The conformations of 13-vertex ML2C2B10 metallacarboranes: Experimental and computational studies

Article abstract of DOI:10.1021/ja067698m

The docosahedral metallacarboranes 4,4-(PMe₂Ph) ₂-4,1,6-closo-PtC₂B₁₀H₁₂, 4,4-(PMe₂Ph)₂-4,1,10-closo-PtC₂B ₁₀H₁₂, and [N(PPh₃)₂][4,4-cod-4,1, 10-closo-RhC₂B₁₀H₁₂] were prepared by reduction/ metalation of either 1,2-closo-C₂B₁₀H ₁₂ or 1,12-closo-C₂B₁₀H₁₂. All three species were fully characterized, with a particular point of interest of the latter being the conformation of the {ML₂} fragment relative to the carborane ligand face. Comparison with conformations previously established for six other ML₂C₂B₁₀ species of varying heteroatom patterns (4,1,2-MC₂B₁₀, 4,1,6-/WC2Bi0, 4,1,10-MC₂B₁₀, and 4,1,12-MC₂B₁₀) reveals clear preferences. In all cases a qualitative understanding of these was afforded by simple MO arguments applied to the model heteroarene complexes [(PH₃)₂PtC₂B₄H₆] ^2- and [(PH₃)₂PtCB₅H ₆]^3-. Moreover, DFT calculations on [(PH₃) ₂PtC₂B₄H₆]² in its various isomeric forms approximately reproduced the observed conformations in the 4,1,2-, 4,1,6-, and 4,1,10-MC₂B₁₀ species, although analogous calculations on [(PH₃)₂PtCB₅H ₆]^3- did not reproduce the conformation observed in the 4,1,12-MC₂B₁₀ metallacarborane. DFT calculations on (PH₃)₂PtC₂B₁₀H₁₂ yielded good agreement with experimental conformations in all four isomeric cases. Apparent discrepancies between observed and computed Pt-C distances were probed by further refinement of the 4,1,2- model to 1,2-(CH₂) ₃-4,4-(PMe₃)₂-4,1,2-closo-PtC₂B ₁₀H₁₀. This still has a more distorted structure than measured experimentally for 1,2-(CH₂)₃-4,4-(PMe ₂Ph)₂-4,1,2-closo-PtC₂B₁₀H ₁₀, but the structural differences lie on a very shallow potential energy surface. For the model compound a henicosahedral transition state was located 8.3 kcal mol^-1 above the ground-state structure, consistent with the fluxionality of 1,2-(CH₂)₃-4,4-(PMe ₂Ph)₂-4,1,2-closo-PtC₂B₁₀H ₁₀ in solution.

Full text of DOI:10.1021/ja067698m

Published on Web 02/22/2007
The Conformations of 13-Vertex ML2C2B10 Metallacarboranes:
Experimental and Computational Studies
Kelly J. Dalby,† David Ellis,† Stefan Erhardt,† Ruaraidh D. McIntosh,†
Stuart A. Macgregor,*,† Karen Rae,† Georgina M. Rosair,† Volker Settels,†
Alan J. Welch,*,† Bruce E. Hodson,‡ Thomas D. McGrath,*,‡ and
F. Gordon A. Stone‡
Contribution from the School of Engineering and Physical Sciences, Heriot-Watt UniVersity,
Edinburgh EH14 4AS, U.K., and Department of Chemistry and Biochemistry, Baylor UniVersity,
Waco, Texas 76798-7348
Received October 27, 2006; E-mail: a.j.welch@hw.ac.uk
Abstract: The docosahedral metallacarboranes 4,4-(PMe2Ph)2-4,1,6-closo-PtC2B10H12, 4,4-(PMe2Ph)2-
4,1,10-closo-PtC2B10H12, and [N(PPh3)2][4,4-cod-4,1,10-closo-RhC2B10H12] were prepared by reduction/
metalation of either 1,2-closo-C2B10H12 or 1,12-closo-C2B10H12. All three species were fully characterized,
with a particular point of interest of the latter being the conformation of the {ML2} fragment relative to the
carborane ligand face. Comparison with conformations previously established for six other ML2C2B10 species
of varying heteroatom patterns (4,1,2-MC2B10, 4,1,6-MC2B10, 4,1,10-MC2B10, and 4,1,12-MC2B10) reveals
clear preferences. In all cases a qualitative understanding of these was afforded by simple MO arguments
applied to the model heteroarene complexes [(PH3)2PtC2B4H6]2- and [(PH3)2PtCB5H6]3-. Moreover, DFT
calculations on [(PH3)2PtC2B4H6]2- in its various isomeric forms approximately reproduced the observed
conformations in the 4,1,2-, 4,1,6-, and 4,1,10-MC2B10 species, although analogous calculations on
[(PH3)2PtCB5H6]3- did not reproduce the conformation observed in the 4,1,12-MC2B10 metallacarborane.
DFT calculations on (PH3)2PtC2B10H12 yielded good agreement with experimental conformations in all four
isomeric cases. Apparent discrepancies between observed and computed Pt-C distances were probed
by further refinement of the 4,1,2- model to 1,2-(CH2)3-4,4-(PMe3)2-4,1,2-closo-PtC2B10H10. This still has a
more distorted structure than measured experimentally for 1,2-(CH2)3-4,4-(PMe2Ph)2-4,1,2-closo-PtC2B10H10,
but the structural differences lie on a very shallow potential energy surface. For the model compound a
henicosahedral transition state was located 8.3 kcal mol-1 above the ground-state structure, consistent
with the fluxionality of 1,2-(CH2)3-4,4-(PMe2Ph)2-4,1,2-closo-PtC2B10H10 in solution.
including the synthesis of supraicosahedral carboranes,7 are
laying the foundations for the eventual synthesis of larger and
larger heteroboranes to complement the theoretical work.
Introduction
There is significant current interest in supraicosahedral
heteroborane chemistry, an area of study which began more than
three decades ago with the synthesis1 and subsequent structural
study2 of the first such species, the 13-vertex cobaltacarborane
4-Cp-4,1,6-closo-CoC2B10H12. For many years large boron-
based polyhedra have been of interest to computational chemists3
and have represented a significant challenge to experimentalists.
Although well over a hundred 13-vertex metallacarboranes are
now known, there is still only a handful of 14-vertex species4
and only two 15-vertex species.5 However, recent advances,6
The usual structure adopted by 13-vertex metallacarboranes
is the docosahedron, I (Chart 1). This structure (of idealized
C2V symmetry) has two degree-six (with respect to the polyhe-
dron) vertices (vertices 4 and 5) and one degree-four vertex
(vertex 1); all the remaining vertices are of degree-five. Recently,
we have reported a variation of this basic shape.8 In the
(4) (a) Evans, W. J.; Hawthorne, M. F. J. Chem. Soc., Chem. Commun. 1974,
38. (b) Maxwell, W. M.; Bryan, R. F.; Grimes, R. N. J. Am. Chem. Soc.
1977, 99, 4008. (c) Maxwell, W. M.; Weiss, R.; Sinn, E.; Grimes, R. N. J.
Am. Chem. Soc. 1977, 99, 4016. (d) Pipal, J. R.; Grimes, R. N. Inorg.
Chem. 1978, 17, 6. (e) Ellis, D.; Lopez, M. E.; McIntosh, R.; Rosair, G.
M.; Welch, A. J. Chem. Commun. 2005, 1917.
(5) (a) McIntosh, R. D.; Ellis, D.; Rosair, G. M.; Welch, A. J. Angew. Chem.,
Int. Ed. 2006, 45, 4313. (b) Deng, L.; Zhang, J.; Chan, H.-S.; Xie, Z. Angew.
Chem., Int. Ed. 2006, 45, 4309.
† Heriot-Watt University.
‡ Baylor University.
(1) Dunks, G. B.; McKown, M. M.; Hawthorne, M. F. J. Am. Chem Soc. 1971,
93, 2541.
(2) Churchill, M. R.; DeBoer, B. G. J. Chem. Soc., Chem. Commun. 1972,
1326.
(3) (a) Brown, L. D.; Lipscomb, W. N. Inorg. Chem. 1977, 16, 2989. (b)
Bicerano, J.; Marynick, D. S.; Lipscomb, W. N. Inorg. Chem. 1978, 17,
2041. (c) Bicerano, J.; Marynick, D. S.; Lipscomb, W. N. Inorg. Chem.
1978, 17, 3443. (d) Lipscomb, W. N.; Massa, L. Inorg. Chem. 1992, 31,
2297. (e) Schleyer, P. v. R.; Najafian, K.; Mebel, A. M. Inorg. Chem. 1998,
37, 6765. (f) Balakrishnarajan, M. M.; Hoffmann, R.; Pancharatna, P. D.;
Jemmis, E. D. Inorg. Chem. 2003, 42, 4650. (g) Wang, Z. X.; Schleyer, P.
v. R. J. Am. Chem. Soc. 2003, 125, 10484.
(6) For example: (a) Ellis, D.; Lopez, M. E.; McIntosh, R.; Rosair, G. M.;
Welch, A. J.; Quenardelle, R. Chem. Commun. 2005, 1348. (b) Deng, L.;
Chan, H.-S.; Xie, Z. J. Am. Chem. Soc. 2006, 128, 5219.
(7) (a) Burke, A.; Ellis, D.; Giles, B. T.; Hodson, B. E.; Macgregor, S. A.;
Rosair, G. M.; Welch, A. J. Angew. Chem., Int. Ed. 2003, 42, 225. (b)
Deng, L.; Chan, H.-S.; Xie, Z. Angew. Chem., Int. Ed. 2005, 44, 2128.
(8) McIntosh, R.; Ellis, D.; Gil-Lostes, J.; Dalby, K. J.; Rosair, G. M.; Welch,
A. J. Dalton Trans. 2005, 1842.
9
3302
J. AM. CHEM. SOC. 2007, 129, 3302-3314
10.1021/ja067698m CCC: $37.00 © 2007 American Chemical Society

Conformations of 13-Vertex ML2C2B10 Metallacarboranes
A R T I C L E S
Chart 1
Chart 2
(the isomer 4,1,2-MC2B10); separated by a single B atom, IV
(the isomer 4,1,6-MC2B10); or separated by two B atoms, V
(the isomer 4,1,10-MC2B10) (see Chart 2). In the fourth type of
MC2B10 metallacarborane the metal is bound to a CB5 ligand
face, VI (the isomers 4,1,8- or 4,1,12-MC2B10).
Several years ago one of us was involved with studies into
the conformational preferences and slipping distortions of {ML2}
fragments (M ) group 10 metal; L ) 2e ligand, e.g., PR3) with
respect to five-atom carborane ligand faces (C2B3 faces with
both adjacent and nonadjacent C atoms, and the CB4 face) in
icosahedral metallacarboranes.14 In the present paper we de-
scribe, by a joint experimental and computational study, the
ways in which {ML2} fragments now bond to the four different
types of six-atom carborane faces (III-VI above) in supra-
icosahedral metallacarboranes, extending and complementing
the earlier work.
henicosahedron, II, the 2-5 connectivity of the docosahedron
is missing, and the structure (of idealized Cs symmetry) has a
trapezoidal 1295 face.
In all known MC2B10 13-vertex metallacarboranes the metal
atom occupies vertex 4, one of the positions of highest
connectivity, and one carbon atom occupies vertex 1, the lowest-
connected position, a simple consequence of the relative
electronegativities of metals, carbon, and boron. Thus, for
MC2B10 species isomers are possible, based on the location of
the second cage C atom. Since 13-vertex metallacarboranes are
usually prepared from reduction and subsequent metalation of
closo-C2B10 carboranes, kinetic isomers of MC2B10 configuration
arise from the use of different carborane isomers: (i) reduction
of 1,2-closo-C2B10 carborane causes the C atoms to spontane-
ously separate,9 affording [7,9-nido-C2B10]2-, and metalation
then affords 4,1,6-MC2B10 metallacarboranes;10 (ii) reduction
Experimental Section
11
of 1,7-closo-C2B10 carborane also yields [7,9-nido-C2B10]2-
,
Synthesis: General. Experiments were performed under dry,
oxygen-free N2 using standard Schlenk techniques, with some subse-
quent manipulation in the open laboratory. Solvents were freshly
distilled over CaH2 (CH2Cl2) or Na wire (THF, 40-60 petroleum ether)
or stored over 4 Å molecular sieves (CDCl3, CD2Cl2) and were degassed
(3 × freeze-pump-thaw cycles) before use. Preparative thin layer
chromatography (TLC) employed 20 cm × 20 cm Kieselgel 60 F254
glass plates. For compounds 3 and 6 IR spectra were recorded from
CH2Cl2 solutions using a Perkin-Elmer Spectrum RX FT spectropho-
hence the same 4,1,6-MC2B10 species; (iii) in contrast, reduction
of 1,12-closo-C2B10 carborane affords [7,10-nido-C2B10]2- and
thus 4,1,10-MC2B10 clusters upon metalation.6a Thermolysis of
these kinetic metallacarborane isomers causes rearrangement to
thermodynamically preferred ones; heating 4,1,6-MC2B10 results,
for some metals, in progressive conversion to 4,1,8- and then
4,1,12-MC2B10 isomers,12 and thermolysis of 4,1,10-MC2B10 has
been shown to produce the 4,1,12- isomer.6a A final known
isomer is 4,1,2-MC2B10. This form is afforded either by direct
insertion of a highly nucleophilic metal fragment into a 1,2-
closo-C2B10 cage13 or by reduction and then metalation of a
C,C-tethered 1,2-C2B10 species.8
1
tometer, and H NMR spectra were recorded at 200.1 MHz (Bruker
AC200 spectrometer) and 31P and 11B spectra at 162.0 and 128.4 MHz,
respectively (Bruker DPX400 spectrometer) from CDCl3 solutions at
room temperature (Heriot-Watt University). For compound 8 IR spectra
were recorded from CH2Cl2 solutions using a Bruker IFS25 FT
spectrophotometer, and NMR spectra were recorded from CD2Cl2
solutions using a Bruker AMX360 spectrometer at 360.1, 115.5, and
90.6 MHz for 1H, 11B, and 13C, respectively (Baylor University).
Elemental analyses were determined by the appropriate departmental
or commercial services. The starting materials 1,2-(CH2)3-1,2-closo-
Focusing only on the metal coordination sphere we therefore
now have access to four types of supraicosahedral MC2B10
metallacarboranes. In three of these the metal fragment is bonded
to C2B4 ligand faces with the cage C atoms either adjacent, III
17
C2B10H10,15 [codRhCl]2,16 and (PMe2Ph)2PtCl2 were prepared by
(9) (a) Getman, T. D.; Knobler, C. B.; Hawthorne, M. F. Inorg. Chem. 1990,
29, 158. (b) McKee, M. L.; Bu¨hl, M.; Schleyer, P. v. R. Inorg. Chem.
1993, 32, 1712. (c) Hermansson, K.; Wo´jcik, M.; Sjo¨berg, S. Inorg. Chem.
1999, 38, 6039.
literature methods or slight variations thereof. All other reagents and
solvents were supplied commercially and used as received.
4,4-(PMe2Ph)2-4,1,6-closo-PtC2B10H12. 1,2-closo-C2B10H12 (0.045
g, 0.31 mmol) and freshly cut sodium (0.015 g, 0.68 mmol) were stirred
in THF (∼20 mL) for 18 h. The resulting solution of Na2[C2B10H12]
was separated from excess sodium by cannula into a cooled (0 °C)
suspension of (PMe2Ph)2PtCl2 (0.15 g, 0.31 mmol) in THF (∼20 mL).
The reactants were allowed to warm to room temperature and stirred
for 18 h. Volatiles were removed in vacuo, leaving a yellow solid which
was subsequently dissolved in CH2Cl2 (20 mL), filtered, and concen-
trated. Preparative TLC eluting with CH2Cl2/40-60 petroleum ether
(1:1) afforded, as major product, a pale-yellow band (Rf ) 0.15)
subsequently shown to be 4,4-(PMe2Ph)2-4,1,6-closo-PtC2B10H12, com-
(10) Representative examples: (a) Hewes, J. D.; Knobler, C. B.; Hawthorne,
M. F. J. Chem. Soc., Chem. Commun. 1981, 206. (b) Khattar, R.; Knobler,
C. B.; Hawthorne, M. F. J. Am. Chem. Soc. 1990, 112, 4962. (c) Khattar,
R.; Knobler, C. B.; Hawthorne, M. F. Inorg. Chem. 1990, 29, 2191. (d)
Khattar, R.; Knobler, C. B.; Johnson, S. E.; Hawthorne, M. F. Inorg. Chem.
1991, 30, 1970. (e) Khattar, R.; Manning, M. J.; Knobler, C. B.; Johnson,
S. E.; Hawthorne, M. F. Inorg. Chem. 1992, 31, 268. (f) Carr, N.; Mullica,
D. F.; Sappenfield, E. L.; Stone, F. G. A.; Went, M. J. Organometallics
1993, 12, 4350. (g) Li, S.; Mullica, D. F.; Sappenfield, E. L.; Stone, F. G.
A. J. Organometal. Chem. 1994, 467, 95. (h) Mullica, D. F.; Sappenfield,
E. L.; Stone, F. G. A.; Woollam, S. F. Can. J. Chem. 1995, 73, 909. (i)
Chui, K.; Yang, Q.; Mak, T. C. W.; Xie, Z. Organometallics 2000, 19,
1391. (j) Wilson, N. M. M.; Ellis, D.; Boyd, A. S. F.; Giles, B. T.;
Macgregor, S. A.; Rosair, G. M.; Welch, A. J. Chem. Commun. 2002, 464.
(k) Burke, A.; Ellis, D.; Ferrer, D.; Ormsby, D. L.; Rosair, G. M.; Welch,
A. J. Dalton Trans. 2005, 1716.
(11) Dunks, G. B.; Wiersema, R. J.; Hawthorne, M. F. J. Am. Chem. Soc. 1973,
95, 3174.
(14) (a) Mingos, D. M. P.; Forsyth, M. I.; Welch, A. J. J. Chem Soc., Chem.
Commun. 1977, 605. (b) Mingos, D. M. P.; Forsyth, M. I.; Welch, A. J. J.
Chem. Soc., Dalton Trans. 1978, 1363.
(15) Paxon, T. E.; Kaloustian, M. K.; Tom, G. M.; Wiersema, R. J.; Hawthorne,
M. F. J. Am. Chem. Soc. 1972, 94, 4882.
(16) Chatt, J.; Venanzi, L. M. J. Chem. Soc. 1957, 4735.
(17) Jenkins, J. M.; Shaw, B. L. J. Chem. Soc. (A) 1966, 770.
(12) (a) Dustin, D. F.; Dunks, G. B.; Hawthorne, M. F. J. Am. Chem. Soc. 1973,
95, 1109. (b) Burke, A.; McIntosh, R.; Ellis, D.; Rosair, G. M.; Welch, A.
J. Collect. Czech. Chem. Commun. 2002, 67, 991.
(13) Barker, G. K.; Garcia, M. P.; Green, M.; Stone, F. G. A.; Welch, A. J. J.
Chem. Soc., Chem. Commun. 1983, 137.
9
J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007 3303

A R T I C L E S
Dalby et al.
Table 1. Crystallographic Data
pound 3, (0.38 g, 20%). C18H34B10P2Pt requires C 35.1, H 5.58.
Found: C 35.2, H 5.60. IR, νmax at 2500 cm-1 (B-H). 11B{1H} NMR:
δ -0.98 (2B), -2.60 (1B), -4.54 (3B), -9.46 (1B), -13.85 (2B),
and -15.00 δ(11B) ) -11.62. 1H NMR, δ 7.45-7.25 (m, 10H, C6H5),
3.95 (br s, 2H, CcageH), and 1.7-1.3 (br m, 12H, P(CH3)2). 31P{1H}
NMR: δ -17.61 (s, JPtP 3894 Hz) and -20.82 (s, JPtP 2800 Hz).
4,4-(PMe2Ph)2-4,1,10-closo-PtC2B10H12. 1,12-closo-C2B10H12 (0.10
g, 0.69 mmol) was cooled to -80 °C. Ammonia (∼20 mL) and freshly
cut sodium (0.13 g, 5.6 mmol) were added, and the resulting dark-
blue solution was stirred for 2 h. Following removal of the ammonia
in vacuo the residue was extracted into THF (20 mL) and transferred,
by cannula, to a frozen solution of (PMe2Ph)2PtCl2 (0.37 g, 0.69 mmol)
in THF (20 mL). The reaction mix was allowed to warm to room
temperature and stirred for 30 min (warning - prolonged stirring
leads to significant decomposition!). Volatiles were removed in vacuo,
and the yellow residue was extracted into CH2Cl2 (20 mL), filtered,
and then purified by preparative TLC eluting with CH2Cl2/40-60
petroleum ether (7:3) to afford, as a narrow yellow band (Rf ) 0.29),
the compound 4,4-(PMe2Ph)2-4,1,10-closo-PtC2B10H12, compound 6,
(0.02 g, 10%). IR, νmax at 2529 cm-1 (B-H). 11B{1H} NMR: δ 15.91
(2B), 11.04 (1B), 0.17 (1B), -8.38 (2B), -16.30 (2B) and -22.83
(2B), δ(11B) ) 0.36. 1H NMR: δ 7.6-7.1 (m, 10H, C6H5), 4.3 (br s,
1H, CcageH), 2.7 (br s, 1H, CcageH) and 1.9-1.2 (br m, 12H, P(CH3)2).
31P{1H} NMR: δ -13.16 (s, JPtP 3013 Hz).
3
6
8‚
1/2CH2Cl2
formula
Mr
system
space group
a/Å
C18H34B10P2Pt C18H34B10P2Pt C46.5H55B10ClNP2Rh
615.58
monoclinic
P21/c
13.6694(9)
10.5522(7)
17.3895(12)
90
100.028(3)
90
2470.0(3)
4
1.655
5.816
1200
33.22
34349
9430
0.0229
615.58
monoclinic
C2/c
20.335(3)
10.4382(14)
13.5990(19)
90
122.496(3)
90
2434.6(6)
4
936.22
triclinic
Ph
10.056(5)
14.692(7)
17.296(13)
113.67(4)
100.35(4)
98.26(3)
2235(2)
2
b/Å
c/Å
R/deg
â/deg
γ/deg
V/Å3
Z
Dc/Mg m-3
µ/mm-1
F(000)
θmax
1.679
5.901
1200
36.29
63206
5859
1.391
0.550
966
27.64
data measured
unique data, n
Rint
45526
10159
0.0664
0.0464
max, min. tfs
0.498, 0.315 0.702, 0.566 0.921, 0.803
R1a
0.0182
0.0397
1.184
320
0.0378
0.0523
0.923
169
0.0601
0.1199
0.998
596
wR2a
S a
variables, p
∆Fmax, ∆Fmin/e Å-3 2.087, -1.294 3.261, -1.891 1.801, -1.745
[N(PPh3)2][4,4-cod-4,1,10-closo-RhC2B10H12]. Similarly, from
1,12-closo-C2B10H10 (0.20 g, 1.39 mmol) and [codRhCl]2 (0.34 g, 0.69
mmol), followed by metathesis with [N(PPh3)2]Cl (0.80 g, 1.39 mmol)
was isolated yellow [N(PPh3)2][4,4-cod-4,1,10-closo-RhC2B10H12], salt
8, (0.50 g, 40%) as the major mobile product following column
chromatography (eluting with CH2Cl2/40-60 petroleum ether, 7:3).
C46H54B10NP2Rh requires C 61.8, H 6.1, N 1.6. Found: C 62.0, H 6.2,
N 1.7. IR: νmax at 2531 cm-1 (B-H). 11B{1H} NMR (CD2Cl2) δ -4.5
(2B), -6.0 (2B), -12.3 (1B), -13.5 (1B), -16.2 (2B) and -17.2 (2B),
δ(11B) ) -11.4. 1H NMR (CD2Cl2): δ 7.8-7.5 (m, 30H, C6H5), 5.06
(vbr, 1H, CcageH), 4.35 (br, 4H, dCH), 2.22 (br, 8H, CH2), and 2.05
(vbr, 1H, CcageH). 13C{1H} NMR (CD2Cl2): 133.8-126.5 (C6H5), 79.0
(br, Ccage), 77.1 [d, J(RhC) 10 Hz, dCH], 42.4 (br, Ccage) and 32.2
(CH2).
2
2
2
a R1 ) ∑(|Fo| - |Fc|)/∑|Fo|, wR2 ) [∑[w(Fo - Fc )2]/∑w(Fo )2]1/2, S
2
2
) [∑[w(Fo - Fc )2/(n - p)]1/2, where n is the number of data and p the
number of parameters. R1 and wR2 quoted for all data.
riding positions, with C-H 0.95 (phenyl), 0.99 (CH2) or 1.00 (cod-
CH) Å, and UH set to 1.2 × UeqC. The CH2Cl2 of solvation of 8 is
disordered about an inversion center and was refined with C-Cl
restrained to 1.76(2) Å. Table 1 contains further experimental details.
Calculations. All geometries were optimized at the B3LYP20 level
with the basis set for all atoms of double-ú quality. For carbon,
hydrogen, and boron Pople’s 6-31G(d,p) basis set was used, while for
platinum and phosphorus the Stuttgart relativistic ECP21 was employed
with an additional d-polarization function on phosphorus.22 Frequency
calculations were performed to determine the nature of the optimized
geometries. The Gaussian 03 program23 was used for all calculations.
Optimized geometries are displayed via Mercury.24
Crystallography. Intensity data were collected on single crystals
of 3, 6, and 8‚1/2CH2Cl2 on a Bruker X8 APEX2 diffractometer,18 with
crystals mounted in inert oil on a glass fiber and cooled to 100 K (3
and 6, Heriot-Watt University) or 110 K (8, Baylor University) by an
Oxford Cryosystems Cryostream. The structures were solved by direct
methods and refined by full-matrix least-squares.19 All non-H atoms
were refined with anisotropic displacement parameters. In compound
6 the cage is disordered about a crystallographic C2 axis, and the
disorder was modeled by two intersecting half-occupancy pentagonal
pyramids for the lower cage belt (B5B9B12B13B8B11), co-incident
at the (crystallographically equivalent) B8 and B9 positions. In spite
of this disorder C1 (≡ C10) was unambiguously identified as a carbon
atom. For 3 and 8 cage-bound H atoms were located in difference
Fourier maps and freely refined, but with thermal parameters set to
1.2 × Ueq of the attached B or C atom. In the case of 6, some cage BH
atoms were located and positionally refined, but restrained to B-H
1.10(2) Å, while others were set in calculated positions; the position
of the single CH was also calculated, with C-H 1.10 Å; again, thermal
parameters for all cage H atoms were set to 1.2 × Ueq of the attached
B or C atom. Note that, although B11 is disordered over two positions,
its attached H atom sits on the crystallographic C2 axis. In 3 and 6,
methyl H atoms (C-H 0.98 Å) and phenyl H atoms (C-H 0.95 Å)
were placed in calculated but riding positions, with UH set to 1.5 or
1.2 × UeqC, respectively. For 8, non-cage H atoms were set in calculated,
Results and Discussion
This paper is concerned with the conformations of nine 13-
vertex metallacarboranes, compounds 1-9. Six of these com-
pounds (1, 2, 4, 5, 7, and 9) have already been described in the
literature (although their conformations were not discussed), and
three (3, 6 and 8) are new. The compounds are summarized in
Table 2 and sketched in Chart 3.
1. Synthesis and Characterization. 4,1,2-MC2B10 Com-
pounds. The synthesis, spectroscopic characterization, and
structural studies of the platina- and nickelacarboranes 1,2-
(CH2)3-4,4-(PMe2Ph)2-4,1,2-closo-PtC2B10H10, 1, and 1,2-(CH2)3-
4,4-dppe-4,1,2-closo-NiC2B10H10, 2, (dppe ) Ph2PCH2CH2PPh2)
(20) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 1372. (b) Becke, A. D. J. Chem.
Phys. 1993, 98, 5648. (c) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B
1988, 37, 785. (d) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
(21) (a) Bergner, A.; Dolg, M.; Kuechle, H.; Stoll, H.; Preuss, H. Mol. Phys.
1993, 80, 1431. (b) Kaupp, M.; Schleyer, P. v. R.; Stoll, H.; Preuss, H. J.
Chem. Phys. 1991, 94, 1360. (c) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer, R.
M. J. Phys. Chem. 1993, 97, 5852.
(22) Ho¨llwarth, A.; Bo¨hme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi, A.; Jonas,
V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking, G. Chem. Phys.
Lett. 1993, 208, 237.
(23) Frisch, M. J., et al. Gaussian 03, reVision C.02; Gaussian Inc., Wallingford,
(18) Bruker AXS APEX2, version 1.0-8; Bruker AXS Inc.: Madison, WI, U.S.A.,
2003.
(19) Sheldrick, G. M. SHELXTL, versions 5.1 and 6.12; Bruker AXS Inc.:
Madison, WI, U.S.A., 1999 and 2001, respectively.
CT, U.S.A., 2004.
(24) Mercury, version 1.4.1; Cambridge Crystallographic Data Center: Cam-
bridge, UK, 2006.
9
3304 J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007

Conformations of 13-Vertex ML2C2B10 Metallacarboranes
A R T I C L E S
Table 2. Summary of Compounds Described
4,1,2-[M]C2B10
4,1,6-[M]C2B10
4,1,10-[M]C2B10 4,1,12-[M]C2B10
(ligand type III) (ligand type IV)
(ligand type V)
(ligand type VI)
[M] ) P2Pta
1
2
3
4
5
6
7
8
[M] ) dppeNi
9
[M]- ) codRh
a P ) PMe2Ph
Chart 3
Figure 1. Experimental conformations of the {MP2} fragments in (left) 1
(θ ) 61.1°) and (right) 2 (θ ) 64.6°).
Figure 2. Perspective view of compound 3. Thermal ellipsoids are drawn
at the 50% probability level, except for H atoms which have an artificial
radius for clarity.
dacarborane salt [N(PPh3)2][4,4-cod-4,1,6-closo-RhC2B10H12],
5.26 Sodium reduction of 1,2-closo-1,2-C2B10H12 followed by
treatment with (PMe2Ph)2PtCl2 and workup involving TLC
affords the yellow compound 3 in reasonable yield. The
11B{1H} NMR spectrum of 3 reveals six resonances between δ
0 and -15, with relative integrals 2:1:3[2+1 co-incidence]:1:
2:1 (from high frequency to low frequency) and with a weighted
average 11B chemical shift, δ(11B) , of -11.6. In the 1H NMR
spectrum is a single resonance for cage CH atoms, δ 3.95,
whereas the 31P{1H} NMR spectrum reveals two singlets (with
attendant 195Pt satellites) at δ -17.6 and -20.8. All these data
are consistent with a (time-averaged) structure in solution of
Cs symmetry, with the PtP2 plane lying on the molecular mirror
plane.
Compound 3 was further characterized by a crystallographic
study. Figure 2 contains a perspective view of the compound,
and Table 3 lists selected molecular parameters. The cage of 3,
like that of 4 and 5, is a docosahedron, but is considerably
distorted by, in particular, a very long Pt4‚‚‚C6 distance,
2.813(13) Å. Note that the Pt4-C6 vector, in projection (see
Figure 3, left) lies effectively perpendicular to the PtP2 plane.
B-B distances within 3 follow the trends already established10k,25
for 4,1,6-MC2B10 polyhedra, with connectivities involving the
degree-six B atom B5 being relatively long, particularly so if
were described recently.8 The cages in both these metallacar-
boranes are docosahedral with formal connectivities between
C2 and B5, although that for 1 is relatively long, 2.055(12) Å,
consistent with a distortion toward a henicosahedral cage.
Figure 1 shows views of the central ML2C2B4 portions of 1
and 2, illustrating the conformations of the {ML2} fragments
relative to the C2B4 ligand face. A convenient (arbitrary)
quantitative measure of these conformations is θ, the dihedral
angle between ML2 and MC1B10B11 (compounds 1-5, 9) or
MC1C10B11 (compounds 6-8) planes, where 0 < θ < 90°
represents, in projection, one ligand L lying in the B3 quadrant,
and 90 < θ < 180° represents L in the B7 quadrant. For
compounds 1 and 2 θ is 61.1° and 64.6°, respectively.
All attempts to prepare [1,2-(CH2)3-4,4-cod-4,1,2-closo-
RhC2B10H10]-, the rhodacarborane anion analogue of 1 and 2,
were unsuccessful.
4,1,6-MC2B10 Compounds. We have prepared the platinum
complex 4,4-(PMe2Ph)2-4,1,6-closo-PtC2B10H12, 3, to comple-
ment its previously described analogues, namely the nickela-
carborane 4,4-dppe-4,1,6-closo-NiC2B10H12, 4,25 and the rho-
(25) Laguna, M. A.; Ellis, D.; Rosair, G. M.; Welch, A. J. Inorg. Chim. Acta
2003, 347, 161.
(26) Hodson, B. E.; McGrath, T. D.; Stone, F. G. A. Organometallics 2005,
24, 1638.
9
J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007 3305

A R T I C L E S
Dalby et al.
Table 3. Selected Molecular Parameters (Å, deg) in
4,4-(PMe2Ph)2-4,1,6-closo-PtC2B10H12, 3
C1-B2 1.555(2)
C1-B3 1.573(2)
Pt4‚‚‚C6 2.8131(13) B8-B11
Pt4-B7 2.2631(14) B8-B13
1.766(2)
1.758(2)
1.730(3)
1.750(2)
1.827(2)
1.816(2)
1.785(2)
1.802(2)
1.768(3)
2.2710(3)
2.3129(4)
C1-Pt4 2.1652(12) Pt4-B10 2.3692(14) B9-B11
C1-B5 1.742(2)
B2-Pt4 2.4438(15) B5-B9
B5-B8
1.867(2)
1.957(3)
B9-B12
B10-B12
B10-B13
B11-B12
B11-B13
B12-B13
Pt4-P1
B2-B5 2.259(3)
B2-C6 1.623(2)
B2-B9 1.932(2)
B5-B11 1.804(2)
C6-B9 1.662(2)
C6-B10 1.693(2)
B3-Pt4 2.3961(14) C6-B12 1.647(2)
B3-B5 1.884(2)
B3-B7 1.856(2)
B3-B8 1.752(2)
B7-B8
1.783(2)
B7-B10 1.862(2)
Pt4-P2
B7-B13 1.764(2)
P1-Pt4-P2 92.363(14)
the other B atom is also connected to the degree-six Pt atom
and one or more cage C atoms. Thus, B2-B5, 2.259(3) Å, is
the longest, followed by B9-B5, 1.957(3) Å, B3-B5,
1.884(2) Å, and B8-B5, 1.867(2) Å. In 3 the conformation of
the PtP2 plane is defined by θ ) 148.7°.
Figure 4. Perspective view of compound 6 showing, for clarity, only one
of the disordered pentagonal pyramids used to model the lower part of the
molecule. Thermal ellipsoids as for Figure 2. Note that the polyhedral
numbering in this Figure is conventional and not necessarily that in the
corresponding CIF.
communication.6a Similar treatment of the [7,10-nido-C2B10H12]2-
anion with (PMe2Ph)2PtCl2 or with [codRhCl]2 (the latter
followed by cation metathesis) yields the yellow platina- and
rhodacarborane species 4,4-(PMe2Ph)2-4,1,10-closo-PtC2B10H12,
6, and [N(PPh3)2][4,4-cod-4,1,10-closo-RhC2B10H12], 8, fol-
lowing chromatographic workup. The isolated yield of 6 is
relatively poor, but the compound is comparatively unstable.
The yield of 8 is reasonable, and certainly 8 is the major reaction
product.
The Cs symmetry of the cage of 3 implied by the NMR data
can be readily understood in terms of the double diamond-
square-diamond fluxional process first suggested by Hawthorne
for the CpCo system12a and recently confirmed computationally
by us for the Sn10j and (η-C6H6)Ru10k systems. This process
effectively generates a mirror plane through atoms Pt4, B2, B11,
and B7 (the integral-1 B atoms), which relates pairwise B3 and
B10, B5 and B9, and B8 and B12 (the integral-2 B atoms) and
renders equivalent the two cage C atoms. To maintain the
inequivalence of the P atoms all that is required is a synchronous
libration of the PtP2 unit about the Pt4B2B7 plane (see Figure
3). Full rotation of the {PtP2} fragment, which would render
the P atoms equivalent, presumably occurs at higher tempera-
tures, but this was not explored as attempts to isomerize 3 by
thermolysis (see later) led only to decomposition.
Compound 4 crystallizes (as its 1:1 CH2Cl2 solvate) in two
crystalline modifications, 4R and 4â, but the cages of the two
forms are practically superimposable. The angle θ, defined as
above, is 134.6° for 4R and 136.7° for 4â. Figure 3 (left and
center) shows views of the MP2C2B4 cores of 3 and 4R,
comparing the orientation of the MP2 planes relative to the
carborane ligand faces. To describe the conformation of the
{codRh} fragment in the anion of 5 we define X1 and X2 as
the midpoints of the cod CdC bonds. Figure 3 (right) is a view
of the RhX2C2B4 central part of the anion of 5, and θ, the
dihedral angle between the RhX2 and RhC1B10B11 planes, is
115.1°.
NMR spectra of 6 are fully consistent with a Cs symmetric
compound. In the 11B{1H} spectrum are six resonances, 2:1:1:
2:2:2 (high to low frequency), between δ +16 and -23, with
δ(11B) ) 0.4, and the 31P{1H} spectrum yields a singlet (with
attendant 195Pt satellites) at δ -13.6. There are two cage CH
1
resonances in the H spectrum, at δ 4.3 and 2.7, implying that
these lie on the mirror plane and not across it. A crystallographic
study of 6 is fully consistent with these conclusions, but
unfortunately suffers from partial disorder of the lower {B6H6}
fragment, which has been modeled as two interpenetrating
pentagonal pyramids (see Experimental Section). Figure 4 is a
perspective view of one molecule of 6 in which only one of
these lower pentagonal pyramids is shown. Table 4 lists key
molecular parameters. Crystallographically, the disordered
model has precise C2 symmetry, but (both) the disordered
components also have effective Cs symmetry, at least as far as
the P2PtC2B10 fragments are concerned. Although the disorder
is also manifest in the upper C2B4 portion of the carborane ligand
4,1,10-MC2B10 Compounds. The nickelacarborane 4,4-
(dppe)-4,1,10-closo-NiC2B10H12, 7, afforded by nickelation of
reduced 1,12-closo-C2B10H12, has been reported in a recent
Figure 3. Experimental conformations of the {ML2} fragments in 3 (left; θ ) 148.7°), 4R (center; θ ) 134.6°) and 5 (right; θ ) 115.1°).
9
3306 J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007

Conformations of 13-Vertex ML2C2B10 Metallacarboranes
A R T I C L E S
Table 4. Selected Molecular Parameters (Å, deg) in
4,4-(PMe2Ph)2-4,1,10-closo-PtC2B10H12, 6
C1-B2 1.599(3) B3-B5
C1-B3 1.611(4) B3-B8
C1-Pt4 2.357(2) B5-B8
C1-B5 1.635(6) B5-B9
B2-Pt4 2.337(2) B5-B11 1.777(7) B11-B12
B2-B5 2.055(5) B6-B12 1.694(5) B11-B13
B2-B6 1.853(4) B7-B13 1.750(5) B12-B13
B2-B9 1.801(3) B8-B11 1.783(7) Pt4-P1
2.117(6) B8-B13
1.781(3) B9-B12
1.983(7) C10-B12
1.881(6) C10-B13
1.714(6)
1.666(6)
1.815(5)
1.747(5)
1.762(7)
1.762(9)
1.751(7)
2.2937(5)
P1-Pt4-P1A 91.87(2)
in that the formally degree-four and degree-five atoms C1 and
C10 are crystallographically equivalent, the disorder does not
influence the broad conclusion regarding the conformation of
the PtP2 unit with respect to the C2B4 ligand face (Figure 5
left). θ for 6 is 87.7°. Detailed analysis of the structure of 6 is
not warranted because of the crystallographic disorder, save to
note very long B2-B5 and B3-B5 and long B8-B5 and B9-
B5 distances. These typical aspects of 4,1,10-MC2B10 com-
pounds are more fully discussed below.
Figure 6. Perspective view of the anion of 8, with thermal ellipsoids as
for Figure 2.
Table 5. Selected Interatomic Distances (Å) in the
-
[4,4-cod-4,1,10-closo-RhC2B10H12
]
Anion of 8
The 11B{1H} NMR spectrum of 7 was relatively uninforma-
tive (only three broad peaks with relative integrals 4:1:5), but
importantly, there is only a single resonance in the 31P{1H}
NMR spectrum.6a The crystallographically determined molecular
structure reveals a docosahedral cage with the phosphine ligands
symmetrically disposed about an effective mirror plane passing
through C1, Ni4, C10, and B11. This is clearly visible in Figure
5 center, a view of the nickel co-ordination sphere. As
anticipated for such an orthogonal arrangement, the parameter
θ for compound 7 is 89.4°.
C1-B2
C1-B3
1.527(5) Rh4-B7
2.235(4) B9-B11
1.729(5)
1.718(5)
1.534(5) Rh4-C10 2.302(3) B9-B12
C1-Rh4 2.177(3) B5-B8
1.885(5) C10-B12 1.706(5)
1.915(5) C10-B13 1.695(5)
1.782(5) B11-B12 1.774(5)
1.763(5) B11-B13 1.778(5)
1.669(5) B12-B13 1.743(5)
1.756(5) Rh4-C41 2.169(3)
1.781(5) Rh4-C42 2.167(3)
1.674(4) Rh4-C45 2.151(3)
1.770(5) Rh4-C46 2.158(3)
1.756(5) C41-C42 1.404(4)
1.729(5) C45-C46 1.406(4)
C1-B5
1.761(5) B5-B9
B2-Rh4 2.266(4) B5-B11
B2-B5
B2-B6
B2-B9
2.072(5) B6-B9
1.813(5) B6-C10
1.834(5) B6-B12
B3-Rh4 2.350(4) B7-B8
B3-B5
B3-B7
B3-B8
1.967(5) B7-C10
1.858(5) B7-B13
1.786(5) B8-B11
The 11B{1H} NMR spectrum of 8 differs from that of 6 in
that the relative integrals are slightly altered (2:2:1:1:2:2) and
the whole spectrum is compressed and shifted to low frequency
(δ -4 to -18, δ(11B) ) -11.4), the latter anticipated for an
anionic metallacarborane. In the 1H spectrum the two cage CH
resonances are again well separated, δ 5.1 and 2.1, and the
different nature of the two cage {CH} units is further reinforced
by very different chemical shifts in the 13C{1H} spectrum, δ
79.0 and 42.4.
The docosahedral nature of the cage and the 4,1,10-RhC2
heteroatom pattern are confirmed by a crystallographic study
(Figure 6 and Table 5). This confirms that the anion has near
Cs symmetry about the plane defined by Rh4, C1, C10, and
B11. In common with recent structural studies on related
(ordered) 4,1,10-[M]C2B10 species ([M] ) CoCp, Ru(p-cymene),
Nidppe),5a the B2-B5 and B3-B5 connectivities in 8 are very
Rh4-B6 2.284(4) B8-B13
long, 2.072(5) and 1.967(5) Å, respectively, and the B8-B5
and B9-B5 distances long, 1.885(5) and 1.915(5) Å, respec-
tively. Defining X1 as the midpoint of the C41dC42 bond and
X2 as the midpoint of the C45dC46 bond, we note that the
RhX2 plane is slightly off perpendicular to the RhC1C10B11
plane, with θ ) 104.2° (Figure 5 right). Pairwise across the
effective mirror plane the B atom which is more closely trans
to the alkene ligand is the more strongly bound to Rh, i.e., Rh4-
B2 < Rh4-B3 and Rh4-B7 < Rh4-B6. The shortest Rh-B
connectivity is to B7, and is accompanied by a small but clear
trans influence in the Rh-Cene bonding, with Rh-C45,46
measurably shorter than Rh-C41,42.
The [N(PPh3)2]+ cation in 8 is ordered, with N-P distances
of 1.563(3) and 1.575(3) Å, and P-N-P ) 145.05(18)°. Salt
Figure 5. Experimental conformations of the {ML2} fragments in 6 (left; θ ) 87.7°), 7 (center; θ ) 89.4°) and 8 (right; θ ) 104.2°).
9
J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007 3307

A R T I C L E S
Dalby et al.
Chart 4
Chart 5
Figure 7. Experimental conformation of the {NiP2} fragment in 9. θ )
90.0°.
8 cocrystallises with 1/2CH2Cl2, disordered about a crystal-
lographic inversion center.
atoms. Clearly the metal LUMO has no conformational prefer-
ence. The HOMO and HOMO-1 (XII), which is pure metal d
character, have orthogonal conformational preferences; however,
because of its hybridization, the HOMO has a greater interaction
with the cage, and the overall conformation is set by the metal
HOMO/cage LUMO interaction. For VII and IX as drawn
above, the cage LUMO has a left-right nodal plane, and for
VIII a top-bottom nodal plane, leading to conformations in
the cases of VII and IX and | in the case of VIII;
4,1,12-MC2B10 Compounds. The nickelacarborane 4,4-
(dppe)-4,1,12-closo-NiC2B10H12, 9, is produced when the cor-
responding 4,1,10-isomer, compound 7, is heated in refluxing
toluene.6a Compound 9 has been characterized crystallographical-
ly.6a Although the structural determination suffers from partial
disorder in one-half of the dppe ligand, it is nevertheless possible
to define the conformation of the (major) NiP2 plane relative
to the plane through NiC1B10B11 as θ ) 90.0° (Figure 7).
2. Conformations. The crystallographic studies described
above reveal what appears to be a clear preference for the
conformations of {ML2} fragments bonded to each of the four
kinds of six-atom ligand faces in 13-vertex 4,1,x-MC2B10
metallacarboranes. For ligand type III (4,1,2-MC2B10) θ is 61.1°
(compound 1) and 64.6° (2). For ligand IV (4,1,6-MC2B10) we
observe θ ) 148.7° (3), 134.6°, 136.7° (4), and 115.1° (5).
Ligand type V (4,1,10-MC2B10) favors effectively perpendicular
arrangements, with θ ) 87.7° (6), 89.4° (7), and 104.2° (8),
and this situation is maintained in the one example of ligand
type VI (4,1,12-MC2B10), with θ ) 90.0° in compound 9. In
this section we explore these conformations computationally.
In earlier studies of the conformations of {ML2} fragments
in icosahedral metallacarboranes we described the conformations
as parallel (|) or perpendicular ( ) with respect to the mirror
plane through the cage. When there are two adjacent cage C
atoms in the ligating face, a conformation is observed (VII,
Chart 4).14b However, when the cage C atoms are nonadjacent,
the alternative | conformation is seen (VIII).27 Finally, in the
case of a single C atom in the ligating face, the conformation
returns to (IX).28
(3) the magnitude and direction of the slip distortion of the
{ML2} fragment with respect to the cage is that which
maximizes both sets of frontier orbital interactions.
These early calculations were useful in that they rationalized
the observed conformations in terms of simple symmetry-based
arguments, appropriate since in the icosahedron all vertices are
of degree-five. This, however, is not the case in the 13-vertex
docosahedron and henicosahedron. Nevertheless, to allow
optimal comparison with the icosahedral system we wanted to
explore the conformations of ML2 fragments with respect to
six-atom carborane rings in which the B and C atoms were
equally connected, and thus initial calculations focused on the
18-electron heteroarene complexes, [(PH3)2PtC2B4H6]2- and
[(PH3)2PtCB5H6]3-. We attempted to optimize both | and
structures for all three isomers of the C2B4 species and for the
CB5 species, using DFT.
When the two C atoms are adjacent, i.e., the analogue of the
C2B10 ligand type III, and structures were minimized in Cs
symmetry, the
form (Figure 8, left) was found to be a
minimum. The | isomer (Figure 8, right) lies 4.4 kcal mol-1
above the form, but the former is not a local minimum, having
two imaginary frequencies which correspond to PH3 rotation.
If the symmetry constraint is relaxed, however, a further
minimum with one C atom eclipsed by a PH3 group (Figure 8,
center) is located just +0.68 kcal mol-1 above the structure.
Note that the minimum ( ) conformation corresponds to a form
with θ ) 60° and recall that in compounds 1 and 2 real examples
of metallacarboranes incorporating the type III C2B10 ligand
measured θ values were 61.1° and 64.6°.
In all cases we rationalized these observations by the results
of MO calculations (at the extended Hu¨ckel level) which
revealed14 that:
(1) in carborane ligands the frontier molecular orbitals are
localized primarily on the B atoms in the open face;
(2) partitioning the electrons as d10-ML2 (and consequently
neutral nido-C2B9 or monoanionic nido-CB10 ligands), the metal
fragment LUMO is an s-pz hybrid (X, Chart 5) and the HOMO
an in-plane d-p hybrid (XI), both reinforced away from the P
In Figure 9 (left) is shown the minimum energy structure,
optimized in Cs symmetry, for the [(PH3)2Pt(C2B4-1,3)]2-
species, the analogue of the platinacarborane 3 containing the
C2B10 carborane type IV. This | form is 19.1 kcal mol-1 more
(27) (a) Green, M.; Spencer, J. L.; Stone, F. G. A.; Welch, A. J. J. Chem. Soc.,
Chem. Commun. 1974, 571. (b) Green, M.; Spencer, J. L.; Stone, F. G. A.;
Welch, A. J. J. Chem. Soc., Dalton Trans. 1975, 179. (c) Welch, A. J. J.
Chem. Soc., Dalton Trans. 1975, 1473.
(28) Carroll, W. E.; Green, M.; Stone, F. G. A.; Welch, A. J. J. Chem. Soc.,
Dalton Trans. 1975, 2263.
stable than the alternative
form (Figure 9, right) which
corresponds to a transition state for {Pt(PH3)2} rotation. The |
9
3308 J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007

Conformations of 13-Vertex ML2C2B10 Metallacarboranes
A R T I C L E S
Figure 8. Computed [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] conformations of [(PH3)2Pt(C2B4H6-1,2)]2-: (Left) Minimum energy. (Center) +0.68 kcal
mol-1. (Right) +4.4 kcal mol-1
.
Figure 9. Computed [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] conformations of [(PH3)2Pt(C2B4H6-1,3)]2-: (Left) Minimum energy. (Right) +19.1 kcal
mol-1
.
Figure 10. Computed [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] conformations of [(PH3)2Pt(C2B4H6-1,4)]2-: (Left) Minimum energy. (Right) +22.0 kcal
mol-1
.
geometry of this model compound corresponds to θ ) 120° in
4,1,6-MC2B10 metallacarboranes, in broad general agreement
with experimental θ values of 148.7° (compound 3), 134.6° and
136.7° (compound 4), and 115.1° (compound 5).
There are two mirror planes of symmetry perpendicular to
the C2B4-1,4 heteroarene ring (the analogue of carborane type
has PtP2 and PtC2 planes orthogonal and lies 22.0 kcal mol-1
below its rotamer in which the {(PH3)2Pt} fragment eclipses
the C atoms (Figure 10, right). Again, the latter does not
correspond to a local minimum, having two imaginary frequen-
cies, one corresponding to {Pt(PH3)2} rotation and the other to
PH3 rotation. In C1 symmetry exactly the same minimum as
that shown in Figure 10 (left) is found, with the form of Figure
10 (right) collapsing to this by simple {Pt(PH3)2} rotation. The
minimum structure corresponds to θ ) 90° in metallacarboranes
V), rendering the descriptors | and
inappropriate and
conferring potential C2V symmetry on the [(PH3)2PtC2B4H6]2-
ion. In C2V symmetry the minimum structure (Figure 10, left)
9
J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007 3309

A R T I C L E S
Dalby et al.
Figure 11. Results of attempted optimization [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] of [(PH3)2PtCB5H6]3-: (Left and center) with Cs symmetry imposed.
(Right) with no imposed symmetry. Although the form on the right is ∼50 kcal mol-1 more stable than the others, the original structure with a CB5
metallabonded ring has collapsed to give a distorted closo 7-vertex platinacarborane with only a CB4 metallabonded ring.
Figure 12. Qualitative representation of the splitting of the e1g pair of MOs of [B6H6]4- on carbon atom substitution, leading to (lower part) predicted
conformations for the three isomers of [(PH3)2PtC2B4H6]2- and for [(PH3)2PtCB5H6]3-
.
incorporating the ligand type V, in excellent agreement with
measured θ values of 87.7°, 89.4°, and 104.2° in compounds
6, 7, and 8, respectively.
ring. This form (Figure 11, right) has effective Cs symmetry
but cannot be accessed without removal of the symmetry
constraint.
Finally, we attempted to optimize | and
forms of
A qualitative understanding of the [(PH3)2PtC2B4H6]2- con-
formations can be gained by considering how the degeneracy
of the e1g pair of π MOs of [B6H6]4- is lifted upon carbon
substitution (Figure 12). Since C is more electronegative than
B, that orbital with the greater coefficient on C will be
preferentially stabilized, becoming the HOMO of [C2B4H6]2-
(or [CB5H6]3-). Conversely, the component of the e1g pair with
the smaller coefficient on C will form the LUMO of the
heteroarene. By analogy with the previous work14 on icosahedral
platinacarboranes, we would expect the conformation to be set
by the metal HOMO/heteroarene LUMO interaction, leading
to the predicted conformations of [(PH3)2PtC2B4H6]2- shown
in Figure 12, all of which are confirmed by calculation and
supported by the crystallographic work on compounds 1-8. This
[(PH3)2PtCB5H6]3- in which the {(PH3)2Pt} fragment is bonded
to a CB5 ring, the analogue of carborane type VI. However,
this species, a model for both 4,1,8- and 4,1,12-MC2B10 metalla-
carboranes (the latter represented experimentally by 9), did not
optimize as expected. If Cs symmetry is imposed, two structures
are found, one | and the other , but neither is a minimum,
each having two imaginary frequencies. The former (Figure 11,
left) has the B atom opposite C substantially displaced out of
the ring toward Pt, whereas in the latter (Figure 11, center) the
CB5 ring has a chair conformation with the B opposite C bent
away from Pt and an unrealistic B-H vector. In C1 symmetry,
a minimum is found (∼50 kcal mol-1 below the two former
structures), but this structure has collapsed from a nido species
with a CB5 metallabonded ring into a distorted closo 7-vertex
platinacarborane in which the metal is attached to only a CB4
analysis also predicts a conformation for [(PH3)2PtCB5H6]3-
Although this model compound was not stable under optimiza-
.
9
3310 J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007

Conformations of 13-Vertex ML2C2B10 Metallacarboranes
A R T I C L E S
Figure 13. Optimized structure [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] of 4,4-(PH3)2-4,1,2-closo-PtC2B10H12 in C1 symmetry: (Left) Perspective view.
(Right;) View showing the conformation. θ ) 68.4°.
Table 6. Comparison of Selected Molecular Parameters (Å, deg)
in (Theoretical) 4,4-(PH3)2-4,1,2-closo-PtC2B10H12 and
(Experimental) 1,2-(CH2)3-4,4-(PMe2Ph)2-4,1,2-closo-PtC2B10H10, 1
(PH3)2PtC2B10H12
1a
Pt4-C1
Pt4-C2
Pt4-B3
Pt4-B6
Pt4-B7
Pt4-B10
C2-B5
Pt4-P1
Pt4-P2
P1-Pt4-P2
θ
2.511
2.906
2.341
2.312
2.307
2.253
1.797
2.335
2.353
98.2
2.381(8)
2.626(8)
2.322(8)
2.310(7)
2.305(7)
2.265(7)
2.055(12)
2.3046(18)
2.3001(17)
96.07(6)
60.8
68.0
a Data taken from ref 8.
tion, such a conformation is displayed by the analogous real
compound 9.
While we have noted good general agreement between the
experimental conformations in compounds 1-8 and those
predicted for the appropriate models with C2B4 heteroarene
ligands, we recognize that these simple models are somewhat
inappropriate in that, in the real systems, the two cage C atoms
are not equivalent, one being of degree-four and the other
degree-five. We have therefore extended the computational work
to include optimizations of the 4,1,2-, 4,1,6-, and 4,1,10- isomers
of the model 13-vertex metallacarborane, (PH3)2PtC2B10H12,
more closely to mimic the real systems. We have also optimized
4,4-(PH3)2-4,1,12-closo-PtC2B10H12, which could not satisfac-
Figure 14. Henicosahedral transition state computed [B3LYP/6-31G(d,p);
Pt, SDD; P, SDD+d] for 1,2-(CH2)3-4,4-(PMe3)2-4,1,2-closo-PtC2B10H10.
a clear distortion toward henicosahedron II, the computed
structure is more like docosahedron I, albeit with a very
extended Pt4-C2 connectivity. Nevertheless, the computed
conformation (right side of Figure 13), with θ ) 68.4°, is in
reasonable agreement with that found, θ ) 61.1° (Figure 1).
Seeking further to understand these differences between
computed and experimental results, we have also performed
calculations on a much more representative species, 1,2-(CH2)3-
4,4-(PMe3)2-4,1,2-closo-PtC2B10H10, in which the tether is
included and a more realistic phosphine is used. In C1 symmetry
this optimized with much the same cage structure as that found
for the simplified species, 4,4-(PH3)2-4,1,2-closo-PtC2B10H12,
i.e. with long Pt4-C2 (2.98 Å) and short C2-B5 (1.83 Å).
However, in separate calculations we also noticed that artificially
lengthening C2-B5 and concomitantly shortening Pt4-C2 to
torily be modeled by [(PH3)2PtCB5H6]3-
.
Figure 13 shows the structure optimized in C1 symmetry for
4,4-(PH3)2-4,1,2-closo-PtC2B10H12, and Table 6 provides a
comparison of key molecular parameters for this optimized
structure and for that of the analogous compound 1,2-(CH2)3-
4,4-(PMe2Ph)2-4,1,2-closo-PtC2B10H10, 1.8 There is generally
good agreement between calculated and found molecular
structures, except that Pt4-C1 is somewhat overestimated in
the calculation, by ∼0.13 Å, and more seriously, C2-B5 is
∼0.26 Å too short and Pt4-C2 is ∼0.28 Å too long, in the
computed model. Thus, while the experimental structure shows
9
J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007 3311

A R T I C L E S
Dalby et al.
Figure 15. Optimized structure [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] of 4,4-(PH3)2-4,1,6-closo-PtC2B10H12 in C1 symmetry: (Left) Perspective view.
(Right) View showing the conformation. θ ) 141.1°.
Figure 16. Optimized structure [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] of 4,4-(PH3)2-4,1,10-closo-PtC2B10H12 in Cs symmetry: (Left) Perspective view.
(Right) View showing the conformation. θ ) 90.0° by symmetry.
values typical of those observed experimentally8 results in only
a marginally less stable molecule, perhaps suggesting that the
observed structure is influenced by solid-state effects to some
extent. Moreover, a relatively soft potential surface for this
deformation is entirely consistent with the facile diamond-
trapezium-diamond process we have previously proposed8 to
account for the fluxionality of compound 1 (and its Nidppe
analogue) in solution, as evidenced by 11B and 31P NMR
spectroscopy. In this process the docosahedral ground-state
structure transforms to the Cs symmetric henicosahedron via,
predominantly, C2-B5 extension and Pt4-C2 contraction. The
henicosahedron could either collapse back to the original
docosahedron or to its mirror image in which C1-B9 shortened
and Pt4-C1 lengthened. Since we could not arrest the flux-
ionality of 1 at low temperature we surmised that the activation
energy for the process was unlikely to be greater than ∼10 kcal
mol-1 29
For 1,2-(CH2)3-4,4-(PMe3)2-4,1,2-closo-PtC2B10H10
.
a
henicosahedral transition state was subsequently located 8.3 kcal
mol-1 above the ground-state structure, entirely in keeping with
this suggestion. This transition state is shown in Figure 14. It
has nearly perfect Cs cage symmetry, broken only for the
molecule as a whole by a staggered arrangement of the
{Pt(PMe3)2} unit. Pt4-C2 () Pt4-C1) is 2.52 Å and C2-B5
() C1-B9) is 2.46 Å. Reassuringly, a similar henicosahedral
transition state was also located for the simplified model
compound, 4,4-(PH3)2-4,1,2-closo-PtC2B10H12, 7.5 kcal mol-1
(29) Abel, E. W.; Bhargava, J. K.; Orrell, K. G. Prog. Inorg. Chem. 1984, 32,
1.
(30) The United Kingdom Chemical Database SerVice; Fletcher, D. A.;
McMeeking, R. F.; Parkin, D. J. Chem. Inf. Comput. Sci. 1996, 36, 746.
9
3312 J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007

Conformations of 13-Vertex ML2C2B10 Metallacarboranes
A R T I C L E S
Figure 17. Optimized structure [B3LYP/6-31G(d,p); Pt, SDD; P, SDD+d] of 4,4-(PH3)2-4,1,12-closo-PtC2B10H12 in C1 symmetry: (Left) Perspective
view. (Right) View showing the conformation. θ ) 84.4°.
Table 7. Comparison of Selected Molecular Parameters (Å, deg)
in (Theoretical) 4,4-(PH3)2-4,1,6-closo-PtC2B10H12 and
(Experimental) 4,4-(PMe2Ph)2-4,1,6-closo-PtC2B10H12, 3
Table 8. Selected Molecular Parameters (Å, deg) for the Cs
Symmetric Optimized Structure of
4,4-(PH3)2-4,1,10-closo-PtC2B10H12
(PH3)2PtC2B10H12
3
C1-B2
C1-Pt4
C1-B5
B2-Pt4
B2-B5
B2-B6
B2-B9
1.533
2.315
1.783
2.412
2.022
1.927
1.793
Pt4-B6
Pt4-C10
B5-B9
2.344
2.497
1.897
1.766
1.770
1.682
1.773
B9-B11
1.770
1.750
1.680
1.790
1.760
2.334
97.28
B9-B12
Pt4-C1
Pt4-B2
Pt4-B3
Pt4-C6
Pt4-B7
Pt4-B10
Pt4-P1
Pt4-P2
P1-Pt4-P2
θ
2.180
2.499
2.445
2.956
2.244
2.475
2.300
2.378
97.25
141.3
2.1652(12)
2.4438(15)
2.3961(14)
2.8131(13)
2.2631(14)
2.3692(14)
2.2710(3)
2.3129(4)
92.363(14)
148.4
C10-B12
B11-B12
B12-B13
Pt4-P1
B5-B11
B6-B9
B6-C10
B6-B12
P1-Pt4-P1A
Table 9. Selected Molecular Parameters (Å, deg) for the C1
Symmetric Optimized Structure of
4,4-(PH3)2-4,1,12-closo-PtC2B10H12
C1-B2
C1-B3
C1-Pt4
C1-B5
B2-Pt4
B2-B5
B2-B6
B2-B9
B3-Pt4
B3-B5
B3-B7
B3-B8
1.529
1.539
2.384
1.784
2.500
2.007
1.920
1.770
2.451
2.015
1.930
1.783
Pt4-B6
Pt4-B7
Pt4-B10
B5-B8
B5-B9
B5-B11
B6-B9
B6-B10
B6-C12
B7-B8
B7-B10
B7-B13
2.250
2.297
2.301
1.901
1.886
1.768
1.770
1.847
1.679
1.784
1.794
1.752
B8-B11
B8-B13
B9-B11
B9-C12
B10-C12
B10-B13
B11-C12
B11-B13
C12-B13
Pt4-P1a
1.746
1.756
1.762
1.683
1.704
1.770
1.722
1.786
1.692
2.338
2.341
95.25
above the ground-state structure, implying that, at least in terms
of the activation barrier, these simplified models represent
appropriate approximations.
The structure calculated (in C1 symmetry) for 4,4-(PH3)2-
4,1,6-closo-PtC2B10H12 is shown in Figure 15 and is in good
agreement with that already described for compound 3. Table
7 compares key molecular parameters. The greatest discrepan-
cies are in Pt4-C6 and Pt4-B10 separations, too long in the
calculations by ∼0.15 and 0.10 Å, respectively. Nevertheless,
the very long Pt-C6 distance and severe buckling of the C2B4
carborane ligand face are both faithfully reproduced in the
calculation, which also successfully predicts (in the correct
sense) unequal Pt-P distances. The calculated (θ ) 141.1°)
and observed (θ ) 148.7°) conformations also agree well, and
it is particularly noteworthy that the conformation of this
platinacarborane is much better reproduced in the model using
a C2B10 ligand rather than the simplified (heteroarene) C2B4
ligand.
Pt4-P2a
P1-Pt4-P2
a P1 trans to B3-B7 and P2 trans to B2-B6, as in compound 9.
C10 2.484 Å), in the diffraction study C1 and C10 are disordered
about a crystallographic C2 axis, and a single, average Pt-C
distance is measured, 2.357(2) Å. Nevertheless, the overall
comparison between experimental and computational structures
is clearly very good. In the crystallographic study θ ) 87.7°,
while in the computational study it is required to be precisely
90.0° by the Cs symmetry imposed.
In Figure 16 are perspective and plan views of the structure
of 4,4-(PH3)2-4,1,10-closo-PtC2B10H12 optimized in Cs sym-
metry, while Table 8 lists selected molecular parameters. The
appropriate comparison here would be with the structure
determined for 6, but unfortunately, as noted above, this suffers
from crystallographic disorder which masks subtle variations
within the cage. Thus, for example, while the computational
study predicts different Pt-C distances (Pt4-C1 2.291 Å, Pt4-
Figure 17 shows the structure optimized for 4,4-(PH3)2-4,1,12-
closo-PtC2B10H12, and Table 9 lists important derived molecular
parameters. Although we do not have an experimental 4,1,12-
closo-PtC2B10 compound with which to compare this compu-
tational result (the obvious precursor, compound 6, is relatively
unstable and does not survive even gentle thermolysis), it is
encouraging that the main features of the only two 4,1,12-
MC2B10 species to have been crystallographically characterized,
9
J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007 3313

A R T I C L E S
Dalby et al.
4-Cp*-4,1,12-closo-CoC2B10H1210k (Cp* ) η-C5Me5) and 4,4-
dppe-4,1,12-closo-NiC2B10H12, 9,6a viz. long B2-B5, B3-B5,
and B5-B9 distances, are reproduced by the calculation.
Moreover, the calculated θ value, 84.4°, is in excellent agree-
ment with that measured for 9, 90.0°.
version of docosahedral enantiomers, consistent with our
inability to arrest the fluxionality of 1,2-(CH2)3-4,4-(PMe2Ph)2-
4,1,2-closo-PtC2B10H10. Calculations on (PH3)2PtC2B10H12 also
allowed the structure of the experimentally unavailable 4,1,12-
PtC2B10 cluster to be explored and proved superior to experiment
in the case of the severely crystallographically disordered 4,1,10-
PtC2B10 species.
Conclusions
Recent developments in supraicosahedral metallacarborane
chemistry have allowed access to homologous series of com-
pounds with four different types of six-atom faces presented to
the metal fragment; including the three new compounds reported
herein, we now have a total of nine crystallographically
characterized examples of species with {ML2} fragments bound
to these six-atom faces. In these species clear patterns are visible
in the conformations of the {ML2} fragments with respect to
the carborane ligand.
Initially we explored these conformations for C2B4 carborane
faces by DFT calculations on the model compounds [(PH3)2PtC2-
B4H6]2- and successfully rationalized the results by qualitative
MO arguments. Unfortunately, analogous calculations on
[(PH3)2PtCB5H6]3- as a model compound with a CB5 ligand
face were unsuccessful. However, further calculations on the
polyhedral model compounds (PH3)2PtC2B10H12 successfully
reproduced the observed conformations for all four types of
carborane ligand.
Acknowledgment. We thank the EPSRC (D.E., S.E.), the
Carnegie Trust (R.D.McI.), and the Robert A. Welch Foundation
(B.E.H., T.D.McG.; Grant AA-0006) for support. At Heriot-
Watt we thank Dr. A. S. F. Boyd for NMR spectra, Mr. G.
Smith for mass spectra, Ms. C. Graham for elemental analysis,
and we acknowledge provision of the diffractometer from
EPSRC Grant GR/R99065. At Baylor the diffractometer was
purchased with funds from the National Science Foundation
Major Research Instrumentation Programme (Grant CHE-
0321214). We acknowledge use of the EPSRC Chemical
Database Service at Daresbury.30 V.S. is the recipient of a
Socrates scholarship from the Philipps-Universita¨t Marburg,
Germany.
Supporting Information Available: Full details of the
crystallographic analyses of 3, 6, and 8‚1/2CH2Cl2 as a CIF;
Cartesian coordinates of all model compounds shown in Figures
or otherwise discussed in the text along with complete ref 23
as a text file. This material is available free of charge via the
Internet at http://pubs.acs.org
In some cases the calculations led to apparent overestimation
of Pt-C distances, and in investigating this for the realistic
model compound, 1,2-(CH2)3-4,4-(PMe3)2-4,1,2-closo-PtC2B10H10,
we were able to compute a low activation energy for intercon-
JA067698M
9
3314 J. AM. CHEM. SOC. VOL. 129, NO. 11, 2007