DiffTools Instructionsv3-8

p1 of 29

DiffTools Script Package for DigitalMicrograph

D. R. G. Mitchell

www.dmscripting.com, April 2014

version 3.8

p2 of 29

Introduction

DiffTools is a suite of scripts which add functionality for proceesing and analysing selected

area electron diffraction patterns to DigitalMicrograph (DM) (Gatan Inc.). The suite of scripts can be downloaded freely from the internet (details follow). Following installation, a

DiffTools menu appears which give access to a range of menu commands and user-friendly

dialogs. A detailed description of the function of each of these scripts follows. If you find

these tools useful and they contribute to your published work, please consider citing the

following article:

DiffTools: Software Tools for Electron Diffraction in DigitalMicrograph, D. R. G. Mitchell,

Microscopy Research and Technique, 71 (2008) 588-593.

Acknowledgements.

In version 3.0 of this software an Auto centre finding routine was incorporated into the Locate SADP Centre script. This routine appears by kind permission of Dr Tim Petersen of

Monash University.

Updates included in this version and known issues.

Version 3.8 contains new functionality compared with v3.7, and users are encouraged to

update. The Measure Spacings and Angles script now has the ability to label patterns with the

inter-spot angle and length ratio and to outline the zone axis cell. A new Point and Measure

script for rapid (rough) measurement of patterns is provided, which does not require the centre

locating. It has been tested on GMS 2.30, but should be compatible with all previous versions

including GMS 1.7 onwards.

In GMS 2 the scriptability of the Results window, used for displaying data outputs, was

removed by Gatan. Many of the DiffTools scripts would ensure that the Results window was

open before outputting data to it. In order to maintain compatibility with GMS 2, this check

has now been removed. If using GMS versions earlier than 2, users should ensure that the

Results window is displayed (Window/Show Results Window) before running these scripts. No error will occur, but any data output will not be visible.

The commands for configuring the script window used in DiffTools for outputting some tabulations of data changed in GMS 2. In GMS 1.8 DiffTools scripts temporarily deactivate line numbering in script windows so that tabulated data appears correctly. In GMS 2 the code

for controlling this option is not backwardly compatible with GMS 1.8 and so has not been

used. Users of GMS 2 may therefore find that their tabulations of d-spacing data have

numbered lines. This option can be deactivated manually by selecting File/Global Info and then clicking on the Scripting tab, then unchecking the Show Line Numbers option. You may wish to turn this back on when writing scripts, as the line numbers are very useful.

The uninstaller script works correctly in both GMS environments. However, in GMS 2, the

script can not shut down DigitalMicrograph automatically after uninstalling DiffTools. The

uninstaller script will prompt the user to shut down manually after uninstalling the DiffTools

suite.

p3 of 29

Protecting your software.

WARNING: While every care has been taken to ensure that this software is bug-free and will

not damage your system, the possibility does exist. Proceed at your own risk. You MUST

make a copy of your DigitalMicrograph preference files before installing this software. If

problems do occur you can quickly reinstall your system with your saved preferences.

Without a saved set of preferences, this could be a very involved process and might even

involve a Gatan engineer call out.

DigitalMicrograph saves various bits of information in different preference files in

/ProgramFiles/Gatan/DigitalMicrograph/Prefs. Note In Windows 7, or if you installed

DigitalMicrograph for your use only, the preference file may live within your user folder do a search on DigitalMicrographCF.8 to find it.

If you install any standalone scripts, these get stored in the DigitalMicrographCF.8 preference file. All the information in the Global Info, which includes script, camera, GIF

preferences etc gets saved in the DM Preferences.7. There is also a Microscope Data preference file which stores microscope hardware configuration data. These files are very

important, and if they get corrupted, they may cause DM to become erratic or a connected

microscope to become inaccessible. It is very good practice to make regular backups of the

Prefs folder (DM should not be running at the time). Always make a fresh copy before

making any changes to your system such as installing new plugins or other software. Call the copy Copy of Prefs YYMMDD - add a date signature. These folders can live inside the DigitalMicrograph folder as long as you change the name from Prefs. However, it is also a good idea to keep a copy on a separate computer, in case you lose a hard disk. This does not

preclude the need for a sensible overall backup strategy for your computer.

If you do develop problems with your system, Gatan generally advise to make a copy of your

preferences, uninstall DM (except for the licenses), reinstall DM, then copy your old

preferences across. This is fine as long as you know you have a good copy of preferences.

Make a copy of your Prefs folder now.

Sourcing/Installing the Software.

The DiffTools suite of scripts consists of the following:

DiffTools Uninstaller an uninstaller script which will automatically remove all of the DiffTools components.

DiffTools.gtk a package which adds all the scripting functionality.

DiffTools Manual instructions for installing and using DiffTools.

Two example diffraction patterns single crystal silicon and polycrystalline gold.

The above files are available as a zipped archive. It can be downloaded freely from Dave

Mitchells DigitalMicrograph Scripting Website.

http://www.dmscripting.com/difftools.html

Installation is very similar in both GMS environments. The only difference is the location of

the plugins folder inside the Gatan folder where the DiffTools package needs to be installed.

The respective paths (in XP) are:

GMS 1.8 and earlier: C:\Program Files\Gatan\DigitalMicrograph\PlugIns

GMS 2: C:\Program Files\Gatan\Plugins

p4 of 29

Installation Option 1: If a previous version of DiffTools is installed and the current set of

DiffTools preferences is to be preserved for use with the new version:

1. Shut down DigitalMicrograph. 2. Navigate to the plugins folder as shown above. 3. Remove the DiffTools.gtk plugin from the Plugins folder store it in a temporary

folder, in case you need to revert to it.

4. Copy the new DiffTools.gtk package file from the downloaded archive into the Plugins folder.

5. Restart DigitalMicrograph and the revised script menu should appear. 6. All previous preferences/crystal structures etc will still be accessible.

Installation Option 2: If a previous version of DiffTools is installed and the current set of

DiffTools preferences: is NOT to be preserved for use with the new version:

1. Open the DiffTools Uninstaller script in DigitalMicrograph, and ensure all unsaved work is saved and closed.

2. With the uninstaller script front-most (title bar is blue), hit CONTROL + ENTER to run the script.

3. Follow the prompts DigitalMicrograph will shut down (in GMS 2 youll be prompted to shut it down manually). All preference files and the DiffTools plugin

will be deleted.

4. Copy the DiffTools.gtk package file from the downloaded archive into the Plugins folder as shown above.

5. Restart DigitalMicrograph and the revised script menu should appear.

Installation Option 3: If this is a new installation of DiffTools:

1. Shut down DigitalMicrograph. 2. Navigate to the plugins folder as shown above. 3. Copy the DiffTools.gtk package file from the downloaded archive into the Plugins

folder.

4. Restart DigitalMicrograph and the DiffTools menu should appear.

Uninstalling the Software.

The uninstaller script (DiffTools Uninstaller) can be used to remove all traces of DiffTools

from DigitalMicrograph. The files removed are the DiffTools.gtk plugin file from the Plugins

folder (see Sourcing and Installing the Software). It will also remove all DiffTools

preferences from the Global Info. These include any custom crystal structures, camera length

calibrations etc which have been created. The uninstaller script should only be used when you

wish to remove all traces of DiffTools from DigitalMicrograph permanently. For updating to

a newer version of the software, the preceding instructions should be followed.

Software Compatiblity/Disclaimer.

This software should be compatible with all versions of GMS 1.7 2.30 inclusive. Please report any bugs or requests for improvements to the author:

[email protected] (remove the nospam to make this email address work).

The software is provided on an as-is basis. No warranty as to its performance is given, nor any

responsibility accepted for any losses arising from failings in it

p5 of 29

This software is designed ostensibly for single crystal selected area electron diffraction

patterns. Users may find CHT Diffraction Analysis by the same author, of use in measuring ring patterns from polycrystalline materials. The software can be downloaded free of charge

from the same location as this software. The operation of this script is described in D. R. G.

Mitchell, Ultramicroscopy, 108, (2008) 367-374. At present, this script is only compatible

with GMS 1.7-1.8. It will not run under GMS 2.0, as a dynamic linked library requires

recompilation.

Setting Preferences.

There are a large number of preferences which control how the various components in

DiffTools behave. These include default values for dialogs, calibration values etc. These

settings can be changed by the user. Introduced in v3.0 is a DiffTools Preferences script which allows the preferences to be displayed and edited in a user-friendly dialog. Preferences

appear under tabs in the dialog, which relate to specific scripts. Some preferences (Image Size

and Profile Aspect Ratio) appear under the Locate Centre tab, but are in fact common to

several of the other scripts. Preferences can also be found in DigitalMicrograph under

File/Global Info/General/Tags/Diff Tools/XXX where XXX is the relevant preference or

cluster of preferences. Preferences can be edited here, but it is easier and safer to do this via

the DiffTools Preferences script as this applies traps to ensure values entered are within acceptable ranges. Details on preferences for specific scripts are given at the end of their

descriptions below.

The default positions of the dialogs generated by the DiffTools scripts are set when the dialog

is closed (by clicking top right). The dialog will reappear at the location at which it was

closed. In some instances closing the dialog by clicking top right with the ALT key held

down, will save the current settings as the new defaults. A prompt will appear if the script has

this capability.

p6 of 29

The DiffTools Menu.

After installation, a new menu appears in DigitalMicrograph called DiffTools. The menu is

shown in Fig. 1.

Fig. 1 The DiffTools Menu

The menu functions are described in detail below.

1. Calibrate SADP The Calibrate SADP dialog is shown in Fig. 2.

Fig. 2 The Calibrate SADP... dialog

The camera lengths which appear in the dialog, and the calibration values used, are those for a

JEOL 2010F with imaging carried out on a Gatan Imaging Filter (GIF 2000) using a 1k x 1k

camera. Calibration data for a JEOL 2000FXII using a DualVision camera in the 35mm port

(1300 x 1030) is also included. TEM C-F contain dummy data. Calibration data for up to six

p7 of 29

different TEMs can be entered into this script. Users will need to calibrate their systems and

edit the data in the script accordingly (see below). Even where a user has an identical

configuration to one of those for which default data is provided, it is still recommended that

calibration be carried out.

The dialog also displays the maximum resolution of the camera as CCD X and CCD Y for

each set of calibration data. The script will expect the image being calibrated to have either

the same resolution as that specified in the dialog, or to be some integer divisible of it due to binning. Where the correct sized image or its binned counterpart is detected, then the

calibration will be scaled to suit (on the basis of the x axis dimension). However, if scaling of

the calibration is not required, selecting a camera length from the pulldown menu with the

ALT key held down, will apply the unbinned calibration to the image ie no scaling will occur.

If the image is not the correct size then an option to apply a Scaled or Direct calibration is given. If, for example, the Scaled option is selected for a 333 pixel wide image (original CCD X size = 1000 pixels), the calibration will scale on the x-axis and report a binning of 3x.

Choosing the Direct option will apply the unbinned calibration and an image binning of 1x will be reported. This may introduce an error, but where the image size differs from the

original CCD size (in x) by only a few pixels, the error is negligible.

The calibration is done by selecting the tab for the microscope of interest (2010F, FXII etc)

then selecting the appropriate camera length from the pulldown menu. The foremost image is

then calibrated.

Camera length calibration can be done by capturing diffraction patterns from a polycrystalline

gold film at various camera lengths at the maximum CCD resolution. The objective lens

voltage must be constant (eucentric position) throughout. The radius of each diffraction ring

(in pixels) is measured and assigned to its corresponding gold d-spacing in ngstroms. Note

it is essential that d-spacings are in ngstroms. A plot of 1/d-spacing vs ring radius (pixels)

yields a straight line plot (provided the assignment of d-spacings to rings is correct). The

slope of this plot is the calibration factor to be entered into the tag corresponding to the

camera length used. The script Gold Calibrator (see below) enables much of this calibration process to be automated.

If the Calibrate SADP script is run with the ALT key held down, the Calibrate SADP Settings dialog appears showing all the data (microscope name, CCD X and Y, camera length labels, calibration values and calibration errors) for the six TEMs. This data can be edited to

suit. Camera lengths should be entered as shown in the default data eg A. 1.5cm . . By

prefixing the camera lengths with A., B., C. etc the order in which the camera lengths appear

in the pulldown menu of the main dialog is preserved they are sorted alphabetically.

Analysis of gold diffraction patterns used for camera length calibration can be carried out

with the Gold Calibrator script. This carries out a linear regression of the data and reports the calibration factor required here. It also reports the error in the slope. This value can be

entered into the Error/% field in the Calibrate SADP Settings dialog next to the corresponding camera length. If this parameter has a non-zero value, then it is propagated to

all patterns subsequently calibrated with this script. The value appears as a tag on the image

(SADP Calibration:Calibration Error). If measurements are then made on such patterns with

the Measure Spacings and Angles script, d-spacings are reported this percentage error. The problem with using the calibration error is that it tends to underestimate the actual error

in measurement by up to a factor of four. This is because the calibration error does not include

the additional errors arising from location of the pattern centre and the centroids of diffraction

spots. These errors are very user and pattern dependent and are difficult to quantify. An

p8 of 29

alternative approach is to simply set Error/% to some arbitrary value such as 1%. The appropriate value can be assessed by analysing d-spacings from known phases and increasing

the error until all the measured d-spacings ( the error) agree with those expected. If the value

of Error/% is left at its default (zero) value for a particular camera length, then Measure Spacing and Angles will not report any errors for patterns calibrated at that camera length.

Pressing the Save button will replace the existing calibration data with that shown in the dialog. Note, the default data can be restored by opening the Global Info (File/Global Info),

opening the DiffTools tag and deleting the SADP Calibration tag. The next time the Calibrate SADP script is run, the default set of calibrations will be recreated.

The Calibrate SADP Settings dialog also has buttons called Save Tags and Load Tags. Save Tags will save the current calibration values in the Global Info as a settings file. (Note any unsaved changes made in the settings dialog are not picked up by this action, so any

changes should be saved (Save), and then the tags saved (Save Tags)). In this way, if the

calibration data is lost for any reason, the entire setup can be quickly restored. Restoration is

achieved by pressing Load Tags and pointing at the saved settings file to load the data into the dialog. These saved data are then automatically loaded and saved into the Global Info. The

settings file can also be used to copy calibrations between PCs without having to manually re-

enter data. Simply copy the settings file to the target PC, run the Calibrate SADP script with the ALT key held down, and press the Load Tags button and select the settings file. Data in the file will be loaded into the settings dialog and is also saved into the Global Info. The

newly saved settings will be available when SADP Calibration is next run.

The Clear Fields button will clear all the fields in the SADP Calibration Settings dialog.

1a. Calibrate SADP Preferences.

User editable preferences: SADP Calibrations

Note: Setting the values for Calibrate SADP is done via a dialogue which appears when the script is run with the ALT key held down. The only preference relating to Calibrate SADP which appears in the DiffTools Preferences is the number of camera lengths, ie how many

camera length entries appear in the Calibrate SADP (ALT) dialog.

There are six sets of preferences six microscope configurations. For a given microscope configuration eg A which corresponds to the left-most tab in the dialog, there are camera lengths (Camera Lengths A), TEM details (TEM A) and the calibration errors (Camera

Length A Errors). Edit these in the (ALT) dialog and when finished, click Save.

User editable preferences:Default Values

Number of Camera Lengths. This preference is accessed when the DiffTools Preferences script is run and the Calibrate tab is selected. This parameter sets the number of rows for data entry in the SADP Calibration Settings dialog which appears when the SADP Calibration script is run with ALT held down. The default value is 15.

2. Remove or Restore Calibration

This script removes or restores the calibration from the front-most image. It will work on

images and diffraction patterns. For simple measurements it can be easier to work in pixels

than reciprocal Angstroms. Removing a calibration results in the calibration data being

written to a temporary tag on the image. In this way, the script can restore these values, when

the script is next run. Obviously, the script can only restore a calibration on an image which

p9 of 29

has had its calibration removed previously by this script ie where the relevant tag exists on the image.

2a. Remove SADP Calibration Preferences :

User Editable Preferences: None

3. Gold Calibrator

The purpose of the Gold Calibrator script is to automate the analysis of calibration diffraction

patterns obtained from polycrystalline gold. The script dialog is shown in Fig. 3.

Fig. 3 Gold Calibrator dialog.

The starting point for using this script is a radial intensity profile from a polycrystalline gold

diffraction pattern. The pattern should be captured under carefully standardised conditions specimen at eucentric height, objective lens at a standard setting. The profile can be generated

using the Locate SADP Centre script to find the centre (see below), and then the Rotational Average script to generate the profile (see below). The pattern and resulting profile need not be calibrated, since the Gold Calibrator script will simply read the peak positions in pixels.

The position of the peaks in the profile can be determined automatically by having the profile

front-most then clicking on the Find Peaks button. This uses a top hat filter to locate the peaks. The settings (Hat, Brim, Threshold) and operation of the top hat filter are described for

the standalone version of this script THF Peak Locator (see below). If the Label checkbox is set, the peaks in the profile will be labelled as will data points in the linear regression.

The Find Peaks function should locate most of the peaks in the profile. Adjusting the Threshold parameter will change the sensitivity. It is not essential to find all peaks. Very weak peaks or unresolved shoulders on peaks can be ignored, especially if there is any

uncertainty in their positions. The peak finding routine may detect false peaks, such as the

edge of the beam stop. Any such false detections must be deleted from the marked up profile

by clicking on the small red cross on the vertical dashed line marking the peak. This should

then change from a cross to a red square. The marked peak can then be deleted by pressing

delete.

p10 of 29

Once the peaks in the profile have been satisfactorily marked, the data can be read into Gold

Calibrator by pressing the Data button. If more than 16 peaks are marked only the first 16 are read in. By default, the first peak position is assigned to the gold (111) spacing. This first

data point is used to generate an approximate calibration factor for automatic fitting of the

remaining peaks. The first peak can be assigned to a different gold spacing by holding down

the ALT key and pressing either the up or down arrow adjacent to the first d-spacing field.

This will assign to the data point the d-spacing of the next available gold spacing (up or

down) and all the other data will be refitted on the basis of the new approximate calibration of

the first data point.

The fitting process plots 1/d-spacing (in -1

) versus diffraction ring radius (pixels). If the

measured radii are assigned to the correct gold spacings then the plot will be a straight line

and the slope of that line is the calibration factor required by the SADP Calibration script described earlier. The fitting of the gold spacings is carried out automatically and the plot

shown in the dialog. The equation of the fit and the correlation coefficient are also displayed.

Manual adjustment of the fit is possible by pressing the up or down arrows next to all the data

points. This will select the next gold spacing in the library of 43 gold spacings built into the

script. The effect of this change is displayed in the plot and in the various fit parameters

shown in the results section of the dialog. The Error parameter is particularly useful for judging how well the data are fitted. This value tends to a minimum when the first data point

is correctly assigned (to gold (111)) and all other peaks are correctly fitted. If one or more

data points do not fit well, this may be due to some error in determining their peak position examine the radial intensity plot. If that is the case, delete the markers of the relevant peaks,

press the Clear button to reset the dialog, then press Data followed by Fit.

The slope of the plot is displayed in the Calibration/-1/pxl field. Also shown is the standard error (Error/%) in the slope as a percentage. These two values are used in the SADP Calibration script. The Volts/kV field refers to the microscope voltage and the CCD Pixel/m to the pixel size of the camera. These two values are used only for computing the camera length. If you do not require a camera length these fields can be ignored. The CCD

pixel size information can be found within the Global Info of DigitalMicrograph which is

connected to the microscope. Open the Global Info (File/Global Info), click on General/Tags,

then navigate to the following tag group:

Camera Manager

Camera Configuration List

[0]

CCD

Pixel Size (um): 14+i14

This is the configuration for an Ultrascan (2k x 2k) Camera under GMS 1.8.3. The pixel size

is 14 x 14m. We also have a second camera (Erlangshen) on this machine The details for this

camera appear under the [1] tag group, which appears below the [0] group shown above. The

pixel dimensions are 6.25 x 6.25m. The location of the information may vary with DM

vintage and camera type.

If you cannot locate this information you should consult the documentation which came with

your camera or ask your local Gatan representative. More recent versions of DM allow users

to edit a calibration lookup table in diffraction mode. In this way, patterns are calibrated as

they are acquired. This calibration requires a camera length value. Be aware that the camera

length displayed on the TEM is for film only . The camera length for the camera may be quite

different, especially where the camera field of view is very different to that of film.

p11 of 29

This script can also be used for calibrating data measured from film. The diffraction ring

diameters are measured, then halved to determine the ring radius. These values (in mm) are

entered into the Data fields. The CCD Pixel size is set to 1000m.

The values currently displayed in the dialog can be saved as new default values by holding

down the ALT key when closing the dialog. A prompt appears asking if the current settings

are to be saved as new defaults. Settings can also be saved under the Gold Calib tab in the DiffTools Preferences script.

3a. Remove SADP Calibration Preferences :

User Editable Preferences: None

The settings under DiffTools Preference/Gold Calib are as follows: Label Font Size: The size of the labels which appear in the linear regression plot in the

dialog: The default value is 14.

Spot Size: The size of the plotted points in the linear regression plot in the dialog: The default

size is 3.

Label Check Box (0-1): The default setting for the label check box at launch. Can be 0 (Off)

or 1 (On): The default value is 1.

Hat: The value in the Hat field. Sets the hat width in pixels for the top hat filter: The default

value is 11.

Brim: The value in the Brim field. Sets the brim width in pixels for the top hat filter. The

default value is 5

Threshold: The value in the Threshold field. Sets the threshold of the top hat filter: The

default value is 0.001.

Voltage: The value in the Volts field. Defines the microscope voltage used (in kV): The

default value is 200.

CCD Pixel Size: The value in the CCD Pixel field. Defines the size of the pixels in the CCD

camera (in m). The default value is 6.25.

4. Locate SADP Centre Any measurements on diffraction patterns with this software are dependent upon knowing the

pattern centre (ie the position of the central (000) beam) accurately. Recently, an accurate and

robust (CHT) method of locating the centre and measuring the pattern automatically was

developed by this author Circular Hough Transform Diffraction Analysis (D. R. G. Mitchell, Ultramicroscopy, 108 (2008) 367-374). For good quality powder diffraction patterns

where accurate d-spacing information is required, this method is robust. However, it does take

a few minutes to run, and in instances where approximate d-spacings will suffice, or where

the pattern is not amenable to CHT analysis as it is from a single crystal/very coarse grained

material or where the approximate centre needs to be located prior to using the Rotate and Inset script (see later), a quick manual method of defining the centre is desirable. This script provides a tool for doing this with reasonable accuracy (probably to within 5 pixels or

so). The dialog is shown in Fig. 4.

p12 of 29

Fig. 4 Locate SADP Centre dialog

Pressing the Get Image button will source the front-most image which should be a diffraction pattern. The Centre check box will draw a small red region of interest (ROI) at the geometric centre of the pattern. When set, the Mask check box will draw a series of concentric rings about the centre. The Rings box sets the number of rings in the concentric mask. The purpose of this mask is to assist the user in locating the position of the central

(000) beam by aligning the rings with diffraction rings or a series of spots. The concentric

rings are magenta and every nth ring is cyan to aid mask and ring alignment. The selection of which nth rings are highlighted in cyan can be set in the preferences see below. The direction arrows move the centre and mask in steps of 5 pixels. With the SHIFT key held

down this reduces to single pixel steps. The position of the centre in X and Y is shown below

the movement buttons. The size (arrow) buttons allow the size of the central red ROI to be

adjusted.

There are three options for finding the centre:

Manual this takes whatever the current centre is set at and marks the image with a tag which contains the details of this position. A number of other scripts in DiffTools will look

for, and use, this tag.

C. of G. centre of gravity. Pressing this button will reposition the centre to the centre of gravity of the image. This approach is fairly unreliable at locating the central beam position in

patterns which are not centro-symmetric. These may include off-axis patterns, high index

patterns, patterns with a beam stop in, or those with strong comet tails from electronic CCD

shuttering, Kikuchi lines etc. If the centre based on the centre of gravity is accepted the image

will be tagged appropriately. If the centre of gravity is not sufficiently accurate the C. of G. button should be pressed while holding down the ALT key. This applies a threshold to the

image removing all intensity below the mean value of the image. Typically this will eliminate

weak features such as Kikuchi lines, and leave only the high intensity diffraction spots for

analysis

p13 of 29

Auto begins by calculating the first moment of the image and uses this as an approximate centre. The image is then rotated + and about an angle which can be set in the preferences. The difference between these two rotated images is a function of how close the guessed centre

is to the true centre. The difference will be at a minimum when the true centre is guessed. A

coarse matrix of guessed centres about the first centre is tested in this manner, and the centre

providing the best fit is retained. In the second pass a finer matrix of points centred on the

new centre is tested, and the centre further refined. This process is repeated though six passes

it typically takes about a minute or so. The ROI and Mask (if displayed) are moved to the new centre and the co-ordinates are displayed. The option to accept the chosen centre is given.

Clear will restore the pattern to its original condition and will delete any annotations/images created by this tool. If the centre has been marked, a tag containing the centre position

information will remain on the image for use by other tools in this suite. The centre position

tag can be removed by holding down the ALT key and clicking on Clear. This will perform the normal clear operations and also remove the centre position tags from the pattern.

4a. Locate SADP Centre Preferences : User Editable Preferences: Default Values

The settings under DiffTools Preference/Locate Centre are as follows: Mark the Centre (0-1) determines whether the centre check box is on or off at launch. The default value is 1 (on). Allowable values are 0 or 1.

Number of Mask Rings the value which appears in the Rings field. The default value is 9 Centre ROI Width the size of the red ROI in pixels which marks the centre. The default value is 10.

Default Image Size determines the vertical height (in pixels) at which images are displayed on the screen. Set this to suit your display. This setting is common to all scripts in this suite.

The default value is 512.

Default Aspect Ratio the aspect ratio (X/Y) of intensity profiles. This setting is common to all scripts in this suite which create such profiles. The default value is 1.6.

Highlight nth Ring every nth ring in the concentric mask will highlighted in cyan, while the rest of the mask is magenta. To turn the highlighting off, set this value to zero. The default

value is 3 every 3rd ring is highlighted. Auto Rotation Angle the net angle over which the + and images are rotated. The default value is 15 degrees, which corresponds to +7.5 and -7.5 rotations of the respective images.

5. Rotate and Inset

The change in lens configuration between imaging and diffraction mode may introduce a

rotation between an image and its corresponding diffraction pattern. If a Gatan Imaging Filter

(GIF) is used a further rotation between the imaging plane and the camera is introduced. To

determine crystallographic directions in images, this rotation must be determined and

compensated for. Traditionally MoO3 is used to calibrate the rotation (see Transmission Electron Microscopy by D. B. Williams and C. B. Carter p152). The result of a calibration should be a table of rotations as a function of magnification.

This script allows the user to inset one image (typically an SADP) into another with a rotation

at a specified angle applied. An option to crop the image, to remove the black space from

rotation is also provided. The default position of the inset image is set with the position check

box and the size is set by the scale factor which is the size of the inset image relative to the parent image. The dialog is shown in Fig. 5.

p14 of 29

Fig. 5 Rotate and Inset dialog

The diffraction pattern is sourced with the SADP button. As this script will rotate and excise the central region of the image (if Crop is selected), it is important that the centre of the pattern has been defined with Locate SADP Centre prior to running this script. An alert will be shown if the position of the central beam (000) has not been defined. If this centre has

not been set, the script will still operate, but the geometric centre of the SADP will be used

instead. The result of this is that if the SADP is not centred within its frame, the pattern will

appear off-centred as an inset. An image is selected with the Image button. Thumbnails of the SADP and image are shown as are their image labels.

The position of the inset in the image top right, top left etc is set with the Position checkbox. To change the position, uncheck the current position and check the new one.

Rotating an image will introduce black space at its corners. This is most evident at a rotation

of 45 degrees. An option to crop the image to retain the largest possible square region of the

SADP (which does not contain any of the added black space) is provided when the Crop option is checked. The angle of rotation specifies how the diffraction pattern is rotated relative

to the image. Positive angles correspond to rotating the pattern anticlockwise relative to the

image and negative angles vice versa. If you simply wish to inset one image into another

without rotation, set the angle to zero. The Scale parameter is the displayed size of the inset SADP relative to the displayed size of the image. The Inset SADP button will carry out the rotation and inset. The inset image is a floating layer and can be repositioned, stretched,

scaled and have its histogram/gamma etc adjusted to taste. If subsequently saved as a Gatan

(.dm3) file, the floating layer will be retained. If exported as a TIFF via File/Save Display as TIFF, the inset becomes flattened into a single layer image and can no longer be changed independently of the parent image. Note that if the File/Save As . . TIFF option is used, then only the parent image will be saved and the inset image is lost. If the ALT key is held

down when the Inset SADP is pressed the inset pattern is also displayed as a standalone image. This can be useful for creating centred copies of patterns where the largest square

region of the pattern, centred on the pattern centre, is required. The size of the excised image

can be controlled to some extent, by turning on the Crop option and setting appropriate rotation angles an angle of 0 giving the largest d-spacing range, and around 45 the smallest depending on the location of the pattern centre.

p15 of 29

5a. Rotate and Inset Preferences: User Editable Preferences: Default Values

The settings under DiffTools Preference/Rotate and Inset are as follows: Crop Inset Image (0-1) determines whether the Crop check box is set to on (1) or off (0) when the dialog is launched. The default value is 1. Acceptable values are 0 and 1.

Inset Position (1-4) determines which Position checkbox is set when the dialog is launched. 1=top right, 2=top left, 3=bottom right, 4=bottom left. The default value is 1.

Acceptable values are integers in the range 1 to 4 inclusive.

Inset Scaling (0.1-0.8) determines the default value in the Scale field when the dialog is launched. The Scale value is the displayed size of the inset/parent image. The default value is 0.35. The range of acceptable values is 0.1-0.8 inclusive.

Inset Angle - determines the default value which appears in the Angle box. The default value is -47 degrees.

6. Rotational Average

This enables rotational averaging to be performed on the front-most image. This should be a

diffraction pattern which has had its centre defined with Locate SADP Centre. The script carries out rotational averaging over an angular range defined by the start and end angles. It

also permits selection of an arc of diffracted intensity by defining both angular and radial

ranges over which to analyse. In this way diffracted intensity around a particular diffraction

feature can be analysed. The dialog is shown in Fig. 6.

Fig. 6 Rotational Average dialog

Get SADP will select the front-most diffraction pattern and enable the dialog. Clear will remove all images created with this script and reset the dialog. The angular range over which

averaging will be performed is specified by the Start and End angles. The 12 oclock position on patterns (relative to the centre) corresponds to zero degrees, and angles increase in

a clockwise manner. The start angle may be either positive or negative. A start position at 9

p16 of 29

oclock may therefore be specified as either 270 or as -90. End angles may only be positive in the range 0-360. Values in all the fields may be changed by typing in values directly or by

using the + and - buttons. These buttons change the value by 5. The magnitude of the change can be modified by holding down various keys: SHIFT = +/- 1, ALT = +/- 10.

When the Get SADP button is pressed a mask of arrows is superimposed onto the image (Fig. 7), which identify the angular range to be analysed. The start angle position is shown by

the cyan arrow and the end angle position by the magenta arrow. For straightforward

rotational averaging, this is all that needs be specified. Pressing the Rot. Av. Button will generate a rotational averaged intensity profile of the pattern over this angular range. Note,

the pattern will be averaged over a radial distance defined by the centre to nearest edge

distance. The length of the arrows has no influence on the radial range over which averaging

occurs. If the Normalise check box is checked, the intensity values in intensity profiles and 3D surface plots will be normalised to the maximum intensity, which will have a value of

100; otherwise the true intensities will be displayed. Note, due to limitations in the

interpolation function used to carry out the rotational averaging, errors of around 1.2% in the

intensities may be introduced. This should be borne in mind when attempting accurate

quantitative work.

Fig. 7 SADP marked up by the Rotational Average script. Rotational averaging will be carried out over the angular range defined by the cyan arrow (start angle=90) and the

magenta arrow (end angle=300). The yellow arrows mark the angular range to be rotationally

averaged. This occurs over a radial range defined by the centre to nearest edge distance.

The following example (ordered TiO2 grown on MgO Figs. 8a-c) demonstrates the use of the arc segment extraction function. The Range parameter enables a range of radii to be selected to capture the full width of the diffraction feature. This is invoked by selecting the

Arc Select checkbox, and causes a pair of rings to be drawn on the pattern. The mean radius of the annulus defined by these rings is that specified by the Length parameter. The Range parameter, which sets the width of the annulus, can be varied between 2 and 50, although the

upper limit can be changed in the DiffTools Preferences. Both the arrow length and the

diameter of the outer ring are limited so that they can not run off the edge of the image.

Therefore, when working with features which are very close to the edge of the image,

adjusting either the Length or Range parameters may cause the value of the other to be limited.

p17 of 29

Fig. 8a shows a pattern from textured anatase (TiO2) grown on (001) MgO. A feature in the

second inner-most diffraction has been chosen by selecting the Arc Select check box. The angular limits of 5 and 40, have been set. Pressing the Rot. Av. button extracts this arc segment, warps it into a straight image, then averages it in the radial direction to produce an

intensity profile as a function of angle. The resulting plot is shown in Fig. 8b. This shows that

the arc of diffracted intensity in the main diffraction feature is made up of a number of intense

features, suggesting that a number of orientation variants are present. From this plot it is

possible to measure the angles between those variants.

It is also possible to display the selected arc segment as a 3D surface plot (Fig. 8c), by

selecting the Surf. Plot checkbox. Pressing the Rot. Av. Button will then cause the rotational average of the pattern (between the start and end angle), the average intensity along

the arc segment, and the 3D surface plot of the arc segment to be displayed.

a) b)

c)

Fig. 8. a) Diffraction pattern from textured TiO2 grown on an MgO substrate. The second

innermost diffraction ring has been selected with the Arc Select feature and an arc defined between 5 and 40; b) Averaged intensity profile (averaged in the radial direction) along the

arc segment defined in a); c) 3D surface plot of the arc segment defined in a), created by

selecting the Surf. Plot checkbox.

p18 of 29

6a. Rotational Average Preferences: User Editable Preferences: Default Values

Start Angle the starting angle for rotational average etc (in degrees). The default value is 0. End Angle the end angle for rotational average etc (in degrees). The default value is 359. Radius Length the length parameter which sets the length of the arrows in the mask, used to define the radius of interest and identify the angular range of measurement. The default value

is 200 (pixels).

Radius Range defines the range or width of the annulus created with the Arc Select function. The default value is 20 (pixels).

Minimum Radius Range The minimum value of the Radius Range parameter. The default value is 2 (pixels).

Maximum Radius Range The maximum value of the Radius Range parameter. The default value is 50 (pixels).

Inter arrow Angle Specifies the angular separation of yellow arrows drawn on the image. This setting is not displayed on the dialog. The default value is 15 (degrees).

7. THF Peak Locator

The THF Peak Locator dialog is shown in Fig. 9.

Fig. 9 THF Peak Locator dialog.

This dialog uses a top hat filter (THF) to process the front-most profile. It will only work on

line profile images such as intensity profiles. Use the Rotational Average script to create an appropriate profile from a diffraction pattern. The dialog provides access to the top hat

filtering parameters hat width, brim width and threshold. One advantage of using rotational averaging to generate a profile is that the noise level is significantly reduced compared to a

simple linear profile. In very noisy profiles, a significant number of false peaks may get

identified. The default values for these three parameters work well with most profiles. They

may need adjusting for noisy profiles. Here, increasing widths and the threshold value may

help. If peaks are missed then decreasing the threshold value may help. For poorly resolved

doublets or shoulder peaks using narrow brim and hat widths may be useful.

The dialog provides options to identify peak maxima or minima (Maxs, Mins). It can also add labels (Label) to identified peaks which give their peak position. The Text check box

p19 of 29

gives the option to output the measured data as a tab delimited text file. The left most column

of line numbers does not get saved with the data and can be ignored. To save the data click on

the Save button at the bottom of the text window and select Text .txt as the format. The data saved are the peak positions, the reciprocal values of the peak positions, the peak

intensities and the normalised peak intensities at the channels corresponding to the peaks

(normalised to the most intense feature found). If the profile is created from a calibrated

diffraction pattern then these data correspond to reciprocal spacing, d-spacing and peak

intensity respectively.

With a profile foremost, pressing the Measure button will create a copy of the foremost profile and name it Image Name (THF). Top hat filtering of this profile is then carried out, and based on the THF parameters, peaks/troughs will be identified. When the Report button is pressed the positions of all the identified peaks are reported in the Results window. If the

Text option is checked, the same data are also displayed in a new window as tab-delimited text.

7a. THF Peak Locator Preferences User Editable Preferences: Default Values

THF Hat Width the width of the Hat parameter in the THF Peak Locator dialog. The default value is 11.

THF Brim Width the width of the Brim parameter in the THF Peak Locator dialog. The default value is 5.

THF Threshold the value of the threshold parameter in the THF Peak Locator dialog. The default value is 0.001

Note the size and aspect ratio of the profile will be the same as the original profile.

8. Invert SADP.

This script works on images, but not on line profiles. It will simply invert the intensities of the

foremost image, so that a diffraction pattern will appear as black diffraction rings/spots on a

white background. Weak diffraction features are generally more visible in this mode when

displayed in hard copy form. The relative intensities of diffraction features are unchanged.

8a. Preferences

User Editable Preferences: None.

9. Reveal Weak Reflections

This script works on images but not line profiles. It copies the foremost image and applies an

exponent to it. The default value is (1/3) a negative cube root. The new image is renamed with the value of the exponent used appended to it. This negative cube root serves to 1) invert

the image intensities and 2) apply a cube root which greatly compresses the contrast range and

greatly increases the visibility of very weak features. Fig. 10 shows the effect of applying a)

an intensity inversion to the image and b) the negative cube root. The weakest outer

reflections are far more visible in b). Obviously this script changes the relative intensity levels

in a diffraction pattern, precluding subsequent measurements of intensities. However, it is

useful for illustrating weak reflections due to phenomena such as ordering, second phase

precipitation etc.

p20 of 29

a) b)

Fig. 10 Effect of applying a) Inverting SADP and b) Reveal Weak Reflections scripts on a diffraction pattern of gold. Weaker reflections are more apparent in b)

9a. Reveal Weak Reflections Preferences

User Editable Preferences: Defaults

Exponent the exponent applied to the pattern. The default value is an inverse cube root (-0.333).

10. Measure Spacings and Angles

This is a utility which can be used for measuring the d-spacings of spots and rings in

diffraction patterns with sub-pixel precision. It can also determine the angles between spots

and their corresponding d-spacings (if the pattern has been appropriately calibrated). The ratio

of the reciprocal spacings of the two selected spots is also reported. Even in the absence of a

calibration, the interspot angle and this ratio enables reflections and zone axes to be

determined. The dialog is shown in Fig. 11.

p21 of 29

Fig. 11 Measure Spacings and Angles dialog

This tool requires that the SADPs centre has been located beforehand. If the centre is not set, the script will proceed and use the geometric centre of the pattern. This is useful if

measurements on Fast Fourier Transforms (FFTs) are required, where the centre of the FFT is

the geometric centre.

There are two measurement modes: Centre d and dd. With Centred selected,

pressing Mark will apply three regions of interest (ROI) to a copy of the foremost image. The size of the copied image is scaled to match the vertical size of the front-most image.

Stretch the vertical size of this to the required height before pressing Mark. The copied image has the same name as the original with (Angles) appended to it. This image can be saved at the end of the measurement session. The original image is not changed. This script

will work on both normal contrast patterns (white diffraction features on a black background)

and on patterns which have inverted contrast due to application of the Invert SADP script.

The centre of the pattern is marked with a blue ROI which cannot be moved. A red and green

ROI are also added and these can be manually positioned on diffraction spots to be measured.

Fig. 12a shows a pattern where the red ROI has been moved close to one spot, The green ROI

is in its initial location. It is not necessary to position the ROI over the spot of interest with

any accuracy. The centre of gravity button (C. of G.) will automatically relocate the ROI to centre it on the spot. This should be fairly accurate even where streaking in the spots is

present, provided the streaking or other intensity variation is centro-symmetric about the spot.

Fig. 12b shows the effect of pressing the C. of G. button on the position of the ROIs. Spot labelling has been turned on by checking the Label Spots option. The font size of the labels in set in the Label Size field. The red ROI, which was close to a spot has become centred on it. The green ROI is in a region of uniform (low) intensity and has not moved. The user would

need to relocate this ROI onto a spot and then press C. of G. again to centre it. The centre of gravity method is iterative. It measures the centre of gravity within the ROI moves the ROI to that centre recalculates the centre of gravity etc. The number of times this iteration occurs is set in the CoG Number of Interations preference (see the end of this section). The default

p22 of 29

number of iterations is 20, and execution takes about a second or so. A smaller number of

iterations may help speed things up, but the ROI may then not be at the true centre of gravity,

if its starting position was a long way away from the centre of gravity. If this is a possibility,

simply pressing C. of G. again will refine the ROI positions further, if they are not currently at the centre of gravity.

a) b)

Fig. 12 Diffraction pattern after selecting the Mark option in Measure Spacings and Angles.a) The user has positioned the red ROI over a spot of interest approximately; b) After pressing the C. of G. button, the red ROI is then automatically centred on the spot. The green ROI is too far from a spot to locate it it will need to be repositioned closer to a spot of interest.

If the ALT key is held down when the C. of G. button is pressed, the red and green ROIs are thresholded before their centre of gravity is determined. The threshold point is the mean value

of each region, with everything below the mean being set to zero. This effectively eliminates

weak features such as comet tails and Kikuchi lines, leaving only the most intense part of the

diffraction spot upon which to determine the centroid.

When the Measure button is pressed, the lengths of the vectors from the centre of the pattern to the ROI centres are calculated and are shown in the Red and Green fields of the dialog. If the location of the ROIs has been set using the C. of G. button the reported position is the calculated centre of gravity. If the ROI has been manually positioned and the Measure option selected, the reported position is the geometric centre of the ROI. The option to locate

spots automatically and manually was included to allow for streaky or diffuse spots where

automatic location may fail. If labelling is selected the colour of the label reflects whether the

measurement was based on manual location (magenta label) or via a centre of gravity

approach (white label). If a position is determined via centre of gravity and then measured, the

d-spacing will be displayed in the label (white text) on the image. If the Measure button is pressed again without moving the ROI, the spot is remeasured and is treated as a manual

measurement (the label is magenta). Small differences between these two measurements are

due to sub-pixel differences between the centre of gravity position (non-integer values) and

the centre of the ROI (integer values).

The reciprocal distances (d-spacings if the pattern is calibrated) are shown in the 1/Red and 1/Green fields. If an error value has been entered into the Calibrate SADP script calibration values, then this value will be used to calculate the error in both the vector

p23 of 29

determination and the corresponding reciprocal thereof the d-spacing. The errors are shown in the results window as percentages of the main values. See the Calibrate SADP section for information on how to include errors into this measurement. If the error is left at its default

(zero) value, then no error values will be reported when using this script. The interspot angle

is shown in the Angle/deg field. Finally the ratio of the two distances on the pattern are shown in the Red/Green field. These data are also reported in the Results window.

Measurement of further spots simply entails moving the ROIs to be centred approximately on

the spots of interest, pressing the C. of G. button to centre them, and finally pressing the Measure button to generate the data. Clear will close the measured image. The option to save the labelled image is given.

If the second measurement option is selected (dd), the measurement effectively ignores

the centre position and makes measurements between pairs of reflections across the centre.

Pressing the Mark button adds a Red and Grn ROI, which should be manually positioned over reflections as described above. Pressing C. of G. will then centre these ROIs on their reflections. It also creates two new ROIs at mirror image locations across the centre. Note the

centre location does need to be defined to carry out this measurement. However, the location

need only be approximate as it is only used to position approximately the two mirror image

ROIs (Der and Nrg) over the equivalent reflections to those marked by the user (equivalent

reflections means if the user marks a (100) reflection with the Red ROI, then the equivalent

reflection ( 100) is marked with the Der ROI). The mirror image ROIs are then centred on the

equivalent reflections using the centre of gravity approach. This method relies on the C. of G. button being pressed. Only then will the Measure button be enabled. Manual positioning of ROIs for measurement can not be performed in this mode.

Pressing the Measure button causes the vector lengths to be extracted by measuring the distance between pairs of marked reflections (ie the Red-Der and Grn-Nrg distances) and

halving them. The centre of the pattern is calculated from the crossover point between the

Red-Der and the Grn-Nrg vectors. This centre is used when calculating the interspot angle ie the angle subtended by Red Crossover Point Grn. The originally marked (approximate) centre is ignored for this measurement. The originally defined centre can be refined to use the

centre calculated as the crossover point. Holding down the ALT key when pressing Measure will cause the position of the original centre and the new crossover point to be displayed

along with the option to Refine the centre to the crossover point or keep the existing centre.

This is an alternative method with which to locate the centre accurately. Simply use the

Locate SADP Centre tool to set the centre approximately, then carry out the above procedure and refine the centre. Maximum accuracy in the crossover point should be obtained

when the Red-Der and Grn-Nrg vectors cross each other at right angles, and when using

higher index reflections, which are furthest from the pattern centre. If the Label option is checked when carrying out measurements in dd mode, d-spacing values will be added as

labels to spots. The labels are cyan in colour to differentiate them from measurements made in

Centred mode (C. of G. = white labels, Manual=magenta labels).

The four check boxes change the way the pattern is marked up during measurement:

Angles: when selected this will draw lines from the centre to each spot, and the angle between

them will then be labelled.

Ratios: when selected the ratio of the length of the red to the green vector is added to the

display.

Cell: when selected a line is drawn between the two spots. This is useful for outlining the

symmetry of the cell in a zone axis pattern.

p24 of 29

Yellow: the above three functions will be drawn in different colours for each pair of spots

measured. If Yellow is selected all subsequent mark-ups will be done in yellow. This provides

higher contrast for publication. Alternatively, if white is preferred, click and drag across all of

the markings to select them, the right click and select: Foreground Color and select white or

the colour of choice.

10a. Measure Spacings and Angles Preferences User Editable Preferences: Default Values

Centre ROI Size the size (in pixels) of the blue ROI which marks the centre. The default value is 20.

Spot ROI Size the size (in pixels) of the red and green ROIs (in pixels) used to mark the spots of interest. The default value is 20.

CoG Number of Iterations the number of iterations used when the C. of G. button is pressed. The centre of gravity (CoG) of the red and green ROIs is determined and the ROIs

are then moved to be centred on the calculated value. This is repeated the number of times set

by this value. The default value is 20.

Add Labels (0-1). sets the default condition of the Label Spots check box at launch. 1=on, 0=off, the default is 1. When the dialog is closed the current setting for this becomes the new

default value.

Angle Measurement Mode (0-1) Sets the default condition for which measurement mode (Centred (0) or dd (1) is used. The default value is Centred (0). When the dialog is

closed the current setting for this becomes the new default value.

Label Size sets the default value for the Label Size field which controls the font size (in points) of the labels. The default value is 18. Acceptable values are 8-72 inclusive. When the

dialog is closed the current setting for this becomes the new default value.

11. Point and Measure . . .

This is a quick and dirty script for measuring diffractions patterns crudely. The script will select the front-most pattern. A caption on the pattern informs the user to select spots by

moving the cursor over them and hitting the SPACE bar to mark them. If the pattern centre

has not been defined previously, then the first point marked is assumed to be the centre and

will be labelled with [000]. Every subsequent spot will be marked with the d-spacing. The

script makes no adjustment for incorrect location of the centre or the spot. The script is not

intended for accurate measurement, but simply to eyeball a pattern and see whether the d-spacings are in the ballpark of a target value.

11a. Point and Measure Preferences

User Editable Preferences: None

12. Background Stripper

Background stripper enables the background intensity of a profile or spectrum to be modelled

and then stripped away. The dialog is shown in Fig. 13. The script acts on whichever image is

front-most. Only 1D images, such as intensity profiles can be processed with this script. There

are four background models available. In all cases at least one region of interest (ROI) must

be fitted to the front-most intensity profile. In the case of the polynomial model two ROIs

p25 of 29

must be present and for the spline fitting, two or more ROIs must be fitted. It is essential that

the ROIs are added from left to right and that they do not overlap.

Where multiple profiles of very similar data are to be processed, the ROIs need only be fitted

to the front-most profile. Pressing the ROI>> button will propagate the ROIs to all open profiles (2D images will be ignored). The ROIs can be propagated by channel number or by

the calibrated values, by selecting the Channel or Calibn option in the dialog.

There are four background models which can be selected with the Background Model pulldown menu:

Log-normal. Electron diffraction pattern backgrounds can be modelled as a log-normal curve, though generally only over a limited range of reciprocal space values. When this option

is selected, the log-normal parameters (Scaling, Sigma, Range) become available, and a log-

normal curve is fitted. This is scaled to match the intensity at the lower limit of the ROI. The

curve shape can be adjusted by modifying the log-normal parameters. These values can be

entered either directly into the relevant fields, or the values can be changed by clicking the +

and buttons. The magnitude of the changes produced by these buttons can be increased by a factor of 10, by holding down the SHIFT key, and decreased a factor of 10 by holding down

the ALT key. The ROI should be fitted to the region over which it is intended to background

strip. Moving the ROI will not change the curve fit, only the range over which the strip

occurs. The log-normal fit will most likely be poor if the region of reciprocal space selected is

too large.

Fig. 13 Background Stripper screen shot

Polynomial: To fit a polynomial curve two ROIs must be fitted to the front-most intensity profile. The three points to which the polynomial is fitted are the intensity values at the left

hand ROI lower and upper limit, and the right hand upper limit (the lower limit of the right

hand ROI is ignored). The polynomial fit can be dynamically adjusted by moving the ROIs.

Care must be taken to select the ROI limit (left or right) of choice. If the ROI limit is missed

and the background is clicked and then dragged, a new ROI will be added, existing ROIs will

be deleted and the dynamic updating of the polynomial will be broken. If this happens, simply

reapply the ROIs, reselect Polynomial from the pulldown menu and try again. This also applies for spline and linear fits, which achieve dynamic updating via ROI repositioning.

Spline. To fit a cubic spline curve two or more ROIs must be present on the front-most image. The spline curve is fitted through the intensity values at the upper and lower limits of

p26 of 29

all fitted ROIs. The spline fit will be updated dynamically when any of the ROIs limits are

moved. The spline is perhaps the most versatile background fitting routine. However, care

must be taken not to create unrealistic backgrounds. These may touch the minima between all

peaks in the profile, but form a very convoluted background which has no physical meaning.

Aim to create a smooth background curve with no inflections.

Linear. To fit a linear background, one ROI must be present on the front-most image. The linear fit is updated dynamically when the ROI limits are moved. Clearly a linear fit can not

be used on steeply sloping and curving profiles, but it may be of use in EELS or EDS spectra

over narrow spectral ranges where the background is very flat.

Once the background has been applied and adjusted, it can be removed from the profile by

selecting Strip. The stripped profile has the value displayed in the Offset field added to it. This may be useful for overlaying multiple profiles with various offsets to allow comparison.

If the ALT key is held down whilst pressing Strip, then both the stripped profile and the background fit will be displayed as separate images.

Reset will remove any fitted background from the image, and will reset the dialog. If ALT is held down whilst pressing Reset then in addition to the above, the dialog values will be restored to the default values.

When the dialog is closed by clicking top right, the dialog position is memorised and the

dialog will appear at the same screen location when the script is next run. If ALT is held down

whilst closing the dialog, then a prompt will appear asking if the current settings of the dialog

should be saved as the new defaults. Dialog preferences for this script may also be edited in

DiffTools Preferences, by clicking on the Bkgd Strip tab.

12. Background Stripper Preferences User Editable Preferences: Default Values

ROI by Channel or Calibn (0-1) sets the method by which the ROIs are calibrated - either by Channel (0) or by Calibn (1). Sigma the value in the log-normal Sigma field. Default value is 0.66. Range the value in the log-normal Range field. The default value is 5. Offset the value in the Offset field. This is the offset applied to stripped backgrounds. The default value is 10.

Note: Scaling is not saved as a default value, since the scaling is calculated each time the log-

normal curve is applied.

13. Crystal Tools This dialog provides a basic crystallographic calculator to aid in the indexing of diffraction

patterns. It provides a simple interface and is easy to use. It does not supplant the need for

full-strength commercial diffraction programs, but avoids the need for switching in an out of

DigitalMicrograph when doing simple calculations. The dialog is shown in Fig. 14.

The crystal system is set with the radio buttons on the left. The lattice parameters and unit cell

angles are set in the top right hand corner. Once this information has been entered,

calculations can be performed. The pulldown menu below the crystal selection buttons allows

selection of data for a number of predefined crystal systems. There are approximately twenty

of these crystal structures/data sets included with the script. However, this crystal data set can

be readily expanded to include users own crystallographic data.

p27 of 29

To enter custom crystallographic data. Set the pulldown menu to blank so that no structure is

selected. Then select the crystal lattice, set the lattice parameters and the unit cell angles.

Below the Hex < - > Rhomb check box is a text window wherein the name of the new crystal can be entered. Crystal structures are sorted alphabetically in the pull down menu, and

so some consistency with naming structures is required eg label elements with their symbol followed by their name (eg Au-Gold). Compounds can be given a common mineral name (eg Rutile) or formula (TiO2). Probably formulae are best, as some compounds do not have

simple names. Also, in the case of polymorphs/allotropes consider adding a qualifier (mineral

name or crystal structure) to the name (eg TiO2 Rutile, Co-hex) to help differentiate the polymorphs/allotropes. Pressing the + button will show the details of the structure that has been entered. If this is all correct pressing Save will save it. The structure is added to the structure catalogue associated with this script. However, it will not be available in the pull

down menu until the script is next launched. Calculations based on the newly entered data can

still proceed. The Hex < - > Rhomb check box will report the equivalent hexagonal data when working with rhombohedral structures and vice versa.

Fig. 14 Crystal Tools dialog.

To avoid losing crystal data inadvertently, the data should be saved as a settings file every

time a new crystal is added to it. The data can be saved by holding down the ALT key and

presing the + button. This brings up a dialog to save the entire set of crystal data in a file called CrystalTools_Settings_20100607. The reverse date stamp at the end of the file name will change with the date. If the crystal data are inadvertently lost from the Global Info, they

can be restored by loading the information from a settings file by holding down the ALT key

and pressing Clear. This brings up a dialog to select a Crystal Tools Setting file. This method can also be used to transfer suites of Crystal Tools crystal data to other PCs.

Four main types of calculation are possible with Crystal Tools : d-spacings, interplanar

angles, interdirection angles and zone axis calculations. For a given set of crystallographic

data, the four panels providing access to these calculations are accessible by clicking on the d

Spacings, Planes, Directions and Zones tabs.

Calculations are performed by pressing the Calc button. At present creating tabulations of data is only implemented for d spacings. Other tabulations functions may be added later if

requested. To create a tabulation of d spacings enter/select the crystallographic data, select the

p28 of 29

d Spacing tab, then press Table. A prompt to select a maximum value of h or k or l or a minimum d spacing value will appear, and the table will then be computed.

With the d Spacings tab selected entering the three Miller indices (hkl) for the reflection of interest and pressing Calc will calculate the d-spacing. If the crystal structure is hexagonal, then the display will request a Miller-Bravais indice (hkil) (i=-(h+k)). Only the h, k and l

indices need be entered, the i index is calculated automatically.

With the planes tab selected the prompt changes to Miller (or Miller-Bravais) indices.

Pressing Calc computes the interplanar angle for these two planes. Similar calculations can be performed for directions.

The Zones tab computes the zone axis for two planes, based on the Weiss law. For consistency, the h2,k2,l2 values must be of a reflection which is anticlockwise on the pattern

relative to h1,k1,l1.

13a. Crystal Tools Preferences User Editable Preferences: Crystal Tools

Crystal tools preferences are not displayed in DiffTools Preferences.

Crystals - The catalogue of crystallographic structures and data is found in the Crystals tag group within the Crystal Tools tag. Typically you would need to edit these data tags only to completely remove a structure which was no longer wanted or to edit incorrect data. Ensure

the Crystal Tools dialog is not open when editing the catalogue. If you incorrectly name a

structure (eg if you called it CaTiO4 instead of CaTiO3) you can not change crystal names.

The only option is to delete the incorrect crystal (see below) then re-enter the data from within

the Crystal Tools dialog.

To remove a crystal, open the Crystals tag group in the Global Info (File/Global Info/Tags/DiffTools/Crystal Tools/Crystals) and right click on the crystal you wish to delete select delete.

To edit the data for a crystal, open the Crystals tag group and right click on the parameter you wish to change and select Edit. Enter the new information. The parameters are : a,b, c in ngstroms, alpha, beta, gamma in degrees and Lattice. Lattice takes the following values:

0=cubic, 1=tetragonal, 2=orthorhombic, 3=hexagonal, 4=rhombohedral, 5=monoclinic and

6=triclinic.

14. Tilt Angle Calculator This computes the net tilt between two sets of goniometer positions when using a double tilt

holder. Simply enter the initial and final tilts in X and Y and press calculate to compute the

net tilt in degrees. This is useful for tilting between zones axes. Here, a knowledge of the net

tilt can help confirm zone axis identification.

14a. Tilt Angle Calculator Preferences

User Editable Preferences: None

15. DM Calculator

p29 of 29

This a scientific calculator, useful for various calculations.

15a. DM Calculator Preferences

User Editable Preferences: None

16. DiffTools Preferences ...

This displays the preferences for the DiffTools scripts (Fig. 15).

Fig. 15 DiffTools Preferences dialog

Descriptions of the preferences for each of the scripts appear at the end of the relevant section

above.

16a DiffTools Preferences

User Editable Preferences: None