Y. Kim

Quantitative chemical lattice imaging: theory and practice

Abstract We describe how the composition of materials may be quantitatively mapped with near-atomic resolution and sensitivity. The procedure combines chemical lattice imaging (which sensitively records the compositional information) with vector pattern recognition (which efficiently extracts and quantifies the local information content). We describe the theoretical underpinnings which allow the local information content to be related to the local sample composition, and the procedure by which this may be realized in practice. Single and double atom substitutions in individual atomic columns of semiconductors can thus be detected at about 60% and 90% confidence levels, respectively. Potential complications due to nonlocal effects (Fresnel fringing, dynamical scattering), geometrical sample imperfections (thickness changes, surface roughness), radiation damage, and photographic nonlinearities are shown to be negligible.

Quantifying the Information Content of Lattice Images

Quantitative information may be extracted from local areas of images that consist of one or more types of unit cell. Fourier-space analysis, real-space intensity analysis, and real-space vector pattern recognition are discussed. The pattern recognition approach efficiently exploits the available information by representing the intensity distribution within each unit cell of the image as a multidimensional vector. Thus, the amount and the effect of noise present are determined, statistically significant features are identified, and quantitative comparisons are made with model images. In the case of chemical lattice images, the position of a vector can be directly related to the atomic composition of the unit cell it represents, allowing quantitative chemical mapping of materials at near-atomic sensitivity and resolution. More generally, the vector approach allows the efficient and quantitative extraction of information from images, which consist of mosaics of unit cells.

Low temperature interdiffusion in the HgCdTe/CdTe system, studied at near-atomic resolution

By combining chemical lattice imaging and vector pattern recognition we determine, as a function of annealing temperature, the composition of individual atomic planes across each HgCdTe/CdTe interface of a multiquantum well stack. The resultant composition profiles, which directly reveal the chemical change across each interface at near atomic resolution, are analyzed in terms of linear and nonlinear diffusion theory, to deduce the interdiffusion coefficient and its activation energy. We find the interdiffusion coefficient to be nonlinear, and a sensitive function of the interface depth beneath the surface.

Quantitative chemical mapping: Spatial resolution

Abstract We show that the spatial resolution of quantitative chemical mapping is determined by the larger of two factors: the periodicity of the chemically sensitive reflections, or the size of the unit cell adopted for pattern recognition. Under appropriate chemically sensitive imaging conditions, nonlocal effects due to dynamical scattering and Fresnel fringing are insignificant, even when artifically large discontinuities in the sample projected potential are introduced.

Importance of steps in heteroepitaxy: The case of aluminum on silicon

We have observed dominant epitaxial growth of Al(100) films on chemically cleaned, hydrogen terminated, off‐oriented Si(111). The films were deposited by thermal evaporation at room temperature in ultrahigh vacuum. X‐ray diffraction shows sharp and intense Al(200) diffraction, enhanced by postdeposition annealing. Crystal quality and the dominance of Al(100) structure depend strongly on the substrate treatment and the off‐cut angle, both of which control the steps on the Si(111) surface. The steps were found responsible for the epitaxial alignment of the film and the substrate lattices. Details of this alignment were observed in transmission electron microscopy cross‐sectional images of the interface. Our findings are in contrast to previously published results which indicate epitaxial growth of Al(111) on Si(111).

Interfacial Stability and Interdiffusion Examined at the Atomic Level

Using chemical lattice imaging in combination with vector pattern recognition, we obtain quantitative profiles of the chemical change across single interfaces with atomic plane resolution. We thus study interdiffusiuon across single GaAs/AlGaAs interfaces as a function of temperature, depth of interface beneath the surface, and doping. Since our technique is sensitive to interdiffusion coefficients as small as 10 −20 cm 2 /s, we can study atomic level changes at a single interface at the low temperatures used for many device processing steps (∼700C). Our results show interdiffusion, and hence the layer stability depend not only on temperature and doping, but also on the distance of the interface from the surface. The implications of these results for the stability of multilayered structures are discussed.

Microstructure of the emitter polycrystalline silicon/silicon interface in bipolar transistors after rapid thermal annealing

We use high‐resolution transmission electron microscopy to determine the microstructure of the emitter polycrystalline (poly) Si/Si interface of real bipolar transistor devices after rapid thermal annealing. Our results quantify the size and density of interfacial oxide voids (pinholes), the areal fraction of the interface covered with voids, and the amount of epitaxial regrowth after annealing in the range 1000–1150 °C. Correlation of these results with the characteristics of the devices shows that the most dramatic electrical changes occur before 2% of the poly‐Si/Si interfacial area is covered with oxide voids.

HgCdTe/CdTe Multiple Quantum Wells: Growth, Stability, and Optical Properties

We have grown HgCdTe/CdTe multiple quantum wells by molecular beam epitaxy which show room temperature photoluminescence and sharp absorption steps at mid-infrared wavelengths. Quantitative chemical mapping, performed by transmission electron microscopy, indicates minimal interdiffusion during growth. Annealing experiments performed at higher temperatures show that the interdiffusion coefficient is a strong function of the depth of the interface below the surface. Absorption spectra have been accurately modeled with a square well/envelope function approach. The films have been used to passively mode lock color center lasers and produce pulses as short as 120 fsec near 2.7 μm.

Quantitative Hrtem: Measuring Projected Potential, Surface Roughness and Chemical Composition

We describe how general lattice images may be used to measure the variation of the potential in crystalline solids in any projection, with no knowledge of the imaging conditions. This approach is applicable to structurally perfect samples, in which interfacial topography, or changes in composition are of interest. We present the first atomic-level topographic map of a Si/SiO 2 interface in plan-view, and the first microscopic compositional map of a Si/GeSi/Si quantum well in cross-section.