Copyright(c)JCPDS-International Centre for Diffraction Data 2000,Advances in X-ray Analysis,Vol.43 212 APPLICATION OF Ni/C-GÖBEL MIRRORS AS PARALLEL BEAM X-RAY OPTICS FOR AND RADIATION T. Holz, R. Dietsch, H. Mai, L. Brügemann * Fraunhofer Institute Material and Beam Technology, Winterbergstraße 28, D-01277 Dresden, Germany * BRUKER Analytical X-Ray Systems, D-76181 Karlsruhe, Germany ABSTRACT Pulsed Laser Deposition (PLD) has been successfully used for the preparation of multilayers having X-ray optical quality. Outstanding features of the PLD-process are high thickness uniformity, precision of deposition process, formation of smooth interfaces and suppression of columnar growth of thin films. High quality layer stacks with laterally graded thickness distributions across 4 -wafers were used to produce s for Mo Kα and Cu Kα radiation. Reflectivities up to 90% depending on period thickness and an excellent energy resolution are obtained. The observed of 1 st BRAGG-orders of typically (2Θ) 0.06 (Cu Kα) and (2Θ) 0.04 (Mo Kα) requires a very precise shaping of the parabolical mirror to achieve a homogeneous intensity profile of the parallel beam. The performance of such PLD-manufactured s in a Twin (TGM) arrangement (Cu Kα and Mo Kα) is demonstrated by selected measurements characterizing the excellent beam quality. This TGM arrangement consists of two s placed in the primary and in the diffracted beam path, resp. For Cu Kα radiation the high monochromized intensity of 10 9 cps is combined with low beam divergence ( φ < 0.02 ) and superior suppression of Cu Kβ radiation ( I(Cu Kα 1 ) : I(Cu Kβ) = 10 6 ) to realize a versatile experimental setup for X- ray reflectometry. The determination of layer thicknesses can be carried out regardless of sample displacement. INTRODUCTION Göbel-Mirrors have induced a decisive step in the improvement of X-ray diffractometry investigations [1]. The parallel beam geometry permits an accurate lattice parameter determination regardless of sample displacement or surface roughness, a measurement without preceding sample preparation and a reduced time of measurements. The aim of investigations presented in this paper is the extension of the improvements in diffractometry to the field of X-ray reflectometry, too.
Copyright(c)JCPDS-International Centre for Diffraction Data 2000,Advances in X-ray Analysis,Vol.43 213 Fig. 1: Setup of Twin s Outstanding properties of Göbel Mirrors of the Ni/C-type are the low divergence, the high reflectivity and the suppression of unwanted incident radiation [2,3]. Referring to maximum intensity the best suited optical component for a primary is a second one on the diffracted beam side. Because of the face to face arrangement of both mirrors, it should be called Twin s (Fig. 1). This setup realizes the parallel beam concept also on the diffracted beam side [4]. EXPERIMENTAL Experiments were carried out with a D5005 diffractometer ( BRUKER AXS, Karlsruhe) equipped with a 6 -sample stage and a knife edge collimator (KEC). The vertical position of sample stage and KEC are adjustable within µm-range precision. Details of the primary and diffracted beam side are shown in Fig 2a/b. The type of X-ray tubes, the used generator settings and the slit sizes are listed in Tab. 1. Copper foils were used to attenuate the high intensities, which were detected by a scintillation counter. Mo Cu X-ray tube focus size generator settings (U/I) K FF Mo 2 KE 8mm x 0.04mm 50kV/40mA K FL Cu 2 KE 12mm x 0.04mm 40kV/40mA incident beam side (see Fig 2a) 0.2mm 0.6mm/4 Soller (horz.) 1mm 1mm/4 Soller (horz.) diffracted beam side (see Fig. 2b) 0.6mm 0.05mm 1mm 0.05mm Tab. 1: Parameters of experimental setup
Copyright(c)JCPDS-International Centre for Diffraction Data 2000,Advances in X-ray Analysis,Vol.43 214 Fig. 2a: Photograph of primary beam side Fig. 2b: Details of diffracted beam side Parallel beam widths of (0.8±0.2)mm for Cu Kα and (0.4±0.1)mm for MoKα are realized for the used Ni/C s TGM - CHARACTERISTICS OF THE BEAM To demonstrate the characteristic beam properties the following measurements were executed: detector scan through the incident beam (Fig. 3a/b) rocking scan measurement of the silicon 004 reflection (Fig. 4a/b) Θ-2Θ scan measurement of the silicon 004 reflection (Fig. 5a/b) These measurements are suited to characterize intensity, divergence and suppression of unwanted incident radiation, such as the Kβ. 11 8 cps] 2 1 2Θ=0.021 8 cps] 10 9 8 7 6 5 4 3 2 2Θ=0.027 1 0-0.04-0.02 0.00 0.02 0.04 Fig. 3a: detector scan through incident beam (Mo-TGM) 0-0.04-0.02 0.00 0.02 0.04 Fig 3b: detector scan through incident beam (Cu-TGM)
Copyright(c)JCPDS-International Centre for Diffraction Data 2000,Advances in X-ray Analysis,Vol.43 215 20 4 cps] 6 3 2 1 ω=0.012 cps] 6 15 10 5 ω=0.014 15.1 15.12 15.14 15.16 15.18 15.2 w [degree] Fig. 4a: Rocking scan of the silicon 004 reflection (2Θ=30.28 ) 34.5 34.52 34.54 34.56 34.58 34.6 w [degree] Fig. 4b: Rocking scan of the silicon 004 reflection (2Θ=69.122 ) 10 6 10 5 10 4 10 3 Mo Kb Ka 1,2 10 7 10 6 10 5 10 4 10 3 Kb Cu W La Ka 1,2 10 2 10 2 26 27 28 29 30 31 Fig.5a: Θ-2Θ scan measurement of the silicon 004 reflection 62 64 66 68 70 72 Fig. 5b: Θ-2Θ scan measurement of the silicon 004 reflection The detector scan in Fig. 3a demonstrates the combination of highest intensities of more than 200,000,000 cps with low divergence for high angular resolution. The small of the intensity distribution suggests nearly ideally parabolically shaped Ni/C s. In fact the divergence is very low as demonstrated in Fig. 4a. The of the rocking curve of the silicon 004 reflection is 0.012. This is near to the theoretical limit, determined by the focus and detector slit sizes. As shown in Fig 5a Mo Kβ radiation is almost completely suppressed. The intensity ratio of I(1) / I(Mo Kb) = 90,000 was derived from the Θ-2Θ scan measurement of the silicon 004 reflection. Because of the different types of X-ray tubes (line focus sizes, generator settings) and parallel beam widths the recorded intensities of Cu- and Mo-TGM are not directly comparable.
Copyright(c)JCPDS-International Centre for Diffraction Data 2000,Advances in X-ray Analysis,Vol.43 216 Highest intensities of more than 1,000,000,000 cps are combined again with low divergence (Fig. 3b). The of the rocking curve of the silicon 004 reflection is equal to 0.014 (Fig. 4b). The resulting excellent angular resolution is presented in Fig. 6. Note, despite of the low acceptance of Si 004 reflection the peak intensity is equal to 2x10 7 cps. The main difference between Cu- and Mo- radiation results from the particular X-ray optical properties of nickel, i.e. the K absorption edge between Cu Kα and Cu Kβ, which influences the Kβ suppression. The intensity ratio reaches therefore I(CuKa 1 ) / I(CuKb) = 1,000,000 as derived from Fig. 5b. REFLECTOMETRY A multilayer with a total layer stack thickness of 240nm was used to test succesfully the angular resolution of the TGM. The Ni/C multilayer consisted of N=75 periods of period thickness d=3.20nm with a carbon top layer. As shown in Fig. 6 Kiessig fringes are clearly resolved when using an angular step width of (2Θ)=0.002. reflectometry () 0.5 1 1.5 2 2.5 3 3.5 Fig. 6: Resolution of Kiessig fringes in X-ray reflectometry of a multilayer having a total layer stack thickness of N x d = 240nm [sample: Ni/Cmultilayer, period number N=75, period thickness d=3.20nm] The measurements in Fig 7 were executed to confirm the independence of peak intensities and angular peak positions from the sample surface adjustment with reference to the parallel X-ray beam. Three different sample stage positions (-200µm, 0, +200µm) were selected. The KEC was adjusted to a constant distance of about 50µm above the multilayer surface. Intensities and peak positions show no shift up to the 5 th order BRAGG-peak. Therefore a significant easier sample alignment procedure can be used in TGM arrangements. Merely a single rocking scan allows to set the alignment position in X-ray reflectometry. Sample alignment is no longer a time-consuming procedure and a source of errors.
Copyright(c)JCPDS-International Centre for Diffraction Data 2000,Advances in X-ray Analysis,Vol.43 217 10 7 10 6 10 5 10 4 10 3 10 2 10 1 reflectometry () graded Ni/C-multilayer sample stage position H=+200µm H=0 H=-200µm Fig. 7: Independence of peak intensity and angular position of specimen position demonstrated by measurements at three different sample stage positions 2.5 5 7.5 10 12.5 15 17.5 CONCLUSIONS The results show the outstanding beam properties of this special arrangement, which combines highest intensities, brilliant suppression of unwanted incident radiation with very low beam divergence. Reflectometry with TGM will be a powerful tool in thin film analysis in the future. ACKNOWLEDGEMENTS Financial support of this work has been provided by the OEF program of the Fraunhofer Gesellschaft, Germany. REFERENCES [1] M. Schuster, H. Göbel, J. Phys. D 28 (1995) A270 [2] T. Holz, R. Dietsch, H. Mai, L. Brügemann, S. Hopfe, R. Scholz, R. Krawietz, B. Wehner, Advances in X-Ray Analysis vol. 41 (1999) 346-355 [3] T. Holz, R. Dietsch, H. Mai, L. Brügemann, proc. EPDIC-6, 22.-25.-08.98, Hungary, to be published in Mat. Sci. Forum [4] R. Stömmer, R. Höpler, M. Schuster, H. Göbel, Advances in X-Ray Analysis, vol. 41 (1999) 336-345