JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 171R ATOMIC SPECTROMETRY UPDATE-ATOMIC EMISSION SPECTROMETRY Barry L. Sharp* Chemistry Department L ough boro ugh University o f Technology Lo ugh bo ro ugh L eices tershire UK LE77 3TU Simon Chenery Analytical Geochemistry Group British Geological Survey Keyworth Nottingham UK NE 12 5GG Raymond Jowitt British Steel Technical Teesside Laboratories P. 0. Box 7 7 Grangetown Middlesbrough Cleveland UK TS6 6UB Simon T. Sparkes and Andrew Fisher Department of Environmental Sciences University of Plymouth Drake Circus Plymouth Devon UK PL4 8AA Summary of Contents 1 Arcs Sparks Low-pressure Discharges and Lasers 1.1 Arcs 1.2. Sparks 1.3. Low-pressure Discharges 1.3.1. Glow discharge lamps 1.3.2. Hollow cathode discharges 1.3.3.Other sources 1.4. Lasers 2 Inductively Coupled Plasmas 2.1 Fundamental Studies 2.2. Sample Introduction 2.2.1. Nebulizers 2.2.2. Flow injection 2.2.3. Chromatography 2.2.4. Electrothermal vaporization 2.2.5. Solid sampling procedures 2.2.6. Chemical vapour generation 2.3.1. Torch and generator design 2.3.2. Spectrometers 2.3.3. Instrument control and chemometrics 2.3. Instrumentation 3 Microwave-induced Plasmas 3.1 Fundamental Studies 3.2. Instrumentation 3.3. Sample Introduction 3.3.1. Direct nebulization 3.3.2. Electrothermal vaporization 3.3.3. Chemical vapour generation 3.3.4. Direct analysis of solids 3.4.1. Instrumentation 3.4.2. Gas chromatography-microwave-induced plasma applications 3.4.3. Supercritical fluid chromatography 3.4. Chromatography 4 Direct Current Plasmas This review describes developments in all aspects of atomic emission spectrometry including fundamental processes and instrumentation reported in the Atomic Spectrometry Updates References in JAAS Volume 7 (93/1012-93/862) and Volume 8 (94/1-94/960).The full references names and addresses of authors can be readily found from the Atomic Spectrometry Update References in the relevant issues of JAAS. However as an additional service to readers an abbreviated form of each reference quoted (except those to Conference Proceedings) is given at the end of the review. The trend towards the use of array detectors for emission spectrometry both CCD and CID types noted last year has accelerated with the availability of commercial instrumentation. These have undoubtedly encouraged developments in chemometric techniques for spectral reduction and offer convenient means for spatial mapping of plasma character- istics. There is also evidence that axial viewing of plasmas is gaining in popularity perhaps as a response to the requirement to optimize light throughput for the echelle spectrometers that are usually used in conjunction with array detectors.172R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL.9 Laser ablation sample introduction is an active area of research with more use being made of UV lasers particularly excimer lasers with outputs down to 193 nm (ArF). Improvements in laser design have led to the introduction of field- based instrumentation for geochemical applications. Analyses based on the emission from individual particles produced by laser ablation have been described which complements the continuing work on the fate of individual aerosol droplets in plasmas.In contrast it appears that laser enhanced ionization spectrometry is likely to remain an area for research rather than routine application. Sample introduction always provides a large number of reports and it is interesting to note that the ultrasonic nebulizer is now accepted as a reliable means of improving detection limits at least for low matrix samples The thermospray which potentially offers similar performance has yet to make the transition to general acceptance. The coupling of chromatographic techniques to atomic spectrometric detectors is now well established and although attention has inevitably moved towards mass spectrometric detection the stubbornly mono-isotopic element As retains its attraction for optical spectrometry.1. ARCS SPARKS LOW PRESSURE DISCHARGES AND LASERS Two atomic emission reviews have been produced the first (93/1165) giving 127 refs. related to the use of plasmas as spectroscopic sources and the second (94/670) covering the development of AES in China from 1990 to 1992 citing 412 refs. 1.1. Arcs Fundamental aspects of plasmas have been reported in three papers diagnostics based on line profile measurements of self- reversed lines (93/4046); non-linear interference effects and the role of ion dynamics in the kinetic theory of Stark broadening (94/958); and aspects of signal treatment (93/4041). Chemical reactions in d.c.arc discharges have been reviewed by Hu et al. (93/2136) and the application of the sulfidation reaction to the direct determination of Bi Mo Pb Sn and W in geological samples reported by Liu (93/3 132). Halogen buffer (93/2170) was mixed with powdered rocks and minerals prior to introduction into an arc discharge and ZnCl (94/669) was added prior to ashing foamed plastic used for extraction of Au. Caesium chloride together with PdCl was used (94/850) to achieve 0.01 ppm detection limits for Co Cr Fe Mn and Ni in trimetallic anhydride. Palladium chloride was also used in the determination of impurities in high-temperature super- conducting materials (93/800). Graphite containing 10% NaCl has been used by Li and Zhou (93/616) in the determination of REE impurities in high-purity holmium oxide.Trace element determination in human hair after ashing (94/836) in zinc tungstate after mixing with C-BaO-CaCO at a ratio of 18 1 1 (93/619) and in sodium chloride (93/703) by ETV into an arc have been reported. Preconcentration by distilling off the matrix in the case of red phosphorus (93/736) and oil (93/475) and by sublimation onto a cold finger in the case of iodine (94/343) was used to obtain ppb level LODs for trace elements. The determination of trace levels of toxic elements in environmental samples (93/C1577) has illustrated the potential of a charge injection device (CID) as an alternative to photographic emulsion for the simultaneous measurement of spectral line and background intensities. Direct specimen excitation has been used by Severin et al. (93/2642) for the determination of Ce in low-alloy steel whereas Strasheim and Bohmer (93/1951) have studied the changes that occurred on the surface of ferrous samples prior to quasi- stationary conditions being established.1.2. Sparks Analytical implications of plasma dynamics in the high voltage spark discharge have been reported by Bye and Scheeline (93/C1394). The spatial and temporal dependence of analyte excitation in several analytical matrices were investigated and compared with earlier large bandpass echelle measurements. Pomeroy et al. (92/2055) used an Cchelle spectrometer with a CID array detector in conjunction with an expert system for the qualitative and quantitative analysis of steel and alu- minium.Water analysis has been carried out by Lucht and Salje (93/2689) using an HPLC system to inject sample at 20pl min-' into a spark plasma operating between copper electrodes the upper electrode being a heated copper plate to which a stainless-steel tube was soldered. The well established rotating disc electrode technique for wear metals in oils has been modified by Kauffman (93/2686) to enable the detection of particles > 45 pm. The sample was placed on the flat surface of the rotating-platform electrode ashed in a furnace at 400 "C for 30s and then inserted into the analysis chamber of the emission spectrometer. 1.3. Low Pressure Discharges 1.3.1. Glow discharge lamps The advantages and disadvantages of the use of GD as an analytical source for spectrometry have been reviewed by Harrison (92/4598) whilst a more specific review of 61 refs.discussing layer-by-layer analysis has been prepared by Drobysher (93/792). Fundamental mechanisms of excitation and ionization have been reported by Hess et al. (93/C1500) for both d.c. and pulsed GDs. Wagatsuma and Hirokawa (94/622) made obser- vations of singly ionized Cu emission lines in the visible wavelength region produced by an argon-helium GD plasma and Weston et al. (93/C1594) gave details of data acquisition and evaluation by a computer controlled Langmuir probe system. Spatial measurements were made by Kuraica et al. (93/1646) who reported radial distribution measurements of electron density and temperatures in the plane cathode GD whilst Rusnak and Vicek (94/874) studied the distributions of excited states of hydrogen and nitrogen in the cathode region of a G D used for the formation of nitride on the surface of steel.Temporal signal profiles of analytical species in modu- lated GD plasmas have been investigated by King and Pan (93/C1433 93/3433) using AES AAS and MS measurements which for argon atoms maximized within 2 ms following termination of the discharge power. Various aspects of gas compositions used in GDs have been reported. De la Cal et al. (93/3601) characterized a He-CH d.c. GD plasma and proposed a simplified kinetic model which accounted for all their observations. Chambers et al. (93/3437) investigated the role of gas dynamics in negative ion formation in a GD ionization source sampling from the atmosphere. The source was used in conjunction with MS.Mixtures of Ne Xe and halogens were used by Galaritskii (93/4093) to study excitation efficiency a mixture of 4.8 Torr (1 Torr = 133.322 Pa) Xe with 0.8 Torr C1 was found to produce the maximum UV radiation yield. Hieftje et al. (93/C1475) reported on a gas sampling GD for the quantification of element ratios among C C1 F and S in organic compounds quoting high excitation energy and low background as the features of GD which made it useful for this application. Ulgen et al. (93/3370) were not content with the levels of argon line background and obtained a 6- to 9-fold improvement in detection limits for sputtered species by using a pulsed supply of square wave form at 400JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL.9 173R and 700 V together with a lock-in amplifier. As Ar emissions predominate at the low voltage these intensities were factorized and used in background subtraction to obtain net analyte signals at the higher voltage. Four presentations relating to the r.$ GD source have been made by Marcus et al. (93/C1358 93/C1360 93/C1496 93/C3026). These dealt with the conflicts involved in the development of the source (93/C3026) figures of merit for r.f. GD-AES (93/C1496) plasma parameter effects on crater shapes in the r.f. GD (93/C1358) and sputtering characteristics of glasses and ceramics using the same source (93/C1350). The effects of parameters such as r.f. power gas pressure and sample thickness on the AE and mass spectra of sintered ceramics have been investigated (93/591 93/C1389) and alternative gases were evaluated by Caruso et al.(93/C1498). Heintz and Hieftje (93/C1499) reported on their recently started work on a magnetically enhanced rJ GD in which they compared the performance of two designs in relation to the basic source without a magnetic field. Reference was made to the prolific work of Sacks et al. who have produced two further papers (93/1945 93/3421) in which the effect of the addition of CF4 at a ratio of 1 1 with Ar on the sputtering behaviour of Ag Al As Cu and Zn was reported. It was also shown that the CF4 addition resulted in significant quenching of numerous spectral features at both high and low pressures. Harrison et al. (93/C1380 94/604) have also considered the GD as a reactive cell using AES and MS to study the presence of metal argides and the effects of reactive contaminants such as air and water vapour.Microwave boosted GDs have been used to provide improved sensitivity to compensate for the dilution effects of mixing non- conductors with a conductive matrix (93/C1480) and by Tomellini (93/3993) in the direct determination of trace elements in surface layers and bulk materials. Steers and Thorne (94/40) have used UV/VIS FT spectrometry to study the excitation of Cr and Fe spectra in a microwave-boosted GD source and their results emphasized that excitation tem- peratures deduced from data on a limited number of lines are meaningless. Gas-jet assisted GD-AAS has been used by Dean et al. (92/4624) and in Korea (93/3263) as a technique for the analysis of conducting samples by AAS.Subsequent work on plastics and ceramics (93/C1600 93/C1626) has also been reported by Dean’s group. Banks and Blades (93/1648) have investigated directed support gas flows for improved sampling efficiency but point out the disadvantage of reduced depth projiling resolution. Practical applications of quantitative depth profiling analysis by GD-AES have been described by Mitchell and Shirley (93/C1481) and a study of the influence of anode geometry on electric field distribution and crater profile using a GD has been made by Demeny et al. (93/1019). Glick and Hieftje (93/723) have embarked on a futuristic analytical scheme using an artificial neural network and multi- variate calibration of a GD for the classification of alloys.Atomic spectra for seven elements in 37 nickel-based and 15 iron- based alloys were acquired with a pho to-diode array spectrometer. A more traditional approach was used by Lundholm and Baltzer (93/3994) in the determination of N in steel by GD-AES. 1.3.2. Hollow cathode discharges The Memphis State University group have been most active with two papers (93/2211,94/634) and four conference presen- tations (93/C1359 93/C1363 93/C1364 93/C1447). Improved analytical precision was achieved by introduction of a current controlled switch (93/C1447) 93/2211) together with a mech- anically stable photometer (93/C1363) and optimization of sputtering prior to sample introduction (93/C1359). Temporal profiles of emission signals for pulse widths from 15 to 500 s for Al C Mo Nb Ti and stainless-steel microcavity HCs were also presented (93/C1364).Studies of the axial evolution of the negative glow in a HC discharge (94/634) using a vidicon video camera confirmed that the optimum pulse width for analytical AES was 7-11 p. A study of matrix eflects on REEs in HC discharges has been made by Mierzwa and Zyrnicki (93/2088) who used solutions of halides of the studied elements in association with calcium barium and strontium matrices. It was concluded that calcium should be removed from a sample prior to the determination of REE traces. You and Marcus (93/C1477) have developed a thermal concentric nebulizer for the intro- duction of volume-limited solutions and Papp (93/3660) reported on an electrothermal and/or HC combined spectro- scopic atomizer and/or radiation source for qualitative and quantitative elemental microanalysis.The excitation mechanism of metal vapour spectra in an HC has been studied by Chera et al. (93/3330). Klemp et al. (93/4036) have characterized a low-pressure HC device for element-selective GC detection and Jin et al. (93/3122) have used an HCL for the trace determination of sulfide and SO by vapour molecular absorption spectrometry. Wilson (94/673) has reported the merits of a boosted HC lamp as a source for AAS and Sansonetti et al. (93/4050) have produced an atlas of the spectrum of a Pt-Ne HC reference lamp. 1.3.3. Other sources Flame characteristics have been investigated by Pupysher et al. (93/3202) using a thermodynamic simulation of element atom- ization in air-C,H N20-C2H2 N,O-propane and air- methylacetylene flames.Temperatures and compositions were calculated for a wide range of fuel-oxidant ratios and equilib- rium concentrations of different forms of 58 elements were established. The ionization behaviour of the alkaline-earth metals in air-C2H and N,O-C,H2 flames was considered by Luecke (93/3967) and the spatial distribution of flame emission intensities reviewed by Sohma (93/4055). Flame emission has been proposed for the determination of Na in cement presented as a slurry (93/2123) and for La in REE concentrates by measurement of the La0 band at 441.8 nm (93/2728). Real-time production control information has been obtained by Wendt and Persson (94/835) from spectroscopic measurements of off-gas flame emissions enabling the desired Cu content of white metal to be achieved with an accuracy of +0.2% in the range 74-78%.In situ measurements were also made by Sorenson et al. (93/1937) in scattered light for the measurement of soot-cluster monomer particle radius and to count the number of monomers per cluster in a methane-oxygen flame. Charalampopoulos et al. (93/2711) further investigated light scattering from flame par- ticles with particular attention to the role of iron pentacarbonyl vapour additions. Calloway and Jones (93/C1450) have reported on the use of a flame as an emission source for AAS and Karanssias et al. (93/3391) used an oxygen-hydrogen flame to introduce slurries of marine sediment into an ICP. Systems for sample introduction into James were reported (93/C1621 93/1952,93/3659) as were flame photometric detec- tors for chromatography (94/858 94/860).Sturgeon et al. (92/2750 92/4642 93/C1473) have further characterized the FAPES technique dealing with the d.c. self- bias potential which develops in the asymmetric r.f. system (93/C1473) the influence of generator frequency (92/2750) and the excitation and detection of molecular species (92/4642). Two different geometries of a capacitively coupled r.J plasma source for use with furnace atomization have been charac- terized by Blades et al. (92/2753,93/542 93/C1356,93/C1553) Helium plasmas could be sustained over a much wider power range 5-150 W than argon plasmas. Gilchrist et al. (94/595) also used argon and helium plasmas in a capacitively coupled version of FAPES and found better sensitivity and S/N for Ag Au Cd Pb Sb Sn TI and Zn with argon.Riby and Harnly174R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 (93/C1474) characterized a helium plasma in hollow-anode FANES sustaining a plasma at pressures up to 900Torr although analytical results were only obtained at pressures up to 600Torr. The FANES technique was compared with ICP-MS for the analysis of laser-ablated biological material by Hoffmann et al. (93/C3034). Plasma characteristics involving a range of gases have been reported. Mehdi et al. (94/620) used optical emission to study an r.f. magnetron sputtering discharge at an argon pressure of 7-150 mTorr as did Ropcke et al. (94/594) to measure relative concentration distributions and energies of species in low- pressure argon discharges.Jones and Carnahan (93/1650) produced an explanation for non-metal ion emission behaviour in argon and helium discharges whilst Alekseer et al. (94/956) determined trace levels of Ne in helium. A neon plasma was investigated by Gavrilov (93/3656 93/4084) and a spectro- scopic study of the plasma produced through bombardment of Mg with Kr+ ions was carried out by Belyaer et al. (94/882). Plasmas involving common gases have also been examined. The temperature distribution in carbon dioxide plasma arc welding (94/809) IR spectroscopic detection of ions and free radicals in a hydrogen discharge plasma (94/916) and optical emission spectra excited by an r.f. oxygen plasma used to evaporate yttrium barium copper oxide have all been studied.Optical emission characteristics and measurements related to the study and control of processing by plasma deposition have been reviewed by Barney et al. (93/3471) and reported by Ishii et al. (93/3319) for titanium nitride films Pan et al. (93/3574) for sodium deposition Peignon et al. (93/3608) for ion etching of tungsten and Durrant et al. (93/3766) for fluorinated poly- mer films. A plasma gun source for the direct atomization of the most refractory materials has been characterized by Goldberg et al. (92/2393) and used in conjunction with pulsed and steady state microwave plasmas (93/C1391) and capacitively coupled r.f. plasmas (93/C1393). A microwave discharge was also used to produce the active nitrogen afterglow used by Yu et al.(93/3281 94/839 94/848) for the determination of Hg and Zn in water by metastable energy transfer emission Spectrometry. Sheeline et al. (92/2607 93/C1392) have again reported on their study of a theta-pinch discharge emission source covering various aspects of quantification. Spatial and temporal tem- perature studies of electrothermal chemical plasmas using atomic spectroscopy have been reported by the Army Ballistics Research Laboratory of the USA (94/921). Typical events involved the deposition of about 300 J of energy yielding temperatures from 10000 to 50000 K the latter being the result of shock wave heating of the supersonic exit flow from a polyethylene capillary. Emission spectra from a high-current line plasma have been reported for the first time (94/789).The formed-ferrite plasma source was driven by a 4.6 pF capacitor bank charged to 25 kV and spectra in the region 120-290 mm were measured. What applications could such a source be useful for? 1.4. Lasers The most significant review of the year was by Darke and Tyson (94/280). They have culled a wide spectrum of infor- mation from the literature on laser-solid interaction and its significance to analytical spectrometry. This review will provide a valuable source of information for several years to come. Thiem et al. (93/478 93/3779) have twice reviewed the use of lasers in atomic spectrometry during this update period once in detail (208 refs. in 93/478) and again with reference to recent advances (19 refs. in 93/3779). Majidi and Joseph (93/3514) reviewed the use of laser induced plasmas for spectroscopy (93 refs.) for a wide variety of applications both industrial and environmental concluding that the major advantage of the technique is its ability to readily sample solids.Sjoestroem and Mauchien (94/687) focused on trace element determination using spectroscopic techniques dependent on resonant absorp- tion of laser radiation by atoms. The 74 refs. provided a very useful bibliography for this specialized field. Other reviews were 9312666 and 94/825. The ready availability of excimer laser systems operating in the UV and array-based solid state detector systems is revol- utionizing direct LA-AES. The excimer laser seems to provide an excellent excitation source and the array detector systems allow simultaneous collection of background signals.Workers at the University of Massachusetts Lowell have continued their investigations into the potential of the ArF excimer laser for LA-AES. They demonstrated (93/3189) that when an ArF laser operated at 193nm was focused onto selected metal targets higher excitation temperatures and ion populations resulted compared with conventional plasmas. Quantitative emission measurements with respect to both space and time were reported (93/3297 93/3257). These studies suggested a confined plasma production with a peak emission after approximately 20 s and a lifetime of less than <lOOs. Simeonsson and Miziolek (93/2882) have also looked at the fundamental properties of the micro-plasma produced by an ArF laser in various carbon-based atmospheres (CO COz methanol and chloroform) and measured similar ionization/ excitation temperatures ( 15000-20000 K) and electron densi- ties ( 1017-1018 ~ m - ~ ) but different breakdown thresholds. The aim of the study was to provide a firm basis for laser microplasma-gas chromatography.Mauchien et al. (93/C 1564) used a 400 mJ XeCl laser with a time-gated multichannel spectrometer and a precise positioning device to produce an experimental apparatus that would maximize reproducibility . It was demonstrated with different aluminium alloys that there were no matrix effects a reproducibility of 1.3% at the 1OOOOpgg-' was possible and this is normally limited by sample heterogeneity. Excimer lasers are not just being used to sample solids. Ng et al. (93/3254) used an ArF excimer laser (193 nm wavelength) with LA-AES to analyse liquid aerosols produced by a conventional concentric nebulizer and spray chamber with an argon carrier gas flow rate of 0.51 min-'.Emission signals lasted for 35-50 ps after each laser pulse. The excitation temperature decreased from 3994 K at 1 ps to only 3607 K at 35 ps after the laser pulse. Detection limits for the nine elements determined varied between 0.3 g 8-l for Li and 20g 8-l for Sr. Nyga and Neu (94/906) developed a double- pulse LA-AES technique with an excimer laser for sampling liquids. They observed that ablated liquids normally produce broad and quenched spectral lines. However with the double pulse system the first pulse generated a cavitation bubble on the surface of the liquid providing a gaseous environment for the micro-plasma produced by the second pulse.The result was sharp atomic and ionic line atomic emission spectra. Uebbing et al. (93/2120) have also demonstrated a double pulse LA-AES system using a 1064 nm Nd:YAG to overcome some deficiencies of a single pulse system. The first pulse was fired at a solid producing a micro-plasma. A second more energetic laser was then fired into the microplasma parallel to and 1.5mm above the surface of the test material. Linear calibration graphs were then obtained for A1 and Mn in glass and steel and for Mg and Mn in glass copper and aluminium using internal standardization. The effect of the atmosphere on the spectroscopic conditions of an LA microplasma is also known to be important. Sdorra and Niemax (93/3125) as part of their continuing basic investigations into LA-AES studied the effects of pressure and composition of bufSer gases (air Ar He N and Ne) as well as laser energy on ablation crater diameter depth and mass loss.The variation in plasma tem- perature and relative electron densities with time were also measured for each gas at a fixed pressure and laser energy. It was demonstrated that Ar was the best buffer gas for elemental analysis of solid samples by LA-ICP-AES although Ne might have advantages in certain instances. Further to their work on buffer gases Sdorra et al. (93/3115) compared the determi-175R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 nation of A1 and Mn in steel and borax glass by LA-AES using a Nd YAG laser at both its fundamental IR frequency ( 1064 nm) and its frequency quadrupled into the UV (266 nm).Unexpectedly when using UV irradiation results were inferior to IR irradiation and the authors considered that the use of internal standards was not possible. They concluded that an additional atomization step was required before the UV system would be analytically useful. Vaskovskii et al. (94/881) investi- gated the effect of air pressure on the dynamic and optical characteristics of a micro-plasma produced by a CO laser. They also studied the morphology of the test material after ablation. Majidi and Joseph (93/C1556) evaluated the effect of various concentrations and pressures of a helium atmosphere on the emission spectra of a laser induced micro-plasma. The laser plasma excitation of electrothermally atomized species was compared for helium argon and nitrogen atmospheres.Lee et al. (93/3267 93/C1565) used air argon and helium atmospheres in the pressure range 10-760 Torr when observing the micro-plasma produced from a Cu target by their ArF excimer laser. Observations on reducing the atmospheric press- ure included a decrease in plasma size; an increase in atom line intensity; decrease in plasma temperature; and a movement of the maximum away from the test material surface for air and argon. In contrast Owens and Majidi (92/2752) investi- gated the effect of high pressure (up to 2300 Torr) Ar He and N2 on the A1 I1 emission at 281.6 nm. Ablation of A1 embedded in resin was performed with a Nd:YAG laser. Calibration graphs were not linear but the results were reproducible.An LOD of 17 pg g-' was obtained. A theoretical and experimen- tal investigation into the effect of atmosphere (He Ar and N,) and gas pressure (severalTorr to atmospheric) on micro- plasma initiation when using an IR laser at 10.6 pm was performed by Hermann et al. (93/3602). The plasma was studied by both space- and time-resolved AES. Theoretical predictions agreed reasonably well with experimental data if the vaporization-initiated plasma breakdown mechanism was used. An explanation of the influence of the atmospheric gas on the laser light to metal surface energy transfer was given. Keefer (94/920) used a modern digital signal image processing system to acquire spectral images of a micro-plasma produced by a 1.5 kW CW C02 laser in an Ar atmosphere.Using the measured temperature field the laser power absorption and thermal plasma emission could be calculated at any point in the plasma to provide a detailed understanding of the energy conversion process. Ultra-fast LA has been reviewed (27 refs.) by Von Linde (93/411 l) and Mehlman et al. (94/922) reported experimental results from the 'Table Top Terawatt' laser. Emission in the far UV range from laser irradiation of targets was recorded. Test materials were silicon wafers coated with A1 layers of variable thickness (10-500 nm). The laser energy penetration depths obtained for 1064 nm irradiation were 30-70 nm but only 25-40 nm at 532 nm. It is likely to be some time before systems of this kind are available generally. The use of LA to mobilize material for vapour deposition is of increasing interest particularly with reference to ceramic superconductors.Jiao et al. (94/843) used plasma emission spectra and time-of-flight (TOF) mass spectrometry to study laser evaporation and deposition of the YBa2Cu,0 - supercon- ductor. Fan et al. (93/3285) investigated laser-induced plasma emission spectra of YBa2Cu30 and observed component atoms singly and doubly charged ions as well as monoxide molecules. Mueller et al. (94/917) compared ablation of diamond-like carbon silicon and copper test materials using a pulse length of 30ns and 500 fs for a 248nm wavelength laser. Optical emission spectra indicated a higher contribution of C + ions rather than C2 and larger molecules for the shorter wavelength laser.Using the 30ns ablation cluster formation was quite obvious using TOF-MS and this coincided with the deposition of micrometre-sized particulates. There was no evidence for cluster formation using the shorter wavelength laser. Vega et al. (94/791) used an excimer laser to ablate a germanium target in an oxygen atmosphere at various pressures. The oxygen content of the deposited film increased to the full stoichiometric value (GeO,) at pressures higher than 5 x lop3 mbar (1 bar= lo5 Pa). Observation of the spectral emission showed no evidence of oxidized species even at high oxygen pressures and the authors concluded that the oxidizing reactions mainly take place at the substrate site. The applications of LA-AES have been very varied.Wisbrun et al. (94/915) investigated laser-induced breakdown spec- troscopy for the detection of heavy metals in environmental samples. Detection limits were usually below those needed for regulatory levels. This particular application required optimiz- ation of the experimental set-up to overcome matrix effects resulting from water and organic fibres as well as problems resulting from the mechanical properties of the test material and the particular distribution of the contaminants in the sample. Aguilera and Campos (93/3153) determined carbon in steel using time-resolved spectroscopy of a plasma produced by a focused Nd:YAG laser in a nitrogen atmosphere. They achieved good agreement with conventional techniques a precision of 1.6% and a detection limit of 65 pg g-I.Carlhoff and Kirchhoff (93/3990) reported on the use of LA-AES for the direct analysis of molten steel inside a converter with a fibre optic to take the light from the plasma to an optical multichannel analyser. The authors claimed the application of this method reduced consumable costs and improved steel quality. The more general application of this particular system to the detection of trace elements down to 10-100 pg g-' in matrices as diverse as steel rubber and rocks was described by Lorenzen et al. (93/2075). With the reduction in size of laser systems there is now a welcome move of LA-AES away from the laboratory and out into the field where speed of results outweighs the limitation of the technique in terms of absolute accuracy and precision.Hardjoutomo et al. (93/3320) designed and constructed a transversely excited atmospheric pressure (TEA) CO laser with a 180 mJ output energy and 50 ns pulse duration especially for jield-based laser microprobe spectrochemical analysis of geological test materials. The use of helium as an atmospheric gas improved the S/N ratio by reduction of the continuous spectrum of the plasma and minimum detectable concen- trations of 50 pg g-' for a Zn I line and 500 pg g-' for an F I1 line were demonstrated. Another field-based system has been devised by Cremers and Kane (93/C1558) for the determi- nation of lead in paint. Various Pb lines were investigated for possible spectral interferences and the 220.35 nm line chosen as optimal. This yielded a detection limit of 1% of Pb in paint corresponding to a surface density of 0.06 mg cmP2.Calibration was performed for between 1 and 11% Pb as normally found in Pb based paints. Sequential laser sampling allowed depth profiles of lead/no-lead layers to be built up. An unusual application of laser-induced breakdown was particle detection in liquids. Fujimori et al. (93/4054) applied this technique to polystyrene particles in water and were able to detect particles as small as 0.02 pm. They observed a change in laser breakdown threshold and plasma emission delay time with particle size and concluded that it should be possible to measure both the concentration and size of particles in fluids by the proposed method. Coupled techniques can sometimes be used to overcome specific difficulties in LA-AES.Majidi et al. (92/2384,93/C1563) coupled an electrothermal vaporization step with excitation from laser ablation and trace element detection by AES. Test solution (5 1) was deposited in the graphite furnace then dried ashed and atomized. At the start of the atomization cycle a pulsed Nd YAG laser was focused and fired along the axis of the furnace into the gas phase. The atomic emission spectra were collected through the dosing hole orthogonal to the laser light. Further work (93/3620) extended this to using a frequency doubled Nd YAG laser (532 nm) for excitation. Similar work176R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 has been carried out by Ridge and Crouch (93/C1562) but using a carbon rod atomizer rather than a graphite furnace.Goldberg et al. (93/C1561) built on their previous work using a theta pinch magnetic field to re-excite a laser-produced plasma for atomic emission studies. This time they used a higher frequency (120kHz) but a lower intensity (20 kG peak) magnetic field. The higher frequency induced current allowed the separation of emission due to the decaying laser plasma from emission due to the interaction of the pulsed magnetic field with the atom/ion cloud. This also allowed the probing of the ionic environment in the decaying plasma. Results on the effect of the atmosphere as well as temporal and spatial emission were presented. Laser enhanced ionization spectrometry remains a topic of research but has never achieved widespread acceptance. Turk (93/2209) compared single and double resonance LEI of phosphorus monoxide in an air-C2H2 flame for the determi- nation of phosphorus.Detection limits of 200 and 30 ng g-' of P were obtained by single and double resonance LEI. The sensitivity of the double resonance was four times that of the single. Axner and Rubinsztein-Dunlop (93/2878) evaluated two step LEI in flames and laser-induced fluorescence (LIF) in a graphite furnace for the determination of Cr in water. They achieved LODs of 1.4 and 0.3 ngml-'. The effect of large amounts of Na on the LEI was investigated. They concluded that detection limits were limited by contamination of the burner and graphite furnace. Yan et al. (93/676) discussed the interference of the OH radical on LEI spectrometry for some trace element determinations in an air-C2H2 flame.The deter- mination of Pb at 288.305 nm was given as an example proving to be a particular problem at low analyte concentration and when a fuel rich flame was operated. Axner et al. (93/2879) used LEI to investigate the lifetimes of atomic metastable states of Au in an air-C,H flame. From this they were able to map the local stoichiometric conditions of the flame. Lee et al. (94/9 19) have investigated resonant multi-photon ioniz- ation (REMPI) both theoretically and practically for the Hg atom. Ayala et al. (93/C1536) evaluated resonance line lasers (RLLs) as excitation sources for analytical spectrometry. Three metal halide RLLs were constructed and characterized GaT In1 and TII. The potential advantages of these sources included narrow line widths natural locking to an atomic transition and the irradiance and coherence of a laser.They investigated the fundamental figures of merit for such sources including spectral output and source lifetime. The application of RLLs was demonstrated for LEI and LIF with ngml-' LODs for both techniques being obtained. It might seem attractive to try LEI type experiments with the ICP. However Turk et al. (93/1954) have previously reported problems with background signals but have now attempted laser-induced ionization in a power-modulated ICP. Detection limits were improved but were still less than adequate. Resonance ionization either with detection by atomic or mass spectrometry has great potential by virtue of the selec- tivity of the excitation processes.But much fundamental data is needed to make this a practical analytical tool. The National Institute of Standards and Technology (NIST) USA has established a data service to provide fundamental information for resonance ionization spectrometry (RIS) and resonance ionization mass spectrometry (RIMS). This service includes atomic data appropriate resonance ionization schemes and operating details. Saloman (94/633) has published NIST's fourth data sheet for Ag Be In Li K Rb Ti and V which includes an update on Ni. Ray et al. (93/3310) scanned the spectral region 6150-670 nm using a dye-laser pumped by a Nd:YAG laser 176 resonances were observed for U. Most were assigned to three photon resonant-ionization processes but some resonances suggested that four photon processes were occuring. Sampling of a solid test material can be achieved directly by resonant laser ablation as used by Borthwick et al.(93/2590) to measure ppm levels of A1 in steel. This procedure was used to measure ion to neutral atom yields at low laser fluences. This group have also developed a novel ablation chamber (93/Ct614) where the laser light is transmitted directly onto the sample using short lengths of optical fibre. Devyatykh et al. (93/2601) determined Fe in high-purity aluminium fluor- ide by electrothermal vaporization in a graphite cell and three- step resonance ionization mass spectrometry. Theoretical absol- ute LODs of 4 x lo-'' pg were calculated. Katsuragawa et al. (93/3686) have developed a simple low cost pulsed atomizer for RIS of Mo enabling them to measure isotope shifts with a 200 g sample.Ma et al. (93/3788) used resonance ionization TOF-MS to determine Ru with a detection limit of 50 ppt 20 times better than using non-resonant schemes. Perhaps one of the most exotic concepts recently has been demonstrated by Matsuo et al. (93/2944). See also Gill et al. Spectrochem. Acta. Part B 1991,46,1227-1235. Sampling and ionization of a test material was performed using laser ablation behind a ring electrode. The singly charged ions produced are then confined using either hyperbolic and cylindrical electrodes for up to 20 min in a He buffer gas. Doubly charged ions were confined for several seconds while some ions like Nd+ and Ta' were highly reactive with background gaseous molecules. Once trapped the ions were studied using either a quadrupole mass spectrometry or laser-induced fluorescence spectrometry. 2.INDUCTIVELY COUPLED PLASMAS 2.1. Fundamental Studies Boumans as an early pioneer of ICP-AES was in an ideal position to view the developments and trends in plasma spectrometry over the last 40 years (94/590). This paper provides the reader with an ideal opportunity to put ICP-AES into past present and future context. More specifically Boumans (93/C1403) reviewed multi-element line selection from first principles discussing true LODs in spectral inter- ference situations and means for improving these and line selection. Also discussed were spectral simulations and auto- matic line selection by expert systems. The future was seen as real-time line selection using multiple lines measurements and the implementation of chemometric approaches.De Loos- Vollebregt and van Veen (93/C1402) have already gone some way to implementing Boumans' view of the future using Kalman filter based software for the determination of different elements in uranium tungsten and environmental test mate- rials. They concluded that this approach led to easier line selection and an improvement in LODs when spectral inter- ference caused problems. Simulation studies demonstrated that the Kalman filter could also have benefits for the new gener- ation of Cchelle spectrometers with charge coupled device (CCD) or photodiode array detectors. Miller-Ihli (93/2988) compared ICP-AES with ICP-MS ( 13 refs.) highlighting fea- tures and identifying their limitations.Scholze et al. (93/C3031) also compared the two techniques specifically for the determi- nation of certain elements in soils. They were particularly interested in the possibilities of solid sample introduction by slurry nebulization or laser ablation as alternatives to digestion and aqueous nebulization. In this period Miyazaki has twice reviewed (92/4404 30 refs. and 93/4060 8 ref.) the principles and characteristics of ICP-AES and its application to the determination of trace elements in water. The fundamental properties of the ICP continue to provide a rich area of investigation. Galley and Hieftje (94/581) pro- pounded the use of a new spatial reference point for analytical atomic and ionic emission measurements. Conventionally this has been the height above the top of the r.f.load coil always a dubious reference with the differences in torch box design,JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 177R however readily found in any system. They suggest that an ideal spatial reference point should have the following charac- teristics tracking of positional trends in the number density of atomic and ionic species or their emission; no influence on the spatial distribution of analyte emission; require no additional sample preparation or adulteration; and be simple to establish measure and track. The suggested reference point was the ‘bullet’ shaped emission pattern from the OH bandhead at 306nm. It has the advantage over the addition of other elements in that it is always present it moves position with changes in power but tracks both ionic and atomic emission.The addition of an EIE enhances its intensity again following atomic and ionic trends. It should be possible to follow the ‘bullet’ in one dimension with some commercial equipment although it is suggested that a simple video camera combined with a narrow band emission filter should allow two- dimensional tracking. Marawi et al. (93/2057) directly com- pared plasma temperature measurements made from the N2+ rotational temperature method and the optical pyrometry method. The use of molecular gas ions such as N2+ or the OH radical as temperature probes has been common practice in the past. These workers followed the method of Strother et al. (Appl. Spectrosc. 1991 45 1031) making use of Planck’s radiation law with two-colour IR optical pyrometry to measure plasma temperature.The pyrolytic probe was placed 15 mm above the load coil and recorded temperatures of 1700-2200 K for r.f. powers ranging from 0.75-2.00 kW. The ICP tempera- ture was found to be linearly dependent on the forward power of the generator. Ogilvie et al. (93/3248 and 93/C1591) used a novel method to identify the mechanism by which ionic lines are excited. They noted that perturbations in atomic resonance line emission are caused by incompletely atomized droplets of solution and associated with these droplets are regions of elevated neutral atom density and depressed temperature. Therefore they were able to observe that (i) charge transfer excitation of ionic lines was only weakly dependent on tempera- ture and therefore positively correlated with atomic resonance line emission and (ii) electron impact excitation was strongly temperature dependent and showed a negative correlation.Rayson (93/C3002) investigated Penning ionization and charge transfer as possible mechanisms for the production of excess populations of excited analyte ions. Pulsed-laser excitation experiments demonstrated that Penning ionization is relatively unimportant whereas droplet-induced fluctuations as used above provided correlations between ground-state atom and excited-state ion populations. Travis et al. (93/C1388) have made preliminary studies on the precision and accuracy of the determination of spectral line wavelengths by ICP-FT spec- trometry. The aim of the work is to establish a suite of wavelengths in the ICP which are known to sufficient accuracy to be used for periodic re-calibration of the FT spectrometer and for calibration transfer to dispersive ICP atomic emission spectrometers. The effect of an introduced matrix on plasma properties has led to several detailed studies.Tripkovic et al. have investigated the influence of matrix elements easily ionizable (Li) and non- easily ionizable (Zn and Ba) on atomic emission both exper- imentally (94/297) and theoretically (94/298). Apparent exci- tation and ionization temperatures electron number density and LTE were measured and calculated. Results with and without the presence of the matrix were compared and it was found that the Li changed the ionization temperature. A theoreti- cal calculation procedure based on the minimization of free energy in the ICP was applied in the temperature range 1000-9000 K assuming chemical equilibrium and the influence of matrix elements (Li Ba and Zn) on the emission intensities of Ca and Cd was investigated.Karyakin and Simonova (93/3972) also investigated the influence of matrix elements on trace element emission both theoretically and experimentally. They concluded that the matrix affects not only plasma tempera- ture and electron density but also allows the collision of major element atoms with analyte atoms resulting in a lowering of analyte emission. Chomet (93/2630) studied the easily ionizable element matrix effect on both ICP-AES and ICP-MS using a wide range of matrix concentrations.The observable effect at low concentrations depended on plasma-generator impedance matching. Phenomena observed with both AES and MS detec- tion systems showed that the alkali element effect occurs during the uolatilization/atomization phase and resulted from a change in the plasmas thermal condition. Wu and Hieftje (93/C1590) tried to separate the influence of solvent and EIEs on matrix effects. Two dimensional CCD imaging spectrometry was employed to measure analyte emission intensity and excitation temperature with and without desolvation and EIE addition. These measurements have demonstrated the efficacy of desolv- ation in reducing the EIE interference. Sesi et al. (93/C1593) reported the effects of EIEs on the fundamental parameters of the ICP.Gas temperatures electron temperatures and electron number densities were measured by laser-light Thompson and Rayleigh scattering. Other work has included the simulation of ionization and excitation processes using the Monte Carlo method (93/C3050). Attempts to improve on conventional argon ICP-AES have taken many forms the simplest has been the use of mixed-gas plasmas. Wagatsuma and Kichinsuke (93/3357) observed an increase in the intensity of Zn emission lines in an Ar-He plasma. They concluded that this was the result of an increase in plasma temperature in the central channel only and that this resulted from the higher heat conductivity of helium. This mixed-gas pair has also been evaluated by Sheppard (93/3276). Du et al. (93/3163 and 93/4007) have used an air-Ar ICP to support the introduction of petroleum distillation residues dis- solved in xylene.It showed several advantages over an all Ar plasma including no carbon deposition a reduction in spectral interferences by molecular bands and better LODs. Heavy metals were detected in the samples using AES. Gomes et al. (93/2085) calculated the temperature of an air ICP by measur- ing the widths of a rotationally non-resolved band of the NO system at different plasma heights. In contrast Morgan et al. (93/3221) used an ICP with a Fourier transform infrared (FTIR) spectrometer to observe the near infrared (NIR) emission spectra of carbon and oxygen on the introduction of lower alcohols. A 4% addition of nitrogen to the plasma was found to enhance the resultant signals.Other approaches to improve performance have included that of Thompson et al. (93/2084) who investigated the use of a specialized time-resolued ICP atomic emission spectrometer. A fast multichannel analogue-to-digital converter allowed tran- sient signals (<1 ms) from several elements to be acquired simultaneously. This was used to study the signals arising from the injection of 1-5 pm solid particles produced either by LA or slurry nebulization. Work at the University of Massachusetts has continued on sealed ICP sources for AES (93/2045,93/2046 93/2047,93/3247). Experimental variables for the sealed system were optimized for As and P. Important parameters included plasma power gas mixture discharge container geometry and operating pressure (93/2045). Analysis of silane (93/2046) identified traces of Fe Ge Mg Sn Ti and Zr while arsine (93/2047) was analysed for C Fe Ge Mg Mo Ni Sn and V impurities.Both static and flowing modes were compared with a conventional 40.68 MHz ICP discharge (93/3247). In the sealed system the white noise level was lower and the noise arising from the formation of vortices in a conventional plasma was absent. Borer and Hieftje (94/295 and 94/296) have evaluated a pressure-differential tandem plasma source for AES. The first stage was an ICP an effective atomizer because it has a high thermal mass and gas kinetic temperature while allowing rapid and versatile sample introduction. The second stage was a reduced pressure MIP. A reduced pressure dis- charge might be expected to be more homogeneous and less noisy.Unfortunately the system exhibited no improvement in178R JOURNAL O F ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 LODs over conventional ICP-AES but the authors were positive about future possibilities. 2.2. Sample Introduction 2.2.1. Nebulizers A considerable amount of work has been reported in this area during this review period. Chen (93/3336) has conducted a review (73 refs.) on the advances in sample introduction tech- niques for ICP-AES. The techniques included those for gas liquid solid and suspended samples. A discussion of some of the problems associated with each technique was also pre- sented. Jin et al. (93/614) reviewed liquid sampling into AES AAS and AFS (170 references). This paper included gas propelled nebulizers spray chambers desolvation arrange- ments micro-sample nebulization organic solvent sampling and liquid sample non-nebulization methodology. Recently the use of ultrasonic nebulizers (USNs) has increased considerably. This is presumably because of the availability of stable convenient commercial equipment that gives high- efficiency nebulization and thus improved LODs and sensi- tivity.This increase in sensitivity has been demonstrated by several workers (93/3794 93/2208 93/C1580 and 93/2042). Typical improvements in LODs and sensitivity were by 1 and 2 orders of magnitude respectively. Conflicting reports of the response times were made. For example in one paper it was reported to be 2-3 times longer than when using a cross-flow nebulizer (93/3794) but in another (93/2042) it was reported as being comparable to a pneumatic nebulizer. Long term stability has also been investigated by Shkolnik et al.(93/C1448). The factors governing the performance of a nebul- izer e.g. carrier gas flow rate or sample uptake rate have been optimized (93/862). Eleven analytes were determined in urine and LODs were 28 times better than those obtained using a Meinhard nebulizer. A paper that contradicts those that claim increased sensitivity for ultrasonic nebulization has been pro- duced by Brenner and Ehrlich (93/C1371). This lack of improvement was attributed to the detrimental effects of high concentrations of salts and major cations entering the plasma source. It was also suggested that interferences by EIEs resulted in inaccurate results.In another paper Brenner et al. (93/2044) found that a USN used with a 40.67 MHz plasma generator gave 10-fold enhancements in LODs (down to the pg I-' level). The use of a Trassy-Mermet sheath gas assembly and a 0.3 mm injector tube allowed the analysis of saline waters. Memory effects were described as minimal and the analysis of CRMs gave accurate results. In a similar paper (94/599) geological materials were analysed for REEs. Various sample decomposition methods were attempted and it was found that the content of some of the analytes varied significantly in the final analysis solution depending on the method used. Aerosol transport eflects have been studied by Tarr et al. (93/721) who used laser-scattering particle size and vapour and aerosol mass transport measurements to characterize the performance of a USN.It was found that the aerosol generated had a broader droplet size distribution but also yielded a greater mass transport of solvent and analyte when compared with a pneumatic nebulizer. The same group has compared the transport effects in dribble and jet USNs (93/2043). Desolvation using a heated spray chamber followed by a condenser reduced solvent loading of the plasma but signifi- cant analyte losses in these regions were observed. The jet USN was found to show greater repeatability and better peak shape for FI measurements. The sensitivity of dry and wet aerosols has been studied by Weber et al. (93/808). Dry aerosols of Ag were produced by a spark discharge and transported by argon at 1.41 min-' to the plasma. Alternatively the dry aerosol was passed through a USN for mixing with dilute HNOJ to give a wet aerosol.The USN was also used to nebulize standard solutions of Ag. This meant that the effects of the nebulizer and the plasma on the signal could be separated. It was found that larger water droplets decreased the signal by 43%. Clifford et al. (93/1641) used dual-beam light scattering interferometry to measure particle size particle velocity distributions size-velocity corre- lation particle number density and volume flux of desolvated aerosols. Several acids and salts were passed through the USN and it was found that the mean diameter and volume flux increased as the acid or salt level rose. A comparison of a commercial and an inexpensive humidifier based USN was also made.Botto (93/3406) used a USN with a plasma formed from Ar and 02. Working conditions for water and for organic solvents were tabulated. The system was applied to the determination of tetraethyllead in aviation fuel. Limits of detection and enhancement factors were also given. A novel micro-flow USN that delivered solvent at rates of pl min-' has been described in a continuation of the work by Tarr et al. (94/693). The Sauter mean diameter of the aerosol generated was very low (2 pm) and the analyte transportation efficiency was close to 100%. Acceptable stability precision and reproducibility were observed and LODs with a flow rate of 10 pl min-l were comparable with those from a pneumatic nebulizer operating at 1 ml min-'. The device was considered to be readily transferable to ICP-MS and microwave instru- ments and would allow coupling with chromatographic techniques such as capillary electrophoresis.Thermospray nebulizers have again been studied by several groups of workers. This is because of their improved analyte transport S/N ratios and LODs. Veber et al. (93/3154) com- pared the matrix effects produced by a fused silica aperture thermospray with other nebulizer types. A calcium matrix provided a comparable amount of interference between the thermospray and the USN which in turn was more than for a pneumatic nebulizer. Conver et al. (93/C1379) used this nebulizer and found it to be highly stable and to have low background contamination. A simple therrnopneurnatic sample introduction system for ICP-AES has been developed by Krasil'shchik and Voropaev (94/763).The LODs obtained were 5-10 times lower in pure water and 2-5 times lower in natural waters when compared with conventional pneumatic nebulization. Precision was between 3 and 5%. Concentric nebulizers have again been studied extensively. The noise characteristics of a Meinhard nebulizer have been investigated by Pang et al. (93/2041). It was found that the primary source of the noise arose from the pump. The use of a pulse free pump (e.g. a double head reciprocating HPLC pump) allowed the discrete frequency noise to be eliminated. The use of an appropriate spray chamber also decreased the amount of noise. A Meinhard nebulizer has also been studied by Tan et al. (93/C1413). Parameters such as the flow rates sample uptake rates and geometry of the nebulizer were all studied.Decreasing the sample flow rate from 2.3 to 0.2ml min-' has been found to result in a 4-fold increase in sensitivity if a Meinhard nebulizer was used (92/2560). It was also found that precision was improved 5-fold if natural aspiration uia a capillary was used rather than a peristaltic pump. Several other types of nebulizer have been investigated. A semi-demountable recirculating nebulizer that allowed 0.5 ml of solution to be nebulized continuously over a period of 4 minutes has been evaluated by Qin et al. (93/3289). The system was tested for parameters such as LODs precision and memory effects. A high efficiency high solids nebulizer has been reported by Meyer (93/C1418). It has been used to determine metals in a variety of matrices including organic solvents pharmaceutical preparations and urine. A comparison of performance with a conventional concentric nebulizer demonstrated that it gave improved LODs.Ilic et al. (93/834) used a cross-flow nebulizer to determine U in different organic solvents. The LODs for the U at the 385.958 nm line were 0.3 1.2 and 1.1 pgrn1-l in water xylene and kerosene respectively. Wear metals in lub- ricating oils have been determined by on-line dilution withJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 179R kerosene (92/4606). Using a V-groove nebulizer particle sizes of up to 10 pm could be tolerated. A V-groove nebulizer and a heated spray chamber have been used to determine metals in molten waxes (94/572).Standard additions was used as the method of calibration. The dynamic range was 0-1Opg g-l. Detection limits were 0.31 0.56 0.47 and 0.87 pg g-' for Cd Cu Fe and V respectively. Direct injection nebulizers (DINS) are still receiving attention. A DIN has been used by Wiederin and Pinkson to determine trace metals in H F samples (93/C1420). A volume of 20 p1 was necessary to provide a steady signal. The aerosol particle size distribution from a DIN has been measured by Wiederin and Houk (93/2119). Photographs and laser scattering data have shown that the DIN produced a finer aerosol with a narrower drop-size distribution than a conventional concentric nebulizer operating with the same flow rates. However the drops were larger than those emanating from a concentric nebulizer coupled with a Scott-type double-pass spray chamber.Shepherd et al. (93/C1453) have used a DIN to introduce volatile hydrocarbon samples into the plasma. It was found that this overcame the problem of excess solvent loading of the plasma giving different results compared with those involving calibration against non-volatile solutions. The effect of surfactants on the analytical performance of an ICP has been evaluated by Bertagnolli et al. (93/3193). Triton X-100 UltraWet 60L and BRIJ-35 were selected to be studied and the analytes of interest were Cu Fe and Mn. The three surfactants had similar effects on analyte transport (k it was improved) but it was found that the use of Triton X-100 led to plugging of the cross-flow nebulizer.Various acids (nitric hydrochloric and sulfuric) and 1 1 mixtures of them caused signal depression (93/3469). This was attributed to a difference in nebulization efficiency caused by a different viscosity surface tension and droplet size distribution. Pulse nebulization has been used to introduce micro-volume samples (93/C1620). Nebulization times of 0.02-10 s and volumes of 2-1000 pl were tested. The precision of the analytical measurements was claimed to be 0.02-0.05% and increased as volume decreased. Several papers comparing nebulizers have also been pro- duced. Droplet size measurements of various nebulizers have been made by Zarrin et al. (93/2411). Electrospray thermospray and pneumatic nebulizers were all studied to determine whether the liquid composition and flow rate had any effect on the droplet size.Browner et al. (93/3962) have used laser scattering on several different nebulizers to study the effect of analyte and solvent transport efficiencies on signal intensity. For most of the nebulizers tested there was a linear relationship between the analyte mass transport and the net intensity of the signal. Several theoretical papers on nebulizers have also been produced although most of them have been conference presen- tations. Monte Carlo techniques have been applied to the simulation of the generation the losses and size distribution of aerosols (93/C3048). These parameters have then been used to calculate the mass transfer rate which in turn allows optimization of the working conditions. Another model has been produced which described the effects of aerosol particle size on discrete sample introduction (93/C1528).The effects of organic solvents on signal intensity in ICP-AES have been discussed by Xin (93/C3071). It was found that increased nebulization efficiency led to increased signal and that this increased efficiency was a function of viscosity density and surface tension of the solvent. Bochert and Dannecker (94/854) have used ICP-AES to analyse single aerosol particles in monodisperse test aerosols. High precision (geometric standard deviation 1.02) LODs in the femtogram range and linear calibrations for particles in the 1-10pm range were all obtained. Spray chambers and desolvation devices have been studied more in this review period than they have for some time.Wu and Hieftje (93/3424) have developed a new spray chamber for ICP spectrometry that produced a 30 YO increase in sample utilization efficiency and 2-3 times less sample clean-out time at half the cost of a conventional Scott type double-pass spray chamber. In addition it also improved S/B ratios LODs and precision. Another spray chamber has been developed by Gregoire et a!. (93/1067). This was a modified Scott type that allowed change from simple liquid nebulization to electro- thermal vaporization or LA without the need for plasma shut-down. Losses due to condensation on curved or irregular surfaces were minimized because of the straight-line gas flow geometry. 2.2.2. Flow injection As ever there have been a number of reviews of this popular method of sample introduction. A literature survey of on-line preconcentration techniques containing 82 refs.has been made by Carbonell et al. (93/771). Liquid-liquid extraction column methods and precipitation were discussed. Tyson continues to produce reviews in this area (93/861 379 refs.) and in another paper he discussed recent and future developments for improving precision and accuracy (93/273 1). The use of micro-columns of ion-exchange or chelation media is still proving popular as a means of preconcentrating analytes prior to their determination. Schramel et al. (93/3114) used a commercial preconcentration system with EDTA-cellulose as a packing to preconcentrate Cr Cu Fe Mn Ni V and Zn in biological and environmental materials. Detection limits were improved by at least 7-fold compared with continuous nebuliz- ation.Precision was typically 0.5-1.5% RSD. Caroli et al. (92/2515) used iminodiacetic acid-ethylcellulose micro- columns to preconcentrate Cd Co Cu and Pb from waters and urine. An acetate buffer (pH 5.5; 2 mol 1-') was used to retain the analytes on the column and nitric acid (2 mol 1-l) was used to elute them. Detection limits were improved by at least an order of magnitude for Cd and Pb. Throughput was 10-12 samples h -I. 8-Hydroxyquinoline (oxine) or its deriva- tives has also been used. Schramel et al. (94/736) used oxine- cellulose to preconcentrate 11 analytes (Al Cd Co Cr Cu Fe Mn Ni Pb V and Zn). Parameters such as the pH of the mobile phase and the concentration of acid eluent were optim- ized. The effects of potential interferences were examined and CRMs analysed.The majority of results were satisfactory and possible reasons for the exceptions (Cr and Fe) were given. Peng et al. (93/C3074) used oxine loaded on a mixture of activated carbon and silica gel to preconcentrate Al Cd Cu Fe and Mn and separate them from potentially interfering matrix ions. A desolvation system was also used to further increase the sensitivity. The LODs were at the ngml-' level and precision (n=6) was in the range 2.6-4.3% RSD for a 120 p1 injection. Anion exchange micro-columns have also been used. Israel et al. (94/569) used an automated ion-exchange device to separate trace constituents from matrix alkali metal salts. Arsenic Cr Mo S and Se have been preconcentrated on 'resin DZ96) by Liu et al.(94/721). The paper described an FI system and optimized the working parameters. Sample throughput was 14 h-' and LODs were 13.0 3.64 2.04 6.19 and 25.2 ng ml-' for As Cr Mo S and Se respectively. Basic alumina has been used by Yamada et a!. (93/1012) to enrich S (as sulfate) from iron samples and separate it from the matrix. Interferences were evaluated and CRMs analysed. The LOD was 0.3 pg of sulfur per gram of iron. Gomez and McLeod (94/568) enriched Au on columns of sulfydryl cotton with KCN as eluent and on Amberlyst A26 resin with NH as eluent. For a sample volume of 10 ml and an eluent volume of 250 pl preconcentration factors of 50 and 40 were obtained on the sulfydryl cotton and Amberlyst respectively. The LOD was 1 pg 1-' for both columns and the precision at the 50 pg I-' level was 1.2-3.5%.As an application Au in waste water was determined. Gold has also been preconcentrated by Guo and Tang (93/C3060). These authors used 'Levextrel' resin180R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 and an eluent of 2.5% thiourea. A 20-fold enhancement factor was achieved with a sampling frequency of 60 h-'. Preconcentration of mercury using complexation of the metal ion with 1,5 bis (di-2-pyridy1)methylene thiocarbonohydrazide followed by on-line extraction into IBMK has been reported by Canada-Rudner et al. (94/579). Vapour generation of the preconcentrated mercury was achieved by mixing with SnC1 in DMF. Potential interferences were evaluated and optimum conditions described.The LOD was 2 ng ml-' the calibration was linear between 10 ng ml-' and 5 pg ml-' and the precision was 3.3% RSD. One of the more novel methods of analyte preconcentration prior to atomic spectrometric determination has been devel- oped by Pretty and Caruso (93/3222). These workers used on- line anodic stripping voltammetry to preconcentrate Pb and T1 by factors of over 60. The technique was reported as being rapid and gave acceptable precision. Electrochemical preconcen- tration has been described by Deng et al. (93/C3055). Analytes were deposited on a mercury plated glassy carbon electrode in a flow-through cell and then released for detection by electro- lytic dissolution. Electrolytic dissolution has also been used by the same group of workers for the multi-element analysis of aluminium alloys (92/2762).The electrolyte used was 1 moll-' HN03 and the electrolysis was at 1150 mA cm- for 20s. Chromium Cu Fe Mg Mn Si and Zn were the analytes determined. Results for the analysis of CRMs agreed well with the certified values. Donnan dialysis has continued to be used by some workers (93/3438 and 93/C1421). This technique has provided precon- centration factors of over 200 for an 8 min dialysis period allowing p1 I-' levels to be determined. Detection limits may be improved further if the dialysis time the final solution temperature and the length of the cation-exchange tubing are increased. Dilution by dialysis has been used to determine Ca K Mg and Na in wines (92/2645). The linear ranges for the method (in pg ml-') were 7-120 for Ca 9-1500 for K 3-180 for Mg and 5-300 for Na.Sampling rates were 120-150 h-' and precision was less than 4% RSD. On-line digestion has again been studied by some workers (94/578). In a continuation of previous work slurried samples were transported to a microwave oven nitric acid added and the system sealed prior to exposure to microwave radiation for 5 min. The sample was then diluted to volume and analysed. Analysis of CRMs yielded good recoveries for all samples except coal. This was attributed to incomplete dissolution. On-line dilutions have been made using computer-controlled hardware to analyse minor constituents of nitric acid-per- chloric acid digestions of plant materials (93/2051). In addition standard additions analyses were also achieved by merging the sample zone with an aliquot delivered from a trapped standard zone.Analysis of plant RMs gave results in agreement with the certificate values. Hill's group determined impurities in organometallic com- pounds such as trimethylgallium and methyllithium by FI-ICP-MS and AES (93/2056). The sample (10-25 pl) was injected into 2% HNO desolvated using a membrane drier tube and admitted to the plasma. Recoveries for Al Cu In Pb and Zn ranged from 92.6 to 107.6%. The effect of matrix acid on the determination of transition metals by FI-ICP- AES was evaluated by Chen et al. (93/3555). Emission sensi- tivity was reduced by the presence of some of the acids (HCl HNO and H2S04) and by some metals (Ca Fe Na and Zn) but electron number density and excitation temperature were unaffected.A micro-computer controlled refractor plate has been used to improve background correction FI transient signals (93/1961). 2.2.3. Chromatography The coupling of chromatographic techniques with ICP-AES has been discussed in general reviews by Chan (92/4590) (68 refs.) and by Hill et al. (93/3219) (127 refs.). Jinno (93/3406) has reviewed the coupling of SFC to ICP-AES (35 refs.). Compared with previous years there have been fewer publi- cations possibly because for speciation work the development effort is concentrating upon ICP-MS. Reports concerning element speciation for this review period focus upon As and were typically based upon the HPLC separation of the species followed by HG (94/483 94/596). Detection limits for As"' AsV monomethylarsonic acid and dimethylarsinic acid were quoted at the sub ng ml-' level.Unfortunately conventional HG is not appropriate for import- ant arsenic species such as arsenobetaine which do not reduce to a volatile form. Rubis et al. (93/C3035) have described the preliminary development of a UV photolysis method for the degradation of arsenobetaine and arsenocholine which may provide a relatively straightforward adaptation of these methods for these compounds. Ion chromatography has been used to separate impurities from molybdenum and tungsten matrices (94/733) impurities in uranium oxide (93/3337) and for preconcentration of analyte elements in samples from biological matrices (93/2591) prior to determination by ICP-AES. 2.2.4. Electrothermal vaporization A review (131 refs.) of ETV into ICPs has been presented by Carey and Caruso (94/827).The advantages of the technique viz ( i ) separating the analytes from the matrix and (ii) removing the solvent which allows the plasma to use more energy for atomization ionization and excitation and (iii) the current state of the technology and possible future directions were all discussed. The use of chemical modifiers to assist in the volatilization of analytes is still proving to be a popular technique. The most popular of the modifiers still appears to be fluorinating agents such as polytetrafluoroethylene. A slurry of this modifier has continued to be used by Bin et al. (93/C3073) to assist in the vaporization of refractory analytes such as Nb Ta U and Zr. Calibrations covered 3 orders of magnitude no memory effects were observed and absolute LODs were 50 90 150 and 16 pg for Nb Ta U and Zr respectively.The same workers have presented a paper that employed a similar technique for determining Ti and V in coal (93/3533). The LODs were 0.8 and 1.5 ngml-' and precision was 1.9 and 2.8% RSD for Ti and V respectively. The procedure was validated by the analysis of certified coal (NIST 1635). Jiang (93/C3068) has again used this technique to improve LODs reduce matrix effects and determine B Cr Mo Ti and V in environmental samples. Huang et al. (93/2240) used the same method to determine 16 REEs. Detection limits were 0.1 ng-1 pg. A number of different modifiers have been evaluated by Nickel et al. (93/3367) for the analysis of ceramic powders.The modifiers included different combinations of KF (C2F4)" Na2B407 BaCO Ba(NO& BaO AgC1 CoF and Pb(B03)2. Techniques such as electron microscopy and X-ray analysis were used to evaluate the efficiency of the modifiers. The mixed modifier BaO and CoF (1 1) helped to volatilize all the impurities. Tungsten vaporizers continue to be used. Okamoto et al. (93/3359) determined Cd and Pb in biological and environmen- tal samples using a modified commercial atomizer and diam- monium hydrogen phosphate as a chemical modifier. Detection limits were quoted as 0.28 and 2.8 ngml-' for Cd and Pb respectively. The calibrations for both elements were linear up to 10 pg ml-' and precision was <8% RSD. Mei et al. (93/3138) have used a tungsten coil atomizer to analyse rice digests for Co Cr Cu Fe Mg Mn and Ni.In another paper (93/3242) the authors use this technique to analyse rice for the same elements plus eight REEs. Detection limits for the REEs were 10-9-10-11 g which exceed those obtained byJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 181 R simple ETAAS and conventional ICP-AES. Precision was typically <6% RSD. Several papers have been published on the more theoretical or mechanistic aspects of ETV-ICP-AES. Verrept et al. (94/753) used Cd Cu and Pb to investigate several of the parameters that may potentially affect the results. Observation height was independent of the sample matrix but the matrix affected the optimal carrier gas flow rate greatly. The forward power did not influence the signal but the higher the power the more matrix loading was permissible.Transport efficiencies for pneu- matic nebulization liquid ETV and solid sampling ETV were also compared. Efficiency was found to be increased for the ETV techniques in comparison with nebulization but there was no difference between the two types of ETV. Byrne et al. (93/3231) have investigated the mechanism of sodium chloride interference on Mn determinations by ETV-ICP-MS. It was concluded that it was a vapour phase interference caused by the formation of MnC1 during atomization. The addition of ascorbic acid increased the volatility of the chloride and thus facilitated interference free determination of the Mn. The interest in ETV as a method of sample introduction means that modifications of old or the manufacture of new instrumentation is fairly common.Golloch et al. (93/3504) developed a vaporizer that had a maximum of 100 heating stages and could reach temperatures of up to 2800 K. Verrept et al. (93/C1367) have modified a commercial atomizer to transport the analytes to the ICP. They also optimized the operating parameters and analysed CRMs for Cd Cu and Pb. Ren and Salin (93/3407) have also modified a commercial vaporizer. By adding a sheath gas between the analyte flow and the wall of the tube analyte condensation decreased transport efficiency increased and matrix effects were reduced. With 10 pl sample injections LODs for Cd Cu Mn Pb and Zn were 1-6 ppb; precision was 3 to 6% RSD. The same workers have also discussed the relative merits of ETV-ICP- AES and described the determination of Pb in soils (93/C1406).The standard additions method of calibration for solid sampling ETV-ICP-AES has been described by Boonen et al. (94/580). Analysis of biological and environmental CRMs for Cd Cu and Pb yielded results in good agreement with the certified values. 2.2.5. Solid sampling procedures Interest in laser ablation solid sampling for ICP-AES continues to wane despite its increasing use as a sampling tool for ICP-MS. However reports have concentrated on using the advantages of laser ablation namely rapid analysis spatial resolution and the ability to sample materials that are difficult to dissolve. Moenke-Blankenburg (93/3713) provided a com- prehensive review (122 references) of laser ablation as a solid sampling mechanism for ICP-AES and ICP-MS.The same group (92/4628) also conducted a detailed comparison of the analysis of jiuorophosphate glasses by laser ablation-ICP- AES/MS conventional dissolution followed by ICP-AES and classical methods. They concluded that the accuracy and precision of the methods were comparable. Xu and Tian (93/C3065) studied the ablation of Zr in reference materials using three different laser systems. (i) A conventional pulsed Nd YAG using either free running or Q-switched mode. Problems were found with mineral zircon (ZrSiO,) which gave lower values for zirconium than chemically produced ZrO in a standard. Q-switched mode was observed to be more accurate and precise. (ii) A dual laser system with a CO laser fusing the sample with a flux followed by a pulsed Nd YAG laser for ablative sampling.Good results were obtained but some problems with deposits on the ICP torch were found. (iii) A Nd YAG laser with acousto-optic Q-switching. Compared with a conventional Q-switched laser the acousto-optic method produces a higher repetition rate (1-5 kHz to 1-20 Hz) lower pulse power (104-106 W) and a longer duration of pulse ( 10-4-10-7 to lo-' ns). The authors were confident that this system fulfilled all their requirements. Koskelo and Cremers (93/C1566) suggested that the major advantage of LA-ICP- AES analysis of metal particles on air-filters was one of speed and efficiency. The analysis of a complete 37 mm diameter filter in less than thirty seconds rather than over an hour for a conventional digestion and analysis could be very important for worker health and safety. Optimization of instrumentation detection limits precision and the use of carbon from the ablated filter medium as an internal standard were discussed. Thompson et al.(94/940) performed a geochemical reconnais- sance suruey of a mineralized area in NE Wales using a multi- elemental data set produced by the LA-ICP-AES analysis of stream-sediment pebble coatings. These manganese and iron oxide coatings are known to provide a natural preconcentration mechanism for both base metals and pathfinder elements. The survey revealed anomalous concentrations of As Cd Co Cu Mo Pb and Zn delineating the copper mineralized Coed-y- Brenin diorite and Dolgellau gold belt. Problems associated with LA include finding suitable solid standards and matrix effects associated with the LA process.Mochizuki et al. (93/706) used a novelfour tube ICP torch so as to introduce both laser-ablated solids and nebulized solu- tions for calibration purposes into the same bulk plasma environment. Water was aspirated while the laser-ablated material was introduced. Differences in transport efficiency were compensated for by internal standardization to a matrix element. Selective vaporization was minimized by using a Q-switched laser with a high peak power. The method was applied to low-alloy steel reference materials and also aluminium- and titanium-based alloys. Chan et al. (93/1948) observed matrix effects when laser ablating high temperature superconductors such as Bi-Sr-Ca-Cu-0.Both KrF excimer picosecond and Nd-YAG nanosecond lasers with ICP-AES photodiode array detection were used. There was an enrichment of volatile Bi,03 and CuO in the vapour phase but melted droplets and the ablation crater showed enrichment of Ca and Sr. Thompson et al. (93/C1602) commented on the particles produced by LA and the matrix effects possibly arising from production of particles with a different composi- tion to that of the bulk or with more than one composition. Observations were made using time resolved-ICP-AES and electron microscopy. Fewer reports on slurry nebulization have been noted with work concentrating on improving performance and overcom- ing problems. Sample preparation of solids is vitally important to quantitative elemental determinations by slurry nebuliz- ation-ICP-AES.Halicz et al. (94/570) addressed this problem by employing a 3-dimensional turbulent mixer mill with the popular zirconia bead and bottle method to produce micron and sub-micron particles of geological test materials. Using a V-groove type nebulizer and a Trassy-Mermet type torch system major and trace elements were determined using aqueous standards. In some cases incomplete recoveries were observed and factors were calculated to correct for this. Alternatively matrix-matched geological RMs were used to overcome this. There has been much interest recently for all forms of argon ICP in the addition of molecular gases to improve performance. Ebdon and Goodall (93/2086) used hydrogen to aid volatilization of solid particles introduced by slurry nebulization and they observed an improvement in accuracy which corresponded to an elimination of interferences when analysing highly refractory particles.The increase in temperature from 2200 to 3900K was attributed to an increased energy transfer from the toroidaI to the annular region of the plasma. This was considered to be a consequence of the higher thermal conductivity of hydrogen. Lobinski et al. (93/773) demonstrated the analysis of refractory ZrO,. Calibration was performed by standard additions with detection limits between 0.03 to 10 pg g-' obtained for 11 elements. Results were shown182R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 to be comparable with those produced by digestion if the particle size was kept below 10 pm.Coedo et al. (94/598) continuing from work reported last year compared the determination of boron in reference steels by spark ablation sampling with microwave digestion followed by pneumatic nebulization. Both methods were optimized for the analyte and matrix accuracy. was satisfactory for both methods but spark ablation proved superior in terms of precision (0.5-1% RSD compared with 1-3.5% RSD) and detection limit (0.65 pg 8-l compared with 2.6 pg g-I). Uchida (94/700) used r.f. sputtering to solid sample doped lead zircon- ate titanates. The material was sputtered onto a quartz plate then dissolved off using mineral acids. Analytical results agreed well with original target composition and the precision was better than 8% for trace elements. Blain and Salin (93/540) using a simpler approach directly introduced pellets of sediment reference material mixed with graphite in a 1 5 ratio into an ICP-AES instrument.This technique gave sharp intense peaks for many environmentally sensitive elements including As Cd Hg Pb Se and Sn. Vaporization of refractory elements was facilitated by addition of a AgCl matrix modifier. The authors further explored the use of this matrix modifier (93/3249) with standard additions being used for calibration. Metal oxide powders were found to be suitable for spiking for Co Cr Ni and V. Accuracy and reproducibility were adequate for trace element analysis but less suitable for high concentrations. Fujimoto et al. (93/3197) investigated the use of a directly inserted thin stemmed graphite cup.The system was designed with a low thermal capacity and a high thermal conductivity for the analysis of trace elements in organic and aluminium nitride matrices. Calibration curves were obtained using aqueous solutions. To improve the sensitivity of refractory or carbide-forming elements the halogenating agents freon or hydrochloric acid were added to the carrier gas. Suganuma et al. (94/908) used a similar system to analyse micro amounts of a glass for Na. Developments in other solid sampling techniques have included a high-energy plasma gun that operates in a low- volume chamber and allows rapid transport of ablated material to an ICP-AES instrument. The transport properties of this system were investigated by McKinstry and Goldberg (93/C1417) and some initial quantitative determinations on refractory reference materials made.Wang and Jia (92/2965) devised a thermo-chemical distillation method for treating solid pulverized river sediment with a flux/reactant and transporting the resulting vapours to an ICP-AES intstrument. Detec- tion limits at the ppt level were claimed for environmentally sensitive elements such as As Cd and T1. 2.2.6. Chemical vapour generation A critical review of hydride generation (134 references) focus- ing upon As and Se has been published (93/641). Sahayam et al. (94/621) have described a modified sample introduction system for hydride generation using a co-axial jet that gave a modest improvement in signal intensities for Bi. A generally accepted convention for hydride generation is that because of the differing chemistries for production of the volatile species it is normally a single element technique.The development of simultaneous techniques for the determination of hydride forming elements has been described by various workers (92/2644 93/686 93/689 93/C3067 94/300). In all of these studies attention was paid to interferences. The most active group in this field that of Sanz-Medel at Oviedo have published some interesting work using a didodecyldimethyl ammonium bromide medium. Methods were developed for As (94/325) where tolerance to interferences was improved and for Cd (94/602) using a sodium tetrahydroborate reduction. Using this approach the authors demonstrated a modest improvement in the LOD to 1 ngml-l. Other work published by this group has included the determination of Pb using a continuous-flow potassium dichromate-lactic acid system (94/597) that gave an LOD of 2 ng m1-l.Other reports concerning vapour generation included the determination of As in hair (93/3262) the determination of Ge in groundwater (93/3128) and the determination of Se ( 94/8 1 ). 2.3. Instrumentation 2.3.1. Torch and generator design There has been a slight increase in the number of papers being published in this area. A review ( 5 refs.) on ICP-AES instrumen- tation has been presented by Steiner (93/3216). Topics included were user objectives instrumentation design criteria echelle spectrometers noise studies instrument performance and sample introduction. A linear flow torch (UFT) has been evaluated by Sesi et al.(94/279). The operating conditions were simplex optimized and the torch was then compared with a conventional tangen- tial flow torch. The data obtained from the LiFT were found to be more precise have better long term stability and the torch consumed less gas and gave rise to less noise and background molecular band emission than the conventional torch. Detection limits for the two torches were comparable. Another LiFT has been described by Rayson and Shen (93/1957). The torch operated at reduced power levels and again used less argon than conventional torches. Additionally the new LiFT gave improved S/N ratios better detection limits and had a similar dynamic range but was slightly more susceptible to interferences due to calcium emission signals. Many of the torches seem to have been developed to cope more easily with the introduction of organic matrices.Lim et al. (93/3261) extended the torch and modified the injector. A cooling system enabled an improvement in S/B of 20-30% to be achieved. Detection limits in xylene were reported to be comparable to those obtained in aqueous solution although memory effects were problematic. A low-power air-argon plasma (50 50) was used with a 40.67 MHz generator by Tang et al. (93/3163). Molecular emissions arising from hydro- carbons (e.g. C2 and CN) were eliminated in such a plasma. Detection limits for some analytes in IBMK were measured and were found to be better for most atomic and some ionic lines when compared with those obtained from a pure argon ICP. The role of the auxiliary gas flow in organic sample introduction was found by Pan et al.(93/3390) to be far more significant than in aqueous nebulization. Changing the flow rate altered the distribution of the solvent loading which in turn changed the plasma excitation temperatures. A new demountable torch has been developed by Eames et al. (93/3191). It consisted of a PTFE base into which the outer and intermediate fused-silica tubes were shrink-fitted. The inner tube was co-axially aligned by means of a PTFE bush which had a three start helical U-shaped groove machined in its outer surface to increase gas swirl velocities. This had the effect of stabilizing the plasma. The silica inner tubes may readily be changed and there was an option to use alumina inserts to provide a torch that is resistant to HF.Ross et al. (92/1262) evaluated a 13 mrn torch using an ICP- mass spectrometer. The torch consumed substantially less argon than a conventional torch (9.35 1 min-’) and was found to produce sensitivities ionization temperatures oxide-ion ratios and doubly charged ratios similar to a conventional torch. Better precision was obtainable with the smaller plasma. Rayson and Shen (93/2114) have modified an ICP torch so that atomic absorption measurements could be made. A power- modulated source generated a plasma in a ‘see-through’ torch. An HCL was focused through the base of the discharge and a time-gated Boxcar averager was used for data aquisition. An end-on viewed ICP with a modified torch and optical system has been described by Yang and Nygaard (93/C3070).The i.d. of the central channel was increased to 3mm thus183R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 decreasing the sample gas velocity and increasing the analyte residence time in the observation zone. Ten-fold improvements in LODs were observed and substantially decreased chemical interferences by the alkali metals were observed. A torch that operated at very high temperature using a mixture of Ar and He has been developed by Romanosky et al. (94/913) to analyse fossil fuel process streams. The torch runs with a conventional annular flow coolant gas no auxiliary gas and the inner tube has been modified so that the nebulizer gas may be admitted at a temperature of 650°C. Multi-element detection was achieved by using several monochromators simultaneously.2.3.2. Spectrometers This has been a relatively quiet area in this review period with only 14 papers being produced. Computerized multichannel AES has been reviewed (36 refs) by Bilhorn et al. (93/4001). The characteristics of a commercial spectrometer with a charge injection device has been described in 2 papers (93/C1392 and 93/C3006). The detector was described as having similar detection limits sensitivities and linear ranges to a photomul- tiplier tube whilst having the capability of observing the background and analytical signals simultaneously. In addition the spectrum between 170 and 800 nm may be stored for later access. This will provide an invaluable aid to practising ana- lysts. In another paper (93/C1382) a CID was used as a detector for the comparison of argon and helium plasmas.The system was used in conjunction with Abel inversion to spatially map the plasmas at various forward powers gas flow rates and sample uptake rates. High quality images were produced because of the solid state stability and anti-blooming capability of the CID. A charge-coupled device (CCD) was used to facilitate real- time internal standardisation (94/593). Precision values of <0.1% RSD were obtained when a signal generated from a linear combination of signals coming from a multi-element line set was used as the reference signal for internal standardiz- ation. This spectrometer has recently become commercially available and it is likely that numerous papers will be produced in the near future.A laboratory produced photodiode array (PDA) spec- trometer has been characterized by Li et a!. (93/507) for a number of elements (Al Ba Co Cu Ga K Mn Na Pr Y and Zn). Limits of detection and some applications were assessed. In addition the PDA was used with an optical fibre probe to measure the spatial distribution of temperature in the ICP. A PDA in conjunction with an FT interferometer has been used by Clarke and Adams (93/3354) for simultaneous multi-element AES. The system was used to determine K Li and Na simultaneously at 13158 14925 and 16978 cm-I respectively in a propane-air flame. Calibrations were linear to 100ppm for each of the analytes and limits of detection were 0.2 1.0 and 1.6 ppm for K Li and Na respectively. An kchelle-based ICP spectrometer that employed twin nebulizers and spray chambers with a single torch has been described in two papers (93/3205 and 93/3460).This patented multiplexed tandem sample introduction system reduces 'dead time' and allowed the analysis of 42 samples h-l compared with 33 samples h-' for a conventional simultaneous spec- trometer. The system was reportedly extremely stable accurate and precise. Another Cchelle spectrometer has been described by Steiner (93/3216). The instrument performance noise characteristics various sample introduction methods and a sample desolvation system were all discussed. Combined ICP-MS and ICP-AES spectrometers have been described by Denton (93/C1398) and by Ayala et al. (93/C3011). In the latter paper the dynamic range of detection was extended and atom-ion equilibria were explored. The importance of wavelength positioning accuracy for multi-component analyses by ICP-AES has been described by Yang et al.(94/748). Serious errors in the estimated analyte concentration may occur if the monochromator does not position itself precisely. A theoretical approach to solving this problem was compared against experimental results. Verrept et al. (93/438) modified a commercial spectrometer by using a quartz refractor plate to measure background corrected transient signals. The computer controlled refractor plate performed scans over a small wavelength interval. The software then allowed background corrected peak areas to be calculated for electrothermal atomization and other methods that produce transient signals.The stability and the resolving power of the instrument were not impaired and the precision for Sr at 10 to 100 pg ml-' in waters was 1.2-0.3%. 2.3.3. Instrument control and chemometrics Chemometric methods for improving the resolution of spec- trometers has proved to be a popular topic in this review period with several papers being produced by Chinese workers. Zhang et al. (93/699 and 93/3286) have used a polynomial smoothing method to correct for interferences arising from severe overlapping of lines. The same authors have also used factor analysis to correct for spectral overlap (94/743). A data matrix was composed from a pure spectrum and a spectrum of the mixture. This data matrix was decomposed by target transformation factor analysis to a spectra matrix and a concentration matrix.The concentration of the component of interest in the mixture may then be obtained from the concen- tration matrix and the interference from the other component is eliminated. As an application the spectral interference exerted by Y on the determination of Cu and A1 was eliminated. Yang et al. (94/849) used Kalrnan Jiltering to resolve closely spaced lines. They found that the resolving power of the filter could be strengthened by decreasing the step size in the scans. Difference in the line profiles of the analyte ion and the interfering ions allowed the filter to resolve coincident lines more easily. The effects of wavelength positioning errors on the results obtained by Kalman filtering ICP-AES have been determined by Yang et al.(93/3422). It was found that a positioning accuracy of 0.1 pm was necessary to obtain accurate and precise measures of analyte concentration. In a related paper (93/3 166) the reliability of Kalman filtering results was evaluated using two methods the NAC criterion (which is based on auto-correlation analysis of the innovation sequence) and the innovation number. Both methods compensate for wavelength positioning errors but the NAC gives information on each individual result whereas the innovation number must have other data to perform the same task. Kalman filtering has also been used for several applications based papers. Brindle and Zheng (94/291) found an improvement in accuracy for the determination of transient signals produced by a computer-controlled hydride generator.The filter was used to remove the white noise and thus determine the signal under the noise even at a ratio of 1 :4. Thirty trace elements in uranium were determined without matrix separation because of Kalman filtering (93/3921). This would not have been possible in conventional ICP-AES. Ytterbium has been deter- mined in vanadium by Ma et al. (93/C3046). This was made possible by separating the V" 328.94 from the Yb" 328.94 line. Detection was by a photodiode array detector made in-house. Relative errors were found to be 1-3%. Errors in apparent analyte concentrations caused by wavelength positioning errors have also been investigated by the use of multi-component analysis (MCA) techniques (94/741). By assuming Gaussian line profiles a model was constructed and the results tested experimentally.In general good agreement between theoretical and experimental measurements was made. It was concluded that MCA techniques have no advantage over conventional correction methods unless they can bring about a reducion in the positioning error. Multivariate methods have also been reported by a few184R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 workers. For example Principal Components Analysis (PCA) has been used by Yates and Aries (93/C1425) to interpret multi-element analytical data. It was found that although standardization within the PCA was not required the use of standards in the data set checked the validity of the procedure. Ivaldi et al. (93/3165 and 93/C1400) used multivariate methods (the method of least squares) to enhance the information obtained from a CCD. The method of least squares smoothing has also been applied by Zhang et al. (93/431) to correct for the V 328.939 nm and Y 308.203 nm line interferences on the determination of Yb at 328.937nm and A1 at 308.215nm respectively.A comparison of Multiple Linear Regression (MLR) with a neural network has been made by Schierle and Otto (94/31). The network was trained from emission spectra from As and Cd and the concentration of both analytes were determined successfully in mixed spectra. It was concluded that close relations exist between the MLR procedure and the operation of the perceptron of the network. Otto et al. have also used a neural network in conjunction with the method of least squares to obtain qualitative determination of sixty-eight analytes (93/3112).The method yielded acceptable results although problems did occur when samples rich in tungsten were analysed. Nikdel (93/C 1374) applied a neural network named ANN (artificial neural network) to the analysis of orange juice. This network mimics the brain's problem solving process and builds a system that makes new decisions classifications and forecasts using pattern recognition and fuzzy logic. Neubock et al. (94/683) have also used fuzzy logic to achieve automated qualitative analysis. Background equivalent concentrations were used to calculate a Bayes' probability that measured the usefulness of a peak for identifying a certain element. Evaluation of the performance of the system showed it to function satisfactorily.Optimization procedures have again proved popular. A review with 4 references describing strategies for multi- parameter optimization has been presented by Noelte (93/3 104). A simplex-optimized program for the determination of temperatures in reduced pressure ICPs has been described by Turner and Fannin (93/3425). The results for several Ar-He reduced-pressure plasmas were tabulated. The optimization of analytical techniques for eliminating matrix effects has been studied by two groups of workers. Spectral interferences arising from Eu have been eliminated by Lu and Zou (93/C3072); while Sun and Zhang (93/C3051) have removed interferences caused by Al Ca Mg K Na and Zn. A program designed to facilitate optimization has been described by Borer et al.(93/1643). The algorithm enabled the net signal background S/B ratio S/B noise ratio precision of the background and precision of the signal to be calculated. Processing time was reduced and the optimization studies were easier and more informative. Several pieces of software have been developed that are designed to maximize analyte sensitivity improve precision and enhance the reliability of the determination in general. Lorber (93/C1399) achieved precision better than 5% RSD and eliminated spectral interferences using a chemometrics approach. Meanwhile Yan et a!. (93/C3054) developed some software that specialized in the detection and processing of multichannel transient signals arising from FI peaks. Huang et al. (94/668) analysed high-purity nitrogen trioxide by ICP- AES removing spectral interferences by an interference coefficient correction method.Recoveries of REEs were 85-115%. Ma et al. have also analysed a mixture of REEs using a computer spectrum stripping method (93/C3049). The method was described as simple rapid and capable of cor- recting for several types of background interference. A program designed to assist in the collection reduction and analysis of echelle spectra from a CCD has been developed by Miller and Scheeline (94/624). Wavelength calibration was better than 41 pixel across a 576x384 pixel array. A user friendly spectroscopic data system covering several techniques (e.g. IR NMR MS) has been described by Hearmon (94/638). It was based on a hierarchical network with high speed local networks and designed to accept data from a wide range of instruments.Webb and Salin (94/42) have developed a computer-assisted method for line selection. The system may be taught the composition of a sample and from this data decisions of which lines to use were made in less than one second. Instrument control procedures have been developed by several workers. A computer supported ICP-AES spectrometer was described by Bortlisz (93/3203). This instrument was capable of determining 33 analytes simultaneously. As an application the analysis of water waste water and water treatment sludges was undertaken. Drift diagnostics have been described by Carre et al. (93/2040). The behaviour of Ba" 455.403 Zn" 206.200 and Ar' 404.442 nm lines as a function of parameters such as power sample uptake rate and gas flow rate was measured.The most probable causes of drift were changes in energy transfer efficiency of sample introduction and surprisingly too long a warm up period. Recent improvements in the factors governing basic ICP stability have been discussed by Dahlquist et al. (93/C3010). These workers state that precision in older instru- ments (which was often at the 0.8% level) had different contribu- tory factors than modern instruments whose precision is 0.2% or better. The errors associated with calibration detection limits accuracy and sensitivity (quality control or quality assur- ance) have been investigated by several authors. The effect of parameters such as power observation height carrier gas flow rate and sample uptake rate on RSD was measured by Krasil'shchik et a!.(92/2561). In addition the dependence of calibration curve linearity and range and the errors of determi- nation were also investigated. The linearity of calibration curves has also been discussed by Miller (93/2705). He discussed the method of using the product moment correlation coefficient (r) the coefficient of determination (r2) and the coefficient of non- determination (1 -r2) for measuring linearity. The quality con- trol procedures used by the US Geological Survey National Water Quality Laboratory were described by Zayhowski and Bushly (93/C1427). As should be the norm calibration check standards blanks and standard reference water samples were run periodically. In addition blind quality control samples were also run.A collaborative study of ICP analysis has been reported by Ambrose and Jowitt (93/3997). In this paper the instrumental performance criteria were quoted and the compre- hensive evaluation system was described. The main source of error was the interlaboratory bias. Errors and detection limits have also been discussed from a theoretical and practical viewpoint by Prudnikov et al. (93/C1619). They investigated instrumental non-instrumental and systematic errors and esti- mated the instrumental and non-instrumental detection limits. The alloy was digested in HC1-H202 and an internal standard (La Sc or Y) added. The best internal standard was La. Analytes such as Dy Fe Mn and Tb were determined giving recoveries of 2 99.7% and precisions of between 0.04 and 0.35% for 5 replicates.The accuracy of wavelength tables has been commented on by Doidge and Nham (93/1958). In this paper several incon- sistencies in the wavelength value between different tables is highlighted. An informative 'tutorial review' on background and back- ground correction in emission spectrometry has been presented by Dawson et al. (93/3220). It includes the general principles of background correction source-generated background and its correction and instrument-generated background and its correction.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 3. MICROWAVE-INDUCED PLASMAS 185 R 3.1. Fundamental Studies Probably the most interesting paper published during this review period was that of Besner et al.(94/617) who have described the spatial distribution of excited species in an Ar surface wave plasma. These workers used tomographic recon- struction to create three dimensional distributions of the excited species in Ar Ar-0 Ar-H and Ar-N plasmas at frequencies of 27-2450 MHz and at pressures of 10-600 Torr. The work demonstrated that the radial distribution of excited atoms related to the radial distribution of the electric field the operating frequency and on the radial distribution of the electron number density. The relative influence of the electric field on excited atoms was found to be greater with increasing frequency whereas the influence upon the electron number density was less significant. It was indicated that future work would include studies of rotational temperatures excitation temperatures and electron number densities in order to model the plasma over this pressure range.Such measurements for a 2450 MHz Ar surface wave plasma at atmospheric pressure have been described by Cotrino et al. (93/2111). Spatially resolved studies of a low pressure He plasma have been described by Lei et al. (93/3284) and similar work is also being performed by Goode and Emily (93/C1555). Masamba and Winefordner (94/746) have determined excitation and rotational temperatures in a He/H CCP as a function of the concentration of H in the plasma. These authors also studied potential matrix effects caused by Na and PO,3- upon Ca emission. Boss and co-workers (93/C1387,93/C14417 93/C1551 93/C2998) have outlined their investigations of atomization and excitation mechanisms in MIPs used as chromatographic detectors. This work is particularly directed to the dissociation of hydrocarbon molecules in such devices and indicated that the typical power used to sustain a plasma is close to that required to decompose the components of a typical GC peak. Other reports included the effect of water on an He MIP (93/C1554) and the investigation of the 100-200 nm spectral region of an He MIP (93/C1552).3.2. Instrumentation Barnes et al. have evaluated a stripline source MIP and compared it against a Beenakker cavity MIP (93/2073). The stripline source was tolerant of aerosol sample introduction stable over a wide range of operating conditions and inter- ferences were less severe. Matusiewicz has described a novel combined generator/cavity (93/1649).The device was operated at 2450MHz and could use Ar He N 0 and air at atmospheric pressure. Notably this source was tolerant of up to 10 ml min-' of liquid sample in a flow of 1 1 min-' plasma support gas. The same author has also described a liquid cooled torch where erosion of the silica by the plasma was eliminated (94/745). Borer et al. (93/1953) have described a novel method for stabilizing an MIP power supply that resulted in a substantial reduction in the noise. Stabilization was achieved by adjusting the line voltage via a controllable inductance placed in series with the variable transformer at the generator input. 3.3. Sample Introduction 3.3.1. Direct nebulization Ultrasonic nebulizers have been used by several groups to introduce liquid samples into various MIP devices.Carnahan's group have studied the determination of C P and S in aqueous solution using their 1.6 kW He MIP (93/3392). Detection limits of 0.4 1.4 and 9.7 pg g-' respectively were reported. The group at Jilin University have reported the determination of P with an LOD of 0.03 pg g-' using an 80 W MIP in Ar or an LOD of 0.0045 pg g-' using He as the plasma gas (93/3279). The same group using a heating/condensing/desic- cating desolvation apparatus (94/724) achieved LODs of 0.23 pg g-' Br 0.12 pg g-' C1 and 0.06 pg g-' I (94/886). Not content with this work they have also investigated a pulsed ultrasonic nebulizer (93/3622). The nebulizer was turned off before the sample was introduced and then pulsed on.This reduced the sample volume and improved stability for samples containing high levels of dissolved solids. An evaluation of FI sample introduction for a 200 W Ar MIP has been prepared by Yolanda et al. (94/704). Ultrasonic nebulization was compared with pneumatic nebulization for a range of sample loop sizes. Using an ODS microcolumn pure ethanol as eluent and a 1 min preconcentration at 1 ml per minute the LOD for Cu in synthetic sea-water was 0.16 ng rn1-I. 3.3.2. Electrothermal vaporization In a continuation of work described above ETV has been compared with direct nebulization for sample introduction into a high power MIP (93/3393). Using emission lines in the near IR rather than the visible spectrum gave an LOD for S of 3 to 5 pg g-'. Matrix interferences from Na and K could be removed by temperature programming of the ETV device.Ali and Winefordner have described a tungsten filament vaporizer for sample introduction into a CCP (93/3146). Detection limits were 1-104 pg for a range of elements. Other reports concerned the determination of Cu Zn and Cd (94/107) and of Zn (93/3950). 3.3.3. Chemical vapour generation Bulska et al. have compared hydride generation for As Sb and Se using Ar and He plasmas operating in either a TMOlo resonator or a surfatron (94/702). Both systems were optimized using elemental Hg vapour and LODs were found to be lower using the surfatron device. The determination of mercury has been reported by Fukushi et al. (93/3465) with an LOD of 11 pg ml-' for inorganic Hg. Taketoshi and Wasa (93/3492) used a tin@) chloride reduction and an atmospheric pressure Ar plasma to achieve a 0.04 ng ml-' LOD for Hg in waste water.The continuousflow generation of halogens has been described by Camufia et al. (94/631). In a study of instrument design operating conditions and vapour generation chemistries methods for the determination of Br C1 and I were developed and the use of FIA investigated. The determination of Br in pharmaceutical materials has been described by Caldaza et al. (93/3083) using an atmospheric pressure Ar MIP surfatron and an LOD of 2pg g-' was achieved. Nakahara et al. (93/2122) have described the determination of I in sea-water with reduction to I- by ascorbic acid prior to oxidation to I using 1 mmol 1-' NaNO in 5 mmol 1-' H2S04.3.3.4. Direct analysis of solids In the percentage increase in publications for this review period this represents the most rapidly growing area of research using MIPs. Masamba et al. (93/3 190) have described an intriguing approach to the analysis of steel samples by direct insertion of a cylindrical sample into the inner tube of a two tube tangential torch installed in a CCP. Calibration was performed using NIST CRMs and LODs for Pb and Zn were 0.08 pg g-' and 5 pg g-' respectively. Ciocan et al. (93/3959 94/289) have described a laser ablation-MIP system using a low pressure Ar MIP. Applications described included the analysis of pure metals alloys glass and ceramics. Time resolved data were recovered and calibration was performed using reference materials.186R JOURNAL OF ANAL,YTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL.9 3.4. Chromatography 3.4.1. Instrumentation Riva et al. (93/3630) have compared GC-MIP with more conventional GC detectors and concluded that MIP detection could be more effective and economical for the determination of S-containing compounds in crude oil. Similar conclusions were reached by Eckert-Tilota et al. (93/817). Torch design has been studied by Alvarez-Bolainez et al. (93/2213). A water cooled capillary torch gave greater sensitivity than a tangential flow torch when using a 2450 MHz Ar plasma. Park and Yo0 (93/3468) have described a cylindrical MIP cavity that gave LODs of 0.46 pg s-' for Br and 0.51 pg s-' for S at plasma gas flow rates between 10 and 20 ml min-'. A semi-automated FI system for the on-line preparation of water samples for the determination of organotin compounds has been described by Szpunar-Lobinska et al.(94/707). The ionic organotin compounds were adsorbed onto an ODS column derivatized on-column using sodium tetraethylborate and eluted with methanol. The procedure gave an LOD of 0.1 ngml-'. Factorial design and response surface methodology have been used to optimize a surfatron GC-MIP system for the determi- nation of Br and C1 (93/2072). Detection limits of 32 pg s-' Br and 25 pg s-' C1 were obtained and were similar to those obtained with a univariate search optimization. 3.4.2. Gas chromatography-microwave induced plasma applications An extensive review of GC-MIP applications ( 140 references) was given by Bulska (92/4620).Recent developments in the technique have been discussed (28 references) by Lobinski and Adams (94/941) and the application to the analysis of pet- roleum described briefly (4 references) by Kosman (93/3975). Kovac and Ramus (93/2071) have investigated the potential for using a single calibration substance to calibrate response for a wide range of compounds. Variation in element response between the calibrant and determinand compounds was 3 to 6% depending on the analyte and the workers concluded that for some applications such errors were acceptable (see also section 3.1. above). The use of GC-MIP systems has been receiving increasing attention for the determination of metal containing compounds. Examples from Uden's group included the measurement of Al B Cr Ga Mn Re Pa Pt Ti and V in a range of synthetic compounds (93/3610 93/3503) and for Cu and Ni in metal chelates (94/859).Other work from this group concerning natural samples included a study of insoluble fractions of sediments and coals for As N 0 P S and Se using pyrolysis GC (93/2068). The application of GC-MIP in trace element speciation studies has been reported by several workers. Dirx et al. (94/775) and Gremm et al. (93/4121) have reported the determi- nation of organotin compounds in environmental samples. Liu et a!. (94/867) also described the determination of organotin species but in this study SFE was used in the sample prep- aration. Lobinski and Adams (93/3117) have described the determination of organolead compounds with a detection limit of 0.1 ng 1-' for water samples.Other work published in this review period included the determination of S in coal following SFE (93/3547) and the determination of pesticides (93/3264). 3.4.3. Supercriticul fluid chromatography Ducatte and Long have described an Bchelle-based SFC-MZP system for a range of metal and non-metal analytes (93/C1488). The advantages of a SFC-MIP-MS system for the determi- nation of halogenated compounds has been described by Olson and Caruso (93/2070). The LODs obtained using this system were 0.75 pg for Br and 15 pg for C1; reproducibility was 5%. 4. DIRECT CURRENT PLASMAS Publications citing the use of the DCP are fewer in this review echelle grating spectrometer that is normally coupled with this period. Work published included that of Brindle (93/2716) emission source has been exploited for the determination of concerning the determination of Sb by vapour generation and REEs in ores (92/2230 93/3559).Finally the DCP has been the improvement of accuracy for data from transient signals used to determine Se in snails with an LOD of using Kalman filtering (94/291). The high resolution of the 0.07 pg g-' (94/684). LOCATION OF REFERENCES The full list of references cited in this Update have been published as follows 93/998-93/C1354 J. Anal. At. Spectrom. 1993 8(3) 137R-149R. 93/C1355-93/2093 J. Anal. At. Spwtrom. 1993 8(4) 169R-194R. 9312094-9312710 J. Anal. At. Spectrom. 1993 8( 5) 239R-262R. 931271 1-93/3353 J. Anal. At. Spectrom. 1993 8(7) 313R-336R. 9313354-9314131 J. Anal. At. Spectrom.1993 8( 8 ) 377R-404R. 94/1-941614 J. Anal. At. Spectrom. 1994,9( l) 1R-23R. 94/615-941960 J. Anal. At. Spectrom. 1994 9( 2) 73R-85R. Abbreviated forms of the literature references quoted (excluding those to Conference Proceedings) are given on the following pages for the convenience of the readers. The full references names and addresses of the authors and details of the Conference presentations can be found in the appropriate issues of JAAS cited above.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 Abbreviated List of References Cited in Update 187R 9211262 Appl. Spectrosc. 1991 45 190. 9212055 Appl. Spectrosc. 1991 45 1120. 9212230 Chem. Listy 1991 85 654. 9212384 Anal. Chem. 1991 63 1600. 9212393 Anal. Chem. 1991 63 2357. 9212515 Anal. Chim.Acta 1991 248 241. 9212560 Vysokochist. Veshchestva 1991 3 219. 9212561 Vysokochist. Veshchestva 1991 3 226. 9212607 J. Anal. At. Spectrom. 1991 6 553. 9212644 At. Spectrosc. 1991 12 199. 9212645 At. Spectrosc. 1991 12 204. 9212750 Appl. Spectrosc. 1991 45 1413. 9212752 Appl. Spectrosc. 1991 45 1463. 9212753 Appl. Spectrosc. 1991 45 1468. 9212762 Anal. Chim. Acta 1991 251 187.9212965 Guangpuxue Yu Guangpu Fenxi 1991 11 36. 9214404 Kogyo Yosui 1991 395 65. 9214590 Analyst 1992 117 571. 9214598 J. Anal. At. Spectrom. 1992 7 75. 9214606 J. Anal. At. Spectrom. 1992 7 127. 9214620 J. Anal. At. Spectrom. 1992 7 201. 9214624 J. Anal. At. Spectrom. 1992 7 229. 9214628 J. Anal. At. Spectrom. 1992 7 251. 9214642 J. Anal. At. Spectrom. 1992 7 339. 931431 Spectrosc. Lett.1992 25 375. 931438 Anal. Chim. Acta 1992 257 223. 931475 Zavod. Lab. 1991 57(12) 25. 931478 Microchem. J. 1992 45 1. 931507 Guangpuxue Yu Guangpu Fenxi 1991 11 22. 931540 Spectrochim. Acta Part B 1992 47,399.931542 Spectrochim. Acta Part B 1992,47,493.93/591 Anal. Sci. 1991 7 537. 931614 Fenxi Shiyanshi 1991 10 42. 931616 Fenxi Shiyanshi 1991 10 35. 931619 Fenxi Shiyanshi 1991 10 65. 931641 Pure Appl. Chem. 1992 64 227. 931676 Fenxi Huaxue 1991,19 1141.931686 Fenxi Huaxue 1991 19 1285. 931689 Fenxi Huaxue 1991 19 1333. 931699 Fenxi Huaxue 1992 20 180. 931703 Fenxi Huaxue 1992 20 348. 931706 Bunseki Kagaku 1992 41 49. 931721 Appl. Spectrosc. 1991 45 1424. 931723 Appl. Spectrosc. 1991 45 1706. 931736 Yejin Fenxi 1991 11 30. 931771 Fresenius’ J. Anal. Chem. 1992 342 529.931773 Fresenius’ J. Anal. Chem. 1992 342 563. 931792 Zh. Prikl. Spektrosk. 1992,56 7. 931800 Zh. Anal. Khim. 1991 46 2447. 931808 Anal. Chem. 1992 64 672. 931817 J. Chromatogr. 1992 591 313. 931834 J. Radioanal. Nucl. Chem. 1992 158 23. 931861 Spectrochim. Acta Rev. 1991 14 169. 931862 Spectroscopy (Eugene Oreg.) 1992 7 37. 9311012 J. Anal. At. Spectrom. 1992 7 661. 9311019 J. Anal. At. Spectrom. 1992 7 707. 9311067 Can. J. Appl. Spectrosc. 1992 37 115. 9311165 IEEE Trans. Plasma Sci. 1991 19 1090. 9311641 Spectrochim. Acta Part B 1992 47 1107. 9311643 Spectrochim. Acta Part B 1992 47 1135. 9311646 Spectrochim. Acta Part B 1992 47 1173. 9311648 Spectrochim. Acta Part B 1992,47 1203.931 1649 Spectrochim. Acta Part B 1992 47 1221. 9311650 Spectrochim. Acta Part B 1992 47 1229.9311937 Appl. Opt. 1992 31 6547. 9311945 Appl. Spectrosc. 1992 46 749. 9311948 Appl. Spectrose. 1992 46 1025. 9311951 Appl. Spectrosc. 1992 46 1117. 9311952 Appl. Spectrosc. 1992,46 1134.9311953 Appl. Spectrosc. 1992 46 1162. 9311954 Appl. Spectrosc. 1992 46 1217. 9311957 Appl. Spectrosc. 1992,46 1245. 9311958 Appl. Spectrosc. 1992 46 1301. 9311961 Appl. Spectrosc. 1992 46 864. 9312040 J. Anal. At. Spectrom. 1992 7 791. 9312041 J. Anal. At. Spectrom. 1992 7 799. 9312042 J. Anal. At. Spectrom. 1992 7 807. 9312043 J. Anal. At. Spectrom. 1992 7 813. 9312044 J. Anal. At. Spectrom. 1992 7 819. 9312045 J. Anal. At. Spectrom. 1992 7 825. 9312046 J. Anal. At. Spectrom. 1992 7 833. 9312046 J. Anal. At. Spectrom. 1992 7 833. 9312047 J. Anal. At. Spectrom.1992 7 839. 9312051 J. Anal. At. Spectrom. 1992 7 865. 9312056 J. Anal. At. Spectrom. 1992 7 895. 9312057 J. Anal. At. Spectrom. 1992 7 899. 9312068 J. Anal. At. Spectrom. 1992 7 979. 9312070 J. Anal. At. Spectrom. 1992 7 993. 9312071 J. Anal. At. Spectrom. 1992 7 999. 9312072 J. Anal. At. Spectrom. 1992 7 1007. 9312073 J. Anal. At. Spectrom. 1992 7 1013. 9312075 J. Anal. At. Spectrom. 1992 7 1029. 9312084 J. Anal. At. Spectrom. 1992 7 1099. 9312085 J. Anal. At. Spectrom. 1992,7 1103. 9312086 J. Anal. At. Spectrom. 1992 7 1111. 93/2088 J. Anal. At. Spectrom. 1992 7 1121. 9312111 Spectrochim. Acta Part B 1992 47 425. 9312114 Spectrochim. Acta Part B 1992 47 553. 9312119 Appl. Spectrosc. 1991 45 1408. 9312120 Appl. Spectrosc. 1991 45 1419. 9312122 Appl. Spectrosc. 1991 45 1561.9312123 Appl. Spectrosc. 1991 45 1568. 9312136 Yankuang Ceshi 1991 10 50. 9312170 Fenxi Shiyanshi 1992 11 38. 9312208 At. Spectrosc. 1992 13 99. 9312209 Anal. Chem. 1991 63 1607. 9312211 Anal. Chem. 1991 63 1933. 9312213 Anal Chem. 1992 64 541. 9312240 Fenxi Huaxue 1992 20 287. 9312411 J. Aerosol. Sci. 1991 22 S343. 9312590 Spectrochim. Acta Part B 1992 47 1259. 9312591 LC-GC 1991 9 704 707 710 712. 9312601 Vysokochist. Veshchestva 1991 6 168. 9312630 Report 1990 ISAL-90-0069 ETN-91-98894; Order No. N91-19915. 9312642 Zauod. Lab. 1991,57,37.93/2666 Ann. Phys. (Paris) 1991,16 63. 9312686 J. Lubr. Eng. 1990 46 173. 9312689 LaborPraxis 1992 16 21 26. 9312705 Spectrosc. Int. 1991 3 41. 9312711 Appl. Opt. 1992 31 6519. 9312716 Analyst 1992 117 1603. 9312728 Analyst 1993 118 193.9312731 Anal. Proc. 1992 29 436. 9312878 Appl. Opt. 1993 32 867. 9312879 Appl. Opt. 1993 32 899. 9312882 Appl. Opt. 1993 32 939. 9312944 Hyperfine Interact. 1992 74 269. 9312988 Spectroscopy (Eugene Oreg.) 1992 7 12 14 18. 9313083 Talanta 1992 39 341. 9313104 LaborPraxis 1992,16,332,334,338.93/3112 Fresenius’ J. Anal. Chem. 1992 343 561. 9313114 Mikrochim. Acta 1992 106 191. 9313115 Mikrochim. Acta 1992 108 1. 9313117 Anal. Chim. Acta 1992 262 285. 9313122 Talanta 1992 39 967. 9313125 Mikrochim. Acta 1992 107 319. 9313128 Fenxi Ceshi Tongbao 1992 11 64. 9313132 Lihua Jianyan Huaxue Fence 1992 28 282. 9313138 Fenxi Huaxue 1992 20 932. 9313146 Anal. Chim. Acta 1992 264 319. 9313153 Appl. Spectrosc. 1992 46 1382. 9313154 Appl. Spectrosc.1992 46 1525. 9313163 Spectrochim. Acta Part B 1992 47B 1353. 9313165 Spectrochim. Acta Part B 1992,47 1361.9313166 Spectrochim. Acta Part B 1992 47 1055. 9313189 Spectrosc. Lett. 1991 24 1173. 9313190 Appl. Spectrosc. 1992 46 1741. 9313191 Appl. Spectrosc. 1992 46 1745. 9313193 At. Spectrosc. 1993 14 4. 9313197 Bunseki Kagaku 1992 41 609. 9313202 Zh. Anal. K him. 1992 47 1378. 9313203 Gewaesserschutz Wasser Abwasser 1991 124 89. 9313205 Int. Lab. 1993 23 25 28. 9313216 Analusis 1992,20 M38.9313219 J. Anal. At. Spectrom. 1993 8 499. 9313220 J. Anal. At. Spectrom. 1993 8 517. 9313221 J. Anal. At. Spectrom. 1993 8 539. 9313222 J. Anal. At. Spectrom. 1993 8 545. 9313231 J. Anal. At. Spectrom. 1993 8 599. 9313242 Fresenius’ J. Anal. Chem. 1992 344 54. 9313247 Spectrochim.Acta Part B 1992 47 1373. 9313248 Spectrochim. Acta Part B 1992,47 1389.9313249 Spectrochim. Acta Part B 1992 47 1471. 9313254 Anal. Chim. Acta 1992 269 123. 9313257 Anal. Lett. 1992 25 2143. 9313261 Anal. Sci. Technol. 1991 4 267. 9313262 Anal. Sci. Technol. 1992 5 51. 9313263 Anal. Sci. Technol. 1992 5 185. 9313264 Analusis 1992 20 S9. 9313267 Appl. Spectrosc. 1992,46 1597. 9313276 Diss. Abstr. Int. B 1992 52 3576. 9313279 Fenxi Huaxue 1992,20 1065. 9313281 Fenxi Shiyanshi 1992 11 36. 9313284 Guangxue Xuebao 1992,12,491.93/3285 Guangpuxue Yu Guangpu Fenxi 1992 12 57. 9313286 Guangpuxue Yu Guangpu Fenxi 1992,12,63.93/3289 Guangpuxue Yu Guangpu Fenxi 1992 12 123. 9313297 Khim. Tekhnol. Vody 1992 14 740. 9313310 J. Opt. SOC. Am. B Opt. Phys. 1992 9 1979. 9313319 Nippon Seramikkusu Kyokai Gakujutsu Ronbunshi 1992 100 1184.9313320 Opt. Laser Technol. 1992 24 273. 9313330 Rev. Roum. Phys. 1992 37 31. 9313336 Yankuang Ceshi 1992,11,16.93/3337 Yuanzineng Kexue Jishu 1991,25 8. 9313354 Analyst 1993 118 229. 9313357 Anal. Sci. 1993 9 83. 9313359 Anal. Sci. 1993,9 105. 9313367 Spectrochim. Acta Part B 1993 48 25. 9313370 Spectrochim. Acta Part B 1993 48 65. 9313390 J. Anal. At. Spectrom. 1992 7 1231. 9313391 J. Anal. At. Spectrom. 1992 7 1243. 9313392 J. Anal. At.188R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY JUNE 1994 VOL. 9 Spectrom. 1992 7 1249. 9313393 J. Anal. At. Spectrom. 1992 7 1253. 9313406 J. Anal. At. Spectrom. 1993 8 51. 9313407 J. Anal. At. Spectrom. 1993 8 59. 9313421 Appl. Spectrosc. 1992 46 1762. 9313422 Appl.Spectrosc. 1992 46 1816. 9313424 Appl. Spectrosc. 1992 46 1912. 9313425 Appl. Spectrosc. 1992 46 1929. 9313433 Anal. Chem. 1993 65 735. 9313437 Anal. Chem. 1993,65,778.93/3438 Anal. Chem. 1993 65 857. 9313460 Am. Lab. (Shelton Conn.) 1992 24( 18) 205 20L. 9313465 Anal. Lett. 1993 26 325. 9313468 Anal. Sci. Technol. 1992 5 263. 9313469 Anal. Sci. Technol. 1992 5 277.9313471 Annu. Tech. Con5 Proc.-Soc. Vac. Coaters 1992 (35) 319. 9313492 Chem. Express 1993 8 13. 9313503 Chromatographia 1992 34( 5-8) 269. 9313504 CLB Chem. Labor Biotech. 1993,44,75,79.93/3514 Crit. Rev. Anal. Chem. 1992 23 143. 9313533 Fenxi Huaxue 1993 21 6. 9313547 Fuel 1993 72 225. 9313555 Guangpuxue Yu Guangpu Fenxi 1992 12 49. 9313559 Guangpuxue Yu Guangpu Fenxi 1992 12 71.9313574 Huadong Huagong Xueyuan Xuebao 1992 18 268.9313601 J. Appl. Phys. 1993 73 948. 9313602 J. Appl. Phys. 1993 73 1091. 9313608 J. Electrochem. SOC. 1993 140 505. 9313610 J. High Resolut. Chromatogr. 1992 15 669. 9313620 J. Trace Microprobe Tech. 1992 10 207. 9313622 Jilin Daxue Ziran Kexue Xuebao 1992 2 106. 9313630 Lab. 2000 1991 5 34. 9313656 Opt. Spektrosk. 1992 73 226. 9313659 Braz. Pedido PI BR 91 04,352 (Cl. GOlN30/24) 10 Mar 1992 Appl. 91/4 352 03 Oct 1991; 6pp.9313660 Teljes HU 60,542 (Cl. GOlN21/74) 28 Sep 1992 Appl. 89/2,336 10 May 1989; 13 pp. 9313686 Rev. Sci. Instrum. 1993 64 265. 9313713 Spectrochim. Acta Rev. 1993 15 1. 9313766 Thin Solid Films 1992 220 295. 9313779 Trends Anal. Chem. 1993 12 18. 9313788 Yuanzi Yu Fenzi Wuli Xuebao 1992 9 2346. 9313794 Zenkoku Kogaiken Kaishi 1992 17 133.9313921 Anal. Chem. 1992,64( 15) 1643. 9313950 Fenxi Huaxue 1992 20 535. 9313959 Spectrochim. Acta Part B 1992 47 611. 9313962 Spectrochim. Acta Part B 1992 47 659. 9313967 Spectrochim. Acta Part B 1992 47 741. 9313972 Zh. Anal. Khim. 1992,47 312.9313975 Am. Lab. (Shelton Conn.) 1992 24,28T 28V 28X-Z. 9313990 Comm. Eur. Communities [Rep.] EUR 1992 EUR 14113 Prog. Anal. Chem. Iron Steel Ind 150. 9313993 Comm. Eur. Communities [Rep.] EUR 1992 EUR 14113 Prog. Anal. Chem. Iron Steel Ind. 218. 9313994 Comm. Eur. Communities [Rep.] EUR 1992 EUR 14113 Prog. Anal. Chem. Iron Steel Ind. 228. 9313997 Comm. EUR. Communities [Rep.] EUR 1992 EUR 14113 Prog. Anal. Chem. Iron Steel Ind. 377. 9314001 Cornput.-Enhanced Anal. Spectrosc. 1992 3 281.9314007 Fenxi Shiyanshi 1992 11 68. 9314036 J. Chromatogr. Sci. 1992 30 136. 9314041 J. High Temp. Chem. Processes 1992,1,91.93/4046 J. Quant. Spectrosc. Radiat. Transfer 1992 47 325. 9314050 J. Res. Natl. Inst. Stand. Technol. 1992 97 1. 9314054 Jpn. J. Appl. Phys. Part I 1992 31 1514. 9314055 Kagaku Kogaku 1992 56 561. 9314060 Kogyo Yosui 1992 405 48. 9314084 Opt. Spektrosk. 1992 72 16. 9314093 Pis’ma Zh. Tekh. Fiz. 1992 18 73. 9314111 Top. Appl. Phys. 1992 70 225. 9314121 Water Rex 1992 26 1163. 94/31 Fresenius’ J . Anal. Chem. 1992 344 190. 94/40 J. Anal. At. Spectrom. 1993 8 309. 94/42 Spectrochim. Acta Part B 1992 47 E1587. 94/81 Chem. Anal. (Warsaw) 1992,37 319.94/107 Gaodeng Xuexiao Huaxue Xuebao 1993 14 44. 941279 J. Anal. At. Spectrom. 1993 8 65. 941280 J.Anal. At. Spectrom. 1993 8 145. 941289 J. Anal. At Spectrom. 1993 8 273. 941291 J. Anal. At. Spectrom. 1993 8 287.941295 J. Anal. At. Spectrom. 1993,8,333.94/296 J. Anal. At. Spectrom. 1993 8 339. 941297 J. Anal. At. Spectrom. 1993 8 349. 941298 J. Anal. At. Spectrom. 1993 8 359. 941300 J. Anal. At. Spectrom. 1993 8 367. 941303 Spectrochim. Acta Part B 1992 47 843. 941325 Talanta 1992 39 1517. 941343 Anal. Lett. 1992 25 2329. 941364 Bioresour. Technol. 1992 42 183. 941483 Mikrochim. Acta 1992,109 39.941568 J. Anal. At. Spectrom. 1993,8,461. 941569 J. Anal. At. Spectrom. 1993 8 467. 941570 J. Anal. At. Spectrom. 1993 8 475. 941572 J. Anal. At. Spectrom. 1993 8 487. 941578 J. Anal. At. Spectrom. 1993 8 687. 941579 J. Anal. At. Spectrom. 1993 8 705. 941580 J. Anal. At. Spectrom. 1993 8 711. 941581 J. Anal. At. Spectrom. 1993,8 715. 941590 J. Anal. At. Spectrom. 1993 8 767. 941593 J. Anal. At. Spectrom. 1993 8 795. 941594 J. Anal. At. Spectrom. 1993 8 803. 941595 J. Anal. At. Spectrom. 1993 8 809. 941596 J Anal. At. Spectrom. 1993 8 815. 941597 J. Anal. At. Spectrom. 1993,8,821.94/598 J. Anal. At. Spectrom. 1993 8 827. 941599 J. Anal. At. Spectrom. 1993 8 833. 941602 J. Anal. At. Spectrom. 1993 8 847. 941604 J. Anal. At. Spectrom. 1993 8 859. 941617 Spectrochim. Acta Part B 1993 48 985. 941620 Spectrochim. Acta Part B 1993 48 1023. 941621 Spectrochim. Acta Part B 1993 48 1035. 941622 Spectrochim. Acta Part B 1993 48 1039. 941624 Spectrochim. Acra Part B 1993,48 E1053.941631 Spectrochim. Acta Part B 1993 48 1115. 941633 Spectrochim. Acta Part B 1993 48 1139. 941634 Anal. Chem. 1992 64 2751. 941638 Anal. Proc. 1993 30 199. 941668 Fenxi Shiyanshi 1993 12 61. 941669 Fenxi Shiyanshi 1993 12 79. 941670 Fenxi Shiyanshi 1993 12 83. 941673 Int. Labmate 1992 17 61. 941683 Microchem. J. 1992 45 343. 941684 Microchem. J. 1992 45 356. 941687 Spectrochim. Acta Rev. 1993 15 153. 941693 Anal. Chem. 1993 65 1689. 941700 Anal. Chim. Acta 1993,276 161.941702 Anal. Chim. Acta 1993,276,377.941704 Anal. Chim. Acta 1993 277 1. 941707 Anal. Chim. Acta 1993 278 99. 941721 Fenxi Huaxue 1993 21 328. 941724 Fenxi Huaxue 1993 21 388. 941733 Fresenius’ J. Anal. Chem. 1993 345 547. 941736 Fresenius’ J. Anal. Chem. 1993 345 600. 941741 Spectrochim. Acta Part B 1993 48 359. 941743 Spectrochim. Acta Part B 1993 48 403. 94/745 Spectrochim. Acta Part B 1993 48 515. 941746 Spectrochim. Acta Part B 1993 48 521. 941748 Spectrochim. Acta Part B 1993 48 543. 941753 Spectrochim. Acta Part B 1993 48 671. 941763 Zh. Anal. Khim. 1993 48 101. 941775 Anal. Sci. 1993 9 273. 941789 Appl. Surf Sci. 1993 69 185. 941791 Appl. Surf Sci. 1993 69 403. 941809 Avtom. Svarka 1992 4 48. 941825 CLB Chem. Labor Biotech. 1993 44 166. 941827 Crit. Rev. Anal. Chem. 1992 23 397. 941835 Extr. Metall. Copper Nickel Cobalt Proc. Paul E. Queneau Int. Symp. 1993,2 1651.941836 Fenxi Huaxue 1993,21,246.94/839 Gaodeng Xuexiao Huaxue Xuebao 1993,14,25.94/843 Guangxue Xuebao 1992,12,1117. 941848 Huanjing Wuran Yu Fangzhi 1992 14(4) 39. 941849 Huaxue Xuebao 1993 51 283. 941850 Huaxue Yu Nianhe 1992 3 166. 941854 J. Aerosol Sci. 1992 23 S417. 941858 J. Chromatogr. 1993 633 151. 941859 J. Chromatogr. 1993 636,271.941860 J. Chromatogr. 1993,636,277.941867 J. High Resolut. Chromatogr. 1993 16 106. 941874 J. Phys. D Appl. Phys. 1993 26 585. 941881 Kvantovaya Elektron. (Moscow) 1993 20 51. 941882 Laser Part. Beams 1992 10 737. 941886 Microchem. J. 1993 47 278. 941906 Opt. Lett. 1993 18 747. 941908 Jpn. Kokai Tokkyo Koho JP 05 34,287 [93 34,2871 (Cl. GOlN21/73) 09 Feb 1993 Appl. 91/209,816 26 Jul 1991; 8 pp.941913 Prepr. Pap. Am. Chem. SOC. Div. Fuel Chem. 1993 38 272. 941915 Proc. SPIE-Int. SOC. Opt. Eng. 1993 1716 2. 941916 Proc. SPIE-Int. SOC. Opt. Eng. 1993 1858 72. 941917 Proc. SPIE-Int. SOC. Opt. Eng. 1993 1858,464. 941919 Report 1990 Order No. N92-11349 145 pp. 941920 Report 1991 Order No. AD-A242941 65 pp. 941921 Report 1992 BRL-TR-3324; Order No. AD-A248541 23 pp. 941922 Report 1992 NRL/MR/468 1-92-6941; Order No. AD-A247827 18 pp. 941940 Trans. Inst. Min. Metall. Sect. B 1992 101 B9. 941941 Trends Anal. Chem. 1993 12 41. 941956 Zavod. Lab. 1992,58( 12) 26.941958 Zh. Eksp. Teor. Fiz. 1993,103(2) 417.