Regioselective borylation of the C-H bonds in alkylamines and alkyl ethers. Observation and origin of high reactivity of primary C-H bonds beta to nitrogen and oxygen

Article abstract of DOI:10.1021/ja503676d

Borylation of aliphatic C-H bonds in alkylamines and alkyl ethers to form primary aminoalkyl and alkoxyalkyl boronate esters and studies on the origin of the regioselectivity of these reactions are reported. The products of these reactions can be used directly in Suzuki-Miyaura cross-coupling reactions or isolated as air-stable potassium trifluoroborate salts. Selective borylation of the terminal C-H bond at the positions β to oxygen and nitrogen occurs in preference to borylation of the other terminal C-H bonds. Experimental studies and computational results show that C-H bond cleavage is the rate-determining step of the current borylation reactions. The observed higher reactivity of C-H bonds at the terminal position of ethylamines and ethers results from a combination of attractive Lewis acid-base and hydrogen-bonding interactions, as well as typical repulsive steric interactions, in the transition state. In this transition state, the heteroatom lies directly above the boron atom of one boryl ligand, creating a stabilizing interaction between the weak Lewis acid and Lewis base, and a series of C-H bonds of the substrate lie near the oxygen atoms of the boryl ligands, participating in a set of weak C- H···O interactions that lead to significant stabilization of the transition state forming the major product.

Full text of DOI:10.1021/ja503676d

Article
pubs.acs.org/JACS
Regioselective Borylation of the C−H Bonds in Alkylamines and Alkyl
Ethers. Observation and Origin of High Reactivity of Primary C−H
Bonds Beta to Nitrogen and Oxygen
Qian Li,†,§ Carl W. Liskey,‡,§ and John F. Hartwig*,†
†Department of Chemistry, University of California, Berkeley, California 94720, United States
‡Department of Chemistry, University of Illinois, Urbana, Illinois 61801, United States
S
* Supporting Information
ABSTRACT: Borylation of aliphatic C−H bonds in alkyl-
amines and alkyl ethers to form primary aminoalkyl and
alkoxyalkyl boronate esters and studies on the origin of the
regioselectivity of these reactions are reported. The products of
these reactions can be used directly in Suzuki−Miyaura cross-
coupling reactions or isolated as air-stable potassium
trifluoroborate salts. Selective borylation of the terminal C−
H bond at the positions β to oxygen and nitrogen occurs in
preference to borylation of the other terminal C−H bonds.
Experimental studies and computational results show that C−H bond cleavage is the rate-determining step of the current
borylation reactions. The observed higher reactivity of C−H bonds at the terminal position of ethylamines and ethers results
from a combination of attractive Lewis acid−base and hydrogen-bonding interactions, as well as typical repulsive steric
interactions, in the transition state. In this transition state, the heteroatom lies directly above the boron atom of one boryl ligand,
creating a stabilizing interaction between the weak Lewis acid and Lewis base, and a series of C−H bonds of the substrate lie near
the oxygen atoms of the boryl ligands, participating in a set of weak C−H···O interactions that lead to significant stabilization of
the transition state forming the major product.
INTRODUCTION
■
Alkylboron reagents are versatile synthetic intermediates that
undergo a variety of transformations.1 These transformations
include recently developed cross-coupling and amination
reactions of aliphatic organoboron compounds,2 as well as
classic oxidations of alkylboron reagents to alcohols. Traditional
methods to form these aliphatic organoboron reagents include
the addition of a reactive organometallic species, such as an
alkyl lithium or alkyl Grignard reagent, to an electrophilic
boron compound and the hydroboration of olefins.3 Both of
these methods require a functionalized precursor, such as an
alkyl halide or an olefin. Furthermore, the hydroboration of
enol ethers or simple acyclic enamines derived from aldehydes
to form alkoxyboranes and aminoboranes often results in
elimination products.4−6
During the past decade, our laboratory reported the
functionalization of aliphatic C−H bonds with metal-boryl
catalysts, and high selectivity was observed for functionalization
of terminal C−H bonds over internal C−H bonds. Rhodium
and ruthenium catalysts that contain pentamethylcyclopentyl-
dienyl ligands (Cp*) catalyzed the formation of 1-alkyl
boronate esters from alkanes and diboron or borane reagents.
The selectivity of these aliphatic C−H borylation processes is
controlled by the steric properties of the substrate. We recently
Aminoalkyl and alkoxyalkyl trifluoroborate salts have been
shown recently by Molander and co-workers to undergo
Suzuki−Miyaura cross-coupling.7−15 These reagents were
synthesized from potassium bromomethyltrifluoroborates by
nucleophilic displacement reactions (eq 1)16 or by conversion
of the bromoethyl ethers into the corresponding organoboron
species through a Cu-catalyzed borylation reaction (eq 2).11
Although these reactions occur in good yields, prefunctional-
ized alkyl halide starting materials are required. A synthesis of
alkyl boronates directly from alkylamines or ethers would avoid
the need for halogenated reagents.17
Received: April 13, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
delineated the origin of this steric control, which results from a
more favorable equilibrium for activation of primary over
secondary C−H bonds and a lower barrier for formation of the
B−C bond by reductive elimination from a primary alkyl
complex than from a secondary alkyl complex.18
complexes containing a series of dative nitrogen ligands related
to Me4phen were also conducted, but these reactions occurred
in lower yields than those with the catalyst containing
Me4phen. (For details, see the Supporting Information.)
Although competitive borylation of the secondary C−H
bonds occurred, borylation of the primary C−H bond occurred
in high yield.25 The same reaction catalyzed by [Cp*RhCl2]2
gave the product from borylation of the methyl group in a low
(<30%) yield.20
We attempted to isolate the product from borylation of N-
methylmorpholine by several methods, including standard
distillation and chromatography, as well as conversion to the
potassium trifluoroborate salt, but the isolated material was
contaminated with impurities. However, the crude product of
this borylation reaction was suitable for cross-coupling with aryl
halides. The reaction of 4-bromoanisole with the product from
borylation of N-methylmorpholine, prepared by C−H
borylation under the above conditions followed by evaporation
of the volatile components, proceeded in good yield (83%
isolated yield based on ArX) in the presence of Pd(dba)2 (5
mol %), Q-Phos (5 mol %), and CsF. Only the aminomethyl
boron compound reacted; no product from coupling of the
secondary boronate ester side product was observed.
In this article, we report the synthesis of aminoalkyl- and
alkoxyalkylboron reagents by Ir-catalyzed C−H borylation
reactions. The products of these C−H bond functionalization
reactions can be converted to stable trifluoroborate salts (eq 3)
or used directly in Suzuki−Miyaura cross-couplings (eq 4).
These C−H borylation reactions occur with an unusual but
pronounced preference for the formation of product from the
C−H bond located at the β positions to oxygen and nitrogen
over the other C−H bonds in ethers and amines and over the
C−H bonds in alkanes. These reactions also occur in good
yield at the methyl C−H bonds of N-methyl cyclic amines.
Experimental mechanistic and computational studies are
consistent with irreversible and rate-determining C−H bond
cleavage through a transition state containing a weak, stabilizing
Lewis acid−base interaction and a series of weak C−H···O
interactions between the C−H bonds of the substrate and the
oxygen atoms of the boryl ligand. These attractive interactions
contribute significantly to the observed selectivity.
The scope of the one-pot borylation and arylation of
methylamines under the developed conditions is summarized in
Table 1. This reaction sequence occurred with a number of
cyclic methylamines, including N-methyl-substituted morpho-
line, piperidine, piperazine, and pyrrolidine (entries 1−5). Both
electron-rich (entries 6 and 7) and electron-deficient aryl
bromides (entries 8−11) underwent the coupling reaction with
the aminomethyl boronate esters in good yields. Reactions of
these boronate esters with aryl chlorides also occurred,
although the yields were slightly lower than those of reactions
conducted with aryl bromides (entries 7−10). Heteroaryl
bromides also underwent the cross-coupling reaction with the
borylation product (entries 12 and 13). The C−H borylation
occurred with commercially available [Ir(COD)OMe]2 in place
of the trisboryl catalyst precursor, but higher loadings of the
catalyst components were required (entry 2).
Borylation of Cyclic and Acyclic Ethyl- and Propyl-
amines. Most classes of C−H bond functionalization reactions
occur at the C−H bonds α to a nitrogen atom, due to
electronic effects.26 To assess the relationship between the
relative reactivity of different C−H bonds of amines and ethers
and the selectivity of prior systems, we evaluated the reactions
of longer-chain N-alkyl cyclic amines and acyclic amines.
Borylation of these substrates occurred with the same catalyst
as we used for the borylation of N-methylamines to form the
corresponding 1-aminoalkyl boronate esters in good yields
(Table 2). The examples in Table 2 show that borylation of the
ethyl groups in amines occurs preferentially over borylation of
methyl, propyl, and butyl groups in amines.
EXPERIMENTAL RESULTS
■
Previously, we reported the borylation of primary aliphatic C−
H bonds catalyzed by Cp*Rh and Cp*Ru complexes. However,
the small scale and high temperatures (150 °C) of these
reactions limited their synthetic utility.19−21 Recently, we
reported a more active iridium catalyst containing 3,4,7,8-
tetramethylphenanthroline (Me4phen) as ligand for C−H bond
functionalization with main group reagents;22 this system
catalyzes the borylation of secondary C−H bonds in cyclic
ethers23 and cyclopropanes24 and the primary C−H bonds in
methylsilanes25a to form the corresponding boronate ester
products. To determine the potential of this new iridium
system to catalyze the borylation of primary C−H bonds of
common organic compounds in a more practical fashion, we
investigated the borylations of alkyl ethers and alkylamines on a
preparative scale. The high reactivity of different ethers and
amines toward this iridium catalyst led us to study the origins of
selectivity in the borylation of these reagents and to compare
the factors controlling selectivity in this system to those
controlling selectivity in reactions catalyzed by Cp*Rh
complexes.
Development of the Borylation of Methylamine C−H
Bonds. We began our investigation of the borylation of ethers
and amines by studying the C−H borylation of N-methyl cyclic
amines. These amines contain just one type of primary C−H
bond and could, therefore, be selective for the formation of a
single product. The reaction of bis-pinacolatodiboron (B2pin2)
as limiting reagent with neat N-methylmorpholine in the
presence of (η6-mes)IrBpin3 (mes = mesitylene) and Me4phen
as ligand generated the aminomethyl boronate ester product in
88% yield based on B2pin2 (eq 5). Reactions catalyzed by
To facilitate isolation of the aliphatic boronate esters, the
crude products were converted to the corresponding air-stable
potassium trifluoroborate salts. In some cases, the yields
exceeded 100%, based on the quantity of diboron reagent.
These values indicate that the HBpin byproduct also reacts with
the amines to form some of the functionalized product (Table
2, entries 1−4).
The C−H borylation of diethylamines, as well as triethyl-
amine, occurred in good yield (entries 1−4) to form
aminoethyl boronate esters. The C−H borylation of Et2NMe
occurred exclusively at the primary C−H bonds of the ethyl
B
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. One-Pot Borylation and Arylation of Methylamines
Table 2. Ir-Catalyzed C−H Borylation of Alkylamines
a
Reaction conditions: B2pin2 (1.0 equiv), amine (9.0 equiv), (η6-
mes)IrBpin3 (2 mol %), 3,4,7,8-Me4phen (2 mol %), neat, 120 °C, 24
h. Yield determined by gas chromatographic analysis with isododecane
as internal standard. The yield is based on the moles of product per
mole of B2pin2; a yield above 100% reflects the conversion of the
b
HBpin byproduct to the alkylBpin. Isolated as trifluoroborate salt.
a
Conditions: KHF2 (4.5 M, aq, 4.0 equiv), MeOH, 22 °C, 3 h. Some
isolated products contain small quantities (about 8−15%) of KBF4
impurity. The yields reported are corrected for the amount of this
impurity, based on 11B NMR spectroscopy. For details, see the
Reaction conditions: 1. ArX (1.0 equiv), B2pin2 (1.3 equiv), amine
(12.0 equiv based on ArX, 9.0 equiv based on B2pin2), (η6-mes)IrBpin3
(4 mol % to B2pin2), 3,4,7,8-Me4phen (4 mol %), neat, 120 °C, 24 h;
2. Ar-X (1.0 equiv), Pd(dba)2 (5 mol % to B2pin2), Q-phos (5 mol %),
CsF (3 equiv), dioxane:H2O (5:1), 100 °C, 24 h. The reported yield in
this table is the isolated yield based on ArX. Secondary C−H
borylation products cannot be avoided in each reaction, but they can
be easily separated with the primary C−H borylation product by
c
d
Supporting Information. 4 mol % catalyst. Contained 10% of an
inseparable impurity by 1H NMR spectroscopy; see Supporting
Information for details.
b
following Suzuki cross-coupling. Reaction conducted with B2pin2 (1.5
equiv), [Ir(COD)OMe]2 (5 mol %), and 3,4,7,8-Me4phen (10 mol %).
Reaction conducted for 48 h.
c
Borylation of Ethers. An Ir-catalyzed borylation of ethers
that is analogous to the Ir-catalyzed borylation of amines would
be useful because alkoxyalkyl boron compounds are valuable
synthetic intermediates. Thus, we evaluated the site selectivity
for borylation of the C−H bonds in alkyl ethers.
group over the more acidic and weaker C−H bonds of the N-
methyl position (entry 1). Borylation of Et2NnPr and Et2NnBu
occurred preferentially at the ethyl groups over the longer-chain
alkyl groups to give the two types of isomeric products in 83:17
and 86:14 ratios, respectively (entries 3 and 4). The reaction of
MeNnPr2 occurred selectively at the propyl group, showing that
the methyl C−H bonds of N-methylamines are the least
reactive of the primary C−H bonds in alkylamines (entry 5).
Similarly, borylation of N-ethylmorpholine occurred to give
one major product resulting from the borylation of the primary
C−H bond in the ethyl group (1:(2+2′) = 94:6, entry 6). This
selectivity for the primary C−H bond is higher than that for
borylation of the methyl group of N-methylmorpholine (1:
(2+2′) = 80:20). Although borylation at an ethyl group was
Borylation of the primary C−H bonds of alkyl ethers
occurred in good yields with the same catalyst that led to the
functionalization of the primary C−H bonds in alkylamines
(Table 3). The catalyst loadings required for borylation of alkyl
ethers were lower than those required for borylation of methyl-
and ethylamines, and the reactions occurred with high
selectivity for the primary C−H bonds of ethyl groups. Alkyl
ethyl ethers underwent the borylation reaction with just 1 mol
% of the catalyst in good yields. Again, the yields exceeded
100% in some cases based on the quantity of diboron reagent,
indicating that the HBpin byproduct reacts with the ethers to
form some of the functionalized product (entries 1−4).
Borylation of dibutyl ether occurred in good yield, but a higher
catalyst loading was required for this reaction than was required
for reactions of ethyl ethers (entry 5).27 Pivalate-protected
alcohols also underwent C−H borylation at the ethyl group of
n
shown to occur in preference to that at a propyl group, Pr3N
did undergo borylation in good yield with 4 mol % catalyst
loading (entry 7). The origin of this selectivity for reaction at an
ethyl group will be discussed later in this paper.
C
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
alkoxymethyl boronate was less than that at 10 h, suggesting
that decomposition of the product from borylation at the
methyl group occurs.
Table 3. Ir-Catalyzed C−H Borylation of Alkyl Ethers
Like the borylation of ethylamines, the borylation of ethyl
ethers occurred preferentially at the ethyl group over longer-
chain alkyl groups. The reaction of ethyl n-propyl and ethyl n-
butyl ether occurred with 85:15 and 87:13 selectivity for
reaction at the ethyl group over the longer-chain alkyl groups,
respectively. In addition, the borylation occurred with
selectivity for functionalization at the ethyl group of ethers
that contain three different types of primary C−H bonds. For
example, borylation of ethyl sec-butyl ether occurred with 86:14
selectivity for a primary C−H bond of the ethyl group over the
less hindered of the two types of primary C−H bonds in the
sec-butyl group (entry 9). Finally, borylation of ethyl menthyl
ether occurred exclusively on the ethyl group (entry 10).
Transformations of the Borylation Products. The
boronate esters generated from borylation of C−H bonds in
ethyl and butyl ethers can be converted into further
functionalized products resulting from cross-coupling and
oxidation. The pinacolate esters of ethoxy- and butoxyboronic
acids underwent Suzuki−Miyaura cross-coupling with aryl
bromides under conditions reported recently by Steel, Marder,
and Liu (eqs 6 and 7).28 In addition, oxidation of the product
from borylation of dibutyl ether occurred under standard
conditions in good yield (eq 8).
a
Reaction conditions: B2pin2 (1.0 equiv), ether (9.0 equiv), (η6-
mes)IrBpin3 (1 mol %), 3,4,7,8-Me4phen (1 mol %), neat, 120 °C, 24
h. Yield determined by gas chromatographic analysis with isododecane
as internal standard. The yield is based on the moles of product per
mole of B2pin2; a yield above 100% reflects the conversion of the
b
HBpin byproduct to the alkylBpin. Isolated as trifluoroborate salt.
Conditions: KHF2 (4.5 M, aq, 4.0 equiv), MeOH, 22 °C, 3 h. Some
isolated products contain small quantities (about 8−15%) of KBF4
impurity. The yields reported are corrected for the amount of this
impurity, based on 11B NMR spectroscopy. For details, see the
c
d
e
f
Supporting Information. 36 h. 4 mol % catalyst. 100 °C. 3 h.
g
h
Isolated as boronate ester. 10 h.
Competition Experiments. To gain a clearer view of the
relative rate of borylation of a primary C−H bond in an ether
versus an alkane, we conducted the reaction of a mixture of n-
BuOEt and octane (Et2O is too volatile to ensure the correct
concentrations under these conditions). The C−H borylation
occurred preferentially with the C−H bonds of the ether over
the C−H bonds of octane to give an 84:12:4 ratio of the
functionalized products, reflecting a 96:4 ratio of products from
functionalization of the ether over those from functionalization
of the alkane (eq 9). Thus, the primary C−H bonds in both
alkoxy groups of the ether are substantially more reactive
toward this borylation process than is the primary C−H bond
of an alkyl group.
We also evaluated the relative rate of borylation of a methyl
C−H bond in an ether versus that in an amine by conducting
the reaction of a mixture of n-PrOEt and n-PrNEt2. The rate of
borylation of the C−H bonds in the ether was similar of that in
the amine. A 44:56 ratio of the products was formed from
functionalization of the ether and amine (eq 10; the ratios
the ether chain, illustrating the potential of the borylation
reaction to form α,ω-functionalized products (entry 6).
Methyl ethers also underwent C−H borylation. These
reactions were selective for borylation at the methyl group
(entries 7 and 8) over a tert-butyl group or a cycloalkyl group.
The products of the borylation of methyl ethers were somewhat
unstable to the reaction conditions, leading to a lower yield of
the products than was obtained from the borylation of ethers
containing longer alkyl chains. This instability was corroborated
by the product distribution from the borylation of two methyl
positions of n-BuOMe over time. During the first 6 h at 100 °C,
the selectivity for borylation of n-BuOMe was 85:15 in favor of
the product from borylation of the methyl group over the
primary C−H bond of the butyl group. This selectivity is
similar to that for the reaction of n-BuOEt, indicating that the
rates of borylation of methoxy and ethoxy groups are similar to
each other. However, after 10 h, the ratio of the two products
was a lower 62:38, and after 16 h the absolute amount of
D
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Regioselectivity of the Borylation of Diethyl Methyl-
amine. To gain further information on the origin of the
regioselectivity of this iridium-catalyzed borylation, we
conducted the borylation of 3-methylpentane, which has a
structure that is similar to that of diethyl methylamine. The
major product resulted from borylation of the terminal C−H
bond in the ethyl group over borylation of the methyl group to
give two constitutional isomers in a ratio of 20:1 (Et:Me,
Scheme 2). This result suggests that steric hindrance is one
Scheme 2. Borylation of 3-Methylpentane
within the ether and amine were ethyl ether:n-propyl ether =
37:7; ethylamine:n-propylamine = 47:9)
Because the terminal C−H bonds in ethyl ethers and
ethylamines are more reactive than other C−H bonds in ethers
and amines, we determined whether these aliphatic C−H
bonds were even more reactive than aromatic C−H bonds. To
do so, we studied the borylation of ethyl phenyl ether. This
reaction formed the product from borylation of the aromatic
C−H bonds, with an o:m:p ratio of 5:47.5:47.5 (eq 11). This
result shows that the reactivity of a methyl C−H bond β to
oxygen in an ether lies between that of an unactivated alkyl C−
H bond and that of an unactivated aryl C−H bond.
Kinetic Isotope Effects (KIEs). To gain insight into the
origin of the site selectivity, we sought to determine if the C−H
bond-cleavage step was irreversible and controlling the site
selectivity or if the C−H bond-cleavage step was reversible, and
a step after C−H bond cleavage was controlling the selectivity.
To do so, we measured the KIE for borylation of methyl and
ethyl ethers. A primary KIE of 2.9 at 100 °C was observed for
the C−H borylation of cyclohexyl methyl ether and cyclohexyl
methyl ether-d3 conducted in separate vessels. A similar primary
KIE of 2.4 was observed for the C−H borylation of diethyl
ether and diethyl ether-d10 in separate vessels (Scheme 1). This
primary KIE implies that the C−H bond-cleavage step is
irreversible and the step during which site selectivity is
determined.
factor that causes the methyl group to be less reactive than the
ethyl group in diethyl methylamine toward the present iridium-
catalyzed borylation. Other factors that influence the
regioselectivity will be discussed in detail in the section
describing computational studies.
COMPUTATIONAL RESULTS
■
To understand the origins of the selectivity for borylation of
C−H bonds located α and β to the oxygen and nitrogen atoms
of ethers and amines, we evaluated the steps that cleave the C−
H bonds and form the C−B bonds by DFT calculations. We
sought to reveal the origin of the differences in rates for C−H
bond cleavage and C−B bond formation at the α and β
positions.
To do so, we first computed the overall potential energy
surface for the C−H borylation of triethylamine. These
calculations were conducted with the Me4phen-ligated
iridium-trisboryl complex (Figure 1) as the active catalyst.
This complex is the Me4phen analogue of the trisboryl
intermediate in the C−H borylation of arenes catalyzed by
[Ir(OMe)(COD)]2 and di-tert-butylbipyridine.29,30 We pro-
pose that the reaction occurs by a similar boron-assisted
Scheme 1. Kinetic Isotope Effects for Borylation of Methyl
and Ethyl Ethers
Figure 1. Proposed mechanism of the borylation of triethylamine.
E
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Potential energy surface for the borylation of triethylamine.
Figure 3. Structures of the transition state for C−H bond cleavage 3-TS and of the Ir−H intermediate 4. The hydrogen atoms in 3,4,7,8-Me4phen
and Bpin ligands are deleted for clarity. The data in parentheses are the bond order for the corresponding bond.
oxidative addition of the C−H bond and reductive elimination
of the C−B bond that has been computed to occur for the
borylation of arenes.31,32 A combination of product displace-
ment by B2pin2, oxidative addition of the diboron compound,
and reductive elimination of HBpin would regenerate the
catalyst.
Potential Energy Surface for the Borylation of
Triethylamine. We computed the potential energy surface
for the borylation of triethylamine by the mechanism shown in
Figure 1, and this surface is represented in Figure 2. The key
steps in this borylation process are ligand exchange, C−H bond
cleavage, C−B bond formation, and regeneration of the
iridium−trisboryl complex.
A single turnover begins with ligand exchange between the
substrate and the iridium complex 13 containing bound
product (generated by the previous catalytic cycle) to yield
the iridium complex 14 containing bound substrate. To release
the reactive iridium complex 1, the substrate must dissociate
from the iridium center in complex 14 to generate complex 1
and the free substrate. This ligand exchange step is endergonic
by 4.8 kcal/mol.
The key bond-cleaving and bond-forming steps occur after
this initial generation of 16-electron complex 1. In contrast to
the rate-determining step of the rhodium-catalyzed borylation
of alkanes,18 the rate-determining step of the iridium-catalyzed
C−H borylation of the amine is the C−H bond-cleavage step,
not the C−B bond-forming step. The overall activation free
energies of the C−H bond-cleavage (via 3-TS) and C−B bond-
forming (via 6-TS) steps are 35.1 and 28.7 kcal/mol,
respectively.33 The C−H bond-cleavage step is assisted by
the adjacent B1 atom (Figure 3) to generate an Ir(V) species 4,
in which the activated hydrogen atom is partially coordinated to
the same boron atom B1 as in 3-TS. The C−B bond-forming
step generates complex 7 containing bound alkyboronate ester.
To regenerate the active iridium−trisboryl complex, a ligand
exchange between complex 7 and B2pin2 occurs. This step is
endergonic by 5.6 kcal. After this ligand exchange, the oxidative
addition of B2pin2 and the reductive elimination of HBpin
F
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
occur. These steps are facile and have free-energy barriers of
just 6.9 and 2.4 kcal/mol from intermediates 8 and 10,
respectively. After the release of HBpin, a more stable complex
13 forms between the iridium center and the product.
In our prior paper on the borylation of the secondary C−H
bonds in cyclic ethers catalyzed by the same iridium catalyst, we
speculated that the high selectivity for borylation of the
secondary C−H bond β to the oxygen atoms of cyclic ethers
results from the coordination of the Lewis basic oxygen atom of
the cyclic ether to the Lewis acidic boryl ligand to form a five-
membered ring in the transition state.23 However, we had not
obtained any experimental data or conducted computations to
evaluate this hypothesis.
Our current computations reveal a more complex set of
interactions between the substrate and the catalyst that controls
selectivity. As shown in Figure 3, the transition state of the C−
H bond-cleavage step, 3-TS, adopts a structure in which the
iridium center, the C−H bond undergoing cleavage, the Lewis
acidic boron atom B1, and the Lewis basic nitrogen atom in the
substrate lie in nearly the same plane. This coplanarity is
revealed by the dihedral angles of φ(B1,Ir,C,N), φ(B1,Ir,C,H),
and φ(H,Ir,C,N), which are 0.2°, −0.2°, and 0.4°, respectively.
This unique structure alignment allows the Lewis acidic boron
atom B1 to play two important roles in the C−H activation
step.
First, the boron atom B1 interacts with the hydrogen atom of
the activated C−H bond. The distance of the hydrogen atom of
the activated C−H bond and the B1 atom in the boryl ligand is
1.97 Å. According to the natural bond orbital (NBO) analysis,
the bond order between the hydrogen atom and the boron
atom is 0.1265, indicating that the C−H activation step is
assisted by this boryl ligand.34
Figure 4. C−H···O interactions in the transition state for C−H bond
cleavage, 3-TS.
between the strengths of the C−H···O interactions between the
C−H of the Me4phen ligand and the boryl oxygen atoms in the
different transition states appear to affect the selectivity of the
C−H bond-cleavage steps (vide infra).
Second, the Lewis acidic B1 atom in the boryl ligand
interacts with the Lewis basic nitrogen atom of the triethyl-
amine substrate in the transition state 3-TS. The orientation of
the B1 and N atoms in 3-TS suggests that a similar
coordination occurs between the Lewis basic nitrogen atom
of the amine and the Lewis acidic boryl ligand because the
nitrogen atom in triethylamine lies directly above the plane of
the B1 atom in the boryl ligand. The dihedral angle of
φ(B1,Ir,C,N) is only 0.2°. However, this stabilizing effect in the
transition state for C−H bond cleavage is low because the
distance between the B1 and N atoms is long (3.55 Å, with a
bond order of only 0.0079 between N and B). Indeed, NBO
second-order perturbation analysis indicates that the stabilizing
effect from coordination of the amine to the B1 center in the
transition state 3-TS is small (0.6 kcal/mol).
Regioselectivity of the Borylation of Diethyl Methyl-
amine. Because borylation at the terminal C−H bond of the
ethyl group in diethyl methylamine occurred exclusively over
borylation of the methyl group, whereas borylation of the ethyl
group in 3-methylpentane was only 20 times more reactive than
that of the methyl group, we considered that attractive
interactions like those in the transition state computed for
the borylation of triethylamine might account for the
particularly high selectivity for reaction at the ethyl group in
mixed N-alkyl ethylamines. Because the C−H bond-cleavage
step is irreversible, we analyzed the differences in energies and
structures of the transition states for C−H bond cleavage of the
different primary C−H bonds in diethyl methylamine to
understand the regioselectivity of the borylation of mixed
alkylamines.
This alignment of atoms in 3-TS creates a structure with a
series of C−H···O interactions35 between the substrate and the
oxygen atoms in the boryl ligand. The distances between the
hydrogen and the oxygen atoms listed in Figure 4 are all shorter
than the sum of their van der Waals radii (2.72 Å).36 Using
NBO second-order perturbation analysis, we computed the
stabilizing energy provided by these C−H···O interactions
between the triethylamine substrate and the boryl ligands.
These values are 0.7 (C1−H1···O1), 1.0 (C2−H2···O1), and
1.2 kcal/mol (C3−H3···O2). The sum of the energies of these
C−H···O interactions is larger than the energy of the weak
Lewis acid−base interaction mentioned above. In addition to
the C−H···O interactions between the substrates and the boryl
ligands, we identified significant C−H···O interactions between
the Me4phen ligand and the boryl ligands (2.8 kcal/mol for
C4−H4···O2, and 0.7 kcal/mol for C5−H5···O3). Differences
We located four transition states for C−H bond cleavage at
the primary C−H bonds of the methyl and ethyl groups of
diethyl methylamine. The two transition states for activation of
the primary C−H bonds of the methyl group (15-TS-Me-a)
and the primary C−H bonds of the ethyl group (15-TS-Et-a)
with the amine oriented proximal to one of the boryl ligands
are shown at the top of Figure 5. The barrier for C−H bond
cleavage at the ethyl group via 15-TS-Et-a was computed to be
7.3 kcal/mol lower than that for C−H bond cleavage at the
methyl group via 15-TS-Me-a. Significant differences were
identified between the structures of these two transition states.
A weak Lewis acid−base interaction and several C−H···O
interactions between the substrate and the boryl ligands were
observed in transition state 15-TS-Et-a, but neither class of
interaction was found in the computed transition state 15-TS-
Me-a. Moreover, in 15-TS-Me-a, the C−H···O interactions
G
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 5. Transition states for C−H bond cleavage at the terminal C−H bond in methyl and ethyl groups of diethyl methylamine.
between the Me4phen ligand and the two boryl ligands in the
same plane are weaker than those in 15-TS-Et-a.
ligands (4.2 kcal/mol in 15-TS-Me-b vs 3.3 kcal/mol in 15-TS-
Me-a). In contrast to the relative energies of the transition
states for cleavage of the methyl C−H bonds, the computed
barrier for cleavage of the primary C−H bond of the
ethylamino moiety via 15-TS-Et-b is 1.2 kcal/mol higher in
free energy than that via 15-TS-Et-a. Because 15-TS-Et-a has
greater steric interactions, this result further implies that the
weak Lewis acid−base interaction between the nitrogen atom
in the ethylamine moiety and boron atom B1 in the boryl
ligand and the set of C−H···O interactions stabilizes the
transition state for cleavage of the primary C−H bond in the
aminoethyl moiety.
We also located two transition states for cleavage of the
methyl and ethyl C−H bonds of the diethyl methylamine in
which the amine lies in a position that avoids steric hindrance
between the substrate and the boryl ligand. These two
transition states are shown at the bottom of Figure 5 and are
labeled as 15-TS-Me-b and 15-TS-Et-b. The computed barrier
for the C−H bond cleavage at the methyl group via 15-TS-Me-
b is 4.1 kcal/mol lower than that via 15-TS-Me-a. The lower
activation energy for C−H bond cleavage via 15-TS-Me-b
results from less steric hindrance between the methylamino
moiety of the substrate and the boryl ligands, as well as stronger
C−H···O interactions between the Me4phen and the boryl
To assess further the origins of the selectivity for reactions
occurring at different primary C−H bonds of mixed alkyl-
H
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 6. Transition states for cleavage of the terminal C−H bonds in 3-methylpentane.
amines, we calculated the transition states for cleavage of the
primary C−H bonds in 3-methylpentane. The transition state
for cleavage of the C−H bond of 3-methylpentane at the ethyl
group via 16-TS-Et-b (Figure 6, bottom right), which contains
the substrate lying between the Me4phen ligand and the boryl
ligand, is lower in energy than the transition state 16-TS-Et-a
(Figure 6, top right), in which the substrate is positioned above
the plane of the boron in the boryl ligand. These relative
energies are the opposite of the relative energies of the
analogous transition states for cleavage of the primary ethyl C−
H bonds in the amine. Because there is no Lewis acid−base
interaction involving the alkane, these relative energies are
consistent with an influence of the Lewis acid−base interaction
on the energy of the transition state for cleavage of the ethyl
C−H bond of the amine.
Like the relative energies for transition states 16-TS-Et-a and
16-TS-Et-b, the transition state for C−H bond cleavage at the
more hindered methyl group of this alkane via 16-TS-Me-b
(Figure 6, bottom left), in which the substrate lies between the
Me4phen ligand and the boryl ligand, is lower in energy than
the transition state 16-TS-Me-a (Figure 6, top left), in which
the substrate is positioned above the plane of the boron in the
boryl ligand. As a result of the absence of the effect of attractive
Lewis acid−base and C−H···O interactions, the calculated
activation free energy for cleavage of the ethyl C−H bond
(position 2 in Scheme 2, via 16-TS-Et-b) is lower than that for
cleavage of the methyl C−H bond (position 1 in Scheme 2, via
16-TS-Me-b) by only 2.1 kcal/mol. This difference in
activation free energies is consistent with the observation of
two constitutional isomers in a 20:1 ratio (Et:Me) (Scheme 2).
Thus, our computational results analyzing the origin of the
higher reactivity of the terminal C−H bonds at the β position
of the nitrogen atom than that at the α position toward iridium-
catalyzed borylation indicate that both repulsive and attractive
interactions affect the rates and selectivities of the C−H bond
functionalization. In each transition state, there is a steric
repulsion between the boryl ligand and the alkyl groups of the
amine or alkane. However, a detailed comparison of the origins
of selectivity for reaction of an alkane and an alkylamine
suggests that attractive interactions contribute to the difference
in free energy barriers to formation of isomeric products.
In each transition state, there is an attractive C−H···O
interaction between the 2- and 9- hydrogens of the
phenanthroline ligand and the oxygen atom of a boryl group,
and the strength of this interaction varies with the structural
changes in the transition state of the C−H bond-cleavage step.
In the most favorable transition states for C−H bond cleavage
of the ethylamines, a series of attractive C−H···O interactions
between the C−H bonds in the substrate and the oxygen atoms
of the boryl ligands were found by computation. The
I
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
stabilization energies of these C−H···O interactions sum to 2.9
and 2.1 kcal/mol in the transition state for cleavage of the
primary ethyl C−H bonds of triethylamine and diethylmethyl-
amine, respectively. These computed transition states also
contain weak Lewis acid−base interactions between the boron
atom of a boryl group and the nitrogen atom of the amine.
In contrast, these stabilizing C−H···O interactions and weak
Lewis acid−base interactions are not observed in the computed
transition state for cleavage of the methyl C−H bond of diethyl
methylamine. The absence of these interactions contributes to
making the transition state for cleavage of this primary C−H
bond higher in energy than that for cleavage of the primary C−
H bond of the ethyl group. A similar combination of
interactions is likely affecting the rates and selectivities of the
reactions of ethers.
UC Berkeley) for support of this work, Johnson Matthey for
[Ir(COD)OMe]2 and [Ir(COD)Cl]2, and AllyChem for
B2pin2. C.W.L. thanks Abbott Laboratories and the NSF
graduate research fellowship program for predoctoral fellow-
ships.
REFERENCES
■
(1) Hall, D. G. Boronic Acids; Wiley-VCH: Weinheim, 2011; Vol. 2.
(2) Jana, R.; Pathak, T. P.; Sigman, M. S. Chem. Rev. 2011, 111, 1417.
(3) Matteson, D. S. Tetrahedron 1989, 45, 1859.
(4) Pasto, D. J.; Cumbo, C. C. J. Am. Chem. Soc. 1964, 86, 4343.
(5) Brown, H. C.; Sharp, R. L. J. Am. Chem. Soc. 1968, 90, 2915.
(6) Fisher, G. B.; Juarezbrambila, J. J.; Goralski, C. T.; Wipke, W. T.;
Singaram, B. J. Am. Chem. Soc. 1993, 115, 440.
(7) Molander, G. A.; Canturk, B. Org. Lett. 2008, 10, 2135.
(8) Molander, G. A.; Gormisky, P. E.; Sandrock, D. L. J. Org. Chem.
2008, 73, 2052.
CONCLUSION
(9) Raushel, J.; Sandrock, D. L.; Josyula, K. V.; Pakyz, D.; Molander,
G. A. J. Org. Chem. 2011, 76, 2762.
■
We reported the C−H borylation of alkylamines and ethers to
produce aliphatic boron compounds that can be difficult to
produce through traditional methods. These studies show that
the high reactivity observed for borylation of the β-position of
cyclic ethers extends to borylation of acyclic ethers and amines.
The aminomethyl boronate ester product from borylation of N-
methyl groups directly undergoes Suzuki−Miyaura cross-
coupling reactions, while the products from borylation of the
acyclic alkylamines and alkyl ethers were readily converted to
the stable trifluoroborate salts.
(10) Colombel, V.; Rombouts, F.; Oehlrich, D.; Molander, G. A. J.
Org. Chem. 2012, 77, 2966.
(11) Fleury-Bregeot, N.; Presset, M.; Beaumard, F.; Colornbel, V.;
Oehlrich, D.; Rombouts, F.; Molander, G. A. J. Org. Chem. 2012, 77,
10399.
(12) Fleury-Bregeot, N.; Raushel, J.; Sandrock, D. L.; Dreher, S. D.;
Molander, G. A. Chem.Eur. J. 2012, 18, 9564.
(13) Molander, G. A.; Shin, I. Org. Lett. 2012, 14, 4458.
(14) Molander, G. A.; Shin, I. Org. Lett. 2012, 14, 3138.
(15) Molander, G. A.; Vargas, F. Org. Lett. 2007, 9, 203.
(16) Molander, G. A.; Sandrock, D. L. Org. Lett. 2007, 9, 1597.
(17) For a recent palladium-catalyzed directed borylation of primary
C−H bonds, see: Zhang, L.-S.; Chen, G.; Wang, X.; Guo, Q.-Y.;
Zhang, X.-S.; Pan, F.; Chen, K.; Shi, Z.-J. Angew. Chem., Int. Ed. 2014,
53, 3899.
Experimental data and computational studies indicate that
C−H bond cleavage is the rate-determining step of this Ir-
catalyzed borylation reaction of alkyl C−H bonds. Computa-
tional studies indicate that the higher reactivity of C−H bonds
at the terminal position of ethylamines and ethers observed
experimentally results from a transition state structure for
cleavage of the primary ethyl C−H bond in which the
heteroatom lies directly above the boron atom in the boryl
ligand. The weak attractive Lewis acid−base interaction in this
structure, along with attractive C−H···O interactions between
the substrate and the boryl ligand, are likely leading to the faster
rates for borylation of amines and ethers than for borylation of
alkanes. We assert that such attractive interactions will likely be
found to affect the regioselectivity and stereoselectivity of a
range of reactions of substrates containing functional groups as
methods to compute such interactions become used more
widely.
́
(18) Wei, C. S.; Jimenez-Hoyos, C. A.; Videa, M. F.; Hartwig, J. F.;
Hall, M. B. J. Am. Chem. Soc. 2010, 132, 3278.
(19) Chen, H. Y.; Schlecht, S.; Semple, T. C.; Hartwig, J. F. Science
2000, 287, 1995.
(20) Lawrence, J. D.; Takahashi, M.; Bae, C.; Hartwig, J. F. J. Am.
Chem. Soc. 2004, 126, 15334.
(21) Murphy, J. M.; Lawrence, J. D.; Kawamura, K.; Incarvito, C.;
Hartwig, J. F. J. Am. Chem. Soc. 2006, 128, 13684.
(22) For use of this catalyst for the silylation of C−H bonds, see:
Simmons, E. M.; Hartwig, J. F. Nature 2012, 483, 70.
(23) Liskey, C. W.; Hartwig, J. F. J. Am. Chem. Soc. 2012, 134, 12422.
(24) Liskey, C. W.; Hartwig, J. F. J. Am. Chem. Soc. 2013, 135, 3375.
(25) (a) Ohmura, T.; Torigoe, T.; Suginome, M. J. Am. Chem. Soc.
2012, 134, 17416. For recent examples of Ir-catalyzed sp3 C−H
borylations, see: (b) Kawamorita, S.; Murakami, R.; Iwai, T.;
Sawamura, M. J. Am. Chem. Soc. 2013, 135, 2947. (c) Mita, T.;
Ikeda, Y.; Michigami, K.; Sato, Y. Chem. Commum. 2013, 49, 5601.
(d) Kawamorita, S.; Miyazaki, T.; Iwai, T.; Ohmiya, H.; Sawamura, M.
J. Am. Chem. Soc. 2013, 134, 12924.
ASSOCIATED CONTENT
* Supporting Information
■
S
Experimental procedures, spectra for all new compounds, and
coordinates for the computed structures. This material is
available free of charge via the Internet at http://pubs.acs.org.
(26) Campos, K. R. Chem. Soc. Rev. 2007, 36, 1069.
(27) Borylation of nBu2O conducted with the commercially available
iridium precursor [Ir(cod)OMe]2 was also studied. The reaction can
be assembled in the glovebox or outside the glovebox using standard
Schlenk techniques. The reactions assembled in the drybox or outside
the drybox occur in yields that are comparable to each other. In
addition, we conducted the reactions on gram scale using [Ir(cod)
OMe]2 as the catalyst precursor. For details, see the Supporting
Information.
AUTHOR INFORMATION
Corresponding Author
jhartwig@berkeley.edu
■
Author Contributions
§Q.L. and C.W.L. contributed equally.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank the NSF (CHE-01213409 to J.F.H. and CHE-
0840505 to Molecular Graphics and Computation Facility at
■
J
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(28) Yang, C. T.; Zhang, Z. Q.; Tajuddin, H.; Wu, C. C.; Liang, J.;
Liu, J. H.; Fu, Y.; Czyzewska, M.; Steel, P. G.; Marder, T. B.; Liu, L.
Angew. Chem., Int. Ed. 2012, 51, 528.
(29) Boller, T. M.; Murphy, J. M.; Hapke, M.; Ishiyama, T.; Miyaura,
N.; Hartwig, J. F. J. Am. Chem. Soc. 2005, 127, 14263.
(30) Liskey, C. W.; Wei, C. S.; Pahls, D. R.; Hartwig, J. F. Chem.
Commun. 2009, 5603.
(31) Vanchura, B. A.; Preshlock, S. M.; Roosen, P. C.; Kallepalli, V.
A.; Staples, R. J.; Maleczka, R. E.; Singleton, D. A.; Smith, M. R. Chem.
Commun. 2010, 46, 7724.
(32) Tamura, H.; Yamazaki, H.; Sato, H.; Sakaki, S. J. Am. Chem. Soc.
2003, 125, 16114.
(33) The overall free energy of activation for C−H bond-cleavage
(via 3-TS) of triethylamine is calculated to be 35.1 kcal/mol at 393 K.
This value is somewhat higher than the measured experimental
activation bairrier of 31.0 kcal/mol. To assess the effect of the basis
sets on this difference, we computed the activation energy using the
lanl2tz-f basis set for iridium and 6-311++g** basis set for other
atoms. Because our reactants consist of 120 atoms, it was not feasible
to optmize all structures with these higher basis sets. However, with
these basis sets, we computed the single-point energies for C−H
activation of triethylamine and for C−B reductive elimination of the
functionalized product from the structures that had been optimized
with lanl2dz for iridium and 6-31g** for other atoms. The overall
activation free energies of the C−H bond-cleavage (via 3-TS) and C−
B bond-forming (via 6-TS) steps were computed to be 31.8 and 25.7
kcal/mol, respectively. The computed free energy of activation at these
levels is remarkably close to the estimated experimental value.
Moreover, the free energy of activation for the C−H bond cleavage
is higher than that of the C−B bond-forming steps from the
calculation at the higher level. However, the computed difference in
free energy (3.2 kcal/mol) from the single-point calculation with the
higher basis sets agreed less with the 20:1 experimental ratio than did
the calculation with the lower basis set, although the difference was
small (1.1 kcal/mol). For consistency, and because of the small
differences, we have reported the values for the ground state and
activation energies of structures with energies and geometries
optimized using the lanl2dz basis set for iridium and 6-31g** basis
set for the other atoms.
(34) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F.
NBO, version 3.1; University of Wisconsin: Madison, WI, 1996.
(35) (a) Taylor, R.; Kennard, O. J. Am. Chem. Soc. 1982, 104, 5063.
(b) Desiraju, G. Acc. Chem. Res. 1996, 29, 441. (c) Johnston, R. C.;
Cheong, P. H.-Y. Org. Biomol. Chem. 2013, 11, 5057.
(36) Bondi, A. J. Phys. Chem. 1964, 68, 441.
K
dx.doi.org/10.1021/ja503676d | J. Am. Chem. Soc. XXXX, XXX, XXX−XXX