Advanced semiconductor detection devices incorporate surface texturing to reduce reflection of the incident radiation, and thus, enhance optical absorption through scattering. Using micromachining techniques, three different silicon surfaces were fabricated, optically characterized, and analyzed in terms of their ability to scatter incident optical energy. The fabricated surfaces consist of: randomly sized and spaced pyramids (RSSPs), deep vertical- wall grooves (DVWGs), and porous silicon. The DVWG structures consist of interdigitated, 270 micrometers deep, 25 micrometers wide, and 1000 micrometers long grooves separated by 5 micrometers wide walls. The RSSP textured surfaces consist of pyramids with random 0.5-12.0 micrometers square base widths and heights, but otherwise consistent shape and symmetry. The pyramid walls make an angle of 54.74 degrees with respect to the sample surface. Porous silicon samples consist of surfaces with etched random pores that are 0.2-5 micrometers in depth, 1-5 micrometers in length, and 0.1-5 micrometers in width. Utilizing a laser scatterometry, the bidirectional reflectance distribution function (BRDF) of silicon textured surfaces has been measured at commercially available laser wavelengths of 1.06 and 10.6 micrometers . A highly- polished, single-crystal silicon wafer was used as a reference surface. The three micromachined surfaces showed an enhanced scatter at 1.06 micrometers as demonstrated by a reduced specular peak and increased average BRDF. The RSSP textured surface also demonstrated a low BRDF at 10.6 micrometers incident laser wavelength.

A large variation in bidirectional reflectance distribution function (BRDF) for certain GaAs surfaces is proposed as the basis for a variable BRDF reference material. Such a material would provide a range of reference values as opposed to the discrete values of diffuse reference materials. The effect is reasonably stable and uniform, and is most likely the result of a weak grating effect caused by surface striations.

Total integrated scatter measurements, the earliest well defined measurement relating light scatter to surface roughness, is now being used in modern production facilities as a means of monitoring product surface roughness. The two applications reviewed here are computer disks, where the issue is a well defined roughness (as opposed to the smoothest possible surface) and an emerging issue with roughness specifications for the backsides of silicon wafers. The paper describes a scanning instrument that allows sample uniformity to be revealed and thus the manufacturing process investigated.

Infrared spectra of two 'black' baffle coatings, which were made both with and without large SiC scattering centers, demonstrate the strong attenuation of specular reflectance caused by scattering. This attenuation is described quantitatively by a reflecting layer model for rough, thick absorbing coatings. Infrared spectra and BRDFs of a very rough aluminum surface, which is a nearly perfect diffuse reflector, demonstrate how the surface scatter changes with wavelength from a region where specular reflection dominates to the region where scatter dominates.

The literature now includes a number of examples where light scatter (BRDF) measurements have been used to determine the surface power spectral density function of smooth, clean reflectors. But most of this data is for front surface metal mirrors and semiconductors. Black glass has been considered for use as a BRDF standard, and there are industry applications (computer disks and front panel displays) that could benefit from the same type of characterization from glass and ceramic surfaces. This paper addresses some of the issues involved with making surface roughness measurements on these surfaces. For example, clear glass will scatter a visible beam from its bulk and second surface as well as its front surface. In addition, calculation of the polarization constant Q must be handled in a more accurate manner. Data from several samples will be analyzed.

One of the most interesting phenomena associated with the scattering of light from a randomly rough surface is that of enhanced backscattering. This is the presence of a well-defined peak in the retroreflection direction in the angular distribution of the intensity of the incoherent component of the light scattered from such a surface. It is due primarily to the coherent interference of each multiply reflected optical path with its time-reversed parter. The enhanced backscattering of light from a randomly rough surface was predicted in the scattering of P- polarized light from a 1D random metal surface, when the plane of incidence was perpendicular to the generators of the surface. The calculation employed lower order perturbation theory and was limited to weakly corrugated surfaces. To test the theory, recent experimental investigations of scattering from very shallow randomly rough characterized Gold surfaces shows the enhanced backscattering phenomenon. The far-field measurements and characterization of the surfaces are presented and compared with analytical results. It is believed that the mechanism responsible for the enhanced backscattering phenomenon is due to the large slope of the very smooth metallic surface.

Recent progress in the machining of optical surfaces promises to significantly reduce the time and cost of manufacturing optical elements. Specific reference is made to a new kind of machining process called deterministic microgrinding. Optical surfaces made by machining processes like single-point diamond turning, or deterministic microgrinding exhibit residual cutting tool marks that result in scattering effects which can significantly degrade optical performance. However, for some infrared applications, post-polishing may not be required and thus resulting in substantial cost savings. In this paper surface scattering theory has been implemented to model the image degradation effects of residual surface irregularities for optical surfaces exhibiting: i) azimuthal tool marks (diamond turning), ii) radial tool marks (deterministic microgrinding) and, iii) random roughness caused by conventional grinding and polishing. Intercomparison of these three processes provides new insight into the scattering behavior and fabrication tolerances for these very different manufacturing processes.

The impact of surface characterization on various aspects of microelectronic device performance and manufacture is briefly reviewed. Optical scattering measurements have been carried out for the surface characterization of AlN, MgO, Al₂O₃ substrates, single crystal silicon wafers and physcial vapor deposited films. Scattering light was recorded up to 80 degree from the specular direction. The angle-resolved scattered light sensitively reveals the characteristics of surface in the spatial frequency domain. Quantitative parameters of surface roughness have been derived from power spectral density distributions and diffraction theory for highly polished smooth surfaces. Instrumentation for full surface characterization has been coupled with computer operation. It has the features of high-speed and noncontact and is inherently area averaging. It lends itself particularly well to online product of high- speed and inprocess monitoring of a manufacturing process. Delicate surfaces can be rapidly 'finger printed'.

There are many areas of science and technology where the scattering of electromagnetic waves by clusters or merging particles are of interest. The merging particles under study might be inclusions in high-density composites, liquid drops, biological cells, macroscopic ceramic particles, etc. As intersecting particles are bounded by a complex physical surface, the problem of scattering from these particles valid for any degree of merging, including touching, and for arbitrary materials of the constituents, has received limited attention. Here we present solutions which are valid and exact in the long wavelength limit compared with the size of intersecting spherical particles and cardioidal particles of similar dimensions. Both shapes are almost coincident everywhere except in the region of intersection. We treat the case when the waves are polarized along the common axis (longitudinal field). The solutions of Laplace's equation are integrals (spheres) or sums (cardioids) over continuous or discrete eigenvalue spectra respectively. The spectral dependencies of the resulting extinction coefficients and the scattering for the spherical and cardioidal particles are quite distinct. There is an enormous difference in the magnitude of absorption responses. Overall the cardioidal particle behaves as if it is almost invisible in terms of effects on the external field for a very broad band of optical frequencies. THe latter result was checked for a number of dielectric permittivities and seems to be universal. It scatters far more weakly than the isolated sphere. In constrast the intersecting sphere has an extinction band which is broad and is much enhanced at longer wavelegnths relative to the simple sphere. This result has significant implications for the design of surfaces with minimum scattering.

In the smooth-surface limit, the angular distribution of the intensity of light reflected and scattered from a rough surface depends only on the root-mean-square value of its surface roughness and is independent of the details of the distribution of the surface height fluctuations. In the case of rougher surfaces, the effects of the shape of the height distribution function which do appear are usually evaluated using an assumed Gaussian height distribution, even though real surfaces can be, and frequently are, non-Gaussian. This paper describes the results of an analytic study of the effects of non-Gaussian height distributions on the reflection and scattering properties of moderately-rough surfaces. As an example we compare predictions based on a Gaussian distribution with those of a one-sided exponential distribution. Large differences are found in and near the specular core, which eventually disappear in the scattering tail. Suggestions for follow-on experimental and analytic studies are given.

Measurement of small particle contamination, via laser particle scanners, is an important tool for maximizing throughput in semiconductor production lines. As line widths shrink, the required minimum observable particle diameter drops in proportion. Unfortunately, the scatter signal from a small particle on a surface falls off very quickly with decreased diameter, even faster than would be predicted by a simple Rayleigh scatter model of an isolated particle. This paper reviews a technique to measure the differential scattering cross-section of very small surface bound particles. The objective is to provide a means of obtaining data for producing the next generation of particle scanners.

As Lawrence Livermore National Laboratory moves forward with the design of the National Ignition Facility (NIF) in the Inertial Confinement Fusion (ICF) program, issues relating to the detection and measurement of laser-induced damage on large optics must be addressed. Currently, microscopy is used to evaluate surface quality and measure damage thresholds on small witness samples. In order to evaluate large areas, an automated system was constructed which can scan optics with dimensions as large as 1 meter and weighing as much as 400 pounds. The use of microscopy as the main test diagnostic has been replaced with an optical scatter detection system. Now large areas can be rastered, and maps can be generated, reflecting inherent and laser-induced scatter in multilayer optical coatings and bulk materials. The integrated scattered light from a test piece is measured in transmission using a HeNe laser as the probe source. When the probe beam is overlapped on a pulsed, high power, Nd:YAG laser beam, damage related scatter may be measured. This technique has been used for: 1) mapping of inherent scatter in an optic, 2) on-the-fly damage detection during a high fluence raster scan of an optic, and 3) single site damage evaluation for the determination of a laser damage threshold. Damage thresholds measured with the scatter diagnostic compare within measurement error to those attained using 1.00 x microscopy.

Single crystals of KH2PO4 (KDP) and (DxHlx)2PO4 (DKDP) will be used for frequency conversion and as part of a large aperture optical switch in the proposed National Ignition Facility (NW) at the Lawrence Livermore National Laboratory (LLNL). These crystals must have good optical properties and high laser damage thresholds. Currently these crystals have a lower laser damage threshold than other optical materials in the laser chain which has forced designers to limit the output fluence of the NIF in order to avoid damaging the crystals. Furthermore, while more efficient frequency conversion schemes are being explored both theoretically and experimentally, the advantages of these schemes can not be fully realized unless the damage thresholds of the conversion crystals are increased. Over the past decade, LLNL has generated an extensive data base on the laser damage in KDP and DKDP crystals both at the first and third harmonics of Nd-YAG.1 While the damage thresholds of these crystals have increased over this time period due, in part, to better filtration of the growth solution,2 the damage thresholds of the best crystals are still far below what is expected from theoretical limits calculated from the band structure of perfect crystals. Thus damage in KDP and DKDP is caused by defects in the crystals. We also rely on a process called laser conditioning to improve the damage thresholds of the crystals. Unfortunately, little is understood about the mechanism of laser induced damage, the conditioning process in the crystals, or the defects which are responsible for damage. We have recently implemented a scatter diagnostic for locating and studying defects in crystals and as a tool for studying the mechanism of laser damage and laser conditioning.

An angular scatterometer (goniometer) was designed and built which is capable of measuring the S₁₁ and S₁₂ components of the scattering matrix on single particles as a function of the scattering angle (Theta) for different values of the azimuthal angle (phi) . The particle is electrostatically charged and is levitated in an electrodynamic levitator (EDL). The EDL is positioned in such a way that the particle is in the center of a semicircular detector ring which can be rotated around the particle so that different scattering planes can be measured. Spherically shaped porous carbon particles (spherocarbs) serve as candidate particles for experimental and theoretical investigations of scattering in various planes. Results are compared to theoretical modeling of stochatistic rough particles in order to assess the influence of shape and surface roughness.

A variety of important optical properties can be determined from spectroscopic analysis of diffuse reflectance of surfaces. The design of a small user friendly, lightweight, field hardened, computer controlled device for performing infrared spectroscopic analysis of trace contaminants on surfaces is described. The device employs a miniature Fourier transform infrared spectrometer with very efficient diffuse reflectance optics and a portable computer to provide reflectance spectra of surfaces measured relative to some idealized surface. These spectra yield qualitative and quantitative chemical information from a host of surfaces that has imminently practical applications in the determination of surface identification, contamination, and degradation.

This paper describes the construction of two variants of a portable glare measurement system and the evaluation of their ability to determine the scattering effect that canopies have upon laser sources. The emphasis of this work was upon the development and demonstration of a portable glare measurement system based around a CCD camera system with high dynamic range. Two fundamental questions needed to be answered. Firstly, did the CCD have the required dynamic range and spatial resolution and, secondly, could it be made into a fully portable system. Demonstration portable glare measuring devices have been constructed using a high dynamic range Wright Instruments AT1 electronic camera and the results produced by these devices are compared to those taken from a large, laboratory-bound glare measurement system. It was found that, although performing to its specification, the camera alone did not have the required sensitivity to clearly distinguish the scatter function from the background noise while not being saturated by the central part of the scatter profile. However by also using neutral density filters to effectively enhance dynamic range, the camera was adequate for a scanning-type protable glare measurement system which could by fully portable.

Near or even beyond the Rayleigh-limit it is not possible to compute the surface PSD analytically. Since stray light measurements are an affordable approach for quality control in industry it has to be investigated, in which case they can be applied. Therefore a statistical method is proposed to determine, whether there is a significant correlation between roughness and scattering, or not. The method is proposed to determine, whether there is a significant correlation between roughness and scattering, or not. The method is tested with a newly developed rugged stray light sensor, several samples, and comparative measurements with an optical profiler.

Infrared laser light scattering is a powerful tool for investigation of inhomogeneities in the bulk semiconductor materials. For sample illumination the diode pumped Nd:YAG laser emitting monomode 1.06 micrometers beam is used. The laser beam waist inside the semiconductor samples does not excess 50 micrometers . The scattered centers inside the sample are observed perpendicularly to the direction of illuminating beam using microscope with infrared CCD camera. To obtain 2D image of scanned plane the sample is moved horizontally by the scanning stage driven by computer. Controlled changing of scanning plane enables the investigation of the sample in the third direction. The scanning and scattered image processing are controlled by computer. The device is tested on GaAs wafers.

The BRDF of a MgF₂ protected Al mirror contaminated with dust particulates due to exposure to the laboratory environment has been measured and is presented for wavelengths of 633 nm, 325 nm, 121.6 nm, and 74 nm. This experimental data is compared with theoretical predictions arising from the OPALS modeling software. This model calculates the BRDF based on the measured particlate distribution found on the surface, and the optical constants of the contaminant. The OPALS software shows promise as a useful tool in the design phase of optical instruments: for drawing up contamination budgets and for incorporation into stray light analysis predicting instrument performance.