (d, 3J(HH) ) 5, 3H, dCMe)
7.94 (br, 1H, dCH), 7.53-7.16 (m, 15H, Ph),
3.43 (br, 3H, NMe), ca. 2.52 (sh, 1H, cage CH),
2.48 (d, 3J(HH) ) 5, 3H, dCMe)
190.0 (br, CO), 110.2 (NtC), 47.3 (br, cage C),
18.3 (d, J(PC) ) 29, PCH2), 8.3 (d, J(PC) ) 3, PCH2CH3),
4.6 (NCCH3)
193.1 (d, J(PC) ) 21, CO), 173.6 (NdC), 53.5 (NMe),
48.0 (br, cage C), 20.8 (dCMe)
ca. 193.1 (d, J(PC) ≈ 20, CO), 174.3 (NdC), 54.1 (NMe),
48.0 (br, cage C), 20.9 (dCMe)
a Chemical shifts (δ) in ppm, coupling constants (J) in Hertz, measurements taken at ambient temperatures in CD2Cl2. b Resonances for terminal BH
protons occur as broad unresolved signals in the range δ ca. -1 to +3. c 1H-decoupled chemical shifts are positive to high frequency of SiMe4. d The signal
for B-C could not be located.
practical terms, this manifests itself as stronger bonding of
stable than the dicarbonylnickel-dicarbollide species with
these ligands to the nickel center, so that 1a is much more
respect to CO loss.
Inorganic Chemistry, Vol. 44, No. 22, 2005 8137

McGrath et al.
Table 3. 11B and 31P NMR Dataa
compd
11Bb (δ)
31Pc (δ)
1a
1b
1c
1d
1e
4.5, 0.0 (2B), -4.3, -9.0 (2B), -11.9 (2B) -17.1 (2B)
ca. -4.0 (sh) -4.5 (2B), -6.6, -14.5 (2B), -17.1 (2B), -19.9 (2B)
-2.0 (sh), -3.0 (2B), -6.0, -12.8 (2B), -15.7 (2B), -19.1 (2B)
1.5 (2B), -1.9, -5.4, -8.3 (2B), -14.5 (2B), -18.0 (2B)
-1.5 (sh), -2.3 (2B), -6.2, -11.9 (2B), -14.9 (2B), -19.0 (2B)
26.6
37.3
1f
-0.2 (sh), -1.4 (2B), -5.9, -10.9 (2B), -14.3 (2B), -18.6 (2B)
1g
1h
2a
2b
3a
4
5a
5b
5c/5dd
0.2, -2.2 (2B), -5.5, -11.6 (2B), -14.4 (2B), -18.5 (2B)
1.1, -1.6 (2B), -5.2, -11.0 (2B), -13.9 (2B), -18.2 (2B)
-2.4, -3.4, -7.7, -11.4, -15.2 (2B), -16.1 (2B), -19.1 (B-CN + 1B)
-0.2, -1.5, -7.1, -9.5, -13.7 (2B), -14.7 (2B), -18.2 (B-CN + 1B)
22.9 (B-OEt2), -3.4, -6.0, -8.4, -12.7, -13.8, -15.9, -17.8, -18.7, -23.2
23.1 (B-OEt2), 4.8 (B-C6H4), -6.4, -9.2 (2B), -11.9, -15.4, -17.3, -18.3, -23.9
9.6 (B-imine), -2.6, -5.5, -6.6, -11.4, -13.9 (2B), -15.2, -17.6, -20.1
ca. -1.1 (br, B-NCMe), -2.0, -4.2, -6.8, -11.1, -14.3 (2B), ca. -14.5 (sh), -18.2, -19.9
9.6 (B-imine), 6.9 (B-imine*), -1.5, -4.8, -6.5, -8.3, -9.4, -10.3, -13.3, -15.0, -16.7, -17.5, -20.0
27.7
60.3
30.2
29.8
37.7
a Measurements were taken at ambient temperature in CD2Cl2. b 1H-decoupled chemical shifts (δ) are positive to high frequency of BF3‚OEt2 (external).
Signals are of unit integral except where indicated; where peaks are ascribed to more than one boron nucleus, this may result from overlapping peaks and
does not necessarily indicate symmetry equivalence. c 1H-decoupled chemical shifts (δ) are positive to high frequency of 85% H3PO4 (external). d Occurs as
a mixture of isomers in the approximate ratio 2:1 (5c/5d - see text). Where discernible, peaks due to the isomer 5d are marked with an asterisk; no attempt
has been made to quantify peak integrals.
on the other by two essentially linear CO groups (Ni-C(11)
1.793(2) Å and Ni-C(12) 1.784(2) Å). The metal vertex is
approximately symmetrically bonded to the five coordinating
atoms of the carborane face, being closest to B(3) and B(6)
(Ni-B(3) 2.077(2) Å and Ni-B(6) 2.1128(19) Å) as a result
of a slight folding of the CB4 face along the B(3)‚‚‚B(6)
vector, which sends C(1), B(7), and B(11) away from the
metal (Ni-C(1) 2.2010(17) Å, Ni-B(7) 2.176(2) Å, and Ni-
B(11) 2.1365(19) Å). In addition, it is noted that the {Ni-
(CO)2} plane is essentially orthogonal to the notional mirror
plane of the carborane moiety (i.e., through C(1), B(9), and
B(12)). These findings are consistent with the results of
molecular orbital calculations that predict this conformation
to provide the most efficient overlap of the metal fragment
frontier orbitals with those of the nido-carborane ligand.23
During the synthesis of compound 1b, it was observed
that a second nickelacarborane was always isolated in low
Figure 1. Structure of the anion of compound 1a showing the crystal-
lographic labeling scheme. In this and subsequent figures, thermal ellipsoids
are drawn with 40% probability and, for clarity, only chemically significant
variable yields (ca. 3-5%) following chromatographic
workup. This species, [2,2,7-(CNBut)3-closo-2,1-NiCB10H10]
hydrogen atoms are shown. Selected distances (Å) and angles (deg) are as
(2a, see Chart 2 for the structure), was apparently formed
follows: Ni-C(1) 2.2010(17), Ni-B(3) 2.077(2), Ni-B(6) 2.1128(19), Ni-
on the chromatography column, as prior IR spectroscopic
analysis of the reaction mixture showed no evidence of it.
Formation of this complex is readily understood in terms of
protonation of 1b by the silica, thereby removing a boron-
bound hydride, and then attachment of a CNBut molecule
at the resulting vacant site. It was noted that compound 1b
was somewhat unstable on silica, and thus, any slight
decomposition would liberate isocyanide for formation of
2a. Attempts to prepare 2a rationally by deliberate proto-
nation of 1b in the presence of CNBut met with only modest
success and afforded only low yields of the desired product.
Intractable mixtures of other species, along with uncom-
plexed carboranes, were also obtained. It is well known that
metal-bound isocyanide ligands are themselves susceptible
to protonation,24 and presumably, such reactions complicate
matters here.
B(7) 2.176(2), Ni-B(11) 2.1365(19), Ni-C(11) 1.793(2), Ni-C(12)
1.784(2), C(11)-O(11) 1.132(2), C(12)-O(12) 1.133(2); C(12)-Ni-C(11)
99.64(9), O(11)-C(11)-Ni 175.9(2), O(12)-C(12)-Ni 178.1(2).
All four compounds 1a-d show characteristic peaks in
the appropriate ratios in their 1H and 13C{1H} NMR spectra
corresponding to their respective nickel-bound ligands, to
the carborane CH unit, and, of course, to the cation. Thus,
the cage CH moieties all resonate with small typical ranges,
δ 2.01-2.75 (1H) and δ 42.4-47.5 (13C).22 Their 11B{1H}
NMR spectra confirm the cluster to be mirror symmetric on
the NMR time scale, with each compound showing six peaks
in the ratio 1:2:1:2:2:2.
A single-crystal X-ray diffraction study on compound 1a
revealed the structure shown in Figure 1. As expected, the
anion consists of a nickel center that is bonded on one side
by the η5-CBBBB face of a {nido-7-CB10H11} ligand and
(23) Mingos, D. M. P.; Forsyth, M. I.; Welch, A. J. J. Chem. Soc., Dalton
Trans. 1978, 1363.
(22) Brew, S. A.; Stone, F. G. A. AdV. Organomet. Chem. 1993, 35, 135.
8138 Inorganic Chemistry, Vol. 44, No. 22, 2005

Synthesis of Nickel-Monocarbollide Complexes
Chart 2
Figure 2. Structure of compound 2b showing the crystallographic labeling
scheme. Selected distances (Å) and angles (deg) are as follows: Ni-C(1)
2.201(3), Ni-B(3) 2.086(3), Ni-B(6) 2.089(3), Ni-B(7) 2.123(3), Ni-
B(11) 2.137(3), Ni-C(10) 1.819(3), C(10)-N(10) 1.159(4), Ni-C(20)
1.830(3), C(20)-N(20) 1.168(3), B(7)-C(30) 1.532(4), C(30)-N(30)
1.139(3); C(10)-Ni-C(20) 95.20(12), C(30)-B(7)-Ni 113.6(2), N(10)-
C(10)-Ni 178.6(3), N(20)-C(20)-Ni 179.0(3), N(30)-C(30)-B(7)
178.7(3).
in the 11B{1H} NMR spectra (Table 3), with cluster asym-
metry shown by the presence of 10 separate signals (some
are coincident), of which one resonance in each case remains
a singlet in fully 1H-coupled 11B NMR spectra. These
singlets, at δ -19.1 (2a) and -18.2 (2b) correspond to the
BsCNR nucleus and are in positions typical for similarly
substituted compounds.10
Although the IR and NMR spectra for compounds 2a and
2b indicated the presence of a boron-bound isocyanide and,
specifically, the 11B NMR data indicated that this substitution
had occurred at a boron atom lying off the notional mirror
plane through the {closo-2,1-NiCB10} core, the exact site of
B-CNR attachment was only confirmed following an X-ray
diffraction study on compound 2b. This afforded the structure
shown in Figure 2, which clearly shows a {Ni(CNXyl)2}
moiety (Ni-C(10) 1.819(3) Å and Ni-C(20) 1.830(3) Å)
that is η5 coordinated by the open face of a nido-carborane
ligand. The carborane itself has been modified by substitution
at B(7). This vertex, in a â position with respect to the carbon
In contrast with 1b, compound 1c was never accompanied
by a tris(isocyanide) species during its synthesis and isolation.
This is reasonable, as 1c was clearly more stable on silica
than 1b, and the relatively poorly electron-donating CNXyl
groups render the anion of 1c much less nucleophilic.
However, protonation of 1c using H[BF4]‚OEt2 in the
presence of CNXyl did afford [2,2,7-(CNXyl)3-closo-2,1-
NiCB10H10] (2b) in reasonably good yields (42%). The best
yield of compound 2a (11%) was in fact obtained by
preparing a CH2Cl2 solution of 1b containing an excess of
the appropriate isocyanide and stirring this mixture with silica
overnight.
Each of the IR spectra of compounds 2a and 2b (Table 1)
shows a characteristic band for the boron-bound CtNR
group that is some 70 cm-1 higher in frequency than those
for the nickel-bound isocyanides. Likewise, 1H and 13C{1H}
NMR spectra (Table 2) confirm the presence of two different
isocyanide moieties in a 2:1 ratio in each of compounds 2a
and 2b. The boron-bound CtN groups appear as broad
resonances at δ 128.1 (2a) and 141.0 (2b) in their 13C{1H}
NMR spectra, deshielded by around 16 ppm as compared to
the peaks for the nickel-bound CtN units at δ 144.1 (2a)
and 157.2 (2b), respectively. Cage substitution is also evident
atom in the Ni-bound CBBBB face, bears the third CNXyl
ligand (B(7)-C(30) 1.532(4) Å). This two-electron donor
has replaced a formerly boron-bound hydride ligand, reduc-
ing the overall molecular anionic charge by one unit, so that
2b and, likewise, 2a are neutral and zwitterionic. This site
of substitution is typical of such systems,4 with this â-B-H
being the most hydridic and, therefore, the most susceptible
to abstraction by a proton. Dimensions within the two types
of CtN groups correlate well with the IR spectroscopic
parameters: the Ni-bound isocyanides have longer CtN
(24) Recent examples include: Seino, H.; Nonokawa, D.; Nakamura, G.;
Mizobe, Y.; Hidai, M. Organometallics 2000, 19, 2002. Pombeiro,
A. J. L. Inorg. Chem. Commun. 2001, 4, 585. Kang, D.-H.; Trenary,
M. J. Am. Chem. Soc. 2001, 123, 8432. Adams, C. J.; Anderson, K.
M.; Bartlett, I. M.; Connelly, N. G.; Orpen, A. G.; Paget, T. J.
Organometallics 2002, 21, 3454. Knorr, M.; Jourdain, I.; Lentz, D.;
Willemsen, S.; Strohmann, C. J. Organomet. Chem. 2003, 684, 216.
Hou, H.; Gantzel, P. K.; Kubiak, C. P. Organometallics 2003, 22,
2817.
bonds (C(10)-N(10) 1.159(4)
Å and C(20)-N(20)
1.168(3) Å) than the boron-bound isocyanides (C(30)-N(30)
1.139(3) Å), a feature readily rationalized by Ni f CtN(π*)
back-bonding, which is not available to the boron vertex.
Inorganic Chemistry, Vol. 44, No. 22, 2005 8139

McGrath et al.
1
As in 1a, the {Ni(CN)2} plane is approximately orthogonal
to the plane through C(1), B(9), and B(12), with some slight
twisting away from the ideal orientation, presumably to
relieve steric crowding between the three bulky isocyanide
ligands.
Thus, H and 13C{1H} NMR spectra showed that both 3a
and 4 incorporated an ether molecule: the 1H NMR spectra
contain multiplets at δ 4.50 and 4.62 (3a) and 4.41 (4) and
at δ 1.46 (3a) and 1.27 (4), in positions quite typical for
B-OCH2 and B-OCH2CH3 protons, respectively. Likewise,
the 13C{1H} spectra showed resonances for these moieties
at δ 78.8 (3a) and 77.7 (4) and at δ 13.3 (3a) and 13.1 (4),
respectively. Their 11B{1H} NMR data also confirmed
B-OEt2 substitution by the loss of cluster symmetry (10
inequivalent boron atoms, with one coincidence for 4) and
the appearance of a strongly deshielded resonance, at δ 22.9
(3a) and 23.1 (4), which remains a singlet upon retention of
proton coupling and which has a chemical shift characteristic
for a B-OEt2 unit in such systems.8,10,13,15 However, there
It was of interest to further investigate the protonation
reactions of compound 1a and its analogues beyond those
preliminary studies that gave rise to compounds 2a and 2b.
It has previously been observed that whereas most {closo-
MCB10H11} anions (M ) Mo,8,10 W,10 Fe,8,13 and Co15)
undergo cage boron hydride abstraction upon protonation,
for M ) Pt protonation typically occurs at the metal center,
forming Pt-H species.6,25 Although compound 1a is quite
stable, in contrast with the above reactions of 1b and 1c,
attempted protonation of compound 1a led to decomposition
via rapid CO evolution and formation of an intractable
mixture in which the only identifiable species was the [closo-
2-CB10H11]- anion, detected by 11B{1H} NMR spectroscopy.
It is thought that in any initially formed products, which
would presumably be neutral, the CO ligands are extremely
labile and, thus, rapidly lost.
Seeking derivatives of 1a that would be amenable to
further reactivity studies, the species [N(PPh3)2][2-CO-2-L-
closo-NiCB10H11] (L ) PEt3 (1e), PPh3 (1f), CNBut (1g),
and CNXyl (1h)) were prepared by treatment of 1a in CH2-
Cl2 with donor ligands L in the presence of Me3NO.
Characterizing data for compounds 1e-h are presented in
Tables 1-3. The significant lowering of the CO stretching
frequency for complexes 1e-h confirms their stronger Ni
f CO(π*) back-bonding upon introduction of the more
powerful donors L. Like compounds 1a-d, all four of the
species 1e-h show typical resonances in their 1H, 13C{1H},
and 31P{1H} NMR spectra corresponding to the various
groups present. In addition, they again show a 1:2:1:2:2:2
intensity ratio pattern in their 11B{1H} NMR spectra,
consistent with a time-averaged mirror symmetry due to free
rotation of the {Ni(CO)L} fragment with respect to the
is also a second singlet, at δ 4.8, in the fully 1H-coupled 11
B
NMR spectrum of 4, indicating the presence of a second
boron-bound substituent. Also notable is the 31P{1H} NMR
spectrum of 4, which shows a resonance at δ 60.3,
significantly shifted from that (δ 37.3) for the parent 1f. This
contrasts with the hardly different values of δ 26.6 and 27.7,
respectively, when compound 1e is converted to 3a. This
observation was suggestive of the possibility that compound
4 perhaps possesses a different structure than that which
would be expected.
1
As the H and 13C NMR data for 4 lacked any peaks
attributable to the second substituent, it was not readily
identified. Initially, it was thought that it might be a boron-
bound chloride moiety, as there is good precedent for a boron
vertex to acquire a halide substituent upon hydride abstraction
in halogenated solvents.26 However, the 11B chemical shift
for this boron atom is a little lower than would be expected
for a boron atom bearing such a substituent26 and, moreover,
the microanalytical data were not consistent with the presence
of a chloride.
Seeking to positively identify this boron-bound group, an
X-ray diffraction study was undertaken. The structure of 4
determined thereby is shown in Figure 3, in which the diethyl
ether group is clearly bonded to B(7) (B(7)-O(1) 1.552(3)
Å), as expected. An ortho-cycloboronation, involving one
of the phosphine phenyl rings, has also taken place adjacent
to the point of OEt2 attachment, and this is revealed as the
second boron vertex substituent. Thus, compound 4 is
formulated as [2-CO-2,11-{µ-PPh2(C6H4-o)}-7-OEt2-closo-
2,1-NiCB10H9]. The terminal H atom at B(11) has been
replaced (B(11)-C(22) 1.593(4) Å), and significantly, this
boron atom is the second of the two â-B-H vertexes with
ligating CBBBB face of the carborane cage. Perhaps surpris-
ingly, complex 1g was the least tractable of 1e-h, and
considerable difficulty was encountered in isolating it in pure
form. This was apparently due to its facile conversion to 1b
along with considerable decomposition. We note that a Pt
analogue of 1f was reported some years ago from the reaction
of [nido-7-CB10H11]3- with [PtCl2(CO)(PPh3)]. This plati-
nacarborane has spectroscopic properties very similar to those
of 1f.6
respect to the carbon atom in the nickel-bound CBBBB belt.
Although such cycloboronation of phenyl rings has
multiple precedents in boron cluster chemistry,27 these
processes very often occur during thermal processes, with
H2 elimination between cage B-H and phenyl C-H and
concomitant B-C bond formation. In the present case, the
reaction might be viewed as a “protonation-induced” elec-
trophilic aromatic substitution. If the first step of the reaction
is the initial protonation of the anion of 1f at a BH vertex,
In view of the observed formation of compounds 2a and
2b from 1b and 1c, it was expected that protonation of 1e
and 1f in the presence of Et2O would simply result in hydride
abstraction, leading to the B-OEt2 complexes [2-CO-2-PR3-
7-OEt2-closo-2,1-NiCB10H10] (R ) Et (3a) and Ph (3b)),
respectively.8,10,13,15 Perusal of the spectroscopic data for the
products (Tables 1-3) showed that 3a was indeed formed,
but it was immediately clear that 3b was not produced and
that another compound, complex 4, was instead produced.
(25) Blandford, I.; Jeffrey, J. C.; Redfearn, H.; Rees, L. H.; Rudd, M. D.;
(26) For example: Hata, M.; Kautz, J. A.; Lu, X. L.; McGrath, T. D.; Stone,
Stone, F. G. A. J. Chem. Soc., Dalton Trans. 1998, 1669.
F. G. A. Organometallics 2004, 23, 3590.
8140 Inorganic Chemistry, Vol. 44, No. 22, 2005

Synthesis of Nickel-Monocarbollide Complexes
yet undetected side products will shed further light on the
processes at play here.
In contrast with our attempts to append OEt2 substituents,
we have found in earlier studies that hydride abstraction
reactions similar to those above in the presence of MeCN
can behave somewhat differently. Although straightforward
B-NCMe substitution is known,15 a more common outcome
was B-imine formation.10,13,15 The latter typically occurred
when an anionic metallacarborane was treated with CF3SO3-
Me in CH2Cl2sNCMe: the Me+ cation initially and
preferentially methylates NCMe to form a nitrilium cation
[MeNtCMe]+, and this then acts as a hydride abstractor to
ultimately yield a B-{N(Me)dC(H)Me} product.13 In the
present system, treatment of 1e and 1f with CF3SO3Me in
CH2Cl2sNCMe (5:1) also afforded B-imine species.
However, in both cases, the reaction is complicated by
the presence of multiple products. From 1e is obtained the
expected imine compound [2-CO-2-PEt3-7-{N(Me)dC(H)-
Me}-closo-2,1-NiCB10H10] (5a) along with the simpler
acetonitrile-substituted complex [2-CO-2-PEt3-7-NCMe-
closo-2,1-NiCB10H10] (5b). In contrast, with 1f, a mixture
of the isomers [2-CO-2-PPh3-7-{(X)-N(Me)dC(H)Me}-
closo-2,1-NiCB10H10] (X ≡ E (5c) and Z (5d)) is formed.
All of these compounds were readily identified by their NMR
characteristics, save that it is impossible to determine the
stereochemistry of the imine moiety in 5a. Although the two
isomers 5c and 5d resisted chromatographic separation, the
1H and 13C NMR parameters could be differentiated and
assigned with the aid of 1H-1H and 13C{1H}-1H correlation
spectroscopy. The two isomers are also formed in the
approximate ratio 2:1, and this assisted in making NMR
assignments.
Figure 3. Structure of compound 4 showing the crystallographic labeling
scheme. Only the major component of the disordered ethyl group is shown,
and for clarity, all but the ipso carbons of the two C6H5 rings are omitted.
Selected distances (Å) and angles (deg) are as follows: Ni-C(1) 2.180(3),
Ni-B(3) 2.072(3), Ni-B(6) 2.123(2), Ni-B(7) 2.136(2), Ni-B(11)
2.116(2), Ni-P 2.1812(9), Ni-C(11) 1.774(2), C(11)-O(11) 1.143(2),
B(7)-O(1) 1.552(3), B(11)-C(22) 1.593(4); P-Ni-C(11) 96.01(8), O(1)-
B(7)-Ni 119.16(15), C(22)-B(11)-Ni 113.56(15), O(11)-C(11)-Ni
178.0(2), C(21)-P-Ni 106.45(8), C(31)-P-Ni 117.06(11), C(41)-P-
Ni 117.22(8), C(22)-C(21)-P 113.60(18).
eliminating H2 and forming a naked, “electrophilic” Bδ+
center, the latter might then be envisaged as attacking the
adjacent phenyl ring to ultimately form the B-C bond and
eliminate H+. Formation of the five-membered NiPCCB ring
would act as a driving force. The nickelacarborane so formed
retains an anionic charge and, hence, would be susceptible
to further boron hydride abstraction and, thence, substitution
by OEt2 to give 4. Such mechanistic details are, of course,
somewhat speculative, but in any event, this scenario requires
that it is more than likely that OEt2 attachment occurs after
the cycloboronation. Otherwise, the intermediate would be
neutral 3b, which would be rather more resistant to further
hydride abstraction. Moreover, it is notable that 3b is not
observed in the product mixture and that reducing the amount
of added acid only serves to reduce the yield of 4.
Conversely, the reaction might proceed via initial formation
of 3b, which then spontaneously eliminates H2: the latter
might be a simple consequence of C-H and B-H proximity,
but if so, it is surprising that it is not observed in other
systems described herein. Notwithstanding these observa-
tions, it is clearly a complex system and it may be that as
An X-ray diffraction study on a crystal of 5c determined
the structure shown in Figure 4. NMR analysis (1H) of the
same crystal confirmed it to be the major isomer in the 5c/d
mixture, and thus, the imine stereochemistry in the two
isomers was established. The two methyl groups of the {N-
(Me)dC(H)Me} moiety are clearly transoid (the torsion of
C(5)-N-C(3)-C(4) is -179.8(5)°), and thus, 5c contains
an (E)-imine. This two-electron donor substituent is bonded
to B(7), as seen in compounds 2b and 4, and the B(7)-N
distance is 1.578(9) Å, comparable to the same parameter
in other metallacarborane-imine species.10,13,26
All three compounds 5a, 5c, and 5d show a broad
resonance to relatively low field (δ 7.82 (5a), 7.76 (5c), and
1
7.94 (5d)) in their H NMR spectra; this is assigned to the
dCH proton of the imine group. The NMe protons of the
same group resonate at δ 3.70 (5a and 5c) and 3.43 (5d),
while those of the dCMe unit are seen at δ 2.54 (5a), 2.39
(5c), and 2.48 (5d). Several of the imine moiety resonances
show additional structure due to coupling with other protons
(27) Examples include: Crook, J. E.; Greenwood, N. N.; Kennedy, J. D.;
McDonald, W. S. Chem. Commun. 1983, 383. Kukina, G. A.;
Sergienko, V. S.; Porai-Koshits, M. A. Koord. Khim. 1985, 11, 400.
Elrington, M. J.; Greenwood, N. N.; Kennedy, J. D.; Thornton-Pett,
M. J. Chem. Soc., Dalton Trans. 1986, 2277. Ferguson, G.; Faridoon;
Spalding, T. R. Acta Crystallogr. 1988, C44, 1371. Bould, J.;
Greenwood, N. N.; Kennedy, J. D. J. Chem. Soc., Dalton Trans. 1990,
1451. Ferguson, G.; Jennings, M. C.; Lough, A. J.; Coughlan, S.;
Spalding, T. R.; Kennedy, J. D.; Fontaine, X. L. R. J. Chem. Soc.,
Chem. Commun. 1990, 891. Dou, J.-M.; Hu, C.-H.; Li, W.; Yao, H.-
J.; Jin, R.-S.; Zheng, P.-J. Polyhedron 1997, 16, 2323. Bould, J.; Clegg,
W.; Teat, S. J.; Barton, L.; Rath, N. P.; Thornton-Pett, M.; Kennedy,
J. D. Inorg. Chim. Acta 1999, 289, 95.
in the ligand, but broadening due to nearby quadrupolar 11
B
and 14N nuclei obscures much of this information. Unfor-
tunately, therefore, insufficient data is available in terms of
chemical shift and coupling constant values to definitively
assign the imine stereochemistry in 5a. In their 11B{1H}
NMR spectra, all three of these compounds display one peak
that is relatively deshielded and which remains a singlet upon
Inorganic Chemistry, Vol. 44, No. 22, 2005 8141

McGrath et al.
bound groups in place, there is scope for extensive further
functionalization of the cluster, a feature that is the subject
of much current interest.13,26
Experimental Section
General Considerations. All reactions were carried out under
an atmosphere of dry, oxygen-free dinitrogen using standard
Schlenk line techniques. Solvents were distilled from appropriate
drying agents under nitrogen prior to use. Petroleum ether refers
to that fraction of boiling point between 40 and 60 °C. Chroma-
tography columns (typically ca. 15 cm in length and ca. 2 cm in
diameter) were packed with silica gel (Acros, 60-200 mesh).
Preparative thin-layer chromatography (TLC) was performed on
20 cm × 20 cm glass plates (250 µm thickness silica gel GF254
,
Merck). Filtration through Celite typically employed a plug ca. 5
cm in length and ca. 2 cm in diameter. NMR spectra were recorded
at the following frequencies (MHz): 1H, 360.1; 13C, 90.6; 11B,
115.5; 31P, 145.8. The reagent [NMe4][closo-2-CB10H11]28 was
prepared according to the literature method; all other materials were
used as received.
Figure 4. Structure of compound 5c showing the crystallographic labeling
scheme. For clarity, only the major component of the disordered imine group
and only the ipso carbons of phenyl rings are shown. Selected distances
(Å) and angles (deg) are as follows: Ni-C(1) 2.196(3), Ni-B(3)
2.090(3), Ni-B(6) 2.101(3), Ni-B(7) 2.154(3), Ni-B(11) 2.125(3), Ni-P
2.2142(8), Ni-C(2) 1.757(3), C(2)-O(2) 1.149(3), B(7)-N 1.578(9),
N-C(3) 1.256(12), N-C(5) 1.486(9), C(3)-C(4) 1.477(7); N-B(7)-Ni
116.1(3), C(3)-N-C(5) 116.3(6), C(3)-N-B(7) 127.6(6), C(5)-N-B(7)
116.1(7), N-C(3)-C(4) 125.5(5).
Synthesis of [N(PPh3)2][2,2-L2-closo-2,1-NiCB10H11] (L ) CO,
CNXyl, and CNBut; L2 ) cod). (i) A THF (50 mL) solution of
[Ni(CO)4] was prepared in situ by bubbling CO through a
suspension of [Ni(cod)2] (0.500 g, 1.81 mmol) at -50 °C with
vigorous stirring. To this was added [NMe4][closo-2-CB10H11]
(0.370 g, 1.80 mmol), and stirring continued at the same temperature
for 16 h. (During this period, the reaction mixture appeared to be
exceptionally sensitive and often significant amounts of metallic
nickel could be deposited at this stage, drastically reducing the
ultimate yield.) After being warmed to ambient temperature, solid
[N(PPh3)2]Cl (1.050 g, 1.83 mmol) was added and the mixture
stirred for an additional 2 h. Removal of volatiles in vacuo gave a
dark residue that was extracted with CH2Cl2 (ca. 50 mL). The
extract was filtered (Celite), concentrated to ca. 10 mL, and applied
to the top of a chromatography column. Elution with CH2Cl2-
petroleum ether (4:1) removed an orange fraction, which was
collected and evaporated in vacuo to give orange microcrystals of
[N(PPh3)2][2,2-(CO)2-closo-2,1-NiCB10H11] (1a) (up to ca. 0.976
g). Further elution, using neat CH2Cl2, on occasion gave small
quantities of impure 1d, but this species was far more conveniently
prepared as described later.
(ii) Solid [Ni(cod)2] (0.670 g, 2.44 mmol) was suspended in THF
(30 mL), and the mixture was cooled to -80 °C. To this was added
CNBut (555 µL, 0.408 g, 4.91 mmol), followed by solid [NMe4]-
[closo-2-CB10H11] (0.500 g, 2.43 mmol). After being warmed to
room temperature, the mixture was stirred for 16 h. To the resulting
dark red-orange suspension was added a solution of [N(PPh3)2]Cl
(1.40 g, 2.44 mmol) in CH2Cl2 (5 mL), and stirring continued for
an additional 1 h. After evaporation in vacuo, the dark orange
residue was extracted with CH2Cl2 (50 mL), the extract was filtered
(Celite), and all volatiles were removed in vacuo. The resultant
mixture was redissolved in CH2Cl2 (5 mL) and subjected to column
chromatography. (Some darkening that indicated decomposition was
evident upon initial application to the silica.) Elution with CH2-
Cl2-petroleum ether (3:2) gave an orange band from which was
obtained [2,2,7-(CNBut)3-closo-2,1-NiCB10H10] (2a) (variable, ca.
0.005-0.010 g) as orange-red microcrystals, following removal of
the solvent in vacuo. Further elution with neat CH2Cl2 gave an
orange band from which was obtained crude [N(PPh3)2][2,2-
retention of proton coupling. These signals are at δ 9.6 (5a
and 5c) and 6.9 (5d); on this basis, it is tempting to assign
an E geometry to the imine in 5a, but as other factors are
also involved in determining the chemical shift, we are
reluctant to make such an assignment.
For compound 5b, the NMR spectra are essentially
straightforward. The protons of the NCMe ligand resonate
as a singlet at δ 2.63, with the carbon atoms of this unit
1
seen at δ 110.2 (CtN) and 4.6 (Me). The H and 13C{1H}
NMR spectra for this compound merit little further comment.
Similarly, the 11B{1H} NMR spectrum of 5b shows 10
signals of unit integral (two are coincident), consistent with
the absence of mirror symmetry. As with compounds 5a,
5c, and 5d, the signal at highest frequency in this spectrum
(δ -1.1) is a singlet in the fully coupled 11B spectrum and
is assigned to the B-NCMe nucleus.
Conclusion
We have prepared the anion [2,2-(CO)2-closo-2,1-
NiCB10H11]-, which represents for the nickel triad the parent
compound of half-sandwich, carbonyl-metal-monocarbol-
lide clusters. This anion is an analogue of the presumed [Ni-
(CO)2(η-C5H5)]+ cation, but it is notable as a very stable,
anionic Ni(II)-carbonyl complex. Similarly, the anion of 1d
is an example of a stable and anionic Ni(II)-olefin species.
Several related species and derivatives have been prepared,
and some further derivative chemistry has been explored.
The reactivity demonstrated for compounds 1a-h showed
that a range of substituents can be introduced onto the boron
vertexes of the metallacarborane cluster. With these boron-
(28) Hyatt, D. E.; Scholer, F. R.; Todd, L. J.; Warner, J. L. Inorg. Chem.
1967, 6, 2229.
8142 Inorganic Chemistry, Vol. 44, No. 22, 2005

Synthesis of Nickel-Monocarbollide Complexes
(CNBut)2-closo-2,1-NiCB10H11] (1b). The latter was purified by
further column chromatography, again eluting with CH2Cl2, to give
pure 1b (0.870 g) as an orange microcrystalline powder. Such a
purification step always also afforded a few more milligrams of
2a each time.
(iii) Similarly, from [Ni(cod)2] (1.00 g, 3.64 mmol), CNXyl
(0.955 g, 7.28 mmol), and [NMe4][closo-2-CB10H11] (0.750 g, 3.65
mmol) in THF (100 mL), stirred for 40 h, with subsequent addition
of [N(PPh3)2]Cl (2.15 g, 3.75 mmol) in CH2Cl2 (10 mL), was
obtained [N(PPh3)2][2,2-(CNXyl)2-closo-2,1-NiCB10H11] (1c) (2.94
g) as orange microcrystals. (No 2b was observed by this route,
and the product workup needed only a single chromatographic step,
eluting with neat CH2Cl2.)
(iv) The solids [Ni(cod)2] (0.500 g, 1.82 mmol) and [NMe4]-
[closo-2-CB10H11] (0.375 g, 1.82 mmol) were suspended in THF
(50 mL) at -80 °C, and the mixture was allowed to warm to room
temperature before it was stirred for 16 h. Solid [N(PPh3)2]Cl (1.2
g, 2.1 mmol) was added and stirring continued for an additional 1
h. Evaporation in vacuo gave a dark brick red oily residue that
was extracted with CH2Cl2 (50 mL), and the extract was filtered
(Celite) and concentrated by evaporation to ca. 5 mL. The latter
was transferred to a chromatography column, from which elution
with CH2Cl2-petroleum ether (4:1) gave a dark red band that
yielded brick red microcrystals of [N(PPh3)2][2,2-(cod)-closo-2,1-
NiCB10H11] (1d) (0.777 g) after removal of the solvent in vacuo.
All of compounds 1a-d isolated as described above were
sufficiently pure for further reactions. Analytically pure samples
were obtained by recrystallization from CH2Cl2-petroleum ether
at -30 °C.
good microanalytical results could not be obtained. Its identity is,
nevertheless, reasonably confirmed by its spectroscopic data (Tables
1-3).
(iv) Similar to the formation of 1f, compound 1a (0.200 g, 0.25
mmol) with CNXyl (0.034 g, 0.26 mmol) and Me3NO (0.020 g,
0.27 mmol) gave orange microcrystals of [N(PPh3)2][2-CO-2-
CNXyl-closo-2,1-NiCB10H11] (1h) (0.081 g) after column chroma-
tography using neat CH2Cl2. On occasion, a small quantity (ca.
10-15%) of 1c was also isolated from this reaction.
Synthesis of [2,2,7-L3-closo-2,1-NiCB10H10] (L ) CNBut and
CNXyl). (i) To a solution of compound 1b (0.150 g, 0.17 mmol)
in CH2Cl2 (15 mL) was added CNBut (30 µL, 0.022 g, 0.26 mmol)
and silica gel (ca. 2 g), and the mixture was stirred for 16 h. After
removal of the solvent in vacuo, the silica mixture was transferred
to the top of a chromatography column. Elution with CH2Cl2-
petroleum ether (3:2) afforded an orange band from which was
obtained [2,2,7-(CNBut)3-closo-2,1-NiCB10H10] (2a) (0.008 g) as
orange-red microcrystals. Compound 2a, as noted earlier, is often
also observed as a side product in the synthesis or reactions of 1b.
(ii) A solution of 1c (0.500 g, 0.50 mmol) and CNXyl (0.066 g,
0.50 mmol) in CH2Cl2 (50 mL) was treated with H[BF4]‚OEt2 (54
wt % in Et2O, 0.1 mL), and the mixture was stirred overnight. The
resulting dark orange solution was evaporated to dryness, the residue
was extracted with CH2Cl2-petroleum ether (1:1, 10 mL), and the
filtered extract was transferred to the top of a chromatography
column. Elution with the same solvent mixture gave a single orange-
red band that was collected and evaporated to dryness, yielding
orange-red crystals of [2,2,7-(CNXyl)3-closo-2,1-NiCB10H10] (2b)
(0.123 g).
Synthesis of [2-CO-2-PEt3-7-L-closo-2,1-NiCB10H10] (L )
OEt2, N(Me)dC(H)Me, and NCMe). (i) A solution of compound
1e (0.200 g, 0.23 mmol) in CH2Cl2-OEt2 (4:1, 25 mL) was cooled
to 0 °C, and CF3SO3CH3 (0.1 mL) was added. The resulting mixture
was stirred for 6 h at room temperature and then evaporated in
vacuo. The resulting dark yellow residue was extracted with CH2-
Cl2-petroleum ether (1:1, 5 mL), and the extract was applied to a
chromatography column. Elution with CH2Cl2-petroleum ether (7:
3) gave a pale yellow band that was collected and after removal of
the solvent gave pale yellow microcrystals of [2-CO-2-PEt3-7-OEt2-
closo-2,1-NiCB10H10] (3a) (0.034 g).
(ii) To a solution of 1e (0.090 g, 0.10 mmol) in CH2Cl2-NCMe
(5:1, 25 mL) was added CF3SO3CH3 (0.1 mL), and the mixture
was stirred for 0.5 h at ambient temperature. After the solvent was
removed from the reaction mixture, the resultant was applied to a
chromatographic column. Elution with CH2Cl2-petroleum ether (4:
1) gave a yellow-orange band that was shown (11B NMR) to contain
two species. Preparative TLC of this mixture using CH2Cl2-
petroleum ether (2:1) gave yellow and orange bands that were
collected and identified as [2-CO-2-PEt3-7-{N(Me)dC(H)Me}-
closo-2,1-NiCB10H10] (5a) (Rf 0.65, ca. 0.010 g, 25%) and [2-CO-
2-PEt3-7-NCMe-closo-2,1-NiCB10H10] (5b) (Rf 0.40, ca. 0.005 g,
13%), respectively.
Compounds 3a, 5a, and 5b were all somewhat unstable in the
solid state (especially the latter two) and even more so in solution.
This prevented us from obtaining good microanalytical data for 5a
and 5b.
Synthesis of [2-CO-2,11-{µ-PPh2(C6H4-o)}-7-OEt2-closo-2,1-
NiCB10H9]. To a solution of compound 1f (0.100 g, 0.10 mmol)
in CH2Cl2-OEt2 (5:1, 25 mL) was added CF3SO3CH3 (0.5 mL),
and the mixture was stirred for 12 h. After this time, removal of
the solvent in vacuo afforded a dark orange residue that was
transferred to the top of a chromatography column. Elution with
CH2Cl2-petroleum ether (2:1) gave a dark red-orange band that
Synthesis of [N(PPh3)2][2-CO-2-L-closo-2,1-NiCB10H11] (L )
PEt3, PPh3, CNBut, and CNXyl). (i) Compound 1a (0.500 g, 0.64
mmol) was dissolved in CH2Cl2 (50 mL), and PEt3 (125 µL, 0.10
g, 0.85 mmol) was added. The mixture was stirred for 16 h,
evaporated to dryness in vacuo, and then redissolved in CH2Cl2
(ca. 5 mL) and applied to a chromatography column. Elution with
CH2Cl2 gave a bright yellow-orange band from which was obtained
[N(PPh3)2][2-CO-2-PEt3-closo-2,1-NiCB10H11] (1e) (0.353 g) as a
yellow-orange microcrystalline solid by evaporation in vacuo.
(ii) Compound 1a (0.600 g, 0.76 mmol) was dissolved in CH2-
Cl2 (25 mL), and to this was added PPh3 (0.200 g, 0.76 mmol),
followed by Me3NO (0.060 g, 0.80 mmol). The mixture was stirred
overnight and then concentrated to ca. 5 mL by evaporation in
vacuo. The resulting solution was transferred to the top of a
chromatography column. Elution with CH2Cl2-petroleum ether (4:
1) removed a dark green band that was shown (11B NMR) to contain
[closo-2-CB10H11]- as the only carborane-containing species, and
so it was discarded. Further elution with CH2Cl2-THF (100:1)
removed a red fraction, from which removal of the solvent afforded
[N(PPh3)2][2-CO-2-PPh3-closo-2,1-NiCB10H11] (1f) (0.610 g) as
orange-red microcrystals.
(iii) To a cooled (-30 °C) solution of 1a (0.200 g, 0.25 mmol)
in CH2Cl2 (10 mL) was added CNBut (29 µL, 0.021 g, 0.25 mmol),
followed by Me3NO (0.020 g, 0.27 mmol). The mixture was
allowed to warm to room temperature and was stirred for 16 h.
After removal of all volatiles in vacuo, the crude product was
redissolved in CH2Cl2 (2 mL) and chromatographed. (As with 1b,
some decomposition was evident upon application to silica.) Elution
with CH2Cl2 gave two orange bands, of which the first was
identified (IR and 11B NMR) as 1b. From the second band was
obtained [N(PPh3)2][2-CO-2-CNBut-closo-2,1-NiCB10H11] (1g) (0.091
g, 43%) as an oily orange solid after removal of the solvent in
vacuo. IR (CH2Cl2): νmax/cm-1 2174 m (CN), 2031 m (CO).
Compound 1g proved to be rather unstable over time, and hence,
Inorganic Chemistry, Vol. 44, No. 22, 2005 8143

McGrath et al.
Table 4. Crystallographic Data for 1a, 2b·CH2Cl2, 4, and 5c·0.5CH2Cl2
1a
2b‚CH2Cl2
4a
5c‚0.5CH2Cl2
formula
fw
C39H41B10NNiO2P2
784.48
C29H39B10Cl2N3Ni
667.34
C24H33B10NiO2P
551.28
C23.5H33B10ClNNiOP
578.74
space group
a, Å
b, Å
P21/c
P21/c
C2/c
P1h
14.542(2)
18.249(4)
15.482(3)
90
101.796(8)
90
4021.9(13)
4
1.296
12.558(3)
18.357(4)
16.282(4)
90
111.944(4)
90
3481.5(14)
4
1.273
32.049(6)
11.7723(17)
18.731(4)
90
124.065(19)
90
5854.3(19)
8
1.251
9.7282(15)
12.0324(16)
13.360(3)
100.440(9)
97.678(18)
107.274(10)
1439.0(4)
2
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
F
calc, g cm-3
T, K
1.336
173(2)
110(2)
173(2)
110(2)
reflns measured
indept reflns
Rint
95236
13022
0.0620
0.598
21177
7962
0.0359
0.736
53532
8905
0.0603
0.739
14377
6514
0.0369
0.843
µ(Mo KR), cm-1
wR2 (all data), R1b
0.1092, 0.0421
0.1676, 0.0547
0.1299, 0.0481
0.1062, 0.0424
a Cocrystallizes with an unidentified solvate (see text). The formula and values for fw, Fcalc, and µ given here ignore the presence of this solvate. b wR2
) [∑{w(Fo - Fc )2}/∑w(Fo )2]1/2; R1 ) ∑||Fo| - |Fc||/∑|Fo| with Fo > 4σ(Fo).
2
2
2
Compound 2b cocrystallized with one molecule of CH2Cl2 as
solvate, located in general space. The molecule was disordered over
two distinct positions, and the two were fully refined as above with
complementary occupancies, with their respective ratios being ca.
59:41 at convergence. The C-Cl distances were restrained toward
sensible values (1.76(2) Å) using the DFIX card in SHELXL-97.30
The lattice of 4 also contained solvent molecules, but in this
case, they could not definitively be identified. It is thought that the
solvate consisted of a mixture of highly disordered fractional
pentane and CH2Cl2 molecules, lying close to a crystallographic
C2 axis (Wyckoff position e). In practice, these ultimately were
successfully modeled as seven fractional carbon atoms with refining
occupancies and anisotropic temperature factors; no attempt was
made to include hydrogen atoms. Additionally, one of the ethyl
groups of the cage-bound OEt2 molecule was disordered over two
positions. The two components were included with refining
complementary occupancies, in the approximate ratio of 67:33 at
convergence, and otherwise were treated as described above.
One-half molecule of CH2Cl2 cocrystallizes with one formula
unit of 5a, lying close to a crystallographic inversion center
(Wyckoff position d) such that there are two alternative positions
for the CH2 moiety. This solvate was otherwise ordered and treated
normally, as above. The imine group was also disordered over two
alternative locations, the two being approximately related by a
notional reflection in the plane through Ni, B(7), and B(9).
Positional and displacement parameters for all C, N, and H atoms
in the two components were refined as described above, and the
parts had refining complementary occupancies that converged at
the approximate ratio of 74:26.
was collected and purified by preparative TLC. Elution of the latter
using CH2Cl2-petroleum ether (5:2) gave an orange-red band (Rf
) 0.72) that was collected and identified as [2-CO-2,11-{µ-PPh2-
(C6H4-o)}-7-OEt2-closo-2,1-NiCB10H9] (4) (0.027 g).
Synthesis of [2-CO-2-PPh3-7-{N(Me)dC(H)Me}-closo-2,1-
NiCB10H10]. A solution of 1f (0.250 g, 0.25 mmol) in CH2Cl2-
NCMe (5:1, 20 mL) was treated with CF3SO3CH3 (0.25 mL), and
the mixture was stirred overnight. Volatiles were removed in vacuo,
and the resulting dark red residue was subjected to column
chromatography. Elution with CH2Cl2-petroleum ether (3:2) gave
an orange band from which removal of the solvent in vacuo gave
red-orange microcrystals of a mixture of [2-CO-2-PPh3-7-{(X)-
N(Me)dC(H)Me}-closo-2,1-NiCB10H10] (X ≡ E (5c), Z (5d))
(0.051 g total).
Structure Determinations. Experimental data for compounds
1a, 2b, 4, and 5a are recorded in Table 4. Diffracted intensities for
1 and 4 were collected at 110(2) K on a Bruker-Nonius X8 APEX
CCD area-detector diffractometer, while those for 2b and 5a were
acquired at 173(2) K on a Siemens SMART CCD area-detector
diffractometer. Both machines employed graphite-monochromated
Mo KR X-radiation (λ ) 0.71073 Å). In the former two cases,
several sets of narrow data “frames” were collected at different
values of θ, for various initial values of φ and ω, using 0.5°
increments of φ or ω; in the latter two, frame sets were collected
for 0.3° increments of ω at four settings of φ. Frames were
integrated using SAINT;29 the substantial redundancy in data
allowed an empirical absorption correction (SADABS)29 to be
applied, based on multiple measurements of equivalent reflections.
All structures were solved by direct methods and refined by full-
matrix least-squares analysis on all F2 data using SHELXTL
versions 5.03 and 6.12.30 All non-hydrogen atoms were successfully
refined with anisotropic thermal parameters. The locations of the
cage carbon atoms were verified by examination of their isotropic
thermal parameters, in conjunction with comparisons of B-B and
C-B internuclear distances. All hydrogen atoms were set riding
on their parent atoms with fixed isotropic thermal parameters
calculated as Uiso(H) ) 1.2Uiso(parent) or Uiso(H) ) 1.5Uiso(parent)
for methyl hydrogens.
Acknowledgment. We thank the Robert A. Welch
Foundation for support and Dr. John C. Jeffery, Bristol
University, U.K., for assistance with the structure determina-
tions of 2b and 5c. The Bruker X8 APEX diffractometer
was purchased with funds received from the National Science
Foundation Major Research Instrumentation Program (Grant
CHE-0321214).
Supporting Information Available: Full details of the crystal
structure analyses in CIF format. This material is available free of
charge via the Internet at http://pubs.acs.org.
(29) APEX 2, version 1.0; Bruker AXS: Madison, WI, 2003-2004.
(30) SHELXTL, versions 5.03 and 6.12; Bruker AXS: Madison, WI, 1995
and 2001.
IC051115C
8144 Inorganic Chemistry, Vol. 44, No. 22, 2005