Electrocatalysis on silver and silver alloys for dichloromethane and trichloromethane dehalogenation

Article abstract of DOI:10.1016/j.electacta.2003.12.061

The reduction of volatile organic chlorides has been investigated on different electrode materials, in pure aprotic and mixed aqueous-aprotic solvents, by means of cyclic voltammetry (CV) and of preparative electrolyses. Silver resulted in all cases the most electrocatalytic material, but some AgBi and AgSn alloys can be considered as promising alternatives. The CV features of the reduction of trichloromethane show two main reduction steps, mainly governed by the accessibility of the electrode active sites. The reduction end products of preparative electrolyses are methane and traces of higher partially dehalogenated hydrocarbons. Oligomeric products sensibly lower the reaction rate in pure aprotic solvent, while totally disappear in presence of water. Production of methane, generally accompanied by chloromethane, is observed, although to a lower extent, also at working potentials much more positive than the ones corresponding to dichloromethane reduction.

Full text of DOI:10.1016/j.electacta.2003.12.061

Electrochimica Acta 49 (2004) 4035–4046
Electrocatalysis on silver and silver alloys for dichloromethane
and trichloromethane dehalogenation
Sandra Rondinini∗, Alberto Vertova
Department of Physical Chemistry and Electrochemistry, University of Milan, Via Golgi 19, 20133 Milan, Italy
Received 11 November 2003; received in revised form 17 December 2003; accepted 17 December 2003
Available online 2 June 2004
Abstract
The reduction of volatile organic chlorides has been investigated on different electrode materials, in pure aprotic and mixed aqueous-aprotic
solvents, by means of cyclic voltammetry (CV) and of preparative electrolyses. Silver resulted in all cases the most electrocatalytic material,
but some AgBi and AgSn alloys can be considered as promising alternatives. The CV features of the reduction of trichloromethane show two
main reduction steps, mainly governed by the accessibility of the electrode active sites. The reduction end products of preparative electrolyses
are methane and traces of higher partially dehalogenated hydrocarbons. Oligomeric products sensibly lower the reaction rate in pure aprotic
solvent, while totally disappear in presence of water. Production of methane, generally accompanied by chloromethane, is observed, although
to a lower extent, also at working potentials much more positive than the ones corresponding to dichloromethane reduction.
© 2004 Elsevier Ltd. All rights reserved.
Keywords: Electrocatalysis; Silver; Dehalogenation; Polychloromethanes
1. Introduction
1. The energy requirements are much lower on Ag than on
glassy carbon (GC) or mercury electrodes, on account of
the much lower (less negative) reduction potentials. For
example the difference between the voltammetric peak
potentials, ꢀEAg–GC = EAg − EGC, range from 0.3 to
1.4 V in dependence on the substrate [9,10].
2. Ag tends to promote inter-molecular couplings (e.g. di-
merization and adduct formation) against intra-molecular
reactions (e.g. elimination and hydrodehalogenation).
3. The extent of the energy gain (item 1) and the branching
ratio between the two pathways (item 2) are modulated
by the properties of each single “actor” and their modi-
fications by mutual interactions.
In the past years the striking electrocatalytic properties
of silver toward organic halides electroreductions have been
disclosed [1–14] and exploited in synthetic [2–6] and en-
vironmentally oriented [7,8,10] process. At the same time,
the combined results of cyclic voltammetry and preparative
electrolysis experiments on both polycrystalline and single
crystal [11–14] electrodes have provided a deeper insight of
the silver–substrate–medium overall system, which led us to
formulate [9] and substantiate [14] the confinement of the
substrate in a reaction cage, a condition depictable also in
terms of an attenuated radical intermediate, R · · · X · · · Ag,
to highlight that the reduction process is governed by three
actors: namely, the silver surface, the organic moiety, and
the bridging halide group.
Fig. 1 collects the EAg versus EGC values for a representa-
tive set of the data obtained inspecting over 50 compounds
[8–10,15]. Glassy carbon was selected as reference electrode
material on account of its inertness toward organic halides
electroreductions; hence the reactivity scale on GC can be
assumed as representative of the intrinsic reactivity of the
RX substrate. As expected, the general order of reactivity for
the leaving halide group (I > Br > Cl) is retained also on sil-
ver, but the reactivity of iodo-derivatives is so enhanced that
the potential spread observed on GC and finely modulated
The effects of the surface structure [5,12,14], of the react-
ing media [5,6,11,13], and of the substrate molecular struc-
ture [5,8,9] have therefore been extensively studied to obtain
the following general framework:
∗
Corresponding author. Fax: +39 02 503 14203.
E-mail addresses: sandra.rondinini@unimi.it (S. Rondinini),
alberto.vertova@unimi.it (A. Vertova).
0013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2003.12.061

4036
S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
-2.6
-2.4
-2.2
-2.0
-1.8
-1.6
-1.4
-1.2
10
18
9
2
1b
5
12
17
15
1a
14b
7
14a
14c
6
16
4
8
3c
3a
-2.4
11
3d
3b
13
-1.0
-2.0
-2.2
-2.6
-2.8
-3.0
-3.2
EGC / V
Fig. 1. Peak potential values measured on silver, EAg (V), against the parallel values measured on glassy carbon, EGC (V), for the following organic
compounds. Symbols denote: circles, iodo; squares, bromo; and triangles, chlorderivatives, respectively. 1a: trichloromethane first peak; 1b: trichloromethane
second peak; 2: dichloromethane; 3a: 1-I-butane; 3b: 2-I-butane; 3c: 1-I,2-methyl-propane; 3d: 2-I,2-methyl-propane; 4: 1-I-hexane; 5: 1-Br-hexane; 6:
1-Br,6-OH-hexane; 7: 1-Br,8-OH-octane; 8: 1-I-adamantane; 9: 1-Br-adamantane; 10: 2-Br-adamantane; 11: acetobromoglucose; 12: acetochloroglucose;
13: 2,4,6-tribromo-phenol; 14a: 2-Br-phenol; 14b: 3-Br-phenol; 14c: 4-Br-phenol; 15 4-Cl-phenol; 16: 1-metoxy,4-I-benzene; 17: 1-metoxy,4-Br-benzene;
18: 1-metoxy,4-Cl-benzene. 1a, 1b and 2, this work; 3–6, 8–10, Ref. [9]; 7, Ref. [15]; 11–12, Ref. [6]; 13–18, Refs. [8,9].
by the organic structure, is compressed (flattened) in a nar-
row range to testify the leading role of the silver–halogen
interactions. This striking leveling effect is much more at-
tenuated for bromo- and chloro-compounds, for which is
then possible to evidentiate a finer reactivity scale in depen-
dence on the presence of other adsorption auxiliary groups
[8–10].
This potential series, representative of the relative activa-
tion energy for the formation of the attenuated radical inter-
mediate, R · · · X · · · Ag, can be nonetheless altered by the
competitive adsorption of other substances, namely, the sol-
vent and the supporting electrolyte ions. While the range of
solvents used so far has been restricted, mainly for solubil-
ity reasons, to acetonitrile (ACN) and ACN/H2O mixtures,
the role played by the supporting electrolyte has been inves-
tigated [5,6,11,13], with special attention to the halide salts
[13].
that during oxidative treatments, including the electro-
chemical oxidation, traces of highly toxic compounds like
dioxins and/or PCB’s might be produced, together with a
non-negligible amount of volatile organic halides (VOH)
which can then escape with the other gaseous products. In
turn, VOH constitute a class of highly toxic pollutants non
only of the gas emissions, but also of drinking waters, as
they include many popular chlorinated chemicals of both
civil and industrial use.
In this context, the extension of our investigations toward
these halides, whose short carbon chain (one or two car-
bon atoms) is accompanied by the high reactivity of the
C–X bond, e.g. iodides are well known alkylating agents,
was straightforward. The interest was also stimulated by the
rather rich electrochemical literature on these compounds
[16–23], which reported, on different cathode materials, ei-
ther dimerization products in non protic media [16,18,19]
or hydrodehalogenation reactions [17,20–23] in water and
in acidified non-aqueous solvents.
Taking into account this high reactivity, we chose as
model compounds the chloride derivatives, and in particu-
lar dichloro- and trichloromethane, mainly by considering
their wide use in organic chemistry. Moreover, the selected
molecules, with their minimal organic structure and the
geminal halide groups, constitute excellent substrates to fur-
ther investigate the rules of selection between the 1e−/2e−
pathways.
Two rather popular silver alloys, namely, AgBi and AgSn
low melting point alloys, known to improve mechanical and
chemical properties of silver, have been also investigated,
with the long-range goal of increasing life-time and shelf-life
of the working cathodes.
Parallelly, the R · · · X · · · Ag intermediate evolution to-
ward 1e− couplings or 2e− intra-molecular reactions is ev-
idenced by the results of preparative electrolyses, as sum-
marized in Table 1.
It is evident that the 1e− inter-molecular pathway is fa-
vored in the case of aliphatic compounds not particularly
hindered by a bulky structure and possibly supported by
the presence of groups specifically interacting with silver,
while the 2e− intra-molecular pathway, and in particular
the hydrodehalogenation, is the unique reactions route for
halophenols. This last result is particularly interesting from
an environmental point of view, since the safe electrochem-
ical degradation of such bio-toxic compounds to the more
biodegradable phenols might constitute a preliminary step
of an integrated waste treatment. In fact, it is well known

S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
4037
Table 1
Synopsis of preparative electrolyses on silver cathodes for different substrates
Substrate
Operating conditions
Reaction pathways
νe (F/mol)
Reference
3-Br-phenol
4-Br-phenol
4-Br-phenol
2,6-Di-Br-phenol
2,6-Di-Br-phenol
2,4-Di-Br-phenol
2,4-Di-Br-phenol
2,4,6-Tri-Br-phenol
2,4,6-Tri-Br-phenol
1-Br-adamantane
1-Br-adamantane
2-Br-adamantane
TEAP/ACN
TEAP/ACN
Hydrodehalogenation
Hydrodehalogenation
Hydrodehalogenation
Hydrodehalogenation
Hydrodehalogenation
Hydrodehalogenation
Hydrodehalogenation
Hydrodehalogenation
2
2
2
2
2
2
2
2
[8,10]
[6,8,10]
[10]
[8,10]
[10]
[6,8,10]
[10]
[8,10]
[8,10]
[9]
TEAP + TEABr/ACN
TEAP/ACN
TEAP + KCl/ACN/H2O 2/9
TEAP/ACN
Na2HPO4 + TEABr/H2O
TEAP/ACN
TEAP/H2O:ACN = 1:1
TEAP/ACN:toluene = 3:2
TEAP/ACN:toluene = 3:2
TEAP/ACN:toluene = 3:2
TEAP/ACN:toluene = 3:2
TEAP/ACN:toluene = 3:2
TEAP/ACN
Hydrodehalogenation
2
Hydrodehalogenation/dimerization
Hydrodehalogenation/dimerization
Hydrodehalogenation/dimerization
Hydrodehalogenation/dimerization
Hydrodehalogenation/dimerization
␣,␣ꢀ:␤␤ꢀ:␣␤ = 1:1:2 dimers
1.9
1.97
1.7
1.5
1.7
≈1
[9]
[9]
[9]
[9]
1-I-adamantane
1-Br-adamantane + 2-Br-adamantane
Tetra-O-acetyl-␣-glucopyranosylbromide
(ABG)
Tri-O-acetyl-␣-d-fucopyranosylbromide
6-Deoxy-6I-1,2:3,4-di-O-isopropylidene-␣-
d-galactopyranose
[3,5,6]
TEAP/ACN
TEAP/ACN
␣,␣ꢀ:␤␤ꢀ:␣␤ = 1:1:2 dimers
≈1
≈1
[3,6]
[3,6]
Dimerization (non-anomeric dimer)
2. Experimental
Composition of the final alloy was checked with SEM-
EDX measurements, both before and after CV experiments.
All substrates, being of analytical grade, were used as
received from the producers (Aldrich and Merck).
2.2. Preparative electrolysis experiments
Preparative electrolyses were carried out in a two-
compartment cell, divided by an ion-exchange membrane.
In only one case (first electrolysis) an anion-exchange
membrane (Sybron® MA3475) was used, while in all other
cases a Nafion®120 cationic membrane was adopted. The
catholyte and anolyte volumes were 200–300 cm3. The gas
output of the cathodic compartment was equipped with a
condenser operated at −35 ◦C with a mixture of solid CO2
and acetone, to guarantee the total reflux of all the volatile
products, with the only exception of CH4 and CH3Cl. The
operating conditions for each substrate are summarized in
Table 3. The cathode consisted either of a rectangular sil-
ver plate or, in one case, graphite rod. As anodes, a silver
plate in ACN saturated with tetraethylammonium bromide
(TEABr), to provide a harmless counter reaction (namely,
Ag + Br−→ AgBr + e−), or a Pt foil in aqueous elec-
trolyte (HClO4 0.5 M, Na2SO4 1 M or KOH 0.5 M) were
alternatively used.
An AMEL 555B potentiostat/galvanostat was used to run
electrolyses either in potentiostatic or galvanostatic mode, a
Luggin capillary allowing safe control of the silver cathode
potential versus a SCE.
The electrolyses were carried out in HPLC grade acetoni-
trile (ACN) or in 50% (vol) admixture with highly deionized
water (MilliQ® Millipore System) with either 0.5 M TEABr
or 0.1 M TEATFB as supporting electrolyte. The progress
of the electrolytic process was monitored by gas chromatog-
raphy using a HP 5890 gas chromatograph, equipped with
a capillary column (HP-5; cross-linked 5% PH ME silox-
ane, 30 m) a FID detector connected with a computer. The
2.1. Cyclic voltammetry (CV) experiments
Voltammetric investigations on CHCl3 and CH2Cl2
(Table 2) were carried out by an AMEL 5000 potentio-
stat/galvanostat driven by Corrware Scribner Associates
Inc., in the 100–1000 mV/s range of sweep rates; the sol-
vents were HPLC grade acetonitrile (ACN), and highly
deionized water (MilliQ® Millipore System), either pure or
in admixture, with 0.1 M (TEAP) or tetraethylammonium
tetrafluoborate (TEATFB) as supporting electrolytes, and
with an aqueous saturated calomel electrode (SCE) as a
reference electrode. Ag working electrodes consisted of
platinum wires (0.5 mm diameter, 1 cm length), covered
with silver by cathodic deposition against a silver anode
from a 10 g dm−3 KAg(CN)2 bath, at 0.5 mA cm−2 depo-
sition current density. Normal cleaning procedure implied
sonication in the operating solvent for 5 min [5,12]. Other
working electrodes were an AMEL 492 “GC” glassy carbon
electrode (∅ 0.3 cm; optimized cleaning procedure: gen-
tle rubbing with cotton gauge soaked in slurry of alumina
powder 0.05 ␮m, followed by 5ꢀ ultrasound treatment in the
relevant solvent), and a series of silver alloys with Bi and Sn
obtained by electrodeposition in the following conditions:
• Ag/Sn: bath 25 g/l SnCl2·2H2O, 200 g/l K4P2O7, 4.7 g/l
AgI, 330 g/l KI; current density 6 mA cm−2
.
• Ag/Bi: bath 15 g/l KAg(CN)2, 21 g/l BiNO4, 60 g/l
KNaC4H4O6·4H2O, 26 g/l NaOH, current density for
3% Bi 1 mA cm−2, for 30% Bi 5 mA cm−2, for 50% Bi
8 mA cm−2
.

4038
S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
Table 2
Reduction peak potentials of trichloromethane and dichloromethane
Electrode material
Substrate
Supporting electrolyte
Ep1 (V)
Ep2 (V)
Ag
Ag
Ag
Ag
Ag
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
0.1 M TEAP in ACN
−1.50
−2.12
−2.23
−2.05
−2.1a
n.d.
0.1 M TEAI in ACN
−1.52
0.1 M TEAP + 10−3 M KCN in ACN
−1.61
0.1 M TEATFB in ACN
−1.45a
0.1 M TEATFB in ACN/H2O 1/1
−1.23 → −1.33a;
−1.14 → −1.24a
−1.60a
AgBi 3 at.%
AgBi 3 at.%
AgBi 30 at.%
Bi
AgSn 64 at.%
Sn
GC
GC
GC
GC
GC
Ag
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CHCl3
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
CH2Cl2
0.1 M TEATFB in ACN
0.1 M TEAP in ACN
−2.24a
−2.5
−1.64
0.1 M TEATFB in ACN
0.1 M TEATFB in ACN
0.1 M TEATFB in ACN
0.1 M TEATFB in ACN
0.1 M TEATFB in ACN
0.1 M TEAP in ACN
0.1 M TEABr in ACN
0.1 M TEATFB in ACN/H2O 1/1
0.1 M TEATFB in ACN/H2O 1/3
0.1 M TEAP in ACN
0.1 M TEAI in ACN
0.1 M TEATFB in ACN
0.1 M TEAP in ACN
0.1 M NaClO4 in ACN
0.1 M TEATFB in ACN
0.1 M TEAP in ACN
−1.7a
−2.3a
−1.87a
−1.65a
−2.3a
−2a
n.d.
−2.14
−2.67
−2.70
−2.63
−2.11
−1.93
−2.11a
−2.02a
−2.20
−2.22
−2.5
Ag
AgBi 3%
AgSn 64%
Sn
GC
GC
−2.23
−2.5
−2.75
−2.66
Ep1 (V) denotes the first reduction peak of CHCl3 and Ep2 (V) denotes either the second reduction peak of CHCl3 or the single reduction peak of
CH2Cl2. Scan rate 100 mV/s and T = 25 1 ◦C.
a
T ≈ 30 ◦C.
signal analysis was carried out using the software provided
by Azur® (DATALYS, San Martin D’Heres). The products
were identified by comparison with commercial standards
(Aldrich) either in gas chromatographic analysis or in CV.
While the differences observed on GC for the different
conditions are not accompanied by modifications of the peak
shape, the well documented dependence of reduction po-
tentials of organic halides on the silver surface preparation
and modification [12,13] is confirmed also for the present
compounds, and in particular for trichloromethane. In fact,
as shown in Fig. 3, the first cycle might significantly dif-
fer from the subsequent ones, not only for the obvious sta-
bilization of the concentration profiles, but also in connec-
tion with surface “aging”, or, more specifically, adsorption
competition among the various solution components. Inter-
estingly, one can individuate up to three peaks assignable to
trichloromethane, the first being the more evident the cleaner
the electrode surface (Fig. 3a, freshly prepared surface), the
intermediate being always present in both the first and the
last cycle (Fig. 3a and b; (b) cleaned but aged surface), the
last appearing only in presence of some strongly adsorbed
specie (e.g. I− or CN−) (Fig. 3c and d).
3. Results and discussion
3.1. Cyclic voltammetry
Fig. 2 shows the typical CV’s recorded on silver and
on GC in the case of ACN + TEAP solutions. On both
electrodes trichloromethane presents two peaks, while
dichloromethane presents only one peak in the accessible
potential range, almost coincident with the second peak of
CHCl3. At the same time, is also evident the large shift of
peak potentials between Ag and GC (more than 600 mV
for both compounds), so that on GC the reduction of
dichloromethane is too close to the background current to
give a fully resolved peak, as already noted also by [23]. An
even larger difference between silver and GC was recorded
in ACN + TEATFB solutions, as reported in Table 2, which
collects the peak values, Ep (V), for the two substrates
investigated together with the relevant experimental condi-
tions. In particular, Table 2 presents the data obtained on
Ag, AgBi, and AgSn alloys (together with the pure other
metal), and GC, and, for the same electrode material (Ag
and GC) also in different supporting electrolytes.
To better elucidate these observations a screening analysis
of the shift of current and potential peak values with vary-
ing scanning rate (ν = 100–1000 mV/s) was performed and
the results can be summarized as given below (see Fig. 4a
and b).
3.1.1. Peak 1a at ≈ −1.2 V
Ip1a is linear with ν1/2, while Ep1a progressively increases
with ν, so that Ep1a versus log(ν) is curvilinear. This peak,
when present, is present only in the first cycle, disappears

Table 3
Synopsis of preparative electrolyses for the electroreduction of trichloromethane
Cathode
materials
Catholyte
Substrate
Anodic Compartment
Condition
Products
Solvent
ACN
Supporting
electrolyte
Electrode
Solvent
Supporting
electrolyte
1
2
Ag Foil
CHCl3 0.2 M
CHCl3 0.2 M
TEATFB 0.2 M
Ag wire
ACN
TMACl
0.2 M
Galvanostatic: 80 mA; max
Ew = −1.8 V
Oligomers, CH4, CH3Cl
+ CH2Cl2 + mixture of C2
compounds in trace
Oligomers, CH4, CH3Cl
+ CH2Cl2 + mixture of C2
compounds in trace
Ag foil/
graphite rod
ACN
TEABr 0.2 M
Ag wire
ACN
TEABr
0.2 M
Galvanostatic: 100–300 mA
3
Ag foil
Ag foil
Ag foil
Ag foil
Ag foil
Ag foil
Ag foil
Ag foil
CHCl3 0.1 M
CHCl3 0.1 M
CHCl3 0.1 M
CHCl3 0.1 M
CHCl3 0.1 M
CHCl3 0.1 M
CHCl3 0.1 M
CHCl3 0.01 M
ACN/H2O 1:1
ACN/H2O 1:1
ACN/H2O 1:1
ACN/H2O 1:1
ACN/H2O 1:1
ACN/H2O 1:1
ACN/H2O 1:1
ACN/H2O 1:1
TEABr 0.2 M
TEABr 0.2 M
TEABr 0.2 M
TEABr 0.2 M
TEABr 0.2 M
TEABr 0.2 M
TEATFB 0.1 M
TEATFB 0.1M
Ag foil
Ag foil
Pt foil
Pt foil
Pt foil
Pt foil
Pt foil
Pt foil
ACN H2O 1:1
ACN H2O 1:1
H2O
TEABr
0.2 M
TEABr
0.2 M
Galvanostatic: 100–400 mA;
max Ew = −1.2 V, high T
Galvanostatic: 300–400 mA;
max Ew = −2.0 V
CH4, CH3Cl
4
CH4, CH3Cl; CH2Cl2 in
trace
CH4, CH3Cl; final catholyte
pH 2
CH4, CH3Cl; final catholyte
pH 11
CH4, CH3Cl; final catholyte
pH 11
CH4, CH3Cl; final catholyte
pH 11
CH4, CH3Cl; final catholyte
pH 11
5
HClO4 0.5 M
Galvanostatic: 300–400 mA;
max Ew = −1.6 V
6
H2O
KOH 0.5 M
KOH 0.5 M
KOH 0.5 M
KOH 0.5 M
Na2SO4 1 M
Galvanostatic: 400–700 mA;
max Ew = −2.0 V
7
H2O
Galvanostatic: 400 mA
8
H2O
Galvanostatic: 400 mA
Galvanostatic: 240–400 mA
Potentiostatic: 1.26 V
9
H2O
10
H2O
CH4, CH3Cl; final catholyte
pH 5

4040
S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
Fig. 2. Cyclic voltammetry of CHCl3 (solid line) and CH2Cl2 (dashed line) on (a) Ag and (b) GC in 0.1 M TEAP in ACN. Scan rate 200 mV/s.
T ≈ 25 1 ◦C. Reference electrode: aqueous SCE.
in any subsequent cycle, and reappears only after electrode
depolarization.
current is a linear function of CHCl3 concentration (a vertical
solid line denotes its position on Fig. 3). Ip1b is linear with
ν
1/2, while Ep1b depends on ν in a variable way. In fact,
3.1.2. Peak 1b at ≈ −1.5 V
when both peaks 1a and 1b are present, Ep1b varies linearly
with log(ν), and the slope of about 30 mV for a 10-fold
increase in ν points to νeα ≈ 1 and hence to a two-electron
This is the main one, always present at any cycle, and
useful also for quantitative determinations since the peak

S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
4041
step (α ≈ 0.5). Otherwise, in the absence of peak 1a, the
p1b versus log(ν) characteristics shows a much higher slope
(νeα = 0.1–0.4).
3.1.4. Peak 2 at ≈ −2.1 V
E
This is the second main peak of trichloromethane and
is almost coincident with the only visible reduction peak
of dichloromethane. Both current and voltage character-
istics are usually highly linear with ν1/2 and log(ν), re-
spectively, while the dEp2/dlog(ν) generally increases from
dichloromethane (≈59 mV/decade) to trichloromethane
(≈66 mV/decade) and their mixtures (≈88 mV/decade). In
the presence of a poisoned surface, or upon addition of
3.1.3. Peak 1c at ≈ −1.6 V
In the presence of strongly adsorbed species, a more neg-
ative peak is observed in addition to, or in lieu of, peak 1b.
Again, Ip1c is linear with ν1/2, and Ep1c is linear with log(ν),
the slope being usually higher than 30 mV/decade.
1.4E-03
2
1.2E-03
1a
1b
1.0E-03
8.0E-04
6.0E-04
4.0E-04
2.0E-04
0.0E+00
-2.0E-04
0.5
1.0
1.5
2.0
2.5
3.0
(a)
-E / V
1.2E-03
1.0E-03
8.0E-04
6.0E-04
4.0E-04
2.0E-04
2
1b 1c
0.0E+00
-2.0E-04
0.5
1.0
1.5
2.0
2.5
3.0
(b)
-E / V
Fig. 3. Cyclic voltammetry of CHCl3 on Ag in 0.1 M TEAP in ACN for freshly prepared (a) and clean but aged (b) surface, in the absence (a, b) and
in presence of added iodides (c) and cyanides (d). Solid line: first cycle; dashed line: “steady-state” cycle. Scanning rate: 500 mV/s. T ≈ 25 1 ◦C.
Reference electrode: aqueous SCE. The vertical solid line denotes the position of peak 1b.

4042
S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
1.6E-03
1.4E-03
1.2E-03
1.0E-03
8.0E-04
6.0E-04
4.0E-04
2.0E-04
1c
2
0.0E+00
-2.0E-04
0.5
1.0
1.5
2.0
2.5
3.0
(c)
-E / V
3.5E-03
3.0E-03
2.5E-03
2.0E-03
1.5E-03
1.0E-03
5.0E-04
1c
2
0.0E+00
-5.0E-04
0.5
1.0
1.5
2.0
2.5
3.0
(d)
-E / V
Fig. 3. (Continued ).
iodide or cyanide further increase up to 180 mV/decade is
observed.
Summing up these results, the following picture for the
first reduction step of trichloromethane in ACN can be of-
fered: on a fresh electrode surface the first electrochemical
+ chemical step gives rise to a radical or anionic alkyl group
(in dependence on a one- or two-electron reduction) and to
an adsorbed chloride ion (peak 1a). Upon increasing the po-
tential, and hence the reduction rate, the surface becomes
increasingly covered by chloride ions and further reduction
can proceed only through less accessible reacting sites (peak
1b). Since only partial Cl− desorption is achieved at the
negative end point of the potential range, the lower energy
sites are not available in the subsequent cycles and peak
1a disappears, while peak 1b clearly shows the features of
a purely diffusive irreversible peak as the chloride surface
The use of the 50 vol% H2O/ACN solvent mixture pro-
duces a drastic decrease of the accessible potential range,
then limiting the observations to the first reduction step of
CHCl3, which includes in a rather broad peak the finer struc-
ture observed in ACN, as shown in Fig. 5. Anyhow, this
single peak shows again some differences between the first
and the last cycle, and in particular: for first cycle Ep ver-
sus log(ν) is highly curvilinear, and a slight curvature is de-
tected also in the Ip versus ν1/2 characteristics, while the
“steady-state” cycle shows higher linearity in both charac-
teristics, and in particular dEp2/dlog(ν) is close to the theo-
retical 30 mV/decade for a two-electron step with α = 0.5.

S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
4043
0E+00
-1E-03
-2E-03
-3E-03
-4E-03
-5E-03
-6E-03
10
15
20
25
30
35
v
1/2 / mV1/2s-1/2
(a)
-1.2
-1.3
-1.4
-1.5
-1.6
-1.7
-1.8
2.00
2.25
2.50
log(v) / v in mVs -1
2.75
3.00
(b)
Fig. 4. Peak current intensity (a) and peak potentials (b) as functions of the scanning rate for peaks 1a (full squares); 1b (empty diamonds) and 1c
(empty triangles).
coverage approaches unity. On an aged surface, the avail-
ability of low energy reacting sites is reduced, so that peak
1a is lower and appears to merge with peak 1b, thus affect-
ing both the Ep value and the dEp/dlog(ν) slope. If chloride
must compete with stronger adsorbable species, like iodide
and cyanide whose adsorption also extends to a wider po-
tential range, an even higher energy is needed to reach the
electrode surface (peak 1c). Note that in this last case also
a substantial decrease of the current intensity between the
first and the second cycle is observed, in line with the fast
“aging” of the electrode surface.
marked, even if at these negative potentials a partial desorp-
tion of chlorides must be accounted for.
The addition of water, a proton donor, certainly eases the
hydrogen abstraction by the alkyl radical or anion, thus in-
creasing the rate of the chemical reaction and hence produc-
ing a positive shift of the entire I/E characteristics. However,
even in the presence of water there is no net indication that
the involved processes are pure two-electron transfers, to
hint that the latent radical intermediate [R · · · X · · · Ag] can
still, at least partially, evolve toward a one-electron pathway.
Alloying silver with bismuth, known to increase hard-
ness and hence durability, or tin, which increases shelf-life
by drastically decreasing silver tarnishing, reduces, as ex-
The second reduction step then occurs on an already poi-
soned surface, therefore its diffusive characteristics are more

4044
S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
2.5E-03
2.0E-03
1.5E-03
1.0E-03
5.0E-04
0.0E+00
-5.0E-04
0.5
0.7
0.9
1.1
1.3
1.5
1.7
1.9
2.1
-E / V
Fig. 5. Cyclic voltammetry of CHCl3 on Ag in 0.1 M TEAP in ACN/H2O 1:1. Solid line: first cycle; dashed line: “steady-state” cycle. Scanning rate:
100 mV/s. T ≈ 25 1 ◦C. Reference electrode: aqueous SCE.
pected, the electrocatalytic properties of the electrode mate-
rial. However, it should be noted that the reduction potentials
on the alloys are generally comparable to the ones recorded
on a cyanide or iodide poisoned silver surface, hence, as in
the case of the AgSn 64 at.%, the small increase in energy
requirements is accompanied by a substantial reduction of
silver content. With these promising premises, we plan to
further investigate the possible applications of these alloys,
especially when shifting from pure ACN to water the use
of a tetraalkylammonium salt is no more mandatory and the
possible formation of alkyl-Bi or alkyl-Sn compounds can
be totally avoided.
run 10 additional determinations were performed on peri-
odically collected samples and on the final reaction mixture
by CV and potentiometric titrations for quantitative estima-
tion of trichloromethane and of chlorides, respectively, as
depicted in Fig. 6.
On the basis of the obtained results the following remarks
can be formulated.
The electrolyses performed at potentials more negative
than the first reduction step gave always methane as main
product, with chloromethane as possible initial coproduct
and only traces of dichloromethane. In pure ACN (runs 1
and 2), however, the production of oligomers rapidly poi-
sons the electrode, drastically reducing the current intensity
and hence the conversion rate. The periodic cleaning of the
electrode or the polarity inversion technique did not produce
any substantial improvement. Switching to a water-rich sol-
vent mixture (50 vol% ACN/H2O) resulted in a total disap-
pearance of oligomers, giving rise to methane as main end
product, with a current yield inversely proportional to hy-
drogen ion concentration, by considering the larger gas evo-
lution in run 5 with respect to runs 6–9. In the same solvent,
limiting the electrode potential to a more positive value (at
about the peak potential) a 100% current efficiency was ob-
tained for the consumption of the substrate and the parallel
formation of chloride ions with an overall electrode reaction
stoichiometry of 1.5 F/mol with respect to each compound
(see Fig. 6). Interestingly a small but significative amount
of methane and of chloromethane was also detected. This
is not totally unexpected, since other authors [17,20–22] re-
ported the easy production of methane from trichloro- or
dichloromethane in water or acidic solutions, but unlike their
experimental conditions, we can absolutely rule out the con-
comitant production of hydrogen.
3.2. Preparative electrolyses
Table 3 summarizes the preparative electrolysis tests per-
formed, together with the relevant experimental conditions
and main reaction products. The process was driven either
potentiostatically or galvanostatically with limitation of the
potential of the working electrode (Ew) where specified.
A two-compartment, membrane divided cell was used, the
preferred membrane being cationic on account of its lower
electric resistance. As anode we used either a sacrificial
Ag electrode in ACN + TEABr background electrolyte or
a Pt foil in aqueous electrolyte. In any case, no effect of
the anodic reaction over the cathodic process was detected,
if one excludes the increase of hydrogen evolution during
run 5, in which perchloric acid was used as anolyte. The
electrolyses where performed almost exclusively on a mas-
sive silver cathode, with the exception of run 2, in which a
graphite rod was also used. All the electrolyses were per-
formed on trichloromethane as substrate. The electrolysis
products were detected mainly by gas chromatography; for

S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
4045
0.016
0.014
0.012
0.010
0.008
0.006
0.004
0.002
0.000
50
40
30
20
10
0
0
50
100
150
200
250
Q / C
Fig. 6. Current intensity (I (mA), empty squares, left axis), cCHCl (full triangles), cCl− (full circles) and cCHCl + cCl− (full diamonds) as functions of
3
3
the quantity of charge (Q (C)) passed during electrolysis 10. Concentrations (c (mol dm−3)) on right axis.
4. Conclusions
thors wish to thank Ms. Theodora Valkova (Marie Curie
Fellow HPMT-GH-01-00314-07) and Mr. Luca Conti for
their contribution to the experiments.
The electrocatalytic role of silver in the reduction of
volatile organic halides has been investigated and applied to
the reduction of volatile organic chlorides, and in particular
to dichloro- and trichloromethane. The cyclic voltammetry
experiments evidentiate that within each main reduction
step a finer distinction can be made in dependence on the
Ag accessible surface states. These effects are stronger in
pure acetonitrile than in water/acetonitrile mixtures. Some
silver alloys were also investigated to improve life-time
and shelf-life of the electrode material; all the tested al-
loys resulted less electrocatalytic than silver, but some
AgBi and AgSn alloys can be considered promising alter-
natives. The reduction products of preparative electrolyses
point to a mixed 1e−/2e− pathway at low electrode po-
tentials, even in the presence of water, while at higher
current densities (that is at higher electrode potentials) the
main end product is methane, which, in any case, is al-
ways produced in any condition. Further investigations are
presently in progress to fully understand this result, which
is highly desirable from the point of view of environmental
applications.
References
[1] M. Benedetto, G. Miglierini, P.R. Mussini, F. Pelizzoni, S. Rondinini,
G. Sello, Carbohydr. Lett. 1 (1995) 321.
[2] S. Rondinini, P.R. Mussini, G. Sello, E. Vismara, J. Electrochem.
Soc. 145 (1998) 1108.
[3] M. Guerrini, P.R. Mussini, S. Rondinini, G. Torri, E. Vismara, Chem.
Commun. 15 (1998) 1575.
[4] S. Rondinini, P.R. Mussini, G. Cantù, G. Sello, Phys. Chem. Chem.
Phys. 1 (1999) 2989.
[5] S. Rondinini, P.R. Mussini, F. Crippa, M. Petrone, G. Sello, Collect.
Czech. Chem. Commun. 65 (2000) 881.
[6] S. Rondinini, P.R. Mussini, F. Crippa, G. Sello, Electrochem. Com-
mun. 2 (2000) 491.
[7] G. Fiori, P.R. Mussini, S. Rondinini, A. Vertova, Ann. Chim. (Rome)
91 (2001) 151.
[8] S. Rondinini, P.R. Mussini, M. Specchia, A. Vertova, J. Electrochem.
Soc. 148 (2001) D102.
[9] S. Rondinini, P.R. Mussini, P. Muttini, G. Sello, Electrochim. Acta
46 (2001) 3245.
[10] G. Fiori, P.R. Mussini, S. Rondinini, A. Vertova, Ann. Chim. (Rome)
92 (2002) 963.
[11] S. Ardizzone, G. Cappelletti, P.R. Mussini, S. Rondinini, L.M.
Doubova, J. Electroanal. Chem. 532 (2002) 285.
[12] S. Ardizzone, G. Cappelletti, P.R. Mussini, S. Rondinini, L.M.
Doubova, Russ. J. Electrochem. (translation of Elektrokhimija) 39
(2003) 170.
[13] P.R. Mussini, S. Ardizzone, G. Cappelletti, M. Longhi, S. Rondinini,
L.M. Doubova, J. Electroanal. Chem. 552 (2003) 213.
[14] S. Ardizzone, G. Cappelletti, L.M. Doubova, P.R. Mussini, S.M.
Passeri, S. Rondinini, Electrochim. Acta 48 (2003) 3789.
Acknowledgements
The financial support of MIUR–Cofin 2002 and of EC
Program Improving Human Research Potential and the
Socio-economic Knowledge Base—IHP contract number
HPMT-CT-2001-00314 is gratefully acknowledged. Au-

4046
S. Rondinini, A. Vertova / Electrochimica Acta 49 (2004) 4035–4046
[15] M. Specchia, Thesis, The University of Milan, 2000.
[16] S. Wawzonek, R.C. Duty, J. Electrochem. Soc. 108 (1961) 1135.
[17] G. Horanyi, K. Torkos, J. Electroanal. Chem. 140 (1982) 329.
[18] A. Kotsinaris, G. Kyriakou, C. Lambrou, J. Appl. Electrochem. 28
(1998) 613.
[20] T. Fotiadis, G. Kyriacou, C. Lambrou, S. Hadjispyrou, J. Electroanal.
Chem. 480 (2000) 249.
[21] N. Sonoyama, K. Hara, T. Sakata, Chem. Lett.
131.
2 (1997)
[22] N. Sonoyama, K. Ezaki, T. Sakata, Adv. Environ. Res. 6 (2001) 1.
[23] C. Costentin, M. Robert, J.-M. Savéant, J. Am. Chem. Soc. 1250
(2003) 10729.
[19] N. Georgolios, G. Kyriacou, G. Ritzoulis, J. Appl. Electrochem. 31
(2001) 207.