Michael D. Schaeberle

Novel Route to Faster Fourier Transform Infrared Spectroscopic Imaging

Fourier transform infrared (FT-IR) spectroscopic imaging microscopy couples a focal plane array (FPA) detector, integrated within an infrared microscope assembly, to an interferometer for attaining a multiplex/multichannel signal detection advantage. While this configuration should enable the acquisition of spatially resolved spectra over the entire field of view of a sample in the time that it takes a conventional FT-IR spectrometer to record a single spectrum, data acquisition in an imaging modality is an intrinsically slower process. We present a novel collection technique for step-scan, micro-imaging spectrometers that both allows large numbers of samples to be imaged rapidly and provides higher signal-to-noise ratios (SNRs) for given experimental time intervals. For example, data may be collected in as little as one minute, while SNRs greater than 800 are achieved for data acquired in less than 10 min. Imaging data acquired in the proposed, more rapid approach exhibit no loss in fidelity compared to data recorded by the conventional imaging techniques.

Raman Chemical Imaging of Microcrystallinity in Silicon Semiconductor Devices

Silicon integrated circuits are fabricated by the creation of complex layered structures. The complexity of these structures provides many opportunities for impurities, improperly annealed dopants, and stress effects to cause device contamination and failure. Nondestructive metrology techniques that rapidly and noninvasively screen for defects and relate silicon device structure to device performance are of value. We describe the first use of a liquid crystal tunable filter (LCTF) Raman chemical imaging microscope to assess the crystallinity of silicon semiconductor integrated circuits in a rapid and nondestructive manner without the need for sample preparation. The instrument has demonstrated lateral spatial resolving power of better than 250 nm and is equipped with a tunable imaging spectrometer having a spectral bandpass of 7.6 cm−1. The instrument rapidly produces high-definition Raman images where each image pixel contains a high-quality Raman spectrum. When combined with powerful processing strategies, the Raman chemical imaging system has demonstrated spectral resolving power of 0.03 cm−1 in a test silicon semiconductor wafer fabricated by using ion implantation. In addition, we have applied Raman chemical imaging for volumetric Raman imaging by analyzing the surface distribution of polycrystalline thin film structures. The approaches described here for the first time are generally applicable to the nondestructive metrology of silicon and compound semiconductor devices.

Theory and Application of Gain Ranging to Fourier Transform Infrared Spectroscopic Imaging

Gain ranging is incorporated into the data acquisition and processing protocol for a Fourier transform infrared (FT-IR) imaging spectrometer employing a focal plane array (FPA) detector. A model for predicting the signal, noise, and signal-to-noise ratio (SNR) for an FPA in terms of the dynamic range of the analog-to-digital converter (ADC) is presented. Conventional gain ranging theory, modified to account for variation of noise with gain by incorporating a linear model for noise prediction, is shown to provide excellent agreement with observed values for the SNR advantage afforded by gain ranging. SNR improvement, as affected by collection parameters, was shown to be limited by the noise characteristics of the FPA detector. The advantages of gain ranging were demonstrated by spectroscopic imaging of a thin section of human prostate tissue.

Visible spectroscopic imaging studies of normal and ischemic dermal tissue

We describe a non-invasive, in vivo hyperspectral imaging method for visualizing the spatial distribution of dermal tissue oxygenation. Real-time images of the dermis are acquired both at multiple, contiguous wavelengths and at relatively narrow spectral bandwidths to generate a data cube consisting of one spectral and two spatial dimensions. For data collection, the sample area is illuminated by radiation, which is delivered by liquid light guides from a quartz tungsten halogen source. Reflected light from the sample is first passed through a liquid crystal tunable filter and then imaged onto a silicon charged coupled device detector. The subsequently digitized data are presented in terms of spectral images reflecting multivariate least squares analyses based upon linear combinations of oxy- and deoxyhemoglobin reference spectra. The generated gray scale images directly represent the varying spatial distributions of dermal tissue oxygenation. As an example, imaging data are obtained from normal tissue and induced ischemic tissue for which both the venous and arterial blood flow was artificially occluded.

Effect of Focal Plane Array Cold Shield Aperture Size on Fourier Transform Infrared Micro-imaging Spectrometer Performance

Micro-imaging spectrometers incorporating focal plane array (FPA) detection require careful demarcation of cold shield aperture size for both optimal performance and prevention of errors. This study examines the effects of changing the diameter of the cold shield aperture on the intensity and spatial homogeneity of the incident radiation. A uniform polystyrene film was repeatedly imaged by using cold shields of varying aperture sizes. It is shown that a smaller than optimal aperture size leads to image edge clipping, resulting in an inefficient use of the array, lower overall signal, spectral distortions, and higher noise characteristics. Use of an aperture size larger than required causes a decrease in the effective dynamic range of measurements, resulting in higher noise levels. The advantages and necessity of optimizing imaging spectrometer performance by employing a cold shield with an appropriately sized aperture are discussed.