High Resolution Electron Microscopy

Electron microscopy is a vital characterization tool for relating structure to the other three corners of the materials science and engineering tetrahedron, namely synthesis/processing, properties/performance and theory/modeling. This has become increasingly true with the shrinking scale and increasing complexity of structural and electronic components. The Electron Microscopy Laboratory (EML) is therefore a key part of the efforts of LosAlamos in the broad area of materials science research. Let us examine the past, present, and future of this important activity.

The EML was significantly expanded in 1988 with the installation of two state-of-the-art transmission electron microscopes, one analytical electron microscope (AEM) and one high resolution transmission electron microscope (HRTEM), respectively a Philips CM30 300kV AEM and a Philips CM30st HRTEM, plus a Camscan scanning electron microscope (SEM). Over the years, we have added Kevex energy dispersive x-ray spectrometers for the AEM and SEM, a Gatan parallel electron energy loss spectrometer (PEELS) also for the AEM, a Dingley camera for orientation determine in texture analysis for the SEM, and various image processing, enhancement and computing features for the HRTEM. These are being achieved using (1) a Synoptics sysTEM unit for image processing, which can be done on-line via a Gatan video camera, and (2) a MacTempas multi-slice computer program for HREM image simulation. These systems have been under continuous development over the past few years as new techniques have come to fruition and a wide variety of materials science problems associated with advanced materials and processing have been tackled. With the completion of the Materials Science Laboratory (MSL), we have been in a unique position to make a dramatic advance in our EM capabilities. We have installed three new instruments based on field-emission (FE) electron source technology - a FE-HRTEM, a FE-STEM, and a FE-SEM. All three instruments are operational and performing in a spectacular fashion.

There is nothing particularly new about the FE source itself. It has stringent vacuum requirements which have tended to make it incompatible with regular electron microscopy. VG built their FEG-STEM many years ago starting with a UHV system; the STEM performed brilliantly but was very user-unfriendly. However, they have finally developed a new unit, the HB-601, which is reasonably friendly and more importantly, is capable of both high resolution imaging (by Z contrast imaging) and fine probe chemical analysis. Several manufacturers have now developed a FE-SEM which is fairly straightforward because of the modest voltages (30kV) needed for a SEM; vacuum problems have been solved by having small apertures between the specimen chamber and the FE source. An important side benefit of this development is that operation at low voltages (down to 500V) is simple and resolution is only slightly degraded so that non-conducting organic and inorganic materials can be imaged without coating. The FE-HRTEM has been most difficult because of the higher voltages involved (300kV). The impetus to finding a solution was provided by the Brite Euram project whereby Philips cooperated with several European Universities (Delft, Tubingen and Antwerp) plus later involvement from AT&T Bell Labs. The critical components of the project are the FE source itself, the special stable specimen stage, the ultra-high resolution objective pole-piece, and the image retrieval system. Meanwhile JEOL and Hitachi realized the significance of these happenings and developed their own instruments. The FE source is a key component in all these new instruments. The importance of the field emission is that it provides a much brighter source, a smaller virtual source, a smaller intense probe, and a coherent beam of electrons for holography and high resolution applications. These instruments will provide a quantum jump in our capabilities for atomic-level structural and chemical analysis of features such as grain boundaries, interfaces and dislocations in advanced materials. It is our intention to develop these facilities to their fullest extent for the characterization of defects in materials such as ceramics, intermetallics, composites, superconductors, and electronic materials.

The FE-SEM (JEOL JSM-6300FXV), FE-STEM (VG HB601) and FE-HRTEM (JEOL JEM 3000F) have all now been delivered and installed in the MSL. The PEELS system on the AEM has been upgraded and is being installed on the FE-HRTEM because of its superior probe size and voltage stability. A new PEELS system is also being installed on the FE-STEM. We have selected new or upgraded EDS systems for the various SEM's and TEM's; these are or will be networked via Ethernet. Networking in a practical way has been talked about for years but has never been useful. The EDS systems are all now based on workstations, PCs, or MAC's and so networking of the EDS systems, the central computer system, and individual computer systems is workable. In fact the FE-SEM has digitized image storage and processing; printing is achieved on a dye-sublimation printer and film is totally eliminated; we have already demonstrated networking capability and remote operation. The JEM 3000F also has a CCD Multi-scan camera and software for image processing and analysis plus software for autotuning the microscope. Again networking is being used for image transfer, storage, and printing.

It is important to appreciate that the generally accepted resolution in a HRTEM is determined by the spherical aberration coefficient, Cs, of the objective lens; the best Cs available for a 300kV microscope is ~0.7 mm, giving a resolution of ~1.7Å. However, for a FE source, the beam is highly coherent and there is information in the diffracted beams out to 1 Å and beyond. The phase of this information (contrast transfer function-CTF) is oscillating rapidly in this region, making interpretation difficult, but computing techniques have been developed to extract meaningful information out to the 1Å level. One method is to take a focal series around the so-called Lichte focus where the CTF is oscillating at a minimal rate; computer programs have been developed by the Antwerp group (Van Dyck et al.) to take a focal series of 20 digitized micrographs and extract a map of the projected potential which represents the atomic positions. There is also a joint project between Michael O'Keefe at LBL and Ondri Krivanek at Gatan to design an improved program along the same lines. The interest is to use this for atomic distributions around dislocations, interfaces, and other defects.

The techniques developed for high resolution structural and chemical imaging are being applied to a wide variety of problems in the arena of advanced materials and processing. Most applications involve dislocations, grain boundaries, interfaces and other defects in ceramics, intermetallics, composites, multilayers, superconductors, and electronic materials. Close interaction are being maintained with the materials modeling community. This is particularly valuable for the assessment of atomic arrangements around defects where the observations can be compared with calculations and used to feed back information for improved interatomic potentials.

It is anticipated that there will be increasing demands for microstructural analysis at the atomic level. Electronic devices are already at the sub-micron level and are being inexorably pushed toward the nanometer range. This means that knowledge of chemistry and structure at the atomic level will be essential. Present capabilities in instrumentation are for chemical analysis via EDS and PEELS at the nanometer level. Incremental improvements in probe size and probe current from better FE sources plus improved detector sensitivity will lead to 1Å spatial resolution in chemical analysis to go with existing 1Å spatial resolution in structure. We also anticipate the increasing importance of texture analysis and plan to order a new FE-SEM with electron back-scattered diffraction pattern capability (EBSP) for mapping the orientation of grains in polycrystalline materials. Again the spatial resolution of this technique needs to be pushed. Another area of growing importance in instrumentation is in situ experimentation for both SEM's and TEM's. Static observation is the common mode for electron microscopy but dynamic observations during heating, cooling, straining, irradiating, or environmental reaction experiments are obviously of great significance. With improved ccd camera design there is no reason why this cannot be done at the atomic level. We also plan to develop holographic techniques for the FE-HRTEM so that holography becomes routine rather than a curiosity.

We also anticipate total integration of microscopes and analytical systems into computer systems and networks. Darkrooms will become obsolete. Standard 1024x1024 images will become 2048x2048, then 4096x4096, etc. Each pixel will also contain a PEELS and/or EDS spectrum. For this, massive storage will become necessary; presently 1MB images are common, in the future 1GB images will be common. In the near term, in fact, we plan to develop a pixel-by-pixel system for gathering PEELS spectra on the FE-STEM. We also anticipate that quantitative information will become increasingly possible from atomic resolution data. This includes both chemical and structural information around defects such as interfaces and dislocations; such quantitative information can then be compared with theory and modeling and used to improve interatomic potentials.

So far as research opportunities are concerned, the need for understanding behavior at the atomic level in advanced materials pushes the state-of-the-art in high resolution electron microscopy and the results in turn push the development of new materials. This will be particularly true in the arena of nanostructured materials as they approach atomic dimensions in thin films and multilayers. Such devices include not only standard silicon technology but also integrated ferroelectrics, superconducting, magnetic, and mechanical devices. The need for atomic resolution is by no means confined to thin film devices. In structural and electronic materials, behavior is determined at the atomic level for example, the behavior of point defects, dislocations and grain boundaries in high temperature materials; the interaction of domain boundaries with defects in ferromagnetic and ferroelectric materials; phase transformations during heating and cooling; in situ dynamical environmental reactions; influence of local electronic bonding on the tribological and electronic properties of hard diamond-like materials. These and many other research opportunities would be difficult to realize without the existence of a central state-of-the-art EM facility which will in turn not only foster the numerous multi-disciplinary collaborations but also bridge the gap among the different materials related groups and divisions within Los Alamos.

Next Research Topic | Topic Index

LANL | DOE | UC
Phone Book | Search

L O S  A L A M O S   N A T I O N A L   L A B O R A T O R Y
Operated by the University of California for the US Department of Energy