Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://danp.sinp.msu.ru/LIIWM/Nimb356_mar07_373Mashkova.pdf
Äàòà èçìåíåíèÿ: Sun Oct 21 20:33:20 2007
Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 22:23:48 2012
Êîäèðîâêà:
N IM B
Beam Interactions with Materials & Atoms

Nuclear Instruments and Methods in Physics Research B 256 (2007) 363­367 www.elsevier.com/locate/nimb

Sputtering of HOPG under high-dose ion irradiation
A.M. Borisov a, E.S. Mashkova
a

a,*

, A.S. Nemov a, Yu.S. Virgiliev

b

Institute of Nuclear Physics, Moscow State University, 119992 Moscow, Russia b NIIgrafite, 111141 Moscow, Russia Available online 24 January 2007

Abstract The dependences of sputtering yield Y of highly oriented pyrolytic graphite under high fluences (1018­1019 ion/cm2) 30 keV N× irra2 diation at ion incidence angles from h = 0 (normal incidence) to h =80° at room temperature (RT) and T = 400 °C have been measured to trace the radiation damage influence on angular behavior of sputtering yield. A difference has been found between angular dependences of sputtering yields at RT, when the irradiation leads to a high degree of disorder, and at temperatures, larger than the temperature Ta responsible for annealing the radiation damage at continuous ion bombardment. ñ 2006 Elsevier B.V. All rights reserved.
PACS: 79.20.þm; 79.20.Ap Keywords: HOPG; High-dose ion irradiation; Sputtering; Developed surface topography

1. Introduction Carbon-based materials are widely used in nuclear reactors and thermonuclear devices, in elements of space technical equipment. They are often allocated in a separate class of materials in fundamental studies of mechanisms of radiation damage and their influences on the physical properties of materials [1,2]. HOPG is characterized by the highest degree of three-dimensional ordering [3]. The density, parameters of the crystal lattice, preferable orientation in a plane (0 0 0 1) and an anisotropy of the physical properties of the HOPG are close to those for natural graphite mineral. In particular, HOPG crystal structure is characterized by an arrangement of carbon atoms in parallel layers, and in each layer the atoms form a grid of correct hexagons with distances between atoms equal 0.1415 nm. The distance between layers is equals 0.3354 nm that gives a theoretical value of density q = 2.265 g/cm3. HOPG is often used in research of anisotropy effects in carbon-based materials under neutron irradiation in nuclear reactors, and high-temperature plasma in fusion reactors. Besides,
*

Corresponding author. Tel.: +7 495 9393 904; fax: +7 495 9390 896. E-mail address: esm@anna19.npi.msu.su (E.S. Mashkova).

the HOPG are used as monochromators of X-ray radiation in X-ray photoelectronic spectroscopy, in diffraction analyzers, in X-ray spectrometers and polarimeters, in researches of neutron radiation, etc. It is known, that many processes of interaction of ions with solids, including sputtering, ion-induced electron emission, and ion scattering, are sensitive to the degree of ordering in crystal lattice [4­7]. In particular, the studies of the angular dependences of ion-induced electron emission yield c for highly oriented pyrolytic graphite UPV1T, unlike for polycrystalline isotropic graphites, have shown distinctions in the form of the angular dependences c(h) at different irradiation temperatures [8]. The influence of orientation of anisotropic pyrolytic graphite on sputtering yield has been observed in [9]. Namely, it has been found, that the sputtering yield under 20 keV Ne+ ion irradiation at RT and normal incidence for the samples which have been cut parallel to the basic plane is larger, than for the ones which have been cut perpendicular to this plane. Earlier we measured the dependences of sputtering yield Y on ion incidence angle h at RT for isotropic polycrystalline graphites (MPG-LT, POCO-AXF-5Q) under 30 keV N× ion irradiation [10]. The aim of the present work is 2 an experimental study of the angular dependence of

0168-583X/$ - see front matter ñ 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.nimb.2006.12.029


364

A.M. Borisov et al. / Nucl. Instr. and Meth. in Phys. Res. B 256 (2007) 363­367

sputtering yield of highly-oriented pyrolytic graphite under 30 keV N× ion irradiation at different irradiation tempera2 tures, a comparison the obtained data with similar ones for polycrystalline graphites and the data obtained in computer simulation, and a study of crystalline structure and topography of surface layers. 2. Experimental The experiment was carried out on mass-monochromator of the Institute of Nuclear Physics, Moscow State University [5]. The samples were fixed in the collision chamber which scheme was plotted in [11]. The target holder allowed variation of ion incidence angles, h, from 0° up to 89° with step of 0.5°, and also to change target's temperature from RT up to 1000 °C. The samples of UPV-1T (NIIgrafite production, Moscow, Russia) had the form of rectangular sheets with thickness of 3 mm, width 22 mm and length 80 mm. Disorientation of basic plane (0 0 0 1) in samples did not exceed 50 0 . The irradiation was produced by 30 keV N× ions, ion current was 0.1­0.3 mA/ 2 cm2, the cross-section of ion beam was 0.35 cm2. The total ion fluences were 1018­1019 ion/cm2. Sputtering yield Y was determined by weight loss measurements assuming molecular ion dissociation by interaction with the solid. The error of measured yield was estimated to be 15% due to the necessity to take into account the gas adsorption. Preliminary processing of the samples included vacuum annealing at T = 350­400 °C. Thus a partial outgasing of the material has occurred. The restoration of weight of the samples after an annealing at an exposition to air lasted about 1 h. The measurement of the sputtering yields of carbon-based materials by weight loss also includes a procedure of stabilization of their weight on air after ion irradiation. Besides, the measurements for each fixed angle h were produced sometimes until the sputtering yield Y
4.0 3.5

(really, relative weight loss) ceased to depend on the number of the irradiations, i.e. corresponded to steady-state conditions, cf. [12]. The example of such data is presented in Fig. 1. The ion-induced electron emission yield, c, was determined as the ratio of the electron current to the primary ion current with the instrument error $2.5%. Crystalline structure of the surface layers was analyzed by the method of reflection high energy electron diffraction (RHEED) in EMR-102 (Russian model) operated at 50 kV and electron beam current 50 lA. The analysis of a surface topography was made by scanning electron microscopy (SEM) using LEO-1430vp. 3. Results and discussion The study of the UPV-1T samples before and after ion irradiation by means of electron diffraction has shown the distinction of surface layer structures at different irradiation temperatures and ion incidence angles, but more complex, than has been observed under similar conditions for polycrystalline graphites [10]. Namely, it has been established, that for non-irradiated samples a diffraction pattern contains point reflexes of type (0 0 2l) which correspond to a graphite monocrystal with an orientation of c-axis close to the normal to sample surface. The irradiation at the temperatures close to RT results, both at normal and at oblique ion incidence, to appearance of an amorphous halo in the diffraction patterns, testifying in favor of a strong disordering in the crystal structure. The diffraction patterns for the samples irradiated at elevated temperatures differ both from the initial picture, and from the picture in case of irradiation at RT and besides depend on ion incidence angle and the geometry of obtained diffraction pattern. At normal and near normal ion incidence and elevated temperatures three spread rings were observed, corresponding to the three most intensive rings for polycrystalline graphites. It should be noted that the surface topography under steady-state conditions is strongly developed and differs in case of an irradiation at RT and elevated temperatures, see Fig. 2 in [13]. In this connection, at research of diffraction at grazing ion incidence one needs to take into account, that in this case some needle-like cones develop at the surface (see Fig. 2) and both the cone mantle and the cone butt-ends are irradiated under essentially different local angles of ion incidence. A cone mantle is irradiated under the local angles close to the nominal ion incidence angles. Accordingly the needlecone butt-ends are irradiated under ion incidence angles close to normal. The electron diffraction of the butt-ends of the cones developed at elevated temperatures has been found similar to those in case of normal ion incidence. When probing electron beam was directed towards to the ion beam, i.e. when the structure of topographical elements flat mantle was analyzed, the ring sets of point reflexes from graphite prismatic planes were observed that testified in favor of change of the UPV-1T initial structure caused by twinning in graphite at ion bombardment, cf. [14].

30 keV N2 - UPV-1 T, = 50

+

o

Sputtering yield Y, atom/at.ion

3.0 2.5 2.0 1.5 1.0 0.5 0.0 -10 0 10 20 30 40 50 60 70 80 90 100

first irradiation second irradiation third irradiation

Time t, min
Fig. 1. Sputtering yield (weight of the sample) stabilization of UPV-1T in air after 30 keV N× ion bombardment for four consecutive irradiations at 2 RT.


A.M. Borisov et al. / Nucl. Instr. and Meth. in Phys. Res. B 256 (2007) 363­367

365

The experimentally obtained dependence of sputtering yield Y on ion incidence angle h on the (0 0 0 1) plane of UPV-1T at RT-irradiation is presented in Fig. 3. In the same figure the corresponding data obtained earlier for polycrystalline graphites MPG-LT (q = 1.7 g/cm3) and POCO-AXF-5Q (q = 1.82 g/cm3), and also data of computer simulation using the program TRIM.SP (the version trvmc 95) for the smooth target containing carbon, nitrogen and oxygen in the ratio C:N:O = 77:19:4, taken from the analysis of RBS-spectra of the irradiated polycrystalline graphite, are presented [10]. The comparison of the Y(h) for UPV-1T with corresponding angular dependences for polycrystalline graphites shows coincidence within the limits of the experimental error in the range of ion incidence angles h = 0­60°, that is caused, as the RHEED shows, by disordering of surface layers of irradiated targets. At h > 60° the sputtering yields for UPV-1T continue to remain in the good agreement with the corresponding yield for MPG-LT. As to the comparison with data for POCO one can see, that Y for POCO is considerably above the corresponding values for UPV-1T. These differences can be connected with the distinction in surface topography of materials, developed under high-dose ion bombardment. While the nominal ion incidence angle increases, in some angular range, the local incidence angles for topographical elements dominate which are larger, than nominal ones. As a result, the sputtering yields appear to be larger than for a smooth surface at the same ion incidence angles, compare the experimental data with the results of computer simulation for the smooth surface, presented in Fig. 3. As nominal ion incidence angle increases the situation arises when the local angles of incidence on cone mantle, smaller than the nominal ones, begin to dominate, that leads to reduction in the sputtering yields in comparison

Fig. 2. SEM micrographs of UPV-1T surfaces, irradiated under 30 keV N× ions: (a) h =0°, RT; (b) h = 0°, T = 400 °C; (c) h =60°, RT; 2 (d) h =60°, T = 400 °C. The tilt is 30°.

Nevertheless, the greatest restoration of the initial structure of the material at a continuous irradiation at elevated temperatures had been observed at grazing ion incidence (see corresponding diffraction patterns in [15]), that correlates, according to estimations of displacements per atom dpa, m(h), with decreasing m at more than an order of magnitude in the given range of ion incidence angles, cf. [8].

Fig. 3. Angular dependences of Y at RT irradiation. Data for MPG-LT, POCO and computer simulation are taken from [10,15,17].


366
2.5

A.M. Borisov et al. / Nucl. Instr. and Meth. in Phys. Res. B 256 (2007) 363­367

30 keV N2 - HOPG (UPV-1T)
2.0

+

Sputtering yield Y, atom/at.ion

1.5

RT

1.0

range of h = 10­40° the sputtering yields measured at RT and at 400 °C, are close. The local minimum at h = 0° may be caused by the transparency of UPV-1T for the projectile beam, as the direction is parallel to low index crystallographic directions [0 0 0 1] that is a typical display of anisotropy of sputtering yield for single crystals [4]. The minimum at h =60° also can be connected to a transparency of the target due to joint effect of twinning and formation of specific (needle-like cone) topography of the surface at high-dose ion bombardment. In reality, the projectile ion reflection from the steep slopes of the cones reduces the sputtering from the cone mantle, while recapture prevents the most ejected particles from leaving the target, and therefore reduces the sputtering yield, cf. [18]. 4. Conclusion The angular dependences of the sputtering yields Y(h) of highly oriented pyrolitic graphite UPV-1T has been measured under 30 keV N× ion bombardment at RT and 2 T = 400 °C. At RT a typical function for random solids: Y $ 1/cos h has been found in ion incidence angular range h = 0­80°. At T = 400 °C the curve Y(h) is non-monotonic and has two minima: at normal incidence and at h = 60°. The RHEED demonstrates that at RT irradiation there are amorphous halo in all studied angular range. At elevated temperatures (T > Ta) there are either typical polycrystalline rings or sets of point reflexes depending on surface topography analyzed by the electron beam. The non-monotonic feature of Y(h) at T > Ta may be due both to the UPV-1T lattice ordering and to the specific needle-like cone surface topography developed during the sputtering. A preliminary developing of surface topography may decrease the HOPG erosion under irradiation at some temperature regime in several times. Acknowledgements The work was sponsored by the Moscow Government. The authors are grateful to E.A. Pitirimova for performing RHEED, M.A. Timopheev for the SEM analysis of topography of the samples and E.S. Parilis for helpful discussions. References
[1] P. Ehrhart, W. Schilling, H. Ullmaier, Radiation damage in crystals, Encyclopedia of Applied Physics, Vol. 15, VCH Publishers, 1996, pp. 429. [2] T.D. Burshell, MRS Bull. 22 (4) (1997) 29. [3] A.S. Fialkov, Uglerod, mezhsloevye soedineniya i kompozity na ego osnove (Carbon and Carbon Based Intercalation Compounds and Composites), Aspect Press, Moscow, 1997 (in Russian). [4] R. Behrish (Ed.), Sputtering by Particle Bombardment I, Springer, Berlin, 1981.

0.5

400 C

o

0.0 0 20 40 60 80

Angle of incidence , deg.
Fig. 4. Angular dependences Y(h) for UPV-1T for RT and T = 400 °C.

with the results of computer simulation for smooth surfaces, cf. [16]. An attempt undertaken in the [17] to correct the simulation data for a smooth surface taking into account the distribution of local ion incidence angles determined using laser goniophotometry for the samples of graphite MPG-LT, shows (see dotted curve in Fig. 3) a reasonable agreement with the experimental data. At research of ion-induced electron emission from highly-oriented pyrographite UPV-1T it has been revealed, unlike for polycrystalline graphites, a distinction in the form of angular dependences of ion-induced electron emission yield c both at RT and at temperatures above a temperature Ta (for UPV-1T Ta > 125 °C, see Fig. 1 in [8]) at which for carbon-based materials a jump in the electron escape is observed, caused by the ion-induced change in the material structure order degree, due to competing processes of accumulation and annealing of radiation damage at continuous ion bombardment. It has been proved, that such behavior of c can be connected both to the change in the secondary electron path length and to the change in the transparency of graphite crystal lattice for the projectile particles in the process of annealing the radiation damage with the increase of irradiation temperature. Sputtering of highly oriented pyrolytic graphite also shows, that the character of behavior of the angular dependences of sputtering yield Y at RT and elevated temperatures is essentially different, see Fig. 4. One can see, that the curve Y(h) at T = 400 °C is non-monotonic and shows two minima: at normal ion incidence (h = 0°) and at h = 60°. At h = 0° when the direction of ion beam coincides with the direction along the axis c, the sputtering yield decreases by 1.3 times as compared to the case of RT irradiation. At h = 60° Y decreases more than by three times. In the


A.M. Borisov et al. / Nucl. Instr. and Meth. in Phys. Res. B 256 (2007) 363­367 [5] E.S. Mashkova, V.A. Molchanov, Medium-Energy Ion Reflection from Solids, North-Holland, Amsterdam, 1985. [6] B.A. Brusilovsky, Appl. Phys. A 50 (1990) 111. [7] E.S. Parilis, L.M. Kishinevsky, N.Yu. Turaev, B.E. Baklitzky, F.F. Umarov, V.Kh. Verleger, S.L. Nizhnaya, I.S. Bitensky, Atomic Collisions on Solid Surfaces, North-Holland, Amsterdam, 1993 (Chapter 11). [8] A.M. Borisov, E.S. Mashkova, A.S. Nemov, E.S. Parilis, Vacuum 80 (2005) 295. [9] E. Hechtl, J. Bohdansky, J. Nucl. Mater. 122­123 (1984) 1431. [10] A.M. Borisov, W. Eckstein, E.S. Mashkova, J. Nucl. Mater. 304 (2002) 15. [11] E.S. Mashkova, V.A. Molchanov, I.M. Fayazov, W. Eckstein, Poverkhnost 2 (1994) 33 (in Russian).

367

[12] J.M. Barnett, M.J. Pellin, W.F. Calaway, D.M. Gruen, Phys. Rev. Lett. 69 (1989) 562. [13] Yu.S. Virgiliev, A.M. Borisov, E.S. Mashkova, A.S. Nemov, E.A. Pitirimova, A.F. Khokhlov, Poverkhnost 4 (2004) 13 (in Russian). [14] D.J. Bacon, A.S. Rao, J. Nucl. Mater. 91 (1980) 178. [15] A.M. Borisov, Yu.S. Virgilev, E.S. Mashkova, A.S. Nemov, E.A. Pitirimova, Poverkhnost 1 (2006) 7 (in Russian). [16] M. Kuestner, W. Eckstein, V. Dose, J. Roth, Nucl. Instr. and Meth. 145 (1998) 320. [17] A.M. Borisov, E.S. Mashkova, A.S. Nemov, S.A. Kamneva, V.A. Kurnaev, N.N. Trifonov, VANT 2 (2004) 65 (in Russian). [18] W.O. Hofer, in: R. Behrisch, K. Witmaack (Eds.), Sputtering by Particle Bombardment III, Springer, Berlin, 1991 (Chapter 2).