Electrochemical dicarboxylation of conjugated fatty acids as an efficient valorization of carbon dioxide

Article abstract of DOI:10.1039/c3ra00129f

Carbon dioxide was electrochemically incorporated in internal conjugated dienes and the process was optimized to achieve satisfactory yields (>70%) even for less reactive substrates. Reactions were performed galvanostatically in an undivided cell at room temperature with a magnesium or aluminium sacrificial anode. Using an optimized electrosynthetic method for the dicarboxylation of 1,3-cyclohexadiene (optimal electrode material, CO₂ pressure, amount of charge), the effect of molecular configuration and alkyl substitution on the reactivity of conjugated double bonds towards carboxylation was studied. Use of a bubble reactor at atmospheric pressure instead of a higher pressure reactor, and lowering of the current density made it possible to effectively perform the double carboxylation of internal conjugated double bonds in open chains. Conjugated linoleic acid methyl esters were used in this reaction for the first time and by searching for the optimal reaction conditions (solvent, supporting electrolyte, reactant concentration, amount of charge, current density) yields approaching 80% of the corresponding fatty triacid product could be obtained, at current efficiencies over 50%.

Full text of DOI:10.1039/c3ra00129f

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Electrochemical dicarboxylation of conjugated fatty
acids as an efficient valorization of carbon dioxide
Cite this: RSC Advances, 2013, 3,
4634
Roman Matthessen,a Jan Fransaer,b Koen Binnemansc and Dirk E. De Vos*a
Carbon dioxide was electrochemically incorporated in internal conjugated dienes and the process was
optimized to achieve satisfactory yields (.70%) even for less reactive substrates. Reactions were performed
galvanostatically in an undivided cell at room temperature with a magnesium or aluminium sacrificial
anode. Using an optimized electrosynthetic method for the dicarboxylation of 1,3-cyclohexadiene (optimal
electrode material, CO2 pressure, amount of charge), the effect of molecular configuration and alkyl
substitution on the reactivity of conjugated double bonds towards carboxylation was studied. Use of a
bubble reactor at atmospheric pressure instead of a higher pressure reactor, and lowering of the current
density made it possible to effectively perform the double carboxylation of internal conjugated double
bonds in open chains. Conjugated linoleic acid methyl esters were used in this reaction for the first time
and by searching for the optimal reaction conditions (solvent, supporting electrolyte, reactant
concentration, amount of charge, current density) yields approaching 80% of the corresponding fatty
triacid product could be obtained, at current efficiencies over 50%.
Received 9th January 2013,
Accepted 30th January 2013
DOI: 10.1039/c3ra00129f
www.rsc.org/advances
Despite the high thermodynamic stability of CO2, it can
easily be activated by a one-electron reduction at an electrode.
Introduction
Carbon dioxide (CO2) is one of the major contributors to the
greenhouse effect and its increasing concentration in the
atmosphere requires fast and efficient measures.1 CO2 can
also be considered as a non-toxic, non-flammable, low-cost and
renewable carbon source for the synthesis of organic
chemicals.2 Due to its high thermodynamic stability, harsh
reaction conditions are often needed for reactions with CO2 in
traditional organic synthesis. Sodium salicylate synthesis3 and
the hydrogenation of carbon dioxide to methane4 are examples
of industrial processes which require high temperatures to
activate CO2. Research on the chemical fixation of carbon
dioxide in organic chemicals is usually limited to rather reactive
substrates like epoxides,5–11 alcohols,12–14 amines,12,14,15 term-
inal alkynes,16 and butadiene,17,18 and even then, elevated
reaction temperatures and/or complex catalyst systems are
required. Organometallic transition metal compounds can be
used to activate carbon dioxide by the formation of metal–CO2
complexes. When using stoichiometric amounts of these
transition metals, CO2 can be incorporated into unsaturated
substrates, with electron-rich Ni(0) complexes being the most
successful.19
These electrochemical processes can often be conducted at
room temperature since the energy of the electrons is
determined by the applied voltage.2 The electroreduction
takes place on a cathode surface, in this way minimizing the
need for homogeneous organometallic complexes. On the
other hand, these complexes can still be used in catalytic
amounts as redox mediator to increase the CO2 reduction
efficiency and influence the product selectivity.2 Since elec-
tricity will be increasingly of sustainable origin in the future,
organic electrosynthesis is a promising technology for envir-
onmentally friendly chemical processes. A recent review by
Frontana-Uribe and colleagues lists an impressive series of
such transformations.20 Electroreduction can be used to
transform CO2 into valuable products. One option is the
synthesis of fuels like formic acid, methanol or methane via
two-, six- and eight-electron reductions, respectively.2,21,22 This
could be a way of storing electric energy from intermittent
renewable sources, like solar or wind energy.23 Another
approach, investigated in this work, is the fixation of CO2 in
organic chemicals by means of an energy-efficient one-electron
reduction step to produce carboxylic acids, which are
important intermediates in the synthesis of polymers and
pharmaceuticals. Carbon dioxide has been incorporated in
organic compounds such as arylalkenes,24–31 alkynes,24,32–38
aromatic ketones,39–44 halides45–50 and epoxides.51–53 The
electrochemical carboxylation of conjugated dienes has been
investigated in less detail,24,28,54–60 with only one paper
reporting on the use of an acyclic aliphatic internal conjugated
aUniversity of Leuven, Centre for Surface Chemistry and Catalysis, Department of
Microbial and Molecular Systems, Kasteelpark Arenberg 23, P.O. box 2461, B-3001
Leuven, Belgium. E-mail: Dirk.DeVos@biw.kuleuven.be
bUniversity of Leuven, Department of Metallurgy and Materials Engineering,
Kasteelpark Arenberg 44, P.O. box 2450, B-3001 Leuven, Belgium
cUniversity of Leuven, Department of Chemistry, Celestijnenlaan 200F, P.O. box
2404, B-3001 Leuven, Belgium
4634 | RSC Adv., 2013, 3, 4634–4642
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
diene as reactant,56 and most emphasis being paid to the
dicarboxylation of butadiene.54,55,59,60 Faradaic yields for
carboxylating conjugated double bonds are generally poor,
despite use of high CO2 pressures, addition of Ni(0) complexes
as redox mediators or use of reaction temperatures below 0 uC
to increase CO2 solubility.24,55,56 At a CO2 pressure of 3 MPa,
the dicarboxylated product of 1,3-cyclohexadiene was obtained
with a current efficiency of only 35%.55 When using a Ni-
functionalized with a polar group along the chain, with 10-
or 12-hydroxystearic acids as examples of compounds widely
used in stabilization of pigment dispersions. Both CO2 and
linoleic acid are renewable reactants, no additional catalyst is
used and the reactions are performed at atmospheric pressure
and room temperature. These factors contribute to make this a
clean and environmentally benign process.
N,N,N9,N99,N99-pentamethyldiethylenetriamine
(Ni-PMDTA)
catalytic system, a product yield of 60% was obtained for the
reaction with 1,4-diphenylbutadiene.24 Even when working at
210 uC a maximum current efficiency of only 26% was reached
in the electrocarboxylation of methyl sorbate.56
Experimental
Chemicals
The use of dry chemicals is crucial in view of the high water
sensitivity of the electrosynthesis procedure. Water causes an
oxidative deactivation of the anode material and a cathodic
formation of hydrogen and formic acid, thus lowering the
current efficiency of the electrochemical process.
Tetrabutylammonium bromide (TBABr), tetrabutylammonium
iodide (TBAI) and tetrabutylammonium hexafluorophosphate
(TBAPF6) were purchased from Sigma-Aldrich (¢99.0%) and
dried at 60 uC in vacuo for 15 h. N,N-Dimethylformamide
(DMF), acetonitrile (CH3CN) and tetrahydrofuran (THF) were
purchased from Acros Organics (Extra Dry, AcroSeal1) and
were used without further purification. A mixture of fatty acids
containing 61% conjugated linoleic acid (Nouracid1 DE 656)
was obtained from Oleon (Oelegem, Belgium). This mixture
was methylated with acetyl chloride in dry methanol with a
molar ratio methanol/fatty acid of 72 : 1 and acetyl chloride/
fatty acid of 6.6 : 1. After continuous stirring for 20 min at 70
uC, a full conversion to the methylated product was obtained.
HCl was neutralized with an aqueous solution of NaHCO3. The
methyl esters were extracted with ethyl acetate and this
organic phase was dried over MgSO4. Ethyl acetate was
removed under reduced pressure to obtain a mixture of fatty
acid methyl esters containing 61% conjugated linoleic acid
methyl esters (CLAME). Other dienes were purchased from
Sigma-Aldrich and were used without further purification. The
purity of CO2 was 99.7% (Air Liquide).
This paper reports on a significant improvement of the
carboxylation yield of internal conjugated dienes by optimiz-
ing electrosynthesis conditions like electrode material, CO2
pressure and current density. The effect of alkyl substitution
and molecular configuration of the conjugated dienes on the
susceptibility towards electrocarboxylation is investigated. All
reactions were carried out in a simple undivided cell with a
sacrificial magnesium or aluminium anode. The electroche-
mical carboxylation with sacrificial anodes was first intro-
61–63
´
duced by the groups of Perichon
and Silvestri.27,64,65 The
use of such anodes has both advantages and disadvantages.
Advantages are the easy product recuperation of the insoluble
metal salts that are formed,66 and the fact that a high
selectivity for dicarboxylation can be obtained, which is a
critical problem in systems with
a
stable anode.59,60
Disadvantages of sacrificial anodes are the production of salt
waste and the gradual consumption of the anode. Using an
undivided cell also avoids the high ohmic resistance intro-
duced by a membrane separating the anolyte from the
catholyte and reduces the cell manufacturing cost.67
A
particular application field for the dicarboxylation is the
modification of renewable resources, such as vegetable and
animal oils and fats. In unsaturated fatty acids, the carboxylic
group, the double bond(s) and the allylic C–H bonds may be
electroactive.68 Reports have appeared on anodic coupling of
fatty acids by decarboxylation,69 anodic acetoxylation of the
double bond,70 and anodic oxidation of allylic hydrogen atoms
to form ketones.71 However, no research on the cathodic
electrocarboxylation of fatty acids has been reported yet.
In this work, a system is developed by which conjugated
linoleic acid methyl esters (CLAME, 1g) can be dicarboxylated
with good yields (Scheme 1). The fatty triacids (2g) formed in
this reaction may have value as emulsifiers in the food and
pharmaceutical industries, and in dispersion technologies in
general. There are important applications for fatty acids
Electrosynthesis procedure
Prior to the electrosynthesis, the anode and cathode were
cleaned with diluted hydrochloric acid, followed by washing
with distilled water and acetone, and then dried with hot air.
The electrocarboxylation reactions were performed in
a
stainless-steel undivided cell with a cylindrical cathode mesh
(wet surface area of 10 cm2) and a central sacrificial anode rod
(wet surface area of 4 cm2). If required, the cell can be
pressurized. A schematic picture of this setup is shown in
Fig. 1. In a typical experimental procedure, dry solvent (3.5
mL), supporting electrolyte (0.175 mmol) and diene substrate
Scheme 1 Electrocarboxylation of conjugated linoleic acid methyl esters.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 4634–4642 | 4635

View Article Online
Paper
RSC Advances
8-Methyl-4-carboxymethyl-3,7-nonadienoic acid dimethyl
ester (methylated 2b). 1H NMR (CDCl3, 300 MHz): d 1.56 (s,
6H, –CH3), 2.04–2.18 (m, 4H, –CH2–), 3.02–3.06 (d ,2H, –CH2
(–CHL)(–COOCH3)), 3.06 (s, 2H, –CH2(–C¡)(–COOCH3)), 3.68
(s, 6H, –COOCH3) 5.31–5.62 (m, 2H, –CHL); GC–MS (m/z, %):
254 (M+, 0.8), 239 (0.8), 222 (8.9), 207 (8.9), 194 (16.3), 180
(16.3), 166 (35.8), 134 (8.9), 121 (13.0), 107 (8.9), 93 (5.7), 85
(13.0), 73 (100.0), 59 (9.8), 41 (16.3).
4,8,12-Trimethyl-5-carboxy-3,7,11-tridecatrienoic acid
1
dimethyl ester (methylated 2c). H NMR (CDCl3, 300 MHz): d
1.57 (s, 6H, (CH3)2.CL), 1.61–1.73 (m, 6H, CH3–C¡), 2.02–
2.19 (m, 6H, –CH2–), 3.05 (d, 2H, –CH2(–COOCH3)), 3.18 (t, 1H,
.CH(–COOCH3)), 3.68 (s, 6H, –COOCH3), 5.31–5.61 (m, 3H,
–CHL); GC–MS (m/z, %): 322 (M+, 4.1), 290 (6.5), 262 (35.0), 247
(9.8), 235 (10.6), 219 (35.8), 207 (48.8), 175 (61.8), 159 (30.9),
147 (43.1), 135 (59.3), 121 (93.5), 107 (65.0), 93 (45.5), 81
(100.0), 69 (77.2), 55 (69.9), 41 (78.9).
8,12-Dimethyl-4-carboxymethyl-3,7,11-tridecatrienoic acid
1
dimethyl ester (methylated 2d). H NMR (CDCl3, 300 MHz): d
Fig. 1 Schematic representation of the experimental setup used for electro-
carboxylation reactions. (A) pressure reactor mode; (B) bubble reactor mode; (a)
high-pressure stainless-steel cell; (b) PTFE tubing; (c) PTFE liner; (d) cylindrical
cathode; (e) rod-shaped anode; (f) cell holder; (g) sample solution; (h) magnetic
stirrer.
1.57 (s, 6H, (CH3)2 . CL), 1.63 (s, 3H, CH3–C¡), 2.02–2.19 (m,
8H, –CH2–), 3.03–3.07 (d ,2H, –CH2(–CHL)(–COOCH3)), 3.07 (s,
2H, –CH2(–C¡)(–COOCH3)), 3.68 (s, 6H, –COOCH3), 5.31–5.61
(m, 3H, –CHL); GC–MS (m/z, %): 322 (M+, 4.1), 290 (6.5), 262
(35.0), 247 (9.8), 235 (10.6), 219 (35.8), 207 (48.8), 175 (61.8),
159 (30.9), 147 (43.1), 135 (59.3), 121 (93.5), 107 (65.0), 93
(45.5), 81 (100.0), 69 (77.2), 55 (69.9), 41 (78.9).
were added to the cell. This reaction mixture was flushed with
CO2 for 5 min to remove all dissolved oxygen. Afterwards, CO2
was charged into the cell to the desired pressure (setup A) or
continuous bubbling was maintained during the whole
electrosynthesis using a mass flow controller (MFC) (setup
B). The galvanostatic electroreduction was carried out under
continuous stirring at room temperature, using a dc regulated
power supply TS3022S (Thurlby Thandar Instruments, UK).
After the desired amount of charge was supplied the solvent
was distilled at reduced pressure, the residue was acidified
with dilute hydrochloric acid (2 M), and the free acids were
extracted with ethyl acetate (2 6 3 mL). The organic phase was
washed with distilled water and dried over anhydrous MgSO4.
After evaporation of the extraction solvent, a yellow viscous
liquid was obtained, containing the product.
1-Methyl-4-isopropyl-2-cyclohexene-1,4-dicarboxylic acid
methyl ester (methylated 2e). 1H NMR (CDCl3, 300 MHz): d
0.99–1.03 (d, 6 H, (CH3)2.C–), 1.45 (s, 3H, –CH3), 1.54–1.76 (m,
4H, –CH2–), 2.03–2.28 (m, 1H, –CH(,(CH3)2)), 3.67 (s, 6H,
–COOCH3), 5.40–5.54 (m, 2H, –CHL); GC–MS (m/z, %): 254
(M+, 0.8), 222 (13.8), 194 (100.0), 179 (26.8), 162 (13.8), 147
(17.1), 135 (53.6), 119 (15.4), 107 (13.8), 93 (46.3), 77 (13.0), 59
(9.7), 43 (13.8).
2,5-Dimethyl-3-hexenedioic acid dimethyl ester (methylated
2f). 1H NMR (CDCl3, 300 MHz): d 1.26 (d, 6H, –CH3), 3.12–3.18
(m, 2H, .CH(–COOCH3)), 3.68 (s, 6H, –COOCH3), 5.63–5.66
(m, 2H, –CHL); GC–MS (m/z, %): 200 (M+, 10.5), 168 (22.8), 141
(56.9), 125 (18.7), 109 (22.8), 85 (100.0), 67 (17.1), 59 (32.5), 41
(16.3).
9,12-Carboxy-10-octadecenoic acid trimethyl ester (methy-
lated 2g). 1H NMR (CDCl3, 300 MHz): d 0.88 (t, 3H, –CH3),
1.11–1.43 (m, 12H, –CH2–), 1.43–1.55 (m, 4H, –CH2
(–CH2CH,)), 1.55–1.67 (m, 2H, –CH2(–CH2COOCH3)), 1.67
–1.79 (m, 4H, –CH2(–CH,)), 2.30 (t, 2H, –CH2(–COOCH3)),
2.95–3.00 (m, 2H, .CH(–COOCH3)), 3.67 (s, 9H, –COOCH3),
5.50–5.54 (m, 2H, –CHL).
Characterization of products
After electrosynthesis, the acid products were methylated
using acetyl chloride in dry methanol as described earlier. The
products were identified by 1H NMR and mass spectra. 1H
NMR spectra were recorded on an AVANCE 300 (300 MHz,
Bruker, Germany) spectrometer in CDCl3 with Me4Si as an
internal standard. Mass spectra were obtained on a 5973-N
spectrometer connected with an HP 6890 gas chromatograph
(Agilent, USA) with an apolar HP-5 column.
Results and discussion
2-Cyclohexene-1,4-dicarboxylic acid dimethyl ester (methy-
lated 2a). 1H NMR (CDCl3, 300 MHz): d 1.58–1.72 (m, 4H,
–CH2–), 3.11–3.19 (m, 2H, .CH(–COOCH3)), 3.68 (s, 6H,
–COOCH3), 5.42–5.54 (m, 2H, –CHL); GC–MS (m/z, %): 198
(M+, 0.8), 166 (47.2), 139 (27.6), 123 (2.4), 107 (31.7), 79 (100.0),
67 (4.8), 59 (16.3).
Electrocarboxylation of 1,3-cyclohexadiene
In order to evaluate the relative reactivity of various diene
types, one first needs to establish the optimal reaction
conditions for a reference compound. 1,3-Cyclohexadiene
(1a) was chosen as a model compound for a short optimization
of the electrosynthesis parameters. The choice of electrode
material is of great importance in the electrocarboxylation
4636 | RSC Adv., 2013, 3, 4634–4642
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
Table 1 Electrocarboxylation of 1,3-cyclohexadiene: influence of various reac-
tion parametersa
Entry Cathode pCO2 (bar) Qb (F mol21
)
Yieldc (%) gd (%)
1
2
3
4
Ta
Pt
5
5
5
5
5
5
3
10
20
5
5
5
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
3.0
2.0
2.5
4.0
18
28
63
88
96
79
89
85
57
79
93
99
12
19
42
58
64
52
60
57
38
79
75
49
Cu
SS
Ni
Ni
Ni
Ni
Ni
Ni
Ni
Ni
5
6e
7
Scheme 2 Cathodic formation of oxalate, carbon monoxide and carbonate.58
8
9
10
11
12
positive effect of Mg2+ ions on the Ni(0)-catalyzed electro-
chemical incorporation of CO2 into unsaturated bonds was
found.32 Employing a nickel cathode eliminates the need for
an additional metal catalyst, thus simplifying the process.
With a nickel cathode and a magnesium anode as the most
promising combination, the effect of CO2 pressure on the
carboxylation yield was investigated (Table 1, entries 5, 7–9).
The electrosynthesis was carried out under different CO2
pressures with a constant current density of 10 mA cm22 until
an optimum CO2 pressure of 5 bar was found. When the CO2
pressure was lowered to 3 bar, a decrease in current efficiency
was observed. A possible reason is the lower concentration of
dissolved CO2 at reduced pressures. This gives rise to an
increased voltage between cathode and anode when working at
a
Solvent: DMF; supporting electrolyte: TBABr (0.05 mol dm23);
anode: Mg; reactant concentration: 0.2 mol dm23; current density:
10 mA cm22
.
Amount of charge. Yield calculated based on initial
b
c
d
e
amount of 1,3-cyclohexadiene. Current efficiency. Anode: Al.
process. Table 1 (entries 1–6) shows the results obtained with
different cathodes (Ta, Pt, Cu, stainless steel, Ni) and anodes
(Mg, Al) in DMF as the solvent. It has been reported that
tantalum in combination with a nickel(0) catalyst has potential
for CO2 incorporation;69 the high overvoltage for hydrogen
formation could be interesting in lowering the water sensitivity
of these reactions. Platinum, which easily produces hydrogen
from water, was used as reference. However, both Ta and Pt
cathodes gave 2a in very poor yields. Electroreduction of CO2
in protic non-aqueous media mainly forms carbon monoxide
at a Pt or Ni cathode, while a Cu cathode mostly gives
hydrocarbons like methane and ethylene, and Fe and Cr
cathodes predominantly form oxalic acid.2 When using these
metals as cathodes in the aprotic electrocarboxylation of 1,3-
cyclohexadiene the yield increases in the following order: Pt ,
Cu , stainless steel (Fe/Cr/Ni = 74 : 18 : 8) , Ni. Nickel gives
similar CO2 reduction products as platinum in protic media,
but gives excellent diene electrocarboxylation yields in aprotic
media. These results show that there is no direct relation
between the product selectivities of these cathodes in both
types of reaction media. The superior results with nickel
suggest that the electrocarboxylation on the cathode surface is
not only dependent on the rate of electroreduction of CO2
itself, but that a mechanism is operative in which dienes are
activated towards CO2 incorporation, with nickel being a
superior surface for this process. This is in agreement with
literature results where nickel(0) complexes were used to
improve the electrocarboxylation of conjugated dienes.24,54,58
A possible explanation for the beneficial effect is the strong
but reversible adsorption of conjugated dienes on nickel.72
Magnesium as an anode material gave better results than
aluminum when working with a CO2 pressure of 5 bar and at a
current density of 10 mA cm22. Magnesium is more easily
oxidized than aluminium, thus requiring a lower cell potential,
which, under these reaction conditions, eliminates undesir-
able redox reactions like decomposition of the supporting
electrolyte or solvent. Using a nickel cathode and a magne-
sium anode seems the best combination to favor dicarboxyla-
tion. This result also agrees well with literature where a
a
constant current density, which in turn causes the
aforementioned side reactions. Increasing the CO2 pressure
to 10 bar and higher also decreased the carboxylation yield. At
these elevated pressures there is an increased production of
oxalates, carbon monoxide and carbonates (Scheme 2). The
corresponding oxalate and carbonate magnesium salts pre-
cipitate on the electrode surface during electrosynthesis, and
this in turn gradually raises the cell potential.
The electrocarboxylation yield for 1,3-cyclohexadiene was
further optimized by passing various amounts of charge
through the cell (Table 1, entries 5, 10–12). Dicarboxylation
of conjugated dienes is a two-electron reduction, which means
that the theoretical charge required is 2 F mol21. Since the
diene concentration decreases as the reaction proceeds, the
formation of undesired products like oxalate, carbon mon-
oxide and carbonate gradually increases. Hence raising the
amount of charge applied to the system maximizes the yield of
desired product, while lowering the current efficiency. An ideal
compromise between yield and current efficiency was found
when using 2.5 F mol21
.
Electrocarboxylation of various conjugated dienes
Next, the optimized system for 1,3-cyclohexadiene carboxyla-
tion was used to examine the effect of molecular configuration
and alkyl substitution on CO2 incorporation into conjugated
double bonds of mostly renewable compounds. Besides 1,3-
cyclohexadiene (1a), myrcene (1b), a- and b-farnesene (1c +
1d), a-terpinene (1e), 2,4-hexadiene (1f) and CLAME (1g) were
used as substrates. These reactants, their corresponding
products and reaction conditions are listed in Table 2.
Special attention was paid to the importance of controlling
CO2 pressure and current density in optimizing the dicarbox-
ylation of less reactive conjugated dienes. Before focusing on
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 4634–4642 | 4637

View Article Online
Paper
RSC Advances
Table 2 Electrocarboxylation of different conjugated dienes: influence of various reaction parametersa
Entry
1
Reactant
Dicarboxylated product
jb (mA cm22
)
Yieldc (%)
96
gd (%)
10
64
2
3
10
10
93
72
62
48
4
10
21
14
5
10
10
5
52
28
74
34
19
49
6e
7e
8
10
2
2
—
15
46
—
10
37
9
10e
a
Solvent: DMF; supporting electrolyte: TBABr (0.05 mol dm23); cathode: Ni; anode: Mg; pCO2 = 5 bar; reactant concentration: 0.2 mol dm23
.
b
f
c
d
e
Current density. Yield calculated based on initial amount of conjugated diene. Current efficiency. pCO2 = 1 atm (10 mL min21).
Representative formula: the starting compound contains the following fatty acid methyl esters: 0.2% C8, 0.1% C10, 0.1% C11:1, 0.1% C12,
1.7% C16, 1.6% C18, 6.5% C18:1, 26.4% C18:2 (non-conjugated), 61.0% C18:2 (conjugated), 0.4% C18:3, 0.2% C22:1, 0.5% C18:1 OH, 1.2%
other.
the results, it is relevant to sketch the mechanisms proposed
so far for the electrocarboxylation of conjugated dienes.55,56
There are two different hypotheses on the reduction mechan-
ism as shown in Scheme 3. In one view the cathodic reduction
of CO2 is followed by an addition of the [CO2N]2 anion radical
to the diene (mechanism I). The alternative is the electro-
reduction of the diene to an allyl radical anion which performs
a nucleophilic attack on carbon dioxide (mechanism II).56 It
has been suggested that both mechanisms may be operative
simultaneously.55 Since oxalic acid is formed by the dimerisa-
tion of two [CO2N]2 anion radicals and is observed in all
reactions as trace product, it is very likely that at least
mechanism I is operative.
dienes could be explained by a difference in adsorption
strength on the cathode surface. It is well known that dienes
display a strong adsorption on Ni compared to monoenes.73 In
1,3-Cyclohexadiene with its fixed cyclic conformation con-
taining two cis bonds is overall more susceptible towards CO2
incorporation than the mixture of 2,4-hexadiene isomers (E,E/
Z,E/Z,Z = 35 : 43 : 22) (Table 2, entries 1, 5), even at various
CO2 pressures and current densities (Fig. 2). The greater
reactivity of 1,3-cyclohexadiene compared to the 2,4-hexa-
Scheme 3 Reaction mechanisms (I) and (II) for the electrochemical dicarboxyla-
tion of internal conjugated dienes.55
4638 | RSC Adv., 2013, 3, 4634–4642
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
drastic decrease in carboxylation efficiency (entries 1, 4).
Actually, the effect of alkyl substitution is not limited to the
steric hindrance around the CO2 incorporation site but also
comprises the electron-donating properties of these alkyl
chains, making the double bonds less susceptible to a
nucleophilic attack by a [CO2N]2 anion radical (mechanism I)
or for a one-electron reduction (mechanism II).
When CLAME was used as the reactant in these optimized
conditions, no carboxylation product was observed at all
(Table 2, entry 8). Steric hindrance and the electron-donating
characteristics of the long alkyl chains may explain this lack of
reactivity. In the reactions of internal dienes like 2,4-
hexadiene, a-terpinene and CLAME, when working under a
Fig. 2 Competitive reactions of 1,3-cyclohexadiene and 2,4-hexadiene. Solvent:
DMF; supporting electrolyte: TBABr (0.05 mol dm23); cathode: Ni; anode: Mg;
diene concentrations: 0.1 mol dm23; amount of charge: 1 F (mol diene)21; yield
calculated based on initial amount of diene. (A) pCO2 = 5 bar, current density:
10 mA cm22; (B) pCO2 = 1 bar, CO2 flow rate: 10 mL min21, current density: 5
CO2 pressure of 5 bar and a current density of 10 mA cm22
,
there was clear evidence for the precipitation of a magnesium
carbonate/oxalate layer on the anode material which caused
deactivation. Changing the setup from a pressure reactor to a
bubble reactor (Fig. 1), while keeping the same current
density, gave even worse results for the dicarboxylation of
2,4-hexadiene (entry 6). At these low CO2 concentrations,
maintaining a current density of 10 mA cm22 required too
high cell potentials causing electrolysis of the conducting salt.
Working at a lower current density with a CO2 pressure of 5
bar in the CLAME reaction gave only a minor increase in
carboxylation yield (entry 9). A carbonate/oxalate layer was still
formed on the anode, drastically increasing the voltage
between the electrodes. When both the current density and
CO2 pressure were lowered, a significant increase in carbox-
ylation efficiency was observed for both 2,4-hexadiene and
CLAME (entries 7, 10). The rationale is that these less reactive
substrates require a longer reaction time, which is obtained by
lowering the current density, this way avoiding high cell
potentials. Another factor is the easy cathodic formation of
oxalates, carbon monoxide and carbonates at elevated CO2
pressures (Scheme 2). These undesired reactions obviously are
even more favored when working with less reactive substrates
like 2,4-hexadiene and CLAME. When working at atmospheric
CO2 pressure in a bubble reactor, the formation of these by-
products is disfavored so the current efficiency increases and
the electrodes remain active during the whole electrosynthesis
process. An extra advantage of the bubble reactor setup
compared to the pressure reactor setup is the enhanced
turbulence facilitating mass transport around the electrodes.
mA cm22
.
comparison with cyclohexadiene, with its fixed cyclic con-
formation, the adsorption equilibrium seems less favorable for
the linear hexadienes, because the loss of the rotation around
the central C–C bond causes a significant decrease in entropy
upon adsorption. This line of thought would suggest that
dienes undergo the carboxylation according to mechanism II
while adsorbed on the surface, either combined or not with
mechanism I (Scheme 3). From the changing E,E/Z,E/Z,Z ratio
of the remaining hexadiene isomers in the initial phase of the
reaction, one calculates that the E,E-configuration is 1.5 times
less reactive than the Z,E-configuration, which in turn is 1.6
times less reactive than the Z,Z-configuration. Thus, the diene
configuration has a strong stereoelectronic effect on the rate of
the dicarboxylation. The competitive reaction of 1,3-cyclohex-
adiene together with the linear hexadienes shows that the
relative reactivity of 1,3-cyclohexadiene versus the E,E/Z,E/
Z,Z-2,4-hexadiene isomers is 3.5, 2.4 and 1.3 respectively. The
fact that 1,3-cyclohexadiene is more reactive even than Z,Z-2,4-
hexadiene proves that besides the configuration of the diene,
also the fixed conformation, as in 1,3-cyclohexadiene,
increases the reactivity.
The dicarboxylation yield for myrcene is comparable with
the one obtained for 1,3-cyclohexadiene (Table 2, entries 1–2).
On one hand, the terminal double bond of myrcene has an
increased reactivity towards CO2 incorporation owing to the
absence of steric hindrance at the non-substituted LCH2
groups. On the other hand, the rotational freedom of the
conjugated system likely results in a weaker adsorption of
myrcene than cyclohexadiene. Farnesene yields less dicarboxy-
lated product than myrcene which is most likely due to the
presence of a sterically hindered internal double bond in
a-farnesene (1c) (entries 2–3). In the reactions of myrcene and
farnesene, a small fraction of dimeric products was formed.
This is in support of mechanism II with activation of the
diene, at least for these two reactants; neither can mechanism
II be discarded for other reactions, even if no dimerisation
products are formed. The electrocarboxylation of a-terpinene
gave poor yields. When this result is compared with that for
1,3-cyclohexadiene it is clear that alkyl substitution on the 1-
and 4-positions of the conjugated double bond causes a
Electrocarboxylation of CLAME
With the aim of further increasing the yield of 2g, the
influence of some key factors on the process has been studied.
All further electroreductions are carried out at atmospheric
CO2 pressure with a Ni cathode and a Mg anode.
The effect of different solvents on the electrocarboxylation
yield of CLAME was examined. Acceptable yields were
obtained when CH3CN was used as solvent (Table 3, entry 2).
The main reason for the increase in yield is the better
solubility of CLAME, making the reactive groups more
accessible for CO2 incorporation. Other benefits of this solvent
compared to DMF are the larger electrochemical window, the
higher CO2 solubility, the lower viscosity and the limited
toxicity. The reaction of CLAME in THF yielded no carboxyla-
tion product (entry 3), which is probably due to the narrower
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 4634–4642 | 4639

View Article Online
Paper
RSC Advances
(Table 3, entries 2, 8 and 9). As mentioned before, the practical
amount of charge should be higher than the theoretical charge
of 2 F mol21. In this electrocarboxylation reaction of CLAME,
the optimal balance between yield and current efficiency was
Table 3 Electrocarboxylation of conjugated linoleic acid methyl esters: influence
of various reaction parametersa
Entry cb (mol dm23
)
Qc (F mol21
2.5
)
jd (mA cm22 Yielde (%) gf (%)
)
found when using 3 F mol21
.
1g
0.2
0.2
0.2
0.2
0.2
0.1
0.3
0.2
0.2
0.2
0.2
0.2
2
2
2
2
2
2
2
2
2
1
3
5
46
65
—
45
61
49
45
51
71
31
79
32
37
52
—
36
49
39
36
51
47
21
53
21
It seems that the carboxylation yield is very dependent on
the current density that is applied (Table 3, entries 9–12). Here
also an optimum can be found for the present reaction
conditions. When using an optimal current density of 3 mA
cm22, a yield of y80% and a current efficiency above 50%
were reached. Lowering the current density to 1 mA cm22
decreases the cell potential to values below what is needed for
the electrocarboxylation. Increasing the current density to 5
mA cm22 increases the voltage between the electrodes leading
to undesired reactions.
To emphasize the major advance in comparison with earlier
studies, 2,4-hexadiene and CLAME were subjected to the most
efficient electrocarboxylation method published until now for
conjugated dienes, developed by the group of Jiang.55 When
using that method, 2,4-hexadiene was dicarboxylated with a
current efficiency of only 21%, while with CLAME, no
dicarboxylation product was observed at all. While Jiang’s
method is suitable for sterically unhindered dienes, like 1,3-
butadiene, or dienes activated by aromatics, like 1,4-diphe-
nylbutadiene, it is not adapted to aliphatic internal conjugated
dienes. As a result of the method proposed herein, it is now
possible to efficiently carboxylate 2,4-hexadiene and conju-
gated linoleic acid methyl esters with current efficiencies of 49
and 53%, respectively.
2
2.5
2.5
2.5
2.5
2.5
2.5
2.0
3.0
3.0
3.0
3.0
3h
4i
5j
6
7
8
9
10
11
12
a
Solvent: CH3CN; supporting electrolyte: TBABr (0.05 mol dm23);
cathode: Ni; anode: Mg; pCO2 = 1 atm (10 mL min21). CLAME
concentration. Amount of charge. Current density. Yield
b
c
d
e
f
calculated based on initial amount of CLAME. Current efficiency.
g
j
h
i
DMF as solvent. THF as solvent. TBAI as supporting electrolyte.
TBAPF6 as supporting electrolyte.
electrochemical window of this solvent. Room temperature
ionic liquids (RTILs) like 1-methyl-1-propylpyrrolidinium
bis(trifluoromethylsulfonyl)imide and 1-butyl-3-methylimida-
zolium tetrafluoroborate were used as solvent. Working with
RTILs as solvent in the dicarboxylation of internal conjugated
dienes gave no product, although these solvents can be very
useful when starting from activated olefins.25 The main
downside of ionic liquids is their high viscosity which easily
causes diffusion-limited reactions. To by-pass this problem,
increased temperatures up to 50 uC were used, which lowered
the CO2 solubility and in turn gave increased cell potentials
when working with less reactive substrates like CLAME.
Because of this increased cell potential, some decomposition
of the ionic liquid was observed. A more general disadvantage
when employing RTILs in electrocarboxylation reactions with
sacrificial anodes is the cumbersome isolation of the product
salts due to the non-volatile nature of these solvents.25,44,46,52
The electrosynthesis was performed with TBAI, TBAPF6 and
TBABr as supporting electrolytes. Tetrabutylammonium was
chosen as a cation for its high redox stability. As shown in
Table 3 (entries 2, 4 and 5) the yield of 2g is influenced by the
nature of the anion. The total yield decreases in the following
order: TBABr . TBAPF6 & TBAI.
An optimal CLAME concentration was found at 0.2 mol
dm23. When the electroreduction was conducted with a lower
concentration, the yield decreased due to the formation of by-
products like oxalates, carbon monoxide and carbonates and
the electrochemical decomposition of the supporting electro-
lyte. Working at a higher concentration also resulted in a
decrease of carboxylation efficiency through the precipitation
of product salts between cathode and anode leading to an
increased cell voltage and a less successful electrosynthesis
(Table 3, entries 2, 6 and 7). The results show that using the
optimal substrate concentration under well-defined reaction
conditions is of great importance.
Conclusions
Electroreductive synthesis allows the production of new and
complex organic molecules through redox-umpolung of an
electroactive group, thus turning an electron-deficient electro-
phile into an electron-rich nucleophile. This synthetic strategy
uses electrons as clean reagents to generate reactive species
under mild conditions in contrast with traditional organic
chemistry where dangerous and polluting redox agents or high
reaction temperatures are required for activation of CO2.20
A
very simple and efficient electrochemical route for the
dicarboxylation of conjugated linoleic acid methyl esters has
been developed. Ni as cathode, Mg as anode, CH3CN as solvent
and TBABr as supporting electrolyte were found to be the best
experimental parameters for this reaction. A yield of 79%
could be achieved at room temperature by supplying 3 F mol21
at 3 mA cm22 to a 0.2 mol dm23 concentration of CLAME,
while bubbling CO2 at atmospheric pressure through the
reaction mixture. With both reactants being from renewable
sources, no additional catalyst being used and mild reaction
conditions being employed, this process agrees well with the
principles of green chemistry.
By varying the amount of charge passed through the cell, the
yield and current efficiency of 2g formation were examined
4640 | RSC Adv., 2013, 3, 4634–4642
This journal is ß The Royal Society of Chemistry 2013

View Article Online
RSC Advances
Paper
27 S. Gambino, G. Filardo and G. Silvestri, J. Appl.
Electrochem., 1982, 12, 549.
28 D. Ballivet-Tkatchenko, J. C. Folest and J. Tanji, Appl.
Organomet. Chem., 2000, 14, 847.
29 H. Wang, M. Y. Lin, H. J. Fang, T. T. Chen and J. X. Lu,
Chin. J. Chem., 2007, 25, 913.
30 G. Q. Yuan, H. F. Jiang, C. Lin and S. J. Liao, Electrochim.
Acta, 2008, 53, 2170.
Acknowledgements
This research is funded by the Industrieel Onderzoeksfonds
KU Leuven (project IKP/10/005). Dirk De Vos is grateful to
KULeuven for support through the Metusalem grant CASAS,
and in the frame of IAP 7 Supramolecular Chemistry and
Catalysis. The authors are grateful to Oleon for donating the
conjugated linoleic acids.
31 H. Wang, Y. F. Du, M. Y. Lin, K. Zhang and J. X. Lu, Chin. J.
Chem., 2008, 26, 1745.
´
´
32 S. Derien, E. Dunach and J. Perichon, J. Am. Chem. Soc.,
References
1991, 113, 8447.
¨
33 F. Koster, E. Dinjus and E. Dunach, Eur. J. Org. Chem.,
1 J. T. Kiehl and K. E. Trenberth, Bull. Am. Meteorol. Soc.,
1997, 78, 197.
2001, 2507.
34 S. Saito, S. Nakagawa, T. Koizumi, K. Hirayama and
Y. Yamamoto, J. Org. Chem., 1999, 64, 3975.
2 M. Jitaru, J. Univ. Chem. Technol. Metallurgy., 2007, 42, 333.
3 Carbonation of alkali metal phenates, R. L. Briggs,
assigned to Dow Chemical Co., US Pat. 4 171 453, 1979.
4 S. K. Hoekman, A. Broch, C. Robbins and R. Purcell, Int. J.
Greenhouse Gas Control, 2010, 4, 44.
5 A. Zhu, T. Jiang, B. Han, J. Zhang, Y. Xie and X. Ma, Green
Chem., 2007, 9, 169.
6 T. Sakai, Y. Tsutsumi and T. Ema, Green Chem., 2008, 10,
337.
´
´
35 S. Derien, J. C. Clinet, E. Dunach and J. Perichon, J. Org.
Chem., 1993, 58, 2578.
36 G. Q. Yuan, H. F. Jiang and C. Lin, Tetrahedron, 2008, 64,
5866.
37 C. Li, G. Yuan and H. Jiang, Chin. J. Chem., 2010, 28, 1685.
˜
´
´
38 E. Dunach, S. Derien and J. Perichon, J. Organomet. Chem.,
1989, 364, C33.
39 G. Yuan, Z. Li and H. Jiang, Chin. J. Chem., 2009, 27, 1464.
40 A. A. Isse, A. Galia, C. Belfiore, G. Silvestri and A. Gennaro,
J. Electroanal. Chem., 2002, 526, 41.
41 K. Zhang, H. Wang, L. Wu, J. Zhang and J. Lu, Chin. J.
Chem., 2010, 28, 509.
42 O. Scialdone, C. Amatore, A. Galia and G. Filardo, J.
Electroanal. Chem., 2006, 592, 163.
43 Q. Feng, K. Huang, S. Liu, J. Yu and F. Liu, Electrochim.
Acta, 2011, 56, 5137.
44 S. F. Zhao, L. X. Wu, H. Wang, J. X. Lu, A. M. Bond and
J. Zhang, Green Chem., 2011, 13, 3461.
45 A. J. Aishah, M. A. Hartini, S. Normala, A. M. Norhuda, H.
H. N. Hanis, H. M. Razif and T. Sugeng, J. Nat. Gas Chem.,
2007, 16, 273.
7 K. Motokura, S. Itagaki, Y. Iwasawa, A. Miyaji and T. Baba,
Green Chem., 2009, 11, 1876.
8 D. J. Darensbourg, A. Horn Jr. and A. I. Moncada, Green
Chem., 2010, 12, 1376.
9 H. Zhou, Y. M. Wang, W. Z. Zhang, J. P. Qu and X. B. Lu,
Green Chem., 2011, 13, 644.
10 Z. Z. Yang, Y. N. Zhao, L. N. He, J. Gao and Z. S. Yin, Green
Chem., 2012, 14, 519.
11 A. Ion, V. Parvulescu, P. Jacobs and D. De Vos, Appl. Catal.,
A, 2009, 363, 40.
12 K. Tomishige, H. Yasuda, Y. Yoshida, M. Nurunnabi, B. Li
and K. Kunimori, Green Chem., 2004, 6, 206.
13 B. M. Bhanage, S. I. Fujita, Y. Ikushima and M. Arai, Green
Chem., 2003, 5, 340.
46 D. F. Niu, J. Zhang, K. Zhang, T. Xue and J. Lu, Chin. J.
Chem., 2009, 27, 1041.
14 A. Ion, C. Van Doorslaer, V. Parvulescu, P. Jacobs and D. De
Vos, Green Chem., 2008, 10, 111.
47 D. F. Niu, L. P. Xia, A. J. Zhang, G. R. Zhang, Q. Y. Tan and
J. X. Lu, Tetrahedron, 2008, 64, 10517.
15 A. Ion, V. Parvulescu, P. Jacobs and D. De Vos, Green Chem.,
2007, 9, 158.
48 Y. Hiejima, M. Hayashi, A. Uda, S. Oya, H. Kondo,
H. Senboku and K. Takahashi, Phys. Chem. Chem. Phys.,
2010, 12, 1953.
49 Q. Feng, K. Huang, S. Liu and X. Wang, Electrochim. Acta,
2010, 55, 5741.
50 P. L. Fabre and O. Reynes, Electrochem. Commun., 2010, 12,
1360.
51 P. Tascedda, M. Weidmann, E. Dinjus and E. Dunach, Appl.
Organomet. Chem., 2001, 15, 141.
16 D. Yu and Y. Zhang, Green Chem., 2011, 13, 1275.
17 A. Behr and G. Henze, Green Chem., 2011, 13, 25.
18 G. Yuan, I. Satoshi, U. Hirokazu and S. Fumie, Inorg. Chim.
Acta, 1994, 222, 145.
19 X. Yin and J. R. Moss, Coord. Chem. Rev., 1999, 181, 27.
20 B. A. Frontana-Uribe, R. D. Little, J. G. Ibanez, A. Palma and
R. Vasquez-Medrano, Green Chem., 2010, 12, 2099.
21 A. S. Agarwal, Y. Zhai, D. Hill and N. Sridhar,
ChemSusChem, 2011, 4, 1301.
52 H. Yang, Y. Gu, Y. Deng and F. Shi, Chem. Commun., 2002,
274.
22 Y. Chen and M. W. Kanan, J. Am. Chem. Soc., 2012, 134,
1986.
53 Y. Wang, G. Q. Yuan, Y. C. Zeng and H. F. Jiang, Chin. J.
Chem., 2007, 27, 1397.
23 G. A. Olah, G. K. S. Prakash and A. Goeppert, J. Am. Chem.
Soc., 2011, 133, 12881.
54 J. Bringmann and E. Dinjus, Appl. Organomet. Chem., 2001,
15, 135.
´
´
24 S. Derien, J. C. Clinet, E. Dunach and J. Perichon,
Tetrahedron, 1992, 48, 5235.
55 C. H. Li, G. Q. Yuan, X. C. Ji, X. J. Wang, J. S. Ye and H.
F. Jiang, Electrochim. Acta, 2011, 56, 1529.
56 K. Zhang, Y. Xiao, Y. Lan, M. Zhu, H. Wang and J. Lu,
Electrochem. Commun., 2010, 12, 1698.
25 H. Wang, G. Zhang, Y. Liu, Y. Luo and J. Lu, Electrochem.
Commun., 2007, 9, 2235.
26 H. Senboku, H. Komatsu, Y. Fujimura and M. Tokuda,
Synlett, 2001, 3, 418.
This journal is ß The Royal Society of Chemistry 2013
RSC Adv., 2013, 3, 4634–4642 | 4641

View Article Online
Paper
RSC Advances
57 G. Yuan, L. Li, H. Jiang, C. Qi and F. Xie, Chin. J. Chem.,
2010, 28, 1983.
58 H. Horberg, D. Schaefer and B. W. Oster, J. Organomet.
Chem., 1984, 266, 313.
65 G. Silvestri, S. Gambino, G. Filardo, G. Greco and
A. Gulotta, Tetrahedron Lett., 1984, 25, 4307.
66 G. Silvestri, S. Gambino and G. Filardo, Acta Chem. Scand.,
1991, 45, 987.
67 J. Chaussard, M. Troupel, Y. Robin, G. Jacob and P. Juhasz,
J. Appl. Electrochem., 1989, 19, 235.
59 V. A. Grinberg, T. A. Koch, V. M. Mazin, E. I. Mysov and S.
R. Sterlin, Russ. Chem. Bull., 1999, 48, 294.
60 D. Pletcher and J. Tietje Girault, J. Appl. Electrochem., 1986,
16, 791.
¨
68 H. J. Schafer, Eur. J. Lipid Sci. Technol., 2012, 114, 2.
69 A. Weiper-Idelmann, M. aus dem Kahmen and H.
¨
J. Schafer, Acta Chem. Scand., 1998, 52, 672.
´
61 O. Sock, M. Troupel and J. Perichon, Tetrahedron Lett.,
70 H. J. Schaefer, M. Harenbrock, E. Klocke, M. Plate and
A. Weiper-Idelmann, Pure Appl. Chem., 2007, 79, 2047.
71 T. Breton, D. Liaigre and E. M. Belgsir, Tetrahedron Lett.,
2005, 46, 2487.
72 S. Koritala, C. R. Scholfie, E. Selke and H. J. Dutton, J. Am.
Oil Chem. Soc., 1973, 50, 310.
1985, 26, 1509.
62 J. F. Fauvarque, C. Chevrot, A. Jutand, M. Francois and
J. Perichon, J. Organomet. Chem., 1984, 264, 273.
63 M. Troupel, Y. Rollin, J. Perichon and J. F. Fauvarque,
Nouveau Journal de Chimie, 1981, 5, 621.
64 G. Filardo, S. Gambino, G. Silvestri, A. Gennaro and
E. Vianello, J. Electroanal. Chem., 1984, 177, 303.
´
´
´
´ ´
73 A. Molnar, A. Sarkany and M. Varga, J. Mol. Catal. A: Chem.,
2001, 173, 185.
4642 | RSC Adv., 2013, 3, 4634–4642
This journal is ß The Royal Society of Chemistry 2013