Synthesis and fundamental properties of stable Ph3SnSiH 3 and Ph3SnGeH3 hydrides: Model compounds for the design of Si-Ge-Sn photonic alloys

Article abstract of DOI:10.1021/ic900612s

The compounds Ph₃SnSiH₃ and Ph₃SnGeH ₃ (Ph = C₆H₅) have been synthesized as colorless solids containing Sn-MH₃ (M = Si, Ge) moieties that are stable in air despite the presence o

Full text of DOI:10.1021/ic900612s

6314 Inorg. Chem. 2009, 48, 6314–6320
DOI: 10.1021/ic900612s
Synthesis and Fundamental Properties of Stable Ph3SnSiH3 and Ph3SnGeH3
Hydrides: Model Compounds for the Design of Si-Ge-Sn Photonic Alloys
Jesse B. Tice,* Andrew V. G. Chizmeshya, Thomas L. Groy, and John Kouvetakis
Department of Chemistry and Biochemistry, Arizona State University, Tempe, Arizona 85287
Received March 30, 2009
The compounds Ph3SnSiH3 and Ph3SnGeH3 (Ph = C6H5) have been synthesized as colorless solids containing
Sn-MH3 (M = Si, Ge) moieties that are stable in air despite the presence of multiple and highly reactive Si-H and
Ge-H bonds. These molecules are of interest since they represent potential model compounds for the design of new
classes of IR semiconductors in the Si-Ge-Sn system. Their unexpected stability and high solubility also makes
them a safe, convenient, and potentially useful delivery source of -SiH3 and -GeH3 ligands in molecular synthesis. The
structure and composition of both compounds has been determined by chemical analysis and a range of spectroscopic
methods including multinuclear NMR. Single crystal X-ray structures were determined and indicated that both
˚
˚
˚
˚
compounds condense in a Z = 2 triclinic (P1) space group with lattice parameters (a = 9.7754(4) A, b = 9.8008(4) A,
˚
o
o
o
c = 10.4093(5) A, R = 73.35(10) , β = 65.39(10) , γ = 73.18(10) ) for Ph3SnSiH3 and (a = 9.7927(2) A, b = 9.8005(2) A,
o
o
o
˚
c = 10.4224(2) A, R = 74.01(3) , β = 65.48(3) , γ = 73.43(3) ) for Ph3SnGeH3. First principles density functional theory
simulations are used to corroborate the molecular structures of Ph3SnSiH3 and Ph3SnGeH3, gain valuable insight into
the relative stability of the two compounds, and provide correlations between the Si-Sn and Ge-Sn bonds in the
molecules and those in tetrahedral Si-Ge-Sn solids.
Introduction
To date, a limited number of Si1-xSnx samples within the
range of 0.10 < x <0.25 has been demonstrated via re-
actions of SnD4 and trisilane, while the analogous growth
of Ge1-xSnx has been limited to Sn fractions of less than
20 at. %. This suggests that the fabrication of higher Sn
contents approaching 50% in the binaries may require the
development of alternate lower temperature routes ideally
involving single source precursors with direct M-Sn
bonds (M = Si, Ge). Such compounds are likely to be
more reactive and able to furnish the entire molecular
core intact to the crystal lattice of the product. In this
regard, pure heteroatom stannyl-hydrides H3MSnH3 with
preformed Sn-M bonds represent, in principle, the most
suitable precursors for this application. Prior work has
established that access to these molecules is vitiated by
their inherent instability with respect to decomposition
into MH4, elemental Sn, and H2.6 One possible solution to
this problem would be to increase the molecular stability of
the tin functionality in the compound by replacing one or
more of the weak Sn-H bonds with suitable organic
groups such as phenyls (Ph=C6H5). Particularly attractive
targets include the (H3-xPhx)SnSiH3 and (H3-xPhx)
SnGeH3 series of molecules since these are likely to
thermally eliminate the phenyl and the H ligands as
benzene byproducts leading to the delivery of the desired
Recently, we reported the fabrication of new Ge1-xSnx
binary alloys with Sn contents up to 20 at.% grown directly
on Si via reactions of SnD4 and Ge2H6.1 Si incorporation into
these materials to form Si1-x-yGexSny (Sn < 12% at. %)
ternaries was subsequently achieved using new silylgermanes
(GeH3)xSiH4-x as the source of Si and Ge.2,3 The ternary
materials exhibit independently tunable bandgaps and lattice
parameters thereby making it possible to envision entirely
new families of optoelectronic devices ranging from commu-
nication applications (modulators to quantum cascade la-
sers) to sustainable energy systems such as high efficiency
photovoltaics. The emergence of the SiGeSn system have also
led to the development of Si1-xSnx alloys as a logical off-
shoot.4 The latter cover a wide range of band-gaps that are
predicted to become direct for x = 0.30-0.50 leading to
additional applications in high performance IR systems.1,5
*Towhom correspondence shouldbeaddressed. E-mail: jesse.tice@asu.edu.
(1) Bauer, M.; Taraci, J.; Tolle, J.; Chizmeshya, A. V. G.; Zollner, S.;
Smith, D. J.; Menendez, J.; Hu, C.; Kouvetakis, J. Appl. Phys. Lett. 2002, 81
(16), 2992–2994.
(2) Fang, Y-Y; Xie, J.; Tolle, J.; Roucka, R.; D’Costa, V. R.; Chizmeshya,
A. V. G.; Menendez, J.; Kouvetakis, J. J. Am. Chem. Soc. 2008, 130(47),
16095–16102.
(3) Kouvetakis, J.; Menendez, J.; Chizmeshya, A. V. G. Ann. Rev. Mat.
Res. 2006, 36, 497–554.
(4) Tolle, J.; Chizmeshya, A. V. G.; Fang, Y.-Y.; Kouvetakis, J.; D’Costa,
V. R.; Hu, C.-W.; Menendez, J.; Tsong, I. S. T. Appl. Phys. Lett. 2006, 89
(23), 231924/1–231924/3.
(6) Wiberg, E.; Amberger, E.; Cambensi, H. Z. Anorg. Allg. Chem. 1967,
351(3-4), 164–179.
(5) Soref, R. IEEE J. Sel. Top. Quantum Electron. 2006, 12, 1678.
r
2009 American Chemical Society
pubs.acs.org/IC
Published on Web 06/04/2009

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 6315
Si-Sn and Ge-Sn composition in the solid. In our prior
work we have successfully applied a similar strategy to
produce Ge1-xSnx alloys using PhSnD3, as a viable Sn
source in the deposition of thin films via elimination of
benzene according to the following reaction: PhSnD3 f
D2 +Ph(D)+Sn.7
Et3SnGe(C6F5)2H analogue were also reported exclusively
on the basis of FTIR characterizations.16,17 On the basis of
the aforementioned studies it appears that Sn-Si or Sn-Ge
metalorganics containing SiHx and GeHx moieties are rela-
tively unstable or inconclusively characterized, particularly
when multiple hydride bonds are present.
Historically, the creation of molecules containing Si-Sn
bonds and metal hydride functionalities MHx (where x =
1-3) has proven to be very difficult as evidenced by the
paucity of references in the literature concerning their synth-
eses and characterizations. In this regard, the possible for-
mation of Me3SnSiH3 (Me = CH3) and (Me3Sn)3SiH was
reported on the basis of 1H NMR data.8 However no product
yields were given since the materials were highly sensitive to
light and presumably could not be readily isolated for further
analysis. In subsequent work, the presence of trace amounts
of Me3SnSiH3 was observed as a byproduct during the
synthesis of Me3SnSiF3 from a reaction of Me3SnH and
Si2F6.9 We note that further efforts to isolate and characterize
Me3SnSiF3 in viable yields have not been reported, and its
application as a source of pure Si-Sn will likely be hampered
by the presence of thermally reactive CH3 ligands, poten-
tially leading to C contamination. Related compounds with
single Si-H bonds such as HSi(SnMe3)3 and Bu3SnSi(H)Me2
(Bu = n-C4H9) were prepared and characterized by physio-
chemical methods; however, their molecular structures were
not reported.10,11 In more recent work, the successive buildup
of SiH3 groups around a central tin atom was purportedly
achieved using reactions of NaSiH3-n(SiH3)n with SnH4 to
yield silyl-substituted sodium stannides with a proposed
formula of NaSnH3-n(SiH3)n.12 These materials were found
to be marginally stable in glyme solutions on a time scale
sufficient to obtain NMR data but were never isolatedin pure
form. Finally, a more stable series of hydrido-substituted
stannylsilanes including the HMe2SiSnPh2Me, HSiR2Bu2Sn-
Cl (R = i-prop, t-but), and H-SiR02-R2Sn-Z (R0 = Me,
i-prop, t-but, R = Me, t-but, and Z = H, Me) have also
been reported.13,14
In this paper, we report the synthesis of the simple and
remarkably robust Si-Sn and Ge-Sn compounds with
compositions Ph3SnSiH3 and Ph3SnGeH3. These represent
the first example of molecules within this class containing
intact silyl and germyl groups that are stable in air. Also,
these compounds can be obtained in single crystalline form
thereby enabling their unequivocal identification by X-ray
diffraction (XRD) determination of their molecular struc-
tures. Theoretical simulations were used to corroborate the
molecular structures, elucidate their intrinsic thermochemical
properties, and predict their relative stability in relation to
both standard states and hypothetical SiSn and GeSn com-
pounds.
Results and Discussion
The preparation of Ph3SnSiH3 was initially explored using
metathesis reactions involving H3Si(SO3CF3) and common
alkali silanide salts such as KSiH3. The main drawback with
the use of the latter in such nucleophilic addition reactions
was the limited stability of the salt in tetrahydrofuran (THF)
solutions resulting in its decomposition into higher order
silanides and unstable Si-H polymers. In the case of the
triflate, H3Si(SO3CF3), the primary limitation precluding its
successful use is the rapid decomposition of the compound
above -10 °C to form highly explosive SiH4 byproducts. In
view of these difficulties, we next employed H3Si(SO3C4F9)
(a butane analogue of the silyl triflate) which as shown below,
represents a highly stable and therefore more practical
source of SiH3 compared to both H3Si(SO3CF3) and KSiH3.
Under optimal conditions the stoichiometric reaction of
H3Si(SO3C4F9)18 and Ph3SnLi19 readily produced the tar-
get compound as a crystalline solid via elimination of Li-
(SO3C4F9) (see eq 1):
While, to the best of our knowledge, the above examples
represent the only extant organotin silanes in the literature,
instances of the analogous organotin germanes containing
Ge-H bonds are even rarer. For example Me3SnGeH3
was prepared in minor amounts and was only characterized
by 1H NMR.15 Attempts to prepare the ethyl deriva-
tive Me3SnGeEt2H (Et = C2H5) and the closely related
Ph3SnLi þ H3SiðSO3C4F9ÞfPh3SnSiH3
þ LiðSO3C4F9Þ
ð1Þ
A minor limitation associated with the use of the H3Si
(SO3C4F9) was its propensity to react slowly with the THF
solvent to form polymeric byproducts leading to its gradual
degradation under prolonged reaction times. This difficulty
was overcome by using stringent reaction conditions invol-
ving the slow addition of H3Si(SO3C4F9) to a THF solution
of Ph3SnLi at -35 °C followed by stirring the resulting
suspension at 22 °C for 1 h. This led to the formation of
the desiredproductin∼40% yield. We note that our attempts
to produce Ph3SnSiH3 using solutions of Ph3SnLi in solvents
other than THF (such as diethyl ether and glymes) produced
large amounts of SiH4 resulting from the decomposition
of the H3Si(SO3C4F9) starting material. Additionally, we
(7) Taraci, J.; Zollner, S.; McCartney, M. R.; Menendez, J.; Santana-
Aranda, M. A.; Smith, D. J.; Haaland, A.; Tutukin, A. V.; Gundersen, G.;
Wolf, G.; Menendez, J.; Kouvetakis, J. J. Am. Chem. Soc. 2001, 123(44),
10980–10987.
(8) Amberger, E.; Muehlhofer, E. J. Organomet. Chem. 1968, 12(1), 55–62.
(9) D’Errico, J. J.; Sharp, K. G. J. Chem. Soc., Dalton Trans: Inorg. Chem.
1989, 9, 1879–18781.
(10) Heyn, R. H.; Tilley, T. D. Inorg. Chem. 1990, 29(20), 4051–4055.
(11) Lahournere, J. C.; Valade, J. C. R. Sea. Acad. Sci., Ser. C: Sci. Chim.
1970, 270(25), 2080–2082.
(12) Lobreyer, T.; Sundermeyer, W.; Oberhammer, H. Chem. Ber. 1994,
127(11), 2111–2115.
(13) Uhlig, F.; Kayser, C.; Klassen, R.; Hermann, U.; Brecker, L.;
Schuermann, M.; Ruhland-Senge, K.; Englich, U. Z. Natur. B: Chem. Sci.
1999, 54(2), 278–287.
(14) Englich, U.; Prass, I.; Schollmeier, T.; Uhlig, F. Monatsh. Chem.
(17) Bochkarev, M.; Maiorova, L.; Razuvaev, G. Zh. Obshch. Khim.
2001, 627(3), 453–457.
1980, 50, 903–907.
(15) Angus, P. C.; Stobart, S. R. J. Chem. Soc., Dalton Trans: Inorg.
Chem. 1973, No. 21, 2374–2380.
(18) Tamborski, C.; Ford, F. E.; Soloski, E. J. J. Org. Chem. 1963, 28,
181–184.
(16) Bravo-Ahivotovskii, D.; Biltueva, I.; Vyazankina, O.; Vyazankin, N.
(19) Lobreyer, T.; Oeler, J.; Sundermeyer, W. Chem. Ber. 1991, 124(11),
Izv. Akad. Nauk SSSR Ser. Khim. 1985, 1214.
2405–2410.

6316 Inorganic Chemistry, Vol. 48, No. 13, 2009
Tice et al.
observed no product formation when an excess of either
Ph3SnLi or H3Si(SO3C4F9) was employed indicating that any
non-stoichiometric mixture increases the decomposition rate
of both reactants. Collectively the above results indicate that
the metathesis mechanism leading to the desired product is
only operative over a very narrow range of reaction condi-
tions, most notably temperature, solvent type, and reactant
ratio. The Ph3SnSiH3 product was separated and isolated
from the main reaction byproduct Li(SO3C4F9) and a side
reaction impurity Ph4Sn as a colorless solid exhibiting a sharp
melting point of 86 °C.
Table 1. Summary of Data for Ph3SnSiH3 and Ph3SnGeH3
Ph3SnSiH3
Ph3SnGeH3
NMR Data
δ MH3 (ppm)
δ29Si (ppm)
119Sn (ppm)
1JSi-H (Hz)
2JSn-H (Hz)
3.66
-106
-133.6
200
3.50
-134.2
56
82
FTIR Data
ν C-H (cm-1
)
)
3100-3000
2164, 2130
3100-3000
The GeH3 analogue [Ph3SnGeH3] of the above Ph3SnSiH3
compound was synthesized via a stoichiometric combina-
tion of freshly prepared BrGeH3 and Ph3SnLi dissolved in
THF (eq 2):
ν Si-H (cm-1
ν Ge-H (cm-1
ν Sn-M (cm-1
)
)
2080, 2045
273
315
Elemental Analysis
calculated C and H
observed C and H
56.70, 4.76
56.92, 4.95
50.77, 4.26
50.63, 4.27
Ph3SnLi þ BrGeH3fPh3SnGeH3 þ LiBr
ð2Þ
The reaction is stirred for 15 h to produce Ph3SnGeH3
(50% yield) which was isolated from the main reac-
tion byproduct LiBr and a side reaction impurity Ph4Sn as
a colorless solid exhibiting a sharp melting point of 82 °C.
Note that shorter reaction times (3-5 h) led to minimal
amounts of product while longer times did not enhance the
overall yield.
Crystal Data
chemical formula
formula weight (amu)
DCALC (g/cm3)
C18 H18 Si Sn
381.1
1.486
C18 H18 Ge Sn
425.7
1.647
3
˚
volume (A )
setting/space group
851.95
triclinic, P1
9.7754
9.8008
10.4093
73.347
65.388
73.184
1.558
2
123
2.51
6.41
857.97
triclinic, P1
9.7927
9.8005
10.4224
74.001
65.482
73.427
3.195
2
123
4.18
11.01
˚
a (A)
˚
b (A)
˚
c (A)
Single crystals of the Ph3SnSiH3 and Ph3SnGeH3 were
grown from a hexane solution of the compounds and subse-
quently used to perform a complete characterization by
XRD, multinuclear NMR, IR spectroscopy, combustion
analysis, and mass spectrometry. Collectively, the results
indicate that the molecules are isostructural and consist of
MH3 and SnPh3 groups connected with single Sn-M bonds.
This close correspondence in molecular architecture affords a
straightforward comparison of the structural and spectro-
scopic properties as summarized in Table 1. Accordingly,
here we provide a detailed interpretation of the spectro-
scopic data for Ph3SnSiH3 and refer the reader to Table 1
and the experimental section for further details regarding the
Ph3SnGeH3 analogue.
R (deg)
β (deg)
γ (deg)
μ (cm-1
)
Z
measurement T(K)
R1a (%, all data)
wR2b (%, all data)
P
P
P
P
a R1 = ||Fo| - |Fc||/ |Fo| ꢀ 100. b wR2 = { w(Fo2 - Fc2)2/
w
(Fo2)2}1/2 ꢀ 100.
1950-900 cm-1, respectively, while the SiH3 asymmetric/
symmetric stretching modes were observed at 2164/
2130 cm-1 as expected. The Si-Sn bond stretch was
observed at 314 cm-1 and closely matches the phonon
frequencies as measured by Raman in the diamond cubic
Si1-xSnx alloys. The C and H combustion analysis yielded
values essentially identical to those expected (within ∼3%).
The Ph3SnSiH3 compound decomposes in the mass spectro-
meter to yield the (C6H5)3Sn ion as the highest peak at ∼ 350
amu indicating that the Si-Sn bond is unstable under these
conditions. In contrast, the parent ion of the Ge analogue
Ph3SnGeH3 was seen at 426 amu demonstrating that the
intact Ge-Sn bond is stronger than its Si-Sn counterpart.
This is also consistent with the bonding properties predicted
by quantum chemical calculations of the compounds,
as described below.
1
The H NMR data of Ph3SnSiH3 revealed a singlet at
3.66 ppm corresponding to the SiH3 protons and multiplets
at 7-7.5 ppm due to the ortho, meta, and para protons the Ph
group. Satellite peaks induced by the 117/119Sn nuclei are
found to be symmetrically distributed by 56 Hz with respect
to their 1H singlet, in analogy with those previously reported
for HSi(SnMe3)3 and the HSiR2’SnR3 compounds.10,14
The corresponding29Si satellite peaks were observed at 200 Hz
which is a typical value for silyl-germyl hydrides such as
(GeH3)xSiH4-x
.
20 The 119Sn chemical shift at ∼ -134 ppm is
characteristic of multiple electron withdrawing phenyls
bonded to the Sn atom while the 29Si shift at -106 ppm is
1
typical for conventional silylgermanes. The 2D H COSY
Crystallographic analyses of Ph3SnSiH3 and Ph3Sn-
GeH3 indicates that the compounds are monomeric in
the solid state. The MH3 ligands are bonded exclusively to
the Sn center which in turn is surrounded by three phenyl
rings arranged in a “paddle” or propeller-like geometry.
Both compounds crystallize in a triclinic unit cell of P1
symmetry (see Table 1) containing two identical molecules
(Z = 2) such as that shown in Figure 1(a) with a thermal
ellipsoid model (50% probability). Table 2 contains
selected bond distances and angles for the molecular units.
spectrum indicated no connectivity between the Si-H singlet
of the SiH3 (3.66 ppm) and the C-H multiplets of the SnPh3
(7-7.5 ppm) while the 29Si-1H HMQC spectrum con-
firms that the Si atom at -106 ppm is bonded exclusively
to the protons at 3.66 ppm. Collectively the correlated
NMR data corroborate the proposed molecular structure
of Ph3SnSiH3. The IR spectra showed the character-
istic vibrations of the phenyl ring at 3100-3000 cm-1 and
(20) Ritter, C. J.; Hu, C.; Chizmeshya, A. V. G.; Tolle, J.; Klewer, D.;
Tsong, I. S. T.; Kouvetakis, J. J. Am. Chem. Soc. 2005, 127(27), 9855–9864.
˚
˚
The Si-Sn (2.556 A) and Ge-Sn (2.583 A) bond distances

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 6317
Figure 1. (a) Molecular structure of Ph3SnSiH3. (b) Coordination of phenyl groups around the -SiH3 group within the solid Ph3SnSiH3 hydride,
˚
illustrating a “cage-like” structure around the Si atom with a mean diameter of ∼3 A.
˚
density functional theory-based simulation, which allows
a comparison of the molecular properties on an equal
footing. The simulations focus on the individual mole-
cules rather than the extended solid state structures
since the latter are composed of weakly bound mono-
meric units and a common packing motif. We estimated
the relative stability of the title compounds using
two decomposition pathways. The first, given by eq 3,
describes the hypothetical and complete dissociation
of the molecules to form stoichiometric SiSn and GeSn
zinc blende phases via elimination of stable benzene
byproducts:
Table 2. Selected Inter-Atomic Distances (A) and Angles (deg) for Ph3SnSiH3
and Ph3SnGeH3
Ph3SnSiH3
Sn(1)-Si(1)
Sn(1)-C(11)
Sn(1)-C(21)
Sn(1)-C(31)
Si(1)-H(1)
2.556(8)
2.140(3)
2.147(3)
2.145(3)
1.36(4)
1.36(4)
1.32(5)
116.88(7)
C(31)-Sn(1)-Si(1)
C(21)-Sn(1)-Si(1)
Sn(1)-Si(1)-H(1)
Sn(1)-Si(1)-H(2)
Sn(1)-Si(1)-H(3)
H(1)-Si(1)-H(2)
H(1)-Si(1)-H(3)
H(2)-Si(1)-H(3)
107.12(7)
108.15(7)
114.0(16)
108.3(15)
110(2)
105(2)
115(3)
104(3)
Si(1)-H(2)
Si(1)-H(3)
C(11)-Sn(1)-Si(1)
Ph3SnGeH3
Sn(1)-Ge(1)
Sn(1)-C(11)
Sn(1)-C(21)
Sn(1)-C(31)
2.583(7)
2.149(4)
2.142(4)
2.145(4)
C(11)-Sn(1)-Ge(1)
C(21)-Sn(1)-Ge(1)
C(31)-Sn(1)-Ge(1)
115.88(11)
107.73(11)
106.72(12)
Ph3SnMH3fM0:5Sn0:5ðsÞ þ 3C6H6ðgÞ
þ ΔHðaÞ M ¼ Si, Ge
ð3Þ
Here we consider the most plausible and perhaps
simplest decomposition pathway although other mec-
hanisms involving ligand exchange are also possible. To
explore this notion we conducted a preliminary experi-
ment in which we heated the bulk solid to temperatures
near ∼150 °C in a glass vessel under an atmosphere of
N2. This produced a solid residue and a mixture of
volatile byproducts such as benzene and variable
amounts of Ph2GeH2 where the ratio of the compounds
could be controlled by the reaction conditions. This
result provides a proof-of-concept corroboration of
the decomposition described by eq 4, and suggests that
reaction kinetics can be further tuned to promote the
benzene decomposition channel. We anticipate that
future gas phase decomposition reactions under kineti-
cally controlled CVD conditions will likely lead to the
desired result.
are typical for crystalline compounds containing M-Sn
bonds.21-23 The C-Sn bond distances range from 2.12 to
˚
2.14 A and are nearly identical to those found in related
organometallics of Sn including SnPh4.24 The inherent
internal structure of the phenyl rings remain undistorted
as evidenced by the close similarity of ortho, para, and
meta bond lengths. The various M-Sn-C and C-Sn-C
bond angles range from 107 to 116° revealing a nominal
deviation from tetrahedral geometry about the Sn center.
A close examination of the packing diagram indicates that
the phenyl rings essentially form arrange a protective cage
composed of hexagonal panels (paddles) encapsulating
the MH3 groups in the extended structure as illustrated in
Figure 1(b). This arrangement may explain the unusual
stability of the solid state samples and their lack of any
discernible decomposition in the presence of oxygen and
moisture for prolonged time periods of over at least
several months.
We also consider a second more thermodynamically
relevant reaction based on the decomposition of the
compounds into conventional standard states for all of
the constituent elements, as shown in eq 4, thus approxi-
mating the standard heat of formation:
Simulation Studies of Structure and Stability. To
further corroborate the structural trends discussed above
and examine the relative stability of the Ph3SnSiH3 and
Ph3SnGeH3 compounds we carried out a systematic
Ph3SnMH3fMðsÞ þ SnðsÞ þ 18CðsÞ þ 9H2ðgÞ
(21) Baumgartner, J.; Fischer, R.; Fischer, J.; Wallner, A.; Marschner, C.;
Flrke, U. Organometallics 2005, 24(26), 6450–6457.
(22) Sharma, H. K.; Cervantes-Lee, F.; Parkanyi, L.; Pannell, K. H.
þ ΔHðbÞ M ¼ Si, Ge
ð4Þ
Organometallics 1995, 15, 429–435.
(23) Parkanyi, L.; Kalman, A.; Pannell, K.; Sharma, H. J. Organomet.
Chem. 1994, 484, 153–159.
In the above reactions we first computed the ground state
electronic energies of the elemental Si, Ge, Sn, (diamond
structures), C (graphite), and the binary phases SiSn and
(24) Chieh, P. C.; Trotter, J. J. Chem. Soc. A 1970, 6, 911–914.

6318 Inorganic Chemistry, Vol. 48, No. 13, 2009
Tice et al.
GeSn (zinc blende). We employed the VASP code25 to
optimize the structures using the generalized gradient
approximation (GGA)26 for exchange and correlation.
All elements were treated using standard PAW pseudo-
potentials27 supplied with the package, and d-electron
valence states were included for the heavier Ge and Sn.
An energy cutoff of 450 eV was used in all calculations,
and 80-190 irreducible k-points were used for the reci-
procal space quadratures. Stringent energy convergence
tolerances and expanded Fourier grids yielded optimized
structures with residual atomic forces and cell stresses less
than 10-5 eV/A and 0.001 kB, respectively. The above
˚
computational conditions yielded ground state energies:
E[C(s)]=-9.1860 eV, E[Si(s)]=-5.4308 eV, E[Ge(s)]=
-4.4884 eV, E[Sn(s)] = -3.8033 eV, E[SiSn(s)] =
-8.9590 eV, and E[GeSn(s)]=-8.2632 eV. We note that
the reference state for these energies is that of the spin-
compensated neutral atom constituents.
Figure 2. Graphical depiction of the electronic energies involved in the
disproportionation and formation reactions described by eq 3 and eq 4,
respectively.
and translational effects for both the gas phase and solid
state systems. In the context of reactions such as those
described by eq 3 and eq 4, which involve the condensa-
tion of large gas phases species into solid form, the change
in zero-point energy (ZPE) represents a dominant
contribution. To compute the latter for the gas phase
molecules we employed the Gaussian03 code using the
same DFT functional as in the VASP calculations de-
scribed above, while for the solids the ZPE correction was
obtained from the phonon spectrum of the ground state
solids using a well-established procedure.28 In particular,
the dynamical matrix of the solids was calculated using
special “frozen phonon” displacements in a supercell,
large enough to converge the relevant interaction energies
(to be described in detail elsewhere). These methodologies
provided the following ZPE estimates (in kJ/mol) for the
extended solid phases: Si(5.92), Ge(3.37), Sn(2.18), SiSn
(8.09), GeSn(5.59), and C(graphite,6.38), while for the
gas phases species we obtained: Ph3SnSiH3(779.8),
Ph3SnGeH3(775.2), C6H6(281.6), and H2(27.5). Combin-
ing the preceding values with the electronic reaction
energies described above (Figure 2), we find ZPE-
corrected disproportionation reactions energies for
Ph3SnSiH3 and Ph3SnGeH3 to be -181 kJ/mol and
-242 kJ/mol, respectively. The corresponding heats of
formation are found to be +409 kJ/mol for Ph3SnSiH3
and +412 kJ/mol for the Ge analogue. Perhaps not
surprisingly, the latter value is quite similar to the stan-
dard heat of formation of related compounds such as
tetraphenyl germanium (+445 kJ/mol) and triphenyl
phenylethynyl germanium (+579 kJ/mol).29 We note that
the inclusion of ZPE corrections has not altered the basic
trends illustrated in Figure 2; the driving force for the
formation of MSn alloys is larger for the Ge compound
than for its Si analogue, and the heat of formation
of Ph3SnGeH3 is predicted to be marginally larger
(∼ 3 kJ/mol) than that of Ph3SnSiH3. The optimized
structures of Ph3SnSiH3 and Ph3SnGeH3, including their
key bond lengths and bond angles are shown in Figure 3
while Table 3 compares the measured and simulated M-Sn
bond length data for the compounds with those in the
corresponding SiSn and GeSn extended solids. As can be
seen from the data, all simulated bond lengths are typi-
cally 1-2% larger than the experimental counterparts,
For the gas phase reactants (Ph3SnSiH3, Ph3SnGeH3)
and products (H2, C6H6) the electronic energies of the
optimized structures were obtained using the same
computational settings as described for the solids
(e.g., cutoffs, k-point grids and tolerances), but by placing
the individual molecules into a large supercell of fixed
˚
edge length 10-15 A to eliminate self-interactions
between molecules. The structure of a given species is
then obtained by optimizing only the atomic degrees of
freedom yielding the following molecular energies:
E[Ph3SnSiH3(g)] = -234.5933 eV, E[Ph3SnGeH3(g)] =
-233.3235 eV, E[H2(g)]=-6.7845 eV, and E[C6H6 (g)]=
-76.0555 eV. In the foregoing and subsequent
discussions, the individual electronic energies of the
molecules and solids are given in eV (96.485 kJ/mol),
while energy differences (reaction energies) are given
in kJ/mol.
Using the above data the reaction energies to form
solid SiSn and GeSn from Ph3SnSiH3 and Ph3SnGeH3,
respectively, are estimated from eq 3 to be -244 kJ/mol
and -300 kJ/mol. These large values indicate a strong
tendency of both molecules to disproportionate and
produce the target solids, and suggest that the driving
force to form the GeSn phase from Ph3GeSnH3 is greater
than that of the Si analogue by ∼0.58 eV, or about
56 kJ/mol. Finally, the negative energy values obtained
from eq 4 provide crude approximations to the heats of
formation of Ph3SnSiH3 and Ph3SnGeH3 of +101 and
+133 kJ/mol, respectively, indicating that the latter
compound is slightly more stable. The relationship
between the reaction energies in eq 3 and the forma-
tion heats (eq 4) is illustrated graphically in Figure 2,
which shows that the energy differences between
the Ph3SnSiH3 and Ph3SnGeH3 is larger than either
that of the alloys (eq 3) or the standard state pro-
ducts (eq 4).
While the above data provides some basic insight into
molecular stability and disproportionation reaction
trends, a quantitative description of the thermochemistry
requires a systematic inclusion of vibrational, rotational,
::
(25) (a) Kresse, G.; Furthmuller, J. Phys. Rev. B 1996, 54, 11169.
(b) Kresse, G.; Hafner, J. J. Phys: Condens. Matter 1992, 6, 8245.
(26) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson,
M. R.; Fiolhais, C. Phys. Rev. B 1992, 46, 6671.
(28) Atsushi. T.; Fumiyasu, O.; Tanaka, I. Phys. Rev. B, 2008, 78,
(27) Blochl, P. E. Phys. Rev. B 1994, 50, 17953.
134106-1-9.

Article
Inorganic Chemistry, Vol. 48, No. 13, 2009 6319
molecules and solid state analogues is essentially
corroborated by our simulations.
Conclusion
We have demonstrated for the first time the synthesis
of robust and air-stable Ph3SnSiH3 and Ph3SnGeH3 com-
pounds containing both direct Sn-Si and Sn-Ge bonds and
the typically highly reactive GeH3 and SiH3 moieties intact.
We ascribe this unexpected stability in part to the cry-
stallographic packing in the solid state structures where the
latter groups are protected by encapsulating phenyl cages.
The compounds were fully analyzed by chemical and spectro-
scopic methods to establish their elemental composition
and bonding properties. XRD single crystal determi-
nation of the molecular structures revealed that the solids
are isostructural. Theoretical simulations indicated that
the compounds should disproportionate readily to form
the corresponding SnSi and SnGe stoichiometric solids via
elimination of stable C6H6 byproducts. These experiments
will be pursued as part of a comprehensive study in the near
future to produce Sn-rich semiconductors. We also estab-
lished experimental and theoretical correlations between the
molecular Si-Sn and Ge-Sn bonds in the title compounds
and those in the extended binary solids thereby providing
guidance for future studies involving the development
of semiconductors based on the Si-Ge-Sn optoelectronic
system.
Figure 3. Molecular structure of Ph3SnMH3 molecules obtained from
simulation (bond lengths given in Angstrom). The primary differences
occur in the Sn-M and M-H bonds, while the remaining structure
includingthe internal structure andrelative orientation ofthe phenyl units
remains essentially unchanged (all C-C-C angles within the rings exhibit
an average value of 120.0 ( 0.6°).
Table 3. Comparison of Observed and Calculated Values M-Sn Bond Lengths in
the Title Compounds and Extended SiSn and GeSn Solid Alloysa
˚
˚
˚
˚
M-Sn (A) M-Sn (A)
M-Sn (A) M-Sn (A)
system
simulated experimental system simulated experimental
Experimental Section
Ph3SiSnH3
Ph3GeSnH3
2.582
2.621
2.556
2.583
SiSn
2.625
2.582
2.56*
2.631
2.61*
General Considerations. All manipulations were carried out
under inert atmosphere (N2 or He) using standard high vacuum
line and drybox techniques. Dry, air-free solvents were distilled
from either anhydrous CaCl2 or sodium benzophenone ketyl
prior to use. The NMR spectra were collected on a Varian Inova
500 MHz spectrometer. The samples were dissolved in deuter-
ated toluene, and the Me4Si or Me4Sn compounds were used as
an internal reference. The FTIR spectra were recorded on a
Nicolet-Magna IR 550 spectrometer either as a Nujol mull
between KBr plates or in a 10 cm gas cell fitted with KBr
windows. Elemental analyses were performed by Columbia
Analytical (Tucson, Arizona). Mass spectrometry (Direct In-
sertion EIMS) data were obtained using a JEOL JMS-GC Mate
II spectrometer in the ASU departmental mass spectrometry
facility. Lithium metal, HSO3C4F9, PhSiCl3 (Aldrich), Ph3SnCl
(Alfa Aesar), and electronic grade GeH4 gas (donated by
Voltaix, Inc.) were used as received. Bromine (Alfa Aesar)
was stored over P2O5 prior to use. The starting materials
triphenyltin lithium, silyl nonafluorobutanesulfonate, bromo-
germane, and phenylsilane were prepared according to litera-
ture procedures.18,19,30,31
2.599*
2.684
2.657*
GeSn
a The experimental values for the latter are obtained from a linear
interpolation of the elemental Si, Ge and R-Sn crystal structures. The
values containing the asterisk represent alloy bond lengths scaled by a
factor 0.99.
which is a well-known artifact of the GGA26 used in DFT.
Nevertheless, these data corroborate the observed trend
that the M-Sn bond length in Ph3SnGeH3 is larger than
that in Ph3SnSiH3, as expected. We note that “experi-
mental” Si-Sn and Ge-Sn bond lengths for the hypothe-
tical SiSn and GeSn alloys (shown in Table 3) were
deduced from the known values of the Si, Ge, and Sn
diamond structures by linearly interpolating between the
homonuclear bond lengths in the constituent elemental
solids (Vegard’s law).
It is generally well-established that the bond length
between a given pair of atoms in a regular solid are
slightly dilated (∼1%) in comparison to the same bond
in a molecular setting. For example, the observed Si-Si
Synthesis of Ph3SnSiH3. To a stirring solution of Ph3SnLi
(4.68 g, 13 mmol/35 mL THF), a stoichiometric amount of H3Si-
(SO3C4F9) (4.32 g, 13 mmol) was added dropwise at -35 °C. The
reaction flask was warmed to 23 °C, and the solution stirred for 1
h. The volatiles were removed in vacuo, and the resultant
solid was extracted with dry toluene to dissolve the product
and minor amounts of tetraphenyl tin. The latter impurity
was precipitated out by cooling to -30 °C and separated
from the solution via filtration. The toluene was then removed
from the filtrate in vacuum to yield a white solid which was
recrystallized from hexane at -30 °C to produce transparent
˚
bond length 2.33 A in disilane Si2H6 is about 1% smaller
than that in the corresponding cubic silicon structure
˚
(2.352 A). A similar reduction in bond length is found
˚
˚
between elemental Ge (2.45 A) and digermane (2.41 A).
On the basis of this observation, we have included in
Table 3 that the bond lengths scaled by the factor of 0.99
for the stoichiometric SiSn and GeSn alloys (indicated by
asterisks) which, agree remarkably well with the observed
Si-Sn and Ge-Sn bond lengths in the Ph3SnSiH3
and Ph3SnGeH3 molecules, respectively. Finally, we note
that the same scaling and correspondence between the
(29) Carson, A. S.; Jamea, E. H.; Laye, P. G.; Spencer, J. A. J. Chem.
Thermodyn. 1988, 20, 1223–1229.
(30) Swiniarski, M. F.; Onyszchuk, M. Inorg. Synth. 1974, 15, 157–161.
(31) Zech, J.; Schmidbauer, H. Chem. Ber. 1990, 123, 2087–2091.

6320 Inorganic Chemistry, Vol. 48, No. 13, 2009
Tice et al.
1
crystals of Ph3SnSiH3 2(2.00 g, 40% yield). H NMR: δ 3.66
and associated software.32,33 Three sets of 606 frames (0.3 wide in
ω with 10 s exposure) were taken at j = 0°, 120°, and 240° for
both materials. The resulting 1818 frames for each compound
were then integrated (SAINT) and corrected for absorption
(SADABS). This procedure produced 6935 and 6689 measured
reflections for Ph3SnSiH3 and Ph3SnGeH3, respectively. The
space group was determined to be P1 (triclinic, No. 2) for both
materials using the XPREP module contained in SHELXTL,
giving 3014 and 3030 unique reflections for Ph3SnSiH3 and
Ph3SnGeH3, respectively. Both structures were solved using stan-
dard runs of XS and refined using XL. After an initial refinement
of the anisotropic thermal parameters for the heavy atoms, the
hydrogens were then added geometrically to the phenyl groups.
Subsequent refinement produced peaks in the Fourier difference
map at expected positions and electron densities for the H atoms
of the SiH3 group of Ph3SnSiH3. The positions and isotropic
thermal parameters of the latter were then included in the refine-
ment parameters resulting in a final R1-value of 0.0251 for 193
parameters and 2912 reflections having Iobs g 2σ(I). For the
Ph3SnGeH3 compound the analogous H peaks were not visible
in the difference map about the Ge atom. In this case hydrogen
atoms were simply added geometrically and allowed to refine as a
1
(s, JSi-H = 200 Hz, JSn-H = 56 Hz, 3H, Si-H3), δ 7.13
(m, 9H, m-Ph, p-Ph), 7.52 (m, 6H, o-Ph, JSn-H = 4 Hz). 29Si
3
NMR: δ -106. 119Sn NMR: δ -133.6. Anal. Calcd for
C18H18SnSi: C, 56.70; H, 4.76. Found: C, 56.92; H, 4.95. Mp =
86 °C. EIMS (m/e): isotopic envelope centered at 351 (M+
- SiH3). IR (film, cm-1): 3135 (vw), 3066 (w), 3047 (w), 3020
(w), 3002 (w), 2164 (s), 2130 (s), 1955 (w), 1883 (w), 1852 (w),
1769 (vw), 1640 (vw), 1579 (w), 1479 (s), 1429 (vs), 1376 (w),
1331 (m), 1301 (m), 1262 (w), 1192 (vw), 1156 (vw), 1072 (s),
1022 (m), 999 (m), 973 (vw), 918 (m), 845 (vs), 729 (vs), 697 (vs)
453 (s), 439 (s), 314 (m).
Synthesis of Ph3SnGeH3. BrGeH3 (1.70 g, 11 mmol) was
condensed at -196 °C into a 100 mL Schlenk flask containing
a THF solution of Ph3SnLi (4.1 g, 11 mmol/50 mL THF). The
flask was first warmed to -40 °C and stirred for 1 h at this
temperature to produce a dark solution. The latter was then
warmed to room temperature and stirred for an additional 14 h
after which the volatiles were removed in vacuo to leave behind
a white solid. This was extracted with toluene to dissolve
the product and minor amounts of tetraphenyl tin impurity.
Crystalline Ph3SnGeH3 was obtained (2.40 g, 50% yield)
following the same procedure described above for the Ph3Sn-
ꢀ
˚
rotatingrigidgroupatafixeddistanceof1.55AfromGe, resulting
in the final R1-value of 0.0418 for 182 parameters and 2710
reflections having Iobs g 2σ(I).
SiH3 analogue: 1H NMR: δ 3.50 (s, JSn-H = 82 Hz, 3H,
2
Ge-H3), δ 7.11 (m, 9H, m-Ph, p-Ph), 7.49 (m, 6H, o-Ph).
119Sn NMR: δ -134.2. Anal. Calcd for C18H18SnGe: C, 50.77;
H, 4.26. Found: C, 50.63; H, 4.27. Mp = 82 °C. EIMS (m/e):
isotopic envelope centered at 426 (M+). IR (film, cm-1): 3135
(vw), 3066 (w), 3047 (w), 3016 (vw), 2080 (s), 2045 (s), 1960 (w),
1880 (vw), 1830 (vw), 1769 (vw), 1644 (vw), 1581 (w), 1482 (m),
1431 (s), 1381 (vw), 1336 (m), 1301 (m), 1262 (w), 1195 (vw),
1162 (vw), 1077 (s), 1027 (m), 1002 (m), 888 (w, br), 778 (vs), 733
(vs), 703 (vs) 455 (s).
Acknowledgment. This research was supported by the
U.S. Air Force Office of Scientific Research (AFOSR
MURI, FA9550-06-01-0442), Voltaix Corporation, and NSF
(IIP 075049).
Supporting Information Available: Crystallographic data for
Ph3SnSiH3 and Ph3SnGeH3 (CIF). This material is available
free of charge via the Internet at http://pubs.acs.org.
X-ray Structure of Ph3SnSiH3 and Ph3SnGeH3. Colorless
crystalline blocks of Ph3SnSiH3 (0.34 mm ꢀ 0.34 mm ꢀ
0.22 mm) and Ph3SnGeH3 (0.27 mm ꢀ 0.21 mm ꢀ 0.15 mm)
were mounted on glass fibers using Apiezon grease for subse-
quent data collection. The data for both compounds were
collected at 123 K using Bruker APEX type diffractometers
(32) SMART (Version 5.632), SAINT (Version 6.45a), SHELXTL (Ver-
sion6.14), and SADABS(Version2.10); BrukerAXS Inc.: Madison, WI, 2003.
(33) APEX2 (Version 2009.1-0), SAINT (Version 7.60a), SHELXTL
(Version 2008/4), and SADABS (Version 2008/1); Bruker AXS Inc.:
Madison, WI, 2009.