Suzuki cross-couplings of (hetero)aryl chlorides in the solid-state

Article abstract of DOI:10.1039/c2nj40064b

The ultrasound-assisted cross-linking of chitosan with hexamethylene diisocyanate with the simultaneous incorporation of Pd(OAc)₂ resulted in a catalyst which is suitable for the solid-state Suzuki cross-coupling of poorly reactive (hetero)aryl chlorides with phenylboronic acid. Reactions were carried out solvent-free in a planetary ball mill allowing the catalyst to be recycled several times.

Full text of DOI:10.1039/c2nj40064b

View Online / Journal Homepage
Dynamic Article Links
NJC
Cite this: DOI: 10.1039/c2nj40064b
www.rsc.org/njc
LETTER
Suzuki cross-couplings of (hetero)aryl chlorides in the solid-state
Giancarlo Cravotto,*a Davide Garella,a Silvia Tagliapietra,a Achim Stolle,b
Stefan Schußler,b Silke E. S. Leonhardtb and Bernd Ondruschkab
Received (in Montpellier, France) 29th January 2012, Accepted 15th March 2012
DOI: 10.1039/c2nj40064b
The ultrasound-assisted cross-linking of chitosan with hexamethylene
diisocyanate with the simultaneous incorporation of Pd(OAc)2
resulted in a catalyst which is suitable for the solid-state Suzuki
cross-coupling of poorly reactive (hetero)aryl chlorides with
phenylboronic acid. Reactions were carried out solvent-free in a
planetary ball mill allowing the catalyst to be recycled several times.
chitosan/Pd(II) catalyst that has already been tested as a
heterogeneous catalyst in microwave-assisted reactions.17
Although chitosan (CS) modification requires some tricky
procedures,18 this material is cheap, stable and biodegradable,
making it suitable for Pd-catalyzed Suzuki reactions.19–21
Despite the fact that several metal-catalyzed reaction protocols
for ball milling synthesis have been developed, only in one case
the possibility of catalyst recycling was included.22 These authors
described the synthesis of oximes with Bi2O3 as a catalyst by
grounding in a mortar the reacting mixture with a pestle.
Introduction
Reactions in ball mills have attracted serious attention in
organic synthesis circles during the last decade.1,2 This method
allows reactions to be carried out in the solid state and possibly
reduces the solvent amount3–6 which makes it attractive for
other application fields as well.7 Pd-catalyzed cross-couplings
via the Suzuki-,8–10 Heck-,11 or Sonogashira-type reactions,12,13
as well as Cu-catalyzed click reactions (CuAAC),14 and the
homo-coupling of terminal alkynes,15 make use of metal salts or
metal complexes which are difficult to separate from the solid
reaction mixtures afterwards.
Experimental
Materials
Commercially available reagents and solvents were used without
further purification. Medium molecular weight CS was purchased
from Sigma Aldrich.
Catalyst preparation
Additionally, the application of aryl chlorides instead of
bromides or iodides as cross-coupling partners has only so far
afforded zero or minor yields. An example of this is a Suzuki
reaction in the presence of KF–Al2O3 using a planetary ball
mill.10 This inorganic solid material was used for the in situ
generation of the base needed for the reaction and Pd(OAc)2
was used as the catalyst.10,16 The synthetic procedure was very
efficient with aryl iodides and bromides but failed in the case of
the poorly reactive aryl chlorides.10 Similar observations were
reported when a more reactive reagent system Pd(PPh3)4 and
K2CO3 (catalyst and base, respectively) was used.8 All the
results reported in the literature gave no reaction or very low
yields with these substrates.8–10,16
CS (5.0 g) was dissolved in 0.1 N HCl (350 mL) in a 1 L round-
bottom cylindrical flask at 60 1C under sonication with a
titanium horn (19.8 kHz, 70 W) for less than 1 h. A solution of
Pd(OAc)2 (2.23 mmol, 500 mg) in 0.1 N HCl (50 mL) was
added after 5 min at 50 1C while the sonication was run at
reduced power (50 W). Hexamethylene diisocyanate (HMDI;
37.0 mmol, 6 mL) was added dropwise at 50 1C and was
sonicated for 10 min at 60 W at which point complete gelation
occurred. The gel was heated for 2 h at 50 1C and a yellowish
solid material formed which was then transferred to a mortar
and triturated with a 0.5 N NaOH solution until pH 10 was
reached. The solid residue was filtered, washed with water,
acetone and diethylether (each solvent 150 mL), and dried
overnight under vacuum. This procedure afforded 8.23 g of the
Pd(II)-loaded urethane/urea-bridged CS derivative 8 catalyst
HMDI-CS/Pd.
With the aim of finding conditions which may widen the
application of this method, we realized that the physicochemical
properties of the catalyst could play a fundamental role in the solid
state. Our candidate for Pd-catalyzed reactions is a cross-linked
Catalyst characterization
a Dipartimento di Scienza e Tecnologia del Farmaco,
University of Torino, Via P. Giuria 9, 10125 Torino, Italy.
E-mail: giancarlo.cravotto@unito.it; Fax: +39 011 6707687;
Tel: +39 011 6707684
FT-IR spectra were recorded on a Perkin-Elmer model 2000
spectrometer in the 650–4000 cmꢀ1 range at a resolution of
2 cmꢀ1 using the ATR-method. TG-curves were recorded on
the TA Instrument TGA 2050 (Shimadzu). The analyses
were carried out with a constant heating rate of 10 K minꢀ1
b Institute for Technical Chemistry and Environmental Chemistry
(ITUC), Friedrich-Schiller University Jena, Lessingstr. 12,
07743 Jena, Germany
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem.

View Online
from 50 to 700 1C under atmospheric conditions. The metal
content of the catalysts was determined by ICP-MS on a
Quadrupole-ICP-MS X Series II (Thermo Fisher Scientific)
after the solid catalyst samples were digested in HNO3 and
aqua regia. The oxidation state of the adsorbed palladium was
determined by measuring the binding energy (Eb) of the Pd
3d-core level by X-ray photoelectron spectroscopy (XPS) with
a Quantum 2000 (PHI Co., Chanhassen, MN, USA) using a
focused monochromatic Al Ka-source (1486.7 eV). Spectra
were normalized to the C 1s peak.
Scheme 1 Suzuki cross-coupling reaction of aryl halides in a plane-
tary ball mill (cf. Table 1).
Catalytic testing
Ball milling was conducted using a planetary ball mill (PM100;
Retsch GmbH) equipped with a single milling beaker (stainless
steel, 125 mL). For the experiments 15 ꢁ 10 mm milling balls
of the same material were applied. The prepared catalyst (1 g),
aryl halide (0.25 mmol), phenyl-boronic acid (0.27 mmol, 33 mg),
and K2CO3 (0.5 mmol, 69 mg) were added to the milling balls
one after the other. The reaction mixture was subsequently milled
at 600 minꢀ1 for either 90 or 120 min. Every 30 min the direction
of rotation was inverted. The reaction products were dissolved in
EtOAc (3 ꢁ 15 mL). The supported catalyst, insoluble in the
organic solvent, was separated by filtration and was washed with
acetone (3 ꢁ 15 mL), dried under vacuum and reused in a new
reaction. The products were analyzed by GC-MS. The yields
determined by GC-MS were comparable with the yields of
the isolated products. Purification was performed via flash-
chromatography on a silica column (CombyFlash Rfs Teledyne
ISCO) using hexane–ethyl acetate mixtures as eluents. Conversions
and yields were initially determined by GC-MS and confirmed by
chromatographic separation (+2%).
Fig. 1 FTIR-spectra of pure CS and cross-linked catalyst
HMDI-CS/Pd.
3320 cmꢀ1 which is caused by OH-, NH2- as well as CONH-
vibrations (amide = non-deacetylated glucosamine groups).19
After cross-linking with HMDI and in situ deposition of
Pd, the band appears less broad. This indicates that the
diisocyanate reacted with the hydroxyl and amino groups of
the polymer. The resulting urea- and urethane-linkages also
contain the CONH-function, which are responsible for this
band. It should be mentioned that isocyanate-groups do not
necessarily react with the polymer, they can also be hydrolyzed
by the solvent which would result in new amine-functions.
Another indicator for successful cross-linking is the band at
2930 cmꢀ1 which refers to CH2-vibrations. After polymer
cross-linking, there is a higher number of CH2-groups in the
polymer, resulting in a stronger and sharper absorption band
in the IR-spectrum. New bands appear at approximately
1600 and 1260 cmꢀ1 which are assigned to the urea- and
urethane-functions (CQO– and C–N-vibrations) in the
polymer. The absorption at 1026 cmꢀ1 indicates C–O–C
vibrations which are typical for carbohydrates. This band
slightly decreases in intensity. This could be caused by the
chain-cleavage that may occur in the acidic solvent that was
used during the cross-linking under power ultrasound.
Product analysis
GC-MS measurements were performed on an Agilent Technologies
6850 Network GC System with a 5973 Network Mass Selective
Detector using a capillary column (length 30 m, i.d. 0.25 mm, film
thickness 0.25 mm). GC conditions were as follows: injection split
1 : 20, injector temperature 250 1C, detector temperature 280 1C.
The temperature program ran from 50 1C (3 min) to 80 1C
at 3 1C minꢀ1 and 80 1C to 300 1C at 10 1C minꢀ1 (30 min).
Helium was used at 25 kPa as the carrier gas.
Results and discussion
The ultrasound-assisted synthesis of the CS catalyst HMDI-
CS/Pd resulted in a solid material which was characterized
using FT-IR, XPS-spectroscopy, and thermal analysis (TGA,
DTA). The metal content of the catalysts, prior to and after
their application in Suzuki cross-couplings (Scheme 1), was
determined with ICP-MS analysis. The addition of HMDI
to the CS solution containing Pd(OAc)2 resulted in the cross-
linking of the polymer chains, a regular distribution of the
metal and a considerable catalytic activity.17
Further evidence of successful cross-linking with HMDI can
be found in the TGA-analysis (Fig. 2). At the beginning of
the measurement, a mass loss of approximately 10% was
observed. This is assigned to absorbed water. It is important
to note that the two curves show different slopes. Pure CS loses
the absorbed water faster than the cross-linked polymer. In
fact, the modified polymer has a network-like structure which
is responsible for the slower release of water. Moreover, the
cross-linking caused an increased amount of polar groups in
the polymer, which can absorb more water and bind it via
hydrogen bonds. The TGA-curve shows that the cross-linked
polymer indeed exhibits a greater (but slower) weight loss.
The metal loading of the catalysts was checked by ICP-analyses.
Due to impurities, non-modified CS exhibited a Pd-loading of
Structural properties of the catalyst
To compare the properties of CS and the catalyst, FT-IR and
thermal analysis of both materials have been carried out.
Fig. 1 shows the overlay of the infrared spectra for pure CS
and HMDI-CS/Pd. Pure CS shows a broad band at around
c
New J. Chem.
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Online
(4a–d, Ullmann reaction; Scheme 1).23 The amount of biphenyl
(3e) from the homo-coupling of 1 was very low (o1%) for
most of the reactions. The amount of homo-coupled products
4a–d is considerably higher for the present CS-based catalyst
system (Table 1) than for other Suzuki cross-coupling procedures
9,10
in ball mills using Pd(OAc)2
or Pd(PPh3)4.8 As expected, the
reactions of aryl iodides (2a,b) and aryl bromides (2c,d) lead to
higher reaction rates, resulting in quantitative conversion after
90 min, whereas the chlorides afforded longer reaction times
(120 min). The increased formation of homo-coupling products 4
was evident in recycling experiments. Thus, with p-iodoanisole
(2a), the yield of 4a was 35% in the second run, compared to
25% after the first. m-Iodophenol (2b), however, exclusively gave
the cross-coupling product 3b. Similarly to 2a, p-bromoaceto-
phenone (2c) and p-bromoanisole (2d) gave 84% and 72% yields
of cross-coupling adducts (3c and 3d respectively) and 15%, 22%
of homo-coupling products (4c, 4d respectively) (Table 1). In a
trial with 2c, no reaction occurred in the absence of 1 and only a
5% yield of 4c was isolated. The higher reactivity and selectivity
of aryl bromides towards Suzuki cross-coupling, when compared
to iodides, have also been reported by other authors.8–10 The
formation of the dehalogenated aromatics was observed among
the side reactions.
Fig. 2 TGA- and DTA-curves for pure CS and HMDI cross-linked
catalyst HMDI CS/Pd; black arrows indicate type of signal: relative
mass or DTA.
o9.4 ꢁ 10ꢀ4 wt%, which is a number of magnitudes lower
than found in the prepared catalyst.
Analysis of the Pd-loaded polymer directly after catalyst
preparation and after the fourth reaction cycle (cf. Table 1)
resulted in Pd-contents of 2.3 ꢂ 0.3 and 2.1 ꢂ 0.3 wt%,
respectively. Thus, the catalyst system proves its stability with
respect to metal leaching. However, catalyst recycling entailed an
average weight loss of 3.4% relative to the initial catalyst mass.
XPS-analysis of the oxidation state resulted in a similar Eb, for
the Pd 3d5/2 core level, to the value reported in the literature.17
Besides the reactions with aryl iodides and bromides, note-
worthy results were also obtained with the less reactive aryl
chlorides 2e–i (Table 1). To the best of our knowledge this is
the first report in which aryl chlorides have been cross-coupled
with 1 in a ball mill affording 410% yield.
Catalytic testing
Reports by Schneider and Ondruschka10 as well as Klingensmith
and Leadbeater9 reported yields of only 6% and 2% for the
cross-coupling of chlorobenzene (2e) and p-chlorotoluene with
1, respectively. Moreover, we exclusively recovered cross-
coupling (3e–i) products in these new experiments (Table 1).
Obviously, the stronger C–Cl bond in 2e–i favours oxidative
addition to the catalyst over the dehalogenation accompanied
by homo-coupling seen in 2a–d. High conversions and high
yields were obtained with 2e and p-chloronitrobenzene (2f).
With p-chloroacetophenone (2g), the reaction proceeded
slowly with partial conversion but with high chemoselectivity,
while higher conversions and yields were detected with the
m- (2h) and o-isomer (2i). We suppose that these slight differences
could arise from the coordinating effect of the carbonyl group
on the palladium, stronger in the ortho and meta positions,
thus favouring the oxidative addition. In the reaction with 2f,
the catalyst was recycled three times with a loss in activity.
ICP-analyses of the catalyst after its application revealed that
leaching is not so significant as to explain the activity loss.
Catalytic tests were carried out by comminuting phenylboronic
acid (1) and several aryl halides (2a–i) together with the catalyst
and K2CO38 which was used as the base in a stainless steel milling
beaker mounted on a planetary ball mill. The reaction with
unloaded CS samples led to no conversion at all.17,19 Solid-state
reactions in the presence of HMDI cross-linked CS/Pd(II)
afforded the cross-coupled products (3a–i) and biphenyls
which originated from homo-coupling of the aryl halide
Table 1 Suzuki reaction of aryl halides (2a–i; Scheme 1) with
phenylboronic acid (1) in a ball mill using HMDI-CS/Pd as catalysta
Entries
2a–i
Time/ X2
min (%)
Yield
(%)
Homo-coupling
(%)
Hal
I
R
a
ab
b
c
d
e
p-MeO 90
90
499 74
499 64
90 90
499 84
499 72
499 99
499 84
25
35
I
m-OH
p-Ac
90
90
Br
Br
Cl
Cl
15
22
p-MeO 90
H
p-NO2
120
f
120
120
120
120
120
120
120
fb
fc
fd
g
h
i
95
82
65
29
40
78
75
64
50
29
36
39
Cl
Cl
Cl
p-Ac
m-Ac
o-Ac
X2 = conversion of 2.a Reaction conditions: 1 g catalyst, 0.27 mmol 1,
0.25 mmol 2, 0.5 mmol K2CO3; planetary ball mill, milling beaker
(stainless steel, 125 mL), milling balls (stainless steel, 15 ꢁ 10 mm);
600 minꢀ1. Distribution from GC-MS analysis as a mean of three runs.
Scheme 2 Solid state Suzuki cross-coupling of heteroaryl chlorides
(5) with phenylboronic acid (1) in a planetary ball mill (for conditions
see Table 1; X = conversion).
b
c
Catalyst application: 2nd run. Catalyst application: 3rd run.
Catalyst application: 4th run.
d
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem.

View Online
Weight loss during catalyst recycling is probably responsible
for the observed effect. Using heterocyclic 2-chloropyridine
(5a) and 4-benzoxazole (5b) in the reaction with 1 afforded the
cross-coupling products 6a and 6b in 79 and 62% yields,
respectively (Scheme 2).
2 A. Stolle, T. Szuppa, S. E. S. Leonhardt and B. Ondruschka,
Chem. Soc. Rev., 2011, 40, 2317.
3 K. Tanaka and F. Toda, Chem. Rev., 2000, 100, 1025.
4 K. Tanaka, Solvent-free Organic Synthesis, Wiley-VCH, Weinheim,
2nd edn, 2008.
5 G. Kaupp, J. Phys. Org. Chem., 2008, 21, 630.
6 M. O0Brien, R. Denton and S. V. Ley, Synthesis, 2011, 1157.
7 S. L. James, C. J. Adams, C. Bolm, D. Braga, P. Collier, T. Friscic,
F. Grepioni, K. D. M. Harris, G. Hyett, W. Jones, A. Krebs,
J. Mack, L. Maini, A. G. Orpen, I. P. Parkin, W. C. Shearouse,
J. W. Steed and D. C. Waddell, Chem. Soc. Rev., 2012, 41,
413–447.
The side reaction of aryl–aryl coupling can be slightly
reduced when working under argon atmosphere. This, how-
ever, is a task that cannot easily be when using a planetary ball
mill. In comparison to results previously published for the
cross-coupling of bromo-N-heteroarenes in a similar ball mill,8
the yields obtained with the present catalyst are considerably
higher, although the reaction times are longer.
8 S. Feldbæk Nielsen, D. Peters and O. Axelsson, Synth. Commun.,
2000, 30, 3501.
9 L. M. Klingensmith and N. E. Leadbeater, Tetrahedron Lett.,
2003, 44, 765.
10 F. Schneider and B. Ondruschka, ChemSusChem, 2008, 1, 622.
11 E. Tullberg, D. Peters and T. Frejd, J. Organomet. Chem., 2004,
689, 3778.
Conclusions
12 D. A. Fulmer, W. C. Shearhouse, S. T. Mendonza and J. Mack,
Green Chem., 2009, 11, 1821.
13 R. Thorwirth, A. Stolle and B. Ondruschka, Green Chem., 2010,
12, 985.
14 R. Thorwirth, A. Stolle, B. Ondruschka, A. Wild and
U. S. Schubert, Chem. Commun., 2011, 47, 4370.
15 R. Schmidt, R. Thorwirth, T. Szuppa, A. Stolle, B. Ondruschka
and H. Hopf, Chem.–Eur. J., 2011, 17, 8129.
16 F. Bernhardt, R. Trotzki, T. Szuppa, A. Stolle and B. Ondruschka,
Beilstein J. Org. Chem., 2010, 6, No.7.
17 K. Martina, S. E. S. Leonhardt, B. Ondruschka, M. Curini,
A. Binello and G. Cravotto, J. Mol. Catal. A: Chem., 2011, 334,
60.
In summary, we have reported an interesting protocol for
Suzuki cross-couplings under ball milling conditions using a
HMDI cross-linked CS/Pd-derivative as catalyst. The catalyst
was prepared in water via sonochemical cross-linking with
HMDI in the presence of the metal salt. This catalyst can be
reused three times and shows low metal leaching during its
application in the Suzuki cross-coupling of aryl chlorides. The
average to high yields that were obtained with several aryl
chlorides, compared with the poor results reported in the
literature, may open the door to reactions with poorly reactive
substrates in the ball mill and possibility to catalyst recycling.
18 G. Cravotto, S. Tagliapietra, M. Trotta and B. Robaldo, Ultrason.
Sonochem., 2005, 12, 95.
19 S. E. S. Leonhardt, A. Stolle, B. Ondruschka, G. Cravotto,
C. De Leo, K. D. Jandt and T. F. Keller, Appl. Catal., A, 2010,
379, 30.
20 H.-F. Zhang, L. Zhang and Y.-C. Cui, React. Funct. Polym., 2007,
67, 322.
Acknowledgements
Financial support from MIUR (PRIN 2008 ‘‘A Green
Approach to Process Intensification in Organic Synthesis’’)
is gratefully acknowledged.
21 X. Xu, P. Liu, S. Li, P. Zhang and X. Wang, React. Kinet. Catal.
Lett., 2006, 88, 217.
22 L. Saikia, J. M. Baruah and A. J. Thakur, Org. Med. Chem. Lett.,
2011, 1, 12.
23 G. Cravotto, M. Beggiato, A. Penoni, G. Palmisano,
S. Tollari, J.-M. Leveque and W. Bonrath, Tetrahedron Lett.,
2005, 46, 2267.
Notes and references
1 B. Rodrıguez, A. Bruckmann, T. Rantanen and C. Bolm, Adv.
Synth. Catal., 2007, 349, 2213.
c
New J. Chem.
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012