In Situ Thin Film Crystallization Studies Using High Temperature Grazing Incidence X-ray Diffraction (htgixrd)

We describe the experimental procedure and use of high-temperature X-ray diffraction techniques combined with grazing-incidence optics for in situ crystallization studies on several polycrystalline thin films; this technique is referred to as High-Temperature Grazing Incidence X-Ray Diffraction or (HTGIXRD). Temperature calibration techniques, sample positioning issues, and limitations of this analysis technique are discussed. Applications of this technique to crystallization of several ceramic thin films include qualitative investigations such as monitoring phase formation and structural transitions as well as quantitative investigations of crystallization kinetics and crystallite size measurements. Introduction Characterization of thin films is of great interest due to the increased use of film structures in modern electronic devices. X-ray diffraction techniques have been employed to study diffusion and stress-strain behavior in multilayered thin-film structures using Bragg-Brantano techniques. [ l] With the recent development of grazing-incidence X-ray diffraction, studies such as stress-depth profiling of polycrystalline thin-film structures have been performed.[2] The grazing or pseudo-parallel beam optics allows good resolution of diffraction peaks from a low divergent X-Ray beam at small (< 2”) grazing angles not obtainable by using conventional Bragg-Brentano optics.[3] This allows for better surface characterization of thin films. In this paper we propose extending the use of grazing-incidence X-ray diffraction to high-temperatures for in situ analysis of thin films. We shall refer to this technique as High-Temperature Grazing Incidence X-Ray DifIi-action or (HTGIXRD). The motivation for HTGIXRD analysis stemmed from a desire to obtain in situ hightemperature diffraction information from shallow surfaces. In situ X-Ray diffraction experiments are required to study the crystallization kinetics in thin films. This method can be used to determine the effects of precursor solution chemistry on thin-film kinetics as a function of temperature or to determine the temperature of a structural phase transition. The diffractometer, grazing incidence optics and high-temperature stage are available from most equipment manufacturers, making this technique relatively cost effective. Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997 Instrumentation A Scintag Xi diffractometer equipped with a line focus Cu target sealed X-Ray tube, 0.3” incident beam divergence slits, 1” incident Soller slits, 0.4” receiving Soller slits and a Peltier cooled solid state detector was used to collect diffraction data. (More intensity could be obtained with larger active area detectors, however, the sample size tends to limit intensity. If the sample is too large it may not have even temperature distribution.) The high-temperature stage was a Beuhler (Model HDK 1.4) with Pt strip and surround heaters. The maximum operating temperature in air was approximately 1600°C. A type S (Pt / Pt 10% Rh) O.Olmm diameter thermocouple was used to monitor the strip heater temperature. In order to avoid significant downtime due to potential failures at the welded region on the bottom surface of the heater, it was decided that a small hole be drilled in the center of the Pt strip heater. The thermocouple bead was then placed in the hole and peened into place at the top surface of the strip. This not only made it easier to change thermocouples, if required, but also allowed better control of temperature nearer to the film sample. Calibration would then be required for each thermocouple replaced. In the presence of silicon (or SiOz), platinum based thermocouples may suffer degradation due to reaction of the silicon with the platinum.[4] Therefore caution should be taken in analysis of silicon based thin film samples above 850°C. Experimental Procedure SrBizTa,Og “SBT” thin film samples were prepared on Si substrates with platinum electrodes according to a sol-gel spin technique.[5] The Pb(Zr0.3Ti0.7)03 “PZT 30/70” film was deposited upon a MgO substrate (platinum electrode) using an IMO process described elsewhere.[6] ZrO;! films were prepared on Silicon substrates via a sol-gel technique similar to the SBT films.[7] The thin film wafers were cut to dimensions (1Omm x 1Omm x OSmm) in order to maintain a constant thermal mass. Duplicates of each sample were prepared and initial trial runs were performed on the first few samples to obtain approximate transition temperatures. In this way it was possible to minimize the time for analysis and maximize the obtained information of subsequent runs. Film samples were placed at the center of the strip heater directly above the thermocouple. All diffraction patterns were collected in an air atmosphere. The Pt heater was first aligned by observing the intensity of the direct beam (40kV/3OmA) and then moving the heater stage into the beam until the intensity was reduced by half at the Peltier detector; aluminum strips were placed at the detector to reduce the intensity of the direct beam. Then the stage (omega axis) was rocked back and forth until an intensity maxim was observed. At this point the heater was parallel relative to direct beam. Changing the grazing incidence angle and choosing the divergence slits limits both the X-Ray beam area and height irradiating the sample; this affects both intensity and resolution. The width (or footprint) and height (thickness) of the beam can be easily calculated. See figure 1. Table I shows values for several Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997 divergent-slit / grazing-angle combinations. In order to maximize the intensity, a 2” grazing angle with 0.3 divergent slits was selected for our standard experimental procedures. During the final alignment, a Si substrate with Si 640b powder on its surface was placed on the heater and the goniometer was set to observe the Si (111) peak at -28.4’ 20. The heater stage was then adjusted to maximize this peak intensity. Figure 1. A schematic diagram of the pseudo-parallel optics geometry showing the height (II) and width (W) for the X-Ray beam. Table I: Beam Heights and Widths for combinations of grazing angles and divergence slits Grazing Angle (deg) Divergence Slit (deg) Beam Width (mm) Beam Height (mm) 2 0.3 30 0.50 2 0.1 10 0.17 1 0.3 60 0.50 1 0.1 20 0.17 Measuring Circle X-r Soiler slits (0.4”) w = Rsin(G)sin(D) sin(a)sin( f3) Temperature calibration was done using melting point standards. These materials were placed on a the surface of a platinized Silicon film and subsequently heated. The observed melting point temperatures were compared to the actual known melting points and a highly linear calibration curve was obtained. Additional calibrations were run biweekly during the course of experimentation with essentially no observable change in the temperature calibration. Figure 2 shows a calibration curve using selected melting point standards. The phase transition of KzSO4 Copyright 0 JCPDS-International Centre for Diffraction Data 1997 Copyright (C) JCPDS-International Centre for Diffraction Data 1997 from orthorhombic to hexagonal was observed at 583°C which was in excellent agreement with published results[S] confirming the accuracy of the calibration curve to an error of about 5°C. K,SO, Orthorhombic Hexagonal 0 200 400 600 60