1 The ECN module. Note

Transcription

1 Version NOVA ECN tutorial 1 The ECN module The ECN is an optional module for the Autolab PGSTAT. The ECN module provides the means to perform Electrochemical Noise measurements (ECN). Electrochemical noise corresponds to seemingly random fluctuations in current and potential generated by stochastic phenomena occurring at the electrochemical interface. The fluctuations of potential and current signals that arise directly from the electrochemical reactions, taking place on the electrode surface, can be measured using the optional ECN module. The ECN measurement is non-invasive because no external perturbation is applied and the current is monitored through a so-called Zero Resistance Ammeter (ZRA). ECN can be used to monitor localized corrosion (pitting), uniform corrosion through measurement of the Noise Resistance, and the deterioration of paints on metal substrates. The same technique can also be used to monitor galvanic coupling in the presence of an electrolyte. Electrochemical noise measurements are also possible without the use of the dedicated ECN module. However, the resolution that can be achieved with the ECN is significantly better than without the use of this module. Under normal operating conditions, the potential resolution for an ECN measurement is ~ 30 µv without the ECN module and as low as 0.76 µv with the ECN module. Scope of the tutorial The aim of this tutorial is to explain how to perform electrochemical noise measurements, with or without the dedicated ECN module. The settings of the ECN module are explained and example procedures for ECN measurements are provided in the tutorial. 1 Page

2 2 Hardware setup In order to use the ECN module, the hardware setup in NOVA must be configured accordingly (see Figure 1). Figure 1 Selecting the ECN module When electrochemical noise measurements are performed without the use of the dedicated ECN module, no additional modules need to be selected. 2 Page

3 3 Connections to the Autolab When electrochemical measurements are performed in combination with the Autolab, a special connection scheme has to be used. Depending on the hardware configuration, two possible connections to the cell can be used. 3.1 Connections with the ECN module When the ECN module is used, a special cable provided with the module must be used to connect the module to the electrochemical cell (see Figure 2). Figure 2 The special ECN cable provided with the module The first working electrode is connected to the WE banana connector provided by the PGSTAT and the red connector from the ECN cable. The black banana plug from the special ECN cable is connected to the reference electrode. Additionally, a connection between the ground of the potentiostat and the second working electrode is required 1 (see Figure 3). 1 This connection can be established using the ground connector on the front panel of the instrument, as shown in Figure 3 or using the ground cable embedded into the cell cable, if applicable. 3 Page

4 E on ECN RE, S and CE from PGSTAT not used Reference electrode Working electrode 1 Working electrode 2 Figure 3 Overview of the connections when using the ECN module The RE, S and CE provided by the Autolab PGSTAT are left disconnected. 4 Page

5 3.2 Connections without ECN module The alternative setup does not use the ECN module, but the connections provided by the Autolab. The WE and S banana connector from the potentiostat are connected to the first working electrode. The RE banana plug is connected to the reference electrode and an additional ground cable is connected to the second working electrode 2 (see Figure 4). CE from PGSTAT not used Working electrode 2 Working electrode 1 Reference electrode Figure 4 Overview of the connections when the ECN module is not used The CE connection provided by the Autolab PGSTAT is left disconnected. 2 This connection can be established using the ground connector on the front panel of the instrument, as shown in Figure 4 or using the ground cable embedded into the cell cable, if applicable. 5 Page

6 3.3 Measuring the ECN signal The ECN module provides one signal, ECN(1).Potential, that can be selected using the Signal sampler (see Figure 5). Figure 5 The signal provided by the ECN is available in the sampler When electrochemical noise measurements are done in combination with the ECN module, the potential difference between the reference electrode and working electrode #1 will be sampled through the ECN module. This signal corresponds to ECN(1).Potential. When electrochemical noise measurements are performed without the ECN module, the potential difference between the reference electrode and working electrode #1 can be sampled through the WE(1).Potential signal. The current is measured as WE(1).Current regardless of the experimental setup. The ECN module has an additional amplifier (four times) and an offset DAC on board. This means that ECN measurements with the ECN module can be performed at the highest resolution. In gain 100, this will be as low as 0.76 µv with the ECN, compared to 30 µv without this module. 6 Page

7 3.4 Experimental considerations ECN measurements are always performed without any applied potential or impressed current. For this reason it is important to leave the CE connector from the Autolab disconnected at all times. The measurements must also be performed at open circuit (cell off). The ECN module has an extra offset compensation DAC on-board which can be used to compensate the DC potential measured during an experiment (see Figure 6). Front panel 12 Bit offset DAC ADC164 E 4 x amplification Figure 6 The ECN module has a 12 Bit offset compensation DAC on-board This means that the potential noise EE nnnnnnnnnn will be measured around a compensated potential, EE cccccccc, given by: EE cccccccc = EE nnnnnnnnnn oooooooooooo DDDDDD(VV) 7 Page

8 When the offset DAC is used, the potential measured by the ECN module can be set close to 0 V and maximum amplification of the signal can be used, resulting in a resolution of 0.76 µv in gain 100. When the offset DAC is not used, only gain 1 or gain 10 can be used in practice. The gain is automatically selected by the ADC164. Gain 1 is used when the measured signal exceeds ± 1 V and gain 10 is used when the measured signal exceeds ± 100 mv. For values below 100 mv, gain 100 can be used. The input range of the ECN module is ± 2.5 V. The offset potential is set through the Autolab control command (see Figure 7). Figure 7 The Autolab control command can be used to set the offset DAC value during a measurement 8 Page

9 4 Current range restrictions Current measurements during ECN experiments must be sampled using a so-called Zero Resistance Amperometer (ZRA). The Autolab PGSTAT provides this type of circuit for measuring the current, but only for the current ranges below 1 ma. Automatic current ranging is therefore possible only up to the 100 µa current range. The procedure validation process will display an error message when current ranges higher than 100 µa are used in an experiment in combination with the ECN module (see Figure 8). This error cannot be ignored and the procedure cannot be started. Figure 8 An error message is displayed by the procedure validation when current ranges higher than 100 µa are used in combination with the ECN module This error situation can only be detected when the ECN module is used. When ECN measurements are performed without the ECN module, this error will not be displayed. It is therefore important to keep the current range restriction in mind at all times. 9 Page

10 5 Cell switch restriction ECN measurements, regardless of the experimental setup, must always be carried out with the cell switched off. During validation, the cell status is checked. If the cell is not switched off, a warning will be displayed in the validation report (see Figure 9). Figure 9 A warning is displayed during validation if the cell is not switched off during an ECN measurement It is possible to continue the measurement, but the data could be invalid. 6 ECN measurements In practical situations, ECN measurements are carried out in time. The duration of the experiment and the interval time depends on the experimental conditions, but usually a small interval time is used in order to measure the high-frequency noise contributions in the system. Typically, the measurements can be performed using the Record signals (>1 ms) command. With this command, current and potential transients can be recorded and displayed real-time, with an interval time down to 1.33 ms. Faster transients can be recorded using the chrono methods command 3. When the measurement is finished, the data is analyzed using dedicated analysis tools. 3 Real-time plotting and automatic current ranging is not possible when using the Chrono methods command. 10 Page

11 7 ECN data analysis Electrochemical noise data is generally analyzed by computing the spectral density of the measured data. This can be achieved by transforming the time domain information to a frequency domain spectrum, using a Fast Fourier Transformation (FFT). Traditional FFT analysis assumes that the data outside of the measured time segment is either zero or that the data in this segment repeats periodically. This hypothesis is not valid for electrochemical noise data. In order to satisfy these requirements and to avoid edge effects in the data, it is common practice to apply a windowing function on the time domain data. This calculation involves the multiplication of the time domain data by a function which is zero at the extremes of the time domain data and rises smoothly to unity value in its centre. A large number of functions are available in the literature 4. In NOVA, five different windowing functions are available: Square Bartlett Hanning Hamming Blackman An alternative method known as the Maximum Entropy Method (MEM) can also be used. This method does not require any assumption concerning the data outside of the measured time segment. This method can be used to compute the spectrum most consistent with the available data. The MEM requires the user to specify a number of coefficients. The higher the number of coefficients, the noisier the computed spectrum will become. Furthermore, it is common practice to calculate statistical indicators like the noise resistance, localization index (or pitting index), skewness and kurtosis. The noise resistance, RR nn is derived from the standard deviations of the measured potential and current, σσ VV and σσ ii : RR nn = σσ VV σσ ii 4 W. H. Press, S. A. Teukolsky, W. T. Vetterling, B. P. Flannery, Numerical Recipes The Art of Scientific Computing, 3 rd edition, Cambridge University Press, Page

12 The pitting index, or localization index, is derived from the standard deviation of the measured current, σσ ii and the root mean squared value of the current, ii RRRRRR : PPPP = σσ ii = ii RRRRRR NN jj=1 NN jj=1 ii 2 jj ii ii jj 2 The values of this indicator can be between 0 and 1. A value close to 0 is observed for systems in which the measured current values show only small deviation with respect to the average current value. On the other hand, the pitting index will be close to 1 when the individual current values are significantly deviating from the average current value. This value is therefore an indication of the distribution of the current values recorded during the ECN experiment. Kurtosis and skewness are defined as: Kurtosis = 1 NN NN XX ii XX ii=1 Skewness = 1 NN NN XX ii XX ii=1 σσ 4 σσ 3 These two factors characterize some aspects of the distribution of the ECN data. For a normal distribution, Kurtosis has a value of 3. Larger values indicate a sharper distribution, whereas smaller values are indicative of a flatter distribution. Skewness characterizes the symmetry of the measured data around the average value. A value of 0 indicates that the distribution is symmetrical about the mean value. A positive or negative value indicates that the distribution is tailing towards larger values or smaller values, respectively (see Figure 10). Figure 10 - Skewness is an indication of the symmetry of the distribution (left negative skewness, right positive skewness) For more details on the analysis of electrochemical noise, the reader is invited to consult the literature on the subject since the discussion of these methods falls outside of the scope of this tutorial. More information on the mathematical background of these methods can be obtained on request. 12 Page

13 To analyze ECN data using the tools described in this section, right-click the data in the analysis view and select the ECN Spectral noise analysis from the context menu or click the button in the quick access toolbar (see Figure 11). Figure 11 Adding the ECN spectral noise analysis tool to ECN data using the right-click menu (top) or using the quick access toobar (bottom) A new item will be added to the data (see Figure 12). Three power spectrum density (PSD) plots are automatically generated, for potential, current and impedance, respectively. The different statistical indicators derived from the experimental data are also automatically generated and displayed below the plots (see Figure 12). Figure 12 The ECN spectral noise analysis tool generates power spectrum density plots and calculates the statistical indicators: Skewness and Kurtosis for Potential, Skewness and Kurtosis for Current, Noise resistance and Pitting index 13 Page

14 The PSD plots are, by default, obtained with the FFT method. To change these settings, click the ECN Spectral noise analysis item in the data explorer. The control parameters of this analysis tool will be displayed in the dedicated frame on the righthand side of the plot (see Figure 13). Figure 13 The ECN Spectral noise analysis control panel is displayed on the right-hand side The following parameters can be adjusted in the control panel (see Figure 14): Figure 14 Detailed view of the ECN Spectral noise analysis parameters Subtract baseline: specifies if the baseline should be subtracted from the measured data. When this option is checked, a linear regression is used to subtract the baseline from the measured potential and current values. Window function: specifies the type of windowing function used for the FFT or the MEM power spectrum density calculation (the square window is selected by default). Noise analysis method: specifies the technique used in the calculation of the power spectrum density plots (FFT or MEM). MEM coefficients: specifies the number of coefficients to be used in the MEM calculation. The default value is 20 and this parameter only applies to the MEM method (this parameter is greyed out for the FFT method). Figure 15 shows typical power spectrum density (PSD) plots obtained using the FFT technique on the data shown in Figure 13. The Square window is used and the baseline is subtracted. 14 Page

15 Figure 15 - Typical PSD plots (blue plot potential, red plot current) obtained using the FFT method with the data shown Figure 13 The PSD data is displayed as a log/log plot. The values shown on the Y axis are expressed as amplitudes, in V 2 or A 2 per Hz. 15 Page

16 The impedance PSD plot shows the ratio of the potential PSD and the current PSD (see Figure 16). Figure 16 The impedance PSD plot obtained using the values shown in Figure 13 The impedance value observed in Figure 16 at the low frequency limit is close to the value of the noise resistance shown in Figure Page

17 8 An ECN measurement on the dummy cell An ECN tutorial folder is located in the Program Files\Metrohm Autolab\Nova 1.11\Shared Databases\Tutorials folder (see Figure 17). Using the database manager, set the ECN folder as the Standard database. Figure 17 Loading the ECN tutorial database The ECN tutorial contains two procedures (see Figure 18). Figure 18 The two ECN tutorial procedures 17 Page

18 8.1 ECN tutorial 1 ECN measurement using the ECN module Load procedure ECN tutorial 1 Electrochemical noise measurement with ECN module in the procedure editor (see Figure 19). Figure 19 Overview of the ECN tutorial #1 procedure The procedure itself measures the ECN(1).Potential and the WE(1).Current according the following sequence: 1. The ECN offset DAC value is reset to 0 V through the Autolab control command 2. The ECN(1).Potential is sampled for 1 s, using an interval time of 0.1 s 3. The average value of the ECN(1).Potential is calculated 4. The Average potential noise value is used a the new offset DAC value for the rest of the measurement 5. The potential and current noise signals are measured in the second timed procedure. that ECN(1).Potential is measured around 0 V because the offset is compensated by the onboard DAC. 6. At the end of the measurement, the correct ECN(1).Potential is recalculated and plotted on the graph. This measurement can be performed on the Autolab dummy cell. Connect the cell cables from the PGSTAT and from the ECN module as shown in Figure 20. This connection scheme corresponds to the one displayed in Figure Page

19 To E ECN CE R2 100Ω C1 R4 1µF WE(+S) (a) 1MΩ RE R1 100Ω C2 1µF R3 WE(+S) (b) WE R7 10kΩ 1MΩ C3 1µF To E ECN WE(+S) (e) R5 WE(+S) (c) 1kΩ C4 R6 1µF 5kΩ WE(+S) (d) DUMMY CELL2 Figure 20 The connections used for the ECN Tutorial #1 procedure Make sure that the CE, RE and S from the PGSTAT are disconnected. When the measurement starts, a message box is displayed (see Figure 21). Figure 21 A message is displayed at the beginning of the measurement 19 Page

20 Click OK to continue. The measurement will proceed for about a second and the ECN(1).Potential will be measured. The measured values will be used to determine the average value and the corresponding offset DAC value. When this is done, a second message box will be displayed (see Figure 22). Figure 22 A second message is displayed when the offset DAC has been set Click OK to continue the measurement. The current and potential will then be measured for 60 s and plotted in the measurement view. When the measurement is finished, the correct ECN potential will be recalculated and plotted on the same graph. Figure 23 shows a typical plot obtained with the Autolab dummy cell (a). The measurement is performed with cell off at all times. Figure 23 The measurement data obtained on the dummy (Blue curve: WE(1).Current; Red curve: ECN potential compensated by the offset DAC; Green curve: the recalculated ECN potential) 20 Page

21 The green curve corresponds to the recalculated ECN potential while the red curve corresponds to the compensated ECN potential actually measured. Because the offset DAC is used to compensate the DC component of the ECN potential, the highest possible resolution can be achieved. Figure 24 shows a detail of the measurement shown in Figure 23 that the resolution in this measurement can be as low as 0.76 µv. Figure 24 The resolution is 0.76 µv in gain 100 When this measurement is performed on real samples, the noise fluctuations are quite large and an amplification of 100 is only possible if the measured signals remain within a limit of ± 100 mv. Figure 25 shows a practical example obtained with a stainless steel (SS316) electrode and an aluminium electrode in 1 M NaCl solution at room temperature. The reference electrode used was a Ag/AgCl (KCl Sat d). The ECN potential is around -0.8 V, the compensated ECN potential fluctuates around 0 V. 21 Page

22 Figure 25 A practical example obtained with the ECN module In this measurement, a resolution of 7.6 µv was obtained in the noise measurement. 8.2 ECN tutorial 2 ECN measurement using the PGSTAT Load procedure ECN tutorial 2 Electrochemical noise measurement with PGSTAT in the procedure editor (see Figure 26). Figure 26 Overview of the ECN tutorial #2 procedure In this procedure, the potential will be measured by the differential amplifier of the PGSTAT rather than the ECN module. This measurement can be performed on the Autolab dummy cell. Connect the cell cables from the PGSTAT as shown in Figure 27. This connection scheme corresponds to the one displayed in Figure Page

23 CE R2 100Ω C1 R4 1µF WE(+S) (a) 1MΩ RE R1 100Ω C2 1µF R3 WE(+S) (b) WE S RE R7 10kΩ 1MΩ C3 1µF WE(+S) (e) R5 WE(+S) (c) 1kΩ C4 R6 1µF 5kΩ WE(+S) (d) DUMMY CELL2 Figure 27 - The connections used for the ECN Tutorial #2 procedure When the measurement starts, a message box reminding the connections to be used in the measurement is displayed (see Figure 28). Figure 28 A message is displayed at the beginning of the measurement Click OK to continue. The measurement will start and the potential and current will be monitored during 60 seconds and plotted in the measurement view. Figure 29 shows a typical plot obtained with the Autolab dummy cell (a). 23 Page

24 Figure 29 The measurement data obtained on the dummy (Blue curve: WE(1).Current; Red curve: WE(1).Potential) The potential values recorded through the PGSTAT show a considerable lower resolution when compared with the data shown in Figure 23. In this set of data, the resolution of the potential signal is 3 µv (measured in gain 100). When this measurement is performed on real samples, the noise fluctuations are quite large and an amplification of 100 is not possible, unless the potential value is very close to 0 V. This means that the practical resolution will be 30 µv, in gain 10 and 300 µv in gain 1. Figure 30 shows a practical example obtained without the use of the ECN module with a stainless steel (SS316) electrode and an aluminium electrode in 1 M NaCl solution at room temperature. The reference electrode used was a Ag/AgCl (KCl Sat d). The potential is around -0.8 V. 24 Page

25 Figure 30 A practical example obtained without the ECN module This tutorial has illustrated the use of the ECN module for noise measurements and an alternative setup without the ECN module. The benefits of the ECN module are the additional 4 times amplification of the signal and the presence of an offset DAC which ensures that the signals are always measured with the highest possible resolution. 25 Page

26 Hardware specifications The ECN module is an optional extension of the Autolab PGSTAT. It provides the means to perform electrochemical noise measurement and provides the highest possible potential resolution. The ECN module is compatible with all the Autolab PGSTAT instruments with the exception of the PGSTAT100, PGSTAT100N and PGSTAT302F and the non-modular instruments (PGSTAT101, PGSTAT204, µautolab II and III). The ECN module is also not compatible with the Multi Autolab instruments. ECN measurements are also possible without the ECN module, but using this dedicated module a practical improvement by a factor of 40 or 400 can be achieved for the measured potential resolution. Table 1 provides an overview of some specifications of the ECN module and the alternative specifications that can be obtained with using the Autolab PGSTAT to perform these measurements. Hardware ECN module PGSTAT Input range ± 2.5 V ± 10 V Maximum potential resolution 760 nv (Gain 100) 30 µv (Gain 10) Potential offset compensation Yes No Potential accuracy 300 µv 150 µv Table 1 Overview of the specifications of the ECN module and the alternative PGSTAT Warning Electrochemical noise measurements are always performed with the cell switched off. 26 Page