close

Вход

Забыли?

вход по аккаунту

?

Electrodes for Potential Measurements in Aqueous Systems at High Temperatures and Pressures.

код для вставкиСкачать
U
E
-
Number 3
March 1987
Pages 161-274
Volume 26
International Edition in English
Electrodes for Potential Measurements in Aqueous Systems at
High Temperatures and Pressures
By Leonard W. Niedrach"
The measurement of pH, redox potentials, and corrosion potentials at high temperatures
and pressures is often desirable for the control and monitoring of industrial processes; e.g.,
for controlling water chemistry in conventional and nuclear power generators, for process
control in the chemical industry, and for monitoring performance in petroleum extraction
and refining. Such measurements are also needed for research on the thermodynamics of
high-temperature aqueous solutions including those associated with geothermal and hydrothermal reactions. Zirconia electrodes for these purposes are described, and their application is illustrated with examples. While not yet routinely available on a commercial basis,
useful versions can be readily fabricated in the laboratory. Some of the pitfalls encountered
in making such measurements at elevated temperatures are also discussed because great
care is often required to assure that valid and useful data are being obtained.
1. Introduction
Of considerable concern in the operation of high-temperature aqueous systems is the control of p H and the redox potential because of their importance as process variables in such areas as paper manufacturing, catalysis, polymerization, and petroleum refining. While in some cases
adequate information can be obtained from measurements
on cooled side streams, it is often more useful to perform
the measurements at the process temperature. Under such
conditions less time lag is involved, and, more important,
unpredicted shifts in equilibria are avoided. The importance of making measurements under process conditions is
particularly evident in the case of corrosion potentials. In
the nuclear industry, for example, it has been demonstrated that maintaining the corrosion potential of stainless
steel below a critical value can be effective in mitigating
I*]
Dr. L. W. Niedrach
General Electric Company
Corporate Research and Development
P.O. Box 8, Schenectady, N Y 12301 (USA)
Angew. Chem In1
Ed. Engl. 26 (1987) 161-169
stress corrosion cracking. For adequate control such measurements must be made under the process conditions.
In recent years much work has been published aimed at
the development of suitable electrodes for such measurements. As might be expected, much of the impetus has
come from the nuclear industry through a widespread interest in mitigating corrosion problems through an understanding of the theoretical principles and thermodynamic
parameters involved. This article reviews some of these developments and describes some of the more useful electrodes. Their applications are illustrated with examples.
2. Description of Electrodes
2.1. Conventional Reference Electrodes
In thinking of a reference electrode one normally focuses on a type that is essentially insensitive to the environment being monitored. Such electrodes are generally
fabricated with a highly reversible couple such as the mercury/mercury(I) chloride (calomel) couple or the silver/
silver chloride couple in contact with a stable, aqueous po-
0 YCH Verlag~gesellschufimbH. 0-6940 Weinherm. 1987
0570-0833/87/0303-0l61 $ 02.50/0
161
tassium chlorir'e solution, and they are capable of establishing a liquid junction with the environment being monitored with only a minor offset or liquid-junction potential
at the interface. For applications at ambient temperature
the saturated o r 3 M calomel and silverfsilver chloride
electrodes are classical examples. Such high concentrations of KCI cannot be employed at high temperatures,
however, because of the strong complexing action on mercury(!) and silver chlorides and the ensuing changes in
chloride activity. It has been found, however, that lower
concentrations of KCI d o function satisfactorily at elevated temperatures, and considerable effort has been devoted to the development of practical reference electrodes
based upon the silver/silver chloride couple.
Two variants of the silver/silver chloride electrode have
proven most popular for high-temperature applications.
One, the isothermal type, operates entirely at the temperature and pressure of the environment being monitored. The
other, the thermal junction type, operates with a thermal
junction across the electrolyte bridge, and the reference
couple itself is held at ambient temperature. Both have
proven quite satisfactory in practice.
Numerous investigators have described a variety of versions of the isothermal-type ele~trode.l'-~'
That of Indig et
al.I3l will serve as an example for more detailed discussion.
As shown in Figure 1, a silver wire coated with silver chloride is contained within a Teflon housing adapted to a
Conax pressure fitting. The filling sofution of 0.01 M KCI
communicates with the external environment through a
porous plug. Related structures employing only water as
the internal filling solution have been found to perform satisfactorily for measurements in the high-purity water employed in nuclear reactor^.^^.^' In this case an excess of
silver chloride is added to the Teflon housing and its dissolution in the water within the housing establishes a steady
reference potential.
The thermal-gradient-type electrode has also been investigated by a number of workers.16-81An example, as described by Danielson,161is shown in Figure 2. It is assem-
n
:KING
TEFLON COATED,
TAPERED, AND
SILVER P L A T E D
SILVER /STAINLESS
STEEL ROD
001 M
SHRINK
TEFLON
TUBING
Fig. I. Isothermal-type reference electrode after lndig et al. [3]. Standard
Conax fittings establish the pressure boundary. Care must be exercised that
the porous zirconia frit forms the only junction with the environment being
monitored.
bled within a Teflon tube, which also serves as the electrolyte bridge to the point of interest within the environment
being monitored. To permit operation in pressurized systems, the Teflon bridge is retained within metal tubing.
The potential offset resulting from the temperature gradient is a function of the temperature difference between
the solution and the reference couple as well as the concentration of the salt in the bridge. These effects have been
studied by Macdonald et aI.,(" who have tabulated relevant
offset data.
In our experience, both types of reference electrode
function quite satisfactorily. It must be emphasized, how-
CONAX
ELECTRICAL GLAND
1/8" INNER-
TEFLON TUBING
ADAPTER
TEFLON SLEEVE
TO CLOSE TUBE
AFTER FILLING
SILVER HOOK TO HOLD
TEFLON TUBE AND
ZIRCONIAYARN
FOR FILLING
1/8" INNERDIAMETER
ZlRCONlA YARN
162
Fig. 2. Thermal-gradient-type reference electrode after Danielson [6]. A porous zirconia
plug (not shown) forms the junction with the
environment being monitored. The zirconia
yarn serves to bridge any gaps in the electro118" SILVER ROD
lyte that may form during operation without a n
overpressure.
t
Angew. Chem. Inr. Ed. Engl. 26 (1987) 161-169
ever, that both d o involve porous junctions between the
reference couple and the environment being monitored. In
principle, then, exchange of solutes can occur between the
two and this can result in drift in the reference potential as
chloride is lost or gained by the filling solution. The rate of
such drift is dependent upon the magnitude of the concentration gradients and the porosity of the junction.[51 It is
also important to consider the expansion of water upon
heating to higher temperatures. Thus, upon cooling after
use at a high temperature, water will be drawn in from the
external environment to further change the chloride concentration within the electrode chamber. This effect is
smaller with a thermal-gradient electrode than with an isothermal electrode because less of the filling solution goes
through the temperature cycle in the former case.
Another possible source of drift is associated with perfusion of dissolved hydrogen or oxygen into the reference
chamber from the environment. Hydrogen can reduce
silver chloride to release hydrochloric acid into the solution, thereby contributing to drift by increasing the chloride ion concentration. Conversely, oxygen can oxidize
silver in the presence of chloride to form silver chloride
and thereby cause drift in the opposite direction. Perfusion
is markedly limited by the outer steel tubing employed
with the Danielson electrode. A variant of the Indig electrode, in which the Teflon housing is replaced with a zirconia tube, reduces perfusion with that type of electrode.[']
2.2. Indicator Electrodes
For the monitoring of the redox potential of a n environment, a platinum electrode can be used in conjunction
with an appropriate reference electrode. Design can vary
significantly, the major concern being a reliable seal
against the pressure and adequate insulation of the support so that the working area is reasonably well defined.
An example is shown Figure 3. A Conax compression seal
with a Teflon insert is used to penetrate the pressure barrier and Teflon is used to insulate all but the area designed
to contact the environment. Alternatively, a flag-type electrode supported by a Teflon-insulated wire passing
through the compression seal may be used.
2.3. pH Electrodes
A wide variety of electrodes have been investigated in
order to measure p H at elevated temperatures. They include the classical hydrogen electrode,['0.' I ] the palladium
hydride electrode,'". 13] the glass electrode, and metal-metal oxide couples.['41 For general application, the first suffers from a need to know the concentration of dissolved
hydrogen in the environment being monitored and from
interference from other redox-active species. Metal oxide
and palladium hydride electrodes also suffer from interference from other redox-active species. U p to temperatures
of about 150°C the glass electrode can be employed, but at
higher temperatures it is severely attacked even by highpurity water. This electrode, however, responds through
changes in a membrane potential rather that through the
Anqew Chem In1 Ed Engl 26(1987) 161-169
respome of electron-transfer couples and is therefore independent of redox-active species in the solution.
TEFLON SEAL
U
Fig. 3. Metal electrode after Indig et al. [3]. The stainless steel support rod is
threaded on the lower end to accept the metal electrode. Care must be exercised to assure a tight seal of the metal electrode against the Teflon.
It was more recently found that stabilized zirconia can
serve as a membrane-type electrode, much like the glass
ele~trode,"~.
16] but without the major disadvantage of the
latter (excessive solubility at high temperatures). This has
proven to be the most satisfactory type of pH sensor available for high-temperature studies. Representative designs
are shown in Figure 4. In early work sensors were fabricated in a fashion somewhat akin to that of the conventional glass electrode in that a buffered saline solution in
contact with a silver wire coated with silver chloride served
as the internal element within the membrane sheath"'.
(cf. Fig. 4a). Although this type of internal element has
been used by other^,'"^ it has been observed that a dry mix
of a metal and its oxide-or two oxides of the same metal
in different valence states-is preferable."81 Such a design,
in which the copper-copper(1) oxide couple serves as the
internal element, is shown in Figure 4b. Mercury/mercury( 11) oxide has also served satisfactorily['6.19] and the use
of silver/silver oxide has also been reported.["] In another
satisfactory approach the internal element has been prepared with platinum left exposed to air.['''
In the design shown in Figure 4b, a 1/4-in.-outer-diameter tube[*] consisting of zirconia stabilized with yttria (8
wt-%) is retained in a Conax type EG-125 fitting with a
seal consisting of Teflon, Vespel, silver, and alumina elements as shown. The dry mix of copper and copper(1) oxide powders (1 : 1 by weight) is packed around a copper
wire in a one to two inch long section near the closed end
of the zirconia tube. This defines the active region of the
['I Obtainable from Corning Glass Works/Zircoa,
Solon, O H (USA).
163
b)
a)
ALUMINA INSULATOR
ALUM IN A I N SU L AT0 R
SILVER DISKS
TEFLON SEAL
PRESSURE F I T T I N G
BUFFERED SALINE
ELECTROLYTE
z IRCON I A
’MEMBRANE
‘
SILVER CHLORIDE
COATING
MIX OF CU + Cue0
+
1“
Fig. 4. Schematic diagrams of pH sensors with zirconia “membranes.” The diagrams are essentially to scale and none 0 1 the dirnrnaions
are critical. Clearance between the zirconia tube and the lower portion of the Conax fitting must be provided, if necessary, by enlarging
the bore of the fitting. The Teflon and Vespel contacting the zirconia must be machined to fit the outside diameters of the individual
tubes, which show considerable variation. For a) and b), see text.
sensor. A backing of glass wool is then added to help confine the powder. Finally, the seal is assembled and tightened with a torque of about 5.7 Nm. This design offers
several advantages .over the earlier aqueous system: (1) it
can be readily prepared by simple mixing of the powders
and packing into the tube; (2) it is extremely stable and
reproducible; (3) it permits ready designation of the active
region of the sensor because the wall is not wetted with a
conducting film; and (4) seal fabrication is simplified because of the absence of an internal aqueous phase.
This arrangement has proven highly satisfactory for laboratory work, but because of the thermal properties of the
polymeric elements and their tendencies to flow and/or relax, care must be taken to keep them thoroughly confined.
For this reason a modification with a brazed metal-to-ceramic seal has been devised.[”’
Because it is relatively new, a brief discussion of the operation of the zirconia pH sensor is perhaps in order. For
the sensor shown in Figure 4b, several interphase potentials are involved in series between the external terminal
(copper wire) and the solution being monitored. These and
the thermodynamic balances involved in the potential-determining reactions at the various interfaces are summarized in Figure 5 according to the approach of Vettev.[22J
The balances are derived from the fact that at equilibrium
the algebraic sum of the electrochemical potentials (,q,)of
the reacting species is zero:
164
where
and where ,p,is the chemical potential of reacting species s
in phase i, , z , is the charge on the species, ,# is the Galvani
(inner) potential in phase i, ,v,is the stoichiometric factor
of the species involved in the potential-determining reaction, and F is Faraday’s constant.
In summing the interphase potentials between phases 1
and 4, the net change is equivalent to that for the simple
copper/copper(l) oxide couple (see Fig. 5):
2Cu
+ HZO
-
CuzO + 2H+ + 2 e -
(3)
The thermodynamics of this reaction is well established.
The membrane may therefore be looked upon as a protective sheath that prevents the metal and its oxide from
interacting directly with the solution or redox-active species in the solution. Whereas a copper/copper(i) oxide mix
could be exposed to water and dilute acids and bases without significant attack, such would not be the case with
strongly acidic or basic solutions. Under any conditions,
use of the unsheathed metal/metal oxide mix would necessitate alternative protection from direct interaction with
hydrogen or oxygen.
Angew. Chem. Int. Ed. Engl. 26 11987) 161-169
This is thus equivalent to the potential of the following half couple: Cu"+ % H 2 0 = % C u 2 0 + H ++ e -
ate hydrogen. The platinum flag then clearly showed its
sensitivity to this change in the redox environment. The
oxygen-ion-conducting ZrOz membrane, however, retained
an essentially constant potential. The small, transient increase in potential is believed to have been a response to a
real transient in pH associated with the change in the passivation reaction on the stainless steel.
400
Fig 5. lnterphase potentials and related thermodynamic data for the zirconia
pH sensor.
:f
T= 285 "C
RUN Zr0,-19
300
3. Applications
t
>
3.1. Measurement of pH
E
L
W
0
PT REDOX PROBE
f 8 p p m 02)
n
Zirconia sensors are particularly of interest for the measurement of pH at relatively high temperatures and pressures; e.g., 285°C at 83 bar. When used under these conditions with a n appropriate silver/silver chloride reference
electrode, they have invariably performed well in both
pure water and simulated brines for as long as 40 days.
Such performance has been thoroughly documented elsewhere,['8.231and representative response data obtained in
conjunction with a reference electrode of the isothermal
type are shown in Figure 6.
Also shown in Figure 6 are the pH responses of a hydrogen electrode and an oxygen electrode under the same
conditions. In the latter two cases the same platinized platinum flag was used; however, the redox status of the solutions was changed by equilibrating with eithera hydrogennitrogen or oxygen-nitrogen mixture. With the hydrogen
electrode the response paralleled that of the zirconia electrode since both electrodes are nearly reversible. The response of the oxygen electrode deviates from the theoretically predicted values because of its poorer kinetic behavior.
The data in Figure 7 further emphasize that the zirconia
electrode, like the glass electrode, does not respond to
changes in the redox potential of the environment. These
data were obtained while 0.0005 M sulfuric acid, initially
aerated, was being fed to the stainless steel autoclave."']
On switching to a nitrogen purge of the feed stream, the
potentials of the autoclave and the platinum flag (oxygen
electrode) immediately started to fall while that of the zirconia pH sensor remained constant. After several hours a
more rapid drop occurred in the potentials of the autoclave and the platinum electrode. This is attributed to a
change in the passivation reaction for the stainless steel
autoclave. lnstead of the passive film being maintained by
the reduction of oxygen to water, water is reduced to liberAngew. Chem. Ini. Ed. Engl. 26 11987) 161-169
0
(r
c
0
w
-100
-2ootW
w
-300
LT
-400
\
LT
v,
-500
ZrO, pH SENSOR
(Y
1
-700
- 800
-900
-1000
\
t\
P: RED:
PrBE
(160 p p b H p )
3
4
5
6
PH
7
8
9
,
10
Fig. 6 . Variations in the potentials of several electrodes with pH. Reference
electrode: silver/silver chloride.
w
-
WITH NITROGEN
-
lh
Y
II
< -200
H
Lcl
TIME
-
Fig. 7. Behavior of a Z r 0 2 pH sensor, the oxygen electrode, and the stainless
steel autoclave as the redox potential of the environment is changed at
285°C The transient perturbation of the membrane potential at the Z r 0 2 pH
sensor IS believed to reflect a real pH transient associated with a change in
the corrosion film on the steel autoclave as the corrosion potential changes.
Reference electrode: silver/silver chloride.
165
An additional example demonstrating the versatility of
the zirconia pH sensor is the collection of data in Figure 8
obtained by monitoring pH changes in simulated geothermal brines upon changing the carbon dioxide content and
upon introducing hydrochloric acid or base into the solution. The upper portion of the figure shows the changes
observed in the solution at 25°C as measured with a commercial glass electrode (Ingold) while the lower portion
shows the measurements at 285°C. For the latter, two reference electrodes, 13RR and 14RR, of the thermal-gradient
type were employed. As much as 100 ppm of sulfide ion
(as sodium sulfide) was also introduced into the solution at
various times. This had no effect on the instrumentation or
the electrodes apart from that caused by the pH changes.
Further examples of the performance of this combination of zirconia pH electrode with a thermal-gradient-type
reference electrode are shown in Figure 9 where measured
pH values are compared with calculated values for 0.2and 20-percent brine solutions. In both cases the agreement is excellent.
As noted above, in experiments of this type there is always some chance for bias to creep into the measurements
through offsets in either the sensing or reference electrode.
It is therefore desirable to calibrate the combination of
sensor electrode and reference electrode with known
standards. In the brine work, for example, slow drift in the
calibration always occurred. Upward displacement of the
calibration curve occurred with a 20-percent brine and
RUN EB-6
OPERATION ON BRINE (0.2%NaCl/0.001M NaHC0,I
O.Z%NaCI,NO CO,,NO 5,-
-----
DAY
-
Fig. 8. Potentials of the zirconia pH sensor at 285°C and glass electrodes at 25°C versus silver/silver chloride as reference
electrode during run GB-6. Two high-temperature reference electrodes, 13RR and 14RR, were used.
a1
10
t9
0
" , 8
CD
(u
/
10
T9
0
L 8
m
3
166
7
I
I
I
I
I
I
I
I
Angew. Chem. Int. Ed. Engl. 26 (1987) 161-169
KCI and resorting to only the dissolution of an excess of
silver chloride gave a more stable potential.[',241
Even under these conditions the zirconia pH sensor and
the R / 0 2electrode['1 have been shown to provide the most
stable potential. Figures 10 and 11 show that the most stable correlations are obtained for the potential of the oxygen electrode vs. the ZrOz pH sensor. The stability of the
R / H 2 electrode also falls into this category and in addition is more reliable in establishing a predictable potentia1.'2'1 Several types of reference electrode are thus available that can be used as cross-checks on one another if sufficient environmental information is available (namely, the
p H and the concentration of dissolved hydrogen or oxygen).
downward displacements with a 0.2-percent brine. Since
the original filling solution in the reference electrode was
0.1 M KCI in all cases, the displacements reflected diffusion of chloride into or out of the reference electrode as a
result of the concentration gradients between the internal
filling solution and the environment being monitored. This
suggests that selection of a filling solution having a chloride concentration close to that of the brine should minimize drift of this type.
3.2. Measurement of Corrosion and Redox Potentials
Corrosion o r redox potentials are often measured in order to maintain their values within some favored range.
Often these limits are specified in terms of potentials relative to that of the standard hydrogen electrode (SHE).
Such values can be derived from measurements against
any of the above electrodes, which may serve as reference
electrodes when sufficient ancillary information is available. For the conventional type of reference electrode, such
as the silver/silver chloride electrode, only the chloride ion
content within the electrode chamber must be known. With
a knowledge of the p H of the environment, a p H electrode
may also serve as a reference electrode. If, in addition, the
hydrogen or oxygen activity of the environment is known,
the R / H 2 (or R/02) electrode may also serve.
In principle the problem of measuring corrosion o r redox potentials should equate with that of measuring the
pH. In each case one is interested in the potential between
a sensing electrode and a reference electrode. A fundamental difficulty arises, however, in that a convenient, definitive means of calibrating the measurement is available
only when measuring pH. In that case a conventional scale
has been adopted and convenient standard buffers are
available for frequent calibrations to assure the accuracy
of the potential measurement even though the reference
and/or the sensing electrodes may have suffered drift.
Such standards are not available for corrosion and redox potential measuremen'ts. In particular, corrosion potentials are expected to vary considerably with the experimental conditions, the age of the corrosion film, the composition of the metal, etc. For this reason the most stable
reference electrode available should be employed for such
measurements. If one wishes to be more conservative, at
least one additional reference electrode of another type
should be employed as a check. To underscore this point
further, two reference electrodes of the same type cannot
be considered as reliable checks upon one another; both
could respond to the environment in identical, erroneous
ways, o r exhibit essentially identical drifts with time.
In connection with the stability of reference electrodes,
we have made a number of intercomparisons of the relative stabilities of silver/silver chloride electrodes, zirconia
pH sensors, and R / H 2 and R/02 electrodes.[''] As noted
in Section 2.1, silver/silver chloride electrodes must be expected to show some drift with time because of their diffusion junctions, especially if temperature cycling of the solution occurs. Such drift has clearly been demonstrated
with fillings of 0.01 and 0.1 M KCI."] In agreement with
Leibovitz et al.J4] we have found that elimination of the
Angew Chem. I n ! . Ed. Engl. 26 (1987) 161-169
4M)
I
-f
w
E
PT REDOX PROBE
I
300-
W
a
0
a
-
I-
V
1
W
W
T=285OC
W
V
z
;
-200 1
3
0
10
20
TIME
30
40
fDAYSl
Fig. 10. Potentials of the platinum electrode and the zirconia pH sensor (reference electrode: silver/silver chloride). The solution was aerated so that the
platinum was functioning as an oxygen electrode. The silver chloride electrode contained no KCI. The similar temporal changes shown by the platinum and zirconia electrodes suggest that the major cause of the variations
lies with the silver/silver chloride electrode. White symbols: Run Zr02-19;
black symbols: Run Zr02-20.
-
PT R E D O X P R O B E
a
w
LL
0
$1
L
"
500
T=285 C'
N
-
>
v)
LJ
300
0
I
1
10
20
TIME [DAYS]
-
1
I
30
40
Fig. 11. Potential of the platinum electrode in the aerated solution (reference
electrode: zirconia pH sensor). The excellent stability implies that both electrodes are very stable. White triangles: Run Zr02-19; black triangles: Run
Zr02-20.
The focus of interest to date has been in measuring corrosion potentials of stainless steel and other structural materials in the high-purity cooling water of boiling-water
167
reactors at about 285°C. Routine measurements of conductivity and p H at ambient temperature assure that the
water is indeed pure and therefore has a pH value close to
the theoretical neutral point of 5.65 at 285°C. Routine
monitoring of hydrogen and oxygen concentrationswhich are the redox-potential-determining species in this
case-then permit the calculation of the potential of the
hydrogen or oxygen electrode.
The potentials of the various electrodes under consideration vs. the S H E potential at 285°C have been calculated.
With freshly prepared reference electrodes, satisfactory
agreement can be obtained with measurements between
various pairs of reference electrodes, generally to within
k 20-30 mV. With the passage of time, however, drift will
sometimes be encountered, especially in measurements involving the silver/silver chloride electrode with the KCI
filling.
Based upon our experience, the zirconia p H electrode is
preferred for measurements with such relatively pure water
systems as the co'oling water of boiling-water reactors. This
has indeed been our choice for studies of corrosion potentials of stainless steel under so-called normal and hydrogen water conditions. Under normal water conditions,
oxygen tends to be somewhat in stoichiometric excess of
hydrogen and the platinum electrode behaves much as an
oxygen electrode. Under hydrogen water conditions, hydrogen is in excess and the platinum electrode behaves as
a hydrogen electrode. The behavior of stainless steel is
more c o m p l i ~ a t e d . ' ~These
~'
conditions are of practical interest because the more reducing conditions result in suppression of stress corrosion crackingf2@and much present
work is aimed at establishing the value of a critical potential as well as electrochemical means for monitoring the
reactor piping.
The data in Figure 12, obtained during studies of crack
growth rates of 304 stainless steel under hydrogen water
conditions,[z41illustrate the way the corrosion potential responds to changes in the water chemistry. The data were
obtained using a zirconia p H electrode, with a copper/
copper(x) oxide internal
as the reference electrode. The water chemistry was also followed continuously
by analyses of pH, hydrogen and oxygen concentration,
and conductivity of the effluent from the autoclave at am-
bient temperature. Because of the reversibility of the R / H Z
electrode, its behavior under changing conditions provided
a check on the zirconia electrode during operation under
hydrogen water conditions. A silver/silver chloride reference electrode was not used in the work.
In this case the potential measurement was employed as
a routine process-control monitor. Shown in Figure 12 are
corrosion potentials measured at 285°C for water nominally containing 160 ppb dissolved hydrogen and no oxygen. At one point, immediately following replacement of a
tank of forming gas (nominally 10 percent H2 in N2) used
to maintain the hydrogen level in the feed stream, the corrosion potential began to rise significantly. Subsequent
analysis indicated that the drift had resulted from oxygen
contamination in the forming gas to a sufficient extent to
introduce 50 ppb into the feedwater. Since the observed
effect possibly reflected an influence of the hydrogen-oxygen recombination reaction on the mixed potential of the
corroding specimen, additional studies were undertaken.
These examples, drawn from the area of nuclear reactor
technology, are perhaps representative of somewhat ideal
conditions for such measurements coupled with relatively
straightforward means for checking on the validity of the
measurements. In situations where complex solutions at
high ionic strength are involved, the required additional
information concerning pH and redox activity may be too
difficult to obtain and correlate to permit valid intercomparisons of the various potential measurements. Under
such circumstances extreme caution is required in the interpretation of measured values.
4. Summary and Outlook
Electrodes have been described for the measurement of
p H and of redox and corrosion potentials at elevated temperatures. While not yet routinely available o n a commercial basis, useful versions can be readily fabricated in the
laboratory. In making such measurements at elevated temperatures many pitfalls can be encountered, and great care
is required to assure that valid and useful data are being
obtained. When possible, direct calibrations to check performance, such as the measurement of pH, should be
HYDROGEN CONCENTRATION
0 FORMING GAS
L'
0
168
10
20
TlMEIhl
-
30
Fig. 12. Transients in the corrosion potential of stainless
steel caused by forming gas (N, containing 10% H2) contaminated with 0.125% oxygen. The oxygen content of the
gas would introduce 50 ppb 0, into the water. This amount
is stoichiometrically equivalent to only 4 ppb H2and would
not be expected a priori to have such pronounced effects on
the corrosion potential of steel.
Angew. Chem. hi.Ed. Engl. 26 (1987) 161-169
made. When such calibration is not possible, it is desirable
to employ at least two reference electrodes of different
types as a means of protection against undetected, insidious drift of a single reference electrode.
The customized electrodes of today are providing a basis
for establishing instrumentation and procedures for their
more routine application in the future. As their usefulness
becomes more established, commercial versions should become more readily available, and general application
should expand.
Received: November 22, 1985;
revised: April 28, 1986 [A 606 IE]
German version. Angew. Chem. 99 (1987) 183
[ I ] M. Indig, D. A. Vermilyea, Corrosion (Houstonj 27 (1971) 312.
121 A. K. Agrawal, R. W. Staehle, Corrosion (Houston) 33 (1977) 418.
131 M. Indig, A. R. Mcllree, Corrosion (Houston) 35 (1979) 288.
[4] J. Leibovitz, W. R. Kassen, W. L. Pearl, S. G. Sawochka: Improved Electrodes for B W R In-Plant ECP Monitoring, EPRI Report No. NP-2524.
Research Project 706-1, July 1982.
[ 5 ] L. W. Niedrach, W. H. Stoddard: The Development o f a High-Temperarure pH Electrode for Geothermal Fluids. Final Report-Task I. Report
No. PNL 3857 UC-66, Marz 1981. This report was prepared for the Pacific Northwest Laboratory, which is operated by the Battelle Memorial
Institute under contract for the United States Department of Energy,
Division of Geothermal Energy (prime contract No. DE-AC-06-76RLO-1830).
[6] M. J. Danielson, Corrosion (Houston) 35 (1979) 200.
Angew. Chem. I n f . Ed. Engl. 26 (1987) 161-169
[7] D. D. Macdonald, A. C. Scott, P. Wentrcek, J . Electrochem. SOC. 126
(1979) 908.
[Sl P. Andresen, M. E. Indig: Inoestigation ofBeneJ2s of B W R Oxygen Control. EPRI Report No. NP-2424-LD. Research Project T I 15-3, June
1982.
191 D. F. Taylor, C. A. Caramihas, J. Electrochem. SOC.I29 (1982) 2458.
[lo] R. S. Greeley, W. T. Smith, R. W. Stoughton, M. H. Lietzke, J. Phys
Chem. 64 (1960) 652, 1445.
[ I l l R. E. Mesmer, C. F. Baes, F. H. Sweeton, J Phys. Chem. 74 (1970)
1937.
[I21 J. V. Dobson, B. R. Chapman, H. R. Thirsk in R. W. Staehle, D. d e G.
Jones, J. E. Slater (Eds.): High Temperature. High Pressure Electrochemistry in Aqueous Solutions. National Association of Corrosion Engineers
(NACE), Houston, TX 1973, p. 341.
[I31 D. D. Macdonald, P. R. Wentrcek, A. C. Scott, 1. Electrochem. SOC.127
(1980) 1745.
[I41 J. V. Dobson, P. R. Snodin, H. R. Thirsk, Electrochim. Acta 21 (1976)
527.
[IS] L. W. Niedrach, Science 207 (1980) 1200.
[I61 L. W. Niedrach, J. Electrochem. SOC. 127 (1980) 2122.
[I71 T. Tsuruta, D. D. Macdonald, J. Electrochem. Soc. 129 (1982) 1221.
[IS] L. W. Niedrach, J . Electrochem. Sor. 129 (1982) 1445.
[19] S. Hettiarachchi, P. Kedzierzawski, D. D. Macdonald, 1. Electrochem.
SOC.132 (1985) 1866.
1201 M. J. Danielson, 0. H. Koski, J. Myers, J . Electrochem. SOC. 132 (1985)
2037.
[21] L. W. Niedrach, W. H. Stoddard, Ind. Eng. Chem. Prod. Res. Den. 22
(1983) 594.
[22] K. J. Vetter: Elektrochemische Kinetik. Springer, Berlin 1961, p. 16.
[23] L. W. Niedrach, W. H. Stoddard, J. Electrochem. SOC.131 (1984) 1017.
I241 L. W. Niedrach, W. H. Stoddard, Corrosion (Houston) 41 (1985) 45.
[25] L. W. Niedrach, W. H. Stoddard: Corrosion (Houston) 42 (1986) 696.
[26] R. L. Jones, A. Machiels, M. Naughton, J. T. A. Roberts, Corrosion '84.
Paper No. 167. March 1984.
169
Документ
Категория
Без категории
Просмотров
0
Размер файла
730 Кб
Теги
measurements, potential, electrode, temperature, high, pressure, aqueous, system
1/--страниц
Пожаловаться на содержимое документа