Bis(imino)acenaphthene (BIAN)-supported palladium(ii) carbene complexes as effective C-C coupling catalysts and solvent effects in organic and aqueous media

Article abstract of DOI:10.1039/c4cy00192c

The synthesis and catalytic properties of two new 1,2-acenaphthenyl N-heterocyclic carbene-supported palladium(ii) catalysts are presented. The acenaphthenyl carbene has been prepared with mesityl or 1,5-diisopropyl N-aryl substituents. Comprehensive catalytic studies for the Suzuki coupling of aryl halides with aryl boronic acids have been conducted. In general, the diisopropyl-functionalised catalyst showed superior selectivity and reactivity. A comparison of the catalytic performances in dichloromethane, toluene and water at low temperatures (30-40 °C) is also presented. Both catalysts were proficient in the homogeneous Suzuki coupling of aryl iodides, bromides and chlorides with boronic acids in dichloromethane. Similar reactions in water led to the formation of insoluble colloidal catalytic species that still exhibited high activity in the Suzuki reaction with aryl chlorides. Reactions performed in toluene showed intermediate results; partial catalyst decomposition led to concomitant homogeneous and heterogeneous catalysis. The heterogeneous palladium precipitates could be easily recovered by filtration and reactivated for subsequent use. Activation energies determined for aryl bromide-based Suzuki reactions were found to be in the range of 159-171 kJ mol^-1 in organic solvents and 111-116 kJ mol^-1 in water. The corresponding activation energy for the aryl chloride was found to be 322 kJ mol^-1 in water. the Partner Organisations 2014.

Full text of DOI:10.1039/c4cy00192c

Catalysis
Science &
Technology
View Article Online
View Journal | View Issue
PAPER
Bis(imino)acenaphthene (BIAN)-supported
palladium(II) carbene complexes as effective C–C
coupling catalysts and solvent effects in organic
and aqueous media†
Cite this: Catal. Sci. Technol., 2014,
4, 1456
Katherine A. Crawford, Alan H. Cowley and Simon M. Humphrey*
The synthesis and catalytic properties of two new 1,2-acenaphthenyl N-heterocyclic carbene-supported
palladium(II) catalysts are presented. The acenaphthenyl carbene has been prepared with mesityl or
1
,5-diisopropyl N-aryl substituents. Comprehensive catalytic studies for the Suzuki coupling of aryl
halides with aryl boronic acids have been conducted. In general, the diisopropyl-functionalised catalyst
showed superior selectivity and reactivity. A comparison of the catalytic performances in dichloromethane,
toluene and water at low temperatures (30–40 °C) is also presented. Both catalysts were proficient in
the homogeneous Suzuki coupling of aryl iodides, bromides and chlorides with boronic acids in
dichloromethane. Similar reactions in water led to the formation of insoluble colloidal catalytic species
that still exhibited high activity in the Suzuki reaction with aryl chlorides. Reactions performed in toluene
showed intermediate results; partial catalyst decomposition led to concomitant homogeneous and
heterogeneous catalysis. The heterogeneous palladium precipitates could be easily recovered by filtration
and reactivated for subsequent use. Activation energies determined for aryl bromide-based Suzuki
Received 13th February 2014,
Accepted 1st March 2014
DOI: 10.1039/c4cy00192c
−1
−1
reactions were found to be in the range of 159–171 kJ mol in organic solvents and 111–116 kJ mol in
−1
www.rsc.org/catalysis
water. The corresponding activation energy for the aryl chloride was found to be 322 kJ mol in water.
6
–9
and allyl moieties.
Suzuki catalysts often employ bulky
Introduction
phosphine ligands that result in optimal catalytic activity and
selectivity. However, phosphine ligands are typically suscep-
The Suzuki–Miyaura cross-coupling reaction is used extensively
in industrial processes for the formation of carbon–carbon
10–12
tible to oxidation and metal-ligand degradation.
As a
1
bonds. This reaction plays an important industrial-scale role
consequence, N-heterocyclic carbenes (NHCs) have attracted
interest as more stable supporting ligands. NHCs feature a
significantly stronger electron donating character than phos-
phines, which enhances the stability of resulting complexes
towards heat, air and moisture. The strong σ-donating ability
of the NHCs also promotes oxidative addition of aryl halides
for the production of natural products, pharmaceuticals, and
2–4
agrochemicals.
Typically, Suzuki coupling is performed in
an organic solvent using homogeneous catalysts that are
designed to generate exclusively the hetero-coupled biaryl
product. The C–C coupled products are obtained by the treat-
ment of an aryl halide with an organoborane in the presence
13,14
and can assist in reductive elimination of biaryl products.
5
of a suitable base (Scheme 1). Complexes based on group X
Recent advances in ‘green’ Suzuki coupling have focused
on developing catalysts that are easily recoverable and can
operate in environmentally benign solvents. Some pertinent
metals are the most proficient in this task and can be ren-
dered soluble in appropriate organic solvents using a range
of spectator ligands such as 3-chloropyridine, triethylamine,
1
5–17
examples include, heterogenised catalysts (e.g., nanoparticles),
1
8–20
metallopolymers,
and single-site catalysts grafted onto
21–23
insoluble support media (e.g., silica or alumina).
For example,
Department of Chemistry, The University of Texas at Austin, Welch Hall 2.204,
Sekar et al. generated active catalysts for Suzuki, Heck, and
1
05 E. 24th St A5300, Austin, Texas 78712-1224, USA. E-mail: smh@cm.utexas.edu;
Tel: +1 512 471 0312
Electronic supplementary information (ESI) available: Synthetic procedures,
Sonogashira cross-coupling reactions using covalently-bonded
1
5
†
Pd–C(binaphthyl) nanoparticles. These nanoparticles formed
quantitatively C–C bonds even after several cycles of catalyst
UV-Vis for catalysts 1 & 2; TEM images of palladium nanoparticles; additional
catalytic activity data under alternative reaction conditions; additional selec-
tivity data for 1 & 2; Arrhenius plot of the arylchloride. CCDC 976184. For ESI
and crystallographic data in CIF or other electronic format see DOI: 10.1039/
c4cy00192c
1
5
use and recovery. Meanwhile, Ma et al. have prepared active
Pd(II) Suzuki catalysts by direct anchoring to SBA-16-type
silica, which afforded an easily recoverable and recyclable
1456 | Catal. Sci. Technol., 2014, 4, 1456–1464
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
were obtained using aryl chlorides at temperatures as low as
40 °C.
Results and discussion
1. Catalyst synthesis and structures
The present work has focused on the catalytic activities of
two related Pd(II)–carbene complexes with the objective of
expanding the scope of this family of catalysts. The carbene
ligands in question (A-I and A-II) were prepared using the
3
0
previously reported route for B-II (Scheme 2). The silver(I)–
carbene complexes B-I and B-II were treated directly with
Scheme
reaction.
1 Generic mechanism for the Suzuki–Miyaura coupling
PdCl
2
(PPh
3
)
2
in toluene to give the corresponding Pd(II) com-
PPh (1) and (IPr)PdCl PPh (2) by trans-
plexes, (IMes)PdCl
2
3
2
3
metalation and elimination of AgCl (Scheme 2; see ESI† for
further synthetic details). The single crystal X-ray diffraction
structure of 2 was obtained by growing suitable crystals by slow
evaporation of CH Cl in air (Fig. 1 and Table 1).‡ The metal
site Pd1 in 2 displays a somewhat distorted square-planar
environment, most likely imposed by the steric bulk of the
N-diisopropyl aryl groups that are rotated approximately per-
heterogeneous composite catalyst that showed excellent activ-
ity even after prolonged recycling.
2
1
2
2
The pursuit of greener Suzuki catalysts has also resulted
in the development of molecular complexes that are catalyti-
cally active in environmentally-benign and renewable solvents,
such as water. Morales-Morales et al. utilized water-soluble
pincer ligands in aqueous media to convert p-substituted aryl
bromides to the corresponding hetero-coupled biphenyl prod-
3 2
pendicular to the C N ring plane, which is commonly
31
observed for such Brookhart-type complexes. The Pd1–C1
bond distance is 2.04 Å while the trans-oriented phosphine
Pd1–P1 bond is 2.32 Å. The average Pd–P bond length
obtained from a survey of similar NHC–phosphine Pd(II)
structures in the CCDC is somewhat shorter, at 2.267 Å. This
may be indicative of a particularly strong trans-influence
exerted by the acenaphthene carbene ligand. As expected, the
fused five- and six-membered aromatic acenaphthene structure
in 2 displays a distinct curvature with a maximum out-of-plane
24
ucts in appreciable yields. Such aqueous reactions raise
an important question: do homogeneous molecular catalysts
remain intact during Suzuki reaction conditions, or do the
catalysts decompose to generate new (presumably heteroge-
neous) catalyst species? A technique that is commonly employed
to address this question involves the injection of excess Hg(0)
into the reaction mixture to poison any leached Pd(0) spe-
2
5,26
cies.
Alternatively, the likely identity of the active Suzuki
cross-coupling species can be determined via a filtration test,
in which the reaction is interrupted and separated into a
supernatant layer and a precipitate; both phases are then
independently probed for the presence of trace metal species
2
7,28
and catalytic reactivity.
In the present work, two new 1,2-acenaphthenyl N-hetero-
cyclic carbene-supported palladium(II) catalysts have been
synthesized with the objective of assessing their comparative
performance in the Suzuki reaction, when conducted in polar
and nonpolar organic solvents, and in aqueous media. In
general, it was found that these catalysts consistently behaved
as stable homogeneous catalysts in organic solvents, and
decomposed to give less well-defined, but still highly active
heterogeneous catalyst species in water. In fact, reactions
performed in water significantly enhanced the activity of the
original Pd(II) catalysts, based on measured rate constants
and activation energies. In addition, both catalysts studied
here were able to activate aryl chloride feedstocks in all
solvents. Consideration of the trend in activation energies for
−
1
the aryl halides (C–Cl > C–Br > C–I; 407, 346, 280 kJ mol ,
respectively) reveals that aryl chlorides are significantly
the least reactive, yet they are industrially the most attractive
because heavy halogenated by-products can be avoided.
In this study, appreciable yields of cross-coupled products
Scheme 2 Syntheses of palladium catalysts 1 and 2.
‡ CCDC 976184 contains the CIF data for catalyst 2.
2
9
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 1456–1464 | 1457

View Article Online
Paper
Catalysis Science & Technology
a
Table 2 Suzuki–Miyaura biaryl coupling of aryl iodides by 1 and 2
Substrate
Catalyst
Entry
1a
Conversion (%)
Shetero (%)
1
81.5
4.6
100
77.3
23.3
100
99.7
91.2
97.1
99.6
97.3
98.9
1
1
b
c
2
2a
2
2
b
c
a
Reaction conditions: aryl iodide (0.216 mmol), phenylboronic acid
CO (0.647 mmol), 1 or 2 (1.0 mol% Pd), solvent
Cl , c-H O; 3.0 mL), 40 °C, 20 h; all reactions were
(0.259 mmol), K
a-toluene, b-CH
2
3
Fig. 1 Crystal structure of 2.
(
2
2
2
monitored by GC.
Table 1 Selected bond distances and angles for 2
Suzuki–Miyaura cross-coupling reactions were expanded
to the use of aryl bromides under otherwise identical reaction
conditions; the results of these reactions are summarized
in Table 3. 1 and 2 exhibited comparable percentage conver-
sions and selectivities of p-bromobenzaldehyde (Table 3;
entries 3–8) to those of analogous reaction conditions with
the iodo-precursor. Again, the highest conversion was obtained
Length (Å)
Angles (°)
Pd1–Cl1
Pd1–Cl2
Pd1–C1
Pd1–P1
C2–C3
2.295(2)
2.321(2)
2.041(7)
2.324(2)
1.356(12)
Cl1–Pd1–Cl2
C1–Pd1–Cl1
C1–Pd1–Cl2
C1–Pd1–P1
N1–C1–N2
172.82(10)
92.2(2)
90.4(2)
173.3(2)
107.3(7)
2
for H O. Interestingly, the catalytic activities appeared to be
displacement of 0.18(2) Å degrees with respect to the C1–N1–
C2–C3–N2 mean plane.
Tu et al. recently reported the synthesis of a related
NHC complex that was shown to be an active catalyst for
unaffected by the functional group position on the bromo-
benzaldehydes. The bromoanisoles revealed a substantial
dependence on the nature of the solvent as well as on the
location of the methoxy functional group. In aqueous solution,
the bromoanisoles yielded slightly lower conversions and
selectivities than those of the analogous bromobenzaldehydes
(Table 3; entries 9c, 10c, 13c, & 14c). Selectivity for the desired
hetero-coupled products was generally high (70–100%) for
aldehyde-substituted precursors, regardless of the substitu-
tion pattern, solvent or catalyst employed (Table 3; entries 3–8).
However, much greater variation in selectivity was observed
for methoxy-substituted precursors (0–99%). Most notably,
catalyst 1 did not yield any of the hetero-coupled biphenyl
3
2,33
Buchwald–Hartwig aminations
and Suzuki–Miyaura coupling
34
reactions. Merino et al. have developed a similar Ru-based
3
5
complex that is capable of ring closing metathesis. The
catalytic activities of 1 and 2 with respect to Suzuki–Miyaura
cross-coupling reactions were initially explored with aryl iodides
(
0.216 mmol), phenylboronic acid (0.259 mmol) and K
2
CO
3
3
(0.647 mmol) in toluene, CH Cl , or H O (3.0 cm ). These
2
2
2
solvents were chosen as examples of nonpolar and polar
aprotic solvents, and polar protic solvents, respectively. Prior
to injection of the catalyst, all reaction mixtures were stirred
at 40 °C until all solids had dissolved.
product from m-bromoanisole in H O, yet catalyst 2 was
2
highly selective under the same conditions (Table 3; entries
11c & 12c). It is difficult to identify clear trends in the
observed reactivity; the variation in selectivities is likely due
to inter-play of steric, electronic and solvation effects. In
general, both catalysts were significantly less active when
tested in both polar and apolar organic solvents. Less than
50% conversion to the hetero-coupled product was obtained
when the bromoanisoles were used (Table 3; entries 9–14a, b)
with the exception of 1 in CH Cl (Table 3; entry 13b). It is
2
. Suzuki coupling by catalysts 1 and 2 with aryl iodide,
bromide and chloride precursors
The catalyst loadings were determined using UV-vis spectro-
photometry (Fig. S1 & S2†), and 1 or 2 was rapidly injected
into the reaction mixture at t = 0. The reaction mixtures were
then stirred at 40 °C for 20 h. A summary of the outcomes
of these reactions are presented in Table 2. 1 and 2 both
converted p-iodobenzaldehyde into the desired hetero-coupled
product with greater than 91% selectivity (determined by GC
analysis; Table 2). The observed catalytic activities revealed a
significant dependence on the nature of the solvent: the cata-
lysts exhibited extremely low activities in CH Cl (Table 2;
2
2
of particular note that the catalysts were considerably more
selective in CH Cl than in toluene.
2
2
Interestingly, the catalysts were not only more active in
reactions employing p-functionalized aryl halides, but they
were also significantly more selective. p-Bromotoluene and
p-bromoethylbenzene were selected to assist in the examina-
tion of the apparent regioselectivity (Table 3; entries 15c & 16c).
In contrast to the bromobenzaldehydes and the bromoanisoles,
high conversion completeness of the substrate to the desired
2
2
entries 1b & 2b), but achieved 77–81% conversions in toluene
(Table 2; entries 1a & 2a) and complete conversion in H O
2
(Table 2; entries 1c & 2c).
1458 | Catal. Sci. Technol., 2014, 4, 1456–1464
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
a
Table 3 Suzuki–Miyaura reactions of aryl bromides by 1 and 2
and selectivities at lower catalyst loadings (0.1–1.0 mol%).
The catalysts appeared to be more active in aqueous media
than in either organic medium, for which the catalytic activi-
ties were only moderate. The activities in apolar toluene were
generally high when halogenated benzaldehydes were employed,
but the overall conversion values decreased dramatically for
coupling reactions performed using methoxy- and aliphatic-
substituted reagents, under otherwise identical conditions.
Interestingly, the reactivity pattern observed in more polar
CH Cl was generally reversed; the highest conversions
Substrate
Catalyst
Entry
3a
Conversion (%)
Shetero (%)
1
92.3
17.4
96.1
84.3
59.0
92.3
77.8
60.4
98.1
72.5
79.9
92.9
87.5
82.2
100
86.6
81.4
99.3
31.7
46.4
93.6
37.3
43.1
91.3
44.9
2.8
33.1
39.8
52.9
62.9
47.9
60.9
100
46.1
34.3
90.8
43.9
93.0
100
84.4
95.9
100
79.8
90.4
100
3
3
b
c
b
2
1
2
1
2
1
2
1
2
1
2
1
2
1
2
4a
4
4
b
c
b
2
2
5a
83.4
97.6
95.7
69.3
88.3
97.1
92.1
99.8
100
86.6
98.5
96.1
13.4
91.5
84.6
14.3
67.9
95.5
31.0
76.3
0
22.5
77.3
99.1
24.4
95.9
99.5
22.1
78.9
89.0
58.5
92.4
100
54.5
88.5
89.1
39.7
90.2
100
were measured for methyl- and methoxy-substituted reagents.
Hence, there is a clear solvent dependence upon reactivity
5
5
b
c
b
6a
(
Tables 2 & 3; entries a & b). However, both catalysts were
6
6
b
c
highly selective for the homo-coupled products. 1 and 2 acti-
vated C–I or C–Br bonds most readily for para-substituted
reagents, presumably due to a combination of preferential
steric and electronic effects of the incoming precursors. The
steric demands of the bulky aldehyde and methoxy- func-
tional groups may also protect the C–Br bond from engaging
in competing oxidative addition reactions.
Following the optimization of 1 and 2 for the Suzuki
coupling of aryl bromides, it was decided to explore the use
of aryl chlorides. Aryl chlorides are the most important indus-
trial target, since they are both cost effective and generate
significantly less harmful/corrosive by-products than the anal-
ogous aryl bromide and iodides. In our studies, the higher
bond dissociation energy of the C–Cl bond resulted in the
need for significantly longer reaction times than those for
C–Br to achieve comparable overall conversions at the same
reaction temperature (Table 4). Catalyst 1 activated the
b
7a
7
7
b
c
b
8a
8
8
b
c
b
9a
9
9
b
c
10a
1
1
0b
0c
11a
1
1
1b
1c
12a
1
1
2b
2c
13a
1
1
3b
3c
14a
1
1
4b
4c
a
Table 4 Suzuki–Miyaura coupling of aryl chlorides by 1 and 2
15a
1
1
5b
5c
16a
50.5
100
96.6
44.3
68.9
100
41.2
60.1
98.6
Substrate
Catalyst
Entry
19a
Conversion (%)
Shetero (%)
1
1
6b
6c
1
42.1
18.0
87.8
44.6
59.7
100
40.6
15.5
76.6
42.0
39.6
96.9
29.7
13.2
83.9
45.2
64.7
96.1
7.2
1
1
9b
9c
15.8
27.0
23.4
100
92.3
1.4
17a
1
1
7b
7c
2
1
2
1
2
20a
2
2
0b
0c
18a
25.5
81.6
76.7
1
1
8b
8c
21a
1b
2
21c
4.7 c
a
b
Reaction conditions: aryl bromide (0.270 mmol), phenylboronic acid
CO (0.811 mmol), 1 or 2 (1.0 mol% Pd), solvent
Cl , c-H O; 3.0 mL), 40 °C, 20 h; all reactions were
20.6
(
0.324 mmol), K
2
3
22a
22b
22c
23a
8.9
(
a-toluene, b-CH
2
2
2
28.4
78.8
11.4
12.1
89.1
26.0
79.8
86.5
b
monitored for conversion by GC. 0.1 mol% Pd.
2
2
3b
3c
24a
hetero-coupled product were evident for reactions performed
in both H O and CH Cl (Table 3; entries 15–18b, c). However,
2 2 2
2
2
4b
4c
these catalysts were found to be significantly less active in
toluene, since 1 and 2 not only exhibited less than 50% conver-
sion, but were also selective for the undesired homo-coupled
product (Table 3; entries 15–18a). Both 1 and 2 could activate
C–Br bonds in all three solvents with similar conversions
a
Reaction conditions: aryl chloride (0.357 mmol), phenylboronic acid
CO (1.07 mmol), 1 or 2 (1.0 mol% Pd), solvent
Cl , c-H O; 3.0 mL), 40 °C, 48 h; all reactions were
(
(
0.429 mmol), K
a-toluene, b-CH
2
3
2
2
2
b
c
monitored by GC. 2.0 mol% Pd. 79.3% selectivity for bifunctionalized
homo-coupled biaryl product.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 1456–1464 | 1459

View Article Online
Paper
Catalysis Science & Technology
carbon–chloride bonds with ranges of 13–42% and 76–88%
conversion for organic and aqueous solvents, respectively
(15 min, 8.5 krpm) to isolate any solid residues, which were
then separated from the organic supernatant. After cooling,
the solid and liquid phases were then analysed by ICP-MS to
determine the %Pd content of each (Table 5). For both 1 and
2, the Pd elemental analysis indicated that only a minority of
(Table 4; entries 19, 21 & 23). In contrast, 2 resulted in con-
version ranges of 40–65% and 96–100%, respectively (Table 4;
entries 20, 22 & 24).
As was observed for the use of brominated reagents (vide
supra), a broad range of selectivity values were measured
using ortho-, meta- and para-substituted chlorobenzaldehydes.
However, catalyst 2 was consistently more selective than 1 in
all solvents. In fact, catalyst 1 only gave appreciable yields of
2 2
metal had precipitated from reactions performed in CH Cl ,
suggesting that the molecular catalysts retained their origi-
nal structures when employed in CH Cl (Table 5; entries
2
2
25b & 26b), however, both catalysts appeared to undergo
significant precipitation when operated in H O or toluene,
2
the hetero-coupled product when operated in H O. In contrast,
which is indicative of either oligomerization or decomposi-
tion due to detachment of Pd from the carbene ligands.
Subsequently, the isolated solid and liquid phases were
employed in repeat catalytic runs under the original reaction
conditions, but using newly added organic reagents. This was
achieved by direct re-use of the supernatant, or by the addi-
tion of fresh solvent to dissolve/suspend the residual solids.
The products of the repeat reactions were monitored by GC
(Table 5). Interestingly, conversion was observed for both the
recycled supernatant and solids in all cases. The solid residues
2
2
was moderately, or highly selective in CH
ingly, the rarely-observed bifunctionalized homo-coupled product
4,4′-biphenyldicarboxaldehyde) was obtained from a 2.0 mol%
loading of 1 with 79.3% selectivity from m-chlorobenzaldehyde
Table 4; entry 21c). In the para-position, 1 was 89.1% selec-
tive for the desired product (Table 4; entry 23c). Conversely,
always created the hetero-coupled product as the major
2 2 2
Cl or H O. Interest-
(
(
2
product in greater than 96% conversion with selectivities
of 92, 79, and 87% for the ortho (Table 4; entry 20c), meta-
(Table 4; entry 22c), and para- (Table 4; entry 24c) positions,
2
obtained from initial reaction in H O or toluene each showed
respectively. From this data, there is no obvious trend that
relates total conversion (reaction rate) to the corresponding
degree of product selectivity. This infers that catalyst structure
and/or solvation effects play important roles in the reaction.
high catalytic activity upon re-use (Table 5; entries 25a, c &
26a, c), presumably due to the poorly-defined, yet catalytically-
active heterogeneous (colloidal or nanoparticulate) species that
had been generated. This type of complex catalysis has been
36–40
documented in several previous instances;
a recent review
by Ananikov specifically describes the inter-play of well-
defined homogeneous and less defined heterogeneous cata-
lysts in C–C coupling reactions. Both the supernatant and
3
. A quantitative assessment of the effect of solvent system
upon Suzuki coupling by catalysts 1 and 2
41
The above studies proved that catalysts 1 and 2 were able to
convert a broad range of aryl iodides, bromides and chlorides
into the desired functionalized biphenyl products with varying
degrees of proficiency in the three contrasting solvent sys-
tems. As might be expected, the trends in activity appeared
to depend in part on the electronic influence of R group on
the aryl halide precursors. For example, whereas OMe sub-
stituents greatly deactivated the C–X bond activation, aldehyde
functionalized aryl halides were much more easily converted
to the corresponding biphenyl aldehydes in high yields by 1
and 2, especially for para-substituted precursors. However,
the polarity of the solvent was clearly most highly influential
in terms of both reaction selectivity and reactivity. In general,
selectivity toward hetero-coupled biaryl products in the Suzuki
reaction was significantly enhanced for both 1 and 2 in sol-
the solid residues obtained from the initial reactions in
toluene converted p-bromobenzaldehyde to p-biphenylaldehyde
with high selectivity upon re-use (Table 5; entries 25a & 26a).
This indicates that the molecular species was only partially
decomposed to give heterogeneous species along with molec-
ular species. Repeat reactions in CH Cl only showed appre-
2 2
ciable conversion for the supernatant, suggesting that 1 and
2 remained mostly solubilized (Table 5; entries 25b & 26b).
2
The reverse outcome was observed for reactions in H O,
Table 5 Recyclability of 1 and 2 in Suzuki–Miyaura biaryl coupling of
a
the p-bromobenzaldehyde supernatant (sup) and precipitate (ppt)
%
Pd
Conversion %
b
b
Catalyst
Entry
sup
ppt
sup
ppt
vents with larger dielectric constants (H
2
O = 80.4; CH
2
Cl
2
=
1
25a
25b
15
97
23
52
73
26
—
85
3
77
48
27
74
—
87.8
50.4
55.4
81.0
54.5
0.2
92.5
12.5
98.6
78.3
4.1
9
.1; toluene = 2.4). It became evident that 1 and 2 were more
2
5c
26a
6b
selective for the hetero-coupled product in solvents with
higher polarity; therefore, further studies were made to gain
a better understanding of the possible mechanistic pathways
responsible for catalytic enhancement effects in certain solvent
systems. In particular, a more thorough characterization of the
active catalysts and post-catalytic species was carried out.
The stabilities of 1 and 2 were first assessed in toluene,
2
2
26c
100
51.5
c
Pd–PVP
27
0.06
a
Reaction conditions: p-bromobenzaldehyde (0.270 mmol),
phenylboronic acid (0.324 mmol), K CO (0.811 mmol), Pd (1 mol%),
solvent (a-toluene, b-CH Cl , c-H O; 3.0 mL), 40 °C, 20 h; all reactions
were monitored by GC. % Pd of supernatant and precipitate deter-
mined by ICP-MS. p-Bromobenzaldehyde (0.270 mmol), phenylboronic
acid (0.324 mmol), K
(0.537 mmol), 40 °C, 20 h.
2
3
2 2 2
b
CH Cl
2 2
2
and H O in Suzuki reactions with p-bromobenzaldehyde
c
and phenylboronic acid. Immediately after cessation of the
Suzuki reaction, the hot reaction mixtures were centrifuged
2 3 2
CO (0.811 mmol), H O (3.0 mL), Pd–PVP
1460 | Catal. Sci. Technol., 2014, 4, 1456–1464
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Table 6 Rate constants for 1 and 2 for p-bromobenzaldehyde in H
2
Oa
suggesting that the catalysts were almost completely decomposed
(
Table 5; entries 25c & 26c). For example, the relative palladium
Catalyst
Entry
0.10 mol%
0.25 mol%
0.50 mol%
content detected by ICP-MS after catalysis by 1 in CH Cl
2
2
1
2
28
29
0.3046
0.8979
0.4407
0.9601
0.5462
0.9667
was 97% in the supernatant and only 3% in the precipitate.
This result is in stark contrast to 23% versus 77% (super-
natant versus precipitate) that was observed after catalysis had
a
Reaction conditions: p-bromobenzaldehyde (3.24 mmol), phenylboronic
CO (9.73 mmol), H O (15.0 mL), 40 °C; all reactions
were monitored by GC.
acid (3.89), K
2
3
2
been conducted in H O (Table 5; entry 25c).
2
The fact that similar overall conversions and coupling
selectivities were observed in the Suzuki reaction using 1 or 2
in three highly contrasting solvents is an example of how
particular attention must be paid to identify the composition
than 0.1 mol% usually resulted in incomplete conversion
(Fig. S4†). As expected for truly catalytic systems, higher
catalyst loadings required proportionately less time to reach
complete conversion. Above 0.25 mol%, the effect of higher
loading was minimal since all reactions reached completion
in less than 3 h for both 1 and 2 (Table 6; entries 28 & 29).
The data summarized in Table 6 imply a pseudo-first order
reaction with respect to the p-bromobenzaldehyde concen-
tration. Also, catalyst 2 was significantly more effective than
2
8,42–46
of the ‘real’ active species.
In the present study, it could
easily be incorrectly assumed from the raw conversion data
that 1 and 2 were proficient in all solvent systems. However, it
is very apparent from the above hot isolation tests that the
predominant catalytic species are different when operated
in H O or CH Cl . Meanwhile, an ill-defined mixture of homo-
geneous and heterogeneous species appeared to co-exist in
toluene. In an attempt to further probe the identity of the
catalytically-active heterogeneous species, pre-formed 4.5 nm
Pd nanoparticles (Fig. S3†) capped with poly(vinylpyrrolidone)
as a stabilizer were employed in the Suzuki reaction under
otherwise identical reaction conditions to those used in the
previous reactions. The Pd–PVP nanoparticles were employed
2
2
2
2
1 in H O for all catalyst loadings studied.
Examination of the temperature dependences for 1 and 2
at 24, 30, 35, and 40 °C were also made by means of kinetic
2 2 2
experiments in toluene, CH Cl , and H O. The results are
summarized in Fig. 3 and the corresponding rate constants
are presented in Table 7. This study provided some further
interesting insights into the comparative identities and reac-
tivities of the dominant active Pd species in each solvent
system. Reactions performed at low temperature were most
sensitive to the nature of the solvent. Specifically, both 1
and 2 affected the coupling of p-bromobenzaldehyde with
phenylboronic acid in H
tivity (Fig. 3; left column). In contrast, reactions in toluene
and CH Cl at the same temperature gave low yields for 1,
while 2 was found to be inactive. The reaction temperature
was then increased in 5 °C steps to 40 °C, which resulted in
increased activity for all three solvent systems, albeit with a
modest reduction in coupling selectivity (Fig. 3; centre and
right columns).
The fact that no turnover was observed at 30 °C in the
organic solvents, while appreciable reactivity was observed
above 35 °C, suggested that a kinetic activation barrier had
been overcome between 30 and 35 °C. This type of temperature-
dependent behaviour is commonplace for molecular species.
in H O-based Suzuki reactions as an alternative source of
2
pre-formed heterogeneous Pd catalysts. These results showed
coupling catalysis in the presence of the Pd–PVP nanoparticles
without the requirement for either 1 or 2; the associated
reactivity and product selectivity was also very similar to
that observed when the molecular catalyst 2 was employed
exclusively (Table 5; entry 27). This is convincing additional
evidence for the hypothesis that heterogeneous Pd(0) species
2
O at 30 °C, with high coupling selec-
2
2
28
in H
2
O are indeed catalytically active for the Suzuki reaction.
The activation energies for 1 and 2 in H O, CH Cl and
2
2
2
toluene were also measured in order to provide kinetic insight
into the nature of the different active species in each solvent.
This was achieved by measuring the rate of coupling for
p-bromobenzaldehyde as a function of catalyst loading (Fig. 2
&
Table 6). Complete conversion of p-bromobenzaldehyde with
phenylboronic acid to yield the desired hetero-coupled
product was achieved in H O for all catalyst loadings in the
2
range 0.1–0.5 mol% in less than 5 h. Catalyst loading lower
Meanwhile, since appreciable turnover was observed in H O
2
at temperatures as low as 24 °C (Fig. S4†), it is clear that a
different, more easily activated catalytic species was respon-
sible for the observed reactivity.
It is also noteworthy that distinct induction periods of ca.
1
–5 h were observed for reactions involving 1 and 2 in H
lower temperatures, resulting in s-shaped reactivity profiles
Fig. 3; black data). This is presumably because at lower tem-
2
O at
(
peratures, the generation of the active heterogeneous Pd species
occurred more slowly. Meanwhile, reactions performed at 35
or 40 °C with 1 and 40 °C with 2 did not show any apprecia-
ble induction period. These observations are also consistent
with the supposition that molecular catalysis dominated
in organic solvents while newly-generated heterogeneous
Fig. 2 Conversion of p-bromobenzaldehyde with phenylboronic acid
as a function of time in Suzuki–Miyaura coupling by 1 and 2 at 40 °C
with 0.10, 0.25, and 0.50 mol% catalyst loading in H O.
2
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 1456–1464 | 1461

View Article Online
Paper
Catalysis Science & Technology
Fig. 3 Reaction of p-bromobenzaldehyde with phenylboronic acid as a function of time in Suzuki–Miyaura coupling by 0.5 mol% of 1 (top row)
and 2 (bottom row) at 30, 35, and 40 °C in toluene (red), CH
2
Cl
2
2
(blue), and H O (black).
a
Table 7 Rate constants of 1 and 2 for p-bromobenzaldehyde
Catalyst
Entry
30a
24 °C
30 °C
35 °C
40 °C
45 °C
1
—
—
0.0610
—
—
0.2393
0.3492
0.2144
—
0.0037
0.2183
0.4091
1.3739
0.2656
0.3678
0.5999
0.2614
1.8077
2.8619
0.5206
1.3877
1.7767
1.5986
—
—
—
2.9481
4.8967
—
3
3
0b
0c
2
31a
3
3
1b
1c
0.1442
a
Reaction conditions: p-bromobenzaldehyde (3.24 mmol), phenylboronic
CO (9.73 mmol), solvent (a-toluene, b-CH Cl , c-H O;
5.0 mL); all reactions were monitored by GC.
Fig. 4 Arrhenius plots for the activation of p-bromobenzaldehyde by
and 2 in toluene (red), CH Cl (blue), and H O (black).
acid (3.89), K
1
2
3
2
2
2
1
2
2
2
a
Table 8 Activation energies for 1 and 2 for p-bromobenzaldehyde
catalysts were dominant in H O: single-site homogeneous
2
species should be immediately highly active above a mini-
mum temperature threshold, but could also become rapidly
poisoned. In contrast, decomposition of the molecular species
−1
Catalyst
Entry
32a
Activation energy (kJ mol )
1
159.2
166.3
111.3
169.8
171.2
115.9
32b
in H O to generate the active heterogeneous Pd species
2
32c
33a
appeared to require a significant amount of time as reflected
in the induction phase.
2
3
3
3b
3c
Activation energy data for the Suzuki coupling reaction
was obtained in the three solvent systems with the slope-ratio
method. Arrhenius plots created using the temperature-
dependent activity data for 1 and 2 (Fig. 4 & Table 8) show
that the Suzuki reaction rates were faster in organic solution
a
Calculated from the slope of the Arrhenius plots in Fig. 4.
Comparative activation energies for the Suzuki coupling of
than in H O. However, the corresponding activation ener-
p-chlorobenzaldehyde in toluene, CH Cl , and H O solutions
2
2
2
2
−
1
gies in organic solvents (159–171 kJ mol ) were significantly
larger than the comparable activation energies in H O (111–
16 kJ mol ). This substantial difference in energy could be
could not be obtained because the reactivities were minimal
at lower temperatures. The C–Cl bond was most effectively
activated by the supposed heterogeneous Pd(0) species in
2
−
1
1
attributed to the different physical nature of the active cata-
lysts, in which the homogenous molecular catalysts 1 and 2
experienced a larger energy barrier for the oxidative addition
of the aryl halide precursors.
H O. Kinetic trials were performed in aqueous solution using
2
p-chlorobenzaldehyde, in which both the catalyst mol% and
the temperature were varied (Fig. 5). While the reactions of
both 1 and 2 resulted in complete conversion of the chlorinated
1462 | Catal. Sci. Technol., 2014, 4, 1456–1464
This journal is © The Royal Society of Chemistry 2014

View Article Online
Catalysis Science & Technology
Paper
Fig. 5 Conversion of p-chlorobenzaldehyde with phenylboronic acid
as a function of time in Suzuki–Miyaura coupling by catalysts 1 (left)
2
and 2 (right) at 40 °C in H O.
Fig. 6 Conversion (black) and selectivity (red) of p-chlorobenzaldehyde
with phenylboronic acid as a function of time in Suzuki–Miyaura
2
coupling at 40 °C with 1.5 mol% of 1 in H O. Open and closed circles
precursors in H O, 2 was the most proficient. Accordingly,
denote selectivities of the hetero- and homo-coupled products,
2
the rate constants for 2 increased drastically as the catalyst
loadings increased (Table 9).
respectively.
In contrast to p-bromobenzaldehyde, 1 and 2 were found
to be initially selective for the homo-coupled product, but as
the reaction proceeded, both catalysts became more selective
for production of the desired hetero-coupled product. After
approximately 30 h, the reaction selectivity had reached a
steady-state, favouring approximately 80% of the hetero-coupled
product. The origin of this selectivity evolution over such a
significant period of time is most likely due to the evolving
nature of the active catalyst species, as molecular 1 or 2 are
converted to eventually stable heterogeneous catalysts. It is
somewhat more difficult to determine the identity of tran-
sient intermediate Pd catalysts (such as small Pd clusters)
that may exist in solution prior to steady-state selectivity
being reached. A so-called ‘cocktail-type’ mixture of active
Pd-based species could dominate during the first 25–30 h
Fig.
phenylboronic acid as a function of time using 0.25 mol% of 2 at 24,
0, 35, and 40 °C in H O.
7 Suzuki–Miyaura coupling of p-chlorobenzaldehyde with
3
2
4
1,47
of reaction (Fig. 6).
A representative example of a typical
reaction progression is depicted in Fig. 6 (see ESI† for all
other selectivity curves).
Table 10 Rate constants for 2 using p-chlorobenzaldehyde in H
2
Oa
Based on the rate constant obtained from 0.25 mol% of
catalyst 2, it was decided to double and quadruple the cata-
lyst loading to 0.5 and 1.0 mol% of 1 and 2, respectively
Catalyst
Entry
36
24 °C
30 °C
35 °C
40 °C
2
0.0031
0.1476
1.0439
2.2459
a
(Fig. S5 & S6†). As expected, the rate of 2 was significantly
Reaction conditions: p-chlorobenzaldehyde (4.28 mmol), phenylboronic
CO (12.86 mmol), H O (15.0 mL), 40 °C; all reactions
were monitored by GC.
acid (5.14), K
2
3
2
faster and more selective in the activation of aryl chlorides,
which was consistent with the data gathered for the corre-
sponding aryl bromides. Although these observations follow
classical kinetic control, it was somewhat surprising that
higher catalyst loadings resulted in lower rate constants
potassium carbonate, (Fig. S7†). These data suggested that
the phenylboronic acid and the potassium carbonate did not
affect the rate determining step of this mechanism. The
temperature dependence of this system was also examined at
(Table 9). In an attempt to understand this trend, the reac-
tion mechanism was further investigated by varying the num-
ber of stoichiometric equivalents of phenylboronic acid and
24, 30, 35, and 40 °C using 0.25 mol% of catalyst 2 (Fig. 7 &
Table 10). The hetero-coupled aryl product was only produced
in complete conversion at 35 and 40 °C and the corresponding
rate constants were found to increase with increasing tempera-
ture (Table 10).
Table 9 Rate constants of 1 and 2 of p-chlorobenzaldehyde in H
2
Oa
b
b
b
b
b
b
Catalyst Entry 0.05
0.10
0.25
0.50
1.0
1.5
The activation energy for p-chlorobenzaldehyde in H O
was determined on the basis of the slope of the Arrhenius
1
2
34
35
—
—
—
0.0141 0.0193 0.0206
2
0.0598 0.0919 2.2459 0.0270 0.0547
—
a
plot (Fig. S14†). The activation energy was determined to be
Reaction conditions: p-chlorobenzaldehyde (4.28 mmol), phenylboronic
CO (12.86 mmol), H O (15.0 mL); all reactions were
−
1
321.8 kJ mol , which is in accord with the current literature
acid (5.14), K
monitored by GC. mol%.
2
3
2
b
values for similar systems.
This journal is © The Royal Society of Chemistry 2014
Catal. Sci. Technol., 2014, 4, 1456–1464 | 1463

View Article Online
Paper
Catalysis Science & Technology
1
7 S. S. Shendage, U. B. Patil and J. M. Nagarkar, Tetrahedron
Lett., 2013, 54, 3457.
Conclusions
In conclusion, two new palladium(II) carbene complexes were
synthesized and studied for their catalytic activities in the
Suzuki reaction. Their comparative behaviour in toluene,
18 B. Karimi and P. Fadavi Akhavan, Chem. Commun., 2011, 47, 7686.
19 B. Karimi and P. F. Akhavan, Inorg. Chem., 2011, 50, 6063.
20 J. Langecker and M. Rehahn, Macromol. Chem. Phys., 2008,
209, 258.
2 2 2
CH Cl and H O were assessed. Both 1 and 2 were able to
activate C–I, C–Br, and C–Cl bonds in polar and nonpolar
21 S. Wei, Z. Ma, P. Wang, Z. Dong and J. Ma, J. Mol. Catal. A:
Chem., 2013, 370, 175.
aprotic solvents, as well as in H O. However, it was shown
2
that catalyst decomposition was prevalent in toluene and
H O, yielding new, presumably heterogeneous, yet equally
2
22 J. Flapper and J. N. H. Reek, Angew. Chem., Int. Ed., 2007,
46, 8590.
active catalyst species. In CH Cl , the molecular catalysts
23 A. Khalafi-Nezhad and F. Panahi, J. Organomet. Chem.,
2012, 717, 141.
24 C. Crisóstomo-Lucas, R. A. Toscano and D. Morales-Morales,
Tetrahedron Lett., 2013, 54, 3116.
25 D. R. Anton and R. H. Crabtree, Organometallics, 1983, 2, 855.
26 P. Foley, J. Am. Chem. Soc., 1980, 102, 6713.
27 K. Yu, W. Sommer, J. Richardson, M. Weck and C. Jones,
Adv. Synth. Catal., 2005, 347, 161.
28 G. Kyriakou, S. K. Beaumont, S. M. Humphrey, C. Antonetti
and R. M. Lambert, ChemCatChem, 2010, 2, 1444.
29 T. D. Sheppard, Org. Biomol. Chem., 2009, 7, 1043.
30 R. R. Butorac, S. S. Al-Deyab and A. H. Cowley, Molecules,
2011, 16, 2285.
2
2
were largely stable and recyclable. In toluene solution, both
forms of the catalyst were found to co-exist. Kinetic studies
revealed that the average activation energy for dissociation of
−
1
the C–Br bond in organic solvents was 166 kJ mol com-
−
1
pared to only 113 kJ mol for H O. The latter result infers
2
that the heterogeneous species generated from the original
Pd(II) carbene complexes were more reactive for the Suzuki
coupling reaction, and that catalyst decomposition was not
actually detrimental in the Suzuki–Miyaura cross-coupling
reaction. The results of this study exemplify how well-defined
molecular and less well-defined heterogeneous Pd species can
provide misleadingly similar results in solution-phase catalysis.
3
1 S. A. Svejda, M. Brookhart and C. Hill, Organometallics,
999, 18, 65.
Acknowledgements
1
3
3
2 T. Tu, W. Fang and J. Jiang, Chem. Commun., 2011, 47, 12358.
3 W. Fang, J. Jiang, Y. Xu, J. Zhou and T. Tu, Tetrahedron,
The authors thank Ms. Stephany Garcìa for supplying the
Pd–PVP catalysts and Dr. Vincent M. Lynch for assisting in
X-ray crystallography, and the Robert A. Welch Foundation
2
013, 69, 673.
4 T. Tu, Z. Sun, W. Fang, M. Xu and Y. Zhou, Org. Lett.,
012, 14, 4250.
5 E. Merino, E. Poli, U. Díaz and D. Brunel, Dalton Trans.,
012, 41, 10913.
3
3
(F-1738, F-003) for funding.
2
Notes and references
2
1
2
3
4
5
N. Miyaura and A. Suzuki, Chem. Rev., 1995, 95, 2457.
J. Magano and J. R. Dunetz, Chem. Rev., 2011, 111, 2177.
C. Torborg and M. Beller, Adv. Synth. Catal., 2009, 351, 3027.
J. P. Corbet and G. Mignani, Chem. Rev., 2006, 106, 2651.
K. C. Nicolaou, P. G. Bulger and D. Sarlah, Angew. Chem.,
Int. Ed., 2005, 44, 4442.
M. Chen, D. A. Vicic, M. L. Turner and O. Navarro,
Organometallics, 2011, 30, 5052.
N. Marion, O. Navarro, J. Mei, E. D. Stevens, N. M. Scott and
S. P. Nolan, J. Am. Chem. Soc., 2006, 128, 4101.
36 H. M. De Vries, F. J. Parlevliet, L. S. De, J. H. M. Mommers,
H. J. W. Henderickx, A. M. Monique and G. De Vries,
Adv. Synth. Catal., 2002, 344, 996.
37 A. H. M. de Vries, J. M. C. A. Mulders, J. H. M. Mommers,
H. J. W. Henderickx and J. G. de Vries, Org. Lett., 2003, 5, 3285.
38 A. F. Schmidt and V. V. Smirnov, J. Mol. Catal. A: Chem.,
2003, 203, 75.
39 A. F. Schmidt and V. V. Smirnov, Top. Catal., 2005, 32, 71.
40 N. T. S. Phan, M. Van Der Sluys and C. W. Jones, Adv. Synth.
Catal., 2006, 348, 609.
41 A. S. Kashin and V. P. Ananikov, J. Org. Chem., 2013, 78, 11117.
42 J. A. Widegren and R. G. Finke, J. Mol. Catal. A: Chem.,
2003, 198, 317.
6
7
8
9
M. G. Organ, S. Avola, I. Dubovyk, N. Hadei, E. A. B. Kantchev,
C. J. O'Brien and C. Valente, Chem.–Eur. J., 2006, 12, 4749.
X. X. Zhou and L. X. Shao, Synthesis, 2011, 3138.
1
1
0 R. Martin and S. L. Buchwald, Acc. Chem. Res., 2008, 41, 1461.
1 T. E. Barder, S. D. Walker, J. R. Martinelli and S. L. Buchwald,
J. Am. Chem. Soc., 2005, 127, 4685.
2 M. Miura, Angew. Chem., Int. Ed., 2004, 43, 2201.
3 G. C. Fortman and S. P. Nolan, Chem. Soc. Rev., 2011, 40, 5151.
4 H. Li, C. C. C. Johansson Seechurn and T. J. Colacot,
ACS Catal., 2012, 2, 1147.
5 D. Ganapathy and G. Sekar, Catal. Commun., 2013, 39, 50.
6 M. Samarasimhareddy, G. Prabhu, T. Vishwanatha and
V. Sureshbabu, Synthesis, 2013, 45, 1201.
43 S. S. Pröckl, W. Kleist, M. A. Gruber and K. Köhler, Angew.
Chem., Int. Ed., 2004, 43, 1881.
44 A. F. Schmidt, A. A. Kurokhtina and E. V. Larina, Kinet. Catal.,
2012, 53, 84.
45 K. Köhler, W. Kleist and S. S. Pröckl, Inorg. Chem., 2007,
46, 1876.
46 S. M. Nobre, S. I. Wolke, R. G. da Rosa and A. L. Monteiro,
Tetrahedron Lett., 2004, 45, 6527.
1
1
1
1
1
47 V. P. Ananikov and I. P. Beletskaya, Organometallics, 2012,
31, 1595.
1464 | Catal. Sci. Technol., 2014, 4, 1456–1464
This journal is © The Royal Society of Chemistry 2014