DICMAPI Seminars 3 rd -4 th December 2014, Naples Practical impedance measurement Michele Curioni

DICMAPI Seminars 3 rd -4 th December 2014, Naples Outline How potentiostats / frequency response analysers work Issues – Non stationarity – Non linearity Meaning of Settings Choosing the best electrodes configuration

DICMAPI Seminars 3 rd -4 th December 2014, Naples Operational Amplifier IN + IN - OUT The amplifier is a device that outputs a current such as the voltage difference between IN+ and IN- is 0

DICMAPI Seminars 3 rd -4 th December 2014, Naples Current Measurement - Zero Resistance Ammeter (ZRA) IN + IN - OUT Rm (RANGE) I IN I OUT V=I IN x Rm

DICMAPI Seminars 3 rd -4 th December 2014, Naples Current Measurement - Zero Resistance Ammeter (ZRA) with Autorange IN + IN - OUT Rm1 I IN I OUT V=I IN x Rm Rm2 Rm3

DICMAPI Seminars 3 rd -4 th December 2014, Naples I IN I OUT V=I IN x Rm To data logger IN + IN - OUT Rm (RANGE) Cf (Bandwidth) Bandwidth

DICMAPI Seminars 3 rd -4 th December 2014, Naples Simplified Potentiostat Wk Ref Aux Cell IN + IN - OUT Control V

DICMAPI Seminars 3 rd -4 th December 2014, Naples Simplified Potentiostat Wk Ref Aux Cell IN + IN - OUT Control V=1V 0 V between these two points 1V

DICMAPI Seminars 3 rd -4 th December 2014, Naples Potentiostat with ZRA IN + IN - Rm (RANGE) Cf (Bandwidth) Measure I Wk Ref Aux Cell IN + IN - OUT Control V

DICMAPI Seminars 3 rd -4 th December 2014, Naples Realistic Potentiostat IN + IN - Rm (RANGE) Cf (Bandwidth) Measure I WkWk Re f Aux Cell IN + IN - OUT Control V IN + IN - OUT V buffer

DICMAPI Seminars 3 rd -4 th December 2014, Naples Bandwidth and Filters IN + IN - Rm (RANGE) Cf (Bandwidth) Measure I I Filtered I Filters ‘smooth’ the recorded signal Bandwidth determines how fast the machine responds to fluctuations RC Filter

DICMAPI Seminars 3 rd -4 th December 2014, Naples Frequency response analyzer (FRA) FRA physically performs multiplication and averaging Multiplier Integrator

DICMAPI Seminars 3 rd -4 th December 2014, Naples Potentiostat+FRA IN + IN - Rm (RANGE) Cf (Bandwidth) Measure I Wk Ref Aux Cell IN + IN - OUT Control V IN + IN - OUT V buffer

DICMAPI Seminars 3 rd -4 th December 2014, Naples Settings

DICMAPI Seminars 3 rd -4 th December 2014, Naples Solartron Modulab

DICMAPI Seminars 3 rd -4 th December 2014, Naples Solartron 1280

DICMAPI Seminars 3 rd -4 th December 2014, Naples Ivium

DICMAPI Seminars 3 rd -4 th December 2014, Naples Measurement Time Measurement time is primarily determined by the lowest value of frequency and averaging method. Secondarily, it is determined by the number of points per decade acquired Measurement time is important: if it is necessary to follow the time evolution of a corrosion process, the measurement time must be much shorter than the rate of change.

DICMAPI Seminars 3 rd -4 th December 2014, Naples Measurement Time 1 Point per decade ExpFreq (Hz)Period (s)N of averages Time x Point (s) Min Time (sec) Actual Time(sec) Points per decade ExpFreq (Hz)Period (s)N of averages Time x Point (s) E E Min Time (sec) Actual Time(sec) Some machines allow to reduce averages or points per decade at low freq.

DICMAPI Seminars 3 rd -4 th December 2014, Naples Issues During EIS Measurements

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non Stationarity (EN) I(t) Ref V(t) I(t) Cathodic Site Anodic Event Active Anodic Area Active Cathodic Area Electrolyte External Circuit

DICMAPI Seminars 3 rd -4 th December 2014, Naples Electrochemical Noise

DICMAPI Seminars 3 rd -4 th December 2014, Naples Corrosion Event and EIS FRA physically performs multiplication and averaging Multiplier Integrator

DICMAPI Seminars 3 rd -4 th December 2014, Naples Corrosion Event and EIS Applied V I Response I noise I measured FRA physically performs multiplication and averaging

DICMAPI Seminars 3 rd -4 th December 2014, Naples Corrosion Event and EIS Applied V I Response I noise I measured If this signal is multiplied and integrated the result is unreliable

DICMAPI Seminars 3 rd -4 th December 2014, Naples Corrosion Event and EIS Applied V I Response I noise I measured At high frequency the problem is not substantial

DICMAPI Seminars 3 rd -4 th December 2014, Naples Noise and Frequency Noisy outputOK For this reason the spectra for corroding system are acquired starting form HIGH frequency and finishing at LOW frequencies

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non stationarity

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non stationarity

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non stationarity

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non stationarity / Approaches to improve data Increase integration time – Experiment might become very long, no guarantee of success if intrinsically too noisy – Balance increased acquisition with reduced point/decade Increase minimum-frequency value – Simply avoid acquiring noisy part of the spectrum and keep experiment short Increase amplitude – Might be successful but might introduce corrosion at low freq and issues of non-linearity If unsuccessful better to use electrochemical noise

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non linearity Applied V Linear resp. Non Linear resp. Amplitude and Phase are obtained (not noisy) but they are not correct. Measured response contains harmonics

DICMAPI Seminars 3 rd -4 th December 2014, Naples Non linearity Behaviour above CANNOT be obtained by any combination of

DICMAPI Seminars 3 rd -4 th December 2014, Naples Approaches to reduce issue Reduce amplitude – Signal to noise ratio decreases – Might need to increase integration time Don’t use impedance/be aware of the limitation of the result Validate results with complementary approaches

DICMAPI Seminars 3 rd -4 th December 2014, Naples Electrodes Configuration 3-electrodes configuration 2-electrodes configuration Modified 2-electrodes configuration

DICMAPI Seminars 3 rd -4 th December 2014, Naples Potentiostat + FRA Pot.stat+FRA Working Electrode Reference 2 or Sense (not always present) Reference 1 Counter Electrode or Auxiliary Electrode

DICMAPI Seminars 3 rd -4 th December 2014, Naples 3-Electrodes Configuration Cell Pot.stat+FRA WE RE2 RE1AUX WE RE CE By using this configuration: -A constant DC potential is maintained between RE and WK, and the AC signal is superimposed to that DC potential -The signal is applied symmetrically with respect to that potential, regardless of the corrosion processes on the surface Consequently: -There must be confidence that the corrosion potential does not drift significantly during the measurement, otherwise significant faradic current is passed -If there is a drift of corrosion potential as a function of time, such drift is suppressed by the measurement.

DICMAPI Seminars 3 rd -4 th December 2014, Naples Examples of Potential Drift Pure Mg AA2024T3 AZ31 – Mg Alloy

DICMAPI Seminars 3 rd -4 th December 2014, Naples Potential drift/measurement time/ DC polarization (3-elec) AA2024T3 Bad OK

DICMAPI Seminars 3 rd -4 th December 2014, Naples AA2024T3 Bad Tragic Potential drift/measurement time/ DC polarization (3-elect)

DICMAPI Seminars 3 rd -4 th December 2014, Naples AA2024T3 Possible Solution (3-elect) Impedance Measurement Free corrosion

DICMAPI Seminars 3 rd -4 th December 2014, Naples 2-Electrodes Configuration Cell Pot.stat+FRA WE RE2 RE1AUX WE1 WE2 By using this configuration: -A constant DC potential (generally 0 V) is maintained between WK1 and WK2. If one uses identical electrodes for WK1 and WK2, maintaining 0V between them does not induce additional corrosion -The signal is distributed among the two electrodes, therefore if asymmerty develops, the impedance of the non- corroding electrode dominates (electrodes are in series) Consequently: -A slow drift in potential is not an issue -The natural potential drift is not perturbed by the measurement -The anodic current on one electrode equals the cathodic current on the other

DICMAPI Seminars 3 rd -4 th December 2014, Naples 2-Electrodes Conf By using this configuration: -A constant DC potential (generally 0 V) is maintained between WK1 and WK2. If one uses identical electrodes for WK1 and WK2, maintaining 0V between them does not induce additional corrosion -The signal is distributed among the two electrodes, therefore if asymmetry develops, the impedance of the non- corroding electrode dominates (electrodes are in series) Consequently: -A slow drift in potential is not an issue -The natural potential drift is not perturbed by the measurement -The anodic current on one electrode equals the cathodic current on the other. 0 V WE 1 WE 2

DICMAPI Seminars 3 rd -4 th December 2014, Naples Modified 2-Electrodes Configuration Cell Pot.stat+FRA WE RE2 RE1AUX WE1 WE2 By using this configuration: -A constant DC potential (generally 0 V) is maintained between WK1 and WK2. If one uses identical electrodes for WK1 and WK2, maintaining 0V between them does not induce additional corrosion -The signal is distributed among the two electrodes, therefore if asymmerty develops, the impedance of the non- corroding electrode dominates (electrodes are in series) Consequently: -A slow drift in potential is not an issue -The natural potential drift is not perturbed by the measurement -The anodic current on one electrode equals the cathodic current on the other Voltmeter

DICMAPI Seminars 3 rd -4 th December 2014, Naples

Summary Potentiostats/FRAs are analogue device based on operational amplifier. It is useful to understand basic functioning principle to determine settings in different machines Non-stationarity and non-linearity can be issues during impedance measurement. Settings can be adjusted to reduce/minimize issues in some circumstances Electrodes configuration is important: – 3-electrodes configuration imply fixed potential – 2- electrodes configuration does not imply fixed potential, but requires the assumption of identical behaviour