Complexes of the gilman reagent with double bonds across the π-σ Continuum

Article abstract of DOI:10.1021/om300622c

By using rapid injection NMR at low temperatures, a variety of π-complexes of lithium dimethylcuprate(I) with C-C, C-N, and C-S double bonds have been prepared and characterized. Complexation is generally accompanied by large upfield changes in chemical shift for the substrate carbon atoms bonded to copper. In the case of α,β-unsaturated carbonyl compounds, the changes for the carbonyl carbons are much smaller in magnitude, which is consistent with the usual η² representation of these structures. It is possible for one ligand to displace another, and in this way an order of stability can be elucidated. Treatment of selected π-complexes with chlorosilanes or cyanosilanes gives Cu^III intermediates.

Full text of DOI:10.1021/om300622c

Article
pubs.acs.org/Organometallics
Complexes of the Gilman Reagent with Double Bonds across
the π−σ Continuum
Steven H. Bertz,* Stephen K. Cope,† Richard A. Hardin, Michael D. Murphy, Craig A. Ogle,*
David T. Smith, Andy A. Thomas,‡ and Tara N. Whaley
Department of Chemistry, University of North CarolinaCharlotte, Charlotte, North Carolina 28223, United States
S
* Supporting Information
ABSTRACT: By using rapid injection NMR at low temperatures, a
variety of π-complexes of lithium dimethylcuprate(I) with C−C, C−
N, and C−S double bonds have been prepared and characterized.
Complexation is generally accompanied by large upfield changes in
chemical shift for the substrate carbon atoms bonded to copper. In
the case of α,β-unsaturated carbonyl compounds, the changes for the
carbonyl carbons are much smaller in magnitude, which is consistent
with the usual η2 representation of these structures. It is possible for one ligand to displace another, and in this way an order of
stability can be elucidated. Treatment of selected π-complexes with chlorosilanes or cyanosilanes gives CuIII intermediates.
β-carbon of 1b affords 1d directly. While he established the role
of lithium−carbonyl complexes in the conjugate addition
mechanism, House dismissed the idea of cuprate−olefin
complexes on the basis of Occam’s razor.
INTRODUCTION
■
The history of organocopper chemistry has been replete with
controversies, as the questions that arose were important and
difficult. One of the earliest concerned the mechanism of the
conjugate addition (1,4-addition) reaction of organocuprate(I)
compounds (“Gilman reagents”). House championed a single-
electron transfer (SET) mechanism wherein an electron is
shuttled from the cuprate, R2CuLi, to the α,β-unsaturated
carbonyl compound (e.g., 1a, Scheme 1a), which is activated by
Li+ complexation.1 The resultant radical anion 1b then
combines with the oxidized copper species to form CuIII
intermediate 1c, which undergoes reductive elimination to
enolate 1d. Alternatively, transfer of the R-radical to the
The prospect of intermediate π-complexes rose like the
phoenix when a group at Chalmers University of Technology
prepared and characterized cinnamate−dimethylcuprate com-
plexes in a seminal study.2 Several years later, a Bell
Laboratories group studied lithium−carbonyl and copper−
olefin complexes in the reaction of Me2CuLi with 10-methyl-
Δ1,9-2-octalone, a highly hindered 2-cyclohexenone derivative.3
They proposed that aggregation might be responsible for
multiple π-complexes, which was confirmed by more detailed
NMR studies.4 German chemists extended the Swedish work
on cinnamate esters by replacing the phenyl with an alkynyl
group.5 A breakthrough was achieved by using rapid injection
NMR techniques:6 it was possible to prepare the Me2CuLi
complex of the highly reactive parent compound, 2-cyclo-
hexenone (vide supra).7 A number of π-complexes of activated
olefins with homocuprates,3,5,7 as well as mixed cuprates,8 have
now been reported.
Scheme 1a. Electron Transfer Mechanisms for
Organocuprate Conjugate Addition
Mechanisms involving π-complexes in the conjugate addition of
organocuprates to α,β-unsaturated carbonyl compounds are shown
in Scheme 1b. Cuprate−olefin π-complex 1 (e.g., R = Me, Chart 1)
can undergo oxidative addition to give CuIII intermediate 1c, which
advances via reductive elimination to enolate 1d. Alternatively, the
Berlan mechanism features a simple olefin insertion reaction to
afford α-cupriocarbonyl intermediate 1e, which progresses via trans-
metallative rearrangement to enolate 1d.9 Both η1 (σ-allyl10−12
)
and η3 (π-allyl11,12) d8 CuIII intermediates have been prepared,
Special Issue: Copper Organometallic Chemistry
Received: July 4, 2012
Published: November 1, 2012
© 2012 American Chemical Society
7827
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
and their structures are based on a square-planar geometry,
rather than the trigonal one envisioned by earlier authors (cf.
Schemes 1a and 1b).
Scheme 1b. π-Complex Mechanisms for Organocuprate
Conjugate Addition
Herein we report the details of our investigations into
complexes of the Gilman reagent with C−C, C−N, and C−S
double bonds, which reveal their structures and stabilities and
inter alia help refine the mechanistic schemes. We describe the
preparation of these complexes using rapid injection method-
ology, their structural characterization using 1D and 2D NMR,
and the exploration of their reactivity patterns.
RESULTS AND DISCUSSION
■
While it is possible to prepare relatively stable cuprate−olefin
complexes by injecting the substrate from a gastight syringe
into a solution of organocuprate in an NMR tube, cooled in a
dry ice/acetone bath, and then transferring the tube to the
precooled probe of an NMR spectrometer,3 a much more
elegant and effective method is now available. Rapid injection
NMR makes it possible to prepare air-sensitive and thermally
unstable species in an NMR tube as it spins in a spectrometer
probe at a controlled temperature under a flow of dry nitrogen.6
This technique minimizes side reactions that may occur during
the preparation of delicate species such as cuprate−olefin
π-complexes or CuIII intermediates. Even when a species is very
unstable, it is possible to put limits on the rate of reaction. For
example, when we injected methyl vinyl ketone into a solution
of lithium methyl(trimethylsilylmethyl)cuprate(I), no π-com-
plex between the reactants was observed, and the product
appeared during the first few seconds.8
Chart 1. Complexes from Unsaturated Carbonyl
Compounds
Unless noted otherwise, all reactions reported here were run
in freshly distilled deuterated tetrahydrofuran (THF-d8) by
injecting a solution of substrate into a solution of cuprate at
−100 °C, as described above. Organocuprate reactions are very
sensitive to experimental variables, which must be carefully
controlled.13 For example, virgin NMR tubes are sine qua non
for experiments with cuprate−olefin π-complexes (see Exper-
imental Section).
Tetrahydrofuran was chosen as the solvent in order to minimize
the complexity of the systems under scrutiny, in terms of both the
copper reagents and their complexes. A “side-by-side” comparison
of the reactions of Bu6Li with CuC15N in (deuterated) ether
versus THF was conducted by using 1H, 6Li, 13C, and 15N NMR
spectroscopy, and the spectra were in general much more complex
in ether, owing to higher degrees of aggregation.14 Spectra of
cuprate−olefin complexes were also more complex in ether.4
Finally, the fact that subsequent reactions of the π-complexes are
much faster in ether limits the time available for characterization in
this solvent, especially for 2D NMR methods. (N.B., Gschwind
has written an exceptionally detailed review.4)
The carbon atoms, Cα and Cβ, and associated hydrogen
atoms, Hα and Hβ, at the termini of a conjugated double bond
(DB) are subsumed under DBα and DBβ, respectively, and
methyl groups attached to copper, MeCu, are further specified as
Meα and Meβ, where the former is closer to DBα and the latter
to DBβ. Calculations support an approximately square-planar
arrangement of Cα, Cβ, Meα, and Meβ.15
Unsaturated Carbonyl Complexes. Early studies of
organocuprate π-complexes involved substrates with electronic
or steric stabilization (cf. Introduction). With our rapid
injection technology, we have been able to explore much
more structural diversity, for example, π-complexes formed
rapidly at −100 °C between Me2CuLi and α,β-unsaturated
aldehydes (1−3), ketones (4−10), and esters (11−17), as
7828
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
a b
,
1
Table 1. H and 13C NMR Data for Complexes from Unsaturated Carbonyl Compounds
ID
MeαH
MeαC
MeβH
MeβC
DBαH
DBαC
DBβH
DBβC
CO
Δα
Δβ
ΔCO
1
−0.90
−0.98
−0.62
−0.73
−1.07
−0.98
−1.20
−1.11
−1.12
−1.16
−1.12
−1.11
−1.15
−1.26
−0.87
−0.79
−1.05
−1.12
−1.16
−0.93
−0.83
−0.51
−0.29
−0.54
−0.53
−0.19
0.00
−1.60
−11.68
−1.06
−4.67
−3.05
−2.44
−4.12
−5.02
−5.56
−7.08
−5.30
−6.04
−4.47
−5.32
−3.55
−3.62
−8.55
−7.90
−7.53
−3.38
−5.30
−3.61
−5.26
−5.83
−2.31
−0.03
−0.21
−0.41
−0.37
−0.33
−0.13
−0.31
−0.42
−0.10
−0.24
−0.22
−0.22
−0.38
−0.26
−0.59
−0.41
−0.45
−0.54
−0.48
−0.54
−0.57
−0.51
−0.29
−0.49
−0.48
−0.19
−5.81
1.30
3.76
3.76
4.36
4.49
3.56
3.60

85.52
87.67
75.90
81.19
74.08
78.27
83.42
81.54
77.45
75.82
75.87
82.06
80.42
78.70
71.54
75.26
69.38
52.94
54.19
50.47
67.75
50.94
50.12
51.04
56.43
2.09, 2.74
2.59, 
4.22, 
3.71, 
2.41, 2.12
2.29, 2.24
2.55, 2.29
2.74, 2.22
, 3.26
, 3.19
, 3.11
, 
47.18
60.77
62.95
65.99
43.78
44.22
51.89
51.45
61.50
61.50
61.65
70.73
70.59
58.80
63.10
63.85
64.20
42.20
42.76
58.65
63.59
50.94
50.12
51.04
50.87
52.26
171.90
167.90
170.45
175.08
185.74
181.73
185.59
183.02
194.75
193.34
193.85
191.70
191.44
177.13
184.53
181.54
178.99
175.32
174.50
169.38
168.22
174.73
175.42
174.92
172.40
171.57
−50.94
−45.67
−53.42
−48.13
−64.41
−60.22
−61.61
−63.49
−52.67
−54.30
−54.25
−40.60
−42.24
−47.72
−56.39
−52.67
−51.62
−75.79
−74.54
−65.26
−63.18
−82.77
−80.82
−83.28
−82.00
−85.41
−88.86
−91.71
−91.44
−88.40
−86.89
−86.45
−75.19
−75.63
−90.15
−90.15
−90.00
−93.37
−93.51
−86.78
−81.54
−80.79
−80.86
−87.71
−87.15
−84.09
−75.26
−82.80
−80.82
−83.28
−79.57
−85.41
−21.40
−24.08
−24.25
−19.62
−13.31
−17.32
−13.92
−16.49
−3.90
−5.31
−4.80
−7.46
−7.72
−11.08
−14.27
−17.26
−9.41
10.80
2
cd
,
3a
3b
4a
4b
5a
5b
6a
6b
4.52
cd
,
5.20
−7.88
−6.42
−10.02
−9.93
−0.57
−1.85
1.56

3.77
3.68
3.76
3.70
3.63
4.30
4.36
4.53
5.13
3.02
2.67
3.33

e
6c
7a
5.22
7b
2.99
, 
c
8
0.47
3.29, 
4.14, 
4.09, 
4.40, 
2.16, 2.01
2.26, 2.06
, 4.10
4.04, 
3.35, 
 2.88
c
9a
9b
10
1.89
c
c
4.11
3.99
11a
11b
12
−8.22
−6.71
2.06
9.98
9.74
13
1.83
2.41
14
−5.30
−3.61
−5.04
−3.93
−2.31
3.35
2.88
3.49
3.88
3.67
9.67
15
9.38
f
16
3.49, 
3.50, 
 3.67
9.69
g
17
10.90
18
52.26
b
4.75
a
d
g
c
All shifts are in ppm, measured at −100 °C unless otherwise noted. Cuprates prepared from CuI unless otherwise noted. Cuprate from CuCN.
e
f
Measured at −70 °C (too broad at −100 °C). Cuprate (Me2CuMgCl·MgCl2) from CuCl. Second CO peak at 172.32 ppm (ΔCO = 7.09 ppm).
Second CO peak at 173.39 ppm (ΔCO = 7.61 ppm).
illustrated in Chart 1. Parent compounds and substituted
derivatives were prepared, and both cyclic and acyclic ketones
and esters were investigated. The preparations of complexes
16−18 that contained potentially reactive functional groups were
carried out via direct and indirect routes, as described in detail in
a later section. Important NMR shifts are listed in Table 1.
A π-complex of a cyclic ketone (6, 7), ester (12), or
anhydride (18), where the double bond is endocyclic, is
constrained in the s-trans conformation. On the other hand, a
complex of a cyclic substrate such as 13, where the double bond
is exocyclic, is fixed in the s-cis conformation. In cases such as
these, the presence of multiple π-complexes is assumed to be
the result of aggregation, as with 6a/6b.7 Addition of LiI to the
mixture increased the proportion of minor species 6b;
therefore, we considered it to be an aggregate with LiI and
assumed 6a to be LiI-free.
2-Cyclohexenone gave two π-complexes with lithiocuprate
Me2CuLi·LiI or Me2CuLi·LiCN,7 but only one π-complex, 6c,
with a magnesiocuprate, Me2CuMgCl·MgCl2. The NMR data
for 6c do not indicate an aggregation state. It is not a stretch to
propose that magnesiocuprate π-complexes such as 6c are
intermediates in the Kharasch reaction of α-enones with Grignard
reagents, where a catalytic amount of copper salt suffices to change
the regioselectivity from 1,2- to 1,4-addition.16
Figure 1. HMBC plot for crotonophenone π-complex 8, correlating
H's of cuprate Meα (−1.26) and Meβ (−0.26) with substrate Cβ (58.80
ppm). The H's of Meα also correlate with the C of Meβ.
Figure 2. NOESY plot for crotonophenone π-complex 8, showing
strong interactions of the H's of cuprate Meβ (−0.26) with substrate
methyl H's (1.35) and Hβ (3.29 ppm).
7829
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
Acyclic substrates with more than one π-complex can in
principle be the result of differences in conformation or
aggregation (or both). In the case of 1, NOESY established
that the conformation of the acrolein moiety was s-trans, as
there was a strong cross-peak between the aldehydic hydrogen
and one of the β-hydrogens, but no interaction between the
former and the α-hydrogen. In the case of cinnamaldehyde,
NOESY indicated a mixture (ca. 1:1) of s-cis and s-trans com-
plexes 3a and 3b, respectively. In contrast, methyl vinyl ketone
complexes 4a and 4b both had the s-cis conformation, as did
both benzalacetone complexes 9a and 9b, which implied aggre-
gation with the lithium halide or pseudohalide (LiI or LiCN,
respectively). Exchange (ROESY) in the case of 11a and 11b
precluded using NOE to assign stereochemistry.
It must be emphasized that we do not have direct evidence
for the structures of the aggregates, and in this regard we
depend upon the experimental studies of the Gschwind
group,4 as well as theoretical work by the Nakamura group.17
In the examples reported here, the aggregates were usually
present to the extent of 10−30%. In the case of benzalacetone,
as many as six complexes were observed at −100 °C; how-
ever, most of them could not be characterized completely,
owing to low concentrations. Upon warming to −80 °C, two
significant species remained, and they persisted upon cooling
back to −100 °C, where they were characterized. In the two
systems studied in detail, 6a/6b and 9a/9b, the results were
very similar using Me2CuLi prepared from either CuI or
CuCN, and only minor differences in chemical shifts were
noted.
Typical 2D NMR plots are shown in Figures 1−4. HMBC
and NOESY were especially valuable because they proved that
MeCu and certain substrate C or H atoms were integral parts of
the same molecule. Furthermore, NOESY proved that the
MeαCuMeβ moiety was on the least hindered face of the
substrate in 13 (cf. Figure 4) and 36 (vide infra). The α and β
positions of 17 were assigned with the aid of HMBC. They
were not resolved in 14−16 and 18.
Figure 4. NOESY plot for π-complex 13, showing strong H−H
interactions between cuprate Meα (−0.83) and ring O2CH (5.44) and
between cuprate Meβ (−0.57) and substrate Hβ (4.04 ppm).
interesting to compare values of ΔCO for the carbonyl groups.
The methyl groups on copper in the complexes can be
compared directly, since the chemical shifts for the starting
cuprates do not change significantly (13C NMR ca. −10 ppm,
1H NMR ca. −1.4 ppm).
The chemical shifts of the olefinic carbon atoms, Cα and
Cβ, are sent dramatically upfield upon coordination by copper,
especially Cβ. All of the upfield shift changes for Cα fall in the
range Δα ≈ −60 30 ppm (ca. mean twice the standard
deviation), and those for Cβ in the range Δβ ≈ −85 10 ppm.
The values of ΔCO populate the ranges −25 to −19 ppm for
aldehydes, −18 to −3 ppm for ketones, and 2 to 11 ppm
for esters. The magnitudes (absolute values) of Δα and Δβ
are much larger than those of ΔCO (|Δβ| > |Δα| ≫ |ΔCO|),
which leads us to conclude that 1−18 are all η2- and not
η3-complexes.18
Unsaturated Nitriles and Sulfones. Nitriles and sulfones
with conjugated C−C double bonds also formed π-complexes.
A mononitrile (19), a 1,2-dinitrile (20), and four 1,1-dinitriles
(21−24) were examined, and all of them gave π-complexes that
were stable enough to characterize. The HMBC plot for
benzylidenemalononitrile π-complex 22 in Figure 5 helped
locate Cα, and it also proved that the dimethylcuprate and
malononitrile moieties were part of the same molecule.
Phenyl vinyl sulfone formed a stable π-complex (25), as
did its β-chloro and β-methyl derivatives. Injection of phenyl
A useful parameter for understanding the change in bonding
upon complexation is the difference in chemical shift, ΔN = δN,C
− δN,U, where δN,C is the chemical shift of nucleus N in complex
C, and δN,U is the chemical shift of the corresponding nucleus in
uncomplexed substrate U. We focus on the values of ΔN for the
carbon atoms attached to copper, especially Δα and Δβ for the
carbon atoms Cα and Cβ of the double bonds. It is also
Figure 3. HMBC plot for π-complex 13, correlating substrate Hβ
(4.04) with cuprate Meα C (−3.38), ring allylic C (72.87), Ph ipso-C
(128.96), and carbonyl C (168.22 ppm).
Figure 5. HMBC plot for π-complex 22, correlating substrate Hβ
(3.44) with cuprate Meα C (−7.63), substrate Cα (17.58), both CN
carbons (120.56, 120.77), and o-Ph C atoms (126.23 ppm).
7830
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
(E)-2-chlorovinyl sulfone (Scheme 2) into 1 equiv of
Me2CuLi·LiI at −100 °C gave π-complex 26, which went on
to phenyl (E)-propenyl sulfone upon warming to −40 °C. In
the presence of a second equivalent of Me2CuLi·LiI, this
product gave π-complex 27 at this temperature. Chart 2
summarizes the structures and Table 2 the NMR data for the
unsaturated nitriles and sulfones.
The upfield chemical shift changes, Δα ≈ −65 10 ppm, for
α,β-unsaturated nitrile complexes are approximately the same
as for α,β-unsaturated carbonyl complexes, and the magnitudes
of Δβ ≈ −110 20 ppm for the former are significantly larger
than those for the latter. (Note that fumaronitrile complex 20 is
a special case, owing to symmetry in the substrate.) The values
of ΔCN ≈ 5−12 ppm for the nitriles are relatively small and
positive, similar to ΔCO ≈ 2−11 ppm for esters (Table 1),
which suggests they have similar bonding patterns.
The values of Δα and Δβ for α,β-unsaturated sulfones are
very similar to those for α,β-unsaturated ketones. This
observation supports η2-structures for both kinds of
complexes.
Compounds with C−N Double Bonds. In 1979, it was
discovered that CuI activates the C−N double bond for the
addition of an alkyllithium to the carbon atom.19 One possible
explanation is coordination of Cu+ to the lone pair on the imine
nitrogen in analogy with the activation of a C−O double bond
by coordination of Li+ to a lone pair on oxygen. A second
possibility is initial formation of an organocopper(I) reagent,
which then adds to the C−N double bond. Attempts to observe
complexes between cuprates and simple imines were not
successful; therefore, we looked for special cases, which would
hold synthetic as well as theoretical interest.
Scheme 2. Preparation of Substituted Sulfone Complexes
The choice of examples with C−N double bonds was largely
dictated by the stability of the substrates. The preparation
of 2-azachalcone π-complex 28 was routine under the usual
conditions. The isomeric 3-azachalcone was not stable, and it
was necessary to use the 4-fluoro derivative, which gave
complex 29. The azachalcone complexes slowly decomposed at
−100 °C with the liberation of ethane, presumably from
reductive elimination of the two methyl groups on copper. In
contrast, chalcone complex 10 was stable at this temperature,
and warming to −10 °C yielded only 1,4-addition product and
no ethane.12
Chart 2. Complexes from Unsaturated Nitriles and Sulfones
Isothiocyanates are a fascinating class of compounds with
both C−N and C−S double bonds. The C−S double-bond
complexes are discussed in the next section. Methyl
isothiocyanate gave C−N complex 30, which was stable at
−100 °C. Warming the sample to −60 °C resulted in addition
(Scheme 3). No C−S complex (vide infra 33) was observed in
this case. In contrast, phenyl isothiocyanate gave transient C−N
complex 31, which rearranged to a stable C−S complex (vide
infra 34).20 Stacked H NMR plots for this transformation are
1
shown in Figure 6. Chart 3 summarizes the structures and
Table 3 the NMR data for the complexes with C−N double
bonds.
Values of Δα = −65 ppm for 29 and Δβ = −100 ppm for 28
are larger in magnitude than the corresponding values, Δα =
−52 ppm and Δβ = −81 ppm, for chalcone complex 10. The
magnitudes of ΔCO for complexes 28 (−8 ppm), 29 (−17
ppm), and 10 (−9 ppm) are relatively small, which suggests
they involve dihapto-coordination.
a b
,
1
Table 2. H and 13C NMR Data for Complexes from Unsaturated Nitriles and Sulfones
ID
MeαH
MeαC
MeβH
MeβC
DBαH
DBαC
DBβH
DBβC
CN
129.27
Δα
Δβ
ΔCN
12.16
19
20
21
22
23
24
25
26
27
−0.88
−0.34
−0.96
−0.57
−0.67
−0.71
−0.69
−0.58
−0.69
−8.37
−5.60
−10.51
−7.63
−8.10
−8.75
−8.18
−9.32
−9.59
−0.48
−0.34
−0.31
−0.34
−0.37
−0.32
−0.42
−0.17
−0.40
−7.86
−5.60
3.71
2.08
2.28

42.31
24.06
22.90
17.58
17.61
16.51
62.25
63.73
66.56
1.89, 1.91
2.28, 
, 
22.20
24.06
64.09
61.34
61.43
50.54
37.60
60.63
51.38
−65.97
−95.31
−62.01
−65.27
−60.31
−69.45
−77.22
−71.32
−66.11
−116.23
−95.31
−118.54
−99.87
−98.93
−109.44
−91.18
−76.85
−91.91
124.92, 124.92
119.92, 119.92
120.77, 120.56
120.86, 120.63
121.92, 122.37
N/A
10.57, 10.57
6.68, 6.68
5.60, 6.48
5.12, 5.87
6.60, 8.15
N/A, N/A
N/A, N/A
N/A, N/A
1.97

3.44, 
3.42, 
3.64, 
1.75, 2.03
4.34, 
2.71, 
1.51

−3.24
−6.26
−0.19
−1.66

3.12
3.40
3.13
N/A
N/A
a
b
All shifts are in ppm, measured at −100 °C. Cuprates prepared from CuI.
7831
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
The large, positive value of ΔCN = 83 ppm for complex 30 is
similar to ΔCS = 85 ppm for complex 32 from carbon disulfide
(next section).
Compounds with C−S Double Bonds. We discovered
that carbon disulfide formed a complex, 32, with lithium
dimethylcuprate(I) during our investigation into reactions of
lithium tetramethylcuprate(III),21 the CuIII analogue of the
Gilman reagent. While Me4CuLi did not react with CS2, we
noticed that a small amount of residual starting material,
Me2CuLi, was rapidly replaced by a new species, identified as
32 (Scheme 4a). Therefore, we investigated the reactions of
Me2CuLi with diverse thiocarbonyl compounds, as summarized
in Chart 4 and Table 4. (Scheme 4b, methanol reactions, will be
discussed in the next section.)
Carbon disulfide and isothiocyanates are the only substrates
that exhibit large downfield changes in chemical shift upon
complexation (cf. 30, 32, and 34). Dithioester complexes 37−
40 exhibit large upfield changes, ΔCS ≈ −148 6 ppm, and
thioketone complexes even larger ones, ΔCS = −157 ppm for
35 and ΔCS = −182 ppm for 36. The latter has the largest
absolute value we have measured to date, and all the values of
ΔCS for thioketone and dithioester complexes are significantly
greater in magnitude than any value of Δα or Δβ for C−C or
C−N double-bond complexes.
Scheme 3. Reactions of Isothiocyanates with Lithium
Dimethylcuprate(I)
Ligand Exchange Reactions. The relative stabilities of
representative complexes were determined by adding a second
substrate to a preformed complex. For example, addition of
methyl dithio-4-trifluoromethylbenzoate to the methyl dithio-
benzoate−cuprate complex 38 resulted in a smooth conversion
to complex 40 and free methyl dithiobenzoate. In this way, the
order of stability for C−S double-bond complexes was
determined to be 40 > 38 > 39 > 37 > 32. We note that the
electron-withdrawing trifluoromethyl group has a stabilizing
effect, whereas the electron-donating methoxy group is
destabilizing relative to H.
A similar study with C−C double bonds in α-enones resulted
in the order 4 > 10 > 9 > 6. Methyl dithioacetate displaced
methyl vinyl ketone from π-complex 4 to give thiocarbonyl
complex 37; therefore, the weakest dithioester complex was
more stable than the strongest α-enone complex.
The α,β-unsaturated nitriles constitute an especially interest-
ing series. Acrylonitrile gave a very weak complex, 19. The
acrylonitrile ligand was even displaced by 2-cyclohexenone,
which formed the weakest α-enone complex, 6. On the other
hand, fumaronitrile complex 20 was one of the strongest, and
fumaronitrile displaced benzylidenemalononitrile from 22, the
most stable malononitrile complex. The complete order of
stability was 20 > 22 > 23 > 24 > 21 > 19. Fumaronitrile also
displaced diethyl fumarate from complex 14 to give 20
quantitatively.
Figure 6. Stacked 1H NMR plots for the conversion of C−N complex
31 (−0.03, −0.47 ppm) to C−S complex 34 (0.43, −0.79 ppm). (The
total time was 40 min.)
Chart 3. Complexes of Compounds with C−N Double
Bonds
It must be emphasized that when we discuss “stability” in this
section, it applies only to competition between ligands, and it is
entirely distinct from other kinds of reactivity such as thermal
stability. For example, the weakest π-complex studied here was
19; nevertheless, it was thermally stable up to 25 °C, while
most of the other complexes went on to products at lower
temperatures.
Treatment of C−S double-bond complex 32 or 40 with methanol
(4 equiv) at −100 °C did not lead to decomposition. The same
result was obtained for Me4CuLi at this temperature (Scheme 4b).
ab
,
1
Table 3. H and 13C NMR Data for Complexes of Compounds with C−N Double Bonds
c
ID
MeαH
MeαC
MeβH
MeβC
DBαH
DBαC
DBβH
DBβC
CO, CN
Δα
Δβ
ΔCO, ΔCN
28
29
30
−1.07
−0.43
−0.48
−0.47
−1.05
10.86
−1.48
−2.76
−0.26
−0.09
0.52
5.02
5.96
−1.25
2.15
b


4.73

66.36

173.33, 
174.09, 
N/A, 207.85

−100.30

−7.66, N/A
−17.04, N/A
N/A, 82.85
5.25
N/A
N/A
93.52
N/A
N/A
−65.15
N/A
N/A
N/A
c
N/A
N/A
N/A
N/A
d
d
d
d
31
−0.03
N/A, 
N/A
N/A, 
a
All shifts are in ppm, measured at −100 °C. Cuprates prepared from CuI. For 30 and 31; for 28 and 29 see DB entries. Transient intermediate;
13C NMR not measured.
7832
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
1
Scheme 4. Treatment of Me2CuLi and Me4CuLi with (a)
Carbon Disulfide and (b) Methanol
Table 4. H and 13C NMR Data for Complexes of
a b
,
Compounds with C−S Double Bonds
ID
MeαH
MeαC
MeβH
MeβC
CS
ΔCS
32
c
−0.76

−10.54

0.37

20.10

279.19

85.29

33
34
35
36
37
38
39
40
41
42
43
−0.79
−0.55
−0.83
−0.96
−0.68
−0.74
−0.52
−0.91
−1.00
−0.76
−9.09
−9.02
−6.76
−8.58
−9.31
−9.46
−9.18
−9.23
−9.34
−9.51
0.43
10.81
9.13
3.26
7.17
9.35
9.19
9.18
6.50
4.06
8.84
193.88
81.89
91.22
84.44
81.91
82.08
79.71
116.79
111.57
84.69
61.43
−0.33
−0.15
−0.16
−0.29
−0.29
−0.27
−0.59
−0.03
0.12
−157.49
−181.93
−152.08
−148.27
−144.90
−149.33
−95.42
−103.40
−143.92
a
b
All shifts are in ppm, measured at −100 °C. Cuprates prepared from
CuI. Not observed.
c
Scheme 5. Preparation of Hydroxy Complex 16 by Direct
and Indirect Routes
Chart 4. Complexes of Compounds with C−S Double Bonds
significant methane generation. These experiments support our
proposal that an associative mechanism is involved.20
Theoretical studies also support this conjecture.23
Ligand exchange has proven to be a useful synthetic technique.
For example, injection of ethyl 2-hydroxyethyl fumarate (EHF)
into the usual Me2CuLi solution at −100 °C resulted in a mixture
(ca. 1:1) of hydroxy complex 16 and lithiated substrate (Li-EHF,
Scheme 5). In contrast, when chalcone complex 10 was prepared
first and then treated with EHF, the product was predominately
16 (95%), which was stable enough for the usual spectral char-
acterization. If there were significant equilibrium concentrations of
Me2CuLi in solutions of 16, we would expect this product to be
transient at best.
Amide complex 17 was prepared in reasonably good yield
(75%) using our standard methodology and in very good yield
(92%) via the displacement of chalcone from its π-complex 10.
Two N−H peaks (ca. 3:1) were present in the H NMR
spectrum of 17, which indicated at least two conformers were
present.
Monoethyl fumarate quenched the cuprate solution imme-
diately upon injection. It was not possible to make a monoethyl
fumarate complex by indirect means, either.
Injection of solutions of maleic anhydride into solutions of
Me2CuLi·LiI at −100 °C gave π-complex 18 in variable yields,
which depended upon how long the injector was cooled in the
probe before injection. In contrast, injection of maleic
In contrast, Me2CuLi reacted rapidly with methanol (4 equiv)
to produce methane under the same conditions. During the
first few seconds, the homocuprate was converted to an Ashby
cuprate, Me3Cu2Li (MeCu + Me2CuLi),22 which also reacted
with methanol at −100 °C.
When carbon disulfide complex 32 was treated with methyl
dithio-4-trifluoromethylbenzoate in the presence of methanol
(4 equiv) at −100 °C, the ligand exchange proceeded to
complex 40 and carbon disulfide as usual, without the
appearance of methane. We believe that a dissociative
mechanism involving free Me2CuLi would have resulted in
1
7833
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
Scheme 6. Preparation of Maleic Anhydride Complex 18 by
Direct and Indirect Routes
Chart 5. First Examples of η1 σ-allyl and η3 π-allyl CuIII
Intermediates
respectively, were observed along with broadening of the peak
for the methyl carbon, which may be attributed to a small,
unresolved cis coupling to 13CN.
Similarly, cis (2JCC = 1.4 Hz) and trans (2JCC = 17.7 Hz)
couplings were observed for the complex of methyl dithio-4-
trifluoromethylbenzoate with 13CH3(2-thienyl)CuLi.8 For the
complex of 13CS2 with Me2CuLi, we measured cis coupling
(2JCC = 4.5 Hz) to the downfield methyl group and trans
coupling (2JCC = 28.8 Hz) to the upfield methyl.20
anhydride into a solution of 19 resulted in the rapid displacement
of acrylonitrile and formation of complex 18 (Scheme 6).
Coupling Constants. Whereas methyl vinyl ketone (MVK)
failed to yield a stable π-complex (cf. 44, Scheme 7) with
MeCuCNLi,8 fumaronitrile afforded complex 45, which was
stable at −100 °C, provided excess substrate (2 equiv) was
used. With a stoichiometric amount of substrate, mixed cuprate
complex 45 metathesized to homocuprate complex 20. This
process is discussed in detail in an accompanying communi-
cation.24
Conversion of CuI to CuIII Complexes and the Key to
Stability. While they frequently had been postulated as
intermediates in reactions involving copper,27 the first CuIII
intermediate identified for a major synthetic reaction was
reported in 2007.10 Intermediate 46 (Chart 5) resulted from
the reaction of π-complex 6 with trimethylsilyl cyanide
(TMSCN) at −100 °C. This σ-allyl species was characterized
as a pseudo-square-planar complex with the aid of two-bond
coupling across copper (vide supra). The related π-allyl CuIII
complex was not observed, owing to the lack of a β-phenyl
substituent (vide infra).
Scheme 7. Reactions of MVK and Fumaronitrile with
In our study of the reactions of organocuprates with allylic
halides and acetates,11 we found that a phenyl substituent
helped stabilize η3 CuIII complex 47 (Chart 5). This inspired us
to examine cinnamaldehyde, crotonophenone, benzalacetone,
and chalcone π-complexes (3 and 8−10, respectively). When
these η2 CuI complexes, prepared from CuCN, were treated
with trimethylsilyl cyanide (Me3SiCN, TMSCN), they gave the
respective η1 CuIII complexes 48a−d (Scheme 8).
MeCuCNLi and Me2CuLi·LiCN
Scheme 8. Reactions of Selected π-Complexes with TMSCN
and TMSCl
Two-bond J-coupling across copper was introduced as a
structure probe for organocopper compounds in 1991; for
example, 13CH3Cu13CNLi was shown to have one methyl and
one cyano group bonded to the same copper atom (2JCC
=
21.6 Hz).25 Both JCC and JCP have been applied to structure
problems.26 When η1 CuIII complex 46 (Chart 5) was prepared
using 13CH3Li and Cu13CN, both small cis (2JCC = 5.4 Hz) and
large trans (2JCC = 35.4 Hz) coupling constants were measured
between the 13C-labeled methyl groups on copper and 13CN; the
methyl−methyl cis coupling was also small (2JCC = 2.9 Hz).10
2
2
On the other hand, when they were prepared from CuI,
treatment with trimethylsilyl chloride (Me3SiCl, TMSCl) or
other chlorosilanes (Et3SiCl, Me2PhSiCl) gave η3 CuIII
complexes 49a−d (or analogous complexes), respectively. When
they were prepared from CuCN and then treated with
chlorosilanes, they gave mixtures of 48 and 49.12
When Cu13CN was used as the precursor for 45, cis (2JCC
=
4.4 Hz) and trans (2JCC = 11.5 Hz) couplings to Cα and Cβ,
7834
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
a b
,
1
Table 5. H and 13C NMR Data for η1 and η3 CuIII Complexes
ID
MeαH
MeαC
MeβH
MeβC
DBαH
DBαC
DBβH
DBβC
C−O
CN
Δα
Δβ
ΔCO
c
48a
48b
0.21
0.04
14.07
13.32
14.15
15.28
−5.81
0.76
0.65
0.70
0.89
0.06
0.30
0.03
0.20
21.12
17.72
21.62
21.70
9.75
5.66
6.15
5.56
6.67
5.31
5.63
5.36
6.44
120.77
129.29
117.87
122.79
96.36
N/A
N/A
N/A
N/A
4.55
3.67
4.26
4.55
N/A
N/A
N/A
N/A
78.51
128.94
139.68
137.48
125.45
124.69
152.79
152.72
153.70
152.87
N/A
N/A
N/A
N/A
N/A
20.46
N/A
N/A
N/A
N/A
15.56
N/A
c
c
N/A
c
48c
48d
0.13
N/A
0.26
N/A
d
49a
49b
49c
49d
0.05
−45.76
ef
f
f
f
f
f
f
f
f
,
−0.51
−0.26
−0.38





N/A



3.74
9.29
7.73
9.07
95.59
96.20
74.84
78.37
139.37
134.44
N/A
24.05
26.82
11.74
14.17
−45.16
−44.55
N/A
a
b
c
All shifts are in ppm, measured at −100 °C. Cuprates prepared from CuI unless otherwise noted. Cuprate prepared from CuCN.
d
e
f
(CD3)2(C6H5)SiCl added, as (CH3)3Si peak interfered. (CH3CH2)3SiCl added, as (CH3)3Si peak interfered. Fleeting intermediate; 13C NMR not
measured.
In the case of crotonophenone, η3-complex 49b was unstable
alkyl halides.32 This reaction has many complications,33 but it is
useful in special cases such as allylic halides and carboxylates.11
In 1988, Bertz, Dabbagh, and Williams reported a new kind
of reaction of Gilman reagents, namely, carbophilic addition to
thiocarbonyl compounds, for example, thioketones and
dithioesters.20 The “double-barreled” carbophilic addition to
dithioesters was made possible by the intermediacy of reactive
mixed thiocuprates (Posner reagents).34 This reaction has
found some interesting applications;35 however, it is included
here mainly for theoretical reasons. In analogy with the
Kharasch reaction, where copper salts change the regioselec-
tivity from 1,2- to 1,4-addition, they change the regiochemistry
of the reactions of Grignard and lithium reagents with
thiocarbonyl compounds from thiophilic to carbophilic
addition.20
Scheme 9 lays out a mechanism for the conversion of a
typical dithioester A to thiolate product E, which gives doubly
substituted thiol F upon workup. We observe intermediate B,
but not conjectural intermediate C or D, which intimates that
their reactions are very fast. Furthermore, the substrate must
remain complexed to copper throughout. Simple thioketones
are very unstable,36 and if a significant amount of D were lost
via dissociation, the overall yield of product would be
measurably diminished. In one paradigmatic reaction, methyl
dithiopropionate gave 96−100% yields of 2-methyl-2-butane-
thiol with Me2CuLi, prepared from a variety of copper(I) salts
(CuI, CuCN, CuBr·SMe2, CuOTf (copper(I) triflate)).20b
This reaction is an example of geminal alkylation. Our first
attempt at this genre involved the reactions of lithium and
and rapidly went on to 1,4-addition product at −100 °C.12 In
the case of cinnamaldehyde, anti−syn complex 49a isomerized
to the corresponding syn−syn complex 49a′. We use the
chemical shifts for 49a in order to compare them with those for
49c and 49d, which also have the anti−syn conformation. In
contrast to 49b, complexes 49a′, 49c, and 49d were stable for
extended periods at −100 °C. It appears that a β-phenyl
substituent is the key to stability.
Important NMR data for complexes 48 and 49 are
summarized in Table 5. For the calculation of Δ-values, we
use the chemical shifts for 3a, 9a, and 10, which all have the s-
cis conformation. The conversion of these complexes to η3 CuIII
complexes 49a, 49c, and 49d, respectively, results in Δα ≈ 20−
27 ppm and Δβ ≈ 11−16 ppm, which move in the opposite
direction from the Δ-values for the starting π-complexes (Table
1). Moreover, the larger effect is observed for Cα instead of Cβ.
The magnitudes of ΔCO ≈ −45 1 ppm are significantly larger
than those of Δα and Δβ (|ΔCO| > |Δα| > |Δβ|), and they are
substantially greater than the magnitudes of ΔCO for 3a (−24
ppm), 9a (−14 ppm), and 10 (−9 ppm).
Strictly speaking, 46, 48, and 49 are intermediates in the
synthetically valuable and theoretically interesting silyl-assisted
conjugate addition of organocuprates.28 Nevertheless, the facile
conversion of cuprate−olefin π-complexes to CuIII intermedi-
ates tends to support the oxidative addition/reductive
elimination mechanism (Scheme 1b) for the usual conjugate
addition reaction of organocuprates to α,β-unsaturated carbon-
yl compounds.
PERSPECTIVE AND CONCLUSIONS
■
The first organocuprates were prepared from CuCN and
Grignard reagents, as reported by Konduirov and Fomin in
1915.29 These presumptive cuprates decomposed without
isolation or characterization to hydrocarbon products. The
first organocuprate(I) compound to be characterized as such
was Me2CuLi, prepared by Gilman in 1952;30 however, its
chemistry was not explored until more than a decade later. In
1966, House and co-workers showed that this “Gilman reagent”
gave 1,4-addition (“conjugate addition”) with α-enones,31 and
they speculated that organocuprates were intermediates in the
Kharasch reaction, the 1,4-addition of Grignard reagents to α,β-
unsaturated carbonyl compounds catalyzed by copper salts.16
We have provided experimental support for this conjecture by
preparing a cuprate−olefin π-complex (6c) starting from CuCl
and MeMgCl.
Scheme 9. Double-Barreled Carbophilic Addition of
Cuprates to Dithioesters
Shortly after the House paper appeared, Corey and Posner
reported the cross-coupling reaction of organocuprates with
7835
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
copper reagents with ditosylhydrazones, which were not
synthetically useful, owing to competing side reactions.19
Three apparently unique reactions of organocuprates are (1)
conjugate addition to activated olefins, (2) cross-coupling with
alkyl halides, and (3) carbophilic addition to thiocarbonyl com-
pounds. It is possible to unite all three reactions into one “unified
theory” by showing that they proceed via similar CuIII inter-
mediates, as illustrated for the cyano-Gilman reagent in Scheme
10, where 1, 2, and 3 are the CuIII intermediates in paths 1, 2, and
3, respectively. In the case of path 1, both η1 and η3 CuIII inter-
mediates have now been prepared,10,12 as described above. Like-
wise, CuIII intermediates have been observed for path 2.33 In the
case of path 3, CuIII intermediate 3 has not yet been observed;
nevertheless, we note it is analogous to 1, where the enolate has
been replaced by a thiolate. (N.B., cuprate−thiocarbonyl com-
plexes may also be considered CuIII species.20)
approximate position of complexes on the π−σ continuum is
supported by the observation of two-bond coupling across
copper for the “tighter” complexes, such as those of carbon
disulfide, fumaronitrile, and methyl dithio-4-trifluoromethyl-
benzoate. We have not been able to detect J-coupling across
copper in α-enone π-complex 6a or 6b with 13C-labeled methyl
groups. While we have included a π−σ scale, we do not mean to
imply quantitative information is presently available.
Three Controversies. The first great controversy of
organocuprate chemistry concerned the mechanism of the
conjugate addition reaction, as discussed in the Introduction.
By using rapid injection NMR, we have now been able to make
a panoply of cuprate−olefin π-complexes and study their
properties. This body of knowledge provides a solid foundation
for future mechanistic investigations.
The second great controversy concerned the existence of
“higher order cyanocuprates”, which were first questioned by a
Bell Laboratories chemist.37 Thanks to a number of
experimental techniques (NMR, IR, EXAFS, and single-crystal
X-ray),38 as well as theoretical calculations,39 we now know that
they do not exist; nevertheless, it is worth considering why
people believed in them in the first place. In many examples,
the cyano-Gilman reagents, R2CuLi·LiCN, were more reactive
than the previous iodo-Gilman reagents, R2CuLi·LiI. Con-
sequently, we are well aware of the danger inherent in the use
of reactivity to assign structures, and we only use it to highlight
major differences, for example, the use of methanol to
differentiate qualitatively “loose” versus “tight” complexes
(i.e., very weak vs very strong bonding).
Scheme 10. Unified Theory of Organocuprate Reactivity
Figure 8 provides a simple way to understand the problem
with “higher order cyanocuprates”. In structures A and B the
negative charges from the methyl groups are balanced by the
positive charges on CuI or CuIII, respectively, and the remaining
negative charge is localized on N, the most electronegative
atom, and balanced by a Li+ ion. The CuI atom in A is d10, and
coordination of two lone pairs gives a total of 14 electrons; the
CuIII atom in B is d8, and coordination of four lone pairs gives a
total of 16 electrons. In C and D, we attempt to add an extra Me
anion to the Cu atoms in A and B, increasing the canonical
electron counts to 16 and 18, respectively, which appear accept-
able. However, calculations for C show that no low-energy con-
figuration is available.39 This is essentially an application of the
Pauling electroneutrality principle.40 Our efforts to prepare D have
proven unsuccessful, as were many attempts to prepare C.37
Finally, we come full circle by settling a third controversy,
concerning the hapticity of bonding in cuprate−olefin
complexes, which we have supported as intermediates in the
conjugate addition reaction. It has been suggested that they are
We have established that η2-complexes are intermediates in
conjugate addition (path 1) and carbophilic addition (path 3). The
elucidation of the exact nature of these η2-complexes may require
bringing other techniques to bear (e.g., X-ray crystallography).
In the meantime, we have conjectured that α-enones are on the
π-side and thiocarbonyls are on the σ-side of the π−σ continuum for
double-bond complexes of transition metals.20 The α,β-unsaturated
nitriles studied here appear to span virtually the entire range.
In Figure 7 we indicate the relative stabilities of a dozen of
the >40 complexes studied here by ordering their ligands. The
Figure 7. Relative stabilities of selected substrates in dimethylcuprate complexes. Each is displaced from its complex by those to the right of it. The
scale is not quantitative.
7836
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
as when virgin NMR tubes were used. In fact, were it not for rapid
injection NMR, we would not have seen this π-complex at all.
Me2CuLi was prepared by adding 2.00 equiv of MeLi·THF-d8 in
benzene-d6 (ca. 1 M) to a suspension of CuI or CuCN (30.0 μmol,
Aldrich, best grade available) in 420 μL of THF-d8 (freshly distilled
from Na/K) in a septum-sealed NMR tube, cooled to −80 °C in a dry
ice/acetone bath. The suspension of copper salt was sonicated for 1
min at 0 °C before addition of the lithium reagent at −80 °C, and the
resulting cuprate solution was agitated for 10 s outside the cold bath
with a vortex mixer (VWR). The sample was checked for purity by 1H
NMR at −100 °C. When a satisfactory sample was obtained, the
septum was removed in a stream of dry nitrogen, and the tube was
immediately lowered into the probe, which was filled with the dry
nitrogen used to spin the sample. The substrate, dissolved in THF-d8,
1
was then injected (60 μL per standard injection), and single-pulse H
Figure 8. Ate-complexes of CuI and CuIII: A and B are known
compounds; C and D are hypothetical “higher order cyanocuprates”.
NMR spectra were obtained periodically (2−20 s, depending on the
rate of reaction). H NMR (13C NMR) spectra were referenced to
1
benzene at 7.34 (128.59) ppm.
η3 CuIII complexes.18 Figure 9a illustrates how the chemical
shifts of relevant carbon atoms change upon complex
formation. It is clear that the carbonyl group (CO) remains
essentially the same and that the olefin is the focus of bond
formation. In contrast stands Figure 9b, where the carbonyl
group is significantly affected by π-allyl complex formation.
Therefore, regardless of where organocuprate−double-bond
complexes may be on the π−σ continuum, we conclude that
they all involve dihapto-coordination.
In this paper we have presented our results for organocuprate
complexes with C−C, C−N, and C−S double bonds, which are
intermediates in reactions of synthetic and theoretical interest.
A copious amount of NMR data strongly supports the
conclusion that all of them are fundamentally based on
dihapto-coordination to copper, although they vary dramati-
cally in the strength of the ligand−copper bond from very weak
to very strong, indeed. The weaker complexes are useful for the
synthesis of stronger, functionalized ones that are difficult to
prepare directly. Finally, this study makes significant con-
tributions to understanding the three most important
controversies in the history of organocopper chemistry.
ASSOCIATED CONTENT
* Supporting Information
Typical 1D and 2D NMR plots. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: sbertz1@uncc.edu; cogle@uncc.edu.
■
Present Addresses
†Department of Chemistry, University of California at San
Diego, La Jolla, California 92093, United States.
‡Department of Chemistry, University of Illinois at Urbana−
Champaign, Champaign, Illinois 61820, United States.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the National Science Foundation for financial
support through grants 0353061 and 0321056 and Prof. Dieter
Seebach for the generous gift of the substrate for complex 13.
Donna Dorton prepared one of the π-complexes.
EXPERIMENTAL SECTION
■
REFERENCES
NMR spectra were obtained with a JEOL ECA-500 spectrometer. All
NMR tubes (Norell) were dried in an oven (150 °C) and cooled
under Ar. In order to get consistently good results, it is critically
important that the NMR tubes be unused. We have experimented with
used NMR tubes that had been cleaned thoroughly by washing with
concentrated nitric acid, deionized water, concentrated ammonium
hydroxide, deionized water, and finally acetone. The lifetime of
π-complex 6 in these tubes was measured in seconds rather than hours,
■
(1) House, H. O. Acc. Chem. Res. 1976, 9, 59−67. For an example of
a cuprate reaction that has been shown to involve electron transfer, see
ref 33c.
(2) (a) Hallnemo, G.; Olsson, T.; Ullenius, C. J. Organomet. Chem.
1984, 265, C22−24. (b) Lindstedt, E. L.; Nilsson, M.; Olsson, T. J.
Organomet. Chem. 1987, 334, 255−261. (c) Christenson, B.; Olsson,
T.; Ullenius, C. Tetrahedron 1989, 45, 523−534.
Figure 9. 13C NMR chemical shift ranges for CO (red), Cα (green), and Cβ (blue) of the substrate and Meα (top magenta) and Meβ (bottom
magenta) of the cuprate before and after (a) complexation giving η2-complexes and (b) silylation converting η2- to η3-complexes.
7837
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838

Organometallics
Article
(3) (a) Bertz, S. H.; Smith, R. A. J. J. Am. Chem. Soc. 1989, 111,
8276−8277. See also: (b) Vellekoop, A. S.; Smith, R. A. J. J. Am.
Chem. Soc. 1994, 116, 2902−2913.
(22) Ashby, E. C.; Watkins, J. J. J. Am. Chem. Soc. 1977, 99, 5312−
5317.
(23) Gartner, T.; Yoshikai, N.; Neumeier, M.; Nakamura, E.;
̈
Gschwind, R. M. Chem. Commun. 2010, 46, 4625−4626.
(24) Bertz, S. H.; Browder, K. L.; Hardin, R. H.; Murphy, M. D.;
Ogle, C. A.; Thomas, A. A. Organometallics 2012, DOI: 10.1021/
om3006646.
(4) Gschwind, R. M. Chem. Rev. 2008, 108, 3029−3053.
(5) (a) Canisius, J.; Mobley, T. A.; Berger, S.; Krause, N. Chem.
Eur. J. 2001, 7, 2671−2675. (b) Krause, N.; Wagner, R.; Gerold, A. J.
Am. Chem. Soc. 1994, 116, 381−382. (c) Krause, N. J. Org. Chem.
1992, 57, 3509−3512.
(25) Bertz, S. H. J. Am. Chem. Soc. 1991, 113, 5470−5471.
(26) Bartholomew, E. R.; Bertz, S. H.; Cope, S.; Dorton, D. C.;
Murphy, M.; Ogle, C. A. Chem. Commun. 2008, 1176−1177. See also
refs 5, 8, 10−12, 20, and 21.
(6) (a) McGarrity, J. F.; Prodolliet, J. J. Org. Chem. 1984, 49, 4465−
4470. (b) McGarrity, J. F.; Ogle, C. A.; Brich, Z.; Loosli, H. R. J. Am.
Chem. Soc. 1985, 107, 1810−1815.
(27) Hickman, A. J.; Sanford, M. S. Nature 2012, 484, 177−185.
(28) (a) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26, 6015−
6018. (b) Corey, E. J.; Boaz, N. W. Tetrahedron Lett. 1985, 26, 6019−
6022. (c) Alexakis, A.; Berlan, J.; Besace, Y. Tetrahedron Lett. 1986, 27,
1047−1050. (d) Horiguchi, Y.; Matsuzawa, S.; Nakamura, E.;
Kuwajima, I. Tetrahedron Lett. 1986, 27, 4025−4028. (e) Nakamura,
E.; Matsuzawa, S.; Horiguchi, Y.; Kuwajima, I. Tetrahedron Lett. 1986,
27, 4029−4032. (f) Bertz, S. H.; Smith, R. A. J. Tetrahedron 1990, 46,
4091−4100. See also: (g) Bertz, S. H.; Miao, G.; Rossiter, B. E.;
Snyder, J. P. J. Am. Chem. Soc. 1995, 117, 11023−11024.
(29) Konduirov, N. V.; Fomin, D. A. Zh. Russ. Fiz.-Khim. O-va, Chast
Khim. 1915, 47, 190−198.
(7) (a) Bertz, S. H.; Carlin, C. M.; Deadwyler, D. A.; Murphy, M. D.;
Ogle, C. A.; Seagle, P. H. J. Am. Chem. Soc. 2002, 124, 13650−13651.
For oscillations in this system, see: (b) Murphy, M. D.; Ogle, C. A.;
Bertz, S. H. Chem. Commun. 2005, 854−856.
(8) (a) Bertz, S. H.; Hardin, R. A.; Murphy, M. D.; Ogle, C. A.;
Richter, J. D.; Thomas, A. A. Angew. Chem., Int. Ed. 2012, 51, 2681−
2685. For TMSM cuprates, see: (b) Bertz, S. H.; Eriksson, M.; Miao,
G.; Snyder, J. P. J. Am. Chem. Soc. 1996, 118, 10906−10907. (c) For a
̈
chiral mixed cuprate−alkene complex, see: Eriksson, J.; Davidsson, O.
Organometallics 2001, 20, 4763−4765. For a mixed cuprate−alkyne
complex, see: (d) Nilsson, K.; Ullenius, C.; Krause, N. J. Am. Chem.
Soc. 1996, 118, 4194−4195.
(30) (a) Gilman, H.; Jones, R. G.; Woods, L. A. J. Org. Chem. 1952,
17, 1630−1634. See also: (b) Gilman, H.; Straley, J. M. Rec. Trav.
Chim. Pays-Bas. 1936, 55, 821−834.
(9) (a) Berlan, J.; Besace, Y. Tetrahedron 1986, 42, 4767−4776.
(b) Berlan, J.; Battioni, J. P.; Koosha, K. Bull. Soc. Chim. Fr. 1979, part
2, 183−190.
(31) House, H. O.; Respess, W. L.; Whitesides, G. M. J. Org. Chem.
1966, 31, 3128−3141.
(10) Bertz, S. H.; Cope, S.; Murphy, M.; Ogle, C. A.; Taylor, B. J. J.
Am. Chem. Soc. 2007, 129, 7208−7209. See also ref 15b.
(11) Bartholomew, E. R.; Bertz, S. H.; Cope, S.; Murphy, M.; Ogle,
C. A. J. Am. Chem. Soc. 2008, 130, 11244−11245 and references
therein.
(32) (a) Corey, E. J.; Posner, G. H. J. Am. Chem. Soc. 1967, 89,
3911−3912. (b) Corey, E. J.; Posner, G. H. J. Am. Chem. Soc. 1968, 90,
5615−5616. See also: (c) Whitesides, G. M.; Fischer, W. F.; San
Filippo, J.; Bashe, R. W.; House, H. O. J. Am. Chem. Soc. 1969, 91,
4871−4882.
(12) Bertz, S. H.; Hardin, R. A.; Murphy, M. D.; Ogle, C. A.; Richter,
J. D.; Thomas, A. A. J. Am. Chem. Soc. 2012, 134, 9557−9560.
(13) (a) Bertz, S. H.; Human, J.; Ogle, C. A.; Seagle, P. Org. Biomol.
Chem. 2005, 3, 392−394. For complications in cuprate preparation,
see: (b) Bertz, S. H.; Gibson, C. P.; Dabbagh, G. Tetrahedron Lett.
1987, 28, 4251−4254. For complexity in asymmetric induction, see:
(c) Bertz, S. H.; Dabbagh, G.; Sundararajan, G. J. Org. Chem. 1986, 51,
4953−4959. See also ref 7b.
(33) (a) Bertz, S. H.; Cope, S.; Dorton, D.; Murphy, M.; Ogle, C. A.
Angew. Chem., Int. Ed. 2007, 46, 7082−7085. See also: (b) Gartner,
̈
T.; Henze, W.; Gschwind, R. M. J. Am. Chem. Soc. 2007, 129, 11362−
11363. For cuprate−iodocyclohexane reaction, see: (c) Bertz, S. H.;
Dabbagh, G.; Mujsce, A. M. J. Am. Chem. Soc. 1991, 113, 631−636.
(34) (a) Posner, G. H.; Whitten, C. E.; Sterling, J. J. J. Am. Chem. Soc.
1973, 95, 7788−7800. (b) Posner, G. H.; Brunelle, D. J.; Sinoway, L.
Synthesis 1974, 662−663. (c) Posner, G. H.; Whitten, C. E. Org. Synth.
1976, 55, 122−127.
̈
(14) (a) Bertz, S. H.; Nilsson, K.; Davidsson, O.; Snyder, J. P. Angew.
Chem., Int. Ed. 1998, 37, 314−317. For aggregation in ether, see:
(b) Bertz, S. H.; Vellekoop, A. S.; Smith, R. A. J.; Snyder, J. P.
Organometallics 1995, 14, 1213−1220. For aggregation in dimethyl
sulfide, see: (c) Bertz, S. H.; Dabbagh, G.; He, X.; Power, P. P. J. Am.
Chem. Soc. 1993, 115, 11640−11641.
(15) (a) Snyder, J. P. J. Am. Chem. Soc. 1995, 117, 11025−11026.
(b) Hu, H.; Snyder, J. P. J. Am. Chem. Soc. 2007, 129, 7210−7211.
See also ref 17.
(35) (a) Jenny, C.; Wipf, P.; Heimgartner, H. Helv. Chim. Acta 1986,
69, 1837−1843. For extension to selenium, see: (b) Fujiwara, S.; Asai,
A.; Shin-ike, T.; Kambe, N.; Sonoda, N. J. Org. Chem. 1998, 63, 1724−
1726.
(36) Kroto, H. W.; Landsberg, B. M.; Suffolk, R. J.; Vodden, A. Chem.
Phys. Lett. 1974, 29, 265−269. See also ref 20c.
(37) (a) Bertz, S. H. J. Am. Chem. Soc. 1990, 112, 4031−4032. See
also: (b) Bertz, S. H.; Miao, G.; Eriksson, M. Chem. Commun. 1996,
815−816. (c) Bertz, S. H.; Chopra, A.; Eriksson, M.; Ogle, C. A.;
Seagle, P. Chem.Eur. J. 1999, 5, 2680−2691.
(16) Kharasch, M. S.; Tawney, P. O. J. Am. Chem. Soc. 1941, 63,
2308−2315.
(17) Yoshikai, N.; Nakamura, E. Chem. Rev. 2012, 112, 2339−2372.
(18) Yamanaka, M.; Kato, S.; Nakamura, E. J. Am. Chem. Soc. 2004,
126, 6287−6293. See also ref 17.
(38) (a) Krause, N. Angew. Chem., Int. Ed. 1999, 38, 79−81 and
references therein.
For a bona fide higher order organocuprate
+
(Ph3CuLi2·Ph2CuLi = Ph3CuLi3 Ph2Cu−) see: (b) Bertz, S. H.;
Dabbagh, G. J. Am. Chem. Soc. 1988, 110, 3668−3670. (c) Olmstead,
M. M.; Power, P. P. J. Am. Chem. Soc. 1989, 111, 4135−4136.
(39) (a) Snyder, J. P.; Bertz, S. H. J. Org. Chem. 1995, 60, 4312−
4313. (b) Snyder, J. P.; Spangler, D. P.; Behling, J. R.; Rossiter, B. E. J.
Org. Chem. 1994, 59, 2665−2667. (c) Snyder, J. P.; Tipsword, G. E.;
Spangler, D. P. J. Am. Chem. Soc. 1992, 114, 1507−1510.
(19) (a) Bertz, S. H. J. Org. Chem. 1979, 44, 4967−4969. For a
useful cuprate geminal alkylation, see: (b) Bertz, S. H. Tetrahedron
Lett. 1980, 21, 3151−3154. See also ref 20.
(20) (a) Bertz, S. H.; Moazami, Y.; Murphy, M. D.; Ogle, C. A.;
Richter, J. D.; Thomas, A. A. J. Am. Chem. Soc. 2010, 132, 9549−9551.
For invention based on reduction potentials, see: (b) Bertz, S. H.;
Dabbagh, G.; Williams, L. M. J. Org. Chem. 1985, 50, 4414−4415. For
thiophilic addition of alkyllithiums, see: (c) Bailey, W. F.; Bartelson, A.
L.; Wiberg, K. B. J. Am. Chem. Soc. 2012, 134, 3199−3207. For
carbophilic addition of allyllithium, see: (d) Shi, J.; Heimgartner, H.
Helv. Chim. Acta 1996, 79, 371−384.
(40) A stable ionic structure must preserve local electroneutrality;
that is, an ionic structure will be stable to the extent that the sum of
the strengths of the electrostatic bonds to an anion equals its charge.
Pauling, L. J. Am. Chem. Soc. 1929, 51, 1010−1026.
(21) (a) Bartholomew, E. R.; Bertz, S. H.; Cope, S. K.; Murphy, M.
D.; Ogle, C. A.; Thomas, A. A. Chem. Commun. 2010, 46, 1253−1254.
(b) Bertz, S. H.; Murphy, M. D.; Ogle, C. A.; Thomas, A. A. Chem.
Commun. 2010, 46, 1255−1256.
7838
dx.doi.org/10.1021/om300622c | Organometallics 2012, 31, 7827−7838