Nickel-catalysed anti-Markovnikov hydroarylation of unactivated alkenes with unactivated arenes facilitated by non-covalent interactions

Article abstract of DOI:10.1038/s41557-019-0409-4

Anti-Markovnikov additions to alkenes have been a longstanding goal of catalysis, and anti-Markovnikov addition of arenes to alkenes would produce alkylarenes that are distinct from those formed by acid-catalysed processes. Existing hydroarylations are either directed or occur with low reactivity and low regioselectivity for the n-alkylarene. Herein, we report the first undirected hydroarylation of unactivated alkenes with unactivated arenes that occurs with high regioselectivity for the anti-Markovnikov product. The reaction occurs with a nickel catalyst ligated by a highly sterically hindered N-heterocyclic carbene. Catalytically relevant arene- and alkene-bound nickel complexes have been characterized, and the rate-limiting step was shown to be reductive elimination to form the C–C bond. Density functional theory calculations, combined with second-generation absolutely localized molecular orbital energy decomposition analysis, suggest that the difference in activity between catalysts containing large and small carbenes results more from stabilizing intramolecular non-covalent interactions in the secondary coordination sphere than from steric hindrance.

Full text of DOI:10.1038/s41557-019-0409-4

Articles
https://doi.org/10.1038/s41557-019-0409-4
Nickel-catalysed anti-Markovnikov hydroarylation
of unactivated alkenes with unactivated arenes
facilitated by non-covalent interactions
Noam I. Saperꢀ ꢀ1, Akito Ohgi2, David W. Small1, Kazuhiko Semba2, Yoshiaki Nakaoꢀ ꢀ2* and
John F. Hartwigꢀ ꢀ1*
Anti-Markovnikov additions to alkenes have been a longstanding goal of catalysis, and anti-Markovnikov addition of arenes
to alkenes would produce alkylarenes that are distinct from those formed by acid-catalysed processes. Existing hydroaryla-
tions are either directed or occur with low reactivity and low regioselectivity for the n-alkylarene. Herein, we report the first
undirected hydroarylation of unactivated alkenes with unactivated arenes that occurs with high regioselectivity for the anti-
Markovnikov product. The reaction occurs with a nickel catalyst ligated by a highly sterically hindered N-heterocyclic carbene.
Catalytically relevant arene- and alkene-bound nickel complexes have been characterized, and the rate-limiting step was shown
to be reductive elimination to form the C–C bond. Density functional theory calculations, combined with second-generation
absolutely localized molecular orbital energy decomposition analysis, suggest that the difference in activity between catalysts
containing large and small carbenes results more from stabilizing intramolecular non-covalent interactions in the secondary
coordination sphere than from steric hindrance.
inear alkylbenzenes are fundamental precursors to a variety which is often faster into metal–ligand bonds of more electron-poor
of industrially relevant surfactants, detergents, plastics metal centres than of more electron-rich metal centres32.
L
and fine chemicals1 with a global market value of over
A catalytic cycle that operates by an alternative mechanism could
2
US$7.75 billion (ref. ). Although termed ‘linear alkylbenzenes’, enable higher levels of regioselectivity for the linear product and
the commercial process for their formation leads to mixtures of bypass the counterbalancing electronic effects on multiple steps.
isomeric branched alkylbenzenes from the rearrangement of car- Previous research by our groups led to a nickel-catalysed, linear-
bocationic intermediates in the alkylation step3. Because branched selective hydroarylation with electron-deficient arenes33 and various
alkylarenes are more resistant to biodegradation4, their use as sur- electronically activated heteroarenes34, but the reactions of electron-
factants and detergents has led to the extensive pollution of rivers, neutral arenes, such as benzene or alkylarenes, occurred with low
lakes and oceans5,6.
turnovers. Thus, an analysis of the factors controlling activity was
Anti-Markovnikov transition-metal catalysed hydroarylation needed to achieve the first addition of an unactivated arene to a ter-
could lead to n-alkylbenzenes, but the reaction of unactivated minal alkene in good yield with high linear selectivity. In the most
alkenes with simple unactivated arenes lacking a directing group7–10 favourable case, this reaction would occur with a catalyst based on
has not been reported with high selectivity for the linear product an earth-abundant, non-precious metal.
(Fig. 1a). The iridium, ruthenium and platinum systems reported
We report here undirected hydroarylations of electron-neutral
by Periana11–15, Gunnoe16–18 and Goldberg19,20 and their co-workers arenes with unconjugated terminal alkenes that occur in good yields
all catalyse the reaction of benzene (1) with propylene (2) to provide with exceptionally high linear/branched selectivity (>50:1 in most
alkylarenes 3 and 4 with moderate activity, but give nearly 1:1 ratios cases, Fig. 1c) catalysed by a nickel complex containing an outsized
of the constitutional isomers (Fig. 1b). Gunnoe and co-workers N-heterocyclic carbene (NHC). The reaction occurs via the forma-
recently reported a two-step synthesis of n-alkylarenes by rhodium- tion of an alkylnickel–aryl intermediate by an unusual ligand-to-
catalysed oxidative alkenylation and hydrogenation, but the process ligand hydrogen transfer (LLHT) to the coordinated alkene33,35 and
involves two steps and a copper salt as the co-catalyst21,22
.
rate-determining reductive elimination. The results of computa-
The regioselectivity of published undirected hydroarylations is tional studies imply that the attractive interactions of the large NHC
controlled by the regioselectivity of insertion of the alkene into the ligand, rather than steric hindrance, lead to the high activity of the
metal–aryl bond (1,2 vs 2,1 addition)13,17,23–25. By this mechanism, nickel–carbene complex.
high selectivity for the linear product requires that insertion form
a branched alkylmetal intermediate, but such an intermediate is Results and discussion
typically less stable than the linear alkylmetal isomer. Moreover, Reaction development. Initial studies33 showed that the combina-
the rates of these processes are likely limited by the counterbal- tion of [Ni(COD)2] (COD, cycloocta-1,5-diene) and the common
ancing electronic effects on the oxidative addition of an aryl C–H NHC ligand IPr (L1; see Fig. 2a) led to the addition of benzene (1)
bond, which is often faster to more electron-rich metal centres to 1-decene (5) with a high linear/branched ratio of 19:1, but with
than electron-poor metal centres24,26–31, and migratory insertion, a turnover of less than one. To increase the TONs, we evaluated
1Department of Chemistry, University of California, Berkeley, Berkeley, CA, USA. 2Department of Material Chemistry, Graduate School of Engineering,
Kyoto University, Kyoto, Japan. *e-mail: nakao.yoshiaki.8n@kyoto-u.ac.jp; jhartwig@berkeley.edu
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
a
Benzene underwent addition to a series of unactivated alkenes
(Table 1) to give the product with a linear/branched selectivity
>50:1 or with undetectable amounts of branched isomer in all cases.
R
M
H
R
+
+
+
R
6
6
H
H
Undirected
H
Because [L4–Ni(η -C6H6)] and [L5–Ni(η -C6H6)] catalyse in parallel
theisomerizationofalkenes, thereactionofbenzene(1)withinternal
hydroarylation
1
Linear
Branched
Me
b
H
Me
Refs. 11–24
+
a
Me
[L–Ni(η6-C6H6)] (3 mol%)
H
NaH (5 mol%)
1
2
3
4
C8H17
H
Na(acac) (5 mol%)
O
O
+
C8H17
H
O
C
Neat, 120 °C, 24 h
N
N
N
N
O
O
O
O
NCMe
Ph
1
5
6
Ir
N
N
Ru
N
Ph
9 equiv.
1 equiv.
Pt
O
O
O
O
SMe2
Ir
N
Ph
Ph
B
Me
Ph
Ph
Ph
Me
H
O
O
R1
=
N
N
R1
Me
MeO
Ph
Me
R1
TON = 13
l/b = 1.6:1
TON = 10
l/b = 1.6:1
TON = 16
l/b = 1:1
Me
L
Ph
Ph
c
L1
IPr
3%
L2
IPr*
13%
L3
IPr*OMe
13%
NHC
Ni
R2
H
Ar1
Ar2
R1
R2
Ar1
Ar2
R1
+
R
R
H
Ar1
Ar2
MeO
MeO
NHC
Ni
R1
Ligand-to-ligand
Ar1
Ar2
hydrogen transfer
(LLHT)
L4
L5
Ar1: R = Me
Ar2: R = Et
m-XylIPr*OMe
56%
DepIPr*OMe
54%
H
R2
l/b > 50:1
l/b > 50:1
b
L4
Fig. 1 | transition metal-catalysed hydroarylation of unactivated alkenes
with unactivated arenes. a, Linear and branched isomeric products formed
by undirected transition-metal-catalysed hydroarylation. b, State-of-the-art
catalytic systems for the hydroarylation reaction of benzene with
propylene with the turnover numbers (TONs) and linear/branched (l/b)
ratios indicated. c, The linear-selective hydroarylation reaction described
in this report.
%VBur with r = 3.5 Å
Ni
45.7%
42.7%
L4
m-XylIPr*OMe
L1
IPr
N1
C1
N2
%VBur with r = 5.5 Å
Ni
45.6%
L1
IPr
56.5%
L4
m-XylIPr*OMe
For L4:
the hydroarylation of 1-decene (5) with benzene (1) catalysed by
6
4
4
5.00
3.75
2.50
1.25
0
3.00
2.25
1.50
0.75
0
6
[L–Ni(η -C6H6)] complexes as catalyst precursors (Fig. 2a)36. The
3
2
yields were more reproducible, particularly at low catalyst load-
ings (vide infra), when catalytic amounts of NaH and Na(acac)
Expand radius
of sphere
2
1
0
0
6
–0.75
–1.50
–2.50
–3.00
were added. Although the reaction catalysed by [L1–Ni(η -C6H6)]
–1.25
–2.50
–3.75
–5.00
–1
–2
–3
–2
–4
–6
provided only trace amounts of hydroarylation product 6, the reac-
tions catalysed by the complex of the more sterically encumbered
ligand IPr* (L2)37 and its more σ-donating analogue IPr*OMe (L3)
occurred with several turnovers38. Further altering the substitution
of the aromatic rings on the sidearms of the NHC with 3,5-dimeth-
ylphenyl groups (L4, m-XylIPr*OMe)39 or with 3,5-diethylphenyl groups
(L5, DepIPr*OMe) led to catalysts that gave much higher yields of the
–4
–4 –3 –2 –1
0
1
2
3
4
–6 –4 –2
0
2
4
6
Distance from centre of sphere (Å)
Distance from centre of sphere (Å)
r = 3.5 Å
r = 5.5 Å
42.7%
56.5%
6
Fig. 2 | Reaction development and characterization of [L4–Ni(η -C6H6)].
hydroarylation product (56% for L4 and 54% for L5 after 24h). a, Identifying NHC ligands to achieve hydroarylation. Yields and linear/
Alkene isomerization during the reaction by an independent pro- branched (l/b) ratios were determined by GC or NMR spectroscopic analysis
cess33 decreased the concentration of the terminal alkene and reac- of the crude reaction mixture. b, ORTEP diagram of [L4–Ni(η
6
-C6H6)]
tion rates over time (Supplementary Figs. 4 and 5), but the catalysts (thermal ellipsoids are shown at the 50% probability level), calculated
were stable and 75 and 84% isolated yields of 6 were obtained, %V
Bur values and steric map. An expanded radius of 5.5ꢀÅ for the %VBur
respectively, with L4 or L5 as ligand after 5days (Table 1). The linear/ calculation was necessary to account for the remote steric environment
branched selectivity when using L4 or L5 was unprecedently high at of L4. The origins of the steric maps depicted are centred 2.0ꢀÅ away from
>50:1 (see above, Fig. 2a). To directly compare the relative rates of the carbene carbon atoms with the +z axis defined by the carbene carbons
catalysts containing different NHC ligands, we measured the initial and the midpoints between the two backbone carbons, and the (x,z) planes
6
rates of the hydroarylation reaction catalysed by [L–Ni(η -C6H6)] are defined by the two backbone carbons and the carbene carbon atoms.
(Supplementary Fig. 17). The initial rate of hydroarylation with L4 Additional details of the calculations of the %V
Bur values can be found in
as ligand was found to be 23 times greater than the initial rate of
hydroarylation with L1 as ligand and two times greater than the ini- omitted from the ORTEP diagram, and NHC sidearms are represented in
tial rate of hydroarylation with L3 as ligand.
wire format.
the Supplementary Information. For clarity, all hydrogen atoms have been
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
Table 1 | Scope of the hydroarylation reaction
[L–Ni(η6-C6H6)] (3 mol%)
NaH (5 mol%)
R2
H
Na(acac) (5 mol%)
R2
R1
+
R1
Neat, 120 °C, 24 h
H
Arene
Alkene
9 equiv.
1 equiv.
Scope of alkene coupling partner (arene = C6H6)
Me
Scope of arene coupling partner (alkene as indicated)
Me Me
Me Me
Me
Me
m
Me
t-Bu
C8H17 Me
C3H7
C8H17
8
+
C8H17
C3H7
H
p
H
Me
H
Me
Me
Me
5
7
9
10
15
16
17
L4: 75% (5 d)
L5: 84% (5 d)
L4: 76% (5 d)
L4: 98% (5 d)
L4: 90%
L5: 90%
L4: 91%
L5: 91%b
m/p = 2.2:1a
L4: 46%
L4: 51%
Me Me
m
Me
t-Bu
F3C
F
t-Bu
TMSO
C8H17
Me Me
TBSO
C8H17
Si
o
H
H
H
TMSO Me
Me Me
F
F
CF3
20
L4: 95%
11
L4: 96%
L5: 92% (4 h)
12
13
L4: 92%
L5: 90%
18
19
L4: 95%
L5: 96%c
L4: 84%
L5: 89% (8 h)
L4: 88%
L5: 59%c
m/o = 1.7:1a
Entry [Ni] (mol%) NaH/Na(acac) Yielda (%)
H
H
F3C
F3C
C6H13
t-Bu
1
2
96 (1 h)
90 (1 h)
3
3
✓
–
Me
Me
Me
Me
H
14
L5
✓
–
3
4
0.3
0.3
85
6
~ 280 turnovers
21
22
23e
L4: 86%a
L5: 90%c,d
aDetermined by GC or NMR spectroscopic analysis of the crude reaction mixture. b18 equiv. of arene were added. c3 equiv. of arene were added. dNaH and Na(acac) were excluded. eMesitylene was added as
the reaction solvent.
alkenes, such as trans-4-octene (7), also occurred to form the 22 containing a pendant alkene provided tetrahydronaphthalene 23
n-alkylarene products in good yields. Consistent with this observa- in 86% yield (by 1H NMR spectroscopy); no cyclized product from
tion, a mixture of cis- and trans-2-hexene (8) and cis- and trans- the potential intramolecular reaction to form a five-membered ring
3-hexene (9) reacted to give the n-alkylarene product in 98% yield. was observed.
Terminal alkenes bearing substituents at the α-position that inhibit
or prevent isomerization reacted to full conversion within 24h. For Investigation of the reaction mechanism. Having developed a
example, alkene 10 bearing a tertiary carbon at the α-position and method for the hydroarylation reaction between unactivated arenes
alkene 11 bearing a quaternary carbon at the α-position reacted to and alkenes that occurs in good yields with high anti-Markovnikov
give the resulting alkylarene products in >90% yield. Both protected regioselectivity, we sought to understand the origins of the high
primary alcohols (12) and vinyl siloxanes (13) were tolerated under activity of the catalysts containing ligands L4 and L5. To do so, we
the reaction conditions.
determined experimentally the resting state of the catalyst with
The reaction of tert-butylethylene (14) with benzene occurred unhindered and hindered alkenes and the rate-determining step of
in nearly quantitative yield after only 1h under the standard con- the reaction, and we used a computational energy decomposition
ditions (Table 1, entry 1). The reaction with 3mol% catalyst with- analysis40 to identify the intramolecular interactions leading to the
out added NaH and Na(acac) provided the product in nearly the high activity stemming from the large ligands.
same yield as that obtained with added NaH and Na(acac) (Table 1,
Arene-bound NHC-ligated nickel complexes were prepared on
entry 2). However, the basic additives increased the turnover num- the gram scale by the addition of the free carbene ligand to a Ni0
bers of reactions conducted with low catalyst loadings. With basic source in C6H6 as solvent under H2 pressure36,41. The solid-state struc-
6
additives, the reaction with only 0.3mol% catalyst formed the addi- tures of these complexes (Fig. 2b for [L4–Ni(η -C6H6)]) illustrate
tion product in 85% yield (TON=283, Table 1, entry 3), but with- the steric impact of the NHC ligand. In contrast to the symmetrical
6
out these additives, the reaction with 0.3mol% catalyst formed the coordination of benzene in [L2–Ni(η -C6H6)], the angle between
6
alkylarene in only 6% yield (Table 1, entry 4; for further details, see the carbene carbon and the benzene ligand in [L4–Ni(η -C6H6)] is
the Supplementary Information). The basic additives likely remove significantly distorted from linearity (169.4°). This difference sug-
trace water, which affects reactions with low loadings.
gests that steric hindrance of the NHC ligand prevents a complete
6
The hydroarylation of unactivated alkenes also occurred with η -interaction.
a variety of electron-neutral, electron-rich and electron-deficient
To quantify the steric properties of L4 further, the percent buried
arenes (Table 1). The reaction of alkene 14 with toluene provided volume (%VBur)42,43 and steric map of L4 were calculated (Fig. 2b)44.
n-alkylarene 15 in good yield as a mixture of m- and p-alkylarene With the standard parameters of Dorta et al.45 for the radius (r)
isomers. Alkene 11 reacted with xylene isomers at the most sterically of the sphere surrounding the metal centre of 3.5Å, the proximal
accessible position to give products 16 and 17 in moderate yields. %VBur of L4 was smaller than that of L1 by 3%. However, with a
Reactions with more electron-deficient fluoroarenes also proceeded radius of 5.5Å to account for the remote steric environment of L4,
in high yields (18−21). Finally, the intramolecular reaction of arene the %VBur of L4 was greater than that of L1 by >10%. This difference
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
a
d
[
13CL4-Ni(η6-C6H6)] (3 mol%)
NaH (10 mol%)
13CL4
13CL4
13CL4
D
Na(acac) (10 mol%)
Ni
Ni
Ni
D5
C6D6
+
+
+
C8H17
t-Bu
N
N
Neat, 120 °C
D6
C8H17
C8H17
t-Bu
1-d6
9 equiv.
5
24
δcarbene = 204.7 ppm
26
1-d6
14
[26]
=
[14]Keq
b
[
13CL4-Ni(η6-C6D6)]
L4
Ni
L4
4.0
Pentane, under N2
25 °C
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0
Ni
+
Me
N2
Ni
N1
C1
Me
Me
2
25
50%
y = 1.10 × 10–1x – 2.77 × 10–1
R 2 = 9.99 × 10–1
C4
C5
C2
(excess)
C3
c
Addition of
t-Bu
0
10
20
30
40
13CL4
[14] (mM)
Ni
D6
e
L4
Ni
L4
Pentane, under N2
25 °C
Ni
N2
+
t-Bu
N1
C1
13CL4
N
N
t-Bu
Ni
C2
Ni
N3
14
(excess)
26
59%
N
N
N4
C3
t-Bu
26
f1 (ppm)
Fig. 3 | Observation and isolation of catalyst resting states. a, Observation of the catalyst resting state in the hydroarylation reaction of 1-d6 with
unhindered alkene 5 by 13C NMR spectroscopy. The resting state with unhindered alkene 5 was found to be complex 24. b, Preparation of bis-olefin
complex 25 and the ORTEP diagram of the X-ray diffraction structure of 25 (thermal ellipsoids are shown at the 50% probability level). c, The equilibrium
6
between [13CL4–Ni(η -C6H6)] and the mono-olefin complex 26 was monitored by 13C NMR spectroscopy. The spectra resulting from the addition of
6
14 (0−60ꢀequiv.) to a solution of [13CL4–Ni(η -C6H6)] in C6D6 are shown. d, Expression for the equilibrium between arene- and alkene-bound 13CL4–Ni
complexes and a plot of the ratio of nickel complexes with respect to [14]. We estimate the error in the integral ratios to be <5% based on the signal-
to-noise ratio of the 13C NMR spectra. e, Preparation of mono-olefin complex 26 and ORTEP diagram of the the X-ray diffraction structure of 26 (thermal
ellipsoids are shown at the 50% probability level). For clarity, all hydrogen atoms have been omitted from the ORTEP diagrams, and NHC sidearms are
represented in wire format.
is slightly greater than the difference between IPr and its N-(2,4,6- (Supplementary Figs. 7 and 8). The bis-propylene analogue 25 was
6
trimethylphenyl) analogue, known as IMes (8% difference in buried isolated in 50% yield from the reaction of [L4–Ni(η -C6H6)] and pro-
volume within 3.5Å for L–AuCl)42, and correlates with the large dif- pylene (2; Fig. 3b) and fully characterized by NMR spectroscopy and
ference in activity between the catalyst containing L1 and that con- single-crystal X-ray diffraction. The 13C NMR spectrum of the reac-
taining L4 in the hydroarylation reaction. However, the large size of tion of the more hindered tert-butylethylene (14) obtained at 25°C
L4 cannot be solely responsible for the high reactivity of the catalyst contains a single resonance at 198.7ppm (Supplementary Fig. 11)
containing L4 in alkene hydroarylation because the reactivities of and the colour of the solution was yellow, matching the spectrum
complexes of L2 and L3 possessing similar steric properties (see the and colour of the alkene complex 26 generated in situ from [13CL4–
6
Supplementary Information for all %VBur values) are closer to that of Ni(η -C6H6)] and 14 (vide infra). However, the 13C NMR spectrum
the catalyst containing L1 than to that containing L4.
recorded at 100°C contains a resonance at 196.2ppm, and the colour
6
Monitoring the reaction catalysed by the complex containing L4 of the solution was red, matching those of [13CL4–Ni(η -C6D6)].
with a 13C-labelled carbene carbon (13CL4) by NMR spectroscopy Studies of the relative stabilities of the alkene and arene complexes
revealed the resting states of the hydroarylation reactions with by varying the amount of 14 in pentane under nitrogen (Fig. 3c,d)
alkenes of varying sizes. The 13C NMR spectrum of the hydroaryla- showed that both [L4–Ni(η -C6H6)] and the alkene complex 26
6
tion reaction between 1-d6 and the long-chain alkene 5 recorded were present at equilibrium at lower concentrations (0.007–0.2M)
at 100°C contains a resonance at 204.7ppm (Fig. 3a), matching of alkene, but the mono-alkene dinitrogen complex 26, which
that of bis-alkene complex 24 generated independently in solu- was identified by NMR and IR spectroscopy and single-crystal
6
tion by combining 2.2equiv. of alkene 5 with [13CL4–Ni(η -C6H6)] X-ray diffraction (Fig. 3e), was the only species observed at higher
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
concentrations of alkene (0.3–0.5M). Thus, the catalytically active
species in the reactions of unhindered alkenes (such as 5) are
bis-alkene complexes, whereas those in the reactions of hindered
alkenes (such as 14) at elevated temperatures are an equilibrium
mixture of mono-alkene and arene complexes.
a
[L4–Ni(η6-C6H6)] (3 mol%)
NaH (10 mol%)
H
R
Na(acac) (10 mol%)
+
R
Neat, 120 °C
H
1
Alkene
To identify the steps of the catalytic cycle, the initial rates of the
hydroarylation of both unhindered alkene 5 and hindered alkene 11
with benzene (1) were measured at varying alkene concentrations
(Fig. 4a). The initial rate of the hydroarylation reaction with termi-
nal alkene 5 was inverse first order in the concentration of 5 (Fig. 4a,
top), indicating that one equivalent of the unhindered alkene 5 dis-
sociates from the bis-alkene resting state 24 prior to the binding of
the arene and the rate-determining step of the reaction. The order
in hindered alkene 11 depended on the concentration of 11 (Fig. 4a,
bottom), which is consistent with the existence of an equilibrium
between the catalytic resting states as a function of the nature and
(excess)
6
5
4
3
2
1
0
y = 1.5x – 0.15
R
2 = 0.99
C8H17
5
0
1
2
3
4
1/[5] (M–1
)
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
concentration of the alkene, as described above (see Fig. 3d)17,46
.
The first-order dependence of the reaction rate on [11] at low [11]
indicates that replacement of the arene with one alkene precedes
the rate-determining step of the catalytic process, whereas the zero-
order dependence of the reaction rate on [11] at high [11] indicates
that the resting state shifts to the mono-alkene complex 26 and that
the highest-energy transition state contains an aryl or arene unit.
Two sets of experiments revealed reversible steps within the
catalytic cycle. First, the reaction of 1-d6 with alkene 11 led to
43% incorporation of deuterium at the 2-alkenyl position in unre-
acted 11 after 60% conversion (Fig. 4b). No deuterium incorpora-
tion into the terminal position of the alkene was observed. Second,
the kinetic isotope effect (KIE) determined from separate reactions
C8H17
Me Me
11
0
0.1
0.2
0.3
0.4
0.5
0.6
[11] (M)
b
[L4–Ni(η6-C6H6)] (3 mol%)
NaH (5 mol%)
C6D5
C8H17
of alkene 11 with either 1 or 1-d6 was only 1.3 0.1 (Fig. 4c)14,47
.
D
Na(acac) (5 mol%)
C8H17
This KIE value contrasts with the measured KIE values of 2.1–2.5
for irreversible C–H activation during ruthenium-catalysed alkene
hydroarylation17,27,28,31. The results of both experiments imply that
the transfer of the arene hydrogen to the alkene by one or multiple
steps is reversible. An overall catalytic cycle for the hydroarylation
of unactivated alkenes consistent with our experimental results and
previous reports33 is shown in Fig. 5a.
D6
+
+
H/D
Neat, 120 °C, 1 h
60% conversion
C8H17
1-d6
11
43% D incorporation
c
[L4–Ni(η6-C6H6)] (3 mol%)
NaH (10 mol%)
C6H5
C8H17
Computational studies using density functional theory (DFT)
provided further insight into the mechanism of C−H bond cleav-
age (Fig. 5b, see the Supplementary Information for computational
details). C−H activation could occur by oxidative addition of the
C−H bond in benzene or by an alternative mechanism involv-
ing the direct transfer of the C–H bond of a coordinated arene to
the bound alkene, known as ligand-to-ligand hydrogen transfer
(LLHT)35. Previous computational studies on the nickel-catalysed
hydroarylation of alkynes, hydroarylation with more acidic arenes,
and hydroarylation with less hindered ligands indicated that the
barrier to direct oxidative addition of the C−H bonds in benzene
and other arenes to NHC-ligated Ni0 complexes is higher than that
of a one-step transfer of the hydrogen from a coordinated arene to
the coordinated π-system of the alkene or alkyne39,48. The barrier we
computed for the LLHT process between benzene (1) and propene
(2) ligands with the catalyst containing L4 (L4TSLLHT, 21.3kcalmol−1),
relative to the combination of benzene and the bis-alkene ground
state (GS), was 5.2kcalmol−1 lower in energy than that for the LLHT
H
Na(acac) (10 mol%)
C8H17
or
or
+
Neat, 120 °C
D6
C6D5
C8H17
11
kH
= 1.3 0.1
D
kD
Fig. 4 | Mechanistic experiments. a, Dependence of the initial rate of the
hydroarylation on alkene concentration for unhindered alkene 5 (top) or
hindered alkene 11 (bottom). Error bars indicate a 10% error in the initial
rate. b, Deuterium incorporation was observed at the 2-alkenyl position
of unreacted alkene 11 in the hydroarylation of 11 with 1-d6. c, A KIE
experiment for the hydroarylation of alkene 11 was conducted in separate
vessels, and the KIE was found to be 1.3ꢀ ꢀ0.1. This result indicates that
hydrogen atom transfer is reversible.
process with the catalyst containing the less hindered L1 (L1TSLLHT
,
alkyl aryl complex bound by L1 (L1TSRE). This difference is consis-
26.5kcalmol−1) and 3kcalmol−1 lower than the barrier for oxidative tent with the much higher activity of [L4–Ni(η -C6H6)] as catalyst
6
6
addition of the aryl C–H bond (see the Supplementary Information for the hydroarylation reaction compared with [L1–Ni(η -C6H6)].
for further details).
One might envision this difference in barriers to result from a ste-
However, the highest-energy transition state of the catalytic ric effect on the rate of reductive elimination. However, computation
process deduced from experiment (vide supra) and computation is of the geometry and non-covalent interactions provided a much dif-
the reductive elimination to form the alkyl−aryl C−C bond from ferent picture of the origin of the high activity of the complex of L4
T-shaped 27, not the LLHT. The computed barrier (relative to the compared with that of the smaller L1. Most striking, the geometrical
ground state) for reductive elimination from the alkyl aryl complex parameters around the nickel atom in L4TSRE were found to be almost
bound by L4 (L4TSRE) is 29.2kcalmol−1, which is 3.9kcalmol−1 lower identical to those in L1TSRE (Fig. 5c), suggesting that an effect beyond
than the computed barrier for the reductive elimination from the simple steric effects within the transition state controls the rate.
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
a
c
NHC
Ni
NHC
NHC
Ni
– Alkene, N2
+ PhH
– PhH
Ni
C1
N2
– PhH
+ 2 alkene
R
R
R
Ni
C3
Resting state is concentration-
dependent for hindered alkenes
Resting state for
C2
unhindered alkenes
– Alkene
+ Alkene
L1TSRE
C6H6
C6H5(CH2)2R
NHC
Ni
+
Ligand-to-ligand hydrogen
transfer
R
H
Ligand exchange
R
NHC
Ni
NHC
Ni
C1
H
Ni
R
R
C3
C2
NHC
Ni
Rate-limiting
reductive elimination
Isomerization
L4TSRE
R
b
d
L2-L1
L3-L2
L4-L3
L4-L1
∆∆E‡
–5.6
–0.1
–1.1
–6.7
‡
∆∆∆Edist
Distortion
NHC
TSLLHT
Transition state for
ligand-to-ligand
–1.1
–4.5
0.3
–1.5
0.4
–2.4
–4.3
33.1
Ni
H
∆∆∆Eint
Interaction energy
–0.3
hydrogen transfer
29.2
∆∆∆EPauli
Pauli repulsion
∆∆∆Eelstat
2.6
0.8
1.8
2.0
5.2
‡
26.5
21.3
NHC
Ni
21.7
–4.9
–1.2
–0.8
–0.2
–3.7
–4.3
Electrostatic
NHC
Ni
21.1
∆∆∆Edisp
London dispersion
–2.9
NHC
∆∆∆Ect
Charge transfer
Ni
–1.1
0.1
–0.2
0
–0.4
–1.7
0.1
TSRE
Transition state for
reductive elimination
GS
∆∆∆Epol
Polarization
0
NHC = IPr (L1)
NHC = m-XylIPr*OMe (L4)
27
0.0
(kcal mol–1
)
Stabilizing
Destabilizing
Fig. 5 | Computational investigations. a, Proposed mechanism for the nickel-catalysed hydroarylation of unactivated alkenes supported by mechanistic
experiments and computations. b, Computed energetics for the hydroarylation reaction with catalysts containing L1 and L4. c, DFT-optimized geometries
of tSRe. d, Energy decomposition analysis for the changes between the ground state and transition state for reductive elimination as a function of the
carbene ligand. ΔEdist is the difference in energy between the most stable form of the free carbene ligand and free nickel fragment and the energy of
the two components in their geometries of the complex, ΔΔEdist is the difference in distortion energies between the ground state and transition state
and ΔΔΔEdist is the difference in ΔΔEdist for the pairs of ligands. ΔEint is the interaction energy of the two fragments in their distorted geometries and is
decomposed as the sum of ΔEPauli, ΔEelstat, ΔEdisp, ΔEct and ΔEpol. In a similar manner, ΔΔE values are the difference in energies between the GS and the tS,
and ΔΔΔE values are the difference in the component ΔΔE for pairs of ligands.
To decipher the origin of the different barriers for reductive other ligands (ΔΔΔEdist,L4-L1 =−2.4kcalmol−1). For all of the four
elimination from the complexes containing L1 and L4, we used the NHC ligands, the NHC fragment is less distorted from the free car-
distortion/interaction or activation strain model pioneered by Houk bene in the transition state for reductive elimination than it is in
and co-workers49 and the second-generation absolutely localized the ground state. However, this difference between the distortion
molecular orbital energy decomposition analysis (ALMO-EDA) energy of the carbene in the transition and ground states is great-
developed by Head-Gordon and co-workers40. Selected energy val- est for L4 and contributes to a lower barrier for the reaction of the
ues from this analysis for the complexes of the four ligands L1−L4 complex containing L4 than for the reactions of the complexes con-
are summarized in Fig. 5d. A comparison of the difference in distor- taining L1−L3 (Supplementary Table 6).
tion energies between pairs of complexes shows that the peripheral
This energy decomposition analysis also shows that the aryl
methyl groups in L4 (ΔΔΔEdist,L4-L3 =−1.5kcalmol−1), in addition to groups in ligands L2−L4, which are not present in L1, lead to inter-
the aromatic rings (ΔΔΔEdist,L2-L1 =−1.1kcalmol−1), lead to distor- action energies that cause the barriers for reactions catalysed by
tion energies that favour reaction with L4 over reaction with the complexes of L4 (and L2 and L3) to be lower than those catalysed by
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
the complex bearing the more common ligand L1 (ΔΔΔEint,L4-L1
=
and many of the reactions occurred in good yields. Even internal
−4.3kcalmol−1). An analysis of the major contributions to the dif- alkenes reacted to give n-alkylarene products. The unparalleled
ference between the interaction energies in the transition state for selectivity and activity of this reaction was enabled by a nickel cata-
reductive elimination with one ligand and those in the transition lyst containing an extremely large N-heterocyclic carbene ligand
state for reductive elimination with another ligand can be seen in that undergoes C–H activation by a mechanism that involves the
Fig. 5d. The largest differences between these values for L4 com- formation of a linear alkyl−metal complex that undergoes reductive
pared with L1 are the Pauli repulsive (steric effects), electrostatic elimination to form the new carbon−carbon bond in the alkylarene
and London dispersion terms. Assessment of each pairwise differ- product with rates that are enhanced by intramolecular non-cova-
ence in these contributions to the interaction energies for the reac- lent interactions. The results of our computational studies imply
tions catalysed by the complexes of the series of ligands reveals the that the conventional view of how steric bulk favours reductive
structural elements of the ligands that lead to these values. This elimination does not apply to this system. Instead of accelerating the
analysis shows that the Pauli repulsion term strongly influences reductive elimination by steric repulsion, the multiple aryl groups in
the barrier and results principally from the larger steric impact of ligand L4 of the most active catalyst lead to favourable electrostatic
L4. Particularly striking is that this Pauli term increases the bar- interactions, and the large number of methyl groups in L4 leads to
rier for the reaction with the large ligand relative to that for the favourable London dispersion interactions and lower distortion
reaction with the smaller ligand, rather than decreasing the barrier energy. We anticipate that this reactivity and the analysis of its ori-
(ΔΔΔEPauli,L4-L1 =5.2kcalmol−1). Typically, one presumes that steric gins should aid the development of new methods for the functional-
effects cause reductive elimination involving the coupling of two ization of alkenes with additional strong, unactivated C–H or X–H
ligands at a single metal centre to be faster for complexes bearing bonds catalysed by complexes of nickel ligated by N-heterocyclic
more hindered ancillary ligands50.
carbenes as well as by complexes of metals other than nickel with
In contrast, attractive electrostatic and London dispersion terms appropriate properties of the ancillary ligands.
reduce the energy of the transition state containing L4 compared
with the ground state more than they reduce this difference in Online content
Any Nature Research reporting summaries, source data, extended data,
energy for the complexes containing the other ligands. More spe-
supplementary information, acknowledgements, peer review information; details
cifically, the values in Fig. 5d show that that presence of the eight
of author contributions and competing interests; and statements of data and code
aryl groups in ligands L2−L4 lead to electrostatic interactions that
availability are available at https://doi.org/10.1038/s41557-019-0409-4.
are approximately 4kcalmol−1 larger than the electrostatic interac-
tions in the complexes of L1 (ΔΔΔEelstat,L2-L1 =−4.9kcalmol−1 and
Received: 3 June 2019; Accepted: 12 December 2019;
Published: xx xx xxxx
ΔΔΔEelstat,L4-L1 =−3.7kcalmol−1). A similar pairwise comparison of
the London dispersion effects shows that the difference in values
between the system containing L4 and the system containing L1
(ΔΔΔEdisp,L4-L1 =−4.3kcalmol−1) results mainly from the presence
of the 16 methyl groups on L4 that are not present in L1−L3.
Although the results above demonstrate that the peripheral
methyl groups in L4 cause the difference between the stabilizing dis-
persive interactions in the ground and transition states containing
L4 to be larger than this difference between those containing L3 (ΔΔ
ΔEdisp,L4-L3 =−2.9kcalmol−1), we further probed the origin of this dif-
ference in dispersive interactions with ligands L3 and L4. The methyl
groups in L4 could participate in stabilizing dispersive interactions
with each other and with other groups in the complex, or they could
cause the positions of the aryl groups and the structure of the core
of the complexes of L4 to be altered from those of L1−L3 in a way
that leads to a larger difference in dispersion interactions between
the ground and transition states bearing L4. To distinguish between
these two possibilities, we performed the same energy decomposi-
tion analysis of the ground and transition states with L3 (lacking
the peripheral methyl groups), but placing the atoms in the same
positions as those in the lowest-energy geometry of L4. The results
of this analysis are included in the Supplementary Information and
show that the differences in dispersion interactions between the
ground and transition states of the complexes containing L4 and
L3 in the same geometry (the minimum-energy geometry of L4,
ΔΔΔEdisp,L4-L3 in L4 geometry =−2.9kcalmol−1; Supplementary Table 12)
are similar to those of the ground and transition states containing
L4 and L3 in their respective minimum-energy geometries. In other
words, the stabilizing dispersion interactions in the complexes of L4
are due to direct interactions with the methyl groups, not to changes
in the ligand geometry imparted by the methyl groups in L4. Finally,
a graphical plot of the non-covalent interactions51,52 present in L4TSRE
(Supplementary Fig. 18) corroborates the presence of significant sta-
bilizing interactions involving the methyl groups.
References
1. Kocal, J. A., Vora, B. V. & Imai, T. Production of linear alkylbenzenes.
Appl. Catal. A 221, 295–301 (2001).
2. Linear Alkyl Benzene Market Size, Share & Trends Analysis Report By
Application (Heavy Duty Laundry, Laundry Powders, Washing Liquids,
Industrial Cleaners, Household Cleaners), And Segment Forecasts, 2012–2020
(Grand View Research, 2017).
3. Röper, M., Gehrer, E., Narbeshuber, T. & Siegel, W. Acylation and alkylation
in Ullmann’s Encyclopedia of Industrial Chemistry (Wiley-VCH, 2000).
4. de Almeida, J. L. G., Dufaux, M., Taarit, Y. B. & Naccache, C. Linear
alkylbenzene. J. Am. Oil Chem. Soc. 71, 675–694 (1994).
5. Ishiwatari, R., Takada, H., Yun, S.-J. & Matsumoto, E. Alkylbenzene pollution
of Tokyo Bay sediments. Nature 301, 599–600 (1983).
6. Macı́as-Zamora, J. V. & Ramı́rez-Alvarez, N. Tracing sewage pollution using
linear alkylbenzenes (LABs) in surface sediments at the south end of the
Southern California Bight. Environ. Pollut. 130, 229–238 (2004).
7. Murai, S. et al. Efcient catalytic addition of aromatic carbon-hydrogen bonds
to olefns. Nature 366, 529–531 (1993).
8. Jun, C.-H., Hong, J.-B., Kim, Y.-H. & Chung, K.-Y. Te catalytic alkylation
of aromatic imines by Wilkinson’s complex: the domino reaction of
hydroacylation and ortho-alkylation. Angew. Chem. Int. Ed. 39,
3440–3442 (2000).
9. Gao, K. & Yoshikai, N. Cobalt–phenanthroline catalysts for the ortho
alkylation of aromatic imines under mild reaction conditions. Angew. Chem.
Int. Ed. 50, 6888–6892 (2011).
10. Schinkel, M., Marek, I. & Ackermann, L. Carboxylate-assisted ruthenium(II)-
catalyzed hydroarylations of unactivated alkenes through C–H cleavage.
Angew. Chem. Int. Ed. 52, 3977–3980 (2013).
11. Matsumoto, T., Taube, D. J., Periana, R. A., Taube, H. & Yoshida, H.
Anti-Markovnikov olefn arylation catalyzed by an iridium complex.
J. Am. Chem. Soc. 122, 7414–7415 (2000).
12. Periana, R. A., Liu, X. Y. & Bhalla, G. Novel bis-acac-O,O–Ir(III) catalyst for
anti-Markovnikov, hydroarylation of olefns operates by arene CH activation.
Chem. Commun. 24, 3000–3001 (2002).
13. Oxgaard, J., Muller, R. P., Goddard, W. A. & Periana, R. A. Mechanism
of homogeneous Ir(III) catalyzed regioselective arylation of olefns.
J. Am. Chem. Soc. 126, 352–363 (2004).
14. Bhalla, G., Liu, X. Y., Oxgaard, J., Goddard, W. A. & Periana, R. A. Synthesis,
structure, and reactivity of O-donor Ir(III) complexes: C–H activation studies
with benzene. J. Am. Chem. Soc. 127, 11372–11389 (2005).
15. Bhalla, G., Oxgaard, J., Goddard, W. A. & Periana, R. A. Anti-Markovnikov
hydroarylation of unactivated olefns catalyzed by a bis-tropolonato
iridium(III) organometallic complex. Organometallics 24, 3229–3232 (2005).
Conclusion
In this work the first highly anti-Markovnikov hydroarylation of
unactivated alkenes with unactivated arenes has been accomplished,
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
16. Lail, M., Arrowood, B. N. & Gunnoe, T. B. Addition of arenes to ethylene and
propene catalyzed by ruthenium. J. Am. Chem. Soc. 125, 7506–7507 (2003).
17. Lail, M. et al. Experimental and computational studies of ruthenium(II)-
catalyzed addition of arene C−H bonds to olefns. Organometallics 23,
5007–5020 (2004).
18. McKeown, B. A., Prince, B. M., Ramiro, Z., Gunnoe, T. B. & Cundari, T. R.
PtII-catalyzed hydrophenylation of α-olefns: variation of linear/branched
products as a function of ligand donor ability. ACS Catal. 4, 1607–1615 (2014).
19. Luedtke, A. T. & Goldberg, K. I. Intermolecular hydroarylation of unactivated
olefns catalyzed by homogeneous platinum complexes. Angew. Chem. Int. Ed.
47, 7694–7696 (2008).
20. Clement, M. L., Grice, K. A., Luedtke, A. T., Kaminsky, W. & Goldberg, K. I.
Platinum(II) olefn hydroarylation catalysts: tuning selectivity for the anti‐
Markovnikov product. Chem. Eur. J. 20, 17287–17291 (2014).
21. Webster-Gardiner, M. S. et al. Catalytic synthesis of “super” linear alkenyl
arenes using an easily prepared Rh(I) catalyst. J. Am. Chem. Soc. 139,
5474–5480 (2017).
22. Chen, J. et al. Catalytic synthesis of superlinear alkenyl arenes using a Rh(I)
catalyst supported by a “capping arene” ligand: access to aerobic catalysis. J.
Am. Chem. Soc. 140, 17007–17018 (2018).
23. Oxgaard, J. & Goddard, W. A. Mechanism of Ru(II)-catalyzed olefn insertion
and C−H activation from quantum chemical studies. J. Am. Chem. Soc. 126,
442–443 (2004).
24. Oxgaard, J., Periana, R. A. & Goddard, W. A. Mechanistic analysis of
hydroarylation catalysts. J. Am. Chem. Soc. 126, 11658–11665 (2004).
25. Suslick, B. A., Liberman-Martin, A. L., Wambach, T. C. & Tilley, T. D. Olefn
hydroarylation catalyzed by (pyridyl-indolate)Pt(II) complexes: catalytic
efciencies and mechanistic aspects. ACS Catal. 7, 4313–4322 (2017).
26. Mann, G., Shelby, Q., Roy, A. H. & Hartwig, J. F. Electronic and steric efects
on the reductive elimination of diaryl ethers from palladium(II).
Organometallics 22, 2775–2789 (2003).
27. Foley, N. A. et al. Comparative reactivity of TpRu(L)(NCMe)Ph (L = CO or
PMe3): impact of ancillary ligand L on activation of carbon−hydrogen bonds
including catalytic hydroarylation and hydrovinylation/oligomerization of
ethylene. J. Am. Chem. Soc. 129, 6765–6781 (2007).
28. Foley, N. A., Ke, Z., Gunnoe, T. B., Cundari, T. R. & Petersen, J. L. Aromatic
C−H activation and catalytic hydrophenylation of ethylene by
TpRu{P(OCH2)3CEt}(NCMe)Ph. Organometallics 27, 3007–3017 (2008).
29. Foley, N. A., Lee, J. P., Ke, Z., Gunnoe, T. B. & Cundari, T. R. Ru(II) catalysts
supported by hydridotris(pyrazolyl)borate for the hydroarylation of olefns:
reaction scope, mechanistic studies, and guides for the development of
improved catalysts. Acc. Chem. Res. 42, 585–597 (2009).
30. Joslin, E. E. et al. Catalytic hydroarylation of ethylene using TpRu(L)(NCMe)
Ph (L = 2,6,7-trioxa-1-phosphabicyclo[2,2,1]heptane): comparison to
TpRu(L′)(NCMe)Ph systems (L′ = CO, PMe3, P(pyr)3, or P(OCH2)3CEt).
Organometallics 31, 6851–6860 (2012).
31. Burgess, S. A. et al. Hydrophenylation of ethylene using a cationic Ru(ii)
catalyst: comparison to a neutral Ru(ii) catalyst. Chem. Sci. 5, 4355–4366 (2014).
32. Malinoski, J. M. & Brookhart, M. Polymerization and oligomerization of
ethylene by cationic nickel(II) and palladium(II) complexes containing bidentate
phenacyldiarylphosphine ligands. Organometallics 22, 5324–5335 (2003).
33. Bair, J. S. et al. Linear-selective hydroarylation of unactivated terminal and
internal olefns with trifuoromethyl-substituted arenes. J. Am. Chem. Soc.
136, 13098–13101 (2014).
35. Guihaumé, J., Halbert, S., Eisenstein, O. & Perutz, R. N. Hydrofuoroarylation of
alkynes with Ni catalysts. C–H activation via ligand-to-ligand hydrogen transfer,
an alternative to oxidative addition. Organometallics 31, 1300–1314 (2012).
36. Hoshimoto, Y., Hayashi, Y., Suzuki, H., Ohashi, M. & Ogoshi, S. One-pot,
6
single-step, and gram-scale synthesis of nononuclear [(η -arene)Ni(N-
heterocyclic carbene)] complexes: useful precursors of the Ni0–NHC unit.
Organometallics 33, 1276–1282 (2014).
37. Berthon-Gelloz, G. et al. IPr* an easily accessible highly hindered
N-heterocyclic carbene. Dalton Trans. 39, 1444–1446 (2010).
38. Meiries, S., Speck, K., Cordes, D. B., Slawin, A. M. Z. & Nolan, S. P.
[Pd(IPr*OMe)(acac)Cl]: tuning the N-heterocyclic carbene in catalytic C–N
bond formation. Organometallics 32, 330–339 (2013).
39. Okumura, S. et al. para-Selective alkylation of benzamides and aromatic
ketones by cooperative nickel/aluminum catalysis. J. Am. Chem. Soc. 138,
14699–14704 (2016).
40. Horn, P. R., Mao, Y. & Head-Gordon, M. Probing non-covalent interactions
with a second generation energy decomposition analysis using absolutely
localized molecular orbitals. Phys. Chem. Chem. Phys. 18, 23067–23079 (2016).
41. Saper, N. I. & Hartwig, J. F. Mechanistic investigations of the hydrogenolysis
of diaryl ethers catalyzed by nickel complexes of N-heterocyclic carbene
ligands. J. Am. Chem. Soc. 139, 17667–17676 (2017).
42. Clavier, H. & Nolan, S. P. Percent buried volume for phosphine and
N-heterocyclic carbene ligands: steric properties in organometallic chemistry.
Chem. Commun. 46, 841–861 (2010).
43. Hillier, A. C. et al. A combined experimental and theoretical study
examining the binding of N-heterocyclic carbenes (NHC) to the Cp*RuCl
5
(Cp* = η -C5Me5) moiety: insight into stereoelectronic diferences between
unsaturated and saturated NHC ligands. Organometallics 22, 4322–4326
(2003).
44. Falivene, L. et al. SambVca 2. A web tool for analyzing catalytic pockets with
topographic steric maps. Organometallics 35, 2286–2293 (2016).
45. Dorta, R. et al. Steric and electronic properties of N-heterocyclic carbenes
(NHC): a detailed study on their interaction with Ni(CO)4. J. Am. Chem. Soc.
127, 2485–2496 (2005).
46. Matsumoto, T., Periana, R. A., Taube, D. J. & Yoshida, H. Regioselective
hydrophenylation of olefns catalyzed by an Ir(III) complex. J. Mol. Catal. A
180, 1–18 (2002).
47. Simmons, E. M. & Hartwig, J. F. On the interpretation of deuterium kinetic
isotope efects in C–H bond functionalizations by transition-metal complexes.
Angew. Chem. Int. Ed. 51, 3066–3072 (2012).
48. Tang, S., Eisenstein, O., Nakao, Y. & Sakaki, S. Aromatic C–H σ-bond
activation by Ni0, Pd0, and Pt0 alkene complexes: concerted oxidative addition
to metal vs ligand-to-ligand H transfer mechanism. Organometallics 36,
2761–2771 (2017).
49. Bickelhaupt, F. M. & Houk, K. N. Analyzing reaction rates with the
distortion/interaction-activation strain model. Angew. Chem. Int. Ed. 56,
10070–10086 (2017).
50. Hartwig, J. F. in Organotransition Metal Chemistry: From Bonding to Catalysis
Ch. 8 (Univ. Science Books, 2010).
51. Johnson, E. R. et al. Revealing noncovalent interactions. J. Am. Chem. Soc.
132, 6498–6506 (2010).
52. Contreras-García, J. et al. NCIPLOT: a program for plotting noncovalent
interaction regions. J. Chem. Teory Comput. 7, 625–632 (2011).
34. Schramm, Y., Takeuchi, M., Semba, K., Nakao, Y. & Hartwig, J. F. Anti-
Markovnikov hydroheteroarylation of unactivated alkenes with indoles,
pyrroles, benzofurans, and furans catalyzed by a nickel–N-heterocyclic
carbene cystem. J. Am. Chem. Soc. 137, 12215–12218 (2015).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2020
NAtuRe CHeMIStRY | www.nature.com/naturechemistry

NATurE CHEMisTry
Articles
(S10-OD023532). NMR spectroscopy was performed in the College of Chemistry’s NMR
facility funded in part by the NIH (S10-OD024998).
Data availability
Crystallographic data for the structures reported in this article have been deposited
at the Cambridge Crystallographic Data Centre, under deposition numbers 1901576
6
([L4–Ni(η -C6H6)]), 1901577 (26) and 1901578 (25). Copies of the data can be obtained
Author contributions
free of charge via https://www.ccdc.cam.ac.uk/structures/. All other data supporting the
findings of this study are available within the Article and its Supplementary Information,
or from the corresponding author upon reasonable request.
All authors conceived and designed the experiments. N.I.S., A.O. and K.S. performed the
experiments. N.I.S. and D.W.S. performed the computations. All authors participated in
discussion and N.I.S. J.F.H. and Y.N. co-wrote the manuscript.
Acknowledgements
Competing interests
We thank S. Arlow, C. Karmel and J. Wang for helpful discussions. We thank Y. Schramm
for preliminary experiments. We acknowledge N. Settineri for X-ray crystallographic
analysis. We thank M. Head-Gordon and M. Loipersberger for discussions on EDA
calculations. This work was supported by the Director, Office of Science, of the U.S.
Department of Energy under contract no. DE-AC02- 05CH11231, by the National
Science Foundation (graduate research fellowship to N.I.S.) and by the Japan Society
for the Promotion of Science (JSPS KAKENHI Grant Number JP15H05799). X-ray
diffraction data were collected using an instrument funded by the NIH (S10-RR027172).
Computations were performed on a computation cluster funded by the NIH
The authors declare no competing interests
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41557-019-0409-4.
Correspondence and requests for materials should be addressed to Y.N. or J.F.H.
Reprints and permissions information is available at www.nature.com/reprints.
NAtuRe CHeMIStRY | www.nature.com/naturechemistry