Q&A for CorroZoom Webinar by G.S. Frankel
- I agree with your assessment in terms of critical steps that lead to pitting in a single phase alloys. But, in the case of a multiphase material in which pit initiation may occur at a phase boundary, how can the CPT be treated as a material characteristic? Is it the matrix material, the second phase or the phase boundary that is critical? (Question by Ron Latanision)
Answer: In this model, we made some simplifications. We are ignoring the initiation sites and the phases. But at a small scale, the model should work fine, especially regarding to the metastable to stable transition. We have shown in the talk that the CPT cannot be considered as an intrinsic material property. It is more like the temperature where the first stable pit initiate. For a multiphase alloy, the CPT is more determined by the most susceptible microstructural features, such as the inclusions.
- What was the surface roughness of SS304 vs SS316 for the polarization curve slide? And what was the electrolyte in this case? (Question by Vladimir Totolin)
Answer: Both samples were polished to 1200 grit, and the test solution was 0.6 M NaCl.
- What is the bump in the pencil electrode curves? (Question by Jamie Noel); What is the peak before Esat is reached for small pits due to? (Question by James Earthman)
Answer: The current peak at the transition point is generated by the supersaturation created by the thinning of the salt film during the downward scan. The supersaturation level increases with increasing downward potential scan rate, thus the aggressiveness of the pit solution increases. As a result, the amplitude of the peak also increases with increasing scan rate. Detailed explanation can be found in Part V of our series of papers: https://doi.org/10.1149/2.0431911jes
- Why is the Tafel slope for dissolution (116 mV/decade) so high? (Question by Rob Kelly)
Answer: When plotting the downward scan curve in semi-log format in the charge transfer region, we will obtain a much lower Tafel slope. However, under that condition, concentration of the pit solution at the pit surface is not constant, and it will keep decreasing during the downward scan. The decrease of concentration decreases the aggressiveness of the pit solution, thus it accelerates the decrease of the pit current during the downward scan and a lower apparent Tafel slope is observed. In our work, the Tafel slope was obtained at a constant concentration (saturated) and the effect of changing pit environment was excluded, thus the Tafel slope is higher.
- Have you considered the effect of flow rate on these pitting critical parameters? (Question by Maryam Eslami)
Answer: The flow rate influences the diffusion process, thus it will influence critical parameters proposed in our model. However, this issue could be handled by making some simple modifications.
- Is there any experimental phenomenon that cannot be well explained by this new framework? (Question by Yangting Sun)
Answer: We believe that many aspects of pitting corrosion can be explained by our framework. However, we do not address the critical issues associated with metastable pit initiation. Also, some systems, like iron in alkaline solutions, exhibit behavior that is not typical for stainless steels, such as a downward scan that is neither positive or negative hysteresis. Instead, the current on the downward scan might exhibit what might be called mixed hysteresis, crossing the upward scan several times. This is hard to explain with our framework.
- What about the MnS inclusions, that often control Epit? (Question by Dan Blackwood) Do inclusions impact on (metastable) pit formation? (Question by Matthias Schulz)
Answer: MnS inclusions can influence the pit initiation rate, and their size can influence the size of metastable pit, thus they also influence the transition of metastable pit to stable pit. Reducing the density and the size of inclusions can improve the pitting resistance.
- How would surface roughness affect these experiments? (Question by Mohammad Umar Farooq Khan)
Answer: For a 1D pit experiment, the surface roughness is not an important factor. We only need to pre-dissolve the 1D pit exceed a certain pit depth to create 1D diffusion, which is about 2 times the diameter. For potentiodynamic polarization tests on bulk sample, the pitting potential usually increases with decreasing of the surface roughness, which probably can be explained by the effect of the surface roughness on the effective diffusion length of the pit.
- The model does not appear to ascribe any role to oxide formation at the bottom of the pit i.e. the dissolution rate at the bottom of the pit is purely active dissolution. Would this model be able to explain the role of inhibitive species in the solution e.g., nitrate or nitrite? According to this model, nitrate would have to inhibit dissolution rate at the bottom of the pit. Is there an experimental evidence for this? (Question by Narasi Sridhar)
Answer: Oxide formation at the pit bottom is important and plays a role, but only at the Ccrit. When the concentration is above the Ccrit, the passive film is unstable and is not able to form on the pit surface, so we only need to consider the dissolution at the pit surface. Additionally, our model uses the Ccrit as the basic criterion to identify the status of pit surface (active or passive). We treat the oxide formation as the consequence of Csurf falling below the Ccrit, but we do not deal with the detailed oxide formation process. Instead, we deal with the issue of the conditions at which Csurf can be maintained at or the Ccrit, which is why we come up with the concept of idiss,max to deal with the competition between dissolution and diffusion. The inhibition effect of species, such as nitrate and nitrite, can be evaluated using idiss,max, which we have already studied and will be reported on in the future.
- Would it be good to study the effect of salt purity on these studies? (Question by Mohammad Umar Farooq Khan)
Answer: Based on our cryo-salt film work, the salt film on a pit in SS304 is a mixture of cations salt with a single crystal structure. Ni cations seem to be able to substitute for the host Fe cations in the salt crystal. At steady state, the salt purity should be determined by the composition of the alloy. It is not possible to change the salt purity without changing the alloy composition. But the purity and the porosity of the salt film may influence the potential field across the film, which may influence the thickness of the salt film at given pit depth.
- Some film and salt film may be compact enough, some maybe not and very easy for ions diffusion and other elements diffusion out. Have you found any difference for various materials, such as SS316, SS304, Pure Fe, etc.? (Question by En-Hou Han)
Answer: This cryo salt film experiment is very challenging, and we haven’t performed it on other alloys yet.
- Jerry, can this framework be used to address Cl-SCC? (Question by Earl Johns)
Answer: We are working on this now.
- Preferred, because kinetically stronger, is the chromium dissolution (chromium -> iron -> nickel). What is your opinion on a thesis that this faster dissolution of chromium and iron results in an enrichment of nickel in the hole bottom? This was proven (I think) by Elsener by means of XPS in the pit bottom. (Question by Andreas Heyn)
Answer: This might be true if the pit growth is under charge transfer control. But if the pit growth is under diffusion control, the dissolution of the metal is actually suppressed with the presence of the salt film, and the dissolution of different alloy elements should be congruent (proportion to the composition) under this condition.
- If I combine your insights with those of Nick Birbilis, to what extent would your model(s) be applicable for the prediction of pit growth behaviour of additive manufactured materials? Given the fact that after 3D printing, we often have very different microstructures with possibly much less, or otherwise different, precipitates. (Question by Axel Homborg)
Answer: For sure, this framework can be used to explain some phenomena exhibited by 3D printed alloys. We are also working on this now.
- Can also Hydrogen evolving to contribute to SCC or HE? (Question by Javier Sanchez)
Answer: Yes, it can.
- What is meant by electro polished surface at the pit shown with salt film? Was the surface electro polished the experiment or was the material polished after the experiment? (Question by Julien Oltze)
Answer: When a salt film precipitates on the pit surface, the metal dissolution inside the pit is under diffusion control. Under this condition, if there is a protrusion on the pit surface, the effective diffusion length at this position will be shorter than other areas, the dissolution rate will be higher, and the protrusion will dissolve back down to create a smooth pit surface, which looks like it was polished. In contrast, if the pit grows under charge-transfer-control, the metal dissolution rate will be influenced by the crystal orientation, thus a crystallographic morphology will be created. This occurred during the experiment, not afterwards.
- Jerry, you mentioned stable pitting as a predecessor for crack formation; is the implication that metastable pits are typically not geometrically conducive to initiate cracking? (Question by William Weimer)
Answer: According to fracture mechanics principles, stress intensity will increase with defect size. It is unlikely that the micron sized hole left by a metastable pit can act as an effect stress concentrator needed for crack initiation.
- What is the smallest stable pit according to your calculations in stainless steel? Would this model fit also to duplex stainless steel? (Question by Cem Örnek)
Answer: The smallest pit depth required for stable growth or the critical pit depth for pit growth stability (rcrit) depends on the temperature and applied potential. The calculation of this parameter also needs the input of the several other parameters, such as Ccrit and ∆Ga*, which will be presented in future papers. However, rcrit under given conditions can be measured directly using 1D pit experiment. This model can be used for multiphase alloys such as duplex stainless steel, but one must consider the differences in passive film breakdown and idiss,max for the different phases.
- Jerry, you mention that pit can grow under charge transfer, but the ionic mobility is still under diffusion control. How to reconcile this point? (Question by VS Raja)
Answer: The ionic mobility and rate-controlling step are two different concepts. The rate-controlling step can be either charge transfer or diffusion. However, under either condition, transport of the ionic species in the pit solution is mainly carried by diffusion. When we say that the pit growth is under charge transfer control, it indicates that the diffusion-limited rate of metal cation transport in the solution is higher than the charge transfer rate at the metal/solution interface at the pit bottom.
- Could you comment on the effect of gravity (sample orientation) on pit stability and salt film formation? How important is gravity in modeling? (Question by Anonymous Attendee)
Answer: The effect of gravity depends on sample orientation – for a sample not facing upwards, gravity will influence the ion transport process through the natural convection created by the higher density of the concentrated pit solution. This effect will influence the critical parameters, but it could be handled by our model with some necessary modifications.
- Why a different slope is observed for deeper pit (higher than 700um) in the Esat vs pit depth curve? If this is because of hydrogen reduction reaction as you explained in the framework, can this be validated? (Question by Ke Wang)
Answer: Our explanation is that, for deep pits, which have low limiting currents, the local cathodic reaction (i.e. hydrogen evolution reaction) can approach the magnitude of the anodic reaction. Therefore, the measured anodic current density is smaller than the actual anodic dissolution current density. Validation was provided by the cryo-salt film work, which required very deep pits, and each exhibited a large void that we attributed to an H2 bubble.
- Pitting mechanisms/reactions in different media such as stainless steels (e.g. SS316) pitting in NaCl solution (with O2) and in FeCl3 solution can be rather different (e.g. in terms of the existence of deposit film and cover...). Do you think the framework is appliable to both cases and in more general terms? (Question by Mike Tan)
Answer: It isn’t clear what exact differences you are referring to in your comment, but we think the framework is applicable to such conditions. Our work was done under potentiostatic control, but open circuit conditions with oxidizers should be similar if there is sufficient cathodic current available. Ferric ions are strong oxidizers with higher solubility than O2 so their limiting current density could be higher than for ORR. However, ferric ions create acidity through hydrolysis, which will affect passive film stability and pit initiation processes.
- In the presentation, you showed activation-controlled and diffusion-controlled. How does this work apply for Ohmic controlled region? i.e. what would be critical parameters like potential, current density for Ohmic controlled region. (Question by Anonymous Attendee)
Answer: There is no such thing as ohmic control of an electrochemical reaction. The charge transfer process is either under activation (charge transfer) or mass transport control. However, the ohmic potential drop in solution will influence the potential at the pit surface, so in a way there is no pure charge transfer control. It is probably best to refer to it as charge-transfer/ohmic control.
- Is there a way to compute Ccrit based on thermodynamic calculations like Csat? (Question by KAMALNATH KADIRVEL)
Answer: Ccrit might be estimated using acidification pitting theory combined with the Pourbaix diagram. However, we think that measurements will be more accurate.
- Do you think this model could extend to corrosion-resistant alloys (NiCrMo Alloys) ? (Question by Edgar Hornus)
Answer: Yes! There is no reason why not.
- What would you consider the best to develop pitting resistant stainless steels, reducing inclusions or increasing nitrogen content? Nowhere these issues figure in PREN criteria! (Question by VS Raja)
Answer: Developing corrosion resistant alloys should take into account both passive film breakdown and pit growth stability. Reducing the amount of inclusions can decrease the rate of pitting initiation, and reducing the size of inclusions can decrease the metastable pit size, thus inhibiting the metastable to stable pit transition. Alloying with interstitial nitrogen probably significantly decreases idiss,max, thus decreasing pit growth stability. All of these approaches are required for alloys with the highest pitting resistance.
- How well the 1-D experiments would match with actual pit growth scenario? What can be the possible cases where this type of experimental system won't work? (Question by Anonymous Attendee)
Answer: 1D pit geometry should be similar to real pits – it’s just that the diffusion process and analysis are simplified. For real pits, the actual pit geometry should be considered to analyze the diffusion process. 3-D diffusion is tractable.
- What are the conditions metastable pits need to satisfy to be shown, is there a metastable critical pitting temperature? In this regard, why 316L is more likely to show metastable pitting than other SS alloys, is there any particular microstructural fact should be considered? Finally, what is the pO2 role in this whole history? (Question by David Bastidas)
Answer: Metastable pitting can occur at very low temperature, and the temperature can influence the initiation rate of metastable pitting events. A critical metastable pitting temperature might exist, but there is no clear evidence for it. The particular SS316 material used in our study has a high density of inclusions, which can act as pit initiation site, thus it exhibits a high metastable pitting initiation rate. It is interesting that its pitting resistance is higher than the other alloys even though the metastable pitting initiation rate was high. Our explanation is that the SS316 has a much lower idiss,max (thus pit growth stability), thus pits cannot stabilize at low potentials. pO2 can affect the open circuit potential and the ORR limiting current density. Certainly, a limitation in the cathodic current can limit pit growth as modeled nicely by Rob Kelly.
- Any comments on conditions when a < asat but idiff,crit< idiss,max <ilim ? the first condition is for the diffusion control process, however, the second condition is for the charge transfer process; so, which one plays a significant role?
Answer: These two conditions cannot be achieved at the same time. When a < asat, the idiss,max will be larger than idiff,crit (for a covered pit).
- Where does in your ideas enter the chemistry? Aggressive anions like halides cause pitting, others not. I have the impression that the aggressiveness is strongly related to the possibility of anions like Flouride Chloride etc. to enhance the cation transfer from the surface of a passive film to the electroyte catalytically, Thus pitting might be discussed as a competition between dissolution and passive film formation. (Question by Henning Strehblow)
Answer: A sufficient aggressive pit environment is required for pitting corrosion, which is defined by the Ccrit. Our framework addresses the question of what conditions can maintain the aggressiveness at or above Ccrit. When the Csurf is larger than Ccrit, the passive film is not able to form, so it is not necessary to consider the competition with passivation under this condition. Ccrit should be the critical parameter related to the chemistry. We consider how this Ccrit be maintained, rather than how the passive film forms or metal dissolves. We do not address why chloride stabilizes pitting but sulfate, for instance, does not. That might be handled by atomistic modeling.
- Should migration transport not be considered? this would increase the idff? (Question by Anonymous Attendee)
Answer: Transport by migration inside the pit solution should be small relative to diffusion, so we ignored it for simplicity. But our papers emphasize that the D in the Fick’s first law diffusion equation is the effective diffusion coefﬁcient of metal cations that takes into account the effect of concentration-dependent diffusivity in the pit and the possible inﬂuence of electromigration.
- Do you think ranking of stainless steel alloys with respect to their localized performance based on CPT vs. repassivation potential would be the same? If not, why? (Question by Guru Prasad Sundararajan)
Answer: For most situations (if they are tested under the same condition), the ranking by CPT and Erp should be the same. However, scatter in the data will result in extreme cases. For example, when the passive film on an alloy is very protective (such as the passive film on sputtered stainless steel), its CPT can be very high, much higher than other similar alloys. However, its Erp might be lower because CPT is not only determined by pit dissolution rate (i.e. idiss,max), but also influenced by the passive film breakdown. In contrast, the Erp is only determined by the pit dissolution rate (at the same pit depth.)
- For the pit stability criteria, you mentioned idiss,max >= idiff,crit for a stable pit. Shouldn't the actual criteria be "dC/dt>=0" where is "C" is the amount of the metal cations in the pit? (Question by KAMALNATH KADIRVEL)
Answer: When the pit is repassivated, the dC/dt also equals 0. Additionally, it is hard to assess dC/dt in a pit.
- Pitting and crevice corrosion are said to have the same type of mechanism. Could this be applied to crevice corrosion as well? (Question by Daniel Klint)
Answer: Yes! See recent papers by R.S. Lillard in J. Electrochem. Soc.
- The CPT can be obtained by measuring Eb at different temperatures, which is also called potentiodynamic CPT. Does the explanation on potentiostatic CPT also apply to the potentiodynamic CPT? (Question by Yangting Sun)
Answer: This is an insightful question and is one of the areas we are working on to advance our framework.
- What does chromium ions from pit try to do, repassivation? (Question by Deepashri Nage)
Answer: Cr3+ in the pit solution helps to create an acidic and Cl--enriched pit environment, which will increase the aggressiveness of the pit solution and promote the further dissolution.
- The idiss,max of Mo-containing 316L is larger than that of 304. Does the Mo ions diffuse from pit solution to bulk solution faster? Or the dissolution of Mo from the base metal into the pit solution faster? (Question by Yangting Sun)
Answer: idiss,max of Mo-containing 316L is smaller than that of 304. See detailed explanation in Part VI of our series of papers recently published in Corros. Sci.
- Does artificial convection affect the salt formation and consequently the results? (Question by Martien DEFFO)
Answer: Yes, it will. Artificial convection will influence the ion transport process, thereby influencing the critical parameters. This can be handled by our framework with some necessary modifications.
- What factors decide the time and size of metastable pitting? How to predict it? (Question by Yangting Sun)
Answer: We cannot answer this question definitively, but it seems that the size of inclusions and strength of the pit cover (as affected for instance by surface pretreatment such as oxidation or pre-passivation) are critical. The critical parameter of acrit might be used for the prediction.
- Does your framework also for other alloys, e.g. Al- or Mg-alloys or Titanium? (Question by Fritz Hunkeler)
Answer: The major difference of the localized corrosion between the stainless steels and Al alloys is that there is copious hydrogen evolution in active pits in Al alloys. Therefore, application of this framework to the localized corrosion of Al alloys must consider the contribution of convection generated by hydrogen evolution. However, the fundamental principles of our framework should be the same. We still need to consider the balance between the production of the aggressive local environment and the transport out of the pit. Instead of pure diffusion, we need to consider the convection, which is messy to deal with.
- Have you looked at how tensile stress affects the pitting process as described by your model? (Question by James Earthman)
Answer: See the work from A. Rota and H. Bohni in the mid 80s. My recollection is that there was not much affect unless the localized attack was intergranular.
- The Metastable Pit has a thin oxide layer over it that will rupture at some point. Is rupture chemically or mechanically based? Can you test? Sounds like rupture influences metastable pit size which influences transition to stable pitting. Can we make the oxide layer cap rupture more or less easily as desired? (Question by Raymond Santucci)
Answer: The nm thick passive film cover spans the micrometer diameter of a metastable pit, so it is a tenuous situation for the pit cover. We expect that both pit solution chemistry and mechanical factors (such as the osmotic pressure difference between inside and outside of the pit) will influence the rupture events of the pit cover. A mechanically robust pit cover can help maintain the aggressive environment of the pit solution and allow the metastable pit to grow for extended period, thus promoting the transition to stable pit growth. Considering pitting resistance from the pit stability aspect, a fragile pit cover is desired to decrease the stability of metastable pit. However, there is a trade-off because, if the passive film is very robust, its resistance to breakdown will increase, which is good for improving the pitting resistance. A robust passive film usually generates a robust pit cover, which, as mentioned above, will promote the metastable/stable pit transition and decease the pitting resistance. The passive film properties are determined by the composition and structure of the oxide film, thus the composition of the alloy. Also, agitation of the bulk solution can usually promote the rupture of the pit cover and thereby reduce the tendency for pit stabilization.
- It is well known that an SS with higher PREN has higher pitting corrosion resistance. How does alloying addition of Cr, N, Mo affect pitting tendency? How does their presence affect transition from metastable to stable pitting corrosion in concentrated solution? (Question by Supratik Roychowdhury)
Answer: Cr, N, and Mo are beneficial elements to improve the pitting resistance, either through improving the resistance to passive film breakdown or inhibiting the metastable to stable pit transition. N and Mo can decrease the idiss,max of the alloy, thus inhibit the metastable/stable pit transition.
- As you said, pitting usually starts at the material inclusions. Would you expect the dissolution to start at the interface due to electrochemistry or is it more dependent on the distortion of the passive film due to the topography? (Question by Sören Lentzsch)
Answer: We have no evidence supporting one or the other. It could be that, for inclusions like MnS, pitting may start from the chemical dissolution of MnS, then the dissolution of metal.
- Can your model be used to predict repassivation temperature as an analogy to repassivation potential? (Question by Jose Vera)
Answer: Yes! Repassivation temperature is a real parameter that has been measured. It is straightforward to apply the framework to address it.
- The model assumes hemisphere. Is it just because it is analytically simple or that is what practically being observed often? (Question by Hang Ren)
Answer: Hemispherical pits are often observed, but not always perfect hemisphere. Additionally, a hemispherical geometry is also simple for analysis.
- I have a question regarding the dependence of pit shape on dissolution rates. At higher dissolution rates, when a salt film would be expected to form, is it fair to say that hemispherical pits would also form vs. A faceted pit? Also, what controls the growth of the hemispherical pit to be either primarily lateral growth or depth growth (long cylindrical channels?) (Question by Leo Monaco)
Answer: When salt film precipitates, polished pit surface are often observed, and the pit should be hemispherical. Lateral or deep growth depends on the difference of actual potential applied at the different position of the pit surface, which is influenced by the salt film thickness at different positions of the pit surface.
- I was wondering if we consider the effect of stress and multiphase alloy, then how the pit initiation or stabilization will affect the critical rate? (Question by PRANSHUL VARSHNEY)
Answer: These issues have been addressed above.
- I have a question about the outliers. You talked about the outliers based on only 5 replicated data; however, increasing the number of data could remove the outlier and change the shape of distribution. Do you have any comments on that part? (Question by Sina Matin)
Answer: Not sure which outliers you are talking about. If you mean the CPT curve of 316L in 0.6 M NaCl, we do not think the curve with the larger CPT is an outlier. A phenomenon that cannot be explained by the common understanding does not mean it is outlier.
- How to use the r criteria in evaluating the pitting corrosion? (Question by Longlin Lei)
Answer: rcrit can be used as a criterion for pit growth stability. Under a given temperature and potential, a larger rcrit indicates a higher pitting resistance.
- Something to think about in interpreting the frozen relics of the strong electrolyte solutions in the pit is the metal ion activity, which will be very low with a strong dependence on valance. Cr, for instance will have a more enhanced solubility than Fe or Ni, and If Cr6+ is generated, it will behave as an anion! (Question by Mark Jaworowski)
- Could you recommend a particular paper with a focus on investigating the characteristics of the salt layer inside the corrosion pit on steel, aluminum, or zinc substrates? (Question by Donald Vonk)
Answer: The related articles can be found in the introduction part of our recent cryo-salt film paper. https://doi.org/10.1016/j.corsci.2020.108812
- Further clarification regarding the oxidation effects on pitting corrosion? Material structure, phases and grain do they have more effect to pitting? (Question by Manish Shinde)
Answer: As described, prior air oxidation will thicken the passive film, which reduces the frequency of breakdown but stabilizes the passive film covers on metastable pits. Also, as mentioned, each phase in for instance a duplex stainless steel will have its own idiss,max and the most susceptible phase will pit first
- It would be interesting to see studies of pitting on electropolished nitinol. I have data where this material is used for 10s of minutes well above its pitting potential without pitting - it actually appears to anodize rather than pit. This is hard for me to explain. (Question by Shawn Kelley)
Answer: This is an example of pitting being controlled by initiation, not pit growth stability. The same is exhibited by sputter deposited SS thin films. However, if a pit initiates at the high potentials that you can scan to, then look out – it will grow at an extremely rapid rate.
- I just wonder if you considered in any point the role of microorganisms and their metabolites? I am actually doing research on that subject and see some synergies in the topic. (Question by Beatriz Watts)
Answer: Good luck with that – it sounds interesting but really complicated. We have not addressed MIC effects on pitting.
- Why the distribution of Cr3+ doesn't repeat with that of Fe and Ni? (Question by Menghao Liu)
Answer: Probably because Cr3+ has a different valence state than Fe2+, so it cannot substitute for the Fe2+ in the FeCl2·4H2O salt crystal lattice.
- Does pitting corrosion occur in medium with proteins? especially on 316L SS. (Question by Jean Geringer)
Answer: We have not addressed this. It is possible that proteins are inhibitors for pitting of SS3316L.
- How best can I estimate/measure pit depth? (Question by NANDA KUMAR T T)
Answer: It is not easy to estimate pit depth from charge passed for a bulk sample with many pits. You would need lots of assumptions. Pit depth can be measured by different techniques, but we often use an optical profiler based on interferometry.