Transmission Electron Microscopy (TEM) is a reliable tool for chemical and structural studies of nanostructured systems. The shape, size and volumes of nanoparticles on surfaces play an important role in surface chemistry. As nanostructured surfaces become increasingly important for catalysis, protective coatings, optical properties, detection of specific molecules, and many other applications, different techniques of TEM can be used to characterize the properties of nanoparticles on surfaces to provide a path for predictability and control of these systems. This dissertation aims to provide fundamental understanding of the surface chemistry of Electroless Metallization onto Polymeric Surfaces (EMPS) through characterization with TEM. The research focuses on a single EMPS system: deposition of Ag onto the cross-linked epoxide "SU8", where Au nanoparticles act as nucleation sites for the growth of Ag nanoparticles on the polymer surface. TEM cross sections were analyzed to investigate the morphology of the Au nanoparticles and to determine the thicknesses of the Ag nanoparticles and of the Ag layers. A method for the direct measurement of the volume and thickness of nanomaterials has been developed in the project using High-Angle Annular Dark-Field (HAADF) Scanning Transmission Electron Microscopy (STEM). The morphology of Au and Ag NPs has been studied to provide reliable statistics for 3-D characterization. Deposition rates have been obtained as a function of metallization conditions by measuring the composition and thickness of the metal for EMPS. In the present work a calibration method was used to quantify the sensitivity of the HAADF detector. For thin samples a linear relationship of the HAADF signal with the thickness of a material is found. Cross-sections of multilayered samples provided by Triquint Semiconductors, FL, were analyzed as calibration standards with known composition in a TECNAI F30 transmission electron microscope to study the dependence of the HAADF detector signal on sample thickness and temperature. Dynamical diffraction processes play an important role in electron scattering for larger sample thicknesses. The HAADF detector intensity is not linearly dependent on sample thicknesses for thick samples. This phenomenon involves several excitation processes including Thermal Diffuse Scattering (TDS) which depends on temperature-dependent absorption coefficients. Multislice simulations have been carried out by Python programming using the scattering parameters (2) available in the literature. These simulations were compared with experimental results. Wedge-shaped Focused Ion Beam (FIB) samples were prepared for quantitative HAADF-STEM intensity measurements for several samples and compared with these simulations. The discrepancies between the simulated and experimental results were explained and new sets of absorptive parameters were calculated which correctly account for the HAADF-STEM contrasts. A database of several pure elements is compiled to illustrate the absorption coefficients and fractions of scattered electrons per nanometer of the sample. In addition, the wedge-shaped FIB samples were used for studying the HAADF-STEM contrasts at an interface of a high- and a low-density material. The use of thick samples reveals an increased signal at the interfaces of high- and low-density materials. This effect can be explained by the transfer of scattered electrons from the high density material across the interface into the less-absorbing low-density material. A ballistic scattering model is proposed here for the HAADF-STEM contrasts at interfaces of thick materials using Python. The simulated HAADF-STEM signal is compared with experimental data to showcase the above phenomenon. A detailed understanding of the atomic number contrast in thick samples is developed based on the combination of experimental quantitative HAADF-STEM and simulated scattering. This approach is used to describe the observed features for Ag deposition on SU8 polymers.

The sizes, shapes, volumes and compositions of nanoparticles are very important parameters determining many of their properties. Efforts to measure these parameters for individual nanoparticles and to obtain reliable statistics for a large number of nanoparticles require a fast and reliable method for 3-D characterization. In this dissertation, a direct measurement method for thicknesses, volumes or compositions of nanomaterials by quantitative atomic number contrast in High-Angle Annular Dark-Field (HAADF) Scanning Transmission Electron Microscopy (STEM) is presented. A HAADF detector collects electrons scattered incoherently to high angles. The HAADF signal intensity is in first-order approximation proportional to the sample thickness and increases with atomic number. However, for larger sample thicknesses this approach fails. A simple description for the thickness dependence of the HAADFSTEM contrast has been developed in this dissertation. A new method for the calibration of the sensitivity of the HAADF detector for a FEI F30 transmission electron microscope (TEM) is developed in this dissertation. A nearly linear relationship of the HAADF signal with the electron current is confirmed. Cross sections of multilayered samples provided by TriQuint Semiconductors in Apopka, FL, for contrast calibration were obtained by focused ion-beam (FIB) preparation yielding data on the interaction cross section per atom. iv To obtain an absolute intensity calibration of the HAADF-STEM intensity, Convergent Beam Electron Diffraction (CBED) was performed on Si single crystals. However, for samples prepared by the focused ion beam technique, CBED often significantly underestimates the sample thickness. Multislice simulations from Dr. Kirkland’s C codes are used for comparison with experimental results. TEM offers high lateral resolution, but contains little or no information on the thickness of samples. Thickness maps in energy-filtered TEM (EFTEM), CBED and tilt series are so far the only methods to determine thicknesses of particles in TEM. In this work I have introduced the use of wedge-shaped multilayer samples prepared by FIB for the calibration of HAADF-STEM contrasts. This method yields quantitative contrast data as a function of sample thickness. A database with several pure elements and compounds has been compiled, containing experimental data on the fraction of electrons scattered onto the HAADF detector for each nanometer of sample thickness. The use of thick samples reveals an increased signal at the interfaces of high- and low-density materials. This effect can be explained by the transfer of scattered electrons from the high density material across the interface into the less-absorbing low-density material. The calibrations were used to determine concentration gradients in nanoscale Fe-Pt multilayers as well as thicknesses and volumes of individual Au-Fe, Pt, and Ag nanoparticles. Volumes of nanoparticles with known composition can be determined with accuracy better than 15%. Porosity determination of materials becomes available with this method as shown in an example of porous Silicon.