Palladium-Catalyzed, Site-Selective Direct Allylation of Aryl C-H Bonds by Silver-Mediated C-H Activation: A Synthetic and Mechanistic Investigation

Article abstract of DOI:10.1021/jacs.6b10220

We describe a method for the site-selective construction of a C(aryl)-C(sp³) bond by the palladium-catalyzed direct allylation of arenes with allylic pivalates in the presence of AgOPiv to afford the linear (E)-allylated arene with excellent regioselectivity; this reaction occurs with arenes that have not undergone site-selective and stereoselective direct allylation previously, such as monofluorobenzenes and non-fluorinated arenes. Mechanistic studies indicate that AgOPiv ligated by a phosphine reacts with the arene to form an arylsilver(I) species, presumably through a concerted metalation-deprotonation pathway. The activated aryl moiety is then transferred to an allylpalladium(II) intermediate formed by oxidative addition of the allylic pivalate to the Pd(0) complex. Subsequent reductive elimination furnishes the allyl-aryl coupled product. The aforementioned proposed intermediates, including an arylsilver complex, have been isolated, structurally characterized, and determined to be chemically and kinetically competent to undergo the proposed elementary steps of the catalytic cycle.

Full text of DOI:10.1021/jacs.6b10220

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Palladium-Catalyzed, Site-Selective Direct Allylation of Aryl C–H Bonds by
Silver-Mediated C–H Activation: A Synthetic and Mechanistic Investigation
Sarah Yunmi Lee, and John F. Hartwig
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 31 Oct 2016
Downloaded from http://pubs.acs.org on October 31, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 7
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Palladium-Catalyzed, Site-Selective Direct Allylation of Ar-
yl C–H Bonds by Silver-Mediated C–H Activation: A Synthet-
ic and Mechanistic Investigation
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Sarah Yunmi Lee and John F. Hartwig*
Department of Chemistry, University of California, Berkeley, CA 94720, United States
ABSTRACT: We describe a method for the site-selective construction of C(aryl)-C(sp3) bond by the palladium-catalyzed direct
allylation of arenes with allylic pivalates in the presence of AgOPiv to afford the linear (E)-allylated arene with excellent regiose-
lectivity; this reaction occurs with arenes that have not undergone site- and stereoselective direct allylation previously, such as
monofluorobenzenes and non-fluorinated arenes. Mechanistic studies indicate that AgOPiv ligated by a phosphine reacts with the
arene to form an arylsilver(I) species, presumably through a concerted metalation-deprotonation (CMD) pathway. The activated
aryl moiety is then transferred to an allylpalladium(II) intermediate formed by oxidation addition of the allylic pivalate to the Pd(0)
complex. Subsequent reductive elimination furnishes the allyl-aryl coupled product. The aforementioned proposed-intermediates,
including an arylsilver complex, have been isolated, structurally characterized, and determined to be chemically and kinetically
competent to undergo the proposed elementary steps of the catalytic cycle.
INTRODUCTION
The higher reactivity of electron-deficient arenes than elec-
tron-rich arenes toward C–H bond functionalization is often
explained by the involvement of a concerted metalation-
deprotonation (CMD) step for cleavage of the C–H bond by
metal carboxylates. 9 Thus, the relative reactivity of the arene
parallels the relative acidity of the aryl C–H bonds. The re-
quirement that the arene possess multiple strongly activating
groups, presumably, results from the need for an acidic C–H
bond to enable cleavage of the C–H bond by a palladium(II)
species by a CMD pathway.1d
The formation of carbon–carbon bonds by direct functional-
ization of aryl C–H bonds is a powerful approach for the syn-
thesis and derivatization of aromatic compounds.1 Common
strategies to facilitate the C–H activation of arenes and control
the site-selectivity of direct functionalizations of arenes focus
on the use of 1) arenes bearing a directing group that can co-
ordinate to the metal catalyst or 2) highly electron-deficient
arenes, such as polyfluoroarenes. However, the site-selective
functionalizations of the C–H bonds in simple arenes lacking a
directing group and lacking a series of strongly activating
groups remains challenging.1c Although there have been re-
ports on site-selective functionalizations of simple arenes to
Recent studies by Larrosa10 and Sanford11 independently on
Pd-catalyzed direct functionalizations of aryl C–H bonds in
the presence of silver carboxylates reveal that the silver car-
boxylate can cleave the C–H bonds in arenes bound to
Cr(CO)3, polyfluoroarenes, and acidic heteroarenes, such as
thiophenes; the resulting arylsilver(I) complex is proposed to
2
,
3
form C(aryl)–C(sp ) bonds,2 simple arenes have not been
reported to undergo site-selective formation of C(aryl)–C(sp3)
bonds by reactions with C(sp3)-electrophiles.4 Yet, such reac-
tions would be valuable because the substrate scope and site
selectivity would complement those of electrophilic aromatic
substitution (Friedel-Crafts reactions) in which electron-rich
arenes are more reactive than electron-poor arenes.5
,
transfer its aryl moiety to a palladium intermediate.12 13 How-
ever, isolation of a phosphine-ligated arylsilver(I) species,
investigation of its ability to undergo transmetallation of the
aryl group to palladium, and evaluation of its competency as a
reaction intermediate in the catalytic process have not been
described. Moreover, determination of whether the Ag(I) sys-
tem can cleave the C–H bonds in less activated arenes as well
as extensions of this step to reactions that form C–C bonds
between sp2 and sp3 sites have not been reported.
In this vein, the direct allylation of simple arenes would be a
useful process because the allyl group in the product could
undergo further functionalization. Although the allylation of
polyfluoroarenes and the allylation of arenes containing direct-
,
ing groups have been described,6 7 the allylation of arenes that
We report the site-selective formation of C(aryl)-C(sp3)
bonds by a palladium-catalyzed direct allylation of mono-
fluorobenzenes and non-fluorinated arenes with allylic
pivalates in the presence of a silver(I) additive to generate
linear (E)-allylated arenes. Detailed mechanistic studies are
consistent with a synergistic catalytic cycle involving a (π-
allyl)palladium complex formed from the oxidative addition of
allylic pivalate to bisphosphine-Pd(0) and an arylsilver species
ligated by a phosphine resulting from silver-mediated cleavage
of relatively unactivated aryl C–H bonds. Isolation of the al-
lylpalladium and arylsilver species and studies of their reactiv-
ity support the proposed cycle.
are less electron poor and that lack directing groups have not
been reported. The allylation of arenes with allylic esters could
occur by a few apparently well-established steps. The reaction
could occur by oxidative addition of the allylic ester to form
an allylpalladium carboxylate8 and cleavage of the C–H bond
of an arene by the allylpalladium carboxylate complex in a
fashion proposed for reactions of arylpalladium carboxylate
complexes; the proposed mechanism for Pd-catalyzed direct
arylations of arenes with aryl halides typically includes the
cleavage of an aryl C–H bond by a palladium intermediate,
such as
a phosphine-ligated arylpalladium carboxylate
LArPd(OCOR).1,9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 7
RESULTS AND DISCUSSION
butylarylphosphines (entries 11–16), occurred in lower yields
1
2
3
4
5
6
7
8
Reaction Development. Our studies began with the mecha-
nistic hypothesis that the allylation of arenes could occur by a
pathway often invoked for the direct arylation of arenes. Al-
lylic esters readily undergo oxidative additions to Pd(0) spe-
cies, and the carboxylate ligand on palladium resulting from
this oxidative addition could trigger the cleavage of a C–H
bond of the arene by a CMD pathway.9 Reductive elimination
would form the allylarene.
On the basis of this initial mechanistic hypothesis and prior
studies suggesting the viability of this pathway for the allyla-
tion of polyfluoroarenes,7a,c we sought to identify conditions
for the direct allylation of monofluoroarenes. We investigated
a series of palladium precursors, ligands, bases and additives
for the direct coupling of neat fluorobenzene 1a with cinnamyl
electrophiles. These studies showed that the linear (E)-
allylated fluorobenzene (3a) formed as a single isomer in 82%
yield when catalyzed by Pd(OAc)2 and di-tert-butyl 2-
anisylphosphine (L) 14 in the presence of Cs2CO3 as a base and
AgOPiv as a stoichiometric additive (Table 1, entry 1). The
allylation occurred selectively at the position ortho to fluorine.
of 3a than did the reaction catalyzed by Pd(OAc)2 and L. Fur-
thermore, the allylation did not proceed without a silver addi-
tive (entry 17). Reactions with Ag(I) salts besides AgOPiv,
such as Ag2CO3 or AgOTf, led to a lower yield of 3a (entries
18 and 19) than did those with AgOPiv. The allylation with
cinnamyl pivalate generated 3a in higher yield than did those
with cinnamyl electrophiles containing other leaving groups
(entries 20–23).
The reaction of fluorobenzene with the branched isomer of
cinnamyl pivalate (2a) under the standard conditions afforded
linear allylarene 3a in 81% yield as a single product (eq 1).
This result suggests that the allylation reaction occurs through
a (π-allyl)palladium intermediate.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scope of Direct Allylation of Arenes. Under our standard
conditions for the allylation of arenes, various allylic pivalates
coupled with fluorobenzene 1a to generate linear (E)-
allylarenes with excellent site-selectivity (Table 2).15 A single
isomer of the corresponding allylarene was obtained with cin-
namyl pivalates containing para-, meta, or ortho-substituents
(entries 2–7), or an extended π-system (entry 8). In addition,
the allylation with 2-methylallyl chloride proceeded to form 3i
as a single product (entry 9). Finally, the reaction of a trisub-
stituted allylic pivalate formed a mixture of E and Z isomers
(3:1) of allylarene 3j (entry 10).
Table 1. Effect of Reaction Parameters on Palladium-
Catalyzed Allylations of Fluorobenzene with Cinnamyl
Pivalatea
Table 2. Scope of the Allylation of Fluorobenzene with
Allylic Pivalatesa
aReaction conditions: 1a (1.40 mL), 2 (0.35 mmol, 1.0 equiv),
Pd(OAc)2 (10%), L (20%), AgOPiv (1.0 equiv), and Cs2CO3 (2.4
aReaction conditions: 1a (0.20 mL), 2 (0.05 mmol, 1.0 equiv),
Pd(OAc)2 (10%), L (20%), AgOPiv (1.0 equiv), and Cs2CO3 (2.4
b
c
equiv) at 120 ºC for 17 h. Yield of purified product. 2f contained
11% of (Z)-isomer. dDetermined by GC analysis. e2-Methylallyl chlo-
ride was used instead of 2-methylallyl pivalate. 3j was obtained as a
b
c
equiv) at 120 ºC for 17 h. LG = leaving group. Determined by GC
f
analysis.
3:1 mixture of E and Z isomers.
Table 1 shows the influence of a series of reaction parame-
ters on the yield. The reaction catalyzed by Pd[P(t-Bu)2(2-
OMeC6H4)]2 (PdL2) formed 3a in a yield that was comparable
to that obtained with Pd(OAc)2 and phosphine L as catalyst
(entry 2); no allylation was observed in the absence of either
Pd(OAc)2 or L (entries 3 and 4). The allylation catalyzed by
complexes of other phosphine ligands, including PPh3 (entries
5), trialkylphosphines (entries 6–10), and di-tert-
The direct allylations of various arenes with cinnamyl
pivalate 2a also occurred under the standard conditions (Table
3). Ortho-substituted monofluorobenzenes such as 1-
fluoronaphthalene and 1-fluoro-2-methylbenzene, as well as 1-
fluoro-4-methylbenzene, were suitable substrates for the al-
lylation, affording the corresponding allylarenes as single
products (entries 1–3). However, para-methoxy fluorobenzene
ACS Paragon Plus Environment

Page 3 of 7
Journal of the American Chemical Society
gave two constitutional isomers (entry 4). The observation of
the product allylated at the position ortho to OMe as a minor
1
isomer is also consistent with the CMD pathway; CMD pro-
cess with anisole has been reported to proceed at the ortho
position.2f,9 Finally, the allylation of para-trifluoromethyl
fluorobenzene underwent C–C bond formation to yield a sin-
gle isomer of allylarene 5e (entry 5).
In addition to the allylations of monofluoroarenes, the al-
lylations of non-fluorinated arenes with 2a occurred (Table 3,
entries 6–9). The allylation of 1,3-benzodioxole proceeded
selectively at the ortho position to form 5f (entry 6). The al-
lylation of chlorinated arenes also occurred, furnishing al-
lylarenes 5g and 5h as the major products (entries 7 and 8).16
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Furthermore,
the
reaction
of
1-methoxy-4-
(trifluoromethyl)benzene produced allylation product 5i, albeit
in low yield (entry 9).
Table 3. Scope of the Allylation of Arenes with Cinnamyl
Pivalatea
major productb
F
entry
major productb
O
Figure 1. ORTEP diagram of complex 6 (O1–Pd1 2.117(14) Å;
P1–Pd1 2.332(6) Å; C1–Pd1 2.094(2) Å; C2–Pd1 2.145(2) Å;
C3–Pd1 2.259(2) Å) (C3–C2–C1 118.8(2)°; C15-P1-Pd1 109.0(7)
°; C10-O1-Pd1 112.1(13)°) (ellipsoids are shown at 50% proba-
bility, and hydrogen atoms are omitted for clarity).
entry
O
Ph
Ph
1
6
5f
, 66%
5a
, 64%
Cl
The competency of complex 6 to be an intermediate in the
catalytic process was investigated. The stoichiometric reaction
of complex 6 with fluorobenzene in the presence of AgOPiv at
120 ºC formed allylarene 3a in 78 % yield after 17 h (eq 3).
This yield is similar to the 82% yield of the catalytic reaction
(Table 1, entry 1). No allylarene 3a, as determined by gas
chromatography, was formed when AgOPiv was omitted from
the stoichiometric reaction of complex 6 with 1a (eq 3). This
lack of formation of allylarene is consistent with the lack of
product formed in the catalytic reaction without silver(I) addi-
tive (Table 1, entry 17). Monitoring of the reaction of eq 3 by
31P NMR spectroscopy showed that complex 6 was fully con-
verted within 10 min to a new palladium species and at this
time, allylarene 3a was formed in only 36% yield. As dis-
cussed later in this paper, the palladium species formed by this
reaction is the resting state of the catalyst (12) in the direct
allylation process.17
Ph
7c
F
Me
OMe
5g
2
Ph
, 79%
(major:minore = 6:1)
5b
, 48%
Cl
F
Ph
8c
Ph
d
5h
, 36%
CF3
OMe
R
Ph
5c
, R = Me: 74%
3
4
9
5d
, R = OMe: 83%
(o-F:o-OMe = 3:1)
5i
, 29%
CF3
d
5e
5
, R = CF3: 87%
(o-OMe:o-CF3 = 16:1)
aReaction conditions: arene (1.40 mL), 2a (0.35 mmol, 1.0 equiv),
Pd(OAc)2 (10%), L (20%), AgOPiv (1.0 equiv), Cs2CO3 (2.4 equiv)
at 120 ºC for 17 h. bYield of purified product. cAg2CO3 was used
instead of AgOPiv. dDetermined by 1H NMR spectroscopy. eCin-
namylated at the position ortho to Cl and para to OMe.
Mechanistic Studies. (π-Allyl)palladium intermediate. A
proposed mechanism for the palladium-catalyzed direct func-
tionalization of arenes with organic electrophiles generally
begins with the oxidative addition of the organic electrophile
to Pd(0).7a,c,8 Indeed, treatment of Pd[P(t-Bu)2(2-OMeC6H4)]2
(PdL2) with 5 equivalents of cinnamyl pivalate 2a at room
temperature generated (π-cinnamyl)palladium pivalate 6 (eq
2). The structure of this complex was confirmed by X-ray
crystallography. The solid-state structure contains a π-
cinnamyl ligand and an ƞ1-pivalate ligand in addition to the
phosphine (Figure 1). No interaction of the palladium center
with the ortho-methoxy group on the ligand was observed.
The difference in conversion of allylpalladium 6 and for-
mation of allylarene 3a suggests that the allylarene 3a formed
by heating 6 with fluorobenzene (1a) might not form directly
from 6. Instead, it could be generated by production of the free
cinnamyl pivalate (2a) and the active catalyst, and the active
catalyst could mediate a process to form allylarene 3a from
arene 1a and allylic pivalate 2a without the intermediacy of
allylpalladium complex 6. Therefore, this stoichiometric reac-
tion was not sufficient to confirm the intermediacy of complex
6 in the allylation process.
To distinguish between the potential role of allylpalladium 6
as an intermediate or a source of an allylic ester, we monitored
the initial formation of allylation products during the stoichi-
ometric reaction of complex 6 with 1a in the presence of 1
equivalent of allylic pivalate 2g (eq 4). If the allylic pivalate
reacted with the arene directly, then the major allylarene prod-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 7
uct at early times would be derived from the allyl group on 6. To gain information on the C–H activation process by the
1
2
3
4
5
6
7
8
However, if complex 6 is a precatalyst and the allylarene does
not form from 6, then the major allylarene product at early
times would be derived from the free allylic pivalate. After 1
min, 1.7% of allylarene 3a formed and no allylarene 3g de-
tectable by gas chromatography was formed. After ~6 min
14% of allylarene 3a and 1.1% of allylarene 3g had formed
(Figure S1 in the Supporting Information). This result implies
that complex 6 does lie on the reaction pathway and that 3a is
generated by the reaction of complex 6 with fluorobenzene 1a,
rather than the reaction of 1a with cinnamyl pivalate 2a in a
process catalyzed by a palladium complex generated from 6.
Ag complex, we prepared the phosphine-ligated silver carbox-
ylate by treatment of AgOPiv with 1 equivalent of Pt-Bu2(2-
OMeC6H4) (L) in C6D6 at room temperature. The complex
1
formed within 10 min, as determined by H and 31P NMR
spectroscopy.18 This complex exists as two unequally populat-
ed rotamers in C6D6 at room temperature with the ratio of
1:1.7 (7a:7b), due to the restricted rotation around the C(aryl)-
P bond of L. X-ray crystallographic analysis of the L-ligated
AgOPiv (7) revealed that this complex also crystallizes as two
rotamers. Both rotamers are monomeric with a κ2-pivalate
ligand (Figure 2).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Role of Ag(I) salts in direct allylation of arenes. The role of
Ag(I) salts in the direct allylation process was unusual. Be-
cause the addition of Ag(I) salts to our Pd-catalyzed allylation
enabled the process to occur with arenes that were unreactive
in the absence of the silver carboxylate (Table 1, entries 1 and
17), we hypothesized that this additive might be involved in
the step that cleaves the aryl C–H bond.10-13
Table
4.
H/D
exchange
experiments
of
1-
fluoronaphthalenea
Figure 2. ORTEP diagram of L-ligated silver pivalate (7) (P1–
Ag1 2.349(12) Å; O1–Ag1 2.135(3) Å; O2–Ag1 2.714(4) Å; P2–
Ag2 2.346(1) Å; O5–Ag2 2.305(4) Å; O4–Ag2 2.447(5) Å) (O1–
Ag1–P1 171.4(10)°; O2–Ag1–P1 136.5(9)°; O1–Ag1–O2 52.0(1)
°; O5–Ag2–P2 148.4(1)°; O4–Ag2–P2 156.0(1)°; O5–Ag2–O4
55.0(1)°) (ellipsoids are shown at 50% probability, and hydrogen
atoms are omitted for clarity).
Monitoring of the catalytic reaction under the standard con-
ditions by 31P NMR spectroscopy showed that the L-ligated
Ag complex 7 was present throughout the process. Moreover,
the initial rate of the catalytic allylation process was higher for
reactions conducted with higher concentrations of phosphine
L (Figure S3). In addition, the kinetic isotope effect (KIE) of
the direct allylation of fluorobenzene 1a and fluorobenzene-d5
with cinnamyl pivalate in separate reaction vessels was 2.3.19
Collectively, these observations are consistent with the pro-
posal that cleavage of the aryl C–H bond occurs by the L-
ligated AgOPiv 7 and that this C–H bond cleavage step strong-
ly affects the overall rate of the reaction.
We hypothesized that the reaction of the L-ligated AgOPiv
(7) with arene would form arylsilver(I) intermediate 8 (eq 5),
which could transfer an activated aryl moiety to a palladium
intermediate. Because arylsilver species are rare and no phos-
phine-ligated arylsilver complexes have been prepared previ-
ously,20 we sought additional information on the structure,
stability, and reactivity of this proposed intermediate. The
arylsilver complex 8a was prepared from the reaction of AgBr
and aryllithium 9 in the presence of L (eq 6).21 Arylsilver 8a is
stable at 4 °C under a nitrogen atmosphere, in the dark, for at
least one month.
aReaction conditions: 1b (0.10 mmol, 1.0 equiv), D2O (10 equiv),
Pd(OAc)2 (20%), L (40%), AgOPiv (1.0 equiv), Cs2CO3 (2.4 equiv)
b
1
in 1,4-dioxane (0.10 mL) at 120 ºC for 17 h. Determined by H and
19F NMR spectroscopy.
To test this hypothesis, we studied H/D exchange reactions
of 1-fluoronaphthalene 1b with 10 equivalents of D2O (Table
4; see supporting information for details). In the presence of
the combination of Pd(OAc)2, L and AgOPiv, 44% deuterium
incorporation occurred selectively at the position ortho to the
fluorine substituent of 1b at 120 °C after 17 h (entry 1). No
deuteration was observed when the reaction was conducted
without added AgOPiv (entry 2), and only a trace amount of
deuterated [D]-1b was observed without the added phosphine
L (entry 3). In addition, the H/D exchange reaction in the
presence of AgOPiv and L in the absence of Pd(OAc)2 result-
ed in 67% of deuterium incorporation at the site ortho to the
fluorine of 1b (entry 4). These results suggest that the combi-
nation of phosphine L and AgOPiv, rather than the palladium
species, are responsible for C–H activation of the arene.
ACS Paragon Plus Environment

Page 5 of 7
Journal of the American Chemical Society
or CsHCO3 and that the silver intermediate would partition
1
2
3
4
5
6
7
8
9
almost equally between transmetallation to form the allylarene
and regeneration of the silver carboxylate. This partial
reversibility of the C–H bond cleavage step accounts for the
modest KIE of 2.3.
MeO
Li
MeO
Ag
L (1 equiv)
(6)
L
AgBr
Et2O
r.t. 1.5 h
F
F
9
8a
, 87%
Although we were not able to obtain crystals of complex 8a
suitable for X-ray diffraction, we characterized an analog of
8a containing P(2-anisyl)3 as the ligand by X-ray diffraction
(Figure 3). The ORTEP diagram of this complex is shown in
Figure 3 and consists of a monomeric structure with a linear
disposition of the phosphine and aryl groups around the silver.
The Ag–C(aryl) distance of this complex (2.12 Å) is shorter
than that in mesitylsilver (2.20 Å), an arylsilver(I) complex
which lacks a phosphine ligand and is tertrameric.22
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Proposed synergistic catalytic cycle. The mechanism in
Figure 4 for the allylation of arenes with palladium(0), phos-
phine ligand and AgOPiv is consistent with all of our data. In
this mechanism, L-AgOPiv 7 reacts with the arene to form
arylsilver intermediate 8 in an endoergic step, presumably by a
pivalate-assisted CMD mechanism. In parallel, oxidative addi-
tion of the allylic ester to the Pd(0) complex bound by L
(PdL2) forms the allylpalladium complex 6. The aryl group is
transferred from silver to 6. The resulting allylpalladium aryl
complex 11 undergoes reductive elimination to form the al-
lylarene product and PdL2.
Figure 3. ORTEP diagram of the P(2-anisyl)3-ligated silver-
aryl complex (P1–Ag1 2.378(9) Å; C1–Ag1 2.123(3) Å) (P1–
Ag1–C1 178.5(1)°) (ellipsoids are shown at 50% probability,
and hydrogen atoms are omitted for clarity).
Ph
PivO
Ph
PdII
6
I
L
-Ag OPiv
L
OPiv
- L
ArH
7
The reaction of Ag complex 8a with cinnamylpalladium 6
formed allylarene 10 in 85% yield within 1 h at room tempera-
ture. PdL2 was the palladium-containing product, as deter-
mined by 31P NMR spectroscopy (eq 7); the formation of PdL2
indicates that phosphine L binds tighter to the palladium(0)
center than it binds to the silver(I) center. The high yield of
allylarene from reaction of 8a with 6 implies that the
transmetallation of the aryl group from silver to palladium and
the following reductive elimination are chemically and kinet-
ically competent to be part of the catalytic cycle.
PdL2
PivOH
I
L
-Ag -Ar
L
+
Ph
11
8
PdII
L
Ar
Ar
Ph
Figure 4. Outline of a possible mechanism for Pd-catalyzed and
Ag-mediated direct allylation of arenes with allylic pivalates.
Resting state of the catalyst. Although our C–H allylation is
not an oxidative process, we examined the possibility that the
Ag(I) salt serves as an oxidant in the catalytic system.1 Treat-
ment of bisphosphine-Pd(0) (PdL2) with 3 equivalents of
AgOPiv afforded L-AgOPiv 7 and a new palladium(II) com-
plex, which resonated in the upfield region of the 31P NMR
spectrum at –23.1 (eq 9). Single-crystal X-ray diffraction
showed that this new complex is the pivalate-bridged dimeric
palladacycle 12, resulting from the intramolecular activation
of the ortho C–H bond of the phosphine L (Figure 5).24 Com-
plex 12 also forms from the reaction of Pd(OPiv)2 and 1
equivalent of L at room temperature within 10 min in C6D6 (eq
10).
For cleavage of the arene C–H bond by L-AgOPiv 7 to
contribute substantially to the rate of the direct allylation
process, the reaction of AgAr 8a with palladium complex 6
should occur with a rate that is similar to or greater than the
reaction of AgAr 8a with PivOH (a reverse of the arene C–H
activation). To reveal these relative rates, we treated arylsilver
8a with a 1:1 ratio of allylpalladium 6 and PivOH in the
presence of 2.4 equivalents of Cs2CO3. This reaction formed
allylarene 10 in 47% yield and arene 1c in 31% yield after 30
min, and 55% yield and arene 1c in 36% yield after 1 h (eq
8).23 This result indicates that the rate of the reaction of 8a
with 6 is similar to that of the reaction of 8a with pivalic acid
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 7
CONCLUSIONS
We have discovered a method for the highly site-selective
1
formation of C(aryl)–C(sp3) bonds by the Pd-catalyzed and
Ag-mediated allylation of aryl C–H bonds with allylic
pivalates to furnish linear (E)-allylarenes in good yields. This
process occurs with arenes that have not been described to
undergo site-selective direct allylations previously, including
those containing fluoro, chloro, trifluoromethyl, or methoxy
groups. The observed site-selectivity suggests that the C–H
bond cleavage proceeds through a concerted metalation-
deprotonation pathway at the most acidic C–H bond.
The available mechanistic data are consistent with a syner-
gistic catalytic process wherein the cleavage of an aryl C–H
bond occurs by a phosphine-ligated AgOPiv complex; the
resulting arylsilver species then transfers an aryl group to the
(π-allyl)palladium pivalate species formed by oxidative-
addition of an allylic pivalate to Pd(0). Subsequent C–C bond-
forming reductive elimination affords the allylation product
and regenerates the Pd(0) species. The proposed palladium
and silver intermediates have been synthesized, unambiguous-
ly characterized by X-ray crystallography, and shown to be
chemically and kinetically competent to be intermediates by
the suggested elementary steps of the catalytic cycle. The im-
plications of these findings to other palladium-catalyzed reac-
tions that occur with silver additives and that have been com-
puted to occur by clusters containing Ag and Pd as well as the
applications of Ag(I)-mediated C–H activation to new C–H
functionalization processes are underway.
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ASSOCIATED CONTENT
Supporting Information
Figure 5. ORTEP diagram of dimeric palladacycle 12 (Pd1–
Pd2 3.248(6) Å; C1–Pd1 1.984(3) Å; P1–Pd1 2.237(9) Å; O2–
Pd1 2.130(2) Å; O4–Pd1 2.105(2) Å) (P1–Pd1–C1 69.3(9)°; Pd1–
C1–C6 108.2(2)°; C1–C6–P1 97.1(2)°; C6–P1–Pd1 85.4(1)°; O4–
Pd1–O2 90.8(8)°) (ellipsoids are shown at 50% probability, and
hydrogen atoms are omitted for clarity).
Experimental procedures and compound characterization data.
This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
jhartwig@berkeley.edu
Palladacycle 12 is the resting state of the palladium catalyst
during the allylation process. This complex and the L-ligated
silver complex 7 were the major phosphine-ligated complexes
in the system, as determined by monitoring the catalytic
reaction of fluorobenzene with cinnamyl pivalate by 31P NMR
spectroscopy (Figure S5). We have also determined that
palladacycle 12 catalyzes the allylation of 1a with 2a to form
allylarene 3a (eq 11). The yield (80%) of this process is
similar to that of the reaction catalyzed by the combination of
Pd(OAc)2 and L or by PdL2 (Table 1, entries 1 and 2).
Moreover, the initial rate and induction period (~8 min) for the
reaction catalyzed by palladacycle 12 (eq 11) are similar to
those of the reactions catalyzed by the combination of
Pd(OAc)2 and L or by PdL2 (Table 1, entries 1 and 2) (Figure
S2). This result indicates that palladacycle 12 does not lie on
the reaction pathway but that it generates the active Pd species,
presumably a Pd(0) species bound by the phosphine, in the
catalytic cycle of the allylation and is not formed irreversibly
as an inactive catalyst.23,25 The mechanism by which 12 forms
a Pd(0) species is not clear.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
This work was supported by the Director, Office of Science, of
the U.S. Department of Energy under contract No. DE-AC02-
05CH11231 and the National Institutes of Health (postdoctoral
fellowship to S.Y.L.: F32-GM113404). We thank Dr. Antonio
DiPasquale for X-ray crystallographic analysis (NIH Shared In-
strumentation Grant S10-RR027172) and Thomas J. O’Connor
(supported by UC Berkeley Amgen Scholars program and
CENTC program) for assistance on the initial investigation.
S.Y.L. thanks Taegyo Lee for insightful discussions and David M.
Peacock for assistance on NMR experiments and for the gift of di-
tert-butylarylphosphines used for initial studies.
REFERENCES
(1) For reviews on C–H functionalizations, see: (a) Lyons, T. W.;
Sanford, M. S. Chem. Rev. 2010, 110, 1147. (b) Yu, J.-Q., Shi, Z.,
Eds. Topics in Current Chemistry: C−H Activation, 1st ed.; Springer:
Berlin/Heidelberg, 2010. (c) Kuhl, N.; Hopkinson, M. N.; Wencel-
Delord, J.; Glorius, F. Angew. Chem., Int. Ed. 2012, 51, 10236. (d) He,
M.; Soulé, J.-F.; Doucet, H. ChemCatChem 2014, 6, 1824. (e) Liu,
ACS Paragon Plus Environment

Page 7 of 7
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
C.; Yuan, J.; Gao, M.; Tang, S.; Li, W.; Shi, R.; Lei, A. Chem. Rev.
2015, 115, 12138.
(11) Lotz, M. D.; Camasso, N. M.; Canty, A. J.; Sanford, M. S.
Organometallics 2016, DOI: 10.1021/acs.organomet.6b00437.
(12) For examples on arene C–H activation by silver carbonate, see:
(a) He, C.; Min, Q.; Zhang, X. Organometallics 2012, 31, 1335 (b). Li,
H.; Liu, J.; Sun, C. L.; Li, B. J.; Shi, Z. J. Org. Lett. 2011, 13, 276.
(13) For examples on Pd-Ag bi- or mulitmetallic intermediates in
arene C–H cleavage, see: (a) Anand, M.; Sunoj, R. B.; Schaefer, H. F.
J. Am. Chem. Soc. 2014, 136, 5535. (b) Yang, Y.; Cheng, G.; Liu, P.;
Leow, D.; Sun, T.; Chen, P.; Zhang, X.; Yu, J.; Wu, Y.; Houk, K. N. J.
Am. Chem. Soc. 2014, 136, 344. (c) Anand, M.; Sunoj, R. B.;
Schaefer, H. F. ACS Catalysis 2016, 6, 696.
(14) (a) Empsall, H. David; Shaw, Bernard L.; Turtle, Brian L. J.
Chem. Soc., Dalton Trans. 1976, 1500. (b) Bornand, M.; Torker, S.;
Chen, P. Organometallics 2007, 26, 3585.
(15) (a) Allylic electrophiles containing heterocycles such as furan
or thiophene were not compatible with our allylation process. (b) The
reactions with allylic electrophiles that bear β-hydrogens upon form-
ing an allylpalladium intermediate (e.g., (E)-but-2-en-1-yl pivalate)
resulted in producing β-hydride elimination products.
(2) For selected examples on direct arylations of arenes, see: (a)
Lafrance, M.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 16496. (b)
Ball, L. T.; Lloyd-Jones, G. C.; Russell, C. A. Science 2012, 337,
1644. (c) Ricci, P.; Krämer, K.; Cambeiro, X. C.; Larrosa, I. J. Am.
Chem. Soc. 2013, 135, 13258. (d) Wagner, A. M.; Hickman, A. J.;
Sanford, M. S. J. Am. Chem. Soc. 2013, 135, 15710. (e) Durak, L. J.;
Lewis, J. C. Organometallics 2014, 33, 620. (f) Ricci, P.; Krämer, K.;
Larrosa, I. J. Am. Chem. Soc. 2014, 136, 18082.
(3) For selected references on alkenylation and alkynylation of
arenes, see: (a) Vora, H. U.; Silvestri, A. P.; Engelin, C. J.; Yu, J.-Q.
Angew. Chem., Int. Ed. 2014, 53, 2683. (b) de Haro, T.; Nevado, C. J.
Am. Chem. Soc. 2010, 132, 1512.
(4) (a) Shi, Z.; He, C. J. Am. Chem. Soc. 2004, 126, 13596. (b)
Ying, C.-H.; Duan, W.-L. Org. Chem. Front. 2014, 1, 546.
(5) G. A. Olah, Friedel–Crafts and Related Reactions, Vol. 2,
Wiley-Interscience, New York, 1964.
(6) For selected recent examples on direct allylation of arenes bear-
ing a directing group, see: (a) Wang, H.; Schröder, N.; Glorius, F.
Angew. Chem., Int. Ed. 2013, 52, 5386. (b) Ye, B.; Cramer, N. J. Am.
Chem. Soc. 2013, 135, 636. (c) Liu, W.; Richter, S. C.; Zhang, Y.;
Ackermann, L. Angew. Chem., Int. Ed. 2016, 55, 7747. (d) Yamagu-
chi, T.; Kommagalla, Y.; Aihara, Y.; Chatani, N. Chem. Commun.
2016, 52, 10129. (e) Choi, M.; Park, J.; Sharma, S.; Jo, H.; Han, S.;
Jeon, M.; Mishra, N. K.; Han, S. H.; Lee, J. S.; Kim, I. S. J. Org.
Chem. 2016, 81, 4771.
(7) For examples on direct allylation of polyfluoroarenes, see: (a)
Fan, S.; Chen, F.; Zhang, X. Angew. Chem., Int. Ed. 2011, 50, 5918.
(b) Yao, T.; Hirano, K.; Satoh, T.; Miura, M. Angew. Chem., Int. Ed.
2011, 50, 2990. (c) Yu, Y,-B.; Fan, S.; Zhang, X. Chem. Eur. J. 2012,
18, 14643. (d) Jiang. H.; Yang, W.; Chen, H.; Li, J.; Wu, W. Chem.
Commun. 2014, 50, 7202. (e) Xie, W.; Chang, S. Angew. Chem., Int.
Ed. 2016, 55, 1736; for an example of electron-deficient azoles, see:
(f) Makida, Y.; Ohmiya, H.; Sawamura, M. Angew. Chem., Int. Ed.
2012, 51, 4122.
(8) (a) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry,
2nd ed.; Wiley-Interscience: New York, 1987; pp 322-333. (b) Hand-
book of Organopalladium Chemistry for Organic Synthesis; Negishi,
E.-i. Ed.; Wiley Interscience: New York, 2002.
(9) For reviews and leading references, see: (a) Lapointe, D.;
Fagnou, K. Chem. Lett. 2010, 39, 1118. (b) Ackermann, L. Chem.
Rev. 2011, 111, 1315–1345. (c) Gorelsky, S. I.; Lapointe, D.; Fagnou,
K. J. Org. Chem. 2012, 77, 658. (d) Lafrance, M.; Rowley, C. N.;
Woo, T. K.; Fagnou, K. J. Am. Chem. Soc. 2006, 128, 8754−8756. (e)
García-Cuadrado, D.; de Mendoza, P.; Braga, A. A. C.; Maseras, F.;
Echavarren, A. M. J. Am. Chem. Soc. 2007, 129, 6880. (f) Guihaumé,
J.; Clot, E.; Eisenstein, O.; Perutz, R. N. Dalton Trans. 2010, 39,
10510.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(16) Biaryls resulting from the direct arylation of arenes with aryl
chlorides were obtained as side products.
(17) Complex 6 with 1 equivalent of L at 120 °C in fluorobenzene
forms a 4:1 mixture of 6 and PdL2 after 10 min. In the presence of
AgOPiv, PdL2 is oxidized to generate dimeric palladacycle 12 (eq 9).
(18) For examples on silver carboxylates ligated by a phosphine,
see: (a) Ng, S. W.; Othman, a. H. Acta Crystallogr. Sect. C Cryst.
Struct. Commun. 1997, 53, 1396. (b) Paramonov, S. E.; Kuzmina, N.
P.; Troyanov, S. I. Polyhedron 2003, 22, 837. (c) Partyka, D. V; Deli-
gonul, N. Inorg. Chem. 2009, 48, 9463.
(19) Simmons, E. M.; Hartwig, J. F. Angew. Chem., Int. Ed. 2012,
51, 3066.
(20) Schmidbaur, H.; Bayler, A.; Synthesis and uses of organosilver
compounds, in The Chemistry of Organic Derivatives of Gold and
Silver, Patai. S.; Rappoport, Z., eds., Wiley, Chichester, UK, 1999, pp.
211-226.
(21) Leusink, A. J.; Koten, G. V.; Noltes, J. G. J. Organomet.
Chem. 1973, 56, 379.
(22) (a) Gambarotta, S.; Floriani, C.; Chiesi-Villa, A.; Guastini, C.
J. Chem. Soc., Chem. Commun. 1983, 1087. (b) Meyer, E. M.; Gam-
barotta, S.; Floriani, C.; Chiesi-Villa, A.; Guastini, C.
Organometallics 1989, 8, 1067.
(23) When the reaction was conducted in the absence of Cs2CO3,
20 % of allylarene 10 and 66 % of arene 1c were formed after 1 h.
(24) For references on the related palladacycle resulting from the
intramolecular C–H activation of P(t-Bu)3, see: (a) Geisseler, H.;
Gross, P.; Guckes, B. German Patent GWXXBX 19647584, 1998;
CAN 129:28078. (b) Henderson, W. H.; Alvarez, J. M.; Eichman, C.
C.; Stambuli J. P. Organometallics 2011, 30, 5038.
(25) Stambuli’s study on the related palladacycle with P(t-Bu)3
established that this complex could form PdP(t-Bu)3]2 upon thermal
decomposition (ref 24b).
(10) Whitaker, D.; Burés, J.; Larrosa, I. J. Am. Chem. Soc. 2016,
138, 8384.
Table of Contents Graphic:
ACS Paragon Plus Environment