Thermo Fisher (TMO) forges ahead with initiatives to strengthen foothold in the high-potential electron microscopy market. Thermo Fisher to Buy Gatan, Boost Electron Microscopy Suite - June 29.
In this section, we will use discuss the Gatan implementation of energy-filtering and spectroscopy instrumentation. Other systems have similar controls and considerations.Before you begin an experiment, it is important to familiarize yourself with the various controls, parameters, and adjustments of the Gatan imaging filter system. This will help you understand the basic functionality that you will use to collect energy-filtered TEM images, elemental maps, and electron energy loss spectra from your transmission electron microscope (TEM) sample. Optical column. Entrance aperturesThe GIF entrance aperture determines what portion of the TEM beam will contribute to EFTEM images and electron energy loss spectra. Each GIF has four entrance apertures that are selectable under pneumatic control.
The largest entrance aperture (5 mm – GIF Tridiem ® and ® SE, or 9 mm – GIF Quantum ER system) you can use for all GIF imaging applications. This aperture and the TEM magnification fully determine the specimen field of view the GIF will capture.
![Gatan microscopy suite 7 Gatan microscopy suite 7](http://www.gatan.com/sites/default/files/styles/300w/public/GMS3.gif?itok=YnK2_ncf)
Imaging apertures for the GIF Quantum are square and are not useable for spectroscopy. The two smaller apertures (2.5 and 5 mm for GIF Quantum system) are useful for spectroscopy applications. The smallest of these apertures cut out the portion of the electron beam that contributes to small residual aberrations of the GIF prism focus (see below) and thereby helps achieve the best possible energy resolution in the electron energy loss spectra. Apertures also control the collection angle and signal intensity in electron energy loss spectroscopy (EELS) experiments. Finally, there is a special mask aperture with a precision fabricated square array of holes.
Use this solely for automated tuning and alignment of your GIF. Multipole opticsSophisticated multipole optical elements (e.g., deflectors, quadrupoles, sextupoles) pre-shape the beam to compensate for the residual focus errors of the GIF prism. Although complex, the fine adjustments of this multipole system are fully automated via the Tune GIF function of the ® (GMS) software. Energy dispersing magnetic prismThe heart of the GIF is its magnetic prism. This element generates a magnetic field that bends the electron beam when it enters the GIF through 90°. This disperses the electrons along a vertical line in accordance to their energies.
Zero-loss electrons, which have the most energy, are bent the least and appear at the bottom of the vertical distribution. Core-loss electrons, which have lost the most energy and are the least energetic, are bent the most and appear at the top of the distribution. The prism also acts as a lens with focusing properties. In this respect, its main function is to form a sharp (demagnified) image of the final TEM projector system crossover at the plane of the energy selection slit. Energy selection slitThe energy selection slit focuses the energy dispersed electron beam. During unfiltered imaging and spectroscopy applications, the slit is fully retracted.
For EFTEM applications, the slit is inserted then the width is adjusted to select the electron energy loss range you desire in each EFTEM image. Typically, at 200 kV, the width of the energy selection varies from 2 – 100 eV, which represents a physical slit opening of 5 – 250 µm on the GIF Quantum system. Energy offset controlsWhile you can adjust the slit to select a variable energy range, the centroid of that energy range varies when you electron optically move the dispersed electron beam up and down relative to the slit edges. You can accomplish this in any of three ways.Spectrum offset (TEM high tension) – The spectrum offset function increases the TEM high tension by an amount equal to the average energy loss you wish to have in an EFTEM image.
Primary electrons that now have a higher energy will lose energy in the sample and pass through the slit at the original ZLP energy. This is the recommended offset method for all EFTEM imaging applications. Note: This feature is not available on dedicated spectrometers.Energy shift (prism current) – The energy shift function varies the strength of the prism magnetic field when you change the current in the drive coil.
This method is recommended when a large energy offset is required (up to 10 keV) and for monochromated systems operating at low (.
AbstractNumerous digital image analysis techniques have been developed with regard to transmission electron microscopy (TEM) with the help of programming. DigitalMicrograph (DM, Gatan Inc., USA), which is installed on most TEMs as operational software, includes a script language to develop customized software for image analysis.
Based on the DM script language, this work provides a script source listing for quantitative strain measurements based on a geometric phase analysis. Keywords: DigitalMicrograph script, Transmission electron microscopy, Quantitative strain measurement, Geometric phase analysis. INTRODUCTIONTransmission electron microscopy (TEM) is a powerful tool for analyzing a broad range of materials with a variety of analysis techniques. Conventional TEM analysis techniques have been improved continuously since TEM was developed in 1930s. One recent advance is the incorporation of aberration correctors, which drastically enhances the resolution of TEM.
On the other hand, numerous data analysis techniques based on post-image processing have also been introduced in an effort to obtain more information from raw TEM data, which is scarcely distinguishable by empirical methods or manual calculations (;;;; ).Several tools, such as Matlab, C/C and Python, can be used to develop software for advanced image processing from TEM images. DigitalMicrograph (DM, Gatan Inc., USA) also provides a scripting language called DM script for developing data analysis software for TEM. The syntax of DM is very similar to that of C/C. Unlike other programming languages, however, DM script simultaneously serves as a compiler to create an object file, as a linker to create an executable file, and as a runtime environment. More precisely, DM works as not a low-level language but as an interpreter for customized scripts. For this reason, its computation speed is relatively slow. Nevertheless, DM script provides a convenient programming environment with a variety of functions to manipulate images and to access the TEM instrument for customized operations.
Several analysis techniques have already been reported based on the DM script (; ). Kim and colleagues recently introduced new algorithms to quantify the symmetry recorded in convergent beam electron diffraction (CBED) patterns and to perform scanning CBED for symmetry mapping (;;; ). Also, the DM script database contains a considerable number of applications for customized software scripting using the DM script language (FELMI-ZFE Internet).Using the advantages of DM script, this work provides a DM script source listing for quantitative strain measurements. Quantitative strain, or displacement measurement, is a widely used image processing technique for materials.
Algorithms for strain mapping are mainly classified into two groups in terms of real and reciprocal space. The Peak Pairs algorithm (PPA) calculates the local displacement using the intensity maximum of the peaks between each pair of neighboring lattice points in real space ( ). Another algorithm known as the geometric phase analysis was proposed. Geometric phase analysis (GPA) is based on phase retrieval calculated in the Fourier (reciprocal) space. While the two techniques are complementary to each other, PPA can fail to calculate the local displacement with a complex lattice structure ( ).
In addition, GPA is created relatively easily with DM script language considering the computation process of the DM script. As mentioned earlier, DM script only works as an interpreter, unlike other low-level programming languages. It therefore requires a significant amount of computation time for a complex loop structure. The PPA algorithm requires a pixel-by-pixel operation to find neighboring lattice points such that the computation time increases rapidly, even for a small two-dimensional image. A dynamic-link library (dll) file may need to be created using C in order to create the script for PPA on DM. In contrast, the DM script has the basic functions for basic Fourier/ inverse Fourier transform and image processing, which are the major functions for the GPA algorithm.
This results in a reasonable computation time reduction on DM. In this report, a DM script source listing for GPA is created and listed in the Appendix (online only at to be used directly on DM. THE PROGRAMshows the GPA software installed on DM. In the Appendix, the script source listing can be found and can simply be installed on DM included in Gatan Microscopy Suite (GMS, version 2.11, 32-bit architecture; Gatan Inc.).
GMS can be freely downloaded from Gatan’s website.A fast Fourier transform or power spectrum image is initially calculated from an HR image in order to measure the local strain, rotation and displacement. For the power spectrum image, the software provides a “Gaussian edge smoothing” function to make spots more distinguishable in the calculated power spectrum. From the power spectrum, any two nonlinear g-vectors can be chosen using two virtual apertures (the oval annotations in DM). Shows an example of a simulated HR image and is the power spectrum calculated from with a Gaussian edge smoothing value of 50. In order to determine two g-vectors, two non-linear spots are selected with the oval annotations (yellow circles) as shown in and are then stored in the software using the “Set g-vectors” function. In this process, two reciprocal lattice vectors are determined by the distance from the center of the power spectrum to the intensity maximum of the peaks. The intensity maximum can be calculated with sub-pixel accuracy based on the center of mass of the intensity of the peaks (the CenterOfMass subroutine in the script source; please see the appendix).Once two g-vectors are selected, a phase map, P( r), can be generated with the “Run” button in the phase image construction menu.
Shows the generated P( r) images for the selected g-vectors. From the first calculated P( r) images, the selected g-vectors can be refined. As shown in, a region-of-interest tool automatically appears on one of the P( r) images.
The user then needs to choose a proper reference area and then refine the g-vectors using the “Refine-g” button. New P( r) images are then reloaded with the refined g-vectors using the “Run” button in the phase image construction menu. These procedures should be repeated until the g-vectors have the same values.
Show the final P( r) images after several refinement procedures.Based on the refined P( r) images, the local strain (?), rotation ( ω) can be calculated using the “strain” button in the mapping menu (see ). Show the calculated strain maps of? Xy ( ) and the rotation matrix ω xy ( ). The local strain fields are observed in the calculated strain maps.
A user can also obtain the map of u( r) by using the “displacement” button in the mapping menu. The two-dimensional displacements, u( r), are used to calculate the strain maps with equation (9) to (11).
Display u( r) for the x and y direction, respectively. The unit of u( r) is determined by the unit of used images. Thus, are represented based on the unit of pixel because the lattice vectors calculated from the simulated HR image have the unit of pixel.Other images of the raw phase image, masked power spectrum, amplitude map, and full/reduced Bragg-filtered images can be also obtained from the software as shown in. From the raw phase image, the local reciprocal lattice g-vectors are directly calculated by differentiation: ∇P’ g( r)=2π g( r). This is then used to calculate the phase image P g( r) by subtracting 2πi g r. The corresponding algorithm can be found in the Appendix (please see the “//Phase image, P(r)”).
While the amplitude map and full/reduced Bragg-filtered images are not directly related to the calculation of strain map, they can be applied to image an antiphase domain boundary ( ) or dislocations ( ). In order to obtain the images shown in, a user needs to check “display options” in the phase image construction menu. FiguresDialogue box for the geometric phase analysis software installed on DigitalMicrograph (Gatan Inc., USA).(A) Simulated high-resolution image for geometric phase analysis.
(B) Gaussian-edge-smoothed power spectrum calculated from(A, B) Calculated P( r) image for two g-vectors. (C, D) P( r) images with the refined g-vectors. A region-of-interest box automatically appears on one of the P( r) images to refine the selected g-vectors.Calculated strain maps for? Xy (C); rotation map for ω xy (D); two dimensional-displacement u( r) along x (E) and y (F).Raw phase map P’ (A), amplitude map A (B), full Bragg-filtered map H (C), and reduced Bragg-filtered H (D) for the lattice reciprocal vectors of 1 and 2.