Synchrotron-based high-resolution photoemission spectroscopy study of ZIRLO cladding with H2O adsorption: Coverage and temperature dependence
Fig. 1 shows the Zr 3d, O 1 s, C 1 s, and Sn 3d HRPES spectra obtained from ZIRLO before the deposition of H2O. Previous study reveals that tin element alone was detected by X-ray photoelectron spectroscopy (XPS), among the minor alloying elements present in ZIRCALOY-4 cladding10. When we performed preliminary experiments using commercial XPS (VG Scientific ESCALAB 220i-XL), the Fe 2p and Nb 3d peaks were not observed in ZIRLO; moreover, the Nb 3d signal was not detected in our HRPES spectrum. Therefore, to enhance the surface sensitivity without considering the detection of the Fe 2p signal, we lowered the photon energy as much as possible, resulting in a photon energy of 710 eV. Furthermore, as the primary analysis in this study involves the Zr 3d spectra, we reduced the photon energy up to 400 eV to obtain the most sensitive Zr 3d HRPES profiles. Fig. 1a displays the Zr 3d spectra acquired at photon energies of 400, 630, and 710 eV, respectively. The Zr 3d profiles obtained at photon energies of 630 and 710 eV were similar; however, that obtained at 400 eV was different. The electron inelastic-mean-free-paths (IMFPs) at photon energies of 400, 630, and 710 eV for Zr 3d, and 710 eV for C 1 s and Sn 3d were calculated to be 7.1, 11.4, and 12.8 Å, and 14.2 and 8.3 Å, respectively, using the NIST electron inelastic-mean-free-path database (Version 1.2) with an algorithm developed by Tanuma, Powell, and Penn23,24. In addition, the IMFP at photon energy of 710 eV for O 1 s was approximately estimated as 7 Å using the universal curve19. In particular, the IMFPs for Zr 3d indicates that the spectrum at a photon energy of 400 eV contains more surface information compared to the others. Therefore, only the Zr 3d spectrum measured at a photon energy of 400 eV will be considered henceforth. The detailed analysis of the Zr 3d spectra with the peak fitting results is presented in Fig. 2.
Fig. 1b displays the O 1 s HRPES profiles, in which three peaks can be observed: O1 at 530.0 eV, O2 at 531.4 eV, and O3 at 533.0 eV. The O1, O2, and O3 peaks are assigned to the O2-, hydroxyl OH−, and the chemisorbed H2O features because the binding energies associated with them appear at 530.0–530.2 eV, 531.4 eV, and 533.4 eV, respectively, in literature25,26,27. The C 1 s HRPES profile is depicted in Fig. 1c, where two distinct peaks can be observed at 281.5 eV and 284.6 eV. We assigned the peak at 281.5 eV to zirconium carbide produced by the reaction between zirconium metal and the adsorbed hydrocarbon; this value is similar to the reported binding energy of 281.6–282.0 eV in previous investigations10,28. In addition, the peak at 284.6 eV was assigned to adventitious carbon composed of various hydrocarbon species based on the previously reported value (284.6 eV)29. As shown in Fig. 1c, the full width at half maximum (FWHM) of adventitious carbon at 284.6 eV is broader compared to that of zirconium carbide at 281.5 eV, which may be due to the existence of various types of hydrocarbons. In the Sn 3d HRPES profile (Fig. 1d), two types of Sn 3d5/2 peaks at 484.1 and 486.0 eV, and 3d3/2 signals at 492.5 and 494.4 eV, respectively, appeared with binding energy separation of 8.4 eV between them. In general, the binding energy of Sn 3d5/2 is used to analyze the Sn 3d HRPES profile. It is known that the binding energies of Sn metal and the SnO2 features in ZIRCALOY-4 cladding occur between 483.9–484.7 eV and 486.0–487.2 eV, respectively10,18. Therefore, we assigned the two Sn 3d5/2 peaks at 484.1 and 486.0 eV to Sn metal and the SnO2 compositions, respectively. Thus, by analyzing the HRPES spectra (Fig. 1) obtained from ZIRLO, we established the presence of O2-, hydroxyl OH–, chemisorbed H2O, zirconium carbide, adventitious carbon, Sn metal, and the SnO2 species.
The fact that such features appeared despite our cleaning process suggests that they may be due to the intrinsic oxygen and carbon in ZIRLO, which was also observed in prior XPS studies on ZIRCALOY-4 cladding10,21.
Fig. 2a depicts the Zr 3d HRPES profile, shown in Fig. 1a, with the peak fitting results. Because the binding energy of Zr 3d5/2 is typically utilized to analyze the Zr 3d HRPES profile, we explain the zirconium species using the binding energy of Zr 3d5/2. In Fig. 2a, six peaks can be observed: Zr1 at 179.3 eV, Zr2 at 180.4 eV, Zr3 at 181.3 eV, Zr4 at 182.4 eV, Zr5 at 183.1 eV, and Zr6 at 184.2 eV. We first assigned Zr1 at 179.3 eV and Zr5 at 183.1 eV to the Zr0 (zirconium metal) and Zr4+ (ZrO2) features, respectively, based on their binding energies ranging from 179.1–179.3 eV and 182.9–183.4 eV in literature9,10,18,30. In addition, it is known that the intervals of the binding energies between the Zr+, Zr2+, Zr3+, Zr4+ oxidation states are approximately 1 eV1,31,32. Considering the binding energy of the Zr4+ feature (183.1 eV) in our system, we assigned Zr2 at 180.4 eV, Zr3 at 181.3 eV, and Zr4 at 182.4 eV to the Zr+ (Zr2O), Zr2+ (ZrO), and Zr3+ (Zr2O3) compositions, respectively, which are zirconium suboxides. Moreover, Zr6 at 184.2 eV, of which a small quantity was present, was assigned to zirconium hydroxide (Zr(OH)4) because its reported binding energy (183.6 eV) was similar to our value33,34. Previously, through the analysis of the C 1 s HRPES profile in Fig. 1c, we had established the existence of zirconium carbide in ZIRLO. According to literature, the binding energy of zirconium carbide in the Zr 3d spectrum is 179.2 eV10,28, which is almost the same as that of zirconium metal (179.3 eV). Although the zirconium carbide component should be considered when the Zr 3d HRPES profile is analyzed, we could not confirm whether this peak was due to zirconium metal or zirconium carbide, in agreement with the previous report 10. Hence, we had to unavoidably ascribe the peak at 179.3 eV to zirconium metal.
It has been previously revealed that the order of the surface components in intact ZIRCALOY-4 cladding are as follows: The uppermost hydrocarbon, zirconium hydroxide, zirconium dioxide, zirconium suboxides, and zirconium bulk layers10. In addition, zirconium carbide is expected to exist in the hydrocarbon layer. As a result, based on the analysis of the Zr 3d, O 1 s, C 1 s, and Sn 3d HRPES spectra, along with the prior report, we concluded that the surface species in ZIRLO cladding could be the same as previously reported. Additionally, we propose that Sn metal and SnO2 compositions could exist within the zirconium bulk layer, and that the chemisorbed H2O feature could be present in the uppermost layer.
To confirm the change in the relative proportions of the zirconium features depending on the H2O coverage and annealing temperature, we calculated the ratio of each peak area obtained from the peak fitting results because the ratio of each peak integral in the Zr 3d HRPES spectra corresponds to their relative proportion (Table 1). As the proportion of zirconium hydroxide is negligible, we consider the other populations, herein. As shown in Fig. 2 and Table 1, the relative proportion of zirconium metal is the largest, and gradually decreases in the following order Zr+, Zr2+, Zr3+, and Zr4+, in accordance with their relative trends reported in literature1. After the exposure of ZIRLO to 100 L and 1000 L H2O, the relative proportion of zirconium metal decreased whereas those of the zirconium oxides including the Zr+, Zr2+, Zr3+, and Zr4+ features mostly increased (Fig. 2 and Table 1). This indicates that when H2O is adsorbed on ZIRLO, H2O and zirconium metal in ZIRLO react each other, relatively decreasing and increasing the metal population and the total quantity of zirconium oxides, respectively, in agreement with previous research on H2O adsorbed on pure zirconium and ZIRCALOY-2 samples1. This phenomenon can be explained by the dissociation of H2O into the adsorbed oxygen and molecular hydrogen gas on ZIRLO at room temperature as reported in the prior study of water molecule on Zr(0001)35. We performed annealing experiments on the 1000 L H2O system adsorbed on ZIRLO. As shown in Fig. 2d, the Zr 3d HRPES profile after annealing at 100 °C for 30 min is similar to that before annealing, indicating that any detectable change did not occur due to annealing at this temperature. However, when temperatures of 300 °C and 500 °C were applied for 30 min, the relative percentage of the Zr0, Zr+, and Zr2+ valence states increased, whereas those of the Zr3+ and Zr4+ oxidation states decreased (Fig. 2e,f). According to literature, on annealing at 200 °C or more, the oxidation states of zirconium are converted to lower oxidation states because of the decomposition of the Zr2O3 and ZrO2 compositions, and the depopulation of oxygen in the surface region accompanied by oxygen diffusion into the bulk10,20. Therefore, we concluded that the decomposition of Zr2O3 and ZrO2 and the diffusion of oxygen into the bulk lead to the reduction of the oxidation states of zirconium at 300 °C.
Fig. 3 displays the coverage and temperature dependence of the O 1 s and Sn 3d HRPES profiles. As shown in Fig. 3a, after the adsorption of 100 L and 1000 L H2O on ZIRLO, the peak related to chemisorbed H2O notably increases. This increase is attributed to the excess H2O molecules on ZIRLO, which could be the remaining quantity after sufficient reaction with zirconium metal. After annealing at 100 °C for 30 min, the profile of the O 1 s HRPES spectrum remined unchanged, in accordance with the analytical result of the Zr 3d HRPES spectrum at this temperature. When the sample was annealed at 300 °C and 500 °C for 30 min, the enhanced peaks returned to the state before the adsorption of H2O on ZIRLO, indicating that the excess H2O molecules were completely desorbed at 300 °C. In the Sn 3d HRPES profiles (Fig. 3b), the peaks related to the Sn metal and SnO2 compositions gradually disappear due to the deposition of 100 L and 1000 L H2O on ZIRLO. As the adsorption of H2O on ZIRLO increases the surface thickness, these signals may be reduced because the probing depth for ZIRLO itself relatively becomes shallow. After annealing at 100 °C for 30 min, only the peak at 486.0 eV related to the SnO2 configuration remained indistinctly in the Sn 3d HRPES spectrum. Furthermore, after annealing at 300 °C for 30 min, it completely vanished instead of recovery, despite the desorption of H2O molecules at this temperature. Therefore, we infer that SnO2 composition decomposed and that Sn metal migrated into zirconium bulk at 300 °C, sequentially.