Torsional distortions in trimesitylsilanes and trimesitylgermanes

Article abstract of DOI:10.1016/S0022-328X(98)00754-2

The crystal structures have been solved for allyltrimesitylsilane (3), trimesitylsilane (4), and trimesitylgermane (5). Steric congestion caused by the three mesityl groups causes some lengthening of the Si-C and Ge-C bonds. The C-Si-C and C-Ge-C bond ang

Full text of DOI:10.1016/S0022-328X(98)00754-2

Journal of Organometallic Chemistry 568 (1998) 21–31
Torsional distortions in trimesitylsilanes and trimesitylgermanes
Joseph B. Lambert *, Charlotte L. Stern, Yan Zhao, Winston C. Tse, Catherine E. Shawl,
Kirk T. Lentz, Lidia Kania
Department of Chemistry, Northwestern Uni6ersity, 2145 Sheridan Road, E6anston, IL 60208-3113, USA
Received 14 October 1997; received in revised form 27 May 1998
Abstract
The crystal structures have been solved for allyltrimesitylsilane (3), trimesitylsilane (4), and trimesitylgermane (5). Steric
congestion caused by the three mesityl groups causes some lengthening of the Si–C and Ge–C bonds. The C–Si–C and C–Ge–C
bond angles are increased when X is not H. Distortion is relieved by rotating the plane of the phenyl rings to create a propeller
conformation. The same distortion has been found in unpublished data of the reported crystal structures of trimesitylsilyl azide
(6) and aminotrimesitylgermane (7). Despite the variety of substituents (allyl, amino, azido, H) and the two different central atoms
(Si, Ge), the twist angle of the propeller (compared with a conformation lacking any twist) lies between 41.6 and 48° for all five
systems. © 1998 Elsevier Science S.A. All rights reserved.
Keywords: Germane; Mesityl group; Propeller conformation; Silane; Steric congestion
1. Introduction
tions increase when large substituents replace hydro-
gens in the ortho positions. Mesityl (2,4,6-
trimethylphenyl; Mes), in which the ortho groups are
methyl, has been widely used when steric bulk is re-
quired. When three such groups are present, as in
trimesitylmethyl (Mes3C–), the molecule is expected to
exhibit considerable twisting about the ipso–para axes
in the standard propeller geometry [1].
The propeller shape of (triaryl)M–X molecules (M=
C, Si, Ge, Sn) is dictated by the interactions between
substituents on the adjacent aryl rings, primarily ortho–
ortho (1). If such interactions were absent, the six ortho
carbons would all lie in a single plane just beneath the
plane formed by the three ipso carbons (2).
Trimesitylsilyl (Mes3Si–) and trimesitylgermyl
(Mes3Ge–) have seen increased use as sterically bulky
groups [2]. We have obtained the crystal structures of
three such compounds: Mes3Si-allyl (3), Mes3Si–H (4),
and Mes3Ge–H (5). We report these structures herein
and compare them with two similar structures already
published in part: Mes3Si–N3 (6) [3] and Mes3Ge–NH2
(7) [4]. In particular, we report the respective twist
angles of the arene rings for these five molecules.
In order to minimize the ortho–ortho interactions, the
plane of each arene ring rotates about the ipso–para
axis so that the atoms or groups on the ortho carbons
of a given ring alternate above and below (1) the plane
of the six unrotated ortho carbons (2). These interac-
2. Results
* Corresponding author. Tel.: +1 847 4915437; fax: +1 847
4917713; e-mail: lambert@casbah.acns.nwu.edu
Trimesitylsilane (4) was prepared by the method of
0022-328X/98/$19.00 © 1998 Elsevier Science S.A. All rights reserved.
PII S0022-328X(98)00754-2

22
J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
Fig. 1. Structure of allyltrimesitylsilane (3).
Lappert et al. [5] by the reaction of trichlorosilane
with 2-bromomesitylene and sodium. Silane 4 was
converted to chlorotrimesitylsilane with phosphorus
pentachloride. Reaction of the chlorosilane with allyl-
lithium yielded allyltrimesitylsilane (3) [6]. In a similar
reaction sequence with germanium, trimesitylgermane
(5) was obtained as a by-product. Suitable crystals for
X-ray analysis were obtained for the allylsilane and
the germane by slow evaporation of a solution of
hexane and acetone. Common procedures for produc-
ing crystals only led to twins for trimesitylsilane. Va-
por diffusion produced small crystals that gave weak
reflections that did not permit anisotropic refinement
(except with the silicon atom). Figs. 1–3 present OR-
TEP diagrams of 3–5, respectively. Tables 1 and 2
present the atomic coordinates and significant struc-
tural parameters for the allylsilane 3, Tables 3 and 4
for the silane 4, and Tables 5 and 6 for the germane
5. Tables 7 and 8 contain analogous data for the silyl
azide [3] and the germylamine [4]. Most of these lat-
ter data did not appear in the original publications
but were obtained from the Cambridge Crystallo-
graphic Database. Additional crystal data for 3–5 are
found in Section 4.

J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
23
Fig. 2. Structure of trimesitylsilane (4).
3. Discussion
mesityl–M–mesityl angle) through flattening of the
silicon or germanium tetrahedron, i.e., raising the C–
M–C angle while lowering the C–M–X angle if the X
group is smaller than mesityl (herein, X is allyl, H,
amino, or azido). The C–Si–C angles for the allylsilane
and the silyl azide are respectively, 112.9 and 114.0°,
indicating a significant increase from tetrahedral. The
value in the silane 4 is anomalously small, again possi-
bly due to the poor refinement level. The germane 5,
however, is unperturbed at 109°. Thus the germane,
and probably the silane 4, with only a hydrogen for the
fourth group X do not require angle distortion. The
germylamine, like the two fully substituted silanes, has
an enlarged angle, 113.9°.
The average angle (C–Si–X) from the allyl
methylene carbon through the silicon atom to an ipso
carbon in 3 (Fig. 1: C1Si1C28, C10Si1C28, and
C19Si1C28 is 105.8°. In order for the mesityl groups to
move further apart (C–Si–C is 112.9°), they must get
closer to the fourth group (allyl), which is considerably
smaller than the aryl groups. Thus the angles around
silicon to the mesityl groups increase whereas those to
allyl decrease. The analogous angles also are lower in
the azide 6 (104.4°) [3] and in the amine 7 (104.7°) [4].
Table 9 assembles averages for the most relevant
structural parameters: the M–C bond length, the Ci–
M–Ci% bond angle, and the torsional angles through M
(Ci–M–Ci%–C%o). Distortions of each of these parame-
ters could decrease steric interactions between adjacent
aryl rings. For example, lengthening of the M–C bond
would move the aryl rings further apart. The normal
˚
range for the C–Si bond is 1.86–1.91 A [7], and the
specific length in tetraphenylsilane is 1.872 A [8]. The
˚
˚
values for the allylsilane (1.909 A) and the silyl azide
˚
(1.889 A) are at the upper end of this range, as might
be expected in sterically congested systems. The value
for the silane 4 is about 0.1 A above this range, but this
result may be in part an artifact of the low level of
isotropic refinement. The normal range for the C–Ge
bond is 1.90–2.00 A, and the specific value for te-
traphenylgermane is 1.956 A [9]. The value for the
amine is at the upper end of this range, and that for the
germane is above it. Thus, the M–C bonds for these
mesityl systems tend to be longer than usual.
A second structural distortion to relieve steric inter-
actions is enlargement of the C–M–C bond angle (the
˚
˚
˚

24
J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
Fig. 3. Structure of trimesitylgermane (5).
mesityl groups. A similar distortion is seen in the azide
6, in which the N–N–Si angle at 125.8° is increased by
about 10° [3].
Thus, the C–M–C angles involving the mesityl groups
have expanded at the expense of the C–M–X angles
involving the smaller fourth groups X.
The various torsional or dihedral angles provide the
most interesting insight into the steric effects. If each
mesityl ring were symmetrically disposed as in 2, the
view along the para–ipso axis of one aryl ring into the
plane of the other three atoms attached to M would
resemble 8. In this Newman projection, the ortho car-
bons with their attached methyl groups appear on
either side with a dihedral angle between them of 180°.
The allyl group has other means to minimize steric
strain. The C28–Si1 length is 1.919(3) A, similar to the
˚
bond lengths from silicon to the mesityl ipso carbons
˚
but about 0.03 A longer than the normal length. Hyper-
conjugation also may contribute to this bond lengthen-
ing. The allyl group extends itself by opening up bond
angles. The Si1–C28–C29 angle is very appreciably
increased to 124.7(2)° from 109.5° expected for an sp3
carbon (although hyperconjugation again could con-
tribute), and the C28–C29–C30 angle is slightly in-
creased to 124.8(4)° from 120° expected for an sp2
carbon. These distortions extend the allyl group, so that
it is less bent and can allow closer approach of the

J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
25
Table 2
The dihedral angle between two rings, as defined by the
Selected structural parameters for allyltrimesitylsilane (3)
angle between the Cipso–M axis and the Ci%pso–C%ortho axis
of another ring (the Ci–M–Ci%–Co% dihedral angle), is
30°, as illustrated by the angles h and i in 8. This is the
acute angle between the plane of the projected aryl ring
˚
Bond lengths (A)
Si1–C1
1.911(3)
1.910(3)
1.906(3)
1.919(3)
1.454(4)
1.352(5)
Si1–C10
Si1–C19
Si1–C28
C28–C29
C29–C30
and the M–Cipso or M–C%
axis in the illustrated
ipso
Newman projection.
When the aryl ring twists to minimize the ortho–ortho
methyl interactions, one of these angles (i) becomes
larger than 30° and the other (h) becomes smaller than
30°. If this distortion is relatively large, as in all the
present cases, one of the aryl rings passes to the other
side of one Co–CH3 bond. The other aryl ring moves
towards the perpendicular, as in 9. The values of h and
i are found three times in each molecule, representing
the projection of each of the three mesityl rings onto the
axes defined by the other two in 9. The means of the
three values are listed in Table 9.
The values of the more acute angle, h, range from 3.0
to 21.8°, and those of the less acute angle, i, range from
68.1 to 87° (to 81.8° if the silane 4 is excluded). The
extent of the twisting distortion may be specified by
comparison of the altered angles in 9 with the original
angles in 8. Thus, h moves from 30° past the C–Ar
Bond angles (°)
C1Si1C10
113.2(1)
C1Si1C19
C10Si1C19
C1Si1C28
C10Si1C28
C19Si1C28
113.0(1)
112.4(1)
105.6(1)
107.5(1)
104.4(1)
Dihedral angles (°)
C1Si1C10C11
C1Si1C10C15
C1Si1C19C20
C1Si1C19C24
C10Si1C1C2
12.9(3)
−163.6(2)
73.5(2)
−113.7(2)
66.7(2)
C10Si1C1C6
−115.6(2)
−156.9(2)
15.9(2)
−164.1(2)
13.7(3)
C10Si1C19C20
C10Si1C19C24
C19Si1C1C2
C19Si1C1C6
C19Si1C10C11
C19Si1C10C15
−116.6(2)
66.8(2)
Table 1
Atomic coordinates and Beq for allyltrimesitylsilane (3)
Atom
x
y
z
Beq
bond to an acute angle on the other side. The distortion
thus is the sum h+30. The angle i also begins as 30°
and increases to a large acute angle, so the distortion is
the difference i−30. The mean distortions may be
calculated for each system and are given as ƒ in Table
9. Surprisingly, all six systems have very similar distor-
tions, ranging only from 41.6 to 48° (to 45.0° if the
silane 4 is excluded). Thus, the mesityl rings rotate some
41–45° to minimize ortho–ortho methyl interactions,
irrespective of the metal M or the substituent.
The crystal structures are unremarkable except for the
germane 5. For both the silane 4 and the germane 5, the
hydride is not observed in the X-ray experiment. It is
deduced to be present because no other atom appears at
the fourth coordination site (no unaccounted for elec-
tron density), because the geometry around silicon or
germanium is tetrahedral, and because no counterions
are present. As Fig. 3 shows, there is disorder around
the germanium atom. The disorder places the germa-
nium atom both above and below the plane of the three
ipso carbons to which it is attached. There is no disorder
in the mesityl rings. Thus, exactly half the germanium
tetrahedra point in one direction and half in the other,
resulting in the observed presence of two half germa-
nium atoms, represented in Fig. 3 by Ge1 and Ge2, with
equal occupancy factors. If the germanium atom were a
chiral center, there would be equal amounts of the two
enantiomers.
Si(1)
C(1)
C(2)
C(3)
C(4)
C(5)
C(6)
C(7)
0.39038(9)
0.5238(3)
0.5636(3)
0.6481(4)
0.6983(3)
0.6649(3)
0.5817(3)
0.5256(5)
0.7926(6)
0.5649(5)
0.5013(3)
0.6751(3)
0.7403(4)
0.6451(4)
0.4778(3)
0.4057(3)
0.7991(4)
0.7185(5)
0.2265(4)
0.2978(3)
0.1632(3)
0.0676(3)
0.1010(3)
0.2434(4)
0.3428(3)
0.1194(4)
0.35217(3)
0.4102(1)
0.4032(1)
0.4473(1)
0.4976(1)
0.5034(1)
0.4610(1)
0.3489(2)
0.5433(2)
0.4727(2)
0.2794(1)
0.2713(1)
0.2156(1)
0.1666(1)
0.1748(1)
0.2291(1)
0.3206(1)
0.1071(2)
0.2314(2)
0.3768(1)
0.4169(1)
0.4257(1)
0.3973(1)
0.3629(1)
0.3528(1)
0.4537(1)
0.4032(2)
0.3175(2)
0.3418(1)
0.2899(1)
0.2715(2)
0.26338(5)
0.2189(2)
0.1243(2)
0.0846(2)
0.1349(2)
0.2287(2)
0.2722(2)
0.0632(3)
0.0908(4)
0.3775(3)
0.2874(2)
0.2931(2)
0.2947(2)
0.2950(2)
0.2976(2)
0.2953(2)
0.3020(3)
0.2944(4)
0.3096(3)
0.3725(2)
0.3552(2)
0.4267(2)
0.5168(2)
0.5383(2)
0.4696(2)
0.2623(2)
0.5896(3)
0.5046(2)
0.1564(2)
0.0962(2)
0.0517(4)
2.26(1)
2.50(6)
2.92(6)
3.56(7)
3.43(7)
3.31(7)
2.79(6)
3.73(8)
5.3(1)
3.81(8)
2.16(5)
2.48(6)
3.01(6)
2.83(6)
2.54(6)
2.26(5)
3.50(8)
4.67(10)
3.07(7)
2.30(5)
2.47(5)
2.86(6)
2.92(6)
2.94(6)
2.60(6)
3.32(7)
4.29(9)
3.57(8)
3.22(7)
4.07(8)
7.0(1)
C(8)
C(9)
C(10)
C(11)
C(12)
C(13)
C(14)
C(15)
C(16)
C(17)
C(18)
C(19)
C(20)
C(21)
C(22)
C(23)
C(24)
C(25)
C(26)
C(27)
C(28)
C(29)
C(30)
−0.0138(5)
0.5005(4)
0.1995(4)
0.1647(4)
0.0099(6)

26
J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
Table 3
Table 5
Atomic coordinates for trimesitylsilane (4)
Atomic coordinates and Beq for trimesitylgermane (5)
Atom
x
y
z
Atom
x
y
z
Beq
Si
C1
C2
C3
C4
C5
C6
C7
0.1024
−0.001(2)
0.084(2)
0.0784(4)
0.162(2)
0.132(2)
0.203(2)
0.284(1)
0.308(2)
0.249(2)
0.040(2)
0.350(3)
0.290(3)
0.156(2)
0.250(2)
0.313(2)
0.294(2)
1.191(2)
0.131(2)
0.278(2)
0.366(2)
0.030(2)
−0.073(2)
−0.126(2)
−0.250(2)
−0.304(1)
−0.239(2)
−0.135(1)
−0.087(4)
−0.427(1)
−0.068(2)
0.9803
Ge1
Ge2
C1
C1*
C2
C2*
C3
C3*
C4
C4*
C5*
C5
C6*
C6
C7
C7*
C8
−0.0652(1)
0.0652(1)
0.0236(9)
−0.0236(9)
0.1041(7)
−0.1041(7)
0.1132(7)
−0.1132(7)
0.0492(8)
−0.0492(8)
0.0257(7)
−0.0257(7)
0.0382(7)
−0.0382(7)
0.1753(9)
−0.1753(9)
0.061(1)
0.0807(1)
0.0807(1)
0.1621(7)
0.1621(7)
0.2530(7)
0.2530(7)
0.3142(6)
0.3142(6)
0.2914(7)
0.2914(7)
0.2002(7)
0.2002(7)
0.1356(6)
0.1356(6)
0.2889(7)
0.2889(7)
0.3591(8)
0.3591(8)
0.0355(7)
0.0355(7)
0.24520(9) 2.68(3)
0.25480(9) 2.68(3)
0.882(1)
0.827(1)
0.758(1)
0.748(1)
0.801(1)
0.867(1)
0.827(1)
0.674(2)
0.931(2)
1.068(1)
1.081(1)
1.145(1)
1.204(1)
1.189(1)
1.124(1)
1.030(1)
1.275(1)
1.110(1)
0.971(2)
0.926(2)
0.929(1)
0.979(1)
1.034(1)
1.032(1)
0.862(3)
0.972(1)
1.085(1)
0.3452(4)
0.1548(4)
0.3588(5)
0.1412(5)
0.4248(5)
0.0752(5)
0.4795(4)
0.0205(4)
0.0322(4)
0.4678(4)
0.0980(5)
0.4020(5)
0.3004(5)
0.1996(5)
0.5512(5)
4.3(2)
4.3(2)
3.8(2)
3.8(2)
3.9(2)
3.9(2)
3.8(2)
3.8(2)
3.6(2)
3.6(2)
3.5(2)
3.5(2)
7.0(3)
7.0(3)
7.7(3)
7.7(3)
6.6(3)
6.6(3)
4.9(4)
3.7(2)
3.7(2)
3.7(2)
3.7(2)
3.2(3)
5.8(4)
6.9(3)
6.9(3)
0.065(2)
−0.006(2)
−0.072(2)
−0.065(2)
0.179(2)
−0.009(3)
−0.132(3)
0.052(2)
0.140(2)
0.152(2)
0.096(2)
0.011(2)
0.004(2)
0.227(2)
0.110(2)
−0.074(2)
0.032(3)
−0.047(3)
−0.042(2)
0.037(2)
0.137(2)
0.137(2)
C8
C9
C10
C11
C12
C13
C14
C15
C16
C17
C18
C19
C20
C21
C22
C23
C24
C25
C26
C27
C8*
C9
C9*
C10
C11
−0.061(1)
−0.1229(8)
0.1229(8)
0.0000
−0.0512(5)
0.3946(5)
0.1054(5)
0.2500
−0.078(1)
0.0946(8)
−0.1336(7)
−0.1336(7)
−0.2468(7)
−0.2468(7)
−0.3060(9)
−0.4257(10)
−0.0782(8)
−0.0782(8)
0.3032(4)
0.1968(4)
0.3025(4)
0.1975(4)
0.2500
C11* −0.0946(8)
C12 0.0913(7)
C12* −0.0913(7)
C13
C14
C15
−0.162(4)
0.041(2)
0.237(2)
0.0000
0.0000
0.1996(9)
0.2500
0.3614(5)
0.1386(5)
C15* −0.1996(9)
This disorder may be explained either statically or
dynamically. In the static explanation, the two struc-
tural modes must be present as a 50/50 mixture, for
example, alternating along the stacks of molecular
units. Alternatively, adjacent ribbons could have oppo-
site senses. In the plane used for the depiction in Fig. 3,
the molecules are arranged in columns (Fig. 4). In the
Table 4
Table 6
Selected structural parameters for trimesitylsilane (4)
Selected structural parameters for trimesitylgermane (5)
˚
˚
Bond lengths (A)
Bond lengths (A)
Si–C1
Si–C10
Si–C19
2.08(2)
2.00(2)
1.99(2)
Ge1–C1
Ge1–C1*
Ge1–C10
2.049(7)
2.031(9)
2.06(1)
Bond angles (°)
C1SiC10
C1SiC19
Bond angles (°)
C1Ge1C1*
C1Ge1C10
101.7(7)
105(1)
108.8(9)
106.6(3)
109.8(2)
110.5(3)
C10SiC19
C1*Ge1C10
Dihedral angles (°)
C1SiC10C11
C1SiC10C15
C1SiC19C20
C1SiC19C24
C10SiC1C2
Dihedral angles (°)
C1Ge1C1*C2*
C1Ge1C1*C6*
C1Ge1C10C11
C1Ge1C10C11*
C1*Ge1C1C2
85(1)
−124(2)
79.2(6)
−118.4(8)
5.3(7)
−158.7(5)
2.8(10)
−158.1(6)
−111.9(6)
84.0(5)
−116.8(8)
82.2(6)
2(4)
−164(1)
−157(1)
12(3)
−110(3)
87(1)
C10SiC1C6
C1*Ge1C1C6
C10SiC19C20
C10SiC19C24
C19SiC1C2
C19SiC1C6
C19SiC10C11
C19SiC10C15
C1*Ge1C10C11
C1*Ge1C10C11*
C10Ge1C1C2
C10Ge1C1C6
C10Ge1C1*C2*
C10Ge1C1*C6*
89(1)
−125(3)
−164(1)
13(3)
−161.6(5)
0.8(9)

J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
27
Table 7
Table 9
Selected structural parameters for azidotrimesitylsilane (6) [3]
Summary of structural parameters
6c
5
7d
˚
Bond lengths (A)
3
4
Si–C1
Si–C10
Si–C19
1.887
1.892
1.888
M
X
Si
Si
H
Si
N3
Ge
H
2.047
109.0
3.0
81.8
42.4
Ge
NH2
1.978
113.9
19.1
68.5
43.8
CH2CHꢀCH2
1.909
112.9
14.2
69.0
41.6
˚
M–X (A)
C–M–C (°)
ha (°)
2.02
1.889
Bond angles (°)
C1Si1C10
105
9
87
48
114.0
21.8
68.1
45.0
113.3
C1Si1C19
C10Si1C19
115.1
113.7
ia (°)
b
ƒ
(°)
Dihedral angles (°)
C1Si1C10C11
C1Si1C10C15
C1Si1C19C20
C1Si1C19C24
C10Si1C1C2
68.0
−114.4
−165.7
16.1
−151.7
29.8
a See structure 9.
b The average 0.5[(h+30)+(i−30)].
c Data from the Cambridge Crystallographic Database from Ref. [3].
d Data from the Cambridge Crystallographic Database from Ref. [4].
C10Si1C1C6
to that of Fig. 4. Four rows of molecules are seen. Each
row represents the molecules in the plane of Fig. 4, but
now the molecules above and below that plane are
visible in the rows above and below a given row. The
disorder about germanium is seen from the diamond-
shaped component at the center of each molecule. The
view is along the para–ipso axis of one mesityl group.
The other two mesityl groups extend to the left and
right. Although it is not apparent in this perspective,
the para–ortho axes alternate in front of and behind the
plane of the perspective, as they represent molecules
from adjacent columns of Fig. 4.
The dynamic explanation for the disorder is now seen
from Fig. 5. In this plane the germanium atoms are
positioned over each other from one row to the next. A
hydride could move easily from one germanium atom
to the next in the row above or below. This movement
would precipitate hydride migration along the entire
vertical column, inverting all the germanium tetrahedra.
Both tetrahedral environments thus are populated, re-
sulting in the observed disorder. By either the static or
dynamic explanation, the crystal would exhibit equal
amounts of the two forms on the average and lead to
the structure of Fig. 3.
C10Si1C19C20
C10Si1C19C24
C19Si1C1C2
C19Si1C1C6
C19Si1C10C11
C19Si1C10C15
61.3
−117.0
75.1
−103.5
−158.1
19.4
first vertical column on the left, two mesityl groups
point up and to either side (as in Fig. 3), while the third
mesityl group points downward between the two
mesityl groups of the next molecule. In the second
column from the left, the order is reversed. One mesityl
group points up and two point down and to either side.
This arrangement optimizes the steric fit within the
plane. The alternation continues, with the third column
like the first, the fourth like the second, and so on.
Fig. 5 provides the view from a plane perpendicular
Table 8
Selected structural parameters for aminotrimesitylgermane (7) [4]
˚
Bond lengths (A)
Ge–C1
Ge–C10
Ge–C19
1.973
1.986
1.976
We emphasize that these explanations for the crystal-
lographic disorder are unproved, and other explana-
tions may be viable. One possibility is that the space
group of trimesitylgermane (5) has lower symmetry
than C2/c. Consequently, we also solved the crystal
structure in the space groups Cc and C2. The disorder
did not disappear in either solution. The germanium
atom appears as two centers of electron density, with
each site half occupied. For the C2 case, the residuals
(R=0.083, Rw=0.064) are significantly larger than the
values for the solution in C2/c (R=0.068, Rw=0.056).
In this space group the atoms on the C2 axis cannot be
refined anisotropically. For the Cc case, the residuals
(R=0.065, Rw=0.048) are comparable or better, but
any improvement can be attributed to the additional
parameters required by the model. For both alternative
Bond angles (°)
C1GeC10
C1GeC19
116.6
113.0
112.0
C10GeC19
Dihedral angles (°)
C1GeC10C11
C1GeC10C15
ClGeC19C20
C1GeC19C24
C10GeC1C2
−119.7
66.7
−157.7
24.9
19.8
−158.5
68.2
−109.2
−112.0
70.7
12.5
−161.1
C10GeC1C6
C10GeC19C20
C10GeC19C24
C19GeC1C2
C19GeC1C6
C19GeC10C11
C19GeC10C15

28
J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
Fig. 4. View of one plane in the lattice of trimesitylgermane (5).
space groups, there are parameter correlations between
all atoms. Furthermore, the C2 case has seven atoms
and the Cc has ten atoms with nonpositive definite
thermal coefficients that militate against acceptance. As
a result, both space groups generate several unrealistic
displacements ellipsoids.
The residuals, the parameter correlations, and the
thermal coefficients all favor the C2/c space group.
Moreover, use of space groups with lower symmetry
still results in disorder involving the germanium atom.
This disorder thus appears to be real and requires a
structural explanation such as the one we gave above.
The problems of distinguishing between C2/c and Cc
have been discussed extensively by Marsh [10]. The
poor quality of the data for trimesitylsilane (4) proba-
bly makes it impossible to render a distinction between
the two. The preponderance of evidence for trime-
sitylgermane, however, strongly favors C2/c.
was filled with N2, and trichlorosilane (2.0 ml, 0.02
mol) was added through a syringe. The flask was heated
slowly to 60°C, at which temperature an exothermic
reaction occurred and the benzene began to reflux. The
solution was then stirred overnight. The black suspen-
sion was filtered through a celite pad. The benzene
solution was concentrated by rotary evaporation. The
residue was crystallized from isopropyl alcohol to give
1
4.05 g (52.5%) of a white powder: mp 194°C, H-NMR
(CDCl3) l 2.12 (s, 18H), 2.26 (s, 9H), 5.73 (s, 1H), 6.78
(s, 6H); IR (KBr) Si–H 2140 cm−1, MS (EI) m/z 386
(M+, 2), 267 (25), 266 (91), 251 (46), 235 (32), 160 (33),
148 (16), 147 (97), 146 (100), 145 (32), 119 (24), 105(30).
Anal. Calc. for C27H34Si: C, 83.93; H, 8.81; Si, 7.25.
Found: C, 83.45; H, 8.99; Si, 7.10.
4.2. Chlorotrimesitylsilane [12]
A 100 ml, round-bottomed flask was charged with
trimesitylsilane (1.35 g, 3.5 mmol), PCl5 (1.12 g, 5.4
mmol), CCl4 (30 ml), and a magnetic stirring bar. The
mixture was heated to reflux under N2 for 24 h. The
solution was concentrated by rotary evaporation, and
the residue was dissolved in hexane (50 ml). Methanol
(15 ml) was added slowly to decompose the unreacted
PCl5. The organic layer was separated, dried over
MgSO4, and concentrated to give 1.22 g (83%) of a
4. Experimental section
4.1. Trimesitylsilane (4) [5,11]
A 500 ml, three-necked, round-bottomed flask,
equipped with a rubber septum, a condenser, and a
glass stopper, was charged with powdered Na (4.8 g,
0.21 mol), 2-bromomesitylene (9.2 ml, 0.06 mol), dry
benzene (150 ml) and a magnetic stirring bar. The flask
1
white solid: H-NMR (CDCl3) l 1.96–2.46 (br, 18H),
2.24 (s, 9H), 6.72–6.90 (br, 6H).

J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
29
Fig. 5. View of a plane in the lattice of trimesitylgermane (5) that is perpendicular to the plane in Fig. 4.
4.3. Allyltrimesitylsilane (3)
the preparation of allyltrimesitylgermane by the same
method.
A 100 ml, round-bottomed flask fitted with a rubber
septum and a N2 inlet needle was charged with allyltri-
phenyltin (8.55 g, 21.9 mmol) and a magnetic stirring
bar. Anhydrous THF (50 ml) was added to the flask via
a syringe. Phenyllithium (12.2 ml, 1.8 mol, 22.0 mmol)
in ether and cyclohexane was then added quickly, and a
large amount of precipitate (Ph4Sn) formed immedi-
ately. After 0.5 h, the suspension was transferred
through a wide-bore cannula to an enclosed glass frit
under N2 and was filtered into a 250 ml flask, which
had been charged with chlorotrimesitylsilane (6.00 g,
14.4 mmol) and a stirring bar. The dark red solution
was stirred at room temperature, and the reaction was
monitored by 1H-NMR. After 2 days, some chloro-
trimesitylsilane was found to remain unreacted. Addi-
tional allyllithium (prepared as described above from
2.80 g of allyltriphenyltin and 4.0 ml of 1.8 M PhLi in
25 ml of THF) was added. After 1 day, all the chloro-
trimesitylsilane was consumed. The reaction mixture
was quenched by water and extracted with hexane (200
ml). The organic solution was dried (MgSO4), concen-
trated, and chromatographed over silica gel with hex-
ane as eluent to give a white solid (2.01 g, 33%): m.p.
4.5. Crystal structure of trimesitylsilane (4)
Suitable crystals were extremely difficult to obtain.
Numerous methods for purification and crystallization
were tried, and efforts led only to crystals that were
well formed but not single. Weakly reflecting crystals
finally were obtained by placing 20 mg of 4, dissolved
in 2 ml of butanol, into a 2 d vial. This vial in turn was
placed in a 10 d vial containing 5 ml of water. The
larger vial was capped, so that water vapor could
diffuse to the surface of the butanol. The single crystals
of 4 appeared at the surface of the water-enriched
butanol and were isolated. A white crystal of dimen-
sions 0.35×0.28×0.08 mm was used for the X-ray
crystal determination. Measurements were performed
on an Enraf-Nonius CAD4 diffractometer at −120°C
˚
with Mo–Kh radiation (u=0.71069 A) by the ꢀ–q
scan technique over a 2q range of 4–55°. Cell parame-
ters were determined by least-squares refinement of the
setting angle of 25 high angle reflections. Further ex-
perimental details are given in Table 10. Intensities of
three standard reflections measured every 3 h of X-ray
exposure showed no significant decay. Intensity data
were corrected for Lorentz polarization and absorption
effects. Calculations were performed on a MicroVax
3600 with the teXsan 5.0 crystallographic software
package [13]. The structure was solved by direct meth-
ods (MITHRIL) [14]. The full matrix least-squares
refinement for all non-hydrogen atoms yielded an R
factor of 0.148 (Rw=0.168) for 1131 reflections with
I\3|(I) (2118 total reflections) and 116 variables.
Because of the limited number of reflections, only the
1
160–163°C; H-NMR (C6D6) l 2.11 (s, 9H), 2.23 (s,
18H), 2.47–2.52 (m, 2H), 4.88–5.03 (m, 2H), 5.82–5.98
(m, 1H), 6.74 (s, 6H); 13C-NMR (C6D6) l 21.0, 25.5,
27.5, 116.2, 129.9, 135.7, 137.9, 138.6, 145.0; 29Si-NMR
(CDCl3) l −18.7l; Anal. Calc. for C30H38Si: C, 84.44;
H, 8.98. Found: C, 84.30; H, 8.84.
4.4. Trimesitylgermane (5)
Trimesitylgermane was produced as a by-product of

30
J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
silicon atom was refined anisotropically. The hydrogen
atoms either were located from difference Fourier
maps or were placed at the calculated positions. They
were included in the final stage of refinement as fixed
contributors to the structure factors. Final positional
parameters and derived data are listed in Tables 3 and
4 with reference to Fig. 2.
tallographic software package [13]. Structural parame-
ters are given in Tables 1 and 2 with reference to Fig.
1.
4.7. Crystal structure of trimesitylgermane (5)
A colorless cubic crystal having approximate dimen-
sions of 0.17×0.24×0.26 mm was mounted using oil
(Paratone-N, Exxon) on a glass fiber. All measure-
ments were made on an Enraf-Nonius CAD4 diffrac-
tometer with graphite monochromated Mo–Kh
radiation. Further experimental details are given in
Table 10. The data were collected at a temperature of
−12091°C using the ꢀ–q scan technique to a maxi-
mum 2q value of 47.9°. Of the 2068 reflections that
were collected, 1994 were unique (Rint=0.043). The
intensities of three representative reflections were mea-
sured after every 90 min of X-ray exposure time. No
decay correction was applied. The linear absorption
coefficient, v, for Mo–Kh radiation is 13.4 cm−1. An
analytical absorption correction was applied, which
resulted in transmission factors ranging from 0.72 to
0.80. The data were corrected for Lorentz and polar-
ization effects. A correction for secondary extinction
was applied (coefficient=2.08373e−07). The structure
was solved by direct methods [15] and expanded using
Fourier techniques [16]. Non-hydrogen atoms were
refined anisotropically. Atom H16 (on germanium)
was included from the difference map but not refined,
and the remaining hydrogen atoms were included in
idealized positions. The hydrogen attached to Ge and
the remaining hydrogen on C14 were not included in
structure factor calculations. The germanium was dis-
ordered off the twofold axis. The final cycle of full-
matrix least-squares refinement was based on 1063
observed reflections (I\3.00|(I)) and 134 variable
parameters and converged (largest parameter shift was
0.00 times its estimated S.D.) with unweighted and
weighted agreement factors of R=0.068 and Rw=
0.056. The maximum and minimum peaks on the final
difference Fourier map corresponded to 0.38 and −
4.6. Crystal structure of allyltrimesitylsilane (3)
A colorless crystal was obtained from the slow
evaporation at room temperature of a solution of hex-
ane and acetone. A colorless, tabular crystal with ap-
proximate dimensions of 0.48×0.35×0.20 mm was
mounted on a glass fiber using oil (Paratone-N,
Exxon). All measurements were made on a Siemens
SMART CCD diffractometer with graphite monochro-
mated Mo–Kh radiation. Further experimental details
are given in Table 10. Diffraction intensities were col-
lected at −7591°C using the ꢀ–q scan technique to
a maximum 2q of 56.4°. Of the 16601 reflections that
were collected, 6184 were unique (Rint=0.048). The
linear absorption coefficient for Mo–Kh was 1.1 cm−
1. Azimuthal scans of several reflections indicated no
need for an absorption correction. The data were cor-
rected for Lorentz and polarization effects. The struc-
ture was solved by direct methods and expanded with
Fourier techniques [15,16]. Non-hydrogen atoms were
refined anisotropically. Hydrogen atoms were refined
isotropically. The final cycle of full-matrix least-
squares refinement was based on 3965 observed reflec-
tions (I\3.00|(I)) and 432 variable parameters and
converged with unweighted and weighted agreement
factors of R=0.062 and Rw=0.062. The maximum
and minimum peaks on the final difference Fourier
3
˚
map correspond to 0.28 and −0.26 e/A , respectively.
All calculations were performed using the teXsan crys-
Table 10
Crystallographic parameters
3
4
5
−
3
˚
0.34 e /A , respectively. Structural parameters are
given in Tables 5 and 6 with reference to Fig. 3.
M
X
Si
Allyl
Si
H
Ge
H
Formula
Molecular
weight
C30H38Si
426.72
C
386.65
27H34Si
C27H34Ge
431.16
Acknowledgements
Crystal system Monoclinic
Monoclinic
C-centered
monoclinic
C2/c (No. 15)
10.999(3)
12.268(4)
17.659(4)
104.69(2)
2304(1)
This work was supported by the National Science
Foundation (Grant No. CHE-9302747).
Space group
a (A)
P21/c (No. 14) Cc (No. 9)
˚
8.1709(3)
23.1734(8)
13.7625(5)
102.2541(9)
2546.5(1)
4
11.107(3)
12.171(1)
17.525(3)
104.75(2)
2291(1)
4
˚
b (A)
˚
c (A)
References
i (°)
V (A )
3
˚
[1] For a review of propeller molecules, see K. Mislow, Acc. Chem.
Res. 9 (1976) 26.
Z
4

J.B. Lambert et al. / Journal of Organometallic Chemistry 568 (1998) 21–31
31
[2] For example, M.J. Fink, M.J. Michalczyk, K.J. Haller, R. West,
J. Michl, Organometallics 3 (1984) 793.
March, Acta Crystallogr. B 51 (1995) 897. (c) R.E. March, Acta
Crystallogr. B 53 (1997) 317.
[3] S.S. Zigler, K.J. Haller, R. West, M.S. Gordon, Organometallics
8 (1989) 1656.
[4] M. Rivie`re-Baudet, A. More`re, J.F. Britten, M. Onyszchuk, J.
Organomet. Chem. 423 (1992) 5.
[5] M.J.S. Gynane, M.F. Lappert, P.I. Riley, P. Rivie`re, M. Rivie`re-
Baudet, J. Organomet. Chem. 202 (1980) C5.
[6] J.B. Lambert, Y. Zhao, Angew. Chem. Int. Ed. Engl. 36 (1997)
400.
[7] W.S. Sheldrick, in: S. Patai, Z. Rappoport (Eds.), The Chemistry
of Organic Silicon Compounds, Part 1, Wiley, Chichester, 1989,
p. 245.
[8] C. Glidewell, G.M. Sheldrick, J. Chem. Soc. (A) (1971) 3127.
[9] (a) P.C. Chieh, J. Chem. Soc. (A) (1971) 3243. (b) P. Rivie`re, M.
Riviere-Baudet, P. Satge´, in: G. Wilkinson (Ed.), Comprehensive
Organometallic Chemistry, vol. 2, 1982, p. 502. (c) C. Glidewell,
O.G. Liles, J. Organometal. Chem. 243 (1983) 291.
[10] (a) R.E. March, Acta Crystallogr. A 50 (1994) 450. (b) R.E.
[11] J.J. Jerius, J.M. Hahn, A.F.M.M. Rahman, O. Mols, W.H.
Ilsley, J.P. Oliver, Organometallics 5 (1986) 1812.
[12] S.S. Zigler, L.M. Johnson, R. West, J. Organomet. Chem. 341
(1988) 187.
[13] teXsan, Crystal Structure Analysis Package, Molecular Structure
Corporation, 1985 and 1992.
[14] G.N. Gilmore, MITHRIL: A Computer Program for the Auto-
matic Solution of Crystal Structures from X-ray Data, Univer-
sity of Glasgow, Glasgow, UK, 1983.
[15] G.M. Sheldrick, SHELX86, in: G.M. Sheldrick, C. Kruger, R.
Goddard (Eds.), Crystallographic Computing 3, Oxford, 1985,
pp. 175–189.
[16] P.T. Beurskens, G. Admiraal, G. Beurskens, W.P. Bosman, R.
de Gelder, R. Israel, J.M.M. Smits, DIRDIF94: The DIRDIF-94
Program System, Technical Report of the Crystallography Labo-
ratory, University of Nijmegen, The Netherlands, 1994.
.
.