Stilbene synthesis through decarboxylative cross-coupling of substituted cinnamic acids with aryl halides

Article abstract of DOI:10.1016/j.apcata.2018.04.031

The Pd-catalyzed decarboxylative cross-coupling reaction between cinnamic acid and aryl iodide derivatives was studied using both homogeneous and heterogeneous Pd-catalysts. It was demonstrated that simple Pd(OAc)₂ can catalyze this reaction wi

Full text of DOI:10.1016/j.apcata.2018.04.031

Accepted Manuscript
Title: Stilbene synthesis through decarboxylative
cross-coupling of substituted cinnamic acids with aryl halides
´
Authors: Nelly Rameau, Baptiste Russo, Stephane
Mangematin, Catherine Pinel, Laurent Djakovitch
PII:
S0926-860X(18)30203-5
DOI:
Reference:
https://doi.org/10.1016/j.apcata.2018.04.031
APCATA 16637
To appear in:
Applied Catalysis A: General
Received date:
Revised date:
Accepted date:
2-2-2018
18-4-2018
23-4-2018
Please cite this article as: Rameau N, Russo B, Mangematin S, Pinel C,
Djakovitch L, Stilbene synthesis through decarboxylative cross-coupling of
substituted cinnamic acids with aryl halides, Applied Catalysis A, General (2010),
https://doi.org/10.1016/j.apcata.2018.04.031
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

Stilbene synthesis through decarboxylative cross-coupling of substituted
cinnamic acids with aryl halides
Nelly Rameau, Baptiste Russo, Stéphane Mangematin, Catherine Pinel, Laurent Djakovitch*
Univ Lyon, Université Claude Bernard Lyon 1, CNRS, IRCELYON, UMR 5256, 2 avenue Albert Einstein, F-
69626 Villeurbanne, France.
* corresponding author: laurent.djakovitch@ircelyon.univ-lyon1.fr
Graphical Abstract
Highlights
 Pd-catalyzed decarboxylative cross-coupling of cinnamic acids and aryl iodide was
studied.
 Both stilbenes and corresponding 1,1-biphenyl ethylene were produced at initial time.
 After optimization, homogeneous Pd(OAc)2 gave up to 90% cross-coupling products
combined yields with higher selectivity toward the stilbenes.
 Pd/C and home-made Pd/SiO2 heterogeneous catalysts were also very effective in this
reaction. Pd/SiO2 was found more stable and reusable.
Abstract
The Pd-catalyzed decarboxylative cross-coupling reaction between cinnamic acid and aryl
iodide derivatives was studied using both homogeneous and heterogeneous Pd-catalysts. It
was demonstrated that simple Pd(OAc)2 can catalyze this reaction with useful to high yields
when engaging ferulic acid whatever the nature of the aryl iodide. However, limitations were
found when varying the nature of the cinnamic acid derivatives mainly due to low
decarboxylation process. This could be overcome in some cases by adding Cu(OH)2/1,10-
phenantroline as co-catalyst. In the presence of heterogeneous catalysts, the studies showed
that both Pd/C and home-made Pd/SiO2 catalysts afforded high product yields; however,
1

while less active it was demonstrated that the Pd/SiO2 catalyst could be reused over at least 4
cycles due to higher stability.
Keywords: decarboxylative coupling, stilbene synthesis, cinnamic acid, palladium, homogeneous
catalysis, heterogeneous catalysis.
1. Introduction
Stilbenes are important class of compounds with usually C14 skeleton that found applications in
agrochemical and pharmaceutical industries. For example, natural (poly)phenols stilbenes derivatives
like Resveratrol, Pterostilbene or other Combretastatines A4 and analogues are recognized to have
beneficial health effects such as cancer-preventive, protection against coronary heart disease,
neuroprotective, anti-aging, anti-inflammatory [1, 2]. Other applications include liquid crystals for
non-linear optic, polymer materials and lubricants. Additionally, stilbenes are also known as
substructure of several macrocycles possessing interesting biological activity like the Marchantins and
Riccardins [3-6].
All of these bioactive molecules (and substructures) are naturally available through extraction from
higher plants families and fruits such as grapes and blackberries; however, in limited quantity. To face
with such limitations and answer to the ever increasing demand, the synthesis of these compounds has
attracted much attention in recent decades, and many reviews have been published on this subject [7-
9]. Beside classical total syntheses mainly based on Wittig and Ullmann protocols generally illustrated
through Resveratol [10-13], palladium catalyzed syntheses involving Heck [14-16], Suzuki [7, 17-19],
Negishi [7, 20] or Stille [21, 22] cross-couplings were the object of intensive developments despite
many negative environmental aspects like the use of toxic tin or boron compounds in Stille or Suzuki
couplings, respectively (Figure 1).
2

Figure 1: Common palladium catalyzed synthesis of stilbenes.
To face with these limitations, several authors developed the decarboxylative Heck coupling as an
alternative and more elegant route for the synthesis of stilbene derivatives (Figure 2). This coupling
allows benzoic acid derivatives to react with styrene derivatives in presence of palladium catalyst and
a large excess of Ag2CO3. Although highly investigated, this route is today still limited to the use of
ortho-substituted benzoic acids or α-heteroaromatic acids [23-29].
Figure 2: Decarboxylative Heck coupling toward stilbenes.
Alternative routes involve the decarboxylative coupling of commercially available cinnamic
derivatives with aryl halides. The original methodology reported by Goossen et al. [30, 31] in the
presence of palladium and copper catalysts, was further sporadically developed following this
procedure [32-34] or using excess of silver carbonate in place of the copper catalyst [35, 36]. Recent
progress involves a tandem Heck/Decarboxylation/Heck strategy to produce stilbene derivatives from
aryl halides and acrylic acid [37]. In all these studies, the main limitation remains reaching efficient
decarboxylation of the cinnamic derivatives that resulted in the use of additives, costly ligands or high
temperatures.
Having in hands an efficient methodology to produce styrenes from cinnamic acids like the coumaric,
ferulic or sinapic acids [38] that are readily available in large amount from renewables either as
extracts from annual plants [39-42] or as by-product from lignin depolymerization [43], we were
interested to revisit this methodology for producing stilbenes.
3

2. Experimental section
2.1 General informations: All commercial materials were used as received without further
purification. NMP was purchased from Aldrich as anhydrous NMP in “sure-sealed bottle” and used
without further treatment. Pd/C catalyst (E 105 CA/W 5%Pdwt) was obtained from EVONIK. The
Pd/SiO2 catalyst (8.2 ± 0.1%wt Pd, particle size 5-11 nm) was prepared according the procedure
previously reported by the wet impregnation method of agglomerated silica Aerosil 200 with Pd(acac)2
in toluene followed by calcination under air flow [19, 44]. The palladium content determinations of
the heterogeneous catalyst were performed by ICP-OES spectroscopy from a solution obtained by
treatment of the catalysts with a mixture of HBF4, HNO3 and HCl in a Teflon reactor at 180 °C.
Transmission electron microscopy (TEM) was carried out on a JEOL 2010 microscope with an
instrumental magnification of 50 000x to 100 000x and an acceleration voltage of 200 kV. The point-
to-point resolution of the microscope was 0.19 nm and the resolution between lines was 0.14 nm. The
microscope is equipped with an EDX link ISIS analyzer from Oxford instruments. Energy-dispersive
X-ray microanalysis (EDX) was conducted using a probe size of 25 nm to 100 nm to analyze grains of
the phases. The powder samples were directly deposited on holey-carbon coated copper grid. X-ray
powder diffraction (XRD) data were acquired on a Bruker D5005 diffractometer using Cu K
monochromatic radiation ( = 1.054184 Å). GC analyses were performed on Agilent 4890
chromatograph equipped with a FID detector, a HP-5 column (cross-linked 5% phenyl-
methylsiloxane, 30m x 0.25mm i.d. x 0.25µm film thickness) with nitrogen as carrier gas. Yields were
determined by GC based on the relative area of GC-signals referred to an internal standard (dodecane)
calibrated to the corresponding pure compounds. Purification of products was accomplished by flash
chromatography at a pressure slightly greater than atmospheric pressure using silica (Merck Silica Gel
60, 230 - 400 mesh). Thin layer chromatography was performed on Fluka Silica Gel 60 F254. Liquid 1H
NMR and 13C NMR spectra were recorded in (CD3)2CO at room temperature on a BRUKER AC-250
spectrometer. All chemical shifts were measured relative to residual 1H or 13C NMR resonances in the
deuterated solvents: (CD3)2CO,  2.05 ppm for 1H, 29.8 ppm for 13C. The mass spectra were obtained
on a Shimadzu GC-MS-QP2010S equipped with a AOC-20i+ autosampler and a Sulpelco SLB-5MS
column (95% methylpolysiloxane + 5% phenylpolysiloxane, 30m x 0.25mm x 0.25m) with He as
carrier gas was used. The mass spectra were obtained on a Shimadzu GC-MS-QP2010S equipped with
a AOC-20i+ autosampler and a Sulpelco SLB-5MS column (95% methylpolysiloxane + 5%
phenylpolysiloxane, 30m x 0.25mm x 0.25m) with He as carrier gas was used.
2.2 Typical procedure for the reaction of 1a with 2a
Ferulic acid (349 mg, 1.8 mmol), 4-iodoanisole (351 mg, 1.5mmol), Pd(OAc)2 (22mg, 0.09 mmol),
K2CO3 (414mg, 3 mmol) were added in a 25 mL round-bottom flask. The mixture of solids was
deaerated by a series of vacuum/argon before adding under an argon flow anhydrous NMP (5mL). The
4

reaction mixture was heated up to 80°C for 15h. After cooling to room temperature, the reaction
mixture was diluted with ethyl acetate (20mL), HCl (10%) (5mL) and water (25mL). The aqueous
layer was separated from the organic layer and extracted with ethyl acetate (3x20mL). The combined
organic layers were washed with water and dried over MgSO4 before being filtered and concentrated
under reduced pressure. The residue was purified by flash column chromatography on silica gel using
a mixture of ethyl acetate/petroleum ether as eluent to afford the stilbene 3a as a white solid in 51%
yield. 1H NMR, 13C NMR and MS (from GC-MS) are in full agreement with the literature (CAS
946613-59-8) [45, 46]. Side product 4a was obtained as a yellow oil in 25% yield.
Data for 4a: 1H NMR (250 MHz, Acetone-D6), δ ppm: 7.66 (s, 1H), 7.34 – 7.22 (m, 2H), 6.97 – 6.86
(m, 3H), 6.85 – 6.72 (m, 2H), 5.27 (dd, J = 8.2, 1.5 Hz, 2H), 3.81 (s, 3H), 3.80 (s, 3H).13C NMR (63
MHz, Acetone-d6) δ ppm: 160.45 (C), 150.56 (C), 148.01 (C), 147.50 (C), 134.98 (C), 134.31 (C),
130.17 (CH), 122.10 (CH), 115.47 (CH), 114.33 (CH), 112.48 (CH), 111.47 (CH2), 56.26 (CH3),
55.54 (CH3). C16H16O3 256.11 g/mol, MS (EI): m/z (%) 256.00 (100.00), 240.95 (24.83), 225.95
(12.33), 224.95 (21.79), 153.00 (13.88), 152.00 (14.99), 141.05 (10.71)115.00 (11.51).
When, heterogeneous catalysts were used, the mass of catalyst was adapted regarding the Pd-loading
on the material to reach the Pd-charge in the reaction (i.e. 0.06 equiv. [Pd]).
2.3 Typical procedure for monitoring the reaction of 1a with 2a:
Ferulic acid (349 mg, 1.8 mmol), 4-iodoanisole (351 mg, 1.5mmol), Pd(OAc)2 (22mg, 0.09 mmol),
K2CO3 (414mg, 3 mmol) were added in a 25 mL round-bottom flask. The mixture of solid was
deaerated by a series of vacuum/argon before adding under an argon flow anhydrous NMP (5mL). The
reaction mixture was heated up to 80°C for 15h. Once the reaction was started, time to time, aliquots
(100µL) were removed from the mixture and analysed by GC or GC/MS after adding a calibrated
amount of dodecane used as standard.
2.4 Typical procedure for leaching studies by hot-filtration for the heterogeneous catalysts:
Ferulic acid (349 mg, 1.8 mmol), 4-iodoanisole (351 mg, 1.5mmol), Pd(OAc)2 (22mg, 0.09 mmol),
K2CO3 (414mg, 3 mmol) were added in a 25 mL round-bottom flask. The mixture of solids was
deaerated by a series of vacuum/argon before adding under an argon flow anhydrous NMP (5mL). The
reactor was placed in a preheated oil bath at 80 °C. The reaction was conducted under vigorous stirring
for 20 minutes of reaction. The supernatant solution was filtered through a cannula with a microglass
Whatman filter (in order to remove all fine particles) and then treated for further 5 h under the
standard reaction conditions. The reaction was monitored over the total period by GC and the results
compared to a standard catalytic reaction.
2.5 Typical procedure for the recycling heterogeneous catalysts (Pd/C and PdO/SiO2)
5

For recycling experiments the heterogeneous catalyst used in a first run was separated by
centrifugation, washed with the 2 x 10 mL warmed anhydrous NMP and reused after drying at room
temperature as described in the typical procedure for the reaction of 1a with 2a.
2.6 Data for isolated compounds 3 and 4.
Following isolated compounds gave spectroscopic data in agreement with the literature: 3b (white
solid, 83% yield) (CAS 102145-21-1) [47], 3c (white solid, 65% yield) (CAS 1192280-85-5) [48], 3d
(white solid, 45% yield) (CAS 144336-66-3) [49, 50], 3e (yellow solid, 22% yield) (CAS 124797-64-
4) [51], 3g (white solid, 5% yield) (CAS 1694-19-5) [52, 53], 3h (white solid, 18% yield) (CAS
110993-22-1) [50], 3i (white solid, 50% yield) (CAS 18951-44-5) [54, 55], 3k (yellow solid, 6%
yield) (CAS 15638-14-9) [56, 57], 3l (yellow oil, 18% yield) (CAS 18513-95-6) [58], 3m (white solid,
42% yield) (CAS 746636-48-6) [59], 4g (white solid, 7% yield) (CAS 4333-75-9) [33], 4h (yellow oil,
76% yield) (CAS 439659-96-8) [60], 4i (yellow oil, 48% yield) (CAS 92549-05-8) [61], 4k (yellow
oil, 36% yield) (CAS 4356-69-8) [62, 63], 4l (Yellow oil, 18% yield) (CAS 93652-35-8) [64].
3f: yellow solid, 50% yield. 1H NMR (250 MHz, Acetone-D6), δ ppm: 7.89 – 7.83 (m, 2H), 7.51 –
7.44 (m, 2H), 7.10 – 7.03 (m, 2H), 7.00 – 6.90 (m, 2H), 6.84 (dt, J = 15.1, 1.0 Hz, 1H), 5.98 (s, 1H),
3.84 (s, 3H), 2.51 (s, 3H).13C NMR (63 MHz, Acetone-d6) δ ppm: 197.03 (CO), 148.61 (C), 147.90
(C) , 137.73 (C) , 135.69 (C), 134.95 (CH) , 134.72 (CH) , 132.18 (C) , 128.41 (CH) ,127.26 (CH),
126.05 (CH) , 116.08 (CH) , 113.87 (CH), 56.10 (CH3) , 26.75 (CH3). C16H16O3 268.11 g/mol, MS
(EI): m/z (%) 268.10 (100.00), 253.10 (55.53), 165.10 (26.71), 153.15 (10.46), 152.10 (14.79), 126.65
(18.15), 119.10 (10.29), 105.10 (11.08), 76.05 (15.83), 43.05 (61.24).
4b: yellow oil, 11% yield. 1H NMR (250 MHz, Acetone-D6), δ ppm: 7.34 (td, J = 8.2, 1.8 Hz, 1H),
7.24 (dd, J = 7.7, 1.9 Hz, 1H), 6.99 (td, J = 7.4, 1.0 Hz, 1H), 6.92 (brd, J = 8.4 Hz, 1H), 6.90 (d, J =
1.9 Hz, 1H), 6.81 (d, J = 8.3 Hz, H), 6.74 (dd, J = 8.2, 1.9 Hz, 1H), 5.65 (d, J = 1.4 Hz, 1H), 5.30 (s,
1H), 5.22 (d, J = 1.4 Hz, 1H), 3.85 (s, 3H), 3.68 (s, 3H).13C NMR (63 MHz, Acetone-d6) δ ppm:
157.04 (C), 146.60 (C), 146.14 (C), 145.17 (C), 133.49 (C), 131.23 (CH), 131.15 (C), 128.93 (CH),
120.55 (CH), 119.93 (CH), 113.83 (CH), 113.60 (CH2), 111.14 (CH), 108.74 (CH), 55.86 (CH3),
55.66 (CH3). C16H16O3 256.11 g/mol, MS (EI): m/z (%) 256.10 (95.64), 241.05 (100.00), 213.05
(82.33), 181.05 (33.11), 153.10 (35.65), 152.10 (34.23), 115.05 (24.78), 76.00 (22.07).
4c: yellow oil, 11% yield. 1H NMR (250 MHz, Acetone-D6), δ ppm: 7.26 (t, J = 7.5 Hz, 1H), 6.98-
6.93 (m 1H), 6.93-6.88 (m, 4H), 6.86 (brd, J = 1.5 Hz, 1H), 5.69 (s, 1H), 5.40 (d, J = 1.4 Hz, 1H), 5.38
(d, J = 1.4 Hz, 1H), 3.85 (s, 3H), 3.81 (s, 3H).13C NMR (63 MHz, Acetone-d6) δ ppm: 159.38 (C),
149.71 (C), 146.10 (C), 145.48 (C), 143.17 (C), 133.67 (C), 129.05 (CH), 121.64 (CH), 120.93 (CH),
114.00 (CH), 113.93 (CH), 113.23 (CH), 113.12 (CH2), 110.72 (CH), 55.94 (CH3), 55.24 (CH3).
6

C16H16O3 256.11 g/mol, MS (EI): m/z (%) 256.15 (100.00), 195.10 (13.53), 181.10 (11.25), 153.10
(10.98), 152.10 (15.86), 115.10 (11.18).
4d: yellow oil, 19% yield. 1H NMR (250 MHz, Acetone-D6), δ ppm: 6.95 (d, J = 8.4 Hz, 2H), 6.84 (d,
J = 8.5 Hz, 1H), 6.79 (d, J = 8.4 Hz, 2H), 6.75 (dd, J = 8.4, 1.8 Hz, 1H), 6.70 (d, J = 1.8 Hz, 1H), 5.89
(brs, 2H), 5.59 (d, J = 1.1 Hz, 1H), 5.57 (d, J = 1.1 Hz, 1H), 3.92 (s, 3H).13C NMR (63 MHz, Acetone-
d6) δ ppm: 158.73 (C), 153.59 (C), 148.62 (C), 148.06 (C), 133.39 (C), 132.02 (CH), 131.67 (C),
124.73 (CH), 120.03 (CH2), 116.31 (CH), 115.23 (CH), 114.87 (CH), 55.94 (CH3). C16H16O3 242.09
g/mol, MS (EI): m/z (%) 243.10 (16.97), 242.05 (100.00), 227.05 (20.93), 211.05 (16.24), 199.05
(10.92), 181.05 (21.95), 153.05 (13.56), 152.10 (16.59), 115.05 (12.64), 91.00 (11.36), 76.05 (12.22),
55.00 (11.39).
4f: yellow solid, 7% yield. 1H NMR (250 MHz, Acetone-D6), δ ppm: 7.84 (d, J = 8.6 Hz, 2H), 7.33
(d, J = 8.4 Hz, 2H), 6.80 (d, J = 8.4 Hz, 1H), 6.79 (d, J = 2.0 Hz, 1H), 6.65 (dd, J = 8.5, 1.9 Hz, 1H),
5.93 (s, 1H), 5.62 (s, 2H), 3.91 (s, 3H), 2.54 (s, 3H).13C NMR (63 MHz, Acetone-d6) δ ppm: 197.36
(CO), 153.53 (C), 148.59 (C), 148.08 (C), 141.87 (C), 137.39 (C), 130.76 (CH), 128.49 (CH), 124.71
(CH), 120.05 (CH2), 116.29 (CH), 114.89 (CH), 55.99 (CH3), 26.40 (CH3). C16H16O3 268.11 g/mol,
MS (EI): m/z (%) 269.20 (18.27), 268.25 (100.00), 254.20 (12.81), 253.20 (71.99), 165.15 (29.88),
153.20 (14.01), 152.15 (18.90), 126.65 (18.98), 105.10 (11.79), 76.10 (19.36), 43.10 (84.03).
4m: yellow oil, 8% yield. 1H NMR (250 MHz, Acetone-D6), δ ppm: 7.12 (d, J = 8.5 Hz, 2H), 7.04 (d,
J = 8.5 Hz, 2H), 6.48 (s, 2H), 5.98 (s, 1H), 5.56 (s, 1H), 5.59 (s, 1H), 3.92 (s, 6H), 3.78 (s, 3H). 13C
NMR (63 MHz, Acetone-d6) δ ppm: 153.65 (C), 147.79 (C), 139.34 (C), 134.21 (C), 132.29 (C),
131.96 (CH), 119.62 (CH2), 112.89 (CH), 108.79 (CH), 56.39 (CH3), 55.31 (CH3). C16H16O3 286.12
g/mol, MS (EI): m/z (%) 287.05 (19.11), 286.05 (100.00), 271.05 (9.34), 256.00 (6.51), 255.05
(10.49), 240.05 (6.14), 239.00 (7.64), 225.00 (5.15), 211.00 (13.42), 152.10 (6.18), 140.05 (5.49),
139.05 (8.83), 128.05 (10.65), 127.05 (8.06), 115.05 (6.43).
3. Results and discussion
In order to develop further the synthesis of functional stilbenes from cinnamic acid derivatives, we
explored the decarboxylative cross-coupling between ferulic acid 1a that is readily available in large
amount from market and 4-iodoanisole 2a (Scheme 1), focusing on the formation of stilbene 3a. 1,1-
biphenyl ethylene 4a, can be observed as a co-product of the reaction [33].
7

Scheme 1 : Decarboxylative cross-coupling of ferulic acid and 4-iodoanisole as a benchmark reaction.
Initially, the reaction was carried out using the optimized conditions achieved for the decarboxylation
of cinnamic acids, i.e.: Cu(OH)2 (0.06 equiv.) in presence of phenanthroline (0.12 equiv.) and K2CO3
(2 equiv.) in NMP (5mL) at 150°C for 15h [38]. Under these conditions, neither 3a, nor 4a were
observed (Table 1, entry 1). The only observed products corresponded to those of the proto-
decarboxylation of the ferulic acid and dehalogenation of 4-iodoanisole. Thus, copper used alone as
catalyst was not efficient to perform such decarboxylative cross-coupling. Following the procedures
described previously by Goossen et al. on decarboxylative cross-coupling of benzoic acid with
arylhalides [30, 31], a palladium source was added in catalytic amount to the reaction mixture. Thus,
adding 0.06 equiv. Pd(OAc)2 afforded the desire products in 84% yield with a 3a:4a ratio of 70:30
(Table 1, entry 2). As expected from reported mechanisms, in absence of base (K2CO3) the combined
yields of 3a+4a dropped to 4% (Table 1, entry 3) whereas in absence of copper source the combined
yields remained as high as 78% without noticeable modifications in the 3a:4a ratio (Table 1, entry 4).
However, it was noticed that while phenanthroline ligand was believed to stabilize copper co-catalyst,
in its absence the combined product yields decreased to 63% outlining its role as Pd-species stabilizer
by limiting probably Pd-black formation (Table 1, entry 5).
To conclude on these preliminary studies, the high yield (78%) obtained for 3a/4a using
Pd(OAc)2/phenanthroline/K2CO3 catalytic system encouraged us to optimise further this procedure.
Table 1: Preliminary studies on the decarboxylative cross-coupling reaction of ferulic acid with 4-
iodoanisole.a
Entry
Copper
Palladium Ligand Base
Conv. (%)b
Yield (3a+4a) (%)b Ratio (3a:4a)
Cu(OH)2
1
2
3
4
5
-----
Yes
Yes
Yes
Yes
K2CO3
K2CO3
-----
72
0
-----
Cu(OH)2 Pd(OAc)2
Cu(OH)2 Pd(OAc)2
100
86
84
4
70:30
70:30
70:30
74:26
Pd(OAc)2
Pd(OAc)2
-----
-----
K2CO3
K2CO3
100
99
78
63
No
a Reaction condition: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Cu(OH)2 (0.06 equiv.), Pd(OAc)2 (0.06 equiv.),
phenanthroline (0.12 equiv.), Base (2 equiv.), NMP (5mL), 150°C, 15h. b Conversions and yields determined by GC.
As a first optimization step, we evaluated several Pd-sources including several commercially available
palladium salts and complexes (Table 2, entries 1-6) that were selected according our previous report
on decarboxylative heterocoupling coupling of substituted benzoic acids to biaryls [65], other
parameters like solvent and base, being kept. Whatever the PdX2 salt (X=OAc, Cl, Br, I) used as
8

catalyst precursor, little differences were observed, all leading almost to the same 76-82% combined
yields in expected products with a 3a:4a ratio of ca. 70:30. Noticeably, the Beller-Herrmann’s
palladacycle gave higher selectivity in 3a with a 3a:4a ratio of 80:20 (Table 2, entry 6). The lack of
influence of the palladium source could indicate that, in all cases similar active Pd0-species are
generated from these precursors, more probably in equilibrium with small Pd-particles or Pd-colloids
serving as Pd-storage [66] (see propose mechanisms, Figure 11).
Since little influence of the palladium source was observed, we decided to proceed using the Pd(OAc)2
as catalyst in order to enhance further the yield in stilbene 3a.
Table 2: Influence of the nature of the palladium catalysts on the decarboxylative cross-coupling
reaction of ferulic acid with 4-iodoanisole.a
Entry
[Pd]
Conv. (%)b
Yield (3a+4a) (%)b
Ratio (3a:4a)
1
2
3
4
5
6
Pd(OAc)2
PdCl2
97
98
100
95
97
88
78
76
76
82
77
82
68:32
67:31
70:30
68:32
71:29
80:20
PdBr2
PdI2
Pd2(dba)3
Palladacyclec
a Reaction condition: ferulic acid (1.8 mmol), 4-iodoanisole (1.5 mmol), [Pd] (0.06 equiv.), K2CO3 (2 equiv.), NMP (5mL),
80°C, 15h. b Conversions and yields determined by GC. c NaOAc was used as the base and reaction were performed at
150°C. d after 6h reaction time at 130°C.
Several reaction parameters were evaluated in order to optimize the yields in stilbene/biphenyl 3a+4a,
including the temperature (Table 3), the solvent (Table 4) and the nature of the base (Table 5). Varying
the reaction temperature from 40°C to 150°C, allowed to improve the conversion of 1a after 15h
reaction time from 47% to nearly quantitative that is reached above 80°C. At the same time the
combined yields of 3a+4a increased up to 78% at 80°C before decreasing (Table 2, entry 3); this
reaction temperature afforded also the highest selectivity (80%) in desired products. Little influence of
the reaction temperature on the 3a:4a ratio was observed, the highest ratio in 3a being obtained at
150°C. This result can be explained by considering the formation of both isomers that includes two-
reaction steps, a decarboxylation and a CC-coupling reactions that proceed more probably with
different reaction rates. According to scheme 2, 3a can be obtained via two pathways: route A
involves firstly a protodecarboxylation producing the 4-vinylguaiacol 7a that undergoes a Heck
coupling reaction to give the expected stilbene 3a. Alternatively, 3a can be obtained by route B that
involves a CC-coupling reaction to deliver the intermediate stilbene 5a that gives 3a upon
protodecarboxylation. It is believed that 4a is almost exclusively formed via route C through the
intermediate 6a. From our previous report [38], the protodecarboxylation is relatively slow at low
temperature; therefore, route B and C will be favored producing slightly higher 3a:4a ratios. On the
9

contrary at, and above, 110°C, the protodecarboxylation of 1a is efficient making route A predominant
for enriching the reaction mixture in 3a.
Table 3: Influence of the temperature on the decarboxylative cross-coupling reaction of
ferulic acid and 4-iodoanisole.a
Entry Temperature (°C)
Conv. (%)b
Yield (3a+4a) (%)b
Ratio (3a:4a)
64:36
Selectivityc
1
2
3
4
5
40
47
69
97
98
99
7
15
70
80
78
64
60
48
78
77
63
70:30
80
68:32
110
150
70:30
74:26
a Reaction condition: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Pd(OAc)2 (0.06 equiv.), K2CO3 (2
equiv.) NMP (5mL), 15h. b Conversions and yields determined by GC.c Selectivity = Yield (3a+4a)/Conversion.
Scheme 2: Pathways delivering 3a or 4a in the decarboxylative cross-coupling of ferulic acid and 4-
iodoanisole.
Then, with the aim of improving further the yields in 3a+4a and possibly 3a selectivity, other reaction
parameters (nature of solvent, nature and loading of the base, loading of palladium salt), selected from
our own expertise and review of current literature [67-70], were evaluated at 80°C using Pd(OAc)2 as
catalyst precursor (Tables 4-6). Except in 1,3-propanediol and PEG-2000 (Table 4, entry 4) for which
the yields in excepted coupling products 3a+4a were low (35% with a ratio 3a:4a of 78:22) due to the
protic nature of these solvents, good results were achieved in more classical NMP, DMA and DMF
with combined 3a+4a yields of 78%, 74% and 69% respectively (Table 4, entries 1-3). In these
solvents, the ratio 3a:4a was close to 70:30. Therefore, NMP was used in following optimisations.
Regarding bases (Table 5), clearly the best result was obtained using K2CO3 (Table 5, entry 1); further
optimisation demonstrated that using 0.5-1.5 equiv. provided further yield enhancements up to 88%
with 1 equiv. K2CO3 (Figure 3) with a slight increase of the selectivity toward 3a. Under these
10

conditions (ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Pd(OAc)2 (x equiv.), K2CO3 (1 equiv.)
NMP (5 mL), 80°C, 15h) the loading of palladium catalyst was optimized; using 0.06 equiv. of
Pd(OAc)2 provided the highest yield (88%) in 3a+4a with a 3a:4a ratio of 75:25 (Table 6). Following
the evolution of the reaction mixture versus the time demonstrated that both products, 3a and 4a, were
produced simultaneously with a constant selectivity close to 3a:4a ca. 70:30. While the initial activity
was over Ai=6.0 mol(3a+4a)/molPd/h, it decreased after 1 hour to stabilize after 2 hours to Ades=0.7
mol(3a+4a)/molPd/h probably due to catalyst deactivation as observed from Pd-black formation. Upon the
time, side products, mainly due to the protodecarboxylation, were observed (Figure 4).
Table 4: Influence of the nature of the solvent on the
decarboxylative cross-coupling reaction of ferulic acid and 4-
iodoanisole.a
Entry Solvent
Conv. (%)b
Yield (3a+4a) (%)b
Ratio (3a:4a)
68:32
1
2
3
4
5
NMP
DMF
DMA
97
91
96
78
69
74
35
35
64:36
68:32
Propanediol 96
PEG 2000 98
78:22
79:21
a Reaction condition: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Pd(OAc)2 (0.06
equiv.), K2CO3 (2 equiv.) solvent (5mL), 80°C, 15h. b Conversions and yields determined by
GC.
Table 5: Influence of the nature of the base on the decarboxylative cross-
coupling reaction of ferulic acid and 4-iodoanisole.a
Entry
Base
Conv. (%)b
Yield (3a+4a) (%)b
Ratio (3a:4a)
68:32
1
2
3
4
K2CO3
97
78
61
38
38
NaOAc 75
NaOH 75
Na2CO3 58
73:27
70:30
66:34
a Reaction condition: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Pd(OAc)2 (0.06
equiv.), Base (2 equiv.), NMP (5mL), 80°C, 15h. b Conversions and yields determined
by GC.
Table 6: Influence of palladium loading on the decarboxylative cross-
coupling reaction of ferulic acid and 4-iodoanisole.a
Entry Pd(OAc)2
Conv. (%)
Yield (3a+4a)
Ratio (3a:4a)
-
1
2
3
4
5
6
7
0
72
88
87
93
99
95
96
0
0.01
0.02
0.04
0.06
0.08
0.1
77
81
86
88
89
85
82:18
81:19
78:22
75:25
73:27
71:29
11

a Reaction condition: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Pd(OAc)2,
K2CO3 (1 equiv.), NMP (5mL), 80°C, 15h. b Conversion and yield determined by GC.
100
88
88
84
78
80
60
40
20
0
54
4
0
0,25
0,5
1
1,5
2
K CO (equiv.)
2
3
Figure 3: Influence of base loading on the decarboxylative cross-coupling reaction of ferulic acid and
4-iodoanisole. Reaction condition: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol), Pd(OAc)2 (0.06
equiv.), K2CO3, NMP (5mL), 80°C, 15h. Conversion and yield determined by GC.
100
80
60
40
20
0
100
80
60
40
20
0
4a (%)
3a (%)
4-iodoanisole 2a
Yield (3a+4a) (%)
Yield byproducts (%)
A
i
A
des
0
2
4
6
8
10
12
14
16
0
2
4
6
8
10
12
14
16
[b]
Time (h)
[a]
Time (h)
Figure 4: Reaction monitoring over the time in the decarboxylative cross-coupling reaction of ferulic
acid and 4-iodoanisole at 80 °C. [a] Conversion of 2a and Yield of [3a+4a] versus the time; [b] 3a/4a
selectivity versus the time. Reaction conditions: ferulic acid (0.8 mmol), 4-iodoanisole (0.5 mmol),
Pd(OAc)2 (0.06 equiv.), K2CO3 (1 equiv.), NMP (5mL). Conversion and yield determined by GC.
With these conditions in hand (cinnamic acid (0.8 mmol), aryl iodide (0.5 mmol), Pd(OAc)2 (0.06
equiv.), K2CO3 (1 equiv.) NMP (5 mL), 110°C, 15 h), we then evaluated the scope and limitations of
the reaction using either various aryl iodides or various cinnamic derivatives (Tables 7-8). Exploring
the different isomers of iodoanisole (i.e.: o-, m- and p-iodoanisole), the results demonstrated that the
reactivity follows the general rule, i.e.; ortho >para > meta giving 94%, 78% and 62% 3+4 combined
yields, respectively whereas the 3:4 ratio seems to be influenced by the position of the substituent, the
ortho derivative leading to the lowest 3 selectivity (Table 7, entries 1-3). On the other hand, the
electronic nature of the substituent on the aryl iodide influenced more strongly the outcome of the
reactions. On the contrary to what is generally observed in classical Heck (and associated) coupling
12

reactions, aryl iodides substituted by electron donating groups (OMe and OH) provided higher 3+4
yields (resp. 78% and 64%, respectively Table 7 entries 3 and 4) than those bearing electron
withdrawing groups (NO2, CH3CO) that gave 22% and 57% 3+4 yields, respectively (Table 7, entries
5 and 6) together with high level of nitrobenzene and acetophenone issued from dehalogenation
reaction. This can be related to competitive dehalogenation and cross-coupling reactions both being
catalyzed by Pd-species. Since the rate of the cross-coupling reaction involving decarboxylative step
was relatively low for these highly reactive substrates, then dehalogenation was predominant.
Table 7: Influence of the nature of the aryl iodide on its decarboxylative cross-coupling with the
ferulic acid.a
Entry
1
ArI
Yield (3+4) (%)b
Ratio (3:4)
94
88[3b] :12[4b]
[2b]
[2c]
2
3
4
5
6
62
78
64
22
57
87[3c] :13[4c]
67[3a] :33[4a]
70[3d] :30[4d]
100[3e] :0
[2a]
[2d]
[2e]
87[3f] :13[4f]
[2f]
a Reaction conditions: ferulic acid (1.8 mmol), aryl iodide (1.5 mmol), Pd(OAc)2 (0.06 equiv.), K2CO3 (1
equiv.), NMP (7mL), 110°C, 15 h. b Yields determined by GC.
Next, eight potentially biosourced cinnamic acid derivatives were evaluated for coupling with p-
iodoanisole. They differ by the number and the nature (OH / OMe) of the substituents on the aromatic
ring. Using the reaction conditions reported above, ferrulic acid 1a, p-hydroxycinnamic acid 1i, and m-
hydroxycinnamic acid 1h gave, moderate to low 78%, 75% and 24% 3+4 yields, resp., (Table 8,
entries 1, 3 and 4) while no reaction was observed with the other substituted cinnamic acid derivatives
despite higher temperature (i.e. 130°C). Based on previous studies [38, 65], we assumed that the lack
of reactivity was linked to the difficulty to decarboxylate these substrates. This encouraged us to add
to the reaction mixture [Cu(OH)2/1.10-phenanthroline] as co-catalyst (Cu(OH)2 (0.06 equiv.), 1,10-
phenanthroline (0.12 equiv.)) known to catalyze the decarboxylation reaction. Under these modified
conditions, the yields in cross-coupling products implying 1h and 1i were improved to 94% and 98%,
respectively. Except caffeic acid 1j that did not react and was recovered at the end of the reaction even
13

at 180°C (Table 8, entry 5), all other substrates led to cross-coupling reactions with 4-idoanisole 2a
giving low to moderate yields (Table 8, entries 2, 6-8). In these examples, the main difficulty remained
the low efficiency of the decarboxylation of the cinnamic acids to afford the expected coupling
compounds 3+4 in high yields. Working at higher temperature did not generally improve the situation
since the competitive dehalogenation of 4-iodoanisdole was also strongly accelerated.
Table 8: Influence of the nature of the cinnamic acid on its decarboxylative cross-coupling with the 4-
iodoanisole.a
Yield (3+4) (%)b
A
Ratio (3:4)
A
Entry
Cinnamic acid
B
B
1
2
78
-
80
70[3a] :30[4a]
73[3a] :27[4a]
[1a]
12
94
-
37[3g] :63[4g]
19[3h] :81[4h]
[1g]
3
24
36[3h] :64[4h]
[1h]
4
5
6
75
-
98
0
45[3i] :55[4i]
51[3i] :49[4i]
[1i]
-
-
-
[1j]
-
42
14[3k] :86[4k]
[1k]
7
8
-
-
21
50
-
-
15[3l]:85[4l]
[1l]
87[3m]:13[4m]
[1m]
a Reaction condition: Conditions A : cinnamic acid (1.8 mmol), aryl iodide (1.5 mmol), Pd(OAc)2 (0.06 equiv.), K2CO3 (1
equiv.), NMP (7mL), 130°C, 15 h. Conditions B : cinnamic acid (1.8 mmol), aryl iodide (1.5 mmol), Pd(OAc)2 (0.06 equiv.),
Cu(OH)2 (0.06 equiv.), 1,10-phenantroline (0.12 equiv.), K2CO3 (1 equiv.), NMP (7mL), 130°C, 15 h. b Yields determined by
GC.
14

In addition, we have performed the coupling reaction in the presence of heterogeneous catalysts. Two
types of supported palladium catalysts were evaluated and their characterization is reported.
The Pd/C catalyst type E105 CA/W (3.5%wt Pd determined by AAS) was provided by Evonik and
characterized by TEM and XRD. TEM (Figure 5a) showed a heterogeneous Pd repartition with an
average Pd particle size of 4.7 nm (standard deviation: 3.3 nm). XRD exhibited few crystallites
attributed mainly to PdO species with characteristic 2 at 34.03° (PdO (101)) with an average particle
size of 3-8 nm calculated according to Scherrer equation (Figure 6a). These value are slightly higher
than previous reports on this catalyst (average Pd particle size of 2.6 nm determined by chemisorption)
[71, 72]. Pd at oxidation state Pd0 was also observed with a characteristic 2 at 43.15° (Pd (111)),
while unexpected given the preparation methodology of this Pearlman type catalyst, that was probably
due to partial catalyst reduction during experiment. These data are in agreement with those obtained
from XPS analysis that exhibit peaks at 337.4 eV (70%) and 338.6 eV (8.9%) attributed to Pd2+-
species, together with signals at 335.3 eV (14.4%) attributed to Pd0-species and 336.1 eV (6.7%) that
could be due to either Pd0 or Pd1+-species according a report from Zharmagambetova et al. [73]
(Figure 7).
40
30
25
20
35
30
25
20
15
10
5
15
10
5
0
0
1
6
11 16
0.5
3 5.5 8 10.5
particle size (nm)
[b]
[a]
[c]
particle size (nm)
Figure 5: TEM of [a] the fresh Pd/C catalyst with corresponding histogram of Pd particle size distributions (N =
200) and [b] and [c] fresh Pd/SiO2 catalyst at two magnifications and corresponding histogram of PdO particle
size distribution (N = 200).
15

2500
2000
1500
1000
500
PdO (101)
Pd (111)
0
0
10
20
30
40
2  (°)
50
60
70
80
[a]
2500
2000
1500
1000
500
PdO (101)
PdO (112)
Pd (111)
0
0
10
20
30
40
2  (°)
50
60
70
80
[b]
Figure 6: XRD of the fresh Pd/C [a]and Pd/SiO2 [b] catalysts

16

Figure 7: XPS of the fresh Pd/C catalyst. Deconvolution of the Pd3d5/2 exhibits value at 338.6eV (9%), 337.4eV
(70%), 336.1eV (7%) and 335.3eV (14%). The latter is probably related to partial reduction under electron
beam.
Pd/SiO2 (Pd/SiO2 8.9%wt Pd) was prepared according to procedure reported in the literature by the
wet impregnation method of agglomerated silica Aerosil 200 with Pd(acac)2 in toluene followed by
calcination under air flow[19, 44] and was fully characterized. XRD analysis indicated that
immobilized palladium phase corresponds mainly to crystalline palladium oxide with characteristic
diffraction peaks of 2 at 33.87° (PdO (101)) and 54.70° (PdO (112)) with an average particle size of
6-10 nm calculated according to Scherrer equation (Figure 6b)[74]. A low intensity signal of 2 at
40.12° (Pd (111)) indicated the presence of reduced palladium. However, it cannot be excluded that
this reduction occurs during XRD measurements that is supported by the absence of Pd0 particles in
TEM and agree previous XPS characterizations [44]. TEM of the fresh catalyst clearly showed
particles with an average size of 8 nm with standard deviation of 1.8 nm (Figure 5b and 5c) which
diffractogram exhibited lattice fringes at 2.60 Å, 2.61 Å and 2.65 Å corresponding respectively to (10-
1), (101) and (002) crystallographic planes of PdO.
Both heterogeneous catalysts were evaluated for the decarboxylative cross-coupling reaction of ferulic
acid with 4-iodoanisole. Due to lower activity of these heterogeneous catalysts compared to the
homogeneous ones, the reactions were carried out at 130°C vs 80°C. Regular samplings allowed to
establish the profile of the decarboxylative cross-coupling reaction of ferulic acid with 4-iodoanisole
(conditions: ferulic acid (1.8 mmol), 4-iodoanisole (1.5 mmol), [Pd] (3mol%), K2CO3 (2 equiv.), NMP
(5mL), 130°C, 6 h) (Figure 8). Using Pd/SiO2 catalyst, as expected a short induction period of ca. 5
min, related to surface reaction occurring mainly at corner and edge atoms of the supported large PdO-
particles before delivering dissolved Pd-species in reaction media, was observed before detecting
products. Then the catalyst was active with an Amax= 19 mol(3a+4a)/molPd/h (i.e. corresponding to the
highest rate) affording finally 82% 3a+4b yield after 3 h. Like with homogeneous Pd(OAc)2 catalyst
both coupling products 3a and 4b were produced simultaneously, and the 3a:4a ratio stabilized after
ca. 1 h to 85:25. With Pd/C catalyst, no induction period could be clearly observed and the catalyst
presented a higher activity of Amax= 33 mol(3a+4a)/molPd/h before slowly deactivating after 1 hour. The
absence of induction period using the Pd/C catalyst can be directly related to the average particle size
of the Pd-species onto the carbon support that was shown to be initially smaller (i.e. 4.7 nm for Pd/C
versus 8nm for Pd/SiO2). Finally, 98% yield in 3a+4a was obtained after 6 hours with a 3a:4a ratio of
ca. 75:25 (stabilized after 1 hour). For both catalysts, while more pronounced with Pd/C, 4-
iodoanicole consumption was more rapid than product 3a+4a formation. This can be related to a two
steps formation of 3a/4a that involves two consecutive reactions with different reaction rates.
According to scheme 2, we assume that 3a/4a were mainly obtained through route B and C, i.e. as a
17

sequence involving initial oxidative addition of the aryliodide on Pd(0)-species followed by the cross-
coupling reaction before the decarboxylative step that deliver the products. The first steps are initiated
by Pd(0)-species whereas the last step is mainly catalysed by Pd(II)-species in our case. This was
supported by detection of intermediates 5a and 6a; however in low amount.
100
100
80
60
40
20
0
Amax
Amax
80
60
40
20
0
4-iodoanisole (%)
Yield (3a+4a) (%)
4-iodoanisole 2a
Yield (3a+4a) (%)
0
1
2
3
4
5
6
0
1
2
3
4
5
6
Time (h)
Time (h)
[a]
[b]
Figure 8: Conversions and Yields versus the time for the decarboxylative coupling reaction of ferulic
acid with 4-iodonaisole. a) Pd/SiO2, b) Pd/C. Reaction conditions: ferulic acid (1.75 mmol), aryl
iodide (1.5 mmol), [Pd] (3mol%), K2CO3 (2 equiv.), NMP (5mL), 130°C, 6 h. Yields and conversions
determined by GC.
In order to get better insight regarding the real activity of these two heterogeneous catalyst, hot-
filtration test [75, 76] was performed. Thus, a catalytic run was started as for a standard reaction, and
after 15 min of reaction, the hot reaction mixture was filtered through a celite pad to afford a clear
filtrate that was then treated as a standard catalytic run and its evolution was followed by GC. The
results were compared to that of a standard catalytic run (Figure 9). Initially, the experiment was
conducted using the PdO/SiO2 catalyst. After hot-filtration, the results showed only little progression
of the combined 3a+4a yield from 8 to 10%. However, we observed that the base used, K2CO3, was
not fully soluble in the reaction media; therefore, another experiment was conducted by adding to the
clear filtrate a new charge of K2CO3 (2 equiv.). In that case, the yield of 3a+4a progressed from 8%
just after the end of the filtration procedure to 26% after 4 hours reaction (Figure 6a). The experiment
conducted using Pd/C catalyst (after addition of a new charge of base) indicated more clearly an
increase of the combined yield after hot-filtration as it progressed from 33% after hot-filtration to 84%
after 4 hours (Figure 9b). These results indicated that dissolved palladium released from the support
via the oxidative addition of the aryliodide according the dissolution/redeposition mechanism [15, 77-
79], contributed to the decarboxylative cross-coupling reaction. In both cases, deactivation of the
dissolved Pd-species in the clear filtrate was observed.
18

100
80
60
40
20
0
100
80
60
40
20
0
Hot-filtration (20 min)
Hot-filtration (20 min)
Standard run
Hot-filtration
Hot-filtration with added base
Standard run
Hot-filtration with added base
0
1
2
3
4
5
0
1
2
3
4
5
Time (h)
Time (h)
[a]
[b]
Figure 9: Residual catalytic activity after Hot-filtration  after 15 min reaction versus standard
catalytic  run. [a] Pd/SiO2 and [b] Pd/C. Reaction conditions: ferulic acid (1.75 mmol), aryl iodide
(1.5 mmol), [Pd] (3mol%), K2CO3 (2 equiv.), NMP (5mL), 130°C, 6 h. Yields and conversions
determined by GC.
Further information was provided by determining the level of Pd lixiviation in each case at 20 minutes
of reaction (Table 9). Data indicated that the level of Pd leaching for the Pd/C catalyst was much
higher (2.7%) than that of Pd/SiO2 (0.5%) that reflect, as expected from literature [80], the large
differences in term of particle size. According that origin of leaching, and related catalytic activity,
take place on low-coordinate Pd-atoms that are found mainly at the edge and corner sites of Pd-
particles, this observation is then directly supported by TEM characterizations of prepared catalysts.
Considering that only dissolved species were responsible for observed product formation, the Pd/SiO2
and Pd/C catalysts reached activity of Amax-cor= 3840 and 1260 mol(3a+4a)/molleached-Pd/h, respectively.
However, this simple calculation, while informative, did not reflect the rate of deactivation of the
dissolved Pd-species in solution by palladium black formation that is supposed to be higher in the case
of the Pd/C catalyst given the higher level of metal leaching and the observed palladium aggregation
[15].
Table 9: Determination of dissolved Pd, expressed in ppm and % of introduced Pd.
Leaching
ppm
Entry Catalyst
Amax (mol(3a+4a)/molPd/h)
Amax-cor (mol(3a+4a)/molleached-Pd/h)a
%
1
2
Pd/SiO2
Pd/C
19
33
5
0.5
3840
1260
26
2.7
a Amax-cor = Amax/[leached Pd]
Finally, the recycling of the Pd/SiO2 and Pd/C catalysts was evaluated (Table 10). In order to get
accurate results, an initial loading of 6mol% [Pd] was used to have enough catalytic material to
perform the tests over several cycles. Regardless the catalysts, for the first run, the results after 1 hour
reaction time confirmed the previous experiments, the Pd/C catalyst being more active than the
Pd/SiO2 catalyst (with, resp., 98 and 87 % combined 3a+4a yield). However, since the second run the
behavior changed. While the activity of the Pd/SiO2 catalyst decreased giving 45% combined yield, it
stayed constant over the next two runs. For the Pd/C catalyst, a strong drop of activity was observed
19

since the second run as attested by the limited 23% combined 3a+4a yield that continued to decrease
in the next runs, the catalyst being almost inactive in the third run. These results indicate that, while
highly active, the Pd/C catalyst is not recyclable under our reaction conditions.
Table 10: Recycling Pd/C and Pd/SiO2 catalyst for the decarboxylative cross-coupling reaction of
ferulic acid with 4-iodoanisole.
Leachingb
Run
Yield (3a+4a) (%)a
ppm
(%)
Pd/SiO2
87
Pd/C
98
23
3
Pd/SiO2
Pd/C
22
4
Pd/SiO2
0.16
Pd/C
1.15
0.21
0.26
0.16
1
2
3
4
3
3
2
7
45
0.16
43
5
0.10
50
2
<3
0.37
Reaction conditions: ferulic acid (1.75 mmol), aryl iodide (1.5 mmol), [Pd] (6mol%), K2CO3 (2
equiv.), NMP (5mL), 130°C. a Yields and conversions determined by GC after 1h reaction time. b Pd
leaching expressed in ppm in the bulk solution and percent ratio to introduced palladium amount.
In order to explain these results, additional analysis of the used catalysts were carried out after four
runs. Unfortunately, the remaining potassium salts present at the catalyst surface prevented careful
XRD or XPS analyses. TEM was more informative and pictures of used Pd/C catalyst (Figure 10a)
showed large Pd aggregates up to 100 nm with an average Pd particle size of 22 nm (standard
deviation: 2 nm). Instability of the Pd/C catalyst under the electron beam prevented indexation of the
observed particles. These data contrast with those obtained for the fresh Pd/C catalyst (see Figure 5a).
Therefore, these data indicate that strong agglomeration occurred during the reaction that could
explain the low reusability.
For the used Pd/SiO2 catalyst, TEM analyses of used catalyst showed Pd particles from 3 to 10 nm
with average size of 7 nm with a standard deviation of 1.6 nm (Figure 10b). Therefore, on the contrary
to the Pd/C catalyst no agglomeration of palladium particles occurred during the reaction, explaining
the higher reusability of this catalyst.
20

14
12
10
8
12
10
8
6
6
4
4
2
2
0
0
10 15 20 25 30
particle size (nm)
0.5
3 5.5 8 10.5
particle size (nm)
[a]
[b]
Figure 10: TEM of [a] the Pd/C catalyst after reaction (4 runs) (a) and as received (i.e. fresh) (b) with
corresponding histogram of Pd particle size distributions (N = 200) and [b] the Pd/SiO2 catalyst after reaction (4
runs) with corresponding histogram of PdO particle size distributions (N = 200).
The results observed during the studies could be summarized with the help of Figure 11, in the light of
previous literature reports [38, 65, 67]. Whatever the nature of the palladium source (i.e. Pd-salts,
Pd/C, Pd/SiO2), dissolved Pd(0) and/or Pd(II)-species, stabilized by coordinated solvent, will be
generated in the reaction mixture. These species can undergo to various Pd-materials, active or not.
Depending on the palladium concentration in solution these species can form inactive small Pd-
particles such as Pd-black. This was observed in few reactions initiated by Pd-salts or with Pd/C
heterogeneous catalyst that was evidenced by TEM characterization of used material for the latter (i.e.
Figure 10a versus Figure 5a). The formation of these species is also a source of catalyst deactivation
when using the heterogeneous Pd/C catalyst. Some species, either in native form, or after redox
activation, will undergo as Pd(II)-catalyst to protodecarboxylation cycle. From our results, this cycle is
believed to occur mainly after the Pd-catalyzed Heck-type coupling cycle, which is initiated by Pd0-
species. This assumption is supported by previous publications related to protodecarboxylation of
cinnamyl derivatives [38]. However, these species can also undergo to protodecarboxylation prior the
Heck-type coupling cycle, delivering in the reaction mixture substituted stilbenes that enter then in
classical cross-coupling reaction. This was, for example, believed to occur with ferulic acid that was
found to provide easily the corresponding 4-vinylguaiacol 7 under similar reaction conditions [38].
21

Such reverse order can partially answer to the differences observed in terms of selectivity between
products 3 and 4.
Figure 11: proposed mechanisms in agreement with for the results observed.
4. Conclusion
A procedure for the palladium catalyzed synthesis of stilbenes from cinnamic acid and aryl iodide
derivatives was developed. Detailed study showed that, whatever the reaction conditions, two isomers
were produced, the expected stilbene and the corresponding 1,1-biphenyl ethylene. After optimization
of the reaction conditions, the coupling of ferulic acid and 4-iodoanisole afforded a yield of 78% in the
two cross-coupling products with a ratio stilbene/1,1-biphenyl ethylene ≈ 70/30. Developing this
procedure to other coupling partners revealed that good yields can be obtained varying the nature of
the aryl iodide, but limited possibilities existed when engaging various cinnamic acid derivatives.
Here, clearly the main limitation was related to low efficiency of the palladium catalysts to
decarboxylate the cinnamic substrates; however, the limitation could be overcome in some cases by
adding copper co-catalyst in the reaction mixture following some of our previous reports [38, 65].
For the first time, heterogeneous catalysts (i.e. Pd/C and Pd/SiO2) were engaged in the decarboxylative
cross-coupling reaction of ferulic acid and 4-iodoanisole. Both catalysts gave high product yields;
however, while the Pd/C catalyst was found to be initially more active than the Pd/SiO2 catalyst it led
to strong deactivation since the second reuse due to high rate of palladium leaching during the first run
22

and strong agglomeration of the Pd-particles. On the contrary the Pd/SiO2 catalyst, while leading as
well to some deactivation, could be reused over at least 4 cycles.
Acknowledgements
The authors gratefully acknowledge the National Agency of Research (CHEMLIVAL N° ANR-12-
CDII-0001_01) for funding. N.R. thanks the Ministry of Research and Education for grant.
References
[1]
[2]
[3]
[4]
J.A. Baur, D.A. Sinclair, Nat. Rev. Drug. Discov. 5 (2006) 493-506.
M.-H. Pan, G. Ghai, C.-T. Ho, Mol. Nutr. Food Res. 52 (2008) 43-52.
T. Eicher, S. Fey, W. Puhl, E. Büchel, A. Speicher, Eur. J. Org. Chem. (1998) 877-888.
Y. Kosenkova, M. Polovinka, N. Komarova, D. Korchagina, N. Kurochkina, V. Cheremushkina,
N. Salakhutdinov, Chem. Nat. Compnd. 43 (2007) 712-713.
[5] S. Friederich, U.H. Maier, B. Deus-Neumann, Yoshinori Asakawa, M. H. Zenk, Phytochem. 50
(1999) 589-598.
[6]
L. Fang, H.-F. Guo, H.-X. Lou, Helv. Chim. Acta. 90 (2007) 748-752.
[7]
K. Ferré-Filmon, L. Delaude, A. Demonceau, A.F. Noels, Coord. Chem. Rev. 248 (2004) 2323-
2336.
[8]
K.B. Becker, Synthesis. 1983 (1983) 341-368.
[9]
G.I. Likhtenshtein, Stilbenes Applications in Chemistry, Life Sciences and Materials Science,
Wiley-VCH, 2010, pp. 1-25.
[10]
[11]
[12]
[13]
[14]
[15]
M.B. Andrus, J. Liu, E.L. Meredith, E. Nartey, Tetrahedron Lett. 44 (2003) 4819-4822.
L. Botella, C. Nájera, Tetrahedron. 60 (2004) 5563-5570.
M.B. Andrus, J. Liu, Tetrahedron Lett. 47 (2006) 5811-5814.
A. Farina, C. Ferranti, C. Marra, Nat. Prod. Res. 20 (2006) 247-252.
M. Gruber, S. Chouzier, K. Koehler, L. Djakovitch, Appl. Catal. A: Gen. 265 (2004) 161-169.
L. Djakovitch, K. Koehler, J.G. De Vries, in: D. Astruc (Ed.), Nanoparticles and Catalysis., Wiley-
VCH, Weinheim, 2008, pp. 303-348.
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
M.J. Climent, A. Corma, S. Iborra, M. Mifsud, Adv. Synth. Catal. 349 (2007) 1949-1954.
L. Joucla, G. Cusati, C. Pinel, L. Djakovitch, Tetrahedron Lett. 49 (2008) 4738-4741.
L. Joucla, G. Cusati, C. Pinel, L. Djakovitch, Appl. Catal. A: Gen. 360 (2009) 145-153.
L. Joucla, G. Cusati, C. Pinel, L. Djakovitch, Adv. Synth. Catal. 352 (2010) 1993-2001.
M.S. Kabir, A. Monte, J.M. Cook, Tetrahedron Lett. 48 (2007) 7269-7273.
V. Kogan, Z. Aizenshtat, R. Popovitz-Biro, R. Neumann, Org. Lett. 4 (2002) 3529-3532.
J.-H. Li, X.-C. Hu, Y. Liang, Y.-X. Xie, Tetrahedron. 62 (2006) 31-38.
A.G. Myers, D. Tanaka, M.R. Mannion, J. Am. Chem. Soc. 124 (2002) 11250-11251.
S.-L. Zhang, Y. Fu, R. Shang, Q.-X. Guo, L. Liu, J. Am. Chem. Soc. 132 (2009) 638-646.
P. Hu, J. Kan, W. Su, M. Hong, Org. Lett. 11 (2009) 2341-2344.
Z. Fu, S. Huang, W. Su, M. Hong, Org. Lett. 12 (2010) 4992-4995.
Y. Zhao, Y. Zhang, J. Wang, H. Li, L. Wu, Z. Liu, Synlett. 2010 (2010) 2352-2356.
J. Wang, Z. Cui, Y. Zhang, H. Li, L.-M. Wu, Z. Liu, Org. Bio. Chem. 9 (2011) 663-666.
F. Jafarpour, S. Zarei, M. Barzegar Amiri Olia, N. Jalalimanesh, S. Rahiminejadan, J. Org. Chem.
78 (2013) 2957-2964.
[30]
[31]
L.J. Goossen, G. Deng, L.M. Levy, Science. 313 (2006) 662-664.
L.J. Goossen, N. Rodriguez, B. Melzer, C. Linder, G. Deng, L.M. Levy, J. Am. Chem. Soc. 129
(2007) 4824-4833.
[32]
[33]
G. Cahiez, A. Moyeux, M. Poizat, Chem. Commun. 50 (2014) 8982-8984.
J. Tang, D. Hackenberger, L.J. Goossen, Angew. Chem. Int. Ed. 55 (2016) 11296-11299.
23

[34]
[35]
[36]
2412.
[37]
M. Yamashita, K. Hirano, T. Satoh, M. Miura, Chem. Lett. 39 (2010) 68-69.
Z. Wang, Q. Ding, X. He, J. Wu, Org. Bio. Chem. 7 (2009) 863-865.
X. Wang, S.-y. Li, Y.-m. Pan, H.-s. Wang, Z.-f. Chen, K.-b. Huang, J. Org. Chem. 80 (2015) 2407-
A. Shard, N. Sharma, R. Bharti, S. Dadhwal, R. Kumar, A.K. Sinha, Angew. Chem. Int. Ed. 51
(2012) 12250-12253.
[38]
S. Cadot, N. Rameau, S. Mangematin, C. Pinel, L. Djakovitch, Green Chem. 16 (2014) 3089-
3097.
[39]
B. Bartolomé, C. Gómez-Cordovés, J. Sci. Food Agric. 79 (1999) 435-439.
[40]
D. Hernanz, V. Nuñez, A.I. Sancho, C.B. Faulds, G. Williamson, B. Bartolomé, C. Gómez-
Cordovés, J. Agric. Food Chem. 49 (2001) 4884-4888.
[41] M.M. Pedrosa, M. Muzquiz, C. García-Vallejo, C. Burbano, C. Cuadrado, G. Ayet, L.M.
Robredo, J. Sci. Food Agric. 80 (2000) 459-464.
[42]
[43]
K.-H. Kim, R. Tsao, R. Yang, S.W. Cui, Food Chem. 95 (2006) 466-473.
H.J. Park, K.-H. Park, J.-K. Jeon, J. Kim, R. Ryoo, K.-E. Jeong, S.H. Park, Y.-K. Park, Fuel. 97
(2012) 379-384.
[44]
J. Tarabay, W. Al-Maksoud, F. Jaber, C. Pinel, S. Prakash, L. Djakovitch, Appl. Catal. A: Gen.
388 (2010) 124-133.
[45]
V.M. Bhusainahalli, C. Spatafora, M. Chalal, D. Vervandier-Fasseur, P. Meunier, N. Latruffe, C.
Tringali, Eur. J. Org. Chem. 2012 (2012) 5217-5224.
[46]
J. Hitce, M. Crutizat, C. Bourdon, A. Vives, X. Marat, M. Dalko-Csiba, Green Chem. 17 (2015)
3756-3761.
[47]
[48]
[49]
F. Bertaud, J.P. Croue, B. Legube, Ozone: Sci. Eng. 23 (2001) 139-148.
A. Sharma, N. Sharma, R. Kumar, U.K. Sharma, A.K. Sinha, Chem. Commun. (2009) 5299-5301.
R.I. Campos, X. Wu, M. Elgland, P. Konradsson, P. Hammarstroem, ACS Chem. Neurosci. 7
(2016) 924-940.
[50]
M. Chalal, D. Vervandier-Fasseur, P. Meunier, H. Cattey, J.-C. Hierso, Tetrahedron. 68 (2012)
3899-3907.
[51]
[52]
[53]
[54]
[55]
[56]
[57]
[58]
Z. Li, K. Wu, G. Su, Y. He, Opt. Mater. 20 (2002) 295-299.
L. Djakovitch, K. Koehler, J. Am. Chem. Soc. 123 (2001) 5990-5999.
L. Djakovitch, H. Heise, K. Köhler, J. Organomet. Chem. 584 (1999) 16-26.
A.K. Shil, P. Das, RSC Adv. 5 (2015) 24859-24863.
Z. Chen, M. Luo, Y. Wen, G. Luo, L. Liu, Org. Lett. 16 (2014) 3020-3023.
M.K. Karunananda, N.P. Mankad, J. Am. Chem. Soc. 137 (2015) 14598-14601.
X.-D. Xiao, Y.-L. Bai, J.-Q. Liu, J.-W. Wang, Tetrahedron Lett. 57 (2016) 3385-3388.
G. Valdameri, L. Pereira Rangel, C. Spatafora, J. Guitton, C. Gauthier, O. Arnaud, A. Ferreira-
Pereira, P. Falson, S.M.B. Winnischofer, M.E.M. Rocha, C. Tringali, A. Di Pietro, ACS Chem. Biol. 7
(2012) 322-330.
[59]
V.P. Androutsopoulos, K.C. Ruparelia, A. Papakyriakou, H. Filippakis, A.M. Tsatsakis, D.A.
Spandidos, Eur. J. Med. Chem. 46 (2011) 2586-2595.
[60]
[61]
[62]
[63]
[64]
[65]
[66]
J.G. Cole, P. Wan, Can. J. Chem. 80 (2002) 46-54.
A. Arcadi, S. Cacchi, F. Marinelli, Tetrahedron Lett. 27 (1986) 6397-6400.
A. Gavryushin, C. Kofink, G. Manolikakes, P. Knochel, Tetrahedron. 62 (2006) 7521-7533.
A. Gordillo, E. de Jesus, C. Lopez-Mardomingo, Chem. Commun. (2007) 4056-4058.
M.-Y. Chang, Y.-H. Huang, H.-S. Wang, Tetrahedron. 72 (2016) 3022-3031.
N. Rameau, S. Cadot, A. Paquet, C. Pinel, L. Djakovitch, Top Catal. 57 (2014) 1430-1437.
A.H.M. de Vries, J.M.C.A. Mulders, J.H.M. Mommers, H.J.W. Henderickx, J.G. de Vries, Org.
Lett. 5 (2003) 3285-3288.
[67]
[68]
[69]
N. Rodriguez, L.J. Goossen, Chem. Soc. Rev. 40 (2011) 5030-5048.
J.D. Weaver, A. Recio, A.J. Grenning, J.A. Tunge, Chem. Rev. 111 (2011) 1846-1913.
G.J.P. Perry, I. Larrosa, Eur. J. Org. Chem. 2017 (2017) 3517-3527.
24

[70]
X. Qin, C. Chen, L. Zhang, J. Xu, Y. Pan, H. Zhao, J. Han, H. Li, L. Xu, Tetrahedron. 73 (2017)
2242-2249.
[71]
[72]
[73]
[74]
[75]
[76]
[77]
[78]
K. Köhler, R.G. Heidenreich, J.G.E. Krauter, J. Pietsch, Chem. Eur. J. 8 (2002) 622-631.
R.G. Heidenreich, K. Köhler, J.G.E. Krauter, J. Pietsch, Synlett (2002) 1118-1122.
A.K. Zharmagambetova, V.A. Golodov, Y.P. Saltykov, J. Mol Catal. 55 (1989) 406-414.
K.-T. Li, M.-H. Hsu, I. Wang, Catal. Commun. 9 (2008) 2257-2260.
L. Djakovitch, P. Rollet, Adv. Synth. Catal. 346 (2004) 1782-1792.
P. Rollet, W. Kleist, V. Dufaud, L. Djakovitch, J. Mol Catal. A: Chem. 241 (2005) 39-51.
M. Genelot, V. Dufaud, L. Djakovitch, Adv. Synth. Catal. 355 (2013) 2604-2616.
M. Wagner, K. Köhler, L. Djakovitch, S. Weinkauf, V. Hagen, M. Muhler, Top. Catal. 13 (2000)
319-326.
[79]
L. Djakovitch, M. Wagner, C.G. Hartung, M. Beller, K. Koehler, J. Mol. Catal. A: Chem. 219
(2004) 121-130.
[80]
G. Collins, M. Schmidt, C. O’Dwyer, G. McGlacken, J.D. Holmes, ACS Catalysis. 4 (2014) 3105-
3111.
25