Optical Metrology Adapts for Nanoscale Measurements

As we venture into the nanoscale region, it becomes increasingly necessary to characterize not just new materials, but some of the more familiar ones, because at those dimensions, their characteristics often alter.

Alain Diebold, Empire Innovation Professor of Nanoscale Science at the College of Nanoscale Science and Engineering (Albany, N.Y.), is developing a research laboratory that will investigate future nanoscale metrology needs. He is looking at various optical technologies, both old and new — some in their linear mode and others in a non-linear fashion. “Ellipsometry is an example of a linear optical measurement,” he said. “You send in a certain wavelength, and then look at it coming out.”

Ellipsometry examines the polarization change in light as it reflects off a surface or grating structure. “Interpreting this change provides data about film thickness or, in ellipsometry's case, for purposes of scatterometry; you learn about the line shape and structure of the grating structure CD,” Diebold said. These measurements, commonly used for semiconductors, are also important for nanotech.

Nanoscale's impact on linear optical metrology is well understood. With very thin, single-crystal silicon films, optical properties change because of quantum confinement; there are ways to use this to one's advantage.¹ The same holds true for semiconductor materials; it is only necessary to work out the procedure for the particular single-crystal thin film. What is true for single-crystal, thin semiconductor films also works for scatterometry structures of those nanodimensions.

Although measuring thin metal films is difficult, there may be an easy solution. “They've looked at very thin metal films and understood how to overcome some of the hurdles associated with measuring very thin, polycrystalline metal films, as well as defining the impact of grain size on the dielectric properties of the thin metal film so that it can be better measured optically,” Diebold said.

These techniques could be transferred and easily adapted to semiconductor purposes, and will be an important part of the research conducted in the new lab. Professor Rob Collins of the University of Toledo (Toledo, Ohio) has looked at various materials and determined how to incorporate the impact of grain size on the optical properties of thin films, obtaining good agreement between his approach and the experimental data.

In the non-linear arena, the same light energy — wavelength — goes in and a different wavelength comes out. So why look beyond linear measurements? A principal reason is that interfaces are always challenging to characterize, and second harmonic generation (SHG) is very good at this. SHG does not look at the same frequency of the light that goes in, but twice the frequency — the second harmonic.

Although the SHG signal has considerably less intensity than the one that went in, under certain experimental conditions and polarizations of the incoming and outgoing light, great sensitivity to the interface is obtained. Compared with linear optical methods, the added sensitivity provided by the optically non-linear SHG is tremendous.

A classic example of SHG applications is the SiO₂/silicon interface, which is well characterized. Professor Michael Downer's group at the University of Texas at Austin has extended this to high-k materials on silicon. The same concept is applicable to all kinds of nanoscale samples.

Techniques that have worked well in other disciplines are being investigated to help characterize new and standard materials for nanoscale applications, as well as to perform difficult measurements. (Source: College of Nanoscale Science and Engineering)

An area of study for the new lab will be nanoscale structures. An example of how SHG measurements can be used in this area is silicon nanodots embedded in SiO₂. Nanodots have a tremendous amount of surface area, which makes for a strong second harmonic signal from them, which has been used to characterize many of their optical properties. Experimentally, nanodots have been applied to future memory device concepts; commercially, they are being used to produce efficient and low-power solid-state lighting.

The research's aim will be the investigation of different nanoscale effects in materials, using both linear and non-linear optical methods to determine how these can be used to not only understand these new materials' properties, but also to determine how those properties can be used to do uniformity measurements of these materials' properties, uniformity of structural features and, eventually, of whatever nanoscale measurements are required to enable manufacturing.