Theoretical studies on the addition of polymetallic lithium organocuprate clusters to acetylene. Cooperative effects of metals in a trap-and-bite reaction pathway

Article abstract of DOI:10.1021/ja964208p

Ab initio and density functional theoretical investigations into the nature of the reaction of acetylene with noncluster organocuprate reagents, MeCu, Me₂Cu^-, and lithium organocuprate cluster reagents, Me₂CuLi, Me₂-CuLi·LiCl, and (Me₂CuLi)₂ were carried out, and the function of the mixed metal cluster was probed. Intermediates transition structures (TSs) as well as the structures on the intrinsic reaction coordinate in the C-C bond forming stage of the reaction of lithium cuprate clusters have been calculated with the ab initio method (MP2) and density functional method (B3LYP) using all-electron basis sets for copper. The addition reaction of the simple organocopper reagent, MeCu, can be viewed as a simple four-centered addition reaction consisting of nucleophilic addition of an anionic methyl group, while the addition of the cuprate reagent, Me₂Cu^-, involves transfer of negative charge to the acetylene via the copper atom. In the cluster reaction of Me₂CuLi·LiC1, the lithium atom in the cluster stabilizes the developing negative charge on the acetylene moiety and assists the electron flow from the copper atom. Reductive elimination of the transient Cu(III) species initially gives a 1- propenyllithuim-like structure intermediate (nonstationary point), which then undergoes intramolecular transmetalation to give the final product. 1-propenylcopper. Essentially the same mechanism operates also with Me₂CuLi and (Me₂CuLi)₂, indicating that the Li-Me-Cu-Me moiety incorporated in the mixed organocuprate cluster is essential for the reaction. Experiments showed that the strong solvation of the lithium atom with a crown ether, which sequesters the lithium cation from the cluster, strongly decelerates the carbocupration reaction. Thus, theory and experiments revealed the cooperative function of lithium and copper atoms in the cuprate reactions.

Full text of DOI:10.1021/ja964208p

J. Am. Chem. Soc. 1997, 119, 4887-4899
4887
Theoretical Studies on the Addition of Polymetallic Lithium
Organocuprate Clusters to Acetylene. Cooperative Effects of
Metals in a Trap-and-Bite Reaction Pathway
Eiichi Nakamura,*,† Seiji Mori,† Masaharu Nakamura,† and Keiji Morokuma*,‡
Contribution from the Department of Chemistry, The UniVersity of Tokyo,
Bunkyo-ku, Tokyo 152, Japan, and Cherry L. Emerson Center for Scientific
Computation and Department of Chemistry, Emory UniVersity, Atlanta, Georgia 30322
ReceiVed December 6, 1996X
Abstract: Ab initio and density functional theoretical investigations into the nature of the reaction of acetylene with
noncluster organocuprate reagents, MeCu, Me2Cu-, and lithium organocuprate cluster reagents, Me2CuLi, Me2-
CuLi‚LiCl, and (Me2CuLi)2 were carried out, and the function of the mixed metal cluster was probed. Intermediates
transition structures (TSs) as well as the structures on the intrinsic reaction coordinate in the C-C bond forming
stage of the reaction of lithium cuprate clusters have been calculated with the ab initio method (MP2) and density
functional method (B3LYP) using all-electron basis sets for copper. The addition reaction of the simple organocopper
reagent, MeCu, can be viewed as a simple four-centered addition reaction consisting of nucleophilic addition of an
anionic methyl group, while the addition of the cuprate reagent, Me2Cu-, involves transfer of negative charge to the
acetylene via the copper atom. In the cluster reaction of Me2CuLi‚LiCl, the lithium atom in the cluster stabilizes the
developing negative charge on the acetylene moiety and assists the electron flow from the copper atom. Reductive
elimination of the transient Cu(III) species initially gives a 1-propenyllithium-like structure intermediate (nonstationary
point), which then undergoes intramolecular transmetalation to give the final product, 1-propenylcopper. Essentially
the same mechanism operates also with Me2CuLi and (Me2CuLi)2, indicating that the Li-Me-Cu-Me moiety
incorporated in the mixed organocuprate cluster is essential for the reaction. Experiments showed that the strong
solvation of the lithium atom with a crown ether, which sequesters the lithium cation from the cluster, strongly
decelerates the carbocupration reaction. Thus, theory and experiments revealed the cooperative function of lithium
and copper atoms in the cuprate reactions.
Introduction
judicious selection of the “countercation” M+. For instance,
upon complexation with crown ether, the R2CuLi reagent loses
its ability to undergo conjugate addition to enones4 and to add
to acetylenes (vide infra), while cuprates having a Lewis acidic
metal as M+ (M+ ) Zn(II),5 Ti(IV),6 Zr(IV),6 etc. as well as
neutral additives as such as BF3,7 Me3SiCl,8 etc.) often exhibit
better selectivities than the lithium cuprates. The exact role of
M+, however, remains to be understood.
The prototypical reagent R2CuLi exists predominantly as a
dimer in solution, as revealed by various physical measure-
ments.9 The dimer has been suggested to have a cyclic structure
A consisting of alternating Li and Cu bridged by pentacoordi-
“Of all the transition-metal organometallic reagents developed
for application to organic synthesis, organocopper complexes
are by far the most heavily used and enthusiastically accepted
by the synthetic organic chemist”.1,2 However, the nature of
the reactive species and their reaction pathway are not yet fully
understood. In the 1960s, House demonstrated that the reactive
species in the organocopper reactions are a species having the
“R2CuLi” stoichiometry.3 The simplest organocopper species,
RCu, is not reactive enough to be synthetically useful. After
many years of synthetic exploration, there exists now a variety
of cuprate reagents of the general formula [R1R2Cu]-M+, and
the reactivity of the cuprate reagents have been tuned by
(4) Ouannes, C.; Dressaire, G.; Langlois, Y. Tetrahedron Lett. 1977, 10,
815-818.
† The University of Tokyo.
‡ Emory University.
(5) Nakamura, E.; Sekiya, K.; Arai, M.; Aoki, S. J. Am. Chem. Soc.
1989, 111, 3091-3093. Arai, M.; Kawasuji, T.; Nakamura, E. J. Org. Chem.
1993, 58, 5121-5129.
X Abstract published in AdVance ACS Abstracts, May 1, 1997.
(1) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles
and Applications of Organotransition Metal Chemistry, 2nd ed.; University
Sciencebooks: Mill Valley, CA, 1987; Chapter 14.
(6) (a) Ti(IV): Arai, M.; Nakamura, E.; Lipshutz, B. H. J. Org. Chem.
1991, 56, 5489-5491. (b) Zr(IV): Lipshutz, B. H.; Ellsworth, E. L. J. Am.
Chem. Soc. 1990, 112, 7440-7441.
(2) (a) Reviews: Posner, G. H. Organic Reactions 1972, 19, 1-113.
Normant, J. F. Synthesis 1972, 63-80. Lipshutz, B. H. Sengupta, S. Organic
Reactions 1992, 41, 135-631. Organocopper Reagents; Taylor, R. J. K.,
Ed.; Oxford University Press: UK, 1994. van Koten, G.; Noltes, J. G.
ComprehensiVe Organometallic Chemistry; Wilkinson, G., Stone, F. G. A.,
Eds.; Pergamon: Oxford, 1982; Vol. 2, pp 709-763. van Koten, G. James,
S. L.; Jastrzebski, J. T. B. H. ComprehensiVe Organometallic Chemistry
II; Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.; Pergamon: Oxford,
1995; Vol. 3, pp 57-133. Lipshutz, B. H. ComprehensiVe Organometallic
Chemistry II; Abel, E. W., Stone, F. G. A., Wilkinson, G., Eds.;
Pergamon: Oxford, 1995; Vol. 12, pp 59-130. (b) For a unique chemistry
of organocuprates in fullerene chemistry, see: Sawamura, M.; Iikura, H.;
Nakamura, E. J. Am. Chem. Soc. 1996, 118, 12850-12851.
(3) House, H. O.; Respess, W. L.; Whitesides, G. M. J. Org. Chem. 1966,
31, 3128-3141. Cf.: Gilman, H.; Jones, R. G.; Woods, L. A. J. Org. Chem.
1952, 17, 1630-1634.
(7) Yamamoto, Y. Angew. Chem., Int. Ed. Engl. 1986, 25, 947-959.
Lipshutz, B. H.; Ellsworth, E. L.; Siahaan, T. J. J. Am. Chem. Soc. 1989,
111, 1351-1358.
(8) (a) Chuit, C.; Foulon, J. P.; Normant, J. F. Tetrahedron 1980, 36,
2305-2310. (b) Nakamura, E.; Kuwajima, I. J. Am. Chem. Soc. 1984, 106,
3368-3370. Horiguchi, Y.; Matsuzawa, S.; Nakamura, E.; Kuwajima, I.
Tetrahedron Lett. 1986, 27, 4025-4028. Nakamura, E.; Matsuzawa, S.;
Horiguchi, Y.; Kuwajima, I. Tetrahedron Lett. 1986, 27, 4029-4032. (c)
Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26, 6015-6018 and
6019-6022. (d) Alexakis, A.; Berlan, J.; Besace, Y. Tetrahedron Lett. 1986,
27, 1047-1050. (e) Johnson, C. R.; Marren, T. J. Tetrahedron Lett. 1987,
28, 27-30. (f) Nakamura, E. In Organocopper Reagents; Taylor, R. J. K.,
Ed.; Oxford University Press: UK, 1994; Chapter 6, pp 129-142. (g) Rate
acceleration of Me3SiI, see: Eriksson, M.; Johansson, A.; Nilsson, M.;
Olsson, T. J. Am. Chem. Soc. 1996, 118, 10904-10905.
S0002-7863(96)04208-4 CCC: $14.00 © 1997 American Chemical Society

4888 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Nakamura et al.
nated alkyl or aryl groups, which was identified in the crystals
A number of cuprate reagents of monomeric stoichiometry,
R2CuLi‚LiX (or R2Cu(X)Li2), are currently used in organic
synthesis. It is important to note that, the synthetic copper
reagents being prepared by the reaction of RLi with Cu(I)
halides, they are almost always the reagents of R2CuLi‚LiX
stoichiometry (X ) I,11,22 Br,23,24 and Cl25). The kinetic studies
on conjugate addition15 and the alkylation reactions9b,13,14
indicated that the dimeric R2CuLi remains to be a kinetically
active species even in the presence of LiI, though the reaction
rates were slightly affected by LiI.17a The small LiI effects have
been attributed to the high equilibrium concentration of a pair
of cyclic dimers, (Me2CuLi)2 + (LiI)2 in an ether solution of
the “Me2CuLi‚LiI” reagent.11 Very recent molecular weight
determination indicated that Me2CuLi‚LiI in THF exists in the
monomeric form.22
As summarized above, a dimeric cluster exists in solution
and takes part in the C-C bond forming reaction (e.g., conjugate
addition and alkylation reaction). However, there is little
understanding of the function of the cluster structure in cuprate
chemistry. In light of the importance of cuprate chemistry and
the paucity of information at the molecular level of understand-
ing of clusters, we have carried out quantum mechanical studies
on the dynamic behavior of cluster and noncluster organocu-
prates by ab initio molecular orbital and density functional (DF)
methods. We have studied the reactions of three prototypical
clusters, Me2CuLi, Me2CuLi‚LiCl, and (Me2CuLi)2 as well as
the simplest noncluster compounds, MeCu and Me2Cu-, in order
to address to the fundamental question: what is the role of
clusters in organocuprate chemistry.
In this and the accompanying articles,26 we will describe the
first systematic theoretical studies on the reaction pathways of
the polymetallic lithium organocuprate clusters reacting with
C-C multiple bonds. This article describes (1) 1,2-addition
of cuprates to acetylene along with a discussion on cuprate
models and computational methods and (2) the conjugate
addition of bis-, tri-, and tetrametallic cuprate clusters to acrolein.
While different in the details of the reaction course, the two
reactions were found to share in common important mechanistic
features, a “trap-and-bite” reaction pathway realized by coopera-
tion of the lithium and copper atoms. Though our studies were
mainly focused on the process involving the crucial C-C bond
forming process, and the reaction pathway prior to the C-C
bond formation was also made.
of several cuprates (e.g., R ) Ph, Me3SiCH2).10 NMR studies
1
7
on H, 13C, and Li nuclei carried out for reagents of the R2-
CuLi stoichiometry also supported the presence of the dimeric
aggregate in an ethereal solution.11,12
In the kinetic studies of the conjugate addition and SN2
alkylation of alkyl and aryl halides,9b,13,14 the reaction rates were
found to be first order to both the substrate and the dimer (R2-
CuLi)2. Thus, the dimer R2CuLi has been suggested to be a
kinetically reactive species. In the conjugate addition reaction,
a kinetically observable cuprate/enone complex forms first,
exists in equilibrium with the dimer and the substrate, and then
affords the final product.15 Recent NMR studies successfully
detected what are likely to be this kinetically observable
complexes16,17 and revealed that they are π-complexes between
a cuprate and an enone substrate bound together by donation/
backdonation interaction. The NMR spin coupling studies
showed that this π-complex assumes characters which are
consistent with the long discussed Cu(III) intermediate.18 The
recent isolation and elucidation of the crystal structures of stable
Cu(III) organometallics19,20 (e.g., [(CF3)4Cu]-) gave a strong
support to the possible intervention of a Cu(III) intermediates
in cuprate addition.21
(9) (a) van Koten, G.; Noltes, J. G. J. Chem. Soc., Chem. Commun. 1972,
940-941. (b) Pearson, R. G.; Gregory, C. D. J. Am. Chem. Soc. 1976, 98,
4098-4104.
(10) (a) van Koten, G.; Jastrzebski, J. T. B. H.; Muller, F.; Stam, C. H.
J. Am. Chem. Soc. 1985, 107, 697-698. (b) Lorenzen, N. P.; Weiss, E.
Angew. Chem., Int. Ed. Engl. 1990, 29, 300-302. (c) Olmstead, M. M.;
Power, P. P. Organometallics 1990, 9, 1720-1722. (d) Olmstead, M. M.;
Power, P. P. J. Am. Chem. Soc. 1990, 112, 8008-8014.
(11) Bertz, S. H.; Vellekoop, A. S.; Smith, R. A. J.; Snyder, J. P.
Organometallics 1995, 14, 1213-1220.
(12) (a) Ashby, E. C.; Watkins, J. J. J. Am. Chem. Soc. 1977, 99, 5312-
5317. (b) van Koten, G.; Noltes, J. G. J. Organomet. Chem. 1979, 174,
367-387. (c) Lipshutz, B. H.; Kozlowski, J. A.; Breneman, C. M. J. Am.
Chem. Soc. 1985, 107, 3197-3204.
(13) Johnson, C. R.; Dutra, G. A. J. Am. Chem. Soc. 1973, 95, 7783-
7787.
The description of the theoretical work will be preceded by
two sections dealing with the experimental background as well
as some new experimental results as to the role of the lithium
atom in the cuprate cluster. Comparison of theoretical methods
as applied to dynamic behavior of organocopper reagents is
(14) Spanenberg, W. J.; Snell, B. E.; Su, M.-C. Microchem. J. 1993, 47,
79-89.
(15) Krauss, S. R.; Smith, S. G. J. Am. Chem. Soc. 1981, 103, 141-
148.
(16) (a) Hallnemo, G.; Olsson, T.; Ullenius, C. J. Organomet. Chem.
1984, 265, C22-24. (b) Ullenius, C. Christenson, B. Pure Appl. Chem.
1988, 57-64.
(22) Gerold, A.; Jastrzebski, J. T. B. H.; Kronenburg, C. M. P.; Krause,
N.; van Koten, G. Angew. Chem., Int. Ed. Engl. 1997, 36, 755-757.
(23) (a) House, H. O.; Chu, C.-Y.; Wilkins, J. M.; Umen, M. J. J. Org.
Chem. 1975, 10, 1460-1469. (b) Westmijze, H.; Kleijn, H.; Vermeer, P.
Tetrahedron Lett. 1977, 23, 2023-2026. (c) Bertz, S. H.; Fairchild, E. H.
Encyclopedia of Reagents for Organic Synthesis; Paquette, L. A., Ed.; John
&Wiley: Chichester: 1995; Vol. 2, pp 1312-1315.
(17) (a) Bertz, S. H.; Smith, R. A. J. J. Am. Chem. Soc. 1989, 111, 8276-
8277. (b) Vellekoop, A. S.; Smith, R. A. J. J. Am. Chem. Soc. 1994, 116,
2902-2913.
(18) (a) Krause, N.; Wagner, R.; Gerold, A. J. Am. Chem. Soc. 1994,
116, 381-382. (b) Nilsson, K.; Ullenius, C.; Krause, N. J. Am. Chem. Soc.
1996, 118, 4194-4195.
(19) Willert-Porada, M. A.; Burton, D. J.; Baenziger, N. C. J. Chem.
Soc., Chem. Commun. 1989, 1633-1634. Neumann, D.; Roy, T.; Tebbe,
K.-F.; Crump, W. Angew. Chem., Int. Ed. Engl. 1993, 32, 1482-1483.
Eujen, R.; Hoge, B.; Brauer, D. J. J. Organomet. Chem. 1996, 519, 7-20.
(20) Kaupp, M.; von Schnering, H. G. Angew. Chem., Int. Ed. Engl.
1995, 34, 986.
(21) Appropriateness of the use of Cu(III) formality has been questioned
(Snyder, J. P. Angew. Chem., Int. Ed. Engl. 1995, 34, 80-81) on the basis
of the calculated electron density, which is far less positive (∼+1) than
the formal charge of +3, and this challenge has been discussed from the
more standard inorganic viewpoint (cf. ref 20 and Snyder, J. P. Angew.
Chem., Int. Ed. Engl. 1995, 34, 986-987). As the Snyder proposal would
also pose problems in the use of the fundamental terminology such as
“reductive elimination”, we use here the conventional formal oxidation
nomenclature.
(24) Bertz, S. H.; Dabbagh, G. J. Org. Chem. 1984, 49, 1119-1122.
Bertz, S. H.; J. Am. Chem. Soc. 1990, 112, 4031-4032. Bertz, S. H.;
Fairchild, E. H. Encyclopedia of Reagents for Organic Synthesis; Paquette,
L. A., Ed.; John & Wiley: Chichester, 1995; Vol. 2, pp 1346-1349.
(25) Bertz, S. H.; Fairchild, E. H. Encyclopedia of Reagents for Organic
Synthesis; Paquette, L. A. Ed.; John & Wiley: Chichester, 1995; Vol. 2,
pp 1324-1326. See, also: (a) stoichiometric reagents: Bertz, S. H.; Gibson,
C. P.; Dabbagh, G. Tetrahedron Lett. 1987, 28, 4251-4254. Lipshutz, B.
H.; Stevens, K. L.; James, B.; Pavlovich, J. G.; Snyder, J. P. J. Am. Chem.
Soc. 1996, 118, 6796-6797. (b) Catalytic reagents: Tamura, M.; Kochi, J.
Synthesis 1971, 303-305. Giner, J.-L.; Margot, C.; Djerassi, C. J. J. Org.
Chem. 1989, 54, 2117-2125. Schlosser, M.; Bossert, H. Tetrahedron 1991,
47, 6287-6292.
(26) Nakamura, E.; Mori, S.; Morokuma, K. J. Am. Chem. Soc. 1997,
119, 4900-4910.

Addition of Li Organocuprate Clusters to Acetylene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4889
Scheme 1. Schematic Representation of Two Conventional
Scheme 2
Mechanisms of Addition of R2Cu- to Acetylene
summarized in Appendix. This crucial information has been
lacking in the literature.
Acetylene Carbocupration. Background
Scheme 3
Among various reactions of cuprates, we have selected the
1,2-addition of a methyl cuprate reagent to acetylene as the
initial target of our studies (eq 1).27-29 The reaction is highly
characteristic of organocopper chemistry and has provided
uniquely effective entries to stereodefined vinyl metallic reagents
starting from readily available acetylenic substrates. A related
reaction of carbocupration of cyclopropenes provided a useful
asymmetric transformation.30 The simplicity of the reaction was
also attractive: acetylene is the simplest among various elec-
trophilic substrates in cuprate reactions, and the reaction mode
is well defined (cis addition).27 In order to clarify the role of
each component of lithium organocuprate clusters, we studied
the reactions of cluster and noncluster cuprates.
chemically identified also by electrophilic trapping, which was
taken to suggest a nucleophilic attack of Me2Cu- to the terminal
carbon. In the absence of an external electrophile, the inter-
mediate in each case decomposed to give an allenyl product by
either 1,2-elimination or reductive elimination. Though these
schemes are rational by themselves, it is puzzling why the initial
attack point of the copper changes so dramatically by a mere
change of the leaving group.
There have been two simplified mechanistic schemes pro-
posed for the reaction of a cuprate with acetylene (Scheme 1).31
Regardless of the mechanism, a π-complex B is considered to
form initially. In the first mechanism, the methyl group donates
electrons to the acetylene leading to the straightforward four-
centered addition reaction (C),32 and the Cu(I) oxidation state
is maintained during the reaction. In the second one, the copper
atom donates electrons to the acetylene by d-π* electron
donation.33 This would lead to a transition state such as D,
from which the product forms with migration of the copper
atom.
The two lines of contrasting experimental results on propargyl
methyl ether (Scheme 2)32 and propargylic acetate34 (Scheme
3) have been taken as support of each mechanism. In the methyl
ether case, the vinyl copper product E has been identified
chemically by electrophilic trapping. The reaction is therefore
considered to have proceeded via the first mechanism. In the
acetate reaction,34a an allenyl Cu(III) intermediate G was
Computational Methods
General Methods. We used Gaussian 94 programs for all ab initio
and DF calculations.35 The ab initio Hartree-Fock (RHF) and RMP2-
(FC) methods36 were employed first for the studies on the reaction
pathway of the Me2CuLi‚LiCl complex. The basis set used for the ab
initio calculations (denoted as 631WH) consists of the Wachters
(62111111/411111/311) contracted all-electron basis set37 with one
diffuse d function of Hay38 for copper and 6-31G(d)39 for the rest. For
the Me2CuLi‚LiCl reaction, which was studied most extensively,
intrinsic reaction coordinate (IRC) analysis40 was performed near the
(35) GAUSSIAN 94, ReVision B.2; Frisch, M. J.; Trucks, G. W.; Schlegel,
H. B.; Gill, P. M. W.; Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.;
Keith, T.; Petersson, G. A.; Montgomery, J. A.; Raghavachari, K.;
Al-Laham, M. A.; Zakrzewski, V. G.; Ortiz, J. V.; Foresman, J. B.;
Cioslowski, J.; Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng,
C. Y.; Ayala, P. Y.; Chen, W.; Wong, M. W.; Andres, J. L.; Replogle, E.
S.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Binkley, J. S.; Defrees, D. J.;
Baker, J.; Stewart, J. P.; Head-Gordon, M.; Gonzalez, C.; Pople, J. A.
GAUSSIAN, Inc.: Pittsburgh, PA, 1995.
(27) Review: Normant, J. F. Alexakis, A. Synthesis 1981, 841-870.
(28) Although the reaction takes place in two stages to incorporate two
acetylene molecules into a cuprate species to give dipropenylcuprate (eq
1), we modeled only the first stage. In light of the results described later,
the second stage will proceed in a similar manner.
(29) Alexakis, A.; Normant, J.; Villie´ras, J. Tetrahedron Lett. 1976, 38,
3461-3462.
(30) Nakamura, E.; Isaka, M.; Matsuzawa, S. J. Am. Chem. Soc. 1988,
110, 1297-1298.
(36) MP2(FC) (FC ) frozen core): Møller, C.; Plesset, M. S. Phys. ReV.
1934, 46, 618-622. Pople, J. A.; Krishman, R.; Schlegel, H. B.; Binkley,
J. S. Int. J. Quantum. Chem. Symp. 1979, 13, 225-241.
(31) Cf. Negishi, E.; Takahashi, T. Synthesis 1988, 1-19.
(32) Alexakis, A.; Marek, I.; Mangeney, P.; Normant, J. F. J. Am. Chem.
Soc. 1990, 112, 8042-8047.
(37) Wachters, A. J. H. J. Chem. Phys. 1970, 52, 1033-1036.
(38) Hay, P. J. J. Chem. Phys. 1977, 66, 4377-4384
(33) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1984, 25, 3063-3066.
(34) (a) Luche, J. L.; Barreiro, E.; Dollat, J. M.; Crabbe´, P. Tetrahedron
Lett. 1975, 4615-4618. Dollat, J.-M.; Luche, J.-L.; Crabbe´, P. J. Chem.
Soc. Chem. Commun. 1977, 761-762. (b) Vermeer, P.; Meijer, J.;
Brandsma, L. Rec. TraV. Chim. Pays-Bas 1975, 94, 112-114.
(39) Hehre, W. J.; Radom, L.; von Rague´ Schleyer, P.; Pople, J. A. Ab
Initio Molecular Orbital Theory; John Wiley: New York, 1986; references
cited therein.
(40) Fukui, K. Acc. Chem. Res. 1981, 14, 363-368. Gonzalez, C.;
Schlegel, H. B. J. Chem. Phys. 1989, 90, 2154-2161.

4890 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Nakamura et al.
transition structure (TS) at the MP2/631WH level. The s value refers
to the reaction coordinate41 in unit of amu1/2‚bohr, with its zero taken
at TS.
tures and energies summarized in Figure 8A). Then, we reexamined
the same reaction at the B3LYP/631A level to find that the structures
of the stationary points are very similar to the MP2 ones (Figure 8B).
Good agreement with the experimental structural data was for various
theoretical levels (see Appendix).
Further comparison of the energies at the MP2 and B3LYP levels
was made against the MP3, MP4SDQ, and CCSD(T) level calculations
based on the B3LYP/631A geometries, for the reaction of acetylene
with MeCu, Me2Cu-, and Me2CuLi‚LiCl. The detailed numbers are
shown in Figure 12 and Appendix. The data indicate that the B3LYP
values are closer to the CCSD(T) values than the MP2 values are. The
activation energies are quite different between the MP2 and the B3LYP
methods.
On the basis of the knowledge on the reaction pathway obtained at
the MP2 level, the less-computer intensive hybrid DF method was
examined to find that it is of comparable performance for the reaction
of Me2CuLi‚LiCl (see Appendix).42 Thus, for all reagents examined
here (MeCu, Me2Cu-, Me2CuLi, Me2CuLi‚LiCl, and (Me2CuLi)2),43
geometry optimization was performed by density functional/Hartree-
Fock hybrid method using the Becke’s three-parameter exchange
functional44 plus Lee-Yang-Parr nonlocal correlation functional
(B3LYP)45 with the Ahlrichs “SVP” all-electron basis set46 for copper
and 6-31G(d) for the rest (denoted as B3LYP/631A), where the Ahlrichs
“SVP” basis set is smaller than the Wachters + Hay set above.
Energies of the reactants, complex, and TS in the MeCu, Me2Cu-,
and Me2CuLi‚LiCl additions were evaluated at the MP3, MP4SDQ,47
and CCSD(T)48 levels. For CCSD(T) calculations of complex and TS
in the Me2CuLi‚LiCl reaction, we used Ahlrichs “SVP” set for Cu and
3-21G sets for other atoms (denoted as 321A).
As examined for the reaction of Me2CuLi‚LiCl (Figure 8B vs 8C),
the relativistic effects (B3LYP/631A vs B3LYP/631D) on the structures
of complex and TS and the activation energies were found to be very
small (maximum 1.7% geometry difference). Further details of the
comparison is given later in the Appendix.
Results
Relativistic effects of copper atom was examined for the reaction
of Me2CuLi‚LiCl by using the B3LYP functional with a quasirelativistic
pseudopotential representing 10-core electrons in conjunction with
(311111/22111/411) contracted basis set developed by Dolg et al.49
for copper and 6-31G(d) basis sets for the rest (denoted as B3LYP/
631D).
Experiments. In order to obtain experimental data on the
role of lithium counterion in carbocupration chemistry, a series
of experiments on the effect of added crown ether were
performed. Parallel experiments have already been reported for
the conjugate addition.4 Phenylacetylene reacts smoothly with
1 equiv of Bu2CuLi‚LiI at -40 °C for 4 h to give a 91:9 mixture
of a branched and a linear adduct, respectively, in 56%
combined yield. When the same reaction was carried out in
the presence of 10 equiv of 12-crown-4, the reaction slowed
down considerably (12% under the same conditions) and
completed only after 18 h at -20 to 0 °C (53% yield; 85:15
Stationary points were optimized without any symmetry assumption
unless noted otherwise. All TSs and symmetry-constraint stationary
points obtained with the 631WH and 631A basis were characterized
by normal coordinate analysis at the same level of theory (number of
imaginary frequency ) 0 for minima and 1 for TSs). Natural charges50
are calculated at the same level as the level for geometry optimizations.
The Boys localization procedure51 was performed to obtain localized
MOs from the occupied B3LYP/631A Kohn-Sham MOs52 for MP2/
631WH geometries. The total electron density is not changed by this
transformation.
Comparison of Theoretical Methods. In numerous previous
studies on organocopper reactions, various theoretical methods have
been employed largely without systematic evaluation of their perfor-
mance. Until recently, the ab initio MP2 level calculations have been
accepted as the best available method for organocopper chemistry, yet
density functional (DF) methods have emerged recently as methods of
higher cost performance. However, the reliability of the DF method
in the analysis of dynamic behavior of organocopper reagents has not
yet been proven. In the present studies, we have systematically
examined the performance of the DF method against standard ab initio
methods of established general reliability.
isomer distribution). In similar experiments using a cyclopro-
penone acetal (eq 3), the carbocupration with Bu2CuLi‚LiBr‚-
Me2S took place very slowly in the presence of HMPA or 18-
crown-6 and did not lead to completion under conditions where
Bu2CuLi‚LiBr‚Me2S itself reacted quickly.53 The change of the
diastereoselectivity suggests that the reactive species may change
upon addition of the solvating agents. Thus, it was found that
a crown ether slows down the carbocupration reaction as well
as conjugate addition and acylation reactions, which have been
reported previously.4
We initially performed a series of calculations on the reaction of
Me2CuLi‚LiCl with acetylene at the RMP2(FC)/631WH level (struc-
(41) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990, 94, 5523-5527.
(42) At the other basis sets with B3LYP and MP2 methods. See: (a)
Snyder, J. P.; Bertz, S. H. J. Org. Chem. 1995, 60, 4312-4313. (b) Cui,
Q.; Musaev, D. G.; Svensson, M.; Morokuma, K. J. Phys. Chem. 1996,
100, 10936-10944. (c) Hertwig, R. H.; Koch, W.; Schro¨der, D.; Schwarz,
H.; Hrusˇa´k, J.; Schwerdtfeger, P. J. Phys. Chem. 1996, 100, 12253-12260.
(d) Barone, V. J. Phys. Chem. 1995, 99, 11659-11666.
(43) Bo¨hme, M.; Frenking, G.; Reetz, M. T. Organometallics 1994, 13,
4237-4245.
(44) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(45) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785-789.
(46) Scha¨fer, A.; Horn, H.; Ahlrichs, R. J. Chem. Phys. 1992, 97, 2571-
2577.
(47) Trucks, G. W.; Salter, E. A.; Sosa, C.; Bartlett, R. J. Chem. Phys.
Lett. 1988, 147, 359-366.
(48) Scuseria, G. E.; Scheiner, A. C.; Lee, T. J.; Rice, J. E.; Schaefer
III, H. F. J. Chem. Phys. 1987, 86, 2881-2890. Pople, J. A.; Head-Gordon,
M.; Raghavachari, K. J. Chem. Phys. 1987, 87, 5968-5975.
(49) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987,
86, 866-872.
(50) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88,
899-926. NBO Version 3.1 in Gaussian 94 package; implemented by
Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F.
(51) Boys, S. F. Quantum Theory of Atoms, Molecules, and the Solid
State; Lowdin, P. O., Ed.; Academic Press: New York, 1968; pp 253-
262. Haddon, R. C.; Williams, G. R. J. Chem. Phys. Lett. 1976, 42, 453-
455.
Cuprate Models. Structures of Me2Cu- (linear) and R2CuLi
dimers (A) known from crystallographic studies10,54 suggest that
the cuprate R-Cu-R structure prefers a linear arrangement.
In fact, we found that bending of the linear Me-Cu-Me
(53) Isaka, M. Ph.D. Dissertation, Tokyo Institute of Technology, Chapter
3, 1991; pp 83-84.
(54) Hope, H.; Olmstead, M. M.; Power, P. P.; Sandell, J.; Xu, X. J.
Am. Chem. Soc. 1985, 107, 4337-4338.
(52) Kohn, W.; Sham, L. J. Phys. ReV. 1965, 140, A1133-A1138.

Addition of Li Organocuprate Clusters to Acetylene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4891
Table 1. Comparison of Theoretical Methods for Representative
Structural Parameters of RT1-d [R ) Me: (Me2CuLi)2] and RT1w
{R ) Me: [Me2CuLi(OH2)]2} and RT1 (H, X ) Cl:
Me2CuLi‚LiCl)
structure in free Me-Cu-Me anion to <C-Cu-C ) 116° (the
angle found in the Me2CuLi/acetylene complex later in Figure
5) results in as much as 28.0 kcal/mol increase of energy
(B3LYP/631A//B3LYP/631A). Theoretical studies by Snyder
and Frenking on various possibilities of cuprate structures not
only reproduced the pertinent properties of the crystallographic
structures10,11,43,54,55 but also led to proposals of some others,
such as H. The geometries of our cuprate models are based on
these experimental and theoretical studies.
For the structure of the smallest lithium organocuprate cluster
Me2CuLi, an open and a C2V closed form (see below) have been
considered.43 With the Me-Cu-Me moiety being bent and
strained, the closed structure is not even a local minimum.
Though the Me-Cu-Me group is in its favorable linear
geometry in the open form (as shown later in Figure 5), the
positive charges in the coordination-free Li atom and the
negative charge in one of the two methyl groups are placed far
away from each other and cannot enjoy favorable charge
interaction.56 As a consequence, the Me2CuLi cluster is less
stable than its symmetrical dimer (A, R ) Me).43 One may
postulate that the Me2CuLi minimum cluster exists in a very
polar basic solvent. The consequence of such solvation will
be either the attenuation of the Lewis acidity of the lithium atom
or the formation of naked Me2Cu- as experimentally shown
for the case of crown ether solvation (vide supra).54
model
level
C-Cua
C-Lia
C-Cu-Cb
RT1-d
B3LYP/631A
MP2/631HWc
B3LYP/321A
MP2/321HWc
HF/631WH
B3LYP/631A
MP2/631WH
1.991
1.968
1.987
1.983
2.066
1.994
1.948
2.070
2.043
2.229, 2.207
2.259, 2.240
2.071
2.087
2.140
171.3
174.2
168.4
168.5
177.4
177.9
174.4
RT1w
RT1
a Bond lengthes in Å. b Angles in deg. c Reference 43.
Figure 1. Stationary points of the MeCu addition to acetylene. MeCu,
acetylene, complex, TS, and product were optimized with C3V, D∞h
,
C2V, Cs, and Cs symmetries, respectively. Bond lengths and angles at
the B3LYP/631A are shown in Å and deg, respectively. Energy changes
at the B3LYP/631A//B3LYP/631A level are shown on the arrows in
Especially noteworthy is the recent proposal of a six-centered
monomeric cuprate H for X ) Me, CN, and I.11,55 The structure
of Me2CuLi‚LiCl cluster can be viewed as a composite of half
of the eight-centered structure A and half of a four-centered
lithium halide dimer.57 Both have been supported by crystal-
lography, and, hence, H appears to be a reasonable structure.
Although structures A and H differ in their numbers of copper
atoms (i.e., dimer vs monomer), they exhibit close similarity
for the Li-R-Cu-R-Li moiety (for details see Figures 9 and
10, Table 1). In view of this similarity, we conjectured that
the monomer H will serve as a good model for studies of the
dimer cluster A, if the remaining part (i.e., the upper part) of
the molecule in A and H does not actively take part in the
cuprate reaction. As will be reported below, this assumption
was found to be reasonable. This assumption was also
practically important prior to our finding that B3LYP method
performs very well in the studies on cuprate chemistry42a,55 (vide
supra), since the current level of computational power does not
allow extensive studies on the dimer possessing two copper
atoms at the ab initio level better than that of the MP2 method.
We found this level to be the minimum level of the ab initio
calculations for the studies on cuprate reactions. From the
chemical point of view, it must be noted that studies using a
kcal/mol. The value of imaginary frequency at TS is 494.7i cm-1
.
Natural charges on atoms or groups in boxes are in bold. Total energies
of MeCu and acetylene are -1680.216 88 and -77.325 65 hartrees at
the B3LYP//B3LYP/631A level, respectively.
monomer model such as Me2CuLi and Me2CuLi‚LiCl are no
less important than those using a dimer, as monomer participa-
tion is also widely considered as one possibility (although it is
not clearly supported nor eliminated by experiments).
Therefore, we selected the Snyder model H with X ) Cl as
the first model of lithium organocuprate cluster11,55 and exam-
ined the stationary points in the acetylene reaction and the
structures on the IRC of the carbocupration reaction by the MP2/
631WH method. Such a careful structure search was necessary
owing partly to the lack of precedent on the dynamic studies of
cuprate clusters and partly to the structural and electronic
complexity of the system under consideration. With detailed
information on the reaction of Me2CuLi‚LiCl at the MP2 level,
we then found that the B3LYP method is reliable enough and
used it for all other reactions. As shown below, comparison of
Me2CuLi‚LiCl and dimer indicated that the essence of the
cuprate cluster reactivities do reside in the Li-Me-Cu-Me-
Li part structure of the cluster. The same was found in the
conjugate addition reaction reported in the following article.26
MeCu Addition. We first examined the simplest model, the
MeCu addition to acetylene, with the B3LYP/631A method
(Figures 1 and 2), and located a π-complex, a TS, and a product.
The geometries and energies are very similar to the Hartree-
Fock results reported previously.58 The complex formation takes
place with a very large stabilization energy (23.6 kcal/mol,
B3LYP/631A//B3LYP/631A), and the activation energy of C-C
bond formation was also very high (43.6 kcal/mol, B3LYP/
(55) (a) Snyder, J. P.; Tipsword, G. E.; Splanger, D. J. J. Am. Chem.
Soc. 1992, 114, 1507-1510. (b) Snyder, J. P.; Splangler, D. P.; Behling, J.
R.; Rossiter, B. E.; J. Org. Chem. 1994, 59, 2665-2667. Stemmler, T. L.;
Barnhart, T. M.; Penner-Hahn, J. E.; Tucker, C. E.; Knochel, P.; Bo¨hme,
M.; Frenking, G. J. Am. Chem. Soc. 1995, 117, 12489-12497. Huang, H.;
Alvarez, K.; Cui, Q.; Barnhart, T. M.; Snyder, J. P.; Penner-Hahn J. E. J.
Am. Chem. Soc. 1996, 118, 8808-8816, and 1996, 118, 12252 (correction).
(56) Lambert, C.; von Rague´ Schleyer, P. Angew. Chem., Int. Ed. Engl.
1994, 33, 1129-1140.
(57) Hall, S. R. Raston, C. L.; Skelton, B. W.; White, A. H. Inorg. Chem.
1983, 22, 4070-4073. Amstutz, R.; Dunitz, J. D.; Laube, T.; Schweizer,
W. B.; Seebach, D. Chem. Ber. 1986, 119, 434-443. Raston, C. L.;
Whitaker, C. R.; White, A. H. Aust. J. Chem. 1989, 42, 201-207.

4892 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Nakamura et al.
Figure 2. Energy profile in the MeCu addition to acetylene. Energies
relative to reactants (B3LYP/631A//B3LYP/631A) in kcal/mol are
shown in parentheses.
631A//B3LYP/631A). In the complex, there are only small
changes in structure (and charge) for the acetylene moiety, as
found for the crystal structure of acetylene complexes of Lewis
acidic CuX (X ) Cl, etc.) salt59 and for the previously calculated
structure of [Cu(C2H2)]+.60 The MeCu/acetylene complexation
is a simple electrostatic complex in its character, with the
substantial dipole of Cu+δMe-δ interacting with the π-electron
cloud of acetylene which further polarizes CuMe. The addition
reaction may be best viewed as a simple four-centered addition
reaction (namely, the metal acting as a Lewis acid) as previously
reported.32,58 The high calculated activation energy is consistent
with the fact that alkylcopper compounds alone are not
(experimentally) reactive nucleophilic reagents.
Me2Cu- Addition.61 Interaction of Me2Cu- with acetylene
(Figures 3 and 4) affords a complex with 0.8 kcal/mol
stabilization energy (B3LYP/631A//B3LYP/631A). Though the
stabilization energy is much smaller than that obtained for
MeCu, the structure of the complex indicates tighter binding
between the cuprate and the acetylene (cf. the longer C1-C2
and the shorter C1-Cu bond lengths as well as the acetylene
deformation), and a significant amount of negative charge is
transferred to acetylene. Since linear Me2Cu- has no permanent
dipole, the electrostatic interaction is small as reflected in the
small stabilization energy. The interaction comes mainly from
the Me2Cu- HOMO-2 (mainly corresponding to Cu 3d orbital)-
acetylene LUMO charge transfer interaction. The sp-hybridized
acetylene acts a good electron acceptor. The structure of the
complex and TS are very similar to the Me2Cu-/acetylene part
structure in the π-complexes of Me2CuLi, Me2CuLi‚LiCl, and
(Me2CuLi)2 (vide infra). The activation energy of 20.3 kcal/
mol (B3LYP/631A//B3LYP/631A) is still high, though it is
much lower than that found for MeCu. This is consistent with
the low reactivity of the crown-solvated cuprate reagents.
Me2CuLi Addition. In the reaction of the minimum cluster
Me2CuLi (Figures 5 and 6), the open cluster closes upon
complexation with acetylene as the linear Me-Cu-Me bond
is bent by complexation. The cluster closing causes the lithium
cation to be stabilized by the two negatively charged methyl
groups and therefore results in stabilization of the complex by
as much as 17.8 kcal/mol (B3LYP/631A//B3LYP/631A). Such
a large stabilization is reflection of the instability of the Me2-
CuLi monomer and is not observed in larger clusters (vide infra),
wherein all negative and positive charges are electrostatically
Figure 3. Stationary points of the Me2Cu- addition to acetylene.
Me2Cu-, acetylene, complex, and product were optimized with D3h,
D∞h, Cs, and Cs symmetries, respectively. TS was optimized with no
symmetry. Bond lengths and angles at the B3LYP/631A are shown in
Å and deg, respectively. Energy changes at the B3LYP/631A//B3LYP/
631A level are shown on the arrows in kcal/mol. The value of imaginary
frequency at TS is 392.8i cm-1. Natural charges on atoms or groups in
boxes are in bold. In the TS described in this and Figures 5, 8, and 13,
the front views place C1-C2-H or the plane of the paper, the side
view places C1-C2-H perpendicular to the plane. Total energies of
Me2Cu- and acetylene are -1720.163 36 and -77.325 645 hartrees at
the B3LYP/631A//B3LYP/631A level, respectively.
Figure 4. Energy profile in the Me2Cu- addition to acetylene. Energies
relative to reactants (B3LYP/631A//B3LYP/631A) in kcal/mol are
shown in parentheses.
stabilized already in the starting clusters. The negative charge
developing on an acetylenic carbon terminus is now stabilized
by the lithium cation inside the cluster structure, and the
activation energy decreases to 15.8 kcal/mol. The charge
stabilization by the lithium atom causes the deformation of the
TS of Me2Cu- addition (Figure 5), changing the Me-Cu-Me
angle from 109.9° to 141.3° and lengthening the Cu-CH bond
(2.250 Å). The activation energy of 15.8 kcal/mol is lower than
that of Me2Cu-, comparable with that of (Me2CuLi)2, and higher
than that of Me2CuLi‚LiCl (vide infra). The C3-Cu-C4 bond
angle is much wider (141°) than those in others, which are
consistently between 109° and 126°.
(58) Nakamura, E.; Miyachi, Y.; Koga, N.; Morokuma, K. J. Am. Chem.
Soc. 1992, 114, 6686-6692. Nakamura, E.; Nakamura, M.; Miyachi, Y.;
Koga, N.; Morokuma, K. J. Am. Chem. Soc. 1993, 115, 99-106.
(59) (a) Thompson, J. S.; Whitney, J. F. J. Am. Chem. Soc. 1983, 105,
5488-5490. (b) Munakata, M.; Kitagawa, S.; Kawada, I.; Maekawa, M.;
Shimono, H. J. Chem. Soc., Dalton Trans. 1992, 2225-2230. (c) Brantin,
K.; Ha˚kansson, M.; Jagner, S. J. Organomet. Chem. 1994, 474, 229-236.
(60) Bo¨hme, M.; Wagener, T.; Frenking, G. J. Organomet. Chem. 1996,
520, 31-43.
(61) In the Me2Cu- addition to acetylene, the geometries of complex
and TS as well as the activation energy were found to be unaffected (<1.2%
for geometry and 0.3 kcal/mol for energy) by addition of a diffuse sp
function of exponent R ) 0.438 for carbon (Clark, T.; Chandrasekhar, J.;
Spitznagel, G. W.; von Rague´ Schleyer. P. J. Comput. Chem. 1983, 4, 294-
301).
Me2CuLi‚LiCl Addition
1. MP2 Studies. From the HF studies described in the
Appendix, it became clear that the calculations on the cuprate

Addition of Li Organocuprate Clusters to Acetylene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4893
Figure 7. Structures, energies, and charges in the addition of Me2-
CuLi‚LiCl to acetylene. (A) Energy relative to reactants, (B) C2-C3
bond length, and (C) natural charges for various points on the reaction
pathway as obtained at the MP2/631WH//MP2/631WH level. The
pathway between INT1 and INT2 was obtained by following the IRC
down from TS. Structural correlation between CP and INT1 as well
as between INT2 and PD was confirmed by usual optimization
procedure.
Figure 5. Stationary points of the Me2CuLi addition to acetylene. Me2-
CuLi, acetylene, and complex were optimized with Cs, D∞h, and C2V
symmetry, respectively, and others with no symmetry. Bond lengths
(and bond angles) at the B3LYP/631A level are shown in angstroms
and degrees, respectively. The value of imaginary frequency of TS is
297.2i cm-1. Natural charges on atoms or groups in boxes are in bold.
Energies of Me2CuLi and acetylene are -1727.687 180 and -77.325 645
hartrees at the B3LYP/631A//B3LYP/631A level, respectively.
Scheme 4. Trap-and-Bite Mechanism for the Addition of
Organocuprate to Acetylene
Figure 6. Energy profile in the Me2CuLi addition to acetylene.
Energies relative to reactants (B3LYP/631A//B3LYP/631A) in kcal/
mol are shown in parentheses.
cluster mandate consideration of dynamical electron correlation
effects. The importance in the treatment of these effects is also
in consonance with what is generally agreed for ab initio
calculations on first row transition metals.62 Thus we studied
MP2 calculations of Me2CuLi‚LiCl addition to acetylene. First
with the MP2 method and then with the B3LYP method, both
of which led to the same chemical conclusion.
Figure 7 shows the energies and other parameters for the
structures of the stationary points on the potential surface as
obtained by MP2 optimization.63 A summary of the findings
that will be detailed below is schematized as a five-phase
scheme, a trap-and-bite mechanism (Scheme 4, X ) Cl and
also for Me-Cu-Me (vide infra)). The net process constitutes
the conversion of one six-centered cuprate (RT1) to another
(PD). The above pathway may be viewed as copper-assisted
carbolithiation followed by intracluster transmetalation. In the
first phase of the reaction (phase 1), the cluster traps acetylene
with its copper atom to form CP. Negative charge flows from
Cu(I) to acetylene (Figure 7C) to liberate the C3 methyl group
free from coordination to Li1 (phase 2). INT1 is an example
of the structure connecting CP and TS. Though this process
involves a major bond reorganization, such a process involving
the cleavage of a dicoordinated lithium atom and a pentacoor-
(62) Lu¨thl, H. P.; Siegbahn, P. E. M.; Almlo¨f, J. J. Phys. Chem. 1985,
89, 2156-2161. Siegbahn, P. E. M.; Svensson, M. Chem. Phys. Lett. 1993,
216, 147-154.
(63) We obtained for the pathway between INT1 and INT2 as shown
in Figures 7 and 8A by IRC analysis at the MP2/631WH level. INT1 and
INT2 are nonstationary points and INT1 and INT2 led to CP and PD with
the minimum search, respectively.

4894 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Nakamura et al.
Figure 8. (A) Complexes, TS, and representative structures in the reaction of Me2CuLi‚LiCl with acetylene obtained at the MP2/631WH level.
Bond lengths and angles are shown in Å and deg, respectively. Energy changes at the MP2/631WH//MP2/631WH level are shown on the arrows
in kcal/mol. INT1 is a structure at s ) -6.41 amu‚bohr-1/2 from CP to TS and INT2 is a structure at s ) +7.18 amu‚bohr-1/2 from TS to PD on
the reaction path. The value of imaginary frequency of TS is 391.7i cm-1. The total energies of RT1 and RT2 are -2193.383 13 and -77.066 79
hartrees, respectively. (B) Stationary points in the Me2CuLi‚LiCl addition to acetylene obtained at the B3LYP/631A level. The value of imaginary
frequency of TS is 191.38i cm-1. (C) Complex and TS in the Me2CuLi‚LiCl addition to acetylene obtained at the B3LYP/631D level, which takes
into account relativistic effects on copper.
dinated methyl group in solution is expected to be a very low
energy process as is known for the behavior of alkyllithium
clusters in ethereal solution.64 By the transformation through
a series of structures (INT1 to INT2), the open cluster bites
into the substrate (phases 3 and 4). It is at this stage that new
C-C and C-Li bonds are formed at the expense of a covalent
C-Cu bond. After propenyllithium formation (cf. INT2, a
nonstationary point), intramolecular transmetalation from Li to
Cu (phase 5) makes the energy drop further to produce the
propenylcopper product (PD).
The representative 3D-structures on the reaction pathway of
the Me2CuLi‚LiCl reaction are shown in Figure 8 and energies
in Figure 9. Interaction of RT1 with RT2 bends the linear C3-
Cu-C4 bonding to form a complex (CP, near Cs symmetry)
(Figure 7A and 8A).65 C1-C2-Cu-C4 dihedral angle of 166.5°
indicates nearly square planar geometry consistent with d8 Cu-
(III) formalism.19,20 The conversion from RT1 to CP ac-
companied with a change (though small) of d-orbital population
(64) Cf.: McGarrity, J. F.; Ogle, C. A.; Brich, Z.; Loosli, H.-R. J. Am.
Chem. Soc. 1985, 107, 1810-1815.

Addition of Li Organocuprate Clusters to Acetylene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4895
Table 2. Comparison of Activation Energies (TS - CP) in
kcal/mol for Various Levels of Theoretical Methods at the MP2/
631WH Optimized Geometries of CP and TS
method/basis set
321A
631WH
MP2
MP3
MP4DQ
MP4SDQ
CCSD(T)
B3LYPa
26.6
14.4
21.3
19.9
14.4
18.2
16.2
15.4
12.6
Figure 9. Energy profile in the Me2CuLi‚LiCl addition to acetylene.
Energies relative to reactants (B3LYP/631A//B3LYP/631A) in kcal/
mol are shown in parentheses.
13.6
a Geometry optimization performed at the same level.
TS. Note that, with the usual geometry optimization procedure,
INT1 further goes down hill to CP with smooth migration of
Li1 to C3.
In INT1, the C3-Li1 bond is completely cleaved (4.278 Å)
and Li1 is coordinated to the π-orbital of C1 (negatively charged,
Figure 7C) perpendicular to the C1-C2-H2 plane. Li1-C1-
C2-H2 dihedral angle is -116°. In TS, the free C3 anion will
undergo C-C bond formation with C2. Further along the
reaction path toward TS, cleavage of C3-Cu bond results in
an increase and decrease of negative charge on C1H1 and C3H3,
respectively (Figure 7C).
In TS, a few important events are taking place simultaneously.
One is the rapid formation of C2-C3 bond (Figure 7B) with
retention of the planar coordination geometry of Cu (Figure 8).68
The LMO in Figure 10 (LMO5) illustrates C2-C3 bond
formation. The relatively short Cu-C2 bond (1.844 Å) is
noteworthy and is formed mainly by the copper’s filled 3d
orbitals as shown in LMO4 in Figure 10. The second notable
geometry change is that Li1 is forming a σ-bond with C1.
LMO3 in Figure 10 indicates that the C2’s filled spn orbital
strongly interacts with Li1 as well as with Cu1. According to a
recent proposal,69 the C1 atom can be considered to act as a
donor ligand for a Cu(III) species. Comparison of the localized
MOs in Figure 10 indicates that the molecular orbital interactions
characterizing the donation/backdonation interaction in CP
(LMO1 and LMO2) changes to those characterizing Cu(III)
reductive elimination (LMO3 and LMO4).
The barrier height from CP to TS is 26.6 kcal/mol at the
MP2/631WH//MP2/631WH level and decreases by several kcal/
mol at higher levels (Table 2). The value for the best level
CCSD(T)/631WH can be estimated from MP4SDQ/631WH,
CCSD(T)/321A, and MP4SDQ/321A results to be 19.1 kcal/
mol. The activation energy remains unaffected (MP2 level, 26.5
kcal/mol) by solvent polarity (THF, ꢀ0 ) 7.58, with the self-
consistent reaction field method (based on Onsager model)70
applied without structure optimization). The B3LYP studies
using an explicit solvent molecule on Li1 in the conjugate
addition of (Me2CuLi)2 (the following paper)26 revealed that
solvation effects are small. The role of the bridging Li-Cl-
Li moiety was noted clearly for the TS. Thus, upon hypotheti-
cally removing the Li1-Cl-Li2 bridge, the activation energy
increased to 33.4 kcal/mol (MP2/631WH//MP2/631WH). This
value coincides with the activation energy of 36.9 kcal/mol
(MP2/631WH//B3LYP/631A) obtained for free Me2Cu- and
indicates that the charge stabilization by the Li1 atom contributes
to lower the energy barrier.
Figure 10. Contour plots of localized Kohn-Sham MOs (B3LYP/
631A//MP2/631WH) of (a) CP and (b) TS in the C1-C2-Cu (LMO1-
4) and the C2-C3-C4 (LMO5) planes. Contour levels in e‚au-3 are
from -0.30 to +0.30 at intervals of 0.05. Positive contours are solid,
negative dotted, and nodal lines dashed.
[4s(0.74), 3d(9.64) in RT1 and 4s(0.52), 3d(9.46) in CP] in
natural charge (MP2) suggested the presence of backdonation
from Cu to acetylene. The typical donation-backdonation
interaction33,66 is clearly seen in the occupied localized Kohn-
Sham MOs in Figure 10 (donation in LMO1 and back-donation
in LMO2), and hence the complex should be better viewed as
a cupriocyclopropene rather than a simple π-donation complex.67
For a reaction of such complexity, information from stationary
points (CP, TS, and PD) was not enough to understand the
reaction pathway. We therefore studied the structures along
the MP2-IRC to find that important intracluster transformations
are taking place between the stationary points. Thus, by
following the down-energy IRC from TS toward CP (Figure
7A), we reached INT1 (a nonstationary point, Figure 8A), which
is an example of an intermediary structure connecting CP and
(65) CP at the RMP2(FC) geometry showed slight triplet instability of
RHF wave function (the expectation value of S2 was however only 0.052
in the UHF calculation). RB3LYP/631A wave functions for both CP and
CP-d at the RB3LYP geometries were found to be stable. See: Seeger, R.;
Pople, J. A. J. Chem. Phys. 1977, 66, 3045-3050.
(68) RHF/631WH wave functions for TS at the RMP2/631WH geom-
etries were found to be stable.
(69) Dorigo, A. E.; Wanner, J.; von Rague´ Schleyer, P. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 476-478. Snyder, J. P. J. Am. Chem. Soc. 1995,
117, 11025-11026.
(70) Onsager, L. J. Am. Chem. Soc. 1936, 58, 1486-1493. Tapia, O.;
Goscinski, O. Mol. Phys. 1975, 29, 1653-1661. Wong, M. W.; Wiberg,
K. B.; Frisch, M. J. Chem. Phys. 1991, 95, 8991-8998. Wong, M. W.;
Wiberg, K. B.; Frisch, M. J. J. Comput. Chem. 1995, 16, 385-394.
(66) Dewar, M. J. S. Bull. Chim. Soc. Fr. 1951, C71-C77. Chatt, J.;
Duncanson, L. A. J. Chem. Soc. 1953, 2939-2947.
(67) No cuprate/acetylene complex could be located on the potential
surface of the HF/631WH level calculations.

4896 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Nakamura et al.
In TS, Li1 moves into, and Cu out of, the C1-C2-H2 plane,
altering the nature of bonding between the acetylenic carbons
and the metals. This change continues until INT2 (a nonsta-
tionary structure, Figure 7A, see also Figure 8A), where Li1
forms a σ-bond with sp2-hybridized C1, and Cu coordinates to
the carbon π-orbital perpendicular to the C1-C2-H2 plane. This
event generates propenyllithium π-complexed with CH3Cu
(INT2, Figure 8A). The structural changes accompany the
change of Cu(III) into simple MeCu(I), which is reflected in
the decrease of the positive charge density on Cu (Figure 7C,
i.e., larger in TS, and smaller in INT2 and PD).
The propenyllithium structure INT2 is not a local minimum,
and the usual geometry optimization smoothly converted INT2
to PD. The structural change represents the conversion of INT2
to the propenylcopper π-complexed with Li via intramolecular
transmetalation. The lowering of the energy after the completion
of the C-C bond formation, which is found in Figure 7A, is
due to the formation of covalent C-Cu bond. It is notable that,
throughout the reaction, the structure of the Li1-Cl-Li2 bridge
remains relatively unchanged, and the charges on the Li1-Cl-
Li2-C4H3 moiety show very little change (<(0.04, data not
shown).
The structure of the propenyl cuprate PD at the MP2 level
of theory was essentially the same as that for the HF structure
and show close resemblance to the calculated11,55 and crystal
structures of dimeric lithium cuprate clusters of general structure
A (as later described in the Appendix).10 The C-Cu bond
lengths show close agreement ((3-7%) with experimental
values for the related structure.11,55 PD features the C1-Cu
σ-bonding and π-coordination between Li1 and C1, resembling
the geometry found in the crystal of [{Li(OEt2)}Cu(C6H5)2]2.10b
2. B3LYP Studies. Stationary points obtained by the MP2
method was examined also by the B3LYP method. Thus, CP,
TS, and PD in the Me2CuLi‚LiCl reaction were optimized at
the B3LYP/631A level (Figure 8B). The B3LYP/631A struc-
tures were indeed found to be in good agreement with the MP2/
631WH structure (Figure 8A). The B3LYP/631A bond lengths
reproduced those obtained by the MP2/631WH within 2%,
except for some partial bonds (dotted line) which were off by
<10%. It is reported that relativistic effect for copper atom is
the largest among first row transition metals.71 Hence, we
evaluated the relativistic effects on the copper atom at the
B3LYP/631D level (Figure 8C) to find that the difference of
the bond length between the B3LYP/631A and the B3LYP/
631D methods was less than 1.7%.
The activation energy at the B3LYP/631A//B3LYP/631A
level is 13.6 kcal/mol. The relativistic effect on the activation
energy is comparable (12.7 kcal/mol at the B3LYP/631D//
B3LYP/631D). Though these activation energies are rather
unrealistically small, they are somewhat closer to the best
estimated activation energy of 19.1 kcal/mol (CCSD(T)/631WH/
/MP2/631WH, vide supra) than the MP2/631WH energy (26.6
kcal/mol). The energetics of the Me2CuLi‚LiCl reaction is
shown in Figure 9 and may be compared with the data of other
reactions studied at the same level of theory.
Figure 11. Reactants, complex, TS in the reaction of (Me2CuLi)2 with
acetylene. Bond lengths and angles at the at the B3LYP/631A level
are shown in Å and deg, respectively. Energy changes at the B3LYP/
631A//B3LYP/631A level are shown on the arrows in kcal/mol. The
value of imaginary frequency of TS-d is 272.8i cm-1
.
B3LYP/631A//B3LYP/631A) was found to be higher than the
13.6 kcal/mol value for Me2CuLi‚LiCl (B3LYP/631A//B3LYP/
631A). The lower activation energy in the latter is perhaps due
to better charge stabilization by the Li1 atom which is attached
to the electronegative chlorine atom.
Discussion
The above results represent the first computational simulation
of the reaction pathway of a large polymetallic cuprate cluster
with acetylene. All stationary points CP, TS, and PD in Scheme
4 have been structurally correlated by the IRC reaction-path-
following method and the usual geometry optimization proce-
dure. We first compared the noncluster copper compounds and
the lithium organocuprate clusters and then evaluated the effects
of cluster structure for Me2CuLi‚LiCl and (Me2CuLi)2.
Simple organocopper compounds (RCu) are unreactive
reagents, while organocuprate reagents (R2CuM) are reactive
for the delivery of a nucleophilic R group to acetylene and
cyclopropene. When the countercation M+ is removed by
solvation, the remaining R2Cu- loses its nucleophilicity. As
shown in Figures 2, 4, 6, 9, and 11, the calculated activation
energies (B3LYP/631A//B3LYP/631A) of the addition of vari-
ous organocopper reagents to acetylene depends very much on
the nature of the reagents. Of the five cuprate structures
examined, the energies of MeCu and Me2CuLi are anomalous,
while others show a similar trend. The origins of the anomalies
for MeCu and the minimum cluster Me2CuLi were discussed
above.
(Me2CuLi)2 Addition. The reaction of the dimer (Me2CuLi)2
was next examined for three important stages of the reaction,
RT1-d, CP-d, and TS-d (Figure 11), and surprising similarity
was found with Me2CuLi‚LiCl. Most importantly, the spatial
arrangement of the crucial Li1-C1-C2-Cu1-C3-C4 moiety
of these three stationary points remains essentially the same as
that obtained for the monomer (Figure 8) with only 0-3.8%
differences of the bond lengths (compared at the B3LYP/631A
level). The activation energy for (Me2CuLi)2 (16.1 kcal/mol,
The activation energy of the MeCu addition to acetylene is
very high (43.6 kcal/mol, B3LYP/631A//B3LYP/631A). The
(71) Pyykko¨, P. Chem. ReV. 1988, 88, 563-594.

Addition of Li Organocuprate Clusters to Acetylene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4897
Scheme 5
Figure 12. Potential energy profiles of (A) MeCu, (B) Me2Cu-, and
(C) Me2CuLi‚LiCl addition to acetylene at various levels. Bars with
white circles are at the CCSD(T)/631WH//B3LYP/631A level (A, B)
or at the CCSD(T)/321A//B3LYP/631A level (C), and with black
squares are at the B3LYP/631A//B3LYP/631A level. The others are at
the MP2/631WH//B3LYP/631A level. Energies relative to reactants
in kcal/mol are shown near arrows. The energies with box are at the
B3LYP/631A level.
metal center is acting as a Lewis acid, and the direct cleavage
of the covalent C-Cu bond (estimated to be 55 kcal/mol)72
contributes to the high energy barrier. In the reaction of Me2Cu-
with acetylene, a back-donation process as well as donative
interaction operates to cause charge transfer from the cuprate
to acetylene. The activation energy drops by 23 kcal/mol to
become 20.3 kcal/mol (B3LYP/631A//B3LYP/631A). When
Me2Cu- is incorporated as part of a lithium cuprate cluster, the
activation energy decreases further to become 13.6-16.1 kcal/
mol (B3LYP/631A//B3LYP/631A). The activation energy is
estimated to be 19.1 kcal/mol at the CCSD(T)/631WH level
(vide supra), and this value is reasonable for a reaction taking
place below 0 °C (Table 2).
Scheme 6
It is notable that the basic atomic arrangement found in the
complex and the TS of the reaction of Me2Cu- (cf. Figure 4
and heavy lines in Scheme 5a) is found largely unchanged in
the corresponding structures in the reactions of all the lithium
cuprate clusters, except for that of the minimum cluster Me2-
CuLi (Scheme 5b). This similarity leads to a reasonable
assumption that the geometry of atoms surrounding the copper
atom is controlled mainly by copper’s coordination geometry
and rather little by the other atoms in the cuprate cluster. Thus,
the “bridge” structure (Li1‚‚Li2), which is large enough to fill
in the distance of 3.9-4.05 Å between C1 and C4 atoms,
stabilizes well the two negatively charged sites. In the minimum
cluster, Me2CuLi, the single Li cation is apparently so small
that it can only achieve stabilization by changing the geometry
of the “core structure”. The electronic property of the bridge
will also contribute to the charge stabilization. With its chlorine
atom, the Li-Cl-Li bridge should be a better negative charge
stabilizer than the Li-Me-Cu-Me-Li, and this could be a
reason for the lower activation energy for Me2CuLi‚LiCl. One
recognizes that the TS becomes earlier (longer C2-C3 bond,
and shorter C2-Cu bond) as the activation energy decreases in
the order of (Me2CuLi)2, Me2CuLi, and Me2CuLi‚LiCl.
The role of the counter cation (M+) in R2CuM is summarized
in the conceptual sketch shown in Scheme 6. Thus, the cuprate
anion first donates electrons to acetylene (d-π* interaction), and
the countercation participates in the stabilization of the negative
charge to generate I (depicted as an extreme representation).
This species is unstable (which is in fact a TS as shown in the
present studies) and undergoes reductive elimination to generate
a vinyl metallic species J and MeCu. Intracluster transmeta-
lation affords a vinylcuprate product. The low reactivity of the
crown-solvated R2CuLi strongly suggests that M+ must exist
better as a part of the cuprate cluster.
The above data provide a unified mechanism for the puzzling
behavior of the propagylic acetate and methyl ether in Schemes
2 and 3.32,34b According to the above mechanistic framework,
the initial stage of the reaction will involve reversible electron
donation from the cuprate to the substrate (K in Scheme 7).
When there is a good leaving group such as acetate, electron
donation at the acetylene terminus will kick out the leaving
group to generate an allenyl Cu(III) species L, which then gives
the allene upon reductive elimination. When the leaving group
is a less potent methoxy group (Scheme 7b), the standard
sequence of carbocupration takes place (via M as an unstable
intermediate) to generate a stable vinyl cuprate (O), which then

4898 J. Am. Chem. Soc., Vol. 119, No. 21, 1997
Nakamura et al.
Figure 13. Representative structures in the Me2CuLi‚LiCl addition to acetylene at the HF/631WH level. Bond lengths and angles are shown in Å
and deg, respectively. Energy changes at the HF/631WH level are shown on the arrows in kcal/mol. Total energy of sum of RT1 and RT2 is
-2269.396 25 hartree (HF/631WH//HF/631WH).
Scheme 7
forming a stable intermediate. The negative charge is transferred
from the copper 3d orbital, and the transferred charge becomes
stabilized by the lithium atom (Li1), as shown by LMO3 and
LMO4 in Figure 10. The whole reaction pathway illustrates
the uniqueness of copper among neighboring elements in the
Periodic Table for its two-stage redox properties and its ability
to form a stable ate complex cluster. The Li1‚‚Li2 bridge
structure of reasonable size not only stabilizes the starting cluster
but also assists the electron flow during C-C bond formation.
The M+ and LiX in R2CuM and R2CuM‚LiX are therefore active
participants of the reaction.
In summary, the present studies show that lithium organo-
cuprate clusters have a built-in function to facilitate carbocu-
pration reactions, which is not available for noncluster copper
reagents (RCu and R2Cu-). In light of its predominance in a
solution of a standard Gilman reagent, the dimeric cluster (Me2-
CuLi)2 is a good candidate for the reactive species, as has been
widely believed. On the other hand, the close similarity of Me2-
CuLi‚LiCl and (Me2CuLi)2 suggests that the LiX-complexed
species (Me2CuLi‚LiX) and various other higher homologues
can also act as reactive species if they have a suitably structured
bridge and high equilibrium concentration. The minimum
cluster Me2CuLi species is a less likely reactive species because
of its low equilibrium concentration. The importance of the
Cu(I)/Cu(III) redox process and the active role of M+ as
countercation was also found in the conjugate addition described
in the following article. On the basis of the above results, we
can ascribe the uniqueness of the copper atom to its ability to
easily undergo the Cu(I)/(III) redox process as well as to form
stable polyorgano polymetallic ate complexes. No elements in
the neighborhood of copper in the Periodic Table possess these
two properties together.
undergoes 1,2-elimination. Clearly, the first pathway (Scheme
7a) has much to do with the mechanism of SN2′-substitution
reaction.5,73
The studies in this and the following paper revealed that the
essence of the C-C bond forming reaction is remarkably
insensitive to the difference of cluster structures. In both Me2-
CuLi‚LiCl and (Me2CuLi)2 clusters, lithium atom followed later
by bond reorganization involving the cleavage of labile C-Cu-
(III) bond. In this way, the reaction circumvents the high energy
cost of direct C-Cu(I) bond cleavage (55 kcal/mol).72 As the
electron donation to acetylene takes place, the copper(I) atom
loses electrons to form a copper(III) transition species, which
immediately starts to undergo reductive elimination without
Appendix: Comparison of the Theoretical Methods
In this Appendix, the performance of theoretical methods is
described for the geometries of simple organocopper compounds
and for the energetics of stationary points in the MeCu and
Me2Cu- additions. Overall, the B3LYP geometry reproduces
the MP2 geometry, and the B3LYP energetics is closer to the
CCSD(T) data than the MP2 energetics.
1. Comparison for Stable Structures. Comparison of the
theoretical methods for the stable structures is summarized in
(72) Armentrout, P. B.; Georgiadis, R. Polyhedron 1988, 7, 1573-1581.

Addition of Li Organocuprate Clusters to Acetylene
J. Am. Chem. Soc., Vol. 119, No. 21, 1997 4899
Table 1. Among the C-Cu bond lengths obtained for the linear
cuprate anion Me-Cu-Me-, the HF/631WH value (2.066 Å)
is 6% longer than that of the experimental data (1.935 Å),54 to
which the values at the MP2/631WH (1.929 Å), B3LYP/631A
(1.969 Å), and values in previous MP243,74 and DFT55,75 results
are close. The optimized structure of Me2CuLi‚LiCl (RT1) was
found to be very similar to those obtained for LiI, LiCN, and
MeLi complexes.11,55 The calculated structures of (Me2CuLi)2
(RT1-d) and [Me2CuLi(H2O)]2 (RT1w) were reported previ-
ously.11,43 Other structural data are summarized in Table 1,
which shows that B3LYP method reasonably reproduces the
MP2 geometry.
2. Comparison for Intermediates and TSs. The energetics
of the reaction intermediates of MeCu, Me2Cu-, and Me2CuLi‚-
LiCl additions to acetylene were also examined. The potential
energy profiles calculated at the MP2/631WH, B3LYP/631A,
and CCSD(T)/631WH (for Me2CuLi‚LiCl addition, CCSD(T)/
321A used) levels for B3LYP/631A optimized geometries are
shown in Figure 12.
sion of CuI (52.3 mg, 0.275 mmol) in dry ether (4.5 mL) was
added a 1.55 M hexane solution of n-BuLi (0.355 mL, 0.55
mmol) at -40 °C over 5 min. Resulting black solution was
stirred at -25 °C for 30 min and then cooled to -40 °C again.
Phenylacetylene (55.0 µL, 0.50 mmol) was added to the solution
of the cuprate at that temperature, and the reaction mixture was
stirred at 4 h and then quenched with saturated aqueous NH4Cl.
The reaction mixture was diluted with Et2O and passed through
a pad of silica gel. Evaporation of solvent in Vacuo afforded a
pale yellow oil. Silica gel chromatography (silica gel 2.0 g,
pentane) afforded a mixture of linear and blanched (cf. eq 2)
carbocupration products as a colorless oil (53.3 mg. 0.28 mmol,
56% yield).
Carbocupration in the Presence of 10 Equiv of 12-crown-
4. To a suspension of CuI (31.4 mg, 0.165 mmol) in dry ether
(2.7 mL) was added a 1.55 M hexane solution of n-BuLi (0.215
mL, 0.167 mmol) at -40 °C over 5 min. Resulting black
solution was stirred at -25 °C for 30 min and 12-crown-4 (0.53
mL, 3.3 mmol) was added dropwise. The resulting charcoal
gray suspension was stirred for an additional 30 min and then
cooled to -40 °C again. Phenylacetylene (32.9 µL, 0.3 mmol)
was added to the suspension at that temperature, and the reaction
mixture was stirred for 4 h and then quenched with saturated
NH4Cl. After workup as mentioned above, silica gel chroma-
tography (silica gel 1.0 g, pentane) afforded a 85:15 mixture of
linear and blanched carbocupration product as a colorless oil
(5.9 mg, 0.031 mmol, 12.3% yield). Upon running the reaction
further at -20 to 0 °C for 18 h, we obtained the product in
53% yield. The rate retardation effect of the crown ether was
also observed, when n-BuLi was treated first with the crown
ether before preparation of the cuprate reagent. Physical data
for the carbocupration product: IR (neat, a 91:9 mixture of regio
isomers) 2656.3, 2927.4, 2861.8, 1625.7, 1492.6, 1456.0, 1378.9,
It can be seen that the MP2 complexation energies tend to
be larger than the CCSD(T) ones and that the B3LYP ones
smaller than the latter. Reverse situation is found for the
energies of the TSs. Hence the MP2 activation energies are
always larger than the B3LYP ones.
3. HF Studies on the Reaction of Me2CuLi‚LiCl. Studies
on the cuprate reactivities in the Me2CuLi‚LiCl reaction (Figure
13) were performed at the HF level to find that this level of
calculations is not suitable for the studies of the TSs.
The TS of C-C bond formation (TS1) was characterized by
its single negative frequency (532.2i cm-1) corresponding to
the C2-C3 bond formation (Figure 13). The C1-Cu-C4 bond
is in a near linear arrangement and the C3-Cu bond is totally
cleaved. The TS hence represents the TS for direct insertion
of acetylene to the C-Cu bond. Going down the energy hill
along the IRC, the TS went smoothly to the desired product
PD with 86.7 kcal/mol exothermicity, which confirms that TS1
lies on the HF potential surface leading to C-C bond formation.
On the other hand, we followed the IRC from TS toward the
reactants and could not find such a π-complex as found in the
higher level calculations.
The vinyl cuprate product PD is a mixed cluster composed
of MeLi, LiCl, and CH3CHdCHCu and is similar to RT1 for
the fundamental cyclic structure. The HF structure was found
to be very similar to that found at the MP2 level (vide infra).
The energy difference between the reactants and TS1 is
unrealistically high (40.5 kcal/mol, HF/631WH//HF/631WH)
and even increased to 42.3 kcal/mol at the MP2/631WH//HF/
631WH level, indicating that TS1 cannot exist on the MP2
potential surface. We also searched for other possibilities to
find the second TS (TS2) which is a TS of acetylene insertion
into the weaker Me-Li bond leading to formal MeLi addition.
The energy of the TS is 36.0 kcal/mol (HF/631WH//HF/
631WH), which is still too high.
1
1261.2, 1095.4, 1025.9, 894.8, 777.1, 700.0; H NMR (400
MHz, CDCl3, major isomer) δ 0.89 (t, J ) 7.08 Hz, 3 H), 1.29-
1.51 (m, 4 H), 2.50 (br t, J ) 7.57 Hz, 2 H), 5.05 (br s, 1 H),
5.25 (br s, 1 H), 7.2-7.5 (m, 5 H); 1H NMR (400 MHz, CDCl3,
minor isomer); δ 0.92 (t, J ) 7.0 Hz, 3 H), 1.29-1.51 (m, 4
H), 2.20 (dt, J ) 7.0, 14.5 Hz, 2 H), 6.22 (dt, J ) 6.9, 16.0 Hz,
1 H), 6.37 (distorted br d, J ) 16.0 Hz, 1 H), 7.2-7.5 (m, 5
H); 13C NMR (100 MHz, CDCl3, major isomer) δ 13.91, 22.41,
30.46, 35.07, 112.00, 126.10 (2 C), 127.20, 128.21 (2 C), 128.44,
148.76. Elemental analysis (for a 91:9 mixture of regio isomers)
calcd for C12H16: C, 89.94; H, 10.06. Found: C, 90.21; H,
9.79.
Acknowledgment. We thank Prof. B. H. Lipshutz for helpful
discussion and Dr. M. Isaka for provision of some experimental
data. Financial support from the Ministry of Education, Science
and Culture, Japan to E.N. and from the National Science
Foundation to K.M. is gratefully acknowledged. E.N. thanks
the generous allotment of computer time from the Institute for
Molecular Science, Japan. S.M. thanks JSPS for a predoctoral
fellowship.
In summary, both ab initio MP2 and density functional
B3LYP methods give reasonable geometric data, yet the MP2
method overestimates the activation energy of the C-C bond
formation where the nature of the copper/acetylene complexation
changes dramatically.
Experimental Details for Eq 2. Carbocupration in the
Absence of 12-crown-4 (a Control Experiment). To a suspen-
Supporting Information Available: Cartesian coordinates
of stationary points in the MeCu, Me2Cu-, and Me2CuLi
addition (B3LYP/631A), and CP, TS, PD, CP-d, and TS-d at
the MP2/631WH and B3LYP/631A levels (10 pages). See any
current masthead page for ordering and Internet access
instructions.
(73) Anderson, R. J.; Hendrick, C. A.; Siddall, J. B. J. Am. Chem. Soc.
1970, 92, 735-737. Anderson, R. J. J. Am. Chem. Soc. 1970, 92, 4978-
4979. Herr, R. W.; Johnson, C. R. J. Am. Chem. Soc. 1970, 92, 4979-
4981.
(74) Antes, I.; Frenking, G. Organometallics 1995, 14, 4263-4268.
(75) Sosa, C.; Andzelm, J.; Elkin, B. C.; Wimmer, E.; Dobbs, K. D.;
Dixon, D. A. J. Phys. Chem. 1992, 96, 6630-6636.
JA964208P