Integrating Metal-Catalyzed C-H and C-O Functionalization to Achieve Sterically Controlled Regioselectivity in Arene Acylation

Article abstract of DOI:10.1021/jacs.8b06476

One major goal of organometallic chemists is the direct functionalization of the bonds most recurrent in organic molecules: C-H, C-C, C-O, and C-N. An even grander challenge is C-C bond formation when both precursors are of this category. Parallel to this is the synthetic goal of achieving reaction selectivity that contrasts with conventional methods. Electrophilic aromatic substitution (EAS) via Friedel-Crafts acylation is the most renowned method for the synthesis of aryl ketones, a common structural motif of many pharmaceuticals, agrochemicals, fragrances, dyes, and other commodity chemicals. However, an EAS synthetic strategy is only effective if the desired site for acylation is in accordance with the electronic-controlled regioselectivity of the reaction. Herein we report steric-controlled regioselective arene acylation with salicylate esters via iridium catalysis to access distinctly substituted benzophenones. Experimental and computational data indicate a unique reaction mechanism that integrates C-O activation and C-H activation with a single iridium catalyst without an exogenous oxidant or base. We disclose an extensive exploration of the synthetic scope of both the arene and the ester components, culminating in the concise synthesis of the potent anticancer agent hydroxyphenstatin.

Full text of DOI:10.1021/jacs.8b06476

Subscriber access provided by UNIV OF DURHAM
Article
Integrating Metal-Catalyzed C–H and C–O Functionalization to
Achieve Sterically Controlled Regioselectivity in Arene Acylation
Nicholas A. Serratore, Constance B. Anderson, Grant B. Frost, Truong-Giang Hoang,
Steven Underwood, Philipp Gemmel, Melissa A. Hardy, and Christopher J. Douglas
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b06476 • Publication Date (Web): 10 Jul 2018
Downloaded from http://pubs.acs.org on July 10, 2018
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 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 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.
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 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Integrating Metal-Catalyzed C–H and C–O Functionalization to
Achieve Sterically Controlled Regioselectivity in Arene Acylation
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
‡
Nicholas A. Serratore, Constance B. Anderson, Grant B. Frost‡, Truong-Giang Hoang† , Steven J. Un-
derwood‡, Philipp M. Gemmel, Melissa A. Hardy§, Christopher J. Douglas*
Department of Chemistry, University of Minnesota – Twin Cities, 207 Pleasant Street SE, Minneapolis, Minnesota
55455
ABSTRACT: One major goal of organometallic chemists is
the direct functionalization of the bonds most recurrent in
organic molecules: C–H, C–C, C–O, and C–N. An even
grander challenge is C–C bond formation when both pre-
cursors are of this category. Parallel to this is the synthetic
goal of achieving reaction selectivity that contrasts with
conventional methods. Electrophilic aromatic substitution
(EAS) via Friedel-Crafts acylation is the most renowned
method for the synthesis of aryl ketones, a common struc-
tural motif of many pharmaceuticals, agrochemicals, fra-
grances, dyes, and other commodity chemicals. However,
an EAS synthetic strategy is only effective if the desired
site for acylation is in accordance with the electronic-controlled regioselectivity of the reaction. Herein we report the steric-
controlled regioselective arene acylation with salicylate esters via iridium catalysis to access distinctly substituted benzophe-
nones. Experimental and computational data indicate a unique reaction mechanism that integrates C–O activation and C–H
activation with a single iridium catalyst without an exogenous oxidant or base. We disclose an extensive exploration of the
synthetic scope of both the arene and the ester components, culminating in the concise synthesis of the potent anti-cancer
agent hydroxyphenstatin.
INTRODUCTION
Traditional metal-catalyzed cross-coupling revolutionized
both the fields of organometallic catalysis and organic syn-
thesis (Figure 1, a).1 However, the need for metal/halogen
pre-functionalization (C–M’, C–X) and inevitable production
of stoichiometric inorganic waste presented the oppor-
tunity for the development of more efficient coupling reac-
tions. To this end, the fields of metal-catalyzed C–H,2-10 C–
C,11-14 C–N,15-19 and C–O20-22 (C–Y) functionalization continue
to significantly progress (Figure 1, b). Numerous types of
reactions involving these groups have been developed,
however many still involve one metal/halogen pre-func-
tionalized partner along with stoichiometric exogenous ox-
idants and/or bases. Combining two metal-catalyzed C–Y
activations into a single C–C bond forming reaction is a
unique challenge. Recently, the seminal efforts merging C–
H with C–H23 and C–C24-25 activation have been recognized
for their potential to transform synthetic strategy (Figure
1, c). There are also a few initial reports of merging C–H
with amide C–N26-27 and anhydride C–O activation.28-30 How-
ever, many of these C–H/C–Y activation reactions still re-
quire stoichiometric oxidants and/or bases. The work pre-
Figure 1. Metal-catalyzed C–C bond forming organic bond
functionalization
sented here integrates C–H with ester Cacyl–O activation
(Figure 1, d). Our mechanistic analysis supports both bond
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 10
1
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
Figure 2. Methods for regioselective acylation. (a) Friedel-Crafts acylation biased by electronics. (b) Directing group-controlled
acylation. (c) Potential multi-step steric-controlled acylation, through steric-controlled borylation followed by cross-coupling.
(d) Our work: single-step steric-controlled acylation by sequential Cacyl–O, C–H bond activation.
activation events occurring at a single iridium catalyst in the
catalytic cycle. A conceptual advance in our strategy is using
the ester C–O bond to provide an endogenous oxidant as
well as a base to complete the catalytic cycle.
Organometallic reaction development also has a rich his-
tory of creating new possibilities for synthetic strategy by
achieving reaction selectivity counter to the conventions of
the time. Modern examples include metal-catalyzed reac-
tions designed to overcome the electronic-controlled regi-
oselectivity of electrophilic aromatic C–H substitutions.31-38
In Friedel-Crafts reactions electron-donating groups (EDG)
Scheme 1. Optimized reaction conditions.
direct acylation to the ortho and para positions, in many
cases producing a mixture of products (Figure 2, a).39 Sin-
gle-site acylation via metal-catalyzed C–H functionalization
is typically achieved using a covalently attached or transient
directing group (DG, Figure 2, b). 28-30,40-41 DG’s permit acyl-
ation primarily at the ortho position (within the regioselec-
tivity purview of Friedel-Crafts acylation) but meta selectiv-
ity is also known.42 Steric-controlled regioselective arene
acylation is possible, but would require multiple synthetic
steps and several precious metal catalysts (Figure 2, c).
Specifically, iridium32,35,37-38 or iron-catalyzed34 C–H boryla-
tion may be used to transform the most sterically accessible
C–H bond into a C–B bond. Boryl-group manipulation and
Suzuki-Miyaura cross coupling would complete the acyla-
tion.43-44 We are unaware of any steric-controlled C–H bond
functionalization method for acylation with a single cata-
lyst. Our iridium-catalyzed sequential bond activation of an
ester Cacyl–O bond and an arene C–H bond permits single-
step, steric-controlled, regioselective arene acylation (Fig-
ure 2, d).
product. Optimized reaction conditions (see supporting in-
formation, Tables S2–S8) for the acylation of the solvent, m-
xylene, with phenyl salicylate gave the meta- and or-
tho/para-substituted products (2aa1 and 2aa2) in an 11:1
ratio. Further, the 11:1 ratio did not change over time as ob-
served by 1H NMR, suggesting kinetic selectivity. This regi-
oselectivity directly contrasts with classic Friedel-Crafts ac-
ylation of m-xylene with benzoyl chloride, where the ortho-
/para- and di-ortho-substituted products were greatly fa-
vored over meta-substitution (99:1).46 This comparison
highlights that our new reaction is not dictated by elec-
tronic-controlled regioselectivity.
RESULTS
We find the two mechanistic paths shown for our acylation
reaction to be most plausible (Scheme 2). These paths are
based on our results and analysis that follow. Phenyl salicy-
late 1 coordinates to [Ir(cod)OMe]2 (A): breaking the dimer,
liberating methanol, and forming square planar intermedi-
ate I, the proposed resting-state of the catalyst. Relative to
catalyst loading, approximately two equivalents of methyl
salicylate II were observed, suggesting sacrificial trans-
esterification of 1 with methanol (B). Oxidative addition
into the Cacyl–OPh bond of I (C) forms square pyramidal
complex III.47 This step and all subsequent steps are pro-
posed to be reversible. From III, multiple pathways are
plausible. First, the limiting factor in catalyst lifetime is pro-
During our investigations into iridium-catalyzed Cacyl–O
bond activation of salicylate esters for intramolecular al-
kene oxyacylation,45 we discovered the intermolecular acyl-
ation of the solvent, m-xylene (Scheme 1). The major regi-
oisomer of this acylation was the sterically-favored meta-
substituted product. Control experiments (see supporting
information, Table S1) showed that iridium was required
for product formation and phosphine improved the yield of
2
ACS Paragon Plus Environment

Page 3 of 10
Journal of the American Chemical Society
1
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
Scheme 2. Plausible mechanisms via either a 4-member or 6-member CMD transition state.
posed to be decarbonylation (D) to irreversibly form irid-
ium carbonyl complex IV. TrixiePhos is proposed to inhibit
this decomposition by reversibly displacing COD (E), stabi-
lizing the Cacyl–O oxidative addition intermediate off-cycle
as V.48 III may continue the productive path through coor-
dination of the arene -bond followed by -adduct for-
mation (F) with Caryl–H bond forming VI. This leads to the
key steric-controlled C–H bond activation step, a 1,2-addi-
tion via concerted metalation-deprotonation (CMD) across
the Ir–OPh bond (G). In this step, a lone pair on the phenox-
ide oxygen deprotonates the Caryl–H through a 4-member
transition state, forming phenol and the new Ir–Caryl com-
oxidative addition of Ir into the Ph–O or Me–O bond and
subsequent protonation may account for the formation of
IX. IX may then enter the catalytic cycle by binding to III (K)
to form octahedral complex X. Elimination of phenol (L)
would then form square pyramidal XI. Arene coordination
(M) would produce XII, which may proceed through a less
strained 6-member CMD transition state to form VII. In ei-
ther proposed mechanism, the base required to deproto-
nate the arene is endogenous to the system. In addition, no
exogenous oxidant is required for catalyst turnover due to
the net reduction of the ester to a ketone over the course of
the reaction.
plex VII.49 Reductive elimination (H) forms the new Cacyl
–
We support our proposed mechanism with evidence from a
series of experiments and computations designed to probe
key mechanistic aspects of the reaction. H/D kinetic isotope
effect (KIE) experiments51 were performed with d4-1,2-di-
methoxybenzene and 4-d1-1,2-dimethoxybenzene (Scheme
3, a). These experiments added insight into the details of
arene coordination through C–H cleavage (III → VI → VII).52
The method of initial rates kinetics was used to determine
independent rates for proteo- and d4-1,2-dimethoxyben-
zene. A KIEinital rates = 1.8 ± 0.1 was measured. Product ratios
Caryl bond, giving square planar VIII with new acylation
product 2 bound to iridium. Finally, ligand exchange of 2 for
1 (I) regenerates intermediate I.
An alternative mechanism involving a more favorable car-
boxylate-assisted 6-member CMD transition state was also
considered.50 We envision that a formal hydrolysis of 1 or II
(J) could form salicylic acid IX to serve as the carboxylate.
Either the presence of trace H2O or possibly an alternative
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 10
1
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
Scheme 3. Mechanistic Analysis (a) Kinetic isotope effect (KIE) experiments. (b) Observed KIE relations and explanation. (c)
Intramolecular KIE mechanistic analysis. (d) Intermolecular KIE mechanistic analysis.
were used to determine the KIEs of intramolecular and in-
termolecular competition experiments. The intramolecular
competition with 4-d1-1,2-dimethoxybenzene gave a simi-
lar KIEintramolecular =1.6 ± 0.1, but interestingly the intermo-
lecular competition with a 1:1 mixture of d4- and proteo-1,2-
dimethoxybenzene gave an amplified KIEintermolecular = 4.6 ±
0.2.
perimental KIE values KIEinitial rates = 1.8 ± 0.1 and KIEintramolec-
=1.6 ± 0.1 are within error and therefore in accordance
with this analysis.
ular
In the intermolecular competition reaction, the value of
KIEintermolecular = 4.6 ± 0.2 may arise from differential reversi-
bility of arene coordination to III (Scheme 3, d). This would
result in an increased [VIπ-H] relative to [VIπ-D], which in
turn would lead to an increased [VIσ-H] relative to [VIσ-D].
[VIσ-H]/[VIσ-D] is expressed in terms of forward and reverse
rate constants (see supporting information), derived with
multiple steady-state approximations. The differential re-
versibility may be explained by secondary isotope effects
and/or the preference for H abstraction over D abstraction
leading to a higher rate of d4-1,2-dimethoxybenzene disso-
ciation. This would channel more reacting arene through
the proteo-1,2-dimethoxybenzene path.56 The net effect is
measured as an amplification of (kH/kD)true by a factor of 2.7.
We also note that KIEintermolecular = 4.6 ± 0.2 is comparable to
the intermolecular competition KIE found in arene boryla-
tion via iridium-catalyzed C–H functionalization (5.0 ±
0.4);58 and it contrasts with the analogous Friedel-Crafts KIE
(1.010 ± 0.0007).59
The observation of separate KIE values for the set of exper-
iments suggests a multistep mechanism from arene coordi-
nation to C–H cleavage.53-54 A primary KIEinitial
is evi-
rates
dence that C–H cleavage is the rate-limiting step of the
mechanism and a measure of the true KIE of C–H/D cleav-
age ((kH/kD)true) (Scheme 3, b). Our analysis of the intra-
and intermolecular KIE’s was primarily influenced by the
work of Beak et al.55 and Adam et al.56 In the case of the in-
tramolecular KIE we propose that the selectivity bias for H
over D abstraction arises from the equilibration of the gen-
eral arene complex VI between σ-adducts (Scheme 3, c).
VIσ-H and VIσ-D may interconvert through VIπ.54, 57 This is fur-
ther evidenced by the fact that KIEintermolecular ≠ KIEintramolecular
implying that arene coordination is a separate step from the
intramolecular selectivity-determining step. If VI fully
equilibrates between σ-adducts and C–H/D cleavage is rate-
limiting (kσH, k-σH, kσD, k-σD >> kH, kD) then the observed KIEin-
Experiments suggest phosphine is necessary for high turn-
over of the catalyst, but is not involved in the catalytic cycle.
Acylation of 1,2-dimethoxybenzene with 1, using the opti-
mized reaction conditions, gave >46 turnovers of catalyst
versus only 12 turnovers when phosphine was excluded
should theoretically reflect ((kH/kD)true).55 Our ex-
tramolecular
4
ACS Paragon Plus Environment

Page 5 of 10
Journal of the American Chemical Society
(see supporting information). Without iridium, no reaction When TrixiePhos was present in concentrations ≥1.8 mM,
1
2
3
4
5
6
7
8
is observed, even with phosphine present. After heating
[Ir(cod)OMe]2 with two equivalents of TrixiePhos in 1,2-di-
methoxybenzene to 170 °C for one hour, we did not observe
a change in the 31P NMR spectra of the reaction mixture.
When this same experiment was performed on a reaction
mixture containing salicylate 1, the phosphine signal de-
creased in intensity over twenty hours. This suggests that
phosphine does not bind to the iridium precatalyst. Instead,
phosphine may be in a dynamic equilibrium between cata-
lytic intermediate III and stable, octahedral complex V, po-
tentially slowing the decomposition of catalyst. Phosphine
loading studies showed that excess phosphine led to a lower
yield of product (see supporting information). This would
be expected if phosphine were an inhibitor.
the reaction was inverse order with respect to phosphine.
This region of inverse order is consistent with the off-cycle
role of phosphine in our proposed catalytic cycle.
We hypothesize that decarbonylation is the main catalyst
decomposition pathway. An infrared spectrum of a crude
reaction mixture after twenty hours was taken showing
peaks at 1981 and 2027 cm-1 (see supporting information),
consistent with a catalytically inactive iridium carbonyl
complex IV.60 If decarbonylation to form IV proceeds via a
bimetallic mechanism, we expect that lower concentrations
of III, induced by phosphine, would slow the rate of decar-
bonylation (via VI) relative to the rate of acylation. Our data
are consistent with this hypothesis, suggesting that iridium
sequestration by phosphine leads to higher turnover num-
bers at optimized conditions. In addition, catalyst poisoning
under a CO atmosphere inhibits the reaction. An attempted
acylation of 1,2-dimethoxybenzene with phenyl salicylate
under 1 atm CO results in <10% conversion after 20 hours.
We attempted to increase the persistence of our catalyst by
acylating 1,2,3-trimethoxybenzene with phenyl salicylate
using a reflux apparatus under a flow of N2, hoping to sweep
away CO formed via decarbonylation. We observed similar
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
To elucidate the kinetic order of the components of the re-
action, rate data was collected for the reaction of phenyl sa-
licylate and 1,2-dimethoxybenzene using the method of ini-
tial rates. This data showed zero order with respect to salic-
ylate and 1,5-cyclooctadiene (1,5-COD). The reaction ap-
peared to be greater than first order, however solvent po-
larity effects could have caused an accelerated reaction at
relatively high concentrations of 1,2-dimethoxybenzene.
Figure 3. (a) M15-L Computational Analysis. (b) Higher energy configurations and phosphine analogues.
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
results with this setup as in a sealed vessel. Also, a crucial
control experiment to test reversibility of the reaction was
performed. The reaction of 2 with excess phenol in mesity-
lene produced starting ester 1 suggesting the entire cycle is
reversible (See supporting information).
Page 6 of 10
1
2
3
4
5
6
7
8
A computational investigation into the mechanism of the re-
action was pursued using the Gaussian 09 and Gaussian 16
suites (See supporting information for citations). The M06,
M06-L, and M15-L density functionals were used to perform
these calculations. The 6-31G(d,p) basis sets on H, C, O, and
P atoms and the SDD basis set on iridium were used for ge-
ometry optimization. The 6-311+G(2df,2pd) basis sets on H,
C, O, and P atoms and the SDD basis set on iridium were
used for single point energy calculations with SMD solva-
tion using toluene as the solvent. Toluene was therefore
chosen as the arene reactant for computational simplicity,
although it is likely that the computed barriers would be
lower using a more electron-rich arene such as 1,2-di-
methoxybenzene. In addition, since JohnPhos performed
similarly to TrixiePhos during reaction optimization (see
supporting information) we used it as a TrixiePhos ana-
logue to further reduce computational complexity. Overall,
the functional that best supported our experimental results
was M15-L. This is unsurprising as M15-L was specifically
developed to model transition-metal complex thermochem-
istry.61 Thus, we chose to only discuss these results in the
text.62
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
Comparing I to XIII and XIII2, it is clear that the coordina-
tion of phosphine to iridium is uphill relative to COD as an
ancillary ligand (Figure 3, a and b). Inspection of the
1–3
JohnPhos complexes VJP shows a large amount of steric
congestion around the iridium. This result suggests that
COD must fully dissociate if there is an equilibrium between
III and V. The oxidative addition transition state complex C
has a relative energy of +22.3 kcal which is reasonable given
the reaction temperature is 170 °C. The 4-member transi-
tion state for CMD G is much higher than expected (+53.2
kcal) with the phenoxide deprotonating the arene. Lower
transition state energies are known in the literature using
carboxylate ligands,50 which have the advantage of forming
more energetically favorable 6-member transition states. A
lower energy transition state was sought using a 6-member
scaffold. Two possible transition state structures were pro-
posed, a first using a second phenol to coordinate and act as
a proton shuttle, forming a 6-member transition state (see
supporting information), and a second using the conjugate
base of salicylic acid IX as a ligand. The transition state with
phenol acting as a proton shuttle had a staggeringly high
barrier of 67.1 kcal, which suggests this pathway is not ac-
cessible in our system. The transition state with salicylate
deprotonating the arene N has a significantly lower barrier
of 43.4 kcal, suggesting that a carboxylate could be acting as
our base for CMD. Of particular interest, the reductive elim-
ination product VIII is uphill by ~2 kcal, suggesting that the
reaction is under thermodynamic equilibrium. Further, the
resting state of the catalyst is structure I, however XI cannot
be ruled out based on our computations. The result that the
reaction appears to be thermodynamically uphill led us to
Figure 4. Substrate scope with respect to the arene. Per-
formed with phenyl salicylate 1 (1 equiv), [Ir(cod)OMe]2 (1
mol%), TrixiePhos (3 mol%), 1,5-cyclooctadiene (1,5-COD,
a
1 equiv), in arene (0.1 M). isolated as an inseparable mix-
1
b
ture; product distribution determined by H NMR. Minor
product 2a_3 inseparable from 2a_1; product distribution
determined by 1H NMR. cAfter 20 h, a second charge of cat-
alyst (1 mol% [Ir(cod)OMe]2, 3 mol% TrixiePhos) was
d
added, and allowed to react at 170 °C for 20 h. Unable to
be fully purified, yield is an upper bound.
calculate the theoretical equilibria for a few additional reac-
tions (see supporting information).
With an understanding of the mechanism, we examined the
scope of this reaction. The scope with respect to the arene
was investigated with mono-, di-, and tri-substituted arenes
(Figure 4). Electron-rich arenes react well under these con-
ditions (2aa–2ac, 2af–2ah, 2an, 2ao), especially compared
to electron-poor arenes (2ad, 2ae, 2ai, 2aj). Acylation of
strongly electron-poor arenes or some functionalized
arenes, was not observed. Substitution generally occurs at
the most sterically accessible site. When multiple C–H
bonds in similar steric environments are available, as in un-
symmetrical 1,2-disubstituted arenes 2ah–2aj, acylation at
6
ACS Paragon Plus Environment

Page 7 of 10
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Chart 1. Comparison of Friedel-Crafts acylation electronic-
controlled regioselectivity (purple) to our acylation
method (blue).
the more electron-rich position is preferred. In the absence
of strong electronic effects, mixtures favoring acylation at
the more accessible sites result. For example, acylation of
toluene (2am) provides a 17:17:1 mixture of regioisomers,
favoring the meta and para isomers. As arenes with
stronger electron-donors undergo acylation, as in anisole
(2an) and N,N-dimethylaniline (2ao), the para isomer is fa-
vored and the ortho isomer is not observed. The reaction
tolerates simple arenes with fluoro, trifluoromethyl, meth-
oxy, methyl, methyl ester, and dimethylamino functional
groups. N-methylated heteroarenes (2ar, 2as) were toler-
ated, with acylation of N-methylpyrrole occurring mostly at
the sterically-favored 3-position (10:1 ratio of regioiso-
mers), contrary to Friedel-Crafts acylation.63 In the case of
N-methylindole (2as), the 3-position relative to the nitro-
gen is favored both sterically and electronically, thus only
one product is observed. Interestingly, naphthalene was ac-
ylated almost exclusively at the -position, the opposite se-
lectivity of Friedel-Crafts acylation.64 The contrasting regi-
oselectivity of this reaction is summarized in Chart 1. The
substrates displayed have distinctive electronic/steric-fa-
vored sites. Electron rich 1,3-disubstituted arenes show
universal selectivity at ortho/para positions with Friedel-
Crafts acylation.46, 63–67 Statistically, the electronically pre-
ferred sites for acylation outnumber the sterically accessi-
ble sites for many of the arenes in Chart 1. Yet, the sterically
preferred sites are selectively acylated with our iridium cat-
alyst. Our chemistry shows good selectivity for the steric-
controlled product, although with a noticeable decrease in
selectivity as the electron density of the ring increases. Very
electron poor arenes were not successful with this method-
ology. In addition, arenes with Lewis basic functional
groups (e.g. nitrile, thiophene, carbamates, or acetals) were
not successful.
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
Figure 5. Substrate scope with respect to the salicylate.
Performed with phenyl salicylate 1_ (1 equiv),
[Ir(cod)OMe]2 (1 mol%), TrixiePhos (3 mol%), 1,5-COD (1
equiv), in arene 1b or 1c (0.1 M). For acylation of m-xylene
(2b_), a second charge of catalyst (1 mol% [Ir(cod)OMe]2,
3 mol% TrixiePhos) was added after 20 h, and allowed to
react for an additional 20 h. aisolated as an inseparable mix-
ture; product distribution determined by 1H NMR. b2b_2
observed and inseparable from mixture; product distribu-
tion determined by 1H NMR. c Yield determined by 1H NMR;
product decomposed during attempted purification.
Next, the scope of arene acylation with respect to the salic-
ylate ester was examined (Figure 5). Acylation of m-xylene
with two charges of catalyst gave the corresponding biaryl
ketones in respectable yields. The regioselectivity of the C–
H activation still strongly favored the corresponding 1,3,5–
substituted products, with at least a >10:1 ratio of isomers.
Interestingly, many of the substituted salicylates also
yielded 2aa1, with apparent loss of the functional group R.68
To access optimal reactivity of these salicylates, acylations
were also performed using 1,2-dimethoxybenzene. Due to
1,2-dimethoxybenzene being electron-rich, only one charge
of catalyst is required, and only one acylation product was
observed under these conditions with no observed loss of
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 10
gave hydroxyphenstatin 5 in 48% yield (36% yield over 2
1
2
3
steps). This synthesis demonstrates a new disconnection
available for the synthesis of biologically relevant mole-
cules.
4
5
SUMMARY
6
7
8
9
We have demonstrated the unique mechanistic and syn-
thetic aspects of arene acylation via sequential C–O/C–H ac-
tivation. Observed KIEs and regioselectivity support that
this mechanism is distinct from Friedel-Crafts acylation and
offers a pathway to substitution patterns previously inac-
cessible in a single synthetic step. Steric-controlled acyla-
tion is achieved via a proposed concerted metalation depro-
tonation C–H activation, as supported by experimental and
computational evidence. Simple, electron rich arenes are
the most successful and a broad range of functional groups
are tolerated on the salicylate ester. Steric-controlled acyla-
tion was utilized to concisely synthesize an anti-cancer tar-
get molecule hydroxyphenstatin. Experimental efforts are
underway to study the reverse reaction, change the direct-
ing group, lower the temperature and improve reaction
yields with stoichiometric quantities of arene. The reaction
is a first breakthrough towards forming new Cacyl–Caryl
bonds under steric control in a single synthetic step using
an iridium catalyst.
Figure 6. Total synthesis of hydroxyphenstatin 5. Reagents
and conditions: (a) POCl5 (1.5 equiv), phenol (20 equiv),
100 °C, 20 hr, 75%. (b) [Ir(cod)OMe]2 (3 × 1 mol%), Trix-
iePhos (3 × 3 mol%), and 1,5-COD (1 equiv) in 1,2,3-tri-
methoxybenzene (0.1 M), 170 °C, 60 h, 48%. The second
and third charges of catalyst were added after 20 and 40 h
point.
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
functional group R. Salicylates with methyl, methoxy, hy-
droxy, fluoro, and trifluoromethyl substituents all gave
products in good yields (67–97%).
Hydroxyl groups were tolerated to varying degrees depend-
ing on substitution pattern; product 2ce was isolated in
67% yield but 2cm was not isolable due to instability on sil-
ica and only observed in 23% yield (quantitative NMR).
Likewise, acylation of m-xylene yielded a trace amount of
potential product (2bm), which could not be quantified.
Chloro and nitro substituents were similarly tolerated de-
pending on substitution pattern (2cf, 2ci, 2ck, 2cl). Phenyl
3-methylsalicylate gave product 2cj in 88% yield, an inter-
mediate yield relative to the other methyl-substituted salic-
ylates, 2ca and 2cg. This implies that steric hindrance
around the hydroxyl directing group does not have a strong
effect on reactivity. Amino- or bromo-substituted salicy-
lates did not acylate m-xylene or 1,2-dimethoxybenzene
(Figure 5, 1n–1p); for the amino-substituted salicylate 1p
this is likely due to insolubility of the starting material in the
reaction conditions.
ASSOCIATED CONTENT
Supporting Information.
Experimental Details for the reaction optimization, initial
rates kinetics and kinetic isotope effect data, preparation of
new compounds, tabulated characterization data (1H NMR, 13
NMR, melting points, MS, and IR). This material is available
free of charge via the internet at http://pubs.acs.org.
C
AUTHOR INFORMATION
Corresponding Author
*cdouglas@umn.edu
A noted drawback of this reaction and related C–H boryla-
tion/silylation reactions is that a large excess of arene is
used, typically in solvent quantities. Under standard condi-
tions, acylation with 78 equivalents of 1,2-dimethoxyben-
zene yields 2ag in 93% yield. Acylation reactions using su-
perstoichiometric amounts of 1,2-dimethoxybenzene in
mesitylene showed formation of product 2ag, yielding 30%
product with 5 equivalents of arene. Similarly, acylation
was observed in 52% yield with 10 equivalents of arene (for
additional data, see supplementary information).
ORCID
Christopher J. Douglas: 0000-0002-1904-6135
Constance A. Anderson: 0000-0002-0321-4640
Nicholas A. Serratore: 0000-0002-2602-425X
Grant B. Frost: 0000-0002-7248-1806
Truong-Giang Hoang: 0000-0003-4368-4142
Steven J. Underwood: 0000-0002-2794-7590

With this understanding of the scope of this reaction, we un-
dertook a synthesis of hydroxyphenstatin 4, a potent anti-
cancer and anti-mitotic target (Figure 6).69 The shortest
previous route to synthesize 4 employs a low-yielding, or-
tho- selective Friedel-Crafts acylation of 1,2,3-trihy-
droxybenzene (25% yield).70 Our new acylation reaction
opens a new disconnection of this target molecule. Di-hy-
droxy acid 3 can be synthesized in one step from literature
procedure.71 Salicylate ester 4 was synthesized from 3 using
phenol and phosphorous oxychloride in 75 % yield. Acyla-
tion of 1,2,3-trimethoxybenzene using salicylate ester 4
Present Addresses
†Department of Chemistry, National University of Singapore, 3
Science Drive 3, Singapore, 117543.
Department of Chemistry, University of Chicago, 5735 South
Ellis Avenue, Chicago, IL, 60637.
§Department of Chemistry, University of California–Berkeley,
CA, 94720.
Author Contributions
‡These authors contributed equally.
Notes
8
ACS Paragon Plus Environment

Page 9 of 10
1
Journal of the American Chemical Society
(28) Dong, Z.; Wang, J.; Ren, Z.; Dong, G. Angew. Chem. Int. Ed. 2015,
54, 12664–12668.
The authors declare no competing financial interests.
(29) Zhou, P.-X.; Ye, Y.-Y.; Liu, C.; Zhao, L.-B.; Hou, J.-Y.; Chen, D.-Q.;
Tang, Q.; Wang, A.-Q.; Zhang, J.-Y.; Huang, Q.-X.; Xu, P.-F.; Liang, Y.-
M. ACS Catal. 2015, 5, 4927−4931.
(30) Mamone, P.; Danoun, G.; Lukas J. Gooßen, L. J. Angew. Chem.
Int. Ed. 2013, 52, 6704–6708.
(31) Hartwig, J. F. Chem. Soc. Rev., 2011, 40, 1992–2002.
(32) Press, L. P.; Kosanovich, A. J.; McCulloch, B. J.; Ozerov, O. V. J.
Am. Chem. Soc. 2016, 138, 9487−9497.
(33) Haines, B. E.; Saito, Y.; Segawa, Y.; Itami, K.; Musaev, D. G. ACS
Catal. 2016, 6, 7536−7546.
(34) Dombray, T.; Werncke, C. G.; Jiang, S.; Grellier, M.; Vendier, L.;
Bontemps, S.; Sortais, J.-B.; Sabo-Etienne, S.; Darcel, C. J. Am. Chem.
Soc. 2015, 137, 4062−4065.
(35) Cheng, C.; Hartwig, J. F. J. Am. Chem. Soc. 2015, 137, 592−595.
(36) Cheng, C.; Hartwig, J. F. Science 2014, 343, 853−857.
(37) Ishiyama, T.; Takagi, J., Hartwig, J. F.; Miyaura, N. Angew. Chem.
Int. Ed. 2002, 41, 3056−3058.
2
3
ACKNOWLEDGMENT
The National Science Foundation (CAREER Award for C.J.D.,
CHE-1151547) funded this work. C.B.A. thanks the National
Science Foundation for a Graduate Research Fellowship (Grant
No 00039202). M.A.H. and P.M.G. thank the National Science
Foundation for support through the Summer Lando/NSF REU
program at the University of Minnesota–Twin Cities (CHE-
1359181). We acknowledge the Minnesota Supercomputing In-
stitute (MSI) at the University of Minnesota for providing re-
sources that contributed to the research results reported
within this paper. URL: http://www.msi.umn.edu. We thank
Dr. Ian Tonks for many helpful discussions. We also thank the
reviewers for critically reading and suggesting substantial im-
provements to an earlier version of this manuscript.
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
(38) Cho, J. Y.; Tse, M. K.; Holmes, D.; Maleczka, R. E. Jr.; Smith, M. R.
III. Science 2002, 295, 305–308.
(39) Brown, A. C.; Gibson, J. J. Chem. Soc., Trans. 1892, 61, 367–369.
(40) Wu, X.-F. Chem. Eur.J 2015, 21, 12252–12265.
(41) Huang, Y.; Zhu R.; Zhao, K.; Gu, Z. Angew.Chem. Int. Ed. 2015,
54, 12669–12672.
(42) Akai S., Peat, A. J.; Buchwald, S. L. J. Am. Chem. Soc. 1998, 120,
9119–9125.
(43) Gautam, P.; Bhanage, B. M. J. Org. Chem. 2015, 80, 7810–7815
(44) Lin, X.-F.; Li, Y.; Li, S.-Y.; Xiao, Z.-K.; Lu, J.-M. Tetrahedron 2012,
68, 5806–5809.
(45) Hoang, G. T.; Pan, Z.; Brethorst, J. T.; Douglas, C. J. J. Org. Chem.
2014, 79, 11383–11394.
(46) m-xylene Friedel-Crafts regioselectivity with benzoyl chlo-
ride: Kusama, H.; Narasaka, K. Bull. Chem. Soc. Jpn. 1995, 68, 2379–
2383.
(47) The configuration of oxidative addition intermediate III was
proposed after comparison with III2 (See Figure 3)
(48) The configuration of TrixiePhos complex V was proposed af-
ter comparison with JohnPhos angalogues VJP1-3 (See Figure 3)
(49) Lapointe, D.; Fagnou, K. Chem. Lett. 2010, 39, 1118–1126.
(50) Davies, D. L.; Macgregor, S. A.; McMullin, C. L. Chem. Rev. 2017,
117, 8649−8709.
(51) Experiments and analysis were performed as detailed in:
Simmons, E. M.; Hartwig J. F. Angew. Chem. Int. Ed. 2012, 51, 3066–
3072. (and references within)
REFERENCES
(1) Johansson Seechurn, C. C. C.; Kitching, M. O.; Colacot, T. J.;
Snieckus, V. Angew. Chem. Int. Ed. 2012, 51, 5062–5085.
(2) He, J.; Wasa, M.; Chan, K. S. L.; Shao, Q.; Yu, J.-Q. Chem. Rev. 2017,
117, 8754−8786.
(3) Hummel, J. R. ; Boerth, J. A.; Ellman, J. A. Chem. Rev. 2017, 117,
9163−9227.
(4) Kim, D.-S.; Park, W.-J.; Jun C.-H. Chem. Rev. 2017, 117,
8977−9015.
(5) Newton, C. G.; Wang, S.-G.; Oliveira, C. C.; Cramer, N. Chem. Rev.
2017, 117, 8908−8976.
(6) Labinger, J. A.; Chem. Rev. 2017, 117, 8483−8496.
(7) Murakami, K.; Yamada, S.; Kaneda, T.; Itami, K. Chem. Rev. 2017,
117, 9302−9332.
(8) Leitch, J. A.; Bhonoah, Y.; Frost, C. G. ACS Catal. 2017, 7,
5618−5627.
(9) Gensch, T.; Hopkinson, M. N.; Glorius, F.; Wencel-Delord J. Chem.
Soc. Rev. 2016, 45, 2900−2936.
(10) Chen, X.; Engle, K. M.; Wang, D.-H.; Yu, J.-Q. Angew. Chem. Int.
Ed. 2009, 48, 5094–5115.
(11) Fumagalli, G.; Stanton, S.; Bower, J. F. Chem. Rev. 2017, 117,
9404−9432.
(12) Chen, P.-H.; Billett, B. A.; Tsukamoto, T.; Dong, G. ACS Catal.
2017, 7, 1340−1360.
(13) Souillart, L.; Cramer, N. Chem. Rev. 2015, 115, 9410−9464.
(14) Chen, F.; Wang, T.;Jiao, N. Chem. Rev. 2014, 114, 8613−8661.
(15) Takise, R.; Muto, K.; Yamaguchi, J. Chem. Soc. Rev.2017, 46,
5864−5888.
(16) Dander, J. E.; Garg, N. K. ACS Catal. 2017, 7, 1413−1423.
(17) Liu, C.; Szostak, M. Chem. Eur. J. 2017, 23, 7157–7173.
(18) Wang, Q.; Su, Y.; Li, L.; Huang, H. Chem. Soc. Rev. 2016, 45,
1257−1272.
(52) For simplification of discussion our KIE data analysis is
presented in context of the phenoxide-assisted 4-member CMD
transition state. A carboxylate-assisted 6-member CMD transition
state would also be consistent with the data.
(53) Sibbald, P. A., Rosewall, C. F., Swartz, R. D., Michael, F. E. J. Am.
Chem. Soc. 2009, 131, 15945–15951.
(54) Churchill, D. G.; Janak, K. E.; Wittenberg, J. S.; Parkin, G. J. Am.
Chem. Soc. 2003, 125, 1403–1420.
(19) Ouyang, K.; Hao, W.; Zhang, W.-X.; Xi, Z. Chem. Rev. 2015, 115,
12045−12090.
(20) Cornella, J.; Zarate, C.; Martin, R. Chem. Soc. Rev. 2014, 43,
(55) Song, Z.; Beak, P. J. Am. Chem. Soc. 1990, 112, 8126–8134.
(56) Adam, W.; Krebs, O.; Orfanopoulos, M.; Stratakis, M.;
Vougioukalakis, G. C. J. Org. Chem. 2003, 68, 2420–2425.
(57) Jones, W. D. Acc. Chem. Res. 2003, 36, 140-146.
(58) Boller, T. M.; Murphy, J. M.; Hapke, M.; Ishiyama, T.; Miyaura,
N.; Hartwig, J. F. J. Am. Chem. Soc. 2005, 127, 14263–14278.
(59) Zhu, H.; Meyer, M. P. Chem. Commun. 2011, 47, 409–411.
(60) Vaska, L.; DiLuzio J. W. J. Am. Chem. Soc. 1961, 83, 2784–2785.
(61) Yu, H. S.; He, X.; Truhlar, D. G. J. Chem. Theory Comput. 2016.
12, 1280–1293.
8081−8097.
(21) Wang, R.; Falck, J. R. RSC Adv. 2014, 4, 1062−1066.
(22) Gooßen, L. J.; Gooßen, K.; Stanciu, C. Angew. Chem. Int. Ed.
2009, 48, 3569–3571.
(23) Yang, Y.; Lan, J.; You, J. Chem. Rev. 2017, 117, 8787−8863.
(24) Wei, Y.; Hu, P.; Zhang, M.; Su, W. Chem. Rev. 2017, 117,
8864−8907.
(25) Nairoukh, Z.; Cormier, M.; Marek I. Nat. Rev. Chem. 2017, 1, 1–
16.
(26) Meng, G.; Szostak, M. ACS Catal. 2017, 7, 7251−7256.
(27) Meng, G.; Szostak, M. Org. Lett. 2016, 18, 796−799.
(62) M06 and M06-L computations are found in the supporting in-
formation
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 10
(63) N-methylpyrrole Friedel-Crafts regioselectivity with benzoyl
chloride: He, F.; Wu, H.; Chen, J.; Su, W. Synth. Commun. 2008, 38,
255–264.
(64) Naphthalene Friedel-Crafts regioselectivity with benzoyl chlo-
ride: Earle M. J.; Hakala, U.; Hardacre, C.; Karkkainen, J.; McAuley,
B. J.; Rooney, D. W.; Seddon, K. R.; Thompson, J. M.; Wähälä, K. Chem.
Commun. 2005, 7, 903–905.
1
2
3
4
5
6
7
8
(65) 1,3-Dimethoxybenzene Friedel-Crafts regioselectivity with
benzoyl chloride: Vekariya, R. H.; Aubé J. Org. Lett. 2016, 18,
3534−3537.
(66) 3-Methylanisole Friedel-Crafts regioselectivity with benzoyl
chloride: Miquel, J. F.; Buu-Hoi, N. G.; Royer, R. J. Chem. Soc. 1955,
3417–3419
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
(67) 1,2,3-Trimethoxybenzene Friedel-Crafts regioselectivity with
benzoyl chloride: Sato, H.; Dan, T.; Onuma, E.; Tanaka, H.; Aoki, B.;
Koga, H. Chem. Pharm. Bull. 1992, 40, 109–116.
(68) Experiments to study this apparent defunctionalization are
underway.
(69) Pettit, G. R.; Grealish, M. P.; Herald, D. L.; Boyd, M. R.; Hamel, E.
Pettit, R. K. J. Med. Chem. 2000, 43, 2731–2737.
(70) Radha, P. Synth. Commun. 2003, 33, 3869–3873.
(71) Yanagisawa, A.; Taga, M.; Atsumi, T.; Nishimura, K.; Ando, K.;
Taguchi, T.; Tsumuki, H.; Chujo, I.; Mohri, S. Org. Process Res. Dev.
2011, 15, 376–381.
Graphical Abstract
10
ACS Paragon Plus Environment