·Home ·Table of Contents ·Methods and Instrumentation

Practical Application of X-Ray Diffraction Imaging Kirsten G. Lipetzky and Robert E. Green, Jr. The Johns Hopkins University, Center for Nondestructive Evaluation 3400 N. Charles Street, 102 Maryland Hall Baltimore, Maryland 21218 USA Contact

Single crystal materials are finding more use in industrial applications everyday. From the microelectronics industry which uses silicon and gallium arsenide single crystal substrates in integrated circuit production, to the aerospace and power generation industries which are producing single crystal, nickel based alloy, turbine blades, there is a need to fully characterize these materials. Because the quality assurance requirements for one industry are not necessarily those of another, a characterization technique which can be tailored to the needs of a particular industry is desirable. One technique which has proven to be useful in a variety of single crystal inspection applications is x-ray diffraction topography[1-11].This paper describes the details of an x-ray diffraction topographic system which was used and shows topographic images illustrating the utility of the system as applied to nickel based alloy, turbine blades, quartz resonators, and silicon carbides substrates.
Key Words: Single crystal; Topography; X-ray diffraction; Turbine blade; Quartz; Silicon carbide

INTRODUCTION

In the nondestructive testing field when the word x-ray is mentioned, the concept of radiography immediately comes to mind. While radiography provides valuable information as to density variations within a given material, it does not provide information as to the crystalline nature of the material. That is, is the material single crystalline or polycrystalline, deformed or "strain free", populated with imperfections or low in defects. In some instances, the nature and degree of crystalline perfection is of little significance to the performance of the material, however, in many industrial applications the crystalline perfection is critical.

Conventional methods for examination of single crystal specimens rely on the Laue technique of x-ray diffraction whereby patterns of individual spots (reflections) are obtained from individual planes. In the Laue method, a continuous (white radiation) x-ray source is utilized. The x-ray source is collimated (usually with a pin-hole collimator) and then incident upon the sample when the transmission configuration is used, or through a hole in a film followed by the sample when the back-reflection configuration is used. Transmission Laue diffraction analysis is limited to thin specimens (in general, mt < 1 ; where : m is the mass absorption coefficient of the specimen and t is the specimen thickness).

The Laue technique suffers from several shortcomings when considering the technique for use in quality control of materials. The first shortcoming is that only a small portion of a specimen can be examined at one time due to the small size of the collimated, incident x-ray beam. Moreover, a large portion of the central region of the diffraction pattern is often obscured by a high level of incoherent scattering (Compton scattering), fluorescent radiation and temperature-diffuse scattering, particularly in back-reflection work. Although Laue x-ray diffraction will allow the orientation of a single crystal specimen to be determined, it does not give the complete picture as to the quality and degree of crystallinity of the entire specimen. By scanning a specimen in the x-ray beam more meaningful results could be obtained, but inspection of the entire specimen using the point probe method would require an extremely long time.

One technique which has proven to useful in a variety of single crystal inspection applications is x-ray diffraction topography. X-ray diffraction topography is the name given to several x-ray diffraction techniques which permit the imaging of strains and lattice isorientations associated with surface and internal defects as small as dislocations to be examined. The term topography literally means "to describe a place" (topos = place, graphein = to write) and the diffraction information which is obtained by x-ray topographic methods may be bulk (transmission) or surface (back-reflection) in nature. Because the topographic techniques are based on Bragg diffraction from a periodic crystal, the images are extremely sensitive to crystal imperfections, strains, and rotations since any alteration to the interplanar position and spacing of the crystal will effect a corresponding change in the Bragg condition.

EXPERIMENTAL PROCEDURE

In the topographic set-up used in this investigation, called asymmetric crystal topography (ACT)[12], a slit collimated white radiation x-ray source was incident upon a high quality asymmetrically cut silicon crystal. The silicon crystal served as both a monochromator of K_a and possibly K_b wavelengths, and a beam expander,resulting in an x-ray beam of approximate dimensions 6 cm high by 2 cm wide. A crystal specimen of interest was mounted on a Newport rotation stage and placed in the path of the monochromated and expanded x-ray beam. The sample was rotated in the monochromated and expanded x-ray beam until a Bragg condition was

Fig 1: Schematic of the asymmetric crystal topography (ACT) system.

satisfied for some plane reflection as detected using an image intensifier with a fluorescent screen faceplate placed near the specimen.The beam diffracted from the specimen crystal and captured by the image intensifier was ensured to be monochromatic when the stringent simultaneous diffraction conditions of the first silicon crystal and the sample crystal was satisfied. Diffraction images of interest were permanently recorded on x-ray film (Kodak occlusal DF-50 dental film) with exposure times ranging from 1/2 hour up to 6 hours depending upon the sample investigated and the intensity of the diffracted reflection observed. Due to the relatively weak x-ray source available in the laboratory (copper target tube operated at 50 kV and 32 mA), as well as the thickness of the majority of the samples examined, the back-reflection configuration was utilized in this research.

In the ACT technique each individual topographic image is essentially a large Laue "spot" generated by diffraction from a particular set of "parallel" lattice planes covering a large area of the crystal. The x-ray beam incident on the specimen illuminated a large area unlike conventional Laue pin-hole techniques, and because of the special beam expanding monochromizing silicon crystal, this large incident beam experienced minimal divergence. A schematic of the ACT system showing positions of the highly perfect, asymmetrically cut silicon first crystal (monochromator and beam expander), the second crystal (specimen under investigation), and the x-ray image intensifier (for direct real-time viewing of topographic images) is shown in Figure 1.

EXPERIMENTAL RESULTS

Turbine Blades
The efficiency of modern gas turbine engines, used for both aircraft and ground based electric power generation, increases with increasing combustion temperature. The physical requirements which limit the choice of turbine blade materials for high temperature operation are low density, thermal stability, thermal fatigue resistance, toughness, resistance to high-temperature oxidation, and resistance to creep. Creep caused by dislocation motion is resisted by addition of alloying elements in solid solution and formation of stable hard precipitates, both of which serve as dislocation pinning points. Diffusional creep is resisted by increasing the grain size, directional solidification to produce long grains with boundaries parallel to the applied stress, and most optimally by eliminating the grain boundaries completely, i.e. using single crystal blades. The use of metallic single crystals for structural engineering applications places new requirements on nondestructive evaluation techniques for quality control.

One of the greatest obstacles in the quality assurance process for single crystal nickel based alloy turbine blades is the determination of the overall crystalline perfection of the final blades. The existing method relies upon chemical etching and visual inspection. While the problems associated with visual inspection are self-evident, there are intricacies associated with the etch process. The present research focuses on nondestructive inspection techniques, based on x-ray diffraction, which would eliminate the need for chemical etching and visual examination.

Fig 2: Schematic diagram of the ACT experimental arrangement used for inspection of a "single" crystal, nickel based alloy turbine blade.

Figure 2 shows a schematic of the ACT experimental arrangement used for inspection of a "single" crystal, nickel based alloy turbine blade. The quotation marks appear in the previous statement because the turbine blade actually contains two stray (secondary) crystals on the surface, where the orientation of the major portion of the blade is considered to be the primary crystal. Figure 3 shows a schematic drawing of the turbine blade which was examined and corresponding ACT images obtained from different regions along the airfoil of the turbine blade. As can be seen, the topographic images do not show all of the portion of the blade on which the incident x-ray beam impinged. This is because the entire blade is not a single crystal. In fact by changing the Bragg angle, one can image each of the stray grains independently. In the two topographic images shown in Figure 3, there is an approximate misorientation of 15^°.

Fig 3: (a) Schematic of a turbine blade sample which contains two secondary crystals on the surface of the blade and resulting x-ray topographic images obtained from (b) a triangular shaped "stray" crystal and from (c) the "stray" secondary crystal next to the triangular shaped crystal.

Quartz Resonators
While quartz resonators have been the mainstay of the ultrasonics industry for some time, intracacies exist in the production of quality resonators and therefore fabrication remains somewhat of an art form. Recently the Johns Hopkins University Center for Nondestructive Evaluation has initiated a research program to investigate methods to analyze quartz single crystals for the purpose of assessing the relative crystal quality of the raw material as well as material which has been manufactured into resonators.

One application for which the ACT system was applied to quartz crystals was to determine if the x-ray topographic system could distinguish between a "good" resonator versus a "bad" resonator and futhermore, the possible cause(s) of failure of the "bad" resonator. Two quartz resonators each of 1 1/2 cm diameter and approximately 1 mm thickness were examined in the ACT system. A feature which was present on each sample, a major flat (for crystallographic purposes), was used as a fiduciary marker to determine whether a sample was in the up or down position. Since there was no way to distinguish the front from the back of a given sample, an arbitrary designation was made.

Each sample was examined individually using the back-reflection configuration of the ACT system. In order that direct comparisons could be made between the front and back of a given sample and between the two different samples, it was necessary to ensure that the same Bragg reflection was selected. This was done by recording a selected diffraction pattern from one sample and then, without changing the Bragg angle, removing the first sample and either flipping the sample over to examine the other side or replacing it with the second, making sure that the position of the fiduciary flat coincided. Minor adjustments in the Bragg angle (< 1^°) needed to be made occasionally in order to optimize the diffraction conditions and for a given Bragg condition a total of four topographs were obtained ("good" sample front, "good" sample back, "bad" sample front, and "bad" sample back). Figure 4 shows typical topographic images acquired from the two samples.

As can be seen in Figure 4, distinct differences can be discerned between the topographs of the "good" sample versus those acquired from the "bad" sample; the most notable being the jagged region (of darker contrast) which appears around the circumference of the "bad" quartz resonator. From experience, this region is known to be indicative of twinning which has occurred within the sample and is the most probable reason for the piezoelectric failure mechanism observed in the "bad" sample.

Fig 4: Back reflection, x-ray diffraction topographs obtained from a (a) "good" quartz resonator (front view), (b) "bad" quartz resonator (front view), (c) "good" quartz resonator (back view), and (d) "bad" quartz resonator (back view).

Silicon Carbide Substrates
Silicon carbide has been attracting great interest in the semiconductor industry, particularly because of its potential application in electronic devices capable of operation in adverse environments specifically, high temperature and radiation.[13-14] Recently, researchers at the Center for Nondestructive Evaluation (CNDE) were asked to use the ACT set-up to evaluate the crystalline perfection of several different silicon carbide substrates. The intention was to use the substrates for subsequent epitaxial thin film deposition and the company involved wanted an assessment of the substrates as a predictive measure for the quality of film which might be able to be grown.

Three different samples were provided to CNDE for evaluation. In the "as received" condition, each sample had been "quartered" [see Figure 5 (a)] and was mounted on a static free film which was stretched over an approximately 10 cm diameter ring. In order that the samples remain in the "as received" condition and not have to be removed from the static free film, a mount was made to hold the 10 cm diameter ring onto the Newport rotation stages of the ACT system. Prior to investigation of the samples via ACT, the samples were examined using Laue x-ray diffraction in the back-reflection configuration to determine their crystallographic orientation.Two substrates, of 3 cm diameter each, had a crystallographic orientation of approximately 7^° off the <111>. The third substrate was 4 cm in diameter and had a crystallographic orientation of approximately 16^° off the <111>.The samples were then analyzed in the ACT system. A typical topograph of one of the smaller diameter silicon carbide substrates is shown in Figure 5 (b). As can be seen there are numerous defects and strains present within the substrate, as indicated by the extreme variations in the diffracted intensity of the x-ray beam. A close up of one area of the topograph is shown in Figure 5 (c), further emphasizing the non-uniformity of the diffraction image. A good single crystal would have diffraction contrast which would be a uniform shade of grey or black. [See for example Figure 4 (a) and (c).] While the resulting images were not those which the company hoped to see, the diffraction images did help to resolve issues related to the film deposition process, specifically, alternate substrates were selected.

Fig 5: (a) Schematic drawing of a silicon carbide wafer which was "quartered". (b) Typical topograph acquired from a silicon carbide wafer and ( c) a close up of a region of the same topograph showing the tremendous diffraction contrast which was present.

CONCLUSIONS

X-ray diffraction topography can easily detect and image the presence of defects within a crystal, making it a powerful nondestructive evaluation tool for characterizing industrially important single crystal specimens such as nickel based alloy turbine blades, quartz resonators and silicon carbide substrates. Whether looking for gross defects (crystallographic misorientations) or smaller defects such as dislocations or even atomic (point defects) scale influences, the x-ray diffraction topographic techniques are well suited to a variety of applications. In the examples shown, x-ray topography was utilized to determine whether a turbine blade grown to be a single crystal specimen was in fact a single crystal, to distinguish between "good" and "bad" quartz resonators, and to evaluate the crystalline perfection of silicon carbide substrates for subsequent thin film deposition. Use of x-ray diffractiones did help to resolve issues related to the film deposition process, specifically, alternate substrates were selected.ism observed in the bad sample.topography in industry, would aid in the optimization of crystal growth and processing of these and other industrially important single crystal applications.

REFERENCES

1. Kirsten G. Lipetzky and Robert E. Green, Jr., "Characterization of Single Crystal Mechanical Test Specimens for Advanced Turbine Applications," Nondestructive Characterization of Materials, IX, 661 (1999).
2. Kirsten G. Lipetzky, Robert E. Green, Jr. and Paul J. Zombo, "Reliability Enhancement of Single Crystal Blades Used in Power Generation Gas Turbines," Failure Analysis: A Foundation for Diagnositcs and Prognostics Development, Virginia Beach, Virginia, pp. 61-70 (April 19-22, 1999).
3. Robert E. Green, Jr., John M. Winter, Jr. and Kirsten G. Lipetzky, "Materials Characterization by X-ray Topographic and Tomographic Imaging," Nondestructive Evaluation (NDE) and Material Properties, IV, 21(1999).
4. K. G. Lipetzky, R. E. Green, Jr., R. W. Armstrong and W. T. Beard, Jr., "The Evaluation of Quartz Resonators via X-ray Diffraction Topography," Review of Progress in Quantitative Nondestructive Evaluation, 18, 2079,1999.
5. W.T. Beard, K.G. Lipetzky and R.W. Armstrong, 'Dynamical Effects in High Resolution Topographic Imaging of Electronic Devices, Advances in X-ray Analysis, 41, 203 (1999).
6. Kirsten G. Lipetzky, Robert E. Green, Jr., and Paul J. Zombo, "Development of X-ray Diffraction Methods to Examine Single Crystal Turbine Blades,"Nondestructive Characterization of Materials, VIII, 423 (1998).
7. R.W. Armstrong, W.T. Beard, K.A. Green and X.J. Zhang, "High-Resolution Imaging of Electronic Devices Using Line Modified-Asymmetric Crystal Topography (LM -ACT)," Il Nuovo Cimento 19 D (2-4), 147 (1997).
8. W.T. Beard, Jr., K.A. Green, X.J. Zhang and R.W. Armstrong, "High Resolution Imaging of Electronic Devices via X-ray Diffraction Topography," Applied Physics Letters, 69(4), 488 (1996).
9. K.A. Green, W.T. Beard, X.J. Zhang, and R.W. Armstrong, "Application of Line Modified-Asymmetric Crystal Topography for Qualitative and Quantitative Evaluation of Integrated Circuits," Advances in X-ray Analysis, 38,227 (1995).
10. W.T. Beard, Jr., K.A. Green, X.J. Zhang and R.W. Armstrong, "In-Depth Resolution of Integrated Circuits via X-ray Based Line Modified Asymmetric Crystal Topography," 1994 IEEE International Reliability Physics Proceedings, IEEE Catalog No. 94CH3332-4, 425.
11. Kirsten A. Green and Robert E. Green, Jr., "Application of X-ray Topography for Nondestructive Inspection of Industrial Materials," Review of Progress in Quantitative Nondestructive Evaluation, 13, 571 (1994).
12. W.J. Boettinger, H.E. Burdette, M. Kuriyama and R.E. Green, Jr., "Asymmetric Crystal Topographic Camera,"Rev. Sci. Instrum.,47, 906 (1976).
13. C. Hallin, I.G. Ivanov, T. Egilsson, A. Henry, O. Kordina and E. JanzJn, "The Material Quality of CVD-Grown SiC Using Different Carbon Precursors," J. Crystal Growth, 183, 163 (1998).
14. I.H. Khan, "The Growth and Structure of Single-Crystal Films," in Handbook of Thin Film Technology, Leon I.Maissel and Reinhard Glang, Editors, New York: McGraw-Hill Book Company, 1970.

© AIPnD , created by NDT.net

|Home|    |Top|