Corrosion Measurement Techniques

Download Report

Transcript Corrosion Measurement Techniques

Corrosion Measurement
Techniques
A copy of this presentation is available in
the CAL group in the computers in the
Teaching Lab, or via the WWW at
http://www.cp.umist.ac.uk/CPC/L_Notes
(c) Bob Cottis 1995
Corrosion Measurement
Techniques
Polarization curves
Linear
Polarization Resistance
Open Circuit Potential Decay
AC Impedance Measurement
Electrochemical Noise Measurement
Weight Loss Measurement
(c) Bob Cottis 1995
Polarization Curves
Measurement
Cell
design
Plotting
data
Interpretation
(c) Bob Cottis 1995
methods
Measurement Methods

Objective
– determine current density under steady-state
conditions as a function of potential
– not really practical, as this would strictly require
one sample for each potential
– therefore compromise on ‘closeness’ to true
steady-state
(c) Bob Cottis 1995
Measurement Methods
Connect electrodes to
corresponding terminals
 Potential control
on potentiostat
Reference Electrode Luggin
Probe - connection
allows
reference
for
Potentiostat
AE potential
potential
to be detected
measurement
close to metal surface
RE
WE
Counter Electrode (or
Auxilliary Electrode or
Working
Electrode
- Secondary
Electrode)
Potentiostat
controls
metal
being
studied
provides
current
path
potential
into solution
(c) Bob Cottis 1995
Measurement Methods

Current control
Potentiostat
Current path
V
AE
RE
WE
Counter Electrode
R
Reference
Electrode
Current
controlled
by Luggin
Probeacross
still
only
used
monitor
potential,
control
oftovoltage
Working
Electrode
to (I=V/R)
limit
IR error
notneeded
connected
to potentiostat
resistor
(c) Bob Cottis 1995
Measurement Methods

Swept potential or current
– Use sweep generator to produce slowly changing
potential
– Sweep generator output controls potentiostat
– Record response on chart recorder (or use
computer monitoring)
– Swept current not often used, as it moves through
corrosion potential very quickly
(c) Bob Cottis 1995
Measurement Methods

Potential or current step
– Step potential or current from one value to the
next, allowing time to stabilise at each new value
– Record current or potential
– May be manually controlled, or use computer to
step potential/current and take readings
(c) Bob Cottis 1995
Measurement Methods

Sweep direction
– Aim to perform experiment in such an order that
the initial polarization affects subsequent results
as little as possible
– Options
new specimen for each potential
 one specimen for cathodic polarization, and one for
anodic, both start at corrosion potential
 one specimen, sweep from cathodic to anodic

(c) Bob Cottis 1995
Measurement Methods

Sweep rate (or step rate)
– Ideal, all measurements made at steady-state
– Time-dependent effects include:
Charging of double layer capacitance (I = C dV/dt)
 Mass transport effects (t  L2/D)
 Adsorbed species and surface films (Faraday’s Law)

– Typical sweep rates are of the order of 1 mV/s or
less
(c) Bob Cottis 1995
Questions

Consider the corrosion of iron in aerated neutral
solution, with the following parameters:
–
–
–
–

Cdl = 35 F / cm2
DO2 = 1.2 x 10-5 cm2 /s
Boundary layer thickness,  = 100 m
Number of iron atoms on surface  21019/cm2
Charge on the electron = 1.6 x 10-19C
Calculate
–
–
–
–
Capacitive current at 1 mV/s
Characteristic diffusion time
Limiting current density for O2 reduction (8 ppm O2)
Time to oxidise Fe surface to FeOH (Fe+) at ilim
(c) Bob Cottis 1995
Cell Design
Working electrode
 Reference electrode
 Counter electrode
 Solution
 Mass transport

(c) Bob Cottis 1995
Working Electrode

Requirements
–
–
–
–
–
–
(c) Bob Cottis 1995
reproducible
representative
free of crevices
free of edge effects
free of galvanic effects
free of water-line effects
Working Electrode

Epoxy embedded electrode:
Apply thin layer of epoxy
Weld or solder connecting
to minimise stress and risk
Apply thick layer of epoxy
wire to specimen
of crevice
Pretreatformation
specimen for
to seal connecting tube and
good
adhesion
Carefully
grind
surface
for strength
to expose metal
Clean surface - don’t
use acetone
(c) Bob Cottis 1995
Working Electrode

Stern-Makrides
electrodes:
Metal rod
Retaining nut
Washers
Heavy-walled
glass tube
Lip seal
between
PTFE case
and electrode
PTFE Washer
Electrode
(c) Bob Cottis 1995
Working Electrode

Avesta cell:
Pure
H2O feed
NaCl
Solution
Specimen
(c) Bob Cottis 1995
Filter paper
Reference Electrode
Commonly use Saturated Calomel Electrode
(SCE)
 Properties may degrade with time (and
misuse)

– check one against another (should not be more
than 1 to 2 mV difference)
– do not pass current through the reference
electrode (e.g. do not connect to working or
counter electrode)
– do not allow to dry out
(c) Bob Cottis 1995
Reference Electrode

Solution in SCE (or Ag/AgCl electrode) is
saturated KCl
– beware of chloride contamination of test solution
by Cl- leaking from reference electrode
– make sure solution remains saturated
(c) Bob Cottis 1995
Luggin Probe
A Luggin probe should be used whenever
there is a significant current applied to the
electrode
Electrode

(c) Bob Cottis 1995
Luggin probe allows point at which potential
is measured to be close to electrode surface
(around 3 times tip diameter is best)
Counter electrode
Counter electrode should allow current to
pass with tolerable polarization
 Often claimed that counter electrode should
have much larger area than working
electrode, but this is not often necessary for
corrosion studies
 Usually use platinum or graphite, although
stainless steel can be used in some situations
(e.g. where only anodic polarization of
specimen is used)

(c) Bob Cottis 1995
Solution

Requirements:
– as high a conductivity as possible (add supporting
electrolyte, such as sodium perchlorate?)
– remain the same (pH, composition) throughout
the experiment - ensure that volume is adequate
– oxygen concentration often critical - aerate by
bubbling air or O2 or deaerate with N2 or Ar
– most reactions temperature sensitive, so control,
or at least record, temperature
(c) Bob Cottis 1995
Mass transport

Methods of controlling mass transport
–
–
–
–
(c) Bob Cottis 1995
rotating disk or cylinder
flow channel
jet impingement
gas bubbling
Plotting of Polarization Curves
Comparison of log-i and linear-i plots
 Identification of anodic and cathodic regions
on log-i plots
 Orientation of plots

(c) Bob Cottis 1995
E-i Plot
1.000E-01
8.000E-02
Fe anodic
H cathodic
6.000E-02
O2 cathodic
Net
Current Density
4.000E-02
2.000E-02
0.000E+00
-2.000E-02
-4.000E-02
-6.000E-02
-8.000E-02
-1.000E-01
-1.500
-1.300
-1.100
-0.900
-0.700
-0.500
Potential
(c) Bob Cottis 1995
-0.300
-0.100
0.100
0.300
0.500
E log |i| Plot
1.000E-01
Current Density
1.000E-02
Fe anodic
H cathodic
O2 cathodic
Net anodic
Net cathodic
1.000E-03
1.000E-04
1.000E-05
1.000E-06
-1.500
-1.300
-1.100
-0.900
-0.700
-0.500
Potential
(c) Bob Cottis 1995
-0.300
-0.100
0.100
0.300
0.500
E log |I| - old plotting method
E
log |i|
(c) Bob Cottis 1995
Interpretation of Polarization
Curves
Addition of reactions on log-I graphs
 Tafel regions
 Mass transport control
 Active-passive transition
 Transpassive corrosion
 Pitting Corrosion

(c) Bob Cottis 1995
Tafel regions
A Tafel region is a straight line in the
E-log|i| plot
 For a reliable Tafel slope:

– the line should be straight for at least one decade
(in this context a decade implies a change of current
density by a factor of ten, i.e a difference of 1 in
log i )
– the region should be next to Ecorr
(c) Bob Cottis 1995
E log |i| Plot
1.000E-01
Current Density
1.000E-02
Fe anodic
H cathodic
O2 cathodic
Net anodic
Net cathodic
1.000E-03
1.000E-04
1.000E-05
1.000E-06
-1.500
-1.300
-1.100
-0.900
-0.700
-0.500
Potential
(c) Bob Cottis 1995
-0.300
-0.100
0.100
0.300
0.500
Tafel Extrapolation
Extrapolate anodic or cathodic Tafel region,
or both, back to Ecorr, when the current
density is icorr
 In aerated neutral solutions, where mass
transport limited oxygen reduction is the
main cathodic reaction, the cathodic reaction
does not have a valid Tafel slope, but the
anodic slope can sometimes be used

(c) Bob Cottis 1995
Question
How can we estimate the rate of hydrogen
evolution during free corrosion?
 Estimate the value for the graph shown.

(c) Bob Cottis 1995
E log |i| Plot
1.000E-01
Current Density
1.000E-02
Fe anodic
H cathodic
O2 cathodic
Net anodic
Net cathodic
1.000E-03
1.000E-04
1.000E-05
1.000E-06
-1.500
-1.300
-1.100
-0.900
-0.700
-0.500
Potential
(c) Bob Cottis 1995
-0.300
-0.100
0.100
0.300
0.500
Mass transport control
When the supply of a reactant becomes mass
transport controlled, we observe a limiting
current density
 The most common case occurs for oxygen as a
cathodic reactant in neutral solutions
 NOTE - the diffusion of a reaction product
away from the electrode will not affect the
rate of the forward reaction

(c) Bob Cottis 1995
Solution Resistance Effects
At high currents the potential drop associated
with the solution resistance can be significant
 It is generally referred to as an IR error
 Gives a straight line on E-i plots

(c) Bob Cottis 1995
E log |i| Plot
1.000E-01
Current Density
1.000E-02
Fe anodic
H cathodic
Net anodic
Net cathodic
O2 cathodic
1.000E-03
1.000E-04
1.000E-05
1.000E-06
-1.500
-1.300
-1.100
-0.900
-0.700
-0.500
Potential
(c) Bob Cottis 1995
-0.300
-0.100
0.100
0.300
0.500
E-i Plot
2.000E-02
1.500E-02
Fe anodic
H cathodic
O2 cathodic
Net
Current Density
1.000E-02
5.000E-03
0.000E+00
-5.000E-03
-1.000E-02
-1.500E-02
-2.000E-02
-1.500
-1.300
-1.100
-0.900
-0.700
-0.500
Potential
(c) Bob Cottis 1995
-0.300
-0.100
0.100
0.300
0.500
Active-passive transition

As a passive film develops, it covers the
surface and shuts off the dissolution reaction,
leading to an active-passive transition
E
log |i|
(c) Bob Cottis 1995
Active-passive transition

For stainless steel we sometimes see two
active-passive transitions, on for Chromium,
and one for Iron
E
log |i|
(c) Bob Cottis 1995
Transpassive corrosion
A passive metal (notably Cr and Fe) may start
to dissolve at a very positive potential when a
higher oxidation state (e.g. Cr6+ as chromate)
is formed
 This is known as transpassive corrosion, and
will give something like a second activationcontrolled reaction
 For alloys the behaviour will be complicated
by the differing behaviours of the alloy
components

(c) Bob Cottis 1995
Anodic Polarization Curve for
Stainless Steel
E
Activation-controlled
Active-passive
Transpassive
Oxygen
Active peak
forcurve
iron
Overall
anodic
corrosion
dissolution
transition
reduction
of Cr
log |i|
(c) Bob Cottis 1995
Pitting Corrosion
Pitting shows up as an increasing anodic
current before (at a less positive potential
than) transpassive corrosion or oxygen
evolution, usually preceded by noise
 E-log|i| plot does not follow same path if
scan direction is reversed, but current is
greater (since pit continues to grow)

(c) Bob Cottis 1995
Pitting Corrosion
E
Current
continues
Noise
Pit eventually
spikes
due
re-to
to increase
after
meta-stable
passivates
pitting
reversal of scan
log |i|
(c) Bob Cottis 1995
What is going on?
Stainless Steel in Aerated Sulphuric Acid
Anodic
Cathodic
E
Anodic
Cathodic
log |i|
(c) Bob Cottis 1995
Linear Polarization Resistance
Measurement
Theoretical
basis
Measurement
Interpretation
(c) Bob Cottis 1995
methods
LPRM Theory

For an activation controlled reaction
 E  Eo 

i  io exp 
  
 E  Eo 
di io


 exp 
dE 
  
i


(c) Bob Cottis 1995
Exchange
Equilibrium
current
Tafel slope
based on
potential
density
exponential (i.e. mV
for a change of 1 in
ln(i))
LPRM Theory

Summing for two reactions
di ia ic


dE  a c
  a  c  1
 
 icorr 
  ac  R p

Anodic
partial
current
Because
partial
is taken
Anodic
Tafel
slope
Cathodic
slope
Cathodic
current
cTafel
Tafel
slope
based
on as
a
negative
(=i-i
))
(positive)
(negative)
density
(=
corr
decade
change
in
corr
current
(i.e. a change
Stern-Geary
of 1 in log i )
coefficient
Rearrange and convert to b rather than 
Rp 
(c) Bob Cottis 1995
B
icorr
,
ba bc
B
2.3ba  bc 
LPRM Measurement Methods
Control variable
 Waveform
 Cell configuration
 Sweep rate

(c) Bob Cottis 1995
LPRM Control Variable

Potential control
– potential range can be optimised
– problems with drift of Ecorr

Current control
– potential range depends on Rp
– measurement inherently centred about i = 0
(c) Bob Cottis 1995
LPRM Measurement Waveform

Triangle wave
– can measure di/dt at i = 0
– requires relatively complex instruments

Square wave (switch between +i and -i)
– simple instruments
– simple to automate

Sine wave
– simplest theory for frequency effects
– complex to perform measurement
(c) Bob Cottis 1995
LPRM Cell Configuration

Two electrode
– assume Rp is the same for two similar electrodes
and measure cell resistance (= 2Rp + Rsol)
– easy, no reference electrode required

Three electrode
– use conventional counter, reference and working
electrodes
– provides lower solution resistance, therefore
better for low conductivity solutions
– more complex instrumentation
(c) Bob Cottis 1995
LPRM Recommendations





Use three electrode measurement with triangle
waveform for laboratory studies
Use two electrode measurement with square
waveform for simple corrosion monitoring (use three
electrodes for high resistance solutions)
Use potential control when icorr variation is large
Use current control when Ecorr varies a lot
When both icorr and Ecorr vary use current control, but
adapt current to keep potential range reasonable
(c) Bob Cottis 1995
LPRM Interpretation

Determination of B value
– calculate from Tafel slopes
ba bc
B
2.3ba  bc 
– correlation with weight loss
– arbitrary value
26 mV for activation control
 52 mV for one reaction at limiting current

(c) Bob Cottis 1995
LPRM Sweep Rate
Must be sufficiently slow for current charging
double layer capacitance to be much less than
total current
 Characteristic time given by RctCdl - cycle time
should be at least 3 times this
 Need not be slow enough to allow diffusion
processes to respond (as the basic theory is
not valid for diffusion processes)

(c) Bob Cottis 1995
LPRM Problems

Theoretically, either
– both reactions must be activation controlled, or
– one reaction must be activation controlled and the
other mass-transport limited
In practice it is rare for real systems to meet
these constraints, and application of LPRM is
not theoretically justified
 Solution resistance adds to measured Rp, and
produces lower apparent corrosion rate

(c) Bob Cottis 1995
Equivalent Circuits
An electrical circuit with the same properties
as a metal-solution interface
 The simplest circuit is a resistor, Rct,
corresponding to the polarization resistance,
in parallel with a capacitor, Cdl, corresponding
to the double layer capacitance

Solution
(c) Bob Cottis 1995
Equivalent Circuits
An electrical circuit with the same properties
as a metal-solution interface
 The Randles equivalent circuit adds a series
resistor, corresponding to the solution
resistance

Rct
Rsol
(c) Bob Cottis 1995
Analysis of Solution Resistance

If we analyse the full response to the LPRM
measurement, we can estimate Rsol, Cdl and Rct
The
Estimate
voltage
Cdlacross
from the
Rsol is
exponential
given by Vdecay.
Thesoltime
Cdl)
oexp(-t/R
-1 (37%)
for V to
When
fall tot=eR
solCdl, of
the initial
V=V
value
is RsolCdl
oexp(-1)
i
Vo=iRsol
E
V=iRct
Time
(c) Bob Cottis 1995
Open Circuit Potential Decay

Similar to analysis of LPRM measurement
– charge double layer capacitance by applying a
current or potential
– disconnect charging current
– monitor decay of potential
(c) Bob Cottis 1995
Open Circuit Potential Decay
Charging at current i
Disconnected
Initial voltage drop = iRsol
E
Delayed
voltage drop
= iRct
Time = RctCdl
0.37iRct
Time
(c) Bob Cottis 1995
(c) Bob Cottis 1995