Mechanistic Origins of Regioselectivity in Cobalt-Catalyzed C(sp2)-H Borylation of Benzoate Esters and Arylboronate Esters

Article abstract of DOI:10.1016/j.chempr.2020.11.017

Carbon–hydrogen (C–H) bonds are ubiquitous in organic molecules, and methods for their selective functionalization to more reactive functional groups is a long-standing goal in catalysis, as applied to organic synthesis. Of the established methods involving transition metal catalysts, many employ carefully engineered substrate-catalyst interactions, placing the targeted C–H bond proximal to the metal catalyst, resulting in activation and subsequent functionalization. Here, we report mechanistic investigations describing a conceptual alternative to this approach whereby a cobalt-based borylation catalyst differentiates between subtle electronic differences in C(sp²)-H bonds of benzoate esters and arylboronate esters. These advances motivate studies of catalysts that rely on inherent differences in C–H bond electronics to distinguish chemically inequivalent sites, providing a new tool for organic synthesis. Synthetic and mechanistic investigations into the C(sp²)-H borylation of various electronically diverse arenes catalyzed by bis(phosphine)pyridine (^iPrPNP) cobalt complexes are reported. Borylation of various benzoate esters and arylboronate esters gave remarkably high selectivities for the position para to the functional group; in both cases, this regioselectivity was found to override the ortho-to-fluorine regioselectivity, previously reported for (^iPrPNP)Co borylation catalysts, which arises from thermodynamic control of C(sp²)-H oxidative addition. Mechanistic studies support pathways that result in para-to-ester and para-to-boronate ester selectivity by kinetic control of B-H and C(sp²-H) oxidative addition, respectively. Borylation of a particularly electron-deficient fluorinated arylboronate ester resulted in acceleration of C(sp²)-H oxidative addition and concomitant inversion of regioselectivity, demonstrating that subtle changes in the relative rates of individual steps of the catalytic cycle can enable unique and switchable site selectivities. Most strategies to selectively activate and functionalize a specific C–H bond in an organic molecule rely upon carefully engineered spatial interactions between a substrate and a transition metal catalyst. Here, we report a conceptually distinct alternative strategy, whereby a cobalt catalyst distinguishes between subtly different C(sp²)-H sites of an arene based on electronics rather than sterics. Mechanistic studies elucidated the origins of substrate-controlled regioselectivity in the C(sp²)-H borylation of benzoate esters and arylboronate esters.

Full text of DOI:10.1016/j.chempr.2020.11.017

ll
Article
Mechanistic Origins of Regioselectivity in
Cobalt-Catalyzed C(sp2)-H Borylation of
Benzoate Esters and Arylboronate Esters
Tyler P. Pabst, Linda Quach,
Kaitlyn T. MacMillan, Paul J.
Chirik
pchirik@princeton.edu
HIGHLIGHTS
Electronically derived
regioselectivity in cobalt-
catalyzed C(sp2)-H borylation
Reversible C–H oxidative addition
of aryl esters in a Curtin-Hammett
regime
Arylboronate borylation under
kinetic control of C(sp2)-H
oxidative addition
Sufficiently electron-deficient
arylboronates undergo reversible
oxidative addition
Most strategies to selectively activate and functionalize a specific C–H bond in an
organic molecule rely upon carefully engineered spatial interactions between a
substrate and a transition metal catalyst. Here, we report a conceptually distinct
alternative strategy, whereby a cobalt catalyst distinguishes between subtly
different C(sp2)-H sites of an arene based on electronics rather than sterics.
Mechanistic studies elucidated the origins of substrate-controlled regioselectivity
in the C(sp2)-H borylation of benzoate esters and arylboronate esters.
Pabst et al., Chem 7, 237–254
January 14, 2021 ª 2020 Elsevier Inc.
https://doi.org/10.1016/j.chempr.2020.11.017

ll
Article
Mechanistic Origins of Regioselectivity
in Cobalt-Catalyzed C(sp2)-H Borylation
of Benzoate Esters and Arylboronate Esters
Tyler P. Pabst,1 Linda Quach,1 Kaitlyn T. MacMillan,1 and Paul J. Chirik1,2,
*
SUMMARY
The Bigger Picture
Synthetic and mechanistic investigations into the C(sp2)-H boryla-
tion of various electronically diverse arenes catalyzed by bis(phos-
phine)pyridine (iPrPNP) cobalt complexes are reported. Borylation
of various benzoate esters and arylboronate esters gave remark-
ably high selectivities for the position para to the functional group;
in both cases, this regioselectivity was found to override the ortho-
to-fluorine regioselectivity, previously reported for (iPrPNP)Co
borylation catalysts, which arises from thermodynamic control of
C(sp2)-H oxidative addition. Mechanistic studies support pathways
that result in para-to-ester and para-to-boronate ester selectivity
by kinetic control of B-H and C(sp2-H) oxidative addition, respec-
tively. Borylation of a particularly electron-deficient fluorinated
arylboronate ester resulted in acceleration of C(sp2)-H oxidative
addition and concomitant inversion of regioselectivity, demon-
strating that subtle changes in the relative rates of individual steps
of the catalytic cycle can enable unique and switchable site
selectivities.
Carbon–hydrogen (C–H) bonds
are ubiquitous in organic
molecules, and methods for their
selective functionalization to more
reactive functional groups is a
long-standing goal in catalysis, as
applied to organic synthesis. Of
the established methods involving
transition metal catalysts, many
employ carefully engineered
substrate-catalyst interactions,
placing the targeted C–H bond
proximal to the metal catalyst,
resulting in activation and
subsequent functionalization.
Here, we report mechanistic
investigations describing a
conceptual alternative to this
approach whereby a cobalt-based
borylation catalyst differentiates
between subtle electronic
differences in C(sp2)-H bonds of
benzoate esters and arylboronate
esters. These advances motivate
studies of catalysts that rely on
inherent differences in C–H bond
electronics to distinguish
INTRODUCTION
The site-selective functionalization of C(sp2)-H bonds is a powerful approach for the
construction and diversification of small molecules.1–5 Typically, site selectivity in
metal-catalyzed C–H functionalization reactions is achieved with directing groups
that form discrete metal-ligand bonds during the catalytic cycle and distinguish
the various C–H bonds by proximity. While widely applied, this approach relies on
specific, predisposed functional groups, limiting its synthetic utility. An alternative
strategy exploiting electronically controlled regioselectivity, whereby catalysts
distinguish between possible sites of activation by their electronic properties, is
appealing due to its potentially broad utility and reliance on the inherent properties
of the substrate. In practice, however, such a relationship between site electronics
and regioselectivity in metal-catalyzed C–H functionalization has neither been
demonstrated nor rigorously understood.
chemically inequivalent sites,
providing a new tool for organic
synthesis.
Of the numerous known transformations for the functionalization of C–H bonds, bor-
ylation is among the most attractive due to the number of synthetic applications of
the readily handled arylboronate ester products.6,7 Catalytic C(sp2)-H borylations
typically employ iridium precatalysts such as [Ir(COD)OMe]2 (COD = 1,5-cycloocta-
diene) in combination with 4,40-di-tert-butyl-2,20-bipyridine (dtbpy) or a related
substituted bipyridine ligand. Using either B2Pin2 or HBPin (Pin = pinacolato) as
the boron source, high turnover numbers are achieved in many cases, and
Chem 7, 237–254, January 14, 2021 ª 2020 Elsevier Inc. 237

ll
Article
remarkable chemoselectivity for C(sp2)-H borylation is observed even in the pres-
ence of more polar carbon-halogen bonds.8,9
Selectivity in iridium-catalyzed C(sp2)-H, borylation is principally sterically driven.8–10 Ar-
ene C–H bonds ortho to bulky (CH3 or larger) substituents are inaccessible to the C–H
activating Ir-tris(boryl) intermediate, and no or minimal borylation occurs at these sites;
the remaining, accessible C(sp2)-H bonds are typically functionalized to generate
approximately statistical mixtures of products (e.g., toluene affords a 2:1 mixture of
meta- and para-borylated products, Scheme 1A). Experimental and computational
studies have provided insight into theoriginof the sterically derived regioselectivity.11,12
For all substrates examined experimentally, irreversible C(sp2)-H oxidative addition to Ir-
tris(boryl) has been invoked as the turnover-limiting step. Density functional theory cal-
culations of the oxidative addition revealed late transition states with fully formed
iridium–carbon bonds, and a distortion/interaction model was applied in an attempt
to rationalize the observed regioselectivity of the borylation of various arenes.13 The
transition state interaction energy between the iridium catalyst and the substrate at
each sterically accessible C(sp2)-H bond served as the most effective predictor of the
experimentally observed selectivities. This model was constructed using the assumption
that C(sp2)-H oxidative addition was turnover limiting for all substrates examined, and
that it proved to be an effective predictor of regioselectivity suggests that the mecha-
nism of iridium-catalyzed borylation is consistent over a wide range of arenes despite dif-
ferences in substrate electronic properties.
Established strategies to bias precious metal borylation catalysts away from statisti-
cal, sterically derived regioselectivity depends upon carefully engineered steric or
electrostatic interactions. Numerous Lewis basic directing groups have been em-
ployed to affect ortho-directed borylation.14–22 More elaborate approaches
involving secondary interactions between catalyst and substrate have been required
to achieve meta-selective borylation; such interactions involve hydrogen bonding,23
Lewis acid-base pairing,24 and ion pairing.25–30 Selective borylation at the site para
to a substituent has proven particularly challenging, with limited strategies for this
outcome (Scheme 1B) including the use of exceptionally large phosphine ligands
or Lewis acid co-catalysts.31–33 An example of para-selective borylation of benzoate
esters employed a unique quinolyl-substituted bipyridine ligand engaged in a sec-
ondary-substrate interaction involving a bridging alkali metal.34 Recently, the
laboratories of Smith and Phipps independently introduced a strategy utilizing tight
substrate anion-ammonium ion pairs as a means of blocking most of the sterically
accessible C(sp2)-H bonds on the substrate to favor borylation para to a given
substituent.35,36
In comparison to the strategies that rely upon steric or electrostatic interactions, re-
gioselective borylation methodologies resulting from inherent oxidative addition
preferences of the metal catalyst at a given site, e.g., electronically driven site selec-
tivity, are exceedingly rare. It has been demonstrated that the iridium/bipyridine cat-
alysts fail to distinguish between the electronically distinct C(sp2)-H bonds in
3-substituted fluoroarenes and provide approximately statistical product distribu-
tions in borylation of these substrates.37 Platinum catalysts supported by either
N-heterocyclic carbenes or a [PSiN] pincer ligand were shown to borylate fluorinated
arenes with improved regioselectivity for the position ortho to the fluorine substitu-
ent albeit under harsh conditions and by an unknown mechanism.38,39
1Department of Chemistry, Princeton University,
Princeton, Princeton, NJ 08544, USA
2Lead Contact
Our laboratory has reported that bis(phosphino)pyridine (iPrPNP) cobalt complexes
are active catalysts for the C(sp2)-H borylation of various arenes and
*Correspondence: pchirik@princeton.edu
https://doi.org/10.1016/j.chempr.2020.11.017
238 Chem 7, 237–254, January 14, 2021

ll
Article
A
tBu
tBu
Me
Me
Me
[Ir(COD)Cl]2
dtbpy
N
N
B2Pin2
Ir
80 °C, 16 h
- HBPin
PinB
BPin
BPin
BPin
BPin
C–H Activating Species
sterically accessible
2:1 product ratio (statistical)
B
Me
Me
Si
Me
Me
Me
tBu
tBu
Me
O
O
O
O
Al
O
O
S
tBu
NR2
N
Me
tBu
Me
O
OR
O
Me
Me
O
N
K
Me
Me
PinB
BPin
P
P
Me
Me
Me
[Ir]
N
N
N
N
N
N
Ir
Me
Ir
Ir
PinB
BPin
PinB
BPin
BPin
PinB
BPin
BPin
BPin
Me
Me
Segawa, Itami (2015)
Nakao (2017)
Chattopadhyay (2017)
Phipps, Smith (2019)
C
O
OCH3
H3
C
PiPr2
H
R
R
N
B2Pin2
Co
=
P
BPin
H
O
O
BPin
HN
NH
B
B
5 mol % [Co]
THF, 23 °C, 24 h
- HBPin
Scheme 1. The Selectivity of Metal-Catalyzed C(sp2)-H Borylation in the Absence of Traditional
Directing Groups
(A) Statistical selectivity in the borylation of toluene, representative of selectivities often observed
in standard iridium-catalyzed C(sp2)-H borylation reactions.
(B) Known strategies to affect para-selective C(sp2)-H borylation using specialized catalytic
systems.
(C) Substrate-controlled para selective cobalt-catalyzed C(sp2)-H borylation, where the selectivity
responds to substrate electronics (this work).
heterocycles;40–42 moreover, these catalysts have demonstrated high regioselectiv-
ity for borylation ortho-to-fluorine substituents.37 A comprehensive mechanistic
study established that fluoroarene borylation by the cobalt catalysts occurs by a
mechanism in which C(sp2)-H oxidative addition to a Co(I)-boryl species is fast and
reversible.43 Oxidative addition of the arene substrate is, therefore, under thermo-
dynamic control; metalation at the ortho site is favored due to the thermodynamic
ortho fluorine effect, whereby metal–aryl complexes with ortho fluorine substituents
form stronger M–C bonds.44–46 The ability of cobalt catalysts to distinguish subtle
electronic differences in arene C(sp2)-H bonds motivated studies beyond fluorine
substituents. An improved understanding of electronically driven site selectivity
and its mechanistic origins would enable new synthetically useful, site-selective bor-
ylation methods that rely on the inherent properties of the substrate rather than pre-
installed directing groups or statistical, sterically driven selectivity.
Here we describe the discovery and mechanistic interrogation of cobalt-catalyzed
borylation reactions wherein regioselectivity outcomes were found to be dictated
by the electronic properties of the arene substrate (Scheme 1C). The presence of
an ester or boronate ester functional group resulted in a strong preference for bor-
ylation at the position para to this substituent subsubstiutent a selectivity which
overrode the previously observed thermodynamic ortho-to-fluorine regioselectivity.
This phenomenon was then leveraged toward the synthesis of para-disubstituted
arenes containing ester and boronate functional groups or two chemically distinct
boronate substituents. Mechanistic investigations revealed that benzoate ester
Chem 7, 237–254, January 14, 2021 239

ll
Article
substrates undergo catalytic C(sp2)–H borylation by a mechanism involving fast
C(sp2)–H oxidative addition and which is similar to that which has been determined
to result in ortho to fluorine regioselectivity for fluorinated (non-ester) arenes; how-
ever, the regioselectivity in the case of benzoate ester borylation appears to arise
from a kinetic phenomenon rather than metal–carbon and carbon–hydrogen bond
thermodynamics.In contrast, mechanistic studies of arylboronate ester borylation re-
vealed a distinct mechanism in which C(sp2)-H oxidative addition is turnover limiting
and likely under kinetic control. Subtle substrate modification resulted in accelera-
tion of C(sp2)-H oxidative addition and inversion of regioselectivity to result in selec-
tive borylation ortho to a fluorine substituent rather than para to the boronate sub-
stituent, as a result of thermodynamic control, establishing the relationship between
reaction regioselectivity and the identity of the operative mechanism.
RESULTS AND DISCUSSION
Borylation of Electronically Diverse Arenes
The unique, electronically derived site selectivity observed in fluoroarene borylation with
[(iPrPNP)Co] catalysts inspired evaluation of the influences of various substituents on the
regioselectivity of the reaction. Studies commenced with the borylation of various mono-
substituted difluorinated arenes using the air-stable cobalt precatalyst37 1-(OPiv)2
(Scheme 2). These substrates were selected for their high activity toward CꢀH borylation
as well as the ease of product characterization by 19F NMR spectroscopy. Most impor-
tantly, this class of substrates allowed for rapid assessment of ring substituent effects in
the site selectivity of the reactions. Borylation occurred efficiently at room temperature
over the course of 24 h in THF with B2Pin2 as the borylating agent and 5 mol %
1-(OPiv)2 with20mol%ofHBPinastheactivator.Ineachcase, highyieldsoftheborylated
product(s) were obtained. While the borylation of 2,5-difluorotoluene (2a) proceeded
with marginal selectivity for the position para to the methyl substituent, the analogous tri-
fluoromethyl-substituted compound 2b was functionalized with 73:27 selectivity for the
para product, illuminating the influence of the [CF3] substituent on the site preference
for the reaction. Anisole (2c) and the N-acetyl-protected aniline (2d) derivatives were
borylated primarily at the meta position, though the selectivity for the meta product
was greater (85:15 versus 60:40) for the former. Ester substituents were found to enforce
a significant preference for para-selective functionalization, as the reaction of a 2,5-di-
fluorobenzoate ester (2e) afforded a 96:4 mixture of products favoring that arising from
borylation para to the ester group. Strikingly, borylation of the 2,6-difluorinated isomer
(2f) of this substrate likewise afforded predominantly the para-to-ester functionalized
product in an 80:20 ratio, demonstrating that the para influence of the ester overrode
the previously described ortho-to-fluorine site selectivity with (iPrPNP)Co catalysts. Bory-
lation of trimethylsilyl-substituted 2g resulted in a 65:35 ratio of the para- and meta-func-
tionalized products, respectively. Notably, when 2,5-difluorophenyl pinacolboronate
ester 2h was subjected to the reaction conditions, the para-diborylated product was ob-
tained with almost exclusive selectivity. The reactionof the analogous [BDan]-substituted
(Dan=1,8-diaminonaphthyl)arene2ilikewiseaffordedpredominantlythemixedpara-di-
borylarene product, albeit with slightly diminished (93:7) selectivity. Attempts to borylate
the MIDA (N-methylimidodiacetic boronic acid ester) and trifluoroborate derivatives of
2h and 2j resulted in no detectable amounts of borylation products, likely due to unfavor-
able solubility properties and unwanted side reactivity, respectively.
The observation of high para selectivity for the borylation of arylboronate esters 2h
and 2i and the apparent large size of the boronate ester substituents raised ques-
tions as to the origin of this unique site selectivity. To decouple the influences of ste-
ric and electronic factors on the high selectivities of these reactions, an
240 Chem 7, 237–254, January 14, 2021

ll
Article
O
tBu
H3C
R
R
R
PiPr2
O
O
N
F
F
F
B2Pin2
Co
O
F
F
F
P
BPin
tBu
BPin
3
1-(OPiv)2
2
3'
5 mol % [Co]
20 mol % HBPin
THF, 23 °C, 24 h
- HBPin
CH3
CF3
OMe
NHAc
F
F
F
F
43
27
85
60
F
F
F
F
F
F
57
73
15
40
2a
2b
2c
2db,c
O
O
B
CO2Me
F
SiMe3
CO2Me
F
F
F
F
4
35
20
1
F
F
96
80
65
99
2e
2f
2gb
2h
O
O
O
O
HN
NH
HN
NH
B
B
B
F
F
F
F
F
7
31
9
8
F
F
93
69
91
92
2i
2j
2k
2l
Scheme 2. Borylation of Various Substituted Difluoroarenes with B2Pin2 Using 1-(OPiv)2 as
Precatalyst
aThe values in the table are the ratios of arylboronate products resulting from borylation of the
substrate drawn. All reactions reached >90% conversion unless otherwise noted.
bReaction conducted at 60ꢁC.
cReaction reached 60% conversion.
approximately isosteric acetal-substituted arene (2j) was prepared and borylated un-
der standard catalytic conditions. While 2h was borylated preferentially at the para
position, the modest 69:31 selectivity observed was significantly less than the near-
exclusive para selectivity observed for the reaction of 2h. The superior para selec-
tivity with 2h (99:1 versus 69:31) indicates that, while sterics may influence the selec-
tivity of the reaction to some extent, the electronics of the boryl substituent are
crucial to the high regioselectivity.
The borylation of 2,6-difluoroaryl boronate esters was pursued to determine
whether the observed para directing effect of the boronate ester persisted in the
presence of different C(sp2)-H bonds ortho-to-fluorine substituents. Cobalt-cata-
lyzed borylation of 2,6-difluorophenylBPin (2k) with 1 equiv of B2Pin2 afforded pre-
dominantly the para-borylated product with a 91:9 selectivity over the meta isomer.
Chem 7, 237–254, January 14, 2021 241

ll
Article
Remarkably, the borylation proceeded with >9:1 selectivity for functionalization
para to the initial BPin substituent despite the presence of sterically accessible
C–H bonds ortho-to-fluorine substituents. Likewise, the reaction of [BDan]-
substituted 2l proceeded with 92:8 selectivity in favor of borylation at the para
position.
Scope of Benzoate Ester and Arylboronate Ester Borylation
With the rare observation of para selectivity in C–H functionalization without the aid
of steric directing groups, cobalt-catalyzed C(sp2)-H borylation of various benzoate
esters and arylboronate esters was explored. Notably, the observed para selectivity
is orthogonal to that expected from classical electrophilic aromatic substitution re-
actions, and such methods are reported to yield mixtures of regioisomers favoring
the meta-functionalized product when either benzoate esters or arylboronic acids
are employed as the substrate.47 Given the motivation to explore electronically
derived regiocontrol in C(sp2)-H activation reactions, only arene substrates with mul-
tiple sterically accessible but electronically differentiated C(sp2)-H bonds were
selected for these studies. Namely, all substrates examined have at least two distinct
C(sp2)-H bonds that are not ortho to a sterically demanding substituent (CH3 or
larger). Using 5 mol % of the well-defined dihydride boryl precatalyst 1-(H)2BPin
and 1 equiv of B2Pin2, various substituted benzoate ester substrates were borylated
in excellent yield to selectively generate the corresponding 4-borylbenzoate esters
(Scheme 3). Two 2,5-difluorinated benzoate esters bearing methyl and tert-butyl
substituents on the O-bound portion of the ester were subjected to the reaction con-
ditions to afford the corresponding para-borylated products 3e and 3m with essen-
tially identical regioselectivities, indicating that the steric profile at this position has
little impact upon the outcome. Borylation of a trifluorinated ester afforded predom-
inantly the para-functionalized product 3n in a 92:8 ratio over the minor, meta-bory-
lated product, while a 2,3-difluorinated substrate gave exclusively the para-bory-
lated product 3o. In two cases where para-to-ester selectivity proved to override
the previously described ortho-to-fluorine selectivity, 3f and 3p were obtained
from the reaction of the corresponding benzoate esters. Borylation of these two sub-
strates occurred with approximately 80:20 regioselectivity in favor of the product
arising from C(sp2)-H borylation of the position para to the ester substituent, with
the selectivity for 3f improving to 95:5 upon purification by flash column chromatog-
raphy. Borylation of 2-trifluoromethylbenzoate methyl ester (2q) afforded 3q and 3q0
in a 70:30 ratio.
The synthesis of 1,4-diborylsubstituted arenes with one [BPin] and one [BDan] sub-
stituent was also pursued due to the well-demonstrated differential reactivity of the
two boron substituents, for example, in Suzuki-Miyaura cross-coupling.48–50 Under
identical reaction conditions to those employed for the borylation of benzoate es-
ters, a selection of [BDan]-substituted arenes were borylated to generate the corre-
sponding mixed diborylarenes (Scheme 3).
Borylation of arylboronate ester substrates with multiple aryl fluoride substituents
proceeded efficiently at room temperature and with high regioselectivity for boryla-
tion para to the [BDan] substituent. The products of these reactions were readily iso-
lated by filtration of the crude reaction mixtures through a thin pad of silica followed
by solvent removal and washing the resulting solid with a portion of cold pentane.
The products arising from borylation of 2,5-difluorophenylBDan were isolated in
89% yield and with a 98:2 ratio of the para- and meta-borylated products 3i and
3i0. Borylation of 2,6-difluorophenylBDan(2l) and 2,3,6-trifluorophenylBDan(2r) on
a larger than one-gram scale afforded the corresponding products 3l and 3r in
242 Chem 7, 237–254, January 14, 2021

ll
Article
Scheme 3. Scope of the para-Selective C(sp2)-H Borylation of Benzoate Esters and Arylboronate
Esters Catalyzed by 1-(H)2BPin
Scope of the borylation of arylBDan compounds to generate mixed 1,4-diborylarenes.
aAll reactions were run on 0.50 mmol scale at 23ꢁC for 24 h unless otherwise noted. Isolated yields
are shown unless otherwise noted.
b3.5 mmol scale.
c4.0 mmol scale.
d2.0 mmol scale.
eReaction conducted at 60ꢁC.
fYield determined by 1H NMR spectroscopy.
80% yield (96:4 regioselectivity) and 83% yield (>99:1), respectively. Borylation of
2,3-difluorinated 2s and difluorinated anisole derivative 2t also proceeded with
high yield and excellent selectivity to generate 3s and 3t with near exclusivity in
both cases. For comparison to previously established catalytic systems, the boryla-
tion of 2l was also carried out with [Ir(COD)OMe]2 and dtbpy; this reaction produced
a complex mixture of organoboron products due to multiple borylations of the dia-
minonapthyl substituent in addition to the arene ring (see Supplemental Information
for details). This result demonstrates that the superior activity of the iridium catalysts
is detrimental to regioselectivity as an uncontrolled borylation results when multiple
aromatic rings with different electronic properties are present. Iridium-catalyzed
Chem 7, 237–254, January 14, 2021 243

ll
Article
borylation of the analogous BPin-substituted compound (2k) generated para-di-
borylated 3k as the major product, but with a significant amount of the triborylated
product arising from initial borylation meta to the boryl substituent of the substrate.
Considering the high regioselectivity observed in the borylation of 2k by the (PNP)
Co catalyst, this result demonstrates the heightened sensitivity of the cobalt system
to the electronics of the distinct sites of the arene substrate.
Application of the cobalt-catalyzed method to arenes bearing substituents other
than or in addition to fluorine maintained the high regioselectivity for para-dibory-
larenes, though some substrates bearing fewer fluorine substituents displayed lower
activity. Increased temperatures (see Scheme 3) were necessary for synthetically use-
ful yields in many of these cases, and column chromatography was required for isola-
tion of the products (see Supplemental Information for details). Monofluorinated
arylBDan substrates bearing methoxy (2v) and methyl (2w) substituents underwent
catalytic C(sp2)-H borylation with 89:11 regioselectivity or greater for the position
para to [BDan]. An ethoxybenzene derivative (2x) was borylated with 87:13 selec-
tivity favoring the 1,4-diborylarene product 3x despite the presence of a different
C(sp2)-H bond ortho to a fluorine substituent. Borylation of sulfonamide (2y) and
trifluoromethyl-substituted (2z) substrates without aryl fluoride substituents also
proceeded, albeit with reduced regioselectivity and activity. Curiously, the boryla-
tion of 2-fluoro-6-trifluoromethylphenylBDan (2aa) yielded predominantly the
meta-diborylarene product 3aa0. This anomalous regioselectivity will be discussed
in detail later in the text.
While the cobalt-catalyzed C(sp2)-H borylations of many of the substrates discussed
above (see Scheme 3) proceed efficiently and with high regioselectivites, next-gen-
eration catalysts are required to expand the scope and overall activity. Substrates
containing aryl chloride and bromide moieties were found to be unsuitable for bor-
ylation with the cobalt catalysts; this is consistent with previously published results
indicating that (iPrPNP)Co catalysts are intolerant of aryl halide substituents other
than fluoride (see the Supplemental Information for details).37 Other substrates
were found to be incompatible with the method due to insufficient activation,
including PhBDan and its 2-methoxy-substituted derivative. While the successful re-
actions described above demonstrate the electronic sensitivity of the cobalt cata-
lysts that enable rare and desirable regioselectivities, further development of cobalt
catalysts with superior activity and functional group tolerance is necessary in order to
extend the utility of this method to the borylation of more electron-rich arene sub-
strates as well as those with more diverse functionality.
Mechanistic Investigations of Benzoate Ester Borylation
The high para-to-functional group regioselectivity observed in borylation of select
benzoate esters and arylboronates raised questions about the origins of this rare
regioselectivity and the mechanism operative in the catalytic reactions for each
substrate class. For all mechanistic studies, the well-defined cobalt precatalyst,
1-(H)2BPin was used to eliminate any complications associated with activation of
1-(OPiv)2.51 First, substrate 2f was used for investigations into benzoate ester bory-
lation because of the orthogonal selectivity to that observed from the ortho-to-fluo-
rine cases reported previously.37 Variable time normalization analysis (VTNA), as
described by Bure´ s, was applied to determine the overall rate law of this reaction us-
ing trials with varying initial concentrations of 1-(H)2BPin, arene 2f, and B2Pin2
(Scheme 4A).52 Using this procedure, the rate law described by Equation 1 was
obtained.
244 Chem 7, 237–254, January 14, 2021

ll
Article
A
CO2Me
CO2Me
CO2Me
F
F
F
F
F
F
2.5 - 10 mol% 1-(H)2BPin
B2Pin2
THF, 23 °C, time
- HBPin
BPin
BPin
3f
2f
3f'
Rate = k[1-(H)2BPin]1[2f]0[B2Pin2]0
80
20
B
C
CO2tBu
CO2tBu
H3C
F
F
F
F
PiPr2
5 mol % 1-(H)2BPin
B2Pin2
N
THF, 23 °C, time
- HBPin
Co
F
P
BPin
3ab
CO2Me
2ab
F
kH
para-4f (and isomer)
= 3.3(1)
kD
(<1 half-life)
(23 °C)
H3C
CO2tBu
CO2tBu
H
Co
H
PiPr2
BPin
N
F
F
F
F
5 mol % 1-(H)2BPin
B2Pin2
P
THF, 23 °C, time
- DBPin
D
D
D
D
D
BPin
1-(H)2BPin
(>1 half-life)
2ab-d3
3ab-d2
D
CO2tBu
CO2tBu
F
F
F
F
5 mol % 1-(H)2BPin
3ab
3ab-d2
B2Pin2
THF, 23 °C, time
D
D
- HBPin
- DBPin
(and isomers)
D
2ab-d3
kH
2ab
= 1.5
kD
(23 °C)
CO2tBu
CO2tBu
F
F
F
F
5 mol % 1-(H)2BPin
3ab
3ab-d2
B2Pin2
HBPin
THF, 23 °C, time
(5.0 equiv.)
D
D
- HBPin
- DBPin
(and isomers)
D
kH
2ab-d3
2ab
= 1.1
kD
(23 °C)
Scheme 4. Mechanistic Investigations for the 1-(H)2BPin-Catalyzed C(sp2)-H Borylation of
Benzoate Esters
(A) Borylation of 2f with 1-(H)2BPin and the experimentally determined rate law.
(B) Parallel deuterium kinetic isotope effect experiments for the borylation of 2ab.
(C) Cobalt resting states were observed in situ by 19F and 31P NMR spectroscopy for the borylation
of 2ab.
(D) Competition kinetic isotope effect experiments for the borylation of 2ab.
Rate = k[1-(H)2BPin]1[2f]0[B2Pin2]0
(Equation 1)
As in our previous studies of fluoroarene borylation catalyzed by (iPrPNP)Co com-
plexes, the order in arene substrate was determined to be zero, indicating saturation
behavior. This result is consistent with a reaction mechanism whereby C(sp2)-H
oxidative addition is fast relative to subsequent steps of the catalytic cycle. In light
of the orthogonal regioselectivity of the borylation of 2f compared to previous fluo-
rinated arenes examined, further mechanistic experiments were pursued. For
Chem 7, 237–254, January 14, 2021 245

ll
Article
deuterium kinetic isotope effect (KIE) studies, 2,6-difluorobenzoate tert-butyl ester
(2ab) and its deuterated variant 2ab-d3 were employed, as the hydrogen isotope ex-
change procedure employed for the synthesis of the deuterated substrate does not
tolerate methyl esters.53 Upon measurement of the initial rates of 1-(H)2BPin-cata-
lyzed C(sp2)-H borylation of 2ab and 2ab-d3 in the presence of 1.75 equiv of
B2Pin2 in two separate vessels, a value of 3.3(1) was determined for the ratio of
kH/kD at 23ꢁC (Scheme 4B). Notably, previous mechanistic investigations on the
ortho-selective borylation of fluorinated arenes gave kH/kD values close to unity,
suggesting that C–H bond cleavage did not occur during the rate-determining
step of these reactions.43 The proposed origins of the observed KIEs will be dis-
cussed at length later in this section.
Monitoring the catalytic borylation in situ by 19F and 31P NMR spectroscopy re-
vealed the presence of two Co(I)-aryl species accessible by C(sp2)-H oxidative
addition to 2f followed by reductive elimination of HBPin. In contrast to previous
studies on fluoroarene borylation, where only the ortho-fluorinated Co(I)-aryl spe-
cies was observed spectroscopically during the catalytic reaction, the two species
para- and meta-4f were present in an approximately 2:1 ratio.43 A significant
amount of the Co(I)-boryl species 1-BPin41 was simultaneously observed by 31P
NMR spectroscopy. These three species were observed at the beginning of the
time course, but all three decreased in relative concentration as the reaction pro-
gressed, and 1-(H)2BPin became the only Co species observed in significant
amounts as conversion increased. This behavior has been described previously
and is attributed to the increasing concentration of HBPin, the stoichiometric by-
product of the reaction, which pulls the cobalt catalyst off-cycle when present in
large concentration.43
Temperature effects for the borylation of 2f were also investigated (see the Supple-
mental Information for details). In contrast to the previous mechanistic studies with
fluorinated arenes, where regioselectivity remained constant independent of the
temperature of the reaction,37 the preference for borylation of 2f at the para position
eroded slightly but significantly at higher temperatures and improved to 83:17 when
the reaction was performed at 4ꢁC.
The relatively large, normal deuterium KIE determined for the borylation of 2f is in
contrast to the near-unity values previously obtained for ortho-to-fluorine selective
borylation using the same catalyst; additionally, this data might appear to conflict
with the rate law determined for this reaction, which indicates that the rate of the
transformation is independent of the initial concentration of 2ab. This zeroth-order
in 2ab combined with the spectroscopic identification of Co(I)-aryl species para-
and meta-4f strongly suggest that the C(sp2)-H activation step precedes the turn-
over-limiting step of this catalytic reaction. Another possibility that reconciles all
available data is that the apparent C–H/C–D KIE arises instead from the rate-deter-
mining cleavage of a B–H (or B–D) bond. Indeed, the mechanism that we have pre-
viously proposed for fluoroarene borylation catalyzed by 1-(H)2Bpin involves the re-
action of the Co(I)-aryl species with HBPin to generate a Co(III) intermediate, which
then undergoes reductive elimination of the arylboronate product. When 2f-d3 is
employed as the substrate, DBPin (rather than HBPin) is expected to undergo
oxidative addition with the Co(I)-aryl species; as such, the kH/kD value of 3.3(1)
observed when reactions were performed in separate vessels (vide supra) is not
necessarily derived from the relative rates of C–H and C–D bond cleavage. To
assess this possibility, a competition KIE experiment was performed; notably, par-
allel and competition KIE experiments can yield different values for the same
246 Chem 7, 237–254, January 14, 2021

ll
Article
catalytic reaction, offering valuable insights into the operative mechanism.54 Cata-
lytic borylation at 23ꢁC of an equimolar mixture of 2ab and 2ab-d3 in the presence
of 1 equiv of B2Pin2 as well as 5 mol % of 1-(H)2BPin (Scheme 4D, top) afforded a
value of 1.5 for kH/kD. The erosion of this value compared with that observed in the
parallel KIE experiments may be attributed to the presence of both HBPin and
DBPin in the reaction mixture, as each is proposed to arise from the initial C–H
(or C–D) oxidative addition of 2ab or 2ab-d3 to 1-BPin followed by reductive elim-
ination of the corresponding borane. To further investigate the possibility that the
observed KIEs are associated with a mechanism involving slow B–H oxidative addi-
tion, a similar competition experiment was performed in the presence of 5.0 equiv
of HBPin. Under these conditions, the concentration of HBPin in the
reaction mixture is much higher than that of DBPin at any given time, and no
rate difference should be expected for the formation of 3ab versus 3ab-d2; in
this experiment, a kH/kD value of 1.1 was determined, demonstrating essentially
complete erosion of the preference for borylation of 2ab over 2ab-d3 and providing
additional support for a B–H bond cleavage event as the slow step of the catalytic
reaction.
Given all of the available experimental data, a mechanism for the 1-(H)2BPin-catalyzed
C(sp2)-H borylation of benzoate esters is proposed wherein the fundamental steps
resemble those previously implicated in the ortho-selectivity borylation of fluorinated
arenes, but where differences in the rates of these steps relative to one another result
in selective functionalization para to the ester substituent. The cobalt precatalyst 1-
(H)2BPin eliminates H2 by a multi-step sequence55 to generate the C–H activating boryl
species, 1-BPin. The species then undergoes C(sp2)-H oxidative addition to generate a
Co(III)-aryl–hydride–boryl intermediate with mutually cis boryl and hydride ligands, and
this species then reductively eliminates HBPin to generate Co(I)-aryl species para- and
meta-4f. In contrast to previous studies43 in which only one isomer of the Co(I)-aryl spe-
cies of (iPrPNP) was observed in situ, isomers metalated at both the major and minor sites
of functionalization were observed, indicating that both of these species are kinetically
accessible during catalysis and are generated in comparable amounts. As suggested
by the KIE experiments, subsequent addition of HBPin to either Co(I)-aryl species to
generate the corresponding Co(III) species with mutually cis aryl and boryl ligands is pro-
posed to be the turnover-limiting step. Reductive elimination of the arylboronate ester
product follows, and the resulting hydride species 1-H can then react with either B2Pin2
(to regenerate 1-BPin) or HBPin (to regenerate 1-(H)2BPin).
With knowledge of the mechanism for the para-selective borylation of benzoate es-
ters in hand, one may rationalize the origins of the observed regioselectivity. Consid-
ering that the rate of the reaction is independent of the concentration of arene 2f,
and the two Co(I)-aryl species para- and meta-4f are present and observable by
19F and 31P NMR spectroscopy during the reaction, it appears that the C(sp2)-H
oxidative addition is fast and reversible and culminates in rapid interconversion be-
tween para- and meta-4f. This rapid interconversion of the two Co(I)-aryl species fol-
lowed by turnover-limiting oxidative addition of HBPin constitute a Curtin-Hammett
scenario56 wherein the selectivity of the reaction is determined by the relative rates
of the reaction of HBPin with either para-4f to give the major para-borylated product
or meta-4f to give the minor product.
Various subtle but significant pieces of data distinguish the mechanism implicated in the
selective ortho-borylation of fluorinated arenes from that proposed in this study by which
benzoate esters are preferentially para-borylated. In both cases, a zeroth-order depen-
dence on the concentration of the arene substrate was demonstrated, and in situ analysis
Chem 7, 237–254, January 14, 2021 247

ll
Article
of the resting states of the reaction revealed the formation of Co(I)-aryl species. Howev-
er, in the ortho-to-fluorine case, only the ortho-fluorinated Co(I)-aryl species was
observed, indicating that the regioselectivity of the reaction is determined prior to the
formation of this species. In the para-to-ester case, both possible isomers of the Co(I)-
aryl complex are formed with a slight (approximately 2:1) preference for para-4f over
meta-4f, and this ratio is inferior to that of the eventual arylboronate product distribu-
tion, suggesting that the Co(I)-aryl intermediates may have comparable thermodynamic
stability while a subsequent step of the reaction favors the production of the para-bory-
lated product. In terms of temperature effects, the regioselectivity in borylation of simple
fluorinated arenes previously examined remained constant as a function of reaction tem-
perature. In contrast, borylation of 2f at various temperatures revealed that the para-to-
ester selectivity improves slightly at low temperatures and slowly erodes as reaction tem-
perature is increased, supporting kinetically controlled regioselectivity. Finally, parallel
KIE experiments for a catalytic borylation, which gave ortho-to-fluorine regioselectivity,
afforded a kH/kD value of 1.1(1) that stands in clear contrast to the value of 3.3(1)
obtained in the present para-to-ester case, ostensibly arising from a turnover-limiting
oxidative addition of HBPin (or DBPin) to a Co(I)-aryl species (vide supra). This raises
questions related to the lack of a KIE in the previously examined ortho-to-fluorine
case. It is worthwhile to note that, in a comprehensive computational study concerning
the mechanism of the C(sp2)-H borylation of benzene by (iPrPNP)Co complexes, DFT cal-
culations using the uB97XD functional predict oxidative addition of HBPin to the Co(I)-
aryl to be the turnover-limiting step for this substrate.55 Moreover, the subsequent
reductive elimination of arylboronate product from the resulting complex was found
to occur with a kinetic barrier of 14.8 kcal/mol compared with a barrier of 17.2 kcal/
mol required for the slow step. It is anticipated that the relative barriers to these two
slowest steps of the catalytic cycle may change as a function of the arene substituents
and substitution pattern, and these variations can, in turn, result in a change in the turn-
over-limiting step. For a reaction in which C–B reductive elimination is turnover limiting,
the value of kH/kD is expected to be near unity, while reactions with turnover-limiting B–H
oxidative addition to Co(I)-aryl should afford normal, primary KIE values such as the value
of 3.3(1) observed in the borylation of 2ab and 2ab-d3.
Mechanistic Investigations of Arylboronate Ester Borylation
We next turned our attention to experimental mechanistic studies of arylboronate
borylation that also results in regioselective functionalization at the para position.
For these investigations, 2,6-difluorinated substrate 2k was selected due to its struc-
tural similarity to 2f and because of the pronounced para-to-boronate selectivity
observed despite the presence of different arene sites ortho-to-fluorine substitu-
ents. VTNA was applied for the catalytic C(sp2)-H borylation of 2k and the rate law
expressed in Equation 2 was obtained.
Rate = k[1-(H)2BPin]1[2k]1[B2Pin2]0
(Equation 2)
A first-order dependence on both the cobalt precatalyst and the arene substrate was
observed, supporting a pathway whereby C(sp2)-H oxidative addition is turnover
limiting. To further assess this possibility, the deuterated arene 2k-d3 was prepared
(see the Supplemental Information) and the deuterium KIE was determined. In two
separate vessels, 2k and 2k-d3 were each borylated in the presence of 5 mol %
1-(H)2BPin and 1.75 equiv of B2Pin2, and a value of 2.0(1) was measured for kH/kD ra-
tio at 23ꢁC. A normal value of this magnitude is consistent with turnover-limiting C–H
bond cleavage (Scheme 5B). In addition, in situ monitoring of the borylation reaction
by 1H and 31P NMR spectroscopy established that the cobalt(I)-boryl, 1-BPin,41 is the
248 Chem 7, 237–254, January 14, 2021

ll
Article
A
BPin
BPin
BPin
F
F
F
F
F
F
2.5 - 10 mol% 1-(H)2BPin
B2Pin2
THF, 23 °C, time
- HBPin
BPin
BPin
3k
2k
3k'
Rate = k[1-(H)2BPin]1[2k]1[B2Pin2]0
91
9
B
C
H3C
BPin
BPin
F
F
F
F
PiPr2
5 mol % 1-(H)2BPin
B2Pin2
N
THF, 23 °C, time
- HBPin
Co
P
BPin
BPin
2k
3k
kH
1-BPin
(<1 half-life)
= 2.0(1)
kD
(23 °C)
H3C
BPin
F
BPin
H
Co
H
PiPr2
N
F
F
F
5 mol % 1-(H)2BPin
B2Pin2
P
BPin
THF, 23 °C, time
D
D
D
D
- DBPin
D
BPin
1-(H)2BPin
(>1 half-life)
2k-d3
3k-d2
D
BPin
BPin
BPin
F3C
F3C
F
F3C
F
F
2.5 - 10 mol% 1-(H)2BPin
B2Pin2
THF, 23 °C, time
- HBPin
BPin
BPin
2ac
3ac
3ac'
Rate = k[1-(H)2BPin]1[2ac]0[B2Pin2]0
23
77
E
F
H3C
BPin
BPin
F3C
F
F3C
F
PiPr2
5 mol % 1-(H)2BPin
B2Pin2
N
THF, 23 °C, time
- HBPin
Co
F
P
BPin
CF3
2ac
3ac'
BPin
meta-4ac
(<1 half-life)
kH
= 1.2(2)
kD
(23 °C)
H3C
BPin
F3C
BPin
H
Co
H
PiPr2
N
F
F3C
F
5 mol % 1-(H)2BPin
B2Pin2
P
BPin
THF, 23 °C, time
- DBPin
D
D
D
BPin
1-(H)2BPin
(>1 half-life)
2ac-d2
3ac'-d1
Scheme 5. Mechanistic Investigations for the 1-(H)2BPin-Catalyzed C(sp2)-H Borylation of
Arylboronate Esters 2k and 2ac
(A) Borylation of 2k with 1-(H)2BPin and the experimentally determined rate law.
(B) Parallel deuterium kinetic isotope effect experiments for the borylation of 2k.
(C) Cobalt resting state observed in situ by 31P NMR spectroscopy for the borylation of 2k.
(D) Borylation of 2ac with 1-(H)2BPin and the experimentally determined rate law.
(E) Parallel deuterium kinetic isotope effect experiments for the borylation of 2ac.
(F) Cobalt resting state observed in situ by 19F and 31P NMR spectroscopy for the borylation of 2ac.
Chem 7, 237–254, January 14, 2021 249

ll
Article
catalyst resting state consistent with slow C–H oxidative addition of the arene 2k
(Scheme 5C).
All of the experimental data obtained for the borylation of 2k are consistent with turn-
over-limiting C(sp2)-H oxidative addition, establishing
a change in mechanism
compared with our previous reports of ortho-to-fluorine selective borylation with 1-
(H)2BPin. These previous studies demonstrated that fast and reversible C(sp2)-H oxida-
tive addition under thermodynamic control resulted in the observed selectivity for bor-
ylation ortho-to-fluorine substituents, a consequence of forming the strongest Co–C
bond. It is apparent from the change in the mechanism that the introduction of an induc-
tively donating boronate ester substituent effectively slows C(sp2)-H oxidative addition
relative to the subsequent steps of the reaction.
That the turnover-limiting step of the borylation of 2k was found to be distinct from other
previously studied fluoroarene substrates prompted consideration of the possible ori-
gins of the observed para-to-boron selectivity. As in the case of iridium-catalyzed bory-
lation, turnover-limiting C(sp2)-H oxidative addition implies that this step is irreversible
and, therefore, selectivity-determining, where the selectivity results from the different ki-
netic barriers for C–H bond cleavage. In the case of the borylation of 2k, the C(sp2)-H
bonds ortho to fluorine are stronger than the C(sp2)-H bond para to the boronate ester
substituent due to the established ortho stabilizing effect of fluorine.46 The observed
selectivity for borylation of the sites para to the [BPin] substituent is, therefore, consistent
with kinetically controlled selectivity, as the reaction favors activation of the weakest
available arene C–H bond. It should be noted that, while the regioselectivities for
both iridium- and cobalt-catalyzed borylation of this substrate are presumed to result
from kinetically controlled oxidative addition, the selectivity of the cobalt-catalyzed re-
action for the position para to the boronate substituent is superior under the conditions
examined. The fundamental difference between C(sp2)-H oxidative addition at the two
metals that causes the distinct selectivities is likely a result of the greater ionicity of the
forming metal–carbon bonds with cobalt compared to the analogous bonds with
iridium.46 Development of a more complete explanation for the superior electronic con-
trol observed with cobalt and associated computational investigations are ongoing. In
previous studies of ortho-selective fluoroarene borylation, the observed product ratios
remained constant when the reaction temperature was varied, consistent with a reaction
under thermodynamic control.37 In the borylation of 2k, varying the reaction tempera-
ture from 4ꢁC to 80ꢁC resulted in a gradual erosion of the regioselectivity, as the temper-
ature increased (see the Supplemental Information for details). While the ratio of prod-
ucts improved to 94:6 at 4ꢁC, activity decreased significantly, reaching only 59%
conversion after 1 week. Conversely, heating to 80ꢁC resulted in a rapid borylation
but produced a diminished 81:19 selectivity in favor of the para-diborylated product.
The observed temperature effects further support a kinetically controlled regioselectivity
for the borylation of 2k.
Observation of kinetically controlled para-to-boronate ester selectivity in the
cobalt-catalyzed C(sp2)-H borylation of 2k that overrides the previously reported
ortho-to-fluorine selectivity inspired other means of enabling substrate-controlled
regioselectivity. Of particular interest were arenes with both fluorine and boronate
ester substituents that were sufficiently electron deficient to accelerate C(sp2)-H
oxidative addition and enable thermodynamic rather than kinetic control. Replace-
ment of one fluorine in 2k with a more electron-withdrawing trifluoromethyl
substituent affords arene 2ac, which is anticipated to be more activated
toward C(sp2)-H oxidative addition than is 2k. Borylation of 2ac with 1 equiv of
B2Pin2 in the presence of 5 mol % 1-(H)2BPin at 23ꢁC generated predominantly
250 Chem 7, 237–254, January 14, 2021

ll
Article
the product of borylation ortho to the fluorine substituent in a 77:23 ratio, with the
minor product arising from borylation of the other sterically accessible C(sp2)-H
bond (para to the BPin substituent) (Scheme 5D).
The observed ortho-to-fluorine regioselectivity observed with 2ac prompted an investi-
gation of the mechanism and kinetic behavior of this reaction. Using VTNA, the rate law
for the C(sp2)-H borylation of 2ac with 1-(H)2BPin was determined (Equation 3).
Rate = k[1-(H)2BPin]1[2ac]0[B2Pin2]0
(Equation 3)
In contrast to the rate law in Equation 2 for the borylation of 2k, a zeroth-order
dependence in arene substrate was observed for the borylation of 2ac, indicating
saturation behavior, which is consistent with fast and reversible C(sp2)-H oxidative
addition. This result was corroborated by deuterium kinetic isotope experiments;
in two separate vessels, borylation of 2ac and 2ac-d2 with 5 mol % 1-(H)2BPin pro-
vided a kH/kD value of 1.2(2) at 23ꢁC, indicating that neither C–H bond cleavage
nor B–H bond cleavage occurs in the turnover-limiting step of the reaction (Scheme
5E). In situ monitoring of the catalytic reaction by 19F and 31P NMR spectroscopy es-
tablished the cobalt(I)-aryl complex, ortho-4ac, as the initial resting state (Scheme
5F), and the minor Co(I)-aryl isomer was not observed by 19F NMR spectroscopy
at any point in the time course. These data are consistent with fast and reversible
C(sp2)-H oxidative addition of the substrate, as previously reported for the boryla-
tion of fluorinated arenes. In addition, the selectivity of the borylation reaction
was unchanged from 23ꢁC to 80ꢁC (see the Supplemental Information for details).
In summary, the subtle electronic differences in arene substitution between 2f, 2k,
and 2ac were sufficient to change the selectivity-determining phenomena in the cat-
alytic borylation of these substrates with 1-(H)2BPin (Scheme 6). In the reaction of
benzoate ester substrates such as 2f, it is proposed that C(sp2)-H oxidative addition
and subsequent reductive elimination of HBPin reversibly generate Co(I)-aryl spe-
cies, of which multiple regioisomers are kinetically accessible; the observed para-
to-ester regioselectivity of the catalytic reaction is then derived from a kinetic
preference for the reaction of HBPin with the para-activated Co(I)-aryl intermediate
over the analogous reaction with the meta-activated isomer. With 2k, slow C(sp2)-H
oxidative addition is under kinetic control and favors borylation at the position para
to the boronate ester, while oxidative addition of 2ac proceeds reversibly to
preferentially generate thermodynamically favored ortho-fluorinated cobalt aryl in-
termediates in a non-Curtin-Hammett regime resulting predominantly in borylation
ortho to the fluorine substituent. Importantly, the presence or absence of a Curtin-
Hammett selectivity-determining scenario is dependent upon the disparity between
the rate of interconversion of the relevant equilibrating intermediates and the rate of
the subsequent turnover-limiting step.56 Detailed understanding of the relative rates
of fundamental steps in the catalytic cycles investigated in this study enables
substrate-controlled regioselectivity derived from the electronic properties of chem-
ically inequivalent C(sp2)-H sites, a concept which promises to empower rational
advances in regioselective C–H activation methodology.
EXPERIMENTAL PROCEDURES
Resource Availability
Lead Contact
Further information and requests for resources and reagents should be directed to
and will be fulfilled by the Lead Contact, Paul J. Chirik (pchirik@princeton.edu).
Chem 7, 237–254, January 14, 2021 251

ll
Article
H3C
PiPr2
N
H3C
Co
F
P
BPin
Co
PiPr2
H
CO2Me
HBPin
N
HBPin
CO2Me
F
F
P
F
HBPin
SLOW
FAST
kinetic
control
H3C
BPin
F
F
PiPr2
N
3f
CO2Me
Co
P
F
F
CO2Me
2f
FAST
Curtin-Hammett regime
H3C
H3C
BPin
Co
PiPr2
PiPr2
H
BPin
N
N
2k
F
F
Co
P
P
BPin
SLOW
kinetic
control
BPin
F
F
BPin
3k
FAST
2ac
thermodynamic
control
H3C
H3C
BPin
BPin
PiPr2
PiPr2
F3C
F
N
N
P
Co
Co
P
H
F
BPin
F
CF3
3ac'
BPin
BPin
ortho fluorine
effect-stabilized
CF3
Scheme 6. Mechanistic Explanation for Substrate-Controlled Selectivity as a Result of Arene
Electronics
(A) Experimentally determined rate laws for catalytic borylation and predominant site of C(sp2)-H
borylation for arene substrates examined in this work.
(B) Explanation for observed regioselectivity in cobalt-catalyzed borylation as a consequence of the
relative rates of C(sp2)-H oxidative addition.
Materials Availability
All unique reagents generated in this study are available from the Lead Contact
without restriction.
Data and Code Availability
All datasets generated within these investigations are available in the Supplemental
Information.
SUPPLEMENTAL INFORMATION
Supplemental Information can be found online at https://doi.org/10.1016/j.chempr.
2020.11.017.
ACKNOWLEDGMENTS
Financial support was provided by the NIH (5R01GM121441). T.P.P. thanks Amgen
for financial support as well as Princeton University for an Edward C. Taylor Fellow-
ship. L.Q. acknowledges support from the Alfried Krupp von Bohlen und Halbach
Foundation. We also thank AllyChem for a gift of B2Pin2.
252 Chem 7, 237–254, January 14, 2021

ll
Article
AUTHOR CONTRIBUTIONS
T.P.P., L.Q., and K.T.M. performed the experiments and analyzed the data. P.J.C.
supervised the research and assisted with experimental design and data analysis.
All authors contributed to the preparation and revision of the manuscript.
DECLARATION OF INTERESTS
The authors declare no competing interests.
Received: April 21, 2020
Revised: August 8, 2020
Accepted: November 14, 2020
Published: December 9, 2020
REFERENCES
1. Hartwig, J.F. (2011). Regioselectivity of the
borylation of alkanes and arenes. Chem. Soc.
Rev. 40, 1992–2002.
iridium complexes. Intermediacy and chemistry
of bipyridine-ligated iridium trisboryl
complexes. J. Am. Chem. Soc. 127, 14263–
14278.
homogeneous ortho borylation catalysis.
J. Am. Chem. Soc. 136, 14345–14348.
21. Hale, L.V.A., McGarry, K.A., Ringgold, M.A.,
and Clark, T.B. (2015). Role of hemilabile
diamine ligands in the amine-directed C–H
borylation of arenes. Organometallics 34,
51–55.
2. Mkhalid, I.A.I., Barnard, J.H., Marder, T.B.,
Murphy, J.M., and Hartwig, J.F. (2010). C–H
activation for the construction of C–B Bonds.
Chem. Rev. 110, 890–931.
12. Larsen, M.A., and Hartwig, J.F. (2014). Iridium-
catalyzed C–H borylation of heteroarenes:
scope, regioselectivity, application to late-
stage functionalization, and mechanism. J. Am.
Chem. Soc. 136, 4287–4299.
22. Li, H.L., Kanai, M., and Kuninobu, Y. (2017).
Iridium/bipyridine-catalyzed ortho-selective
C–H borylation of phenol and aniline
3. Hartwig, J.F. (2012). Borylation and silylation of
C–H bonds: a platform for diverse C–H Bond
functionalizations. Acc. Chem. Res. 45,
864–873.
13. Green, A.G., Liu, P., Merlic, C.A., and Houk,
K.N. (2014). Distortion/interaction analysis
reveals the origins of selectivities in iridium-
catalyzed C–H borylation of substituted arenes
and 5-membered heterocycles. J. Am. Chem.
Soc. 136, 4575–4583.
derivatives. Org. Lett. 19, 5944–5947.
23. Kuninobu, Y., Ida, H., Nishi, M., and Kanai, M.
(2015). A meta-selective C–H borylation
4. Xu, L., Wang, G., Zhang, S., Wang, H., Wang, L.,
Liu, L., Jiao, J., and Li, P. (2017). Recent
advances in catalytic C–H borylation reactions.
Tetrahedron 73, 7123–7157.
directed by a secondary interaction between
ligand and substrate. Nat. Chem. 7, 712–717.
14. Boebel, T.A., and Hartwig, J.F. (2008). Silyl-
directed, iridium-catalyzed ortho-borylation of
arenes. A one-pot ortho-borylation of phenols,
arylamines, and alkylarenes. J. Am. Chem. Soc.
130, 7534–7535.
24. Bisht, R., and Chattopadhyay, B. (2016). Formal
Ir-catalyzed ligand-enabled ortho and meta
borylation of aromatic aldehydes via in situ-
generated imines. J. Am. Chem. Soc. 138,
84–87.
5. Wright, J.S., Scott, P.J.H., and Steel, P.G.
(2020). Iridium-catalysed C-H borylation of
heteroarenes: balancing steric and electronic
regiocontrol. Angew. Chem. Int. Ed. https://
doi.org/10.1002/anie.202001520.
15. Ishiyama, T., Isou, H., Kikuchi, T., and Miyaura,
N. (2010). Ortho-C–H borylation of benzoate
esters with bis(pinacolato)diboron catalyzed by
iridium–phosphine complexes. Chem.
Commun. 46, 159–161.
25. Davis, H.J., Mihai, M.T., and Phipps, R.J. (2016).
Ion pair-directed regiocontrol in transition-
metal catalysis: a meta-selective C–H
6. Hall, D. (2011). Boronic Acids: Preparation and
Applications in Organic Synthesis, Medicine
and Materials, Second Edition (Wiley-VCH
Press).
borylation of aromatic quaternary ammonium
salts. J. Am. Chem. Soc. 138, 12759–12762.
7. Campeau, L.-C., Chen, Q., Gauvreau, D.,
Girardin, M., Belyk, K., Maligres, P., Zhou, G.,
Gu, C., Zhang, W., Tan, L., et al. (2016). A robust
kilo-scale synthesis of doravirine. Org. Process
Res. Dev. 20, 1476–1481.
16. Robbins, D.W., Boebel, T.A., and Hartwig, J.F.
(2010). Iridium-catalyzed, silyl-directed
26. Davis, H.J., Genov, G.R., and Phipps, R.J.
(2017). meta-selective C–H borylation of
benzylamine-, phenethylamine-, and
phenylpropylamine-derived amides enabled
by a single anionic ligand. Angew. Chem. Int.
Ed. 56, 13351–13355.
borylation of nitrogen-containing heterocycles.
J. Am. Chem. Soc. 132, 4068–4069.
17. Ros, A., Estepa, B., Lo´ pez-Rodrıguez, R.,
´
´
8. Ishiyama, T., Takagi, J., Ishida, K., Miyaura, N.,
Anastasi, N.R., and Hartwig, J.F. (2002). Mild
iridium-catalyzed borylation of arenes. High
turnover numbers, room temperature
reactions, and isolation of a potential
intermediate. J. Am. Chem. Soc. 124, 390–391.
Alvarez, E., Ferna´ndez, R., and Lassaletta, J.M.
(2011). Use of hemilabile N,N ligands in
nitrogen-directed iridium-catalyzed
borylations of arenes. Angew. Chem. Int. Ed.
50, 11724–11728.
27. Chattopadhyay, B., Dannatt, J.E., Andujar-De
Sanctis, I.L., Gore, K.A., Maleczka, R.E.,
Singleton, D.A., and Smith, M.R. (2017). Ir-
catalyzed ortho-borylation of phenols directed
by substrate-ligand electrostatic interactions: a
combined experimental/in silico strategy for
optimizing weak interactions. J. Am. Chem.
Soc. 139, 7864–7871.
18. Ros, A., Lo´ pez-Rodrıguez, R., Estepa, B.,
´
´
9. Cho, J.Y., Tse, M.K., Holmes, D., Maleczka,
R.E., and Smith, M.R. (2002). Remarkably
selective iridium catalysts for the elaboration of
aromatic C–H Bonds. Science 295, 305–308.
Alvarez, E., Ferna´ndez, R., and Lassaletta, J.M.
(2012). Hydrazone as the directing group for Ir-
catalyzed arene diborylations and sequential
functionalizations. J. Am. Chem. Soc. 134,
4573–4576.
28. Smith, M.R., Bisht, R., Haldar, C., Pandey, G.,
Dannatt, J.E., Ghaffari, B., Maleczka, R.E., and
Chattopadhyay, B. (2018). Achieving high ortho
selectivity in aniline C–H borylations by
modifying boron substituents. ACS Catal. 8,
6216–6223.
10. Chotana, G.A., Rak, M.A., and Smith, M.R.
(2005). Sterically directed functionalization of
aromatic C–H bonds: selective borylation ortho
to cyano groups in arenes and heterocycles.
J. Am. Chem. Soc. 127, 10539–10544.
19. Xiao, B., Li, Y.M., Liu, Z.J., Yang, H.Y., and Fu, Y.
(2012). Palladium-catalyzed monoselective C–
H borylation of acetanilides under acidic
conditions. Chem. Commun. 48, 4854–4856.
11. Boller, T.M., Murphy, J.M., Hapke, M.,
Ishiyama, T., Miyaura, N., and Hartwig, J.F.
(2005). Mechanism of the mild functionalization
of arenes by diboron reagents catalyzed by
20. Ghaffari, B., Preshlock, S.M., Plattner, D.L.,
Staples, R.J., Maligres, P.E., Krska, S.W.,
Maleczka, R.E., and Smith, M.R. (2014). Silyl
phosphorus and nitrogen donor chelates for
29. Mihai, M.T., Davis, H.J., Genov, G.R., and
Phipps, R.J. (2018). Ion pair-directed C–H
activation on flexible ammonium salts: meta-
selective borylation of quaternized
Chem 7, 237–254, January 14, 2021 253

ll
Article
phenethylamines and phenylpropylamines.
ACS Catal. 8, 3764–3769.
C–H borylation of arenes catalyzed by a PSiN-
pincer platinum complex. Chem. Commun. 51,
17662–17665.
48. Noguchi, H., Hojo, K., and Suginome, M.
(2007). Boron-masking strategy for the
selective synthesis of oligoarenes via iterative
Suzuki-Miyaura coupling. J. Am. Chem. Soc.
129, 758–759.
30. Lee, B., Mihai, M.T., Stojalnikova, V., and
Phipps, R.J. (2019). Ion-pair-directed borylation
of aromatic phosphonium salts. J. Org. Chem.
84, 13124–13134.
40. Obligacion, J.V., Semproni, S.P., and Chirik,
P.J. (2014). Cobalt-catalyzed C-H borylation.
J. Am. Chem. Soc. 136, 4133–4136.
49. Yoshida, H., Seki, M., Kamio, S., Tanaka, H.,
Izumi, Y., Li, J., Osaka, I., Abe, M., Andoh, H.,
Yajima, T., et al. (2020). Direct Suzuki-Miyaura
coupling with naphthalene-1,8-
31. Saito, Y., Segawa, Y., and Itami, K. (2015). para-
C-H borylation of benzene derivatives by a
bulky iridium catalyst. J. Am. Chem. Soc. 137,
5193–5198.
41. Obligacion, J.V., Semproni, S.P., Pappas, I.,
and Chirik, P.J. (2016). Cobalt-catalyzed C(sp2)-
H borylation: mechanistic insights inspire
catalyst design. J. Am. Chem. Soc. 138, 10645–
10653.
diaminato(dan)-substituted organoboranes.
ACS Catal. 10, 346–351.
32. Saito, Y., Yamanoue, K., Segawa, Y., and Itami,
K. (2020). Selective transformation of strychnine
and 1,2-disubstituted benzenes by C–H
borylation. Chem 6, 1–2.
50. Mutoh, Y., Yamamoto, K., and Saito, S. (2020).
Suzuki-Miyaura cross-coupling of 1,8-
42. Obligacion, J.V., and Chirik, P.J. (2017).
Mechanistic studies of cobalt-catalyzed C(sp2)-
H borylation of five-membered heteroarenes
with pinacolborane. ACS Catal. 7, 4366–4371.
diaminonaphthalene (dan)-protected
arylboronic acids. ACS Catal. 10, 352–357.
33. Yang, L., Semba, K., and Nakao, Y. (2017). para-
selective C–H borylation of (hetero)arenes by
cooperative iridium/aluminum catalysis.
Angew. Chem. Int. Ed. 56, 4853–4857.
51. Le´ onard, N.G., Bezdek, M.B., and Chirik, P.J.
(2017). Cobalt-catalyzed C(sp2)–H borylation
with an air-stable, readily prepared terpyridine
cobalt(II) bis(acetate) precatalyst.
´
43. Pabst, T.P., Obligacion, J.V., Rochette, E.,
Pappas, I., and Chirik, P.J. (2019). Cobalt-
catalyzed borylation of fluorinated arenes:
thermodynamic control of C(sp2)-H oxidative
addition results in ortho-to-fluorine selectivity.
J. Am. Chem. Soc. 141, 15378–15389.
34. Hoque, M.E., Bisht, R., Haldar, C., and
Chattopadhyay, B. (2017). Noncovalent
interactions in Ir-catalyzed C–H activation: L-
shaped ligand for para-selective borylation of
aromatic esters. J. Am. Chem. Soc. 139, 7745–
7748.
Organometallics 36, 142–150.
52. Bure´ s, J. (2016). Variable time normalization
analysis: general graphical elucidation of
reaction orders from concentration profiles.
Angew. Chem. Int. Ed. 55, 16084–16087.
44. Clot, E., Eisenstein, O., Jasim, N., MacGregor,
S.A., McGrady, J.E., and Perutz, R.N. (2011). C-
F and C-H Bond activation of fluorobenzenes
and fluoropyridines at transition metal centers:
how fluorine tips the scales. Acc. Chem. Res.
44, 333–348.
35. Mihai, M.T., Williams, B.D., and Phipps, R.J.
(2019). Para-selective CꢀH borylation of
common arene building blocks enabled by ion-
pairing with a bulky countercation. J. Am.
Chem. Soc. 141, 15477–15482.
53. Yu, R.P., Hesk, D., Rivera, N., Pelczer, I., and
Chirik, P.J. (2016). Iron-catalysed tritiation of
pharmaceuticals. Nature 529, 195–199.Corpas,
J., Viereck, P., and Chirik, P.J. (2020). C(sp2)–H
activation with pyridine dicarbene iron dialkyl
complexes: hydrogen isotope exchange of
arenes using benzene-d6 as a deuterium
source. ACS Catal. 10, 8640–8647.
45. Selmeczy, A.D., Jones, W.D., Partridge, M.G.,
and Perutz, R.N. (1994). Selectivity in the
activation of fluor inated aromatic
36. Montero Bastidas, J.R., Oleskey, T.J., Miller,
S.L., Smith, M.R., and Maleczka, R.E. (2019).
Para-selective, iridium-catalyzed CꢀH
borylations of sulfated phenols, benzyl
alcohols, and anilines directed by ion-pair
electrostatic interactions. J. Am. Chem. Soc.
141, 15483–15487.
hydrocarbons by rhodium complexes [(C5H5)
Rh(PMe3)] and [(C5Me5)Rh(PMe3)].
Organometallics 13, 522–532.
54. Simmons, E.M., and Hartwig, J.F. (2012). On the
interpretation of deuterium kinetic isotope
effects in C-H bond functionalizations by
transition-metal complexes. Angew. Chem. Int.
Ed. 51, 3066–3072.
46. Clot, E., Me´ gret, C., Eisenstein, O., and Perutz,
R.N. (2009). Exceptional sensitivity of metal-aryl
bond energies to ortho-fluorine substituents:
influence of the metal, the coordination
sphere, and the spectator ligands on M-C/H-C
bond energy correlations. J. Am. Chem. Soc.
131, 7817–7827.
37. Obligacion, J.V., Bezdek, M.J., and Chirik, P.J.
(2017). C(sp2)-H borylation of fluorinated
arenes using an air-stable cobalt precatalyst:
electronically enhanced site selectivity enables
synthetic opportunities. J. Am. Chem. Soc. 139,
2825–2832.
55. Li, H., Obligacion, J.V., Chirik, P.J., and Hall,
M.B. (2018). Cobalt pincer complexes in
catalytic C–H borylation: the pincer ligand flips
rather than dearomatizes. ACS Catal. 8, 10606–
10618.
47. Mortier, J. (2016). Arene Chemistry: Reaction
Mechanisms and Methods for Aromatic
Compounds (Wiley).Harvey, D.R., and Norman,
R.O.C. (1962). The ortho:para-ratio in aromatic
substitution. Part VI. The nitration of
benzeneboronic acid. J. Chem. Soc. 3822–
3823. https://pubs.rsc.org/en/content/
articlelanding/1962/JR/jr9620003818#!
divAbstract.
38. Furukawa, T., Tobisu, M., and Chatani, N.
(2015). C–H functionalization at sterically
congested positions by the platinum-catalyzed
borylation of arenes. J. Am. Chem. Soc. 137,
12211–12214.
56. Seeman, J.I. (1983). Effect of conformational
change on reactivity in organic chemistry.
Evaluations, applications, and extensions of
Curtin-Hammett/Winstein-Holness kinetics.
Chem. Rev. 83, 83–134.
39. Takaya, J., Ito, S., Nomoto, H., Saito, N., Kirai,
N., and Iwasawa, N. (2015). Fluorine-controlled
254 Chem 7, 237–254, January 14, 2021