Most commercially important microelectronic devices (particularly Si MOSFETs) and many experimental devices depend on critical control of the properties of interfaces, thin film layer thicknesses, and critical device dimensions (eg. gate lengths, wall angles, etc.). Many emerging ideas for nanoelectronics require even more aggressive levels of dimensional control (3-D structures with atomic level accuracy).
The related problems of controlling manufacturing processes for device fabrication and of understanding the properties of materials used in devices require high-speed, high-accuracy, nondestructive in situ and ex situ measurements of both wafer and process parameters. My efforts are currently concentrated on the use of spectroscopic ellipsometry and related reflected light techniques to characterize processes and electronic materials. For materials characterization, we also use supporting ex situ characterization methods including scanning electron microscopy, atomic force microscopy, and high-resolution X-ray diffraction.
Most of our current activities are centered on the challenging and industrially important issue of controlling reactive ion etch processes for the fabrication of deep submicron gate MOSFETs. We are using these wafer metrology tools in conjunction with plasma characterization tools including optical emission spectroscopy, UV absorption spectroscopy, and novel broadband RF metrology, to probe the relationships between plasma conditions and the evolution of etched features on the wafer. My students and I have collaborated closely in these efforts with control systems experts including Profs. James Freudenberg, Jessy Grizzle, and Pramod Khargonekar. Our accomplishments include the first published demonstration of multiple-input, multiple-output (MIMO) real-time feedback control of RIE’s, unique high-speed algorithms for continuous time extraction of etch rate and remaining film thicknesses from laser reflectometry data (EKF-R), improved methods for spectral reflectometry measurements of rough films such as polycrystalline Si, and improved methods for the real-time estimation of atomic F concentrations from actinometry data.
A major focus of my groups current efforts on wafer metrology is the promising but extremely challenging problem of measuring patterned wafers. We are currently studying spectroscopic ellipsometry and reflectometry on periodic grating structures using the Rigorous Coupled Wave Analysis (RCWA) method of Moharam and Gaylord. We have shown experimentally that specular mode SE measurements can yield highly accurate measurements of critical dimensions, etch depths, and wall shapes. Our simulations show that this technique will work with good sensivity into the deep submicron regime. We are currently struggling to improve our numerical analysis techniques to make this a more practical method for general use. We are also working on more approximate methods for measurements on less period structures. One of our strong goals is to be able to quantitatively monitor the evolution of very small structures during RIE processes and to then relate variations in this topography to plasma conditions
In my past research efforts, I and my students and coworkers have contributed to improvements in gate dielectric growth and deposition for both Si MOSFETs and for passivating films on GaAs and InP. I was one of the US pioneers in the development of nitrided oxide gate dielectrics for very thin , high reliability Si MOSFETs (including the discovery of the improved resistance of the nitrided oxide interface to ionizing radiation induced damage). It is my eventual goal to use the metrology and control methods we are developing for RIE applications to studying and improving Si MOS insulator systems.
Fred Lewis Terry, Jr.
October 9, 2005