Long-term Corrosion Tests Research Articles

Long-term corrosion tests of Ti3SiC2 and Ti2AlC in oxygen containing LBE at temperatures up to 700 °C

Two MAX-phase materials, Ti3SiC and Ti2AlC, were tested at 550 °C, 650 °C and 700 °C up to 10 000 h in LBE (lead-bismuth-eutectic) containing of 10−6 and 10−8 wt% oxygen. It was found that secondary phases have a strong influence on corrosion effects. Ti3SiC showed a surface disintegration at 550°C/10−6 wt% oxygen after longer exposure, while a 4–7 μm thick TiO2 layer with Pb-Bi inclusions was detected on Ti2AlC. However, Ti3SiC is protected by a double layered oxide with an outer part of TiO2 and a mixed inner layer of SiO2 and TiO2 at the higher temperatures. Ti2AlC formed a TiO2 surface layer containing Al2O3. Some defects could be observed on the Ti3SiC surface in LBE containing 10−8 wt% oxygen at 550 °C and 650 °C. The secondary phases between the Ti3SiC grains showed strong oxidation at 700 °C. Due to the high Al solubility in LBE, Ti2AlC experienced strong dissolution attack after longer exposure times at 650 and 700 °C.

Long-term corrosion behavior of ODS-Eurofer in flowing Pb-15.7Li at 550 °C

Low activation ferritic-martensitic steels (RAFM-steels) are foreseen as structural materials in different blanket designs (HCLL, DCLL, WCLL) with Pb-15.7Li as breeding and partly also as cooling medium. In HCLL and WCLL designs the structural material will be in direct contact with the flowing liquid breeder at operating temperatures up to 550°C. In the past, investigations concerning the corrosion behavior of RAFM-steels like F82H-mod. and Eurofer showed that these alloys are attacked by the flowing breeder. These corrosion tests pointed out, that the corrosion attack depends mainly on flow velocity and temperature. All these alloys were ‘single’ phase ferritic-martensitic steels without any additions. However, structural materials with better strength and creep resistance like ODS-Eurofer or ferritic steels are gathering interest in blanket development and will be required at least for DCLL application due to higher operation temperatures. Nevertheless, reliable data concerning compatibility with Pb-15.7Li are missing until now.In this paper results from long-term corrosion testing of ODS-Eurofer will be reported for exposure times up to 1.5 years at a flow velocity of 0.1m/s. The evaluated data for ODS-Eurofer corrosion will be compared with values of ‘single’ phase ’classical‘ Eurofer. The observed corrosion attack and mechanisms will be discussed in detail considering the testing conditions and the microstructure of the RAFM-steels.

Prediction of Materials and Coating Durability Based on Atmospheric and Laboratoty Tests

All industries are interesting in the durability/service life of products, structures, equipment, plants, etc. One factor affecting this is corrosion resistance. There are many methods for received such data. Methods supporting standardized data are long-term atmospheric corrosion tests, mapping of corrosivity, field tests on real structures including the evaluation of long-term exposed materials and coatings. The choice of suitable accelerated test is very important for receiving the reasonable information.

Long-term corrosion of austenitic steels in flowing LBE at 400 °C and 10−7 mass% dissolved oxygen in comparison with 450 and 550 °C

Long-term corrosion tests for up to ∼13,194 h on 1.4970 (15-15 Ti), 316L and 1.4571 austenitic steels were carried out at 400 °C in flowing LBE (2 m/s) with 10−7 mass% dissolved oxygen. The steels show general slight oxidation (Cr-based oxide film) along with local, pit-type solution-based corrosion attack. The incubation time for pit-type attack is ∼4500 h. After ∼13,194 h, the maximum pit depth observed was ∼14, 23 and 57 μm for 1.4970, 316L and 1.4571, respectively, that corresponds to local corrosion rates of ∼6, 10 and 26 μm/year. At 450 °C and 550 °C, the corrosion rates are ranged in between ∼120−220 μm/year and ∼500–3000 μm/year, respectively. Corrosion appearances and mechanisms are discussed.

Corrosion Behavior of Zirconium, Titanium, and Their Alloys in Simulated Dissolver Solution of Fast Breeder Reactor Spent Nuclear Fuel Using Zircaloy-4 Mock-Up Dissolver Vessel

Long-term corrosion testing of a mock-up dissolver vessel to be employed in the aqueous reprocessing of spent nuclear fuels of fast breeder reactors has been initiated. In this work, a Zircaloy-4 (Zr-4) mock-up dissolver vessel was used as the testing facility to evaluate the corrosion rate of several candidate materials based on zirconium and titanium in the boiling and vapor phases of simulated dissolver solution (SDS) comprising fission and corrosion product ions in 11.5 M nitric acid. Several campaigns of 100, 250, 500, 1000, and 2500 h of operation were completed. The corrosion rates of the candidate materials are expressed both in micrometers per year (μm/yr) and mils per year (mils/yr). Zirconium-702, Zr-4, autoclaved Zr-4, and commercial pure titanium (CP-Ti) exhibited low corrosion rates of 0.08 to 0.23 μm/yr (0.003 to 0.009 mils/yr) in the as-received and welded conditions exposed to the boiling liquid phase of the dissolver solution for 2500 h. Whereas the CP-Ti and CP-Ti weld exhibited marginally higher corrosion rates of 1.0 μm/yr (0.04 mils/yr) and 1.9 μm/yr (0.075 mils/yr), respectively, in the vapor phase of the dissolver solution, the lowest corrosion rate of 0.08 μm/yr (0.003 mils/yr) was obtained for the autoclaved Zr-4 sample exposed to boiling SDS. Scanning electron microscope investigations did not reveal any corrosion attack for the titanium and zirconium samples. Laser Raman spectroscopic analysis confirmed that the origins of passivity of zirconium and titanium samples were due to the formation of ZrO2 and TiO2, respectively. However, the CP-Ti/AISI Type 304L stainless steel (SS 304L) and Zr-4/SS 304L dissimilar welds had undergone severe corrosion. Visual inspection of the Zr-4 dissolver vessel revealed no corrosion attack after operation for 2500 h. The results of this 2500-h campaign would serve as the baseline data for the analysis of future long-term campaigns.

Long-term corrosion tests of nanostructural nitinol of (55.91 wt % Ni, 44.03 wt % Ti) composition under static conditions: Composition and structure before and after corrosion

Nanostructural nitinol is investigated before and after storage for a long time in solutions simulating pH of physiological liquids. The structure and composition are determined by SEM, TEM, and Auger spectroscopy. The material base is the B2 phase in the form of fibers 30–70 nm in diameter. Polishing promotes the formation of a thin homogeneous surface layer of titanium oxide, which thickens after solution storage; and annealing, the layer unchanging, thick, and heterogeneous.

Corrosion behavior of thermal-sprayed WC cermet coatings in SO4 2− environment

A series of the electrochemical and long-term corrosion tests was carried out in a 3.5 wt% Na2SO4 solution on thermal-sprayed WC-17Co and WC-10Co-4Cr cermet coatings in order to examine the effect of composition of binder materials on the corrosion behavior. The results reveal that the overall corrosion resistance of the WC-17Co coating is inferior to that of the WC–Co–Cr coatings due to the corrosion of binder materials which induce WC particles to fall off. CoO and WO3 oxide films form on the surface of WC-17Co coating in Na2SO4 solution electrochemical corrosion process, which will protect the coating in the process of corrosion. Cr2O3 oxide film formed on the WC-10Co-4Cr coating surface has a strong hindered role to corrosion. The corrosion mechanism of WC-17Co coating in Na2SO4 solution is entire corrosion of Co matrix, while it is film-hole corrosion mechanism for WC-10Co-4Cr coating.

Corrosion Behavior of Carbon Steel in Piperazine Solutions for Post-Combustion CO2 Capture

L. Zheng, J. Landon, W. Zou, J. E. Remias, K. LiuCenter for Applied Energy Research, University of Kentucky, 2540 Research Park Drive, Lexington, KY 40511-8410, United StatesAnthropogenic carbon dioxide generation from the combustion of fossil fuels such as coal and natural gas may be a factor to global climate change. One of several mitigation options is to adopt post-combustion CO2 capture and storage, which is a process consisting of the separation of CO2 from industrial and energy-related sources by typically using aqueous solutions of alkanolamines, and then transporting pressurized CO2 to a storage location for long-term isolation from the atmosphere [1-3]. However, one of the issues for the application of this process is corrosion. The documented results have shown that the CO2 captured aqueous solutions of alkanolamines without inhibitor are highly corrosive not only to carbon steels with little alloying of chromium but also to stainless steels with chromium content up to approximately 20 wt.%, which is even worse in the presence of oxygen for post-combustion application [4-6]. This can directly affect the whole system availability as well as the economics due to unplanned downtime, and reduced equipment lifetime. To decrease corrosion in current commercial CO2 capture systems but without affecting its CO2capture efficiency and significantly increasing cost, one of the methods is to apply an alternative solvent with low corrosivity [3-10].Recent studies have shown that piperazine (PZ), a cyclic secondary amine with two nitrogen groups, has some attractive advantages [7], which make PZ a viable alternative solvent. Moreover, PZ has been routinely reported to be an effective promoter of the absorption rate in CO2 capture solutions [6, 8-10]. Investigation of the corrosion of materials in PZ is not only necessary to understand the mechanism of the corrosion of materials in the existing blends with PZ, such as promoted Methyl diethanolamine, but also to develop new PZ activated blends as well as to apply PZ as a single solvent for CO2 capture in the future. In this work, both electrochemical testing and long-term immersion testing (totaling 1050 h) were applied to investigate the corrosion of carbon steel in CO2-loaded concentrated PZ solutions.A106 Grade B carbon steel samples, cut from pipes, were used for corrosion investigation. All of the samples were ground with SiC sand paper, and then ultrasonically cleaned with deionized (DI) water and acetone. The effects of CO2 loading on the corrosion rate of A106 in 30 wt.% PZ in the typical operational range (0.23 ~ 0.43 mol CO2/mol alkalinity) and temperatures from 20 to 80 oC were examined by electrochemical measurements such as potentiodynamic polarization and EIS. In addition, long-term immersion corrosion testing was conducted in a traditional corrosion cell for a continuous duration of 1050 h. A set of samples was removed every 150 h to study the impact of duration on the corrosion. The corrosion rate after each run was calculated on the weight loss method. Fig. 1 shows the corrosion rate of A106 as a function of immersion duration. As can be seen, the corrosion rate decreased sharply from approximately 0.14 mmpy after 150 h to approximately 0.02 mmpy after 600 h. Thereafter, a steady-state corrosion process was observed with increases in duration up to 1050 h. To understand the corrosion mechanism, X-ray Diffraction and Scanning Electron Microscopy/Energy Dispersive Spectroscopy were used to characterize the corroded samples. The details will be discussed in the following meeting.Fig. 1 Corrosion rate of A106 in 0.43 C/N CO2-loaded 30 wt.% PZ at ~100 oC as a function of immersion duration Acknowledgement The authors acknowledge the Carbon Management Research Group (CMRG) members, including Duke Energy, Electric Power Research Institute (EPRI), Kentucky Department of Energy Development and Independence (KY-DEDI), Kentucky Power (AEP), and LG&E and KU Energy, for their financial support. Also, the authors acknowledge suggestions and discussions from Ayokunle Omosebi and Xin Gao, and solution analysis from Zhiao Li and Neal C Koebcke.

Long-term corrosion behavior of CVD SiC in 360 °C water and 400 °C steam

Long-term corrosion tests were performed using chemical vapor deposition (CVD) SiC specimens in 360°C water and 400°C steam to evaluate the corrosion behavior of SiC cladding under normal operation conditions. The CVD SiC specimens exhibited weight loss caused by dissolution in both 360°C water and 400°C steam. For the corrosion test in 360°C water the dissolution rate was much higher in the static autoclave, where the dissolved oxygen content was higher, than in the loops, where the dissolved oxygen content was maintained at approximately 5ppb. The grain boundaries of the SiC dissolved preferentially during the early stage of corrosion, and the grains became thinner and detached from the surface, thereby leading to an acceleration of the weight loss. The pre-oxide was not able to protect the SiC from dissolving in water, even though it was effective, to a certain extent, at mitigating the dissolution of the SiC. In the corrosion test in 400°C steam, the weight loss was significant from the early stage of corrosion. No acceleration in the corrosion rate was observed in 400°C steam for up to 135days, and the weight loss at 90days was similar to that obtained from the 360°C water autoclave test.

Effect of metal cations on corrosion behavior and surface film structure of the A3003 aluminum alloy in model tap waters

The effect of the metal cations, Na+, K+, Ca2+, Mg2+, Zn2+, and Ni2+, on the oxide film structure and morphology changes during long-term immersion corrosion tests of aluminum alloy (A3003) in model tap waters was investigated by X-ray photoelectron spectroscopy (XPS) and scanning electron microscopy. The effect of the metal cations on the corrosion behavior was also investigated with mass change and electrochemical tests. The hardness of the metal cations, X, based on the hard and soft acids and bases (HSAB) concept was applied to explain the effect of metal cations on the passive oxide film structure and corrosion resistance. The mass change rate and corrosion current density decreased with increasing metal cation hardness. The XPS results showed that hard metal cations like Zn2+ and Ni2+ were incorporated in the oxide films, while the four soft metal cations were not incorporated in the oxide films. The results are in good agreement with those which could be expected from the HSAB hardness of the metal cations.

Long-term performance of 1H, 1H′, 2H, 2H′-perfluorooctyl triethoxysilane (POTS) microcapsule-based self-healing anticorrosive coatings

Self-healing anticorrosive coatings were prepared based on the microencapsulation of 1H,1H′, 2H,2H′-perfluorooctyl triethoxysilane (POTS). Toluene played a critical role in the successful synthesis of POTS microcapsules although it was not contained in the resultant capsules. The excellent corrosion-resistant property of a POTS microcapsule-modified silicone elastomer coating was demonstrated by a long-term corrosion test, in which the prepared coating was unscratched and exposed to HCl solution for 1 month, and the metal substrate was inspected after peeling off the coating by scanning electron microscopy and energy dispersive X-ray spectroscopy. In addition, the self-healing behavior of a scribed POTS-based epoxy coating was characterized by an electrochemical impedance spectroscopy measurement when the prepared coating was exposed to NaCl solution. The simulated parametric study from electrochemical impedance spectroscopy equivalent circuit demonstrated the process of corrosion-healing kinetics. The anticorrosive mechanism was proposed based on the self-healing functionality of POTS microcapsules.

Corrosion behavior of thermal sprayed WC cermet coatings containing metallic binders in saline environment

A series of electrochemical and long-term corrosion tests were carried out in a neutral saline (5% NaCl) vapor of 35 °C on thermal sprayed WC cermet coatings containing different kinds of metallic binders in order to examine the effect of composition of binder materials on the corrosion behavior. The experimental results revealed that the overall corrosion resistance of WC-Co coating was inferior to that of WC-Co-Cr coating. For the coatings without Cr, WC-Co, general corrosion occurred in binder materials in addition to galvanic corrosion between WC particles and metallic binders in the neutral environment. By contrast, the formation of passive film in the form of surface oxide in the coatings containing Cr, WC-Co-Cr, suppressed the binder and metallic binders to be eroded. It is found that the chemical composition of metallic binder materials is one of the important factors influencing the corrosion resistance of HVOF sprayed WC cermet coatings in the neutral vapor.

Long-term Corrosion of Zircaloy-4 and Zircaloy-2 by Continuous Hydrogen Measurement under Repository Condition

ABSTRACTCorrosion behavior is a key issue for the waste disposal of irradiated metals, such as hulls and endpieces, and is considered to be a leaching source of radionuclides including C-14. However, little information about Zircaloy corrosion in anticorrosive conditions has been provided.In the present study, long-term corrosion tests of Zircaloy-4 and Zircaloy-2 were performed in assumed disposal conditions (dilute NaOH solution, pH 12.5, 303 K) by using the gas flow system for 1500 days. The corrosion rate, which was determined by measuring gaseous hydrogen and the hydrogen absorbed in Zircaloy, decreased with immersion time and was lower than the value of 2×10−2 μm/y used in performance assessment (1500-day values: 5.84×10−3 and 5.66×10−3 μm/y for Zircaloy-4, 1000-day values: 8.81×10−3 μm/y for Zircaloy-2). The difference in corrosion behavior between Zircaloy 4 and Zircaloy-2 was negligible. The average values of the hydrogen absorption ratios for Zircaloy-4 and Zircaloy-2 during corrosion were 91% and 94%, respectively.The hydrogen generation kinetics of both gas evolution and absorption into metal can be shown by a parabolic curve. This result indicates that the diffusion process controls the Zircaloy corrosion in the early corrosion stage of the present study, and that the thickness of the oxide film in this stage is limited to approximately 25 nm and may therefore be in the form of dense tetragonal zirconia.

Optically transparent Li–Al–CO3 layered double hydroxide thin films on an AZ31 Mg alloy formed by electrochemical deposition and their corrosion resistance in a dilute chloride environment

An optically transparent Li–Al–CO3 layered double hydroxide (LDH) film was formed on an AZ31 substrate by applying DC voltage between the cathode (AZ31) and the anode (stainless steel) in a Li+/Al3+ aqueous solution. The film thickness increased with electrolysis time. The LDH-coated sample exhibited low corrosion rate in NaCl solution, which was consistent with the results from long-term corrosion testing in a high-humidity environment. The LDH film exhibited good corrosion performance, which mainly resulted from its high hydrophobicity and its increased surface pH value during corrosion testing.

Long-Term Operation and Basic Corrosion Test in an Experimental Loop for Liquid Breeder in Korea

Korea has developed a liquid breeder blanket and participated in the test blanket module (TBM) program within the international thermonuclear experimental reactor group with a helium-cooled molten lithium (Li) concept. Based on this concept, helium and liquid Li were used as a coolant and breeder, respectively. Additionally, ferritic-martensitic steel (FMS) was considered as a structural material. However, according to our strategy for developing a liquid breeder TBM and its more relevant DEMO concept, not only liquid Li breeders, but also lead-Li (PbLi) breeders were considered. An experimental loop for a liquid breeder (ELLI) was constructed for the purpose of validating the design and fabrication of our electromagnetic (EM) pump; testing the effects of the magneto-hydro-dynamics; and investigating the compatibility of PbLi using structural materials such as FMS. In the ELLI, Pb-15.7Li, where Li is 15.7 at % (called PbLi hereafter), is used as the breeding material, and the EM pump circulates it through the loop. The maximum operating pressure and temperature in the loop are 0.5 MPa and 550 °C, respectively. In this paper, performance tests with the EM pump were carried out. During the three separate experiments, the EM pump was operated for 250 h with a speed of 0.16 m/s for corrosion tests. For a material of corrosion test, tubular-type specimens and cylindrical-type specimens were fabricated and installed in three test pots of the loop. After installing the specimens into the loop, the corrosion test was performed while the EM pump was operating with a 0.16 m/s flow rate at 340 °C for 250 h.

Corrosion resistance of environment‐friendly sealing layer for Zn‐coated sintered NdFeB magnet

AbstractIn this paper, a protective sealed Zn coating (SZC) was prepared on sintered NdFeB magnet by the combination of electrodeposition and sol–gel method. The unsealed Zn coating (UZC) was also studied for a comparison. The surface morphology of UZC and the cross‐section morphology of SZC were investigated using scanning election microscope (SEM). The microstructure of Zn coating and structure of sealing layer were studied by X‐ray diffraction (XRD) and Fourier transform infrared (FT‐IR) spectrum, respectively. The corrosion characteristics of SZC and UZC in neutral 3.5 wt% NaCl solution were evaluated using electrochemical measurements including open circuit potential (OCP), electrochemical impedance spectroscopy (EIS) and potentiodynamic polarization test, indicating that the anticorrosive properties of SZC coated specimens increased 20 times compared with that of UZC coated specimens. In order to further investigate the anticorrosive properties of SZC, a long‐term immersion test was carried out in neutral 3.5 wt% NaCl solution using EIS. The results of long‐term corrosion test showed that the SZC could provide long‐term protection in neutral 3.5 wt% NaCl solution for NdFeB magnet.

Corrosion behaviour of martensitic and austenitic steels in flowing lead–bismuth eutectic

The LINCE loop is a forced convection loop designed for long-term corrosion tests in lead–bismuth eutectic (LBE) at CIEMAT. The LBE volume of in the loop is 250 l and the maximum flow velocity in the region of specimens is approximately 1 m s −1. An oxygen control system has been implemented in the loop. The corrosion behaviour of AISI 316L and T91 steels was investigated in flowing LBE at temperatures of 575 and 725 K for exposure times of 2000, 5000 and 10,000 h. At 575 K, the results showed a good response, with no weight loss detected in any of the materials after exposure to the flowing LBE up to 10,000 h. A similar behaviour was observed for the specimens tested at 725 K during 2000 and 10,000 h. Specimens extracted at intermediate time (5000 h) showed an anomalous behaviour with important weight loss. These specimens were placed at the bottom of the hot test section, and this position probably made them to suffer an accused process of cavitation–erosion.

Electrochemical corrosion behaviour of plasma electrolytic oxidation coatings on AM50 magnesium alloy formed in silicate and phosphate based electrolytes

PEO coatings were produced on AM50 magnesium alloy by plasma electrolytic oxidation process in silicate and phosphate based electrolytes using a pulsed DC power source. The microstructure and composition of the PEO coatings were analyzed by scanning electron microscopy (SEM) and X-ray Diffraction (XRD). The corrosion resistance of the PEO coatings was evaluated using open circuit potential (OCP) measurements, potentiodynamic polarisation tests and electrochemical impedance spectroscopy (EIS) in 0.1 M NaCl solution. It was found that the electrolyte composition has a significant effect on the coating evolution and on the resulting coating characteristics, such as microstructure, composition, coating thickness, roughness and thus on the corrosion behaviour. The corrosion resistance of the PEO coating formed in silicate electrolyte was found to be superior to that formed in phosphate electrolyte in both the short-term and long-term electrochemical corrosion tests.

Corrosion risk of bonded, post-tensioned concrete elements

Corrosion protection for bonded, internal, posttensioning tendons (PTTs) can be effective. Within structural elements, PTTs can be well protected by a multilayer system that includes sound design of surface treatments; high-quality concrete; corrosion-resistant ducts; high-quality cementitious grout, coatings, and other internal barriers in the prestressing steel; and superior anchorage protection measures. Potential weaknesses exist in the protection system, however, when the concrete has high permeability or when the concrete is cracked. Additional weaknesses can occur when the posttensioning ducts are not adequately spliced and lack adequate impermeable concrete cover or when the portland cement grout contains voids, bleed water, or cracks. Lastly, the prestressing steel may be more susceptible to corrosion if not adequately protected or handled during construction. Over the past 2 decades, numerous countries reported damage and failure of individual tendons in transportation structures. A thorough research program, initiated at the Phil M. Ferguson Structural Engineering Laboratory, undertook 4 experimental research programs: 1) a study of high performance grouts with the goal of improving the corrosion protection of prestressing strands; 2) a long-term macrocell corrosion test series, initiated to investigate corrosion protection for PTTS to precast concrete segmental construction in which half of the macrocell lab specimens underwent internal examinations after 4-1/2 years and the other half after 8 years of exposure testing; 3) long-term beam corrosion tests that examined the effects of posttensioning and crack width on corrosion protection; and 4) a long-term column corrosion test series initiated to examine corrosion protection in vertical elements. This paper documents the results from the macrocell, beam, and column corrosion test series and recommends durability-design guidelines based on the overall research results.

Experimental studies of the interactions between anaerobically corroding iron and bentonite

In the horizontal emplacement concept (KBS-3H) for the disposal of radioactive waste, which is being developed in Sweden and Finland, copper canisters will be surrounded by bentonite buffer and placed in perforated steel containers in long horizontal boreholes in the crystalline bedrock, at a depth of ≈500 m. Under the chemical conditions in a deep repository, it is possible that the release of iron from the steel containers could influence the physico-chemical properties of the bentonite, for example, by exchange of the interlayer ions. In order to gain further insights into this process, an experimental study was undertaken, to investigate the mode of iron uptake into bentonite and the extent of changes induced in the basic physico-chemical properties of bentonite. The samples were taken from long-term anaerobic corrosion tests of carbon steel or cast iron in compacted bentonite (Na/Ca-bentonite: Volclay MX-80, ∼4% Fe 2O 3) in contact with a simple artificial groundwater at 30 °C or 50 °C. A range of analytical techniques was applied to samples of corrosion product on carbon steel and cast iron and to the bentonite surrounding the corroding specimens. Corrosion products and bentonite samples were analysed using scanning electron microscopy (SEM), electron microprobe analysis (EPMA), Raman spectroscopy, X-ray diffraction (XRD) and Mössbauer transmission spectroscopy. In addition, the bentonite samples were analysed using Fourier transform infra-red spectroscopy (FTIR), transmission electron microscopy (TEM) with energy dispersive spectroscopy (EDS) and selected area electron diffraction (SAED), exchangeable cation analysis and cation exchange capacity (CEC) measurements. Hydraulic conductivity and swelling pressure were also measured. From visual observation, the corrosion product formed on the carbon steel in bentonite was less voluminous than that formed on steel in artificial porewaters with no bentonite, although previous work showed that the corrosion rates were slightly higher in the presence of bentonite. The Raman spectroscopy analysis showed that corrosion products on the surface of carbon steel and cast iron consisted of an inhomogeneous mixture of magnetite, hematite and goethite. The predominant species was magnetite. In the bentonite, the concentration of iron decreased with increasing distance away from the iron–bentonite interface, with local iron concentrations as high as 20 wt% in some experiments. The total iron content of the bentonite in contact with corroding carbon steel wires increased by several percentage points during the experiments and the cation exchange capacity of the bentonite was reduced. After contact with corroding steel wires the hydraulic conductivity of MX-80 increased substantially. The results of the analyses were consistent with ion exchange of Fe 2+ ions with Na + ions in the montmorillonite interlayer positions but the exact chemical location of all the additional Fe 2+ ions is currently uncertain. There was no evidence for the transformation of montmorillonite to an iron-rich clay mineral phase.