Synthesis and characterization of Lewis base-free, σ-bonded lithium aryls: A structural model for unsolvated phenyllithium in the solid state

Article abstract of DOI:10.1021/ja962889i

The synthesis and characterization of four Lewis base-free, σ-bonded lithium aryls are reported. This work was undertaken in order to provide a model for the solid-state structure of phenyllithium, which is currently unknown. Nondonor hydrocarbon solubili

Full text of DOI:10.1021/ja962889i

J. Am. Chem. Soc. 1997, 119, 2847-2852
2847
Synthesis and Characterization of Lewis Base-Free, σ-Bonded
Lithium Aryls: A Structural Model for Unsolvated
Phenyllithium in the Solid State
Rudolf J. Wehmschulte and Philip P. Power*
Contribution from the Department of Chemistry, UniVersity of California,
DaVis, California 95616
ReceiVed August 19, 1996X
Abstract: The synthesis and characterization of four Lewis base-free, σ-bonded lithium aryls are reported. This
work was undertaken in order to provide a model for the solid-state structure of phenyllithium, which is currently
unknown. Nondonor hydrocarbon solubility of the four lithium aryls (LiC6H3-3,5-t-Bu2)6 (1), (LiC6H4-4-t-Bu)n (2),
(LiC6H4-4-n-Bu)n (3), and {LiC6H4-4-SiMe2(t-Bu)}n (4) was achieved by the incorporation of meta- or para-substituents
on the aryl rings. This permitted 13C NMR spectroscopy and crystal growth using their solutions. It is proposed
that the absence of bulky ortho-substituents allows association of the lithium aryls to occur in a manner similar to
that of phenyllithium itself. The 13C NMR data for the ipso-carbon atoms suggest an association number of at least
four or, more probably, six in solution. These data are in agreement with the X-ray crystal structure of 1, which is
hexameric, with a distorted octahedral (trigonal antiprismatic) array of lithium ions. Six of the eight Li3 faces are
capped by an aryl group that interacts primarily through the C(ipso) atom. Weaker Li-C(ortho) interactions are
also apparent. This structure is the first of this type for an unsolvated, σ-bonded lithium aryl. Crystal data with Cu
KR (λ ) 1.541 78 Å) radiation for 1 at 130 K: 1, C84H126Li6, M ) 1177.49, a ) 13.516(2) Å, b ) 15.124(3) Å, c
) 20.958(3) Å, R ) 84.084(13)°, â ) 86.249(11) Å, γ ) 68.675(13)°, V ) 3967.7(11) Å3, Z ) 2, space group P1h,
R1 ) 0.085 for 7983 (I > 2σ(I)) data.
Introduction
within the tetramer. The dimers (LiC6H3-2,6-Mes2)2 (B) and
(LiC6H3-2,6-Dipp2)2 (C) display different types of π-bonding
to ortho-aryl substituents of the terphenyl ligands.
The structures of unsolvated, σ-bonded lithium aryls are not
as well investigated as their alkyl counterparts.1 For example,
full details of the solid-state structure of the simplest lithium
aryl, i.e. LiPh, remain unavailable. Part of the reason for this
is the lack of hydrocarbon solubility and low volatility of simple
derivatives such as LiPh, LiC6H4-4-Me, or LiMes (Mes ) 2,4,6-
Me3C6H2-), which makes the growth of crystals suitable for
X-ray crystallography problematical in the absence of donor
solvents such as ethers or amines. Hydrocarbon solubility and
consequent ease of crystal growth may be increased by using
lipophilic aryl ring substituents, and this has allowed the crystal
The observation of π-complexation of the Li+ ion in these2,3
and related4 sterically encumbered lithium aryls suggests that
π-interactions of this type might be a common feature of
σ-bonded lithium aryl structures in general. This hypothesis,
however, has received little support from 13C NMR spectros-
copy.5,6 Thus, Lewis base-free LiPh in the solid state displays
a C(ipso) chemical shift5a that is very close to that of LiPh in
Et2O solution which, by previous NMR spectroscopic studies,6d
was shown to have a structure very similar to that of the tetramer
(Et2O‚LiPh)4 in the crystal phase.7 The X-ray crystal structure
of the latter reveals no Li+-π-aryl ring interactions and has a
distorted cubane arrangement of four tetrahedrally disposed,
ether-solvated Li+ ions and four σ-bonded phenyls, which triply
bridge the Li3 faces through a C(ipso) atom.7 In contrast, exact
details of the Lewis base-free LiPh structure in the solid state
are not currently available.
2
structures of the unsolvated σ-bonded aryl lithiums (LiTrip)4
(Trip ) 2,4,6-i-Pr3C6H2-), (LiC6H3-2,6-Mes2)2 , and (LiC6H3-
2
2,6-Dipp2)23 (Dipp ) 2,6-i-Pr2C6H3-) to be determined. These
are illustrated in Figure 1. In addition, the structure of a weakly
complexed benzene adduct, η6-C6H6‚LiC6H3-2,6-Trip2, which
has a unique, unassociated structure, is known.3 A characteristic
feature of the structures of these sterically crowded lithium aryls
is that the Li+ ion, in addition to being σ-bonded to a phenyl
ring, is also π-complexed, either inter- or intramolecularly, by
nearby aryl ring(s). Thus, the structure of (LiTrip)4 is a tetramer
(structure A, Figure 1) in which the four Li+ ions lie ap-
proximately in a plane, and each Li+ is σ-bound to one aryl
and π-bonded to an aryl ring from a neighboring molecule
To investigate the solid-state structure of uncomplexed LiPh,
four lithium aryl derivatives, which have hydrocarbon-solubi-
lizing substituents only at the meta- or para-positions of the
phenyl ring, have been synthesized. The absence of bulky
substituents at the ortho-positions was expected to allow
(4) Kurz, S.; Hey-Hawkins, E. Organometallics, 1992, 11, 2729.
(5) (a) Johnels, D.; Edlund, U. J. Organomet. Chem. 1990, 393, C35.
(b) Berger, S.; Fleischer, U.; Geletneky, C.; Lohrenz, J. C. W. Chem. Ber.
1995, 128, 1183.
(6) (a) Waack, R.; West, P.; Doran, M. A. Chem. Ind. (London) 1966,
1035. (b) Jones, A. J.; Grant, D. M.; Russell, J. G; Fraenkel, G. J. Phys.
Chem. 1969, 73, 1624. (c) Seebach, D.; Ha¨ssig, R.; Gabriel, J. HelV. Chem.
Acta, 1983, 66, 308. (d) Jackman, L. M.; Scarmoutzos, L. M. J. Am. Chem.
Soc. 1984, 106, 4627. (e) Bauer, W.; Winchester, W. R. ; Schleyer, P.v.R.
Organometallics, 1987, 6, 2371. (f) Wehman, E.; Jastrzebski, T. B. H.,
Ernsting, J.-M; Grove, D. M.; van Koten, G. J. Organomet. Chem. 1988,
353, 133.
X Abstract published in AdVance ACS Abstracts, March 1, 1997.
(1) (a) Setzer, W. N.; Schleyer, P.v.R. AdV. in Organomet. Chem. 1985,
24, 354. (b) Weiss, E. Angew. Chem. Int. Ed. Engl. 1993, 32, 1501. (c)
Beswick, M. A.; Wright, D. S. ComprehensiVe Organomet. Chem. II;
Pergamon Press: New York, 1995, Vol. 1, Chapter 1. (d) Methoden der
Organischen Chemie (Houben-Weyl, Carbanionen, Band E19d, Hanack,
M., Hrsg.), Georg Thieme Verlag: Stuttgart, 1993.
(2) Ruhlandt-Senge, K.; Ellison, J. J.; Wehmschulte, R. J.; Pauer, F.;
Power, P. P. J. Am. Chem. Soc. 1993, 115, 11353.
(3) Schiemenz, B.; Power, P. P. Angew. Chem. Int. Ed. Engl. 1996, 35,
2150.
(7) Hope, H.; Power, P. P. J. Am. Chem. Soc. 1983, 105, 5320.
S0002-7863(96)02889-2 CCC: $14.00 © 1997 American Chemical Society

2848 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Wehmschulte and Power
Figure 1. Schematic drawings of the currently known unsolvated lithium aryls (LiTrip)4 (A), (LiC6H3-2,6-Mes2)2 (B), and (LiC6H3-2,6-Dipp2)2
(C). Carbon atoms are shown as circles, lithium atoms as thermal ellipsoids (30%), and σ-bonds as solid lines. Hydrogen atoms are not shown.
data after addition of a second equivalent Et2O ca. 1.5 h later: 8.26 (s,
o-H), 7.41 (s, p-H), 3.08 (q, OCH2), 3JHH ) 6.9 Hz, 1.50 (s, C(CH3)3),
0.85 (t, OCH2CH3). 13C{1H} NMR: 171.8 (i-C), 147.5 (p-C), 133.8
(o-C), 120.2 (m-C), 65.6(OCH2), 34.9 (C(CH3)3, 32.2 (C(CH3)3), 14.9
(OCH2CH3). 7Li NMR: 2.20. Et2O in C6D6 for comparison: 1H
NMR: 3.25 (q, OCH2), 3JHH ) 6.9 Hz, 1.11 (t, CH3). 13C{1H} NMR:
65.9 (OCH2), 15.5 (CH3).
association of the lithium aryls to occur in a manner similar to
that in phenyllithium itself, while lipophilic substituents at the
meta- or para-positions should provide sufficient hydrocarbon
solubility for solution spectroscopy and crystal growth from their
solutions. The results of these experiments are now reported.
Experimental Details
(THF‚LiC6H3-3,5-t-Bu2)4. Tetrahydrofuran (14.4 µL, 0.18 mmol,
12.8 mg) was added to a slurry of 1 (0.035 g, 0.03 mmol) in ca. 0.4
mL of C6D6 at room temperature to give a clear, colorless solution.
Further equivalents of THF were added such that the ratios of Li:THF
General Procedures. All reactions were performed by using
modified Schlenk techniques under an inert atmosphere of N2 or in a
Vacuum Atmospheres HE43-2 drybox. Solvents were freshly distilled
under N2 from Na/K or Na/K/benzophenone ketyl and degassed twice
before use. NMR spectra were obtained on a General Electric QE-
300 NMR spectrometer and referenced to an internal standard. 7Li
NMR spectra were referenced to the external standard LiCl in D2O
solution. Li(t-Bu) (1.5 M) in pentane, 4-n-BuC6H4Br, 4-t-BuC6H4Br,
1,4-dibromobenzene, and t-BuMe2SiCl were purchased commercially
and used as received. 3,5-t-BuC6H3Br was synthesized by a literature
procedure.8a 4-t-BuMe2SiC6H4Br was synthesized in a manner similar
to 4-Ph3SiC6H4Br,8b by the reaction of t-BuMe2SiCl with LiC6H4-4-
Br, and purified by distillation under reduced pressure (bp 67-69 °C,
0.1 Torr, 59% yield).
(LiC6H3-3,5-t-Bu2)6 (1). A solution of Li(t-Bu) (9.4 mmol) in
benzene (20 mL) was added dropwise to a solution of 3,5-t-Bu2C6H3-
Br (1.28 g, 4.7 mmol) in benzene (30 mL) at room temperature. This
produced slight warming, and a very fine white solid began to
precipitate from the colorless solution after ca. 30 min. After being
stirred for 3 days, the precipitate was allowed to settle, and the colorless,
slightly cloudy supernatant solution was decanted and was concentrated
to ca. 30-40 mL. Cooling in a 6 °C refrigerator for 1 week gave 0.15
g of 1 as large (1-2 mm) colorless prisms, which were suitable for
X-ray crystallography. Concentration of the mother liquor to ca. 5
mL followed by cooling in a 6 °C refrigerator for 4 days gave a further
0.06 g of (LiC6H3-3,5-t-Bu2)6. Yield: 22.8%. Mp: softens at ca. 127
°C and gradually darkens from orange (ca. 195 °C) to black (ca. 290
°C). No melting was observed below 310 °C. Anal. Calcd for
LiC14H21: C, 85.68, H, 10.79. Found: C, 85.90, H, 10.10. 1H NMR
(C6D6, 23 °C): 7.84 (s, o-H, 2H), 7.50 (s, p-H, 1H), 1.28 (s, m-C(CH3)3,
18H). 13C{1H} NMR (C6D6, 23 °C): 170.6 (broad, i-C), 150.7 (p-C),
132.7 (o-C), 122.1 (m-C), 35.0 (m-C(CH3)3), 31.8 (m-C(CH3)3). 7Li
NMR(C6D6, 23 °C): 3.5 (s).
1
were 1:2, 3, 5, and 10. NMR spectra were recorded in each case. H
and 13C{1H} NMR spectra were recorded within 60 min of each THF
addition. These spectra are illustrated in Figure 5 (vide infra). 1H
NMR: 8.29 (s, o-H), 7.43 (s, p-H), 3.12 (m, OCH2), 1.52 (s, m-C(CH3)3,
1.11 (m, CH2). 13C{1H} NMR: 173.4 (i-C), 146.9 (p-C), 134.2 (o-C),
119.5 (m-C), 67.8 (OCH2), 35.0 (m-C(CH3)3), 32.3 (m-C(CH3)3), 25.2
(CH2).
Reaction between 1 and Excess THF (vide infra Figue 5). 1H
NMR: 8.08 (s, o-H), 7.21 (s, p-H), 3.52 (m, OCH2), 1.46 (m, m-C(CH3)3
and CH2). 13C{1H} NMR: 183.6 (i-C), 144.8 (p-C), 136.9 (o-C), 117.3
(m-C), 67.8 (OCH2), 34.8 (m-C(CH3)3), 32.5 (m-C(CH3)3), 25.8 (CH2).
THF in C6D6 for comparison: 1H NMR: 3.55 (m, OCH2), 1.42 (m,
CH2). 13C{1H} NMR: 67.7 (OCH2), 25.8 (CH2).
(LiC6H4-4-t-Bu)n (2). A solution of Li(t-Bu) (20.8 mmol) in benzene
(15 mL) was added dropwise to a solution of 4-t-BuC6H4Br (2.22 g,
10.4 mmol) in benzene at room temperature. The solution gradually
became warm, and a white, very fine precipitate began to form after
ca. 5 min. Stirring was discontinued and the slurry was left standing
at room temperature for ca. 20 h. This solution was then cooled in a
6 °C refrigerator for 5 days. Removal of the turbid, colorless
supernatant liquid gave 0.65 g of microcrystalline 2, which can be
recrystallized from refluxing toluene to afford the product as very small
colorless needles which were dried under reduced pressure. Yield:
44.6%. Mp: turns red at 170 °C; melts and turns dark red at 205-6
°C (dec). 1H NMR (C6D6, 23 °C): 7.86 (d, o-H, 2H), 3JHH ) 6.9 Hz,
7.27 (d, m-H, 2H), 1.25 (s, p-C(CH3)3). 13C{1H} NMR (C6D6, 85 °C),
150.6 (p-C), 140.6 (o-C), 125.8 (m-C), 34.4 (p-C(CH3)3), 31.4 (p-
C(CH3)3). 13C{1H} NMR (C7D8, 110 °C): 165.8 (i-C), 150.8 (p-C),
140.6 (o-C), 125.9 (m-C), 34.7 (p-C(CH3)3), 31.6 (p-C(CH3)3). 7Li
NMR (C6D6, 23 °C, referenced to external LiCl in D2O): 2.7 (s). Anal.
Calcd for C11.17H13.34Li (1/n2‚1/6PhMe): C, 86.80; H, 8.70. Found: C,
86.31, H, 8.52. The compound 2 crystallizes as a solvate from both
toluene and benzene. The ratio of LiC6H4-4-t-Bu to C6H6 or PhMe is
6:1 by 1H NMR. The solvent may be removed under reduced pressure.
(Et2O‚LiC6H3-3,5-t-Bu2)4. Et2O (13.6 µL, 0.13 mmol, 9.6 mg) was
added to 1 (0.025 g, 0.022 mmol) in ca. 0.4 mL of C6D6 in an NMR
tube at room temperature. After ca. 15 min, a clear, colorless solution
resulted. The NMR spectra taken 15 or 60 min after Et2O addition
were identical. 1H-NMR: 8.26 (s, broad, o-H), 7.45 (s, p-H), 2.86 (q,
(LiC6H4-4-n-Bu)n (3). Li(t-Bu) solution (1.5 M;12.4 mL, 18.6
mmol) in n-pentane was added dropwise to a solution of 1.98 g (9.3
mmol) of 4-n-BuC6H4Br in pentane at 0 °C. The slightly cloudy, pale
yellow reaction mixture was slowly warmed to room temperature (ca.
2 h) and stirred for 3 days. The finely divided, off-white solid was
collected on a glass frit, washed with n-pentane (30 mL), and was dried
under reduced pressure. Yield: 1.63 g. The NMR spectra of this solid,
3
OCH2), JHH ) 6.9 Hz, 1.50 (s, C(CH3)3), 0.57 (t, OCH2CH3). 13C-
{1H} NMR: 171.1 (i-C), 147.9 (p-C), 133.4 (o-C), 120.3 (m-C), 65.2
(OCH2), 34.9 (C(CH3)3), 32.1 (C(CH3)3), 14.2 (OCH2CH3). The NMR
(8) (a) Bartlett, P. D.; Roha, M.; Stiles, R. M. J. Am. Chem. Soc. 1954,
76, 2349. (b) Organometallic Syntheses, Eisch, J. J., King, R. B., eds.,
Academic Press, New York, 1981; Vol. 2, pp 93-94.

σ-Bonded Lithium Aryls
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2849
+ Li(n-Bu) hexane8 LiArV + n-BuX
(1)
which is sparingly soluble in hot C6D6 and is a mixture of 3 and ca.
ArX
(X ) Br, I)
20% LiBr on the basis of elemental analysis, are as follows. 1H NMR
3
(C6D6, 80 °C): 7.76 (d, o-H, 2H), JHH ) 7.5 Hz, 7.10 (d, m-H, 2H),
3
3
2.50 (t, R-CH2, 2H), JHH ) 7.5 Hz, 1.57 (tt, â-CH2, 2H), JHH ) 7.5
Hz, 1.30 (tq, γ-CH2, 2H), 0.87 (t, CH3, 3H), 3JHH ) 7.5 Hz. 13C{1H}
NMR (C6D6, 85 °C): 166.2 (i-C), 142.2 (p-C), 140.6 (o-C), 129.2 (m-
C), 36.0 (R-C), 33.8 (â-C), 22.7 (γ-C), 13.9 (CH3). 7Li NMR (C6D6,
80 °C, referenced to external LiCl in D2O): 2.2 (s). Attempts to extract
pure 3 from this solid by using refluxing toluene (80 mL) led to a
clear, red solution, from which 0.29 g of a pale, off-white, microcrys-
talline solid precipitated after cooling to room temperature. This solid,
which appeared to be less soluble in hot C6D6 than the crude product,
was identified as 3 by its conversion to Me3SiC6H4-4-n-Bu in 77%
yield by the reaction with excess Me3SiCl (1.0 mL) in 20 mL of
n-pentane for 3 days at room temperature. Elemental analysis shows
the presence of LiBr even after recrystallization.
[LiC6H4-4-SiMe2-t-Bu]n (4). Li(t-Bu) solution (1.5 M; 9.5 mL, 14.2
mmol) in pentane was added dropwise to a solution of 1.93 (7.1 mmol)
BrC6H4-4-SiMe2-t-Bu in ca. 100 mL pentane at room temperature. After
ca. 30 min, the reaction mixture became pale yellow and a fine white
precipitate began to form. The reaction vessel was then cooled with a
water bath. Stirring was discontinued after 4 h to allow the fine,
voluminous precipitate (a mixture of 4 and LiBr) to settle. Because
after 4 days the precipitate still occupied two-thirds of the reaction
mixture, it was collected on a glass frit and dried under reduced pressure
(crude yield, 0.97 g). Extraction with hot toluene (40 mL, 90 °C, 5
min), followed by decanting the slightly cloudy supernatant liquid and
subsequent cooling to ca. -20 °C for 1 week, resulted in the isolation
of 0.23 g of 4 (contaminated with LiBr) as a colorless microcrystalline
solid (tiny needles), which was washed twice with n-pentane (2 × 20
mL) and dried under reduced pressure. Yield: 16.3%. 1H NMR(C6D6,
22 °C): 7.76 (d, o-H, 2H), 3JHH ) 6.9 Hz, 7.44 (d, m-H, 2H), 0.93 (s,
C(CH3)3, 9H), 0.25 (s, Si(CH3)2, 6H). 13C{1H} NMR(C6D6, 22 °C):
169.8 (i-C), 139.5 (o-C), 136.5 (p-C), 134.7 (m-C), 26.8 (C(CH3)3),
17.2 (C(CH3)3), -6.1 (Si-CH3). 7Li NMR(C6D6, 22 °C): 2.4 (s).
Mp: turns pink at 170 °C; gradually darkens until it melts and turns
dark red at 205-7 °C.
For example, in the case of [LiC6H4-4-t-Bu]n, NMR data
revealed the slow formation of the lithium aryl which initially
remains in solution. After a ca. 50% conversion, however, a
white microcrystalline precipitate began to form and additional
n-Bu signals appeared in the H NMR spectrum. These were
assigned to 4-t-BuC6H4-n-Bu, which is the product of the
competing reaction between LiAr and n-BuBr (eq 2).
1
LiAr + n-BuBr f n-BuAr + LiBrV
(2)
This coupling reaction is apparently suppressed in the case
of simple lithium aryl species such as LiPh or LiC6H4-4-Me by
their insolubility in nondonor hydrocarbons. Thus, they are
removed from the equilibrium by precipitation, which tends to
drive the reaction in eq 1 to completion. For sterically hindered
ortho-substituted aryls (i.e., compounds A, B, or C) this reaction
also appears to be suppressed, but in this case it is probably for
steric reasons. To minimize this coupling reaction the meta-
and para-substituted aryl halides were lithiated by 2 equiv of
the more reactive Li(t-Bu) according to the method of Seebach12
(eq 3).
ArBr + 2Li(t-Bu) f
LiAr + t-BuH + (CH3)2CHdCH2 + LiBr (3)
In this method the reaction between LiAr and t-BuBr is
greatly diminished by the competing reaction between Li(t-Bu)
and t-BuBr to afford LiBr, t-BuH, and Me2CHdCH2. Monitor-
ing these reaction mixtures by NMR spectroscopy showed
clearly the formation of the respective lithium aryls. In the case
of the more soluble species (LiC6H3-3,5-t-Bu2)6 (1), the very
fine (almost colloidal) LiBr precipitate was allowed to settle
for several days, after which time the supernatant liquid was
decanted, concentrated under reduced pressure, and cooled to
ca. 6 °C for 1 week to afford the crystalline product in low
yields (ca. ∼20%). The compound [LiC6H4-4-t-Bu]n (2)
crystallizes directly from the reaction mixture in the form of
small colorless spherical aggregates (up to 1 mm diameter)
which contain some benzene. Recrystallization from hot
benzene or toluene affords small colorless needles of 2 which
are solvates of the formula LiC6H4-4-t-Bu‚1/6C6H6 or LiC6H4-
4-t-Bu‚1/6PhMe, as determined by 1H NMR. The molecules of
the solvent may be evaporated under reduced pressure. The
observation of the ratio of one solvent molecule per 6 equiv of
lithium aryl in the case of both benzene and toluene supports
the formualtion of 2 as a hexamer.
The syntheses of (LiC6H4-n-Bu)n (3) and (LiC6H4-4-SiMe2-
(t-Bu2))n (4) are best conducted in n-pentane. The off-white
solid crude productssa mixture of either 3 or 4 and LiBr13swere
collected on a sintered glass frit, washed once with n-pentane,
and dried under reduced pressure. Dissolving 3 in refluxing
toluene gives a clear, red solution. Interestingly, the LiBr
coproduct appears to be partly soluble under these conditions
as well, and further recrystallization of 3 from toluene failed to
remove all the LiBr. This suggests that it could be associated
with the ArLi framework. However, it is notable that, at present,
organolithium-lithium halide complexes have only been iso-
lated as donor stabilized species as in (LiPh‚Et2O)3‚LiBr,7 {Li-
X-ray Crystallography. Crystals suitable for X-ray crystallographic
studies were grown from benzene (1a) or toluene (1b) solutions, and
separate data sets were collected on each. Both crystal structures
displayed disorder in four of the six crystallographically independent
t-Bu groups (see Supporting Information). The crystals grown from
benzene (i.e., 1a) gave a superior data set and these data were used to
calculate the structural parameters. Crystal data at 130 K are as follows.
1a: C84H126Li6, M ) 1177.49, triclinic, space group P1h, a ) 13.516-
(2) Å, b ) 15.124(3) Å, c ) 20.958(3) Å, R ) 84.084(13)°, â ) 86.249-
(11)°, γ ) 68.675(13)°, V ) 3967.7(11) Å3, Z ) 2, DC ) 0.986 g
cm-3, λ (Cu KR) ) 1.541 78 Å, F(000) ) 1296, µ ) 0.387 mm-1
,
7983 reflections with I > 2σ(I), R1 ) 0.085 and wR2 ) 0.202. 1b:
C84H126Li6, M ) 1177.48, triclinic, space group P1h, a ) 12.954(3) Å,
b ) 13.487(3) Å, c ) 13.505(4) Å, R ) 104.18(2)°, â ) 104.51(2)°,
γ ) 111.19(2)°, V ) 1977.0(9) Å3, Z ) 1, DC ) 0.989 g cm-3, λ (Cu
KR) ) 1.54178 Å, F(000) ) 648, µ ) 0.388 mm-1, 3938 reflections
with I > 2σ(I).
The structures were solved9 by using direct methods. Absorption
corrections (XABS2) were applied to all data.10 Refinement was by
full matrix least-squares methods based on F2 with anisotropic thermal
parameters for non-disordered non-hydrogen atoms. The occupancies
of each disordered methyl carbon site are provided in the Supporting
Information.
Results and Discussion
Synthesis. Initial experiments to test the Schlosser method11
(illustrated by the generalized equation (1)) for the synthesis of
compounds 1-4 showed that metal halogen exchange did indeed
take place.
(12) Seebach, D.; Neumann, H. Chem. Ber. 1974, 107, 847.
(13) Analyses of crude samples of compounds 3 and 4 showed that they
were contaminated with ca. 20-30% LiBr. The carbon and hydrogen
analyses are consistent with the remainder of the sample being the lithium
aryl. Recrystallization reduced, but did not eliminate, the amount of included
LiBr.
(9) SHELXTL-Plus, A Program for Crystal Structure Determination,
Version 5.03, 1994, Siemens Analytical X-Ray Instruments: Madison, WI.
(10) Parkin, S. R.; Moezzi, B.; Hope, H. XABS2: an empirical absorption
correction program. J. Appl. Crystallogr. 1995, 28, 53.
(11) Schlosser, M.; Ladenberger, V. J. Organomet. Chem. 1967, 8, 193.

2850 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Wehmschulte and Power
Figure 3. View of the aryl group interaction with one of the capped
Li3 triangles in 1. No ring substituent atoms are shown.
Figure 2. Thermal ellipsoid plot (30%) of one of the two independent
molecules of 1. Methyl groups and hydrogen atoms are not shown for
clarity.
(c-CH)(CH2)2}2‚(LiBr)2‚Et2O)4,14 or the intramolecularly coor-
dinated LiC6H3{CH2N(Me)CH2CH2NMe2}2-2,6‚LiBr.15 Cooling
to room temperature affords feathery microcrystals, which with
excess Me3SiCl in pentane give 4-n-BuC6H4SiMe3 in 77% yield.
The aryllithium compound 4, however, can be extracted from
this mixture with hot toluene to give, after filtration and
crystallization at -20 °C, dumbell-shaped microcrystalline
aggregates. So far, only crystals of 1 have proven suitable for
X-ray crystallography. However, the availability of these
enabled the first structural characterization of a simple, base-
free, sterically unencumbered σ-bonded aryllithium compound.
Structure. (LiC6H3-3,5-t-Bu2)6 crystallizes from benzene in
the triclinic space group P1h with two crystallographically
independent, trimeric half-molecules in the asymmetric unit.
These half-molecules are chemically equivalent, but one (i.e.
incorporating Li(4), etc.), displays a slightly more irregular
geometry than the other. The other half of each molecule is
generated by a crystallographically required inversion center.
There are no close contacts between the resultant hexamers,
and the shortest intermolecular Li-C distance is 5.269 Å. Each
hexameric molecule (Figure 2) involves a trigonal, antiprismatic
array of lithiums in which there are six short (range 2.491(9)-
2.545(9) Å, avg 2.52(2) Å) and six long (range 3.050(8)-3.270-
(9), avg 3.14(7) Å) Li-Li distances. The six aryl substituents
triply bridge six of the eight triangular Li3 faces and are bound
principally through their ipso-carbon atoms. Two of the
equilateral and opposite faces which involve the six longest Li-
Li edges remain uncapped. The aryl groups do not bridge the
Li3 faces symmetrically since there is one short (range 2.119-
(7)-2.151(7) Å, avg 2.14(1) Å) and two longer (range 2.195-
(7)-2.247(7) Å, avg 2.22(1) Å) sets of Li-C(ipso) distances.
Thus, each Li+ ion has a stronger interaction with a specific
ipso-carbon, i.e. Li(1) with C(1), Li(2) with C(15), Li(3a) with
C(29), etc. Furthermore, the short Li-C interactions have a
greater σ-character since they are more closely aligned to the
aryl ring plane, as evidenced by the relatively low angles (16.9-
24.2°) between the Li-C and C(ipso)-C(para) lines (Figure
3). The remaining two long Li-C bonds at each lithium have
more π-like character with angles between the Li-C and
C(ipso)-C(para) vectors in the range 52.6-77.4°. This bond-
ing configuration, which is very similar in the case of each
lithium, also allows Li-C(ortho) interactions that are in the
Figure 4. Schematic drawing summarizing the important average
interatomic distances (Å) and angles in 1.
range 2.36-2.47 Å. These weaker Li-C(ortho) interactions
are also facilitated by the alignment of the phenyl ring above
each Li3 face. Similarly, relatively close Li-H contacts (in the
range 2.06-2.52 Å) involving ortho-hydrogens of the aryl group
may be calculated. The aromatic rings display the usual
structural deviations associated with substitution of a hydrogen
with an electropositive metal. The C-C-C angle at the ipso-
carbon atoms, which formally carry the negative charge, is
narrowed to 113.9° (avg) and this is accompanied by a slight
lengthening of the C(ipso)-C(ortho) distances to 1.404 Å
(avg). Important structural parameters for 1 are summarized
in Figure 4.
The structure of 1 may be compared to those of the alkyl
hexamers {Li(c-C6H11)}6‚2C6H6,16 {LiCH2(c-CHCMe2CMe2)}6,17
{LiCH2SiMe3}6,18 {Li(i-Pr)}6,19 or {Li(n-Bu)}6.20 These com-
pounds also possess distorted octahedral or trigonal prismatic
structures with uncapped transoid Li3 faces, except in the case
of {Li(c-C6H11)}6‚2C6H6, where there are weak interactions with
benzene. The six long Li-Li interactions in these structures
(16) Zerger, R.; Rhine, W.; Stucky, G. J. Am. Chem. Soc. 1974, 96, 6048.
(17) Maercker, A.; Bsata, M.; Buchmeier, W.; Engelen, B. Chem. Ber.
1984, 117, 2547.
(18) Tecle´, B.; Maqsudur Rahman, A. F. M.; Oliver, J. P. J. Organomet.
Chem. 1986, 317, 267.
(19) Siemeling, U.; Redecker, T.; Neumann, B.; Stammler, H.-G. J. Am.
Chem. Soc. 1994, 116, 5507.
(20) Kottke, T.; Stalke, D. Angew. Chem. Int. Ed. Engl. 1993, 32, 580.
(14) Schmidbauer, H.; Schier, A.; Schubert, U. Chem. Ber. 1983, 116,
1938.
(15) Wehman-Ooyevaar, I. C. M.; Kapteijn, G. M.; Grove, D. M.; Spek,
A. L., van Koten, G. J. Organomet. Chem. 1993, 452, C1-C-3.

σ-Bonded Lithium Aryls
J. Am. Chem. Soc., Vol. 119, No. 12, 1997 2851
Figure 5. 13C{1H} NMR spectra of 1 and 1 + 1, 2, 3, 5, and 10 equiv of THF in C6D6 at room temperature, which illustrate the conversion of
(THF‚LiC6H3-3,5-t-Bu)4 (bottom) to {(THF)2LiC6H3-3,5-t-Bu}2 (top).
are also associated with these uncapped Li3 faces. The shorter
and longer Li-Li distances in 1 may be compared with those
observed in the hexameric alkyls. For instance, the short and
long Li-Li interactions in {Li(i-Pr)}6 are in the ranges 2.388-
(3)-2.404(3) and 2.926(3)-3.014(3) Å, respectively, which are
marginally (ca. 0.1 Å) shorter than those observed in 1. The
Li-C distance in 1 (avg 2.14(1) Å) is among the shortest
observed in any associated base-free organolithium structure1
and may be compared with the range of ca. 2.18-2.20 Å
observed for the shorter Li-C bonds in the hexameric lithium
alkyls mentioned above.16-20 The longer group of Li-C
distances in 1 (avg 2.22(1) Å) is also shorter than those observed
in the hexameric alkyls by a similar amount.
In essence, the structural parameters of 1 show that the Li-C
bonding in 1 is significantly stronger than those in the
corresponding alkyls and makes for a tighter aggregate overall.
This is to be expected in view of the lower coordination number
(i.e. five, cf. six in the alkyls) of the lithium bound ipso-carbon.
In addition, the approximate sp2 hybridization of the carbon
bonding orbital, together with the presence of π-electron density,
further enhance the lithium-carbon interactions.
already established by X-ray crystallography.7 The 13C NMR
data for these compounds in ether solution (with ca. 10% C6D6
reference) show that there is only a small (ca. 2 ppm) difference
in the shifts of ipso-carbons. Attempts to record the spectra of
(LiPh‚Et2O)3‚LiBr in pure C6D6 led to the precipitation of LiBr
and 13C NMR resonances identical to those of (LiPh‚Et2O)4. It
is therefore concluded that the inclusion of small amounts of
lithium halide in 3 and 4 seems to have only a minor influence
of their ipso-carbon chemical shifts. The addition of 1 equiv
of Et2O or THF per lithium to C6D6 solutions of 1 led to only
a slight change of the C(ipso) chemical shift to 171.1 or 173.4
ppm, respectively. Furthermore, these additions resulted in a
dramatic increase in the solubility of 1 in benzene. Additional
evidence of Et2O or THF complexation to the Li+ ions in 1
was provided by shifts in the δ values of the ether or THF
protons and carbons. It may be tentatively concluded that, when
1 is treated with 1 equiv of Et2O or THF per lithium, it forms
1:1 complexes which are tetrameric aggregates with structures
very similar to those established for (Et2O‚LiPh)4 and (Et2O‚
LiC6H4-4-Me)4 in ether.6
It has already been shown by a number of groups that ipso-
carbon chemical shifts of the dimeric or monomeric solvated
lithium aryls are significantly further downfield than the
corresponding tetramers and are in the range of 180-200 ppm.6
Presumably, the greater downfield shift in these dimers or
monomers is due to the lower coordination number (and lower
shielding) of the C(ipso) carbon, which has a coordination
number of four in the dimer and three in the monomer. The
monomers and dimers were invariably obtained by the addition
of a sufficient quantity of a Li+ complexing donor molecule to
break down the aggregation. In a similar manner, the addition
of excess THF to 1 in C6D6 results in a downfield shift of ipso-C
resonance to 183.6 ppm (Figure 5), which is consistent with
the formation of the dimer {(THF)2‚LiC6H3-3,5-t-Bu2}2 (cf.
188.2 ppm for C(ipso) of LiPh in THF) under similar conditions.6c
NMR Studies. 13C NMR spectroscopy of the lithium aryls
1-4 shows that they possess the usual strong downfield5b shift
in their ipso-carbon signals. The high temperatures of the
measurementssdue to the limited solubilitiessresult in a narrow
ipso-carbon signal, suggesting fast exchange of the Li+ ions
and thereby precluding the determination of Li-C coupling
constants. Their ipso-carbon chemical shifts are 1, 170.6 ppm
(C6D6, 23 °C); 2, 165.8 ppm (C7D8, 110 °C); 3, 166.2 ppm
(C6D6, 85 °C); and 4, 169.8 ppm (C6D6, 22 °C), which are
6f
relatively close to those of the solvates (Et2O‚LiC6H4-4-Me)4
(δ ) 170.2 (C(ipso))) and (Et2O‚LiPh)4 (δ ) 174.2 (C(ipso)))
that also have tetrameric structures in ether solution. In the
case of 3 and 4, it could be argued that if lithium halide were
included as part of their structures, the chemical shifts could
be affected. However, we have compared the 13C NMR spectra
of (LiPh‚Et2O)4 and (LiPh‚Et2O)3‚LiBr, whose structures were
2
The C(ipso) chemical shifts of (LiTrip)4 (δ ) 184.0
(C(ipso))) or (LiC6H3-2,6-Dipp2)3 (δ ) 186.0 (C(ipso))) in

2852 J. Am. Chem. Soc., Vol. 119, No. 12, 1997
Wehmschulte and Power
which Li+ is both η1-σ-bonded and η6-π-bonded to different
phenyl rings (structures A and C) also imply that the effective
coordination number of the C(ipso) is lowered (i.e. to three) in
these compounds. In contrast, the chemical shift for 1 is closer
to that of (LiC6H3-2,6-Mes2)2 structure B (δ ) 173.5 (C(ipso))).
In this case, although the ipso-carbons are four-coordinate, it
appears reasonable to believe that the chemical shift so produced
would be similar to that in 1, where the five-coordinate C(ipso)
has a weaker interaction with two Li+ ions.
for the classic tetrameric structures (LiMe)422 or (LiEt)4,23 which
represent an intermediate stage between a molecular and a
polymeric ionic structure. However, with groups that are large
enough to prevent intermolecular interactions but not large
enough to cause significant crowding (e.g. i-Pr, c-C6H11, CH2(c-
CHCMe2CMe2), n-Bu, CH2SiMe3, C6H3-3,5-t-Bu2), hexameric
structures of the lithium derivatives are observed, and this
aggregation number is currently the most commonly seen in
the solid state for unsolvated lithium alkyls or aryls.24
Does LiPh have a hexameric structure like 1 in the solid?
The similarities between the spectroscopy of 1, its ether or THF
adduct, and the corresponding uncomplexed and complexed
LiPh derivatives suggest that the answer is probably yes. The
observation of one-sixth solvent molecule per aryllithium unit
in the benzene and toluene solvates of 2 lends further credence
to this view. However, it has been proposed6f that the
insolubility of LiPh in hydrocarbons is due to intermolecular
interactions (cf. (LiMe)4 structure) between (LiPh)2 dimers that
stack continuously to give polymers with four-coordinate
lithium. Although the results in this paper do not preclude this
possibility, it is probable that the solubility of a putative (LiPh)6
hexamer in such solvents would also be extremely low.
Conclusions
The first structural determination of a base-free, σ-bonded
lithium aryl, which is not sterically encumbered by ortho-
substituents, shows that it is hexameric in the crystalline phase.
Moreover, the primary interaction between the lithiums and the
aryl rings involves the ipso-carbon atom. There are also
secondary interactions involving ortho-carbons, but there are
no η6-aryl ring-Li linkages similar to those seen in the
structures of (LiTrip)4 (A) or (LiC6H3-2,6-Dipp2)2 (C). The
structure is thus quite different from the other three structures
(i.e., A, B, and C) of unsolvated lithium aryls known at present.
In essence, it is more representative of what the structure of a
simple uncomplexed lithium aryl would resemble in the solid.
Solution 13C NMR for 1-4 strongly support a similarly
associated structure in solution in the absence of Lewis bases.
The addition of 1 equiv of Et2O or THF per lithium to C6D6
solutions of 1 result in the formation of complexes, which on
the basis of 13C NMR data probably have the tetrameric
formulae (Et2O‚LiC6H3-3,5-t-Bu2)4 and (THF‚LiC6H3-3,5-Bu2)4
with structures similar to that of (Et2O‚LiPh)4.
Acknowledgment. This research was supported by the U.S.
National Science Foundation and the donors of the Petroleum
Research Fund, administered by the American Chemical Society.
Supporting Information Available: Full tables of the data
collection parameters, atom coordinates, bond distances and
angles, hydrogen coordinates, and anisotropic thermal param-
eters together with notes on structure solution and refinement
(22 pages). See any current masthead page for ordering and
Internet access instructions.
The hexameric structure of 1 underlines the fact that the
aggregation number six (rather than the more widely assumed
four) is the most preferred one for unsolvated lithium alkyls or
aryls. With sterically crowding groups and where there are
intermolecular interactions, lower aggregation numbers (i.e. 2
JA962889I
(22) (a) Weiss, E.; Lucken, E. C. A. J. Organomet. Chem. 1964, 2, 197.
(b) Weiss, E.; Hencken, G. J. Organomet. Chem. 1970, 21, 265.
(23) (a) Dietrich, H. Acta Crystallogr. 1963, 16, 681. (b) Dietrich, H. J.
Organomet. Chem. 1981, 205, 291.
(24) Organolithium compounds with aggregation numbers higher than
six (e.g. Li(n-Pr)8 or Li(n-Pr)9) have been detected in solution at low
temperatures. Nonetheless, at room temperature the hexameric structure
predominates: Fraenkel, G.; Henrichs, M.; Hewittt, J. M.; Su, B. M.; Geckle,
M. J. J. Am. Chem. Soc. 1980, 102, 3345.
20
or 4) are generally found. Thus, [Li(t-Bu)]4 and {LiC-
21
(SiMe3)3}2 are tetrameric and dimeric for steric reasons, and
relatively close intermolecular interactions are in part responsible
(21) Hiller, W.; Layh, M.; Uhl, W. Angew. Chem. Int. Ed. Engl. 1991,
30, 324.