Electrocatalysis on silver and silver alloys for dichloromethane and trichloromethane dehalogenation

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.

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, ꢀE_Ag–GC= E_Ag− E_GC, 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 E_Agversus E_GCvalues 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).

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

E_GC/ V

Fig. 1. Peak potential values measured on silver, E_Ag(V), against the parallel values measured on glassy carbon, E_GC(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

ν_e(F/mol)

Reference

3-Br-phenol

4-Br-phenol

2,6-Di-Br-phenol

2,4-Di-Br-phenol

2,4,6-Tri-Br-phenol

1-Br-adamantane

2-Br-adamantane

TEAP/ACN

Hydrodehalogenation

2

[8,10]

[6,8,10]

[10]

[8,10]

[10]

[6,8,10]

[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

Hydrodehalogenation

2

Hydrodehalogenation/dimerization

␣,␣^ꢀ:␤␤^ꢀ:␣␤ = 1:1:2 dimers

1.9

1.97

1.7

1.5

1.7

≈1

[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

␣,␣^ꢀ:␤␤^ꢀ:␣␤ = 1:1:2 dimers

≈1

[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 CH₄and 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 (HClO₄0.5 M, Na₂SO₄1 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 CHCl₃and 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⁻³KAg(CN)₂bath, at 0.5 mA cm⁻²depo-

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 5^ꢀ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

E_p1(V)

E_p2(V)

Ag

CHCl₃

0.1 M TEAP in ACN

−1.50

−2.12

−2.23

−2.05

−2.1^a

n.d.

0.1 M TEAI in ACN

−1.52

0.1 M TEAP + 10⁻³M KCN in ACN

−1.61

0.1 M TEATFB in ACN

−1.45^a

0.1 M TEATFB in ACN/H₂O 1/1

−1.23 → −1.33^a;

−1.14 → −1.24^a

−1.60^a

AgBi 3 at.%

AgBi 30 at.%

Bi

AgSn 64 at.%

Sn

GC

Ag

CHCl₃

CH₂Cl₂

0.1 M TEATFB in ACN

0.1 M TEAP in ACN

−2.24^a

−2.5

−1.64

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 NaClO₄in ACN

0.1 M TEATFB in ACN

0.1 M TEAP in ACN

−1.7^a

−2.3^a

−1.87^a

−1.65^a

−2.3^a

−2^a

n.d.

−2.14

−2.67

−2.70

−2.63

−2.11

−1.93

−2.11^a

−2.02^a

−2.20

−2.22

−2.5

Ag

AgBi 3%

AgSn 64%

Sn

GC

−2.23

−2.5

−2.75

−2.66

E_p1(V) denotes the ﬁrst reduction peak of CHCl₃and E_p2(V) denotes either the second reduction peak of CHCl₃or 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, E_p(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

I_p1ais linear with ν^1/2, while E_p1aprogressively increases

with ν, so that E_p1aversus 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

CHCl₃0.2 M

TEATFB 0.2 M

Ag wire

ACN

TMACl

0.2 M

Galvanostatic: 80 mA; max

E_w= −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

CHCl₃0.1 M

CHCl₃0.01 M

ACN/H₂O 1:1

TEABr 0.2 M

TEATFB 0.1 M

TEATFB 0.1M

Ag foil

Pt foil

ACN H₂O 1:1

H₂O

TEABr

0.2 M

TEABr

0.2 M

Galvanostatic: 100–400 mA;

max E_w= −1.2 V, high T

Galvanostatic: 300–400 mA;

max E_w= −2.0 V

CH₄, CH₃Cl

4

CH₄, CH₃Cl; CH₂Cl₂in

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

HClO₄0.5 M

Galvanostatic: 300–400 mA;

max E_w= −1.6 V

6

H₂O

KOH 0.5 M

Na₂SO₄1 M

Galvanostatic: 400–700 mA;

max E_w= −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 CHCl₃concentration (a vertical

solid line denotes its position on Fig. 3). I_p1bis linear with

ν

^1/2, while E_p1bdepends 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, E_p1bvaries linearly

with log(ν), and the slope of about 30 mV for a 10-fold

increase in ν points to ν_eα ≈ 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

_p1bversus log(ν) characteristics shows a much higher slope

(ν_eα = 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 ν^1/2and log(ν), re-

spectively, while the dE_p2/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, I_p1cis linear with ν^1/2, and E_p1cis 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 CHCl₃on 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 E_pver-

sus log(ν) is highly curvilinear, and a slight curvature is de-

tected also in the I_pversus ν^1/2characteristics, while the

“steady-state” cycle shows higher linearity in both charac-

teristics, and in particular dE_p2/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/ mV^1/2s^-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 E_pvalue and the dE_p/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 CHCl₃on 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 (E_w) 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

50

100

150

200

250

Q / C

Fig. 6. Current intensity (I (mA), empty squares, left axis), c_CHCl(full triangles), c_Cl− (full circles) and c_CHCl+ c_Cl− (full diamonds) as functions of

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.

Article Doi

Electrocatalysis on silver and silver alloys for dichloromethane and trichloromethane dehalogenation

DOI: 10.1016/j.electacta.2003.12.061

Source and publish data:

Authors:

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

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

Products guided by the article

R&D Labs maybe for 7440-22-4

Relevant to this article

Hot Product