Heavy analogues of the 6π-electron anionic ring systems: Cyclopentadienide ion and cyclobutadiene dianion

Article abstract of DOI:10.1016/j.jorganchem.2007.01.011

The heavy analogues of the anionic 6π-electron systems, lithium 1,2-disila-3-germacyclopentadienide 2^- · [Li⁺(thf)], 1,2-disila-3,4-digerma- and 1,2,3,4-tetrasilacyclobutadiene dianions 7^2- · 2[K⁺(thf)₂

Full text of DOI:10.1016/j.jorganchem.2007.01.011

Journal of Organometallic Chemistry 692 (2007) 2800–2810
www.elsevier.com/locate/jorganchem
Heavy analogues of the 6p-electron anionic ring systems:
Cyclopentadienide ion and cyclobutadiene dianion
*
Vladimir Ya. Lee , Kazunori Takanashi, Risa Kato, Tadahiro Matsuno,
*
Masaaki Ichinohe, Akira Sekiguchi
Department of Chemistry, Graduate School of Pure and Applied Sciences, University of Tsukuba, Tsukuba, Ibaraki 305-8571, Japan
Received 15 October 2006; received in revised form 10 January 2007; accepted 10 January 2007
Available online 18 January 2007
Abstract
The heavy analogues of the anionic 6p-electron systems, lithium 1,2-disila-3-germacyclopentadienide 2ꢀ Æ [Li+(thf)], 1,2-disila-3,4-
digerma- and 1,2,3,4-tetrasilacyclobutadiene dianions 72ꢀ Æ 2[K+(thf)2] and 82ꢀ Æ 2[K+(thf)2], were synthesized by the reduction of the neu-
tral precursors 1, 3 and 4, respectively. 2ꢀ Æ [Li+(thf)], the heavy analogue of the cyclopentadienide ion, is an aromatic compound,
whereas 72ꢀ Æ 2[K+(thf)2] and 82ꢀ Æ 2[K+(thf)2], the heavy analogues of the cyclobutadiene dianion, are both non-aromatic.
ꢀ 2007 Elsevier B.V. All rights reserved.
Keywords: Aromaticity; Cyclopentadienide ion; Cyclobutadiene dianion; Silicon; Germanium
1. Introduction
stable antiaromatic congeners H–J appeared to be a great
problem, which is still not completely overcame. Of all aro-
matic charged species, the 6p-electron systems are of par-
ticular importance. However, even being significantly
stabilized by the aromatic delocalization, such compounds
are by far more reactive than the isoelectronic neutral ben-
zene, which opens the unlimited possibilities for their func-
tionalization providing the great synthetic applicability for
such compounds. It is therefore not surprising, that the
chemistry of charged 6p-electron aromatic systems, first
of all, that of cyclobutadiene dianion, cyclopentadienide
ion and cycloheptatrienylium (tropylium) ion, was one of
the most intriguing topics in the second half of the last
century and was greatly developed during this period.
However, the area of the heavy congeners of the above-
mentioned classes of aromatic compounds, in which the
skeletal C atoms were partially (or fully) replaced by their
heavier congeners of group 14 (Si, Ge, Sn, Pb), was almost
unexplored until the early beginning of 1990s [2]. Of all
these aromatic systems (heavy versions of cyclobutadiene
dianion, cyclopentadienide ion, tropylium ion), the stable
representatives of only second class of compounds, i.e.
the heavy analogues of cyclopentadienide ion, have been
There is a striking relationship between the stability of
the charged ring systems, both cationic and anionic, and
their aromaticity (or antiaromaticity). Thus, the fundamen-
tal Huckel’s rule predicts the important aromatic
¨
stabilization for (4n + 2) p-electron systems, such as cyclo-
propenylium ion A (2p), cyclobutadiene dication B (2p),
cyclobutadiene dianion C (6p), cyclopentadienide ion D
(6p), cycloheptatrienylium ion E (6p), cyclooctatetraene
dication F (6p) and cyclooctatetraene dianion G (10p),
and, by contrast, great antiaromatic destabilization is a
charactersitic feature of 4np-electron systems, such as
cyclopropenide ion H (4p), cyclopentadienylium ion I
(4p), cycloheptatrienide ion J (8p) (Scheme 1) [1]. And
indeed, whereas the chemistry of the aromatic compounds
of the types A–G is broadly and reliably established in
organic chemistry with a variety of the stable representa-
tives reported to date in the literature, the isolation of their
*
Corresponding authors. Tel.: +81 29 853 4523; fax: +81 29 853 4314.
E-mail addresses: vyalee@staff.chem.tsukuba.ac.jp (V.Ya. Lee),
sekiguch@staff.chem.tsukuba.ac.jp (A. Sekiguchi).
0022-328X/$ - see front matter ꢀ 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2007.01.011

V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
2801
Scheme 1.
more or less studied prior to our investigations [3]. Such
pioneering work on the chemistry of the stable anions
and dianions of sila- and germacyclopentadienes was con-
ducted in the research groups of Joo [4a], Boudjouk [4b,4g–
j], West [4c,4d] and Tilley [4e,4f], who investigated their
structures and aromaticity in details. The corresponding
Sn-analogues, stannacyclopentadiene anions and dianions,
have been recently synthesized by Saito and co-workers [5].
The deep theoretical insight to the problem of aromaticity
of the heavy analogues of cyclopentadienide ion was pro-
vided by Schleyer in the series of papers published in the
mid-1990s [6]. The lone example of the metastable silatro-
pylium ion, reported by Komatsu group in 2000, was gen-
erated and characterized below ꢀ50 ꢁC being irreversibly
decomposed at the higher temperatures [7]. To the best of
our knowledge, nothing was known about synthetic and
structural aspects of the heavy analogues of cyclobutadiene
dianion up to date. It is therefore evident, that even now
the field of the heavy analogues of such fundamentally
important species as cyclopentadienide ion and cyclobut-
adiene dianion is still rather ‘‘uncultivated’’ leaving many
synthetic and mechanistic problems to be resolved. In this
paper we present a full account of our recent contributions
to the field, particularly emphasizing the structural pecu-
liarities and aromaticity problems of newly synthesized
heavy analogous of cyclopentadienide ion (which will be
named below for simplicity as the ‘‘heavy cyclopentadie-
nide’’) [8a] and cyclobutadiene dianion (which will be
named as the ‘‘heavy cyclobutadiene dianion’’) [8b].
point the reduction of the recently prepared by us 1,2-
disila-3-germacyclopenta-2,4-diene 1 [9], which incorpo-
rated the Si@Ge double bond into the Si@Ge–C@C diene
system being the most electronically perturbed metallacy-
clopentadiene known to date, was of particular interest.
The right choice of the stoichiometry, namely – 2 equiv-
alents of the alkali metal for 1 equivalent of heavy cyclo-
pentadiene 1, was of crucial importance: utilization of an
excess of reducing reagent resulted in unavoidable overre-
duction to form complicated reaction mixtures. Finally,
potassium graphite KC8 was found to be the most effective
and selective reducing agent for 1, smoothly producing the
t
corresponding potassium salt 2ꢀ Æ K+ along with the Bu2-
MeSiK as the only side product of the reaction. Appar-
ently, the two-electron reduction of the Si@Ge double
bond takes place as the initial step of the reaction, which
is quite reasonable since the LUMO of 1 is largely repre-
sented by the p*-orbital of the Si@Ge bond [10]. The driv-
t
ing force of the next step, b-elimination of Bu2MeSiK, is
evidently the formation of the aromatic 6p-electron system
2ꢀ Æ K+ (its aromaticity will be discussed below). However,
since the isolation of 2ꢀ Æ K+ was precluded due to its poor
crystallinity, we have performed an exchange of the coun-
ter-cation from potassium to lithium by the treatment
of 2ꢀ Æ K+ with an excess of dry LiBr in THF. Following
this procedure, the final product, lithium 1,2,3-tris(di-
tert-butylmethylsilyl)-4-phenyl-1,2-disila-3-germacyclopen-
tadienide 2ꢀ Æ Li+, was isolated as bright-orange plates by
recrystallization from pentane in 34% yield (Scheme 2).
The X-ray analysis revealed the composition of 2ꢀ Æ Li+
as a THF solvate, 2ꢀ Æ [Li+(thf)], featuring several particu-
larly important structural peculiarities (Fig. 1). The
Li+-cation is pentahaptocoordinated to the anionic GeSi2C2-
five-membered ring, thus the overall structure represents a
classical ‘‘half-sandwich’’ complex diagnostic of the cyclo-
pentadienide derivatives of the type C5R5Li [11]. The dis-
tances from the Li atom to the skeletal atoms lie within
the normal range expected for Ge–Li, Si–Li and C–Li
bonds. The analysis of the structural changes taking place
upon the reduction of heavy cyclopentadiene 1 to form the
heavy cyclopentadienide 2ꢀ Æ [Li+(thf)] is especially instruc-
tive. First of all, the planarity of the five-membered ring of
the starting 1 was only very insignificantly affected by the
reduction, as was demonstrated by the comparison of their
2. Results and discussion
2.1. Heavy lithium cyclopentadienide
Typically, group 14 elements (E14) based metalloles (sil-
oles (E14 = Si), germoles (E14 = Ge) and stannoles
(E14 = Sn)) anions and dianions could be generated by
the alkali metal reduction of the exocyclic E14–halogen or
E14–Si bonds [2]. This resulted in the isolation of a variety
of compounds, which were discussed as possessing a signif-
icant amount of aromaticity. However, all these derivatives
included only one element E14, being produced from the
cyclopentadiene derivatives, in which E14 uniformly occu-
pied sp3-position of the cyclopentadiene ring. In such view-

2802
V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
be mentioned, that the lengths of all skeletal bonds in
2ꢀ Æ [Li+(thf)] are intermediate between those of the typical
single and double bonds. Considering all above-mentioned
structural features of 2ꢀ Æ [Li+(thf)], one should recognize
that they provide a solid evidence for the delocalization
of the 6p-electron density over the entire five-membered
ring, which in turn stands for its aromaticity. Evidently,
the extent of aromaticity of 2ꢀ Æ [Li+(thf)] is smaller than
that of the prototypical cyclopentadienide ion CpꢀLi+: in
contrast to CpLi, where the negative charge is equally dis-
tributed between all skeletal C atoms, one cannot assume
such even distribution of electron density over all skeletal
atoms in 2ꢀ Æ [Li+(thf)] because of their different electroneg-
ativities. The most essential part of the negative charge is
expected to be accumulated on Ge atom, the pyramidaliza-
tion of which is most important of all skeletal atoms: Ge1
(342.89ꢁ), Si1 (350.75ꢁ), Si2 (357.03ꢁ), C1 (359.94ꢁ), C2
(359.79ꢁ). The aromaticity of the model compound of
Scheme 2.
t
2ꢀ Æ Li+ (H3Si-groups instead of the real Bu2MeSi-substit-
uents) was finally verified by the most effective aromaticity
probe, nucleus-independent chemical shift (NICS) [13], cal-
˚
culated at 1 A above the ring center: NICS(1) = ꢀ8.4 [12b].
This value is comparable (albeit smaller) to those of the
classical 6p-electron aromatic compounds: C5H5Li
(NICS(1) = ꢀ10.3) and C6H6 (NICS(1) = ꢀ11.2) [12b].
However, in contrast to the crystal structure, the solu-
tion behavior and structure of 2ꢀ Æ Li+ is highly solvent-
dependent. Thus, the heavy cyclopentadienide preserves
its aromaticity in the non-polar solvents (toluene, benzene),
whereas it is unable to benefit from aromatic delocalization
in polar solvents (THF) anymore. The manifestation of
aromaticity of 2ꢀ Æ Li+ in toluene-d8 was convincingly dem-
onstrated by NMR spectroscopy (the spectral characteris-
tics of 2ꢀ Æ Li+ in C6D6 are almost identical to those in
1
toluene-d8, see Section 3). Thus, in H NMR spectrum of
t
2ꢀ Æ Li+ all Bu-groups of silyl-substituents were nonequiv-
Fig. 1. Crystal Maker view of 2ꢀ Æ [Li+(thf)] (hydrogen atoms are not
shown). Selected bond lengths (A): Ge1–Si1 = 2.3220(5), Si1–
Si2 = 2.2403(7), Ge1–C2 = 1.9303(17), Si2–C1 = 1.8269(18), C1–C2 =
1.402(2), Ge1–Si5 = 2.4092(5), Si1–Si3 = 2.3558(7), Si2–Si4 = 2.3550(7),
alent because of the g5-coordination of Li+-ion to the five-
membered ring breaking the symmetry of the molecule.
The tendency in the changes of the 13C NMR resonances
of the skeletal C atoms upon the reduction is in line with
the theoretically predicted one [6b]: namely, CH-atom in
2ꢀ Æ Li+ is shielded comparing to that of the starting 1 [9]
(143.2 vs. 150.0 ppm, Dd = ꢀ6.8 ppm), whereas CPh-atom
is deshielded (181.4 vs. 173.6 ppm, Dd = +7.8 ppm). The
skeletal Si atoms in 2ꢀ Æ Li+ resonate at 54.4 and
69.1 ppm, which values are in between those of the starting
1 [9] (ꢀ45.7 ppm for sp3-Si and 124.0 ppm for sp2-Si), that
is sp3-Si atom is deshielded upon the reduction, whereas
sp2-Si atom is shielded. One should also note, that both
values of 54.4 and 69.1 ppm lie between those of the typical
Si–Si and Si@Si bonds [14]. The most important prove for
the aromaticity of 2ꢀ Æ Li+ in toluene solution was obtained
˚
Ge1–Li1 = 2.760(4),
Si1–Li1 = 2.692(3),
Si2–Li1 = 2.677(4),
C1–
Li1 = 2.259(4), C2–Li1 = 2.284(4), Li1–O1 = 1.869(4). Selected bond
angles (ꢁ): Ge1–Si1–Si2 = 94.49(2), Si1–Si2–C1 = 101.48(6), Si2–C1–
C2 = 122.74(13), C1–C2–Ge1 = 119.72(13), C2–Ge1–Si1 = 97.89(5).
interior bond angles: 539.9ꢁ for 1 [9] vs. 536.3ꢁ for
2ꢀ Æ [Li+(thf)]. Consequently, the skeletal Ge1 and Si2
˚
atoms are situated slightly above (0.1458 and 0.1486 A),
whereas the skeletal Si1, C1 and C2 atoms are situated
˚
slightly below (0.1507, 0.0638 and 0.0798 A) the GeSi2C2-
meanplane. Second, even more important, feature is the
tendency in the changes of the skeletal bond lengths upon
the reduction: all double bonds of the starting 1 [9] are
elongated to form the final 2ꢀ Æ [Li+(thf)] (Ge1–Si1:
7
from its Li NMR spectrum, which revealed a single reso-
˚
˚
2.250(1) vs. 2.3220(5) A, C1–C2: 1.343(5) vs. 1.402(2) A),
nance at ꢀ5.4 ppm, a very high field characteristic for the
aromatic lithium cyclopentadienide derivatives [15]. How-
ever, in highly coordinating THF the behavior of heavy
cyclopentadienide is totally different. The 1H NMR
whereas all single bonds of 1 are shortened (Ge1–C2:
˚
˚
1.972(3) vs. 1.9303(17) A, Si1–Si2: 2.364(1) vs. 2.2403(7) A,
˚
Si2–C1: 1.888(3) vs. 1.8269(18) A) (Fig. 2) [12a]. It should

V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
2803
Fig. 2. Structural changes taking place upon the reduction of 1 to produce 2ꢀ Æ [Li+(thf)] (H and C atoms of the tBu2MeSi-substituents, H atoms of the Ph-
group, H atom on the C1 atom, and the THF molecule coordinated to Li atom are not shown).
ized in terms of the interplay between the two tendencies:
cyclic p-delocalization and thermodynamic stabilization.
If the singly-charged species (cyclopentadienide ion) can
greatly benefit from p-delocalization, which results in the
overall aromatic stabilization of the molecule, the delocal-
ization in the doubly-charged systems (cyclobutadiene
dianion) has a totally opposite, destabilizing, effect due to
Scheme 3.
the unavoidable mixing of the two extra electrons, which
cannot be efficiently separated resulting in a strong Cou-
lomb repulsion [18]. In general, such Coulomb repulsion
spectrum of 2ꢀ Æ Li+ in THF-d8 solution displayed only
three signals for tBu-groups because of the breaking of pen-
tahaptocoordination of Li+-ion to the heavy Cp-ring. The
can overcompensate the stability gained by delocalization,
and consequently the trend for electron localization and
separation may cause the significant departure of the cyclo-
butadiene dianion ring from planarity [18]. However, intro-
duction of the highly positive lithium ions may greatly
counteract this unfavorable tendency, significantly dimin-
ishing the degree of the extra electrons mixing, which
finally results in the much greater extent of the cyclic delo-
calization and, in effect, in the overall aromatic stabiliza-
tion of the system [19]. Recently, the first isolable
dilithium salt of the cyclobutadiene dianion stabilized by
the four p-accepting Me3Si-groups was prepared, and it
was shown to possess a planar four-membered ring in
which the two electrons were fully delocalized as would
be expected for the 6p-electron aromatic system [20]. How-
ever, the situation with the heavy analogues of the cyclo-
butadiene dianion, where the skeletal C atoms are
replaced with the heavier group 14 elements, was still unre-
solved, and no information (either experimental or theoret-
ical) concerning the synthesis and structure of such
derivatives was available in the literature.
For the synthesis of the heavy cyclobutadiene dianions,
at first we designed the appropriate precursors – disil-
adigermetene 3 and tetrabromotetrasiletane 4. The synthe-
sis of the first compound 3 was accomplished by the
unusual ring-enlargement reaction of 3H-disilagermirene
5 with GeCl2–dioxane complex (Scheme 4) [21]. The second
precursor 4 was prepared by the bromination with CHBr3
of tetrasiletane 6, which was synthesized by the reductive
cyclization of RSiHCl2 with lithium naphthalenide
(Scheme 5).
t
resonances of one of the Bu2MeSi-groups are characteris-
tically shifted to higher fields (ꢀ0.33 ppm for Me-group
t
and 0.80 ppm for Bu-groups), indicating the accommoda-
tion of the large part of negative charge on the skeletal
atom next to this tBu2MeSi-group. The resonances of both
skeletal Si atoms of 2ꢀ Æ Li+ now appeared in low-field
characteristic of the doubly-bonded Si atoms: 97.4 and
7
104.9 ppm [16]. The Li NMR spectrum displayed a signal
in the typical region of g1-germyllithiums: at ꢀ0.6 ppm
[17], which is greatly shifted to lower field comparing to
that measured in toluene-d8 (ꢀ5.4 ppm) and clearly outside
the range expected for the aromatic cyclopentadienide
derivatives [15]. All these spectral characteristics of 2ꢀ Æ Li+
undoubtedly evidence for the change of the coordination
mode from the delocalized g5 in nonpolar toluene to a
localized g1 in polar THF (Scheme 3). Thus, one can con-
clude that in THF the heavy cyclopentadienide 2ꢀ acquires
the properties of the localized cyclopentadienide featuring
a negative charge on Ge atom and Si@Si and C@C double
bonds.
2.2. Heavy potassium disiladigerma- and
tetrasilacyclobutadiene dianions
The story of the doubly-charged anionic 6p-electron sys-
tems, such as cyclobutadiene dianion, is greatly different
from that of their singly-charged analogues (cyclopentadie-
nide ion). The origin of such difference can be easily visual-

2804
V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
Scheme 4.
Scheme 5.
The reduction of both 3 and 4 with potassium graphite
Si3(Si4) atoms are 341ꢁ and 326ꢁ, respectively. The Si–K
˚
KC8 (4.2 and 6.9 equiv., respectively) in THF smoothly
and cleanly produced the corresponding dianionic deriva-
tives, 1,2-disila-3,4-digerma- and 1,2,3,4-tetrasilacyclobut-
adiene dianions 72ꢀ Æ 2K+ and 82ꢀ Æ 2K+ (Scheme 6). Both
heavy cyclobutadiene dianions were isolated by the recrys-
tallization from hexane in the form of their THF solvates,
bond distances of 3.3011(8)–3.4044(7) A are normal for
silyl potassium derivatives. These structural features of
8
2ꢀ Æ 2[K+(thf)2] (as well as those of 72ꢀ Æ 2[K+(thf)2], which
are quite similar) definitely do not fit the geometrical crite-
ria of aromaticity in its classical sense (ring planarity, cyclic
bonds equalization, tetrahaptocoordination expected for
the aromatic cyclobutadiene dianion in its D4h conforma-
tion) [19,20]. On the contrary, these structural characteris-
tics are consistent with the non-aromaticity of the
7
2- Æ 2[K+(thf)2] and 82- Æ 2[K+(thf)2], as extremely air- and
moisture-sensitive emerald-green crystals in 70 and 73%
yields.
Since both 72ꢀ Æ 2[K+(thf)2] and 82ꢀ Æ 2[K+(thf)2] are crys-
tallographically isomorphous, below the structural features
of only one of them, 82ꢀ Æ 2[K+(thf)2], will be discussed as
the representative example of both compounds (Fig. 3).
The three structural peculiarities are of decisive impor-
tance: (1) the Si4 four-membered ring is non-planar (fold-
ing angle 34ꢁ), (2) both K+-cations are accommodated
above and below the Si4-ring being dihaptocoordinated at
1,3- and 2,4-positions, (3) the lengths of the skeletal Si–Si
bonds are different: two of them, Si1–Si4 and Si2–Si3, are
˚
nearly equal (2.3301(8) and 2.3300(8) A), whereas Si1–Si2
˚
is shorter (2.2989(8) A) and Si3–Si4 is longer
˚
(2.3576(8) A), therefore the overall shape of the ring is a
folded trapeze. Such geometry of the ring dictates the ori-
entation of the bulky silyl-substituents, which are bent
away from the coordinated K+-cations, thus occupying
the alternating up and down positions. Consequently, all
skeletal Si atoms are appreciably pyramidalized: the sum
of the three Si–Si–Si bond angles around Si1(Si2) and
Scheme 6.

V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
2805
ure of the counter-ions: smaller cations could approach the
cyclobutadiene dianion ring more closely and therefore
more effectively counteract the Coulomb repulsion of the
two negative charges giving the overall stabilization of
the system. Indeed, according to our preliminary results,
the disodium derivative of the heavy cyclobutadiene dian-
ion 82ꢀ Æ 2[Na+(thf)2] has an g4-coordination of both Na
atoms (inversed sandwich structure) and more planar, than
that of 82ꢀ Æ 2[K+(thf)2], Si4-ring [22].
In contrast to the evident similarity of the solid state
structures A, the solution behavior of 72ꢀ Æ 2[K+(thf)2]
and 82ꢀ Æ 2[K+(thf)2] is markedly different, which was
clearly manifested by their 29Si NMR spectra (Scheme 7).
Thus, the resonances of the skeletal Si atoms in 72ꢀ Æ 2K+
were observed in low-field at 113.7 ppm, characteristic of
the doubly-bonded Si atoms [14]. This points out to the
most important contribution of the resonance structure B
with the preferential accommodation of both negative
charges on the more electronegative Ge atoms, leaving
the two skeletal Si atoms doubly-bonded (Scheme 7). That
is, quite similar to the case of the heavy cyclopentadienide
2ꢀ Æ Li+, the difference in electronegativities between the Si
and Ge atoms in 72ꢀ Æ 2K+ favors electron localization.
However, in contrast to 2ꢀ Æ Li+, in which such unfavor-
able electronegativity difference does not overcome the gen-
eral tendency of aromatization of the singly-charged
species, the electronegativity difference totally governs the
situation in highly destabilized doubly-charged 72ꢀ Æ 2K+
breaking the trend of cyclic delocalization and, conse-
quently, making the molecule non-aromatic. In contrast
to 72ꢀ Æ 2K+, the resonances of the skeletal Si atoms of
Fig. 3. Crystal Maker view of 82ꢀ Æ 2[K+(thf)2] (hydrogen atoms are not
shown). Selected bond lengths (A): Si1–Si2 = 2.2989(8), Si2–Si3 =
2.3300(8), Si3–Si4 = 2.3576(8), Si1–Si4 = 2.3301(8), Si1–Si5 = 2.3602(8),
Si2–Si6 = 2.3597(8), Si3–Si7 = 2.3728(7), Si4–Si8 = 2.3719(7). Selected
bond angles (ꢁ): Si2–Si1–Si4 = 88.09(3), Si1–Si2–Si3 = 88.12(3), Si2–Si3–
Si4 = 86.72(3), Si1–Si4–Si3 = 86.74(3). Dihedral angle (ꢁ): Si1–Si2–Si3/
Si1–Si3–Si4 = 34.16(2).
˚
8
2ꢀ Æ 2K+ were measured at 17.0 ppm, which is definitely
outside the range of the doubly-bonded Si atoms. This fact
is to be realized as an indication of more important degree
of delocalization of negative charges in 82ꢀ Æ 2K+ (structure
C), than that in 72ꢀ Æ 2K+ (Scheme 7). This is not surpris-
ing, taking into account that the four-membered ring of
the former compound is constituted of all identical Si
atoms. The planarization of the Si4-ring, which is a neces-
sary prerequisite for the cyclic delocalization, might be
facilitated in solution through the fast ring flipping taking
compounds, at least in the solid state. The requirement for
the commonly accepted magnetic criterion of aromaticity
[13] is also not fulfilled: the calculation of NICS(1) for both
7
2ꢀ Æ 2[K+(thf)2] and 82ꢀ Æ 2[K+(thf)2] (using Me3Si-substi-
tuted model compounds) provided the positive values of
+4.3 and +6.1, indicating the absence of a diatropic ring
current [12b]. It should be noted, that the extent of the
delocalization and aromaticity greatly depends on the nat-
Scheme 7.

2806
V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
Remarkably, the four-membered ring Si2Ge2-core was pla-
nar in contrast to the starting cyclic disiladigermetene 3
[21], which features a folded Ge2Si2-skeleton. The Si@Si
˚
double bond length in 9 of 2.1874(12) A is typical for the
cyclic disilenes [25], and it is expectedly trans-bent
(47.69(8)ꢁ). The methylation of the heavy cyclobutadiene
dianion 72ꢀ Æ 2[K+(thf)2] exclusively proceeded at Ge atoms,
which gives a solid support for the above-discussed thesis
on the predominant localization of the negative charges
on Ge atoms. The Ge-bound Me-groups are regularly
arranged in a trans-fashion to minimize the steric hin-
drances around Ge atoms. Typically for the four-mem-
bered cyclic disilenes [23a,23b,23c], the doubly-bonded
skeletal Si atoms of 9 exhibited a low-field signal at
167.6 ppm.
Scheme 8.
The search for other synthetic applications of the heavy
cyclopentadienide 2ꢀ Æ [Li+(thf)] and heavy cyclobutadiene
dianions 72ꢀ Æ 2[K+(thf)2] and 82ꢀ Æ 2[K+(thf)2], particularly
their utilization as the novel ligands for transition metal
complexes of the new generation, is currently under inves-
tigation in our research group.
3. Experimental
3.1. General procedures
All experiments were carried out using high-vacuum line
techniques or in an argon atmosphere of MBRAUN MB
150B-G glove box. The solvents were predried over sodium
benzophenone ketyl and finally dried and degassed over
potassium mirror in vacuum immediately prior to use.
NMR spectra were recorded on Bruker AC-300FT NMR
(1H NMR at 300.13 MHz; 13C NMR at 75.47 MHz; 29Si
7
NMR at 59.63 MHz; Li NMR at 116.64 MHz) and Bru-
Fig. 4. Crystal Maker view of 9 (hydrogen atoms are not shown). Selected
bond lengths (A): Si3–Si4 = 2.1874(12), Si3–Ge2 = 2.3982(9), Si4–Ge1 =
ker ARX-400FT NMR (1H NMR at 400.23 MHz; 13C
NMR at 100.64 MHz; 29Si NMR at 79.52 MHz) spectrom-
eters. GC–MS spectra were measured on Shimadzu
GCMS-QP5000 instrument and direct mass spectra were
obtained on JEOL JMS SX-102 instrument. UV-spectra
were recorded on Shimadzu UV-3150 UV–Vis spectropho-
tometer in hexane. Elemental analyses were performed at
the Analytical Center of the University of Tsukuba (Tsu-
kuba, Japan).
The starting compounds were prepared according to the
described procedures: 1,2-disila-3-germacyclopentadiene 1
[9], 3H-disilagermirene [26], dichlorogermylene dioxane
complex [27].
˚
2.3967(9), Ge1–Ge2 = 2.4828(5), Si3–Si7 = 2.3653(12), Si4–Si8 = 2.3629(12),
Ge1–Si5 = 2.4295(10), Ge2–Si6 = 2.4288(10), Ge1–C1 = 1.984(3), Ge2–
C2 = 1.989(3). Selected bond angles (ꢁ): Si4–Si3–Ge2 = 93.39(4), Si3–
Ge2–Ge1 = 86.49(2), Ge2–Ge1–Si4 = 86.37(2), Ge1–Si4–Si3 = 93.59(4).
Torsional angle (ꢁ): Si7–Si3–Si4–Si8 = 47.69(8).
place at room temperature. It should be noted, that the
solution structures of both 72ꢀ Æ 2K+ and 82ꢀ Æ 2K+ do
not depend on the polarity of solvents, which was clearly
manifested in the striking similarity of their NMR spectra
measured in both polar THF-d8 and non-polar toluene-d8
(see Section 3).
As expected for the cyclobutadiene dianion derivatives,
both 72ꢀ Æ 2[K+(thf)2] and 82ꢀ Æ 2[K+(thf)2] are excellent sub-
strates for the reaction with a variety of electrophiles to
produce novel inorganic ring systems. For example,
3.2. Synthesis of lithium 1,2,3-tris(di-tert-butylmethylsilyl)-
4-phenyl-1,2-disila-3-germacyclopentadienide
(2ꢀ Æ [Li+(thf)])
7
2ꢀ Æ 2[K+(thf)2] smoothly reacts with Me2SO4 to produce
a single product, 1D-disiladigermetene 9, in which both
Me-groups are bound to Ge atoms, isolated as orange crys-
tals in 95% yield (Scheme 8). The X-ray analysis of 9, rep-
Dry oxygen-free THF (4 ml) was added by vacuum
transfer to a mixture of 1,2-disila-3-germacyclopenta-2,4-
diene 1 (298 mg, 0.35 mmol) and potassium graphite
KC8 (102 mg, 0.76 mmol). The color of the reaction mix-
ture immediately turned to a very dark red. After 30 min
resenting
a
novel hybrid cyclotetrametallene [23],
confirmed its constitution as a cyclic disilene (Fig. 4) [24].

V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
2807
stirring at room temperature, THF was replaced with dry
oxygen-free hexane (4 ml) and graphite was removed by
centrifugation. After evaporation of hexane, dry LiBr
(464 mg, 5.35 mmol) was added to the residue followed
by the vacuum transfer of dry THF (4 ml). After one
day stirring at room temperature in a sealed tube, THF
was again replaced with dry hexane (4 ml) and inorganic
salts were removed by centrifugation. The dark red resi-
due was recrystallized from dry pentane to give
2ꢀ Æ [Li+(thf)] as bright orange plates (92 mg, 34%). M.p.
>135 ꢁC (dec.); 1H NMR (toluene-d8): d 0.19 (s, 3H,
Me), 0.55 (s, 3H, Me), 0.69 (s, 3H, Me), 1.09 (s, 9H,
CMe3), 1.19 (s, 9H, CMe3), 1.24 (s, 9H, CMe3), 1.26 (s,
9H, CMe3), 1.33 (s, 9H, CMe3), 1.34 (s, 9H, CMe3),
7.08 (t, J = 7.4 Hz, 1Hpara), 7.20 (t, J = 7.4 Hz, 2Hmeta),
7.56 (d, J = 7.4 Hz, 2Hortho), 8.21 (s, 1H, C(Ph)@CH);
13C NMR (toluene-d8): d ꢀ4.8 (Me), ꢀ2.9 (Me), ꢀ2.1
(Me), 21.1 (CMe3), 21.4 (CMe3), 21.8 (CMe3), 22.1
(CMe3), 22.6 (CMe3), 23.2 (CMe3), 30.0 (CMe3), 30.2
(CMe3), 30.3 (CMe3), 30.39 (CMe3), 30.44 (2CMe3),
125.0 (Carom), 127.8 (Carom), 128.9 (Carom), 143.2
(C(Ph)@CH), 153.5 (Cipso), 181.4 (C(Ph)@CH); 29Si
NMR (toluene-d8): d 15.5, 23.2, 31.3, 54.4 and 69.1 (skel-
was filtered off and solvent was evaporated in vacuum.
The product Bu2MeSiSiHCl2 was isolated by Kugelrohr
t
distillation as a colorless solid (7.0 g, 57%). B.p. 90–
1
100 ꢁC/10 mmHg; H NMR (C6D6): d 0.04 (s, 3H, Me),
0.97 (s, 18H, 2CMe3), 5.86 (s, 1H, Si–H); 13C NMR
(C6D6): d ꢀ9.2 (Me), 20.9 (CMe3), 28.9 (CMe3); 29Si
NMR (C6D6): d 3.8 (tBu2MeSi), 12.6 (SiHCl2).
3.4. Synthesis of trans,trans,trans-1,2,3,4-tetrakis(di-tert-
butylmethylsilyl)tetrasiletane (6)
A solution of lithium naphthalenide (prepared from
naphthalene (6.2 g, 48 mmol) and lithium (1.4 g,
202 mmol) in 1,2-dimethoxyethane (60 ml)) was added to
t
a solution of Bu2MeSiSiHCl2 (5.7 g, 22 mmol) in THF
(30 ml) at ꢀ78 ꢁC. Then the reaction mixture was stirred
at room temperature for 15 h. After evaporation of
solvent, dry hexane was introduced to the residue and
inorganic salt was filtered off. After evaporation of
hexane, naphthalene was separated by Kugelrohr distilla-
tion under vacuum (50 ꢁC/0.1 mmHg), and the residue
was recrystallized from hexane to give 6 as colorless crys-
1
tals (1.6 g, 42%). M.p. 213–215 ꢁC; H NMR (C6D6): d
7
1
etal Si atoms); Li NMR (toluene-d8): d ꢀ5.4; H NMR
(C6D6): d 0.24 (s, 3H, Me), 0.58 (s, 3H, Me), 0.73 (s,
3H, Me), 1.13 (s, 9H, CMe3), 1.22 (s, 9H, CMe3), 1.26
(s, 9H, CMe3), 1.27 (s, 9H, CMe3), 1.37 (s, 18H,
2CMe3), 7.10 (t, J = 7.4 Hz, 1Hpara), 7.23 (t, J = 7.4 Hz,
2Hmeta), 7.63 (d, J = 7.4 Hz, 2Hortho), 8.30 (s, 1H,
C(Ph)@CH); 13C NMR (C6D6): d ꢀ4.8 (Me), ꢀ3.0
(Me), ꢀ2.1 (Me), 20.9 (4 CMe3), 22.7 (CMe3), 23.0
(CMe3) 30.2 (3CMe3), 30.4 (3CMe3), 124.9 (Carom),
127.7 (Carom), 128.8 (Carom), 143.2 (C(Ph)@CH), 153.5
(Cipso), 181.0 (C(Ph)@CH); 29Si NMR (C6D6): d 15.6,
23.3, 31.2, 54.0 and 68.2 (skeletal Si atoms); 7Li NMR
(C6D6): d ꢀ5.7; 1H NMR (THF-d8): d ꢀ0.33 (s, 3H,
Me), 0.38 (s, 3H, Me), 0.46 (s, 3H, Me), 0.80 (s, 18H,
2CMe3), 1.16 (s, 18H, 2CMe3), 1.19 (s, 18H, 2CMe3),
6.83 (t, J = 7.5 Hz, 1Hpara), 6.99 (t, J = 7.5 Hz, 2Hmeta),
7.46 (d, J = 7.5 Hz, 2Hortho), 8.13 (s, 1H, C(Ph)@CH);
13C NMR (THF-d8): d ꢀ4.4 (Me), ꢀ2.2 (Me), ꢀ0.6
(Me), 21.6 (2 CMe3), 22.0 (2 CMe3), 24.1 (2 CMe3),
30.9 (2CMe3), 31.0 (2CMe3), 31.3 (2CMe3), 122.9 (Carom),
127.0 (Carom), 129.1 (Carom), 150.6 (C(Ph)@CH), 155.6
(Cipso), 198.8 (C(Ph)@CH); 29Si NMR (THF-d8): d 8.0,
0.33 (s, 12H, 4 Me), 1.19 (s, 72H, 4CMe3), 4.03 (s, 4H,
4 Si–H); 13C NMR (C6D6): d ꢀ5.5 (Me), 21.7 (CMe3),
29.9 (CMe3); 29Si NMR (C6D6):
d
ꢀ86.4 (d,
J
¼ 162:6 Hz, Si–H), 13.8 (tBu2MeSi). Anal. Calc.
29Si–1H
for C36H88Si8: C, 57.98; H, 11.89. Found: C, 57.68; H,
11.84%.
3.5. Synthesis of trans,trans,trans-1,2,3,4-tetrabromo-
1,2,3,4-tetrakis(di-tert-butylmethylsilyl)tetrasiletane (4)
A solution of 6 (200 mg, 0.29 mmol) in bromoform
(30 ml) was stirred at 110 ꢁC for 2 h. After the solvent
was evaporated in vacuum, the residue was recrystallized
from hexane to give 4 as colorless crystals (262 mg, 86%).
M.p. 233–234 ꢁC; 1H NMR (C6D6): d 0.62 (s, 12H, 4
Me), 1.29 (s, 72H, 4CMe3); 13C NMR (C6D6): d ꢀ3.1
(Me), 23.3 (CMe3), 31.0 (CMe3); 29Si NMR (C6D6): d 0.1
(Si–Br), 21.3 (tBu2MeSi).
3.6. Synthesis of 1,2-dichloro-1,2,3,4-tetrakis(di-tert-
butylmethylsilyl)-3D-1,2,3,4-disiladigermetene (3)
7
20.4, 28.1, 97.4 and 104.9 (skeletal Si atoms); Li NMR
A mixture of 3H-disilagermirene (60 mg, 0.08 mmol)
and dichlorogermylene dioxane complex (20 mg,
0.09 mmol) was placed in a glass tube, and dry oxygen-
free THF (0.5 ml) was vacuum transferred into this tube.
Reaction immediately took place even at low temperature,
and the color of the reaction mixture changed from dark
red to bright orange. After evaporation of solvent, the
residue was recrystallized from hexane to give 3 quantita-
tively as bright orange crystals. M.p. 158–160 ꢁC; 1H
NMR (C6D6): d 0.36 (s, 6H, 2 Me), 0.54 (s, 6H, 2 Me),
1.15 (s, 18H, 2CMe3), 1.19 (s, 18H, 2CMe3), 1.30 (s,
18H, 2CMe3), 1.37 (s, 18H, 2CMe3); 13C NMR (C6D6):
(THF-d8): d ꢀ0.6; UV–Vis (hexane) kmax/nm (e) 375
(4050), 474 (4020).
3.3. Synthesis of 1,1-di-tert-butyl-2,2-dichloro-1-
t
methyldisilane Bu2MeSiSiHCl2
t
t
A solution of Bu2MeSiNa (prepared from Bu2MeSiBr
(11.2 g, 47 mmol) and Na (9.0 g, 391 mmol)) in heptane
(150 ml) [28] was added at ꢀ40 ꢁC to a solution of HSiCl3
(51.0 g, 377 mmol) in hexane (100 ml). The mixture was
stirred at room temperature for 2 h, then inorganic salt

2808
V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
d ꢀ5.3 (Me), ꢀ3.3 (Me), 22.3 (CMe3), 22.4 (CMe3), 22.46
(CMe3), 22.50 (CMe3), 29.8 (CMe3), 30.0 (CMe3), 30.5
(CMe3), 30.7 (CMe3); 29Si NMR (C6D6): d 12.5 (skeletal
Si), 17.1 and 38.3 (SiMetBu2); UV–Vis (hexane) kmax/nm
(e) 327 (11500), 432 (11600). Anal. Calc. for
C36H84Cl2Ge2Si6: C, 47.96; H, 9.39. Found: C, 47.76; H,
9.12%.
3.9. Synthesis of trans-3,4-dimethyl-1,2,3,4-tetrakis(di-tert-
butylmethylsilyl)-1D-1,2,3,4-disiladigermetene (9)
Excess Me2SO4 (1.59 mmol) was added by vacuum
transfer to the solution of 72ꢀ Æ 2[K+(thf)2] (311 mg,
0.26 mmol) in hexane (2 ml). The color of the reaction mix-
ture quickly changed from dark green to orange. After
evaporation of solvent and excess Me2SO4, the residue
was recrystallized from hexane to afford 9 as orange crys-
3.7. Synthesis of tetrakis(di-tert-butylmethylsilyl)-1,2-
disila-3,4-digermacyclobutadiene dianion dipotassium salt
(72ꢀ Æ 2[K+(thf)2])
1
tals (213 mg, 95%). M.p. 175–177 ꢁC; H NMR (C6D6): d
0.33 (s, 6H, 2 Me), 0.38 (s, 6H, 2 Me), 1.08 (s, 6H, 2 Me,
GeMe), 1.20 (s, 18H, 2CMe3), 1.21 (s, 18H, 2CMe3), 1.22
(s, 18H, 2CMe3), 1.23 (s, 18H, 2CMe3); 13C NMR
(C6D6): d ꢀ5.0 (Me), ꢀ4.0 (Me), 5.5 (Me, GeMe), 21.4
(CMe3), 21.6 (CMe3), 22.2 (CMe3), 22.4 (CMe3), 29.9
(CMe3), 30.28 (CMe3), 30.30 (CMe3), 30.5 (CMe3); 29Si
NMR (C6D6): d 19.0, 23.0, 167.6 (Si@Si); UV–Vis (hexane)
3D-1,2,3,4–Disiladigermetene 3 (190 mg, 0.21 mmol)
and KC8 (120 mg, 0.89 mmol) were placed in a reaction
tube with a magnetic stirring bar. The dry oxygen-free
THF (2 ml) was introduced by vacuum transfer, and the
reaction mixture was stirred at room temperature to give
a dark green solution in 30 min. Then solvent was
removed in vacuum and dry oxygen-free hexane (7 ml)
was introduced by vacuum transfer. After the graphite
was removed by centrifugation, the solution was cooled
to ꢀ30 ꢁC to afford 72ꢀ Æ 2[K+(thf)2] as emerald-green
crystals (176 mg, 70%). 1H NMR (THF-d8): d 0.28 (s,
6H, 2 Me), 0.31 (s, 6H, 2 Me), 1.13 (s, 36H, 4CMe3),
1.15 (s, 36H, 4CMe3); 13C NMR (THF-d8): d ꢀ1.7
(Me), 0.7 (Me), 22.3 (CMe3), 23.3 (CMe3), 31.6 (CMe3),
31.7 (CMe3); 29Si NMR (THF-d8): d 5.3, 20.8, 113.7 (skel-
etal Si atoms). 1H NMR (toluene-d8): d 0.54 (s, 6H, 2
Me), 0.56 (s, 6H, 2 Me), 1.35 (s, 36H, 4CMe3), 1.42 (s,
36H, 4CMe3); 13C NMR (toluene-d8): d ꢀ1.7 (Me), 0.6
(Me), 22.0 (CMe3), 23.1 (CMe3), 31.4 (CMe3), 31.5
(CMe3); 29Si NMR (toluene-d8): d 8.3, 22.6, 111.9 (skele-
tal Si atoms).
k
max/nm (e) 337 (820), 388 (1160), 441 (2890). Anal. Calc.
for C38H90Ge2Si6: C, 53.02; H, 10.54. Found: C, 53.33;
H, 10.28%.
3.10. Crystal structure analyses of the compounds
2ꢀ Æ [Li+(thf)], 82ꢀ Æ 2[K+(thf)2] and 9
The single crystals of compounds 2ꢀ Æ [Li+(thf)],
8
2ꢀ Æ 2[K+(thf)2] and 9 for X-ray diffraction study were
grown from the saturated hexane solutions. Diffraction
data were collected at 120 K on a Mac Science DIP2030
Image Plate Diffractometer with a rotating anode (50 kV,
90 mA) employing graphite-monochromatized Mo-Ka
˚
radiation (k = 0.71070 A). The structures were solved by
the direct method, using SIR-92 program [29], and refined
by the full-matrix least-squares method by SHELXL-97 pro-
gram [30]. The crystal data and experimental parameters
for the X-ray analysis of 2ꢀ Æ [Li+(thf)], 82ꢀ Æ 2[K+(thf)2]
and 9 are listed in Table 1. Crystallographic data of
2ꢀ Æ [Li+(thf)], 82ꢀ Æ 2[K+(thf)2] and 9 have been deposited
with the Cambridge Crystallographic Data Centre.
3.8. Synthesis of tetrakis(di-tert-
butylmethylsilyl)tetrasilacyclobutadiene dianion dipotassium
salt (82ꢀ Æ 2[K+(thf)2])
1,2,3,4-Tetrabromo-1,2,3,4-tetrakis(di-tert-butylmethyl-
silyl)tetrasiletane 4 (48 mg, 0.045 mmol) and KC8 (42 mg,
0.31 mmol) were placed in a reaction tube with a magnetic
stirring bar. The dry oxygen-free THF (1.5 ml) was intro-
duced by vacuum transfer, and the reaction mixture was
stirred at room temperature to give a dark green solution
in 30 min. Then solvent was removed in vacuum and dry
oxygen-free hexane (5 ml) was introduced by vacuum
transfer. After the graphite was removed by centrifugation,
the solution was cooled to ꢀ30 ꢁC to afford 82ꢀ Æ 2[K+(thf)2]
Acknowledgements
This work was supported by the Grants-in-Aid for Sci-
entific Research (Nos. 17550029, 17655014, 18037008,
and 18039004) from the Ministry of Education, Culture,
Sports, Science and Technology of Japan and COE (Center
of Excellence) program.
Appendix A. Supplementary material
1
as emerald-green crystals (37 mg, 73%). H NMR (THF-
CCDC 288755, 238469 and 623571 contain the supple-
d8): d 0.34 (s, 12H, 4 Me), 1.20 (s, 72H, 8CMe3); 13C
NMR (THF-d8): d ꢀ0.4 (Me), 22.6 (CMe3), 31.7 (CMe3);
mentary crystallographic data for 2ꢀ Æ [Li+(thf)],
8
2ꢀ Æ 2[K+(thf)2] and 9. These data can be obtained free of
1
29Si NMR (THF-d8): d 10.0, 17.0 (skeletal Si atoms). H
charge
via
http://www.ccdc.cam.ac.uk/conts/retriev-
NMR (toluene-d8): d 0.55 (s, 12H, 4 Me), 1.40 (s, 72H,
8CMe3); 13C NMR (toluene-d8): d ꢀ0.5 (Me), 22.4
(CMe3), 31.4 (CMe3); 29Si NMR (toluene-d8): d 11.8, 18.4
(skeletal Si atoms).
ing.html, or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax:
(+44) 1223-336-033; or e-mail: deposit@ccdc.cam.ac.uk.
Supplementary data associated with this article can be

V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
2809
Table 1
Crystallographic data and experimental parameters for the crystal structure analysis of 2ꢀ Æ [Li+(thf)], 82ꢀ Æ 2[K+ (thf)2] and 9
Compound
2
ꢀ Æ [Li+(thf)]
42H84GeLiOSi5
8
2ꢀ Æ 2[K+(thf)2]
9
Empirical formula
Formula mass (g molꢀ1
Collection temperature (K)
k (Mo Ka) (A)
Crystal system
Space group
C
825.07
120
0.71070
Monoclinic
P21/c
C52H116K2O4Si8
1108.37
120
0.71070
Triclinic
C38H90Ge2Si6
860.83
120
0.71070
Monoclinic
C2/c
)
˚
ꢀ
P1
Unit cell parameters
˚
a (A)
13.8930(2)
32.5110(8)
11.7920(4)
90
110.6800(10)
90
4983.0(2)
4
13.0830(7)
13.3180(4)
22.2070(10)
89.659(3)
88.870(2)
61.283(3)
3392.7(3)
2
52.7310(12)
12.5420(3)
36.0270(10)
90
120.659(2)
90
20496.0(9)
16
1.116
˚
b (A)
˚
c (A)
a (ꢁ)
b (ꢁ)
c (ꢁ)
3
˚
V (A )
Z
Dcalc (g cmꢀ3
l (mmꢀ1
F(000)
)
1.100
0.763
1796
1.085
0.317
1220
)
1.336
7456
Crystal dimensions (mm)
h Range (ꢁ)
0.50 · 0.10 · 0.05
2.05–27.95
ꢀ18 6 h 6 17, ꢀ42 6 k 6 0,
0 6 l 6 14
53097
0.30 · 0.30 · 0.25
2.53–28.01
0 6 h 6 17, ꢀ14 6 k 6 16,
ꢀ29 6 l 6 29
31087
0.25 · 0.20 · 0.20
2.09–27.89
0 6 h 6 69, 0 6 k 6 16,
ꢀ47 6 l 6 40
97713
Index ranges
Collected reflections
Independent reflections
Rint
11029
0.043
14318
0.0280
23077
0.0600
Reflections used
Parameters
11029
452
1.021
0.0690/1.5463
0.0387
14318
596
1.018
0.0960/1.4696
0.0532
23077
830
1.015
0.0785/0.0000
0.0508
Sa
Weight parameters a/bb
c
R1 [I > 2r(I)]
d
wR2 (all data)
0.1102
0.1535
0.1391
Maximum/minꢀim3 um residual electron
0.454/ꢀ0.620
0.841/ꢀ0.446
0.695/ꢀ0.757
˚
density (e A
)
n
o
0:5
P
2
a
S ¼
½wðF 2o ꢀ F 2c Þ ꢁ=ðn ꢀ pÞ
where n is the number of reflections and p is number of parameters.
2
b
c
w ¼ 1=½r2ðF 2oÞ þ ðaPÞ þ bPꢁ, with P ¼ ðF o2 þ 2F 2c Þ=3.
P
P
R1 ¼ kF oj ꢀ jF ck= jF oj.
wR2 ¼
n
o
0:5
P
P
2
2
d
½wðF 2o ꢀ F c2Þ ꢁ= ½wðF 2oÞ ꢁ
.
(d) R. West, H. Sohn, D.R. Powell, T. Muller, Y. Apeloig, Angew.
¨
Chem., Int. Ed. Engl. 35 (1996) 1002;
(e) W.P. Freeman, T.D. Tilley, G.P.A. Yap, A.L. Rheingold, Angew.
Chem., Int. Ed. Engl. 35 (1996) 882;
(f) W.P. Freeman, T.D. Tilley, L. Liable-Sands, A.L. Rheingold, J.
Am. Chem. Soc. 118 (1996) 10457;
found, in the online version, at doi:10.1016/j.jorganchem.
2007.01.011.
References
[1] F.A. Carey, R.J. Sandberg, Advanced Organic Chemistry, Part A:
Structure and Mechanisms, fourth ed., Kluwer Academic/Plenum
Publishers, New York, 2000.
[2] (a) V.Ya. Lee, A. Sekiguchi, M. Ichinohe, N. Fukaya, J. Organomet.
Chem. 611 (2000) 228;
(g) J.-H. Hong, Y. Pan, P. Boudjouk, Angew. Chem., Int. Ed. Engl.
35 (1996) 186;
(h) S.-B. Choi, P. Boudjouk, P. Wei, J. Am. Chem. Soc. 120(1998) 5814;
(i) S.-B. Choi, P. Boudjouk, J.-H. Hong, Organometallics 18 (1999)
2919;
(b) A. Sekiguchi, V.Ya. Lee, Chem. Rev. 103 (2003) 1429.
[3] (a) J. Dubac, A. Laporterie, G. Manuel, Chem. Rev. 90 (1990)
215;
(j) S.-B. Choi, P. Boudjouk, K. Qin, Organometallics 19 (2000) 1806.
[5] (a) M. Saito, R. Haga, M. Yoshioka, Chem. Commun. (2002)
1002;
(b) E. Colomer, R.J.P. Corriu, M. Lheureux, Chem. Rev. 90 (1990)
265;
(b) M. Saito, R. Haga, M. Yoshioka, Chem. Lett. 32 (2003) 912;
(c) M. Saito, R. Haga, M. Yoshioka, K. Ishimura, S. Nagase,
Angew. Chem., Int. Ed. 44 (2005) 6553;
´
(c) J. Dubac, C. Guerin, P. Meunier, in: Z. Rappoport, Y. Apeloig
(Eds.), The Chemistry of Organic Silicon Compounds, vol. 2, Wiley,
Chichester, 1998 (Part 3, Chapter 34);
(d) R. Haga, M. Saito, M. Yoshioka, J. Am. Chem. Soc. 128 (2006)
4934;
(d) M. Saito, M. Yoshioka, Coord. Chem. Rev. 249 (2005) 765.
[4] (a) W.-C. Joo, J.-H. Hong, S.-B. Choi, H.-E. Son, J. Organomet.
Chem. 391 (1990) 27;
(e) M. Saito, M. Shimosawa, M. Yoshioka, K. Ishimura, S. Nagase,
Organometallics 25 (2006) 2967.
[6] (a) B. Goldfuss, P.v.R. Schleyer, Organometallics 14 (1995) 1553;
(b) B. Goldfuss, P.v.R. Schleyer, F. Hampel, Organometallics 15
(1996) 1755;
(b) J.-H. Hong, P. Boudjouk, J. Am. Chem. Soc. 115 (1993) 5883;
(c) R. West, H. Sohn, U. Bankwitz, J. Calabrese, Y. Apeloig, T.
Mueller, J. Am. Chem. Soc. 117 (1995) 11608;
(c) B. Goldfuss, P.v.R. Schleyer, Organometallics 16 (1997) 1543.

2810
V.Ya. Lee et al. / Journal of Organometallic Chemistry 692 (2007) 2800–2810
[7] T. Nishinaga, Y. Izukawa, K. Komatsu, J. Am. Chem. Soc. 122
(2000) 9312.
[8] Preliminary communications: (a) V.Ya. Lee, R. Kato, M. Ichinohe,
A. Sekiguchi, J. Am. Chem. Soc. 127 (2005) 13142;
(b) V.Ya. Lee, K. Takanashi, T. Matsuno, M. Ichinohe, A. Sekiguchi,
J. Am. Chem. Soc. 126 (2004) 4758.
[17] 7Li NMR resonance of g1-Li r-bonded to a Ge atom in trisgermole
dianion is ꢀ0.93 ppm in THF-d8: J.-H. Hong, Y. Pan, P. Boudjouk,
Angew. Chem., Int. Ed. 35 (1996) 186.
[18] (a) G.v. Zandwijk, R.A.J. Janssen, H.M. Buck, J. Am. Chem. Soc.
112 (1990) 4155;
(b) T. Sommerfeld, J. Am. Chem. Soc. 124 (2002) 1119;
(c) S. Feuerbacher, L.S. Sederbaum, J. Am. Chem. Soc. 125 (2003) 9531.
[19] (a) A.J. Kos, P.v.R. Schleyer, J. Am. Chem. Soc. 102 (1980) 7929;
(b) M. Balci, M.L. McKee, P.v.R. Schleyer, J. Phys. Chem. A 104
(2000) 1246.
[9] V.Ya. Lee, M. Ichinohe, A. Sekiguchi, J. Am. Chem. Soc. 122 (2000)
12604.
[10] The easy reduction with alkali metals of the E@E0 double bonds (E,
E0 = Si, Ge) due to their low-lying LUMOs is well-documented: (a)
V.Ya. Lee, A. Sekiguchi, Chem. Lett. 33 (2004) 84;
(b) A. Sekiguchi, S. Inoue, M. Ichinohe, Y. Arai, J. Am. Chem. Soc.
126 (2004) 9626.
[20] A. Sekiguchi, T. Matsuo, H. Watanabe, J. Am. Chem. Soc. 122
(2000) 5652.
[21] V.Ya. Lee, K. Takanashi, M. Ichinohe, A. Sekiguchi, J. Am. Chem.
Soc. 125 (2003) 6012.
[11] C. Elschenbroich, Organometallics, 3rd ed., Wiley–VCH, Weinheim,
2006.
[22] The details of the crystal structure of disodium derivative
[12] (a) The optimized geometry for the model 2ꢀ Æ Li+ with the H3Si-
8
2ꢀ Æ 2[Na+(thf)2] will be reported in the forthcoming publications.
t
groups instead of the real Bu2MeSi-substituents well reproduced the
[23] Other examples of the cyclotetrametallenes of heavier group 14
elements: (a) M. Kira, T. Iwamoto, C. Kabuto, J. Am. Chem. Soc.
118 (1996) 10303;
crystal structure of 2ꢀ Æ [Li+(thf)], featuring the diagnostic g5-coordi-
nation of Li+-ion and similar trends for the skeletal bond lengths;
(b) The DFT calculations were performed with the GAUSSIAN-98
program package at the B3LYP level using 6-31G (d) basis set for Ge,
Si, O, C, Li, H atoms and LANL2DZ basis set for Ge atoms.
[13] P.v.R. Schleyer, C. Maerker, A. Dransfeld, H. Jiao, N.J.R.v.E.
Hommes, J. Am. Chem. Soc. 1118 (1996) 6317.
(b) N. Wiberg, H. Auer, H. No¨th, J. Knizek, K. Polborn, Angew.
Chem., Int. Ed. 37 (1998) 2869;
(c) A. Sekiguchi, T. Matsuno, M. Ichinohe, J. Am. Chem. Soc. 123
(2001) 12436;
(d) See Ref. [21];
[14] For the chemical shifts of disilenes, see the following reviews: (a) R.
Okazaki, R. West, Adv. Organomet. Chem. 39 (1996) 231;
(b) M. Weidenbruch, in: Z. Rappoport, Y. Apeloig (Eds.), The
Chemistry of Organic Silicon Compounds, vol. 3, Wiley, Chichester,
2001 (Chapter 5).
(e) V.Ya. Lee, K. Takanashi, M. Nakamoto, A. Sekiguchi, Russ.
Chem. Bull., Int. Ed. 53 (2004) 1102.
[24] There are two independent molecules in the unit cell of 9, which
rather insignificantly vary in their geometry, therefore the structural
features of only one of them are discussed.
[15] The aromatic lithium cyclopentadienide derivatives reveal very
diagnostic high field signals in their 7Li NMR spectra due to ring
current effects: (a) ꢀ8.60 ppm for C5H5Li (see Ref. [11]);
(b) ꢀ7.51 ppm for (HMe2Si)5C5Li: A. Sekiguchi, Y. Sugai, K. Ebata,
C. Kabuto, H. Sakurai, J. Am. Chem. Soc. 115 (1993) 1144;
(c) ꢀ5.63 ppm for [g5-Li Æ TMEDA][g5-Li]1/2[g1-Li]1/2[Et4C4Ge]:
S.-B. Choi, P. Boudjouk, J.-H. Hong, Organometallics 18 (1999)
2919.
[25] (a) P.P. Power, Chem. Rev. 99 (1999) 3463;
(b) See Ref. [14b].
[26] V.Ya. Lee, M. Ichinohe, A. Sekiguchi, N. Takagi, S. Nagase, J. Am.
Chem. Soc. 122 (2000) 9034.
[27] S.P. Kolesnikov, I.S. Roqozhin, O.M. Nefedov, Izv. Akad. Nauk
SSSR, Ser. Khim. (1974) 2379.
[28] A. Sekiguchi, T. Fukawa, M. Nakamoto, V.Ya. Lee, M. Ichinohe, J.
Am. Chem. Soc. 124 (2002) 9865.
[16] The resonances of the endocyclic doubly-bonded Si atoms in the five-
membered rings lie in the range 86.0–124.2 ppm: see [9] and A.
Grybat, S. Boomgaarden, W. Saak, H. Marsmann, M. Weidenbruch,
Angew. Chem., Int. Ed. 38 (1999) 2010.
[29] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M.C.
Burla, G. Polidori, M. Camalli, J. Appl. Crystallogr. 27 (1994) 435.
[30] G.M. Sheldrick, SHELXL-97: Program for Crystal Structure Refine-
ment, University of Go¨ttingen, Go¨ttingen, Germany, 1997.