Kinetics and Mechanism of the Palladium-Catalyzed Oxidative Arylating Carbocyclization of Allenynes

Article abstract of DOI:10.1021/jacs.7b10267

Pd-catalyzed C-C bond-forming reactions under oxidative conditions constitute a class of important and widely used synthetic protocols. This Article describes a mechanistic investigation of the arylating carbocyclization of allenynes using boronic acids and focuses on the correlation between reaction conditions and product selectivity. Isotope effects confirm that either allenic or propargylic C-H activation occurs directly after substrate binding. With an excess of H₂O, a triene product is selectively formed via allenic C-H activation. The latter C-H activation was found to be turnover-limiting and the reaction zeroth order in reactants as well as the oxidant. A dominant feature is continuous catalyst activation, which was shown to occur even in the absence of substrate. Smaller amounts of H₂O lead to mixtures of triene and vinylallene products, where the latter is formed via propargylic C-H activation. The formation of triene occurs only in the presence of ArB(OH)₂. Vinylallene, on the other hand, was shown to be formed by consumption of (ArBO)₃ as a first-order reactant. Conditions with sub-stoichiometric BF₃·OEt₂ gave selectively the vinylallene product, and the reaction is first order in PhB(OH)₂. Both C-H activation and transmetalation influence the reaction rate. However, with electron-deficient ArB(OH)₂, C-H activation is turnover-limiting. It was difficult to establish the order of transmetalation vs C-H activation with certainty, but the results suggest that BF₃·OEt₂ promotes an early transmetalation. The catalytically active species were found to be dependent on the reaction conditions, and H₂O is a crucial parameter in the control of selectivity.

Full text of DOI:10.1021/jacs.7b10267

Subscriber access provided by READING UNIV
Article
Kinetics and Mechanism of the Palladium-Catalyzed
Oxidative Arylating Carbocyclization of Allenynes
Teresa Bartholomeyzik, Robert Pendrill, Richard Lihammar, Tuo Jiang, Göran Widmalm, and Jan-E. Bäckvall
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b10267 • Publication Date (Web): 20 Nov 2017
Downloaded from http://pubs.acs.org on November 20, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 15
Journal of the American Chemical Society
1
2
3
4
5
6
7
Kinetics and Mechanism of the Palladium-Catalyzed Oxidative
Arylating Carbocyclization of Allenynes
Teresa Bartholomeyzik,† Robert Pendrill,† Richard Lihammar, Tuo Jiang, Göran Widmalm,* JanꢀE. Bäckvall*
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Department of Organic Chemistry, Arrhenius Laboratory, Stockholm University, SEꢀ10691 Stockholm, Sweden
ABSTRACT: Pdꢀcatalyzed C–C bondꢀforming reactions under oxidative conditions constitute a class of important and widely used
synthetic protocols. This article describes a mechanistic investigation of the arylating carbocyclization of allenynes using boronic
acids and focuses on the correlation between reaction conditions and product selectivity. Isotope effects confirm that either allenic
or propargylic C–H activation, respectively, occurs directly after substrate binding. With an excess of H2O, a triene product is
selectively formed via allenic C–H activation. The latter C–H activation was found to be turnoverꢀlimiting and the reaction is zeroth
order in reactants as well as the oxidant. A dominant feature is continuous catalyst activation which was shown to occur even in the
absence of substrate. Smaller amounts of H2O lead to mixtures of triene and vinylallene products, where the latter is formed via
propargylic C–H activation. The formation of triene occurs only in the presence of ArB(OH)2. Vinylallene, on the other hand, was
shown to be formed by consumption of (ArBO)3 as a first order reactant. Conditions with subꢀstoichiometric BF3·OEt2 gave
selectively the vinylallene product, and the reaction is first order in PhB(OH)2. Both C–H activation and transmetalation influence
the reaction rate. However, with electronꢀdeficient ArB(OH)2, C–H activation is turnoverꢀlimiting. It was difficult to establish the
order of transmetalation vs. C–H activation with certainty, but the results suggest that BF3·OEt2 promotes an early transmetalation.
The catalytic active species were found to be dependent on the reaction conditions and H2O is a crucial parameter in the control of
selectivity.
Scheme 1. Phenylating oxidative carbocyclization of
allenyne 1a under selective conditions
INTRODUCTION
One common characteristic of many Pdꢀcatalyzed oxidative
reactions is the occurrence of a C–H bond cleavage step which
can mechanistically be classified either as a βꢀhydride
elimination or as a C–H activation. This step is of central
importance and has attracted considerable attention recently.1,2
Currently, an increasing number of selective Pdꢀcatalyzed
oxidative transformations are available and accompanying
mechanistic studies have provided information concerning the
underlying processes.3,4 This paper presents results that
provide mechanistic insight into an arylating carbocyclization
reaction, and discusses its different possible pathways and the
factors determining the product selectivity.
We have previously described the PdIIꢀcatalyzed arylating
oxidative carbocyclization of 1,5ꢀallenynes.5,6 The reaction can
give either an arylated triene product 2 or an arylated
vinylallene product 3 (Scheme 1). It was found that conditions
employing BF3·OEt2 as additive provided selectively
vinylallene products 3 for a broad range of allenynes and
arylboronic acids.6 The selective formation of trienes 2 was
accomplished using an excess of H2O. The optimization of the
latter protocol (upper reaction in Scheme 1) showed that for
substrates with a longer propargylic alkyl substituent (R =
alkyl vs. R = H in Scheme 2) it was necessary to replace the
standard oxidant 1,4ꢀbenzoquinone (BQ) with tetraꢀFꢀBQ in
order to obtain satisfying yields. We have furthermore identiꢀ
fied the formation of 4ꢀhydroxyphenoxyꢀsubstituted side
products (5, 6) which explains the low yields of 3 in some
cases (Scheme 2).
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 15
Pd(OAc)2 was used as PdII catalyst precursor throughout this
work.13
The reaction profile for the selective transformation of 1a to
2a under the optimized reaction conditions with 5 equiv of
H2O is given in Figure 1a. Since an excess of H2O is
employed, (PhBO)3 is immediately hydrolyzed to PhB(OH)2.
Besides the formation of arylated triene 2a in 94% yield, we
observed the formation of hydroquinone (HQ) and boric acid
1
2
3
4
5
6
7
8
Scheme 2. Proposed intermediates M1 and M2 in the
formation of arylated carbocycles 2 and 3 and side
products 4–6
E
E
Ph
4
R = H (1a)
ref 6
as byꢀproducts and
a corresponding decrease in the
a
9
concentration of H2O. The formation of benzene by
protodeboronation of PhB(OH)2 only occurred at a very low
concentration of BQ.14
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
E
E
E
E
c for R = H, alkyl
d for R = H
Pd
L
X
0.5 equiv (PhBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
ref 6
E
E
R
E
E
E
E
Ph
R
M1
2
•
•
5.0 equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
Ph
R
1a
2a
E
E
1
E
E
e
•
R
•
Pd
•
R = H, alkyl
ref 6
L
X
R
Ph
M2
3
R = Me (1b)
b
E
E
E
E
•
in situ
HO
HO
O
Ph
O
5
6
(a) (PhBO)3, cat. Pd(OAc)2, BQ, H2O, dioxane, 80 °C
(b) PhB(OH)2, cat. Pd(OAc)2, BQ, excess H2O, dioxane, 80 °C
(c) (PhBO)3, cat. Pd(OAc)2, tetraꢀFꢀBQ, excess H2O, dioxane, 60 °C
(d) (PhBO)3, cat. Pd(OAc)2, BQ, excess H2O, THF, rt
(e) PhB(OH)2, cat. Pd(OAc)2, BQ, cat. BF3—Et2O, THF, rt
X = anionic ligand, L = neutral ligand
Figure 1. Selective reaction to give 2a. a) Reaction profile. b)
Rate profile of product formation. Scale: 0.1 mmol 1a.
Possible pathways that account for the formation of all the
observed products are shown in Scheme 2. A dienylꢀpalladium
intermediate M1 was suggested previously for PdIIꢀcatalyzed
cyclizations of allene substrates.7–9 Based on kinetic isotope
effects (KIEs) we recently proposed the unusual allenylꢀ
palladium intermediate M2 in two related transformations.7,10
This intermediate also rationalizes the generation of 5, which
is the product of a formal oxidative PdIIꢀcatalyzed propargylic
C–H functionalization. In this article, we report on a
mechanistic study of the arylating carbocyclization based on
kinetic analyses. The experimental data presented support the
previously proposed pathways involving allenic or propargylic
C–H activation, respectively.6
Investigation of induction in the formation of triene 2a.
The reaction is associated with an unusually long induction
period extending over the entire reaction time. Initially, the
rate of product formation is zero, and it increases linearly with
time (Figure 1b).
Induction periods in catalytic reactions can be the result of
an ongoing preꢀcatalyst activation. The resulting continuous
change in catalyst concentration complicates the interpretation
of kinetic data.15 To obtain insight into the induction of the
reaction, we designed soꢀcalled preꢀstirring experiments. It
was found that mixing all reaction components except
allenyne 1a and stirring them for 3 h at 25 ˚C prior to the
addition of 1a leads to an initial reaction rate larger than zero
(experiment A in Figure 2a).16 This observation shows that the
catalytic system undergoes changes which are independent of
the substrate.17
RESULTS AND DISCUSSION
The arylating carbocyclization reactions of allenyne 1a were
monitored in situ by 1H NMR spectroscopy.11 During the study
we encountered an inconsistent overall reaction rate related to
the quality of the THFꢀd8 used. The reaction reached a higher
rate when the solvent was freshly distilled compared to when
it had been kept over 3Å molecular sieves. Additionally, the
reaction was significantly slower when peroxyacetic acid was
added to the reaction mixture in order to mimic varying
contents of peroxides.12 To ensure reproducibility, the same
THFꢀd8 was generally used to perform one set of experiments.
2
ACS Paragon Plus Environment

Page 3 of 15
Journal of the American Chemical Society
1
2
standard conditions A:
E
E
3
4
5
6
0.5 equiv (PhBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
•
E
E
1a
'preꢀstirring'
2a
3 equiv H2O
1 mL dry THFꢀd8
25 °C
3 h
dry THFꢀd8 (0.1 M)
Ph
25 °C
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Figure 2. Investigation of catalyst activation with the aid of preꢀstirring experiments. Scale: 0.1 mmol 1a. a) Under standard conditions A,
the shown mixture is preꢀstirred before substrate 1a (0.1 mmol) is added. In experiments B and C, (PhBO)3 or BQ, respectively, were
added together with 1a after the preꢀstirring period. b) Selected region of the H NMR spectrum before and after the addition of 1a
according to conditions A. c) The total volume after addition of 1a was 1 mL in all experiments, and the values refer to the different
volumes during the preꢀstirring period in experiments D–G.
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
Catalyst activation during preꢀstirring was found to require
the presence of both PhB(OH)2 and BQ, as shown in
Figure 2a. When preꢀstirring was performed without added
H2O, the initial rate was lower compared to the reference.12 As
there were small amounts of H2O in the solvent, we speculate
that H2O plays a part in the activation as well.
to the previously observed effect of H2O on the product
distribution.6
a) (PhBO)3 + 3 H2O
3 PhB(OH)2
Keq = 62.6 mMꢀ1 (0.5 M at 25 °C)
THFꢀd8
b)
Ph
OH
B
Ph
O
B
Ph
+ 2 H2O
+ H2O
B
B
+
PhB(OH)2
3 PhB(OH)2
O
OH
The slower overall reaction rate in experiment C (without
BQ) is likely due to partial decomposition of Pd(OAc)2 as
supported by the formation of Pd black. Furthermore, BQ was
found to be an indicator for the process occurring during preꢀ
stirring, since its NMR peak broadens significantly before it
takes its original shape upon addition of 1a (Figure 2b). This is
suggestive of BQ having a role as ligand to palladium in
addition to its function as reoxidant. The overall reaction rates
were higher when the preꢀstirring was performed in smaller
volumes, thus showing that catalyst activation during preꢀ
stirring is concentration dependent (Figure 2c). Additional
experiments showed that the activation process during preꢀ
stirring was not influenced by the concentrations of either
PhB(OH)2 or BQ.12 Therefore, the concentration dependence
of the catalyst activation shown in Figure 2c is the result of the
different Pd(OAc)2 concentration in these experiments.18 Even
though the preꢀstirring strategy has provided a tool to
modulate the initial reaction rates, it did not lead to full
catalyst activation before starting material 1a was consumed.
O
O
B
Ph
Ph
Figure 3. Hydrolysis of (PhBO)3 in THFꢀd8: (a) equilibrium
between phenyl boroxine and phenylboronic acid; (b) twoꢀstep
hydrolysis via (PhBOH)2O; (c) change of concentrations of
boronic species at increasing amounts of added water.
One initial experiment (vide infra) suggests that this state
can be reached by performing multiple consecutive additions
of reactants. However, this strategy would be susceptible to
reproducibility issues and is therefore not a suitable method
for accurate kinetic analysis.
The yields of reaction products 2a, 3a and 4 as a function of
the concentration of H2O are given in Figure 4, and – in
support of our hypothesis – the obtained graph resembles
Figure 3. For small amounts of H2O, which corresponds to
high [(PhBO)3], allene 3a is the major product. With
increasing concentration of H2O (and therefore at higher
[PhB(OH)2]), the yield of allene 3a decreases in favor of
formation of triene 2a. The latter is formed selectively and in
Effect of H2O. Under H2O concentrations below 200 mM
(2 equiv), the anhydride forms of PhB(OH)2 are present
(Figure 3). We speculated that this equilibrium may be related
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 15
1
2
the lower reaction rates at [H2O]total >300 mM could be the
yields of 90−95% for conditions applying >4.5 equivalents of
H2O.
formation of unreactive PdII hydroxy complexes.20b
3
4
5
6
7
8
0.5 equiv (PhBO)3
0.5 equiv (PhBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
E
E
E
E
E
E
E
E
E
E
E
E
5 mol% Pd(OAc)2
1.1 equiv BQ
•
+
•
+
•
•
0.5 equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
H2O
dry THFꢀd8 (0.1 M)
25 °C
Ph
Ph
Ph
Ph
1a
2a
3a
1a
2a
3a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Reaction profile for a reaction with limiting [H2O].
Scale: 0.1 mmol 1a.
In contrast to the observed kinetics for the formation of
triene 2a, the induction is less pronounced under conditions
for selective formation of vinylallene 3a using low
concentrations of H2O (Figure 5). Considering only the phase
of the reaction for which [PhB(OH)2] = 0 and [2a] = const, the
rate of vinylallene formation has a first order dependence on
(PhBO)3 (Figure 6).
Figure 4. The effect of H2O on the product distribution; final
concentrations of products 2a, 3a, and 4 vs. initial [H2O]total (free
[H2O] plus amount consumed in the hydrolysis of (PhBO)3,
determined at <1% conversion). Conditions: 0.1 mmol of 1a, 0–9
equiv added H2O. Full conversion was obtained within 10 h for
>0.5 equiv H2O.19
0.5 equiv (PhBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
E
E
E
E
E
E
The yield of phenylated side product 4 increases at first and
reaches ca. 9% at [H2O]total ≈ 120 mM, but is suppressed at
higher [H2O]. Curiously, the maximum yield of 4 coincides
with the maximal concentration of (Ph(OH)B)2O. From the
same set of experiments it was also found that the rate of
formation of triene 2a (measured as the highest observed rate)
increases with the amount of added H2O until a maximum is
reached at [H2O]total ≈ 300 mM. At higher concentrations of
H2O however, the reaction is gradually slowed down.12
•
+
•
H2O
dry THFꢀd8 (0.1 M)
25 °C
Ph
Ph
1a
2a
3a
When the phenylating carbocyclization is performed with
[H2O]total ≤100 mM, the liberated PhB(OH)2 is consumed prior
to complete conversion of allenyne 1a, leaving (PhBO)3 as the
only identified boronꢀreactant (“limiting conditions”). At this
point, the generation of triene 2a stops abruptly while the
formation of vinylallene 3a continues, although at a slower
rate (illustrated in Figure 5 for 0.5 equiv of added H2O).
Therefore, our results suggest a relationship between the
concentrations of (PhBO)3, (Ph(OH)B)2O and PhB(OH)2 and
the formation of 3a, 4 and 2a, respectively. An alternative
explanation is that the different products are formed in the
presence of different active catalytic species, and that these
species are influenced by the concentration of water. We have
identified both Pd3(µ2ꢀOAc)6 and Pd3(µ2ꢀOH)(OAc)5 in the
reaction mixtures.13,20 With large amounts of added water
(3 equiv), Pd3(µ2ꢀOAc)6 is converted to Pd3(µ2ꢀOH)(OAc)5
within 10 minutes. In contrast, with small amounts of added
water, Pd3(µ2ꢀOAc)6 is present in large quantities throughout
the entire reaction, and only small amounts of Pd3(µ2ꢀ
OH)(OAc)5 are formed. These two species are believed to be
precursors to the active catalytic species in the formation of
the different reaction products.12 A possible explanation for
Figure 6. Rate of formation of 3a vs. [(PhBO)3] for different
[H2O]total (0–0.75 equiv added H2O) illustrating a first order
dependence on (PhBO3). Plotted is the region of the reactions in
which [2a] = const due to the complete consumption of PhB(OH)2
and water (e.g., t = 2.5–12 h in Figure 5). Scale: 0.1 mmol 1a.
Furthermore, the rate of the reaction to give vinylallene 3a
is comparatively higher with higher initial concentration of
H2O. The observed faster rate of formation of 3a could be
explained by a higher degree of catalyst activation in the
presence of more water.21
Effect of boronic acid substituent. Our original report
demonstrated the dependence of the ratio 2/3 on the structure
of the substrate.5 Table 1 provides the corresponding results
regarding the influence of the arylboronic acid. Clearly,
electronꢀdeficient arylboronic acids favor the formation of the
corresponding vinylallenes 3, whereas electronꢀrich
arylboronic acids favor triene 2. Commercially available
4
ACS Paragon Plus Environment

Page 5 of 15
1
Journal of the American Chemical Society
0.5 equiv (ArBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
boronic acids generally consist of a mixture of ArB(OH)2 and
E
E
E
E
E
E
2
(ArBO)3. For more electronꢀdeficient arylboronic acids the
relative amount of ArB(OH)2 is expected to be higher due to
the facilitated hydrolysis of (ArBO)3.22 The results described
above (cf. Figure 4) suggest that (PhBO)3 reacts to give
vinylallene 3a, whereas PhB(OH)2 is responsible for the
formation of triene 2a. If the product distribution were solely
controlled by the ratio of the boronic acid derivatives present,
we would have expected the opposite trend regarding the
substituent effect in Table 1.
•
+
3
4
5
6
7
8
9
•
0.5 equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
Ar
Ar
1a
2
3
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Selectivities for various ArB(OH)2.a,b
1.3 equiv ArB(OH)2
1 mol% Pd(OAc)2
E
E
E
E
E
E
1.1 equiv BQ
•
+
•
THF (0.1 M), rt, 20 h
Ar
Ar
1
2
3
Figure 7. Formation of triene 2 and vinylallene 3 for different
(ArBO)3. Scale: 0.1 mmol 1a.
Influence of other reaction components in the selective
formation of triene 2a. When studying the reaction at
different preꢀcatalyst loadings saturation behavior was
observed (Figure 8). A possible explanation is that formation
of the active catalyst from oligomeric species is slow and not
proportional to the amount of added preꢀcatalyst (vide
supra).12,25,26
Entry
Ar
2 (%)
3 (%)
2/3
1c
2d
3e
4
C6H5
50
68
3
17
6
75:25
92:8
6:94
43:5
91:9
5:95
2ꢀnaphthyl
4ꢀBrꢀC6H4
4ꢀCHOꢀC6H4
4ꢀMeOꢀC6H4
3ꢀNO2ꢀC6H4
45
39
5
30
53
<3
5
0.5 equiv (PhBO)3
6
55
E
E
E
E
Pd(OAc)2
1.1 equiv BQ
(a) Scale: 0.1 mmol 1a. (b) Yields and selectivities were
determined by H NMR analysis of crude reaction mixtures
using anisole as an internal standard. (c) Isolated yields, see
ref 5. (d) 0.1 mmol in 0.5 mL. (e) 9% of 1a remained.
•
1
3.0 equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
Ph
1a
2a
Consequently, we chose 4ꢀNO2ꢀ and 4ꢀMeOꢀphenyl
boroxines and investigated their kinetic behavior under strict
control of the water concentration present (0.5 equiv added
H2O, Figure 7). As for phenyl boroxine, the generation of the
corresponding trienes 2 relied on the presence of ArB(OH)2.
Furthermore, it was confirmed that the electronic properties of
the aryl boroxine play a dominant role with respect to the
product distribution. The use of (4ꢀNO2ꢀC6H4ꢀBO)3 provided
vinylallene 3ab with high selectivity. Changing from
PhB(OH)2 to the 4ꢀNO2ꢀsubstituted derivative substantially
increased the fraction of ArB(OH)2 that is consumed in the
pathway towards vinylallene 3. In contrast, the preference of
ArB(OH)2 to react towards the corresponding triene 2 is
significantly enhanced for the electronꢀrich 4ꢀMeOꢀ
C6H4ꢀB(OH)2. Once the latter is consumed, the reaction
towards the corresponding vinylallene 3ac (thereby consuming
(4ꢀMeOꢀC6H4ꢀBO)3) accelerates. In accordance with the
results for (PhBO)3 (cf. Figure 6), it was verified for an
electronꢀrich boroxine (4ꢀMeO) and an electronꢀdeficient
derivative (4ꢀCF3) that the rate of formation of 3 is consistent
with a first order dependence on [(ArBO)3] after the
Figure 8. Saturation behavior in Pd(OAc)2; ∆(rate)/∆t vs. added
Pd(OAc)2, with rate being ∆[2a]/∆t. Conditions: 1a (0.1 mmol),
1–20 mol% Pd(OAc)2.
Further, the dependence on the concentration of the various
reactants was studied by preꢀstirring identical mixtures, but
varying concentrations of either substrate 1a, (PhBO)3/H2O or
BQ, respectively, after the preꢀstirring period (cf. Figure 2).
All of these experiments provided an initial reaction rate
similar to that obtained under standard conditions (experiment
A in Figure 2a) suggesting that the reaction is of zeroth order
in these reactants. This interpretation was confirmed by an
experiment based on multiple additions of all the reactants
(Figure 9). After the third addition it appears that the reaction
rate had reached its maximum value. The conversion of 1a to
give 2a proceeded with a nearly constant rate and essentially
the same rate just before and after additions B and C –
independent of the change in [1a], [(PhBO)3], and [BQ].
consumption of ArB(OH)2.12,
,
23 24
5
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 15
E
E
1
2
3
4
5
6
0.5 equiv (PhBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
E
E
E
E
1a
(PhBO)3
BQ
1a
(PhBO)3
BQ
•
E
E
(PhBO)3
Pd(OAc)2
BQ
•
H2O
H2O
3.0 equiv H2O
additive
1a
Ph
25 °C
3 h
1a
2a
H2O
dry THFꢀd8 (0.1 M)
25 °C
Ph
dry THFꢀd8
2a
7
pr e-stir ring per iod
addition A
addition B
addition C
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. Rate of formation of 2a vs. released [HOAc].
Conditions: 0.1 mmol 1a; a) no additive; b) 20 mol% HOAc; c)
20 mol% HOAcꢀd4.
Figure 9. Multiple addition experiment; ∆[2a]/∆t vs. time.
Conditions: preꢀstirring period: (PhBO)3 (100 µmol), Pd(OAc)2
(5 µmol), BQ (220 µmol), H2O (650 µmol) in 1 mL THFꢀd8,
25 °C; addition A: 1a (200 µmol); addition B and C: 1a
(200 µmol), (PhBO)3/H2O (to give 300 µmol PhB(OH)2 and
350 µmol free H2O), BQ (to give 220 µmol); additions B and C
were performed before complete consumption of 1a (<15 µmol)
to avoid potential catalyst decomposition.
Kinetic Isotope Effects for reactions with waterꢀ
controlled selectivity. To obtain the phenylated products 2
(triene) or 3 (vinylallene), either an allenic or a propargylic C–
H bond must be broken, respectively.27 Under nonꢀselective
conditions ([H2O]total <300 mM) methyl group deuteration at
either the propargylic (A) or the allenic position (B) resulted in
a significant change of the ratio 2a/3a (Scheme 3).
Since only the change in concentration of added Pd(OAc)2
leads to a clear effect on the reaction rate (Figure 8), we
suggest that the turnoverꢀlimiting step follows an equilibrium
that is controlled by the concentration of Pd catalyst.
Scheme 3. Effect of deuteration on the product ratio
A ꢀ Propargylic deuteration
0.5 equiv (PhBO)3
E
E
E
E
E
E
E E
The reaction of 1a to give 2a is accompanied by the release
of HOAc originating from Pd(OAc)2. The concentration of
liberated HOAc is linearly correlated to the reaction rate
(Figure 10). However, the addition of HOAc (20 mol%) to the
reaction mixture did not cause an increase in the rate of the
reaction suggesting that more HOAc is released when the
reaction is faster, rather than the released HOAc causing a
faster reaction. Addition of HOAcꢀd4 (20 mol%) to the
reaction mixture resulted in more HOAc being released in
relation to the reaction rate. This observation can be attributed
to catalystꢀbound AcO– being replaced by AcO–ꢀd3 and
provides experimental evidence for a reversible release of
acetate during the catalytic cycle.
5 mol% Pd(OAc)2,
1.1 equiv BQ
•
+
+
D3C
Ph
CD3
•
Ph
x equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
D
Ph
CD3
D
1aꢀd3
2aꢀd3
3aꢀd2
4ꢀd3
B
Allenic deuteration
−
E
E
E
E
E
E
E
E
as above
CD3
D
D
•
+
+
•
Ph
CD3
Ph
D
Ph CD3
D3C
CD3 CD3
1a
D
ꢀd6
2aꢀd5
3aꢀd6
4ꢀd5
for 1.0 equiv H2O:
for 1.2 equiv H2O:
2a
49%
3a
2a/3a
4
a)
1aꢀd3 (A)
10%
8%
83:16
reference (1a)
(33%)
(34%)
(7%)
(49:51)
b)
c)
1aꢀd3 (A)
63%
7%
8%
7%
1%
89:11
12:88
1aꢀd6 (B)
50%
reference (1a)
(48%)
(27%)
(8%)
(64:36)
for 1.6 equiv H2O:
1a
d)
ꢀd6 (B)
reference (1a)
21%
40%
<1%
(7%)
34:66
(83:17)
(73%)
(15%)
Scale: 0.1 mmol 1aꢀd3 (A) or 1aꢀd6 (B). Yields in red and blue
refer to products from reactions starting from 1aꢀd3 and 1aꢀd6,
respectively. Yields in black serve as reference and were
interpolated from the data in Figure 4. Yields and the amount of
added H2O were determined from the 1H NMR spectra of the
crude reaction mixture. Full conversion was reached for a) and b)
at 4.5 h and 4 h, respectively, and 27 m
M
and 36 m
M
of C6H6
6
ACS Paragon Plus Environment

Page 7 of 15
Journal of the American Chemical Society
1
2
3
4
5
6
7
were formed, respectively. For c) 15% of 1aꢀd6 remained after full
consumption of the phenyl source at 5 h; 51 m of C6H6 was
formed. For d) 15% of 1ꢀd6 remained after full consumption of the
phenyl source at 6 h; 33 m of C6H6 was formed.
M3 lying far towards intermediate M3. The large isotope
effect measured in individual experiments (cf. Scheme 4b)
identifies the C–H cleavage as the turnoverꢀlimiting step, and
is thus in full agreement with the zero order dependence on the
reactants.
M
M
This means that the selectivity is determined by the C–H
bond cleavage as an early step in the reaction and that
pathways starting with arylpalladation are ruled out.28
Scheme 5. Proposed mechanism for the formation of triene
2a
8
9
For the selective reaction to give triene 2a (i.e. with 5 equiv
of H2O) we determined the KIE in a competitive reaction setup
which gave a large value of kH/kD = 8.7 ± 1 (Scheme 4a).29
Due to the underlying induction process and the continuous
change in rate, it was not possible to conclusively determine
KIE (kH/kD) from individual runs with a conventional reaction
setꢀup. Instead, we applied the preꢀstirring strategy (cf.
Figure 2) to obtain information about the KIE (Scheme 4b).
Since the catalyst activation during preꢀstirring and therefore
the amount of active catalyst in the very beginning is identical,
the ratio of the initial reaction rates for the protonated or
deuterated substrate, respectively, is considered comparable to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a conventional KIE. And indeed, with rateinit,H/rateinit,D
=
9.0 ± 1, a value similar to the competitive KIE (Scheme 4) was
obtained.29
Scheme 4. KIEs in the formation of triene 2a
Only the equilibrium that is crucial to match the experimental
results is indicated.
Kinetic Isotope Effects in the formation of vinylallene
3a. To investigate the KIE in the selective formation of
vinylallene 3a, we used the reaction conditions employing
BF3·OEt2 as an additive (cf. Scheme 1).32 In a competitive
experiment, we obtained
a value of kH/kD = 7.1 ± 1
(Scheme 6a).29 This large competitive isotope effect is in
agreement with a mechanism in which the first irreversible
step involving 1a is propargylic C–H bond cleavage.
Interestingly, under conditions using BF3·OEt2, we observed
a reasonably short induction period (vide infra) allowing for
the determination of kobs,H and kobs,D. From these rate constants,
a KIE of kobs,H/kobs,D = 1.8 ± 0.1 was obtained (Scheme 6b).33
The rate constant kobs,H(D) is determined by the overall rate of
the catalytic cycle, whereas the measured value for the
competitive KIE reports on the relative reactivity of 1a or 1aꢀ
d3 in the step where the C–H(D) bond is broken.34 Therefore,
the deviation between the KIEs from the parallel and
competitive experiments suggests that the C–H bond cleavage
is only to a minor extent turnoverꢀlimiting and that a second
step in the catalytic cycle has a slightly higher activation
barrier.35,36 An alternative explanation would be that
monomeric catalyst is delivered into the cycle via an offꢀcycle
equilibrium, which is influenced by the different C–H(D)
cleavage rates, resulting in a lower “onꢀcycle catalyst”
concentration for the reaction of 1aꢀd3. In this case, kobs,H
would be larger than kobs,D even though C–H(D) cleavage is
not turnoverꢀlimiting.37 This is considered less likely based on
KIE experiments under different conditions which indicate
that the KIE is independent of the “onꢀcycle catalyst”
concentration.12
Scale: 0.1 mmol 1a + 1aꢀd6 (1:1) (a), 1a or 1aꢀd6 (b). kH/kD
=
[(rateH/rateD)] / ([1a]/[1aꢀd6]); rateinit,H(D) is the initial rate after
addition of 1a (1aꢀd6).12
Proposed mechanism for the formation of triene 2a. We
suggest that the formation of triene 2a follows a mechanism
involving allene attack on PdII with concomitant allylic C–H
bond cleavage to give intermediate M1 (Scheme 5).
Subsequent cyclization to M4 and transmetalation with
phenylboronic acid would give M5. A reductive elimination
from the latter releases the product (2a). After reoxidation of
Pd0 to PdII by BQ, the catalyst is available for the next
catalytic cycle. We consider it likely that transmetalation can
occur either in M1 or M4 (the latter option is shown in
Scheme 5). The formation of side product 4 (cf. Scheme 2),
which is formed in small amounts, shows that at least some
early transmetalation occurs.30,31 The observed zeroth order in
the reactants is rationalized by the equilibrium between 1a and
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 15
E
E
E
E
1
2
3
Scheme 6. KIEs in the formation of vinylallene 3a
(PhBO)3, cat. Pd(OAc)2, BQ
10 mol% BF3—Et2O, H2O
•
a) intermolecular competition experiment
•
Ph
dry THFꢀd8 (0.1 M), 25 °C
4
5
6
7
8
9
0.5 equiv (PhBO)3
3a
1a
E
E
E
E
E
E
E
E
5 mol% Pd(OAc)2,
1.1 equiv BQ
•
•
+
+
•
Ph
•
Ph
10 mol% BF3—Et2O
0.5 equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
D
CD3
D
1a
1a
3a
3a
ꢀd3
ꢀd2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
kH/kD = 7.1 ± 1
b) measurement of individual rates
0.5 equiv (PhBO)3
5 mol% Pd(OAc)2
1.1 equiv BQ
E
E
E
E
•
kH/kD = 1.8 ± 0.1
Figure 11. Rate of 1a consumption vs. [PhB(OH)2] illustrating
the first order relation. The experiments are shown for the region
between 10–95% conversion or until the reaction had stopped (see
note 40). Conditions for experiments A–D: 0.1 mmol 1a; A: 2.5
mol% Pd(OAc)2, 1 equiv (PhBO)3, 1.1 equiv BQ, 0.75 equiv H2O;
B: 5 mol% Pd(OAc)2, 1 equiv (PhBO)3, 1.1 equiv BQ, 1.0 equiv
H2O; C: 5 mol% Pd(OAc)2, 0.5 equiv (PhBO)3, 1.1 equiv BQ, 0.5
equiv H2O; D: 5 mol% Pd(OAc)2, 0.5 equiv (PhBO)3, 2.0 equiv
BQ, 1.0 equiv H2O.
Ph
•
10 mol% BF3—Et2O
0.5 equiv H2O
dry THFꢀd8 (0.1 M)
25 °C
(D)H
CH(D)3
H(D)
3a
i) 1a
ii) 1aꢀd3
3aꢀd2
Scale: 0.1 mmol 1a + 1aꢀd3 (1:1) (a), 1a or 1aꢀd3 (b). kH/kD
[(rateH/rateD)] / ([1a]/[1aꢀd3]) for (a) and kH/kD = kobs,H/kobs,D for
=
(b).12
Furthermore, experiments using different amounts of
BF3·Et2O suggested that there is a slight inhibitory effect of
this additive (Figure 12a). This may be related to an
involvement of BF3·Et2O in an offꢀcycle reaction. The fastest
reaction was achieved with ca. 2 mol% of BF3·Et2O.41 The
ratio 2a/3a and also the yield of vinylallene product 3a were
unaffected proving the potency of this additive. The
dependence of the reaction rate on the added Pd(OAc)2 was
studied for 10 mol% added BF3·OEt2, and a saturation
behavior was observed (Figure 12b). This saturation occurred
at even lower catalyst loading than that in the absence of
BF3·OEt2 (cf. Figure 8). This could possibly be the result of
the optimal ratio [Pd]/[BF3·Et2O] being exceeded.42 Since the
fraction of active Pdꢀspecies is not known, it is difficult to
make assumptions about the required stoichiometry between
the Pdꢀcatalyst and the BF3·OEt2 coꢀcatalyst.43
Kinetics in the formation of vinylallene 3a. Under
selective conditions for formation of 3a as given in Scheme
6b, the phenylating carbocyclization reaction of 1a provided a
63% NMR yield of 3a and ≤5% of 2a. There was 15–20% of
material that the observed side products could not account for.
At first, we studied the relationships between the reaction
rate and BQ, 1a, (PhBO)3, and PhB(OH)2 by varying their
initial concentrations (cf. reactions A–D in Figure 11).38 An
overall reaction order of one was established as the
concentration of each of these reactants was linearly related to
the rate througout the reaction.12 Only when plotting the rate
vs. [PhB(OH)2] was the intercept zero, revealing that the
reaction is first order in [PhB(OH)2] (Figure 11).39
0.5 equiv (PhBO)3,
cat. Pd(OAc)2, 1.1 equiv BQ
E
E
E
E
cat. BF3—Et2O, 0.5 equiv H2O
•
Ph
•
dry THFꢀd8 (0.1 M), 25 °C
3a
1a
Figure 12. Influence of [BF3·Et2O] (a) and added [Pd(OAc)2] (b)
on kobs. With rate = kobs [PhB(OH)2]. Conditions: 0.1 mmol 1a; a)
5 mol% Pd(OAc)2, 1–10 mol% BF3·Et2O;41 b) 0.5–5 mol%
Pd(OAc)2, 10 mol% BF3·Et2O.
8
ACS Paragon Plus Environment

Page 9 of 15
1
2
3
4
5
6
7
Journal of the American Chemical Society
0.4 equiv (PhBO)3
5 mol% Pd(OAc)2
2 equiv BQ
The analysis of the reaction progress with regard to the
concentration of H2O is complicated by the interplay of
different equilibria (e.g. the hydrolysis of (PhBO)3) and its
consumption throughout the reaction. If a sufficient amount of
protons are available in the form of H2O, then HQ and B(OH)3
are formed as byꢀproducts of the reaction.44 For less than ca. 1
equivalent of H2O the formation of boronꢀadducts of
deprotonated HQ is expected.45 Figure 13 depicts experiments
with different amounts of (PhBO)3 and H2O. The fastest
reactions are observed for (PhBO)3/H2O = 1:1 (A and B), thus
when only minimal amounts of free H2O are expected. This
result may therefore be indicative of an inhibitory effect of
H2O, possibly due to formation of nonꢀreactive Pd3(µ2ꢀ
OH)(OAc)5.12 It is worth noting that for vinylallene formation,
the effect of H2O on kobs (reaction rate/[1st order reactant])
under conditions with BF3·Et2O is the opposite of that with
limiting H2O (Figure 13 vs. Figure 6).
E
E
E
E
•
•
Ph
10 mol% BF3—Et2O
1 equiv H2O
dry THFꢀd8 (0.1 M), 25 °C
3a
1a
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
E
E
E
E
(PhBO)3, 5 mol% Pd(OAc)2,
BQ, 10 mol% BF3—Et2O, H2O
•
•
Ph
dry THFꢀd8 (0.1 M), 25 °C
3a
1a
Figure 14. a) Reaction profile of selective phenylating
carbocyclization of 1a giving 3a under optimized conditions.
b) Rate profile of product formation. Scale: 0.1 mmol 1a.
Effect of boronic acid substituent in the formation of
vinylallene. According to the KIE measurements using
(PhBO)3, the propargylic C–H bond activation is only to a
minor extent turnoverꢀlimiting in the formation of 3a (cf.
Scheme 6). Since we observed a first order dependence on
[PhB(OH)2] (Figure 11), the transmetalation is to a major
extent turnoverꢀlimiting, probably due to a slightly higher
activation barrier than that for the C–H bond cleavage.46 This
transmetalation step is expected to be sensitive to electronic
changes. We therefore studied the reaction with seven paraꢀ
substituted boroxines under the optimized conditions (Figure
15) and performed additional KIE measurements for the most
electronꢀdonating and electronꢀwithdrawing substituents (4ꢀ
OMe and 4ꢀNO2, respectively) (Scheme 7).
Figure 13. Rate of 1a consumption vs. [PhB(OH)2] illustrating
the effect of (PhBO)3/H2O. Variation in [BQ] is expected to be
irrelevant. The experiments are shown for the region between 10–
95% conversion or until the reaction had stopped, see note 40.
Conditions: 0.1 mmol 1a, (PhBO)3 (A: 1 equiv, B, C, E: 0.5
equiv, D: 0.4 equiv), H2O (A, C, D: 1 equiv, B: 0.5 equiv, E:
3 equiv), BQ (A, B, E: 1.1 equiv, C: 1.5 equiv, D: 2 equiv).
In some of the performed experiments a process was
observed which consumed HQ, but not BQ and led to
decomposition of product 3a into undetectable species. This
decomposition was suppressed by decreasing the
concentration of (PhBO)3 and the product yield was increased
with higher BQ concentration. After optimization of the
reaction conditions based on the kinetic analysis, the reaction
provided an NMR yield of almost 80% with <1% of 2a. The
induction was found to be completed at ca. 20% conversion
(Figure 14).
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 15
1
2
3
4
5
6
7
8
0.4 equiv (ArBO)3
5 mol% Pd(OAc)2
2 equiv BQ
equilibrium between catalystꢀbound and free substrate 1a.
E
E
E
E
With this in mind we consider the two possible mechanisms
for the formation of 3a shown in Scheme 8. The proposed
mechanism A starts with reversible complexation of the active
PdII species by substrate 1a leading to an adduct which
subsequently can undergo C–H bond cleavage to give M2.
After this step, which was found to be partially turnoverꢀ
limiting when using PhB(OH)2, the allenylpalladium
intermediate M2 undergoes cyclization, followed by reductive
elimination from allylpalladium species M7 to provide
product 3a. Transmetalation with PhB(OH)2 may occur either
in intermediate M2 or M6 (the latter option is shown in
Scheme 8A). In the second possible mechanism (Scheme 8B),
the catalytic cycle starts with the transmetalation between the
catalyst and PhB(OH)2 which is followed by reversible
complexation of the substrate, C–H bond cleavage,
irreversible cyclization, and reductive elimination. Both
proposed mechanisms A and B can only account for the
observed first order dependence on [PhB(OH)2] if
transmetalation is slow and to a major extent turnoverꢀ
limiting. Furthermore, the observed zeroth order in starting
material 1a is rationalized by M6 or L2PdX2 being resting
states in mechanism A and B, respectively. The high
regioselectivity in the last step of both alternative mechanisms
(the reductive elimination from a πꢀallyl Pd intermediate) may
be attributed to the formation of the stabilized πꢀsystem.48
Mechanisms A and B can both rationalize the formation of the
side products that were observed in the absence of BF3·Et2O
for substrates with a propargylic CH2 (cf. Scheme 2) or a
propargylic CH group instead of CH3.12 This suggests that an
allenylꢀpalladium intermediate (such as M2 or M10) is also
•
Ar
•
10 mol% BF3—Et2O
1 equiv H2O
dry THFꢀd8 (0.1 M), 25 °C
3
1a
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 15. Substitutent effect in the formation of vinylallenes 3.
a) Rate vs. [ArB(OH)2] for selected boronic acids illustrating the
change in reaction order for 4ꢀNO2ꢀC6H4ꢀB(OH)2. b) The
deviation from linearity in the plot ∆[3]/∆t vs. [ArB(OH)2]
illustrated as the xꢀaxis intercept vs. σpara.
12,47 Scale: 0.1 mmol 1a.
The following yields were obtained (corresponding reaction time
and conversion in brackets): OMe: 70% (11 h, 92%); tꢀBu: 79%
(10 h, 99%); Me: 76% (13 h, 100%); H: 73% (5 h, 91%; cf.
Figure 14); Br: 81% (6 h, 99%); CHO: 80% (5 h, 98%); CF3: 82%
(6 h, 98%); NO2: 77% (5 h, 95%).
Scheme 7. KIEs from individual rates in the formation of
vinylallene 3 using different boronic acids
0.4 equiv (ArBO)3
5 mol% Pd(OAc)2
2 equiv BQ
E
E
E
E
involved in the pathway towards
3 under unselective
•
conditions and can be intercepted by a reaction with HQ to
•
Ar
give 4ꢀhydroxyꢀphenoxyꢀsubstituted compounds 5 and 6.49
10 mol% BF3—Et2O
1 equiv H2O
dry THFꢀd8 (0.1 M), 25 °C
(D)H
CH(D)3
H(D)
i) 1a
3
ii) 1aꢀd3
3ꢀd2
for Ar = 4ꢀMeOꢀC6H4:
kH/kD = 2.1 ± 0.2
kH/kD = 1.7 ± 0.2
kH/kD = 6.5 ± 1.7
for Ar = C6H5:
for Ar = 4ꢀNO2ꢀC6H4:
Scale: 0.1 mmol 1a or 1aꢀd3. kH(D) was obtained as the slope in
∆[3]/∆t vs. [ArB(OH)2] for Ar = C6H5, 4ꢀMeOꢀC6H4 or in
∆[3ab]/∆t vs. [1a] for Ar = 4ꢀNO2ꢀC6H4; 2ꢀd3 was formed in
varying amounts; for Ar = 4ꢀMeOꢀC6H4: 14%, for Ar = C6H5:
11%, for Ar = 4ꢀNO2ꢀC6H4: 8%.12
Electronꢀdonating groups (OMe, tꢀBu, Me) as well as Br as
substituents displayed very similar kinetics as the
unsubstituted phenyl boroxine – a first order dependence on
[ArB(OH)2] and for the methoxy substituent a KIE of kH/kD =
2.1 ± 0.2 (vs. 1.7 ± 0.2 for Ar = C6H5). However, for electronꢀ
withdrawing substituents (CHO, CF3, NO2), and thus with a
facilitated transmetalation, the turnoverꢀlimiting step is shifted
towards the C–H cleavage. This resulted in a first order
dependence on the starting material 1a and a zero order
dependence on the arylboronic acid. For Ar = 4ꢀNO2ꢀC6H4, the
KIE was measured and a large value of kH/kD = 6.5 ± 1.7 was
determined (Scheme 7).12
Proposed mechanism for formation of vinylallenes. The
KIE measurements have clarified that the essential feature for
the formation of phenylated 3a is the propargylic C–H bond
cleavage occurring prior to the cyclization and that there is an
10
ACS Paragon Plus Environment

Page 11 of 15
Journal of the American Chemical Society
1
Scheme 8. Possible mechanisms for the formation of 3a in the presence of BF3·Et2O
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Only those equilibria are indicated that are crucial to match the experimental results.
Most kinetic results for the selective formation of
vinylallenes 3 from substrate 1a are easily rationalized with
either mechanism A or B in Scheme 8 as illustrated by the
following discussion. When electronꢀdeficient arylboronic
acids such as 4ꢀNO2ꢀC6H4ꢀB(OH)2 are used, the rate of the
reaction shows a first order dependence on [1a], and a zero
order dependence on [ArB(OH)2]. This is explained by
facilitated transmetalation shifting the turnoverꢀlimiting step
from transmetalation to C–H bond cleavage (M3→M2 or
M9→M10), and the resting state to the substrateꢀbinding
equilibrium (L2PdX2+1a⇌M3 or M8+1a⇌M9). These
changes are in line with a distinctively larger KIE value for the
4ꢀNO2ꢀC6H4ꢀB(OH)2 than for the more electronꢀrich
derivatives.
explanation for the differences in the yield of 2ꢀd3. In contrast,
these are difficult to account for in mechanism A.
However, on the basis of the present data it is difficult to
conclusively discern mechanisms A from B, and it could well
be that a mixture/crossꢀover of pathways is involved in the
selective formation of vinylallenes 3.12
Discussion of the origin of selectivity. The high control of
product selectivity in the arylating carbocyclization
(Scheme 1) is of great interest for synthetic applications.
Despite the extensive study of this reaction, the obtained
experimental data do not provide any conclusive evidence
regarding the origin of this selectivity and the question
remaining to be answered is why the allene is activated in the
presence of water, whereas the alkyne is activated in the
presence of BF3·Et2O (Scheme 9).
The general influence of BF3·Et2O on the arylating
cyclization (Scheme 8) could be the formation of more
cationic PdII complexes containing more weakly coordinating
counterions X– compared to reaction mixtures without
BF3·Et2O. For example, BF3·Et2O could promote early
transmetalation as suggested in mechanism B, i.e. accelerating
,
it compared to the complexation with substrate 1a.50 51 On the
other hand, BF3·Et2O could be responsible for propargylic C–
H abstraction (vide infra) in either mechanism A or B.
When the reaction was performed with 1aꢀd3, triene 2ꢀd3
was formed as side product. Its amount was found to depend
on the arylboronic acid used, with higher yields for more
electronꢀrich arylboronic acids (Scheme 7). In mechanism B,
the electronic properties of intermediate M9 (which undergoes
either propargylic or allenic C–H bond cleavage) are
influenced by substitution of ArB(OH)2, thus offering an
11
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 15
1
Scheme 9. Summary of proposed mechanisms for the
zeroth order in substrate and BQ, first order in PhB(OH)2 and
2
3
4
5
6
7
8
9
formation of 2a and 3a
inverse order in BF3·OEt2. The energy barriers for C–H
activation and transmetalation should be similar to one another
since the turnoverꢀlimiting step can be altered by changing the
Ar group in ArB(OH)2 from an electronꢀrich to an electronꢀ
deficient one. Additionally, the observed change in reactivity
of differently paraꢀsubstituted ArB(OH)2 indicates that
BF3·OEt2 may promote an early transmetalation.
The present investigation has provided
a
better
understanding of the mechanism for oxidative PdIIꢀcatalyzed
carbocyclizations involving allenes, where arylboronic acids
are used in cascade reactions. These reactions are important in
creating novel carbocycles of relevance for natural products
and biologically active compounds. A detailed knowledge of
the mechanism of these and related reactions, including the
relevance of underlying processes is crucial for the further
development of applications of cyclization chemistry in
selective synthesis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
One explanation could be that in the absence of BF3·Et2O
the reaction is frontier orbitalꢀcontrolled, whereas in the
presence of BF3·Et2O the reaction becomes charge
,
controlled.52 53 In the frontier orbitalꢀcontrolled reaction, the
ASSOCIATED CONTENT
interaction between the LUMO on PdII and the HOMO on the
allene and alkyne, respectively, would determine the outcome
of the reaction. A simple calculation shows that the HOMO of
an allene is higher in energy than the HOMO of an alkyne,54
which would lead to preferential reaction of the allene in the
frontier orbitalꢀcontrolled reaction. Under chargeꢀcontrolled
reaction conditions, the alkyne is expected to react faster than
the allene due to a larger partial negative charge of the triple
bond.
Supporting Information. Experimental and spectral details,
detailed experimental procedures, additional data, and discussions
of the results of kinetic experiments are available free of charge
via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
* jeb@organ.su.se, gw@organ.su.se
S
UMMARY. The mechanistic investigation of the oxidative
Author Contributions
phenylating carbocyclizations using deuterium labeled
derivatives of allenyne substrate 1a provided conclusive
evidence for selectivity being determined in the first
irreversible step after substrate binding. Allenic C–H
abstraction leads to functionalized triene 2a while propargylic
C–H abstraction gives functionalized vinylallenes 3. The KIE
results also demonstrate that the C–H abstraction is turnoverꢀ
limiting for the formation of 2a, whereas in the reaction
employing BF3·OEt2 to give 3a it is partially turnoverꢀlimiting
for electronꢀrich arylboronic acids and fully turnoverꢀlimiting
for electronꢀdeficient boronic acids. Kinetic measurements of
the phenylating carbocyclization without added BF3·OEt2
revealed that there is a long induction period due to the
activation of the Pd preꢀcatalyst, which requires the presence
of both arylboronic acid and BQ. Under these conditions, the
kinetics of the two pathways leading to trienes 2 or
vinylallenes 3, respectively, appear quite different from one
another. The formation of trienes 2 only occur in the presence
of ArB(OH)2, whereas the rate of vinylallene formation
exhibits a first order dependence on (ArBO)3 after the
consumption of ArB(OH)2. When the reaction using (PhBO)3
is performed with an excess of H2O to give selectively triene
2a, the observed zero order in all the reactants agrees with the
C–H cleavage being the slowest step in the catalytic cycle (cf.
Scheme 5). The reaction order in Pd(OAc)2 appeared to be of
complex nature due to a saturation behavior. Although the
active catalytic species was not identified, the conversion of
Pd3(µ2ꢀOAc)6 to Pd3(µ2ꢀOH)(OAc)5 in the presence of water
was observed. Under conditions employing BF3·OEt2 only a
very short induction period was observed, possibly related to
the conversion of Pd3(µ2ꢀOAc)6 to the active species. This
reaction giving phenylated vinylallene 3a was found to be
†These authors contributed equally.
Funding Sources
Financial support from the Swedish Research Council (2016ꢀ
03897 and 2013ꢀ4859) and The Knut and Alice Wallenberg
Foundation is gratefully acknowledged.
ACKNOWLEDGMENT
We thank Dr. Javier Mazuela for help with the preparation of
starting materials and Dr. Youqian Deng for initial control
experiments.
REFERENCES
(1) For selected recent reviews on C–H activation, see: (a)
Ackermann, L. Acc. Chem. Res. 2014, 47, 281. (b) Arockiam, P. B.;
Bruneau, C.; Dixneuf, P. H. Chem. Rev. 2012, 112, 5879. (c) Kuhl,
N.; Hopkinson, M. N.; WencelꢀDelord, J.; Glorius, F. Angew. Chem.,
Int. Ed. 2012, 51, 10236. (d) Ackermann, L. Chem. Rev. 2011, 111,
1315.
(2) For siteꢀselective C–H functionalization, see: (a) Neufeldt, S.
R.; Sanford, M. S. Acc. Chem. Res. 2012, 45, 936 and references
herein. (b) Cook, A. K.; Emmert, M. H.; Sanford, M. S. Org. Lett.,
2013, 15, 5428. (c) Su, Y.; Zhou, H.; Chen, J.; Xu, J.; Wu, X.; Lin,
A.; Yao, H Org. Lett. 2014, 16, 4884. For enantioselective C–H
functionalization, see: (d) Osberger, T. J.; White, M. C. J. Am. Chem.
Soc. 2014, 136, 11176. (e) Chai, Z.; Rainey, T. J. J. Am. Chem. Soc.
2012, 134, 3615. For siteꢀselective βꢀhydride elimination, see: (f)
Werner, E. W.; Sigman, M. S. J. Am. Chem. Soc. 2010, 132, 13981.
(3) (a) Lyons, T. W.; Hull, K. L.; Sanford, M. S. J. Am. Chem. Soc.
2011, 133, 4455. (b) Lotz, M. D.; Remy, M. S.; Lao, D. B.; Ariafard,
A.; Yates, B. F.; Canty, A. J.; Mayer, J. M.; Sanford, M. S. J. Am.
Chem. Soc. 2014, 136, 8237. (c) Leskinen, M. V.; Madarász, A.; Yip,
K.ꢀT.; Vuorinen, A.; Pápai, I.; Neuvonen, A. J.; Pihko, P. M. J. Am.
12
ACS Paragon Plus Environment

Page 13 of 15
1
Journal of the American Chemical Society
Chem. Soc. 2014, 136, 6453. (d) Pereira, K. C.; Porter, A. L.;
(21) (∆[3a]/∆t) / [(PhBO)3] was found to be proportional to the rate
2
3
4
5
6
7
8
9
Potavathri, S.; LeBris, A. P.; DeBoef, B. Tetrahedron 2013, 69,
4429. (e) Cheng, G.ꢀJ.; Yang, Y.ꢀF.; Liu, P.; Chen, P.; Sun, T.ꢀY.; Li,
G.; Zhang, X.; Houk, K. N.; Yu, J.ꢀQ.; Wu, Y.ꢀD. J. Am. Chem. Soc
2014, 136, 894. (f) Engelin, C.; Jensen, T.; RodriguezꢀRodriguez, S.;
Fristrup, P. ACS Catal. 2013, 3, 294.
(4) For purely computational studies, see: (a) Yang, Y.ꢀF.; Cheng,
G.ꢀJ.; Liu, P.; Leow, D.; Sun, T.ꢀY.; Chen, P.; Zhang, X.; Yu, J.ꢀQ.;
Wu, Y.ꢀD.; Houk, K. N. J. Am. Chem. Soc. 2014, 136, 344. (b) Zhang,
S.; Shi, L.; Ding; Y. J. Am. Chem. Soc. 2011, 133, 20218. (c) Sköld,
C.; Kleimark, J.; Trejos, A.; Odell, L. R.; Nilsson Lill, S. O.; Norrby,
P.ꢀO.; Larhed, M. Chem. Eur. J. 2012, 18, 4714.
of formation of Pd3(µ2ꢀOH)(OAc)5 (see Figure S27). In a separate
experiment, it was excluded that [2a] influences the rate of formation
of 3a, see Figure S14.
(22) Tokunaga, Y.; Ueno, H.; Shimomura, Y.; Seo, T. Heterocycles
2002, 57, 787.
(23) In the case of (4ꢀNO2ꢀC6H4ꢀBO)3 the accurate quantification
of [(ArBO)3] is problematic because of overlaying 1H NMR signals in
the aromatic region and therefore (4ꢀCF3ꢀC6H4ꢀBO)3 was chosen as an
exemplary electronꢀdeficient aryl boroxine.
(24) Due to similar concentration profiles for 1a and [(ArBO)3], it
is not possible to unequivocally determine which is the first order
reactant. However, for (4ꢀNO2ꢀC6H4ꢀBO)3 as well as for (C6H5ꢀBO)3 a
first order dependence on 1a can be ruled out.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(5) Deng, Y.; Bartholomeyzik, T.; Persson, A. K. Å.; Sun, J.;
Bäckvall, J.ꢀE. Angew. Chem. Int. Ed. 2012, 51, 2703.
(6) Bartholomezyik, T.; Mazuela, J.; Pendrill, R.; Deng, Y.;
Bäckvall, J.ꢀE. Angew. Chem. Int. Ed. 2014, 53, 8696.
(7) For allenynes, see: (a) ref 5. (b) Deng, Y.; Bartholomeyzik, T.;
Bäckvall, J.ꢀE. Angew. Chem. Int. Ed. 2013, 52, 6283.
(8) For enallenes, see: (a) Franzén, J.; Bäckvall, J.ꢀE. J. Am. Chem.
Soc. 2003, 125, 6056. (b) Persson, A. K. Å.; Jiang, T.; Johnson, M.
T.; Bäckvall, J.ꢀE. Angew. Chem. Int. Ed. 2011, 50, 6155. (c) Jiang,
T.; Persson, A. K. Å.; Bäckvall, J.ꢀE. Org. Lett. 2011, 13, 5838. (d)
Jiang, T.; Bartholomeyzik, T.; Mazuela, J.; Willersinn, J.; Bäckvall,
J.ꢀE. Angew. Chem. Int. Ed. 2015, 54, 6024–6027.
(25) (a) Cunningham, I. D.; Fassihi, K. J. Mol. Catal. A: Chem.
2005, 232, 187. (b) Driver, T. G.; Woerpel, K. A. J. Am. Chem. Soc.
2004, 126, 9993.
(26) It is common that preꢀcatalyst saturation is attributed to an offꢀ
cycle equilibrium with oligomeric forms. However, the reaction
studied here differs from published examples due to the pronounced
induction period indicating that equilibration is slow and that an
equilibrium state is not reached within the reaction time. (a) ref 20b
(b) Rosner, T.; Le Bars, J.; Pfaltz, A.; Blackmond, D. G. J. Am. Chem.
Soc. 2001, 123, 1848.
(9) For dienallenes, see: (a) Löfstedt, J.; Närhi, K.; Dorange, I.;
Bäckvall, J.ꢀE. J. Org. Chem. 2003, 68, 7243. (b) Piera, J.; Persson,
A.; Caldentey, X.; Bäckvall, J.ꢀE. J. Am. Chem. Soc. 2007, 129,
14120. (c) Karlsson, E. A.; Bäckvall, J.ꢀE. Chem. Eur. J. 2008, 14,
9175.
(27) For a KIE study of the related borylating oxidative carboꢀ
cyclization of allenynes, see ref 7b.
(28) A pathway beginning with an arylpalladation followed by a
carbocyclization and ending with a βꢀelimination is ruled out by the
competition experiment. There would be no competitive deuterium
isotope effect observed if the cleavage of the C–H bond would occur
after any irreversible step involving substrate 1a (see ref 7b).
(29) This large value refers to a primary KIE, but is also
affected by a secondary KIE. A possible contribution of hydrogen
tunneling should also be considered. However, a primary KIE may
reach ca. 10 (at room temperature) if bending vibrations are taken into
consideration, see: Bell, R. P. Chem. Soc. Rev. 1974, 3, 513.
(30) A pathway towards 2a involving transmetalation followed by
complexation of 1a cannot be excluded, but is considered less likely
due to the absence of a base.
(31) The formation of 4 is proposed to proceed through a
dienylpalladium intermediate such as M1. Other pathways may also
be possible, see: Deagostino, A.; Prandi, C.; Tabasso, S.; Venturello,
P. Molecules 2010, 15, 2667.
(32) To ensure reproducible conditions, it was necessary to use
(PhBO)3 and a defined amount of H2O (0.5 equiv after a quick
optimization) in anhydrous solvent.
(10) Deng, Y.; Bäckvall, J.ꢀE. Angew. Chem. Int. Ed. 2013, 52,
3217.
(11) The approach to monitor the entire course of the reaction is
also referred to as Reaction Progress Analysis: (a) Blackmond, D. G.
Angew. Chem. Int. Ed. 2005, 44, 4302. (b) Mathew, J. S.; Klussmann,
M.; Iwamura, H.; Valera, F.; Futran, A.; Emanuelsson, E. A. C.;
Blackmond, D. G. J. Org. Chem. 2006, 71, 4711.
(12) For more information see the SI.
(13) Please note that “Pd(OAc)2” is used as a simplification and the
material used consists of the trimeric [(Pd(OAc)2]3. The existence of
1
the trimeric form of Pd(OAc)2 in THFꢀd8 was confirmed through H
NMR diffusion measurement (DOSY) and has also been reported
previously (see ref 20a).
(14) Regarding protodeboronation of PhB(OH)2, see: Kuivila, H.
G.; Reuwer, J. F., Jr.; Mangravite, J. A. J. Am. Chem. Soc. 1964, 86,
2666.
(15) (a) Rosner, T.; Pfaltz, A.; Blackmond, D. G. J. Am. Chem.
Soc. 2001, 123, 4621. (b) Singh, U. K.; Strieter, E. R.; Blackmond, D.
G.; Buchwald, S. L. J. Am. Chem. Soc. 2002, 124, 14104. (c) Strieter,
E. R.; Blackmond, D. G.; Buchwald, S. L. J. Am. Chem. Soc. 2003,
125, 13978.
(33) This value is affected by a secondary KIE.
(34) Simmons, E. M.; Hartwig, J. F. Angew. Chem. Int. Ed. 2012,
51, 3066.
(35) The possibility that the observed KIE results from a C–H(D)
cleavage during the reoxidation of Pd0 (and hence the formation of
HQ from BQ) seems not compatible with the observed zeroth order in
BQ. The data would also be in agreement with two completely
different turnoverꢀlimiting steps for the reactions of 1a or 1aꢀd3,
respectively. Considering the relatively small difference in the
activation barriers for the C–H(D) cleavage, this explanation seems
less likely. See also note 46. The magnitude of KIE = 1.8 itself would
also be in agreement with an equilibrium isotope effect (see ref 34).
However, this scenario cannot explain the difference between KIE
(16) In the preꢀstirring experiments, yields of 2a and selectivities
were marginally improved.
(17) Conversion of Pd3(µ2ꢀOAc)6 to Pd3(µ2ꢀOAc)5OH was
observed during the preꢀstirring period. However, we were unable to
relate this conversion to the degree of activation (cf. SI section
S6.11).
(18) Water can be excluded as an important factor under these
conditions (3 equiv H2O), because increasing the initial water content
in a “regular” reaction from 3 to 6 equiv does not lead to an
accelerated rate (cf. Figure S11).
and KIEcomp
.
(19) For 0.5 equiv H2O <2% of 1a remained after 20 h; for 0.3 equiv
H2O <5% of 1a remained after 23 h; without additional H2O 9% of 1a
remained after 48 h, and 56% of 3a was formed.
(20) (a) Bedford, R. B.; Bowen, J. G.; Davidson, R. B.; Haddow,
M. F.; SeymourꢀJulen, A. E.; Sparkes, H. A.; Webster, R. L. Angew.
Chem. Int. Ed. 2015, 54, 6591. (b) Mekareeya, A.; Walker, P. R.;
CouceꢀRios, A.; Campbell, C. D.; Steven, A.; Paton, R. S; Anderson,
E. A. J. Am. Chem. Soc. 2017, 139, 10104.
(36) For examples where a modest KIE was explained by a
partially rateꢀdetermining step, i.e. that there is a second step with a
similar barrier, see: (a) ref 3f. (b) Bissember, A. C.; Levina, A.; Fu, G.
C. J. Am. Chem. Soc. 2012, 134, 14232.
(37) An analogous rationalization is described in ref 20b.
(38) [PhB(OH)2] was varied by regulating the hydrolysis of
(PhBO)3 with the amount of added water. We did not see a significant
13
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 15
1
2
3
4
5
6
7
8
change in the (PhBO)3/PhB(OH)2 equilibrium related to the presence
of BF3·Et2O, see SI.
(39) In some cases decomposition of product 3a took place after all
starting material was consumed. It is plausible that some
decomposition also occurs during the reaction. Therefore, the reaction
rate is generally given based on the consumption of 1a. Nonetheless,
some randomly chosen examples showed that using the rate based on
formation of 3a led to the same conclusions.
(40) The region below 10% of the maximal rate where noise
becomes dominating was excluded.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(41) Experimentally it was not feasible to test lower [BF3·Et2O]
(1 mol% ≙ 0.12 µL). In the experiment, 0.12 µL BF3·Et2O were
measured. However, according to the integration in the 1H NMR
spectrum, 2 mol% were added.
(42) Alternatively, there could be an offꢀcycle equilibrium
involving active catalyst, see ref 26a,b.
(43) Both Pd3(µ2ꢀOAc)6 and Pd3(µ2ꢀOH)(OAc)5 have been
identified in the reaction mixtures (cf. ref 20). Neither was found to
be the active catalytic species which appeared to form from
Pd3(µ2ꢀOAc)6 rather than from Pd3(µ2ꢀOH)(OAc)5 (see section S7.4 in
the SI).
(44) The reduction of BQ to HQ requires two protons. One of them
may come from the initial C–H abstraction involving the starting
material 1a.
(45) It was not possible to characterize these compounds as they
are of transient nature, and do not have resolved peaks in the 1H NMR
spectrum.
(46) We have confirmed a first order dependence on [PhB(OH)2]
also for the reaction of 1aꢀd3. This is important since it serves as
evidence that the reactions of 1a and 1aꢀd3 follow the same
mechanisms (elementary steps).
(47) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165.
(48) Another explanation is that the initially formed
(σꢀallyl)palladium complex A is highly reactive and undergoes
reductive elimination to give product faster than it rearranges to the
corresponding (πꢀallyl)palladium complex B.
(49) We have not detected a diallene sideꢀproduct that would result
from reductive elimination in intermediate M10 in mechanism B (this
would also apply to the case in mechanism A for transmetalation
occurring prior to cyclization).
(50) In a control experiment the transmetalation between Pd(OAc)2
and PhB(OH)2 (1:1 mixture) in THFꢀd8 was shown to proceed very
slowly and only trace amounts of biphenyl were formed. In the
presence of 20 mol% BF3·Et2O, full conversion was observed after 4
h.
(51) The formation of biphenyl as a side product was not observed.
Possibly, the complexation of substrate is more beneficial compared
to a second transmetalation.
(52) Fleming, I. ”Frontier Orbitals and Organic Chemical
Reactions”, Wiley, 1976.
(53) It is likely that the acetate is efficiently removed from
palladium upon addition of BF3·Et2O, which would lead to a cationic
PdII complex such as [1a(Pd)(κ2ꢀOAc)] with BF3(OAc)– as
counterions.
(54) A DFT calculation of simple allene (CH2=C=CH2) and simple
acetylene (CH≡CH) using B3LYP/6ꢀ31G(d,p) gave HOMOallene
=
–7.16 eV and HOMOacetylene = –7.68 eV.
14
ACS Paragon Plus Environment

Page 15 of 15
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
ACS Paragon Plus Environment