Preparation of lithium stannide mixtures in organic solvents. A alternate source of lithium in organolithium chemistry

Article abstract of DOI:10.1016/j.jorganchem.2004.04.037

Lithium stannides were prepared from lithium naphthalenide and tin (II) chloride or tin (0) powder in THF solvent at room temperature under dry argon atmosphere. They were characterized with elemental analysis, XRD, and solid ^6.7Li NMR. Stabilities and reactivities of lithium stannides prepared from different conditions were tested and showed they were stable for a limited time at low temperatures. Best reactivity was obtained when they were prepared from tin (II) chloride and an excess of lithium naphthalenide. The lithium stannide mixture can reductively cleave carbon-halogen bonds and yield pinacol coupling with aldehydes. Organolithium compounds prepared from lithium stannides and organic halides add to ketones or aldehydes under Barbier conditions.

Full text of DOI:10.1016/j.jorganchem.2004.04.037

Journal of Organometallic Chemistry 689 (2004) 3421–3425
www.elsevier.com/locate/jorganchem
Preparation of lithium stannide mixtures in organic solvents.
A alternate source of lithium in organolithium chemistry
1
Reuben D. Rieke , Jun-sik Lee , Young-Sik Kye , Gerard S. Harbison
2
*
Department of Chemistry, University of Nebraska, Lincoln, Nebraska, NE 68588-0304, USA
Received 22 January 2004; accepted 27 April 2004
Available online 11 September 2004
Abstract
Lithium stannides were prepared from lithium naphthalenide and tin (II) chloride or tin (0) powder in THF solvent at room tem-
6
,7
perature under dry argon atmosphere. They were characterized with elemental analysis, XRD, and solid Li NMR. Stabilities and
reactivities of lithium stannides prepared from different conditions were tested and showed they were stable for a limited time at low
temperatures. Best reactivity was obtained when they were prepared from tin (II) chloride and an excess of lithium naphthalenide.
The lithium stannide mixture can reductively cleave carbon–halogen bonds and yield pinacol coupling with aldehydes. Organolith-
ium compounds prepared from lithium stannides and organic halides add to ketones or aldehydes under Barbier conditions.
ꢀ
2004 Published by Elsevier B.V.
6,7
Keywords: Lithium stannide; Organolithium; Li NMR; Reductive cleavage of C–X bond
1
. Introduction
Recently Yus and Ramon [7] reported a new ap-
proach for the preparation of organolithium compounds
using naphthalene or 4,4 -di-tert-butylbiphenyl and lith-
0
The first organolithium compound was prepared by a
transmetallation reaction of diethylmercury with lithium
metal in 1917 by Schlenk and Holtz [1]. In the 1930s,
Gilman et al. [2], Wittig and Leo [3], and Ziegler and
Colonius [4] independently prepared organolithium
compounds by a metal–halogen exchange reaction of or-
ganic halides with lithium metal. Simple organolithium
compounds, mainly butyllithium and phenyllithium,
which were prepared from above mentioned methods
were used to prepare other organolithium compounds
by metallation (metal–hydrogen exchange) reaction with
acidic organic compounds [5] and halogen–lithium ex-
change reaction with organic halides [5,6].
ium metal. Reductive cleavage of carbon–oxygen, car-
bon–nitrogen, and carbon–sulfur bond and reductive
opening of saturated heterocycles produced organolith-
ium compounds.
Lithium stannides were electrochemically prepared
by Furuya et al. [8], producing several distinct species,
7
identified by deconvolving poorly resolved Li solid-
7
state NMR spectra; the Li resonances were conjecturally
assigned using Knight shifts and T values. We have
1
been able to produce a mixture of lithium stannides as
a black slurry by treating preformed lithium naphthale-
nide with tin (II) chloride or tin powder under stirring
for 10 min at 298 K in THF under argon.
*
Corresponding author.
E-mail address: rrieke@unl.edu (R. D. Rieke).
2. Results and discussion
1
Present address: Rieke Metals, Inc. 1001 Kingbird Road, Lincoln,
Several lithium stannides [9] were prepared by mixing
lithium and tin metal or tin alloy at high temperature in
the previous reports. But the purpose of the previous
NE 68521 USA.
2
Present address: Department of Chemistry, Korea Military
Academy, Seoul, Korea.
0
022-328X/$ - see front matter ꢀ 2004 Published by Elsevier B.V.
doi:10.1016/j.jorganchem.2004.04.037

3422
R. D. Rieke et al. / Journal of Organometallic Chemistry 689 (2004) 3421–3425
reports was analysis of the structure with XRD and no
application in organic chemistry was performed with
lithium stannides.
resolution was markedly improved. Fig. 1(a) shows the
7
Li MAS NMR spectrum of a sample prepared with
3.3 mol Li per mol SnCl , over a 24 h reaction time
(prep. 8 in Table 1). In this spectrum, several distinct,
2
At first, active tin was prepared by reacting 2 equiva-
lents of preformed lithium naphthalenide with tin chlo-
ride and then reacted with n-octyl bromide. But no
reaction was detected. Then, excess (3.3 equivalents)
amount of preformed lithium naphthalenide was reacted
with tin chloride and the resulting black slurry in THF
was reacted with n-octyl bromide and produced several
products including elimination, reduction, and homo-
coupling product.
7
Knight-shifted Li resonances can be observed, flanked
by the usual sidebands at integer multiples of the rota-
tion frequency. However, the spectral resolution is still
comparatively poor, and the relative contributions of
the central 1/2 to ꢀ1/2 and the satellite transitions to
the isotropic peaks cannot accurately be determined.
Overlap of sidebands with isotropic peaks also compli-
6
cates the picture. We therefore turned to Li NMR, with
remarkable results. With the availability of higher mag-
Lithium stannide mixtures were characterized by ele-
6
7
6
netic fields, the less receptive Li nucleus has become an
mental analysis and solid-state Li and Li magic angle
spinning (MAS) NMR. Elemental analysis data showed
a Li:Sn ratio for the slurry varying from 0.8 to 1.8, in-
creasing as the initial lithium was increased. The elemen-
tal composition of the lithium stannide mixture did not
appear to depend on the tin source (tin (II) chloride or
commercial tin (0) powder). When lithium was added
as single large piece with a catalytic amount of naphtha-
lene, the preparation time took 3 h, but not all of the in-
itial lithium was converted into lithium stannide.
XRD measurement showed only tin (0) and did not
detected lithium stannides. The possible explanation
might be the decomposition of lithium stannides in the
air during the sample preparation.
7
attractive alternative to the more abundant Li spin,
whose spectroscopy is complicated by a larger quadru-
6
pole constant. Li has been used to study the state of
the lithium cation in minerals [10], battery materials
6
[11–13], ceramics [14], and glasses [15]. Li, while it is
a spin 1 nucleus, has negligible quadrupole couplings;
6
7
while Li NMR is less sensitive than Li NMR, the ab-
sence of substantial homonuclear dipolar and quadru-
polar interactions, makes the resolution considerably
better (Fig. 1(b)).
This enhanced resolution allows us to determine that
the resonance reported [8] at ꢁ38 ppm and assigned to
LiSn, is in fact two peaks in a 2:1 ratio, at 42 and 31
ppm. Since these peaks seem to vary in synchrony, it
is possible they arise from a single lithium stannide;
and while the evidence this species is LiSn is by no
means conclusive, there certainly exist other compounds
of formula AX with two chemically distinct A and two
distinct X atoms, each in a 2:1 ratio.
7
Li solid-state NMR of the lithium stannides mixture
using static samples at a field of 7 T, showed poorly re-
solved spectra that qualitatively resembled those ob-
tained by Furuya et al. [8]. However, under MAS,
6
Fig. 1(c) and (d) show the Li MAS spectra of the
products of prep. 4 and prep. 11, respectively; these
spectra allow us to identify the resonances tentatively as-
signed to LiSn at 76 ppm; Li Sn at 15 ppm, and Li Sn
2
2
5
2
7
at 3 ppm. Materials with lower Li content could not be
spun in the magnetic field due to their conductivity, but
7
static Li NMR showed them to contain a high fraction
of the putative LiSn . Irrespective of assignment, the
2
well-resolved lithium NMR signals surely indicate
well-defined, crystalline lithium stannides.
To determine the reduction potential of the lithium
stannide mixtures, it was reacted with several molecules
of varying reduction potential to yield an approximate
reduction potential. Benzophenone (ꢀ1.72 eV, DMF),
anthracene (ꢀ1.92 eV, DMF), stilbene (ꢀ2.14 eV,
DMF), and 2-methylpyrazine (ꢀ2.23 eV, DMF) showed
the characteristic color change when they were reacted
with lithium stannide. The redox potential should be
lower than naphthalene (ꢀ2.54 eV in DMF) because
lithium stannide was prepared from lithium naphthale-
nide [16]. From the above data, the redox potential of
the lithium stannide mixture can be approximated to
be in the range from ꢀ2.23 to ꢁꢀ2.54 eV.
7
Fig. 1. (a) Li MAS NMR spectrum of lithium stannides prepared
6
. (b) Li MAS NMR spectrum of lithium
with 3.3 mol Li per mol SnCl
2
6
stannides prepared with 3.3 mol Li per mol SnCl
2
. (c) Li MAS NMR
spectrum of lithium stannides prepared with 4.5 mol Li per mol SnCl
2
.
6
d) Li MAS NMR spectrum of lithium stannides prepared with 2.5
(
mol Li per mol Sn.

R. D. Rieke et al. / Journal of Organometallic Chemistry 689 (2004) 3421–3425
3423
Table 1
Reactivity of lithium stannide prepared from different conditions
a
b
Li
b
Np
c
Entry
Tin source
Method
Time
Yield (%)
1
2
3
4
5
6
7
8
9
SnCl
SnCl
SnCl
SnCl
SnCl
SnCl
SnCl
SnCl
2
2
2
2
2
2
2
2
A
A
B
B
B
B
B
B
A
B
B
B
3.49
3.40
3.36
4.51
4.15
2.98
3.03
3.35
1.07
1.20
2.50
2.15
0.12
3.81
0.12
4.97
4.51
3.30
3.29
3.82
0.13
1.33
2.77
2.34
8 h
1.5 h
88
99
99
10 min
10 min
10 min
10 min
10 min
24 h
d
57
81
e
75
99
f
35
30
37
Sn (0)
Sn (0)
Sn (0)
Sn (0)
12 h
1
1
1
0
1
2
10 min
10 min
10 min
d
55
73
e
a
Method A: Tin compound, lithium, and naphthalene were reacted at the same time. Method B; Tin compound was reacted with lithium
naphthalenide which was prepared right before it was used.
b
Equivalent amount of lithium and naphthalene compare to 1 equivalent of tin compound.
The yields were obtained by GLC using internal standard (n-decane) and anisole.
c
d
1
.25 equivalent of 2-bromoanisole was used.
equivalent of 2-bromoanisole was used.
e
f
1
0.35 equivalent of 2-bromoanisole was used.
The stability of the lithium stannide mixtures was
stannides prepared from 4.5 or 5.0 equivalents of lith-
ium naphthalenides gave better results for the chloride
substrates. For the cyanide-containing aryl bromide,
tested by reacting 2-bromoanisole with 2 equivalents
of lithium stannide which had been stored under various
conditions. Lithium stannide mixtures in THF can be
stored 1 week in a freezer (ꢀ13 ꢁC) without change of
chemical reactivity, but the reactivity is reduced by
THF/HNEt (80/20) solvent mixture was used to pre-
2
vent the attack of the generated organolithium on the
cyanide group. Amino- and hydroxy-containing aryl
halides were also successfully reduced. A large excess
of lithium stannides (6 equivalents) was required for
phenolic halides.
1
8% on 4 weeks storage. Storage for the same duration
in ether or hexane resulted in even further reduced reac-
tivity (57% and 38%). When mixtures were stored free of
solvent, all reactivity was lost within 2 weeks.
The reactivity of the lithium stannide mixtures was
determined using the reductive cleavage of 2-bromoani-
sole to yield anisole. The results are found in Table 1.
The organolithium compounds prepared from lithium
stannides and aryl halides failed to add to ketones or
aldehydes in normal conditions, but the addition reac-
tions were successful under Barbier conditions. The
preliminary addition reactions of organolithium with
aldehydes and ketones are presented in Scheme 1.
In conclusion, this study has yielded a fast and con-
venient method of preparing lithium stannides of excep-
tional reactivity. The lithium stannides react with
aldehydes to yield pinacols, and can readily reductively
cleave carbon–halogen bonds. Preliminary experiments
showed organolithium compounds prepared from
lithium stannide successfully add to aldehydes and
ketones under Barbier conditions.
The most reactive alloy was obtained from SnCl and
2
3
–3.5 equivalents of preformed lithium naphthalenide.
In all cases, lithium stannides produced from SnCl2
yielded better results than the lithium stannides generated
from tin (0) powders.
The lithium stannide mixture successfully cleaves
carbon–halogen bonds at room temperature within
10 min with good to excellent yields (Table 2). Most
bromide substrates gave quantitative yields. Lithium
Table 2
Reductive cleavage of aryl halides
a
Entry
ArX
SnCl
2
:Li:ArX
Time
Yield (%)
3. Experimental
1
2
3
4
5
6
7
2-OMePhBr
3-OMePhCl
3-OMePhCl
2-CNPhBr
1.00:3.44:0.45
1.00:3.41:0.46
1.00:4.54:0.44
1.00:3.39:0.30
1.00:3.50:0.49
1.00:5.05:0.46
1.00:3.50:0.16
10 min
3 h
99
55
72
3
.1. Preparation of lithium stannides
10 min
10 min
10 min
1 h
b
69
Lithium (10.2 mmol) and naphthalene (11.2 mmol)
2-NH
2-NH
2
PhBr
PhCl
2-OHPhBr
99
88
99
were weighed into a 50 ml centrifuge tube. Tin (II) chlo-
ride (3 mmol) was weighed into a 50 ml two-neck round
flask that was equipped with argon gas/vacuum two-way
manifold. All the weighing procedures were done inside
of the dry box. 20 ml of dry THF was added to the
2
7 h
a
The yields were obtained by GLC using internal standard
n-decane or n-nonane) and reduced compounds.
(
b
2
THF–HNEt (80–20%) mixture solvent was used.

3424
R. D. Rieke et al. / Journal of Organometallic Chemistry 689 (2004) 3421–3425
Br
+
fine powder sample was transferred into the 8 mm diam-
R1COR2
eter, 25 mm height dimension of small glass tube. A rub-
ber septum was placed on the glass tube and analyzed
with Bruker 600 MHz solid state NMR.
OH
MeO
Li Sn
MeO
R2
x
THF, RT, 10 min
R1
3
.5. Stability study
R1 = Me
2
R = PhCH2CH2
1a
1b
77%
57%
Lithium stannides were prepared as mentioned be-
1
R2 = Ph
R = H
fore and kept in the freezer (ꢀ13 ꢁC) for a specific time
in different solvent. Then they were warmed to room
temperature and reacted with 2-bromoanisole for
10 min 1 equivalent of 2-bromoanisole was used for 2
equivalents of lithium stannide. The reaction yields were
obtained by GLC using internal standard (n-decane)
and anisole.
Scheme 1.
centrifuge tube and 10 ml of dry THF was added to the
two-neck round flask and both tube and flask were stir-
red for 1 h at room temperature under dry argon gas.
Tin (II) chloride solution was cannulated into the centri-
fuge tube which containing lithium naphthalenide solu-
tion. The mixture solution was stirred for 10 min at
room temperature and then centrifuged for 10 min at
3.6. Reactivity study
2
000 rpm to separate the lithium stannide from the sol-
Lithium stannides were prepared from specific condi-
tions of different time, tin source, stoichiometry, etc. The
lithium stannides prepared from each condition reacted
with 2-bromoanisole with a ratio of 1–2 of 2-bromoan-
isole and lithium stannide except few experiments spec-
ified. The reaction yields were obtained by GLC using
internal standard (n-decane) and anisole.
vent, naphthalene, and LiCl. The upper solution was re-
moved through the cannula and 10 ml of dry THF was
added to the residue. The mixture was stirred for 10 min
and then centrifuged again for 10 min at 2000 rpm to
wash one more time. The upper solution was removed
via cannula and 20 ml of dry THF was added
3
.2. XRD analysis
3.7. Reductive cleavage of aryl halides
Lithium stannides were prepared as mentioned above
Lithium stannides (3 mmol) were prepared as men-
tioned before. Neat aryl halides (1.35 mmol) and inter-
nal standard (n-decane or n-nonane) was added to the
lithium stannides via syringe and the reaction mixture
was stirred for 10 min at room temperature. The hydro-
lyzed aliquot was extracted with ether and dried over
and vacuum dried for 24 h at room temperature under
dry argon atmosphere. Then they were brought into
the dry box and crashed to a fine powder with mortar
and pestle. The fine powder sample was transferred into
the one-side sealed capillary tube under argon gas and
the other side was sealed with para film. The para film
treated side was sealed over the flame. But the signal
from the sample in a capillary tube was too small. The
actual XRD analysis was done after the sample was
treated with silicon grease under atmosphere condition.
MgSO and injected into the GLC. The reduction yield
4
was obtained from the calibration curve.
3.8. Addition of organolithium compounds to the alde-
hydes and ketones
3
.3. Elemental analysis
Lithium stannides (3 mmol) were prepared as men-
tioned before. A mixture of aryl halide (1.35 mmol)
and aldehyde or ketone (1.05 mmol) in THF (5 ml)
was added to the lithium stannides via cannula and
the reaction mixture was stirred for 10 min at room tem-
perature. The reaction mixture was quenched with satu-
rated aqueous solution of ammonium chloride and
organic layer was extracted with ether (2·10 ml) and
Lithium stannides were prepared as mentioned above
and vacuum dried for 24 h at room temperature under
dry argon atmosphere and then sent to analyze with
atomic absorption spectroscopy.
6
.4. Solid Li NMR analysis
,7
3
then dried over MgSO . The organic solvent was evapo-
4
1
.0 M aqueous solution of lithium chloride was used
rated and the product was separated by the flash column
chromatography (ethyl acetate/hexane).
6
,7
as a reference in Li NMR experiment and the chemi-
cal shift of reference was set to 0 ppm. Lithium stannides
were prepared as mentioned above and vacuum dried
for 24 h at room temperature under dry argon atmos-
phere. Then they were brought into the dry box and
crashed to a fine powder with mortar and pestle. The
a-(4-Methoxyphenyl)-a-methylbenzenepropanol (1a):
1
H NMR (CDCl , 300 MHz) d 1.52 (s, 3H), 2.04 (t,
3
J=8.58 Hz, 2H), 2.36–2.62 (m, 3H), 3.68 (s, 3H), 6.83
(app d, J=8.59 Hz, 2H), 7.05–7.11 (m, 3H), 7.16–7.21
1
3
(m, 2H), 7.34 (app d, J=8.59 Hz, 2H); C NMR

R. D. Rieke et al. / Journal of Organometallic Chemistry 689 (2004) 3421–3425
[6] H. Gilman, A.L. Jacoby, J. Org. Chem. 3 (1938) 108.
3425
(
1
CDCl , 75 MHz) d 30.72, 31.03, 46.55, 55.56, 74.70,
3
[
7] (a) M. Yus, Chem. Soc. Rev. (1996) 155;
(b) D.J. Ramon, M. Yus, Eur. J. Org. Chem. (2000) 225.
13.96, 126.14, 126.54, 128.79, 140.34, 142.87, 158.62.
1
4
-Methoxy-a-phenylbenzenemethanol (1b): H NMR
[8] K. Furuya, K. Ogawa, Y. Mineo, A. Matsufuji, J. Okuda, T.
Erata, J. Phys. Condens. Matter. 13 (2001) 3519.
(
1
(
1
CDCl , 300 MHz) d2.74 (s, 1H), 3.70 (s, 3H), 5.66 (s,
3
1
3
H), 6.77–6.82 (m, 2H), 7.17–7.32 (m, 7H); C NMR
[9] (a) a-LiSn: W. Muller, H. Schafer, Z. Naturforsch. B 28 (1973)
246;
CDCl , 75 MHz) d 55.78, 76.22, 114.39, 127.00,
3
(b) b-LiSn:W. Blase, G. Cordier, Z. Kristallogr. 193 (1990) 317;
(c) Li Sn :D.A. Hansen, L.J. Chang, Acta Crystallogr., Sect. B 25
(1969) 2392;
(d) Li Sn :U. Frank, W. Muller, H. Schafer, Z. Naturforsch. B
27.91, 128.49, 128.96, 136.82, 144.66, 159.50.
2
5
5
2
Acknowledgement
30 (1975) 1;
(
(
(
(
e) Li
1975) 6;
f) Li Sn
g) Li13Sn
16.
7
Sn
2
:U. Frank, W. Muller, H. Schafer, Z. Naturforsch. B 30
This work was supported by a research grant from
the National Science Foundation.
7
3
:W. Muller, Z. Naturforsch. B 29 (1974) 304;
:U. Frank, W. Muller, Z. Naturforsch. B 30 (1975)
5
3
[10] L.A. Groat, B.C. Chakoumakos, D.H. Brouwer, C.M. Hoffman,
C.A. Fyfe, H. Morell, A.J. Schultz, Am. Miner. 88 (1) (2003)
References
1
95.
[
[
1] W. Schlenk, J. Holtz, Ber. 50 (1917) 262.
2] (a) H. Gilman, E.A. Zoellner, W.M. Selby, J. Am. Chem. Soc. 54
[11] K.I. Gnanasekar, H.A. Cathrino, J.C. Jiang, A.A. Mrse, G.
Nagasubrahmanian, D.H. Doughty, B. Rambabu, Solid State
Ionics 148 (3 and 4) (2002) 299.
(
(
(
1932) 1957;
b) H. Gilman, E.A. Zoellner, W.M. Selby, J. Am. Chem. Soc. 55
1933) 1252.
[12] Y.J. Lee, C. Eng, C.P. Grey, J. Electrochem. Soc. 148 (3) (2001)
A249.
[
[
[
3] G. Wittig, M. Leo, Ber. Deut. Chem. Ges. 64 (1931) 2395.
4] K. Ziegler, H. Colonius, Justus. Liebigs Ann. Chem. 479 (1930) 135.
5] (a) G. Wittig, U. Pockels, H. Dr o¨ ge, Ber. Dtsch. Chem. Ges. 71
[13] J.F. OÕGara, G. Nazri, D.M. MacArthur, Solid State Ionics 47
(1–2) (1991) 87.
[14] P. Kempgens, R.K. Harris, D.P. Thompson, Solid State NMR 15
(2) (1999) 109.
(
(
1938) 1903;
b) H. Gilman, R.L. Bebb, O. Baine, J. Am. Chem. Soc. 61 (1939)
[15] H. Eckert, Z. Zhang, J.H. Kennedy, Chem. Mater. 2 (3) (1990)
273.
109;
(
(
c) H. Gilman, F.W. Moore, O. Baine, J. Am. Chem. Soc. 63
1941) 2479.
[16] If the reduction potential of lithium stannide is higher than
naphthalene, it will not be prepared from lithium naphthalenide.