TY - JOUR
T1 - Quantification of gold and silver in minerals by laser-induced breakdown spectroscopy
AU - Díaz, Daniel
AU - Hahn, David W.
AU - Molina, Alejandro
N1 - Funding Information:
This research was sponsored by Fundación para la Promoción de la Investigación y la Tecnología (FPIT) del Banco de la República de Colombia (project number 3712 ) and Vicerrectoría de Investigación, Universidad Nacional de Colombia-Sede Medellín (project number 30324 , QUIPU 200000015330 ). Authors acknowledge the Colombian Mining Company Operadora Minera S.A. (Medellín) for providing access to their facilities and supplying ore samples and chemical analyses. D. Díaz thanks the support of Colciencias (Convocatoria No. 567, 2012).
Publisher Copyright:
© 2017 Elsevier B.V.
PY - 2017/10/1
Y1 - 2017/10/1
N2 - The performance of laser-induced breakdown spectroscopy (LIBS) to identify and quantify gold and silver in ore samples was evaluated. Ores from a gold-producing mine and samples artificially doped with Au and Ag solutions to previously defined concentrations (surrogates) were prepared as 50-mm pellets prior to LIBS analysis. Silver detection and intensity measurement was straightforward for concentrations from 0.4 to 43 μg/g and from 1.1 to 375 μg/g in ore and surrogate samples, respectively. Au emission lines were not found after ensemble average or accumulation of 100-single shot LIBS spectra of ore samples containing up to 9.5 μg/g Au. However, the Au signal was present in the spectra of surrogate samples, for which a detection limit of about 0.8 μg/g was determined. When the number of sampling shots in ore samples increased, various single shot spectra registered Au emission lines. The number of spectra containing Au emission lines increased with the number of single shots. Those results, as well as scanning electron microscopy analysis of ore samples, suggest that the discrete analyte distribution as well as the inherent discrete characteristics associated to LIBS made the presence of gold in the LIBS spark an unlikely occurrence. The particle sampling rates (the percentage of laser pulses expected to sample at least one particle) were estimated for gold concentrations of 1.1 and 10.0 μg/g as 0.04% and 0.32%, respectively. A Monte Carlo simulation indicated that > 100 gold-containing particles should be sampled to accurately represent the discrete character of gold in the ore. Sampling 100 such particles requires > 105 laser pulses over a single pellet. Despite the fact that this rather large number of shots makes difficult to conduct conditional analysis on pellets, for some samples that withstood 5000 shots, gold quantification in ores was successfully achieved at concentrations as low as 1 μg/g. Results are encouraging and illustrate the applicability of LIBS to gold and silver in field semi-quantitative analysis.
AB - The performance of laser-induced breakdown spectroscopy (LIBS) to identify and quantify gold and silver in ore samples was evaluated. Ores from a gold-producing mine and samples artificially doped with Au and Ag solutions to previously defined concentrations (surrogates) were prepared as 50-mm pellets prior to LIBS analysis. Silver detection and intensity measurement was straightforward for concentrations from 0.4 to 43 μg/g and from 1.1 to 375 μg/g in ore and surrogate samples, respectively. Au emission lines were not found after ensemble average or accumulation of 100-single shot LIBS spectra of ore samples containing up to 9.5 μg/g Au. However, the Au signal was present in the spectra of surrogate samples, for which a detection limit of about 0.8 μg/g was determined. When the number of sampling shots in ore samples increased, various single shot spectra registered Au emission lines. The number of spectra containing Au emission lines increased with the number of single shots. Those results, as well as scanning electron microscopy analysis of ore samples, suggest that the discrete analyte distribution as well as the inherent discrete characteristics associated to LIBS made the presence of gold in the LIBS spark an unlikely occurrence. The particle sampling rates (the percentage of laser pulses expected to sample at least one particle) were estimated for gold concentrations of 1.1 and 10.0 μg/g as 0.04% and 0.32%, respectively. A Monte Carlo simulation indicated that > 100 gold-containing particles should be sampled to accurately represent the discrete character of gold in the ore. Sampling 100 such particles requires > 105 laser pulses over a single pellet. Despite the fact that this rather large number of shots makes difficult to conduct conditional analysis on pellets, for some samples that withstood 5000 shots, gold quantification in ores was successfully achieved at concentrations as low as 1 μg/g. Results are encouraging and illustrate the applicability of LIBS to gold and silver in field semi-quantitative analysis.
KW - Conditional analysis
KW - Gold
KW - Gold ore
KW - Laser-induced breakdown spectroscopy (LIBS)
KW - Silver
UR - http://www.scopus.com/inward/record.url?scp=85029149919&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85029149919&partnerID=8YFLogxK
U2 - 10.1016/j.sab.2017.08.008
DO - 10.1016/j.sab.2017.08.008
M3 - Article
AN - SCOPUS:85029149919
SN - 0584-8547
VL - 136
SP - 106
EP - 115
JO - Spectrochimica Acta, Part B: Atomic Spectroscopy
JF - Spectrochimica Acta, Part B: Atomic Spectroscopy
ER -