Site-Selective Csp3-Csp/Csp3-Csp2Cross-Coupling Reactions Using Frustrated Lewis Pairs

Article abstract of DOI:10.1021/jacs.1c01622

The donor-acceptor ability of frustrated Lewis pairs (FLPs) has led to widespread applications in organic synthesis. Single electron transfer from a donor Lewis base to an acceptor Lewis acid can generate a frustrated radical pair (FRP) depending on the s

Full text of DOI:10.1021/jacs.1c01622

pubs.acs.org/JACS
Article
3
3
2
Site-Selective Csp −Csp/Csp −Csp Cross-Coupling Reactions Using
Frustrated Lewis Pairs
Ayan Dasgupta, Katarina Stefkova, Rasool Babaahmadi, Brian F. Yates, Niklaas J. Buurma,
Alireza Ariafard, Emma Richards, and Rebecca L. Melen*
Cite This: J. Am. Chem. Soc. 2021, 143, 4451−4464
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: The donor−acceptor ability of frustrated Lewis pairs
(FLPs) has led to widespread applications in organic synthesis. Single
electron transfer from a donor Lewis base to an acceptor Lewis acid
can generate a frustrated radical pair (FRP) depending on the substrate
and energy required (thermal or photochemical) to promote an FLP
3
into an FRP system. Herein, we report the Csp −Csp cross-coupling
reaction of aryl esters with terminal alkynes using the B(C6F5)3/Mes3P
FLP. Significantly, when the 1-ethynyl-4-vinylbenzene substrate was
3
employed, the exclusive formation of Csp −Csp cross-coupled products
was observed. However, when 1-ethynyl-2-vinylbenzene was employed,
3
3
2
solvent-dependent site-selective Csp −Csp or Csp −Csp cross-coupling resulted. The nature of these reaction pathways and their
selectivity has been investigated by extensive electron paramagnetic resonance (EPR) studies, kinetic studies, and density functional
theory (DFT) calculations both to elucidate the mechanism of these coupling reactions and to explain the solvent-dependent site
selectivity.
3
Scheme 1. (A) Previous Work on Metal-Catalyzed Csp −Csp
INTRODUCTION
■
3
2
Cross-Coupling Reactions and FLP-Mediated Csp −Csp
Cross Coupling and (B) This Work on FLP-Mediated,
Frustrated Lewis pairs (FLPs) have garnered much attention
over the last two decades, with numerous FLP systems being
reported in the literature.1 The cooperative reactivity of the
Lewis acidic and basic components has led to a plethora of
different small-molecule activation reactions2 and catalysis.3
Current studies have focused on using FLP systems as
alternatives or complementary systems to the use of
transition-metal catalysts in organic synthesis.4 Recently, new
reactivities of FLPs have been disclosed, indicating that Lewis
acids and bases undergo single electron transfer (SET) events5
depending on the energy required (thermal or photochemical)
to promote the FLP into a frustrated radical pair (FRP)
system. In these instances, an electron is transferred from the
donor Lewis base (LB) to the acceptor Lewis acid (LA) to
generate a reactive FRP. Indeed, we and others have postulated
that such radical reactivity may be taking place in the reactions
of the B(C6F5)3/Mes3P FLP with certain substrates.6 The
radical reactivity of FLPs has the potential to open up new
opportunities for metal-free synthesis. In a previous study of
the B(C6F5)3/Mes3P FLP with diaryl esters and alkenes, we
3
3
2
Solvent-Dependent, Site-Selective Csp −Csp/Csp −Csp
Cross-Coupling Reactions
3
2
observed Csp −Csp coupling reactions. We proposed a radical
mechanism for the reaction based on the observation of
[Ar2CH]• and [Mes3P]•+ in electron paramagnetic resonance
(EPR) studies (Scheme 1A).
Received: February 9, 2021
Published: March 15, 2021
In this current study, we were interested in the reactions of
FLPs with alkynes in the presence of aryl esters (Scheme 1B).
The 1,2-trans-addition of the Lewis acidic and basic
© 2021 The Authors. Published by
American Chemical Society
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464
4451

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
components of FLPs to alkynes is well established7 and has
also been employed in catalytic transformations.8 Interestingly,
when using terminal acetylenes such as phenylacetylene
(PhCCH) with stronger bases such as tBu3P or TMP,
deprotonation occurs instead of 1,2-addition to the alkyne
generating [LBH][PhCCB(C6F5)3] salts.7c In transition-
metal chemistry, the activation of terminal alkynes for cross-
coupling reactions is commonplace in the synthetic chemist’s
toolbox to construct carbon−carbon bonds.9 For example,
palladium- or copper-catalyzed Sonogashira cross-coupling
reactions of terminal alkynes with aryl or alkenyl halides
product, and a stoichiometric or catalytic (10 mol %) amount
of the Lewis acid B(C6F5)3 led to only 22 and 5% isolated
3
yields of the desired Csp −Csp cross-coupled product, 2a
(Table 1, entries 3 and 4). Stoichiometric amounts of both a
3
Lewis acid and Lewis base were required for the Csp −Csp
coupling reaction to attain satisfactory yields of the cross-
coupled products. The B(C6F5)3/Mes3P FLP in toluene at 70
3
°C led to the Csp −Csp cross-coupled product, 2a, being
formed in 54% yield (Table 1, entry 5). The optimum
temperature was 70 °C with both lower (21 °C) and higher
(110 °C) temperatures giving reduced yields (18 and 45%
respectively) (Table 1, entries 6 and 7). More polar THF gave
the highest yield of 83%, with trifluorotoluene (TFT) giving a
71% yield. CH2Cl2 and hexane, on the other hand, showed
poorer or low yields (Table 1, entries 8−11). Interestingly,
other Lewis acid boranes (BF3·OEt2 and BPh3) as well as
Brønsted acids (CF3SO3H) showed no product formation
when combined with Mes3P (Table 1, entries 12−14). Other
basic phosphines including tBu3P, Ph3P and o-tol3P had
complicated reaction mixtures with no or moderate yields of 2a
being formed (Table 1, entries 15−17). Nitrogen Lewis bases
including TMP (2,2,6,6-tetramethylpiperidine), DABCO (1,4-
diazabicyclo(2,2,2)octane), and DMA (4-bromo dimethyl
aniline) led to no product formation (Table 1, entries 18−
20). The optimum conditions for the reaction were therefore
chosen to be the use of the B(C6F5)3/Mes3P FLP in THF at 70
°C for 22−24 h.
have been used for Csp −Csp coupling.10 In this study, we are
3
2
3
interested in the less well reported Csp −Csp coupling reactions.
Typically, palladium11 or earth-abundant metal catalysts such
as iron12 and copper13 are employed for these reactions,
although examples are known with other elements such as
indium14 as well as stoichiometric reactions using Brønsted15
and Lewis acids.16
Herein, we report the high reactivity of frustrated Lewis
3
pairs in selective Csp −Csp coupling reactions between aryl
esters and terminal alkynes or 1-ethynyl-4-vinylbenzene. We
also report solvent-dependent site selectivity when using 1-
3
ethynyl-2-vinylbenzene as a substrate leading to selective Csp −
3
2
Csp or Csp −Csp cross-coupling depending upon the solvent
employed.
RESULTS AND DISCUSSION
■
Reaction Scope. With the optimized reaction conditions
in hand, several aryl esters (1a−l) were tested for the FLP-
Reaction Optimization and Development. Initially, the
3
FLP-mediated Csp −Csp cross-coupling reaction between bis(4-
3
mediated Csp −Csp coupling reaction with various acetylenic
3
compounds (Scheme 2). Diaryl esters bearing electron-
withdrawing/π-releasing (p-F, 1a; p-Cl, 1b), neutral (p-H,
1c), and electron-donating (p-OMe, 1d) groups all worked
well for the reactions when coupled with aryl-substituted
terminal acetylenes with electron-releasing (p-OMe, p-tBu),
neutral (p-H), electron-withdrawing/π-releasing (p-F, p-Cl),
and electron-withdrawing (p-CF3) groups generating products
2a−q in 60−85% yields. Asymmetrical diaryl esters 1e and 1f
Table 1. Optimization of the Reaction Conditions for Csp −
a
Csp Cross-Coupling Reactions
entry
LA
LB
solvent
temp (°C) yield (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
toluene
toluene
toluene
toluene
toluene
toluene
toluene
THF
70
70
70
70
70
21
110
70
70
45
70
70
70
70
70
70
70
70
70
70
0
0
22
5
54
18
45
83
71
38
0
0
0
0
45
0
0
0
0
0
Mes3P
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
BF3·OEt
BPh3
CF3SO3H
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
B(C6F5)3
b
Mes3P
Mes3P
Mes3P
Mes3P
Mes3P
Mes3P
Mes3P
Mes3P
Mes3P
Mes3P
tBu3P
Ph3P
o-tol3P
TMP
DABCO
DMA
3
were also found to undergo the Csp −Csp cross-coupling
reaction, with several alkynes generating C−H-functionalized
products 2r−w in excellent isolated yields (69−80%). We
could also use alkyl/aryl esters containing just one aryl-
stabilizing group. 1g could be cross-coupled with electron-
neutral (phenylacetylene) and electron-releasing (4-ethynyla-
nisole) alkynes to give 2x and 2y albeit in slightly lower yields
of 61 and 65%, respectively. However, when cyclohexyl-
(phenyl)methyl-4-fluorobenzoate (1h) was employed, poor
conversion was observed. Diaryl esters bearing strongly
electron-withdrawing (p-CF3, 1j) groups also failed to react
at all with terminal alkynes. We attribute this to the electron-
deficient nature of the ester which is not Lewis basic enough to
form an adduct with the Lewis acidic borane in the initial step
TFT
CH2Cl2
hexane
toluene
toluene
THF
toluene
THF
THF
THF
THF
THF
1
of the reaction as observed by in situ 11B and H NMR
spectroscopy. The unwillingness of ester 1j to react with
arylacetylenes can also be interpreted from DFT calculations
(see later and SI Figure S180).
a
All of the reactions were carried out on a 0.1 mmol scale, and
reported yields are isolated. All reactions were carried out for 20−22
b
h. 10 mol % B(C6F5)3.
3
After achieving good success for the Csp −Csp cross-coupling
reaction at the benzylic position, we investigated a wider
substrate scope (Scheme 3). To this end, allylic ester (E)-1,3-
diphenylallyl-2,2,2-trifluoroacetate (1k) was used in the C−H
functionalization. To our delight, excellent yields of products
2z−ah (72−89%) were obtained. While benzylic and allenylic
esters worked well, the same was not true for cross-coupling at
fluorophenyl)methyl-4-fluorobenzoate (1a) and phenylacety-
lene was investigated using a range of reaction conditions
(Table 1). As expected, no reaction occurred in the absence of
an FLP (Table 1, entry 1). Reaction with only the Lewis base
component of the FLP (Mes3P) showed no cross-coupled
4452
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
a
3
Scheme 2. Csp −Csp Cross-Coupling Reactions between Esters 1 and Acetylenes
a
All reactions were performed on a 0.1 mmol scale.
4453
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
3
Scheme 3. Csp −Csp Cross-Coupling Reactions between
Scheme 4. Cross-Coupling Reactions between Esters 1a,c,e
and 1-Ethynyl-4-vinylbenzene 4a
a
a
Ester 1k and Acetylenes
a
0.1 mmol scale; reported yields are isolated.
Scheme 5. Cross-Coupling Reactions between Ester 1a and
a
Substrates 4 bearing Internal Alkynes
a
All reactions were performed on a 0.1 mmol scale.
a
Table 3. Solvent-Dependent Site-Selective Studies
a
All reactions were performed on a 0.1 mmol scale.
entry ester ratio of 2:3:5 in THF entry ester ratio of 2:3:5 in TFT
Table 2. Selectivity Reactions between Esters 1a and
Acetylenes/Styrene
1
2
3
4
5
6
7
8
1a
1b
1c
1i
1d
1e
1g
1j
0.2:1:0
0.2:1:0
0.2:1:0
0.4:1:0.1
0:0:0
0.2:1:0
0.5:1:0.1
0:1:1.5
9
1a
1b
1c
1i
1d
1e
1g
1j
1:0:0
1:0:0
1:0:0
1:0:0
1:0:0
1:0:0
1:0:0
0:0:0
a
10
11
12
13
14
15
16
a
1
Ratios were determined from the crude H NMR spectra.
the propargylic position. When the aryl/alkynyl ester 1-phenyl-
3-(trimethylsilyl)prop-2-yn-1-yl 4-fluorobenzoate (1l) was
employed with terminal acetylenes, an inseparable complicated
reaction mixture resulted.
In our previous studies,6 we have shown that reactions of
esters 1 with styrenes in the presence of the same FLP leads to
3
2
Csp −Csp coupled products. We therefore undertook an
experiment to investigate the regioselectivity of the reaction
by reacting the ester starting material with a 1:1 mixture of an
acetylene and a styrene. For this reaction, three outcomes are
entry R1
R2
solvent yield Csp −Csp (%) yield Csp −Csp (%)
3
3
2
1
2
3
4
F
F
H
H
H
H
Ph
TMS
THF
TFT
THF
THF
78 (2a)
72 (2a)
n.d.
n.d.
n.d.
63 (3a)
18 (3a)
3
theoretically possible: (i) formation of the Csp −Csp coupled
58 (2c)
3
2
product, (ii) formation of the Csp −Csp coupled product, or
a
3
3
2
0.1 mmol scale, reported yields are isolated; n.d. = not detected.
(iii) formation of a mixture of Csp −Csp and Csp −Csp coupled
products. Using the optimized reaction conditions, an
4454
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
a
Scheme 6. Products of Solvent-Dependent Site-Selective
Cross-Coupling Reactions
Scheme 7. Possible Reaction Mechanisms
a
a
(A) Diamagnetic pathway. (B) Radical pathway.
The crude 1H NMR spectrum clearly showed a sharp singlet
3
at δ = 5.09 ppm which confirmed the formation of the Csp −
Csp cross-coupled product, 2a, isolated as the major product in
78% yield (Table 2, entry 1). We were not able to detect any
1
characteristic peaks (i.e., a doublet at δ = 4.98 ppm in the H
6
3
2
NMR spectrum) for the Csp −Csp coupled compound. To
investigate the effect of solvent on the reaction, we also
conducted the reaction in TFT, where again 2a was isolated as
the major product albeit with a slightly reduced yield of 72%
3
(Table 2, entry 2). The observation of exclusive Csp −Csp
a
Reactions were carried out on a 0.1 mmol scale, and yields were
coupling is presumably a consequence of the higher reactivity
of the alkyne functionality over the alkene. A similar
competition experiment, using a 1:1 mixture of styrene and
a
b
isolated. Yield in THF. Yield in TFT.
3
2
the internal alkyne diphenylacetylene, gave only Csp −Csp
equimolar mixture of aryl ester 1a, 4-fluorophenylacetylene,
and 4-fluorostyrene were reacted together with B(C6F5)3/
Mes3P (Table 2).
coupling producing 3a in 63% yield (Table 2, entry 3).
Interestingly, TMS-protected alkynes behaved in the same
3
manner as terminal alkynes, predominantly giving the Csp −Csp
4455
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Scheme 8. Mes3P/Diaryl Methylene Cation Equilibria in
Solution
3
crude reaction mixture, which displayed a 1:0.4 ratio of Csp −
3
2
Csp to Csp −Csp cross-coupled products.
To demonstrate the scope for this selectivity, we synthesized
a substrate containing both alkene and alkyne functionalities,
namely, 1-ethynyl-4-vinylbenzene (4a).17 4a was subsequently
reacted with different aryl esters (1a,c,e) in the presence of the
B(C6F5)3/Mes3P FLP. In agreement with the observation
3
above, we observed only the formation of the Csp −Csp
compounds as the major product (2ai−ak; 70−76%) using
the optimized reaction conditions (Scheme 4). In all cases, the
3
2
Csp −Csp coupled product was either not detected or was
observed in <5% yield in both THF and TFT solvents.
As for the intermolecular competition reactions, we also
synthesized internal alkynes in which the acetylenic proton in
4a was replaced by a phenyl or TMS group to explore how this
affected the regioselectivity of the reaction (Scheme 5). Using
the optimized reaction conditions, the reaction of aryl ester 1a
with 1-(phenylethynyl)-4-vinylbenzene (4b) exclusively gave
3
2
the Csp −Csp cross-coupled product in 71% isolated yield from
the reaction with the alkene, whereas when 1a was reacted with
trimethyl{(4-vinylphenyl)ethynyl}silane (4c), reaction at the
alkyne and removal of the TMS group in situ afforded the
3
Csp −Csp cross-coupled product (2ai) in 69% yield, with no
significant reaction at the alkene site observed.
With these results in hand, we further explored the substrate
scope using the 1-ethynyl-2-vinylbenzene (4d).18 4d was
reacted with ester 1a and the B(C6F5)3/Mes3P FLP under the
optimized reaction conditions (THF, 70 °C, 24 h). Contrary
to the reactions above, the reactions were found to be highly
3
2
site-selective for the Csp −Csp coupled product, 3, from
reaction at the alkene functional group. Examining the crude
1H NMR spectrum revealed a 0.2:1:0 ratio of the products
3
3
2
3
(Csp −Csp coupled, 2)/(Csp −Csp coupled, 3)/(Csp −Csp and
3
2
Csp −Csp coupled, 5) (Table 3, entry 1) with isolated yields of
3
2
79% for the Csp −Csp coupled product, 3c, and 16% for the
3
Csp −Csp coupled product, 2al (Scheme 6). Interestingly, when
changing the solvent to TFT, the selectivity was completely
3
reversed, exclusively giving Csp −Csp product 2al from reaction
at the alkyne site.
This was observed by crude H NMR, indicating a 1:0:0
1
Figure 1. CW X-band EPR spectra (T = 298 K) of (A) FLP + ester
1d, (B) FLP + phenylacetylene, (C) FLP + phenylacetylene (black,
experimental; red, simulated), and (D) FLP + phenylacetylene + ester
recorded at (a) t = 0 min and (b) after storage at 77 K for 30 min.
TFT was used as the solvent for all EPR measurements.
ratio of 2/3/5 (Table 3, entry 9) in which 2al could be isolated
in 61% yield (Scheme 6). Remarkably, by simply changing the
solvent we can switch the site selectivity of the reaction.
We next investigated this solvent-dependent site selectivity
for a range of other esters and found the same general trend. In
the following discussion, all reaction product ratios were
coupled product with the loss of the TMS group. Using a 1:1
1
mixture of styrene and trimethyl(phenylethynyl)silane afforded
determined via crude H NMR studies and are listed in Table
3
the Csp −Csp coupled product (2c, 58% isolated) as the major
3, with the corresponding isolated yields for the products
shown in Scheme 6. Initially, we explored the reactions in THF
solvent. When electron-withdrawing 1b (p-Cl) or electron-
neutral symmetrical diaryl esters 1c (p-H) and 1i (p-Me) were
3
2
product with small amounts of the Csp −Csp cross-coupled
product (3a, 18% isolated) formed (Table 2, entry 4). This
1
was also observed by in situ H NMR spectroscopy of the
4456
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 2. DFT-computed reaction pathways for the reaction of I3 with phenylacetylene calculated by SMD/B3LYP-D3/def2-TZVP//SMD/
B3LYP-D3/6-31G(d) in THF. The relative free energies are given in kcal/mol. For this energy profile, structure 1a is set as the reference point as
indicated in Figure S178.
1d, on the other hand, showed no reactivity at all in THF
(Table 3, entry 5). Asymmetrical diaryl ester 1e gave a ratio of
0.2:1:0 for 2/3/5 (Table 3, entry 6), with 2aq and 3h being
isolated in 15 and 73% yields, respectively. Alkyl/aryl ester 1g
also gave the Csp −Csp cross-coupled product as the major
isomer, albeit less selectively, showing a 0.5:1:0.1 ratio of the
three products 2ar/3i/5g in isolated yields of 24% (2ar), 40%
(3i), and 13% (5g) (Table 3, entry 7 and Scheme 6). 1,3-
Diphenylallyl-2,2,2-trifluoroacetate (1k), on the other hand,
showed a preference for the double cross-coupled product,
giving a 0:1:1.5 ratio of 2/3/5 with isolated yields of 22% (3j)
and 41% (5h) (Table 3, entry 8 and Scheme 6).
Table 4. DFT-Computed TS4radical pathway and
TS5cationic pathway for the Reaction of 1a, d, or j with Various
Arylacetylenes (p-XC5H4CCH) Calculated by SMD/
B3LYP-D3/def2-TZVP//SMD/B3LYP-D3/6-31G(d) in
a
3
2
THF
entry
ester (Ar)
X
TS4radical pathway TS5cationic pathway
1
2
3
4
5
6
7
8
9
1a (p-FC6H4)
1a (p-FC6H4)
1a (p-FC6H4)
1a (p-FC6H4)
1a (p-FC6H4)
1d (p-OMeC6H4)
1d (p-OMeC6H4)
1j (p-CF3C6H4)
1j (p-CF3C6H4)
NO2
CF3
H
OMe
NMe2
CF3
OMe
CF3
23.9
26.2
26.3
26.0
27.0
23.7
27.8
24.1
22.2
23.2
20.1
18.6
13.9
8.2
13.0
8.1
25.5
17.5
Subsequently, we repeated the above series of reactions in
TFT, and remarkably, the regioselectivity was altered and the
3
2
selectivity was improved. No Csp −Csp coupled product (3) or
OMe
3
3
2
double Csp −Csp/Csp −Csp coupled product (5) was observed
with any combination of substrates. Rather, a ratio of 1:0:0 of
products 2/3/5 was observed in all cases for esters 1a−e, 1g,
and 1i (Table 3, entries 9−15). This included p-OMe ester 1d
a
The relative free energies are given in kcal/mol.
used, there was a clear preference for reaction at the alkene site
3
3
2
which did not show any product formation in THF. Csp −Csp
leading to Csp −Csp coupled products 3d, 3e, and 3f in ratios
of 0.2:1:0, 0.2:1:0, and 0.4:1:0.1 for 2/3/5, respectively (Table
3, entries 2−4). In all cases, the major and minor regioisomers
cross-coupled products 2al−ar could be isolated in 50−63%
yields (Scheme 6). The only exception was 1,3-diphenylallyl
2,2,2-trifluoroacetate (1k), which gave a complex mixture of
inseparable products when reacted with 1-ethynyl-2-vinyl-
benzene (4d) in TFT, none of which could be identified as 2,
3, or 5 (Table 2, entry 16).
3
2
could be separated. The Csp −Csp cross-coupled products were
isolated in 74% (3d), 71% (3e), and 56% (3f) yields, and the
3
Csp −Csp cross-coupled products were isolated in 15% (2am),
14% (2an), and 18% (2ao) yields. In the case of 1i as the
starting material, double cross-coupled product 5d was
observed in small amounts and could be separated and
isolated in 10% yield (Scheme 6). Electron-rich p-OMe ester
3
Proposed Reaction Mechanism. The Csp −Csp cross-
coupling reaction could be explained by either a single- or a
two-electron pathway.
4457
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 4. Hammett plot for the reaction of I3 with arylacetylenes.
studies have postulated that the B(C6F5)3/Mes3P FLP can
undergo a single electron transfer (SET) process generating
radical ion pair [B(C6F5)3]•−/[Mes3P]•+.21 Slootweg et al.
later postulated that this process is promoted upon irradiation
with visible light (390−500 nm).5b In this pathway (Scheme
7B), we propose that the first step of the reaction is identical to
the diamagnetic pathway whereby B(C6F5)3 activates the diaryl
ester to generate diaryl methylene cation I3 and borate anion
[R1CO2B(C6F5)3]−. The Lewis base then reacts with cation I3,
forming a Lewis acid−base adduct. We propose that this
adduct is in equilibrium with diaryl methylene radical
[Ar2CH]• and phosphonium radical cation [Mes3P]•+, which
are formed from the homolytic cleavage of the C−P bond.
From here, diaryl methylene radical [Ar2CH]• adds to
arylacetylene, leading to a vinylic radical species which, upon
abstraction of a hydrogen atom by [Mes3P]•+, generates the
desired C−C bonded product.
Figure 3. DFT (SMD/B3LYP-D3/def2-TZVP//SMD/B3LYP-D3/6-
31G(d) in THF)-computed energy barrier (TS4 and TS5, Table 3;
entry 1−5) plotted as a function of the Hammett-substituent constant
in a Hammett-style plot. The relative free energies are given in kcal/
mol.
Table 5. Competitive Reaction among 1a, Different
Acetylenes, and the FLP
To understand which pathway is operating, we undertook
extensive electron paramagnetic resonance (EPR), kinetic, and
density functional theory (DFT) studies to understand the
3
reaction mechanism for the Csp −Csp coupling reaction.
entry
X
X′
product ratio
kx/kx′
EPR Studies. The knowledge that the B(C6F5)3/Mes3P
FLP can generate radical species prompted us to undertake an
EPR study to determine if radical species could be observed in
these reactions. As reported previously, no EPR signal could be
detected from the B(C6F5)3/Mes3P FLP in the absence of any
substrates.21 Upon addition of an equimolar ratio of ester 1d to
the FLP in a TFT solution, several EPR signals arising from
multiple paramagnetic species were detected at room temper-
ature (Figure 1A). Upon comparison with previous reports, the
two intense resonance lines in a 1:1 ratio (marked with
asterisks) centered on giso = 2.012 (B ≈ 335 mT) and
separated by a phosphorus hyperfine splitting of aiso(31P) =
670 MHz (23.8 mT) are attributed to the formation of the
[Mes3P]•+ cation.22 A second paramagnetic species was also
detected in this sample, which is characterized by a 1:2:1 triplet
centered on giso = 2.010 and separated by a hyperfine splitting
of 470 MHz (16.7 mT). This EPR profile must originate from
two identical I = 1/2 nuclei, in this case associated with two
equivalent 31P nuclei. The signal is therefore tentatively
assigned to the formation of a [(P(Mes)n=2,3)2]•+ dimer
formed from the association of excess Mes3P with radical
cation [Mes3P]•+ as observed previously.23 This assignment is
1
2
3
OMe
CF3
OMe
H
H
CF3
(2d:2c) 14.3:1
(2b:2c) 0.12:1
(2d:2b) 1: < 0.05
21.9
0.082
First, a diamagnetic pathway could operate (Scheme 7A) in
which the Lewis acid component of the FLP activates the ester
carbonyl atom, leading to the formation of diaryl methylene
cation I3 and the borate anion [R1CO2B(C6F5)3]−. This was
observed in our previous studies when B(C6F5)3 was added to
the diaryl ester in the presence of a nucleophile to trap the
resultant carbocation.19 The diaryl methylene cation and the
Lewis basic component of the FLP can then undergo a 1,2-FLP
addition to the alkyne similar to other FLP 1,2-additions.20
Finally, the elimination of [Mes3PH]+ leads to the C−C-
bonded product and salt [R1CO2B(C6F5)3][HPMes3] (R1 = p-
FC6H4 or CF3). This can be observed in the reaction by
multinuclear NMR spectroscopy showing a clear 1JPH coupling
of 479.5 Hz.
Alternatively, a radical pathway could operate (Scheme 7B),
which may explain the necessity for using Mes3P as a Lewis
base rather than other phosphine or nitrogen bases. Previous
4458
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 5. DFT-computed reaction pathways for the reaction of I3 with 1-ethynyl-4-vinylbenzene (4a) and Mes3P calculated using the SMD/
B3LYP-D3/def2-TZVP//SMD/B3LYP-D3/6-31G(d) level of theory in THF.
supported by our observation that the relative ratios of the
EPR signals of [Mes3P]•+/[(P(Mes)n=2,3)2]•+ are inter-related
(i.e., when a large signal intensity of the monomer is observed,
only a trace of the corresponding dimer is detected). The
varying ratio of these EPR signals under different reaction
conditions demonstrates the conversion between monomer
and dimer via the reaction of the monomer with a second
molecule of phosphine to yield the dimer radical cation, as
previously observed for a series of phosphines,24 and possibly
also the varying stabilities of the two radical species. The EPR
spectrum of [(PMes2)2]•+ has previously been reported,25
characterized by giso = 2.006 and aiso(31P) = 474 MHz (17
mT), and it is noted that literature examples of phosphine
the formation of the carbon-based bismethoxy-diphenyl-
methylene radical formed upon C−O bond cleavage was
observed in this case, perhaps due to the instability of the
radical species.
We then probed the room-temperature EPR spectrum of the
FLP in the presence of phenylacetylene (Figure 1B,C). Under
these experimental conditions, we postulated that another
possible mechanism could be the abstraction of the terminal
hydrogen atom of the acetylene by the [Mes3P]•+ radical
cation to form the diamagnetic [Mes3PH]+ cation and the
corresponding phenylacetylene radical. As can be seen from
the wide field scanning range in Figure 1B, no evidence of
monomer [Mes3P]•+ or dimer [P(Mes)n=2,3 2
]
was observed
•+
•+
in this solution, suggesting the formation of the diamagnetic
[Mes3PH]+ cation. However, no EPR evidence for the
generation of the phenylacetylene radical was obtained. It is
noted in previous literature studies that the terminal
phenylacetylene radical is inherently unstable and is typically
observed only via EPR spectroscopy under controlled
conditions, such as neat liquids sealed under vacuum, or via
matrix isolation methods.27 Notably, the remaining boron
component of the FLP is not involved in the above hydrogen
atom abstraction from the alkyne and may be expected to
dimer cation radicals of divalent (R2P)2 and trivalent
(R3P)2 systems yield very similar EPR spectra,26 dominated
•+
by the 1:2:1 phosphorus hyperfine splitting.
The corresponding anisotropic EPR spectra of the
[Mes3P]•+ monomer and [P(Mes)n=2,3 2
]
dimer species in
•+
frozen solutions are shown in the SI (Figure S176), from which
the principal values of the g and A tensors were determined.
The spin Hamiltonian parameters for all of the paramagnetic
species detected in this work are listed in the SI (Table S1).
Importantly, contrary to our previous reports,6 no evidence for
4459
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
experimental and simulated data, thereby this signal is
tentatively attributed to a carbon-based radical, perhaps
indicating that the reaction could be occurring through a
radical mechanism.
DFT and Kinetic Studies. To examine the contrasting
reaction pathways and to explain the experimental and EPR
observations, we undertook a thorough DFT investigation of
all potential reaction pathways. Calculations were performed at
the SMD/B3LYP-D3/def2-TZVP//SMD/B3LYP-D3/6-
31G(d) level of theory in THF and toluene solvent to
examine the origin of the products. As previously reported by
Slootweg et al., we found that the formation of the frustrated
radical ion pair from the FLP is energetically unfavorable by
35.7 kcal/mol. This corroborates the observation that we and
others6 fail to see any EPR signal in solutions of B(C6F5)3 and
Mes3P. Very recently, Slootweg et al.5a showed that the
coordination of B(C6F5)3 to the diaryl ester increases the
electron affinity of the substrate, and the energy required for
SET from Mes3P to the methylene carbon atom is 40.0 kcal/
mol.5a This is still quite large, and our calculations have
revealed that, regardless of whether a diamagnetic or
paramagnetic reaction pathway is ultimately operative, the
first step in the reaction is the same: B(C6F5)3 activation of the
ester to generate diaryl methylene cation I3 (Scheme 8) and
the borate anion [R1CO2B(C6F5)3]− with an activation energy
of 10.7 kcal/mol. (See SI Figure S178 for the free-energy
profile.) The energies of I3 can be noticeably varied by
changing the substitution at the para position of the aryl esters.
Electron-withdrawing (p-CF3, 1j), electron-withdrawing/π-
releasing (p-F, 1a), and electron-releasing (p-OMe, 1d)
showed different energies for I3 (SI Figure S179−180). As
expected, I31d (−9.4 kcal/mol) is energetically more favorable
than I31j (13.5 kcal/mol) due to the electron-releasing
substituents (p-OMe). Strongly electron-withdrawing substitu-
ents (p-CF3) conversely make I3 formation thermodynamically
less favorable. I3 is then the branching point for the single- and
two-electron pathways (Figure 2). The combination of diaryl
methylene cation I3 with Mes3P can lead to three possible
species in solution (Scheme 8): (i) the frustrated Lewis pair
(uncoordinated I3 + Mes3P), (ii) the Lewis acid−base adduct
(I4), and (iii) the frustrated radical pair (FRP, I5). The energy
difference and reaction barriers between these species are very
low; therefore, it is likely that all three scenarios exist in
equilibrium under the reaction conditions. This supports the
EPR data which shows the formation of [Mes3P]•+ and the
[(P(Mesn=2,3)2]•+ dimer in the reaction of the ester with the
Mes3P/B(C6F5)3 FLP.
Figure 6. DFT calculations of the site selectivity of I3 with 1-ethynyl-
2-vinylbenzene 4d, calculated at the SMD/B3LYP-D3/def2-TZVP//
SMD/B3LYP-D3/6-31G(d) level in THF and toluene.
remain in solution. The intense multiplet signal observed,
reproduced in Figure 1C across a narrow field range, is
therefore assigned to the boron radical anion, [B(C6F5)3]•−. A
satisfactory simulation of the experimental data was achieved
using giso = 2.0114 and incorporating a single boron nucleus,
aiso(10,11B) = 27 MHz, two sets of six equivalent fluorine nuclei
from the ortho and meta positions, with aiso(19F)ortho = 18.28
MHz and aiso(19F)meta = 3 MHz, and three para fluorine nuclei
with aiso(19F)para = 20.2 MHz, which agrees well with previous
literature reports of this species.28 The corresponding
anisotropic spectrum for this sample was unfortunately not
resolved due to a poor-quality glass of the frozen solvent,
thereby preventing the determination of the complete
anisotropic spin Hamiltonian parameters for this radical anion.
Having determined the reactivity of the FLP to the
individual substrates, the EPR spectrum of a full reaction
mixture containing equimolar amounts of Mes3P, B(C6F5)3,
ester, and phenylacetylene was recorded (Figure 1Da). As can
easily be seen, evidence of the [B(C6F5)3]•− radical anion is
clearly observed, but there are no signals attributed to
monomer or dimer phosphorus radicals present. However, it
is noticed that there is additional intensity superimposed in the
center of this signal (giso ≈ 2.001) which must originate from a
second paramagnetic species not previously observed. Upon
storage of this sample at 77 K for 30 min, the signal intensity
originating from the [B(C6F5)3]•− radical anion was lost
completely, leaving only a narrow resonance (Figure 1Db).
Unfortunately, the short lifetime of this radical in solution
prevented full resolution of the hyperfine coupling, but the
narrow spectral width arising from only small hyperfine
couplings is an indication of a carbon-based radical rather
than a boron or phosphorus species (upon consideration of the
theoretical isotropic hyperfine a0 values a0(10B) = 30.43 mT,
a0(11B) = 90.88 mT, and a0(31P) = 474.79 mT). The
experimental spectrum is reproduced again in the SI (Figure
S177) alongside simulations of the styrene and phenyl-
acetylene radicals, using previously reported literature values.29
Gratifyingly, there is reasonable agreement between the
Although the carbon-based radical could not be observed in
this case, we have observed a weak carbon-based radical EPR
signal when reacting other esters with the Mes3P/B(C6F5)3
FLP in the absence of irradiation or heat.6 We then
investigated the addition of the cation (I3) or radical (I6) to
the alkyne. Although both pathways are feasible under the
reaction conditions, the cationic pathway was lower in
activation energy than the radical pathway by about 26.3−
18.6 = 7.7 kcal/mol. In the case of the diamagnetic pathway,
the addition of the I3 cation to the alkyne generates I8 via
TS5. The resulting cation in I8 is highly reactive and is rapidly
trapped by the Lewis base Mes3P generating I9. Finally, anti-
elimination of [Mes3PH]+ generates the cross-coupled
compound and phosphonium borate salt as the final products.
The Lewis base employed is very important for the reaction
to occur as seen in the screening studies. First, the ability to
4460
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
form an FLP (or weak adduct) with both the Lewis acid
borane and the (di)aryl methylene cation (I3) is critical, as
other strong, less hindered Lewis acids such as BF3 do not
work in the reaction. Likewise, smaller phosphines or amines
tend to coordinate more strongly to the carbocation (I3). In
addition to being able to form an FLP, the base also functions
to trap reactive I8. Indeed, one could possibly conceive that
the reaction could proceed with the Lewis acid only, whereby
compared to the TS5cationic pathway energy barrier (range 8.1−
25.5 kcal/mol) when changing the substituent on the
phenylacetylene (Table 4, entries 6−9). As with ester 1a,
both esters disclosed a larger energy difference between the
two pathways when reacted with acetylenic compounds
bearing an electron-donating group (p-OMe). Likewise, a
much smaller energy difference was noted for both esters when
reacted with acetylenic compounds bearing an electron-
withdrawing group (p-CF3). Interestingly, DFT studies showed
that for the case of the reaction of ester 1j with 1-ethynyl-4-
(trifluoromethyl), the radical pathway is slightly energetically
more favorable compared with the cationic pathway (Table 4,
entry 8). These results suggest that for electron-withdrawing
arylacetylenes both paramagnetic and diamagnetic mechanisms
are potentially possible, whereas for electron-rich arylacety-
lenes a purely diamagnetic pathway is operative.
The key difference in the two mechanisms involves the
reaction of either a cationic intermediate or a radical
intermediate with the arylacetylene, generating a new cationic
or radical species. Whether this new intermediate is a cationic
or a neutral radical species can be probed using a Hammett
plot (cf. computational studies) based on substituted
arylacetylenes. To gain this insight into the reaction pathway
and substituent effects also from experimental evidence, we
examined competition reactions among the FLP, aryl ester 1a,
and arylacetylenes p-XC6H4CCH bearing electron-with-
drawing, electron-neutral, and electron-releasing groups. The
Hammett plot requires relative rate constants for the reaction
of different substituted alkynes that were obtained using a
series of competition experiments. Initial competition experi-
ments in the presence of 1.5 equiv of five arylacetylenes p-
XC6H4CCH (X = CF3, F, Cl, H, and OMe) were
unsuccessful. The excess arylacetylene present in the reaction
mixture destroyed the efficacy of the FLP system, producing a
complicated reaction mixture which was not suitable for in situ
NMR analysis. We therefore carried out three binary
competition experiments with two alkynes being present at
1.5 equiv each (Table 5).
−
R1CO2 accepts the proton in the last step (I8 → I10 →
product, Figure 2). However, DFT calculations showed that
this pathway was less favorable thermodynamically, and
experimentally this reaction showed only a 22% yield with
many side products formed.
We were curious to know whether both the paramagnetic
and diamagnetic pathways are operative in parallel or if a
diamagnetic mechanism is purely responsible for the product
formation for all alkynes and the radicals observed from EPR
studies were simply off-pathway intermediates. To establish
this, we undertook further DFT calculations to investigate the
effect of electron-withdrawing, electron-donating, and electron-
neutral substituents on the ester (1) and the alkyne on the
activation barrier for the reaction.
The energy barriers TS4radical pathway and TS5cationic pathway
were calculated (Table 4, see SI Figure S181). Initially we
varied the electronic properties of the alkyne using p-
XC6H4CCH (X = NO2, CF3, H, OMe, NMe2) with ester
1a. As evidence from the DFT calculations, changing the
substitution at the para position of the arylacetylene did not
make significant difference for TS4radical pathway in their
respective energy barrier (23.9−27.0 kcal/mol) (Table 4,
entries 1−5). However, the energy barrier for TS5cationic pathway
changed dramatically (8.2−23.2 kcal/mol). When electron-
withdrawing groups (p-NO2 and p-CF3) on the arylacetylene
were employed, the differences between TS4
and
radical pathway
TS5
are 0.7 and 6.1 kcal/mol (Table 4, entries 1
cationic pathway
and 2). Electronically neutral phenylacetylene exhibits a
TS4radical pathway → TS5cationic pathway difference of 7.7 kcal/mol
(Table 3, entry 3). Electron-donating groups such as methoxy
(TS4radical pathway − TS5cationic pathway = 12.1 kcal/mol) and N,N-
dimethylamine (TS4radical pathway − TS5cationic pathway = 18.8 kcal/
mol), on the other hand, showed a significant energy difference
(Table 4, entries 4 and 5). These observations can be seen in
Figure 3, which shows that for TS4 there is a negligible change
in the energy barrier when changing the electronic properties
of the acetylenic substrate. This is in agreement with little
charge formation on the reaction center in the radical
mechanism. Conversely, for TS5, there is a strong positive
correlation between the TS5 energy barrier and the substituent
σp constant. This is expected for the cationic pathway because
a developing positive charge adjacent to the substituted phenyl
ring will be stabilized by electron-donating groups (e.g., p-
NMe2 or p-OMe) on the acetylene. As can be seen in Figure 3,
TS5 is lower than TS4 for all substituents explored, although
the difference becomes small for strongly electron-withdrawing
groups.
We subsequently computed the energy barrier for the two
transition states by varying the electronic properties of the aryl
ester using electron-donating (p-OMe, 1d) and electron-
withdrawing (p-CF3, 1j) esters with electron-deficient (1-
ethynyl-4-(trifluoromethyl)benzene) and electron-rich (1-
ethynyl-4-methoxybenzene) acetylenes (Table 4, entries 6−
9). For both esters, a smaller change in the TS4radical pathway
energy barrier was observed (range 22.2−27.8 kcal/mol)
Using the optimized reaction conditions, three parallel
reactions were carried out in which equimolar mixtures of (a)
1-ethynyl-4-methoxybenzene and phenylacetylene, (b) 1-
ethynyl-4-(trifluoromethyl)benzene and phenylacetylene, and
(c) 1-ethynyl-4-methoxybenzene and 1-ethynyl-4-(trifluoro-
methyl)benzene were reacted with 1 equiv of ester 1a. The
3
ratios of Csp −Csp cross-coupled products (2d/2c, 2b/2c, and
2d/2b) were determined from the crude reaction mixture
using 19F NMR spectroscopy (Table 5, see SI, Figure S175).
For entries 1 and 2 in Table 5, the relative integrals for the
products were used to calculate the remaining equivalents for
the alkynes after reaction. Using the approach developed by
Ingold and Shaw30 and proposed for one-pot Hammett plots
by Harper and co-workers,31 we obtained relative rate
constants kx/kx′ for the reaction of differently substituted
arylacetylenes with reaction intermediate I3 or its equilibrium
species. Entry 3 confirms that in the competition between 1a
and 1-ethynyl-4-(trifluoromethyl)benzene/1-ethynyl-4-
methoxybenzene, product 2b is undetectable, in agreement
with the >200-fold difference in rate constants deduced from
entries 1 and 2. The relative rate constants in Table 5 are
normalized with respect to the unsubstituted alkyne and can
therefore be used directly to construct a Hammett plot (using
Hammett substituent constants from ref 32). The Hammett
plot (Figure 4) shows a clearly negative slope of −6.6 1.7,
4461
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
suggesting the formation of a positive charge on the new
intermediate, which is indicative of the cationic reaction
mechanism being operative. The Hammett plot is also in
agreement with the Hammett-style plot constructed using the
computational data (Figure 3) for the cationic mechanism.
The slope of the Hammett-style plot for TS5 (Figure 3) is 8.9
1.2 kcal mol−1. At 70 °C, this corresponds to a Hammett ρ
value of 5.7 0.8. The computational data is thus in excellent
agreement with the experimental findings.
Curtin−Hammett principle, the reaction proceeds predom-
inantly via TS5 from rapidly equilibrating I3, I4, I5, and I6.
These rapidly equilibrating species in solution are supported by
the observation of radical species of varying stabilities and
lifetimes in the reaction mixture. Thus, radical species are
formed in the reaction but are not making a substantial
contribution on the reaction pathway to the product, with the
possible exception of arylacetylenes with strongly electron-
withdrawing (e.g., p-NO2, p-CF3) substituents. These obser-
vations will be of importance when designing future reactions
that can switch between one- and two-electron pathways
depending upon the substrate. Moreover, we observed high
site selectivity when ethynyl vinylbenzene substrates were
employed in these reactions. 1-Ethynyl-4-vinylbenzene sub-
strates reacted only at the alkyne site, but 1-ethynyl-2-
vinylbenzene substrates showed high selectivities depending
upon the polarity of the solvent. For 1-ethynyl-2-vinylbenzene
Finally, we turned our attention to explaining the
regioselectivity with compound 4a. The DFT-computed
(SMD/B3LYP-D3/def2-TZVP//SMD/B3LYP-D3/6-
31G(d)) reaction pathways for the reaction of I3 with 1-
ethynyl-4-vinylbenzene (4a) and Mes3P in THF reveal that the
cationic pathway is energetically more favorable (Figure 5).
After the generation of I3, reaction at either the alkyne or
3
3
2
alkene site affords the corresponding Csp −Csp or Csp −Csp
3
cross-coupled product. Although the transition-state energies
in THF, Csp −Csp coupling was observed, resulting in alkene
3
for the I3-alkyne adduct (14.4 kcal/mol) and I3-alkene adduct
functionalization, whereas in toluene or TFT exclusive Csp −
3
(14.3 kcal/mol) are very similar, the Csp −Csp cross-coupled
Csp coupling and alkyne functionalization resulted. The
contrasting selectivity was explained by DFT and computed
transition states in the differing solvents. FLP-mediated C−C
bond-forming reactions are still relatively new, but there is no
doubt that advances will continue to be made in this area. This
reported methodology can be utilized to generate compounds
that can be subsequently employed for the synthesis of useful
novel natural products where metal-free synthesis is highly
desirable for avoiding metal toxicities.15,33
products are thermodynamically more stable (−1.4 kcal/mol)
2
2
than the Csp −Csp cross-coupled product (5.7 kcal/mol),
3
explaining why only the Csp −Csp cross-coupled product is
observed for 1-ethynyl-4-vinylbenzene (4a) (Figure 5). To
confirm this, we also executed single-point benchmark
calculations for this transition state with a different method
and solvent system (SI, Table S2), which showed results
similar to those indicated in Figure 5.
The alternating site selectivity when using 1-ethynyl-2-
vinylbenzene (4d) in differing solvents can also be highlighted
in DFT studies. For the calculations, the mechanism was
studied by utilizing two different solvents (toluene and THF).
Experimentally, both toluene and TFT solvents lead to
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c01622.
■
sı
3
preferential Csp −Csp coupling. DFT calculations for the
Detailed experimental procedures, NMR spectra, EPR
data, kinetic data, and DFT data (PDF)
reaction of I3 with 4d showed that the transition-state energy
for the addition of the diaryl methylene cation to the alkene or
alkyne varies depending upon the solvent (Figure 6). In
toluene solvent, the energy barrier for the addition of I3 to the
more nucleophilic alkyne was 3.7 kcal/mol lower in energy
than the transition state for addition to the alkene (11.7 versus
15.4 kcal/mol). The converse was true for reactions in THF,
whereby the addition of I3 to the alkene was 0.5 kcal/mol
lower in energy than the energy barrier for addition to the
alkyne (14.7 versus 15.2 kcal/mol). We attribute this to the
higher dipole moment of the calculated transition-state
structure of TS3a, which can be stabilized better by the
THF molecule. These results also explain why, in THF, the
AUTHOR INFORMATION
Corresponding Author
■
Rebecca L. Melen − Cardiff Catalysis Institute, School of
Chemistry, Cardiff University, Cardiff CF10 3AT, Cymru/
Wales, United Kingdom; orcid.org/0000-0003-3142-
2831; Email: MelenR@cardiff.ac.uk
Authors
Ayan Dasgupta − Cardiff Catalysis Institute, School of
Chemistry, Cardiff University, Cardiff CF10 3AT, Cymru/
Wales, United Kingdom
Katarina Stefkova − Cardiff Catalysis Institute, School of
Chemistry, Cardiff University, Cardiff CF10 3AT, Cymru/
Wales, United Kingdom
Rasool Babaahmadi − School of Natural Sciences-Chemistry,
University of Tasmania, Hobart, Tasmania 7001, Australia;
orcid.org/0000-0002-2580-6217
Brian F. Yates − School of Natural Sciences-Chemistry,
University of Tasmania, Hobart, Tasmania 7001, Australia
Niklaas J. Buurma − Cardiff Catalysis Institute, School of
Chemistry, Cardiff University, Cardiff CF10 3AT, Cymru/
Wales, United Kingdom
3
reactions were less selective, leading to a mixture of Csp −Csp
3
2
and Csp −Csp products due to the small energy difference
between the two pathways. The reactions in a solvent such as
3
toluene (or TFT) are more selective toward Csp −Csp coupling
due to the larger energy difference between the two pathways.
CONCLUSIONS
We have demonstrated new reactivities of FLPs in the
functionalization of terminal alkynes through Csp −Csp
■
3
coupling reactions with aryl esters. DFT studies found that a
diamagnetic pathway was most likely, although a low-energy
single-electron pathway could operate to some extent. In
particular, DFT studies indicate that the combination of the
Mes3P/diaryl methylene cation led to three species of similar
energy in solution: the FLP (I3), the Lewis acid−base adduct
(I4), and the frustrated radical pair (I5). According to the
Alireza Ariafard − School of Natural Sciences-Chemistry,
University of Tasmania, Hobart, Tasmania 7001, Australia
Emma Richards − Cardiff Catalysis Institute, School of
Chemistry, Cardiff University, Cardiff CF10 3AT, Cymru/
4462
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
516. (e) Shi, L.; Zhou, Y.-G. Enantioselective metal-free hydro-
genation catalyzed bychiral frustrated Lewis pairs. ChemCatChem
2015, 7, 54−56.
Wales, United Kingdom; orcid.org/0000-0001-6691-
2377
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c01622
(5) (a) Holtrop, F.; Jupp, A. R.; Kooij, B. J.; van Leest, N. P.; de
Bruin, B.; Slootweg, J. C. Single-electron transfer in frustrated Lewis
pair chemistry. Angew. Chem., Int. Ed. 2020, 59, 22210−22216.
(b) Holtrop, F.; Jupp, A. R.; van Leest, N. P.; Dominguez, M. P.;
Williams, R.; Brouwer, M. A. M.; de Bruin, B.; Ehlers, A. W.;
Slootweg, J. C. Photoinduced and thermal single-electron transfer to
generate radicals from frustrated Lewis pairs. Chem. - Eur. J. 2020, 26,
9005−9011. (c) Liu, L. L.; Stephan, D. W. Radicals derived from
Lewis acid/base pairs. Chem. Soc. Rev. 2019, 48, 3454−3463.
(d) Schilter, D. Frustration leads to radical behaviour. Nat. Rev.
Chem. 2018, 2, 255. (e) Hamilton, H. B.; Wass, D. F. How important
are radical mechanisms in frustrated Lewis pair chemistry? Chem.
2017, 3, 198−210. (f) Ménard, G.; Hatnean, J. A.; Cowley, H. J.;
Lough, A. J.; Rawson, J. M.; Stephan, D. W. C−H Bond activation by
radical ion pairs derived from R3P/Al(C6F5)3 frustrated Lewis pairs
and N2O. J. Am. Chem. Soc. 2013, 135, 6446−6449. Also see
(g) Warren, T. H.; Erker, G. Frustrated Lewis Pairs II: Expanding the
Scope; Topics in Current Chemistry 334; Springer: Berlin, 2013.
(h) Piers, W. E.; Marwitz, A. J. V.; Mercier, L. G. Mechanistic Aspects
of Bond Activation with Perfluoroarylboranes. Inorg. Chem. 2011, 50,
12252−12262. (I) Aramaki, Y.; Imaizumi, N.; Hotta, M.; Kumagai, J.;
Ooi, T. Exploiting single-electron transfer in Lewis pairs for catalytic
bond-forming reactions. Chem. Sci. 2020, 11, 4305−4311.
Funding
A.D., K.S., and R.L.M. acknowledge the EPSRC for an Early
Career Fellowship for funding (EP/R026912/1). A.D., E.R.,
and R.L.M thank the Leverhulme Trust for a research project
grant (RPG-2020-016). A.A., B.F.Y., and R.B. thank the
Australian Research Council (ARC) for project funding
(DP180100904) and the Australian National Computational
Infrastructure and the University of Tasmania for the generous
allocation of computing time. Information about the data that
underpins the results presented in this article, including how to
access them, can be found in the Cardiff University data
catalogue. This can be found at http://doi.org/10.17035/
d.2021.0130087985.
Notes
The authors declare no competing financial interest.
REFERENCES
■
(1) For the first report, see (a) Welch, G. C.; Juan, R. R. S.; Masuda,
J. D.; Stephan, D. W. Reversible, metal-free hydrogen activation.
Science 2006, 314, 1124−1126. For reviews, see (b) Jupp, A. R.;
Stephan, D. W. New directions for frustrated Lewis pair chemistry.
Trends in Chemistry 2019, 1, 35−48. (c) Scott, D. J.; Fuchter, M. J.;
Ashley, A. E. Designing effective ‘frustrated Lewis pair’ hydrogenation
catalysts. Chem. Soc. Rev. 2017, 46, 5689−5700. (d) Stephan, D. W.
The broadening reach of frustrated Lewis pair chemistry. Science
2016, 354, aaf7229. (e) Stephan, D. W. Frustrated Lewis pairs. J. Am.
Chem. Soc. 2015, 137, 10018−10032. (f) Stephan, D. W.; Erker, G.
Frustrated Lewis pair chemistry: development and perspectives.
Angew. Chem., Int. Ed. 2015, 54, 6400−6441. (g) Stephan, D. W.
Frustrated Lewis pairs: from concept to catalysis. Acc. Chem. Res.
2015, 48, 306−316.
(2) For selected reviews, see (a) Lam, J.; Szkop, K. M.; Mosaferia,
E.; Stephan, D. W. FLP catalysis: main group hydrogenations of
organic unsaturated substrates. Chem. Soc. Rev. 2019, 48, 3592−3612.
(b) Power, P. Main-group elements as transition metals. Nature 2010,
463, 171−177. (c) Stephan, D. W.; Erker, G. Frustrated Lewis pairs:
Metal-free hydrogen activation and more. Angew. Chem., Int. Ed. 2010,
49, 46−76. (d) Stephan, D. W. Frustrated Lewis pairs: a new strategy
to small molecule activation and hydrogenation catalysis. Dalton
Trans 2009, 2009, 3129−3136. Also see (e) Wilkins, L. C.; Melen, R.
L. Small Molecule Activation with Frustrated Lewis Pairs.
Encyclopedia of Inorganic and Bioinorganic Chemistry; John Wiley &
Sons, Ltd., 2017; pp 1−24.
(6) Soltani, Y.; Dasgupta, A.; Gazis, T. A.; Ould, D. M. C.; Richards,
E.; Slater, B.; Stefkova, K.; Vladimirov, V. Y.; Wilkins, L. C.; Willcox,
D.; Melen, R. L. Radical reactivity of frustrated Lewis pairs with diaryl
esters. Cell Reports Physical Science 2020, 1, 100016.
(7) (a) Fasano, V.; Curless, L. D.; Radcliffe, J. E.; Ingleson, M. J.
Frustrated Lewis pair mediated 1,2-hydrocarbation of alkynes. Angew.
Chem., Int. Ed. 2017, 56, 9202−9206. (b) Yanagisawa, T.; Mizuhata,
Y.; Tokitoh, N. Additive-free conversion of internal alkynes by
phosphanylalumanes: Production of phosphorus/Aluminum frus-
trated Lewis pairs. ChemPlusChem 2020, 85, 933−942. (c) Jiang,
C.; Blacque, O.; Berke, H. Activation of terminal alkynes by frustrated
Lewis pairs. Organometallics 2010, 29, 125−133. (d) Dureen, M. A.;
Stephan, D. W. Terminal alkyne activation by frustrated and classical
Lewis acid/phosphine pairs. J. Am. Chem. Soc. 2009, 131, 8396−8397.
(8) Chernichenko, K.; Madarász, A.; Pápai, I.; Nieger, M.; Leskela,
M.; Repo, T. A frustrated-Lewis-pair approach to catalytic reduction
of alkynes to cis-alkenes. Nat. Chem. 2013, 5, 718−723.
(9) For examples, see (a) Beattie, J. W.; Wang, C.; Zhang, H.;
Krogman, J. P.; Foxman, B. M.; Thomas, C. M. Dimerization of
terminal alkynes promoted by a heterobimetallic Zr/Co complex.
Dalton Trans 2020, 49, 2407−2411. (b) Sivaguru, P.; Cao, S.; Babu,
K. R.; Bi, X. Silver-catalyzed activation of terminal alkynes for
synthesizing nitrogen-containing molecules. Acc. Chem. Res. 2020, 53,
662−675. (c) Ansell, M. B.; Navarro, O.; Spencer, J. Transition metal
catalyzed element−element′ additions to alkynes. Coord. Chem. Rev.
2017, 336, 54−77. (d) Trost, B. M.; Masters, J. T. Transition metal-
catalyzed couplings of alkynes to 1,3-enynes: modern methods and
synthetic applications. Chem. Soc. Rev. 2016, 45, 2212−2238.
(e) Halliday, C. J. V.; Lynam, J. M. Gold−alkynyls in catalysis:
alkyne activation, gold cumulenes and nuclearity. Dalton Trans 2016,
45, 12611−12626.
(10) For reviews, see (a) Chinchilla, R.; Nájera, C. The Sonogashira
reaction: A booming methodology in synthetic organic chemistry.
Chem. Rev. 2007, 107, 874−922. (b) Chinchilla, R.; Nájera, C. Recent
advances in Sonogashira reactions. Chem. Soc. Rev. 2011, 40, 5084−
5121. For selected articles, see (c) Zhang, Z.-H.; Dong, X.-Y.; Du, X.-
Y.; Gu, Q.-S.; Li, Z.-L.; Liu, X.-Y. Copper-catalyzed enantioselective
Sonogashira-type oxidative cross-coupling of unactivated C(sp3)−H
bonds with alkynes. Nat. Commun. 2019, 10, 5689. (d) Correia, C. A.;
Li, C.-J. Copper-catalyzed cross-dehydrogenative coupling (CDC) of
alkynes and benzylic C−H Bonds. Adv. Synth. Catal. 2010, 352,
1446−1450.
(3) For selected reviews, see (a) Carden, J. L.; Dasgupta, A.; Melen,
R. L. Halogenated triarylboranes: synthesis, properties and
applications in catalysis. Chem. Soc. Rev. 2020, 49, 1706−1725.
(b) Paradies, J. Mechanisms in frustrated Lewis pair-catalyzed
reactions. Eur. J. Org. Chem. 2019, 2019, 283−294. (c) Scott, D. J.;
Fuchter, M. J.; Ashley, A. E. Designing effective ‘frustrated Lewis pair’
hydrogenation catalysts. Chem. Soc. Rev. 2017, 46, 5689−5700.
(4) For reviews, see (a) Hoshimoto, Y.; Ogoshi, S. Triarylborane-
catalyzed reductive N-alkylation of amines: A perspective. ACS Catal.
2019, 9, 5439−5444. (b) Fyfe, J. W. B.; Watson, A. J. B. Merging new
organoborane chemistry with living radical polymerization. Chem.
2017, 3, 31−55. For selected articles, see (c) Khan, I.; Manzotti, M.;
Tizzard, G. J.; Coles, S. J.; Melen, R. L.; Morrill, L. C. Frustrated
Lewis pair (FLP)-catalyzed hydrogenation of Aza-Morita−Baylis−
Hillman adducts and sequential organo-FLP catalysis. ACS Catal.
2017, 7, 7748−7752. (d) Légaré, M.-A.; Courtemanche, M.-A.;
́
Rochette, E.; Fontaine, F.-G. Metal-free catalytic C−H bond
activation and borylation of heteroarenes. Science 2015, 349, 513−
4463
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
̌
(11) (a) Gazvoda, M.; Virant, M.; Pinter, B.; Kosmrlj, J. Mechanism
(b) Buick, A. R.; Kemp, T. J.; Stone, T. J. Electron spin resonance
spectra of radical anions of styrene and related compounds. J. Phys.
Chem. 1970, 74, 3439−3444.
(28) Kwaan, R. J.; Harlan, C. J.; Norton, J. R. Generation and
characterization of the tris(pentafluorophenyl)borane radical anion.
Organometallics 2001, 20, 3818−3820.
of copper-free Sonogashira reaction operates through palladium-
palladium transmetallation. Nat. Commun. 2018, 9, 4814. (b) Huang,
H.; Liu, H.; Jiang, H.; Chen, K. Rapid and efficient Pd-catalyzed
Sonogashira coupling of aryl chlorides. J. Org. Chem. 2008, 73, 6037−
6040.
(29) For the styryl radical, see (a) Bennett, J. E.; Howard, J. A. ESR
spectrum of the α-styryl radical. Chem. Phys. Lett. 1971, 9, 460−462.
For the phenylacetylene radical, see (b) Sipe, H. J.; Kispert, L. D.;
Hwang, J. S.; Tsonis, C. P. Electron Paramagnetic Resonance studies
of thermally produced radicals in phenylacetylene. J. Phys. Chem.
1982, 86, 4017−4019.
(30) Ingold, C. K.; Shaw, F. R. CCCLXXXVIII.The nature of the
alternating effect in carbon chains. Part XXII. An attempt further to
define the probable mechanism of orientation in aromatic
substitution. J. Chem. Soc. 1927, 0, 2918−2926.
(31) Yau, H. M.; Croft, A. K.; Harper, J. B. One-pot’ Hammett plots:
a general method for the rapid acquisition of relative rate data. Chem.
Commun. 2012, 48, 8937−8939.
(32) Hansch, C.; Leo, A.; Taft, R. W. A survey of Hammett
substituent constants and resonance and field parameters. Chem. Rev.
1991, 91, 165−195.
(33) (a) Miao, M.; Jin, M.; Chen, P.; Wang, L.; Zhang, S.; Ren, H.
Iron(III)-Mediated Bicyclization of 1,2-Allenyl Aryl Ketones:
Assembly of Indanone-Fused Polycyclic Scaffolds and Dibenzo[a,e]-
pentalene Derivatives. Org. Lett. 2019, 21, 5957−5961. (b) Martins,
G. M.; Back, D. F.; Kaufman, T. S.; Silveira, C. C. SeCl2-Mediated
Approach Toward Indole-Containing Polysubstituted Selenophenes.
J. Org. Chem. 2018, 83, 3252−3264.
(12) (a) Sawama, Y.; Goto, R.; Nagata, S.; Shishido, Y.; Monguchi,
Y.; Sajiki, H. Chemoselective and direct functionalization of methyl
benzyl ethers and unsymmetrical dibenzyl ethers by using iron
trichloride. Chem. - Eur. J. 2014, 20, 2631−2636. (b) Xiang, S.-K.;
Zhanga, L.-H.; Jiao, N. sp−sp3 C−C bond formation via Fe(OTf)3/
TfOH cocatalyzed coupling reaction of terminal alkynes with benzylic
alcohols. Chem. Commun. 2009, 2009, 6487−6489.
(13) For a selected review, see (a) Thomas, A. M.; Sujatha, A.;
Anilkumar, G. Recent advances and perspectives in copper-catalyzed
Sonogashira coupling reactions. RSC Adv. 2014, 4, 21688−21698.
For a selected article, see (b) Cao, Y.-X.; Dong, X.-Y.; Yang, J.; Jiang,
S.-P.; Zhou, S.; Li, Z.-L.; Chen, G.-Q. A Copper-catalyzed
Sonogashira coupling reaction of diverse activated alkyl halides with
terminal alkynes under ambient conditions. Adv. Synth. Catal. 2020,
362, 2280−2284.
(14) Nan, G.; Zhou, J. Indium(III) catalyzed direct sp3-sp C−C
bond formation from alcohols and terminal alkynes. Lett. Org. Chem.
2013, 10, 555−561.
(15) Fisher, K. M.; Bolshan, Y. Brønsted acid-catalyzed reactions of
trifluoroborate salts with benzhydryl alcohols. J. Org. Chem. 2015, 80,
12676−12685.
(16) Kabalka, G. W.; Yao, M.-L.; Borella, S. Substitution of hydroxyl
groups with alkynyl moieties using alkynylboron dihalides: An
efficient approach to secondary alkylacetylene derivatives. Org. Lett.
2006, 8, 879−881.
(17) Malkoch, M.; Thibault, R. J.; Drockenmuller, E.;
Messerschmidt, M.; Voit, B.; Russell, T. P.; Hawker, C. J. Orthogonal
approaches to the simultaneous and cascade functionalization of
macromolecules using click chemistry. J. Am. Chem. Soc. 2005, 127,
14942−14949.
(18) Barluenga, J.; Andina, F.; Aznar, F.; Valdés, C. New cascade
processes on group 6 Fischer-type carbene complexes: Cyclo-
propanation and metathesis reactions. Org. Lett. 2007, 9, 4134−4146.
(19) Dasgupta, A.; Stefkova, K.; Babaahmadi, R.; Gierlichs, L.;
Ariafard, A.; Melen, R. L. Triarylborane catalysed alkenylation
reaction of aryl esters with diazo-compounds. Angew. Chem., Int. Ed.
2020, 59, 15492−15496.
(20) Follet, E.; Mayer, P.; Stephenson, D. S.; Ofial, A. R.; Berionni,
G. Reactivity-tuning in frustrated Lewis pairs: Nucleophilicity and
Lewis basicity of sterically hindered phosphines. Chem. - Eur. J. 2017,
23, 7422−7427.
(21) Liu, L.; Cao, L. L.; Shao, Y.; Menard, G.; Stephan, D. W. A
radical mechanism for frustrated Lewis pair reactivity. Chem. 2017, 3,
259−267.
(22) Pan, X.; Chen, X.; Li, T.; Li, Y.; Wang, X. Isolation and X-ray
crystal structures of triarylphosphine radical cations. J. Am. Chem. Soc.
2013, 135, 3414−3417.
(23) Lyons, A. R.; Symons, M. C. R. Unstable intermediates. Part
117 − Electron spin resonance studies of dimeric (σ*) radicals of type
R3X−XR3 with a three electron X−X bond, when X is phosphorus or
arsenic. J. Chem. Soc., Faraday Trans. 2 1972, 68, 1589−1594.
(24) Gara, W. B.; Roberts, B. R. An electron spin resonance study of
the electrochemical oxidation of phosphorus(III) compounds. J.
Chem. Soc., Perkin Trans. 2 1978, 1978, 150−154.
(25) Symons, M. C. R.; Tordo, P.; Wyatt, J. The structure of
diphosphine radical cations. J. Organomet. Chem. 1993, 443, C29−
C32.
(26) Culcasi, M.; Gronchi, G.; Tordo, P. Tetraaryldiphosphine
cation radicals. electrochemical generation and ESR study. J. Am.
Chem. Soc. 1985, 107, 7191−7193.
(27) (a) Sipe, H. J., Jr.; Hwang, J. S.; Tsonls, C. P.; Klsper, L. D.
Electron paramagnetic resonance studies of thermally produced
radicals in phenylacetylene. J. Phys. Chem. 1982, 86, 4017−4019.
4464
https://dx.doi.org/10.1021/jacs.1c01622
J. Am. Chem. Soc. 2021, 143, 4451−4464