Titanium Carbide Coating for Hafnium Hydride Neutron Control Rods: In Situ X-ray Diffraction Study

Abstract

This article considers the possibility of using a magnetron-deposited coating for the protection of hafnium hydrides at high temperatures as a material for neutron control rods. We describe the role of TiC coating in the high-temperature behavior of hafnium hydrides in a vacuum. A 1 µm thick TiC coating was deposited through magnetron sputtering on the outer surface of disk HfH_x samples, and then in situ X-ray diffraction (XRD) measurements of both the uncoated and TiC-coated HfH_x samples were performed using synchrotron radiation (at a wavelength of 1.64 Å) during linear heating, the isothermal stage (700 and 900 °C), and cooling to room temperature. Quadrupole mass spectrometry was used to identify the hydrogen release from the uncoated and TiC-coated hafnium hydride samples during their heating. We found the decomposition of the HfH_1.7 phase to HfH_1.5 and Hf and following hafnium oxidation after the significant decrease in hydrogen flow in the uncoated HfH_x samples. The TiC coating can be used as a protective layer for HfH_x under certain conditions (up to 700 °C); however, the fast hydrogen release can occur in the case of a coating failure. This study shows the temperature range for the possible application of TiC coatings for the protection of hafnium hydride from hydrogen release.

1. Introduction

Nowadays, boron carbide (B₄C) is commonly applied as a material for neutron control rods in nuclear reactors [1,2,3]. B₄C has a high neutron absorption cross-section in the fast energy spectrum; however, it also has a rather high cost, especially in the case of high enrichment by B¹⁰ isotopes. Moreover, B₄C has a short lifetime due to a pellet-cladding mechanical interaction failure caused by helium production in a reaction of B¹⁰(n, α)Li⁷ [4,5]. Extending the resources used for control rods to new materials can result in reduced costs for their manufacturing and disposal [6]. Materials such as Eu₂O₃, dysprosium-based materials, and Hf and its compounds (e.g., HfB₂ and HfH_x) have been intensively investigated as candidates for neutron control rods [7,8,9]. Among these alternatives, hafnium-based materials are promising candidates due to two main advantages. Firstly, Hf and its isotopes have large cross-sections of thermal neutrons, while hydrogen in Hf-H materials can moderate fast neutrons to thermal ones [4]. More importantly, no helium production occurs during reactions between neutrons and Hf isotopes, so control rods made from Hf-based compounds can have a prolonged lifetime compared to used B₄C.

Hafnium has been applied as a material for control and/or emergency rods in the USA, Russia, and Japan, and it has been used in some research nuclear reactors [7,8,9,10,11]. The combination of materials used for control and/or emergency rods has also been considered, where B₄C was partially replaced by new materials. The application of hafnium hydride in a control rod for a large, fast reactor where a B₄C control rod was originally employed was studied in ref. [9]. The possible use of control rods made from hafnium hydride in a large, sodium-cooled, fast breeder reactor was also investigated in [11]. Neutron absorbers based on HfH_1.62 can be feasible in normal situations, but the material’s absorption ability is significantly reduced at high temperatures due to hydride decomposition and the following release of hydrogen [2]. However, a comprehensive study of the properties of Hf-based materials is required to realize a full substitution of B₄C in control and emergency rods. Currently, data concerning Hf-H systems are limited in the literature. In 2021 [12], the high-temperature steam oxidation (up to 1400 °C) of metallic hafnium was studied to assess its potential application as an absorber material in light-water reactors. The heat capacity and thermal conductivity of hafnium hydride with various hydrogen contents (HfH_x with x = 1.6–1.72) from room temperature to 673 K was examined in ref. [13]. Some electronic, mechanical, and thermal properties of Hf-H systems were examined and discussed in ref. [14]. Moreover, the number of phase transitions in Hf-H systems depends on hydrogen concentration. Hafnium has two allotropic forms: hexagonal close-packed (hcp) α-Hf and face-centered cubic (fcc) β-Hf [7]. HfH_x has a tetragonal phase (ε-phase) in a range of 1.82 < x < 2.0 [8,12] followed by a gradual deformation into an fcc structure (δ-phase) at x equal 1.60 … 1.72. The tetragonal structure (deformed cubic δ′-phase) of HfH_x is observed in a range of 1.48 < x < 1.60. While pure hafnium has an hcp structure, recently, the experimental determination of the H–Hf phase diagram using in situ neutron diffraction was performed in [15]. Hence, detailed measurements and the analysis of the high-temperature behavior of Hf-H systems are urgently needed. On the one hand, the crystal structure of HfH_x and hydrogen desorption should be analyzed depending on normal operation and accidental temperatures. On the other hand, possibilities to reduce the hydrogen release from Hf-H materials at high temperatures also should be considered. In [6], aluminum diffusion coating was employed for stainless-steel cladding, with following oxidation, to drastically reduce the hydrogen transfer coefficient. However, it is more efficient to preserve the absorbing characteristics of an Hf-H sample by forming a barrier layer on its surface, preventing hydrogen release. One such approach is the deposition of a protective coating onto the surface of HfH_x tablets to limit the hydrogen release from Hf-H at high temperatures. Carbide coatings can be more appropriate due to hydrogen impermeability, high melting point, and thermal expansion coefficients close to those of HfH_1.66, up to 1000 °C [16]. Moreover, carbides do not interact with liquid sodium at the specified temperatures. At present, TiC coating is well studied, and it is widely applied as a protective material [16,17,18,19,20]. Titanium carbide is also characterized by high resistance to radiation damage that can be beneficial for using it as a protective coating for nuclear applications [21,22,23]. TiC has an fcc lattice with a parameter of 4.31 Å [24], while the HfH_x system (x = 1.60 … 1.72) also has an fcc lattice with an interplanar distance of 4.68 Å. Moreover, TiC and HfH₂ have similar linear expansion coefficients of 4.3 × 10⁻⁵ and 5.9 × 10⁻⁵ °C⁻¹, respectively [25]. Based on the above description of the HfH_x system, TiC was chosen as a candidate material for protective coating. Thus, the aim of this investigation was to determine the effect of TiC coating deposition on structural properties of HfH_x samples and hydrogen release at high temperatures (up to 900 °C) under vacuum conditions. For this, the TiC coating was deposited on an outer surface of the HfH_x samples and, then, in situ X-ray diffraction (XRD) measurements of the uncoated and TiC-coated HfH_x samples during linear heating up to 900 °C were performed in a vacuum.

2. Experimental Details

2.1. Sample Preparation

Disk samples made from a hafnium hydride (HfH_x) with a diameter of 10 mm and thickness of 1 mm were used in this study. Due to the commercial confidentiality, it is not possible to describe the technological processes of the production of the HfH_x samples. The first series of the HfH_x samples remained uncoated (in the as-received state) and no additional mechanical or heat treatment was applied. The second series of samples had the TiC coating, which was deposited using the closed-field dual magnetron sputtering system (MS) equipped with Ti (99.95%) and C (99.9%) targets. The distance from both targets to the samples was equal to 15 cm. Firstly, the samples were washed with isopropyl alcohol to remove any surface contamination, and then they were dried by air for 2 min. Subsequently, the samples were loaded in the vacuum chamber of the installation [26], which was next pumped using a turbomolecular pump TMP–403LS (Shimadzu, Kyoto, Japan) up to a base pressure of 2 × 10⁻³ Pa. The pre-heating of the vacuum chamber up to ~75 °C was necessarily used for decreasing a pressure water vapor. At the next stage, the HfH_x disks were bombarded by 1.25 keV Ar⁺ ions for 10 min to clean their surfaces of any contamination. Argon with the high purity of 99.9999% was used as a work gas. The samples were planetary rotated during both the processes of ion bombardment and coating deposition. The diameter of the planetary-rotated carousel was equal to 380 mm. Before the coating deposition, a quartz thickness gauge (Micron-5, Izovac Ltd., Minsk, Belarus) was used for determining deposition rates of Ti and C separately using the same type of MS, but it was operated in the single sputtering mode. Based on the calculated deposition rates, the power densities were determined for Ti (Q_Ti) and C (Q_C) targets sputtered by the dual configuration of the MS. The deposition conditions for the TiC coating are presented in

Table 1. The deposition parameters for the TiC coating.

QTi, W/cm2	QC, W/cm2	t, h	h, µm	Ub, V	js, mA/cm2	Tmax, °C
15.7	23.6	2.5	1.2 ± 0.1	−50	10	253

Note: Q_Ti—power density for Ti target sputtering; Q_C—power density for C target sputtering; t—deposition time; h—coating thickness; U_b—substrate bias potential; j_s—averaged ion current density on a substrate; T_max—maximal substrate temperature.

.

Initially, the coating was deposited onto a Si (110) substrate for determining the thickness and elemental and phase composition of the obtained coating (the TiC phase), which was performed using an X-ray diffractometer XRD-7000S (Shimadzu, Kyoto, Japan) with CuKα1 radiation (40 kV, 30 mA) and scanning electron microscope Tescan Vega-3 (SEM, Brno, Czech) equipped with an energy dispersive X-ray (EDS) attachment (Figure 1). The PDF4+ database was used for determining the phase composition of the coating.

After the selection of the deposition mode, the TiC coating was deposited on the disk HfH_x samples using two sequential depositions to ensure the coating on the entire surface of the samples.

2.2. In Situ XRD Diffraction of the HfH_x Samples

The in situ XRD measurements were performed using synchrotron radiation at the station “Precision diffractometry II” at the Siberian Synchrotron and Terahertz Radiation Center of the Budker Institute of Nuclear Physics of the Siberian Branch of Russian Academy of Science (Novosibirsk, Russia). The samples were examined using a wavelength of 1.64 Å in a 2θ range of 30–65°. The experiments were conducted using two maximal temperatures of T_max = 700 and 900 °C. Initially, the samples were loaded in a vacuum chamber HTK 2000N (Anton Paar, Graz, Austria), which was next pumped to 10⁻³ Pa using a turbomolecular pump Turbovac 90i/iX (Leybold GmbH, Cologne, Germany). Then, a 10 min pause was applied to determine the residual pressure in the chamber, after which the sample was linearly heated from room temperature (RT) to T_max at a heating rate of 20 °C/min. Upon reaching T_max, the samples were isothermally treated for 10 min, and then they were cooled to RT at a rate of 20 °C/min. The heating of the samples was performed using a Pt (99.9%) heater. To control the temperature, a Pt-Ro thermocouple was welded to the back side of the Pt heater. The accuracy of the temperature measurements was equal to ±3 °C. Diffraction patterns were recorded during the whole experiment.

To determine the phase composition of the samples, the reference intensity ratio method (RIR) was used according to ref. [27]:

$$W_k=\frac{\raisebox{1ex}{$I_k^{max}$}\!\left/ \!\raisebox{-1ex}{${RIR}_k$}\right.}{\sum _i{I}_i^{max}/{RIR}_i}$$

(1)

where W_k—the mass fraction of the analyzed phase, (wt.%); I_k^max—the maximum value of the intensity of the analyzed phase, (arb.units); RIR_k—the reference ratio of intensities of the analyzed phases. The PDF4+ database was used to determine the phase composition of the experimental samples.

A quadrupole mass spectrometer SRS UGA100 (Stanford Research Systems, Sunnyvale, CA, USA) was used for controlling the working atmosphere during the tests.

3. Results

3.1. In Situ XRD Study of the Uncoated HfH_x Sample

In situ XRD patterns of the uncoated HfH_x sample are shown in Figure 2, Figure 3 and Figure 4. Initially (Figure 2), the uncoated HfH_x sample had two phases: Hf and HfH_1.7. Because the HfH_x sample had low content of the metallic Hf phase in the initial state, its reflexes had low intensity (especially at ~50°), so it was difficult to track in a range of the low temperature. Up to 600 °C, only the Hf and hydride phases (HfH_1.7) were present in the diffraction patterns. Then, the formation of the δ′-phase HfH_x with the low hydrogen content (HfH_1.5) was observed. The HfO₂ phase appeared with the increase in temperature from 750 to 900 °C.

The XRD patterns of the uncoated HfH_x sample were not changed during the isothermal treatment and cooling stage (Figure 3 and Figure 4). Only some shifts of the reflexes occurred due to a thermal reduction of crystal lattices of materials during the cooling stage. This sample had the Hf, HfO₂, HfH_1.5 and HfH_1.7 phases in the case of the isothermal treatment at 900 °C and cooling stage from 900 to 30 °C.

Using the RIR method, the phase composition of the uncoated HfH_x sample was calculated for the in situ XRD measurement with T_max of 900 °C (Figure 5). Although the RIR method is quantitative, the calculation results should be analyzed, taking into account its accuracy, which is usually 2–3%. In each experiment, the concentration of the gases in the vacuum chamber was fixed using the quadrupole mass-spectrometer during the high-temperature tests at T_max of 700 and 900 °C. These dependences are presented in Figure 6. The second axis (“Temperature”) is also added in Figure 5 and Figure 6 to better illustrate the results.

According to Figure 5, the fraction of the metallic Hf phase increased from 4 to 7 wt.%, when the uncoated HfH_x sample was heated to 900 °C. At the same time, some decrease in the content of the HfH_1.7 phase (15 wt.%) was found. Oppositely, the HfH_1.5 phase appeared in this temperature range (11 wt.%). Figure 6b shows that the hydrogen signal was very low up to 700–720 °C, and then it smoothly rose. At the higher temperature (>720 °C) and during the isothermal stage (900 °C), a sharp increase in the content of the metallic hafnium, HfH_1.5 and oxide phases was observed, while the concentration of the HfH_1.7 phase noticeably decreased from 81 to 34 wt.%. Slight changes in the phase composition of the sample were observed during the cooling stage of the uncoated HfH_x sample. Based on the recording of gas concentrations, the hydrogen signal abruptly increased and had its maximum at the beginning of the isothermal stage (900 °C).Then, the signal smoothly dropped to the background value at the end of this stage. It is important that the content of the hafnium oxide phase remained the same during the cooling stage and the hydrogen concentration did not noticeably change down to RT.

When the maximum temperature of the in situ XRD measurements was equal to 700 °C (Figure 6a), the increase in the hydrogen concentration was observed during the isothermal stage. The hydrogen signal had other shape in comparison with the previous one (Figure 6b), but it could have been caused by the lower value of the maximal temperature during the test.

3.2. In Situ XRD Study of the TiC-Coated HfH_x Sample

The in situ XRD patterns of the TiC-coated sample during linear heating in a vacuum were different from the data for the uncoated sample. The coated sample initially contained the Hf, HfH_1.7 and TiC phases, and that did not change up to 800 °C (Figure 7). The TiO₂ phase appeared at ~800 °C, corresponding to oxidation of the coating by residual gases in the vacuum chamber.

The XRD patterns of the TiC-coated sample during the isothermal stage (at 900 °C) are presented in Figure 8. The figure demonstrates that the HfH_1.5 phase was formed at the starting period of the isothermal stage (~1 min), and then its intensity somewhat increased. At 2–3 min of the isothermal stage, the formation of the HfO₂ phase was observed. Then, the phase composition of the TiC-coated sample did not change qualitatively (Figure 8 and Figure 9) during the isothermal and cooling stages of the test.

Similarly to the uncoated HfH_x sample, the phase composition of the TiC-coated HfH_x sample was also calculated for the in situ XRD measurements with T_max of 900 °C (Figure 10) using the RIR method. The concentrations of gases in the vacuum chamber were recorded by the quadrupole mass-spectrometer during the in situ measurements at T_max of 700 and 900 °C (Figure 11).

According to Figure 10, a significant increase in TiO₂ content (to 19 wt.%) was observed during the heating stage of the in situ XRD test. At the same time (up to 800 °C), some reduction (16 wt.%) in the HfH_1.7 content was observed. The measurements of the chamber atmosphere using the mass-spectrometer (Figure 11b) showed a stable signal for the hydrogen concentration, and then this signal increased somewhat during the heating stage. In the case of the isothermal stage, the fraction of the HfH_1.7 phase was significantly reduced from 27 to 10 wt.%, while the content of the TiO₂ phase increased (to 27 wt.%). According to the XRD measurements in the indicated temperature range, the metallic Hf and HfO₂ phases were formed in the sample. Figure 11b shows the abrupt rise in the hydrogen signal that occurred at the beginning of the isothermal stage, and then it sharply decreased to the background value. However, the hydrogen signal had another shape in comparison with the data of the uncoated HfH_x sample: two stages can be highlighted. The first stage was the sharp rise and drop of the signal that were observed through one half of the isothermal period. The second stage continued up to the end of the isothermal period; it was characterized by a slower drop in the signal value. Later, the hydrogen signal did not noticeably change. Also, the calculations of the phase composition of the TiC-coated HfH_x sample did not show any significant changes.

Figure 11a shows that the hydrogen had a uniform signal during the in situ XRD measurement with the maximum temperature of 700 °C The hydrogen concentration increased at the end of the isothermal stage, and then it gradually decreased.

4. Discussion

The dependence of the structural properties of the Hf-H system on temperature is strongly important given the many possible applications in the nuclear industry [3,4,9]. The in situ XRD measurements of the uncoated HfH_x during the linear heating up to 900 °C show that the HfH_1.7 phase can decompose to the metallic Hf and HfH_1.5 phases, followed by the hydrogen release that was observed even at 700–720 °C (Figure 5 and Figure 6). It agreed well with the Hf-H diagram [6,28], where the decomposition of the HfH_1.7 phase to a mixture of the Hf and HfH_1.7 phases was described. The in situ XRD test with a maximal temperature of 700 °C determined the behavior of the Hf-H system. Similar regularities were also observed for the 4th group elements of the transition metals (titanium and zirconium and their hydrides) in refs. [29,30].

When the hydrogen flow was decreased, the hafnium intensively oxidized, and the outer HfO₂ layer could work as a barrier for the hydrogen release. On the other hand, the possible re-oxidation of the hafnium hydride into a dioxide (HfO₂) could occur, as was found by XPS in [31]. However, as the HfO₂ phase was detected in the sample only after the hydride decomposition, only low oxidation kinetics could be presumed under the given experimental conditions. Oxidation of the uncoated HfH_x samples can be due to the presence of some residual oxygen in the atmosphere of the chamber and desorption of oxygen from the elements in the vacuum chamber during heating. The obtained results show that some surface treatment of the HfH_x samples for their oxidation can be recommended for forming the protective scale, especially to decrease the hydrogen release.

The deposition of protective coatings also can be used to limit the hydrogen release from the Hf-H system. The TiC coating with a thickness of 1 µm was formed on full outer surface of the HfH_x samples by a magnetron sputtering of Ti and C targets in Ar atmosphere. The decomposition of the HfH_1.7 phase to the Hf and HfH_1.5 phases was also found in the sample beneath the coating. However, the TiC coating can be applied as a hydrogen barrier up to 700 °C and slightly higher that is well observed in the mass-spectrometry measurements (Figure 11). The TiO₂ phase was detected at 800 °C, before the intensive hydrogen release from the TiC-coated HfH_x sample (at 900 °C). The oxidation of the TiC up to a depth of 0.1–1 µm can occur at very low partial pressures of O₂ (10⁻⁴–2 × 10⁻² Pa), but it requires a high temperature (1500 °C) [32]. Considering the conditions in the present experiments (heating up to 900 °C), the TiC had parabolic oxidation kinetics, indicating formation of a protective scale [33] that can strongly limit inward oxygen diffusion. A more abrupt signal of hydrogen concentration was observed in the mass-spectrometry measurements compared to the data for the uncoated sample. Both of these indicated the limitation of the hydrogen flow from the sample using the TiC coating. It is important to point out that the hydrogen signal from the TiC-coated sample had a two-stage shape. This indicates several possible processes resulting in the hydrogen release, such as a coating failure and an outward diffusion of hydrogen. The change from the TiC phase to the TiO₂ results in the high (55% [34]) expansion of the oxidized coating in a volume and can accelerate its failure. The hydrogen diffusivity in hafnium is rather high (6.9 × 10⁻³ cm²/s) [35], and activation energy barriers for the diffusion of hydrogen in the metallic hafnium and hydrides are the same [28]. So, the highest hydrogen flow from the sample occurred during the isothermal stage as the TiC coating failed. The coating failure can also lead to a free pathway of the residual oxygen to the HfH_x sample, resulting in its oxidation and the growth of the protective scale (HfO₂), which can limit the hydrogen release.

The higher temperature of the hydrogen release from the TiC-coated HfH_x sample compared to the uncoated one is a beneficial point for the possible application of the protective coating for nuclear control rods. The oxidation of the TiC coating can cause its failure and should be additionally estimated under normal operation and accidental conditions. Based on the obtained results, the TiC coating can be recommended for use as a protective scale for hafnium hydrides to limit hydrogen release under certain conditions.

5. Conclusions

This study describes the possibility of using TiC coatings on HfH_x samples to decrease hydrogen release at high temperatures. A TiC coating with a thickness of 1 µm was deposited onto the HfH_x samples by magnetron sputtering. Then, in situ XRD measurements of the uncoated and TiC-coated HfH_x samples were performed using synchrotron radiation with a wavelength of 1.64 Å up to two maximal temperatures of 700 and 900 °C to determine a role for the coating. The following conclusions were made.

(1)

In the case of the uncoated HfH_x sample, the HfH_1.7 phase was decomposed to the HfH_1.5 and Hf phases, followed by an intensive hydrogen release starting from 700 °C.

(2)

The formation of an outer HfO₂ layer can limit hydrogen release from hafnium hydrides, but it was only observed after a significant decrease in hydrogen flow from the uncoated HfH_x sample.

(3)

The TiC coating can be used as a protective layer for the HfH_x sample at high temperatures (700 °C). However, this coating was partially oxidized into TiO₂ at higher temperatures, and then, after some period, was not protective.

(4)

Due to the failure of the TiC coating, the fast hydrogen release from the TiC-coated HfH_x samples was observed at 900 °C.