XIII_0690 Determination of Silver in Sulfide Minerals by ETMS Trajce Stafilov", Anna Lazaru a, and Ernst PerniclGib "Institute ofCbemistry, Faculty 0/ Science St. Kiril and Metodij University P.o.s. 162, 91001 Skopje, Macedonia bMax-Planck InstitutfUr Kernphysik Postfacb 103980, D-6900 Heidelberg, Germany INTRODUCTION The interference of various ele­ ABSTRACT ments in the determination of silver It is known that the mineral by ETAAS in geological samples was A method for the determina­ lorandite (flAsS) from the Alshar investigated by a small number of deposit (Republic of Macedonia) tion of silver in sulfide minerals authors (34-35). Fishkova and could be used as a solar neutrino (real gar, orpiment, lorandite, mar­ casite, and stibnite) as well as Vilenkin (34) established that only detector 0). The aim of this study iron interferes at concentrations was to determine the Pb concentra­ dolomite by electrothermal atomic absorption spectrometry is dis­ higher than 100 mg.dnr" . tion in the thallium minerals from cussed. After dissolution of the Alshar, particularly in lorandite, samples, silver was extracted by Because of the differences in which is produced in the nuclear diphenyl thiocarbamate in hydro­ results reported in the literature reaction between solar neutrinos chloric acid media using methyl for the determination of silver at low and 2°511 (present in lorandite): isobutyl ketone, butyl acetate, or concentrations in complex matrices toluene as the solvent. The proce­ using FAAS or EfAAS, we performed 20sn + V -> 205Pb + e- dure was verified by the method a complete investigation as to the Therefore, it is necessary to be of standard additions and with interference of all elements in able to exactly determine the lead standard reference samples. The arsenic-antimony-thallium ores and standard deviation (SD) for 0.5 some sulfide minerals. The extraction concentration in thallium and non­ ng/Ag was 0.01 ng. The relative ofsilver by diphenyl thiocarbamate thallium minerals (2-5), marcasite standard deviation ranged from (FeS), or in dolomite [CaMg(CO~21. 3-5%. The detection limit, calcu­ in hydrochloric add media and differ­ The results provide the necessary lated as three SDs of the blank, ent solvents (MIDI(, butyl acetate, data on the purity of the minerals, was found to be 0.001 ug/g, and toluene) is described here, which is important for different geochemical investigations and also EXPERIMENTAL helps to explain the background of silver by FAAS. Some authors Instrumentation radiation of the Alshar locality. In 03-14) suggested the direct deter­ continuing our previous work on mination of silver in solutions. A Perkin-Elmerf Model Zeeman the determination of these elements To remove the potential interfer­ 3030 atomic absorption spectrome­ (6- 10), we now suggest a new ences in the determination of silver ter, equipped with a Perkin-Elmer method for the determination of sil­ in different geological samples HGA-600 graphite furnace and a ver in the above-mentioned minerals with FAAS, extraction was sugges­ Perkin-Elmer AS-60autosampler, was using electrothermal atomic absorp­ ted using the compounds methyl used. The light source was a silver tion spectrometry (ETAAS). isobutyl ketone (MIBK) 05-16), hollow cathode lamp (perkin-Elmer diethyldithiocarbamate (7), or Model Intensitron'"). Pyrolytically With the first investigations in other organic compounds (18-23). coated graphite tubes with platforms the 1960's, flame AAS was used were used. The instrumental parame­ for the determination of silver in Different approaches are ters are given in Table I. geological samples (11-12). The described for the determination of influence of the matrix in the deter­ silver by ETAAS . Brooks et al. (24), Reagents and Standards mination of silver was reported Persiani and Durking (25), and All reagents and standards were Hamalainen et al. (26) suggested only when the potential interfering of analytical grade. The stock solu­ element was present in high con­ the direct determination of silver tion of silver was prepared by dis­ centrations. Tyndall (11) proposed in solutions of copper are samples. solving AgN03 with deionized a method for the direct determina­ The direct determination of silver water. The concentration of silver tion of silver from copper ore sam­ in solid geological samples using ples with a low Pb concentration. the graphite furnace was suggested Spitzer and Tesik (12) conduded by other authors (27-29). Some that the presence of high amounts authors proposed the separation of of iron influences the determination silver from the matrix (32-35). 158 tomic• eetroseo July/August 1995 TABLE I due was dissolved in a solution so that the concentrations of these Instrumental Parameters of 1 mol.dm' HCl. After dissolu­ elements were similar to the con­ for the Determination tion of the mineral samples, the centrations in the sample solutions. ofSilver by ETAAS solution was filtered and trans­ Figure 1 shows that the interfering ferred into a separatory funnel. elements tend to decrease the inte­ Wavelength 328.1 run Then, 1.5 em> diphenyl thiocar­ grated absorbance of silver at high Spectral slit with 0.7 run bamate (1% in acetone) was concentrations. In order to avoid Calibration mode Peak area added, the solution brought to these interferences and because Lamp current lOrnA 20 cm3 volume with 1 mol.dnr? of the very low concentration of Background Zeeman HCI, and then 5 cm3 MIBK, silver in the samples investigated, correction corrector butyl acetate, or toluene was it is necessary to separate and con­ added. The mixture was shaken centrate the silver in the samples. Matrix modifier NH 4H2P0 4 Aliquot injected for five minutes and silver was Forthisreason,ane~ction into HGA 20 IJL determined by ETAAS in the method is proposed. organic layer, using aliquots Dry In the literature, authors of 20 rnm? for introduction into Temperature 110°C the graphite furnace. describe the analysis of different Ramp time 20 s types of samples using diphenyl Hold time 30 s (b) Stibnite and dolomite: 0.1-0.5 g thiocarbamate for the extraction of powdered mineral sample was silver but using different conditions Char dissolved in 10 cm3 concentrat­ and procedures than we propose in I. Temperature 400°C ed HCI and 1 mL concentrated this work. For instance, Shaburova Ramp time lOs ~02' Then the same procedure et al, (36) used diphenyl thiocarba­ Hold time lOs was followed as described for mate for silver extraction from bro­ II. Temperature 600°C realgar and orpiment. mide with chloroform as the solvent. Bashov and Sokolova (19) Ramp time lOs (c) Marcasite: 0.1-0.5 g powdered proposed the extraction of silver in Hold time lOs mineral sample was dissolved in different silicate ore samples with 5 cm3 concentrated HCl, 5 cm3 Atomization diphenyl thiocarbamate in butyl concentrated HN0 , and a few 3 acetate. In both procedures, FAAS Temperature l800°C of~02' drops The solution was was used. Aruscavage and Campbell Ramp time Os evaporated to dryness, and then (37) suggested the extraction of Hold time 3s the same procedure was fol­ silver with diphenyl thiocarbamate Flow mode Gas stop lowed as described for realgar from silicate samples with butyl and orpiment, but with toluene Cleaning acetate, but from a 20% tartaric acid or butyl acetate as the solvent. Temperature 2650°C solution. Ramp time 1 s (d) Lorandite: 0.1-0.2 g powdered To check whether As, Ca, Mg, Sb, mineral sample was dissolved in Hold time 3 s Fe, and 11 coextract with Ag, solu­ 15 em> concentrated HCl, 5 cm3 Gas Argon tions with the same silver concentra­ concentrated HN0 , and a few 3 tion but with different interfering of~02' drops The solution was element concentrations were pre­ evaporated to dryness and then in this solution was 1000 mg.drrr', pared. Silver was extracted using the the same procedure was fol­ from which all diluted solutions proposed method with MIBK, butyl lowed as described for realgar were prepared. Ore and mineral acetate, or toluene as the solvent. and orpiment, but using butyl samples were taken from the Alshar After extraction, silver was deter­ Mine, Republic of Macedonia. acetate as the solvent. mined in the organic phase by neu­ tron activation analysis, and 11, Fe, Procedure RESULTS AND DISCUSSION Sb, Ca, and Mg were determined by (a) Realgar and orpiment. 0.1-0.5 g The interference of the following FAASafter evaporation and dissolu- G of powdered sample of realgar matrix elements present in ores and tion of the residue. Only 0.3-0.4% of or orpiment was dissolved in minerals was studied: As, Sb, Fe, AI, the total As was found to be extrac­ 3 • 10 cm concentrated HN03 11, Ca, and Mg. Solutions with the ted together with the silver and A few drops of H,O., were same concentration of silver but 0.005-0.01% of the remaining added and the soiutfon evapo­ with various concentrations of the elements. With such depletion rated to near dryness. The resi­ interfering elements were prepared factors, further interferences are not 159 A expected in the determination of sil­ ver. Very good results were obtainec 0.24 for the extraction of silver with diphenyl thiocarbamate from acid 0.22 solutions in the recovery values wtl MIBI<, butyl acetate, or toluene wa 0.20 used as the solvent in the presencjf As, Sb, Ca, and Mg (realgar, orpimit, 0.18 stibnite, and dolomite). When Fe ; present in the sample (marcasite)t is 0.16 not possible to extract silver in MK because iron coextracts with silv. 0.14 r­ Extraction of silver from a Tl-matK ! (lorandite) is possible only with 0.12 ;­ butyl acetate. 0.10 i_-----L_---'-_--l-_----L-_--l...-_ -L-_J......--------'_---'-_-"' Tables II and III show that sis­ o 5 10 15 20 25 30 35 40 45 50 factory results were obtained 'r the m(M):m(Ag).10 3 determination of silver in mirral samples from the Alshar Min(with Fig.