Acid Dissolution Behavior of Ferritic FeCrAl Tubes Candidates for Nuclear Fuel Cladding

For over half a century at least 30 countries have been generating electricity for the civilian grid using energy obtained from nuclear fission. In a reactor, the energy released during the fission of uranium 235 is captured by water to produce steam, which is used to move the turbines that generate electricity. The reactors have been traditionally built using common simple materials such as carbon steels, austenitic stainless steels, and some nickel-based alloys. Zirconium (Zr) alloys such as Zircaloy-2 and Zircaloy-4 were used for the cladding of the urania fuel pellets.¹  The total mass of Zr in each light water reactor could be in the order of 40 tons.² 

After the March 2011 accident at the Fukushima power stations, there was a consensus in the international nuclear materials community to remove all Zr components from inside the reactor^1-2  by manufacturing the fuel rods using accident tolerant materials.^1-7  The safer fuel alternatives are called accident tolerant fuels (ATF) or advanced technology fuels (ATF), and they can be classified into two large groups considering the material of the two components of the fuel rod, (1) newer cladding or (2) newer fuel. The desirable attributes of the ATF cladding are mainly high-strength and -oxidation resistance at accident temperatures to maintain the architecture of the rods and avoid the release of fission products from the rods during the accident. The main desirable attributes of an ATF fuel are enhanced thermal conductivity and resistance to premature fragmentation.¹  In the family of ATF cladding, three main concepts are currently under development:

• Coated zirconium alloys,

• Iron-chromium-aluminum (FeCrAl) monolithic, and

• Silicon carbide composites.

For the ATF fuel pellets, the main concepts under development are:

• Advanced doped UO₂ fuels,

• Uranium nitride (UN), and

• Uranium silicide (U₃Si₂).

The aim of the current manuscript is to advance the characterization of FeCrAl materials for the cladding. Ferritic FeCrAl alloys were never used in a nuclear reactor; therefore, their behavior needs to be assessed in the entire fuel cycle (Figure 1). The specific aim of the current research is to evaluate the relative capability of reprocessing FeCrAl cladding as compared to the traditional Zirc-2 cladding (step 5 in Figure 1). Immersion and electrochemical corrosion tests were conducted on FeCrAl and other engineering alloys in three mineral acids (hydrochloric, sulfuric, and nitric) at 30°C, 60°C, and 90°C.

In the development of ATF, not only the physical, chemical, and mechanical properties of the materials need to be considered but also the economics of the entire fuel cycle as well as the safety of the operation of the nuclear power station. The newer ATF concepts like FeCrAl cladding need to perform well in the entire fuel cycle (Figure 1). The ATF fuels need to be manufactured in an industrial scale in a competitive and efficient manner, then the fuel rods need to withstand the reactor normal operation conditions maybe up to 10 y, then the used fuel needs to be transported safely to cooling pools for maybe another 20 y storage (step 3 in Figure 1), and later probably to a dry cask storage for 100 y or sent to reprocessing to reduce the total volume of the waste and to extract the useable components from the used fuel. Eventually, the long-lasting radioactive components of the fuel may need to be disposed in a geologic repository (step 6 in Figure 1).

Used fuel reprocessing is convenient in order to reduce the total amount of used fuel for disposal as compared to used fuel from a once-through process.⁸  Countries like France, Japan, United Kingdom, Russia, and India favor the separation of useable material from the long-lived and highly radioactive material, which will be going to final geologic disposal.^8-9  The International Atomic Energy Agency (IAEA) reports that the most suitable process to separate the components in the used fuel is the solvent extraction process.⁸  Using solvent extraction, the three main components (uranium, plutonium, and waste—such as fission products and minor actinides) can be separated in a large-scale continuous process involving remote operation.⁸  The IAEA mentions that the PUREX process was perfected during several decades of use for the separation of plutonium for military purposes.⁸  The U.S. Nuclear Regulatory Commission (NRC) also lists the PUREX process as an alternative by which the irradiated fuel may be dissolved in nitric acid and then the uranium, plutonium, and highly radioactive fission products can be separated by solvent extraction.¹⁰  The PUREX process was considered for fuel clad in ferritic ODS steels.¹¹  The authors reported that the dissolution rate of the ferritic steels in nitric acid at 90°C decreased when the concentration of chromium in the ferritic steels increased from 9% to 14% to 18%.¹¹  Recently, the corrosion properties of Al-containing 17Cr ferritic ODS steels were evaluated in typical reprocessing conditions such as systems containing nitric acid.¹²  It was reported that this 17Cr steel remained mostly passive in the nitric acid media probably by the formation of a surface oxide film rich in Cr₂O₃ and Al₂O₃.¹² 

Corrosion studies have also been conducted on the behavior of materials used for the construction of the acid processing equipment.^13-15  Their results show resistance to corrosion by passivation in nitric acid of Zirc-2 and austenitic stainless materials.

There is no information on the dissolution behavior of FeCrAl alloys in common mineral acids, therefore the objective of this research was to study the dissolution of typical cladding tubing having two compositions of FeCrAl (APMT and C26M) (Table 1) in three common mineral acids (H₂SO₄, HNO₃, and HCl) as a function of the temperature (30°C, 60°C, and 90°C) using both standard ASTM G1 and G31 immersion tests as well as electrochemical tests such as ASTM G59, G102, and G106. The dissolution rate of the FeCrAl alloys is compared to the rate of other traditional nuclear materials such as austenitic stainless steels (304SS [UNS S30400]⁽¹⁾ and 316SS [UNS S31600]) and austenitic nickel alloys (Alloy 600 [UNS N06600] and Hastelloy C-276^† [UNS N10276]) (Table 1). All of the materials investigated in Table 1 were tubes.

where Δm is the mass change after each period of immersion, δ the density of the alloy (Table 1), A is the surface area of the exposed specimen, and t is the immersion time.

where k is a conversion factor, and i_corr is the corrosion current density in A/cm² (calculated from the measurements of the resistance to polarization, Rp) (Equation [2]), EW is the equivalent weights for each of the five alloys calculated according to ASTM G102, and d is the density (Table 1).

The electrochemical impedance spectroscopy (EIS) tests were performed at a potential 10 mV above the corrosion potential from 100,000 Hz to 0.01 Hz taking five measurements per decade. Lastly, a potentiodynamic anodic polarization was conducted at a scan rate of 0.167 mV/s from 150 mV below the corrosion potential to +400 mV_SCE.

As the corrosion rate of alloys in mineral acids increases rapidly as the temperature increases, the immersion tests were conducted at 30°C for three periods, 24 h, 72 h, and 168 h, at 60°C for 24 h, and at 90°C for 3 h. Tables 2, 3, and 4 show the corrosion rates values in mpy (mils per year) measured for the seven alloys in the three acids at 30°C. Figure 4(a) shows the 63 (nine for each alloy) tube specimens that were used for immersion corrosion testing at 30°C in sulfuric, hydrochloric, and nitric acid. Figure 4(b) shows the display of the coupons during the immersion of 21 specimens in hydrochloric acid (three specimens for each alloy). The volume of the acid was large enough (3 L) to comply with the coupons exposed surface-to-volume ratio recommended by ASTM G31 standard.

Figure 5 shows the effect of testing time on the corrosion rate of Zirc-2, APMT, and C26M coupons in (a) sulfuric acid, (b) hydrochloric acid, and (c) nitric acid at 30°C. There was a good reproducibility of the corrosion rate results for the three specimens of each alloy. Please note that the Y-axis is in the logarithmic scale, that is, the corrosion rates values between the three types of alloys shown are generally large. Figure 5 shows that the highest corrosion rate in the three acids was always for C26M, probably due to its ferritic nature and its lower content of Cr. The lowest corrosion rate was always for Zirc-2 coupons showing the stability of the surface zirconium oxide (ZrO₂) in the mineral acids. Looking at the data for 24 h immersion, the most corrosive acids for Zirc-2 coupons were nitric and sulfuric. The most corrosive acid for C26M was sulfuric and the least corrosive was nitric. For APMT the most corrosive acid was hydrochloric, and the least was sulfuric and nitric. The corrosion rate data for APMT in hydrochloric acid were less reproducible. Data show the high stability of chromium oxide in nitric acid at the low temperature of 30°C.

In general, Figure 5 also show that as the testing time increased, the corrosion rate slightly decreased. This could be related to the transformation of the surface of the coupons. The corrosion rate is inversely proportional to the area of the specimens (Equation [1]) and for the calculations of corrosion rate in Tables 2 through 4 and Figure 5, the original surface area of the specimens was used. Figure 5 data also show that probably it is sufficient to measure dissolution rates for only 24 h immersion.

Figure 6 shows the corrosion rates for the seven alloys after 24 h immersion in the three acids at 30°C. For most of the Cr-containing alloys, the most corrosive acid was hydrochloric showing the instability of the chromium oxide (Cr₂O₃) film in this acid due to the formation of the soluble species chromium chloride (CrCl₃). The least corrosive acid tended to be nitric because it promotes the formation of the Cr₂O₃ passive film on the surface. Alloy C26M with the lowest content of Cr (Table 1) showed the highest corrosion rate. It was somehow surprising that for C26M, sulfuric acid was the most aggressive. As expected, Hastelloy C-276 had equally good resistance to corrosion in the three mineral acids.

Figure 7 shows the appearance of the surface of two corroded tube specimens after 168 h immersion in hydrochloric acid at 30°C. It is clear the higher intergranular dissolution of the C26M coupon compared to the Type 304SS coupon.

Tables 5, 6, and 7 and Figures 8 (a), (b), and (c) show the effect of temperature on the corrosion rate for the seven alloys tested (Table 1). The corrosion rate increased significantly as the temperature increased. The testing time had to be reduced to 3 h at 90°C, and even in this short time the two FeCrAl (C26M and APMT) completely dissolved in sulfuric acid (Figure 8[a]). Meanwhile, the corrosion rate of Zirc-2 was little affected by the temperature in the range 30°C to 90°C. Nitric acid (Figure 8[c]) seemed to be more corrosive to Zirc-2 while sulfuric acid was most corrosive toward C26M (Figure 8[a]). Hydrochloric acid was most corrosive toward APMT (Figure 8[b]). The corrosion rate of both of the FeCrAl alloy was much higher than the corrosion rate of Zirc-2. For example, sulfuric acid at 60°C, the corrosion rate of C26M was more of four orders of magnitude higher than the corrosion rate of Zirc-2 (0.1 mpy vs. >1,000 mpy) (Figure 8[a]). Similarly, the corrosion rate of APMT in hydrochloric acid at 90°C was also more than four orders of magnitude higher than for Zirc-2 (Figure 8[b]). Even in the less corrosive nitric acid, the corrosion rate at 90°C of C26M was more than three orders of magnitude higher than for Zirc-2 (Figure 8[c]).

Figure 9 shows the instantaneous corrosion rates of the seven tested alloys in 0.5 M H₂SO₄, 1 M HCl, and 1 M HNO₃ at 30°C. There are two data points for each alloy in each acid, and the results were highly reproducible. In general, most of the seven alloys had the highest corrosion rate in hydrochloric acid (similar to immersion tests results in Figure 6). The current cladding material (Zirc-2) had one of the highest resistances to corrosion (together with alloy C-276) in the three acids. The highest corrosion rates in the three acids were for alloy C26M, probably due to its lowest content of Cr (Table 1). For example, C26M dissolved in hydrochloric acid at a rate of four orders of magnitude higher than Zirc-2. APMT had low corrosion rates at 30°C (Figure 9) but its corrosion rate increased significantly at 60°C and 90°C (not currently shown for electrochemical tests).

Figure 10 shows the EIS Bode Diagrams for four alloys (Zirc-2, 304SS, APMT, and C26M) in the three mineral acids at 30°C. The left column shows the values of impedance as a function of the frequency and the right column shows the phase angle as a function of the frequency. Figure 10(a) shows that in sulfuric acid the impedance for C26M at the frequency of 0.01 Hz was only 10 Ω·cm², which is more than four orders of magnitude lower than the impedance of Zirc-2. This agrees well with the corrosion rates in Figures 9 and 6. The plot of phase angle vs. frequency in sulfuric acid in Figure 10(b) shows that C26M was actively corroding while the other three alloys had a small degree of passivation, probably due to their higher Cr content, which is reflected in their higher impedance in Figure 10(a) and their lower corrosion rates (Figure 9).

The impedance Bode plots for Zirc-2 and C26M in Figure 10(c) for hydrochloric acid shows a behavior similar to that in sulfuric acid. However, the impedance behavior of APMT and 304SS in hydrochloric acid was much lower than in sulfuric acid (Figure 10[a]). Hydrochloric acid seems to be more corrosive toward a chromium oxide surface film. This is well represented in the phase angle plots in Figure 10(d).

Figure 10(e) shows that the impedance values in nitric acid at a frequency of 0.01 Hz of the four alloys were closer to each other (less than two orders of magnitude separation) mainly because the impedance was lower for Zirc-2 compared to sulfuric and hydrochloric and higher for C26M (∼1,000 Ω·cm²). Figure 10(f) shows that Zirc-2 may not have a protective oxide on the surface. Figure 10(f) shows the presence of an oxide film protecting 304SS and especially APMT, which is reflected in Figure 10(e) and in the corrosion rate values in Figure 9.

Figures 11, 12, and 13 show, respectively, the potentiodynamic polarizations of the seven alloys (Table 1) in 0.5 M H₂SO₄, 1 M HCl, and 1 M HNO₃ at 30°C. In sulfuric acid (Figure 11), the alloys with >15% Cr were all passive with a high E_corr of approximately +200 mV_SCE and a corrosion current of approximately 0.1 μA/cm², while the C26M showed an E_corr of −500 mV_SCE (no passive film present) and a corrosion current of approximately 1 mA/cm² or four orders of magnitude higher than for the passive alloys. The corrosion potential for Zirc-2 was approximately +20 mV_SCE and the corrosion current was approximately 0.01 μA/cm².

In hydrochloric acid (Figure 12), Zirc-2 had a corrosion current of approximately 0.01 μA/cm² and a corrosion potential of near +100 mV_SCE. The C26M and APMT alloys had a low corrosion potential of near −500 mV_SCE and corrosion currents in the order of 1 mA/cm², which is five orders of magnitude faster corrosion than for Zirc-2. Comparing Figures 11 and 12 it is clear that APMT is less corrosion resistant in hydrochloric than in sulfuric at 30°C. As the potential increased, APMT showed pseudo passivation, however, the C26M showed higher dissolution rates than APMT.

Figure 13 shows that in nitric acid at 30°C, the C26M alloy had a higher dissolution rate than Zirc-2 alloy. For Zirc-2 the E_corr was approximately +400 mV_SCE and the i_corr was approximately 0.1 μA/cm² while for C26M they were −200 mV_SCE and 10 μA/cm², respectively. C26M had a higher corrosion potential in nitric acid compared with sulfuric and hydrochloric, suggesting the presence of an oxide film on the surface in nitric acid.

The comparison of the values of corrosion rate between Figures 6 and 9 (electrochemical vs. immersion) shows a good correlation of the values, even though one represents instantaneous dissolution (Figure 9) and the other is the average over a time of 24 h (Figure 6). For example, for C26M in sulfuric acid, the electrochemical and immersion testing showed a corrosion rate of approximately 1,000 mpy or 25 mm/y. For C26M both techniques showed that the highest corrosion rate was in sulfuric acid and the lowest in nitric acid.

The new data presented in this manuscript regarding the dissolution rate of engineering alloys such as Type 304 and 316 stainless steels as well as Inconel 600^†, Hastelloy C-276, and Zircaloy-2 agree well with previously published data for these common industrial alloys.^16-17  However, there were no data on the corrosion rate of common mineral acids of the alloys in the family of FeCrAl. This current work shows that the powder metallurgy C26M and APMT alloy with a ferritic microstructure have, in general, higher corrosion rates in mineral acids than other alloys such as the traditional metallurgy austenitic Type 304 and 316L (UNS S31603) stainless steels.

For the application related to used nuclear fuel reprocessing through the wet method of solvent extraction, the presence of a FeCrAl cladding will facilitate this process as compared to the current system using a Zircaloy-based cladding, due to its higher solubility and the easier access of the acid to the desired fuel. In any of the three mineral acids tested, the dissolution rate of the FeCrAl material was orders of magnitude faster than for Zirc-2 current cladding material. It is likely also that the powder metallurgy and ferritic nature of C26M and APMT may have contributed to their higher dissolution rate in mineral acids.

It was of interest to evaluate the reprocessing capabilities of the proposed FeCrAl materials for ATF cladding. Electrochemical and immersion corrosion tests were conducted in three mineral acids (sulfuric, hydrochloric, and nitric).

• Results show that at 30°C C26M had orders of magnitude higher dissolution rate than Zirc-2, which may facilitate reprocessing solvent extraction processes.

• The highest dissolution rate of C26M at 30°C was in sulfuric acid and the lowest was in nitric acid. APMT had a higher dissolution rate in hydrochloric than in sulfuric acid.

• Immersion tests showed little effect of testing time, that is, similar corrosion rate data were obtained for 24 h immersion as for 168 h immersion.

• Temperature was a considerable accelerator of the corrosion dissolution rate. By increasing the temperature from 30°C to 90°C, the corrosion rate for C26M and APMT in hydrochloric acid increased two orders or magnitude.

• Results show that both immersion and instantaneous corrosion methods provided equivalent results regarding the values of the dissolution rate of the seven tested alloys. Electrochemical tests are faster to conduct.

The technical expertise of R.J. Blair is highly appreciated. This material is based upon work supported by the Department of Energy [National Nuclear Security Administration] under Award Number DE-NE0009047. The financial support of Global Nuclear Fuel and GE Research is gratefully acknowledged. This report was prepared as an account of work sponsored by an agency of the U.S. Government. Neither the U.S. Government nor any agency thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the U.S. Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the U.S. Government or any agency thereof.