2017 Summer Newsletter
Congratulations to Marie Collin on receiving the price for the best poster at the ‘Scientific Days of Marcoule’
Marie Collin won the price for the best poster at the ‘Scientific Days of Marcoule’, a 2 day local conference (June 22-23, 2017) focused on the research conducted in CEA center. Marie Collin presented the work she performs within the WastePD project on the behavior of nanoconfined water in passivating layer formed during glass corrosion. Congratulations!
CEA developes a comprehensive multi-scale understanding of borosilicate glass corrosion under various environmental conditions and solicitations.
The Gin group at CEA performs comprehensive studies on borosilicate glass corrosion, focusing on the international simple glass (ISG), a 6 oxide borosilicate glass. They are interested in how the glass is passivated by surface layers when it is altered in water highly concentrated in Si. During the last few months, they investigated the ISG corrosion in silica saturated conditions, pH 7, 90°C, with KCl salt, trying to understand the dynamics of water in the passivating film. Under these conditions, the glass dissolution rate dropped by a factor 5000 relative to the initial dissolution rate. After one year of exposure, the ISG formed an alteration layer on the surface, which is a 1.5 µm thick amorphous gel formed by in situ reorganization of the Al-Zr-silicate glassy network following the release of B, Na and Ca. The water content and speciation within this gel layer was characterized. The pore size was found to be 1 nm. Coupons of corroded glass have been immerged at room temperature in a solution of isotopically tagged water molecules H218O for various durations. The oxygen isotopic profiles were recorded in the gel layer with by Time-of-flight-secondary-ion-mass-spectrometry (ToF SIMS). They found that 18O, thus H2O of the bulk solution penetrates through the gel layer up to the reactive interface. Moreover, the increase of 18O within the gel is noticeable even at the shortest contacting time (3 min). The concentration of 18O significantly increases with the contacting time in the tagged solution. As oxygen isotopic ratio is quantitatively determined by this technique and because the O distribution in the gel is known, it is then possible to calculate the fraction of pore water or pore water + OH groups exchanged as a function of time. The result shows that a small fraction pore water (< 5 %) is quickly exchanged (< 3 min) whereas it takes more than 500 hours to complete a full equilibration between the bulk water and the pore water. Their results suggested that the gel displays a few open channels allowing a small fraction of water molecules to diffuse quickly up to reaction front but the gel also strongly hinders the mobility of the largest fraction of pore water.
Wang group at LSU tries to understand the iondine release mechanism of iodine-apitite waste form
As part of the ceramic team under WastePD, LSU focuses on understanding the fundamental mechanism of radionuclide incorporation and confinement for new waste form design, and transport behavior of radionuclide in bulk crystalline ceramics and across solid-solid and solid-liquid interfaces. Crystalline waste forms are investigated using modeling techniques to study the radionuclide incorporation and using leaching experiment to determine the mechanisms that control waste form long-term durability. Both Artificial Neural Network and atomistic scale modeling techniques are used to study the incorporation. The goal is to provide fundamental understanding for designing ceramic nuclear waste forms with improved incorporation, waste loading, stability, and predictable long-term leaching behavior in geological disposal environmental. The outcomes include mechanistic insight of ceramic nuclear waste incorporation and durability, a new design method for ceramic waste forms based on Artificial Neural Network, and fundamental understanding of long-term degradation behavior of ceramic waste forms.
To better understand the iodine release mechanism of the iodine-apatite waste form, the Wang group performed some long-term durability testing. They postulated that iodine is released from host apatite matrix through an ion exchange mechanism. To verify this hypothesis, a 14-day semi-dynamic leaching test was conducted on the iodine-apatite waste in solutions containing 0.1 molar of different anions (CO32-, PO43-, Cl-, and SO42-) (Fig. 1).
After the leaching experiment, the samples were subjected to infrared spectroscopy (IR) and XRD X-ray Diffractometry (XRD) analysis to reveal the changes in surface chemistry. XRD results indicated some deformation of crystal structure, which allowed the lattice to accommodate new ions introduced by the leaching process. This finding is consistent with the ion exchange mechanism proposed by LSU researchers. The speculation will be further validated by performing a Rietveld refinement analysis on the XRD data and SEM/EDS of surface.
Frankel group successfully developed a highly corrosion resistant, single phase, Ni-based high entropy alloys (HEAs)
Frankel group continues to develop a highly corrosion resistant, single phase, Ni-based high entropy alloys (HEAs). In the previous study, they observed some secondary phases in the alloy matrix. The problem was not completely solved after re-solutionizing the sample at a higher temperature. A Ru-rich particle was still observed at the bottom side of the button. The particle was likely a remnant from insufficient mixing during melting, and the voids nearby suggest growth during the annealing. The re-solutionizing process lead to a uniform sample, but newly formed pores (0.5 ~ 1um) were observed. Their distribution suggest that these pores were formed because of the Kirkendall effect caused by the second phase regions in the as-annealed sample. Based on current results, Frankel group decided to directly homogenize the as-cast sample at high temperature. After this homogenization process, a uniform chemical composition distribution was achieved as indicated by the EDS analysis. No second phases were observed. At the same time, the porosity was greatly reduced as compared to the re-solutionized sample, indicating the Kirkendall pores in re-solutionized sample can be removed by direct homogenization of the as-cast sample. At this point, Frankel group successfully fabricated a quality, single phase, Ni-based HEA.
Frankel group also assessed the corrosion resistance of the HEAs by performing potentiodynamic polarizations in 3.5 wt.% NaCl solution at 30 oC. The polarization curves of as-cast, as-annealed and re-solutionized samples, Fig. 2, show that all three samples exhibit similar corrosion resistance, being spontaneously passive with very high breakdown potential. After polarization, the corrosion morphologies were examined with optical microscopy. The corrosion morphologies for all three samples indicated that dramatic current increase at breakdown potential was not derived from pitting, but from transpassive dissolution. The etching preferentially started from crevices, such as the edge of working area or bubbles, then spread across entire surface quickly with the current dramatically increasing. This etching was possibly derived from transpassive dissolution of Cr oxide in the passive film.
In order to obtain the critical pitting temperature (CPT), Frankel group performed potentiostatic polarization on the HEAs in 3.5 wt.% NaCl with temperature scanning, as shown in Fig. 3. No stable pitting breakdown was observed, indicating that the CPT of as-cast sample is above 85 oC and is above 90 oC for the re-solutionized sample. Also, no large metastable pitting transients were observed. The magnified curve in Fig. 3(b) shows the occurrence of just small metastable pitting events (~1 um) and the frequency of events increases with the temperature. This indicates that the Ni-based HEA possesses excellent pitting resistance. Therefore, Frankel group performed further electrochemistry tests in more aggressive environment. However, no pits were observed even in 12 M HCl at 30 ⁰C. They found that the polarization resistance of the HEA increased with time during exposure for 15 h at the OCP in 12 M HCl. The polarization resistance then decreased with further exposure time, possibly due to a change in the surface composition or depassivation by dissolution of the oxide surface layer.
Locke group seeks to understand the long term atmospheric corrosion and stress corrosion cracking of various alloys
Locke group started a long term atmospheric corrosion study in collaboration with Drs. Eric Schindelholz and Charles Bryan at Sandia National Laboratory. Corrosion morphology (pit size distribution, pit number density, and shape) are being characterized as a function of salt surface load density, relative humidity, temperature, and exposure time using optical profilometry, FIB-SEM, microXCT, and robomet (a robotic cross-sectional metallographic examination technique). The 1 week, 2 week and 1 month samples have been removed from the environmental chambers and are in the process of being analyzed.
Locke group is also preparing materials for stress corrosion cracking (SCC) testing, which will be ready within the next month. Materials being tested include 304H, sensitized 304H, Alloy 600 and PH 13-8 Mo H1050. Identical salt loading and atmospheric conditions as the atmospheric corrosion study will be applied to dog-bone style tensile samples used for SCC studies where intermittent high R ripple-style fatigue loading will be applied to mark the progression of fracture after crack initiation has occurred.
Taylor group works to bring together experiment and modeling
Taylor group attempts to model the potentiodynamic polarization curves while taking oxide formation into account. They will obtain experimental data from the OSU and UVa metals team and fit the polarization curves to extract the mechanistic information. Taylor group also built a kinetic Monte Carlo code, which was used to model the chloride ion adsorption and diffusion on Zr(0001) surface.
In a separate study, Taylor group adopted a special quasirandom structure (SQS) approach and coded it in MATLAB to build some Ni-22Cr slab models for exploring alloy design space in Ni based alloy systems. Having this working model for an alloy system will be very important, which will open up many new avenues for research that will have strong ties to the experimental teams.
Seong group develops new algorithms for glass surface characterization
Seong group tested the oxygen speciation algorithm of x-ray photoelectron spectroscopy (XPS) for a corrosion-treated ISG glass surface. They carried out a preliminary XPS analysis of the ISG glass surface treated in a pH 7 aqueous solution saturated with SiO2 with 13 mM KCl at 30 oC for 21 days. The XPS analysis result of the alteration layer formed after this treatment was compared with those of the annealed samples. The surface treatment used in this preliminary test leached out sodium modifier ions and boron network former component; at the same time, a small amount of potassium was deposited in the alteration layer, which is consistent with the ToF-SIMS analysis result of similar samples from CEA. After the concentration of each element was obtained, areal density of bridging oxygen (BO), non-bridging oxygen (NBO), and OH species per nm2 area was calculated using the charge balance algorithm. The BO, NBO, and OH densities in the annealed surface were in reasonable agreement with the molecular dynamic (MD) simulation results. However, the concentration of the OH species came up with a negative number. The negative number might mean the coordination of the aluminum and network connectivity of silicon might have changed, which needs to be confirmed by comparing with MD simulations.
Specular reflection infrared spectroscopy (SR-IR) can provide the structural information of the silicate network in the alteration surface layer. In principle, this requires processing the SR-IR spectra with the Kramers-Kronig (KK) transformation; however, no commercial software can do this transformation properly for the silicate stretch mode region. The Seong group has devised a new algorithm that circumvents the fundamental limitation of the KK transformation and tested it with soda lime silica glass. Using the newly developed algorithm, the real and complex components of the complex refractive index of ISG glass in the mid-IR region were obtained. This is the first refractive index data obtained for ISG glass. Using this data, the Seong group was able to estimate the probe depth at a given incidence angle and wavenumber of IR.
The hydrous speciation is very important to understand the water content and transport in the alteration layer. The Seong group used attenuated total reflection infrared spectroscopy (ATR-IR) and measured the OH stretch and H2O bend modes. They were able to estimate the relative abundance ratio,[SiOH]/[H2O], of the hydrous species from the ATR-IR spectra.
In addition, Seong group explored the possibility of using Raman to monitor changes in the silicate network. The full spectral interpretation of the Raman spectra is not possible at this moment. It is worth noting that the peak positions characteristic to Qn (n=1~4) species reported for the silicate glasses cannot be used for the boroaluminosilicate glass (although such examples are frequently found in the literature). The full interpretation may require first-principles calculations of vibrational modes. The Seong group used principal component analysis (PCA) algorithm to find main factors. The scores of the first principal component (PC1) did not change with the corrosion treatment time. They believe this is because the PC1 represent the bulk spectrum. The scores of the second PC (PC2) varies drastically with the treatment time (Figure 4c). Seong group suggested that these changes may be related to the alteration layers formed at the ISG glass surface.
QuesTek makes the calculation of Pourbaix diagrams easier for CRAs and HEAs
QuesTek have built an archive of Pourbaix diagrams for various metallic materials, including commercial corrosion resistant alloys (CRAs) and HEAs that are being experimentally investigated at OSU and UVa. QuesTek also developed a generic procedure (Figure 4) and code file in ThermoCalc that enables them to rapidly generate more Pourbaix diagrams for metals and alloys in different environments.
The concept of “canary” alloys was proposed at the beginning of the project in light of the extreme difficulty in characterizing and studying experimentally the highly corrosion resistant alloys such as C22 and the Ni-HEA that has been synthesized at OSU. After discussion within the Metals Team, the consensus on designing a canary Ni-HEA with a deliberately-lowered corrosion resistance is to replace the elements that strongly enhance corrosion resistance (i.e., Mo, W, Ru) with those with less potency (i.e., Mn, Co). Following this replacement, Ni and Fe contents are fixed at the original concentration, while Cr is being varied and balanced by Mn and Co where Mn and Co keep a 1:1 ratio for simplicity. OSU will continue to fabricate these canary alloys as designed by QuesTek, and their corrosion performance and machanism will be assessed across the entire Metals Team.
Lian group continues to advance in Cs and Cl nuclear waste sequestration
In the last newsletter, Lian group reported the synthesis, consolidation, and characterization of Cs2SnCl6 as a potential host material for Cs and Cl sequestration. The pure chloride displayed excellent thermal stability up to 600°C and was consolidated into pure pellets through spark plasma sintering (SPS). By revisiting Cs2SnCl6-xIx after the consolidation success achieved in pure Cs2SnCl6 samples, Lian group investigated the potential for defect perovskites to serve as iodine-bearing waste forms.
While Cs2SnI6 achieves exceptionally high iodine mass loading (66 wt% I) it displayed rapid dissolution in water in benchtop tests, leading it to be unsuitable as a host phase for iodine. By comparison, Cs2SnCl6 displayed both greater water resistance and thermal stability. In an effort to enhance the water and thermal stability of iodine-bearing defect perovskites, iodine was incorporated into Cs2SnCl6, producing a solid solution denoted as Cs2SnCl6-xIx. X-ray diffraction (XRD) of the Cs2SnCl6, Cs2SnI6, and Cs2SnCl6-xIx powders suggests the incorporation of iodine into the Cs2SnCl6 matrix and the associated lattice expansion from the replacement of chlorine with iodine.
Lian group also studied the dissolution of polycrystalline phase-pure hollandites and single-crystal surfaces of perovskites. Limited literature has been devoted to the subject of interaction of Cs-containing hollandites with aqueous environment. The (Ba,Cs)-hollandite structure has been well understood – a network of edge and corner shared TiO6 octahedra forming long one-dimensional tunnels that can host a range of alkali or alkaline earth cations. This accommodation of large Ba/Cs cations occurs by partial reduction of Ti4+ to Ti3+ or by partial substitution of Ti4+ by M3+ cations. M3+ typically are Group III elements (Al, Ga etc.) or d-transition elements (Fe, Cr, Ti, V etc.). The variable oxidation states of transition elements in the hollandite network could significantly affect the phase stability and/or durability of the material in a given REDOX environment. Hence, the REDOX characteristics of M3+ is particularly important. In addition to REDOX behavior, the chemical dissolution of the hollandites also needs to be studied to understand the underlying fundamental science. For this study, three different hollandite systems Ba1.15M3+2.3Ti4+5.7O16 (M = Ti, Cr and Al) are selected. This chemistry was chosen to observe the effects of oxidation of octahedral cations (Ti3+/Cr3+) on both chemical and electrochemical dissolution and compare with that of REDOX stable Al composition. To replicate the conditions induced by corrosion of stainless steel canisters, during electrochemical measurements, polished surface of hollandite will be exposed solutions with different levels of Fe(NO3)3.
Lian group is also trying to challenge the prevailing theory regarding the dissolution mechanism of mineral perovskite. It is generally accepted that the alkaline cations (Ca2+/Sr2+) are selectively leached out via ion-exchange process followed by formation of Ti-rich passivation layer. The process can be summarized as an incongruent dissolution process where a passivation layer is formed essentially by means of two separate reactions – ion-exchange and base-catalyzed hydrolysis of Ti-O framework. It can be further understood that external factors such as temperature, pH, leachant chemical composition and concentration affect the kinetics and nature of the dissolution process. However, relatively recent studies on dissolution behaviors of natural minerals such as Na0.39Ca0.59Al1.59Si2.41O8 feldspar and CaCO3 calcite, etc., based on in-situ AFM/TEM observations, contradict the theory of conventional incongruent dissolution. The proposed mechanism from these studies is ‘coupled dissolution and precipitation’ where (i) the stoichiometric dissolution of the surface and (ii) precipitation of a product phase due to an enrichment of interfacial layer occur simultaneously. Hence it is of interest to revisit the dissolution process of perovskite. Commercially available single crystal samples of SrTiO3 perovskite, with very well-defined cleavage surfaces and compositions, will be used for this study. Similar ASTM leaching tests and characterization techniques, as mentioned above, will be performed for these surfaces. In addition, the dissolution of the pristine surface will be studied, in-situ, using AFM (at room temperature). Flow through liquid cell where the flow rate of the leachant can be controlled will be used to observe the surface dissolution process. These types of experiments have been used in several mineral dissolution studies. The idea is to identify the fundamental interfacial processes occurring upon initial contact and the following surface changes. Collective information from in-situ testing and the leaching tests can be critical to identify the overall mechanism and predict the long-term behavior of the material.
UVa focuses on metallurgical characterization on newly synthesized HEAs
Scully group at UVa work actively on the metallurgical characterization on newly synthesized HEAs with emphasize on processing/heat treatment to produce a chemically homogenous single-phase alloy. The HEA was synthesized at OSU and the samples were shared between UVa and OSU for collaborative studies. The sample abstained during the first batch was not single phase. After re-homogenization at UVa, the HEA displayed a homogeneous distribution of the various elements. To determine which elements are present in the air-formed oxide, they performed base line XPS scan and found that all expected elements in the HEA were detected. Most elements were detected in the zero valence state and oxidized conditioned. The results indicate that all elements in the HEA contribute to the oxide with the exception of Ru. A calculated thickness of 1 nm is estimated for the oxide layer based on attenuation of zero valence peaks, although the degree of uniformity of this thickness across the HEA sample is unknown. Scully group also conducted some preliminary corrosion electrochemistry experiments on the HEA and control samples including alloy C22 and alloy 600. All materials exhibited a broad potential range where passivity has a very low current density. Under some testing conditions, the HEA showed even lower passive current density compared to alloy C22. This suggests the HEA can easily form a passive layer. On the other hand, alloy 600 has a small passive window before breaking down and undergoing corrosion at potentials below which transpassivity is observed. These materials also indicate OER and Cr+3 to Cr+6 at high potentials. Both the HEA and C22 exhibit a negative hysteresis in all testing conditions suggesting the passive film repairs itself, preventing pit initiation.
Scully group also studied the electrochemical kinetics of the HEA corrosion. Potentiodynamic polarization curves obtained for the NiCrFeRuMoW HEA and C22 in Na2SO4 solutions were compared by environment, i.e. by varying pH level. In the pH 1 and pH 0 environments, there was an active-to-passive transition initiated following EOCP, followed by a passive regime beginning at approximately -0.25 VMMSE, and finally a transpassive dissolution event occurring at approximately +0.5 VMMSE. Both downward scans in pH 1 and pH 0 for the HEA display a positive hysteresis when compared to the upward scan—a behavior not exhibited in either C22 or in the HEA polarized in pH 4 or 12. This property will be further explored by altering the apex potential of the polarization scan. In pH 12 (Figure 5e), there is a marked increase in cathodic events on the HEA compared to at pH 4, as well as an increased passive potential and lowered potential at which transpassive dissolution occurs. Anodic polarization curves were also used as comparisons between the HEA and C22 alloys within the same environment. In pH 1 and pH 4 the alloys behaved similarly in terms of the measured OCP and the passivation current density. In pH 12, the heightened cathodic events are pronounced compared to C22, which retains transpassive behavior throughout the upward scan.
Computational studies from UNT and PNNL help CEA to better interpret the experimental data on glass corrosion
Jinchend Du’s group at UNT will model the behavior of nanoconfined water using dissociative potential in order to interpret the Gin group’s experimental data showing an ultra low mobility of water in nanoporous passivating layers formed on ISG glass. At the mesoscale John Vienna’s group at PNNL will model the same results using the Kinetic Monte Carlo approach.
As part of the computational synergy effort, the Wang group calculates the structural, mechanical, and thermodynamics properties of zirconia using density function theory (DFT).
The Wang group is actively involved in the computational synergy activity. They geometrically optimized the crystal structure of ZrO2. The results of geometry optimization and elastic constants showed a good fit with the experimental data. The structurally optimized ZrO2 was then utilized to calculate the mechanical and thermodynamics properties such as the density of states, density of photon states, and the temperature dependence of the Debye temperatures.
QuesTek hosted a materials design workshop
On Apr 13, QuesTek hosted a materials design workshop, where postdocs and graduate students from groups across the three material teams in WastePD came visit QuesTek and had a full-day interactive discussion on the systems-based materials design approach. Following two tutorial talks given by QuesTek materials designers, the attendees were divided to three teams based on the class of material they represent and each team went on to produce a system design chart for their material, which was later presented around the room and discussed. After the attendees returned to their groups, they would provide feedback to their PIs on the design charts and systems-based design philosophy. Because these system design charts were more of a product of an intellectual exercise, they were not, and were not meant to be, entirely accurate and complete. Nevertheless, QuesTek believed that they established a robust foothold for embarking on dialogues with other PIs in the Center on how to incorporate the systems approach to our scientific studies and, more specifically, refine the ICME design approach for very corrosion resistant materials.
After the workshop, several rounds of conversations with all three materials thrusts on the systems design charts ensued. Numerous insightful comments, suggestions, questions, as well as concerns were voiced by the PIs. QuesTek has been attempting to take all the feedback into account and has come up with updated versions of the design charts. In recognition of the prominent dependency of a material’s corrosion resistance on its operation environment, a double systems chart scheme that distinguishes the operational impacts on the material structure and properties from those related with material processing has been adopted.
PSU and UVa couple their characterization analysis capabilities to determine the interfacial activities of HEAs
PSU has been collaborating with UVA to measure the passivating oxide layer thickness formed on the surface of HEA under electrochemical conditions using spectroscopic ellipsometry (SE). PSU has designed an in-situ SE sample cell for this experiment. They are also constructing an in-situ cell for sum frequency generation (SFG) analysis.
Sarita Sahu - firstname.lastname@example.org
Sarita Sahu is a new PhD student advised by Jerry Frankel. She will be working on the corrosion interactions between metal and glass. She will also be involved in the high entropy alloy develpment project.
Brandon Free - Free.email@example.com
Brandon will be co-advised between PI’s Locke and Taylor. Brandon will begin work on quantifying small crack growth and crack initiation out of corrosion features and incorporating his results as well as those from T. Weirich into modeling efforts.
- G.S. Frankel, J. Vienna, J. Lian, “WastePD, An Innovative Center on Materials Degradation,” Nature Mat. Deg., 1 (2017) 5;
- G. Yao, Z. Zhang, and J. Wang, “Beta transmutations in apatites with ferric iron as an electron acceptor – implication for nuclear waste form development,” Phys. Chem. Chem. Phy.s, 19, (2017) 25487-25497.
- X. Lu, D. Schreiber, J. Neeway, J. Ryan, J. Du, “Effects of Optical Dopants and Laser Wavelength on Atom Probe Tomography Analyses of Borosilicate Glasses”, J. Amer. Cer. Soc., 100 (2017) 4801.
- A.J. Samin, C.D. Taylor, “A first principles investigation of the oxygen adsorption on Zr(0001) surface using cluster expansions,” Appl. Surf. Sci. 423 (2017), 1035-1044.
- M. Ren, L. Deng, J. Du, “Structure and properties of sodium borosilicate and aluminborosilicate nuclear waste glasses from molecular dynamics simulations”, J. Non-Cryst. Solids, 476 (2017) 87.
- J. Du and J. Rimsza, “Atomistic computer simulations of water interactions and dissolution of inorganic glasses”, Nature Mat. Deg., accepted for publication.
- P.C. Roth, H. Shan, D. Riegner, N. Antolin, S. Sreepathi, L. Oliker, S. Williams, S. Moore, and W. Windl. “Performance analysis and optimization of the RAMPAGE metal alloy potential generation software,” in Proceedings of the 4th ACM SIGPLAN International Workshop on Software Engineering for Parallel Systems (SEPS 2017). ACM, New York, NY, USA, 11-20. DOI: https://doi.org/10.1145/3141865.3141868.
- Gin, S. “Current understanding and remaining challenges in measuring and modeling long-term performance of borosilicate nuclear waste glass” June 21th 2017, NWTRB meeting, Richland WA, USA.
- Collin M., Gin S., Fournier M., Charpentier T. “Impact of alkali in solution on glass alteration and water behavior in nanoporous altered layer” August 2017, Goldschmidt Conference, Paris, France.
- Locke, J., Schaller R., Schindelholz E., Bryan C. “Atmospheric Corrosion and Stress Corrosion Cracking of Corrosion Resistant Alloys Used for Dry Storage of Spent Nuclear Fuel” October 1-5, ECS meeting, National Harbor, MD.