Facile isomerization and unprecedented decarbonation of metallacarboranes with fluorinated aryl substituents

Article abstract of DOI:10.1021/om200691z

Reduction and metalation of 1,2-(4′-F₃CC₆F ₄)₂-1,2-closo-C₂B₁₀H₁₀ affords 4,1,12-MC₂B₁₀ compounds directly at room temperature with no evidence of intermediate 4,1,6- or 4,1,8-analogues. For M = CpCo this species is the only isolable product. The M = Cp*Co analogue is produced together with an icosahedral dimetallacarborane that remarkably has lost a cage C vertex.

Full text of DOI:10.1021/om200691z

COMMUNICATION
pubs.acs.org/Organometallics
Facile Isomerization and Unprecedented Decarbonation of
Metallacarboranes with Fluorinated Aryl Substituents†
Jas S. Ward, Hugo Tricas, Greig Scott, David Ellis, Georgina M. Rosair, and Alan J. Welch*
Department of Chemistry, Heriot-Watt University, Edinburgh EH14 4AS, U.K.
S Supporting Information
b
ABSTRACT: Reduction and metalation of 1,2-(40-F3CC6F4)2-1,2-closo-C2B10H10
affords 4,1,12-MC2B10 compounds directly at room temperature with no evidence of
intermediate 4,1,6- or 4,1,8-analogues. For M = CpCo this species is the only isolable
product. The M = Cp*Co analogue is produced together with an icosahedral
dimetallacarborane that remarkably has lost a cage C vertex.
e recently described the synthesis and characterization of a
series of icosahedral carboranes with fluorinated aryl
overlapping triplets. An X-ray diffraction study6 ultimately re-
W
vealed 1 to be the 4,1,12-MC2B10 isomer 1,12-(40-F3CC6F4)2-4-
Cp-4,1,12-closo-CoC2B10H10 (Figure 1). It is well established
that 4,1,12-MC2B10 metallacarboranes can be prepared from
either 4,1,6-3,7 or 4,1,10-analogues8 by thermolysis (the former
proceeding via the 4,1,8-isomer), but the direct formation of a
4,1,12-MC2B10 isomer from an o-carborane derivative at only
room temperature is unprecedented. We suggest that the initial
product of reduction and metalation of I is, indeed, a 4,1,6-
MC2B10 species but that this spontaneously isomerizes to the
thermodynamically preferred 4,1,12-isomer. Although the pre-
cise mechanism by which 4,1,6-MC2B10 sequentially isomerizes
to 4,1,8- and 4,1,12-forms has not been determined, there is no
reason to believe that the same process does not happen here, in
which case the barriers to isomerization must be considerably
reduced due to either steric or electronic (or both) influences of
the fluorinated aryl substituents.9
C substituents,1 deliberately targeted as potential precursors
for supraicosahedral (hetero)carboranes. Of these the most
promising compound, on the basis of a combination of NMR
spectroscopic, computational, and electrochemical data, was
1,2-(40-F3CC6F4)2-1,2-closo-C2B10H10 (I). We now report that
reduction and subsequent metalation of I with {CpCo} and
{Cp*Co} fragments affords 4,1,12-CoC2B10 supraicosahedral
products which are the result of unprecedented room tempera-
ture (or below) isomerization. The {Cp*Co} reaction addition-
ally results in a monocarbon metallacarborane which is the result
of a unique decarbonation reaction.
Reaction of a THF solution of I, previously reduced with 2
equiv of sodium naphthalenide, with Na[C5H5] and CoCl2, fol-
lowed by aerial oxidation, affords the orange cobaltacarborane 1
as the only isolable product.2 Compound 1 was initially char-
acterized by microanalysis, mass spectrometry, and1H, 11B and 19F
NMR spectroscopy. Conventionally, reduction and subse-
quent metalation of 1,2-closo-C2B10 species affords 4,1,6-MC2B10
docosahedral products3 which are asymmetric in the solid state4
but Cs symmetric in solution via a double diamondꢀsquareꢀdiamond
(DSD) fluxional process that is well understood.3,5 However,
spectroscopic analysis of 1 clearly suggested that it is asymmetric
in solution. Thus, in the 11B{1H} NMR spectrum in acetone-d6
are eight resonances with the relative integrals 2:1:1:1:1:1:1:2
(from high frequency to low frequency) and the 13C spectrum
includes two Ccage resonances at 67.0 and 69.0 ppm. Although
initial inspection of the 19F spectrum suggests only one unique
40-F3CC6F4 substituent, expansion of the multiplet (due to the
CF3 group) centered on ꢀ57.3 ppm clearly shows it to be two
When I is stoichiometrically reduced (2e) and treated with
Na[C5Me5] and CoCl2, three compounds are isolated following
workup.11 The major product (24%), 2, was ultimately shown
simply to be the modified carborane 1,2-(40-F2{C5Me5}-
CC6F4)2-1,2-closo-C2B10H10,12 but minor products 3 (6%) and
4 (8%) are cobaltacarboranes.
Compound 3 was characterized spectroscopically. Although
single crystals of suitable quality for a diffraction study could not
be obtained, we confidently identify 3 as 1,12-(40-F3CC6F4)2-4-
Cp*-4,1,12-closo-CoC2B10H10, the simple Cp* analogue of 1.
Special Issue: F. Gordon A. Stone Commemorative Issue
Received: August 1, 2011
Published: September 19, 2011
r
2011 American Chemical Society
2523
dx.doi.org/10.1021/om200691z Organometallics 2012, 31, 2523–2525
|

Organometallics
COMMUNICATION
Figure 1. Molecular structure of 1 with most atoms labeled. B2 lies
between C1 and B6, B9 between B5 and C12, and B7 between B3 and
B10. Atoms (except H) are shown as 50% probability ellipsoids. Selected
distances (Å): Co4ꢀC1 = 2.069(3), Co4ꢀB2 = 2.156(4), Co4ꢀB6 =
2.127(4), Co4ꢀB10 = 2.125(4), Co4ꢀB7 = 2.170(4), Co4ꢀB3 =
2.247(4).
Figure 2. Molecular structure of 4 with most atoms labeled. The
heteroborane numbering follows the standard convention.19 Atoms
(except H) are shown as 50% probability ellipsoids. Selected distances (Å):
Co2ꢀC1 = 2.081(6), Co2ꢀB3 = 2.065(7), Co2ꢀB7 = 2.107(7), Co2ꢀB11 =
2.089(7), Co2ꢀB6 = 2.029(7), Co8ꢀB3 = 2.056(7), Co8ꢀB4 = 2.036(7),
Co8ꢀB9 = 2.071(7), Co8ꢀB12 = 2.051(7), Co8ꢀB7 = 2.079(7),
B12ꢀO121 = 1.555(7).
This conclusion is based on the very similar 11B NMR chemical
shift ranges of 1 and 3, +8 to ꢀ13 ppm and +9 to ꢀ14 ppm,
respectively, and the notable absence in 3 of a high-frequency
resonance, ca. +21 ppm, characteristic of a 4,1,8-MC2B10 species.3
The detail of the pattern of the 11Bspectrumof3(1:1:1:1:1:2:1:2) is
slightly different from that of 1, but in both cases the integral-2
resonances are merely accidental 1 + 1 coincidences.
architecture by the reduction/metalation approach and toward
an understanding of the sequence of events involved in their
formation.18
’ ASSOCIATED CONTENT
The third reaction product, compound 4, was fully character-
ized. Two different {Cp*Co} fragments are implied by the H
S
Supporting Information. CIF files giving crystallo-
b
1
graphic data for compounds 1, 2, and 4. This material is available
NMR spectrum, which also contains resonances attributable to
THF but at relatively high frequencies, suggesting a formal
[C4H8O]+ unit bound to the cage through a trisubstituted oxygen
atom. The 11B spectrum reveals six resonances (1:1:1:2:2:2) the
highest frequency resonance of which is only a singlet, and this
suggests the presence of only nine B atoms, one of which does not
carry an exo H atom. The exact nature of 4 was established by a
crystallographic study (Figure 2).13
free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*Tel: +441314513217. Fax: +441314513180. E-mail: a.j.welch@
hw.ac.uk.
Compound 4 is the 12-vertex, distorted-icosahedral dimetal-
lamonocarborane 1-(40-F3CC6F4)-2,8-Cp*2-12-THF-2,8,1-closo-
Co2CB9H8, formally a zwitterionic species with charges of
1ꢀ on the cage and 1+ on O121. It is related to the reduced
precursor I by the following five processes, listed in random order
with no implication of a reaction sequence: addition of the first
{Cp*Co} fragment, addition of the second {Cp*Co} fragment,
loss of a {BH} fragment, loss of a {C(40-F3CC6F4)} fragment,
and transformation of {BH} into {B(THF)}. Of these processes
the loss of the {C(40-F3CC6F4)} fragment is the most remark-
able. Carbon vertex loss (decarbonation) is rare in carborane
’ ACKNOWLEDGMENT
We are grateful to the EPSRC for support of this work (project
grant EP/E02971X/1 supporting H.T. and D.E.; DTA award
supporting G.S.).
’ DEDICATION
†In memory of F. Gordon A. Stone, a true giant of organometallic
chemistry.
’ REFERENCES
ꢀ
chemistry. Stíbr et al. have described the formation of arachno-
monocarboranes from [7,9-nido-C2B10H12]2ꢀ in the presence of
Lewis base,14,15 and Xie has reported decarbonation of supraicosa-
hedral carboranes under similar conditions,16 but as far as we are
aware the formation of a closo metallamonocarborane by decarbo-
nation on metalation of reduced dicarborane is unprecedented.
The 2,8,1-M2CB9 icosahedral isomer has only been seen once
previously, the result of spontaneous direct insertion of a pendant
{Pt(dppe)}0 fragment into an 11-vertex closo MnCB9 cluster by
Stone et al.17 Current studies are directed toward the synthesis of
further homo- and heterobimetallic examples of this cluster
(1) Tricas, H.; Colon, M.; Ellis, D.; Macgregor, S. A.; McKay, D.;
Rosair, G. M.; Welch, A. J.; Glukhov, I. V.; Rossi, F.; Laschi, F.; Zanello,
P. Dalton Trans. 2011, 40, 4200–4211.
(2) A dry, oxygen-free THF solution (10 mL) of I (0.40 g, 0.69
mmol) was treated with 1.39 mmol (2.1 equiv) of sodium naphthalenide
as a THF solution at 0 °C. NaCp (1.41 mL of a 2.0 M solution in THF,
2.82 mmol) and then CoCl2 (0.333 g, 3.92 mmol) were added, and the
reactants were stirred at room temperature for 18 h. Air was drawn
through the mixture for 30 min, and all volatiles were removed in vacuo.
The residue was extracted into dichloromethane (20 mL) and the extract
filtered. All volatiles were again removed in vacuo. Naphthalene and any
2524
dx.doi.org/10.1021/om200691z |Organometallics 2012, 31, 2523–2525

Organometallics
COMMUNICATION
unreacted carborane were removed by washing with 40ꢀ60 °C petro-
leum ether. Preparative thin-layer chromatography (TLC) with dichlor-
omethane/40ꢀ60 °C petroleum ether (2/5) as eluent yielded a single
orange band, compound 1 (Rf = 0.63, 0.054 g, 11%). The yield is increased
to 16% if the reaction mixture is heated to reflux before oxidation. Anal.
Found: C, 36.0; H, 2.10. Calcd for C21H15B10CoF14: C, 36.0; H, 2.16.
1H NMR (acetone-d6, 298 K): δ 5.81 (s, C5H5) ppm. 11B{1H} NMR
(acetone-d6, 298 K): δ 7.43 (2B), 4.65 (1B), 3.15 (1B), 0.08 (1B),
ꢀ4.10 (1B), ꢀ5.76 (1B), ꢀ9.00 (1B), ꢀ13.60 (2B) ppm. 19F NMR
(acetone-d6, 298 K): δ ꢀ56.3 (m [2 ꢁ t on expansion], 6F, CF3),
ꢀ133.5 (app s, 4F, o-F), ꢀ142.8 (m, 4F, m-F) ppm. Partial 13C NMR
(acetone-d6, 298 K): δ 69.0 (br, 1C, Ccage), 67.0 (br. 1C, Ccage). EI-MS:
envelope centered on m/z 701 (M+).
(3) Dustin, D. F.; Dunks, G. B.; Hawthorne, M. F. J. Am. Chem. Soc.
1973, 95, 1109–1115.
(4) Churchill, M. R.; DeBoer, B. G. Inorg. Chem. 1974, 13, 1411–1418.
(5) Burke, A.; Ellis, D.; Ferrer, D.; Ormsby, D. L.; Rosair, G. M.;
Welch, A. J. Dalton Trans. 2005, 1716–1721.
(6) Crystal data for 1: C21H15B10CoF14, M = 700.36, monoclinic, Cc,
a = 24.237(3) Å, b = 12.4207(17) Å, c = 9.0234(12) Å, β = 105.478(7)°,
V = 2617.9(6) Å3, Z = 4, Dc = 1.777 Mg mꢀ3, μ = 0.771 mmꢀ1, F(000) =
1376, data to θmax = 25.65° collected at 100(2) K on a Bruker X8
diffractometer using Mo Kα radiation, 4860 independent reflections out
of 17 205 measured, Rint = 0.0499, R1 = 0.0376, wR2 = 0.0603, S = 1.022
for 4115 data with I > 2σ(I), absolute structure parameter 0.251(12).
(7) Burke, A.; McIntosh, R.; Ellis, D.; Rosair, G. M.; Welch, A. J.
Collect. Czech. Chem. Commun. 2002, 67, 991–1006.
(s, 15H, C5Me5) ppm. 11B{1H} NMR (CDCl3, 298 K): δ 27.15 (1B, B12),
19.17 (1B, 17.32 (1B), 4.04 (2B), 0.57 (2B), ꢀ18.74 (2B) ppm. 19F
NMR (CDCl3, 298 K): δ ꢀ56.07 (m, 3F, CF3), ꢀ120.05 (m, 2F, o-F),
ꢀ144.39 (m, 2F, m-F) ppm. EI-MS: envelopes centered on m/z 795 (M+)
and 723 (M+ ꢀ THF).
(12) Crystal data for 2: C36H40B10F12, M = 808.78, triclinic, P1, a =
14.5973(13) Å, b = 16.0164(14) Å, c = 18.3603(16) Å, α = 70.015(4)°, β =
79.973(4)°, γ = 89.535(4)°, V = 3966.3(6) Å3, Z = 4, Dc = 1.354 Mg mꢀ3
,
μ = 0.113 mmꢀ1, F(000) = 1656, θmax = 25.49°, 100(2) K, 18 020/69 711
reflections, Rint = 0.0605, R1 = 0.0835, wR2 = 0.2115, S = 1.146 for 14 072
data with I > 2σ(I).
(13) Crystal data for 4:C32H46B9Co2F7O, M = 794.84, orthorhom-
bic, Pbca, a = 12.8521(5) Å, b = 16.2817(6) Å, c = 33.9667(12) Å, V =
7107.7(5) Å3, Z = 8, Dc = 1.486 Mg mꢀ3, μ = 0.998 mmꢀ1, F(000) =
3264, θmax = 25.68°, 100(2) K, 6737/39 726 reflections, Rint = 0.1167,
R1 = 0.0899, wR2 = 0.1459, S = 1.422 for 4792 data with I > 2σ(I).
ꢀ
(14) Pleꢀsek, J.; Stíbr, B.; Fontaine, X. L. R.; Jelínek, T.; Thornton-Pett,
M.; Heꢀrmꢁanek, S.; Kennedy, J. D. Inorg. Chem. 1994, 33, 2994–3002.
ꢀ
(15) Wille, A. E.; Pleꢀsek, J.; Holub, J.; Stíbr, B.; Carroll, P. J.;
Sneddon, L. G. Inorg. Chem. 1996, 35, 5342–5346.
(16) Zhang, J.; Chan, H.-S.; Xie, Z. Chem. Commun. 2011, 47,
8082–8084 and references therein.
(17) Du, S.; Jeffery, J. C.; Kautz, J. A.; Lu, X. L.; McGrath, T. D.;
Miller, T. A.; Riis-Johannessen, T.; Stone, F. G. A. Inorg. Chem. 2005,
44, 2815–2825.
(18) Scott, G.; Welch, A. J.; Macgregor, S. A. To be submitted for
publication.
(8) Ellis, D.; Lopez, M. E.; McIntosh, R.; Rosair, G. M.; Welch, A. J.;
(19) Grimes, R. N. Carboranes, 2nd ed.; Elsevier: Amsterdam, 2011.
Quenardelle, R. Chem. Commun. 2005, 1348–1350.
(9) It is well-established in icosahedral metallacarborane chemistry
that steric crowding can lead to low-temperature isomerizations,10 and
clearly 1,6-(40-F3CC6F4)2-4-Cp-4,1,6-closo-CoC2B10H10 would be some-
what more sterically crowded than 4-Cp-4,1,6-closo-CoC2B10H12 (circum-
stantial evidence of this is that in 1 the Cp ligand leans away from the
C7F7 substituent on C1, making a dihedral angle of 10.9° with the
B5B8B13C12B9 reference plane, whereas the equivalent angle to the
4
B5B8B13B12B9 plane in 4-Cp-4,1,6-closo-CoC2B10H12 is only 4.3°).
On the other hand, it is not intuitively obvious that 1 would be any less
sterically crowded than the putative 1,8-(40-F3CC6F4)2-4-Cp-4,1,8-
closo-CoC2B10H10 species from which it presumably forms.
(10) E.g.: Baghurst, D. R.; Copley, R. C. B.; Fleischer, H.; Mingos,
D. M. P.; Kyd, G. O.; Yellowlees, L. J.; Welch, A. J.; Spalding, T. R.;
O’Connell, D. J. Organomet. Chem. 1993, 447, C14–C17.
(11) I (0.50 g, 0.87 mmol) in THF (10 mL) was treated with 1.82
mmol (2.1 equiv) of sodium naphthalenide. NaCp* (5.20 mL of a 0.5 M
solution in THF, 2.60 mmol) and CoCl2 (0.293 g, 3.45 mmol) were
added and the reactants stirred at room temperature for 18 h. After 30
min of aerial oxidation volatiles were removed and the residue was
washed with 40ꢀ60 °C petroleum ether. Column chromatography on
silica with dichloromethane/40ꢀ60 °C petroleum ether (1/1) as eluent
afforded pale yellow 2 (0.169 g, 24%) and an orange mixture that was
separated by preparative TLC, with dichloromethane/40ꢀ60 °C pet-
roleum ether (2/3) as eluent, to yield 3 (Rf = 0.75, 0.037 g, 6%) and 4
(Rf = 0.27, 0.058 g, 8%). Data for 2 are as follows. Anal. Found: C, 53.0; H,
4.85. Calcdfor C36H40B10F12: C, 53.5; H, 4.98. 1H NMR(CDCl3, 298 K):
δ 1.85 (s, 12H, C5Me5), 1.46 (s, 12H, C5Me5), 1.25 (s, 6H, C5Me5) ppm.
11B{1H} NMR (CDCl3, 298 K): δ ꢀ1.97 (2B), ꢀ9.89 (8B) ppm. 19
F
NMR (CDCl3, 298 K): δ ꢀ94.92 (m, 4F, CF2), ꢀ135.29 (m, 4F, o-F),
ꢀ138.03 (m, 4F, m-F) ppm. EI-MS: envelope centered on m/z 809
(M+). Data for 3 are as follows. Anal. Found: C, 41.2; H, 3.28. Calcd for
C26H25B10CoF14: C, 40.5; H, 3.27. 1H NMR (CDCl3, 298 K): δ 1.71
(s, C5Me5) ppm. 11B{1H} NMR (CDCl3, 298 K): δ 8.78 (1B), 4.73 (1B),
1.93 (1B), 0.12 (1B), ꢀ2.21 (1B), ꢀ7.98 (2B), ꢀ9.76 (1B), ꢀ14.00
(2B) ppm. 19F NMR (CDCl3, 298 K): δ ꢀ56.55 (m, 6F, CF3), ꢀ131.31
(m, 4F, o-F), ꢀ141.26 (m, 4F, m-F) ppm. EI-MS: envelope centered on
m/z 771 (M+). Data for 4 are as follows. Anal. Found: C, 48.6; H, 5.83.
Calcd for C32H46B9Co2F7O: C, 48.4; H, 5.83. 1H NMR (CDCl3, 298 K):
δ 4.33 (M, 4H, THF), 2.11 (m, 4H, THF), 1.76 (s, 15H, C5Me5), 1.48
2525
dx.doi.org/10.1021/om200691z |Organometallics 2012, 31, 2523–2525