LIBS Info: Element Analysis
| Title | Authors | Material | Detector | Spectrometer | Software |
|---|---|---|---|---|---|
| Pre-flight calibration and initial data processing for the ChemCam laser-induced breakdown spectroscopy instrument on the Mars Science Laboratory rover | D. Vaniman, R.L. Tokar, R. Perez, A. Ollila, H. Newsom, P-Y Meslin, N. Melikechi, J. Mazoyer, N. Lanza, O. Gasnault, Cecile Fabre, M.D. Dyar, D. Delapp, L. Deflores, A. Cousin, B.L. Barraclough, D. Blaney, S. Bender, S.M. Clegg, Ryan B. Anderson, O. Forni, J. Lasue, Sylvestre Maurice, Roger C. Wiens | Geological | ICCD | Czerny-Turner (Unspecified) | |
| Laser: | KGW 1067.0000nm 14.0000mJ 3.000Hz |
||||
| Gate Delay: | -10.000us | ||||
| Gate Width: | 9999.000us | ||||
| Very comprehensive description of the pre-MARS work on the ChemCam. In this paper, 69 geological samples were analysed at a distance of 3.0m using a Mars like atmosphere (in particular 930Pa CO2). To ensure homogeneity at LIBS Beam diameter scale (350-550um), the samples were typically pressed powders. Extensive pre-processing is undertaken, including subtraction of dark spectrum (that is, without laser pulses), denoising using wavelet transforms, continuum removal, live wavelength calibration and correction for instrument response. The paper also compares Total Emission measures, noting particularly poor performance for gypsums. LOD figures are from univariate calibrations, whereas RMSEP values are from PLS2 model. The Authors also did work on improvements in PLS models from removing a standard or a spectrum from the model. | |||||
| Element | Detection Limit (ppm) | Wavelength (nm) | Other Wavelengths (nm) | Calibration Method | Calibration Range (ppm) | Notes |
|---|---|---|---|---|---|---|
| C | 24000.0000 (Calibration Curve Slope) | -10.0000 | N/A | Univariate | -10.0000--10.0000 | As this analysis takes place under Mars conditions [930Pa CO2], the detection limit is higher as the technique needs to discriminate sample carbon from atmosphere signal. |
| Pb | 300.0000 (Minimum Detectable Concentration) | 406.0000 | N/A | None | -10.0000--10.0000 | LOD#1 [note: wavelength is to nearest integeter nm] |
| Ba | 46.0000 (Calibration Curve Slope) | 455.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| Sr | 15.0000 (Calibration Curve Slope) | 408.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| Rb | 11.0000 (Calibration Curve Slope) | 780.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| As | 960.0000 (Minimum Detectable Concentration) | 447.0000 | N/A | None | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| Cu | 424.0000 (Minimum Detectable Concentration) | 327.0000 | N/A | None | -10.0000--10.0000 | LOD#1 [note: wavelength is to nearest integeter nm] |
| Mn | 63.0000 (Calibration Curve Slope) | 403.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| Cr | 9.7000 (Calibration Curve Slope) | 425.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| K | 12.0000 (Calibration Curve Slope) | 770.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 [note: wavelength is to nearest integeter nm] |
| Cl | 35000.0000 (Minimum Detectable Concentration) | 838.0000 | N/A | None | -10.0000--10.0000 | LOD #1 |
| Li | 0.3000 (Calibration Curve Slope) | 671.0000 | N/A | Univariate | -10.0000--10.0000 | LOD#2 |
| H | 320.0000 (Calibration Curve Slope) | 656.0000 | N/A | Univariate | -10.0000--10.0000 | LOD #2 from paper |
| Element | RMSE (ppm) | Wavelength (nm) | Calibration Method | Notes |
|---|---|---|---|---|
| K | 0.3950 % | -10.0000 | PLS | Scaled from K2O |
| Na | 0.4850 % | -10.0000 | PLS | Scaled from Na2O |
| Ca | 3.9700 % | -10.0000 | PLS | Dragged higher by a few significant outliers |
| Mg | 3.0300 % | -10.0000 | PLS | Dragged higher by some large outliers |
| Mn | 0.1300 % | -10.0000 | PLS | Good overall PLS fit |
| Fe | 2.8800 % | -10.0000 | PLS | Note: Scaled from Fe2O3 |
| Al | 2.2200 % | -10.0000 | PLS | Note scaled from AL2O3 figure |
| Ti | 0.7900 % | -10.0000 | PLS | For most samples the Mean error is much lower - dragged higher by several outliers |
| Si | 7.7300 % | -10.0000 | PLS | Uses the same number of PLS factors for all elements. |