1 s2.0 001346869080027L Main
1 s2.0 001346869080027L Main
1 s2.0 001346869080027L Main
Abstract-Impedance measurements in the high frequency range beyond the electrolyte resistance, R,,
currently exhibit one or several loops irrelevant to the electrode process. For low conductivity media
encountered in spreading corrosion studies these relaxations appear in a lower frequency range
(l&l00 Hz) and can overlap the faradaic frequency range, thus leading to possible misinterpretation of
the data. This aspect was investigated in concentrated acetic acid solutions using two different kinds of
reference electrode. Solution resistivity as well as nature of the reference electrode and the distance between
working electrode surface and reference extremity were extensively investigated. The parasitic contribu-
tions of the different elements of the experimental setup on the measured impedance was modelled by a
unique electrical circuit which reproduces with an excellent precision the experimental data. This allowed
us to underline the various contributions of the electrical couplings occurring in the electrochemical cell,
especially between the reference electrode and the other elements.
1125
1126 S. CHECIBRCLIAN
et al.
Ag/AgCl reference which Z’ and Z” are the real and imaginary parts of
electrode 4 Platinum the impedance.
Wire
3. EXPERIMENTAL RESULTS
Rotating
device 3.1. Ag/AgCl electrode and Luggin capillary
Figure 2a shows the progression of the impedance
i diagram in relation to the distance x in an 80% acetic
CSCOOH acid solution. In the high frequency range of the
II
+ 0.1 M LiClO~ Platinum
electrode diagram (2.5-65 kHz) one sees a capacitive loop
whose diameter remains constant whatever x may be
and on which the frequency distribution does not
vary.
At low frequency (less than 100 Hz), the beginning
of a large diameter loop appears. This impedance is
Counter-electrode
relative to the faradaic impedance of the working
electrode. The intersection of its high frequency part
Fig. 1. Experimental device.
with the real axis leads to the electrolyte resistance R, .
May we recall that one can also calculate R, by
electrode was mounted on an ED1 Tacussel rotating comparing the value x with Newman’s equation
device whose speed was set at 1000 rpm for all the based on the distribution of the primary electrical
tests. The cell configuration adopted made it possi- field[7].
ble[2] to apply the analytical equations of Newman[q A third loop appears at the intermediate frequen-
for estimating the electrolyte resistance value at a cies (100 Hz-2.5 kHz): when x is greater than
given point in the cell. 2 mm, it is capacitive and its diameter grows with x.
Two types of reference electrodes were tried out: (i) When x becomes less than 2 mm the capacitive loop
A commercial Ag/AgCl electrode prepared in anhy-
drous acetic acid, saturated with LiCl and AgCl,
extended by a Luggin capillary filled with anhydrous
acetic acid + 0.1 M LiClO,. In order to limit ex-
changes with the solution, the extremity of the capil-
lary was plugged with asbestos fibre: (ii) A platinum
wire 0.5 mm in diameter, inserted in a glass capillary,
only the cross section of the platinum wire being in
contact with the electrolyte. This wire does not in
itself constitute a reference electrode but plays the
role of a low impedance alternating potential probe.
Its connection, via a capacitance, in parallel with the
reference electrode is generally recommended in order
to attenuate measurement difficulties at high frequen-
cies[l, 91. We shall see further on the reasons for the
improvement thus obtained. The extremity of the
reference electrode is located on the axis of the
working electrode. A screw device made it possible by
the use of a dial indicator to precisely check the
distance x between the surface of the working elec-
trode and the extremity of the reference electrode.
2.2. Electrolytes
60C
Two solutions with different resistivities were used: 1
2.3. Regulation-measurements
The measurements were carried out with a Solar-
tron-Schlumberger assembly (1286 potentiostat, 1250
frequency response analyser) controlled by an Apple
IIe personal computer.
The tests were carried out at room temperature.
For each reference electrode/solution couple, we
recorded the impedance diagram of the cell in galv- Fig. 2. Impedance diagram at different distances x between
anostatic mode for different values of x. A pseudo- the capillary tip (Ag/AgCl electrode) and the working
three-dimensional (Z’, Z”, x) display was adopted, in electrode: (a) 80% acetic acid; (b) 100% acetic acid.
Impedance measurements in low conductivity media 1127
Fig. 3. Impedance diagram at different distances x between the extremity of the platinum electrode and
the working electrode: (a) 80% acetic acid; (b) 100% acetic acid.
c2T
? ..zQR~l
Zf 7
Z,=R:+i
jd, ’
zapp.=
($+ I)(_!.&;_;9
(7) that the simulated diagrams are close to the experi-
mental diagrams.
The variation of the distance between the working
electrode and the reference electrode was simulated
RI 1 by modifying the ratio RJR; whilst keeping the sum
R, + R: constant. The values adopted for R, were
*+ (8) chosen from the experimental diagrams shown
B
in Figs 2 and 3 whereas the sum R, + R: was esti-
mated in accordance with the Newman’s equation:
-2-lkC-S
60
i
Iktiz
I I
*
2
3 Z’(kCU
R* (km
f
I I )
WC
m Z’(kra
Fig. 5. Simulated diagrams corresponding with the situation of the Ag/AgCl electrode with the capillary:
(a) 80% acetic acid; (b) 100% acetic acid.
1130 S. CI~ECH~RCLIAN
et al.
Table 1. Values attributed to the various elements of Fig. 4 in the case of the Ag/AgCl electrode with the capillary
AidAtSl 4 R,+R: RI R* R, Cd C, G C,
80%
CH,COOH &2kf2 4kQ 2MQ 2MR 1OOkI-I @IJF 10 pF 30pF IMF
100%
CH,COOH O-O.5MR 2MR 2MfJ IOOMR 1OOk.Q 6OpF 10 pF 3OpF 1MF
Table 2. Values attributed to the various elements of Fin. 4 in the case of the platinum electrode
Platinum R R+R: R:’ R. R, C.4 C, c, c.
80%
CH,COOH &1.5kf-& 4kR 0.1 MR 0.1 M&I 0.1 MR MpF 1OpF 30pF O.lpF
100%
CH,COOH &I .2 MR 2MR 2MQ 1OOMR 0.1 MR 6OUF 10~F 30~F O.luF
R, + R: = p/4,,, in which r,, is the radius of the we estimated the capacitance presented by the glass
working electrode. tube used for the capillary. The latter had an exter-
nal diameter of 4 mm, a wall thickness of 1 mm
and was submerged to about 80mm. By applying
4.2. Digital simulation the equation which allows us to calculate the
Figure 5 represents the diagrams simulating the capacitance of a cylindrical capacitor, we found:
behaviour of the Ag/AgCl electrode. They were ob- C = 32pF. This value matches quite well those
tained by attributing to the elements of the model chosen for C, and C, (considering that the calcu-
the values given in Table 1. In the same way, the lation was approximate, given that the potential
values in Table 2 make it possible to simulate the field around the capillary was probably not
behaviour of the platinum electrode (Fig. 6). The uniform).
choice of these values provides a very good similar- The qualitative interpretation of the shape of the
ity between the diagrams obtained experimentally diagrams observed appears quite clearly if one notes
and those calculated from the chosen model, this that the diameter of the high frequency capacitive
being for both concentrations of acetic acid and loop is equal to: (R, + R:). C,/(C, + C,). Capaci-
both electrodes. Below we give some arguments on tances C, and C, therefore constitute a dividing
the consistency of the values attributed to the ele- bridge acting on the potential V of the counter-
ments of the model. electrode (c$ Fig. 4). The existence of two loops
The values adopted for C,, and R, concern the is a consequence of the antagonistic effect of
faradaic impedance of the working electrode; the capacitances C, and C, on the potential V, measured
values are in fair agreement with those found by the reference electrode. C, introduces at this
during experiments which we have carried out level a part increasing with the frequency of the
elsewhere. voltage of the counter-electrode. In parallel, C, re-
In the case of the Ag/AgCl electrode with the duces V, to zero when the frequency goes towards
capillary the presence of the capacitance C, is infinity.
not justified, and this is why we attribute a high
value to it which gives it a negligible impedance
in the frequency range in which we are interested. 5. CONCLUSION
For the platinum electrode, the value adopted
(0. I pF) corresponds with the double layer capaci- A thorough knowledge of the phenomena ob-
tance of the platinum in acetic acid. Experimental served in the range of high frequencies when
determination of the latter by impedance measure- measuring impedance in low conductivity media
= 41.4 PF cme2 that is for
ment gave us: CdPI,AeOH80s,, is essential for interpreting experimental diagrams.
our 0.5 mm diameter electrode: C = 0.093 ~IF. R:I The role of the conductivity of the electrolyte, the
does not vary when the resistivity of the solution design of the reference electrode and the geometry
changes. This indicates that it is not totally depen- of the cell on the artefacts altering the impedance
dent on the test solution, but rather on the measurements at the high frequency limit of the
electrolyte contained in the capillary. However, in spectrum is correctly described by a unique electrical
the case of the platinum electrode it is not negli- circuit. The values attributed to the elements corre-
gible which makes us think that it may to a large spond with physical consideration and/or direct
extent be localized in the internal circuits of the measurements.
potentiostat. The increases in the resistivity of the solution can
Capacitances C, and C2 are independent of the to a certain extent be compensated by the use of a
resistivity of the solution. We chose to give C, a low impedance potential measurement probe such
higher value than C, to allow for the potentiostat as a platinum wire, keeping in mind that this type
input capacitance. In order to check if the values of electrode does not constitute a real reference
attributed to C, and C2 were consistent with reality, electrode.
Impedance measurements in low conductivity media 1131
(W
Fig. 6. Simulated diagrams corresponding with the situation of the platinum electrode. (a) 80% acetic
acid; (b) 100% acetic acid.