Welcome to the ASU Thermal Emission Spectroscopy Laboratory Spectral Library. This library consists of thermal infrared emission spectra (typically 2000 – 220 cm-1) of a variety of geologic materials. Each spectrum is accompanied by descriptive information, including physical and compositional information, sample quality, and a comments field to describe special circumstances and/or unique conditions. Descriptions of the processes of sample preparation, spectral analysis, and quality determination are provided below.
We have intentionally included a variety of materials and sources of spectra for easy access and wide distribution, while retaining the ability to fully document each individual sample. The emphasis in developing this library has been to obtain spectra of samples for comparison to spectra obtained from planetary and Earth-orbiting spacecraft, airborne instruments, and laboratory measurements. The contents of the library have been driven by the research interests of personnel in the ASU laboratory, as well as by an effort to obtain samples of important minerals underrepresented in those research studies. It is important to note that this library is intended to be dynamic and new samples will be added as they are acquired and processed.
Most samples were prepared through a process of crushing, sieving, and washing, with the final pure sample grains selected by hand. Some samples (dominantly carbonates, phosphates, and sulfates) were only analyzed in hand sample form. Crushed samples are generally 710-1000 microns in size, unless otherwise noted in the database. This size fraction was used because volume scattering effects that are introduced with small particle sizes are minimized, it simulates the roughness of sandy and rocky surfaces, and grains are ideally randomly oriented to prevent preferential viewing of any crystal axis. The purity and composition of each sample was determined by at least one of the following: electron microprobe, x-ray diffraction, bulk chemical analysis, and in rare cases, by spectral comparison to known samples in the literature (these are noted in the “Misc. Comments” field).
Spectral Data Acquisition
Emission spectra were acquired using a Nicolet Nexus 670 interferometric spectrometer equipped with a CsI beamsplitter and an uncooled deuterated triglycine sulfate (DTGS) detector; the spectral range of the instrument is from 2000 — 220 cm-1 (5 — ~45 microns). Both the spectrometer and the sample chamber/glovebox were continuously purged with nitrogen gas during sample analysis to minimize atmospheric H2O and CO2 which also have absorption features in the 2000-220 cm-1 region of the spectrum. The particulate samples were heated in an oven to 80°C to improve the signal to noise ratio during spectral analysis (this temperature is maintained during analysis by placement of the sample cup on a heater element). The samples were raised into a water-cooled sample chamber that closely approximates a blackbody cavity [Ruff et al., 1997]. A total of 270 scans at 2-cm-1 sampling were taken over ~7 minutes and averaged together by the spectrometer. In the case of a hand sample, active heating during measurement is not possible. Hand samples were taken directly from the oven and placed into the sample chamber and 180 scans were taken over a period of ~5 minutes to minimize the effects of sample cooling. The spectral calibration method is a variation of method 1 of Christensen and Harrison  as described in detail by Ruff et al., .
Data and Documentation
Attached to each spectrum are database fields that describe the sample. In some cases, the mineral has been renamed as a result of chemical analysis but the original mineral name, source and sample number have been retained in separate database fields. A sample quality field has been included to provide a quick assessment of sample purity. This field contains four degrees of sample purity: 1) the mineral and spectra are of high purity and quality; 2) there are minor impurities, but overall quality is still good; 3) the sample contains significant impurities which may significantly influence its spectrum (this designation usually results in removal from the library); 4) sample has no compositional information yet, but appears to be of good quality on the basis of spectral comparison to a sample of known purity (i.e., use at your own risk).