ARTICLE

pubs.acs.org/Organometallics

How Racemic Secondary Alkyl Electrophiles Proceed to

Enantioselective Products in Negishi Cross-Coupling Reactions
Xufeng Lin,*,†,‡ Jian Sun,‡ Yanyan Xi,† and Delian Lin†
†
State Key Laboratory of Heavy Oil Processing, College of Chemistry & Chemical Engineering, China University of Petroleum
(East China), Qingdao, 266555, People's Republic of China
‡
Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, People's Republic of China
bS Supporting Information
ABSTRACT: Two mechanisms, namely, the Ni(0)
Ni(II) and Ni-
(I)
Ni(III) mechanisms, for nickel-bis(oxazolinyl)pyridine complex
catalyzed Negishi cross-coupling reaction were investigated with den-
sity functional calculations. The Ni(I)
Ni(III) mechanism, containing
sequential steps of transmetalation
oxidative addition
reductive elim-
ination, is more favorable than the Ni(0)
Ni(II) mechanism, based on
the energetic span model. The enantioselectivity of the coupled product
from a racemic secondary alkyl electrophile was calculated by the
relative reaction rate (rS/rR) of the reductive elimination step that
forms the coupled product in the S-enantiomer over that leading to the R-enantiomer. The rS/rR can be calculated from the relative
free energy of the transition states for these two reductive elimination pathways in the Ni(I)
Ni(III) mechanism. The calculated
enantioselectivity for the model reaction is consistent with the experimental report. The influence of the asymmetric steric hindrance
of the catalyst ligand on the reductive elimination step is also discussed.

’ INTRODUCTION there was no report of mechanistic studies to account for the

Ni or Pd complex catalyzed cross-coupling reactions between stereoconvergence, the involvement of a radical in these systems
organo-electrophiles (typically organohalides) and organo-nu- has been suggested.4b,c
cleophiles (typically alkylmetal halides, e.g., EtZnI) are widely Due to the great significance of asymmetric catalysis, the con-
used methods for building desired molecules by C
C bond version of racemic alkyl halides to high ee value products greatly
formation.1,2 However, the formation of a C(sp3)
C(sp3) bond interests us. The explanation/prediction of ee values calls for
with an alkyl electrophile, especially with a secondary alkyl deep insight into the reaction mechiansm for Ni complex catalyzed
electrophile, is much more difficult than a C
C bond formation Negishi cross-coupling reactions. Although the textbook me-
with an aryl or vinyl eletrophile. In recent years obvious progress chanism for cross-coupling reactions containing sequential steps
has been made in developing cross-coupling reactions of sec- of oxidative addition
transmetalation
reductive elimination
ondary alkyl electrophiles by Fu and co-workers.3
5 They reported has been established for more than three decades,1 several re-
nickel-catalyzed reactions using the bis(oxazolinyl)pyridine ligands searchers6
9 have pointed out that this mechanism is not feasible
that successfully achieved Negishi coupling of unactivated sec- for C(sp3)
C(sp3) cross-coupling reactions. Actually in the
ondary alkyl bromides.3 The variety of catalyst ligand scaffolds expanding numbers of works on cross-coupling reactions, it
and the steric properties of the alkyl electrophiles may lead to can be found in the literature that the textbook mechanism
different product selectivity. The first example of catalytic asym- may not be feasible (or need modification) even for some non-
metric cross-coupling of secondary electrophiles was achieved by C(sp3)
C(sp3) cross-coupling reactions.2d,10,11 The reaction
exploiting the C2 symmetry of the catalyst ligand, when employ- mechanism for C(sp3)
C(sp3) cross-coupling is still elusive in
ing R-bromoamides as substrates.4 In ref 4a they established that the literature. For nickel-terpyridine complex catalyzed Negishi
a family of reaction partners, racemic secondary benzylic halides, alkyl
alkyl cross-coupling reactions, it has been experimentally8
can be coupled with organozinc reagents in very good enantio- and theoretically9 suggested that the sequential steps of oxidative
meric excess (ee), as expressed in reaction A in Scheme 1 (see the addition
reductive
transmetalation working with the Ni-
notations of the catalyst ligand as well). Upon investigation of a (I)
Ni(III) cycles may be plausible.
variety of conditions and substituents R and R2, they determined Similar to terpyridines, Pybox and its derivatives are another
that Negishi reactions of 1-bromoindanes proceed in yields family of trinitrogen ligands. In this work we explored the
ranging from 39% to 89% and the ee value ranging from 90%
to 98%. (R)-(i-Pr)-Pybox was also used for the synthesis of Received: December 28, 2010
(R)-1-methyl-3-oxo-indene, with an ee value of 92%.4b Although Published: June 01, 2011

r 2011 American Chemical Society 3284 dx.doi.org/10.1021/om1012049 | Organometallics 2011, 30, 3284–3292
Organometallics ARTICLE

Scheme 1. Real4a (A) and Model (B) Negishi Cross-Coupling (IRC) computations were done to confirm the transition states con-
Reactions Studied in this Papera necting the appropriate reactants and products.16

’ RESULTS AND DISCUSSION

To save computational resources, Pybox instead of S-Pybox
(see Scheme 1 for the notations) was used to examine the
potential energy surfaces of the Ni(0)
Ni(II) and Ni(I)
Ni(III)
mechanisms in the first two subsections (subsections A and B).
(A) Ni(0)
Ni(II) Mechanism. The textbook1 Ni(0)
Ni(II)
mechanism for reaction B contains the following three steps:
Oxidative addition : NiðPyboxÞ þ Br-R 1 f Br-NiðPyboxÞ-R 1
ð1Þ

Transmetalation : Br-NiðPyboxÞ-R 1
þ CH3 ZnBr f CH3 -NiðPyboxÞ-R 1 þ ZnBr2 ð2Þ

Reductive elimination : CH3 -NiðPyboxÞ-R 1 f NiðPyboxÞ þ R 1 -CH3

ð3Þ
a
R1 and R2 denote R-indenyl and methyl as indicated, respectively. Two
catalyst ligands, namely, Pybox and S-Pybox, as indicated, were exam-
ined in this paper. Figure 1 presents the optimized geometries of species involved
in eqs 1
3, and Figure 2 depicts the free energy profiles of these
potential energy surfaces (PES) of two mechanisms for reaction three steps. For the transmetalation step (eq 2), a reactant
B to model reaction A (Scheme 1) by density functional theory complex (denoted as RC2) and a product complex (denoted
(DFT) calculations. The Ni(I)
Ni(III) mechanism instead of as PC2) were found, and a transition state (TS2) containing a
the Ni(0)
Ni(II) one (textbook mechanism applying to reac- four-membered-ring Br 3 3 3 Ni 3 3 3 C 3 3 3 Zn moiety can be identi-
tion B) is feasible based on the energetic span model. Then, fied. For the reductive elimination step (eq 3) starting from R1-
based on the Ni(I)
Ni(III) catalytic cycle, the relative rates of Ni(Pybox)-CH3, a three-membered-ring Ni 3 3 3 C 3 3 3 C moiety
formation of products (rS/rR) in both S- and R-enantiomers were can be seen in the transition state (TS3) structure. As proposed
calculated. The calculated rS/rR for reaction B is consistent with by Kozuch and co-workers,17 the energetic span (δE) of a
the experimental report by Fu and co-workers.4a We believe that catalytic cycle determines if it is favorable or not. The lower
this work provides deep insight into the mechanism for metal- δE, the more favorable the catalytic cycle is. The δE can be
catalyzed cross-coupling with C(sp3)
C(sp3) bond formation calculated from the energy difference between the turnover
reactions, especially for the enantioselectivity-determining step- frequency determining transition state (TDTS) and the turnover
(s). More importantly, the method for calculation/prediction of frequency determining intermediate (TDI). Although the transi-
the enantioselectivity for this type of reaction will help synthetic tion-state structure for the oxidative addition step (eq 1) was not
chemists design/select their catalyst ligands and the structure of examined, Figure 2 provides a lower limit of δE for the Ni-
reactants for asymmetric synthesis. (0)
Ni(II) catalytic cycle. If the free energy of the transition
state (TS1) for eq 1 is not high enough to be a TDTS, then the
TDI is the RC2 formed from the R1-Ni(Pybox)-Br and
’ COMPUTATIONAL METHODS CH3ZnBr in the transmetalation step, and the TDTS is the
Density functional calculations were performed using the hybrid
transition state of the reductive elimination step (TS3 in eq 3).
B3LYP12 exchange and correlation functionals in order to explore the The energetic span calculated from the free energy of these two
potential energy surfaces of the model Ni-catalyzed Negishi alkyl
alkyl structures is 53.7 kcal/mol. If the free energy of TS1 is high
cross-coupling reactions. The 6-31G* basis set was used for all C, H, O, enough to be a TDTS, the real energetic span will be larger than
N, Ni, Zn, and Br atoms. Five component d functions were employed in 53.7 kcal/mol. The lower limit of δE for the Ni(0)
Ni(II)
the calculations. The B3LYP/6-31G* level of theory is appropriate to catalytic cycle will be compared to that for the Ni(I)
Ni(III)
calculate the PES of Ni-trinitrogen ligand complex catalyzed cross- catalytic cycle, reported later, and conclusions will be drawn then.
coupling reactions, as reported in the previous study.9 (B) Ni(I)
Ni(III) Mechanism. The catalytic cycle working with
The PES for the reactions of interest were explored by optimizing the Ni(I)
Ni(III) mechanism for reaction B, similar to that
geometries in the energy minima for the reactants, the intermediates, reported in our previous work,9 can be written as
and the products, and the first-order saddle points for transition states
using the Gaussian 03 program suite (employing C.02 version).13 Vibra- NiðPyboxÞ-CH3 þ R 1 -Br f NiðPyboxÞ-Br þ R 1 -CH3 ð4Þ
tional analyses were performed to confirm energy minima and first-order
saddle points as well as to obtain the zero-point-corrected energies NiðPyboxÞ-Br þ CH3 ZnBr f NiðPyboxÞ-CH3 þ ZnBr2 ð5Þ
(ZPE) and free energies (at 273.15 K) of the optimized geometries. As
reaction A was carried out in dimethylacetamide (DMA) as solvent, bulk Equation 4 is the combination of the following three steps:
solvation effects were examined with the polarized continuum model
(PCM)14 utilizing DMSO to model DMA.15 Intrinsic reaction coordinate NiðPyboxÞ-CH3 þ R 1 -Br f Br-NiðPyboxÞ-CH3 þ R 1 3 ð6Þ

3285 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292

Organometallics ARTICLE

Figure 1. Optimized geometries at B3LYP/6-31G* level of theory for the species involved in eqs 1 and 2. Hydrogen atoms on Pybox are deleted for

).
clarity in this and the following figures. Key bond lengths are indicated (in Å

Br-NiðPyboxÞ-CH3 þ R 1 3 f CH3 -NiBrðPyboxÞ-R 1 ð7Þ

CH3 -NiBrðPyboxÞ-R 1 f NiðPyboxÞ-Br þ R 1 -CH3 ð8Þ

1
Equation 6 is the Br transfer step, eq 7 the R radical attack
step,18,19 and eq 8 the reductive elimination step that results in
the coupled product. The combination of eqs 6 and 7 can be
considered as the oxidative addition step. The overall reaction of
eqs 4 and 5 can be written as
R 1 -Br þ CH3 ZnBr f R 1 -CH3 þ ZnBr2 ½NiðPyboxÞ-Br as catalyst
ð9Þ

(a) eq 4: Ni(Pybox)-CH3 in Reaction with Alkyl Bromide.

Figure 3 presents the optimized geometries of the two Ni(I)
species involved in eq 4. As the catalyst ligand of Pybox has a C2v Figure 2. Free energy profile of the Ni(0)
Ni(II) mechanism.
symmetry and (S)-R1-Br (see Figure 1) and (R)-R1-Br are the
mirror images of each other, all the reactants, intermediates,
transition states, and products of eq 4 in S-enantiomer form are “-S” was used for all the transition states and Ni(III) intermedi-
all the mirror images of their corresponding R-enantiomer. The ates in the pathways that lead to the coupled product R1-CH3 in
free energy profiles of eq 4 with the S-enantiomer form should be the S-enantiomer form, and here the results of eq 4 are presented
the same as those with R-enantiomer form (vide infra). A suffix of only for the S-enantiomer production first.
3286 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292
Organometallics ARTICLE

Figure 4 presents all the geometries of the transition states, Ni(III)-S intermediate [CH3-Ni(Pybox)Br-R1] proceeding by
intermediates, and products involved in eqs 6
8. The transition TS7-S has a C(1)
Ni bond length of 2.015 Å
. The transition-
state of the Br transfer step (eq 6), TS6-S, occurs at the state structure (TS8-S) of the reductive elimination step has
Br 3 3 3 C(1) distance of 2.613 Å
and the Br 3 3 3 Ni distance of a trigonal C(1)
Ni
C(2) geometry with C(1)
Ni, Ni
C(2), and
2.557 Å
. The C(1), Br, and Ni atoms are roughly in a straight line. C(1) 3 3 3 C(2) distances of 2.150, 2.097, and 2.130 Å
, respectively.
The products proceeding through TS6-S are a R1 radical and the The products proceeding by TS8-S consist of the coupled
Ni(II) species [Br-Ni(Pybox)-CH3], in which the Br
Ni bond product R1-CH3 as the S-enantiomer, as well as another Ni(I)
length is 2.498 Å
. Br-Ni(Pybox)-CH3 also has a square-based species, Ni(Pybox)-Br.
pyramidal ML5 structure, as in the case of I-Ni(tpy)-CH3 in the The free energy profiles of these three steps are presented in
previous work.9 R1 radical attack onto the Ni(II) species occurs Figure 5. The Br transfer step (eq 6) has a free energy of
at the other side of the Pybox plane with respect to the Br transfer activation of 14.2 kcal/mol and is slightly endothermic by
step. The transition state of this step (TS7-S in Figure 4) 2.3 kcal/mol. The R1 radical attack step (eq 7) has a free energy
occurs at the C(1) 3 3 3 Ni distance of 2.852 Å
in the reaction of activation of 14.4 kcal/mol and is slightly exothermic by
coordinate in terms of the C
Ni bond formation. The
(1)
1.5 kcal/mol. So, the overall oxidation addition step (eqs 6 and 7)
has two mildly large free energies of activation, and the free
energy change is very small. The reductive elimination step
(eq 8) needs a higher free energy of activation (20.3 kcal/mol)
than the preceding two steps and is highly exothermic by a free
energy decrease of 45.2 kcal/mol. The large exothermicity of eq 8
makes the overall reaction of eq 4 exothermic by 44.4 kcal/mol.
These data indicate that, for sequential steps of Br transfer,
radical attack, and reductive elimination, the former two steps are
in fast equilibrium, and the reductive elimination step is irrever-
sible and rate-determining in the process of Ni(Pybox)-CH3
reacting with R1-Br. This is dramatically different from the case of
Ni(tpy)-CH3 reacting with propyl/isopropyl iodide, in which the
iodine atom transfer step is rate-determining.9
Figure 3. Optimized geometries of the two Ni(I) species involved in (b) eq 5: Ni(Pybox)-Br Transmetalating with CH3ZnBr.

).
eq 4. Key bond lengths are indicated (in Å For Ni(Pybox)-Br transmetalating with CH3ZnBr to afford

).
Figure 4. Optimized geometries of the transition states and intermediates in eqs 6
8 (or eq 4). Key bond lengths are indicated (in Å

3287 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292

Organometallics ARTICLE

Ni(Pybox)-CH3, a reactant complex (denoted as RC5) and a corresponding to the R-enantiomer. Since one of the main
product complex (PC5), which are more stable than the separated points in this paper is to show how the enantioselecitive product
products, were found. A four-membered-ring transition-state is produced in the presence of the C2 S-Pybox ligand, it is
structure similar to the one for Ni(bipyridine)-I transmetalating necessary to verify that the C2v symmetric ligand of Pybox does
with CH3ZnI was reported by C
ardenas and co-workers.10b As not lead to any enantioselectivity. In this work the transition
the product of this transmetalation step is the reactant of the states, intermediates, and the products involved in eq 4 that lead
bromine transfer step (eq 6), the relative energy of the product in to (R)-R1-CH3 were also examined structurally and energetically.
eq 5 is set to 0 kcal/mol. Therefore, if the free energy profiles for These geometries are, not surprisingly, the mirror images of the
eqs 5
8 (in Figure 7 and Figure 5) are put together, one may see corresponding ones presented in Figure 4 (therefore, not pre-
that the TDI for the Ni(I)
Ni(III) catalytic cycle is RC5 in the sented in this paper), and the free energy profile is the same (the
transmetalation step (eq 5), and the TDTS is TS8-S in the difference is no larger than 0.03 kcal/mol) as the one presented
reductive elimination step (eq 8). This situation is quite similar to in Figure 5. Thus, for the Pybox ligand without a substitution
the Ni(0)
Ni(II) cycle. However, the energetic span for the group, the rates of all the steps producing the coupled product in
Ni(I)
Ni(III) cycle is 42.8 kcal/mol, which is significantly lower the R-enantiomer are the same as those producing the S-enantio-
than the one (with the lower limit of 53.7 kcal/mol) for the mer, which leads to no enantioselectivity for the cross-coupling
Ni(0)
Ni(III) cycle (see subsection A). Therefore, the Ni- reactions. On the other hand, computation of the pathways for
(0)
Ni(II) mechanism is discarded in this paper in the following producing the R-enantiomer with the C2v ligand shows the
part for enantioselectivity calculation. computational error contributes little (<0.03 kcal/mol) to the
(c) No Enantioselectivity of the Coupled Product in the Case difference of energy profiles for both enantiomers when the
of C2v Pybox Used As Ligand. As is mentioned above, all the ligand is replaced with a C2 ligand (vide infra).
reactants, intermediates, transition states, and products of eq (C) Ni(I)
Ni(III) Mechanism in the Case of C2 Symmetric
4 in S-enantiomer form are all chirally symmetric to those S-Pybox Used As Catalyst Ligand. (a) Structural and Energetic
Characterization of eqs 7 and 8 with the S-Pybox Ligand. In the
above sections it has been demonstrated that the textbook mechan-
ism of sequential steps of oxidative elimination, transmetalation,

Figure 5. Free energy profile for the reaction between Ni(Pybox)-CH3 Figure 7. Relative free energy profile for the transmetalation step (eq 5)
and R1-Br (eq 4 or eqs 6
8). in the Ni(I)
Ni(III) mechanism.

Figure 6. Optimized geometries of the species involved in eq 5. Key bond lengths are indicated (in Å
).

3288 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292

Organometallics ARTICLE

and reductive elimination working with the Ni(0)
Ni(II) couple following four equations when the C2 S-Pybox ligand is used:
is not feasible thermodynamically, while the mechanism con-
Br-NiðS-PyboxÞ-CH3
taining sequential steps of oxidative addition, reductive elimina-
tion, and transmetalation working with the Ni(I)
Ni(III) þ R 3 f CH3 -NiBrðPyboxÞ-R 1 ½S-enantiomer, S-NiðIIIÞ-S
1

couple is feasible. The Ni(I) species [Ni(Pybox)-CH3] acts as ðS-7-SÞ

a catalyst for reaction B. The oxidative addition of aryl iodide
occurring at a Ni(I)-binitrogen ligand complex was also pro- Br-NiðS-PyboxÞ-CH3
posed by Phaple et al.10 Since only eqs 7 and 8 account for the
formation of a new bond in the C(1) site in the R1 moiety, only þ R 3 f CH3 -NiBrðPyboxÞ-R 1 ½R-enantiomer, S-NiðIIIÞ-R
1

these two steps were examined in this work as presented ðS-7-RÞ

hereafter.
If the catalyst ligand has a lower symmetry than Pybox, such as
CH3 -NiBrðS-PyboxÞ-R 1 ðS-enantiomerÞ f NiðPyboxÞ-Br
(S)-isopropyl-Pybox (S-Pybox) or (R)-isopropyl-Pybox, the
structures of TS7-S, Ni(III)-S, and TS8-S will no longer be þ ðSÞ-R 1 -CH3 ðS-8-SÞ
chirally symmetric with those of TS7-R, Ni(III)-R, and TS8-R,
due to the presence of asymmetric steric hindrance. Thus, eqs 7 CH3 -NiBrðS-PyboxÞ-R 1 ðR-enantiomerÞ f NiðPyboxÞ-Br
and 8 for the C2v Pybox ligand should be replaced with the þ ðRÞ-R 1 -CH3 ðS-8-RÞ
A prefix of “S-” was used in the notations of the above four
equations and in all the structures involved in these equations for
indicating the ligand used here is S-Pybox. A suffix of “-S” or “-R”
was used to represent the structures or equations that lead to the
coupled product as an S- or R-enantiomer. For example, S-TS8-R
denotes the transition state of eq S-8-R, the C(1) atom of which is
in the R-conformation.
Figure 8 shows the optimized geometries of Br-Ni(S-Pybox)-
CH3 [S-Ni(II)] and the R1 radical, as the reactants. The R1
radical attacks the Ni atom in S-Ni(II) (eqs S-7-S and S-7-R)
on the other side of the Pybox plane with respect to the Br atom.
The Ni atom in S-Ni(II) attacking the C(1) atom in the R1 radical
Figure 8. Optimized geometries of the reactant for the radical attack
at the Zþ side (defined in Figure 8) matches the transition state
S-TS7-S (see Figure 9) at the Ni 3 3 3 C(1) distance of 2.897 Å
. In
step (eqs S-7-S and S-7-R) in the Ni(S-Pybox)-CH3-catalyzed cross-
coupling reaction. For a better description of the two sides of the R1
radical when it goes to the Ni atom (since they lead to different contrast, when the R1 radical attack occurs at its Z
side, the
enantiomers in the coupled product), Zþ and Z
directions were transition state (S-TS7-R in Figure 10) occurs at the Ni 3 3 3 C(1)
defined by putting the benzene ring in the R1 radical on the XY plane distance of 3.090 Å
. S-TS7-R is not the mirror image of S-TS7-S,
of a 3D Cartesian coordinate system as indicated. R1 radical is symmetric and S-TS7-R comes earlier than S-TS7-S along the reaction
with the XY plane. coordinate in terms of the Ni
C(1) bond formation process.

Figure 9. Optimized geometries for the intermediates, the transition states, and the products in the reaction that forms a cross-coupled product in
S-enantiomer form with the ligand S-Pybox.

3289 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292

Organometallics ARTICLE

Figure 10. Optimized geometries for the intermediates, the transition states, and the products in the reaction that forms cross-coupled product in
R-enantiomer form with the ligand S-Pybox.

This implies that radical attack in the Z
direction may take place
more easily than that in the Zþ direction.
The Ni(III) species process that proceeds through the transi-
tion states S-TS7-S is denoted as S-Ni(III)-S (Figure 9), where
the Ni
C(1) bond length is 2.033 Å
and the C(1) atom is in the
S-conformation. The Ni(III) species that proceeds via S-TS7-R is
denoted as S-Ni(III)-R (Figure 10), where the Ni
C(1) bond
length is 2.024 Å
and the C(1) atom is in the R-conformation. The
reductive elimination of S-Ni(III)-S affords the coupled product
(S)-R1-CH3 with a transition state S-TS8-S, in which the
Ni 3 3 3 C(1) and C(1) 3 3 3 C(4) distances are 2.055 and 2.042 Å
,
respectively. The reductive elimination of S-Ni(III)-R affords the
coupled product (R)-R1-CH3, with a transition state, S-TS8-R, in
which the Ni 3 3 3 C(1) and C(1) 3 3 3 C(4) distances are 2.273 and
1.993 Å
, respectively. This geometrical information indicates that
S-TS8-R occurs later than S-TS8-S along the reaction coordinate
in terms of C(1)
C(4) bond formation and Ni
C(1) and Ni

C(4) bond cleavages. This implies that the reductive elimination
of S-Ni(III)-S may occur more easily than that of S-Ni(III)-R.
The free energy profiles shown in Figure 11 support the two
hypotheses in the above two paragraphs. Compared to the free
energy profile for the case of the Pybox ligand (shown with a
dotted line), the use of S-Pybox leads to a mild modification of
this profile. The radical attack step in the Zþ direction has a free Figure 11. Free energy profiles of eqs S-7-S/R and S-8-S/R for
energy of activation of 16.5 kcal/mol, while that in the Z
formation of cross-coupled product R1-CH3 as both S (solid triangles
direction has a free energy of activation of 8.9 kcal/mol. The with dashed line) and R (solid stars with dashed line) enantiomers. Here
S-Ni(III)-R species is kinetically more easily formed than S-Pybox is used as the ligand of the Ni(I) catalyst. The free energy profile
S-Ni(III)-S. S-Ni(III)-R is a bit more stable than S-Ni(III)-S, of eqs 7 and 8 where Pybox is used as the ligand (open circles with dotted
by about 2.8 kcal/mol. However, situation changes in the reductive line) is shown for comparison.
elimination step (eq S-8-S or -R). The free energy of activation in
eq S-8-S for (S)-R1-CH3 formation is 17.7 kcal/mol, and that in S-Pybox is used as the ligand, the pathways that lead to the
eq S-8-R for (R)-R1-CH3 formation is 27.5 kcal/mol. R-enantiomer and S-enantiomer share the same TDI, but have
(b) Enantioselectivity of the Cross-Coupled Product for different TDTSs. The difference in the energetic spans for these
Reaction B. As described in part (b) in subsection B, for the two pathways is determined by the difference of the free energies
Ni(I)
Ni(III) catalytic cycle, the TOF-determining intermedi- for S-TS8-R and S-TS8-S, as can be written as
ate is the reactant complex from Ni(pybox)-Br and CH3ZnBr
(RC5), and the TOF-determining transition state is the transi-
tion state of the reductive elimination step (TS8-S/R). When δER
δES ¼ GS-TS-8R
GS-TS-8S ðIÞ

3290 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292

Organometallics ARTICLE

Figure 12. Conformations of the Ni
C(1) bond in the three kinds of possible first-order saddle points when exploring the transition-state structure of
the reductive elimination step (eq 8). The round cycle represents the C(1) atom. See text for a detailed description.

Possible side reactions of the R1 radical were not considered elimination steps in producing both S- and R-enantiomers, with
here, because they may not influence the overall enantioselec- both the Pybox and S-Pybox ligands.
tivity of the coupled product. Therefore, the S/R ratio can be From inspection of the transition-state structures of the
calculated from the Boltzmann distribution of each transition reductive elimination step with the Pybox ligand (eq 8), TS8-S
state20 for the reductive elimination step (i.e., GS-TS-8S and GS-TS-8R) (see Figure 4) and TS8-R, it was found that the Ni 3 3 3 C(1) bond
at the temperature of 273 K, as the following: adopts the conformation as indicated in conformation 1 in
  Figure 12, where the Ni
C(2)H3 bond lies between the C(1)

GS-TS-8R
GS-TS-8S C(a) and C(1)
C(b) bonds. Actually when exploring the PES,
½S=½R ¼ rS =rR ¼ exp ðIIÞ
RT another two first-order saddle points, which also have imaginary
vibrations of the C(1) 3 3 3 C(2) bond formation/cleavage, were
The enantiomeric excess is
also found. In these two first-order saddle points the Ni 3 3 3 C(1)
½S=½R
1 bond adopts the conformation as indicated in conformations 2
ee ¼ ðIIIÞ and 3 in Figure 12. The Ni
C(2)H3 bond is between the C(1)
H
½S=½R þ 1
and C(1)
C(a) bonds and between the C(1)
H and C(1)
C(b)
in the case where the S-enantiomer is dominant over the R- bonds. These two kinds of first-order saddle points have higher
enantiomer. Using the data in Figure 11 one can easily obtain that energies/free energies than those of TS8-S or TS8-R.
the rS/rR is 8.9  106 with the gas phase model and 4.3  106 Because the indene substrate also has a bulky benzene ring, the
with consideration of the bulk solvation effect (see Computa- Ni
C(1) bond could not adopt the three conformations freely
tional Method), which corresponds to an ee value of more than when the isopropyl groups are present on the S-Pybox ligand.
99%. Although qualitatively this result is consistent with the S-TS8-S can still adopt conformation 1, while S-TS8-R cannot
experimentally measured value (92%),4a there is an overestima- adopt conformation 1 because there is a strong steric hindrance
tion somehow. As it has been demonstrated that the ee value between the benzene ring and the isopropyl group. So S-TS8-R
changes from 0 to close to 1 when the C2v ligand is replaced with has to adopt conformation 2 to lower the steric hindrance, and
the C2 ligand, it is reasonable to anticipate that the incorporation this may lead to a higher energy/free energy of S-TS8-R than that
of solvent molecules may be disadvantageous to some degree to of S-TS8-S (GS-TS8-R > GS-TS8-S), which results in the enantios-
the asymmetry of the ligand, therefore lowering the calculated ee electivity of the coupled product.
value. However, presently the solvation effect is considered only
by the single-point PCM calculation of the geometries optimized
in the gas-phase calculation. We believe the prediction could be ’ CONCLUSION
more accurate if the geometry optimizations are performed with (1) By calculation of the energetic spans for both the Ni-
the PCM calculation. (0)
Ni(II) mechanism, which contains sequential steps
(c) Discussion of Steric Effect of the S-Pybox Ligand for the of oxidative addition, transmetalation, and reductive
Reductive Elimination Step. It has been demonstrated that the elimination, and the Ni(I)
Ni(III) mechanism, which
reductive elimination step is irreversible and rate-determining. contains sequential steps of transmetalation, oxidative
The enantioselectivity of the coupled product may be accounted addition, and reductive elimination, the latter mechanism
for by the free energy difference of the transition states of the was found more favorable than the former one.
reductive elimination steps leading to the S-enantiomer and R- (2) The enantioselectivity of the coupled product is deter-
enantiomer products. It may be helpful to carry out conforma- mined by the difference between the free energies of the
tional analysis for the transition-state structures of the reductive transition states of the reductive elimination steps for
3291 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292
Organometallics ARTICLE

model reaction B. The free energy of activation for the (6) (a) Terao, J.; Watanabe, H.; Ikumi, A.; Kuniyasu, H.; Kambe, N.
reductive elimination of the S-Ni(III)-R species in gen- J. Am. Chem. Soc. 2002, 124, 4222–4223. (b) Terao, J.; Todo, H.;
erating the coupled product as the R-enantiomer is higher Watanabe, H.; Ikumi, A.; Kambe, N. Angew. Chem., Int. Ed. 2004,
than that of the S-Ni(III)-S species in generating the 43, 6180–6182.
S-enantiomer, which may account for the high enantios- (7) (a) Hadei, N.; Kantchev, E. A. B.; O’Brien, C. J.; Organ, M. G.
Org. Lett. 2005, 7, 3805–3807. (b) Hadei, N.; Kantchev, E. A. B.;
electivity of reaction B. The method for calculating/ O’Brien, C. J.; Organ, M. G. J. Org. Chem. 2005, 70, 8503–8507.
predicting the enantioselectivity of the coupled product (8) Jones, G. D.; Martin, J. L.; McFarland, C.; Allen, O. R.; Hall,
may help synthetic chemists design/select their catalyst R. E.; Haley, A. D.; Brandon, R. J.; Konovalova, T.; Desrochers, P. J.;
ligands and reactant substrates. Pulay, P.; Vicic, D. A. J. Am. Chem. Soc. 2006, 128, 13175–13183.
(3) The conversion of racemic alkyl electrophiles to highly (9) Lin, X.; Phillips, D. L. J. Org. Chem. 2008, 73, 3680–3688.
enantioselective coupled product is due to the presence of (10) (a) Phapale, V. B.; Bu~ nuel, E.; García-Iglesias, M.; C
ardenas,
asymmetric steric hindrance of alkyl groups in the S-Pybox D. J. Angew. Chem., Int. Ed. 2007, 46, 8790–8795. (b) Phapale, V. B.;
ligand, which makes the transition state of the reductive Guis
an-Ceinos, M.; Bu~ nuel, E.; C
ardenas, D. J. Chem.—Eur. J. 2009,
elimination step that proceeds to the S-enantiomer dif- 15, 12681–12688.
ferent from that which proceeds to the R-enantiomer. (11) One step of oxidative addition and reductive elimination
reported: Wang, M.; Lin, Z. Organometallics 2010, 29, 3077–3084.
(12) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098–3100. (b) Lee, C.;
’ ASSOCIATED CONTENT Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789. (c) Becke, A. D.
J. Chem. Phys. 1993, 98, 5648–5653.
bS Supporting Information. Tables containing atomic co- (13) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
ordinates, calculated energies without and with zero-point energy
correction (in au), and free energies (in au, at 313.15 K) for all of Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson,
the stationary structures reported in this paper. This material is
G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
available free of charge via the Internet at http://pubs.acs.org. Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.;
Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo, C.;
’ AUTHOR INFORMATION Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Corresponding Author Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
*E-mail: [email protected]. Fax: þ86-532-86981565. S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.;
Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
’ ACKNOWLEDGMENT Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
This research has been supported by the National Science Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
Foundation of China (21003159) and grants from China Uni- 03, Revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
versity of Petroleum (Y0904044, Y0904045). X.F.L. thanks The (14) (a) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981,
University of Hong Kong for the computational resources 55, 117–129. (b) Miertus, S.; Tomasi, J. Chem. Phys. 1982, 65, 239–245.
provided. (c) Cossi, M.; Barone, V.; Cammi, R.; Tomasi, J. Chem. Phys. Lett. 1996,
255, 327–335. (d) Cossi, M.; Scalmani, G.; Rega, N.; Barone, V. J. Chem.
Phys. 2002, 117, 43–54.
’ REFERENCES (15) As the solvent of DMA is not available in Gassuain 03, DMSO
(1) Books, for example: (a) Diederich, F., Stang, P. J. Metal- (dielectric constant ε = 47) was used to model DMA (ε =38).
Catalyzed Cross-Coupling Reactions; Wiley-VCH: Weinheim, Germany; (16) (a) Gonzalez, C.; Schlegel, H. B. J. Chem. Phys. 1989,
1998. (b) de Meijere, A.; Diederich, F. Metal-Catalyzed Cross-Coupling 90, 2154–2161. (b) Gonzalez, C.; Schlegel, H. B. J. Phys. Chem. 1990,
Reactions, 2nd ed.; Wiley-VCH: Weinheim, Germany; 2004. (c) Beller, 94, 5523–5527.
M.; Bolm, C. Transition Metals for Organic Synthesis, 2nd ed.; Wiley-VCH: (17) (a) Kozuch, S.; Amatore, C.; Jutand, A.; Shaik, S. Organome-
Weinheim, 2004. tallics 2005, 24, 2319–2330. (b) Kozuch, S.; Shaik, S. J. Am. Chem. Soc.
(2) For reviews: (a) C
ardenas, D. J. Angew. Chem., Int. Ed. 1999, 2006, 128, 3355–3365. (c) Kozuch, S.; Shaik, S. Acc. Chem. Res. 2011,
38, 3018–3020. (b) Luh, T. Y.; Leung, M. K.; Wong, K. T. Chem. Rev. 44, 101–110.
2000, 100, 3187–3204. (c) C
ardenas, D. J. Angew. Chem., Int. Ed. 2003, (18) Kochi, J. K. Pure Appl. Chem. 1980, 52, 571–605.
42, 384–387. (d) Frisch, A. C.; Beller, M. Angew. Chem., Int. Ed. 2005, (19) Klein, A.; Budnikova, Y. H.; Sinyashin, O. G. J. Organomet.
44, 674–688. (e) Phapale, V. B.; C
ardenas, D. J. Chem. Soc. Rev. 2009, Chem. 2007, 692, 3156–3166.
38, 1598–1607. (20) Balcells, D; Maseras, F. New J. Chem. 2007, 31, 334–343.
(3) (a) Zhou, J.; Fu, G. C. J. Am. Chem. Soc. 2003, 125, 14726–
14727. (b) Powell, D. A.; Fu, G. C. J. Am. Chem. Soc. 2004, 126, 7788–
7789. (c) Zhou, J.; Fu, G. C. J. Am. Chem. Soc. 2004, 126, 1340–1341.
(4) For asymmetric cross-coupling: (a) Arp, F. O.; Fu, G. C. J. Am.
Chem. Soc. 2005, 127, 10482–10483. (b) Fischer, C.; Fu, G. C. J. Am.
Chem. Soc. 2005, 127, 4594–4595. (c) Gonz
alez-Bobes, F.; Fu, G. C.
J. Am. Chem. Soc. 2006, 128, 5360–5361. (d) Smith, S. W.; Fu, G. C.
Angew. Chem., Int. Ed. 2008, 47, 9334–9336.
(5) For cross-coupling of secondary C(sp3) electrophiles with
C(sp2) neucleophiles: (a) Smith, S. W.; Fu, G. C. J. Am. Chem. Soc.
2008, 130, 12645–12647. (b) Lou, S.; Fu, G. C. J. Am. Chem. Soc. 2010,
132, 5010–5011.(c) Lunding, P. M.; Fu, G. C. J. Am. Chem. Soc. 2010,
132, 11027
11029.

3292 dx.doi.org/10.1021/om1012049 |Organometallics 2011, 30, 3284–3292