Formation and Erosion of Wc under W Irradiation of Graphite

The bombardment of C with 100 keV and 1 MeV W at normal incidence is studied as a function of the incident W fluence experimentally and by computer simulation with the program TRIDYN. Calculated oscillations in the amount of retained W and in the target weight change are confirmed experimentally for 100 keV at room temperature. XPS investigations show W2C formation during ion implantation already at room temperature. RBS depth profiles for 1 MeV bombardment show W mobility and surface segregation even at liquid nitrogen (LN2) temperatures. At elevated temperatures W clusters to form nano-particles at the surface and the oscillations in the retained amount of W disappear. INTRODUCTION In a future large fusion machine, ITER [1], three elements have been proposed to serve as first wall materials in different locations; Be for the main vessel wall, and C and W in the divertor. These materials will interact on plasma-facing surfaces due to erosion and redeposition. As shown earlier [2], in the case of bombardment of W with C, erosion occurs at small ion fluences followed by net deposition as C covers the W surface. Recently it has been found in computer simulations, that oscillations in the sputtering yield as a function of fluence can occur [3]. Due to the strong implications of these kinds of effects in fusion plasma machines, it is the aim of this paper to show that these oscillation effects can also be observed experimentally. At temperatures of 100C and above the oscillations disappear. EXPERIMENTAL PROCEDURE The substrates used in this work were 12 x 15 x 0.5 mm rectangles of oriented high purity pyrolytic graphite supplied by Union Carbide Corp. that were cut with the graphite planes either parallel or perpendicular to the sample surface. The blanks were mechanically polished to produce a mirror finish on the active surface. Bombardment with 100 keV W ions was carried out on an Extrion model 200-1000 ion implantation accelerator at room temperature and elevated temperatures up to 700C. 1 MeV implantations were performed on a 1.7 MeV General Ionics Tandetron at room temperature and at liquid-nitrogen (LN2) temperature. Temperatures were monitored by chromel-alumel thermocouples mounted in the sample holders. The vacuum in the chamber was typically in the low 10 -7 mbar range during operation. For all implants the beam was rastered so that the beam was reduced from its unrastered value by approximately 30% in each direction to give reasonable uniformity over the implanted region. The samples were weighed before and after each implantation using a Mettler microbalance with a sensitivity of 1 microgram. The balance was zeroed before and after each measurement, but drift could not be totally eliminated and the error in the measured masses was estimated to be ±3 micrograms. The amount of W retained in the samples was measured quantitatively after each dose using Rutherford backscattering spectrometry (RBS) with 2.3 MeV He ions at a detector angle of 160. The sample was tilted at 60 to the incident beam in the Mat. Res. Soc. Symp. Proc. Vol. 658 © 2001 Materials Research Society