Title :CMCBEDOFF Keywords :CBED, Thickness, TEM Computer Control Computer :IBM PC Operating System :DOS Programming Language :Pascal Hardware Requirements :None Author(s) :M.T. Otten Correspondence Address :Philips Analytical :International Business Centre :Electron Optics,Building AAE :5600 MD Eindhoven,The Netherlands ------------------------------------------------------------------------------------ Operating Instructions for On-Line and Off-Line CBED Thickness Measurements on the Philips CM Microscope M.T. Otten Philips Analytical International Business Centre -- Electron Optics Building AAE 5600 MD Eindhoven THE NETHERLANDS JULY 1989 Operating Instructions for On-Line CBED Thickness Measurement INTRODUCTION This manual describes the operation of program CBED for on-line and program CBEDOFF for off-line measurement of thickness through Convergent Beam Diffraction. In principle, the on-line measurement can be done by using the internal CM measurement in diffraction and converting the d spacings given by the microscope to real distances, but it is very difficult to shift perpendicularly from one fringe to the next, resulting in substantial measurement errors. Program CBED still uses the internal measurement of the CM microscope by Remote Control, but by using the cursor keys of the computer the shift is made automatically perpendicular to the fringes. The necessary data are collected and the thickness determined directly by the program. The instructions necessary for execution are shown on the screen of the Remote Control computer. NOTE: These programs are unsupported software. They may be distributed and/or modified freely. They may not be copyrighted by any individual other than the author or Philips Electron Optics. Philips Electron Optics is under no circumstances responsible for any errors in the software or any damage resulting from its use. Programs CBED and CBEDOFF run on IBM-compatible computers (IBM is a trademark of the International Business Machines Corporation (that obey the conditions specified for Remote Control of the CM microscope. Program CBED has been tested on an XT-compatible. It may not work on an AT-compatible because of the higher processing speed which can produce some problems in com- munication with the microscope. Print-out is formatted for an EPSON dot-matrix printer. The following files are supplied with the program: CBED.INS The text of these operating instructions (WordPerfect format) CBED.COM The program itself CBED.PAS Its source code RCM.INC Include files containing: Remote Control subroutines REMCBED.INC Remote Control subroutines CBEDOFF.COM A similar program for off-line measurements CBEDOFF.PAS Its source code LINEAR.PAS Linear Regression Include file INSTALLATION Installation of the program on a floppy disc or the hard disc is done by using the MS-DOS copy command. For two floppy-disc drive systems: Insert the program disc in drive b and the MS-DOS bootable disc in a: Type COPY B:CBED.COM A: and press the carriage return. Type COPY B:CBEDOFF.COM A: and press the carriage return. Remove the program disc and insert a data disc in drive b: NOTE: The active disc during running of the program may not be write- protected (silver sticker on the side) because the program must be able to write a log file on the disc. For one floppy-disc plus hard disc drive systems: Insert the program disc in drive a: Type COPY A:CBED.COM C: and press the carriage return. Type COPY A:CBEDOFF.COM C: and press the carriage return. Remove the program disc THEORETICAL BACKGROUND High accurate thickness measurements can be done by determining the spacings of the subsidiary fringes visible parallel to the Kikuchi line in Dark-Field discs. The method is outlined in: Kelly, P.M., Jostsons, A., Blake, R.G., and Napier, J.G. (1975), Phys. Stat. Sol. A31, 771. The data required for the measurement are: (1) the high tension (2) the d spacing of the Dark-Field disc used (3) a measurement of the distance between the Bright-Field and Dark- Field discs (4) the distances between the dark subsidiary fringes and the central Kikuchi line For measuring thickness through convergent beam diffraction the formula is used, where Si is the deviation parameter, l the electron wavelength, dhkl the d-spacing of the diffraction used, ni are the distances from the central bright fringe to the dark fringes 1, 2, etc. and 2 qB the distance in the pattern representing the Bragg angle. When (Si/ni)2 is plotted against (1/ni)2, where ni is the number of the dark fringe (1,2,3, etc.), the plot yields a straight line, if ni was chosen correctly. If ni was incorrect (at larger thickness fringes, 1,2,3 gradually disappear so the first dark fringe becomes 2,3,4 etc.), then the plot yields a curved line. The thickness of the specimen is derived from the intercept, which is equal to (1/t)2. The slope of the line is equal to -(1/ )2, where is the extinction distance of the operating diffraction. Because the thickness measurement depends mostly on the higher-order fringes, it is highly accurate and rather insensitive to measuring errors or to the assignment of ni. The slope, however, depends very much on the measurement of the first fringe and is, therefore, much more susceptible to analytical error. In the program the assignment of ni, is done on the basis of the correlation coefficient, which should be -1.0 for a perfect fit to the straight line. A number of assignments are tested and the one with the lower correlation coefficient taken to be the correct one. If the line gives a positive slope, the results are rejected as a bad measurement. RUNNING PROGRAM CBED WARNING: In order to perform the on-line measurement, the following condi- tion must be set on the PRINTING PARAMETERS page of the MICROCONTROLLER: except for OPERATION, all selections of GENERAL DATA, SCANNING DATA (if present) and SERVICE DATA must be off. This is necessary, because the program retrieves the high tension and the d spacing through a log file. If selections other than OPERATION are on, logging takes too long and the program will interrupt data logging, which causes the microscope to be no longer Remote Controllable. In order to make Remote Control again possible when that happens, the micro- scope must be restarted (small RESTART button underneath DATA DIM). At all times the program can be stopped by pressing Ctrl-C (press the Ctrl key and the C at the same time). However, it is advisable to observe the light of the disc drive used for logging to see if any action is taken there. If the light is out, the program can be stopped without any problems. The measurement of the d spacing forms an important part of the measure- ment. The internal diffraction measurement of the CM microscope must there- fore have been calibrated (in view of the stability, this calibration needs to be done only once every couple of months or so) and the specimen must be in the eucentric position. In order to perform a thickness measurement, the operator must first obtain a Convergent Beam Pattern that shows the thickness fringes. It is important that the condition be as near as possible to a two-beam condition (the dark Kikuchi line goes through the middle of the Bright- Field disc). In general, the conditions necessary to have a sufficiently large disc spacing are obtained in nanoprobe mode. Focus the beam on the specimen and press the Diffraction button. If necessary, switch to a larger or smaller C2 aperture (which will affect the size of the discs). Select a camera length (the higher the camera length the more accurate the measurement) such that the Bright-Field disc and the Dark-Field disc are both still visible on the screen. Start the program by typing CBED and a carriage return, while the computer is in MS-DOS. The program will now determine several things, such as the position of the PRINTING PARAMETERS (2 on a TEM, 4 on a STEM) and will switch OPERATION and MEASURING on the PRINTING PARAMETERS page if necessary. As long as the introductory messages are present on the screen, it is not essential that the microscope is already in diffraction with a focussed beam, so if beam damage or contamination is problematic, focusing of the beam can be done afterwards. Do not interfere with the program by touching any knobs on the microscope. Once the necessary data have been collected, a schematic of the Convergent Beam Pattern is shown on the screen. If not already done, focus the beam and go to Diffraction. Since the determination of the centre of the diffraction discs is rather difficult to do with a high degree of accuracy, the program determines the centres from: (1) the intersection of the dark Kikuchi line and the rim of the Bright- Field disc (it doesnÕt matter on which side); (2) the intersection of the bright Kikuchi line and the rim of the Dark- Field disc (on the same side as the previous measurement); (3) the other intersection of the bright Kikuchi line with the rim of the Dark-Field disc. When so instructed by the program, the operator should move these three points to a reference point such as the tip of the beam stop. For good accuracy it is more important that the Kikuchi lines are positioned precisely on the reference point that it is exactly at the rim of the disc. For moving the pattern, the operator must use the cursor keys of the computer. Using the SHIFT knobs directly will interfere with the measurement and produce poor results. The cursor keys are organized as follows: The arrow keys produce a slow shift in the four perpendicular directions. It is possible that the absolute shift direction does not coincide with that of the screen due to the rotation of the diffraction pattern. The Home, End, Page Down, and Up keys produce a rapid shift in the same four directions. The direction of these keys is such that the shift they produce corresponds to the arrow key 1/8 turn anti-clockwise so that: Home = 10 x arrow right Page Up = 10 x arrow up Page Down = 10 x arrow left End = 10 x arrow down (NOTE: The deviation from their normal usage was found to be more logical in operation at the microscope.) Each point is entered by hitting the carriage return. Once the three points have been determined, the program shifts the diffraction pattern so that the reference point is in the centre of the Dark-Field disc. The operator then shifts the fringes one by one to the reference point with the cursor keys. Again the Home, End, Page Up, and Page Down produce the coarser shift. All shifts are now per- pendicular to the fringes, whether horizontal or vertical cursors are used. The position of the fringes are again entered by hitting the carriage return. Once enough fringes have been measured (four or five should, in general, be enough), hit the carriage return again without further shifting and the program will do the fit to the straight line and produce the measurement results. For the measurement it is not important whether the first fringes are included or not, but all fringes measured should be in a complete sequence; e.g., the sequence 2,3,4,5 is allowed but not 1,3,4,6. Once results are obtained, the progrm asks if they should be printed. If yes, the operator can add text to the print-out, which will contain the high tension, electron wavelength, d spacing, fringe measurements, line fit data, thickness, and the extinction distance. After that the program asks if more measurements are to be performed. If not, the program returns control to MS-DOS. RUNNING PROGRAM CBEDOFF Program CBEDOFF is started by typing CBEDOFF and hitting a carriage return while the computer is in MS-DOS. At all times the program can be stopped by pressing Ctrl-C (press the Ctrl key and the C at the same time). In order to perform a thickness measurement, the operator must first have a Convergent Beam Pattern that shows the thickness fringes. It is important that the condition be as near as possible to a two-beam condition (the dark Kikuchi line goes through the middle of the Bright- Field disc). In general, the conditions necessary to have a sufficiently large disc spacing are obtained in nanoprobe mode. The data required for the measurement are: (1) the high tension (in kV) (2) the d spacing of the Dark-Field disc used (in nm) (3) a measurement of the distance between the Bright-Field with Dark- Field discs (in arbitrary units, but the same as of item 4) (4) the distance between the dark subsidiary fringes and the central Kikuchi line (in the same units as item 3) Start the program by typing CBEDOFF followed by a carriage return. The program will display messages explaining the input data required. If those data have been collected, typing Y will continue into the program, N will return to MS-DOS. The program will then ask for the input data. For the measurement it is not important whether the first fringes are included or not, but all fringes measured should be in a complete sequence; e.g., the sequence, 2,3,4,5 is allowed but not 1,3,4,6. Once results are obtained, the program asks if they should be printed. If yes, the operator can add text to the printout, which will contain the high tension, electron wavelength, d spacing, fringe measurements, line fit data, thickness, and the extinction distance. After that the program asks if more measurements are to be performed. If not, the program returns control to MS-DOS.