Determination of the Gold Content in Geogas by Resonance Ionization Mass Spectrometry W. Y. MA*, Q. HUI, M. XUE, W. X. JI AND D. Y. CHEN Department of Physics, T singhua University, 100084 Beijing, China A resonance ionization mass spectrometric technique for the by an electric field. The most efficient ionization of the excited atoms is through a discrete autoionizing state embedded above ultratrace determination of Au in geogas samples was developed. A three-step excitation scheme leading to an the first ionization limit of the atoms, because its crosssection can exceed by several orders of magnitude that of autoionizing state was used for detecting the Au content of 60 geogas samples combined with a graphite furnace non-resonance photoionization.11–13 The ions created are subsequently measured in a mass spectrometer where electrothermal atomizer. In order to increase the selectivity and detection efficiency, a linear time-of-flight mass additional elemental or isotopic selectivity is added.Time-of- flight (TOF), magnetic sector and quadrupole mass spec- spectrometer was built and tested. The experimental results are in good agreement with those obtained by neutron trometers have been used, of which the TOF mass spectrometer is particularly suitable for use in combination with pulsed laser activation analysis. The study of the Au distribution in geogas is of great interest in prospecting for gold deposits. excitation and has the advantage of a high transmission efficiency.Keywords: Gold; geogas; resonance ionization mass This paper reports on the ultrasensitive determination of spectrometry the Au content in geogas samples, based on thermal atomization of a substance in vacuum followed by resonant stepwise A recent method for prospecting for deeply concealed gold photoionization of Au atoms through an autoionizing state deposits involves the determination of the concentration of Au and combined with TOF mass spectrometry. Our purpose was in geogas.Wang et al.1 developed a rapid, inexpensive and to test if this method can be used in searching for buried gold highly efficient method for the dynamic collection of geogas deposits. To our knowledge, this is the first application of from soil that proved to be satisfactoryfor buried gold deposits. RIMS to the detection of Au in geogas. The experimental The problemis that the Au content in geogas has not previously results were also compared with those obtained by neutron been investigated because of its low concentration, lower than activation analysis (NAA). 0.1 ng l-1, and the lack of sensitive analytical methods. Using conventional methods for determining the Au content in this EXPERIMENTAL natural material, preliminary chemical enrichment must be used. The complexity of the multistage chemical treatment of RIMS System samples make such a concentration technique labour consum- The analytical spectrometer for RIMS was arranged according ing and not very reliable. The most severe problem is that to a standard scheme.11 It includes an electrothermal atomizer, until now there has been no suitable preliminary chemical sample autochanger, laser system, TOF mass spectrometer and enrichment method for Au in geogas, because the amount of signal collector; a schematic diagram is illustrated in Fig. 1. sample is very small. Hence there is an urgent need to develop A 20ml volume of sample solution was poured into a a method with high sensitivity and high selectivity for this graphite crucible and air dried, then it was heated by an natural material.electric current to about 1500 °C. To suppress the strong The concept of resonance ionization spectrometry (RIS) was background from thermal ions and electrons there were two first advanced by Hurst and co-workers2,3 and called single- D-shaped boxes at electric potentials of ±24 V above the atom detection (SAD) since a striking example of a single Cs heater.As a result, the residual background ion level was not atom was detected. Combining RIS with mass spectrometry, higher than one ion per second. Three laser beams were resonance ionization mass spectrometry (RIMS) is characterized by a very high sensitivity owing to efficient ionization and detection of the ions produced, and high selectivity owing to multiple resonant transitions and mass-selective detection. The high sensitivity and selectivity of RIMS provide an accurate measurement method for trace amounts with little or no sample preparation.This, in turn, improves the speed of the analysis. This feature is very suitable for ultratrace element determinations in small amounts of sample, such as determining the Au content in geogas and many other hard-to-obtain samples. Many studies4 –10 have demonstrated that RIMS is an extremely powerful tool for detecting trace amounts of most elements in the Periodic Table, and it has wide uses in geochemistry, chemical exploration and many other fields.The RIMS technique includes two processes, resonance ionization and mass analysis. In the resonance ionization process, the specific atoms are excited resonantly by laser radiation to an intermediate state in one or several steps, and Fig. 1 Schematic diagram of the RIMS system. then only the excited atoms are ionized by laser radiation or Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12 (57–59) 57directed to the vacuum chamber and intersected the atomic beam perpendicularly between the two electrodes. The Au atoms in this sample were excited and ionized by laser beams through an autoionizing state. The Au+ ions created were subsequently accelerated under a dc electric field of 3.2 kV. After travelling in the field-free region of the TOF mass spectrometer they were separated from the background ions produced by other elements or molecules with the multiphoton non-resonant ionization process, and were detected with microchannel plates at the end of about a 1.5 m field-free drift tube.The current pulse signal from the microchannel plates was amplified by a fast amplifier and coincided with the delayed gate signal in synchronism with the laser pulse. The output Fig. 2 Autoionizing spectrum in the third excitation step of Au atoms signal was integrated to a charge which was converted into a leading from the 6d2D5/2 state to the autoionizing state with resonance digital form by a QDC (charge-to-digital converter) and was wavelength 588 nm.The solid curve represents a fit by a Shore–Fano finally fed into a microcomputer. The data were acquired and profile and the points represent the experimental data.14 processed automatically. The residual pressure in the chamber was about 10-6 Torr. By means of the sample autochanger, 20 samples can be For trace Au determination, the three-step scheme of measured continuously without affecting the vacuum.As a autoionizing state excitation and ionization was chosen: result, the measurements were carried out efficiently and the experimental conditions were uniform for all samples. 6s2S1/2 CCCC l1=243 nm 6p2P3/2 CCCC l2=479 nm Laser System 6d2D5/2 CCCC l3=588 nm autoionizing state The laser system consists of three tunable dye lasers pumped by an excimer laser to produce laser beams with suitable A first-step laser pulse with wavelength l1=243 nm excited wavelengths and intensities.The excimer laser (EMG202) and the Au atoms from the 6s2S1/2 ground state to the 6p2P3/2 two dye lasers (3002E and 3002EC) were obtained from intermediate state. Then a second-step laser pulse with wave- Lambda Physik (Germany). The third dye laser was made in length l2=479 nm performed further excitation of the atoms our laboratory. The excimer laser (XeCl) can produce laser to the 6d2D5/2 state. Finally, a third-step laser pulse with pulses of 400 mJ energy and 308 nm wavelength with a pulse wavelength l1=588 nm excited the atoms to an autoionizing width of 28 ns. In our experiments, the repetition rate of the state.The energies of the first-, second- and third-step laser excimer laser was 20 Hz. With non-linear processes consisting pulses in the experiment were 10 mJ, 100 mJ and 8 mJ, respectof second harmonic generation and mixing, the dye lasers ively. In the probe volume the typical diameter of the laser produce tunable light from the IR to UV region with intensity beams was about 5 mm, so the energy fluences of the laser 10 mJ–10 mJ and linewidth 0.2 cm-1.radiation were 51 mJ cm-2, 509 mJ cm-2 and 41 mJ cm-2, respectively. These values are sufficient to saturate the resonant transitions. Sample Preparation Aqueous AuCl3 Samples The samples were taken from the surface of a goldmine in the north of China, deeply concealed underground. About 10 l of Aqueous solutions were obtained by dilution of the AuCl3 geogas in soil were extracted by a gas acquirer from a hole standard solutions with deionized water.Calibration curves 50 cm deep underground and Au atoms were absorbed in a were constructed by measuring the AuCl3 standard solutions foamed plastic. Half of each piece of foamed plastic was over the concentration range 0.01–10 ng ml-1 before and after analysed by RIMS and the other half was analysed by neutron analysing the geogas samples every day because of a lack of a activation analysis, reported elsewhere in detail.14 standard geogas sample.The calibration curve given by y= The chemical processing of the sample was very simple: the 2450x+6 shows excellent linear behaviour. The AuCl3 stan- foamed plastic was ashed at 450 °C in a ceramic crucible and dard solution of concentration 0.01 ng ml-1 and a blank then dissolved in 1 ml of pure aqua regia to convert the Au sample of de-ionized water were used for determining the compounds into AuCl3.After drying, the sample can be detection limit of Au. Using the 3s criterion, where s is the preserved for several weeks. A 1 ml volume of 1% aqua regia root-mean-square error of the blank sample, the detection was added to the ceramic crucible to re-dissolve AuCl3 before limit of Au obtained with the AuCl3 standard solution was detection. Only 20 ml of the liquid sample were injected into 0.003 ng ml-1. the graphite crucible for analysis. Geogas Samples For real sample analysis 32 geogas samples taken above the RESULTS AND DISCUSSION goldmine and 28 geogas samples from the surrounding area Autoionizing State of Au were analysed.Each sample was measured three times and the mean values obtained are given in Table 1. The concentration Gold atoms were excited in three steps to an autoionizing state lying 4636.5 cm-1 above the first ionization limit, which of Au in geogas samples taken above the goldmine is in the range 0.12–6 ng l-1, obviously larger than that in samples was found as reported previously.15 This autoionizing state, characterized by a strong and narrow resonance peak, has a from the surrounding area (0.02–0.1 ng l-1).The precision is about 30% for the Au content in geogas at levels as low as very large photoionization yield. Fig. 2 shows the autoionizing spectrum with a resonance wavelength of 588 nm; the solid 0.02 ng l-1 and is better than 15% for Au contents larger than 0.1 ng l-1.curve represents a fit by a Shore–Fano profile.14 58 Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12Table 1 Gold content in geogas in soil (ng l-1 geogas) In conclusion, RIMS is an extremely efficient method for the determination of ultratrace concentrations of elements. Samples from above goldmine Samples from surrounding area Significant attributes include excellent sensitivity and very high Sample no. Au content/ng l-1 Sample no.Au content/ng l-1 selectivity, which provide for accurate measurements with little or no sample preparation and allow the use of very small 1 0.14 33 0.09 amounts of sample. These merits are very important in ultra- 2 0.15 34 0.04 3 0.19 35 0.10 trace element determinations in hard-to-obtain samples and 4 0.20 36 0.10 improve the speed and accuracy of the analysis considerably. 5 0.28 37 0.09 It is clear that RIMS will be useful in prospecting for concealed 6 0.34 38 0.03 deposits and has a wide range of application in many fields. 7 0.44 39 0.03 8 0.63 40 0.04 The authors thank Professor X. J. Xie and X. Q. Wang of the 9 6.0 41 0.02 10 0.64 42 0.03 Geophysical and Geochemical Exploration Institute for pro- 11 0.38 43 0.04 viding the geogas samples and many helpful discussions during 12 0.25 44 0.07 these experiments. 13 0.18 45 0.06 14 0.14 46 0.02 15 0.18 47 0.02 16 0.25 48 0.04 REFERENCES 17 0.30 49 0.05 1 Wang, X. Q., Xie, X. J., and Lu, Y.X. Chin. Geophys. Geochem. 18 0.68 50 0.08 Explor., 1995, 19, 161. 19 0.39 51 0.06 2 Hurst, G. S., Nayfeh, M. H., and Young, J. P., Appl. Phys. L ett., 20 0.23 52 0.09 1977, 30, 229. 21 0.15 53 0.07 3 Hurst, G. S., Nayfeh, M. H., Kramer, S. D., and Young, J. P., 22 0.13 54 0.09 Rev. Mod. Phys., 1979, 54, 767. 23 0.12 55 0.06 4 Hurst, G. S., and Letokhov, V. S., Phys. T oday, 1994, 47, 38. 24 0.15 56 0.08 5 Payne, M. G., Deng, L., and Thonnard, N., Rev. Sci. Instrum., 25 0.18 57 0.10 1994, 65, 2433. 26 0.23 58 0.09 6 Kluge, H. J., Bushaw, B. A., Passler, G., Wendt, K., and 27 0.63 59 0.05 Trautmann, N., Fresenius’ J. Anal. Chem., 1994, 350, 323. 28 2.72 60 0.10 7 Saloman, E. B., Spectrochim. Acta, Part B, 1993, 48, 1139. 29 0.35 8 Saloman, E. B., Spectrochim. Acta, Part B, 1994, 49, 251. 30 0.21 9 Wendt, K., Passler, G., and Trautmann, N., Phys. Scr., 1995, 31 0.18 T58, 104. 32 0.14 10 Perera, I. K., Lyon, I. C., and Turner, G., J. Anal. At. Spectrom., 1995, 10, 273. 11 Hui, Q., Chen, D. Y., Niu, J. G., Cheng, Y., Xu, X. Y., and Zhao, Table 2 Comparison of the results obtained by RIMs and NAA W. Z., Inst. Phys. Conf. Ser., 1990, 114, 297. 12 Ma, W. Y., Hui, Q., Zhao, W. Z., Wen, K. L., Xu, X. Y., and Au content/ng l-1 geogas Chen, D. Y., Mod. Phys. L ett. B, 1991, 5, 1095. 13 Ma, W. Y., Li, L. Q., Zhao, W. Z., and Chen, D. Y., Opt. Sample no. RIMS NAA Commun., 1991, 85, 408. 12 0.25 0.24 14 Wang, X. Q., T he Study of Prospecting for Giant Gold Deposits by 23 0.12 0.12 Wide-spaced Geochemical Sampling in Combination With L SAD 44 0.07 0.06 and Other Analytical T echniques, Chinese Report of National 8th 46 0.02 0.03 Five-year Plan, 1995, 30. 52 0.09 0.10 15 Zhao, W. Z., Xu, X. Y., Ma, W. Y., Cheng, Y., Hui, Q., Wen, 59 0.05 0.06 K. L., and Chen, D. Y., Appl. Phys. B, 1991, 52, 299. 16 Braun, T., and Rausch, H., Anal. Chem., 1995, 67, 1517. It is well known that NAA is a useful method for trace Paper 6/06031E elements determinations with high sensitivity, high accuracy Received September 2, 1996 and good precision.16 Table 2 gives some results for the Au Accepted October 2, 1996 content in geogas samples obtained by our method and by NAA. The results obtained with these two methods show good agreement. Journal of Analytical Atomic Spectrometry, January 1997, Vol. 12 59