Argentum Solutions, Inc.
Sterling guidance on corrosion and materials degradation
| Potential-pH Diagrams |
 |
|
| Intelligent Tools |
 |
|
 |
|
 |
|
| Corrosion Calculator |
 |
|
| Corrosion Economics Estimator |
 |
|
|
TUTORIAL ON POLEXPERTTM AND THE CYCLIC POTENTIODYNAMIC POLARIZATION TECHNIQUE David C. Silverman |
|
Applying Tutorial Principles - Critique of ASTM F2129 The discussions of effects of experimental and environmental variables and features useful for interpreting polarization scans provide the background needed by the corrosion practitioner both to develop appropriate procedures and to critically examine procedures that have already been developed. The purpose of this portion of the tutorial is to show how the information in this tutorial might be applied to examining a procedure for generating a polarization scan to help to improve it. The example used is ASTM F2129-06 - "Standard Test Method for Conducting Cyclic Potentiodynamic Polarization Measurements to Determine the Corrosion Susceptibility of Small Implant Devices". The discussion implicitly assumes that the test environment provides a reasonable model of the actual working environment. Focus is on the following experimental variables and scan characteristics:
The corrosion potential is created by the potentials of all of the electrochemical processes occurring on the alloy surface. Use of the cyclic potentiodynamic polarization scan for corrosion prediction is built on the concept that predictions of behavior of a material in an environment can be made by forcing the material from its steady state interaction with the environment at a constant rate and observing how that material responds as the force is removed at the same constant rate and the material "relaxes" back to its steady state condition. The force in this case is the applied voltage. Features that emerge during this process are used for the prediction. In principle, for the cyclic potentiodynamic polarization scan to predict field performance (e.g. propensity for localized corrosion in the field) the scan should reflect one that would be generated on an electrode made of the same alloy after extended exposure to the field environment. The goal of the laboratory simulation is to get as close as possible. Potentials such as the pitting potential and repassivation potential or protection potential have little meaning in and of themselves. Only their values relative to the corrosion potential and to each other have meaning for prediction. Such predictive capability is demonstrated in the artificial neural network/expert system POLEXPERT. If the open circuit potential is not initially at steady state, its value does not reflect the processes governing the actual corrosion reactions but reactions occurring as the steady state corrosion process is approached. Under these conditions, the difference between the pitting potential or repassivation potential and what is thought to be the corrosion potential can be misleading. Two seemingly identical alloys subjected to what is thought to be the same surface treatment can approach steady state along different potential-time trajectories in the same environment. Reproducibility is affected because there is no guarantee that two seemingly identical alloys (especially alloys exhibiting strongly passive characteristics) would be in identical states after relatively short exposures to the same environment. With respect to the one hour immersion criterion, alloys of interest for medical devices tend to suffer very low general corrosion rates. These more passive type of alloys can require much more than 1 hour to reach a steady state corrosion potential. For example, the open circuit potential of 316ss immersed in salt water under aerated conditions has been shown to change for periods greater than 72 hours. The open circuit potential measured after one hour may not be the corrosion potential needed as a starting point for the polarization scan. If it is not, the calculation and interpretation of differences between the pitting or repassivation potential and what is thought to be the corrosion potential would be suspect. Some compromise may be required because reaching true steady state may not be consistent with completing experimental work in a reasonable amount of time. Reproducible results enabling training of the artificial neural network in POLEXPERT required between 12 and 24 hours to elapse between initial immersion and the start of the polarization scan for more passive systems. Using the difference between the repassivation potential and the initial corrosion potential or between the pitting potential and the initial corrosion potential for prediction implicitly assumes that if the open circuit potential could be measured while generating the scan:
ASTM F2129 provides two criteria for scan rate, 0.167 mV/s or 1 mV/s. If the scan rate is not chosen properly it can skew the polarization scan, detrimentally affecting its appearance and the usefulness its characteristics. The polarization scan should record an applied current that reflects the polarization resistance at each voltage. If the scan rate is too high, surface capacitance (e.g. double layer, oxide, etc.) can interfere resulting in a lower effective polarization resistance, a higher applied current. That increased current can affect the pitting potential and repassivation potential. The result can be a polarization scan that does not reflect the corrosion process. Inappropriate predictions and poor reproducibility can be the result. The reasons are outlined on page 4 of this tutorial that discusses experimental variables in more detail. That information, summarized below, provides the background for assessing scan rate choice. Electrochemical impedance spectra offer a framework for providing a gross estimate the appropriate scan rate. The assumption here is that such spectra can be modeled as containing one or two constant phase elements. If the exponent on the constant phase element is close to 1, the constant phase element can be treated as a capacitance. For simplicity the assumption is made here that the alloy-environment interaction can be modeled by a parallel resistor-capacitor combination in series with a resistor. The model is shown in this figure The pertinent information needed to estimate the appropriate scan rate can be found by focusing on the impedance magnitude versus frequency as shown in this figure One might assume that the polarization resistance for alloys with medical applications lies above at least105 ohm-cm2, most likely much higher. Otherwise, general corrosion might be an issue. The uncompensated solution resistance probably lies between about 10 and 100 ohm-cm2. Capacitances might be assumed to be in the range of 100 µ farad/cm2. Using the table in the section on scan rates as a guide capacitive interferences to the polarization resistance are negligible for scan rates in the range of 0.005 to 0.05 mV/s. The strong possibility exists that at 1 mV/s, the polarization scan could be reflecting effects of capacitance, not corrosion. If so, this scan rate would be inappropriately high. Such low scan rates as estimated above could be impractical because not only would the experimental duration be inordinately large but these low scan rates often cannot be achieved with commercial equipment. The scan rate of 0.167 mV/s would be, for the most part, reasonable from this analysis. Note that scan rates between 0.25 mV/s and 0.5 mV/s were used to generate the data trained in the artificial neural network used in POLEXPERT which included passivating iron, nickel, and titanium alloys. Such training suggests these slightly higher scan rates might be adequate as well enabling a shortening of the time to generate the polarization scan. The current density or potential at which the polarization scan is reversed can play a significant role in the appearance of the polarization scan and the value of the repassivation potential or protection potential. The reason is that the value of the repassivation potential is dictated by the damage to the surface created as the potential is scanned in the anodic direction from the corrosion potential. The standard provides three criteria for the point the scan is reversed (1) a current density two decades greater than the current at breakdown (which is not synonymous with though might include the pitting potential), (2) a potential that is 800 mV relative to a saturated calomel electrode or (3) a potential defined by certain attributes of the polarization scan. As shown by the example in the section on point of scan reversal, choosing the wrong point for reversing the voltage sweep can lead to an erroneous polarization scan and a poor prediction of long term behavior. This example is one of many that show that pushing electrodes into current and voltage regions in which the alloy surface is so changed that it no longer has any relevance to the expected corroding surface (e.g. forcing significant dissolution of chromium or extensive pits) can force interpretations that lead to errors in judgment. The reason is that how the alloy repassivates after being forced to corrode to such extremes may force the reverse scan to reflect phenomena that have no relevance to the application. No standard point of scan reversal can be proposed with complete confidence. Several points can aid in the choice.
4. Scan Features and Methodology Several additional criteria in the standard can be discussed using the principles outlined in this tutorial:
Non-completion of reverse scan The recommendation is made that if the polarization scan shows either a positive hysteresis or no hysteresis, the polarization scan can be stopped anodic to the corrosion potential at elevated potential or current. As shown by the training required for POLEXPERT, the position of the reverse scan, anodic-to-cathodic transition relative to the corrosion potential is a variable that was found to aid in the accuracy of the artificial neural network. In theory, that transition would be expected at the corrosion potential. But, in reality, changes in the surface structure during polarization and even the fact that the scan rate is not zero can cause that potential to be above or below (anodic or cathodic with respect to) the corrosion potential. The difference between this potential and the corrosion potential may provide an additional indication of persistence of passivity. If the polarization scan appears as in the figures "Polarization Scan for Passive Behavior", "Polarization Scan for Localized Corrosion", or "Polarization Scan for Oxidizable/Reducible Surface" the position of the anodic-to-cathodic transition relative to the corrosion potential provides information about the possibility of a passive film at the corrosion potential. The repassivation potential and the anodic-to-cathodic transition potential may be the same if the repassivation potential is considered the voltage at the point that the alloy surface no longer provides electrons. For alloys that can passivate either by a change in oxidation state (ferrous to ferric), a change in the passive layer (greater enrichment of chromium oxide, for example), or rely on alloy content and not necessarily a strong oxide layer for passivation (some higher nickel-based alloys), polarization to more noble potentials relative to the corrosion potential might place the surface in a more passive state than at the corrosion potential at least until the transpassive region is reached. The position of the corrosion potential relative to this anodic-to-cathodic transition provides information on the persistence of this passivity at the corrosion potential. The interpretation depends on where the anodic-to-cathodic voltage is read on the polarization plot. In POLEXPERT, it was read where the cathodic leg re-emerged at 0.1 µamp/cm2. The implication is that continuing the reverse scan until at least the anodic-to-cathodic transition is reached or to where its value can be read can provide additional insight for prediction. Location of "protection potential" An alternative (some consider more modern) descriptor for "protection potential" is "repassivation potential". This potential was originally meant to signify the point below which localized corrosion especially crevice corrosion would not occur. Pitting was assumed not to occur below the pitting potential. Experience has revealed that since this potential is dictated by a combination of the alloy-environment interaction and experimental variables (scan rate, current at scan reversal, etc.), some distance (e.g. 200 mV) should exist between the repassivation and corrosion potentials. That is, if the repassivation potential lies more than that distance (e.g. 200 mV) anodic or noble with respect to the corrosion potential, crevice corrosion is not suggested at the corrosion potential. Verification is always required by alternative experiments. The repassivation potential can lie cathodic to the corrosion potential when hysteresis is negative. Suggesting that the repassivation potential not be determined in this case can be misleading. One question is the point that should be called the repassivation potential. The standard uses the point at which the reverse scan crosses the anodic scan. This choice would be appropriate if crevice corrosion only occurred in the presence of negative hysteresis and the corrosion potential lied cathodic (active) relative to the repassivation potential. But, both the blue and red scans in the figure "Polarization Scan for Oxidizable/Reducible Surface" have been found with alloys suffering crevice corrosion when exposed using coupons with crevice forming washers. The main dictating factor seemed to be that the repassivation potential as defined in this picture was close to the corrosion potential and the anodic to cathodic transition as defined in this picture was anodic or noble with respect to the corrosion potential even though the hysteresis was positive. In this case, defining the repassivation potential or protection potential as the point at which the two curves crossed would have made interpretation and subsequent prediction difficult. Breakdown Potential Use of the term "breakdown potential" to signify the more rapid rise in current at elevated potentials is ambiguous. The phenomenon to which this potential refers in this standard has a number of causes. As shown in this figure for "Polarization Scan for Localized Corrosion", the potential is actually the pitting potential. As shown in this figure for a "Polarization Scan for Passive Behavior" or this figure for a "Polarization Scan for Oxidizable/Reducible Surface" the term would refer to the broad potential range over which the scan enters the transpassive region anodic with respect to the corrosion potential. If the potential is high enough as in the example on page 4 this region does signify destruction of the passive film but not a pitting potential. In many instances, no destruction occurs when this region is entered. Oxygen evolution cannot be assumed. Only when pitting occurs is the boundary sharp enough for a unique potential to be defined as the "pitting potential". Otherwise, modern usage tends to leave the point unnamed. The implication is that the term "breakdown potential" would be better replaced by more appropriate terminology. Previous Page: POLEXPERT (the intelligent prediction tool) |
|
David C. Silverman, Ph.D. - Primary Consultant |
){kind=link}
){kind=link}
){kind=link}
){kind=link}
){kind=link}
){kind=link}