Serendipity strikes again - Efficient preparation of lithium tetramethylcuprate(III) via rapid injection NMR

Article abstract of DOI:10.1039/b924000d

Lithium tetramethylcuprate(iii), Me₄CuLi, the Cu^III analog of the Gilman reagent, has been prepared in high yield from halo-Gilman reagents Me₂CuLi·LiX (X = Cl, Br, I) and 2,3-dichloropropene and found to have surprising thermal stability. The cyano-Gilman reagent (X = CN) follows a different pathway.

Full text of DOI:10.1039/b924000d

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Serendipity strikes again—efficient preparation of lithium
tetramethylcuprate(III) via rapid injection NMRw
Erika R. Bartholomew,a Steven H. Bertz,*b Stephen K. Cope,a
Michael D. Murphy,a Craig A. Ogle*a and Andy A. Thomasa
Received (in College Park, MD, USA) 16th November 2009, Accepted 6th January 2010
First published as an Advance Article on the web 25th January 2010
DOI: 10.1039/b924000d
Lithium tetramethylcuprate(III), Me4CuLi, the CuIII analog of the
Gilman reagent, has been prepared in high yield from halo-
Gilman reagents Me2CuLiꢀLiX (X = Cl, Br, I) and 2,3-dichloro-
propene and found to have surprising thermal stability. The
cyano-Gilman reagent (X = CN) follows a different pathway.
In 1952, Gilman et al. prepared the first organocopper(I) ate
complex, Me2CuILi, by simply adding MeLi (2 equiv.) to CuI
salts such as CuI.1 Unfortunately, there is no CuIII salt
‘‘CuX3’’ that would allow Me4CuIIILi to be prepared by an
analogous displacement route. Nevertheless, by using rapid
injection NMR techniques,2–6 we can now report the efficient
preparation and definitive characterization of the correspond-
ing CuIII ate complex, lithium tetramethylcuprate(III) 1, from
halo-Gilman reagents Me2CuLiꢀLiX 2a–c (X = a, Cl; b, Br; c, I)
and 2,3-dichloropropene.
Fig. 1 Fully-coupled 13C NMR spectra of 13C-labeled 1, (13CH3)4CuLi:
By treating halo-Gilman reagents with allyl chlorides 3a or
3b (Scheme 1), we were able to prepare Z3 CuIII p-complexes
4a and 4b and Z1 CuIII s-complexes 5a and 5b, respectively.6
However, when we attempted to prepare the analogous chloro
derivatives 4c/5c (Z = Cl) from halo-Gilman reagents and
2,3-dichloropropene 3c, the reaction took a completely different
course, owing to b-chloride elimination, and the products were
lithium tetramethylcuprate(III) 1 and allene 6.
experimental (bottom) and simulated (top; see text for parameters).
observed a trace of 1 when we treated 2c with MeI in THF-d8
at ꢁ100 1C,4 and Gschwind et al. attributed a small side-
product peak to 1 in a study of lithium cyano(trimethyl)-
cuprate(III).7 In neither case could the product be characterized.
Complex 1 is characterized by a single peak at ꢁ0.22 ppm in its
1H NMR spectrum and a single peak at 14.76 ppm in its
13C NMR spectrum in THF-d8 at ꢁ100 1C. The structure of 1
was established by obtaining the fully-coupled 13C NMR spec-
trum of 13C-labelled 1, (13CH3)4CuLi, starting from 13CH3Li.
Fig. 1 contains stacked plots of the spectra: experimental (bottom)
and simulated (top, PERCH Solutions software).
The yields under these conditions were modest (ca. 30%),
but much improved over previous attempts. For example, we
The detailed match of the spectra in Fig. 1, including the
barely resolved long-range couplings, provides strong support
for our structural assignment. As in all square planar CuIII
complexes observed to date,3–7 the two-bond 13C–13C trans-
coupling across copper, 2Jtrans = 44.44 Hz, in 1 is significantly
2
larger in magnitude than the cis-coupling, Jcis = ꢁ2.44 Hz.
The values for the one-bond, three-bond and four-bond
3
1H–13C couplings are 1J = 122.37 Hz, Jtrans = ꢁ2.31 Hz,
3Jcis = 3.05 Hz, Jtrans = 0.82 Hz and Jcis = 0.00 Hz. This
method was used by Berger and co-workers to establish the
structures of dimethylcuprate(I)8 and tetramethylzincate(II).9
We have been able to optimize the yield of 1 by considering
the mechanism of its formation (Scheme 2). In our previous
studies of 3a and 3b,6 we did not observe Z2-intermediates
(e.g., 7a or 7b, respectively, not shown, cf. 7c and 7d). We
undertook the study of allylic substrates 3c and 3d containing
electron withdrawing groups in a further attempt to observe
4
4
Scheme 1 Dichotomy in reactions of halo-Gilman reagents 2a–c with
propenyl derivatives 3a/3b versus 3c/3d (cf. Scheme 2).
a Department of Chemistry, University of North Carolina–Charlotte,
Charlotte, NC 28223, USA. E-mail: cogle@uncc.edu;
Fax: +1-704-687-3151; Tel: +1-704-687-2524
b Complexity Study Center, Mendham, NJ 07945, USA.
E-mail: sbertz@complexitystudycenter.org
w Electronic supplementary information (ESI) available: DOSY plots
and calculations. See DOI: 10.1039/b924000d
ꢂc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 1253–1254 | 1253

View Article Online
Scheme 3 Allene formation from 3c and cyano-Gilman reagent 2d.
The reaction of 2d with 3c appears to be related to one
observed by Norinder et al.,11 in which treatment of R2CuLi
with perfluoroallyl iodide gave rise to R–R; however,
perfluoropropadiene was not reported.
Tetrakis(trifluoromethyl)cuprate(III)12 has been considered
a special ‘island’ of organocopper(III) stability. Inter alia we
have shown that fluorine atoms are not sine qua non for a
species such as 1 to have significant stability.
Scheme 2 Mechanism of formation of 1 from 2a–c and 3c or 3d.
Z2-intermediates, e.g., 7c and 7d, in analogy to those observed
in the conjugate addition pathway of a-enones.2 While we did
not observe these Z2-intermediates, either, we did serendipitously
discover a useful new route to 1.
Both lithium dimethylaurate(I) and lithium tetramethylaurate(III)
are known compounds,13 and now the same is true for
lithium dimethylcuprate(I) and lithium tetramethylcuprate(III).
More importantly, 1 is a useful starting material for other
organocopper(III) compounds.14
We did not observe electron-deficient Z3-complexes 4c or 4d,
and we include them on the basis of the presence of 4a and 4b in
our previous study.6 Apparently, they react rapidly with 2 to
give trimethyl derivatives 5c or 5d, respectively. We did not
observe 5c with 2a–c; however, by switching to a poorer leaving
group, viz. fluoride, we were able to observe intermediate 5d at
ꢁ100 1C. Whatever the exact nature of the intermediates might
be, a total of 3 equiv. of 2 are required for each equivalent of 1
produced, since MeCu/LiX is not reactive under our conditions.
We have found that the second and third equivalents of 2 can
be replaced by MeLi, so that the optimal reagent is a 2 : 1
mixture of MeLi and 2, prepared by adding 4 equiv. of MeLi to
CuX (X = Cl, Br, I). For example, injection of a solution of 3c
in THF-d8 into the reagent, CuClꢀSMe2 + 4MeLi,z in THF-d8
at ꢁ100 1C yielded 95% of 1 (based on CuCl) and 90% of 6
(based on 3c), measured by using the double internal standard
method (trimethoxybenzene in the NMR tube and mesitylene in
the injector syringe). The presence of a coordinating ligand such
as dimethylsulfide or trimethylphosphine, introduced with the
CuI precursor, gave a significantly cleaner reaction mixture.
Temperature control is crucial: the yields of 1 from the
optimal procedure were nearly quantitative at ꢁ100 1C; however,
running the reaction at ꢁ80 1C gave a diminished yield of 1
(ca. 70%). Preparation at 0 1C gave a low yield (o20%) of 1.
(The main side-product was ethane.) Nevertheless, when 1 was
prepared at ꢁ100 1C and then warmed to 0 1C, it was
surprisingly stable (half-life, t1/2 = 7 h). At 20 1C, 1 decomposed
to 2 and ethane in a first-order reaction, t1/2 = 0.75 h.
DOSY experiments10 with internal standardsw gave a
measured value of 287 amu, which is in good agreement
with the calculated molecular weight, 290.94 amu, for
Me4CuꢀLi(THF-d8)2. This complex appears to be a solvent
separated ion pair, as HOESY and NOESY experiments did
not detect any interactions between the Cu and Li moieties.
When we injected 3c into the cyano-Gilman reagent,
Me2CuLiꢀLiCN 2d, the resulting reaction efficiently yielded
ethane (80%) and allene (90%), but no 1. Intermediate 8
(Scheme 3) was observed at ꢁ100 1C; the cis-stereochemistry
of the methyl groups sets them up for facile reductive elimination.
Inclusion of 2 equiv. of MeLi (i.e., Me2CuLiꢀLiCN/2MeLi)
afforded 1 (15%) and 8 (38%) in addition to methane (38%)
and ethane (7%) after 1 h at ꢁ100 1C.
In conclusion, this is one of the rare instances where the
obverse and reverse of the copper coin are both interesting. On
the halo-Gilman side, we have developed a high yield prepara-
tion of lithium tetramethylcuprate(III), and on the cyano-
Gilman side we have discovered a new synthesis of allenes.
We thank the NSF for grants 0353061 and 0321056 and
Perch Solutions, Ltd. for NMR simulation software.
Notes and references
z A dry, unused NMR tube was charged with CuClꢀSMe2 (30 mmol).
The tube was evacuated and then filled with nitrogen and sealed with a
septum; 420 mL of freshly distilled THF-d8 was added to the tube
via gas-tight syringe. It was then cooled to ꢁ80 1C, and 120 mL of
1 M MeLiꢀTHF (120 mmol) in benzene-d6 was added. The tube was
sonicated for 0.1 h at 0 1C. The septum was removed and the tube was
immediately lowered into the NMR probe. After cooling to ꢁ100 1C,
3.0 mL of 2,3-dichloropropene (30 mmol) in 60 mL of THF-d8
was injected.
1 H. Gilman, R. G. Jones and L. A. Woods, J. Org. Chem., 1952, 17,
1630–1634.
2 S. H. Bertz, C. M. Carlin, D. A. Deadwyler, M. D. Murphy, C. A. Ogle
and P. H. Seagle, J. Am. Chem. Soc., 2002, 124, 13650–13651.
3 S. H. Bertz, S. Cope, M. Murphy, C. A. Ogle and B. J. Taylor,
J. Am. Chem. Soc., 2007, 129, 7208–7209.
4 S. H. Bertz, S. Cope, D. Dorton, M. Murphy and C. A. Ogle,
Angew. Chem., Int. Ed., 2007, 46, 7082–7085.
5 E. R. Bartholomew, S. H. Bertz, S. Cope, D. C. Dorton,
M. Murphy and C. A. Ogle, Chem. Commun., 2008, 1176–1177.
6 E. R. Bartholomew, S. H. Bertz, S. Cope, M. Murphy and
C. A. Ogle, J. Am. Chem. Soc., 2008, 130, 11244–11245.
7 T. Gartner, W. Henze and R. M. Gschwind, J. Am. Chem. Soc.,
¨
2007, 129, 11362–11363.
8 T. A. Mobley, F. Muller and S. Berger, J. Am. Chem. Soc., 1998,
120, 1333–1334.
¨
9 T. A. Mobley and S. Berger, Angew. Chem., Int. Ed., 1999, 38,
3070–3072.
10 D. Li, G. Kagan, R. Hopson and P. G. Williard, J. Am. Chem.
Soc., 2009, 131, 5627–5634.
11 J. Norinder, J.-E. Backvall, N. Yoshikai and E. Nakamura,
¨
Organometallics, 2006, 25, 2129–2132.
12 D. Naumann, T. Roy, K. F. Tebbe and W. Crump, Angew. Chem.,
Int. Ed., 1993, 32, 1482–1483.
13 G. W. Rice and R. S. Tobias, Inorg. Chem., 1975, 14, 2402–2407.
14 S. H. Bertz, M. D. Murphy, C. A. Ogle and A. A. Thomas, Chem.
Commun., 2010, 46, DOI: 10.1039/b921412g.
ꢂc
This journal is The Royal Society of Chemistry 2010
1254 | Chem. Commun., 2010, 46, 1253–1254