Modelling and experimental analysis of the effect of solute iron in thermally grown Zircaloy-4 oxides

Abstract

Simulations based on density functional theory (DFT) were used to investigate the behaviour of substitutional iron in both tetragonal and monoclinic ZrO 2. Brouwer diagrams of predicted defect concentrations, as a function of oxygen partial pressure, suggest that iron behaves as a p-type dopant in monoclinic ZrO 2 while it binds strongly to oxygen vacancies in tetragonal ZrO 2. Analysis of defect relaxation volumes suggest that these results should hold true in thermally grown oxides on zirconium, which is under compressive stresses. X-ray absorption near edge structure (XANES) measurements, performed to determine the oxidation state of iron in Zircaloy-4 oxide samples, revealed that 3+ is the favourable oxidation state but with between a third and half of the iron, still in the metallic Fe 0 state. The DFT calculations on bulk zirconia agree with the preferred oxidation state of iron if it is a substitutional species but do not predict the presence of metallic iron in the oxide. The implications of these results with respect to the corrosion and hydrogen pickup of zirconium cladding are discussed.

Understanding the role of Fe, Cr and Ni in Zircaloy-2 with special focus on the role of Ni on hydrogen pickup

“…The temperature used in this study is 1500 K (rather than the cladding operating temperature of about 600 K) since below this temperature the tetragonal phase is not stable without stress or doping. Previous studies [19,[33][34][35][36] that are based on similar simulation techniques have also taken this approach and have obtained results consistent with experiment. Furthermore, Bell [37] demonstrated the analogous effects of increased temperature and stress on the predicted defect concentrations.…”

“…Using the formation energies, Brouwer diagrams were generated with concentrations of 10 −4 units of dopant per ZrO 2 , which roughly corresponds to the amount of Fe, Cr, or Ni found in regions outside areas of elevated concentration in Zircaloy-2 and Zircaloy-4 cladding [20,28,45], bearing in mind that the solute solubilities would be lower. It has also been previously shown that identification of dominant defect species is not very sensitive to the imposed dopant concentration during the construction of Brouwer diagrams [19].…”

“…As such, two cluster arrangements were considered: one Fe atom on a Zr site with an associated oxygen vacancy {Fe Zr : V O }, and two Fe atoms on nearest neighbour Zr sites sharing an oxygen vacancy {Fe Zr : V O : Fe Zr }. Detailed results from the simulation of these Fe defects have been published previously [19]. Here we only present the Brouwer diagram relevant to this work.…”

“…Figure 1 shows the predicted defect concentrations in tetragonal ZrO 2 as adapted from [19]. Across all P O 2 , the dominant Fe defect is the trimer defect, with the prefered Fe oxidation state decreasing from 3+ to 2+ as P O 2 decreases.…”

“…Experimental X-ray absorption near edge structure measurements (XANES) have shown that in the oxide layer of corroded zirconium alloys, Fe mainly assumes 3+ charge state, while the 2+ charge state has also been observed in significant amounts [15][16][17][18][19]. Results of density function theory (DFT) simulations suggest the increase in hole concentration as the charge compensation mechanism for the Fe defects may lead to a reduction in HPUF [19]. The same study also showed that the improved corrosion resistance imparted by Fe may be due to the Fe ions binding strongly (> 2 eV) to oxygen vacancies.…”