Electron spin resonance probing of point defects in Si/high-k structuresand functional oxides
Author:
Abstract:
Today, we want our devices based on semiconductor devices to be ever smaller, faster, more multi-functional, and, if possible, also cheaper. For years now industry has been able to meet these demands bringing e.g., tiny cell phones, MP3-players, photo and video cameras, and laptops on the market. The word 'nano' has even surpassed its scientific meaning (unit prefix) and became the magic word to sell electronic devices and even washing-powder. Up to present, the success of semiconductor-based industry has been the scalability of the microelectronics. More powerful chips could be realized through scaling of the metal-oxide-semiconductor (MOS) device dimensions. Using the Si/SiO2 entity as basic building block, this down-scaling was successfully accomplished for gate thicknesses down to the 1-nm range. At this very moment, however, semiconductor industry has reached the end of the road for pure Si/SiO2 based technology. Further down-scaling of the SiO2 gate oxide layer would induce excessive leakage currents because of direct tunneling, and moreover, serious reliability problems. This means that to keep on track with the requested advances in semiconductor devices an alternative "new" gate insulating material will be required for future generations of MOS devices. As an obvious solution, an insulator with a higher value of dielectric constant (k) would allow one to use a physically thicker oxide layer retaining the same gate capacitance. The use of such an alternative insulator instead of the extensively studied SiO2 oxide brings about unforeseen problems associated with, e.g., interlayer formation and dopant penetration. Many candidate materials have been studied, leaving us with more questions than answers. In search for some fundamental answers three of these so-called high-k materials, ZrO2, HfO2, and LaAlO3, were studied in this thesis and are the subject of chapters 3 and 4. Another outcome of the current nano-wave is the rising popularity of, e.g., nanoparticles, nanowires, and nanotubes. Scaling down the SiO2 dielectric to such small dimensions, however, changes its electrical and optical properties, opening a whole new area of research. The characterization of the structure of SiO2 nanoparticles is the subject of chapter 2. The last few years, applied and fundamental research of nano-structures has grown exponentially revealing the upcoming relative impact of defects in the core of the nano-materials (nanoparticles, nanotubes, ultrathin layers) as well as at the interface. The study of point defects subsequently became of general interest since the presence of one single point defect can play a crucial definite part in the optical and electronic properties, such as, e.g., photoluminescence, doping of semiconductor nanoparticles, and conduction (in e.g., carbon nanotubes), of the nano-structures. To gain a better understanding of the detrimental influence of the presence of point defects atomic identification of occurring point defects is of vital importance. Up to the present the only known technique able to reveal the required atomic-scale information is electron spin resonance. Hence, in this thesis we will use electron spin resonance to monitor, characterize, and hopefully identify occurring point defects in a range of materials of interest for future developments in semiconductor nano-technology. The outline of this thesis is as follows: Chapter 1 starts with a summary of some elementary notions of the theory and practice of electron spin resonance experiments. Further a brief overview of the characteristics of some occurring defects is Si/SiO2 and in Si/high-k structures is given. In chapter 2 an extensive electron spin resonance study is presented of fumed silica nanoparticles. Monitoring occurring point defects as a function of post-formation heating and treatment revealed interesting atomic-scale information concerning the particles' network structure. The third chapter reports on the observation of P-impurity related point defects in nm-thick P-implanted HfO2 films on (100)Si and in ZrO2 powder -two oxides prominent in current high-k insulator research. It is shown that the incorporation of P in these high-k oxides results in ESR-active defects possibly acting as hole traps. This finding is important in view of the observed enhanced dopant penetration through HfO2 layers during the necessary dopant activation anneals. The study of the nature and stability of the (100)Si/LaAlO3 interface is the subject of chapter 4. Here it is demonstrated that the interface is abrupt and stable for annealing up to about 800 °C. It is evidenced that upon annealing in the range 800-860 °C a Si/SiO2-type interlayer starts forming. Upon annealing at temperatures higher then 930 °C the interlayer with SiOx nature is found to break up. The latter is possibly related crystallization and possibly silicate formation. In the last chapter the influence of ion implantation in amorphous bulk SiO2 was studied. In cooperation with R. Weeks (Vanderbilt University, USA), R. Magruder (Belmont University, USA), and R. Weller (Vanderbilt University, USA) the densities of observed defects were compared to the optical absorption bands around 4.8 and 5.3 eV. In this manner information could be obtained concerning the source of the optical absorptions. This thesis ends with a summary and general conclusions of the experimental work performed.