Deuterium permeation in Er$_2$O$_3$ thin film fabricated on a type 316L stainless steel substrate

TL;DR

This study demonstrates the fabrication of Er$_2$O$_3$ thin films on stainless steel for hydrogen isotope permeation barriers, showing effective reduction at specific high temperatures but degradation at higher temperatures.

Contribution

It introduces a new Er$_2$O$_3$ thin film fabrication method for permeation barriers and evaluates its performance at high temperatures.

Findings

Permeation reduction most effective at 650°C.

Film quality confirmed by microscopy and spectroscopy.

Degradation occurs above 800°C.

Abstract

A metal-oxide film can be used as a hydrogen-isotope permeation barrier in the fuel circulation system for nuclear fusion. We fabricated Er$_2$O$_3$ thin film on a type 316L stainless-steel substrate by using a metal-organic chemical vapor deposition technique for the purpose of hydrogen-isotope permeation barrier. Electron microscopy based imaging and energy-dispersive X-ray spectroscopy measurements indicate a sound film quality together with X-ray diffraction experiments. We also measured deuterium permeation in the film at high temperatures ranging from 600 $^{\circ}$C to 800 $^{\circ}$C. The permeation reduction was most apparent at 650 $^{\circ}$C. Above 800 $^{\circ}$C, we confirmed that the film was damaged and did not work as a permeation barrier.

Tables

Table 1. Table 1: Permeation reduction factors at different temperatures

temp. ∘C	600	650	700	750	800
from Ref. Wu	291	235	151	-	-
this study	592	881	510	76	23

Full text

nuclear fusion, hydrogen permeation, Er2O3 film, chemical vapor deposition

I Introduction

Hydrogen isotopes of deuterium and tritium are fuels for nuclear fusion experiments. In the fusion fuel delivery and storage systems, such hydrogen-isotopes can permeate through the fuel lines and can be released to outside during the operation of a nuclear fusion reactor, which is an environmental issue. Coated metal-oxide (M-O) films can reduce fuel permeation because the permeation process in M-O films is different from that in a bare stainless steel. It is known that hydrogen easily permeates through SS fuel lines by thermal diffusion. However, when hydrogen atoms meet a surface of a M-O film, the hydrogen atoms tend to break M-O bond and form H-O bond until they replace all of the M-O bondLee ; Wu . Normally, the M-O bonding energy is larger than the H-O bonging energy. To elongate the permeation process, it is desired to have large bonding energy difference between the M-O and H-O bonds. Up to date, Al2O3 film coating was known to be an excellent permeation barrier. However, in the recent years, Er2O3 film has been suggested to be a good candidate Wu ; Levchuk for replacing Al2O3 film as a permeation barrier, because an Er2O3 film can be easily obtained and has a large M-O binding energy in comparison to an Al2O3 film. In this regard, various Er2O3 coating techniques and substrate materials are vastly adopted to improve hydrogen-isotope permeation reductionWu ; Levchuk ; Yao ; Liu ; Chikada1 ; Mao ; Chikada2 ; Wang ; Li . In the recent years, theoretical calculations on hydrogen permeation through grain boundaries on M-O films have been reported by using ab initio density fuctional theory, which indicated that proper analysis of the microstructures of M-O films is essential to understand of the hydrogen permeation mechanismMao .

In this study, we tried to deposit Er2O3 film on a type 316L stainless-steel (SS316L) substrate by using a chemical vapor deposition (CVD) technique. We obtain a 480-nm-thick Er2O3 film, which shows appreciable amount of permeation reduction factor (PRF) for deuterium at temperature below 700 ∘C. With increasing temperature above 700 ∘C, the PRF reduced drastically due to the fast activation of Er2O33 film. We will discuss structural and permeation properties of coated Er2O3 film in the next sections.

II Experiment

A schematic of the CVD system used for this study is shown in Fig. 1. A 50-mm-diameter and 600-mm-long quartz tube was used as the main reactor chamber. The quartz reactor is surrounded by a heater that can heat the reactor up to 1200 ∘C. A 20-mm-diameter with 1-mm-thick SS316L coin shaped substrate was located on a square-shape quartz substrate holder, which is equipped at the reactor center with a vertical angleof 30∘. The substrates were carefully polished to a mirror finish by using a lapping machine. Before being coated, the substrates were heat treated at 700 ∘C in the reactor in hydrogen atmosphere to removed the surface contaminant. Tris(2,2,6,6-tetramethyl-3,5-heptanedionato) erbium, or simply called Er(tmhd)3 was used as a precursor for Er2O3 coating. The Er(tmhd)3 precursor is known as stable while transporting from the sublimator, which is quickly decomposed and formed to Er2O3 above 500 *∘*C in oxygen atmosphere. The sublimator and the transport tubes are composed of stainless steel, which were wrapped with heater stripes. The whole units including sublimator, transport lines and heater stripes were wound by thermal insulating materials to reduce the temperature fluctuation during the process. High purity Ar gas was used as a precursor carrier, which is regulated by mass-flow controller (MFC) with the flow rate of 50 sccm. After passing through the MFC, Ar gas was heated to 180 *∘*C. Therefore, the sublimated precursor’s temperature was maintained at 180 ∘C in the sublimator and the transport lines. The flow rate of O2 gas was 100 sccm. During the coating process, the pressure inside of the reactor was maintained at 10 mbar by using an automatic pressure regulating valve, which is mounted between the chamber and the vacuum pump. The total deposition time was 4 hrs. and after finishing the film deposition process, samples were slowly cooled down by 0.5 *∘*C/min. to room temperature. One thing of note is the clogging of the precursor in the precursor transport line after the film growth. Even though Er(tmhd)3 is known as a stable precursor, a significant amount of the precursor was deposited inside of the transport line. Such precursor clogging in the transport line not only elongated the coating time but also deteriorated the sample quality. Therefore, the used transport line had to be replaced with a new one after every deposition.

The estimated Er2O3 film thickness was measured to be 395 nm by using an ellipsometer (Woolim Model M-2000). For further analysis of the coated Er2O3 film, we performed scanning electron microscopy (SEM) image, energy dispersive spectrometry (EDS), and X-ray diffraction measurements. To obtain the deuterium permeation, we used a home-built hydrogen-isotope permeation measurement system (HPMS). Figure 1b shows a schematic diagram of the HPMS, which consists of a sample mount, a heater and residual gas analysis (RGA) modules. The supplying deuterium gas pressure was maintained at 1 bar during the permeation measurements. To minimize the sample temperature fluctuation, the heater unit is located in a vacuum chamber as a thermal shield. The partial pressure of deuterium at the permeate side was monitored by using the RGA, which is converted to the permeability of deuterium. The detailed development and operation procedure of the HPMS can be found elsewhereLee2 ; Lee3 .

III Results and Discussion

By using an SEM and an EDS, we obtained surface images and atomic composition of the sample before and after the permeation measurements. Figure 2a exhibits an SEM image of a fresh as-grown Er2O3 film before the permeation measurements. It shows granular shape grains with the average size of 35 nm. After the high temperature permeation measurements at 850 *∘*C, as seen in Fig. 2b, the average grain size was increased to $\sim$ 50 nm. However, the film was damaged and no longer worked as a permeation barrier at this elevated temperature.

The EDS analysis shows that the atomic percentile contents of O and Er is 43.8 % and 22.7 %, respectively, for as-grown sample before the permeation experiments. The residual 33.5 % comes from the substrate (SS316L). The content ratio between O to Er is 1.93, which is higher than the stoichiometric ratio 1.5 due to the surplus oxygen contents. After the permeation experiments, the atomic percents of O and Er changed to 34.1 % and 17.0 %, respectively. In this case, the atomic content ratio between O and Er is 2.0. Exposure to high temperature deuterium may reduce the amount of O and Er contents. The similar contents ratio before and after the permeation measurements suggests that the surplus oxygen contents may come from the native oxide formed on the SS316L surface. Even though the sample exposed to the high temperature deuterium, it cannot affect the native oxide on the surface of SS316L substrate, which does not alter the O to Er contents ratio.

The XRD spectra of Er2O3 and SS316L substrate are displayed in Fig. 3. The peaks at 50.7*∘* and 74.6*∘* are from FeCrNiC SS316L substrate. The peak at 43.6*∘* is the combined peak of Er2O3 (431) and SS316L (111) planes, and all other peaks are indexed to be of Er2O3. As seen in the figure, the XRD peak intensities of Er2O3 became sharper (smaller line-widths) and stronger after the permeation experiments compared with those of the as-grown sample. This is due to the improvement of the crystallinity by the long-term exposure at high-temperature during the permeation experiments. However, though the improvement of the crystallinity, the Er2O3 film was significantly damaged due to the large mismatch of the thermal expansion coefficients between the substrate and the coated film at such high temperature as seen in Fig. 2b.

Arrhenius plots of the deuterium permeabilities for both the base and Er2O3-coated SS316L samples are displayed in Fig. 4. Solid square markers are the data measured in this study. To compare our permeability data with those of the similar deposition technique from other research group, data extracted from Ref. Wu are displayed as circular markers. The permeability was recorded when deuterium permeation reached the steady state. We measured permeability at five different temperature points from 600 *∘*C to 800 *∘*C by 50 *∘*C steps. As seen in the figure, the permeability increases with increasing temperature. Below 700 ∘C, the slope and the amount of the permeability is moderate and small, respectively. However, above 700 ∘C, the permeability slope rapidly increases. In comparison, the values measured in this study are significantly lower than the values from Ref. Wu (solid circle). Permeation reduction factor (PRF), which is the ratio between the permeability of SS316L and Er2O3, is summarized in Table 1. A larger PRF value indicates a better permeation protection. The PRF at 600 *∘*C is 592, which increases to 881 at 650 *∘*C, then decreases to 510 at 700 *∘*C. It is not clear such a sudden increase of PRF at 650 *∘*C. One plausible explanation can be an annealing effect. As mentioned above, the permeability values were recorded at the steady state; it takes hours to days to be reached at the steady state depending on the designated temperature. During the 600 *∘*C and 650 *∘*C measurements, the sample was exposed several hours at such elevating temperature exceeding the growth temperature, and the improved crystallinity of the sample during the measurements possibly increases the PRF at 650 *∘*C.

IV Conclusion

We fabricated an Er2O3 film on a SS316L substrate by using a CVD method. After 4-hour growth, the film thickness was measured to be 395 nm by using an ellipsometry. SEM images indicate that grains with average size of $\sim$ 30 nm for an as-grown polycrystal film become large enough to induce crack damages after the deuterium permeation measurements at 850 ∘C. By using the home-built HPMS, we measured the deuterium permeability of Er2O3 film at temperatures between 600 *∘*C and 800 *∘*C by 50 *∘*C steps. The measured permeation reduction was most apparent at 650 *∘*C. Above 700 *∘*C, the permeability decreases rapidly with increasing temperature, and at 850 *∘*C, the film was damaged and no longer worked as a permeation barrier.

Acknowledgments

The present research was supported by the research fund of Dankook University in 2017. One of the author, YK thanks Dr. Y. H. Shin and Prof. S. J. Noh for their experimental supports.