Reactions of (Et2NCH2CH2NEt 2)·H2SiCl2 with selected diorganometallic reagents of magnesium and lithium

Article abstract of DOI:10.1021/om100544f

Addition of the THF-insoluble di-Grignard reagent from 2,2′-dibromo- 4,4′-tert-butylbiphenyl (1) to a solution of [(teeda)·H ₂SiCl₂] in CH₂Cl₂/THF produced 2,7-di-tert-butyl-9H-9-silafluorene (3) in isolated, recrystallized yields of <20%. The related soluble dilithio reagent from bis(2-bromo-4-methylphenyl) methylamine (2) formed in Et₂O, when reacted with [(teeda)·H₂SiCl₂] in CH₂Cl ₂/Et₂O, gave similar yields of 5,10-dihydro-2,5,8- trimethylphenazasiline (4). In the absence of CH₂Cl₂ the major product produced from 1 was the spirocycle 2,2′,7,7′-tetra- tert-butyl-9,9′-spirobi[9H-9-silafluorene] both in a solvent-free form (5′) and as an ethanol solvate (5), both of which were crystallographically characterized. The spirocycle 2,2′,5,5′,8, 8′-hexamethyl-5,10-dihydro-10,10-spirobiphenazasiline (6) was formed from the reaction of the dilithio reagent of 2 in the absence of CH ₂Cl₂.

Full text of DOI:10.1021/om100544f

5708 Organometallics 2010, 29, 5708–5713
DOI: 10.1021/om100544f
Reactions of (Et2NCH2CH2NEt2) H2SiCl2 with Selected
3
Diorganometallic Reagents of Magnesium and Lithium†
Joyce Y. Corey,* Kevin A. Trankler, Janet Braddock-Wilking, and Nigam P. Rath
Department of Chemistry and Biochemistry, University of Missouri;St. Louis, St. Louis, Missouri 63121
Received June 2, 2010
Addition of the THF-insoluble di-Grignard reagent from 2,20-dibromo-4,40-tert-butylbiphenyl (1) to a
solution of [(teeda) H2SiCl2] in CH2Cl2/THF produced 2,7-di-tert-butyl-9H-9-silafluorene (3) in isolated,
3
recrystallized yields of <20%. The related soluble dilithio reagent from bis(2-bromo-4-methylphe-
nyl)methylamine (2) formed in Et2O, when reacted with [(teeda) H2SiCl2] in CH2Cl2/Et2O, gave similar
3
yields of 5,10-dihydro-2,5,8-trimethylphenazasiline (4). In the absence of CH2Cl2 the major product
produced from 1 was the spirocycle 2,20,7,70-tetra-tert-butyl-9,90-spirobi[9H-9-silafluorene] both in a
solvent-free form (50) and as an ethanol solvate (5), both of which were crystallographically characterized.
The spirocycle 2,20,5,50,8,80-hexamethyl-5,10-dihydro-10,10-spirobiphenazasiline (6) was formed from the
reaction of the dilithio reagent of 2 in the absence of CH2Cl2.
The most commonly studied heterocycles of the heavier
elements of group 14 are those containing a silicon heteroatom;
those with two exocyclic hydrides may be viewed as secondary
silanes and, as such, have provided many novel products upon
reaction with transition-metal complexes.1 A common reaction
sequence to silicon heterocycles (and heavier group 14 elements)
requires the synthesis of an appropriate organic dihalide fol-
lowed by conversion to the di-Grignard or dilithio reagent,
which is then quenched with a suitable silane such as H2SiCl2,
HSiCl3, SiCl4, or Si(OMe)4.2 The last three reagents then require
a reducing agent such as LiAlH4 (LAH) to convert the Si-Cl or
Si-OMe to Si-H. The sequence is shown generally in Scheme 1.
When the objective is a silicon heterocycle with two exocyclic
silicon-hydride bonds, the most direct quenching agent for the
diorganometallic intermediate would appear to be H2SiCl2.
However, H2SiCl2 is a gas at room temperature and, since it
ignites on exposure to air, there are safety problems. Boudjouk
and co-workers have reported the disproportionation of HSi-
Cl3 in the presence of N,N,N0,N0-tetraethyl-1,2-ethanediamine
3.3 An X-ray structure verified the six-coordinate geometry
about the silicon center, as illustrated in Figure 1.
The synthetic potential of [(teeda) H2SiCl2] was originally
demonstrated through reaction with PhMgCl in a 1/3 CH2Cl2/
THF mixture to give Ph2SiH2 in a reported yield of 75%. Al-
though the mechanism of the coupling of the Grignard reagent
with the Si-Cl bonds of the six-coordinate complex was not
addressed, the cis orientation of the chlorides suggests that reac-
tion with a diorganometallic reagent (which could be viewed as a
chelating agent) might provide direct access to the target hetero-
cycle illustrated generally in Scheme 1. To test this possibility,
3
the reaction of [(teeda) H2SiCl2] with the diorganometallic
3
reagents formed from 2,20-dibromo-4,40-di-tert-butylbiphenyl
(1) and bis(2-bromo-4-methylphenyl)methylamine (2) to pro-
duce the known silafluorene (3)4 and phenazasiline (4),5 respec-
tively, shown in Figure 2, was examined.
Results and Discussion
To test the possibility of using [(teeda) H2SiCl2] as a precur-
(teeda) to give the solid complex [(teeda) H2SiCl2], which is
3
3
sor for the formation of heterocycles, various reaction conditions
were examined involving changes in solvent or solvent mixtures
as well as direction of addition. Products obtained from the
stable in the absence of air and moisture and is formed in yields
from 86 to 98% when the molar ratio of HSiCl3 to teeda exceeds
reaction of the diorganometallic reagents with [(teeda) H2SiCl2]
3
† Part of the Dietmar Seyferth Festschrift. Dedicated to Professor Dietmar
Seyferth in honor of his outstanding leadership as Editor-in-Chief of
Organometallics.
*To whom correspondence should be addressed. E-mail: corey@
umsl.edu. Fax: (314) 516-5342. Tel: (314) 516-5360.
(1) (a) Corey, J. Y. Adv. Organomet. Chem. 1975, 13, 139. (b) Corey,
J. Y.; Braddock-Wilking J. Chem. Rev. 1999, 99, 175. (c) Corey, J. Y. Chem.
Rev., in press.
(2) One selected example for each silicon monomer are as follows.
H2SiCl2: (a) Corey, J. Y.; John, C. S.; Ohmsted, M. C.; Chang, L. S.
J. Organomet. Chem. 1986, 304, 93. HSiCl3: (b) Chang, L. S.; Corey, J. Y.
Organometallics 1989, 8, 1885. SiCl4: (c) Liu, Y.; Stringfellow, T. C.; Ballweg,
T.; Guzei, I. A.; West, R. J. Am. Chem. Soc. 2002, 124, 49. Si(OMe)4: (d)
Yamaguchi, S.; Jin, R.-Z.; Tamao, K.; Sato, F. J. Org. Chem. 1998, 63, 10060.
(3) (a) Boudjouk, P.; Kloos, S. D.; Kum, B.-K.; Page, M.; Thweatt,
D. J. Chem. Soc., Dalton Trans. 1998, 877. (b) Kloos, S. D.; Boudjouk, P.
Inorg. Synth. 1998, 32, 294.
were examined by GC and GCMS to obtain molecular weights
and thus probable product assignments, particularly in the case
of minor components. The major products were verified by
doping of the GC aliquots with authentic samples as well as by
isolation of solid material and 1H NMR characterization.
To a slurry of the Grignard reagent produced from 2,20-
dibromo-4,40-di-tert-butylbiphenyl (1) with Mg in THF was
added a solution of [(teeda) H2SiCl2] in CH2Cl2. After aqueous
3
workup, distillation provided a fraction that contained 3 (75%)
(4) Chang, L. S.; Corey, J. Y. Organometallics 1989, 8, 1885.
(5) Braddock-Wilking, J.; Corey, J. Y.; French, L. M.; Choi, E.;
Speedie, V. J.; Rutherford, M. F.; Yao, S.; Xu, H.; Rath, N. P.
Organometallics 2006, 25, 3974.
r
pubs.acs.org/Organometallics
Published on Web 08/10/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 21, 2010 5709
Scheme 1
reactions with the diorganometallic systems. Analysis, by
GC, of the crude product showed the presence of C6H5-
C6H5 (m/e 154, 5% by GC), Ph2SiH2 (60% by GC), and
Ph3SiH (m/e 260, 14% by GC), with all three verified by
doping withauthentic samples, aswell asHPh2SiOSiPh2H (m/
e 382, 11% by GC). When the Grignard reagent was prepared
from PhBr and Mg turningsinEt2O and thenaddedtoa slurry
Figure 1. Dichlorodihydro(N,N,N0,N0-tetraethyl-1,2-ethane-
diamine-N0,N0)silicon.3a
of [(teeda) H2SiCl2] in CH2Cl2/Et2O, followed by aqueous
3
workup, the distilled portion contained teeda (26%), biphenyl
(9%), Ph2SiH2 (35%), Ph2Si(OEt)H (13%), and HPh2SiO-
SiPh2H (14%).
When the reaction of 1 or 2 with Mg turnings was monitored
by withdrawing aliquots that were hydrolyzed with D2O and
subjecting them to GC and GCMS studies, at least two side
products were revealed. The 4,40-di-tert-butylbiphenyl formed
from 1 was a mixture of C20H26-nDn (n=0-2), indicating that
proton transfer occurred, most likely to the CMgBr unit. In addi-
tion, another component, C20H26O, was tentatively assigned to
formation of a COMgBr unit that would then form the phenol
upon hydrolysis. Reaction of 2 and Mg similarly gave a mixture
of deuterated N-methyldi-p-tolylamine, C15H17-nN (n=0-2) as
well as C15H17NO assigned to the related phenol. These side
reactions indicate that, effectively, partial removal of the diorga-
Figure 2. 2,7-Di-tert-butyl-9H,9-dihydrosilafluorene (a)4 and
5,10-dihydro-2,5,8-trimethylphenazasiline (b).5
and 4,40-di-tert-butylbiphenyl (referred to henceforth as the
biphenyl) (16%). Recrystallization of the fraction from absolute
EtOH gave 3(91%;∼19% yield based on starting dibromide) as
well as the biphenyl (8.8%). A higher boiling fraction, when
recrystallized twice from EtOH, gave clear crystals of the etha-
nol solvate of the spirocycle 5, as revealed by the X-ray structure.
The 1H NMR spectrum of 5 showed a singlet for the tBu group,
no SiH resonances, and the three characteristic multiplets in
the aromatic region associated with the ring-closed product. In
another experiment performed on a larger scale, the slurry of
the Grignard reagent from 1 in THF as well as a slurry of
nometallic reagent takes place before the reaction with [(teeda)
3
H2SiCl2] has been initiated. Nondistillable reaction residues that
did not melt up to 320 °C and were also insoluble in common
solvents were observed in some reactions.
[(teeda) H2SiCl2] in THF were added simultaneously to a
The ring-closed products have been previously prepared in
low to modest yields (1.5% for 3 from the same di-Grignard
reagent and HSiCl3;4 24% of 4 from the same dilithio
3
reservoir of THF as closely as possible at the same rate. After
workup, distillation provided a forerun that contained 4,40-di-
tert-butylbiphenyl, followed by sublimation of the remainder to
give well-formed blocks of 2,20,7,70-tetra-tert-butyl-9,90-spirobi-
[9H-9-silafluorene] (5; 74% based on the starting dibromide).
5
precursor used in the current study with SiCl4 ). The melting
points and 1H NMR data of the isolated heterocycles in the
current study are comparable to previously reported data.
Removal of the hydrocarbon impurities by recrystallization
was difficult by either the more traditional preparation out-
lined in Scheme 1 or with the use of the teeda complex in the
current study.
1
The spirocycle was characterized by H, 13C, and 29Si NMR
spectroscopy, elemental analysis, and X-ray diffraction.
When the dilithio reagent, produced from 2 in Et2O, was
added to a solution of [(teeda) H2SiCl2] in ∼1/1 CH2Cl2/
3
Et2O, the phenazasiline 4 was isolated in 97% purity (∼16%
yield) after distillation and recrystallization from EtOH.
However, when the diorganometallic was added to a slurry
The crystal structures of both the solvated (5) and unsolvated
forms (50) have also been determined. The molecular structures
of one of the three unique molecules of 50 and one and a half
molecules of 5 (with accompanying disordered solvent) are
shown in Figure 3, and the crystal data and structure refinement
parameters are summarized in Table 1. There are no unusual
features exhibited by either structure. The average Si-C bond
of [(teeda) H2SiCl2] in Et2O, hydrolytic workup and distilla-
3
tion provided a forerun that did not contain 4. The non-
distilled material was recrystallized from cyclohexane to give
the spirocycle 2,20,5,50,8,80-hexamethyl-5,10-dihydro-10,10-
spirobiphenazasiline (6) in approximately 26% yield (based
on the starting dibromide). The spirocycle had previously
been reported in 35% yield from the reaction of the dilithio
reagent generated from 2 with SiCl4.6
0
˚
distances in 5 are 1.869, 1.867, and 1.865 A in the three indepen-
dent molecules and the average for 5 in the one and a half inde-
˚
pendent molecules is 1.865 A. The total of nine C-Si-C angles
in the two forms varies only from 91.49 to 92.46°. The two planes
of the silafluorene are not strictly orthogonal to each other, and
the twist angles between the planes are 84.6-87.1° in 50 and
86.0-88.8° in 5. There are no short intermolecular contacts
between the spirocycle and the ethanol solvate in 5. One
interesting feature of the lattice of 50 (isolated by sublima-
To determine the effect of the absence of CH2Cl2 in
the reaction of PhMgX (X=Cl, Br) with [(teeda) H2SiCl2],
3
PhMgCl in THF was added to a slurry of [(teeda) H2SiCl2] in
3
THF followed by hydrolytic workup, as described for the
3
˚
tion) is a “void” of volume 118 A . The existence of this void
in the lattice can be explained by the fact that no molecule of
(6) (a) Gilman, H.; Zuech, E. A. J. Org. Chem. 1959, 24, 1391. (b)
Gilman, H.; Zuech, E. A. J. Org. Chem. 1962, 27, 2897.

5710 Organometallics, Vol. 29, No. 21, 2010
Corey et al.
Figure 3. (a) Projection view with 50% thermal ellipsoids of the one and a half unique molecules of 5 with the solvent of crystalliza-
˚
tion. H atoms are omitted for clarity. Si-C bond distances range from 1.8586 to 1.8724 A (6 values), and the C-Si-C ring angles
range from 91.5 to 92.5° (3 values). (b) Projection view of one of the three unique molecules of 50 with 30% thermal ellipsoids. H
˚
atoms are omitted for clarity. Si-C bond distances range from 1.865 to 1.873 A (12 values), and the C-Si-C ring angles range
from 91.5 to 92.1° (6 values).
Table 1. Crystal Data and Structure Refinement Details for 5 and 50
ethanol solvate, 5
unsolvated, 50
C40H48Si
556.87
empirical formula
fw
C122H150OSi3
1716.69
100(2)
temp, K
wavelength, A
150(2)
˚
0.710 73
monoclinic
C2/c
0.710 73
monoclinic
P21/n
cryst syst
space group
unit cell dimens
˚
a, A
24.1635(14)
15.3773(9)
29.543(2)
90
107.695(3)
90
10457.8(12); 4
1.090
0.094
19.7846(4)
21.0568(4)
25.5326(5)
90
93.966(1)
90
10611.4(4); 12
1.046
0.090
˚
b, A
˚
c, A
R, deg
β, deg
γ, deg
3
˚
V, A ; Z
calcd density, Mg/m3
abs coeff, mm-1
cryst size, mm3
0.46 ꢀ 0.40 ꢀ 0.28
0.26 ꢀ 0.15 ꢀ 0.15
no. of rflns collected
no. of indep rflns
completeness to θ, % (angle, deg)
abs cor
max, min transmissn
refinement method
no. of data/restraints/params
goodness of fit on F2
final R indices
230 998
122 989
15 726 (R(int) = 0.0479)
100.0% (25.00)
semiempirical from equivalents
0.9738, 0.9583
full-matrix least squares on F2
15 726/2/603
1.018
R1 = 0.0615, wR2 = 0.1182
0.509 and -0.272
23 141 (R(int) = 0.0793)
99.9% (27.00)
none
0.9866, 0.9769
full-matrix least squares on F2
23 141/33/1130
1.041
R1 = 0.0604, wR2 = 0.1359
0.298 and -0.277
-1
˚
largest diff peak and hole, e A
suitable size was available to fill this space in the lattice
when the crystals were obtained by sublimation, but this
space was occupied by solvent when a sample was recrys-
tallized from EtOH.
hypervalent silicon center.3a However, solutions that were
monitored after 1 h of the preparation of the sample exhibited
1
two, unassigned, new H signals downfield from the original
SiH resonance, but the corresponding room-temperature 29Si
NMR spectrum was not reported. No dismutation of [(teeda)
H2SiCl2] in solution has been reported by Boudjouk and co-
The observation of the formation of spirocycles suggests that
both the chloride and hydride ligands in [(teeda) H2SiCl2] are
3
3
replaced. In the solid structure of [(teeda) H2SiCl2], the Si-H
workers,3 but the dismutation of [(tmeda) H2SiCl2] in polar
solvents at elevated temperatures has been observed by
3
3
˚
bonds are shorter by ∼0.17 A than those calculated for H2SiCl2
(also reflected in the high 1JSi-H value of 404 Hz and a higher
Kroke and co-workers,7 as was the dissociation of [(tmeda)
H2SiCl2] into tmeda and H2SiCl2 in solution. Therefore, the
reacting species in a solution of [(teeda) H2SiCl2] could be
3
stretching frequency in the IR spectrum) and the Si-Cl bonds
˚
are ∼0.18 A longer than the value calculated for tetrahedral
3
H2SiCl2,3a which could be consistent with a faster reaction of
either the six-coordinate complex or the tetracoordinate, free
silane H2SiCl2. The isolation of spirocycles indicates that,
whatever species is present, both Si-H and Si-Cl bonds
react, although the order in which this occurs could not be
determined.
the Si-Cl bonds. The actual [(teeda) H2SiCl2] species present
3
in a mixture of CH2Cl2 and THF (at room temperature) has not
been reported. The 1H NMR spectrum of [(teeda) H2SiCl2] in
3
CD3CN and in CDCl3 was obtained at room temperature (one
singlet for the Si-H resonance in each solvent), and the 29Si
NMR spectrum was collected at -30 °C with one resonance
exhibiting SiH coupling reported at high field as expected for a
(7) Fester, G. W.; Eckstein, J.; Gerlach, D.; Wagler, J.; Brendler, E.;
Kroke, E. Inorg. Chem. 2010, 49, 2667.

Article
Organometallics, Vol. 29, No. 21, 2010 5711
used as supplied, and PhMgBr was prepared from PhBr and Mg
turnings in diethyl ether.
Low-resolution mass spectral data (EI, 70 eV) were collected
on a Hewlett-Packard Model 5988A instrument and exact mass
data on a JEOL-JMS-300 instrument (EI). Gas chromato-
graphic analyses were performed on a Shimadzu Model GC-
14A gas chromatograph utilizing a 30 m DB-5 capillary column.
The GC percentages are uncorrected. Melting point determina-
tions were performed on a Mel-Temp electrothermal capillary
melting point apparatus and are uncorrected. Proton, 13C{1H},
and 29Si{1H} nuclear magnetic resonance spectra were recorded
using a Bruker ARX-500 spectrometer equipped with a 5 mm
broadband probe or on an Avance300 instrument equipped with
a 4 nucleus probe. Spectra are referenced internally to residual
solvent peaks (CDCl3 or C6D6) or to external TMS. The DEPT
sequence was utilized in collecting 29Si NMR data. Elemental
analyses were performed by Atlantic Microlab, Inc. The X-ray
crystal structure determination was performed on a Bruker
SMART diffractometer equipped with a CCD area detector.
Formation of the Grignard Reagent from 2,20-Dibromo-4,40-di-
tert-butylbiphenyl (1) and Mg Followed by Reaction with
Figure 4. Numbering used for aromatic proton assignment in
the 1H NMR spectra.
Conclusion
As a precursor to secondary silacycles, the use of the
complex [(teeda) H2SiCl2] as a reactant with a diorganome-
3
tallic does afford the heterocycles 3 and 4 when CH2Cl2 is
part of the reaction mixture, although the isolated yields are
low. The reaction of the dibromide precursors (1 and 2) with
Mg led to undesired byproducts apparently resulting from
reaction with the solvent, as revealed by quenching aliquots
of the solutions with D2O. The N-methyldi-p-tolylamine and
4,40-di-tert-butylbiphenyl that result from this side reaction,
in particular, were difficult to remove from the desired
products. This was also the case when the preparation of 3
was reported previously using a Grignard reagent.4
[(teeda) H2SiCl2]. (a) A solution of 1 (2.01 g, 4.71 mmol) in
3
THF (30 mL containing 0.4 mL of EDB) was added, dropwise,
to a slurry of Mg turnings (0.567 g, 23.3 mmol) in THF (5 mL).
The reaction was slightlyexothermic. About 1/2 h after completion
of the addition, an aliquot was removed and hydrolyzed with
water. A GC/GCMS analysis indicated the presence of 4,40-di-
tert-butylbiphenyl (57%; GC), 2-bromo-4,40-di-tert-butylbiphe-
nyl (14.5%; GC), and 1 (12%; GC) (all three verified by doping
with authentic samples) in addition to a component (12%;
GC) with m/e 282 tentatively assigned to 2-( p-tBuC6H4)-5-tBu-
C6H3OH. The mixture was heated at a gentle reflux for an
additional 2 h. When the mixture was cooled to room tempera-
In addition to the monocycles 3 and 4, the spirocycles
derived from the silafluorene and the phenazasiline, 5 and 6,
respectively, were also obtained in higher yields than for the
monocycles under heterogeneous conditions that exist when
ether is the cosolvent with CH2Cl2 for the teeda complex or
when THF is used without CH2Cl2 as cosolvent. In addition
to the byproducts obtained from the formation of the
diorganometallic reagents, the products observed in the
various reactions with the teeda complex suggest that not
only do the Si-Cl bonds of the complex react but so also do
the Si-H bonds. Products are also observed with melting
points >300 °C whose 1H NMR spectra suggest a polymeric
product that contains at least the skeleton from the starting 1
and 2, indicating another reaction pathway that diverts the
diorganometallic from the formation of the monocycles or
the spirocycles. The reaction of PhMgX (X = Cl, Br) with
the teeda complex under solvent conditions different from
those originally reported3 was also investigated. If only THF
solvent is used in the reaction of PhMgCl with the teeda
complex, the expected product Ph2SiH2 was formed in lower
yields than in the original report. The reaction of PhMgBr
with the teeda complex in CH2Cl2/THF gave Ph2SiH2 in
yields somewhat lower than those reported for PhMgCl, and
when Et2O replaced the THF, the yield of Ph2SiH2 was cut
approximately in half.
ture, athickslurryformed. Asolutionof[(teeda) H2SiCl2] (1.30 g,
3
4.72 mmol) in CH2Cl2 (8 mL) was cannulated into the THF slurry
of the Grignard reagent, whereupon a clear solution formed. The
mixture was stirred overnight before removing the solvent and
adding diethyl ether (30 mL) and HCl (50 mL, 0.2 M). The ether
layer was separated and dried over MgSO4. After removal of the
ether the crude product was vacuum-distilled (Kugelrohr dis-
tillation). For the first fraction (0.547 g, bp 150-190 °C/0.2
mm), GC analysis showed that the sample contained silafluorene
3 (75%) and 4,40-di-tert-butylbiphenyl (16%). The first fraction
was dissolved in absolute EtOH and on cooling to -7 °C provided
0.269 g of 3 as a white solid (91% silafluorene 3 and 8.8% 4,40-di-
tert-butylbiphenyl), mp 101-103 °C (lit.4 mp for 3 containing 3%
4,40-di-tert-butylbiphenyl, 98-99 °C). A second 0.206 g fraction,
with bp up to 250 °C/0.2 mm, was obtained. Recrystallization of
this fraction from absolute EtOH gave 65 mg of a white solid, mp
253-255 °C, which proved to be the spirocycle 2,20,7,70-tetra-tert-
1
butyl-9,90-spirobi[9H-9-silafluorene] (5). H NMR (CDCl3, 300
MHz): δ 1.27 (s, C(CH3)3, 36H), 7.42 (d, J = 2.0 Hz, Ar-H3, 4H),
7.54 (dd, J = 8.0, J = 2.0 Hz, Ar-H2, 4H), 7.82 (d, J = 8.0Hz, Ar-
1
H1, 4H). H NMR (C6D6, 300 MHz): δ 0.99 (s, C(CH3), 36H),
7.46 (dd, Ar-H2, 4H), 7.73 (d, Ar-H3, 4H), 7.92 (d, Ar-H1,
4H) (numbering of the Ar-H protons is the same as depicted in
Figure 4). A second recrystallization from EtOH gave a few shiny
crystals, mp 273-275 °C (as an ethanol solvate), that were used
for the X-ray diffraction study.
Experimental Section
General Procedures. All reactions, unless otherwise noted,
were carried out under an argon atmosphere using standard
Schlenk techniques. The solvents THF and Et2O were distilled
from Na/9-fluorenone,8 and CH2Cl2 was distilled over P2O5
prior to use. Pentane was washed with H2SO4 prior to distilla-
tion from CaH2. The compounds 2,20-dibromo-4,40-di-tert-
butylbiphenyl4 and bis(2-bromo-4-methylphenyl)methylamine6
were prepared by slightly modified literature methods (see the
(b) A solution of 2,20-dibromo-4,40-di-tert-butylbiphenyl
(7.8 g, 18 mmol) in dry THF (20 mL) was added slowly to a
slurry of Mg (7.8 g, 320 mmol, 100 mesh) in THF (20 mL) to
which a small amount of EDB had been added. The mixture was
allowed to react at room temperature with stirring for 16 h and
then cannulated into a pressure addition funnel. A slurry of
Supporting Information), and [(teeda) H2SiCl2]3a was prepared
3
by the literature method. The teeda complex was stored in a
drybox, and portions were transferred when needed. Commer-
cial MeLi (1.6 M in Et2O) and PhMgCl (1.8 M in THF) were
[(teeda) H2SiCl2] (5.0 g, 18 mmol) in THF (40 mL) was placed in
3
a second addition funnel, and the two slurries were added to a
250 mL flask containing THF (40 mL). The mixture was allowed
to react at room temperature for 2 days. After hydrolysis with
saturated NH4Cl solution, the organic layer was dried over
(8) Kamaura, M.; Inanaga, J. Tetrahedron Lett. 1999, 40, 7347.

5712 Organometallics, Vol. 29, No. 21, 2010
Corey et al.
MgSO4 and the volatiles were removed to give a yellow oil.
Kugelrohr distillation provided a forerun, bp up to 100 °C/0.2
mmHg, that contained 4,40-di-tert-butylbiphenyl. Sublimation
through the temperature range 100-250 °C/0.2 mmHg gave
white, well-formed blocks of 2,20,7,70-tetra-tert-butyl-9,90-
spirobi[9H-9-silafluorene] (50): 3.7 g (74% based on the starting
Reaction of PhMgX (X = Cl, Br) with [(teeda) H2SiCl2] in
3
THF. (a) To a slurry of [(teeda) H2SiCl2] (1.00 g, 3.64 mmol) in
3
THF (20 mL) was added, dropwise, a THF solution of PhMgCl
(5 mL, 1.8 M), and the mixture was stirred overnight. After the
usual workup, GC analysis of the crude product showed the
presence of Ph2SiH2 (m/e 184, 60%), Ph3SiH (m/e 260, 14%)
(both verified by doping with an authentic sample), and HPh2-
SiOSiPh2H (m/e 382, 14%). Kugelrohr distillation provided a
fraction, bp 75-97 °C/0.4 mmHg, 0.285 g, that contained,
from GC/GCMS analysis, Ph2SiH2 (76%) and HPh2SiOSiPh2H
(9%). A second fraction, bp 100-190 °C/0.4 mmHg, 0.152 g,
contained the following from GC analysis: Ph2SiH2 (16%), Ph3SiH
(64%), and HPh2SiOSiPh2H (8%). The nondistilled portion, 0.133
g, contained HPh2SiOSiPh2H (60% by GC) and an unidentified
component with a longer retention time (28% by GC).
1
dibromide), mp 243-246 °C. H NMR (CDCl3, 500 MHz): δ
1.28 (s, C(CH3)3, 36H), 7.44 (d, J = 2.0 Hz, Ar-H3, 4H), 7.55
(dd, J = 8.2, 2.0 Hz, Ar-H2, 4H), 7.86 (d, J = 8.2 Hz, Ar-H1,
4H). 13C{1H} NMR (CDCl3, 125.8 MHz): δ 31.6, 35.0, 120.5,
128.4, 131.4, 133.6, 137.7, 150.5. 29Si{1H} NMR (CDCl3, 99.4
MHz): δ -6.52. Mass spectrum: m/e 556 (Mþ). Anal. Calcd for
C40H48Si: C, 86.27; H, 8.69. Found: C, 86.22; H, 8.84.
Reaction of nBuLi with Bis(2-bromo-4-methylphenyl)methyla-
mine (2) Followed by Reaction with [(teeda) H2SiCl2]. (a) A
3
(b) A Grignard reagent was prepared from PhBr (0.80 mL,
1.19 g, 7.6 mmol) in THF (10 mL) that contained EDB (0.2 mL)
and Mg turnings (0.182 g, 7.49 mmol). The PhMgBr was cannu-
solution of 2 (0.496 g, 1.36 mmol) in diethyl ether was cooled in a
dry ice/acetone bath (precipitation occurs), and nBuLi (1.1 mL, 2.5
M) was added. After addition was complete the flask was warmed
to room temperature. An aliquot was withdrawn and hydrolyzed
with D2O (99.96%), and the GC trace showed two major com-
ponents (72% by GC, m/e 211-213, C15H17-nDnN; n = 2 (87%),
1 (7%), 0 (5.7%), and 15% by GC, m/e 227, C15H17NO (suggested
structure, MeN(C6H4Me-p)(C6H3OH-o, Me-p)). The aryllithium
lated into a solution of [(teeda) H2SiCl2] (1.07 g, 3.92 mmol) in a
3
CH2Cl2/THF mixture (5 mL/12 mL), and the flask with the
Grignard reagent was rinsed with additional THF (4 mL), added
to the reaction mixture, and the resultant mixture was stirred over-
night. After aqueous workup with dilute HCl (0.2 M) and extraction
with ether, the ether layer was dried over MgSO4. The volatiles were
removed, and distillation (Kugelrohr) of the remainder provided a
fraction, 0.307 g, bp up to 85 °C/0.4 mmHg (nondistilled portion
weighed 116 mg). Analysis of the distilled fraction by GC indicated
the presence of teeda (12%), biphenyl (12%), and Ph2SiH2 (67%).
Redistillation of this fraction gave a sample that contained biphenyl
(10%), Ph2SiH2 (82%), and HPh2SiOSiPh2H (2.9%).
solution was cannulated into a solution of [(teeda) H2SiCl2] (0.39 g,
3
1.5 mmol) in CH2Cl2/Et2O (8 mL/9 mL), whereupon a precipitate
formed. The mixture was stirred overnight before removing the
solvents and adding diethyl ether (30 mL) and HCl (50 mL, 0.2 M).
The ether layer was separated and dried over MgSO4. After removal
of the ether, the crude product was vacuum-distilled (Kugelrohr
distillation). The first fraction, bp up to 175 °C/0.01 mmHg, 0.154 g,
contained three components >∼6% by GC: MeN(C6H4Me-p)2
(6.4%, m/e 211), phenazasiline 4 (69%, m/e 239), and a component
with m/e 295 (5.5%) assigned to 2,5,8-trimethyl-10-butyl-5,10-
dihydrophenazasiline. The sample was recrystallized from absolute
ethanol to give 4 (97% by GC): 0.051 g, mp 103-103.5 °C (lit.5 mp
(c) A similar reaction on half the scale described in (b) but
with Et2O as solvent gave after workup a fraction, bp up to
85 °C/0.01 mmHg, 0.240 g, that contained the following by GC
and GCMS: teeda (26%), biphenyl (9%), Ph2SiH2 (35%),
Ph2Si(OEt)H (13%), and HPh2SiOSiPh2H (14%) with 0.128 g
nondistilled material.
1
X-ray Structure Determination for 2,20,7,70-tetra-tert-butyl-
9,90-spirobi[9H-9-silafluorene] (50) and Its Ethanol Solvate (5).
Crystals of X-ray diffraction quality were obtained by recrystalliza-
tion (2ꢀ) of5from absolute ethanol, and crystals of 50 were obtained
by sublimation. Crystals of appropriate dimension were used for the
single-crystal X-ray structure determinations using Bruker Kappa
ApexII and Bruker Smart 1K charge coupled device (CCD) detector
systems. All data were collected using graphite-monochromated Mo
103-104.5 °C). H NMR (C6D6, 300 MHz): δ 2.16 (s, C-CH3,
3H), 2.97 (s, N-CH3, 3H), 5.16 (s, SiH2, 2H), 6.71 (d, J = 8.4 Hz,
Ar-H1, 2H), 7.06 (dd, J = 9.0, 2.0 Hz, Ar-H2, 2H), 7.33 (d, J = 2.0
Hz, Ar-H3, 2H) (chemical shift values and multiplicities correspond
to the literature report5).
(b) A solution of 2 (0.700 g, 1.90 mmol) in Et2O (10 mL) was
cooled to -78 °C, at which point 2 begins to precipitate, and
nBuLi (1.5 mL, 2.5 M) was added. The slurry was stirred for 1/2 h
before it was slowly warmed to 0 °C, and the dry ice/acetone
bath was replaced with an ice/water bath. An aliquot hydro-
lyzed with D2O (99.96%) showed two major components:
C15H17-nDnN (m/e 211-213, n = 2 (69%), 1 (12%), 0 (18%);
71% by GC) and 2 (m/e 367, 369, 371; 17% by GC). After an
additional 1/2 h, the lithium reagent was cannulated into a slurry
˚
KR radiation (λ = 0.710 73 A) from a fine-focus sealed-tube X-ray
source. Preliminary unit cell constants were determined with a set of
36 narrow frame scans. Typical data sets consist of combinations of
φ and φ scan frames with typical scan width of 0.5° and a counting
time of 15-30 s/frame at a crystal to detector distance of 4.0-
5.0 cm. The collected frames were integrated using an orientation
matrix determined from the narrow frame scans. Apex II, SMART,
and SAINT software packages9 were used for data collection and
data integration. Analysis of the integrated data did not show any
decay. Final cell constants were determined by global refinement of
reflections from the complete data set. Collected data were corrected
for systematic errors using SADABS10 on the basis of the Laue
symmetry using equivalent reflections.
of [(teeda) H2SiCl2] (0.523 g, 1.92 mmol) in Et2O (20 mL) and
3
the resultant mixture stirred overnight. The solvents were
removed, and diethyl ether (30 mL) and HCl (50 mL, 0.2 M)
were added. The ether layer was separated and dried over
MgSO4. After removal of the ether from the crude product,
distillation (Kugelrohr) provided a fraction, bp up to 130 °C/0.4
mmHg, 0.145 g, which solidified. A GC trace indicated that the
only major component (84%) in the fraction was the starting
dibromide 2. The nondistilled portion, 0.263 g, was dissolved in
CH2Cl2; the CH2Cl2 was then replaced with boiling cyclohexane
and the solution filtered while hot. The filtrate deposited a solid,
2,20,5,50,8,80-hexamethyl-5,10-dihydro-10,10-spirobiphenazasi-
line (6): 0.104 g, mp 227-230 °C (lit.6 mp 230-233 °C after four
recrystallizations). 1H NMR (C6D6, 300 MHz): δ 1.95 (s, CCH3,
12H), 3.23 (s, NCH3, 6H), 6.91 (d, J = 8.5 Hz, Ar-H1, 4H), 7.11
(dd, J = 8.5, 2.2 Hz, Ar-H2, 4H), 7.45 (d, J = 2.2 Hz, Ar-H3,
4H). 1H NMR (CDCl3, 300 MHz): δ 2.19 (s, CCH3, 12H), 3.65
(s, NCH3, 6H), 7.07 (d, Ar-H1, 4H), 7.11(d, Ar-H3, 4H), 7.22
(dd, Ar-H2, 4H). MS (EI, 70 eV): m/e found 446.2168, calcd
446.2178.
Crystal data and intensity data collection parameters are
given in Table 1.
Structure solution and refinement were carried out using the
SHELXTL-PLUS software package.11 The structures were
solved by direct methods in monoclinic space groups C2/c and
P21/n, respectively. The models were refined with full-matrix
P
2 2
least-squares refinement by minimizing w(Fo2 - Fc ) . All non-
hydrogen atoms were refined anisotropically to convergence.
(9) SMART and SAINT; Bruker Analytical X-Ray, Madison, WI.
(10) SADABS; Bruker Analytical X-Ray, Madison, WI, 2008.
(11) Sheldrick, G. M. Bruker-SHELXTL. Acta Crystallogr. 2008,
A64, 112.

Article
Organometallics, Vol. 29, No. 21, 2010 5713
t
One of the Bu groups is disordered over two positions for 50.
Disorder was resolved with 80%/20% occupancy atoms. Geo-
metrical and displacement parameter restraints (SADI, ISOR,
SIMU) were used for the disordered group in 50, whereas the
geometric restraint DFIX was used for the solvent molecule in 5.
All H atoms were added in calculated positions and were refined
using appropriate riding models (AFIX m3). The models were
refined to convergence to the final residual values of R1 = 4.6%
and 6.0% and wR2 = 13.1 and 15.9%, respectively.
Acknowledgment. J.Y.C. wishes to thank Teresa
Bandrowsky and J. B. Carroll for assistance in col-
lection of the NMR data and R. E. Winter and
J. Kramer for collecting the mass spectral data. Fund-
ing from the National Science Foundation (MRI, No.
CHE-0420497) for the purchase of the ApexII diffract-
ometer is acknowledged.
Complete listings of geometrical parameters, positional and
isotropic displacement coefficients for hydrogen atoms, and
anisotropic displacement coefficients for the non-hydrogen
atoms are available as Supporting Information. Tables of
calculated and observed structure factors are available from
the authors in electronic format.
Supporting Information Available: Text giving experimental
details for compounds 1 and 2 and tables, figures, and CIF files
giving crystallographic data for 5 and 50, including crystal data
and structure refinement details, atomic coordinates, and bond
distances and angles. This material is available free of charge via
the Internet at http://pubs.acs.org.