Hydrogen Sorption in Zirconium and Relevant Surface Phenomena

Materials Transactions, Vol. 48, No. 5 (2007) pp. 1012 to 1016 #2007 The Japan Institute of Metals Hydrogen Sorption in Zirconium and Relevant Surface Phenomena Jeshin Park 1, Wonbaek Kim 1; * and Misook Won 2 1 Korea Institute of Geoscience and Minerals Resources, Daejeon 305-350, Korea 2 Korea Basic Science Institute Busan Branch, Busan 609-735, Korea The hydrogen sorption of pure zirconium was investigated as a function of activation temperature. The sorption speed increased with activation temperature up to 750 K and decreased at higher temperatures. The X-ray photoelectron spectroscopy study revealed that oxygen peak virtually diminishes on heating at temperatures above 750 K. The cross-sectional TEM showed that the surface amorphous layer crystallizes on heating at 873 K. The structural transition from amorphous to crystalline state is attributed to the decrease in hydrogen sorption. [doi:10.2320/matertrans.48.1012] (Received September 29, 2006; Accepted February 15, 2007; Published April 25, 2007) Keywords: non-evaporation getter (NEG), sorption rate, activation, deactivation 1. Introduction Non-evaporable getters (NEG) have been widely used in many fields as displays, accelerators and ultra high vacuum systems. 1) Zr-based alloys are typical non-evaporable getter (NEG) materials due to their high sorption properties for various active gases. In order for a getter to absorb the active gases, it should be heated in high vacuum or in inert environment to reveal a clean metallic surface. The process is recognized as activation. The activation temperature is considered as a temperature at which the getter material gains its full sorption capacity and thus is important for effective use of a specific getter material. However, the way to determine the activation temperature (or time) have not been documented well. One suggested method is to examine the disappearance of oxygen peaks during in-situ heating of getter material in XPS 4) (X-ray photoelectron spectroscopy) chamber. A previous study reported a maximum of hydrogen sorption speed with activation temperature for Zr 57 V 36 Fe 7 alloy. 5) The alloy, commercially known as ST707, has relatively low activation temperature ranging from 673 K and 773 K. 2) The alloy is composed of two phases i.e. AB 2 type cubic Laves and hexagonal -Zr solid solution. 3) During series of research on getter materials we also noticed similar decrease in sorption speed for pure Zr as well as various Zrbased alloys including ST 707. We, however, could not isolate reasons for the reduced sorption since the alloy systems were too complicate to analyze properly. Instead, we have chosen pure Zr to elucidate the reasons for the activation behavior in Zr alloys. In this paper, the phenomenon was investigated using XPS and TEM (Transmission Electron Microscopy). 2. Experimental Procedure *Corresponding author, E-mail: wbkim@kigam.re.kr Pure Zr powders were prepared by HDH (Hydride DeHydride) method. The raw material was Zr sponge (99.5%, Sejong Inc.). The average particle size of Zr powers was about 45 microns. The activation behavior of pure Zr was studied by the ultimate pressure measurement of the specimen chamber at room temperature after heating at 373, 473, 573, 673, 773, 873 and 973 K. To minimize the effect of chamber heating, the sample was heated inductively. The duration of heat treatment for activation was 10 minutes in all cases. The surface structure was studied by X-ray photoelectron spectroscopy(escalab 250 XPS Spectrometer) equipped with a ceramic heating stage. The XPS spectrum was obtained using a monochromatic AlK X-ray. The temperature of the samples on a ceramic heating stage was controlled up to 873 K with a temperature controller (RHC, VG Scientifics, UK). The chamber vacuum was maintained at 6:7 10 6 Pa with an ion pump (MidiVac, Varian, USA). The composition of each element was calculated using the sensitivity factors from a library of Scofield Al anode. Since the powder compact was not suitable for the XPS study, Zr sheet was used instead. The sheet sample was heated from 323 to 973 K at intervals of 50 K in the XPS system. The XPS spectrum was recorded after the sample reached to a programmed temperature. The peak intensity of the elements was divided by the corresponding sensitivity factors and plotted as a function of the activation temperature. The hydrogen sorption rate was measured at the molecular flow regime under 5 10 3 Pa as described in detail in ASTM-F798. 6) The basic principle of the method is based on the measurement of total pressure difference between gas inlet and sample chambers and the sorption rate of getter is calculated as S ¼ Q=P 2 and Q ¼ CðP 1 P 2 Þ Here, S is sorption speed of sample, Q sorption quantity, C conductance of orifice, P 1 and P 2 are the pressure of gas inlet and sample chambers, respectively. The sample chamber was made of quartz (diameter 40 mm, length 160 mm) and was evacuated with a scroll pump and a turbo pump. The sample was about 1.5 gram and compressed uniaxially to a pellet of 10 mm diameter. The hydrogen sorption rate was measured at room temperature after activation for 10 minutes at a given temperature. After the measurement, the microstructure was observed by HTEM (High-resolution Transmission Electron Microscopy) and oxygen and nitrogen contents were measured by a gas analyzer (LECO, EF-400).

Hydrogen Sorption in Zirconium and Relevant Surface Phenomena 1013 Fig. 1 Initial hydrogen sorption rates of pure Zr with activation temperature. Fig. 2 Ultimate chamber pressure after activation heating of Zr powder at each temperature. Fig. 3 X-ray photoelectron spectra during in-siut heating of Zr sheet sample. 3. Results and Discussion 3.1 Getter Property Figure 1 shows the initial sorption rate for hydrogen after activation at various temperatures. The hydrogen sorption rate of Zr increases with activation temperature. It suggests that the degree of activation increases with temperature. However, the rate starts to decline at 873 K revealing a maximum. A similar behavior was previously reported for Zr 57 V 36 Fe 7 alloy by C. Benvenuti et al. 3) The sorption rate maximum with activation temperature can be expected if the surface area reduces by sintering. The temperature range, however, was too low for any significant sintering to occur. Our study revealed no signs of macroscopic sintering throughout the experiment. Therefore, it is obvious that other reasons should be found. When a getter is activated by heating in a sealed vacuum chamber, residual gases are removed by absorption reducing or at least suppressing the pressure rise from the chamber outgassing. Once it is activated fully, no further reduction in chamber pressure is expected to occur. Therefore, the completion of activation would be identified by the minimum in ultimate pressure. It usually took more than 10 hours to reach at a stable ultimate pressure. Figure 2 shows the ultimate chamber pressure measured at room temperature after heating at each temperature. It shows that the activation of Zr occurs at temperatures between 573 and 773 K. Fig. 4 Amounts of Zr and oxygen with activation temperature. However, after heating at 873 K, the chamber pressure rises abruptly. Therefore, significant changes of Zr should occur which acts to retard the hydrogen sorption. One easy explanation would be the surface area reduction by sintering at higher temperature. However, no sign of macroscopic sintering or surface area change by BET (Brunauer-Emmett- Teller, Micrometrics, Model ASAP2400) was observed. The activation of a getter has been investigated using X- ray photoelectron spectroscopy. The activation corresponds to the observation that the surface oxide peak disappears while metallic peaks prevails. Figure 3 shows XPS peaks of

1014 J. Park, W. Kim and M. Won (c) Fig. 5 HTEM images and oxygen content at each location of pure Zr before activation and after activation at 773 K for 10 min, and at 873 K (c) for 10 min. oxygen and zirconium. The activation of Zr is identified by the disappearance of oxygen peak after heating at 573 K (Fig. 3). The variation of O1s and Zr3d spectra with activation temperature was plotted in Fig. 4. It shows that oxygen on surface diminishes with concurrent increase in Zr. At temperatures higher than 575 K virtually no oxygen could be identified which indicated the completion of activation. Nevertheless, no further information on the sorption rate maximum could be obtained from the XPS study. 3.2 Relevance of Structural Transition Figure 5(c) is the high-resolution TEM image and EDS profile near the surface. The surface of Zr was covered with about 20 nm thick oxide (298 K). After activation at 773 K, the surface oxide layer was thinned to about 15 nm. It was about 20 nm after activation at 873 K (c). The oxygen content of original sample was 22.82 at% near surface. It decreased to 16.58 at% after activation at 773 K for 10 minutes. It increased to 20.26 at% after heating at 873 K. Figure 6 shows the cross-section HTEM micrographs of the surface structure of non-activated, after activation at at 773 K and at 873 K (c). The oxide layer of original sample was amorphous. It was similar after the activation treatment

Hydrogen Sorption in Zirconium and Relevant Surface Phenomena 1015 (c) Fig. 6 HTEM images of pure-zr before activation and after activation at 773 K for 10 min, and at 873 K (c) for 10 min. at 773 K. However, after heating at 873 K, the amorphous structure is seen to crystallize. Hydrogen molecules are adsorbed on the surface and dissociated into hydrogen atoms. These atoms are absorbed on surface layer of the sample and diffuse into the bulk. Each process is the rate-limiting step for the absorption of hydrogen. It is reasonable to expect the structural change in surface layer as crystallization may have an effect on the absorption of hydrogen. The solubility and diffusivity of hydrogen in amorphous and crystalline phases has been the subject of several research. They utilized electrochemical hydrogen permeation and P-C-T (Pressure-composition-temperature) diagram to evaluate the effect of structure. In this study, the hydrogen sorption rate was determined by measuring pressure drop in closed vacuum chamber containing sample which is analogous to the PCT measurement. It is reported that at same temperature and pressure, the solubility of hydrogen in some amorphous alloys was substantially higher than in crystalline phase. 7 10) The higher solubility of amorphous alloys was explained that hydrogen dissolves in larger sites with lower ground state energy and thus more interstitial sites are available for hydrogen compared to crystalline phase. 9) It may be that the decrease in hydrogen sorption at high temperature is due to the lowered hydrogen solubility in crystallized phase. 4. Summary The activation and sorption properties of pure Zr were investigated. The hydrogen sorption rate of Zr increased with activation temperature reaching at a maximum at 773 K. Above 773 K, the sorption rate decreased again. In-situ XPS measurement showed the disappearance of oxygen peak at all temperatures above 773 K. TEM study revealed that the surface amorphous layer crystallized during heating at above 773 K. The decrease in hydrogen sorption after heating at above 773 K is due to the lower solubility of hydrogen in crystalline phase.

1016 J. Park, W. Kim and M. Won REFERENCES 1) T. A. Giogri, B. Ferraio and B. Storey: J. Vac. Sci. Technol. A3 (1985) 417 425. 2) C. Benvenuti and P. Chiggiato: J. Vac. Sci. Technol. A14 (1996) 1341 1349. 3) M. M. Gunter, D. Herein, R. Schumacher, G. Weinberg and R. Schogl: J. Vac. Sci. Technol. A16 (6) (1998) 3526 3535. 4) V. Matolin, J. Drbohalv and K. Masek: Vac. 71 (2003) 317 322. 5) C. Benvenuti and P. Chggiato: Vac. 44 (1993) 511 513. 6) ASTM F-98-97, American Society for Testing and Materials, Annual Book of ASTM Standards, (1997). 7) A. J. Maeland, L. E. Tanner and G. G. Lobowitz: J. Less-Comm Met. 74 (1980) 279 285. 8) J. J. Kim and D. A. Stevenson: J. Non-Cryst. Solids 101 (1988) 187 197. 9) Y. S. Lee and D. A. Stevenson: J. Non-Cryst. Solids 72 (1985) 249 266. 10) R. Kirchheim, T. Mutschele, W. Kieninger, H. Gleiter, R. Birringer and T. D. Koble: Mat. Sci. & Eng. 99 (1988) 457 462.