18 Analytical Atomic Spectroscopy 1.3 LOW PRESSURE DISCHARGES 1.3.1 Glow Discharge Lamp The stable and reproducible emission characteristics of the GDL offer attractive advantages over many other less-easily controllable forms of excitation (see ARAAS, 1976, 6 , IS). Harrison et al. (344) have illustrated the use of the discharge as an excitation source for OES, a sputter atom-source for AAS and an ionization source for atomic MS.The dischargz allows the control of parameters such as the choice of gas and its pressure in order to optimize excitation conditions, Surface effects in the sample are of special importance and studies using scanning electron microscopy show the formation of spires and cones, which can influence analytical results. The growing versatility of the technique has been demon- strated by several application studies.An automatically stabilized GDL has been used by Czakow (603) for the analysis of powders, metals and solutions. Kenawy et al. (637) and Fijalkowski (638) have determined Br, C1, F, Se and S in non-metallic materials. Butterworth (43) has compared the pa-form- ance of a GDL with that of a spark source for low-alloy steel analysis using a constant-time exposure, i.e., by exposure until a preset integrated-current value was reached, and by the method of Jaeger (ARAAS, 1972, 2, Ref. 18) in which sputtering rate was used as zn internal standard. Calibration curves were rectilinear and the precision obtained compared favour- ably with that obtainable by using the spark source. A similar comparison has been made for a range of matrices including low- and high-alloy steel, aluminium, copper and lead alloys (606).For rare-earths analysis, solutions (10~1) have been evaporated on to a polished copper plate (548); for Sn in tin ores a technique in which the powdered sample was mixed with high-purity Cu in the ratio of 1 : 20, and then compressed and degassed, proved successful in the hands of unskilled staff (631).A glow-discharge atomizer described earlier (Anal. Chem., 1975, 47, 194) has been further improved and characterized as an atom source for multi-element analysis (1224). Excitation temperatures of the order of 5000 K in the Ar discharge were reported; the addition of 1% H, assisted in the production of a reproducible, stable glow. Butler (878) has described a new AFS apparatus incorporating two GDLs.The analyte was atomized cathodically in a first GDL to emit primary radiation; this radiation was focussed through a quartz lens into a second GDL operating in a pulsed mode to produce resonant fluorescent radiation. The fluorescence was passed through an interference filter to a photomultiplier tube and the average fluorescence was recorded.Using this equipment for the determination of Mg and Cu in steel, and Cu in aluminium, detection limits of 20, 10 and 25 ppm respec- tively were obtained; for Cu and Ag in gold, the limits were 2 and 3 ppm, respectively (682). In the latter application the analytical curve tended to bend towards the concentra- tion axis above 0.5% concentration, but this would be reduced by operating the GDL at a power lower than the 0.01 kVA used in these experiments.TV camera used with a GDL: 1574. Other reference of interest -Part Z: Fundamentals and Znstrumentation 19 1.3.2 Cathodic Sputtering and Hollow Cathode Discharges Work of the type first reported by Walsh (Spectrochim. Acta, 1955, 7 , 108) on the atomic absorption measurement of vapours produced by cathodic sputtering has continued at C.S.I.R.O.(684). Internal standards were used for the determination of Fe, Cr, Ni and Cu in metallic and non-metallic samples; the ratio of the absorbance of the analyte to that of the internal standard was compared with the corresponding ratio in the reference material (452). This allowed accurate determinations to be made even when the sputtering rate was quite different from that of the reference sample.The assumption that concentration ratios in the vapour phase were identical to those in the solid phase was supported by experi- mental measurements of Ni/Cu ratios in different alloys. Theoretical studies of mechanisms in the hollow cathode discharge have included work by Semyonova et al. (635), who considered the influence of resonance and recombination processes upon the radiation of Cu, Ag and Au vapours.Torok et al. (915) examined the radial distribution of electron temperature in a liquid-air cooled discharge; Zhiglinskii et al. (1316) proposed a method for determining the temperature of the plasma and the concentration of metal atoms from changes in the intensity distribution of the hyperfine structure of the spectral line.Atomization caused by a sputtering effect as in the HCL discharge should lead to reduced matrix effects; results of matrix effect studies have been reported by McCamey et al. (332), and by Szilvassy (744) who studied the determination of rare earths in oxides. Caroli et al. (869) obtained emission spectra of Cu and steel by using HCL and spark excitation sources and of Cu using a GDL.The HCL radiation source was the most accurate for OES analysis. Pichugin et al. (166) stabilized the discharge in a HCL by packing the analyzed material into a cavity drilled into the bottom of a demountable graphite electrode. Improved detec- tion limits at the ppb level are reported for impurities in graphite powder. Sabatovskaya et al.(1 139) cxamined the influence both of the inside diameter of the cathode and the gas pressure on the detection limits obtained using a HCL. Line intensities increased with decreasing internal diameters and gas pressures, but discharge stability was poor when using a cathode of reduced internal diamcter. A cathode in which the cavity had a ‘stepped’ profile (4 mm at inlet, 1.5 mm at base) ensured good stability and gave a 10-fold improve- ment in detection limits at a gas pressure of 20 Torr.Hollow cathode excitation with vidicon detection offers a promising alternative to convc.Ttiona1 OES with photomultiplier detection, Good results were reported for the simultanecus dctermination of rare earths in minerals when a standard-addition procedure was used (272), and also for Pb and Cd determination in airborne dusts using graphite porous-cup electrodes as the filter material in the sampling device (659, 1646). Other references of interest - Determination of Eu in rare earths by isotope dilution using a HCL discharge: 633. Electron impacr 3xcitation sources: 650. Hanle effect in atomic vapours produced by cathodic sputtering: 380. Surface analysis by ion-beam sputtering: 930.