Thallophilic interactions and Tl-aryl π-interactions are competitive with cation-cation repulsion: [LTl2L]2+ dications as salts of weakly co-ordinating anions

Article abstract of DOI:10.1039/c3dt51233a

The synthesis and characterization by single-crystal X-ray diffraction of two thallium salts of the non-co-ordinating anion [B{3,5-(CF₃) ₂C₆H₃}₄] are reported. They possess Tl-Tl contacts supported by

Full text of DOI:10.1039/c3dt51233a

Dalton
Transactions
View Article Online
View Journal
COMMUNICATION
Thallophilic interactions and Tl–aryl π-interactions are
competitive with cation–cation repulsion: [LTl2L]2+
dications as salts of weakly co-ordinating anions†
Cite this: DOI: 10.1039/c3dt51233a
Received 10th May 2013,
Accepted 21st June 2013
Jennifer Cullinane,a Andrew Jolleysb and Francis S. Mair*a
DOI: 10.1039/c3dt51233a
www.rsc.org/dalton
The synthesis and characterization by single-crystal X-ray diffrac-
With reference to the work reported herein, which uses the
tion of two thallium salts of the non-co-ordinating anion new ligand class, the β-triketimines, perhaps the most relevant
[B{3,5-(CF3)2C6H3}4] are reported. They possess Tl–Tl contacts sup- exceptions to the first generalisation (i.e. those complexes
ported by Tl–aryl interaction. This cation association overcomes showing Tl–Tl interactions but with only terminal ligands) are
both electrostatic repulsion and cation–anion size and charge Tl(I) complexes of β-diketiminates and tris(pyrazolyl)borates.
matching considerations in the [LTl2L]2+ dications so formed.
In the β-diketiminate cases, Tl3 chains with Tl–Tl contacts
of 3.6 and 3.8 were found in [CH{(Me)CN-2,4,6-
Å
While Pyykkö has delineated the theoretical basis for metallo-
philic interactions,1 and explained why their strength should
be maximised for gold, the larger, more diffuse and more
polarisable 6s2 outer shell of Tl(I) has also attracted attention
from supramolecular chemists: the polarized 6s2 shell of
low-co-ordination-number Tl cations will frequently foster
association through Tl(I)–Tl(I) interactions (of 1083 Tl-organic
structures in the CSD,2 167 have Tl–Tl separations below 4 Å)
whose nature is predominantly dispersive, which can give rise
to extensive supramolecular associations, variously reviewed.3
The results we report here differ in two respects from this
prior work. Firstly, most cases showing Tl–Tl contacts have
these supported by bridging ligands. Secondly, in most cases,
some of the ligands carry a negative charge to balance that of
the Tl(I) cation. Notable exceptions to both of these generaliz-
ations are discussed below, in the context of our results, which
represent a rare report of Tl(I) cations bound to a neutral
ligand, L, self-associating with another cation to generate
[LTl2L]2+ dications (Scheme 1), in an apparent contravention of
the laws of electrostatics, since these are the first cases with
close Tl–Tl contacts. The formation of the dications also con-
travenes both the ‘rules of thumb’ of like-charges and of like-
sizes in the formation of ionic crystals.
Me3C6H2}2Tl]3,4a but in the bulkier [CH{(Me)CN-2,6-
Pri2C6H3}2Tl]2 only a very loose dimer with a Tl–Tl contact of
3.908 Å was found.4b Reducing steric compression by switching
to a β-dialdiminate, even retaining the diisopropylphenyl
groups, resulted in a shortening of the Tl–Tl distance to 3.76 Å
in [CH{HCN-2,6-iPr2C6H3}2Tl]2.4c This value is close to that
computed by Pyykkö as the optimum distance at which thallo-
philic attraction is balanced by core–core repulsion.5
Similar Tl–Tl distances are found in some Tl complexes of
tris(pyrazolyl)borates (tp). While there are many examples of
Tltp compounds showing no such interaction, the tetrahedral
cluster [{Tl(tpcpr)}4] {tpcpr = hydrotris(3-cyclopropylpyrazolyl)-
borate}, showed a Tl–Tl distance of 3.65 Å.5a Again, increase in
bulk reduced nuclearity and increased Tl–Tl distance, since
only a dimer was formed in [{Tl(tpp-tol)}2]; {tpp-tol = hydrotris-
(3-p-tolylpyrazolyl)borate}, with Tl–Tl = 3.86 Å.6b
Shorter distances are attained with lower-denticity ligands,
as shown by the tetrahedral cluster [{(Me3Si)3CTl}4], Tl–Tl =
3.32 Å,7a some 0.3 Å shorter than in [{Tl(tpcpr)}4], and in
Power’s attainment of the shortest Tl(I)–Tl(I) distance so far
found, 3.09 Å, with monodentate terphenyl ligands.7b
The β-triketimine ligands used here have much in common
with the two ligand classes discussed above, being κ-3 tri-
dentate (like the tris-(pyrazolyl)borates) derivatives of the
β-diketiminates, obtained by reaction of their lithium salts with
imidoyl chlorides.8 However, as neutral ligands, they differ
from both. There is one report of a pair of neutral tris(pyrazo-
lyl)methyl (tpm)-ligated complexes of Tl(I), the simple octa-
hedral monomer [Tl(tpm)2][PF6] and the loose dimer [{Tl(tpm)-
PF6}2] in which the well-separated Tl(I) cations in the dimer are
linked by two μ-2, cis-κ-2 PF6 ions.9 The only other report of
[{LTl}2]2+ concerns a dipyridyl-aryl-phosphole ligand bridging
aSchool of Chemistry, The University of Manchester, Manchester, UK.
E-mail: mair@manchester.ac.uk; Fax: +44 (0)161 275 4598;
Tel: +44 (0)161 200 4551
bSchool of Chemistry, University of Southampton, Southampton, SO17 1BJ, UK
†Electronic supplementary information (ESI) available: Full details of syntheses,
spectroscopic and crystallographic analysis, and computational studies. CCDC
899511 and 899512. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/c3dt51233a
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.

View Article Online
Communication
Dalton Transactions
Scheme 1
two thallium ions via aryl interactions while keeping the Tl
cations well-separated,10 but only recently has the co-ordina-
tion chemistry of Tl with neutral ligands, studied by employ-
ment of very weakly co-ordinating anions, received sustained
attention. Ethers and arenes are so far the main ligand types
considered,11 and in all cases, Tl–Tl interactions are conspi-
cuous by their absence.12
Two complexes of different neutral tki ligands were
obtained by direct combination with Tl[BArF] ([BArF]
=
[{3,5-(CF3)2C6H3}4B]). This compound is a useful reagent in
halide abstractions, and is normally prepared from TlOEt and
[H(OEt2)][BArF], in turn prepared in dry Et2O solution from
dried NaBArF and dry HCl.12 We found that Tl[BArF] was
obtainable by simple aqueous acetone metathesis using TlNO3
and crude Na[BArF]. Isolation by vacuum filtration in the open
laboratory gave material identical by elemental analysis and
powder X-ray diffraction to that reported. Identical dichloro-
methane solvate crystals also were obtained.13 This offers a
considerable simplification in the synthesis of this useful non-
hygroscopic material.
Fig. 1 Crystal and molecular structure of the centrosymmetric dication in com-
pound 1, [(tkiMe3iPr3Tl)2]·[BArF]2. Hydrogen atoms, the anion, and hexane of
crystallization are omitted. Symmetry-equivalent positions denoted with ‘’’. Key
interatomic distances (Å): Tl1–N1: 2.746(3); Tl1–N2: 2.630(3); Tl1–N3: 2.673(3);
Tl–N mean: 2.673(3); Tl–aryl centroid: 3.154(3); Tl1–Tl1’: 3.649(4).
Complexes 1 and 2 were thus conveniently preparable by solid phase),5 and with that shown in the (neutral) tetrahedral
combination of ligand with Tl[BArF] in dichloromethane tetramer [{Tl(tpcpr)}4],6a but in place of the further clustering
(Scheme 1). Crystals were obtained by layering the solution found in that case, an intermolecular η6 π-complexation com-
with hexane. Structures were determined by single-crystal X-ray pletes the co-ordination sphere of the Tl. The ring-centroid
diffraction measurement. Key interatomic distances are listed to Tl distance of 3.154 Å compares with ranges of 2.72 Å in
in the figure captions.
[Tl(C6Me6)][H2N{B(C6F5)3}2] to 3.09
Å
in [Tl(C6H5Me)2-
A feature common to 1 and 2 is that there are no significant CB11H6Br6] as reported in a recent review of Tl(I) complexes of
interactions between the Tl(I) cations and the BArF anions, arenes with weakly co-ordinating anions.11 However, such
perhaps unsurprisingly for such a weakly co-ordinating and interactions are responsible for pairing of monomers into
diffusely charged anion: the closest Tl–F contacts in 1 and 2 dimers at a ring centroid–Tl distance as high as 3.51 Å in [LiTl-
are 5.91 and 5.99 Å, respectively. These compare with values of {1,8-bis(trimethylsilylamido)naphthalene}].14 This longer dis-
3.05–3.86 Å in the parent uncomplexed ion pair Tl[BArF].12 In tance, though still shorter than the sum of two Tl Van Der
place of cation–anion contacts, the LTl cations form dimers, Waals radii (3.92 Å),15 is associated with use of a highly efficient
with some aryl interaction,10,11 and some Tl–Tl interaction chelating anionic amido electron donor. Absence of such ligands
sufficient to overcome coulombic repulsion at the distances through use of weakly co-ordinating anions promotes closer
observed. Both complexes form centrosymmetric dimers. The arene interaction;11 our examples using neutral tris-(aryl)imine
subtle difference in ligand structure between 1 and 2 offers a ligands appear to lie somewhere between these extremes.
means to intimately probe the nature of the association, and
lends support to the idea that the interactions they share may save for replacement of one of the ligand backbone methyls
have a structure-directing influence.
with a more sterically demanding tBu group. This has
Compound 2, [(tkiMe2tBuiPr3Tl)2]·[BArF]2, is identical to 1
Compound 1, [(tkiMe3iPr3Tl)2]·[BArF]2·[C6H14]‡ (see Fig. 1) the effect of pushing its isopropylphenyl group towards the
shows the shorter Tl–Tl interaction. The Tl–Tl distance is com- thallium, resulting in a short intra-monomer phenyl – ipso
parable with that computed by Pyykkö as optimal in CpTl carbon – Tl contact of 3.24 Å. Yet-shorter ipso C–Tl contacts of
(3.82 Å computed in the gas-phase dimer, 3.65 Å found in the 2.87 Å have been observed in neutral amido–Tl complexes.16
Dalton Trans.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Communication
anions.18 Experimental anion–Au distances in dimers and
chains of [{(MeNH)2C}2Au]+ with chloride, bromide, BF4 and
PF6 anions were 3.40, 3.54, 3.67 and 3.93 Å, respectively.19 At
these distances, correlated with the charge density of the
anions, the explanation of anion-mediated cation aggregates is
reasonable. However, in 1 and 2, where the anion is as diffuse
as [BArF] with its single negative charge spread over 69 atoms,
with all Tl–anion contacts exceeding 5.9 Å, one is tempted to
look elsewhere to bolster what must be a smaller attractive
term; full understanding of all electrostatic and other packing
terms contributing to the experimental lattice energy will be
required to determine the extent to which anion-mediated
cation aggregation contributes to the dimerization. One poss-
ible other place to look is the ‘rules of thumb’ of crystal
packing:20 firstly, ‘like charges pack’ is not followed, with
dimerization creating a dication to pack with two monoanions.
Secondly, ‘like sizes pack’ is not followed. The effective size of
the BArF anion in the crystals was calculated to be 968 Å3 on
average.21 Two of these were found in each cell, taking up
50–60% of the cell volume, hence, size matching for mono-
Fig. 2 Crystal and molecular structure of the centrosymmetric dication in com-
pound 2, [(tkiMe2tBuiPr3Tl)2]·[BArF]2. Hydrogen atoms, the anion, and a hexane of
crystallization are omitted. Symmetry-equivalent positions are denoted with ‘’’.
Key bond distances (Å): Tl1–N1: 2.623(2); Tl1–N2: 2.631(2); Tl1–N3: 2.763(2);
Tl–N mean: 2.672(3); Tl–aryl centroid: 3.104(3); Tl1–Tl1’: 3.7862(2).
The compression is also evidenced by the large sp2 N angle on meric monocations was close, yet the dimers still form.
this imine arm (126°), and the even larger sp2 C angle on the
We have attempted to probe this phenomenon with some
tBu-substituted carbon (130°). Compression of the remaining preliminary computational analysis using GAUSSIAN.22 Com-
co-ordination space around the thallium results in sterically- putational cost was minimised by use of single-point calcu-
promoted loosening of the Tl–Tl contact (to 3.79 Å) and a lations operating on the crystallographic geometry. In order to
slippage of the Tl–aryl interaction. The shortest inter-monomer estimate the true magnitude of coulombic repulsion, the
aryl–Tl contacts are 3.31 and 3.30 Å, to carbons most distal to charges on Tl were assessed using Natural Population Analy-
the centre of the molecule. Overall, however, the shift has caused sis.23 The results of energy and charge calculations on both an
a tightening of the Tl–aryl interaction, as judged by the Tl–ring isolated monomer cation and an isolated dimer dication, at
centroid distance. The Tl–N distances are almost unchanged. SCF-HF and at B3LYP density functional, both using LANL2DZ
These subtle changes in structure, a sterically-promoted inter- basis sets,24 are presented in Table 1.
monomer slippage causing stronger aryl–Tl and weaker Tl–Tl
interactions, can be seen on comparison of Fig. 1 and 2.
The NPA charges are higher than the Mulliken charges,
clustering around +0.9 across the methods and monomer/
One structural feature common to 1 and 2 is the distinct dimer range. They are systematically lower for dimer than for
twisting of the imine planes; the nitrogen lone pairs do not monomer, indicating the effect of π–aryl co-ordination, in all
point towards the metal. This confirms two things: firstly, that cases except the Mulliken DF-B3LYP case. This underlines the
the bonding is predominantly electrostatic, and secondly, that superiority of NPA charge analysis. Taking the most conserva-
the tki ligand class shares with its tpm cousin a reluctance to tive NPA estimate of true “Tl+” charge, +0.82, the electrostatic
splay outwards to accommodate larger metal ions. Instead, repulsion between two such charges at the experimental Tl–Tl
this twisting furnishes a larger void in which the metal ion distance of 3.64 Å is calculated to be +61.76 kcal mol−1. The
may reside.9
computed dimerization energies of the cation from 1 are +40
In these cases Tl–arene interactions foster thallophilic (B3-LYP) and +47 (SCF) kcal mol−1. These smaller endothermic
interactions, in concert balancing the electrostatic repulsion,
while steric repulsion mediates the closeness of approach of
the two cations. Neither attractive term is strong, but in combi-
nation, aided by several smaller CH–π and other weak inter-
Table 1 Summary of computational results
actions which are normally summatively termed ‘crystal
Tlq(M)
Tlq(NPA)
Energy/Hartrees
ΔED/kcal mol−1
packing forces’ they seem able to find balance with the
unfavourable electrostatics. It is noteworthy that it is thallium
cations which assemble without the help of bridging anions, a
feat that, according to computational results of Carvajal
et al.,17 is beyond even the much stronger attractive term
enjoyed by gold(I) species.1 Computational results from Pyykkö
reveal that while excited states have attractive components, the
HF-SCF
Monomer
Dimer
DF-B3LYP
Monomer
Dimer
+47.4
0.82
0.84
0.94
0.87
−1524.6289
−3049.18217
+39.9
0.70
0.65
0.90
0.82
−1536.06077
−3072.05798
Basis set: LANL2-DZ. Tlq(M) = Mulliken charge on Tl; Tlq(NPA) = Natural
Population Analysis charge on Tl; ΔED = energy of dimerization, i.e. 2 ×
ground state of Au2 is entirely repulsive;18 Carvajal’s results
2+
on L2Au+ dimers concur, unless the association is mediated by monomer energy − dimer energy.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.

View Article Online
Communication
Dalton Transactions
values reflect the favourable effect of Tl–aryl interaction, which
is partially modelled by these two methods.11 However, these
widely-used methods fail to predict the formation of these
dimers by some considerable margin. This failure is partly
because of the single-point nature of the calculations, partly
because of the inability of SCF to model configuration inter-
action effects, only partially recovered by DF-B3LYP, and partly
due to the gas-phase nature of the calculation. Current under-
standing ascribes less than 10 kcal mol−1 to Tl–Tl attraction,5
which leaves over 30 kcal mol−1 unaccounted for. Fixing the dica-
tion in a lattice of anions, even at the distances observed, rather
than consideration of a gas-phase species, will recover some of
this discrepancy. The issue is ripe for investigation at the highest
levels of theory not available to our computational resources.
Furthermore, evidence for the importance of Tl–Tl inter-
actions is not restricted to that provided by crystallography and
computation; the luminescence of some gold–thallium organo-
metallic complexes was ascribed to T–Tl interaction, which
persisted upon dissolution.25
4 (a) M. S. Hill, R. Pongtavornpinyo and P. B. Hitchcock,
Chem. Commun., 2006, 3720; (b) M. S. Hill, P. B. Hitchcock
and R. Pongtavornpinyo, Dalton Trans., 2005, 1433;
(c) Y. Cheng, P. B. Hitchcock, M. F. Lappert and M. Zhou,
Chem. Commun., 2005, 752.
5 (a) P. Pyykkö, M. Straka and T. Tamm, Phys. Chem. Chem.
Phys., 1999, 1, 3441; (b) H. Schumann, C. Janiak,
M. A. Khan and J. J. Zuckermann, J. Organomet. Chem.,
1988, 354, 7.
6 (a) A. L. Rheingold, L. M. Liable-Sands and S. Trofimenko,
Chem. Commun., 1997, 1691; (b) G. Ferguson,
M. C. Jennings, F. J. Lalor and C. Starahan, Acta Crystal-
logr., Sect. C: Cryst. Struct. Commun., 1991, 47, 2079.
7 (a) W. Uhl, S. U. Keimling, K. W. Klinkhammer and
W. Schwarz, Angew. Chem., 1997, 209, 64, (Angew. Chem. Int.
Ed. Engl., 1997, 36, 64); (b) R. J. Wright, A. D. Phillips,
S. Hino and P. P. Power, J. Am. Chem. Soc., 2005, 127, 4794.
8 D. Barnes, G. L. Brown, M. Brownhill, I. German,
C. J. Herbert, A. Jolleys, A. R. Kennedy, B. Liu, K. McBride,
F. S. Mair, R. G. Pritchard, A. Sanders and J. E. Warren,
Eur. J. Inorg. Chem., 2009, 1219.
In summary, there is a clear hierarchy of strength of inter-
action governing assembly of the dimeric dications 1 and 2;
firstly, imine co-ordination, then arene co-ordination, then
thallophilic interaction, perhaps aided by the higher lattice
9 D. L. Reger, J. E. Collins, R. Layland and R. D. Adams,
Inorg. Chem., 1996, 35, 1372.
energy generally available to AB2 systems in comparison to AB 10 S. Welsch, C. Lescop, R. Réau and M. Scheer, Dalton Trans.,
ones, other factors being equal. It remains surprising that 2009, 2683.
cation–anion attraction and cation–cation repulsion do not 11 T. Auel and E. L. Amma, J. Am. Chem. Soc., 1968, 90, 5941;
appear to be factors, nor do the ‘rules of thumb’ of like sizes
and charges. [LTl2L]2+ appears possible so long as there is a
bridging aryl to foster the interaction, even in the absence of
close anion contacts. There is now a molecular chemistry to
add to examples of Tl22+ in ionic lattices such as TlZrF5.26 This
result contributes to an isoelectronic series of closed-shell 6s2
H. Schmidbaur, W. Bublak, J. Riede and G. Müller, Angew.
Chem., Int. Ed. Engl., 1985, 24, 414, (Angew. Chem., 1985,
97, 402); Y. Sarazin, D. L. Hughes, N. Kaltsoyannis,
J. A. Wright and M. Bochmann, J. Am. Chem. Soc., 2007,
129, 881; Y. Sarazin, N. Kaltsoyannis, J. A. Wright and
M. Bochmann, Organometallics, 2007, 26, 1881.
2− 27
4+ 29
6+ 30
dimers, Rb2
,
Ba2,28 Tl22+, Pb2
,
and Bi2
.
We are 12 Reviews: H. Schmidbaur and A. Schier, Organometallics,
engaged in collection of data on further examples of this class,
and welcome further theoretical study, with full all-electron
2008, 27, 2361; M. Bochmann, Coord. Chem. Rev., 2009,
253, 2000, and ref. therein.
treatment of dispersive forces, in addition to attempts to fully 13 R. P. Hughes, D. C. Lindner, A. L. Rheingold and
quantify lattice energy considerations. Furthermore, the syn- G. P. A. Yap, Inorg. Chem., 1997, 36, 1726.
thetic search for molecular examples of other members of the 14 K. W. Hellmann, G. Galka, L. H. Gade, A. Steiner,
isoelectronic series, which currently is seen only in ionic
lattices or gas-phase, may be merited.
D. S. Wright and D. Stalke, Chem. Commun., 1998, 549.
15 A. J. Bondi, J. Phys. Chem., 1964, 68, 441.
16 C. H. Galka and L. H. Gade, Inorg. Chem., 1999, 38, 1038.
17 M. A. Carvajal, S. Alvarez and J. J. Novoa, Theor. Chem. Acc.,
2006, 116, 472.
18 M. Barysz and P. Pyykkö, Chem. Phys. Lett., 2000, 325, 225.
19 R. L. White-Morris, M. M. Olmstead, F. Jiang and
A. L. Balch, Inorg. Chem., 2002, 41, 2313; R. L. White-
Morris, M. M. Olmstead, F. Jiang, D. S. Tinti and
A. L. Balch, J. Am. Chem. Soc., 2002, 124, 2327.
Notes and references
‡Nomenclature: β-triketimine = tki; the substituents on the imine carbon atoms
are indicated, followed by an indication of the N-aryl substituents, which in all
cases are ortho. Hence, tkiMe3iPr3
= =
HC{MeCN(2-iPrC6H5)3}; tkiMe2tBuiPr3
HC{MeCN(2-iPrC6H5)}2{tBuCN(2-iPrC6H5)}.
1 P. Pyykkö, Chem. Soc. Rev., 2008, 37, 1967.
20 F. Basolo, Coord. Chem. Rev., 1968, 3, 213.
2 Cambridge Structural Database: F. H. Allen, Acta Crystal- 21 The values, 1: 949; 2: 987 Å3, were determined using the
logr., Sect. B: Struct. Sci., 2002, 58, 380.
3 K. Akhabari and A. Morsali, Coord. Chem. Rev., 2010, 254,
VOIDS function within PLATON: A. L. Spek, Acta Crystal-
logr., Sect. D: Biol. Crystallogr., 2009, 65, 148.
1997; L. H. Gade, Dalton Trans., 2003, 367; P. Gosh, 22 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
A. L. Rheingold and G. Parkin, Inorg. Chem., 1999, 28, 5464;
C. Janiak, Coord. Chem. Rev., 1997, 163, 107; C. Janiak and
R. Hoffman, J. Am. Chem. Soc., 1990, 112, 5924.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
Dalton Trans.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Communication
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, 24 P. J. Hay and W. R. Wadt, J. Chem. Phys., 1985, 82,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, 299.
H. Nakai, T. Vreven, J. A. Montgomery Jr., J. E. Peralta, 25 E. J. Fernandez, J. M. Lopez-de-Luzuriaga, M. Monge,
F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers,
K. N. Kudin, V. N. Staroverov, T. Keith, R. Kobayashi,
J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam,
M. E. Olmos, J. Perez and A. Laguna, J. Am. Chem. Soc.,
2002, 124, 5942; E. J. Fernandez, J. M. Lopez-de-Luzuriaga,
M. E. Olmos, J. Perez, A. Laguna and M. C. Lagunas, Inorg.
Chem., 2005, 44, 6012.
M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, 26 S. P. Gabuda, S. G. Kozlova and R. L. Davidovich, Chem.
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, Phys. Lett., 1996, 254, 89.
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, 27 R. H. Huang, D. L. Ward and J. L. Dye, J. Am. Chem. Soc.,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, 1989, 111, 5707.
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, 28 S. F. Li, X. Xue, G. Chen, D. W. Yuan, Y. Jia and X. G. Gong,
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and J. Chem. Phys., 2006, 124, 224711.
D. J. Fox, GAUSSIAN 09 (Revision B.01), Gaussian, Inc., 29 C. Dybowski, S. P. Gabuda, S. G. Kozlova, G. Neue,
Wallingford, CT, 2010.
23 A. E. Reed, R. B. Weinstock and F. Weinhold, J. Chem.
Phys., 1985, 83, 735.
D. L. Perry and V. V. Terskikh, J. Solid State Chem., 2001,
157, 220.
30 H. A. Harvig, Z. Anorg. Allg. Chem., 1978, 444, 151.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans.