An unexpected transmetalation intermediate: Isolation and structural characterization of a solely CH3 bridged di-copper(i) complex

Article abstract of DOI:10.1039/c6cc01510g

Structural characterization of unsupported, two metal centres bridging methyl groups is rare. They have been proposed as transmetalation intermediates in cuprate chemistry, but as yet no structural evidence has been presented. We have isolated a di-copper(i) complex with solely a methyl ligand bridging two Cu(i) atoms, representing a new bonding mode of CH₃.

Full text of DOI:10.1039/c6cc01510g

ChemComm
View Article Online
View Journal
COMMUNICATION
An unexpected transmetalation intermediate:
isolation and structural characterization of a
Cite this:DOI: 10.1039/c6cc01510g
solely CH3 bridged di-copper(I) complex†‡
Received 18th February 2016,
Accepted 25th February 2016
Roberto Molteni,a Rudiger Bertermann,a Katharina Edkinsb and Andreas Steffen*a
¨
DOI: 10.1039/c6cc01510g
www.rsc.org/chemcomm
Structural characterization of unsupported, two metal centres bridging and mass spectrometric studies have been undertaken to under-
methyl groups is rare. They have been proposed as transmetalation stand the solution behaviour of Gilman cuprates and the alkylation
intermediates in cuprate chemistry, but as yet no structural evidence reaction mechanisms, which is still an active research field.5,6
has been presented. We have isolated a di-copper(I) complex with
However, the reactivity of MeCu has also been tamed by
solely a methyl ligand bridging two Cu(I) atoms, representing a new coordination of various ligands, such as phosphines7 or N-hetero-
bonding mode of CH3.
cyclic carbenes (NHCs),8 which now allows its use as a co-catalyst
for acrylonitrile polymerization9 as well as for E–H bond activation
Organocopper(I) reagents are highly important transmetalation reactions.10 In analogy to Gilman cuprates, the solution behaviour
agents in catalytic cross-coupling reactions, and homo-organo- of phosphine stabilized MeCu is extremely complex, including
cuprates [CuR2]À (R = alkyl, aryl) as their anionic counterparts, a wide range of equilibria involving naked CuMe, {CuMe}n
so-called Gilman reagents, represent the most frequently used oligomers and [CuMe(PR3)n] complexes in various stoichiometries.
transition metal reagents in C–C bond forming reactions in On the basis of in situ NMR spectroscopic studies, dimerization of
organic synthesis.1 Thus, great efforts have been made to [CuMe(PCy3)] and subsequent methyl transfer from one Cu atom
elucidate the structures of such organocopper(I) compounds to another have also been proposed, leading to the presence of
in order to gain a deeper understanding of their reactivity, a cuprate complex in solution, but no structural evidence was
which is mandatory for the design of efficient catalytic cycles available so far.7a,c,d,9 Herein, we report a dicopper(I) complex
involving transmetalation and for control over regio- and bearing a methyl ligand bridging the two copper atoms, which
stereoselectivity of alkylation. The development in this regard has wider implications for understanding cuprate chemistry
is particularly impressive for methyl copper(I), which has originally and organocopper transmetalation reactions.
been described by the group of Gilman in the early 1950s as
[CuMe(PPh3)2] (1) was synthesized following a literature
extremely reactive in solution and potentially explosive, thus no procedure, which involves reductive alkylation of [Cu(acac)2]
structural data have been reported until recently.2 However, the with Al(OEt)Me2 in the presence of excess PPh3, giving 1 as a
formation of its lithium homocuprate, Me2CuLi, usually further yellow powder.7a,b Its further purification has been reported by
stabilized by an additive such as LiX (X = I, CN), gave an easy-to- washing the powder with diethyl ether; however, our attempts
handle methylating agent,2a,3 now widely used in organic synthesis, of purification via recrystallization of 1 led, unexpectedly, to
particularly in 1,4-addition reactions to enones.4 The adduct the isolation of single crystals of [Cu(PPh3)2(m-Me)CuMe] (2)
Me2CuLiÁLiX tends to self-aggregate in solution, and its structure (Scheme 1 and Fig. 1). Specifically, slow diffusion of diethyl
and reactivity highly depend on the solvent, the specific additive, ether into a THF solution of 1 at À30 1C gave 2 in 67% yield,
and the concentration.1c,5 Thus, a number of NMR spectroscopic while at higher temperature only decomposition of 1 is observed.
Changing the conditions of the crystallization by substituting
a
¨
¨
¨
Institut fur Anorganische Chemie, Julius-Maximilians-Universitat Wurzburg,
¨
Am Hubland, 97074 Wurzburg, Germany. E-mail: andreas.steffen@uni-wuerzburg.de
b School of Medicine, Pharmacy and Health, Durham University Queen’s Campus,
University Boulevard, Stockton-on-Tees, TS17 6BH, UK
† Dedicated to Professor Todd B. Marder on the occasion of his 60th birthday.
‡ Electronic supplementary information (ESI) available: Details of synthesis,
NMR spectroscopy, computations and X-ray diffraction, as well as experimental
and simulated NMR and IR spectra. CCDC 1430441. For ESI and crystallographic
data in CIF or other electronic format see DOI: 10.1039/c6cc01510g
Scheme 1 Conditions for isolation of 2 or 3 from dissolved 1.
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
Fig. 2 Comparison of the experimental solid state IR spectrum of 2
(black) with the calculated spectra of [Cu(PPh3)2(m-CH2)CuMe] (red) and
[Cu(PPh3)2(m-CH3)CuMe] (blue).
Fig. 1 Molecular structure of [Cu(PPh3)2(m-Me)CuMe] (2) obtained from
single crystal X-ray diffraction studies. Thermal ellipsoids drawn at the 50%
probability level; H atoms omitted for clarity.
bands above 3000 cmÀ1 (Fig. 2). IR bands in the region between
2700 and 2900 cmÀ1 are typical for a bridging CH3 group.11 The
calculated IR spectrum of [CuI(PPh3)2(m-CH2)CuIIMe] shows five
bands between 2838 and 2932 cmÀ1, but the low energy vibration
at 2710 cmÀ1 was only reproduced by [CuI(PPh3)2(m-CH3)CuIMe],
giving five bands at 2703, 2825, 2857, 2885, and 2901 cmÀ1 with
a shoulder.
diethyl ether with hexane or C6F6 as the antisolvent yields
[CuMe(PPh3)3]7d (3) instead.
According to the single crystal X-ray diffraction data, 2
exhibits a carbon unit bridging two copper atoms, which raises
the question of its identity, i.e. whether indeed a CH3 group
connects two Cu(I) atoms or whether a CH2 moiety is present,
indicating a mixed-valence complex with Cu(I) and Cu(II). Solid
state EPR measurements of a cooled sample at À20 1C gave
no magnetic answer, arguing against a paramagnetic mixed-
valence compound. Compound 2 is highly sensitive to moisture,
oxygen and temperature, and decomposes at room temperature in
solution quickly, albeit more slowly in the solid state. Nevertheless,
several attempts were necessary to perform some quick solid-state
NMR studies. The 13C{1H} CP/MAS solid-state NMR spectrum of
crystalline 2 (Fig. S11, ESI‡) gives broad signals centred at +1,
À5, À8 and À16 ppm, respectively, which can be attributed to
chemically different methyl groups, but may also arise from
partial decomposition or 1J(63,65Cu,13C) couplings. However, the
31P{1H} CP/MAS solid-state NMR spectrum of a freshly prepared
sample was acquired with 16 scans in 3 minutes (Fig. S12, ESI‡).
The spectrum shows two asymmetric quartets for the two
The identity of 2 being resolved as [Cu(PPh3)2(m-Me)CuMe],
its structure can either be understood as coordination of MeCu
to [CuMe(PPh3)2], or as binding of {Cu(PPh3)2}+ to one Cu–Me
bond of linear dimethyl cuprate {CuMe2}À. We prefer the latter
description due to the non-ideally trigonal planar coordination
geometry of Cu2 and the C1–Cu2–C2 angle of 172.1(1)1
(Table 1). In accordance with a donor–acceptor interaction
between the cuprate and cationic {Cu(PPh3)2}+, the Cu2–C2
bond (2.011(2) Å) is significantly increased compared to the
bond between Cu2 and the terminal C1 (1.924(2) Å), which is
within the range of other dimethyl cuprates.3,7c,12 The geometry
of the Cu1 atom is distorted tetrahedron, as can be seen from
the angles around that metal centre given in Table 1. The Cu1–
Cu2 distance of 2.4121(4) Å is very similar to the one found in
{[tBu2P(NSiMe3)2-k2N]Cu}2(m-CPh2) (2.4165(3) Å), a rare example
of a structurally characterized dicopper(I) complex with two
copper atoms bridged only by a m-carbene and not by other
ligands.13 However, 2 is, to the best of our knowledge, the first
dicopper(I) complex bridged only by a methyl group.
inequivalent phosphorus atoms (diso
(
31PA) E À2, diso
(
31PB) E
À4.6 ppm) in the crystal structure. The splittings between the
lines of these quartets increase to a higher field. The observed
asymmetric quartets arise from J and residual dipolar couplings
of the 31P nuclei with the two copper isotopes, 63Cu and 65Cu. A
spin system simulation without dipolar interaction can be found in
It has to be mentioned at this point that the structural proof
of di- or bimetallic complexes of the type {M(m-Me)M} with no
1
the ESI‡ (Fig. S13). In addition, the H BR24 Cramps solid-state
NMR spectrum shows two signals at 0.8 and À0.6 ppm, which can
be assigned to the different methyl moieties in 2 (Fig. S14, ESI‡).
The identity of the bridging alkyl moiety in 2 was further
determined by ATR-FTIR spectroscopy and compared with the
calculated IR spectra (BP86-D3BJ/def2-tzvp/ZORA) of a hypothetical Cu1–P1
CH2 bridged and a CH3 bridged complex. The experimental IR
Table 1 Selected structural parameters of [Cu(PPh3)2(m-Me)CuMe] (2)
Distance (Å)
Cu1–Cu2
Angle (1)
2.4121(4)
2.2580(6)
2.2618(6)
2.137(2)
2.011(2)
1.924(2)
C1–Cu2–C2
P1–Cu1–P2
172.1(1)
122.33(2)
52.05(6)
71.04(7)
56.91(6)
130.02(9)
Cu1–P2
Cu1–C2
Cu2–C2
C2–Cu1–Cu2
Cu1–C2–Cu2
C2–Cu2–Cu1
C1–Cu2–Cu1
spectrum of 2 shows five bands at 2710, 2781, 2830, 2852 and
2885 cmÀ1, the latter with a shoulder, and several overlapping Cu2–C1
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016

View Article Online
ChemComm
Communication
Fig. 4 1H–13C HSQC NMR spectra after dissolving 2 in d8-toluene (left, A)
and d8-THF (right, B) at À40 1C. The projections in the f1-direction show
the 13C-DEPT135-NMR-spectra, respectively.
Fig. 3 Bonding orbitals of the Cu–(m-Me)–Cu triangle motif in 2 obtained
from DFT calculations (BP86-D3BJ/def2-tzvp/ZORA) showing the cuprophilic
interactions.
generated by prior ligand dissociation from other 1, forming
[CuMe(PPh3)] (4). A PCy3 analogue of 4 has previously been
isolated, and the formation of [CuMe(PCy3)2] upon addition of
phosphine has been suggested, which is the inverse reaction to
our proposal.9 Further loss of phosphine could give [CuMe],
which promotes the formation of isolated [CuMe(PPh3)3] (3),
and allows association with 1 giving the isolated complex 2. An
interesting reaction is dimerization of 4, involving phosphine
transfer from one copper atom to the other. Indeed, we were able
to observe this ligand redistribution by dissolving 2 in d8-toluene
solution at À40 1C giving [CuMe(PPh3)] (4), as a 1H–13C HSQC
NMR experiment shows a cross-peak at 0.50/À6.0 ppm, typical for
monophosphine copper methyl complexes (Fig. 4).7c,8,9,17
In contrast, in d8-THF solution at the same temperature the
neutral compound [Cu(PPh3)2(m-Me)CuMe] (2) dissociates into
the ion pair [Cu(PPh3)]+/[CuMe2]À (5) as the main species, giving
rise to a 1H–13C HSQC cross-peak at À0.42/À7.5. However, further
unidentified minor CuMe compounds are also present (1H–13C:
À0.98/À15.5, 0.22/À5). The 31P{1H} NMR spectrum shows two
broad overlapping resonances at À3 and À5 ppm, which are much
broader than the one found in d8-toluene at À4.1 ppm, indicating
interconversion between these complexes (Fig. S17 and S22, ESI‡).
The proposed equilibrium between 2 and 5 has been confirmed
by re-dissolving 5 in d8-toluene, which gave an 1H–13C HSQC
spectrum identical to an original sample of 2 dissolved in the
same solvent (Fig. S19, ESI‡).
other bridging moieties is very rare. We found only one example
in which two transition metals are bridged by solely a methyl
group, i.e. [PtMe(dmpe)(m-Me)Cu(PtBu3)].14 Two other structurally
characterized compounds that are similar, MeLi and Me2Mg
¨
complexes of {Ni(C2H4)2} reported by Porschke and co-workers,
have been debated to contain additional interactions between
the alkali/earth alkali metal and one of the olefin ligands at the
nickel(0) center.15
The potential existence of monomeric cuprates of the type
[L2Cu(m-Me)CuMe] (L = OMe2, SMe2) has been proposed on the
basis of theoretical stability studies.16 An NBO analysis showed
that the calculated structures should gain their stability mainly
from the donor–acceptor interaction between one of the Cu–Me
bonds and the cationic {L2Cu}+ fragment, although cuprophilic
interactions are also present, adding to the stabilization of the
3-center-2-electron-bond.16 In line with that interpretation, the
Mayer bond order in 2 obtained from our DFT calculations is
ca. 1/3 for each of the bonds in the Cu1–Cu2–CH3 triangle,
which is mainly formed by HOMO, HOMOÀ5 and HOMOÀ11
(Fig. 3 and Fig. S8, ESI‡), the latter two involving cuprophilic
interactions.
The fact that either 2 or [CuMe(PPh3)3] (3) is obtained from
[CuMe(PPh3)2] (1), depending on the solvent used for crystallisation,
suggests an equilibrium between a number of species in solution.
This apparently involves, for copper phosphine alkyl compounds,
rare ligand redistribution, i.e. phosphine and methyl transfer
between the two copper atoms (Scheme 2). In order for 1 to
form 3, phosphine association is necessary, which can only be
The isolation and structural characterization of the inter-
mediate [Cu(PPh3)2(m-Me)CuMe] (2) provides a nice snapshot of
the above described methyl and phosphine ligand redistribution
equilibrium. Several implications arise from our findings. The
employment of copper alkyl phosphine complexes allows for-
mation of a variety of species, which can potentially participate
in the reaction of interest. For instance, both cuprates as well as
free phosphines formed from [CuMe(PCy3)] can initiate anionic
acrylonitrile polymerization, indicating that the original compound
as such is not involved in the polymerization reaction.9
Furthermore, bimetallic intermediates, which are stabilized
by weak metallophilic interactions, play an important role in
cooperative bimetallic catalysis. Organic group transfer solely
supported by labile d8–d10 bonds has been proposed mainly on
the basis of kinetic and DFT studies for Sonogashira and Stille
Scheme 2 Possible equilibrium reactions arising from dissolution of 1
leading to the isolation of 2 and 3, and to the observation of 4 and 5.
This journal is ©The Royal Society of Chemistry 2016
Chem. Commun.

View Article Online
Communication
ChemComm
cross-coupling reactions co-catalyzed by Cu(I) or Au(I),18 and the
Negishi coupling reaction is also thought to benefit from Pd(II)–
Zn(II) bond formation.19 Very recently, coupling of alkynes
mediated by dual gold catalysis has been proposed to involve
a di-gold(I) key complex exhibiting d10–d10 aurophilic interactions.20
Despite the fact that aurophilic interactions are much stronger than
cuprophilic interactions (15 vs. 4 kcal molÀ1),21 our findings suggest
that the latter can also foster ligand redistribution and organic
group transfer.
Angew. Chem., Int. Ed., 2002, 41, 3263; (e) X. Y. Yang and P. Knochel,
Synlett, 2004, 81–84.
2 (a) H. Gilman, R. G. Jones and L. A. Woods, J. Org. Chem., 1952, 17,
1630–1634; (b) D. B. Grotjahn, D. T. Halfen, L. M. Ziurys and
A. L. Cooksy, J. Am. Chem. Soc., 2004, 126, 12621–12627.
3 H. Hope, M. M. Olmstead, P. P. Power, J. Sandell and X. Xu, J. Am.
Chem. Soc., 1985, 107, 4337–4338.
¨
4 A. Alexakis, J. E. Backvall, N. Krause, O. Pamies and M. Dieguez,
Chem. Rev., 2008, 108, 2796–2823.
5 R. M. Gschwind, Chem. Rev., 2008, 108, 3029–3053.
6 (a) S. H. Bertz, R. A. Hardin and C. A. Ogle, J. Am. Chem. Soc., 2013,
135, 9656–9658; (b) S. H. Bertz, R. A. Hardin, M. D. Murphy and
C. A. Ogle, Chem. Commun., 2013, 49, 3010–3012; (c) M. D. Murphy,
C. A. Ogle and S. H. Bertz, Chem. Commun., 2005, 854–856; (d) M.
Neumeier and R. M. Gschwind, J. Am. Chem. Soc., 2014, 136,
5765–5772; (e) W. Henze, T. Gartner and R. M. Gschwind, J. Am. Chem.
Soc., 2008, 130, 13718–13726; ( f ) A. Putau and K. Koszinowski,
Organometallics, 2011, 30, 4771–4778; (g) A. Putau, M. Wilken and
K. Koszinowski, Chem. – Eur. J., 2013, 19, 10992–10999.
7 (a) A. Miyashita and A. Yamamoto, Bull. Chem. Soc. Jpn., 1977, 50,
1102–1108; (b) A. Miyashita, T. Yamamoto and A. Yamamoto, Bull.
Chem. Soc. Jpn., 1977, 50, 1109–1117; (c) D. F. Dempsey and G. S.
Girolami, Organometallics, 1988, 7, 1208–1213; (d) P. S. Coan, K. Folting,
J. C. Huffman and K. G. Caulton, Organometallics, 1989, 8, 2724–2728.
8 N. P. Mankad, T. G. Gray, D. S. Laitar and J. P. Sadighi, Organo-
metallics, 2004, 23, 1191–1193.
9 F. Schaper, S. R. Foley and R. F. Jordan, J. Am. Chem. Soc., 2004, 126,
2114–2124.
10 L. A. Goj, E. D. Blue, C. Munro-Leighton, T. B. Gunnoe and J. L. Petersen,
Inorg. Chem., 2005, 44, 8647–8649.
11 (a) E. Diana, O. Gambino, R. Rossetti, P. L. Stanghellini, T. Albiez,
W. Bernhardt and H. Vahrenkamp, Spectrochim. Acta, Part A, 1993,
49, 1247–1259; (b) P. Krohmer and J. Goubeau, Z. Anorg. Allg. Chem.,
1969, 369, 238–248.
12 S. H. Bertz, R. A. Hardin, T. J. Heavey and C. A. Ogle, Angew. Chem.,
Int. Ed., 2013, 52, 10250–10252.
A typical reaction of dimethylcuprate is alkylation of a,b-
unsaturated ketones, such as 3-methyl-cyclo-2-hexen-1-one, which
reacts with Gilman’s reagent within 12 hours in almost quantitative
yield.4 In contrast, 2 shows only 50% conversion to give 3-dimethyl-
cyclohexanone at room temperature as well as at À30 1C within
8 hours, partially due to decomposition. The degree of association
between [CuMe2]À and its counterion was shown to be very
important for the reactivity,1c,5 and that the alkylation reaction
of 4-methyl-cyclo-2-hexen-1-one stops upon adding 15-crown-5.22
The formation of a p-complex between, e.g., Me2CuLi, either as a
monomer or as a contact ion pair, and the substrate is a crucial
step, involving simultaneous coordination of Li+ to the enone
carbonyl oxygen and the cuprate.6c,23 In addition, theoretical
studies suggest that bending of linear [CuMe2]À by a coordination
partner is crucial for weakening the Cu–Me bond and would
therefore increase the reactivity.24 The lower activity of 2 compared
to Gilman’s reagent in the methylation of enones can thus be
explained with a much weaker interaction in solution between
dimethylcuprate and [Cu(PPh3)2]+ than with Li+, which is due to
the higher steric demand for the copper phosphine complex
and its higher stability in solution as an isolated cation.
In conclusion, we have isolated and structurally characterized
the first dicopper(I) complex, [Cu(PPh3)2(m-Me)CuMe] (2), in which
two metal centres are solely bridged by a methyl group and
experience stabilizing metallophilic interactions. Apart from 2,
only one Pt–Cu complex as an example of transition metal {M(m-
Me)M} compounds without further bridging ligands exists.14
Complex 2 is obtained as a result of phosphine ligand redistribution
and represents a transmetalation intermediate on the way from
[CuMe(PPh3)2] (1) to [Cu(PPh3)2]+[CuMe2]À (5), which we observed in
solution. Thus, we were able to show a new bonding mode of CH3
and provide structural evidence for previously proposed solution
equilibria, giving further insight into transmetalation reactions
of organocopper compounds.
13 P. Hofmann, I. V. Shishkov and F. Rominger, Inorg. Chem., 2008, 47,
11755–11762.
14 M.-E. Moret, D. Serra, A. Bach and P. Chen, Angew. Chem., Int. Ed.,
2010, 49, 2873–2877.
¨
15 (a) W. Kaschube, K. R. Porschke, K. Angermund, C. Kru¨ger and
¨
G. Wilke, Chem. Ber./Recl., 1988, 121, 1921–1929; (b) K. R. Porschke,
K. Jonas, G. Wilke, R. Benn, R. Mynott, R. Goddard and C. Kru¨ger,
Chem. Ber./Recl., 1985, 118, 275–297.
16 P. M. Polestshuk, P. I. Dem’yanov and V. S. Petrosyan, Russ. J. Phys.
Chem., 2009, 83, 1913–1923.
17 T. A. Mobley, F. Mu
¨ller and S. Berger, J. Am. Chem. Soc., 1998, 120,
1333–1334.
18 (a) M. H. Perez-Temprano, J. A. Casares and P. Espinet, Chem. – Eur.
J., 2012, 18, 1864–1884; (b) J. del Pozo, D. Carrasco, M. H. Perez-
Temprano, M. Garcia-Melchor, R. Alvarez, J. A. Casares and P. Espinet,
Angew. Chem., Int. Ed., 2013, 52, 2189–2193; (c) D. Carrasco, M. H. Perez-
Temprano, J. A. Casares and P. Espinet, Organometallics, 2014, 33,
3540–3545.
19 J. del Pozo, E. Gioria, J. A. Casares, R. Alvarez and P. Espinet,
Organometallics, 2015, 34, 3120–3128.
20 M. H. Larsen, K. N. Houk and A. S. K. Hashmi, J. Am. Chem. Soc.,
2015, 137, 10668–10676.
We thank the Deutsche Forschungsgemeinschaft (DFG, STE
1834/4-1) for funding. A. S. is grateful to Prof. T. B. Marder for
his generous support.
¨
21 (a) P. Pyykko, Chem. Rev., 1997, 97, 597–636; (b) H. Schmidbaur and
A. Schier, Chem. Soc. Rev., 2012, 41, 370–412.
22 C. Ouannes, G. Dressaire and Y. Langlois, Tetrahedron Lett., 1977,
815–818.
23 (a) M. John, C. Auel, C. Behrens, M. Marsch, K. Harms, F. Bosold,
R. M. Gschwind, P. R. Rajamohanan and G. Boche, Chem. – Eur. J.,
2000, 6, 3060–3068; (b) 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.
Notes and references
1 (a) Z. Rappoport and I. Marek, The Chemistry of Organocopper Compounds,
Wiley, Chichester, Hoboken, NJ, 2009; (b) N. Krause, Modern Organocopper
Chemistry, Wiley-VCH, Weinheim, 2002; (c) N. Yoshikai and E. Nakamura,
Chem. Rev., 2012, 112, 2339–2372; (d) C. Piazza and P. Knochel, 24 S. Mori and E. Nakamura, Tetrahedron Lett., 1999, 40, 5319–5322.
Chem. Commun.
This journal is ©The Royal Society of Chemistry 2016