Protocols for conducting in situ transmission electron microscopy (TEM) reactions using an environmental TEM with dry gases have been well established. However, many important reactions that are relevant to catalysis or high-temperature oxidation occur at atmospheric pressure and are influenced by the presence of water vapor. These experiments necessitate using a closed-cell gas reaction TEM holder. We have developed protocols for introducing and controlling water vapor concentrations in experimental gases from 2% at a full atmosphere to 100% at ~17 Torr, while measuring the gas composition using a residual gas analyzer (RGA) on the return side of the in situ gas reactor holder. Initially, as a model system, cube-shaped MgO crystals were used to help develop the protocols for handling the water vapor injection process and confirming that we could successfully inject water vapor into the gas cell. The interaction of water vapor with MgO triggered surface morphological and chemical changes as a result of the formation of Mg(OH)2, later validated with mass spectra obtained with our RGA system with and without water vapor. Integrating an RGA with an in situ scanning/TEM closed-cell gas reaction system can thus provide critical measurements correlating gas composition with dynamic surface restructuring of materials during reactions.

Electron energy loss spectroscopy (EELS), X-ray photoelectron spectroscopy (XPS), and transmission electron microscopy have been used to study iron catalysts for Fischer–Tropsch synthesis. When silica-containing iron oxide precursors are activated in flowing CO, the iron phase segregates into iron carbide crystallites, leaving behind some unreduced iron oxide in an amorphous state coexisting with the silica binder. The iron carbide crystallites are found covered by characteristic amorphous carbonaceous surface layers. These amorphous species are difficult to analyze by traditional catalyst characterization techniques, which lack spatial resolution. Even a surface-sensitive technique such as XPS shows only broad carbon or iron peaks in these catalysts. As we show in this work, EELS allows us to distinguish three different carbonaceous species: reactive amorphous carbon, graphitic carbon, and carbidic carbon in the bulk of the iron carbide particles. The carbidic carbon K edge shows an intense “π*” peak with an edge shift of about 1 eV to higher energy loss compared to that of the π* of amorphous carbon film or graphitic carbon. EELS analysis of the oxygen K edge allows us to distinguish the amorphous unreduced iron phase from the silica binder, indicating these are two separate phases. These results shed light onto the complex phase transformations that accompany the activation of iron catalysts for Fischer–Tropsch synthesis.

We show heteroepitaxial growth of GaAs on Ge/SiGe grown on nanometer-scale grating structures. Conventional lithography techniques were combined with reactive ion and wet-chemical etching to fabricate 1-D patterns of silicon posts. The quality of the GaAs layers was investigated using high-resolution x-ray diffraction (HRXRD), transmission electron microscopy (TEM), scanning electron microscopy (SEM), photoluminescence (PL) and etch pit density (EPD) measurements. Our results show significant improvement in the quality of heteroepitaxial layers grown on nano patterned structures compared to those on the unpatterned silicon. The optical quality of the GaAs/Ge/SiGe on nano-scale patterned silicon was comparable to that of single crystal GaAs.

We report highest quality Ge epilayers on nanoscale patterned Si structures. 100% Ge films of 10 μm are deposited using chemical vapor deposition. The quality of Ge layers was examined using scanning electron microscopy (SEM), transmission electron microscopy (TEM), and high-resolution x-ray diffraction (HRXRD) measurements. The defect density was evaluated using etch pit density measurements. We have obtained lowest dislocation density (5×105 cm-2) Ge films on the nanopatterned Si structures. The full width half maximum peaks of the reciprocal space maps of Ge epilayers on the nanopatterned Si showed 93 arc sec. We were able to get rid of the crosshatch pattern on the Ge surface grown on the nanopatterned Si. We also showed that there is a significant improvement of the quality of the Ge epilayers in the nanopatterned Si compared to an unpatterned Si. We observed nearly three-order magnitude decrease in the dislocation density in the patterned compared to the unpatterned structures. The Ge epilayer in the patterned Si has a dislocation density of 5×105 cm-2 as compared to 6×108 cm-2 for unpatterned Si.

A model was developed to calculate the elastic fields, including strain energy density, in multilayers grown epitaxially on a planar substrate. This model works well for compliant and non-compliant substrates. In particular we illustrate the model for four layer heterostructure and apply it for graded Ge (SixGe1−x) grown on a planar silicon substrate. Using the equations for static equilibrium and Hooke's law for isotropic materials under a plane stress condition, the elastic fields associated with each layer were calculated. The strain partitioning in this model reduces to the limiting case of a two- layer structure available in the literature. As it turns out here, strain partitioning is a function of the bulk unstrained lattice parameters, elastic constants and thicknesses of the layers. The model was qualitatively verified by comparing the strain energy density with the dislocation density away from a relatively thick substrate. This model helps shed some light on the factors important in achieving defect free multilayers for optoelectronic devices.

We describe novel 2-D structures that facilitate strain relief and allow us to obtain Ge epilayers that are free of defects. These structures can potentially absorb thermal expansion and lattice expansion mismatch as well as enable liftoff of heteroepitaxial layers for subsequent wafer reuse. Conventional lithography techniques were combined with reactive ion and wet-chemical etching to fabricate 2-D patterns of silicon posts. The dimensions of the posts were varied keeping the pitch (center to center distance) constant. Heteroepitaxial growth of Ge/SixGe1−x on these micrometer-scale structures was investigated. While, keeping the growth parameters constant, the geometry of the structures was varied to determine the optimum configuration for the highest quality heteroepitaxial growth. The quality of the Si1−xGex buffer system was investigated using high-resolution x-ray diffraction. Transmission electron microscopy (TEM) was used to analyze the epilayer cross-sections. Surface morphology was analyzed using scanning electron microscopy (SEM), atomic force microscopy (AFM) and optical microscopy. Our results show that the quality of the heteroepitaxial layers improves as the width of the posts in the 2-D pattern was decreased.

Evaporation induced self assembly (EISA) within microdroplets produced by a vibrating orifice aerosol generator (VOAG) has been used to produce monodisperse mesoporous silica particles. This process exploits the concentration of evaporating droplets to induce the organization of various amphiphilic molecules, effectively partitioning the silica precursor (TEOS) to the hydrophilic regions of the structure. Promotion of silica condensation, followed by removal of the surfactant, provides ordered spherical mesoporous particles. Using the VOAG we have produced highly monodisperse particles in the 5 to 10 μm diameter range. The cationic surfactant CTAB typically leads to hexagonal mesostructure with mean pore size of about 2 nm and specific surface area around 900 m2/g. We have also shown that the pore size in CTABtemplated particles can be increased to 3.8 nm by incorporating trimethylbenzene as a swelling agent. The TMB prefentially locates inside and swells the hydrophobic regions of the surfactant mesostructure. Pore size can also be varied by the choice of amphiphile. Hexagonally ordered particles have been produced using the nonionic surfactant Brij-58 and block copolymer F127. These powders possessed mean pore size 2.8 nm and 6.9 nm, respectively. The uptake of alkyl pyridinium chloride molecules have recently been measured, revealing an uptake capacity that is explained by surface adsorption (as opposed to simple pore infiltration). Kinetics of the uptake process are still be analyzed.

A detailed understanding of phase transformation mechanisms in small particles is critical for the design of small metal particle catalysts and nanophase materials. For palladium supported on silica, it is known that under certain conditions, reduction of PdO can lead to the formation of small Pd metal particles containing central faceted voids[l]. In-situ microscopy studies show that the voids form when PdO is reduced at temperatures between 200 and 400°C but collapse when the temperature reaches about 550°C [2]. Here we present in-situ high resolution microscopy showing the details of the void formation process.

Model Pd/SiO2, catalysts were prepared by dispersing Pd metal over silica microspheres using techniques described elsewheretl]. The Pd/SiO2, catalyst were fully oxidize°C for 2 hours. The catalyst was crushed and dispersed over Mo grids and plasma cleaned to remove any residual carbonaceous material. A Philips CM 200 with a slow scan CCD camera and video recording system was used to recording HREM images during in situ reduction at 200°C.

We have used the technique of in situ electron microscopy to study the oxidation and reduction of the palladium (Pd) catalysts. In this study, we have subjected a Pd catalyst to oxidation and reduction cycles and studied the changes in particle structure and morphology with in situ electron diffraction and imaging. The PdO particles can be reduced to Pd metal in situ at temperatures as low as 200°C in an atmosphere of a few Torr of both H2 and O2. We also found that essentially the same reduction occurred in the vacuums of 10−6 to 10−7 Torr in two different electron microscopes. Our in situ reduction studies show that many of the oxide particles form voids when reduced to Pd metal. The decrease in volume that occurs during reduction is often accommodated by a combination of particle shrinkage and void formation. The production of voids does not seem to depend on either the reducing atmosphere or the rate of reduction, although the voids appear to be unstable above 500°C.

The stability of BN thin film coatings (2–5 nm thick) on MgO and TiO2 substrates was investigated using transmission electron microscopy (TEM). The samples were heated in air for at least 16 hours at temperatures ranging from 773 K - 1273 K. On MgO supports, the BN thin film coating was lost by 1073 K due to a solid state reaction with the substrate leading to formation of Mg2B2O5. No such reaction occurred with the TiO2 substrate and the BN was stable even at 1273 K. However, the coating appeared to ball up and phase segregate into islands of near-graphitic BN and clumps of TiO2 (rutile). The oxidizing treatment appears to promote the transformation from turbostratic BN to graphitic BN.

When BN is synthesized via polymeric precursors and applied to ceramic substrates, tough adherent coatings of hexagonal-BN (h-BN) are obtained after annealing at 1200°C in N2. The study of these coatings is facilitated by using nonporous oxide powders containing single crystal particles of submicron size (e.g. cubes of MgO) as model ceramic substrates. These oxide powders permit high resolution TEM examination of the BN coatings with no further sample preparation. In this study, samples of BN/MgO cubes containing 50 wt% BN were heated in air at elevated temperatures for 16 hours to study the oxidation resistance of BN coatings. The BN coating was found to be stable at 600°C, but the 700°C-treated sample showed evidence for partial amorphization of the coating and reaction with MgO. A significant fraction of the MgO in the 800°C-treated sample had transformed to Mg2B2O5. The reaction of MgO with the BN coating under oxidizing ambients leads to loss of the cubic morphology in the precursor powder.

Poly(borazinylamines) have been processed in THF and toluene solvents and aerogel forms produced by supercritical drying methods. The polymer aerogels have been pyrolyzed and porous amorphous and crystalline BN monoliths have been produced. These have been characterized and effects of processing factors on surface area and pore structure are described.

Ceramic coatings on oxides can be studied by high resolution transmission electron microscopy (HRTEM), with minimal sample preparation, if the substrate consists of nonporous particles of simple geometric shape. Interfaces suitable for ‘end-on’ examination by HRTEM can be readily obtained without any necessity for ion-beam thinning. All the interface orientations that are thermodynamically stable are available for examination from a single sample. This technique is of general applicability and can be used for studies of metal-ceramic and ceramic-ceramic interfaces. We have examined the nature of boron nitride interfaces with oxides such as MgO, TiO2 and Al2O3 and find that BN appears to wet the oxide surface and form tough, adherent coatings. The hexagonal crystalline BN grows with the (0001) planes always being locally parallel to the oxide surface in every instance.

A procedure for preparing cross-section samples of specific regions on an integrated circuit die is outlined. The procedure involves the use of glass slabs to prepare the sample stack used for cross-sectioning. The transparent glass slabs on the top surface of the die enable monitoring of the region of interest during the thinning process. This ensures that the region of interest will be located in the electron-transparent section used for TEM imaging. Samples prepared using this technique are sturdy enough to survive successive thinning and observation in a TEM. The glass slides do not give any charging problems and generally thin much slower than the silicon so as to provide added stability to the sample. An additional advantage is that the sample thickness can be accurately measured using the vernier attachment on an optical microscope during the sample thinning process. The imaging of specific NMOS transistors fabricated on buried oxide layers is illustrated.

High fluence ion implantation of N (1x1018/cm2 at 150 keV) has been used to form buried nitride layers in (110) silicon. After annealing at 1200 C for 5 hrs. a continuous, polycrystalline alpha-Si,N- layer (200 nm thick) is observed beneath a surface silicon film 306 nm thick. The upper Si/Si3N4 interface appears to be more abrupt than that observed in (100) silicon with minimal dendritic intergrowth and no evidence for microtwinning in the silicon. Furthermore, a band of nitride precipitates can be detected 500 nm below the continuous nitride layer. These nitride precipitates grow semi-coherently within the silicon with no observable strain or misfit dislocations within the silicon. The nitride precipitates are internally faulted to accomodate the 10% lattice mismatch in the (0001) nitride direction. Short term anneals reveal that the precipitates have fully crystallized within 10 min. at 1200 C while the continuous nitride layer is still amorphous.