Metal-rich, neutral and cationic organotin clusters

Article abstract of DOI:10.1002/anie.200500117

Tin can form examples of rare organically substituted clusters whose structures are related to Zintl salts. The two new types of metal-rich organotin clusters, a neutral Sn₉Ar₃ (see picture) and two ionic [Sn₁₀Ar′₃]⁺ species, were structurally characterized and their bonding situation investigated by density functional calculations. (Figure Presented).

Full text of DOI:10.1002/anie.200500117

Communications
Cluster Compounds
Metal-Rich, Neutral and Cationic Organotin
Clusters**
Anne F. Richards, Barrett E. Eichler, Marcin Brynda,
Marilyn M. Olmstead, and Philip P. Power*
With few exceptions, isolated clusters of the heavier Group 14
elements (Si–Pb) are of the Zintl (Enxꢀ, x = 2, 3, or 4)[1,2] or
organically substituted, that is, EnRm (n = m or n > m; R =
aryl, aryl, silyl, amide),[3,4] types. A notable feature of the two
classes is that they rarely display a structural or electronic
correspondence to each other. This situation is particularly
evident in the structures of the element-rich[5] Group 14
organo clusters EnRm (n > m) such as Sn8{Si(SiMe3)3}6,[6]
Sn8(ArMes
)
4
[7] (ArMes = 2,6-Mes2C6H3; Mes = 2,4,6-Me3C6H2),
2
2
Ge8{N(SiMe3)2}6,[8] Ge6(ArMes )2,[9] Sn4Ge2(ArMes [9] as well as
)
2
2
2
*
*
the recently reported lead clusters Pb10R6 and Pb12R6 (R* =
Si(SiMe3)3).[10] These bear little stoichiometric or structural
xꢀ
xꢀ
resemblance to the common E5 or E9 (E = Si–Pb) Zintl
ions of these elements. A number of publications have shown
that Zintl ion clusters, such as E9xꢀ, can be derivatized with
transition-metal or main-group fragments with preservation
of the original framework.[11–15] Nonetheless, there is only one
example (the anion [Ge9{N(SiMe3)2}3]ꢀ)[16] of a derivatized
Zintl cluster that can be synthesized without use of preformed
E5xꢀ or E9xꢀ units. No stable examples of neutral clusters with
structures analogous to the Zintl ions exist at present.
Similarly, knowledge of cationic clusters of the Group 14
elements is extremely limited. For example, cationic Zintl
clusters of these elements of any kind are currently unknown
and the only well characterized stable cationic organo
Group 14 cluster is the recently described silyl-substituted
germanium species [Ge10(SitBu3)6I][B(C6F4H)4].[17] Herein we
report three new element-rich organo clusters of tin whose
xꢀ
structures are analogous to, or derived from, the E9 Zintl
framework. These are the neutral Sn9(ArTrip )3·4THF
2
(1·4THF; ArTrip = 2,6-(2,4,6-iPr3C6H2)2C6H3) and the cationic
2
clusters [Sn10(ArMes )3] [AlCl4]·PhMe (2·PhMe) and [Sn10-
2
(ArMes )3][GaCl4]·PhMe (3·PhMe).
2
The neutral cluster Sn9(ArTrip )3, which is paramagnetic
2
and has 21 framework electrons, was synthesized by the
[18]
thermolysis of {ArTrip Sn(m-H)}2
in hot toluene and was
2
isolated as dark red, almost black crystals. This method is a
new synthetic route to tin clusters[19] that arose from exploring
[*] Dr. A. F. Richards, Dr. B. E. Eichler, Dr. M. Brynda,
Prof. M. M. Olmstead, Prof. P. P. Power
Department of Chemistry, University of California
One Shields Avenue, Davis, CA 95616 (USA)
Fax: (+1)530-752-8995
E-mail: pppower@ucdavis.edu
[**] The authors thank the NSF-CRC program for financial support. M.B.
was supported by the Swiss National Science Foundation Grant
8220-067593.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
2546
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200500117
Angew. Chem. Int. Ed. 2005, 44, 2546 –2549

Angewandte
Chemie
the decomposition of the hydride in an attempt to isolate the
observed to the tin atoms capping the rectangular prism faces,
Trip2
distannyne ArTrip SnSnAr
.
[20] Elimination of hydrogen does
that is, the substituted Sn1, Sn1A, and Sn5 in 1. DFT
2
occur, but competing Sn C bond cleavage results in ligand
calculations[24] for Sn9Me3 model species for 1 afforded Sn Sn
ꢀ
ꢀ
stripping to afford 1. The arene HArTrip was also isolated from
bond lengths that resemble those experimentally observed,
with the lengths calculated for the four-coordinate tin centers
(average 2.935 ꢀ) falling within the above-mentioned range.
The calculated h/e value, 3.83 ꢀ(avg)/3.25 ꢀ(avg) ꢁ 1.18 is
not as great as that seen in 1. However, it is greater than that
in the [Sn9]3ꢀ Zintl ion (h/e = 1.08).[22]
2
the reaction mixture. Compound 1 was characterized by X-
ray crystallography and by EPR spectroscopy. The latter
afforded a relatively broad signal with a g value near 2.031 in
solution in toluene at room temperature. A representation of
the structure of 1[21] is shown in Figure 1. The Sn9 framework
The cluster salts 2 and 3 were synthesized by the reduction
of Sn(Cl)ArMes with one equivalent of KC8 in the presence of
2
0.5 equivalents of AlCl3 or GaCl3. This synthetic approach
arose out of attempts to isolate clusters with both Group 14
and Group 13 atoms in their frameworks. The inclusion of
Group 13 elements was successful with use of the subvalent
gallium (i) precursor “GaI”,[25] but the use of MIIICl3 (M = Al
or Ga) halides has not afforded “mixed” clusters to date. The
products 2 and 3 were obtained as red–black hexagonal
crystals. Although these crystals have a relatively high melting
point, attempts at redissolving them result in decomposition
with the deposition of elemental tin. Single-crystal X-ray
diffraction established the structure that is illustrated for 2 in
+
Figure 2. The Sn10 framework in the [Sn10(ArMes )3] ion is
2
almost identical in both compounds and the structure is based
on a distorted octadecahedron. It may be viewed as being
ꢀ
composed of Sn2+ ions (Sn5) capping Zintl-like [Sn9(ArMes )3]
2
moieties which are related to the neutral species 1 by a one-
electron reduction. The existence of such monoanionic Zintl-
like ions has precedence in the above-mentioned [Ge9{N-
(SiMe3)2}3]ꢀ ion.[16] The framework electron count for the Sn10
Figure 1. Schematic drawing of 1. Selected bond lengths [ꢀ] and
angles [8]: Sn1-C1 2.21(2), Sn1-Sn2 2.944(2), Sn1-Sn3 2.967(2), Sn1-
Sn4A 2.912(2), Sn2-Sn3 2.992(2), Sn2-Sn4 3.008(2), Sn3-Sn4 2.999(2),
Sn3-Sn5 2.914(2), Sn4-Sn5 2.932(2), Sn5-C37 2.21(3); C1-Sn1-Sn2
122.9(4), C1-Sn1-Sn3 119.5(4), C1-Sn1-Sn4A 116.0(4), Sn2-Sn3-Sn4
60.27(4), Sn3-Sn2-Sn4 59.98(4).
has a tricapped, trigonal-prismatic, closo structure, which is
similar to that first reported for the 21-framework-electron
Zintl ion [Sn9]3ꢀ by Corbett in 1983.[22] The most prominent
structural features of trigonal-prismatic clusters of this type
concern the Sn–Sn separations that define the height (h) and
edge (e) of the trigonal prism and the distance from the
capping to prismatic atoms.[23] In 1, the average values for
these distances are h = 4.12(4), e = 3.000(1), and capping
distance 2.94(3) ꢀ, respectively. These parameters afford an
h/e value of 1.37, which is significantly higher than the range
currently known (1.07–1.17) for [E9]3ꢀ Zintl ions with
21 framework electrons.[2] The h/e value is closest to that
recently observed for the anionic substituted germanium
cluster [Ge9{N(SiMe3)2}3]ꢀ (h/e = 1.27),[11] which also has a
closo structure despite the fact that it has 22 framework
electrons for which a nido, monocapped square-antiprismatic
arrangement is predicted by Wadeꢁs rules. Calculations[11] on
the model species [Ge9H3]ꢀ, however, yielded structural
parameters close to those experimentally observed. In con-
trast, the derivatized Zintl ion [Ge9(BiPh2)2]2ꢀ[14] displays a
nido Ge9 framework that is consistent with its 22 framework
Figure 2. Schematic drawing of the cation in 2 (that in 3 is almost the
same). Selected bond lengths [ꢀ] and angles [8]. Sn1-C1 2.194(11),
Sn2-C25 2.198(12), Sn1-Sn3 2.9174(8), Sn1-Sn4 3.010(1), Sn1-Sn5
3.000(1), Sn2-Sn5 3.118(1), Sn2-Sn4 3.122(1), Sn3-Sn4 3.087(1), Sn4-
Sn5 2.9552, Sn4A-Sn4B 2.8929(2), Sn4-Sn4’ 3.185(2); C1-Sn1-Sn3
112.7(3), C1-Sn1-Sn4 115.2(3), Sn3-Sn1-Sn4 61.64(3), Sn4-Sn1-Sn5
57.91(3), Sn3-Sn1-Sn4 61.64(3), Sn4-Sn1-Sn5 57.91(3), Sn4-Sn1-Sn4B
96.14(5), C25-Sn2-Sn5 102.70(3), Sn4-Sn2-Sn5 56.52(2), Sn4A-Sn2-
Sn4B 53.87(4), Sn4-Sn2-Sn4’ 65.22(6), Sn1-Sn3-Sn4 62.09(3), Sn1-Sn4-
Sn5 61.66(3), Sn4’-Sn4-Sn5 114.45(3), Sn1-Sn5-Sn4 67.74(3), Sn2-Sn5-
Sn4 61.82(3), Sn1-Sn5-Sn2 83.84(4).
ꢀ
electrons. The Sn Sn bond length pattern in 1 is similar to that
seen in closo [E9]3ꢀ Zintl clusters where the shortest bonds
(Sn–Sn bond lengths in the range 2.914(2)–2.967(2) ꢀ) are
Angew. Chem. Int. Ed. 2005, 44, 2546 –2549
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2547

Communications
cation is 22, which is agreement with the view that it is derived
[1] J. D. Corbett, Angew. Chem. 2000, 112, 682; Angew. Chem. Int.
Ed. 2000, 39, 670.
[2] T. F. Fꢂssler, Coord. Chem. Rev. 2001, 215, 347.
from monocapping of the original trigonal-prismatic Sn9-
(ArMes
) unit by a tin atom. This capping causes large
3
2
distortions of the original trigonal-prismatic arrangement
seen in the structures of 1 and the [Ge9{N(SiMe3)2}3]ꢀ ion. The
original parallel trigonal Sn3 faces of the prism, which were
separated by a relatively uniform “height” of approximately
4.0 ꢀ between the respective pairs of tin atoms in 1, become
separated by 3.824, 3.946, and 4.594 ꢀ in 2; the longest
distance (between Sn4 and Sn4B) in 2 is now bridged by the
[3] A. Sekiguchi, S. Nagase, The Chemistry of Organic Silicon
Compounds, Vol. 2, Part 1 (Eds.: Z. Rappaport, Y. Apeloig),
Wiley, Chichester, 1998, chap. 3, pp. 119 – 152.
[4] N. Wiberg, P. P. , Power, Molecular Clusters of the Main Group
Elements (Eds.: M. Driess, H. Nꢃth), Wiley-VCH, Weineim,
2004, chap. 25, pp. 188 – 208.
[5] A Schnepf, Angew. Chem. 2004, 116, 680; Angew. Chem. Int. Ed.
2004, 43, 664.
[6] N. Wiberg, H.-W. Lerner, S. Wagner, H. Nꢃth, T. Seifert, Z.
Naturforsch. B 1999, 54, 877.
[7] B. E. Eichler, P. P. Power, Angew. Chem. 2001, 113, 818; Angew.
Chem. Int. Ed. 2001, 40, 796.
[8] A. Schnepf, R. Kꢃppe, Angew. Chem. 2003, 115, 940; Angew.
Chem. Int. Ed. 2003, 42, 911.
[9] A. F. Richards, H. Hope, P. P. Power, Angew. Chem. 2003, 115,
4205; Angew. Chem. Int. Ed. 2003, 42, 4071.
[10] K. W. Klinkhammer, Y. Xiong, S. Yao, Angew. Chem. 2004, 116,
6328; Angew. Chem. Int. Ed. 2004, 43, 6206.
ꢀ
capping Sn5. Elsewhere within the cluster the Sn Sn bond
lengths are within the range 2.760(2)–3.185(2) ꢀ.
In conclusion, we have isolated and characterized two new
types of organotin clusters that are related structurally to the
xꢀ
Sn9
Zintl anions. The isolation of these compounds
strengthens the notion that clusters of Group 14 elements
can demonstrate a great variety of structural types and firmly
establishes a relationship between the Zintl and organo
xꢀ
clusters. The results suggest that further capping of the Sn9
framework should lead to higher clusters. In addition, the
ability of the terphenyl ligands to stabilize 1–3 as well as other
clusters in which the majority of the metal atoms are
unsubstituted[9] is noteworthy.
[11] B. W. Eichhorn, R. C. Haushalter, J. Chem. Soc. Chem.
Commun. 1990, 937.
[12] D. R. Gardner, J. C. Fettinger, B. W. Eichhorn, Angew. Chem.
1996, 108, 2867; Angew. Chem. Int. Ed. Engl. 1996, 35, 2852.
[13] J. Campbell, H. P. A. Mercier, H. Francke, D. P. Santry, D. A.
Dixon, G. J. Schrobilgen, Inorg. Chem. 2002, 41, 86.
[14] A. Ugrinov, S. C. Sevov, J. Am. Chem. Soc. 2002, 124, 2442.
[15] A. Urgrinov, S. C. Sevov, Chem. Eur. J. 2004, 10, 3727.
[16] A. Schnepf, Angew. Chem. 2003, 115, 2728; Angew. Chem. Int.
Ed. 2003, 42, 2624.
[17] A. Sekiguchi, Y. Ishida, Y. Kabe, M. Ichinohe, J. Am. Chem. Soc.
2002, 124, 8776.
[18] B. E. Eichler, P. P. Power, J. Am. Chem. Soc. 2000, 122, 8785.
[19] The assembly of the Pb10 and Pb12 clusters in ref. [10] is also
believed to proceed through lead hydride intermediates.
[20] L. Pu, A. D. Phillips, A. F. Richards, M. Stender, R. S. Simons,
M. M. Olmstead, P. P. Power, J. Am. Chem. Soc. 2003, 125,
11626.
[21] Crystal data for 1·4THF, 2·PhMe, and 3·PhMe obtained with
MoKa (l = 0.7107 ꢀ) radiation at 90 K: 1: monoclinic, C2/c, dark
red needle, a = 28.176(2), b = 26.4492(2), c = 16.673(1) ꢀ, b =
106.743(1)8, V= 11899.6(14) ꢀ3, Z = 4, R1(obs data) = 0.0826;
wR2(all data) = 0.2296; 2: orthorhombic Cmme, black hexagonal
block, a = 22.694(2), b = 29.964(2), c = 13.391(1) ꢀ, V=
9105.6(13) ꢀ3, Z = 4, R1 (obs data) = 0.0608; wR2 (all data) =
0.1913; 3: orthohrombic Cmme, dark red block, a = 22.622(2),
b = 29.917(3), c = 13.376(1) ꢀ, Z = 4, R1 (obs data) = 0.0645;
wR2 (all data) = 0.1816. CCDC-259809–259811 contain the
supplementary crystallographic data for this paper. These data
can be obtained free of charge from the Cambridge Crystallo-
graphic Data Centre via www.ccdc.cam.ac.uk/data_request/cif.
The structure of 1·4THF is disordered with respect to the
twofold axis passing through Sn5 and C37. Division of the
molecule into two parts was unsuccessful. But only the Sn atoms
were assigned anisotropic thermal parameters. The half mole-
cule of 2,6-Trip2C6H3 was restrained using the SAME feature of
SHELXL which relates its 1,2 and 1,3 positions to those of one-
half of the organic moiety. Two molecules of THF are present, in
the asymmetric unit. The molecules 2·PhMe and 3·PhMe are
isostructural and were solved in the space group Cmme (No. 67,
formerly Cmma). They consist of the cation [R3Sn10]+ at a
crystallographic site with mm symmetry, the anion [AlCl4]ꢀ or
[GaCl4]ꢀ at a site of 222 symmetry and disordered PhMe at a site
of 2/m symmetry. The anion is fully ordered, and PhMe was
refined as a rigid group at the multiplicity of the site (0.25
Experimental Section
All manipulations were carried out under an inert atmosphere in
water-free conditions.
1: {Sn(m-H)ArTrip
} (0.601 g, 0.5 mmol) in toluene (30 mL) was
2
2
heated to 1008C in an oil bath for 1 h, during which time the initial
blue color of the solution became dark red. The solution was cooled
and filtered. The volume of the filtrate was reduced to saturation, but
attempts to grow crystals for X-ray diffraction from toluene solutions
were unsuccessful. With the addition of small quantities (ca. 5%
volume) of THF, crystals of 1·4THF suitable for X-ray studies were
obtained as red needles (0.052 g, 19%); m.p. desolvates at ca. 808,
dec > 1008C. Elemental analysis (%) calcd for C108H147Sn9: C 51.64, H
5.90; found: C 52.16, H, 6.03.
2: A solution of Sn(Cl)ArMes (0.467 g, 1 mmol) and AlCl3 (0.67 g,
2
0.5 mmol) in THF (20 mL) was added dropwise to a well-stirred
suspension of KC8 (0.135 g, 1 mmol) in THF (20 mL) cooled in a dry-
ice/acetone bath. The solution was allowed to reach room temper-
ature within 12 h and filtered. The filtrate was collected, the volatile
materials were removed under reduced pressure, and the residue was
extracted with toluene (30 mL) and filtered. The filtrate was reduced
in volume to incipient crystallization and stored at room temperature
for one week to afford 2 as black (dark red in transmitted light)
hexagonal crystals. (0.028 g, 12%); m.p. 144–1468C. Elemental
analysis (%) calcd for C72H75Cl4AlSn10: C 37.66, H 3.29; found: C
38.10, H 3.34.
3: The preparation of 3 was almost identical to that for 2 except
that GaCl3 was employed instead of AlCl3. The product 3 was
obtained as black (dark red in transmitted light) crystals (0.023 g,
10%); m.p. 185–1868C. Elemental analysis (%) calcd for
C72H75Cl4GaSn10: C 36.97, H 3.32; found: C 37.61, H 3.29.
Received: January 12, 2005
Published online: March 22, 2005
Keywords: cluster compounds · density functional calculations ·
.
structure elucidation · tin
2548
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2546 –2549

Angewandte
Chemie
occupancy) with isotropic thermal parameters. In the cation,
there are two anomalies which lower the symmetry of the
cluster. One of the Sn atoms (Sn5) is clearly at half occupancy
and Sn4 occupies two sites in the ratio Sn4:Sn4’:0.75:0.25, which
was established by refinement. The three atoms Sn4, Sn4A, and
Sn4B are symmetry-related positions for the major atom of the
site shared by Sn4 and Sn4’. The minor atom, Sn4’, is drawn such
that it occupies the side of the cluster where Sn5 is absent to
avoid unrealistically short bonds. In the average structure, Sn4’
would occupy all four positions related by mm symmetry, and
Sn5 (which resides on a mirror plane) would be reflected to the
opposite side as well. These various distortions in the cluster may
be the cause of disorder in the R group beginning with the atom
C1. The atoms that make up this group are disordered with
respect to a mirror plane that passes through the ipso carbon C1
and the para carbon C4 of the C1 aryl ring. This gives the
appearance of large thermal motions perpendicular to this plane.
To better describe this group, the atoms C1 and C4 were shifted
off the mirror and the remainder of the atoms in this group
located from a difference Fourier map. Thus, the atoms C1–C24
were refined at 0.5 occupancy with isotropic thermal parameters
and a SHELXL PART-1 statement.
[22] S. C. Critchlow, J. D. Corbett, J. Am. Chem. Soc. 1983, 105, 5715.
[23] T. F. Fꢂssler, U. Schutz, Inorg. Chem. 1999, 38, 1866.
[24] The geometry optimizations of the model tin clusters Sn9Me3 and
[Sn10Me3]+ were performed using DFT with a hybrid B3LYP
functional, including Beckeꢁs three-parameter nonlocal
exchange potential and the nonlocal correlation functional of
Lee–Yang–Parr . To reduce the computational costs without
neglecting the relativistic effects, the effects of the core electrons
on valence shell for Sn were represented by the quasi-relativistic
effective-core potentials (ECPs). For this purpose two ECPs
were used for tin atoms: Los Alamos LanL2DZ pseudopotential
ECP with a valence basis (Sn 46 electrons core ECP, plus
contracted (3s,3p)![2s,2p]) and the large core CRENBL ECP
augmented with polarization functions for C and Sn. Final
optimization was also performed with an all electron 3-21g*
basis set. The DFT calculations were carried out using the
GAUSSIAN03 package and the representations of the molec-
ular structures and molecular orbitals were generated with the
MOLEKEL program. The optimized geometrical parameters
(interatomic distances [ꢀ] and angles [8]) vary slightly with the
basis set size; the Sn–Sn bond lengths are shorter by an average
of 0.05–0.07 ꢀ in the structure optimized at B3LYP/3-21g* level
compared to the structure optimized using ECPs. These
parameters obtained at B3LYP/3-21g* for Sn9Me3 and
[Sn9Me3]+ are in good agreement with those found in the crystal
structure of 1 and 2. Model of 1, Sn9Me3: Sn1-Sn2 2.937, Sn1-
Sn2A 2.937, Sn1-Sn3 2.935, Sn1-Sn4A 2.937, Sn2-Sn3 3.262, Sn3-
Sn4 3.247, Sn2-Sn4 3.254; torsion angle Sn1A-Sn3A-Sn4-Sn5
33.4. Model of 2, [Sn9Me3]+: Sn1-Sn5 2.965, Sn1-Sn3 2.979, Sn1-
Sn4 3.108, Sn4···Sn4B 4.852; Sn3-Sn1A-Sn4 69.4, Sn1-Sn3-Sn3A
50.8, Sn1-Sn4-Sn2 76.8, Sn1-Sn5-Sn2 81.3. Other relevant bond
lengths and angles from the optimized structures are listed in the
Supporting Information.
[25] A. F. Richards, M. Brynda, P. P. Power, Organometallics 2004, 23,
4009.
Angew. Chem. Int. Ed. 2005, 44, 2546 –2549
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2549