SEMATECH Identifies the Tool to Do the Job

Even analytical tools are pushed to the limit in advanced processing.

Laura Peters, Senior Editor -- Semiconductor International, 1/1/1998

SEMATECH Identifies the Tool to Do the Job

The 1997 version of the National Technology Roadmap for Semiconductors highlights specific areas of concern regarding metrology. As the industry migrates from off-line to in-line and in situ sensor-based process control, it is challenged to develop robust sensors, process controllers and data management tools; improve the detection of impurities (particles, oxygen and metallics) at levels of interest for starting materials; measure frequency-dependent dielectric constant of low-k materials; control new processes such as damascene and copper metalization; and develop reference materials and standards for characterizing gate dielectrics, thin films, thin film interfaces and other process needs.

Potential solutions for crosscutting metrology needs for next-generation devices, including the analysis of particulate, metallic and organic contamination, off-line dopant characterization and centralized analytical facilities for research and development, are summarized in the table.

Most of the existing analytical tools for identifying organic surface contamination are capable of qualitative (and in most cases quantitative) analysis of the specified control levels of 10¹⁴ to 10¹³ carbon atoms/cm² on silicon. These include ion mobility mass spectrometry; thermal desorption gas chromatography/mass spectrometry; surface secondary ion mass spectrometry (SIMS) and time of flight SIMS (ToF-SIMS); and X-ray photoelectron spectroscopy. Heavy ion backscattering spectrometry (HIBS), used to calibrate reference materials for surface metallic contamination, is an example of a method appropriate for a centralized facility, according to the Roadmap. One area of concern is that the existing nondestructive methods for measuring metallic surface contamination, such as total reflection X-ray fluorescence (TXRF) analysis, have detection limits for Fe, Ni, Cu and Na close to the required control levels specified for 0.18 mm manufacturing.

As statistical process control at the specified detection limits usually requires the measurement variability to be one-tenth the control limit, potential solutions shown in the table reflect each tooly ability to meet the needs for wafer particle analysis, surface contamination analysis and doping characterization. Finally, doping technology and technology computer-aided design (TCAD) process simulation require improved one-, two- and three-dimensional dopant profiling to address transient enhanced diffusion effects and process control for ultrashallow junctions.

Table 1. Potential Solutions for Materials and Contamination Characterization