Вы находитесь на странице: 1из 8

Journal of

Electroanalytical
Chemistry
Journal of Electroanalytical Chemistry 600 (2007) 8794
www.elsevier.com/locate/jelechem

The nitrate reduction process: A way for increasing interfacial pH


Myriam Nobial a, Olivier Devos a, Oscar Rosa Mattos b, Bernard Tribollet
a

a,*

UPR 15 du CNRS Laboratoire Interfaces et Syste`mes Electrochimiques, Universite Pierre et Marie Curie, Case 133, 4 place Jussieu,
75252 Paris Cedex 05, France
b
Laboratorio de Corrosao Manuel de Castro, PEMM/COPPE Universidade Federal do Rio de Janeiro, Brazil
Received 20 December 2005; received in revised form 28 February 2006; accepted 7 March 2006
Available online 24 April 2006

Abstract
This paper is devoted to the electrochemical study of the nitrate reduction in aqueous media. The reduction process allows the production of hydroxyl ions at the electrode surface, increasing the local pH. This process provides a useful tool for the precipitation of
metallic oxide/hydroxide such as ZnO. When the interfacial pH was high enough, the chemical equilibrium was shifted towards the crystallization. The nitrate reduction was carried out in the presence of dissolved oxygen. The investigated potential range corresponded to
the mixed reduction of oxygen and nitrate where both reactions produced OH and increased the interfacial pH. In this case, it was
shown that the pH could reach a value higher than 12 in 1 M KNO3 instead of 10.4 in the presence of only oxygen. Thus, nitrate reduction was useful for the precipitation occurring at pH higher than 10.4. The electrochemical mechanism was studied by electrochemical
impedance spectroscopy in a potential range corresponding to the rst step of the oxygen reduction. The stationary currentvoltage
curve did not highlight any current due to the nitrate reduction. The impedance diagrams revealed two capacitive loops characterizing
the reduction process of oxygen, namely the charge transfer process at high frequencies and the diusion process at low frequencies
according to the expectations. In the lowest frequency domain, an inductive loop was observed with an amplitude depending on the
potential. It was shown that this response was due to the presence of adsorbed species which blocked a part of the active surface for
the oxygen reduction. These adsorbed species came from nitrate reduction even if the stationary current of this reaction was negligible
in the corresponding potential range. The decrease of the double layer capacity conrmed this approach.
 2006 Elsevier B.V. All rights reserved.
Keywords: Nitrate reduction; Adsorption; Impedance; Interfacial pH

1. Introduction
Precipitation of dierent materials occurs when the pH
is high enough. The shift to the basic pH can be obtained
by imposing a cathodic reaction on the electrode. These
materials include individual oxides (ZrO2, PbO, ZnO [1
4]), scale deposits as CaCO3, [5,6] as well as compounds
including BaTiO3 [7], superconductors [8] and biomaterials
[9,10]. The deposition is achieved from the metal ions by
electro-generating hydroxyl ions to form metal oxide/
hydroxide lms on a cathodic substrate. Dierent species
for producing hydroxyl anions were proposed in the literature like oxygen [11], water, hydrogen peroxide [12] or
*

Corresponding author. Tel.: +33 1 44274170; fax: +33 1 44274074.


E-mail address: bt@ccr.jussieu.fr (B. Tribollet).

0022-0728/$ - see front matter  2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jelechem.2006.03.003

nitrate ions [1316]. Reduction of these species allows a


controlled pH value to be reached at the electrode/solution
interface. As an example, the pH value corresponding to
the precipitation of calcium carbonate occurs at a pH value
close to 8 depending on the ionic strength of the carbonatefree water [17]. Thus, the reduction of the dissolved oxygen
which produced OH and a maximum pH of 10.4 was sufcient to form the scale deposition [18]. In the case of other
metal oxide/hydroxide compounds, such as ZnO and ZrO2,
the pH must be higher [1]. As a consequence, additional
reduction processes producing hydroxyl ions are necessary,
such as water or other soluble species reduction. Water
reduction produces OH and dihydrogen gas simultaneously. However, the resulting H2 bubbling on a substrate
is not appropriate for precipitate formation. Another possibility is the oxygen saturation of the solution which

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794

2. Experimental
2.1. Current densitypotential and electrochemical
impedance spectroscopy
The experimental set-up was a classical three electrode
cell with a rotating disk gold electrode (with a 5 mm diameter) as working electrode. A large platinum grid was used
as counter electrode. The potential of the working electrode was monitored with respect to a saturated calomel
electrode (SCE) by means of a potentiostat (Sotelem).
The electrolyte solutions were prepared by dissolving
0.1 M KCl (Prolabo), 0.1 M or 1 M KNO3 (Prolabo) in
distilled water. The current densitypotential curves were
plotted with a sweeping rate of 0.5 mV s1 corresponding
to the steady-state. Electrochemical impedance spectroscopy measurements were obtained by means of a PC/4
GAMRY INC. controlled by a software version 4.2
GAMRY INC. The impedance measurements were performed between 65 kHz and 10 mHz with 7 points/decade.
2.2. pHpotential measurements

was xed and a SCE was used as reference electrode.


Before each experiment, the working electrode was cleaned
using an ultrasonic bath during 5 min containing ethanol
and rinsed with distilled water. Two electrolytes were used
for the experiments: 0.1 M KCl and 0.1 M KNO3 solutions. The potential of the grid electrode was monitored
versus SCE with a potentiostat (Sotelem). The current
potential curves were plotted with a sweep rate of 2 mV s1
corresponding to steady-state for pH measurements, simultaneously measured with pH-meter (LPH230T Tacussel).
3. Results and discussion
3.1. Currentpotential curves
Fig. 1 represents current densitypotential curves at
steady-state for various nitrate ion concentrations from
0 up to 1 M KNO3. In 0.1 M KCl solution (without
KNO3) as supporting electrolyte, the measured current corresponds to the dissolved oxygen reduction process in a
cathodic potential range up to 1.2 V/SCE. Fig. 2, which
is a magnication of the rectangular dot region of Fig. 1,
highlights the two steps of the oxygen reduction. It shows
two limiting current densities that appear at E = 0.5 V/
SCE and E = 0.90 V/SCE according to the diusion control of the two steps written in the following equations:
k1

O2 2e 2H2 O $ H2 O2 2OH ;


 k2

H2 O2 2e $ 2OH :

The water reduction which appears at a potential lower


than 1.25 V/SCE (Fig. 1) corresponds to the reaction:
2H2 O 2e $ H2 2OH :

For 0.1 M KNO3 and a potential higher than 0.5 V/SCE,


no inuence of nitrate ions was detected on the current
(Fig. 2). For a potential lower than 0.5 V/SCE a decrease
of the limiting current density was observed and the second
step of the oxygen reduction was not clearly dened. In a
-10
-8
-2

allowed the interfacial pH to reach a maximum value of 12


[11]. Nevertheless, this technique was not useful due to the
continuous O2 bubbling in the bulk. Other papers related
the addition of nitrate ions as a good way to increase the
local pH and control the deposit conditions [1316]. Dierent electrochemical research works were devoted to the
nitrate ions reduction in aqueous solutions in the last two
decade because (i) the electrochemistry of nitrate is important for cleaning and monitoring nitrogen oxide in air and
water [19] and (ii) it leads to the synthesis of useful chemicals like hydroxylamine and ammonia [20]. Although
many papers in the literature are available for processes
in acidic media [19,2125], only few investigated the reduction reaction in alkaline media [20,2629] and no paper
really deals with nitrate ions reduction in neutral media
[30,31]. The role of nitrite ions and the nature of the reduction products remain unsolved, since nitrate ion reduction
is very complex, involving dierent mechanisms with a
large number of stable intermediates. Moreover, Da Cunha
et al. [24] showed that in acidic media, nitrate ions
adsorbed on gold electrodes and the reduction of nitrate
ions produced nitrite ions which were co-adsorbed.
This paper is devoted to the electrochemical study of the
nitrate ions reduction on a gold electrode for the improvement of deposits formed by shifting the pH.

J / mA.cm

88

-6
-4
-2

The working electrode was a gold grid (Goodfellow,


wire diameter: 0.06 mm). A 0.5 mm diameter gold wire
was welded on the periphery of the gold grid for electrical
connection. The grid electrode was placed in contact with
the at end of a combined pH electrode (Metler-Toledo)
and maintained on the lateral wall of the pH electrode with
silicon. The experimental device was described in detail in
Ref. [18]. A large platinum grid used as counter electrode

0
0.0

-0.5

-1.0
-1.5
E / V/SCE

-2.0

Fig. 1. Current densityvoltage curves at steady-state on gold rotating


disc electrode (surface area 0.2 cm2). (h) 0.1 M KCl, (s) 0.1 M KNO3 and
(n) 1 M KNO3, rotation speed U = 1000 rpm. Sweep rate = 0.5 mV s1.

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794

89

-0.3
20

-1

(b)

-1

i / mA

J / mA.cm

-2

(c)
-0.2

(a)

-1

16

-1/2

4.59 mA .s

12

-0.1
(d)

0.0
0.00

-0.25

-0.50

-0.75

-1.00

-1.25

-1.50

E / V/SCE
Fig. 2. Currentvoltage curves at steady-state (zoom of dot square region
in Fig. 1) corresponding to the two steps of dissolved oxygen reduction.
Same experimental conditions and legend as in Fig. 1(a)(d) correspond to
the potential of impedance diagrams presented in Fig. 5. Note that the EIS
was performed in 0.1 M KNO3 (s).

rst approach, this dierence could be due to two dierent


processes: nitrate adsorption blocking a weak percentage
of active surface for oxygen reduction and/or kinetic modication of the second step of oxygen reduction (reaction
2). For a potential lower than 1.15 V/SCE, the current
value is higher than the measured current in the presence
of only oxygen which is attributed to the additional reduction current of nitrate ions.
For 1 M KNO3 and for a potential higher than 0.5 V/
SCE the current was lower than that in the solution without KNO3 and containing 0.1 M KNO3 (Fig. 2). This could
be due to the two processes previously described. For this
concentration, the second step of the oxygen reduction
was also not well dened. For a potential lower than
1 V/SCE the increase of current was due to the nitrate
reduction. Thus, in a potential range from 1 V/SCE down
to 1.5 V/SCE the current density increased according to
the nitrate concentration (Fig. 1). This result was attributed
to the reduction reaction of nitrate ions as in Eq. (4)
whereas the signicant increase of current due to the water
reduction was observed at potentials lower than 1.25 V/
SCE without NO
3 ions.



NO
3 H2 O 2e $ NO2 2OH ;

Fig. 3 represents the inverse of the reduction current measured at a stationary potential of 1.1 V/SCE versus inverse of the square root of the rotation speed U. The
solution was deaerated by bubbling N2 gas to avoid the
contribution of the oxygen reduction current at this potential. The curve could be tted by a dot line which did not
cross the origin. According to the KouteckyLevich equations, the faradaic current can be written as in:
1
1
1
;
I NO3
Ik ID

-1/2

4
3

-1/2

.10 / s

Fig. 3. KouteckyLevich curve on gold rotating disc electrode. 0.1 M


KNO3 in deaerated solution and E = 1.1 V/SCE.

where I NO3 was the nitrate reduction current, Ik the kinetic


contribution and ID the diusion contribution expressed as
follows:
2=3

I D 0:62nFSDNO t1=6 C NO3 U1=2 KU1=2 ;


3

where n is the electron number (n = 2), F the Faraday constant (96500 C), S the surface area (0.2 cm2), DNO3 the diffusion coecient of nitrate ions, t the kinematics viscosity
(0.994 cm2 s1), C NO3 the nitrate ions concentration
(0.1 M) and K1 the slope of the dot line plotted in Fig. 3.
The diusion coecient of nitrate ions was deduced
from the experimental value of the slope as DNO3 1:2
105 cm2 s1 which is in agreement with [32]. Thus, nitrate
ions reduction was a process partially controlled by the
mass transport but the plateau corresponding to the limiting diusion process of NO
3 reduction was hindered by the
water reduction which occurred in the same potential
domain.
3.2. pHpotential curves
Fig. 4 shows the currentpotential and pHpotential
curves in solutions containing dierent nitrate ion concentrations on a gold grid and without stirring, contrarily to
the previous results carried out with a controlled hydrodynamic regime by a rotating disc electrode. The reproducibility of the currentpotential curves was poor when the
mass transport limitation of the oxygen reduction current
took place. Indeed, the oxygen reduction was measured
on a gold grid where natural convection was predominant.
Nevertheless, for lower potential than 1 V/SCE, where
the reduction current of nitrate ions started to be observed,
the current curves had the same shape as that observed in
Fig. 2.
The variations of the interfacial pH simultaneously measured with the reduction current are also plotted in Fig. 4.
According to a previous work carried out without nitrate
ions [18], the interfacial pH increased from 6.1, which rep-

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794


-6
-5

12
pH
10

-3

pH

i / mA

-4

-2

-Im Part [Z(f)] / k.cm

90

0.8
0.6
0.4

(a)

E=-0.25V/SCE

7.2Hz

100kHz

0.2

0.1Hz

0.0
0.0

0.5

1.0
1.5
2.0
2
Re Part [Z(f)] / k.cm

i
-1
6
-0.5

-1.0

-Im Part [Z(f)] / k.cm

0.0

3.0

3
E=-0.35V/SCE

2.5

-1.5

E / V/SCE

Fig. 4. Currentpotential curves in () 0.1 M KCl, (m) 0.1 M KNO3, (r)


1 M KNO3 and interfacial pHpotential curves in () 0.1 M KCl, (h)
0.1 M KNO3, (s) 1 M KNO3.

resented the bulk pH, up to 10.4 at the oxygen reduction


plateau. For the three tested solutions and at 0.9 V/
SCE, the measured interfacial pH was almost identical
due to the negligible contribution of the nitrate reduction
current. An increase of the interfacial pH started at the
same potential value (0.9 V/SCE) for the two nitrate
solutions. From this potential reaction (4) must be taken
into account with a higher increase of pH observed for
higher NO
3 concentration according to the current evolution in Figs. 1 and 4. Note that the pH strongly increased
without nitrate due to the water reduction process from
1.25 V/SCE which was in agreement with the current
measurement in Fig. 2 and Eq. (3). In the presence of
1 M NO
3 , an interfacial pH between 10.4 and 12 was measured in a potential range between 0.9 and 1.2 V/SCE,
not reachable in presence of only dissolved oxygen without
hydrogen gas.

2
1 100kHz

37mHz

27Hz

3mHz

-1

-Im Part [Z(f)] / k.cm

(b)

2
3
4
5
2
Re Part [Z(f)] / k.cm

(c)

E=-0.40V/SCE

4
2 100kHz
17.3Hz

37mHz

3mHz

0
0

4
6
8
10
12
2
Re Part [Z(f)] / k.cm

14

30
E=-0.60V/SCE

3.3. EIS investigation


Fig. 5 represents the impedance diagrams in the Nyquist
plan versus potential for a 0.1 M KNO3 solution using a
disc electrode rotating at 1000 rpm in order to control
the hydrodynamic ux in the vicinity of the electrode.
The measurement was carried out in a potential range corresponding to the rst step of the dissolved oxygen reduction where the contribution of the stationary reduction
current of NO
3 was not clearly observed yet for a rotating
disc electrode (Fig. 2) and a non convective electrode such
as that used for local pH measurement as well (Fig. 4). The
corresponding stationary current density and potential are
given in Fig. 2. In the impedance diagram of Fig. 5a, two
capacitive loops are clearly observed. Other impedance diagrams were plotted at more cathodic potentials which inuence the two capacitive responses. The high frequency (HF)
response is attributed to the charge transfer process corresponding to the oxygen reduction. As the reduction is mass
transport controlled, the second capacitive loop corresponds to the convective diusion process of oxygen. In

-Im Part [Z(f)] / k.cm

(d)

20

10

100kHz
3mHz

50mHz

-10
-10

10
20
30
2
Re Part [Z(f)] / k.cm

40

50

Fig. 5. Impedance diagrams in Nyquist plane in 0.1 M KNO3. Rotation


speed U = 1000 rpm and potential (a) E = 0.25 V/SCE, (b) E =
0.35 V/SCE, (c) E = 0.40 V/SCE and (d) E = 0.60 V/SCE. The
corresponding applied potential is represented in Fig. 2.

Fig. 5, the amplitude of the HF loop becomes negligible


over the amplitude of the low frequency (LF) capacitive
loop when the potential varies until the rst plateau. Then,
an inductive loop is observed, not attributed to the oxygen
reduction mechanism, when the cathodic potential reaches

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794

Z CPE

1
a;
Qjx

(-1)

1E-4

1E-5

Q (F.s

0.35 V/SCE (Fig. 5b). The origin of the loop will be


discussed afterwards. The HF measurements are not in
agreement with pure capacity behaviour. A phase value
lower than 90 was measured allowing a constant phase
element (CPE) to be set in the electrochemical kinetic
model. Fig. 6 shows the imaginary part of the impedance
Zim as a function of frequency in logarithmic coordinates
[33]. A straight line was plotted with a slope value equal
to 0.91. In the case of a pure capacity, a 1 value as slope
should be obtained. The impedance expression of a CPE is:

-6

3.2 10

where a represents the slope of the straight line plotted in


Fig. 6, Q is the CPE parameter corresponding to a capacitance when a = 1 and x is the pulsation (rad s1). With the
a value, Q was calculated from:
 
sin a p2
Q
8
a.
Z im 2pf
The Q value was plotted in Fig. 7 versus the frequency f
in logarithmic coordinates. A constant value of 3.2
106 lF s(a1) was obtained for a frequency higher than
1 kHz.
From E = 0.35 V/SCE, the HF part of the LF capacitive loop tended to be a straight line with a slope slightly
higher than 45. It is well known that the oxygen reduction
is mass transport controlled. As a result, the LF response
was readily attributed to the diusion process of oxygen
from the bulk towards the electrode surface. Then the faradaic impedance corresponding to the oxygen reduction
was a classical charge transfer resistance Rct in series with
a diusion impedance ZD. The CPE was in parallel with
this faradaic impedance.
According to Brug et al. [34] the mean value of the double layer capacity Cdl could be calculated from the CPE
parameters according to:
1a

1
Q C adl R1
s Rct 

91

1E-6
10

100

1000

10000

f / Hz

Fig. 7. Loglog representation of Q versus frequency. Q was calculated by


means of Eq. (8). The gure was plotted using impedance diagram in
Fig. 5a.

where Rs is the solution resistance. As shown in Fig. 5, the


Rct value is much higher than Rs. Thus the inverse of Rct is
negligible over the inverse of Rs. As a consequence, Eq. (9)
can be simplied as follows:
C dl Rs1a=a  Q1=a ;

10

Cdl was calculated for the dierent investigated potentials


(Fig. 8) by using the Q and a values obtained with the
graphical method described in Figs. 6 and 7 and using
Eq. (10). Note that Rs is the real part of the impedance
in the highest frequency range. For a low cathodic potential, the capacity had a reasonable value for a double layer.
When the potential shifts toward the reduction, Cdl rapidly
decreases to reach a low limiting value of 5.6 lF cm2
(Fig. 8). As Cdl is directly proportional to the active surface, its decrease could be due to the presence of adsorbed
species [36], such as NO
2 , which could block a part of the
active surface for the oxygen reduction. This assumption
was conrmed by the inductive response observed in EIS
(Fig. 5). Fig. 9a shows the experimental and tted impedance diagrams at E = 0.25 V/SCE corresponding to the

1
14

slope = - 0.91

-2

Cdl / F.cm

-Im Part [Z(f)]

12

0.1
10
8

0.01
6

10

100

1000

10000

f / Hz

Fig. 6. Loglog representation of the imaginary part of the impedance of


Fig. 5a versus frequency. The slope of the straight line represents the a
coecient dened in Eq. (7).

4
0.2

0.3

0.4
0.5
-E / V/ECS

0.6

0.7

Fig. 8. Representation of Cdl versus potential. Cdl was calculated by


means of Eq. (10).

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794

0.8 E = -0.25V/SCE

(a)

0.4

7.2 Hz

0.0
0.0

-Im Part [Z()] / k.cm

-Im Part [Z(f)] / k.cm

92

0.5

1.0
1.5
2.0
2
Re Part [Z(f)] / k.cm

2.5

3.0

3
(b)

E=-0.35 V/SCE

2
1
27Hz

0
-1
-1

2
3
4
5
6
2
Re Part [Z()] / k.cm

Fig. 9. Experimental and tted impedance diagrams in Nyquist plan in


0.1 M KNO3. U = 1000 rpm and potential (a) E = 0.25 V/SCE (b)
E = 0.35 V/SCE. The position of the applied potential was represented
in Fig. 2.

two capacitive loops without contribution of the inductive


response. The tted curve was calculated by using the
equivalent circuit described previously with the diusion
impedance written as:
p
thd jx=D
Z D RD p ;
11
jxD
where d is the thickness of the diusion layer (lm) and RD
(X cm2) the diusion resistance.
At E = 0.25 V/SCE, the tted impedance diagram is in
agreement with the experimental one in the whole frequency range (Fig. 9a). The tted parameters associated
with the capacitive loops are presented in Table 1 from
0.25 up to 0.6 V/SCE corresponding to the limiting current of the reaction (1). In the whole potential range, Rs is
sensibly constant. The Q and a values allowed the Cdl evolution with potential to be calculated as discussed above
(Fig. 8). According to the expectations, the tted Rct value
which represents the amplitude of the HF loop, decreased
with cathodic potential. For potentials more cathodic than

Table 1
Fitted values corresponding to the impedance of the capacitive loops of
the oxygen reduction
E (V/SCE)
Rs (X cm2)
Q (106 F s(1a))
a
RctO2 (kX cm2)
RD (103 kX cm2)
(d/D1/2) (101 s)
Cdl (lF cm2)

0.25
86
3.2
0.91
1.2
2.5
5.6
12.8

0.3
73
6.0
0.85
1.1
4
5.5
8.0

0.35
82
7.5
0.82
1.0
10
5.6
7.0

0.4
81
7.5
0.80
0.1
16
5.9
6.4

0.45
95
9.3
0.78

0.6
75
20.0
0.70

21
5.4
6.0

49
5.5
5.6

0.4 V/SCE, when the potential was shifted toward the


diusion plateau, the tted curve did not allow Rct to be
estimated with accuracy because it was negligible over the
diusion resistance. The LF domain of the impedance
allowed the time constant d/D1/2 of the diusion impedance
to be calculated. Table 1 shows that this value did not
depend on a variation of potential. With DO2 2
109 m2 s1 [27], the thickness of the diusion layer was calculated as d = 22 lm. Thus, it was shown that a cathodic
potential lower than 0.45 V/SCE identied an electrochemical mechanism characterised by an equivalent circuit
with a CPE in parallel with ZD. The resistance Rct in series
with ZD was negligible in the model of the quasi-reversible
electrochemical system when the potential was close to the
diusion plateau. Diard et al. observed the same phenomenon using K3Fe(CN)6/K4Fe(CN)6 as reversible redox
system [35].
For low cathodic potential (E = 0.25 V/SCE), no
induction response was observed by EIS leading the
reduction process of oxygen to be characterised. Then,
from a potential E = 0.35 V/SCE (Fig. 5b), the presence of an inductive loop indicated the existence of an
adsorption process on the surface of the substrate. The
amplitude of this LF response increased with the cathodic potential (Figs. 5bd) even if no steady current variation was recorded (Fig. 2). This surprising result, not
usually observed in presence of only dissolved oxygen,
was attributed to the nitrate reduction. Nevertheless,
the eect of the NO
3 contribution on the stationary current started to be observable from a much lower potential of 1.15 V/SCE for 0.1 M KNO3 (Fig. 2). De
Vooys et al. [36] have presented nitrate ions reduction
mechanism as in reactions (12)(15). This mechanism
presents two faradaic reactions (Eqs. (13) and (15)) where

three species NO
3 , NO2 and NO were adsorbed. De
Groot and Koper [23] proposed a similar mechanism
except that Eq. (14) was not considered
h1


NO
3aq $ NO3ads ;

12
h2

h1




NO
3 ads H2 O 2e $ NO2ads 2OH ;
h2


NO
2 ads $ NO2aq ;
h2

13
14

h3



NO
2 ads H2 O e $ NOads 2OH :

15

The expression of dhdt1 , dhdt2 and dhdt3 can be written according


to the kinetics equations corresponding to reactions
(12) and (15). h1, h2, h3 are the coverage rate of NO
3,
NO
2 and NO with 0 6 h 6 1. Thereafter Eq. (16) will be
considered to simplify the theoretical approach
h h1 h2 h3 .

16

In the proposed mechanism, reactions (13) and (15) are


faradaic equations. As shown in Fig. 2 no faradaic eect
of nitrate was observed on stationary current density
potential curve up to 0.5 V/SCE in a solution containing

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794

0.1 M KNO3. In this potential domain Eqs. (13) and (15)


are at steady state. Thus the current density due to these
equations is very small and the corresponding impedances
were considered as negligible in the whole frequency
range. So that, for a potential higher than 0.5 V/SCE,
the current density (Fig. 2) was only due to the oxygen
reduction on a surface S(1  h). The expression of the
current of oxygen reduction was written in Eq. (17). Then,
Eq. (16) led Eq. (18) to be expressed, where dh
was dependt
dant on the coverage rate and potential.
i k 1 1  hC O2 0 exp bE;
dh
f h; E;
dt

17
18

C O2 0 is the oxygen concentration at the electrode surface,


k1 is the kinetic constant of Eq. (1) and b = b RT/nF with b
the transfer coecient, T the temperature and R the gas
constant. The dynamical study by EIS allowed Eqs. (17)
and (18) to be written as:
eO
e k 1 1  h exp bE C
~i bk 1 1  hC O2 0 exp bE E
2
~
 k 1 C O2 0 exp bEh;
19




of
of
e
h  E.
20
jx~h  ~
oh E
oE h
Eq. (20) was expressed as follows:

of 
~
h
oE h
 .

e jx  of 
E

21

Then Eq. (19) was divided by and could be written as the


following expression:
eO
e
h~
E
C
k 1 1  h exp bE 2  k 1 C O2 0 exp bE
~i
~i
~i
22

with
Rct bk 1 1  hC O2 0 exp bE.

23

Then Eq. (22) becomes Eq. (24):


Z Rct Z D Rct k 1 C O2 0 exp bE

~
e
h E
e ~i
E

24

with
Z D Rct k 1 1  h exp bE

eO
C
2
~i

25

and
Z D Rct k 1 1  h exp bE

eO
C
2
 :
oe
C O2 
nFD oy 

26

with


e O 
o
C
2
~i nFD
.
oy 
0

Finally, Z could be written by using Eqs. (21) and (24):


Z

Rct Z D
;
Rct of j
1 jxofoE h
oh jE

28

of
j
oE h

and of
j were real parameters which dened the third
oh E
loop in the lowest frequency domain of the EIS diagram
in Fig. 5. To obtain an inductive loop with Eq. (28), it is
of
necessary that oE
jh and of
j have the same sign.
oh E
Fig. 9b shows the experimental and tted diagrams at
E = 0.35 V/SCE in the whole frequency domain. The
inductive loop was tted by using Eq. (28). At this potential, the tted curve shows a good agreement with the
experimental diagram. Note that the amplitude of the tted
inductive loop is lower than that of the experimental one.
As previously shown, various adsorption processes appear
in the nitrate reduction mechanism. Then the experimental
inductive loop was probably due to more than one adsorption process. To have a better t of the experimental inductive response, h1, h2 and h3 parameters should be taken into
account in the electrochemical kinetics model. However,
this rst approach showed that an inductive response could
be observed in presence of adsorbed species blocking the
active surface without any faradaic contribution of these
latter.
4. Conclusion

oh E

1 R1
ct

93

27

Electrochemical investigation of NO


3 reduction was
carried out for the controlled production of hydroxyl ions
in the vicinity of the working electrode. From an aqueous
solution containing 1 M KNO3, it was shown that the
pH could reach a value of 12 at 1.2 V/SCE without dihydrogen bubbles due to the water reduction, which is not
suitable for electrodeposition. This pH was about 2 units
higher than that obtained from only the oxygen reduction
process. This result showed that adding KNO3 in the solution was an appropriate method for the formation of
metallic hydroxide for example, which precipitated in a
pH range higher than 10.4 which was the maximum pH
obtained at the reduction limiting current of dissolved oxygen. The mechanism of NO
3 reduction was studied at
steady state and by EIS in 0.1 M KNO3. The stationary
curves allowed to conrm that the reduction process was
partially controlled by mass transport. Moreover, the EIS
investigation showed the presence of an inductive loop in
the low frequency range due to the adsorption process on
the electrode surface. The inductive response appeared in
a low cathodic potential where no additional current was
observed on the reduction current of dissolved oxygen
and no pH variation as well. This result was conrmed
by the theoretical approach of the impedance diagrams
where the faradaic reactions of nitrate reduction were at
equilibrium. As a consequence, EIS detected adsorbed species at the electrode surface due to the presence of nitrate
ions in the solution in a potential range where no stationary current of NO
3 reduction was involved yet.

94

M. Nobial et al. / Journal of Electroanalytical Chemistry 600 (2007) 8794

References
[1] I. Zhitomirsky, L. Gal-Or, R. Chaim, J. Electrochem. Soc. 138 (1991)
1939.
[2] I. Zhitomirsky, L. Gal-Or, R. Chaim, J. Electrochem. Soc. 138 (1991)
1942.
[3] I. Zhitomirsky, L. Gal-Or, A. Kohn, H. Hennicke, J. Mater. Sci.
Lett. 14 (1995) 807.
[4] I. Masanobu, O. Takashi, J. Electrochem. Soc. 143 (1996) L53.
[5] O. Devos, C. Gabrielli, M. Tlili, B. Tribollet, J. Electrochem. Soc.
150 (2003) C494.
[6] L. Beaunier, C. Gabrielli, G. Poindessous, G. Maurin, R. Rosset, J.
Electroanal. Chem. 501 (2001) 41.
[7] Y. Matsumoto, T. Morikawa, H. Adachi, Hombo, J. Mater. Res.
Bull. 27 (1992) 1319.
[8] S.B. Abolmaali, J.B. Talbot, J. Electrochem. Soc. 140 (1993) 443.
[9] P. Royer, C. Rey, Surface Coating Technol. 45 (1991) 171.
[10] M. Shirkhanzadeh, J. Mater. Sci. Lett. 10 (1991) 1415.
[11] S. Peulon, D. Lincot, Adv. Mater. 8 (2) (1996) 166.
[12] Th. Pauporte, D. Lincot, J. Electrochem. Soc. 148 (2001) C310.
[13] I. Masanobu, O. Takashi, J. Electrochem. Soc. 144 (1997) 1949.
[14] I. Masanobu, O. Takashi, J. Appl. Phys. Lett. 68 (17) (1996)
2439.
[15] I. Masanobu, J. Electrochem. Soc. 144 (12) (1999) 4517.
[16] L. Jaeyoung, T. Yongsug, Electrochem. Solid State Lett. 4 (9) (2001)
C63.
[17] M.M. Tlili, M. Benamor, C. Gabrielli, H. Perrot, B. Tribollet, J.
Electrochem. Soc. 150 (2003) C765.

[18] C. Deslouis, I. Frateur, G. Maurin, B. Tribollet, J. Appl. Electrochem. 27 (1997) 482.


[19] S. Ureta-Zanartu, C. Yanez, Electrochim. Acta 42 (1997) 1725.
[20] S. Cattarin, J. Appl. Electrochem. 22 (1992) 1077.
[21] T.Y. Safonova, O.A. Petrii, J. Electroanal. Chem. 448 (1998) 211.
[22] G. Horanyi, E.M. Ryzmayer, J. Electroanal. Chem. 188 (1985) 265.
[23] M.T. De Groot, M.T.M. Koper, J. Electroanal. Chem. 562 (2004) 81.
[24] M.C.P.M. Da Cunha, M. Weber, F.C. Nart, J. Electroanal. Chem.
414 (1996) 163.
[25] M.C.P.M. Da Cunha, J.P.I. De Souza, F.C. Nart, J. Am. Chem. Soc.
16 (2000) 771.
[26] H.-L. Li, J.Q. Chambers, D.T. Hobbs, J. Appl. Electrochem. 18
(1998) 454.
[27] K. Bouzek, M. Paidar, A. Sadilkova, H. Bergmann, J. Appl.
Electrochem. 31 (2001) 1185.
[28] I.G. Gasella, M. Gatta, J. Electroanal. Chem. 568 (2004) 183.
[29] M.S. El-Deab, Electrochim. Acta 49 (2004) 1639.
[30] F. Bouamrane, A. Tadjeddine, J.E. Butler, R. Tenne, C. LevyClement, J. Electroanal. Chem. 405 (1996) 95.
[31] M.J. Alowitz, M.M. Scherer, J. Am. Chem. Soc. 36 (3) (2002) 299.
[32] Handbook of Chemistry and Physics, 63rd ed., 1982.
[33] M.E. Orazem, N. Pebere, B. Tribollet, J. Electrochem. Soc. 153
(2006) B129.
[34] G.J. Brug, A.L.G. Van den Eeden, M. Sluyters-Rehbach, J.H.
Sluyters, J. Electroanal. Chem. 176 (1984) 275.
[35] J.P. Diard, C. Hecker, J. Electroanal. Chem. 121 (1981) 125.
[36] A.C.A. De Vooys, R.A. Van Santen, J.A.R. Van Veen, J. Mol. Catal.
A 154 (2000) 203.

Вам также может понравиться