A tumizatiurz and Excitatiort 1.5 VAPOUR GENERATION 1.5.1 Hydride Generation In the past year a review of hydride generation methods (see also Section 3.1.1.5) has appeared (2002). The use of hydride generation is no longer confined to the original AAS methods. The introduction of hydrides into the ICP, MIP and d.c. arc plasmas is reported in Sections 1.2.1.4, 1.2.2.2 and 1.2.3, respectively. Labelled 75Se was used to confirm expressions derived to quantify the effects of apparatus design, atomization efficiency and operating conditions on signal absorbance and peak area (967).A patent has been published (1265) that described a new type of heated quartz measurement cell with end pieces made of heat resistant materials such as quartz, ceramic or graphite. The function of the end pieces is to cool down the hot H,, which is formed in the cell, and thus prevent its ignition when it comes into contact with the air.There has been increased interest in the determination of the volatilized metals by nun-dispersive AFS. Nakahara et d. have dcscribcd methods based on NaBH, reduction, and transport of the hydride to an Ar/H, flame where A F was measured non-dispersively following excitation with an EDL.They have applied their apparatus to the determination of As (412), Bi (167, 1589) and Sb (797, 1556), and described in each case, the optimization of conditions and identification of inter-element interferences, Azad et al. (510) reported the30 A 11 a1 y tical A tornic Spectroscopy use of a non-dispersive AFS method for the determination of Se in soil digests.Measure- ments were again made in an Ar/H, flame. Co-precipitation of Se with lanthanum hydroxide or masking with Te have been recommended as methods for the elimination of Cu interference (510, 713). Continuum source AFS has been applied to the determination of metal hydrides formed by NaBH, reduction using dispersive (572) and non-dispersive (wide bandpass interference filter) measurement (573).The Ar/H, flame was used for atomization and a 300 W Eimac Xe arc lamp for excitation. In the non-dispersive measurement system, multi-element determinations were made by separation of the hydride species in a tube filled with gas-chromatographic packing materials. Electrochemical reduction, of As to ASH,, has been used in a non-dispersive AFS method (34), and for the conversion of Sn to SnH, in an AAS method (1835).In both cases a quartz tube atomizer heated to 700 "C was used. Several detailed interference studies have again been reported (1 390, 1882, 1893) and one paper (1390) included some useful hints on methods that can be used to reduce inter-element effects. Taga et al. (1805) described an interesting method which can be used for the determination of Te(IV) and Te(V1) in mixtures.When NaBH, was used on its own, only Te(1V) was reduced to the hydride and determined. Addition of TiCl, as a pre-reductant allowed the determination of both oxidation states of Te. Other references of interest - Antifoaming agent for As in urine determinations: 1980. Automation: 1491.Determination of Sn in a long path absorption tube in a N2/H2 flame: 1779. Use of enhancing reagents in Pb determinations: 632. 1.5.2 Mercury Determination Papers on this subject (see also Section 3.1.1.4) continue to appear in considerable numbers and many of them include minor modifications to the apparatus or procedures used for the cold vapour AAS method, particularly with regard to automation.Two new AFS methods have been described both involving non-dispersive detection of the Hg AF signal (411, 1157). In one method, Hg was electrolytically deposited from solution onto a gold cathode and subsequently released by heating the cathode at 700 "C in a stream of He (1 157). A linear response from 20 pg to 2 p g of Hg with a detection limit of 0.7 pg was achieved.The method was successfully applied to biological material and natural water samples. In the second procedure (41 I), the usual reduction by SnCl, was used and dispersive and non-dispersive AFS methods were compared, The non-dispersive method was linear over the range from the detection limit of 0.05 ng (0.003 ppb) to 1 pg. Non-dispersive AAS methods have been described by Hoffman et al.(1656) and Fuwa et al. (790, 915) that made use of the more sensitive Hg absorption line at 184.9 nm. In both cases the apparatus consisted simply of a Hg lamp [discharge lamp (1656) or EDL (790)], a long-path absorption cell and a V.U.V. sensitive photomultiplier tube. Considerable improvements in sensitivity were claimed in comparison to measurements at the inter- combination line at 253.7 nm.For example (91 5), a characteristic concentration of 0.29 ng (1 5% absorption) and a detection limit of 0.05 ng were achieved for 0.5 ml of solution. The factors affecting the shape of peaks in the cold vapour AAS determination of Hg have been investigated and amalgamation was proposed as a means of overcoming problems due to the variable rate of release of Hg from the reduction cell (1665).Collection of Hg released from the reduction cell on activated charcoal, Au and Ag in various forms has been the subject of renewed interest as a means of increasing the sensitivity of the AAS method. Activated charcoal appears to be the least satisfactory (957,A tomizatioii and Excitatiott 31 1614) as it shows considerable memory effects, slow and incomplete dcsorption, a high affinity for interfering volatile organics (957) and poorer sensitivity than other materials (1614).The use of Ag-coated quartz wool was recommended by two groups of workers (76, 1614) and Au-coated quartz wool by others (1581, 1664). Release of Hg occurred at 650-800 "C. The Au-coated quartz wool was reported to be more efficient for direct collection of Hg from air than was Au wire (1664), and was "nearly" quantitative at flow rates up to 10 1 min-1 and temperatures of 50 "C (1581).Thc use of very thin films of Au on sea sand was preferred t a the same material coated with Ag, as the latter did not retain R,Hg and RHgX was only partially adsorbed (957, 1462). Au or Ag liners applied to graphite furnaces were also found to give improved sensitivity for the determination of Hg by AAS (1566).A home-made electrical furnace was used as an AA detector for the GC determination of alkyl-mercury compounds in fish tissue (1463). A radiotracer study using 197Hg has been used to demonstrate the extent of Hg losses during analysis and the trapping of Hg in a vapour generation apparatus (1943).Other references of interest - Background correction using the wings of the broadened 253.7 ng Hg line: 1941. Comparative study of NaBH, tablets and SnCl, solutions: 1936. Dzterminations of Hg by a Au film Hg detector (resistivity measurement): 2053. 1 S.3 Methods Based on Molecular Absorption Some authors have this year re-discovered the possibility of measuring molecules, particu- larly those containing nitrogen, in the cold vapour or hydride AAS cell by means of their molecular absorption (see ARAAS, 1976, 6, Ref. 1351). Ammonia has been measured at 201 nm with a continuum source or with EDLs with lines in the appropriate region, such as those of As, Pb, Sb or Se (164, 718, 1364). Nitrate and nitrite were determined by prior reduction to NH, (164, 1364). A nitric oxide (NO) hollow-cathode lamp was used as a light source for the measurement of NO in a cold vapour measurement cell (948). The (0, 0) band of the NO molecule at 227nm was used and the light beam was split both in time and space to correct for drift and scatter. Internal calibration was achieved by use of a sealed quartz cell of known length, which could be tilted into the light beam, and contained a fixed concentration of NO. Similar procedures to those described above have been adopted for H,S, SO, and thiosulphate (718, 1364). Boron has been determined by generation of BF, by reaction with calcium fluoride and sulphuric acid. In a high-temperature flame, AA of B, emission of BO, and molecular absorption of BF were all observed (164).