Metal cation-methyl interactions in CB11Me12 - salts of Me3Ge+, Me3Sn +, and Me3Pb+

Article abstract of DOI:10.1021/ja0475205

Oxidation of Me₆M₂ (M = Ge, Sn) and Me₄Pb with the CB₁₁Me₁₂ radical in alkane solvents produced the insoluble salts Me₃M⁺CB₁₁ Me₁₂, ^- characterized by CP-MAS NMR and EXAFS. The cations interact with methyl groups of CB₁₁Me₁₂^- with coordination strength increasing from Pb to Ge. Density functional theory (DFT) calculations for the isolated ion pairs, Me₃M⁺CB₁₁Me ₁₂^- (M = Ge, Sn), revealed three isomers with the cation above methyl 2, 7, or 12, and not above a BB edge or a BBB triangle. The interaction has a considerable covalent component, with the cation attempting to perform a backside S_E2 substitution on the methyl carbon. In a fourth less favorable isomer the cation is near methyl 1, inclined toward methyl 2, and interacts with hydrogens. DFT atomic charge distributions and plots of the electrostatic potential on the surface of spheres centered at the CB ₁₁H₁₂ and CB₁₁Me₁₂-, icosahedra display the effects of uneven charge distribution within the anion and contradict the common belief that the negative charge of the cage anion is concentrated primarily on the cage boron atoms 7-12; in CB₁₁Me ₁₂^-, roughly half is on the cage carbon and the rest on methyls 7-12.

Full text of DOI:10.1021/ja0475205

Published on Web 08/31/2004
-
Metal Cation-Methyl Interactions in CB11Me12 Salts of
Me3Ge+, Me3Sn+, and Me3Pb+
Ilya Zharov,†, Tsu-Chien Weng,‡ Anita M. Orendt,§ Dewey H. Barich,§
James Penner-Hahn,‡ David M. Grant,§ Zdenek Havlas,¶ and Josef Michl*,†
Contribution from the Department of Chemistry and Biochemistry, UniVersity of Colorado,
Boulder, Colorado 80309-0215, Department of Chemistry, UniVersity of Utah,
Salt Lake City, Utah 84112-0850, Department of Chemistry and Biochemistry, UniVersity of
Michigan, Ann Arbor, Michigan 48109-1055, and Institute of Organic Chemistry and
Biochemistry, Academy of Sciences of the Czech Republic, 16610 Prague 6, Czech Republic
Received April 28, 2004; E-mail: michl@eefus.colorado.edu
•
Abstract: Oxidation of Me6M2 (M ) Ge, Sn) and Me4Pb with the CB11Me12 radical in alkane solvents
produced the insoluble salts Me3M+CB11Me12-, characterized by CP-MAS NMR and EXAFS. The cations
-
interact with methyl groups of CB11Me12 with coordination strength increasing from Pb to Ge. Density
functional theory (DFT) calculations for the isolated ion pairs, Me3M+CB11Me12 (M ) Ge, Sn), revealed
-
three isomers with the cation above methyl 2, 7, or 12, and not above a BB edge or a BBB triangle. The
interaction has a considerable covalent component, with the cation attempting to perform a backside SE2
substitution on the methyl carbon. In a fourth less favorable isomer the cation is near methyl 1, inclined
toward methyl 2, and interacts with hydrogens. DFT atomic charge distributions and plots of the electrostatic
potential on the surface of spheres centered at the CB11H12- and CB11Me12- icosahedra display the effects
of uneven charge distribution within the anion and contradict the common belief that the negative charge
of the cage anion is concentrated primarily on the cage boron atoms 7-12; in CB11Me12-, roughly half is
on the cage carbon and the rest on methyls 7-12.
Introduction
and second, counterions and solvents that would be stronger
ligands than the methyl group must be absent.
Coordination of methyl groups to metal cations is of
considerable interest currently1 because of its role in C-H
activation2 and in cation-anion interactions in transition metal
catalysis.3 Although some structures with a cationic transition
metal coordinated to a methyl group are known,4-9 examples
of main group cations interacting with a methyl group are
exceedingly rare.10 Presently, we describe three new examples.
An examination of the coordination of the weak ligand, a methyl
group, to a cationic metal center poses two challenges. First, a
cationic metal center such as R3M+ has to be made available,
The preparation and properties of the heavier congeners of
carbenium ions, R3M+ (M ) Si, Ge, Sn, Pb), in condensed
media have been a subject of extensive studies and controversy
for decades11-13 because of fundamental interest in understand-
ing similarities and differences between carbon and heavier
group 14 elements. The obstacle hindering access to non-
coordinated R3M+ cations is their extreme electrophilicity, which
prevents the use of the usual fluoride-based superacids. In the
presence of counterions or solvents, covalent bonds between
the central atom of the R3M+ cation and the anion or the solvent
are formed with extraordinary ease,14-18 and very weakly
nucleophilic anions19 are required. Salts of simple non-
coordinated group 14 cations such as R3M+ (R ) H or alkyl)
† University of Colorado.
‡ Unversity of Michigan.
§ University of Utah.
Present address: Department of Chemistry, University of Utah.
¶ Academy of Sciences of the Czech Republic.
(1) Hall, C.; Perutz, R. N. Chem. ReV. 1996, 96, 3125 and references therein.
(2) Shilov, A. E.; Shul’pin, G. B. Chem. ReV. 1997, 97, 2879 and references
therein.
(11) Lickis, P. D. In The Chemistry of Organic Silicon Compounds, 2nd ed.;
Rappoport, Z., Apeloig, Y., Eds.; John Wiley & Sons: Chichester, UK,
1998; pp 557-594.
(3) Marks, T. J.; Chen, E. Y.-X. Chem. ReV. 2000, 100, 1391 and references
(12) Reed, C. A. Acc. Chem. Res. 1998, 31, 325.
therein.
(13) Zharov, I.; Michl. J. In The Chemistry of Organic Germanium, Tin and
Lead Compounds; Rappoport, Z., Apeloig, Y., Eds.; Wiley & Sons:
Chichester, 2002; Vol. 2, p 633.
(4) Burns, C. J.; Andersen, R. A. J. Am. Chem. Soc. 1987, 109, 5853.
(5) Stults, S. D.; Andersen, R. A.; Zalkin, A. J. Am. Chem. Soc. 1989, 111,
4507.
(14) Ashradi, M.; Johnels, D.; Edlund, U. Chem. Commun. 1996, 1279.
(15) Arshadi, M.; Johnels, D.; Edlund, U.; Ottosson, C.-H.; Cremer, D. J. Am.
Chem. Soc. 1996, 118, 5120.
(16) Schleyer, P. v. R.; Buzek, P.; Mu¨ller, T.; Apeloig, Y.; Siehl, H.-U. Angew.
Chem., Int. Ed. Engl. 1993, 32, 1471.
(17) Olah, G. A.; Rasul, G.; Prakash, G. K. S. J. Organomet. Chem. 1996, 521,
271.
(18) (a) Lambert, J. B.; Zhang, S.; Stern, C. L.; Huffman, J. C. Science 1993,
260, 1917. (b) Lambert, J. B.; Zhang, S.; Ciro, S. M. Organometallics 1994,
13, 2430.
(6) Berg, D. J.; Andersen, R. A. Organometallics 2003, 22, 627.
(7) Zhang, S.; Piers, W. E. Organometallics 2001, 20, 2088.
(8) Ingelson, M. J.; Mahon, M. F.; Patmore, N. J.; Ruggiero, G. D.; Weller,
A. S. Angew. Chem., Int. Ed. 2002, 41, 3694.
(9) Clarke, A. J.; Ingleson, M. J.; Kociok-Ko¨hn, G.; Mahon, M. F.; Patmore,
N. J.; Rourke, J. P.; Ruggiero, G. D.; Weller, A. S. J. Am. Chem. Soc.
2004, 126, 1503.
(10) Zharov, I.; King, B. T.; Havlas, Z.; Pardi, A.; Michl, J. J. Am. Chem. Soc.
2000, 122, 10253.
9
10.1021/ja0475205 CCC: $27.50 © 2004 American Chemical Society
J. AM. CHEM. SOC. 2004, 126, 12033-12046
12033

A R T I C L E S
Zharov et al.
Scheme 1
have not been prepared and isolated, since suitable solvents and
anions have not been found. Even tert-butyl groups fail to
provide sufficient protection,20 and only even more highly
sterically hindered21-25 and/or electronically stabilized26-28
cations are known in non-coordinated form.29 Although interest-
ing in their own right, such highly hindered R3M+ cations are
useless for the present purpose.
We use the salts Me3M+CB11Me12-, in which the requisite
methyl groups are carried by the anions, using alkanes as weakly
coordinating solvents. This was suggested by the observations
made on n-Bu3Sn+CB11Me12-, which showed clear evidence
of cation-methyl coordination.10 The parent of the anion used
is the deltahedral30 carba-closo-dodecaborate(-) anion CB11H12
-
(1), whose substituted derivatives have received considerable
attention recently.19 The parent CB11H12- (1) was first synthe-
sized by Knoth in 1967,31 and is now accessible in two steps
from cheap bulk chemicals, albeit in poor yield.32 It contains a
highly delocalized charge dispersed over a large icosahedral
cage. It is inert to electrochemical oxidation in acetonitrile33
and stable toward acids and bases. Its B-H bonds show distinct
hydridic character,34 and one can expect its halogenated35 and
methylated36,37 derivatives to show some halide and methide
anion character, and hence to have some coordinating ability
as well. In this regard the trifluoromethylated derivative appears
ideal, but unfortunately, it is shock-sensitive and explosive.38
The CB11Me12- anion (2)36 coordinates only weakly39 and can
only do so through its methyl groups. It has the additional
advantage of carrying a lipophilic barrier around the carborane
cage, alleviating possible solubility problems.40 It is very stable
toward bases and acids, except for neat strong acids such as
hydrofluoric, sulfuric, triflic, and trifluoroacetic.38
related to the choice of a reaction medium. For a long time,
hydride abstraction by the trityl cation41 provided the most
successful way of generating group 14 cations in solution.18,42
More recently, an alternative allyl leaving group approach was
utilized to prepare trimesitylsilylium,21 trimesitylgermylium,43
and trimesitylstannylium43 ions.23 However, in both cases
aromatic solvents appear to be the least coordinating media that
can be used. To provide a reaction environment devoid of all
unsaturation, we have taken advantage of a property of the
-
CB11Me12 anion that can be viewed as its strong but also its
weak point: it is much easier to oxidize than the parent
CB11H12-. The corresponding neutral radical CB11Me12 (3)39
•
is a stable one-electron oxidant that is freely soluble in nonpolar
solvents. Its oxidation potential in acetonitrile, 1.15 V above
the ferrocene/ferrocenium couple, is comparable to that of
Ce(IV) at neutral pH.
The choice of a procedure for the generation of the
Me3M+CB11Me12 salts poses an interesting challenge and is
-
(19) (a) Lupinetti, A. J.; Strauss, S. H. Chemtracts 1998, 11, 565. (b) Reed, C.
A. Acc. Chem. Res. 1998, 31, 133. (c) Krossing, I.; Raabe, I. Angew. Chem.,
Int. Ed. 2004, 43, 2066.
We use the radical 3 to prepare R3M+ cations by oxidation
of neutral and nonpolar R3M-containing precursors.10 This
reaction occurs even in highly nonpolar alkane solvents and
simultaneously provides the desired R3M+ cation and the anion
2 (Scheme 1). The oxidation is believed to proceed in two
single-electron-transfer steps. The first one would produce a
cation radical from an R3M-M′R′n precursor, whose weak one-
electron bond could then dissociate to give the R3M+ cation
and an R′nM′• radical. Depending on the nature of M′, the latter
could be further oxidized in a second single-electron-transfer
step to give the R3M′+ cation, or it could dimerize. Dispropor-
tionation is best avoided, since it produces unsaturated structures.
(20) Xie, Z.; Bau, R.; Benesi, A.; Reed, C. A. Organometallics 1995, 14, 3933.
(21) Lambert, J. B.; Zhao, Y. Angew. Chem., Int. Ed. Engl. 1997, 36, 400.
(22) Kim, K.-C.; Reed, C. A.; Elliott, D. W.; Mueller, L. J.; Tham, F.; Lin, L.;
Lambert, J. B. Science 2002, 297, 825.
(23) Lambert, J. B.; Lin, L.; Keinan, S.; Muller, T. J. Am. Chem. Soc. 2003,
125, 6022.
(24) Sekiguchi, A.; Fukawa, T.; Lee, V. Y.; Nakamoto, M. J. Am. Chem. Soc.
2003, 125, 9250.
(25) Sekiguchi, A.; Fukawa, T.; Lee, V. Y.; Nakamoto, M.; Ichinohe, M. Angew.
Chem., Int. Ed. 2003, 42, 1143.
(26) Sekiguchi, A.; Tsukamoto, M.; Ichinohe, M. Science 1997, 275, 60.
(27) Sekiguchi, A.; Matsuno, T.; Ichinohe, M. J. Am. Chem. Soc. 2000, 122,
11250.
(28) Sekiguchi, A.; Fukaya, N.; Ichinohe, M.; Ishida, Y. Eur. J. Inorg. Chem.
2000, 6, 1155.
(29) Sekiguchi, A.; Ichinohe, M.; Nakamoto, M. Organometallic News 2003,
1, 11.
(30) Grimes, R. N. Carboranes; Academic Press: New York & London, 1970.
(31) (a) Knoth, W. H. J. Am. Chem. Soc. 1967, 89, 1274. (b) Knoth, W. H.
Inorg. Chem. 1971, 10, 598.
(32) Franken, A.; King, B. T.; Rudolph, J.; Rao, P.; Noll, B. C.; Michl, J. Collect.
Czech. Chem. Commun. 2001, 66, 1238.
(33) Wiersema, R. J.; Hawthorne, M. F. Inorg. Chem. 1973, 12, 785.
(34) (a) Jel´ınek, T.; Plesˇek, J.; Herˇma´nek, S.; Sˇt´ıbr, B. Collect. Czech. Chem.
Commun. 1986, 51, 819. (b) Plesˇek, J.; Plza´k, Z.; Stuchl´ık, J.; Herˇma´nek,
S. Collect. Czech. Chem. Commun. 1981, 46, 1748.
(35) (a) Plesˇek, J. Chem. ReV. 1992, 92, 269. (b) Strauss, S. H. In Contemporary
Boron Chemistry; Wade, K., Marder, T. B., Hughes, A., Eds.; Royal Society
of Chemistry: London, 2000; Vol. 253, p 44. (c) Stasko, D.; Reed, C. A.
J. Am. Chem. Soc. 2002, 124, 1148.
(36) King B. T.; Janousˇek, Z.; Gru¨ner, B.; Trammell, M.; Noll, B. C.; Michl, J.
J. Am. Chem. Soc. 1996, 118, 3313.
There is good precedent for the production of cations of group
14 elements by oxidation of neutral precursors. Recently, a series
of complexes [t-Bu3MsNtCR]+ (M ) Si, Ge, Sn) has been
prepared44 by oxidation of the corresponding t-Bu6M2 dimet-
allanes with Ph3C+B(3,5-CF3-C6H3)4- in nitrile solvents. One-
electron oxidation of Me6Sn2, Me3SnGeMe3, or Me3SnSiMe3
by the 10-methacridinium cation in acetonitrile led to the
[Me3SnsNtCMe]+ cation.45 Oxidation of R4Sn, Me3SnR, and
(41) Corey, J. Y. J. Am. Chem. Soc. 1975, 97, 3237.
(42) Xie, Z.; Manning, J.; Reed, R. W.; Mathur, R.; Boyd, P. D.; Benesi, A.;
Reed, C. A. J. Am. Chem. Soc. 1996, 118, 2922.
(37) Clayton, J. R.; King, B. T.; Zharov, I.; Michl, J. Submitted for publication.
(38) King, B. T.; Michl, J. J. Am. Chem. Soc. 2000, 122, 10255.
(39) King, B. T.; Noll, B. C.; McKinley, A. J.; Michl, J. J. Am. Chem. Soc.
1996, 118, 10902.
(43) Lambert, J. B.; Zhao, Y.; Wu, H.; Tse, W. C.; Kuhlman, B. J. Am. Chem.
Soc. 1999, 121, 5001.
(44) Ichinohe, M.; Fukui, H.; Sekiguchi, A. Chem. Lett. 2000, 12, 600.
(45) Fukuzumi, S.; Kitano, T.; Mochida, K. J. Am. Chem. Soc. 1990, 112, 3246.
(40) Posp´ısˇil, L.; King, B. T.; Michl, J. Electrochim. Acta 1998, 44, 103.
9
12034 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004

-
Metal Cation
−
Methyl Interactions in CB11Me12 Salts
A R T I C L E S
Scheme 2
R6Sn2 (R ) Me, Et, n-Bu, Ph, vinyl) by the thianthrene cation
radical in acetonitrile also resulted in the formation of the
corresponding [R3SnsNtCMe]+ cations,46 and the [Ph3Pbs
NtCMe]+ cation was prepared by oxidation of Ph6Pb2 with
AgNO3 in acetonitrile.47 Electrooxidation of various organotin
and organolead compounds has been shown to proceed via
electron transfer as the rate-determining step and to first produce
the cation radicals [R3M-MR3]•+, which then decompose to give
R3M+ and R3M•, followed by further radical oxidation.48 One-
electron oxidation of the highly hindered stable stannyl radical
(t-Bu2MeSi)3Sn• yielded the stable cation (t-Bu2MeSi)3Sn+.24
The next issue is the solubility of the resulting salt in the
alkane solvent, critical if single crystals are to be grown. In a
short communication, we reported that the use of n-butyl
substituents provides sufficient solubility for crystallization of
the resulting salt, n-Bu3Sn+CB11Me12-, from hexane.10 Pres-
ently, however, we deal with the CB11Me12- salts of the simplest
trialkylated cations, Me3E+, accessed by applying the reaction
of Scheme 1 to Me6Ge2, Me6Sn2, and Me4Pb. In this case, the
lower solubility in alkanes prevented us from obtaining single
crystals, and we had to resort to powder EXAFS measurements
to obtain structural information. The results obtained from the
reaction of 3 with Me6Si2 were different in that 3 acted as a
methyl radical transfer agent rather than an electron transfer
agent, and they will be reported separately.49
The white solid 4, obtained in pentane at -78 °C, was
dissolved in dry C6D6, and then the solution was frozen in liquid
nitrogen and transported to an NMR instrument. An 119Sn NMR
spectrum was taken as soon as the solvent in the NMR tube
melted and showed a 119Sn NMR signal at 324.7 ppm. The
benzene solution of 4 was not stable upon warmup and darkened
ca. 10 min after reaching room temperature. Numerous 119Sn
NMR signals at lower chemical shifts were then observed,
together with a clean 11B NMR spectrum of 2, which suggests
that the benzene adduct of the Me3Sn+ cation decomposed,
leaving the anion 2 intact. This was confirmed by the negative-
mode ES/MS of the resulting solution, which showed only the
presence of 2.
Results and Discussion
A similar experiment with CD2Cl2 at -60 °C yielded a
solution that gave a 119Sn NMR spectrum with a signal at 335.9
ppm and a 11B NMR spectrum with signals at -1.96 (1 B) and
9.20 (10 B, unresolved), similar to those of 2. When the
temperature was raised from -60 to 0 °C, the signals of B7-11
and B2-6 became resolved, which suggests that at the lower
temperature the interaction of the Me3Sn+ cation and the anion
2 is stronger than the interaction of the cation with the solvent,
while at higher temperature this order is reversed. The CD2Cl2
solution of 4 was again unstable above 0 °C.
Preparation and NMR Spectra of Me3Sn+CB11Me12- (4).
When a solution of 1 equiv of hexamethyldistannane (Me6Sn2)
in dry pentane was added to a solution of 2 equiv of the radical
3 in dry pentane at -78 °C under argon atmosphere, the blue
color of the radical disappeared within seconds and a white solid
4 formed. This was soluble in polar solvents. The 11B NMR
spectrum of its acetone solution was identical to that of the Cs+
salt of the anion 2. Negative-mode ES/MS of its methanol
solution showed a single anion with m/e 311 (both m/e and
isotopic distribution identical to those of the anion 2), while
positive-mode ES/MS of the same solution showed a cation with
m/e 165 and isotopic distribution characteristic of compounds
with a single Sn atom in the molecule. This corresponds to the
Me3Sn+ cation. In acetonitrile, several complexes of the type
[Me3Sn(CH3CN)m(H2O)n]+ were found in the mass spectrum.
Such behavior of R3Sn+ cations under atmospheric pressure
ionization/electrospray conditions has been reported.50 All this
suggested that 4 indeed was the desired Me3Sn+CB11Me12- salt
(Scheme 2).
The 119Sn NMR shifts found for 4 in C6D6 and CD2Cl2 are
similar to (a) the signal at 348 ppm reported52 for n-Bu3Sn+-
-
B[3,5-(F3C)2-C6H3]4 in CD2Cl2 at -70 °C (this signal disap-
peared above -20 °C); (b) the signal at 360 ppm reported53 for
n-Bu3Sn+[B(C6F5)3H]- in C6D6; and (c) the signal at 322 ppm
reported54 for Me3Sn+SO3F- in HSO3F at -60 °C. They are
somewhat lower than the 434.2 ppm reported55 for Bu3Sn+-
[B(C6F5)4]- in toluene-d8 and the 434 ppm reported56 for
Bu3Sn+[C6F4-1,2-{B(C6F5)2}2(µ-OCH3)]- in toluene-d8.
When dissolved in CD3CN at -40 °C, the white solid 4 gave
a 119Sn NMR signal at 233.1 ppm and 11B NMR signals at -1.07
(1 B), -8.97 (5 B), and -10.86 (5 B). This solution was
unstable above -5 °C. The large displacement of ∼ -100 ppm
in the 119Sn NMR signal of 4 in CD3CN demonstrates that the
interaction of Me3Sn+ with the solvent is quite strong, while
11B NMR shows no interaction with the anion. When a small
amount of water was added to this solution, a new signal at
205.0 ppm appeared, similar to that reported57 for (Et3Sn)2O+H
in C6D6.
This conclusion was further verified by a reaction of 4 with
PhLi in pentane (Scheme 2), which yielded two products. A
pentane-soluble product was isolated by preparative GC in 95%
yield based on Me6Sn2. Its NMR and GC/MS spectra were
identical with those of an authentic sample of Me3SnPh and
agreed with published reports.51 The pentane-insoluble product
was identified by NMR and ES/MS as Li+2.
(46) Loczynski, S.; Boduszek, B.; Shine, H. J. J. Org. Chem. 1991, 56, 914.
(47) Doretti, L.; Faleschini, S. Gazz. Chim. Ital. 1970, 100, 819.
(48) Kochi, J. K. Angew. Chem., Int. Ed. Engl. 1988, 27, 1227.
(49) Zharov, I.; Fete, M. G.; Michl, J. Unpublished results.
(50) Dakternieks, D.; Lim, A. E. K.; Lim, K. F. Phosphorus, Sulfur, Silicon
Relat. Elem. 1999, 150-151, 339. (b) Henderson, W.; Taylor, M. J.
Polyhedron 1996, 15, 1957.
(52) Kira, M.; Oyamada, T.; Sakurai, H. J. Organomet. Chem. 1994, 471, C4.
(53) Lambert, J. B.; Kuhlmann, B. J. Chem. Soc., Chem. Commun. 1992, 931.
(54) Birchall, T.; Manivannan, V. J. Chem. Soc., Dalton Trans. 1985, 2671.
(55) Blackwell, J. M.; Piers, W. E.; McDonald R. J. Am. Chem. Soc. 2002,
124, 1295.
(56) Henderson, L. D.; Piers, W. E.; Irvine, G. J.; McDonald, R. Organometallics
2002, 21, 340.
(51) Bullpitt, M.; Kitching, W.; Adcock, W. J. Organomet. Chem. 1976, 116,
161.
9
J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004 12035

A R T I C L E S
Zharov et al.
of the products showed [(Me3Ge)2OH]+ and Me3Ge+ cations.
Reaction of the white precipitate 5 with PhLi in pentane
(Scheme 2) yielded Me3GePh (89% yield based on Me6Ge2,
isolated by preparative GC; NMR and MS spectra identical to
those reported59) and Li+2 (by NMR and ES/MS). All this
-
suggested that 5 is the anticipated Me3Ge+CB11Me12 salt
(Scheme 2).
Germanium lacks a sensitive and convenient NMR nuclide,60
and there is no easy direct NMR method to confirm the
formation of Me3Ge+ in the above reaction by dissolving 5 in
a way similar to that described for 4, or in the solid state. The
CP-MAS 11B NMR spectrum of 5 was identical to that of Cs+2.
The CP-MAS 13C NMR spectrum of 5 (Figure 1A) was
compatible with the presence of Me3Ge+, and neither of these
spectra showed direct evidence for a coordinating interaction.
The 13C signal of Me3Ge+ at 5.6 ppm is at higher chemical
shift relative to the signals of CR-Ge in neutral germanium
compounds (cf. -1 ppm in Me4Ge, -1.7 ppm in Me3GeCN59),
compatible with the presence of positive charge on the Ge atoms.
It is at lower chemical shift from the corresponding signal for
Me3Sn+ (10.4 ppm), suggesting that the positive charge on the
Ge atom in 5 is smaller than that on the Sn atom in 4, and that
the cation-anion charge-transfer interactions in 5 are stronger
than those in 4. However, the difference could also be due to
increased bond strengths and higher excitation energies in
Me3Ge+ compared to Me3Sn+.
-
Figure 1. CP-MAS 13C NMR of (A) Me3Ge+CB11Me12 (5), (B)
-
-
Me3Sn+CB11Me12 (4), and (C) Me3Pb+CB11Me12 (6).
The CP-MAS 11B NMR spectrum of solid 4 was identical to
that of Cs+2. CP-MAS 13C NMR of 4 (Figure 1B) confirmed
the presence of CB11Me12 and was compatible with the
-
presence of Me3Sn+. The signal of the Me3Sn+ carbon in 4 at
10.4 ppm is at higher chemical shift relative to the signals of
the CR-Sn carbon in neutral tin compounds (cf. -9 ppm for
Me4Sn, 0.3 ppm for Me3SnCl58), which is compatible with the
presence of positive charge on the Sn atom, in agreement with
the 119Sn NMR chemical shift. We realize that other factors
may dominate the chemical shift value, in particular the
decreased excitation energies in the cation.
The reaction of Et4Ge with the radical 3 is much slower than
the reaction of Me6Ge2 but still results in Et3Ge+CB11Me12
-
(as determined by ES/MS).
Preparation and NMR Spectra of Me3Pb+CB11Me12- (6).
Hexamethyldiplumbane (Me3Pb-PbMe3) can be prepared rela-
tively easily61 but is unstable and readily decomposes after
preparation, or upon storing even in the absence of air. Since it
was possible to oxidize Et4Sn and Et4Ge with the radical 3,
oxidation of tetramethylplumbane (Me4Pb), which is a stable
compound at room temperature (but can spontaneously detonate
during a distillation62), was attempted. When 1 equiv of
tetramethylplumbane (Me4Pb) in dry pentane was treated with
1 equiv of the radical 3 in dry pentane at -78 °C, the blue
color of the radical disappeared within seconds, and a white
solid 6 was formed. 11B NMR and negative-mode ES/MS of
its acetone solution clearly showed that the anion 2 was formed
as the only boron-containing product in the above reaction. In
addition, positive-mode ES/MS of its methanol solution showed
the Me3Pb+ cation.
The CP-MAS 119Sn NMR signal for solid 4 was found at
466.2 ppm. While this number is higher than those observed
for the complexes of 4 in solution, it is much lower than the
806 ppm chemical shift reported43 for the Mes3Sn+ cation, and
particularly the 2653 ppm shift reported24 for the non-
coordinated tin atom in (t-Bu2MeSi)3Sn+B(C6F5)4-. Clearly, the
Me3Sn+ cation in solid 4 is significantly coordinated, and the
only coordination possible is to the methyl groups of the anion.
Unfortunately, CP-MAS NMR did not reveal any changes in
the chemical shifts of either methyl group carbons or cage boron
atoms of the anion that would provide further evidence for this
coordination.
Reaction of the white precipitate 6 with PhLi in pentane
yielded Me3PbPh (89% yield based on Me4Pb, isolated by
preparative GC; NMR and MS spectra were identical to those
reported51) and Li+2 (by NMR and ES/MS). All this suggested
Et4Sn was also tested as a precursor for a trialkylstannylium
ion. Its reaction with the radical 3 is slower and not as clean as
in the case of Me6Sn2 but does result in Et3Sn+CB11Me12-, as
judged by ES/MS.
Preparation and NMR Spectra of Me3Ge+CB11Me12- (5).
An analogous oxidation of 1 equiv of hexamethyldigermane
(Me6Ge2) with 2 equiv of the radical 3 in dry pentane at room
temperature produced a white solid 5 within 1 h. The 11B NMR
spectrum of its acetone solution was identical to that of the Cs+
salt of the anion 2. Negative-mode ES/MS of its methanol
solution showed only the anion 2, while positive-mode ES/MS
-
that 6 was the hoped-for Me3Pb+CB11Me12 (Scheme 2).
The white solid 6 obtained in pentane was not soluble in C6D6
but dissolved in dry CD2Cl2. This solution, stable at room
temperature, showed a 207Pb NMR signal at 1007.4 ppm and
11B NMR signals at -0.99 (1 B), -8.37 (5 B), and -9.79
(5 B), similar to those of the Cs+ salt of the anion 2. The 207Pb
(59) Neumann, W. P.; Kuehlein, K. Justus Liebigs Ann. Chem. 1967, 702, 13.
(60) Harris, R. K. In NMR and Periodic Table; Harris, R. K., Mann, B. E.,
Eds.; Academic Press: London, 1978; p 7.
(61) Arnold, D. P.; Wells, P. R. J. Organomet. Chem. 1976, 111, 269.
(62) .Gilman, H.; Jones, R. G. J. Am. Chem. Soc. 1950, 72, 1760.
(57) Lambert, J. B.; Ciro, S. M.; Stern, C. L. J. Organomet. Chem. 1995, 499,
49.
(58) Mann, B. E.; Taylor, B. F. 13C NMR Data for Organometallic Compounds;
Academic Press: London, 1981; pp 64-77.
9
12036 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004

-
Metal Cation
−
Methyl Interactions in CB11Me12 Salts
A R T I C L E S
NMR shift found for 6 in CD2Cl2 is similar to the signal at 980
ppm reported14 for Me3Pb+SO3F- in HSO3F at low temperature.
Although it is relatively easy to observe 207Pb NMR in
solution,63 CP-MAS 207Pb NMR experiments can be quite
demanding,64 and for 6 they have failed. At the same time, the
13C and 11B CP-MAS NMR spectra of 6 were measured
successfully. The CP-MAS 11B NMR spectrum of 6 was
identical to that of Cs+2. The CP-MAS 13C NMR spectrum of
-
6 (Figure 1C) confirmed the presence of CB11Me12 and was
compatible with the presence of Me3Pb+. Neither spectrum
showed any direct evidence for a coordinating interaction. The
13C signal of Me3Pb+ at 31.4 ppm is at a significantly higher
chemical shift than the signals of the CR-Pb carbon in neutral
lead compounds (cf. -3 ppm in Me4Pb, 15.2 ppm in Me3PbBr)59
and also from the corresponding signal for Me3Sn+ (10.4 ppm),
compatible with the notion that the positive charge on the Pb
atom in 6 is larger compared to that on the Sn atom in 4, and
that the cation-anion charge-transfer interactions in 6 are
weaker than those in 4. Once again, it is possible that the sources
of the differences are the weaker C-Pb bonds and the lower
excitation energies in Me3Pb+.
-
The EXAFS Structure of the Me3M+CB11Me12 Solids
(4-6). Our attempts to grow crystals of 4, 5, and 6 from alkane
solvents failed due to the insufficient solubility of these
compounds. Attempts to use slow diffusion of reagents also
failed, and powder diffraction65 showed that all the solids
obtained were predominantly amorphous. Structural information
was, however, obtained from extended X-ray absorption fine
structure (EXAFS) analysis. The refinement procedure was first
calibrated on the tetraphenyl compounds Ph4M (M ) Ge, Sn,
Pb). In each case, the C-M distance determined by EXAFS
agreed within 0.002 Å with that deduced from single-crystal
X-ray diffraction (Table S2, Supporting Information).66
Figure 2. Fourier transforms of the k3-weighted EXAFS data for 4, 5, and
6. All spectra are plotted on the same scale and offset vertically for clarity.
Fourier transforms were calculated over the range 2-15 Å-1 for 4 and 5
and 2-10 Å-1 for 6. For each element, data for Ph4E (pronounced outer-
shell scattering) and Me3E+CB11Me12 are paired to emphasize the
-
difference in outer-shell scattering.
Table 1. Structural Data (Å, Å2, deg) from EXAFS of Solid 4, 5,
-
and 6, and as Calculateda for Isolated Me3M+CB11Me12 Ion Pairs
-
Coordinated in Position 12 of CB11Me12
M
Geb
Sn
Pb
d(M-CR)
d(M‚‚‚C)
d(M‚‚‚X)c
EXAFS
σ2 × 103
calcd
1.94
1.6
1.98 (1.95)
2.12
2.4
2.15
2.17
3.8
2.18
The Fourier transforms of the EXAFS data for 4, 5, 6, and
the Ph4M references are shown in Figure 2, and the EXAFS
data are included in the Supporting Information (Figure S1).
The best fits to the EXAFS data are summarized in Table 1.
EXAFS
σ2 × 103
calcd
2.49
3
2.14 (2.10)
2.77, 3.02
3
2.42
2.9
8
2.63
Two sets of M-C distances, a short one and a long one, were
found by EXAFS of 4, 5, and 6. The first-shell peak is assigned
to the methyl carbon directly attached to the metal. As expected,
it moves progressively to longer distance on going from Ge to
Sn and to Pb (1.94, 2.12, and 2.17 Å, respectively; see Tables
1 and S2). This M-CR(sp3) distance is approximately 0.02 Å
shorter than the M-CR(sp2) distances in the corresponding Ph4M
compounds, and about 0.04 Å shorter than a normal M-CR-
(sp3) bond (1.98, 2.14, and 2.22 Å, respectively67), as would
be expected for a metal atom hybridization close to sp2.
Although this difference is small, previous studies have shown
that the precision of EXAFS is typically 0.004 Å.68 Therefore,
EXAFS
3.63
2.6
4.63
5.2
4.51
6.6
σ2 × 103
calcd d[C(12)‚‚‚B(12)]
calcd d[M‚‚‚C(7,8)]
calcd d[M‚‚‚C(7,8)]
calcd B12C12M
calcd ∑ CRMCR
calcd BCH
1.85 (1.75)
5.07 (4.81)
4.42 (4.11)
180 (171)
340 (342)
86 (90)
1.72
5.22
4.54
180
347
96
1.67
5.24
4.35
173
354
101
a B3LYP/SDD. b MP2/6-31G(d) results in parentheses. c An outer-shell
distance.
given the availability of excellent reference compounds, we are
confident that this represents a real decrease in the M-CR
distance.
(63) Wrackmeyer, B.; Horchler, K. In Annual Reports on NMR Spectroscopy;
Webb, G. A., Ed.; Academic Press: London, 1990; Vol. 22, pp 249-250.
(64) Sebald, A. In AdVanced Applications of NMR to Organometallic Chemistry;
Gielen, M., Willem, R., Wrackmeyer, B., Eds.; Wiley & Sons: Chichester,
UK, 1996; pp 124-157.
(65) Performed by Prof. Peter W. Stephens at the State University of New York,
Stony Brook, NY.
(66) (a) Chieh, P. C. J. Chem. Soc. A 1971, 3243. (b) Belsky, V. K.; Simonenko,
A. A.; Reikhsfeld, V. O.; Saratov, I. E. J. Organomet. Chem. 1983, 244,
125. (c) Preut, H.; Huber, F. Acta Crystallogr. C 1993, 49, 1373.
(67) Allen, F. H.; Kennard, O.; Watson, D. G.; Brammer, L.; Orpen, A. G.;
Taylor, R. J. Chem. Soc., Perkin. Trans 2 1987, S1.
(68) Riggs-Gelasco, P. J.; Mei, R.; Ghanotakis, D. F.; Yocum, C. F.; Penner-
Hahn, J. E. J. Am. Chem. Soc. 1996, 118, 2400.
In all cases, there is clear evidence of additional outer-shell
scattering, indicative of M‚‚‚C interactions, for which the
EXAFS distance determinations will be less accurate. For the
tetraphenyl species, the outer-shell scattering is readily under-
stood as arising from the phenyl carbons located farther than
the ipso carbon. Fits to these data require the use of multiple-
scattering corrections. Data fits obtained by constraining the
phenyl ring to the known geometry and allowing only M‚‚‚C
distances and Debye-Waller factors to vary were in excellent
9
J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004 12037

A R T I C L E S
Zharov et al.
agreement with crystallographic data and provided confidence
that the outer-shell fits for the carboranyl salt data should also
be reliable.
For all three cations, the longer M‚‚‚C distances (2.5-3.0
Å) can only be interpreted as the distance between the Me3M+
-
cation and the carbon atom of a methyl group of the CB11Me12
anion with which the cation interacts. These distances are longer
than the normal M-C bond lengths but are much shorter than
the sum of van der Waals radii of Ge (2.0 Å69), Sn (2.17 Å70),
or Pb (2.2 Å71) and of a methyl group (2.0 Å69). Clearly, the
cation-anion interactions in 4, 5, and 6 are significant. We shall
see below that this is further supported by the observation of
features attributable to M‚‚‚CsB scattering in 4 and 6, and by
the observation of a further outer-shell Ge‚‚‚C scattering in 5
that must arise from an additional cation-anion interaction.
EXAFS of Solid Me3Sn+CB11Me12- (4) and Me3Pb+CB11-
Me12 (6). The first of the two outer-shell peaks seen for the
Figure 3. EXAFS (inset) and corresponding Fourier transforms of n-Bu3-
Sn+CB11Me12 before (blue) and after (green) exposure to air, and for Me3-
Sn+CB11Me12 (red).
-
Sn atom in 4 and the Pb atom in 6 is much more intense than
the second. For 4, the data can be fitted using a single outer-
shell distance of approximately 3.03 Å, with an additional weak
peak at R + R ≈ 4.1 Å that can be modeled as the scattering
from the B atom in a linear Sn‚‚‚CH3sB unit (R is the “EXAFS
phase shift”, whose typical magnitude is 0.4 Å). Although the
R + R ≈ 4.1 Å peak is weak and the Sn‚‚‚B distance thus is
not well defined, the apparent Sn‚‚‚B distance of 4.63 Å would
be consistent with a quite reasonable C-B distance of 1.60 Å.
However, closer examination of the range of possible fits for
4 (see Table S3) showed that, in fact, a variety of nearly identical
fits are possible. All have an Sn-CR distance of 2.12 Å but
have apparent Sn‚‚‚C distances ranging from 2.76 to 3.03 Å,
depending on the starting parameters of the fit. This variability
suggested that additional parameters were required to model
the data. Consistent with this, the Fourier transform of the fit
was not a good match to the data. In particular, the Fourier
transform contained a second outer-shell peak, even though only
a single Sn‚‚‚C distance was used in the fit. A better fit was
obtained if the Sn‚‚‚C shell was split into two Sn‚‚‚C shells
(Table S3 and Figure S2, bottom). These fits were robust with
respect to starting parameters, consistently giving distances of
2.77 and 3.02 Å. Careful examination of the different contribu-
tions to the asymmetric Sn‚‚‚C fit shows that the two different
Sn‚‚‚C shells have EXAFS contributions that are out-of-phase
over much of the range of the data (red and green lines in Figure
S2, bottom). The resulting destructive interference means that
only a single Sn‚‚‚C peak is seen in the Fourier transform,
although two Sn‚‚‚C distances are refined. This destructive
interference greatly complicates the interpretation of the data.
exposed to air. While the two intact samples have similar Fourier
transforms, it is nevertheless clear from a close examination
that there are small but significant differences in the Sn
environments. The nearest-neighbor Sn-CR distance in n-Bu3-
Sn+CB11Me12 is somewhat longer than that in 4, while the
Sn‚‚‚C distance appears to be much shorter than that in 4, as
indicated by the lower R value for the second peak in the Fourier
transform. The change in the Sn-CR distance is confirmed by
quantitative fits, which give a value of 2.16 Å (0.04 Å longer
than that in 4, but still close to the usual70 2.14 Å Sn-C bond
length). The data gave an apparent average Sn‚‚‚C distance of
2.76 Å, nearly identical to the shorter of the two Sn‚‚‚C distances
in 4 and close to the average Sn‚‚‚C distance found crystallo-
graphically. No longer Sn‚‚‚C EXAFS signal is seen in n-Bu3-
Sn+CB11Me12; there is a very small peak that could correspond
to a ca. 3.0 Å Sn‚‚‚C distance, but this peak is near the noise
level of the data. The absence of a detectable ca. 3.0 Å Sn‚‚‚C
signal may be due, at least in part, to interference from signals
expected from the three Câ carbon atoms in the n-butyl chains,
located ∼3.0 Å away from the Sn atom. Thus, taking the short
Sn‚‚‚C distance as a measure of pyramidalization, it would
appear that the Sn atom in 4 has hybridization very similar to
that in n-Bu3Sn+CB11Me12. This is reassuring, given that the
119Sn NMR chemical shifts in the two solids are very similar
(466.2 ppm in 4 versus 461.2 ppm72 in n-Bu3Sn+CB11Me12).
The discrepancy between the Sn-CR bond length in
n-Bu3Sn+CB11Me12 found from the single-crystal X-ray struc-
ture,10 ∼2.07 Å, and the value of 2.16 Å found presently from
EXAFS is probably in part due to librational effects and in part
due to the low accuracy of the X-ray determination, which was
performed on a disordered crystal. We consider the EXAFS
value to be the more accurate. In contrast, the two methods
agree quite well with regard to the long axial Sn‚‚‚C distances.
For 6, the outer-shell Pb‚‚‚C distance is 2.9 Å, and once again
this is accompanied by a very weak distant peak that can be
modeled by a linear Pb‚‚‚CsB interaction. As for 4, the best
fit to this additional interaction gives a reasonable B-C distance
of 1.60 Å.16 Unlike the data for 4, the data for 6 do not show
any evidence for asymmetric Pb‚‚‚C coordination. The data can
To test the accuracy of the fit for 4, EXAFS data were also
collected for n-Bu3Sn+CB11Me12-, whose single-crystal X-ray
structure was determined earlier.10 Because of disorder, the
crystallographic data for n-Bu3Sn+CB11Me12 were of low
accuracy but nevertheless showed that the tin atom has a
distorted trigonal bipyramidal structure, with a Sn-CR bond
length of ∼2.07 Å, and two weak axial Sn‚‚‚C interactions, with
an average distance of 2.81 Å. The Fourier transforms of the
EXAFS data for n-Bu3Sn+CB11Me12 and for Me3Sn+CB11Me12
are compared in Figure 3, together with that for a sample
(69) Bondi, A. J. Phys. Chem. 1964, 68, 441.
(70) Davies, A. G. Organotin Chemistry; VCH: Weinheim, 1997.
(71) Batsanov, S. S. Russ. J. Gen. Chem. 1998, 68, 495.
(72) Zharov, I. Ph.D. Dissertation, University of Colorado at Boulder, 2000; p
53.
9
12038 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004

-
Metal Cation
−
Methyl Interactions in CB11Me12 Salts
A R T I C L E S
Figure 5. Possible structures of 5 proposed on the basis of EXAFS results.
myl cation in 5. If there were a fifth group coordinated to the
germanium atom on the back side, its carbon atom would
presumably be much farther than 2.50 Å from the Ge, and
probably not detectable by EXAFS.
However, it is clear from Figure 2 that there is another outer-
shell peak for 5 at ∼3.1 Å. This peak does not appear to result
from a backside methyl group because it is relatively intense.
EXAFS amplitudes fall off as 1/R2; therefore, given that this
peak is nearly 1.5 times as far away, it would be expected to
be less than half as intense. Another possible explanation is
that this peak is due to a Ge‚‚‚B interaction from a linear
Ge‚‚‚CH3sB structure. A linear arrangement would increase
the amplitude of the outer-shell EXAFS due to multiple
scattering and is probably observed for 4 and 6, as noted above.
However, for 5, the apparent Ge‚‚‚B distance (∼3.5 Å) would
be too short to be consistent with the observed Ge‚‚‚C distance
and the expected BsC distance. The remaining possibility is
to attribute the ∼3.1 Å peak to a shell of Ge‚‚‚C at an average
distance of 3.63 Å, and this can only originate from one or two
of the methyls in the associated CB11Me12- anion (7, or 7 and
8). It requires the Ge-C-B angle to be bent away from 180°,
and the absence of linearity in the Ge‚‚‚CH3sB moiety then
accommodates naturally the absence of detectable Ge‚‚‚B
EXAFS. If the empty metal orbital is pointing in the expected
direction, equal tilting toward two methyls (toward a face-
Figure 4. Possible structures of 4 and 6 proposed on the basis of EXAFS
results.
be modeled perfectly well assuming symmetric ligation by two
Pb‚‚‚C scatterers at 2.9 Å, and attempts to fit the data using
two shells of Pb‚‚‚C scatterers gave unrealistic negative Debye-
Waller factors (Table S3 and Figure S3). It is possible that the
failure to resolve two different Pb‚‚‚C distances in 6 is a result
of the shorter k range (and thus lower resolution) of the data.
However, symmetric ligation by two M‚‚‚C interactions in 6
would be consistent with apparent increase in M‚‚‚C distance
for 6 and would imply a nearly symmetric, trigonal bipyramidal
arrangements for the Me3Pb+ cation with rather weak coordina-
tion to the two axial methyls provided by the CB11Me12- anions
(Figure 4). Still, it is difficult to rule out entirely the possibility
that the Me3Pb+ cation has a strongly unsymmetric structure
similar to that proposed for Me3Sn+. In this case, it is likely
that only the shorter of the axial distances is reflected in the
EXAFS.
EXAFS of Solid Me3Ge+CB11Me12 (5). The outer-shell
centered coordination position on the CB11Me12 anion) is
-
-
scattering is most intense for the germanium compound 5, which
shows two well-defined outer-shell peaks of comparable am-
plitude. The shorter of these, which appears at ∼ 2.1 Å in Figure
2, is well modeled with a Ge‚‚‚C distance of 2.49 Å. The
EXAFS data would be consistent with either one or two axial
interactions at 2.5 Å, representing a distorted tetrahedral and a
trigonal bipyramidal structure, respectively. Distinguishing
between these two possibilities by EXAFS alone is extremely
difficult due to the correlation between the apparent coordination
number and the Debye-Waller factor. However, the unusually
short axial distance of 2.5 Å is inconsistent with a trigonal
bipyramidal structure for the cation in 5. There is insufficient
space around the Me3Ge plane to permit coordination to two
anions, each at 2.5 Å. If the germyl cation had this five-
coordinate arrangement, the axial coordination sites would be
expected to have significantly longer Ge‚‚‚C distances. There-
fore, a distorted tetrahedral geometry with three short interac-
tions and one long interaction is proposed for the trimethylger-
unlikely due to the resulting short distance (4.0 Å) between at
least one of the methyl groups in the Me3Ge+ cation and the
methyls of the anion (structure A in Figure 5). Such a tilt
requires a strong bending with a Ge‚‚‚CH3sB angle of 122°.
We therefore consider a tilt toward one methyl group (toward
an edge-centered coordination position on the CB11Me12- anion)
more likely (structure B in Figure 5, with a 4.7 Å closest distance
between the Me3Ge+ methyls and those of the anion, and a
bending angle of 130°). This tilt produces a five-membered
(Ge‚‚‚CH3sBsBsCH3‚‚‚) ring, which might be rigid enough
to give a detectable Ge‚‚‚CH3 EXAFS peak at 3.6 Å.
-
Computations for the Me3M+CB11Me12 Ion Pairs. Al-
though the structures of ion pairs will not be easily related to
the structures of the solids, and although interactions other than
classical electrostatic forces are not easily calculated with any
accuracy between ions of the size in question, we considered it
useful to perform calculations on the ion pairs to obtain some
insight into the cation-anion interactions in the salts 4-6. First,
9
J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004 12039

A R T I C L E S
Zharov et al.
-
Table 2. Calculated (B3LYP/SDD) Energies and Structures of Isolated Me3M+CB11Me12 Ion Pairs
coordinated methyl group
-
-
Me3Sn+CB11Me12
Me3Ge+CB11Me12
12 (B)
7 (B)
2 (B)
1 (C)
12 (B)
7 (B)
2 (B)
1 (C)
rel E, kcal/mol
q(Me3M+), e
d(M‚‚‚C), Å
d(C-B), Åa
MCB, deg
0
2.4
8.8
20.5
+0.82
3.07
1.50b
131
0
2.8
10.0
+0.70
2.25
1.72
173
22.4
+0.78
2.89
1.50
133
+0.74
2.42
1.72
180
+0.75
2.43
1.71
176
+0.78
2.48
1.68
171
+0.62
2.14
1.85
180
+0.64
2.17
1.81
177
∑
CRMCR, deg
BCH , deg
347
96
348
96
350
99
354
340
86
341
89
345
95
353
111
111c
c
a Average d(C-B) in free 2 is 1.62 Å. b d(C-C).
CCH.
we are interested in the competition among vertex-centered,
edge-centered, and face-centered coordination and the relative
coordinating ability of the four inequivalent vertex types in the
-
CB11 cage, and we are looking for guidance concerning the
nature of the cation-anion interactions. Second, we compare
the interactions with the three different metal cations.
The bulk of our calculations were performed by density
functional theory (DFT), a method that describes attractive van
der Waals interactions poorly or not at all (Tables 1 and 2).
We were only able to afford a calculation by the MP2 method
-
for one of the isomeric ion pair structures of Me3Ge+CB11Me12
.
This procedure is known to somewhat overestimate attractive
van der Waals interactions. In our case, it gives an optimized
structure very similar to that obtained from the DFT calculation,
except that the incipient SE2 substitution on the methyl carbon
has not proceeded quite as far along the reaction coordinate.
This result is reassuring in that it suggests that van der Waals
attractions do not play a dominant role in determining the ion
pair structure. Nevertheless, while we believe that the results
of the computations are quite dependable when it comes to the
structure and properties of the individual partner ions, when it
comes to inter-ion interactions, they are best viewed as
qualitative guides rather than definitive statements.
Relative Coordinating Ability of Methyl Groups in
CB11Me12-. Considerable computational effort was expended
in a search for potential energy minima in which the cation
would be located above a BB edge such as 7-12 or 2-7, or
above a BBB triangular face such as 7,8,12. The latter type of
coordination is known for the case of toluene-Li+ 73 and
phosphine-Ag+ 9 salts. No edge-centered or face-centered
minima were found in the present case, and in the end, careful
geometry optimization always led to one of the minima in which
the metal atom lies above a B vertex. This is the type of
coordination we concentrate on in the following.
Figure 6. Calculated (B3LYP/SDD) structures of Me3Sn+CB11Me12- ion
pairs coordinated to positions 1 (A), 12 (B), 2 (C), and 7 (D) of the
-
CB11Me12 anion.
in position 7 of the carboranyl anion. Our calculated (B3LYP/
SDD) ion-pairing energies of the four separately optimized
-
isomers of an isolated Me3Sn+CB11Me12 (Figure 6) and an
isolated Me3Ge+CB11Me12 ion pair with the cation next to
-
positions 1, 2, 7, and 12 favor the association with the 12-methyl
group over that with the 7-methyl group, but not by a wide
margin (Table 2). The interaction of the cation with the 2-methyl
group is considerably weaker. In all three cases, the metal atom
prefers to lie almost exactly on the local approximate five-fold
axis of the icosahedral anion. In contrast, the cations interact
much more weakly with the 1-methyl group, and the metal atom
is then tilted off the axial position by nearly 50° toward the
2-methyl group. It is not immediately obvious to what degree
these preferences are of purely electrostatic origin and reflect
an uneven distribution of charge in the carborane anion (Tables
3 and 4) and to what degree they are due to distinct coordinating
abilities of the individual vertices.
Electrostatic Interactions. We consider the classical, purely
electrostatic interactions first. These are computed fairly reliably
by any of the methods used. The charge distribution in the
anions 1 and 2 has been the subject of debate for some years.
We have now computed it at the B3LYP/6-31+G(d) level of
theory and present the results in Tables 3 and 4, respectively.
We list both the natural atomic charges defined by Weinhold75
and the Mulliken charges; note the vast superiority of the former,
Although many structures of salts of CB11H12- and its various
derivatives have been published over the years, the relative
coordinating ability of its inequivalent vertices is somewhat
-
uncertain. Electrophilic interactions with the vertices of CB11H12
are generally believed to decrease in the reactivity order 12 >
7-11 > 2-6 . 1, but complexes with coordination both
through positions 1274 and 742 have been reported. Recently,
single-crystal structures of several phosphine-Ag+ 1-H-CB11-
Me11- salts have been described,8,9 and in some the phosphine-
stabilized silver cation is strongly coordinated to a methyl group
(73) King, B. T.; Noll, B. C.; Michl, J. Collect. Czech. Chem. Commun. 1999,
64, 1001.
(74) Crowther, D. J.; Borkowsky, S. J.; Swenson, D.; Meyer, T. Y.; Jordan, R.
F. Organometallics 1993, 12, 2897, and references therein.
9
12040 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004

-
Metal Cation
−
Methyl Interactions in CB11Me12 Salts
A R T I C L E S
-
Table 3. Calculated (B3LYP/6-31+G(d)) Naturala and Mulliken Atomic Chargesb for the Free CB11H12 Anion (1)
12
7
−11
2
−6
1
B
H
-0.161 (-0.077)
+0.055 (-0.006)
-0.106 (-0.083)
-66.07 (-59.51)
-67.06
-0.185 (-0.046)
+0.058 (-0.006)
-0.127 (-0.052)
-63.45 (-60.33)
-64.63
-0.022 (-0.047)
+0.049 (0.001)
0.027 (-0.046)
-59.04 (-63.52)
-59.76
-0.712 (-0.651)c
+0.322 (0.224)
-0.390 (-0.427)
-55.25 (-65.03)
-50.69
vertexd
Coul attre
elst pot.f
a Reference 66. b In elementary charge units; Mulliken charges in parentheses. c Cage carbon atom charge. d Sum of charges on B (or C) and H atoms at
each vertex. e Coulombic attraction (kcal/mol) between a +1 point charge, placed 2.5 Å from the hydrogen atom of a vertex on the B-H (C-H) axis, and
natural (Mulliken) atomic charges in the anion. f Electrostatic potential (kcal/mol) at the point described in footnote d.
-
Table 4. Calculated (B3LYP/6-31+G(d)) Naturala and Mulliken Average Atomic Chargesb for the Free CB11Me12 Anion (2)
position
12
7
2
1
B
+0.102 (-0.757)
-0.911 (-0.987)
+0.236 (+0.206)
-0.204 (-0.369)
-0.101 (-1.126)
-64.53 (-76.53)
-65.75
+0.065 (+0.408)
-0.907 (-1.009)
+0.236 (+0.205)
-0.200 (-0.392)
-0.134 (0.015)
-62.69 (-59.39)
-63.96
+0.253 (+0.880)
-0.929 (-1.032)
+0.241 (+0.211)
-0.207 (-0.398)
+0.046 (+0.482)
-59.55 (-59.99)
-59.15
-0.548 (-2.063)c
-0.639 (-0.996)
+0.243 (0.232)
+0.090 (-0.298)
-0.458 (-2.362)
-52.48 (-86.85)
-50.39
C (methyl)
H
CH3
vertexd
Coul attre
elst pot.
a Reference 66. b In elementary charge units; Mulliken charges in parentheses. Because of the different disposition of the methyl groups at otherwise
equivalent vertices, the charges on the latter vary slightly. Average values are given. c Cage carbon atom charge. d Sum of charges on B, C, and H atoms
at each vertex. e Coulombic attraction (eV) between a +1 point charge, placed 2.5 Å from the methyl group carbon atom of a vertex on the B-C (C-C)
axis, and all natural (Mulliken) atomic charges in the anion.
-
Table 5. Calculated (B3LYP/SDD) Natural Atomica Chargesb for Isolated Me3M+CB11Me12 Ion Pairs
coordinated methyl group
-
-
Me3Sn+CB11Me12
Me3Ge+CB11Me12
position
12
7
2
1
12
7
2
1
c
CH3
12
-0.29
-0.28
-0.29
+0.08
-0.33
-0.28
-0.29
+0.07
-0.29
-0.28
-0.30
+0.05
-0.29
-0.28
-0.31
-0.11
-0.25
-0.27
-0.28
+0.08
-0.32
-0.27
-0.29
+0.07
-0.29
-0.28
-0.25
+0.05
-0.28
-0.28
-0.31
-0.07
7-11
2-6
1
vertexd
12
-0.05
-0.14
+0.14
-0.67
-0.12
-0.13
+0.13
-0.67
-0.07
-0.13
+0.13
-0.70
-0.06
-0.13
0.11
-0.05
-0.14
+0.14
-0.67
-0.11
-0.11
+0.13
-0.67
-0.06
-0.13
+0.17
-0.70
-0.05
-0.13
+0.12
-0.85
7-11
2-6
1
-0.89
a Reference 66. b In elementary charge units. c Average sum of charges on methyl C and H atoms at each vertex. d Average sum of charges on B, C, and
H atoms at each vertex.
which reproduce the calculated electrostatic potential at the
likely locations of the counterions within a few kilocalories per
mole, whereas the latter yield numbers some of which are off
by dozens of kilocalories per mole. We use only the former in
further discussion. The charges calculated for ion pairs of 2 with
Me3Sn+ and Me3Ge+ remain almost the same regardless of the
cation position (Table 5), showing that the anion is not very
polarizable and that it is sensible to base the discussion on the
charge distribution in the isolated anions.
Two qualitative arguments concerning the charge distribution
in 1 and 2 come to mind. At first sight, the general tendency of
positive counterions to associate with positions 7-12 in the
known crystal structures and the preponderance of electrophilic
substitution in these positions suggest that most of the negative
charge is located in positions distant from the cage carbon atom.
Yet, one would expect the most electronegative atom in the
cage, the carbon in position 1, to carry most of the negative
charge.
tronegativities of carbon and boron, in the free CB11H12- anion
(1)76 most of the calculated negative charge resides on vertex
1, which carries 3 times the negative charge of the most negative
boron vertex. Vertices 2-6 are nearly electroneutral. Each single
vertex in the group 7-12 is weakly negative, but together, they
carry about three-fourths of a negative elementary charge.
Interestingly, vertex 12 is slightly less negative than vertices
7-11. Within each vertex, the charge is distributed unevenly.
The hydrogens in positions 2-12 are weakly positive, and the
negative charge on these vertices is carried by the boron.
Together, the boron atoms in positions 7-12 carry little over 1
unit of negative charge. The carbon-bound hydrogen in position
1 is positive (and is also known to be fairly acidic).
-
In the CB11Me12 anion (2), the charge distribution among
the vertices is similar, with the carbon vertex 1 even more
strongly negative. The 1-methyl group is nearly electroneutral,
whereas the 2-12 methyls are strongly and equally negative.
The positive charges on the boron atoms, large negative charges
on the methyl carbons, and positive charges on the methyl
hydrogens remind us that, on boron, methyl is a σ-withdrawing
and π-donating substituent.
The charges listed in Tables 3 and 4 neatly resolve this
apparent contradiction. In agreement with the relative elec-
(75) NBO, Version 3.1: Glendening, E. D.; Reed, A. E.; Carpenter, J. E.;
Weinhold, F. Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988,
88, 899.
(76) (a) Zahradn´ık, R.; Balaji, V.; Michl, J. J. Comput. Chem. 1991, 12, 1147.
(b) McKee, M. L. J. Am. Chem. Soc. 1997, 119, 4220.
9
J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004 12041

A R T I C L E S
Zharov et al.
-
Figure 8. Electrostatic potential (B3LYP/6-31+G(d)) around CB11Me12
.
The 5 Å long color-coded arrows are placed 1.2 Å above the carbon atom
of methyl groups in positions 1, 2, 7, and 12. One arrow is placed between
positions 1 and 2, showing that in this region the electrostatic potential is
higher than that directly above the methyl group in position 1.
the B-C bond at the vertex adjacent to the cation Me3Sn+
(Me3Ge+) is 0.1 Å (0.2 Å) longer than the others, the
coordinated methyl group is flattened, and the cation is distinctly
pyramidalized (Table 2). The geometrical distortions calculated
for the interaction of the cations with the 1-methyl group of
the anion are quite different. Compared to the other isomers,
the M‚‚‚C distance is much bigger (for Ge, by ∼0.7; for Sn, by
∼0.6 Å), the cage-CH3 bond is only slightly stretched (by
0.02 Å), the methyl group remains tetrahedral (∑ HCH is
smaller by ∼31° for Ge and ∼36° for Sn), and the Sn+ center
is less pyramidalized (∑ CRMCR is larger by ∼11° for Ge and
∼6° for Sn). The Me3M+ group is tilted toward the methyl in
position 2 ( CCGe ) 133°, CCSn ) 131°), whereas for the
other three isomers the metal atom M lies in a direction very
close to that of the B-C bond at the vertex. The computed
degree of electron density shift from the anion to the cation is
much less in the 1-methyl isomer than in the others.
The calculated geometrical distortions suggest that the
covalent interaction is of the type expected if the cation
interacted with the carbon atom of the methyl groups in positions
2-12, and with two of the hydrogen atoms of the methyl group
in position 1. When the 12, 7, and 2 isomers of the Me3Ge+ or
Me3Sn+ ion pairs are compared, those with the cation in position
12 have the shortest M‚‚‚CH3 and the longest BsCH3 distance,
the most pyramidalized Me3Sn+ cation, the flattest methyl group,
and the most inter-ion charge transfer (Table 2). Judging by
the nature of the geometrical distortions, the empty orbital of
the metal atom removes electron density from the back lobe of
carbon orbital that is used to attach the methyl group to the
cage boron atom. In contrast, the interaction between the metal
cation and the 1-methyl group seems to involve the CH bonds
of the methyl.
Figure 7. Sphere: Electrostatic potential (B3LYP/6-31+G(d)) around
-
-
CB11H12 (1, left) and CB11Me12 (2, right). The radii of the spheres are
5.2 and 5.7 Å, respectively, 2.5 Å from the hydrogen atoms (1) or methyl
carbon atoms (2). Views along exocyclic bonds; from top to bottom, looking
at positions 1, 2, 7, and 12.
Thus, in both cases, vertex 1 carries not only close to half of
all the negative charge but also a significant local dipole whose
positive end is pointed outward. This makes it difficult for a
positive counterion to derive much stabilization from interaction
with the negative charge on the carbon vertex by approaching
it closely, and it is better off when located at the opposite end
of the cage. This becomes clear when all Coulombic interactions
between the cation and the anion are added quantitatively. This
can be done crudely by adding pairwise interactions between
atomic point charges calculated by the Weinhold procedure, or
accurately by computation of the electrostatic potential, with
similar results (Tables 3 and 4). The relative preferences are
not very different in 1 and 2.
The electrostatic potentials around 1 and 2 calculated for the
surfaces of spheres of a radius appropriate for the Me3Sn+
counterion are presented in Figures 7 and 8. These graphical
presentations provide a particularly instructive picture of the
strong deviation of the electrostatic potential around these anions
from spherical symmetry. In both instances, a clear preference
is given to the location of the positive point counterion in
position 12, closely followed by those in positions 7-11, and
distantly by positions 2-6 and particularly 1.
Covalent Interactions. The calculated structures of the
partner ions leave little doubt that there is more to the interaction
than mere electrostatics. The presence of true chemical bonding
interactions in the computed structures is not in doubt for the
isomers in which the cation is located in positions 2-12, as
The purely electrostatic and the covalent energetic preferences
thus coincide, and both favor position 12, followed by 7, 2,
and 1. A comparison of the relative DFT energies of the four
isomeric Me3Sn+CB11Me12- and Me3Ge+CB11Me12- ion pairs
(Table 2) with the relative energies of the purely electrostatic
9
12042 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004

-
Metal Cation
−
Methyl Interactions in CB11Me12 Salts
A R T I C L E S
Figure 9. Calculated (MP2/6-31G(d) and B3LYP/SDD) structures of Me3M+CB11Me12- (M ) Ge, Sn, Pb) ion pairs with coordination to the methyl group
in position 12 of the carborane.
interactions (Table 4) suggests that the electrostatic effects are
responsible for two-thirds to three-fourths of the preference and
the covalent effects for the rest.
Relative Coordinating Ability of the Metal Atoms in
Me3M+. A comparison of the DFT optimized structures
computed for the ion pairs 4, 5, and 6 is provided in Figure 9
and Tables 1 and 2. The most notable results are the following:
(a) the M‚‚‚CH3sB arrangement is nearly linear; (b) the cations
are pyramidalized and the coordinating 12-methyl group of the
anion is flattened; (c) the B(12)-C(12) bond of the anion is
elongated; and (d) the M‚‚‚C(12) distance is much longer than
in normal covalent MsC bonds but much shorter than the sum
of van der Waals radii of the metal and of a methyl group. The
structural distortions from the free ion structures are calculated
to diminish in the series Ge, Sn, Pb. This is also the order of
decreasing inter-ion charge transfer (Table 2) and agrees with
the conclusion drawn from the solid-state 13C NMR chemical
shifts, namely that the positive charge on CR increases in the
series of Me3Ge+, Me3Sn+, and Me3Pb+. The calculated
-
structure of Me3Ge+CB11Me12 seems to be approaching the
transition-state geometry for backside SE2 substitution on the
carbon of the CH3 group.
The MP2 calculated structure of Me3Ge+CB11Me12- is very
similar to that described above, but the B(12)-C(12)-Ge angle
is 171°, the germyl cation is less pyramidalized, and the B(12)-
C(12) bond is less extended, as if the intended SE2 substitution
on the methyl carbon were somewhat less advanced. The methyl
group coordinated to the germyl cation is still essentially planar,
as would be expected in the transition state.
Comparison of the Computed Structures of Ion Pairs 4-6
with the EXAFS Structures of Solids (Table 1). The computed
structures of the ion pairs are not easily extrapolated to the solid
structures. In addition to the purely electrostatic interactions with
the other ions present, in the solid both the anion and the cation
are capable of specific interactions with more than one coun-
terion, as suggested by the published10 crystal structure of n-Bu3-
Sn+CB11Me12-. As expected, the calculated M-C(12) distances
in the ion pairs are much shorter than those found by EXAFS
for the solids 4, 5, and 6. Optimization on the ion triple
[CB11Me12-Me3Sn+CB11Me12-] with coordination through po-
sitions 2 and 9 (equivalent to 7) of the anion (Figure 10) at the
B3LYP/SSD level resulted in a structure with the three
-
Figure 10. Calculated (B3LYP/SDD) structure of the [CB11Me12
Me3Sn+CB11Me12-] ion triple.
constituents nearly coaxial. The two inequivalent Sn‚‚‚C
distances are 2.74 Å (9-Me) and 2.98 Å (2-Me), now in nearly
perfect agreement with observations on the solid (Table 1). The
cation is calculated to be pyramidalized less than observed in
the crystal of n-Bu3Sn+CB11Me12- (Sn atom 0.08 Å out of the
CR plane; observed, 0.32 Å). This is in line with the conclusion
drawn above from the EXAFS results.
Except for the obvious difference in the Sn-CR and Pb-CR
bond lengths, well reproduced by the calculations, the Me3Sn+
salt 4 and the Me3Pb+ salt 6 are calculated and observed to be
-
very similar (Table 1). Overall, EXAFS of Me3Sn+CB11Me12
(4) and Me3Pb+CB11Me12- (6) suggests that, in these solid, each
cation is coordinated to two anions in infinite columns similar
- 10
to those found in the single crystal of n-Bu3Sn+CB11Me12
.
9
J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004 12043

A R T I C L E S
Zharov et al.
chemical shift in the Pb complex were not successful. Mass spectra
were recorded with a Hewlett-Packard 5989 API/ES/MS. An HPLC
system employing a reverse-phase C18 column (250 × 4.6 mm, 5 µm)
with methanol/water (containing a 1% AcOH/0.7% Et3N buffer) as the
mobile phase was used for monitoring reactions, while larger columns
of the same phase were used for semipreparative separations. All volatile
compounds were analyzed using Varian 3400 analytical GC with a
0.2 mm × 25 m, 0.33 µm RSL-150 5% cross-linked silica capillary
column. Their separations were performed using Varian 3400 prepara-
In that case, assuming additivity of the coordination energies,
our calculations suggest that it is energetically more favorable
by 10.6 kcal/mol for the cations to coordinate to the antipodal
positions 2 and 9 of the anion, rather than the antipodal positions
1 and 12, even if it leaves the best coordinating methyl group
in position 12 unused. Although the calculations are for ion
pairs and observations for a solid lattice, the trends in the two
are similar. The distance from the metal to the carbon of the
coordinated methyl group is computed to be ∼0.2 Å longer in
the Pb-containing free ion pair than in the Sn-containing free
ion pair. The difference of the distances observed in the solid
is a little over 0.1 Å (counting the closer methyl group in 4).
While the EXAFS results for the outer shells in 4 and 6 are
very similar, there is no doubt that the structure of Me3Ge+-
CB11Me12- (5) is quite different. Instead of forming an unsym-
metrical trigonal bipyramid, the Me3Ge+ cation is coordinated
primarily to one and more weakly to one (or, less likely, two)
other methyl group of a single anion. Our calculations suggest
that, in a free ion pair, the germanium would prefer coordination
to a single methyl group on an anion, but in the solid it is
apparently advantageous to coordinate to two; in that case the
primary coordination is expected to be to position 12, and the
weaker coordination to position 7. We have no information on
how these ion pairs pack into the solid lattice. The very
noticeable observed reduction in the Ge‚‚‚CH3 distance relative
to Sn‚‚‚CH3 (by ∼0.3 Å) and Pb‚‚‚CH3 (by 0.4 Å) is quite nicely
reproduced by the calculations and suggests that the coordination
of the Me3Ge+ cation is much stronger than that of Me3Sn+
and Me3Pb+. Presumably, this provokes a much stronger
pyramidalization at the Ge+ atom and sterically precludes any
stabilizing coordination from its back side.
1
tive GC with a /4 in. × 21 ft 5% OV-7 80/100 Chromosorb GHP
packed column. Individual compounds were purified to >99.5% purity
(by analytical GC).
EXAFS Experiments. All carborane species were synthesized and
sealed under a dry argon atmosphere until use. Tetraphenylmetal
compounds were purchased from Aldrich and used without further
purification. Sample powders were ground with dry BN and packed
evenly into a 1 mm thick sample cell with polypropylene windows.
All sample manipulation was performed in a glovebox under a dry
nitrogen atmosphere.
X-ray absorption data were collected at Stanford Synchrotron
Radiation Laboratory (SSRL). Data collection details are summarized
in Table S1 (see Supporting Information). Absorption data were
collected in transmission mode using N2-filled ion chambers for Ge
and Pb, and Ar-filled ion chambers for Sn. X-ray energies were
calibrated using an element foil as an internal standard, with the first
inflection point of the foil defined as 11 103 (Ge, K-edge), 26 714 (Sn,
K-edge), or 13 038 eV (Pb, L3-edge). A Si (220) double-crystal
monochromator was used in all cases.
Data were analyzed using EXAFSpak, written by G. N. George at
SSRL. Averaged absorbance data were processed using a second-order
polynomial pre-edge subtraction, a k3-weighted, four-region, third-order
spline for extended X-ray absorption fine structure (EXAFS) back-
ground removal, and were scaled using a Victoreen background
function. The EXAFS formula used for data fitting is
Experimental Part
n
NiFi(k,r)Si(k)
ø(k) )
sin(2kr + φtot)
∑
General. Standard Schlenk and glovebox techniques were used for
handling air-sensitive reagents. Pentane, hexane, cyclohexane, meth-
ylcyclohexane, benzene, Me6Si2, Me6Sn2, Me6Ge2, and Bu6Sn2 were
dried over CaH2, distilled, and stored under Ar. The solvents were stored
over a potassium mirror. The following compounds were purchased
and used as received: Ph4Ge, Ph4Sn, Ph4Pb, Me3SnPh, PhLi, t-BuLi,
ethyl triflate, calcium hydride, cesium chloride, and silver nitrate.
kr2
i)1
where Ni is the number for scatterer i at the distance of r to the
photoabsorber, Fi(k,r) is the backscattering amplitude, Si(k) is a scale
factor, φtot is the total phase shift of the scattering path, and the sum is
over all types of scattering atoms.
-
EXAFS data were fitted using amplitude and phase parameters
calculated with FEFF 7.02, and calibrated using EXAFS data for GePh4,
SnPh4, and PbPh4 (Table S2, Supporting Information), which gave for
the element threshold energies of 11 113.5 (Ge), 29 213.6 (Sn), and
13 036.8 eV (Pb). The scale factor was held constant at 0.9. These
parameters gave accurate first-shell fitting (see Table S2). These E0
values and scale factor were used to fit the k3-weighted data for the
carborane compounds. The bond distance and the Debye-Waller factor
for each shell were optimized until the square sum of the k3-weighted
differences between raw data and the FEFF model reached minimum.
The bond-valence sum (BVS) analysis78 was applied to each optimiza-
tion result as an accessory tool to choose the best fit. The bond valence
(s) is defined as
Me3NH+CB11H12 was purchased from Katchem Ltd. (Prague, Czech
1
Republic). Solution H, 11B, 13C, 29Si, 119Sn, and 207Pb NMR spectra
were measured with Varian XRS-300 and Varian Unity-500 spectrom-
1
eters. H and 13C chemical shifts were measured relative to the lock
solvent. 11B chemical shifts were measured relative to BF3OEt2, with
positive chemical shifts at higher frequency. B(OCH3)3 was used as an
external standard (18.1 ppm). 29Si, 119Sn, and 207Pb chemical shifts were
measured relative to the corresponding Me4E. Solid-state CP/MAS 11B,
13C, and 119Sn NMR spectra were obtained on a Chemagnetics CMX-
200 spectrometer using a Chemagnetics MAS probe equipped with a
7.5 mm PENCIL rotor. Due to the air sensitivity of the compounds,
the samples were sealed in short sections (approximately 15 mm) of 5
mm o.d. tubing. These tubes were then packed into the rotor surrounded
by KBr to aid in balancing of the samples for high-speed spinning.
Spin rates between 1 and 2.5 kHz were attained. The 11B spectra of
the three different salts (4, 5, 6) were all identical; therefore, no effort
was made to properly reference these spectra. The 13C spectra were
externally referenced using HMB, with the methyl resonance being
17.6 ppm on the TMS scale. The 119Sn spectrum was referenced using
the signal from tetracyclohexyltin, with the Sn resonance taken to be
-97.4 ppm on the SnMe4 scale.77 Attempts to measure the 207Pb
r0 - r
s ) exp
(
)
B
where B ) 0.37, r is the bond length, and r0 is an empirical value. For
a given metal ion, the BVS is the sum of bond valences for each bond
pair to the metal ion. The empirical value of r0 was determined to
minimize the difference between the BVS and the formal oxidation
(77) Harris, R. K.; Sebald, A. Magn. Reson. Chem. 1987, 25, 1058.
(78) Brown, I. D.; Altermatt, D. Acta Crystallogr. 1985, B41, 244.
9
12044 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004

-
Metal Cation
−
Methyl Interactions in CB11Me12 Salts
A R T I C L E S
state for crystallographically characterized compounds from the Inor-
ganic Crystal Structure Database (ICSD).
Me3Sn+CB11Me12- (4) was prepared at room temperature as a white
solid in 20 mL of pentane from 25 mg (0.08 mmol) of the CB11Me12
•
radical and 8.4 µL (13.2 mg, 0.04 mmol) of Me6Sn2 in 97% yield (37
mg). CP-MAS 11B NMR: δ -1.7 (B12), -11.8 (B2-11). CP-MAS 13C
NMR: δ 57.5 (1-C), 14.4 (1-CH3), 10.4 (CH3-Sn), -1.5 (2-12-CH3).
CP-MAS 119Sn NMR: δ 466.8. IR (solid): 582, 687, 773, 869, 950,
1068, 1143, 1247, 1285, 1365, 1461, 2846, 2865, 2939 cm-1. MS/
ES(+) in methanol: base peak at m/e 165 with the expected isotopic
distribution (Me3Sn+). MS/ES(-) in methanol: base peak at m/e 311
with the expected isotopic distribution (CB11Me12-). Anal. Calcd for
C16H45B11Sn: C, 40.45; H, 9.55. Found: C, 40.97; H, 8.99.
Calculations. Geometry optimizations were performed using the
Gaussian 98 program,79 charges were calculated using the NBO
method,75 and molecular modeling and visualization were performed
using Molden 3.680 and Spartan SGI Version 4.1.1 (Wavefunction, Inc).
The basis sets used were SDD (i.e., D95V for H, B, and C, and Stuttgart/
Dresden ECPs for Ge, Sn, and Pb),81 6-31G(d),82 and 6-31+G(d).82
All structures have been optimized with increased integral precision
(Int ) UltraFine) up to tight convergence. For all resulting geometries
the Hessian matrices were analyzed, and if negative eigenvalues
appeared, the geometry was changed until all eigenvalues remained
positive and the geometries corresponded to true potential energy
minima. For each of the four isomers of both Me3Ge+ and Me3Sn+
salts, the optimization was started from three different geometries, with
the cation placed against the B-CH3 (C-CH3) vertex, between two
adjacent vertices, and between three adjacent vertices. In all cases, the
geometries converged to the same structure. In six of them (three for
Me3Ge+ and three for Me3Sn+), the cation ended up placed against
essentially exactly against a boron vertex, with the B-C-M angle
nearly exactly equal to 180°. Only for the isomers with the Me3Ge+ or
Me3Sn+ cation located in position 1 was the C-C-M line significantly
bent, with the metal atom inclined away from its carbon vertex and
toward one of the adjacent B-CH3 vertices.
The 119Sn and 11B NMR spectra of 4 in C6D6, CD2Cl2, and CD3CN
were obtained by preparing 4 in an NMR tube in pentane under argon
atmosphere, removing pentane, adding the dry deuterated solvent, and
sealing the tube. For low-temperature NMR measurements, the sample
preparations were conducted at -78 °C, and the sealed and frozen NMR
tubes were transported to an NMR instrument.
•
An analogous reaction of 25 mg (0.08 mmol) of the CB11Me12
radical and 15.8 µL (18.8 mg, 0.08 mmol) of Et4Sn in 20 mL of pentane
resulted in a white solid after 1 h of stirring at room temperature. ES/
MS of this solid in methanol showed a base peak at m/e 206 (positive
mode) with the expected isotopic distribution (Et3Sn+), and base peak
at m/e 311 (negative mode) with the expected isotopic distribution
(CB11Me12-). A small peak at m/e 325 (negative mode) was also
detected (CB11Me11Et-).
Data for electrostatic potential plots were obtained from the Gaussian
cube file.
Me3Ge+CB11Me12- (5) was prepared at room temperature as a white
•
•
•
solid in 20 mL of pentane from 25 mg (0.08 mmol) of the CB11Me12
Oxidation with CB11Me12 sGeneral Procedures. The CB11Me12
radical and 8.0 µL (9.4 mg, 0.04 mmol) of Me6Ge2 in 96% yield (33
mg). CP-MAS 11B NMR: δ -1.5 (B12), -11.6 (B2-11). CP-MAS 13C
NMR: δ 57.6 (1-C), 14.4 (1-CH3), 5.6 (CH3-Ge), -1.7 (2-12-CH3).
IR (solid): 605, 707, 775, 879, 955, 1078, 1145, 1248, 1295, 1369,
1463, 2844, 2867, 2931 cm-1. MS/ES(+) in methanol: base peak at
m/e 119 with the expected isotopic distribution (Me3Ge+). MS/ES(-)
in methanol: base peak at m/e 311 with the expected isotopic
distribution (CB11Me12-). Anal. Calcd for C16H45B11Ge: C, 44.79; H,
10.57. Found: C, 45.21; H, 10.09.
radical is placed in a flame-dried ampule equipped with a stir bar and
a septum. The ampule is connected to a dry solvent reservoir and to a
vacuum/Ar line by thick-walled latex tubing. The system in evacuated
(10-3 mmHg), the vacuum line is disconnected, and the desired amount
of solvent is distilled into the ampule cooled in liquid nitrogen. The
system is then filled with argon, and the solution is warmed to the
desired temperature with stirring until all radical is dissolved. A material
to be oxidized is added dropwise with vigorous stirring either neat (dry,
degassed) using a syringe or as a dry, degassed solution in an alkane
solvent via a Teflon cannula. After the blue (pentane, hexane) or green
(cyclohexane, methylcyclohexane) color of the radical disappears, the
reaction is complete. Trapping reagents are added to the resulting
solution/solids via a syringe or a cannula. For further sample prepara-
tions, the solvent can be carefully removed from insoluble products
via cannula, or evaporated in the case of soluble products, after which
the ampule can be sealed. Solid-state NMR samples were prepared by
opening ampules in a glovebox, transferring their contents into an NMR
tube equipped with a vacuum connector, and sealing the tubes, first
off the vacuum line and then to the desired length on a lathe. EXAFS
samples were prepared in a glovebox by grinding the salts with dry
BN powder and packing the mixture evenly into a 1 mm thick sample
cell with polypropylene windows.
•
An analogous reaction of 25 mg (0.08 mmol) of the CB11Me12
radical and 15.2 µL (15.1 mg, 0.08 mmol) of Et4Ge in 20 mL of pentane
resulted in a white solid after 2 h of stirring at room temperature. ES/
MS of this solid in methanol showed a base peak at m/e 160 (positive
mode) with the expected isotopic distribution (Et3Ge+), and a base peak
at m/e 311 (negative mode) with the expected isotopic distribution
(CB11Me12-). A small peak at m/e 325 (negative mode) was also
detected (CB11Me11Et-).
Me3Pb+CB11Me12- (6) was prepared at room temperature as a white
•
solid in 20 mL of pentane from 25 mg (0.08 mmol) of the CB11Me12
radical and 17.0 µL (21.5 mg, 0.08 mmol) of Me4Pb in 93% yield (42
mg). CP-MAS 11B NMR: δ -1.5 (B12), -11.6 (B2-11). CP-MAS 13C
NMR: δ 57.6 (1-C), 31.4 (CH3-Pb), 14.7 (1-CH3), -1.8 (2-12-CH3).
IR (solid): 542, 667, 778, 867, 952, 1062, 1145, 1246, 1285, 1366,
1462, 2843, 2867, 2933 cm-1. MS/ES(+) in methanol: base peak at
m/e 253 with the expected isotopic distribution (Me3Pb+). MS/ES(-)
in methanol: base peak at m/e 311 with the expected isotopic
distribution (CB11Me12-). Anal. Calcd for C16H45B11Pb: C, 34.10; H,
8.05. Found: C, 34.66; H, 7.75.
(79) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Ortiz,
J. V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon,
M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.9; Gaussian,
Inc.: Pittsburgh, PA, 1998.
Conclusions
The oxidation of Me6E2 (M ) Ge, Sn) and Me4Pb with
•
CB11Me12 in pentane yields the insoluble salts Me3M+-
CB11Me12-, which have been characterized by chemical trap-
ping, NMR, EXAFS, and calculations. In these solids, the
cationic lead or tin center is coordinated to the methyl groups
of two CB11Me12- anions in a trigonal bipyramidal fashion, the
former most likely symmetrically and the latter unsymmetrically.
The cationic germanium center is coordinated strongly to one
(80) Schaftenaar, G.; Noordik, J. M. J. Comput.-Aided Mol. Design 2000, 14,
123.
(81) H, B, C: Dunning Jr., T. H.; Hay, P. J. In Modern Theoretical Chemistry;
Schaefer, H. F., III, Ed.; Plenum: New York, 1976; Vol. 3, pp 1-28. Ge,
Sn, Pb: Igel-Mann, G.; Stoll, H.; Preuss, H. Mol. Phys. 1988 65, 1321.
(82) H, C: Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972, 56,
2257. B: Dill, J. D.; Pople, J. A. J. Chem. Phys. 1975, 62, 2921. +: Clark,
T.; Chandrasekhar, J.; Schleyer, P. v. R. J. Comput. Chem. 1983, 4, 294.
d: Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
9
J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004 12045

A R T I C L E S
Zharov et al.
-
and weakly to another one or two methyls of a single CB11Me12
ion. When the methyl is carried by a carbon, its interaction with
the metal ion is quite different and primarily involves two of
the methyl hydrogens.
anion. Most of the interaction is electrostatic, but a quarter to
a third can be attributed to covalent bonding. The strength of
the metal-methyl interaction increases in the order Me3Pb+ <
Me3Sn+ , Me3Ge+, and the strength of both the electrostatic
and covalent interaction increases in the position order
Acknowledgment. This work was supported by the National
Science Foundation (CHE-0140478 and CHE-9709195) and by
the Ministry of Education of the Czech Republic (Center for
Complex Molecular Systems and Biomolecules, project
LN00A0032). The EXAFS data were measured at SSRL, a
national user facility operated by Stanford University on behalf
of the U.S. Department of Energy, Office of Basic Energy
Sciences, with additional support from the Department of
Energy, Office of Biological and Environmental Research, and
the National Institutes of Health, National Center for Research
Resources, Biomedical Technology Program.
-
1 , 2 < 7 < 12. Electrostatic potential plots for CB11H12
and CB11Me12- on the surface of a sphere centered at the center
of the icosahedron clearly display a deviation from spherical
symmetry.
When the methyl is carried by a boron atom, the associated
metal cation behaves as if it were attempting a backside SE2
substitution on the methyl carbon; when it is Me3Ge+, it nearly
succeeds in reaching the transition state for extraction of the
methyl group. When the cation is Me3Si+ or Me3C+, the transfer
of the methide anion actually succeeds.49,72 In these cases, we
propose that the coordinative bonding interaction is best
represented by a dotted line between the metal atom and the
carbon of the “methidic” methyl group, since its hydrogens play
a secondary role and are actually pushed aside by the metal
Supporting Information Available: Full experimental details
for all new compounds. This material is available free of charge
via the Internet at http://pubs.acs.org.
JA0475205
9
12046 J. AM. CHEM. SOC. VOL. 126, NO. 38, 2004