and Identification Using Reflectance Spectroscopy Bogdan Lita Brian Curtiss Gary A Fager Lee Feldman Susan M Parks and Kevin B Tanguay Malvern Panalytical 1625 S Fordham St Longmont CO USA ID: 1025201
Download Presentation The PPT/PDF document "Enabling Ground-based Materials Analysis" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. Enabling Ground-based Materials Analysis and Identification Using Reflectance SpectroscopyBogdan Lita, Brian Curtiss, Gary A. Fager, Lee Feldman, Susan M. Parks, and Kevin B. TanguayMalvern Panalytical, 1625 S. Fordham St., Longmont, CO, USA
2. OverviewTopics covered:Wavelength calibrationField Wavelength calibration verification and adjustmentRadiometric calibrationSystem noise characterizationSpectral Resolution performance and improvements
3. Wavelength CalibrationXe, Ar, Kr, Hg-Ne, Hg-Ar atomic emission lines as primary referencesUsing NIST published air wavelengthsTypical line widths are <0.001 nm so no spectral resolution adjustments necessaryThird order polynomials generated for each spectrometer to convert channel number to wavelengthInitial calibration
4. Wavelength CalibrationFeature positions observed in reflectance wavelength standards shift with changes in spectral resolutionShifts on the order of 1-2 nm for a 6 to 15 nm change in spectral resolution are typicalThus, this type of wavelength standard is unable to achieve sub-nanometer wavelength calibration accuracyIt is suitable as a secondary wavelength standard if tied to emission line wavelengthsReflectance wavelength standards
5. Wavelength CalibrationMulti wavelength reflective calibration standard (rare earth element oxides + polyethylene terephthalate) paired to instrumentSharpest peaks are used for wavelength calibration verification and adjustment – minimum of 5 peaks per spectrometerPeak positions measured and recorded immediately following primary wavelength calibration establish “Time 0” feature wavelengthsVerification and adjustment
6. Wavelength CalibrationPeak positions determined repeatedly to within 0.1 nmValidate calibration and only make adjustment(s) if neededA simple linear fit is sufficient to adjust back to the instrument’s “time 0” wavelength calibrationAbsolute differences between initial “Time 0” and newly adjusted wavelengths typically less than 0.1 nmVerification and adjustment
7. Radiometric CalibrationLaboratory radiometric calibration performed using:NIST-traceable 1000W FEL lamps (irradiance source) — several lamps used in rotation to detect output shifts or drift25 cm calibrated Spectralon plaqueCenter point of plaque viewed by fore optic (45° view angle)Radiance is computed from raw DN measurement:DN == raw dark corrected detector countsIT == integration time (VNIR)Gain == gain factor (SWIRs)Radiance and Irradiance calibrations
8. System NoiseNoise Equivalent Delta Radiance (NEdL) is a measure of noise performance independent of source radianceNEdL is measured using a stable ~3000°K source illuminating a Spectralon plaqueNEdL is computed as the standard deviation in radiance units when viewing this sourceNoise Equivalent Delta Radiance
9. System NoiseThe signal-to-noise ratio (SNR) performance of a spectroradiometer is a useful predictor of data qualityThe observed SNR varies with the radiance of the viewed surface and the period of the measurementWhen measuring SNR, variation in both instrument and illumination source contribute to the observed SNRGiven the inherent instability of atmospheric transmission, SNR measured using a solar source is typically dominated by time-varying source radianceIf the goal is to characterize spectroradiometer performance, it is far better to measure the noise in a controlled setting where illumination source instability can be minimized and then calculate SNR using measured radiance spectra for the source of interest.Signal-to-noise ratio
10. System NoiseSpectroradiometer SNR for a given radiance is computed by dividing the radiance by the NEdL:The above SNRs were calculated using an NEdL measured using a 1 second measurement period. Noise performance is typically improved by the square root of the increase measurement period.Computing the signal-to-noise ratio
11. Spectral ResolutionMost field spectroradiometers use array detectors in the visible and near infrared (VNIR) 350-1,000 nm wavelength rangeThe mismatch between the flat array detector and the grating’s curved focal plane result in spectral resolution drop off at both ends of the spectral rangeWhile it is possible to implement a flat field correction when designing a holographic grating, the correction is incompleteVNIR — 350-1000nm
12. Spectral ResolutionBy incorporating an addition optical element in front of the detector the spectral resolution function is flattened:Improving VNIR performance
13. Spectral ResolutionVNIR solar radiance: when compared to a standard design (red), the improved VNIR spectroradiometer design (green) resolves finer features at BOTH ends of spectral range. Note: Solar radiance spectra were acquired in June 2018 in Colorado. Slight difference in radiance are likely due to water absorption or change in atmospheric reflectance. Improved VNIR performance
14. Spectral ResolutionMeasured full range (350-2,500 nm) spectral resolution of a modified ASD FieldSpec4 high-resolution Next Generation (NG) spectroradiometer obtained using noble gas atomic emission lines. The flatter SWIR (1,000-2,500 nm) resolution is because a scanning design results in a fixed detector to grating distance. Full wavelength range, 350-2500 nm
15. January 28, 201915