ISSN:0306-1353    

DOI:10.1039/AA98010FX001

出版商:RSC    

年代:1980

数据来源: RSC

ISSN:0306-1353    

DOI:10.1039/AA98010BX003

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

CHAPTER 1 Atomization and Excitation 1.1 ARCS, SPARKS, LASERS AND LOW-PRESSURE DISCHARGES 1.1.1 Arcs and Sparks 1.1.1.1 Theoretical. One of the most important contributions to the understanding of the nature of arcs and sparks in recent years has been the work published by Waltcrs and his group, based at the University of Wisconsin. Recently, Walters has paid a glowing tribute to an early paper by Kaiser and Wallraff.This was published in German in 1939 and Walters and Parnsworth havc now published a full translation (I 21 8). The experiments reported by Kaiser and Wallraff werc of such a remarkable quality that the passage of time has not diminished thcir value. Although the method of generation of sparks is a little dated, the work is still rclevant today because it addressed the spark source, first in terms of the chemistry it causes and secondly, in terms of the net spectrocheniical eflects of integrating that chemistry. There is much to commend in this excellent papcr to anybody working in the field.In the current year, Walters group has shown that with positionally stable discharges betwccn copper electrodes, thc vapour moves primarily along the inter-electrodc gap in proportion to the current duration (I).For substantial pcriods after the discharge has terminated, a significant degree of ionization and a high concentration of neutral atoms remain in the gap. A further series of experiments was performed to elucidate the physical and chemical nature of' the spark discharge; the influcncc of an external magnetic field was also studied (1224).A Schlicren system producing high quality images was built to facilitate thc observations (845). Other authors have also rcpsrted techniques for the observatioii and control of high voltage sparks (259, C1414) and of a miniature nanosecond spark discharge (29 If. Coleman and co-workers (94'1, C1415) have followcd earlier work (see ARRAS, 19791, 9, 1), in which independent discharges for sampling and excitation werc used.A form of inductive coupling WAS employed to provide thc spectral excitation, which concentrated most of the energy in the atom and ion resonance lines, thereby making them sharper, and almost eliminated the background. The use of arcs in magnetic fie1d.s containcd within graphite cylinders ha:, again been reported by Vukanovic (290, 3SO). It has been suggested that rotating magnetic fields can improve dctcction limits for all cases where the samples do not contain large quantities of elements with low ionization potentials (C1570).An extensive study of electron prcssure, temperature, degree of ionization and the concentration of elements in both thc anode and cathode regions has been made (1219).The temperururcs of an a'rc have been measured €or niatrices of KCI, NaCI, SiO, and CaCO, and werc found t o be 4350, 4570, 5230 and 5780 K, rcspectivcly (303). In a buffered arc, a positive relationship between clectron pressure and temperature was found (1 221). 1.1.1.2 Analytical Techniqzm. Emission spectroscopy using arcs and sparks i s one of the oldest multi-element techniques available, but with the advent of newer sources such as the ICP, its widc range of application is often forgotten.Review papers prescnted by Parsons (C1173) and Norris (C14.53) serve as good reminders of thc continuing role of arcs and sparks. 3132 Analytical Atomic Spectroscopy Analysis by the direct aspiration of solutions into an arc has obvious attractions and a comparison of the performance achieved by an arc with that achieved by an ICP has bcen made (1202).Lower detection limits were obtained with the ICP for ion lines, but on using atom lines the advantage was only retained for elements having a high ionization potential. It has been shown that the addition of CCI, to the analysis solution yielded a 50-fold increase in the rate of volatilization of elements in an arc (1222).An elaborate approach has been proposed by Schmidt and Sacks (@1416), who placed a capillary arc in a N,O/C,H, flame generated on an AA burner. The solutions were aspiratccl into the flame, the function of which was to supply a regular and even flow of solute particles into the arc and additionally to stabilize it. Solutions have also been injected directly into an arc through the lower electrode (C1547).Additional references on the preceding topic - 51 3, C1551, C1552, 1621. A system of direct introduction of powders into the arc has been described (1228). Florian et al. (294) found that it was better to have a continuous rather than an intermittent supply of the sample in the arc gap. The same authors have pointed out the probabl: sources of errors for the more conventional method of analysis in which the sample is compacted into a cavity in the lower electrode (C1543).lmprovcnients of an ordcr of magnitude in detection limits have been achieved (432) by compressing the samples into a channel cut into a graphite electrode and then passing this under the counter electrode at a rate of 0.4-0.6mms-1.For the determination of trace elements, some useful guidelines to the corrcct choice of excitation atmosphere have been identified (C 1545). Pre-conczntration techniques are often necessary and a review of these has been published. (40). The determination of Ag and Au in geological samples has been achieved by chemical extraction and evaporation of the resultant solution onto carbon powder, which was then analysed (39).Krasil’shchik and Voropaev (1539) have used electrolysis to concentrate the elements Al, Ca, Co, Cu, Fe, Mg, Mn, Ni, Pb and V from sodium nitrate solution on to a graphitc electrode. Additional reference on the preceding topic - 212. Carrier distillation techniques are well accepted and there has been a report of their use for the analysis of paint flakes in forensic work (1 108).Rautschkc (293) studied the effect of various additives on the determination of Group IV-VIII elements and found that alkali halides decreased the plasma temperature but C,C1, did not. The importance of selecting the corrcct counter electrode has been emphasized in the analysis of copper alloys using an arc. The choice depended largely upon the alloying elements (1347).In analyses using spark sources, the choice of countcr electrode and the excitation atmosphere were both found to be important (1 537). The accuracy and precision of a microspark technique for the analysis of stainless steels, has been ascertained (1682). The correct choice of spectral line pairs reduced the RSD to 0.04-0.01 for Cr, Mn, M o and Ni and to 0.10-0.1 1 for Si and Ti, these values being poorer than those normally achieved by a modern spark source and direct reading spcctro- meter.By limiting the area available for the spark to a diameter of 3mm instead of the usual 7mm, the necessity for both a high energy prespark and a long preburn time was removed in the analysis of resulphurized steel (C1429).Investigations of the potential of a spark source for the analysis of liquid steel have been reported, but the precision was poorer than that achieved with solid samples (1 128). The high energy preburn technique yielded detection limits for elements in lead of 1-2ppm, and RSDl values at higher concentrations of about 0.01 (1105). It was, however, necessary to recast the samples before analysis using a copper mould.Blocks of 40mm diameter and 6 mm thickness were used. The exploding thin films technique has again been proposed for the determination 01Atomization arid Excitutiorz 33 refractory elements (1230). The samples were held in suspension in alcohol and then placed on a thin film of silver deposited on polypropylcnc or polycarbonate membrane filters.A vidicon was used for the evaluation of the spectra (C1426) for which theoretical models have been proposed (C1473, C1474). For the determination of V, concomitant and chemical compound effects were minimal, provided the particle size was grcatcr than 10 pm ((21466). An increase in analytical accuracy has been achieved by improved interpretation and calibiation of spectrographic plates (C1544, C1554, C1557). 1.1.2 Lasers Lasers were originally used in atomic spectroscopy for both sampling and spectral excitation of the material of interest. The spectral intcnsity of the analytical lines howcvcr was often insufficient and the auxiliary techniques of cross-excitation of the vapour were introduced. Since then the use of the laser solely for sampling the material has incrcased and a widc variety of techniques has been used for the analytical measurements.Mohr (1617) has shown that for emissioiz urznlysis using the conventional cross-excita- tion technique, the composition of the atmosphere in the sampling cliamber should bc optimized. ?"his produced a reduction in self-reversal and absorption, in the interferences, and an improvement in detection limits.The vaporized material has also been introduced into a MIP for which the analytical precision varied from 1.2 to 13.8% for peak-height measurement and from I .3 to 12.1 % for peak-area measurement (1 13'3). Dctection limits varied from 0.9 ppm for Zn in aluminium to 22 ppm for Mn in steel. A solution, which was steadily dropped on to a surfacc, was evaporated using a laser and the vapour conveyed to a plasma discharge tube (1123).Limits of detection for Ba, Ca, Co, Ee, K, Li, Na, Ni and Sr were in the range 1 to 400pg. A bsorptiovz methods have been used by Quentmeyer et 01. (1203) for analysing aluminium and steels. The detection limits achieved for Cr, Mg, and Mn in aluminium were similar to those achieved by FAAS. Small samples in a histochemical application have been analysed by transporting the vaporized material into an electrothernial atomizer (1 379).A similar approach for the analysis of alumina, copper, lead oxide and pyrites was reported to give poor precision (1258). Detection limits have been improvcd by an order of magnitude by supplementing the laser pulse with an electrical discharge (1101). Tlic vapour from samples mixed and pressed with graphite has been found to be in thermal cquilibrium with a mean temperature of 3400 K (882). 1.1.3 Low-pressure Disekarges The kinetic gas temperatures and electron density in the Grjmm type of glow discharge lamp have been measured by the construction of a lamp that enabled the discharge to be viewed side on (1215). A calibration procedure has been repoTted that allowed the accurat? dcterniination of the binary alloys of gold and copper (1235).Du: Gregoria et al. (1536) suggested that to determine elements at high coilcentrations in steel, it is better to use a cunstant charge or the total light emitted as a reference, while for the lower concentrations constant lime exposures are best. Unfortunately, the results achieved were not as good as might have been expected.Earlier work (see ARAAS, 1979, 9, Ref. 8021, which showed that the HCL offers better precision than the GDL for the analysis of steel, has been confirmed (1120). Butler (C1412), in a review on the application of the GDL to ferrous and non-ferrous alloys, concluded that problem areas exist in trace and micro-trace analysis and that developments employing magnetic fields and supplementary discharges may help to overcume thcm.Calculation of the fraction of the elements ionized in the boosted GDL has been used as an aid to line selection (C839) (see also ARAAS, 1979, 9, Refs. 628, 687 and 69 1).34 Analytical Atomic Spectroscopy The arzalysis of izon-conducting powders remains a problem.Limited success has been achieved for the analysis of plant ashes, for which good agreement was obtained with the certificated value for Mg. The precision for P however was greater than 20lOJo and the concentrations of Zn were less than the detection limit of the method (871). The atomic fluorescence resonance detector (ARAAS, 1979, 9, Refs. 812, 11371, has again been applied to the determination of C, P and S in steels (1208).The matrix emission had to be observed to correct for interferences in C and P detzrniinations, but not for S. A similar device has been described (944, 1241) in which the line narrowing normally associated with selective modulation was observed. Detection limits for the determination of elements in aluminium, copper and iron alloys were in the range 1-1OOppm.rn their study of hollow cathode lamp sources, Niemczyk and Mehs (125) suggested that a theoretical model based upon local thermal equilibrium was not appropriate and that more consideration should be given to the radiative processes taking place. The pro- cesses of spectral excitation have also’ been studied by examining the effect of Na on Mg in the presence and absence of magnetic fields (1581). Typical “arc-like” spectra have been observed in a HCL, but superimposed on lower backgrounds which could therefore enable them be used as reference spectra (1556).Broekaert (1211) concluded that the HCL was an excellent source for the analysis or rare earths in solution provided that the operational parameters and the method of sample preparation were carefully selected. A technique utilizing metastabfe tramfer emission in an active N, afterglow plasma (see ARAAS, 1979, 9, 6) has now been interfaced to a gas chromatograph and found to function as a very sensitive detector for a variety of organometallic compounds of As, Hg, Pb, Sc and Sn and as a non-specific detector for organic compounds (C1442). Hydrocarbons separated by GC were found to provide intense CN emissions. A further application has been the determination of Pb in the washings from leaves where the solutions were dried and atomized from a tantalum boat into an 80 W microwave-induced discharge (1 189). A linear response from 1-2000ng with a detection limit of 0.2ng was obtained. Elkatton (C720) has studied pulsed excited N, and Ar/N, mixtures in the afterglow plasma. Other references of interest - Isotope shifts of Yb using a HCL : 1210. Lamps and light sources for analytical atomic spectroscopy : C14061.

ISSN:0306-1353    

DOI:10.1039/AA9801000031

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

34 Analytical Atomic Spectroscopy 1.2 PLASMAS In recent ycars interest in various forms of plasmas as atomization and excitation sourccs has been a dominant feature of both published rcports and conference programmes. The opening of the new decade was marked by the first major international confercnce devoted solely to spectrochemical analysis by plasmas, held in P’uertoi Rico. The greatest volume of literature has concerned the r.f.inductively coupled plasma (ICP). This techniquc is now rapidly developing towards maturity, as witnessed by a large volume of applications litera- ture, reviewed elsewhere in this volume, and the acceptance of ICP methodology in recommended procedures, e.g., by the U.S. Environmental Protection Agency (649, 1050). A rcmaining fruitful area for research seems to be that of sample introduction.Tbe decline in interest concerning microwave plasmas has continued and here the grcatest attcntion seems to be focused on the use o€ the MTP as a chromatographic dztector. Once again reports concerning d.c. arc plasmas have increased and szvcral more fundarncntal investiga- tions, especially of interferences, as well as applications have been published.A tamization and Excitation 35 1.2.1 R.f.Inductively Coupled Plasmas 1.2.1.1 Reviews. It being some 18 years since Greenfield first started analytical experiments with an ICP, and this being the age of majority in England, he was able to review “Plasma Spectroscopy comes of age” (1305, 1708). The attributes of fie ICP were discussed, especi- ally as regards detection limits.Some current controversies were also1 identified and there was an interesting, but inconclusive, comparison of high and low power plasmas. Boumans (371, 938) reviewed the present and future position of ICP-OES: in analytical chemistry. A useful report (68’3) has also appeared of a discussion between users of different plasma instruments and manufacturers.Some insights are afforded on aerosol generation, instrumen- tation and possible future developments. Various leading workers in the field have presented conference reviews ((227, C656, C661, C668, C670, C711, Cll43, C1366, C1381, C1471, C 1749). I .2. I .2 Plasma Characteristics. The present knowledge of excitcetion conditions in the central channel of the ICP is inadequate to explain some interesting observations reported by various workers.Mermet’s review ((2582) of possible mechanisms was therefore timely. It has been again confirmed (C760, 1205) that the Ar ICP is not in local thermal equilibrium (LTE) but when water was added, c g . , by nebulization using the carrier gas, the electron density, electronic excitation and ionization temperatures increased.These observations apparently conflict with expectations that water vapour should consume part of the power input, implying a decrease in temperature. Temperature and pressure measurements in a low-power ICP discharge have been reported (20). Off-axis temperature maxima were reported with the temperature of the central channel being strongly dependent on the injector gas velocity.A preliminary study of some vapour phase reactions of titanium and silicon chlorides in an enlarged plasma augmented by a d.c. arc has been presented (21). Interest in the modelling of ICP discharges has continued. Calculations have shown (C584) that the electromagnetically induced recirculation in the discharge region decreases with increase in the operating frequency over the range 3 to 263 MHz, such that at the latter frequency the injector gas will easily penetrate the fire-ball.The conclusions of a computer simulated model of the ICP reported last year (see ARAAS, 1979, 9, 7) have been reiterated (I 9) and confirmed by experimental study (20). Acknowledging that the ICP i s not in LTE, Lovett (C1039) has suggested that eventual modelling will probably depend on detailed kinetic balancing of the excitation, de-excitation, ionization and recombination processes, A full kinetic model was not proposed pending further experimental studies.The electron concentrations in the spectroscopically important regions of an ICP have been calculated to be about one order of magnitude higher than for LTE (C131883. The diffusion of electrons from the discharge region against the Ar flow towards the coil is intense and appears to be responsible for the relatively high conductivity of cold Ar at the lower edge of the coil.Several studies of the analytical characteristics of the ICP in the vacuum-ultraviolet region of the spectrum have been published and the advantages of this spectral region recognized (C658, C831).These advantages include the avoidance of spectral interferences, e.g., by Fe and Cu on P (C831), the occurrence of the most sensitive lines for certain elements such as I and S (C41 , 170) where a S detection limit of 50 pg ml-1 was obtained at 180.73nm, and the availability of alternative analysis lines, e.g., for By Hg, P (C421, although spectral interferences, e.g., from Ca and Si (1091), may still be encountered.Using a high power demountable torch, Denton et al. (C136, C1041) have investigated promising analytical lines for Br, C, CI, N, 0 and S in the region between 120 and 185 nm. The necessary components of a spectrometer for use at such low wavelengths were discussed,36 Analytical Atomic Spectroscopy including a comparison between dispersive and non-dispersive wavclength isolation.This extension of the application of the ICP may be particularly suitable for chromatographic detection. All the above systems rely upon purged or vacuum spectrometers and a suitably purged optical path. One such optical system, which has been described in dctail (C1397), was capable of operating down to 175 nm where quartz starts to absorb. Many non-metals remain undetectable even in the accessible v.u.v., but Northway and Fry have suggested that these, difficult to excite, elcments may be determined by using non-resonant transitions in the visible and near-infrared regions of the spectrum (C67).Fry’s studies have included measurements of Br, C, C1, F, H, N and 0 using a photodiodc array (C1040).More detailed studies have been reported of : 114 N(1) lines in the region 353.30 to 1065.303 nm in an all Ar plasma (S57), the strongest line, 821.67 nm, yielding a detection limit of 1 ppm; 66 O(1) lines in the region 287.6 to 950.5 nm with a detection limit of 0.5 p g at 777.19 nm (858); 56 F(1) lines in thc region 614.9 to 895 nm with a detection limit of 0.35 pg at 685.6 nm (1 31 1).In the latter case rigorous precautions were necessary to prevent torch erosion. Carbon emission at 247.8 nm and 1913.0 nm has been used by Yatcs er a!. (C728) to determine total organic carbon down to 1 ppm. While the ICP i s capable of dissociating molecules as stable as CO, considerable molecular emission can still be observed. A study of the emission spectra of OH, NO and NH has shown these to be a function of height of observation, input power and surrounding environment (1 648). Molecular species primarily arise from impurities in the Ar, entrainment of air and from the sample.These diffcrent sources yield different spatial distributions (C647). This knowledge can be used to minimize molecular interfcrenccs by obscrving radiation from plasma volumes containing low molecular concentrations. It is increasingly being recognized that defining different regions o f fke discharge i s crucial to our understanding of certain phenomcna.Koirtyohann et a!. (1721) have proposed a system of nomenclature for the different regions of an Ar ICP, e.g., the central channel was said to consist of in ascending order : a pre-heating zone, an initial radiation zone, and a normal analytical zone.The position of thc central of thcsc zones is sensitive to changes in operating conditions. These phenomena have also been found by othcr workers. Horlick and Blades (843) in a spatial study of Ca(I), Ca(1I) and Sr(I1) lines have shown the dangers of drawing conclusions about ICP signal changes from mcasurements at a single height.The effect of excess Li on the above lines varied according to height in the plasma, a strong enhancement being observed close to the coil and minimal enhancement at normal observa- tion heights. Variation of power altered the positions at which these effects were observed. These results are consistent with those reported by Ebdon (C1749) on the variation of thc enhancement of Mn emission by Na with viewing height, power and other operating para- meters in Ar and N, cooled plasmas.Since thc interference was related to five interacting operating parameters, a Simplex optimization procedure was used to minimize interference. A variety of techniques have been used to obtain spatial emission data. Blades and Horlick (C70, C5813, C588) used a self-scanning linear photodiode array to obtain both detailed vertical and horizontal information.Lateral intensities were converted to radial intensities by Abel inversion. A theoretical model, which described emission intensity as a function of analytc excitation, ionization characteristics and source temperature gradient, was built upon the information gained. Bastiaans (C 1 35, C 1042) has madle spatially rcsolved measurements both via Abel inversion of emission signals and vicr laser saturated fluorescence using a tunable dye laser.A nitrogen-lascr pumped tunable dye laser has bcen used to map the relative ionic and atomic fluorescence profiles for Ba in an Ar ICP (1232). At low heights the ion profile resembled a hollow pencil with a double-peaked distribution while thc atom profile appeared complementary to the ion profile.The use of AAS to measure metastableAtomization and Excitation 37 Ar atoms in one position in the ICP (1248) suggests another tool that can be used for mapping. After some years of controversy acceptablc accounts of ICP optimization are now appearing. Greenfield and Burns (3) have discussed necessary and sufficient criteria for the unequivocal comparison of plasma torches as emission sources.Later they used these criteria to compare different torches run under optimized conditions on the same spectro- meter (477). The criterion used was that of signal-to-background ratio (SBR), the alternat- ing variable search method was used to find thz optimum. The comforting conclusion was that low power was optimal for an all Ar plasma on a Scott torch but higher powcr was optimal for a N, cooled plasma on a Grccnfield torch.Ebdon et al. (497, C1749) came to similar conclusions after a number of plasmas had been optimized for different elements using the variable step-size Simplex method. They also uscd this method to evaluate different torches (495). Strasheim (C1144) has used Simplex techniques to optimize SBR for Cu, Ag and Pb in aqueous solutions.Ebdon sounded a note of warning that their Simplex studies had shown that optimum conditions in an all Ar ICP may be shifted in the presence of a matrix (C363), e.g., to higher power for As in the presence of Cr. Similarly it was shown that conditions for optimal SBR may not b2 optimal for minimal intereference (C1749).This conclusion has been confirmed by Kanicky and Toman (C1567), who there- fore sougbt to identify compromise operating conditions for multi-element analysis. In a two-parameter study of power and observation height, Dahlquist et al. (C1382) showcd that 18 diverse elements behaved in a similar fashion with best limits of detection and minimal interference from alkali clemcnts at observation heights below 16 mm.The mutual inter- ferences between alkali elements observed above this height presumably also vary with the closely related parameter, injector gas flow-rate, but unfortunately this was apparently not discussed. The possibility of selection of optimum instrumental parameters for any element with a scqucntial ICP spectrometer is an obvious advantage of such an instrument. It is therefore surprising that a more thorough discussion of optimization methodology and criteria was not made in a confercnce paper (C65) where the general advice given was to retain “standard conditions”.The similar sensitivity of the thrce Cd lines, 214.438, 226.502 and 22&.802nm, presents a particular challenge to optimization.The net line intensity at these three lines has, to complicate the matter further, been shown to vary with power but in a different way with different nebulizers (1227). The signal-to-noise ratio characteristics of the ICP have been investigated and it has been shown that the SD of the analyte signal is linearly related to the signal level (1316). The dominance of analyte flicker as the noise was probably due to the nebulizer-sample delivery system and therefore was not strongly related to the plasma power or correlated with noise on the Ar(1) lines.A study of thc noise power spectra using the fast Fourier transform identified three major noise types : white noise, low frequency noisc and high frequency proportional noise (1 236). The latter decreased with increasing nebulizer flow-rate.The contribution of the different noise types also varied with the observation height. Further confirmation of aerosol borne noise was the striking correlation obtained between sample emission signals and laser light scattered by the aerosol droplets (C 1385). The practical effects of many of these phenomena may be bctter illustratcd when reports arc available of a survey programme of ICP detection limits for “real” samples as well as the usual “distilled water” standards (23).While the majority of plasmas use only Ar, interest is reviving in gases other than argon. Montaser and Fassel (C138’6) have critically compared plasmas cooled by pure Ar and pure N, and various intermediate mixtures using a typical Fassel torch. A comparison of Ar and N, cooled Ar plasmas operated at powers typically optimal for an all Ar plasma showed superior detection limits for the all Ar plasma (1196).Unfortunately, no attempt38 Analytical Atomic Spectroscopy was made to compare the two under their own optimal conditions. Another comparison of Ar and N, cooled plasmas, this time using a Greenfield torch, showed little differciice between thc two plasmas for the analysis of aqucous solutions, but considerable advantages accrued by using a N, coolant for the analysis of organic solvents (C762).With N, coolant, no carbon deposition occurred and stability rcmained excellent. The advantages of the higher powered N, cooled plasma (which i s less perturbed by H,) in hydridc generation procedures and for GC detection have also been stressed (C761).In the same paper, the use of air as the coolant was claimed to increase scnsitivity and to be particularly suited to the use of discrete-sample nebulization by injection cup. The preliminary application of an all N, ICP, first reported last year (ARAAS, 1979, 9, Ref. 1299), has now bcen described (1307, C1387). The ICP was started on Ar and then gradually switched over to N,.Optimal gas flow-rates were different in the two plasmas and the minimum powcr required for the N, discharge was 1.3 kW compared to 0.2 kW for Ar. The continuum background was lower in the N, plasma, which was also closer to LTE. A similar N, plasma, also sustained on conventional instrumentation, has becn reported by Choot and Horlick (C585).Greater sensitivity for ion lines seems to be available generally in an all Ar JCP but atom lines may give significantly improved SBR in a N, ICP. This is in agrccment with trends already well known when N, rcplaces Ar as the coolant in a conventional Ar ICP. Only a few papers have ever been published describing the use of the ICP for specfro- scopies other thmz emission.A report of the ICP as an atom reservoir for AAS has again shown that the available sensitivity is poorer than for flame AAS (1254). A possible use of ICP-AAS, however, might be for metals or samples that are difficult to atomize in flames. Work has also continued on the use of the ICP as an excitation source for AFS (1 186, C1425; see also Section 1.3.5). When a highly scattering sample was aspirated into a separated air /C,H, flame, a scatter-correction procedure based on self-absorption in the ICP was used (1186).High analyte concentrations in the same matrix were determined by mcasuring the emission of the sample in the JCP using a flame resonancc detector. Interest in plmma source mass spectrotmetry is growing. At the present stage it is clear that the analytical potential is still being explored (C636, C640, C641, C1391, C1454; see also ARAAS, 1979, 9, 9).Large excesses (> 500 pg ml-1) of an element of low ionization potential suppressed the count rate of analyte ions (C153). This was partly duc to blocking of the pinhole orifice in the sampling cone prior to entry into the quarupole mass spectro- meter and partly duc to suppression of analyte ionization in the TCP itself.Reaction in thc relatively cool boundary layer and supersonic jet formed during the extraction process influenccd the spectra of some elements that form stable oxides, e.g., the appearancc of YO+ (C1390). Evidence has also bcen obtained confirming suprathermal ionization in thc ICP, suggesting that ICP-MS could be used as a diagnostic tool for investigating plasma characteristics. 1.2.1.3 Sample Introduction.Fortunately much attention is now being paid to methods of sample introduction into the ICP and it is now widely regardcd as the part of ICP instru- mentation most amenable to improvement. Koirtyohann (C663) has reviewed methods for sample introduction, acknowledging the problems but stressing the advantage of the central channel through which sample can be directed.In a review of sampling and analysis strategies for ICP-AES, Wattcrs (C732) has undcrlined the advantages of pneumatic nebul- ization for the analysis of liquids and the disadvantage of dilution when solid samples have to be dissolved. Much more i s now understood about aerosol characteristics following pneumatic nebulization.Browner and co-workers have Continued to publish results from their impres- sive studies (C745, C1046). The usc of a cascade impactor to measure droplet size distribu- tions has been fully describzd (1187, C1463) as has an interesting technique based on theA tomizaiiort and Excitation 39 suppression of the increase of saturated vapour pressure, with decreasing droplet diametcr, by enhancement of the surface tension of the liquid to be nebulized (1198).With such techniques droplet sprays can be characterized and the effect of spray chambers, for example, evaluated (996). The effect of the spray chamber has also been studied by other workers (4). In contrast, in another study a spray chamber of volume 7.5 1 has bcen used to permit detailed study of aerosol generation and loss rates ((2586).These studies have shown the existence of a major aerosol loss process during the actual operation of the pneumatic nebulizer. This has been confirmed by a dual nebulizer experiment in which the aerosol of one nebulizer was destroyed by the other, but only when it was producing a spray and when the two nebulizers fed one spray chamber (C1384).Other conference papers have been presented discussing the fundamental processes of nebulization, aerosol conditioning, trans- port and vaporization (C11, C14, C1044). It was claimed that solute vaporization and ionization suppression effects are largely attributable to processes that occur during acrosol production and transport (C 1044). Mechanisms for nebulizer clogging, which occurs in some systems when samples of high salt content are aspirated, have been suggested (C14).This problem has also bcen discussed by Gustavsson (24) in a useful investigation of factors that affect nebulizer characteristics. The importance of the dispersion of thc acrosol was stressed, but this was measured in only a comparative way using microscope slides.A CW He-Ne laser has been used to produce Mie scattering of nebulized aerosols (C746). This technique showed that water aerosols persisted within the plasma into normal vicwing zones. Renewed interest has also been shown in the use of pneumatic nebulization to inject charged solutions into the ICP (1652). The new form of interference, aerosol ioriic redistribution, first reported last year (see ARAAS, 1979, 9, 10) has been the subject of further reports (C745, 1377, C1456).Minor clement concentrations can be enhanced as a result of ionic redistribution during nebuliza- tion. The extent to which this occurs is critically dependent on both droplet size and the amount of major cation. The concomitant can also increase or reduce the interference effect in the ICP by changing the excitation conditions, as shown by dual nebulizer experiments.Concentric pneumatic nebulizers, often of all-glass construction, have provcd exceed- ingly popular. Meinhard (C747) has described ways in which residues may be removed non-destructively from an all-glass nebulizer to restore its performance to original levels. Such a nebulizer is of course unsuited to the nebulization of hydrofluoric acid media.Goultcr (C 1 392) has described a corrosion resistant concentric nebulizer, spray chamber and torch capabls of analytical work using such media. A platinum/iridium nebulizer has also bcen dcscribed ((263). An injection cup for the analysis of p l samples using a conventional glass concentric nebulizer has been further discussed (1055).This seems more appropriate for use with a higher power ICP as the low power Ar TCP was destabilized by the air entrained between sample injections. Sample sizes of 20 to 50 pl were investigated and serum samples successfully analysed. Another elegant approach to introducing pl volumes into the ICP via a conventional concentric ncbulizer has been reported (531).A microsyringe sample injector was supplied by a flow-in jcction analysis system. The fritted-disc nebulizer of Ape1 (see ARAAS, 1979, 9, 10) has the advantage of high efficiency (> 60%) at very low sample-uptake rates (as low as 1.2 ml h-1). A modification of the original design has been reported (C1047) and further work is being directed to eliminating drift and memory effects. Major advances have bcen reported this year in the design of pneumatic cross-flow nebulizers.A remarkable improvement in precision (to levels typicallly as low as 0.22% RSD) has been achieved (C749). This nebulizer operates at 2001psig Ar pressure and has a fixed geometry. It was necessary to control the coolant Ar flow by a mass flow40 Analytical Atomic Spectroscopy controller to achievc this precision.The nebulizer also aspirated samples with up to 35% dissolved solids without blockage. Fixing the geometry of the cross-flow nebulizer is apparently very advantageous. Boorn and Browner (C54, 15233 have reportcd a compact cross-flow nebulizer of rugged all-glass construction, and with fixed gcometry. Their design also incorporated an impact bead.The construction and performance of a similar one-piecc nebulizer has also been outlined (C15). Another new fixed geometry cross-flow nebulizer described by Wohlers and Hoffman (C61, C751) used a platinum alloy tube for sample introduction and a sapphire jewel for the gas inlet. Gustavsson (24) has shown that the usc of a bevelled cdgc for the liquid capillary is advantageous.The careful control of tip shapes and the permanent alignment of the capillary were both incorporated into the design of another experimental nebulizer (C62). A Pt needle cross-flow ncbulizer and polythene spray chamber for use with HF solutions has been described (C766). The advantages of Babingtoiz-type nebulizers are increasingly bcing recognized. The designs of Wolcott and Sobel (ARAAS, 1979, 9, Ref. 592) and Suddendorf and Boycr (C750; see also ARAAS, 1978, 8, Ref. 1447) havc inspired other attempts to dcsign simple nebuli- zers capable of handling samples with high dissolved, or even suspended, solids. In onc type, the sample flowed over a jewelled orifice from which the injector gas flowcd (C61, C751). This was effective in nebulizing brines (30% NaCI), milk, tomato juice and apple cider.Another Babington-type nebulizer was reported to be superior to conventional pneumatic nebulizers as regards detection limits, precision, memory and analysis time (C752). This nebulizer, which incorporated a glass impact bead, could also handle susper.ded matter up to 1 mm in diameter. Additional reference on the preceding topic - C 1400. While the attractions of ultrasonic nebulizers, particularly that of high efficiency indepcndent of carrier gas-flow rate, are well known, and have again been reviewed (C13, C63, C662), their popularity is limited by memory effects and poor reliability.There have this year, howevcr, been significant practical reports of the use of ultrasonic nebulization for sea-water analysis following ion-exchange preconcentration of the trace metals (C723, 1520).One approach to improving reliability is to protect the transducer from chemical attack and erosion. A method of doing this by bonding the transducer to a glass plate, one half-wavelength thick to provide acoustic coupling, has been patented (655). Taylor (654) has modified a commercial ultrasonic nebulizer so that the transducer may be operated in an upright position and1 with improved water cooling.This device has been applied to a numbcr of environmental analyses (C58). Trassy (653) has patented an ultrasonic nebulizer in which the aerosol produced was surrounded by an annulus of dry gas to prevent the aerosol hitting the hot walls of the injector near the plasma and thus depositing salts.This nebulizer, operated with an automatic computer-compatible filling and cleaning system, has shown excellent stability in the practical analysis of, e.g., blood serum (C748). The desire to use the ICP with solid-sampling, if only to minimize sample prcparation or dilution, has already been alluded to above. One promising approach is to usc a nebulizer to deliver a slurry to the plasma in an analogous way to that previously reported for flames (see ARAAS, 1979, 9, 19).The Babington-type nebulizers described above seem particularly suited to this kind of application, e.g., after the homogenization of plant or animal tissue (C12). Aspects of this approach have been further studied using laser diffraction (C1045). Novel concentric nebulizers have also been used to spray slurries of powdered coals and ceramics (C63, C753). Laser vaporization has been further studied (C46).Careful control of the time interval between shots of the pulsed ruby laser, when vaporizing mctal samples, reduced power variations and hence analytical errors. Reports have again appeared of spark erosion methods for solid sampling (C657, C761, C1417). The sample vapour can beAtomization oizd Excitation 41 externally generated, by a spark or interrupted! arc, and then swept by Ar into the ICP.Non-conductive solids can be sampled if briquetted with either C or Cu. Zil'bcrshtein (644) has again reported his method for solid sampling in which graphite sample holders are inserted into an ICF (see also ARAAS, 1979, 9, Refs. 224 and 875). A not dissimilar device, termed the sample elevator, has also been described (C761).The samples (0.1 mg or 10 pl) were placed in a graphite cup, which was covered with a graphite cap with a hole in it. The cup was pushed into a 3 kW ICP between the lower two windings of the coil. The whole crucible was thus inductively heated and the sample distilled. Electrothermal vaporization of samples into the ICP is particularly suited to small sample volumes.A carbon rod is still the preferred vaporizer (172, 619, C14001, although the filament-in-furnace tcchnique (sec ARAAS, 1979, 9, Ref. 463) has now been applied, with a claimed improvement in speed of analysis and freedom from ash affects ((2826). A slightly modified commercial graphite cup atomizer has also been used (C60). Some results with this latter system suggest that high Na levels may cause a spatial shift in the emission profiles.Kirkbright and co-workers (172, 619) have shown that the enhancements of certain signals by matrix elements were due to improvements in transport efficiency between the vaporizer and the ICP. Radio-tracer studies showed, for example, that Sc enhanced the transport of As and Cd by approximately an order of magnitude.T h i s appears to ol'fer a valuable way of minimizing matrix effects that otherwise might reduce transport efficiency. The application of preconcentration or separation using orgnszic solvents has proved to be more problematical for plasmas than for flames. A theoretical approach to the calcula- tion of aerosol droplet diameters, as reduced by evaporation in a solvent-vapour saturated atmosphere, has been described (1239) and applied to the aspiration of organic solvents into an ICP.The effects of ethanol on spectral crnission intensities and the impedance of an ICP have been studied (812). The plasma resistance increased with ethanol concentration but the effect was smaller as the r.f.power was increased. The difficulties in tuning the matching network of a crystal-controlled generator in the presence of an organic solvent probably arises from the steeper slope of the tuning curve compared with that for aqueous samples. An alternative way of improving sensitivity that is particularly popular, probably because of its suitability for several important but difficult to excite elements, is hydride generation (C657, C761, C7717, C1146, C1451).Details of a method for the simultaneous determination of As, Sb and Bi in soils and sediments via hydride generation and with a rapid throughput has been published (1080). The use of potassium iodide was claimed to give greater tolerance to transition-metal interferences. A continuous-flow arsine generation method has been compared wi.th discrete production of the gas (961).The latter gave a superior limit of detection (0.001 pug As ml-1) and RSD (0.02) and was used to determine As in waste water and steel. Sulphur was determined by generation of H,S, using hydriodic acid, which was swept into an ICP and the total S determinedl in the V.U.V. at 180.7 nm ((2'758, 1043). For the generation of Hg vapour, tin(I1) chloride was preferred to sodium borohydride as the reductant (C7, C750) and presumably in this way the deleterious effect of excess H, on thc plasma was avoided.In a novel but related approach, Black and Browner '(C575) have converted metals in biological samples to volatile fluorinated p-diketonate complexes. The metals can be allowed to react directly with the ligands at ambient or elevated temperatures and the resulting chelates swept into the ICP.The ICP is proving to be a versatile, if expensive, chromatographic detector. Reports have appeared of the ICP employed as a GC detector for S, with a linear dynamic range of 3 to 4 orders of magnitude (C758), C, P and S (C1043) and Pb and Si, the former being as tetra-alkyl compounds in petrols ((2'755).Windsor and Denton have described in detail a coupled GC-ICP instrument and have reported detection limits and linear dynamic ranges for Br, C , Cl, F, Fe, H, I, Pb, Si, and Sn (236). These figures of merit and the42 Analytical Atomic Spectroscopy selectivity compared impressively with flame emission and microwave emission detectors. An example of the analysis of a mixture of tetramethyl tin, benzene, toluene and p-xylene was given.The direct coupling of the outlet of an HPLC to an ICP has also been reported (1190). The utility of the approach for the determination of C, Co, Cu, Fe, Mn, P and Zn was demonstrated and C, Co and P determined in vitamin B,2. This technique could prove particularly useful for the determination of metallo-proteins and for structural studies of enzymes that have metal atoms in functional sites.It would appear that peak broadening in I-IPLC-ICP can be brought within acceptable limits and that detection limits in the ng ml-l range with linear dynamic ranges of 4 orders of magnitude arc possible. Further demonstra- tions of the technique have been the separations of : iron carbonyl complexes; molybdenum carbonyl complexes; ferrocene compounds; arsenic compounds; organoniercurials and organolead compounds in petrol, as described in a comprehensive survey of capabilities by Gast et al.(378). 2.2.2.4 Instrumentation. Complete TCP instruments are dcscribcd in Section 2.4.1, where the continuing growth of interest in rapid-scanning sequential instruments will again be clear.Several commercial systems of this type have bcen described (C10, C30, C64, C66, (2137, C739, C740, C741 and C1398) as well as other ccnnputer-controllcd rapid-scanning mono- chromators (C131, C737, C774, C14'37, 1515 and 1718). Apart from this growth thore has not been notable maturation in the instrumental aspccts of thc TCP. It seems surprising that while such sophisticated electronics are now being applied to the readout side of the spectrometer, so little is being published on the design of r.f.generators. An automated tuning and ignition circuit for a free-running 4 kW r.f generator has been reported (C733) to improve power stability considerably. The number of papers discussing torch design appears to be in dcclinc. The interest reported last year (ARAAS, 1979, 9, 9), centred upon mini-torches and low gas consumption arrangements, has been maintained.Kornblum (C765) reported that the problem of bubbles at the top of the water-cooled jacket in his design of low flow torch had been solved by altering the internal walls of the torch. Thc addition of an auxiliary gas flow greatly improved the SBR.Unfortunately, only 0.1 lmin-1 of carrier gas flow could be used and even with a Babington--type nebulizer difficulties remain. Thcse problems were apparent1 y overcome in another low-flow torch rcported by Lowe and Sutton using an cxtendcd injcctor (C868). A demountable torch of 'Grecnfield-i-jcometry' but incorporating a flared intermediate tube has been designed and evaluated (495).This torch could be operated with either Ar or N, coolant gas over a wide range of flow-rates. The intcrchangeable tubes and replaceable threaded brass jets for gas introduction make this a particularly versatile torch. A stream-lincd torch for work with organic solvents that does not suffer from carbon deposits has been further reported (C764) (see ARAAS, 1979, 9, Ref. 1090).A closed discharge cell has been placed in an r.f. coil and the excited low-pressure ring discharge studied (C147, C1564). By placing freeze-dried samples in such electrodeless lamps, mercury has been determined in a variety of fresh and saline watcrs down to pg 1-I lcvels. An advantage of the technique is the low power required, 84 W at 62.5 MHz. An ccononiically feasible plasmatron has also been described (312).The plasma in a cathodc chamber, at 7.5 atm, was fed with dry aerosol using an Ar flow-rate of 3.5 ml min-1 and opcratcd at a total Ar flow of 1.5 1 min-1 at 260 V and 17 A. Dztection limits were reported for Cr, Ir, Nb, Pd, Pt, Re, Rh, Ru, Ti and W and were comparable to AAS, but matrix effects required supprcssi on. There have been a number of reports of plasmas observed from diffcrcnt viewing angles.Provided an air stream at the tip of the discharge was present, to act as :1 thermal barrier, Dcmers (823) reported a number of advantages of axial '(end-on) viewing. Com-Atomization mid Excitatim 43 promise analytical conditions can be arrived at more easily and detection limits improved. Apparently interference effects and self-absorption were not affected significantly.Ape1 et al. (C735) did however rcport some self-absorption for Cu and Pb when their ICP was viewed axially using an air barrier and mirror system. When viewed axially with an IDES the line intensities in an ICP were improved by a factor of 5 to 15 (C1395, C1737). This system may also allow elucidation of certain plasma mechanisms. Measurements between the turns of an induction coil have also been reported (C1145).Minor modifications to the coil were required but this form of observation could prove exceptionally valuable in seeking explanations of certain plasma phenomena by observing conditions at the onset of the atomization process. The importance of resolution in practical analysis using an ICP hardly needs to be stressed.The combination of an ICP with an echelle spectrometer offers high resolution but at the cost of luminosity. Two interesting conference papers on such combinations have appcarcd (C8, C 1402). Holographic gratings have continued to increase in popularity, particularly because of their low stray-light characteristics (C'742, 859, Cl389). By inter- facing two direct-reading spectrometers to one ICP and linking them both to the same computer, as many as 61 elements have been determined in geological samples with minimal restrictions on spectral-line selection as well as freedom from spectral interference (C44).Plasma spectrometry has now reached the stage where thcrc is considerable emphasis on data acquisition. A computer controlled standardization protocol for an XCP has been devised and tested (C56) and a highly automated, operator unattended, computer controlled ICP system for water analysis described (C75).Maesscn and Balke (C767) have shown that when calibration curves were plotted assuming equal weight of response for various con- centrations, the values of the residuals corrcsponding to the lower concentrztion region gradually appeared to increase with decreasing concentration.A SIT image detector has bccn evaluated for multi-element analysis with an ZCP (844). This enabled spectra to be corrected from stored background spectra. The feasibility of automatic background correction by linear arrays of silicon photodiodes was also investigated (C734), but the system was found to be detcctor-noise limited.Measurement systems based on photodiode array spectrometers and Fourier transform spectrometers have been developed that are capablc of simultaneously acquiring information from large spectral regions (C48, C616). Cross-correlation proved a most effective means of processing the data. Precision was improved in intensity measurements of ICP emission by a procedure in which the line was scanncd 416 times and 90 data points acquired over a range of 0.18nm (1514).The least squares method was then used to fit a smooth Gaussian curve to the points. Various software packagcs for data acquisition and calculation for ICP systems have been reported (C31, C132, C591). I .2.I.5 Interferences. The interesting and apparently complex nature of interferences in the ICP is belatedly becoming recognized.Interferences only lead to errors if uncorrected and the recognition that some problems exist is thus welcome. Koirtyohann (C667) has reviewed interferences in the ICP, acknowledging the existence of physical, chemical and spectral interferences and identifying the latter as the greatest problem. A useful practical account of how inter-element (spectral and background) interferences were corrected in a direct- reader has also appeared (C768).Interferences arising from sample introduction have already been reviewed above (see Section 1.2.1.2). The high excitation energy of the ICP ensures that many lines are emitted and that spectrail line interferences are common. Boumans (C744, C874, 11 16) has continued with his valuable work of predicting such interferences using sensitivity ratios, concentration ratios and resolution.A list of degree of interference for some 900 prominent ICP lines for 6744 A iialy t ical A t omic Spectroscopy chemical elements in terms of the critical concentration ratio has now been published (1585). The scanning of the wavelength rangc of interest may be uscful in idcntifying interferences.A scanning system (C1147) and a software graphics package (C1383) to do this have been described, as has the hardware and software associated with a direct-reading polychromator that allowed the recording and manipulation of spectral information near analyte spectral lines (C738). The use of lines in thc v.u.v., as is used conventionally in spark analysis, has been advocated as a way of avoiding crowded parts of the iron spectrum in which interferences may be encountered in the analysis of steel (C1396).Uncorrected background emission may cause problems with any emission source. Back- ground shifts of a complex nature may occur when samplc is introduced into the plasma. Larson and Fasscl (374, C666) havc shown that spectral line broadening and radiative electron-ion rccombination processes may make significant contributions to the back- ground.Linear Stark broadened lines produccd spectral background at uncxpectcdl y large displacements from the line centre. Thc problems of background Correction when highly structured spectra, e.g., Ar emission lines, are superimposcd upon the continuum have been emphasized (1086).Certain emissions, e.g., from NO, OH, N2+, Ar, C or Si species, may crcate the need to use altcrnativc analytical lines. Papers havc appeared noting the advantagcs of difIerent approaches to background correction, e.g., by wavclength modulation (C592), the usc of video terminal to identify rapidly background correction requirements (C3 1, C6&), and by variable wavelength correction on a rapid scanning monochromator (164).Some of the most interesting developments have been those which have provided more data upon the complex efEccts of easily ionizable elements upon analytc emission. Ebdon (C763) has shown that Na has a significant effect on Mn emission for both ion and atom lines in Ar and N, cooled plasmas. The cffects arc non-lincar, normally significant cnhancc- ments, and severe at conditions of optimal SBR.It was also shown, howcvcr, that the ICP may be reoptimized for minimum intcrferencc with acceptable scnsitivity (C 1749), cmphasiz- ing thc importance of injcctor gas flow-rate and viewing height on these cnhancements. Watson (651) has studied thc cffect of Na on a variety of clcmeiits and also showed that thcsc varied with viewing height.The obscrvcd interfercnccs generally decreased with incrcasing powcr, but it should be rcmembcred that power will also affect thc relative positions of zones in the plasma. The cffccts of Na on Ca(I), Ca(Il), Mg(I), MglII), Fe(I1) and Ba(1I) were shown by Mermet et al. (C783’) to be more severe with a wet, rathcr than desolvated, acrosol.Complex cffects of K on Ba(1) and Ba(I1) emission signals have also bzen observed (1024). Potassium increased the emission of both species but the cnhance- ments varied with power and viewing height. Mutual “ionization interfercnccs” have been observed between alkali clements at viewing heights below 16mm, but abovc this level thc effccts wcrc ncgligiblc (C1382).An cxplanation for many of these obsarvations may be derived from studics of the different rcgions of the plasma as discusscd in Scction 1.2.1.1, but a satisfactory explanation of the mechanisms is still awaited. 1.2.2 Microwave-excited Plasmas Thc dcclinc in publications on niicrowave plasmas has bcen rcmarkablc, cspzcially €or capacitativcly couplcd plasmas where thcre have becn virtually n o new rcports.dc Galan (C711) has com’pared thc ICP and MIP and suggested that thc latter has few advantages except for non-metals. Thc results of this situation can clearly be observed in the growth of papers concerning the usc of the MTP as an clcmcnt specific dctcctor. I .2.2.I Fundameiital Studies. A vcry valuable report of the measurement of several funda- mental parameters in an atmospheric pressure MIP in a TM,,, cavity has appeared (1238).A tomiza,tiovt and Excitation 4s The stability of thc plasma, the intensity of the atomic emission from the Ar support gas and from the analytc species were proportional to the electron density.The latter varied with power and the presence of electrophili,c species, but the electron temperaturc did not as it i s related to the kinetic energy of the fastest moving electrons.The output power from a microwave generator has been shown to be affected by the load impcdance (1225) and consequently a method for measuring incident and reflected power in a NIP was described. The parameters that influence atornizatiolz in the MTP have been studicd (C138). The elxtron dcnsity and kinetic energy were shown to be critical.By holding one of thesc constant their effects were deduced. Samples atomized by a miniature graphite furnace have been introduccd into a MIP and the potential for AES and AFS studied (C8137). The fluoresccncc was excited by a pulsed tunable dye laser. Time-resolved measurements of transient emission signals were made to elucidate mechanisms. Vaporization interferences in a microwave plasma, e.g., the effect of phosphate on Ca and A1 on Ca, have been greatly reduccd by adding CH, to the ,4r (1 336).Spectroscopic measurement confirmed that this also promoted dissociation of refractory monoxides. In contrast, others have found that molccular dopant gases markedly rcduced Hg emission in Ar or He MIPS, prcsumably bccause the efficicncy of thc energy transfer mechanism was impaired (C719).Zander (C718) has reviewcd the plcrsnza, charcrcteristics of a He MIP in the now popular TM,,, cavity. The diszdvantages of plasma wandler upon sample introduction, tubc wall erosion and restriction of sample volume were noted as were the advantages of high energy density. Thc excitation tcmpcrature and electron density were profiled across thc axially viewed plasma.A novcl discharge tube for such a cavity, which allowed the introduction of an ultrasonically nebulized aerosol into a toroidal annulus, has been described (C587). A graphite furnacc atomizer coupled to a microwave generator using a coaxial probe has been used as a novel microwave cavity (C57). Such a device i s an intercsting curiosity because ETA is a favoured form of sample introduction for the MIP.Exploratory work on the coupling of an Ar atniosphcric pressure MTP to a quadrupole mass spectrometer for elemental analysis has been reported (C1162). Apparcntly, problems with clogging of thc orificc ljetween the plasma and spectrometer werc overcome by enlarging thc holc. 1.2.2.2 Sample Zntrodiccrion.The use of the MIP for AES has been revicwed with an emphasis upon samplc introduction problems (C156, 1584); gaseous sample introduction was preferred, for example, a GC effluent or hydride, or vapour from a graphite cord electro- thermal atomizer. Reasonable efficiency of pneumatic nebulization can only be obtained with Ar and thus the advantages of the He MIP, such as the ability to dcterminc halogens, are lost.Chloride has been determined in an MIP by introducing C1, gas, gcncrated from acidified KMnO, solution (1712). A detection limit of 10 ng was reported. A laser beam was used by rshizuka and Clivamino (1 133) to volatilize microgram amounts of metallurgical samples. The vapour was then drawn into an Ar MTP for the determination of Al, Cr, Cu, Fe, Mn, Mo, Ni, Pb, Ti and Zn.There are a number of features of the MIP that make it particularly attractivc as a gas-chromatographic detector. Udcn (C639, C745) has reviewed the wholc field of plasma emission spectroscopic detectors, noting the advantages of the TM,,,, cavity for use with modern ultra-narrow fused silica capillary GC columns. A He MIP has been applied to halogen specific detection following high-resolution gas-chromatographic separation of chlorinated products from the humic materials in natural waters (C1424).Another purpose built system using a He MIP in a TM,,, cavity has been utilized for pesticide analysis and for C and halogen specific detection in a number of model compounds (C756). Preliminary results on a novel GC-MIP system havc also been presented (C759) as well as further results from an established commercial system (C150).The latter instrument has capabilities previously only found in GC-MS systems and has particular application in environmental46 Analytical A tomic Spectroscopy monitoring, pharmaceutical analysis and determining H/D isotope ratios. A comparison of a He MTP detector, or as it is termed “an clectrodcless discharge photodetector,” with other GC detectors has been reported (C143). 1.2.3 D.c. Arc Plasmas The growth of reports concerning plasma arcs in which a portion of the flame-like plasma is transferred away from the primary arc column, known as d.c. arc plasmas, has continued. The majority of the reports are still confcrence papers discussing the application of com- mercially available systcms.Zander (C1369) has rcvicwed the opcrational advantagcs and disadvantages of spectrochemical d.c. plasma jets. 1.2.3.1 Fundamental Studies. The additiori of He to a d.c. plasma has previously bccn reported to give an improvement in pcrformancc (scc ARAAS, 1979, 9, Refs. 55, 603, 1322 and 1978, 8, Ref. 182). A full report of the advantages of this plasma, with 25% Ik a d d d to the Ar flow, altered electrode gcomctry and sample introduction chimncy, has now appcarcd (853).Phosphate intcrfercnce on Ca was apparently rcmovcd in this plasma. Thc mechanism of the enhancement effect of He has becn studied (C714), as has the cffcct of the diamctcr of the chimney on anion interferences (C1448). Hclium has also bcen intro- duced into the Ar stream in a conventional commercial system and again cnhanccmcnts were observed, but thc dischargc bccamc noticeably lcss stable because of the morc diffusc plasma produced (C7 15).Novel variations of d.c. arc plasmas are still being developed. A new miniature version can be operated at less than 7 A and less than 5 1 min-1 gas flow (C59). Samplc was intro- duced by two types of ultrasonic nebulizzr and promising dctcction limits were reported (C1431).Another new source with improved excitation capability has also bccn announced (C141). Comparisons with the ZCP of three electrode d.c. argon plasmas haw b x n madc. A systematic examination of d.c. plasma included somc refercnce to ICP characteristics (88 1). The advantages of the d.c.plasma were seen as low dctcction limits (approaching those of the ICP), high precision, good stability and three orders of magnitude dynamic range (ICP uscrs would generally claim to bctter this). The disadvantages were idxtificd as thc extremely small excitation region, the difficulty of achieving compromise multi-element conditions and alkali-metal interferences. A morc dircct comparison of a d.c. plasma and an ICP using the same echelle spectrometer and data acquisition systcm has also bccn made (C1367).A simultaneous 32-channel cchelle spcctromcter has bcen modified to allow rnicro- computer control and interfacing to a mini-computer (C53). This instrument has been fitted with both an ICP and d,c. plasma source and these may also be computer controlled.Pro- bably a very significant development in this area is the availability of an TCP source for the conventional commercial echellc spectrometer and this has prompted another comparativc paper (C139). The rcsults of this preliminary study appeared to be surprisingly inconclusive. Some interesting accounts of instrumental aspects havc appeared. Pyzik (C69) measured the emission profiles of 10 elements in the excitation region of the plasma, which is 1-2mm in size.The P profile continucd to show increased ihtcnsity into the plasma core but the othcr profiles peaked within the excitation region, so that the order of peak positions along the vertical leg of the invcrted-Y plasma source was : P, B, Mg, Fc, Ti, Si, Ca, Al, Na and K. It has also been found that there is a correlation between ionization potential and the size of the zone of maximum excitation for the alkaline-carth elements (C1432).The vertical position of the zone in which ions were found depcnded parabolically on atomic mass with a minimum for Ca, suggesting that Ar (of the same atomic mass) is playing a major role in the excitation process. A microprocessor-based control unit designed for useAtomization alnd Excitation 47 with the d.c.plasma has been described and demonstrated in various applications (C7 173. It was of particular value in providing a rapid semi-quantitative analysis that allowed the preparation of standards closely matching the sample, for a subsequent definitive analysis. A study of the emission using a Fabry-Perot interferometer has enablcd a number of parameters to be characterized (1 122).Improvements such as the introduction of samples as dry aerosols and methods for the reduction of noise were suggcsted. 1.2.3.2 Sample Znlrocluction. Continuous and discrete sample nebulization into thc plasma have been compared( (C574, C1174, C1430). For the latter a polypropylene funnel at the end of the pneumatic ncbulizcr capillary was fed with discrete drops from a pcristaltic pump.Thc nebulization efficiency, sensitivity, precision, linearity of calibration, sample requirements and interference effects were studied. Interfacing the multi-element capability of the d.c. plasma to HPLC or GC creates a versatile chromatographic defection system. The possibilities of such coupled systems liavc been reviewed (C1368, C1424).A ncw iiiterfacc for open-tubular capillary-column GC operation was described with detection limits at the pg s-1 levcl for compounds containing B, Cr, Fe, Hg, Pb, Si and Sn (C754). The major application of this system was reported to be in the specific element detection of B and Si in chcrnical derivatization GC (C639). A gas chromatograph equipped with a non-destructive Ar ionization dctector has been interfaced directly to a three-clectrode Ar plasma and some preliminary results reported (C1447).The use of thc 180.7, 182.0 and 18'2.6 nm V.U.V. S lincs enabled thc specific deter- mination of S containing compounds (C149, C736, 1312). The performance of the plasma detector was found to be superior to that of a flame photometric detector as a di-tection limit of 3X 10-10 g s-1 and 3 orders uf magnitude linear working range wen obtained.Thc stability of the plasma to quite high liquid input fluxes has enabled interfacing for HPLC detection (C73'6, C754). Size exclusion chromatography has been used to separate species in aqucous solutions, the effluent then being fed, via a refractive index detector, to the plasma (C757). 1.2.3.3 Interferences. Severe interference problems are known to exist in the d.c. arc plasma and various methods have been investigatcd to ovcrcome these. The greatest interest has again centred on the effccts of easily ionizable concoinitmts (see also ARAAS, 1979, 9, 15). These effects have becn reported in samples as diverse as orange juice (C1433) and coal-ash digests (C1434).A systematic study has been mad:: of enhancements for 14 elements duc to salinity and concentration (1223). This enabled the prediction of the different enhancements for ionic and atomic lines. Thcse results have bcen extended using stcpwise multiple-regression analysis to permit the analysis of sea-water (C1449). Johnson et al. (919) have shown that both ionic and atomic lines, for a wide range of elements were enhanced by easily ionized elements (e.g., Cs, Li). The con- comitants did not, however, cause significant changes in the electronic teniperaturc or density, although the rotational temperature was reduced, nor did they alter thc efficiency of sample introduction. These observations clearly show that a simplc explanation based on ionization suppression is not tenable and a tentative mechanism based on increases in the population of metastable Ar species was proposed. Enhancements by K, Li and Na have been observed to be directly related to the concentration of these elements up to a maximum of about 1 moll-1 (C7243. The decline after this level was attributed to spatial distortion of the plasma. Correction procedures for these enhancement effects have been proposed, viz.: use of an internal standard, which may however require computer manipulation of the data as the sensitivity to enhancement of different lines varies (C712), matrix matching (in this casc sea-water) or dramatic dilution (C722), and standard additions for brines (C729) and48 Aiialytical Atomic Spectroscopy for determining Ba in sea-water where Li was also added as a buffer ((2730). A warning was sounded to the unwary by a series of experiments demonstrating that the suppression of Ca emission by Al, which in any case varied with viewing height, also varied with other con- comitants (C1445). In this case Cs did not enhance the signal but caused an additional 5 to 10% suppression. Spectral interferences are encountered as in all emission techniques. The utility of an echclle grating polychromator with a capability for rapid high-resolution wavelength scans, which can be displayed visually for the identification of potential spectral interference problems,, has been stressed1 (C142, C1444). The mathzmatical approach of Kaiser (AXAAS, 1978, 8, Ref. 1033) to evaluating the selectivity of a mcthod has been applied to the d.c. plasma with an echelle monochromator to show that it offers almost ideal selectivity when considering spectral and inter-component interferences (1 106). Stray light cfiects have been observed from Ca and Mg using the 3-electrode plasma, but these could b? corrected by a simple linear equation (36). This report also showed clearly the improvcmcnt of this system over the %electrode plasma. The background characteristics of a d.c. Ar plasma have been studied (C716). The effect of high levels of Cr, F, Nb, Ta, Ti or W as concomi- tants on several useful analytical lines was evaluated and suitable lines for specified analyses identified. Concomitant salts when sprayed into thc plasma have been shown to increase the background both because of stray light effects and by shifting the continuum radiation level (C713).

ISSN:0306-1353    

DOI:10.1039/AA9801000034

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

48 Aiialytical Atomic Spectroscopy 1.3 FLAMES Interest in the processes of nebulization and the development of this and other mcthods for sample introduction to flames have again been reflected in the number of publications in this area. The potential of AFS and of Zeeman methods in AAS has been actively explored and advanccs in thesz areas are considerable. The role of lasers in analytical spectroscopy continues to be extensively investigated and exciting possibilities in atomic and molccular fluorescence spectrometry are envisaged.Perhaps the greatest attention this year has been devoted to interference effects and several important papers have been offered. A welcome departure from the usual observational accounts of interference effects was made in a review by Rubeska (1 304), which dealt comprehensively with the fundamental processes that result in interferences in AAS and AES. 1.3.1 Fundamental Studies No new flame types have been reported this year but work has continued (457) on the relatively new O,/C,H,/He flame discussed last year (ARAAS, 1979, 9, Ref. 1819) and) the much older N,O/propane or butane flame first introduced by Butler and Fulton, ( A y p l .Opt., 1968, 7, 213'1) has been revived for the analysis of copper alloys by AAS (234). Calculated compositions of air/N, and air/C,H, flames have been compared with the results of an experimental mass spectrometric study in which good agreement for molecular but not radical species was obtained ((21571). A laser method has been used to measure free-electron concentration and temperature in an airlnatural gas flame (272).The concentration field of Na in an air/H, flame was studied with an apparatus making use of resonance absorption to produce a monochromatic image of the flame which could be photographed (275). Flume rise velocities have again been studied by the well known single aerosol droplet technique (901) and by a particle track method (849).In the latter it was found that A1Atomization and Excitatioln 49 (10-40pm) or MoS, (12-35pm) particles were suitable but Mg and Sic were not because of their slowness in reaching the velocity of the air/COI flame (see also ARAAS, 1478, 7, Ref. 1187). Additional reference on the preceding topic - C1048. A theoretical model of the typical optical arrangement of a practical AA spectrometer, rather than the idealized form assumed in the Beer-Lambert law, has been devised (C124).This has been used to study the effects of spatial non-uniformity of the atom density in the optical path as well as temporal fluctuations. The former can lead to reduction of the slope and curvature of the calibration graph at high optical densities. Temperature fluctuations were considered to be unlikely in practice to havc significant effects on working curve linearity. The influence of typical AAS geometry on non-linearity of working curves and similar effects produced by analyte concentration fluctuations were also discussed with the aid of a theoretical model (C123).A further study of this matter (1720) has led to a suggested burner design in which the burner shape corresponds to the shape of the optical beam.The theory of a new form of analytical atomic spectroscopy, coherent forward scattering spectroscopy, was discussed and an experimental demonstration given for Pb in the concentration range 0.1-10 ng (879). Encouragingly good agreement has been obtained in a comparison of three different methods of temperuture measurement in laminar O,/H,/N, flames (343).The methods used included Raman scattering, line reversal and the two-line technique. Other studies using Raman scattering (97’5) andl saturatcd laser fluorescence excitation methods (342) of flame temperature measurement have been made. The two-line AES method has been reviewed (328). In a review of noise and signal-to-noise ratio, the advantages of double-beam instru- mentation and of internal-standard operation for minimizing multiplicative noise, ix., that introduced simultaneously with the analyte signal, were identified r1213). A discussion and comparison of noise in single- and dlouble-beam instruments has been presented for air/C,H, and N,O/C,H, flames (1207) (see also ARAAS, 1979, 9, Refs. 929 and 1199).Several spectrochemical techniques, FAAS, FAES and d.c. plasma emission, were characterized for sources of noise that limit precision at concentrations in an “optimal” range well above the detection limit. Graphs of precision versus concentration were presented, and methods for cxtending the useful range to higher concentrations, such as reducing the optical path-length, discussed (1407).Another invcstigation of shortened path length has confirmed that rotated- burner operation does not adlversely affect precision (194). Prediction of noise in AAS has been demonstrated with the aid of experimental noise power spectra in the region 0 to 5 Hz, reported for Cu and Zn analytes in an air/C,H, flame (373). Slot and Meker burners have been compared for precision and noise in the AES mode (455).Interest in Zsemm and other magnetic eflects in atomic spectroscopy continues to grow and has resulted in a number of theoretical and experimental studies as well as new instrumentation. A unified theory of magneto-optical phenomena, using the Jones calculus, has been applied tcr the quantitative description of instrumentation using the Zeeman, Faraday or Voigt effects (1243).The contributions of magnetically ind’uced dichroism and the resonant Voigt effect in the AAS determination of Ag have been investigated (1262). A careful arid revealing comparison of conventional and Zeeman AAS, using apparatus suitable for flame or graphite furnace atomizers, concluded that, of the 44 elements tested, only Bi, Cu, In and Li showed an appreciable (> 2 x 1 loss of sensitivity in the Zeeman mode.This study did, however, also demonstrate that for certain elements at nioderatc concentrations, and for many at high concentrations, severe curvature or even double-valued analytical curves were obtained with Zeeman methods. Despite this limitation however, the authors were satisfied that in general the Zeeman approach provided superior background50 Analytical A tomic Spectroscopy correction and demonstrated it with several difficult matrices, with only a small loss of sensitivity or analytical range (940).Both theoretical and experimental methods have been applied to investigating analytical curve shape in Zeeman AAS (814, $60). AAS measurement of Li isotopic abundance : 1510.Internal standard AAS operation for Ca : C85. Relative atomic transition probabilities : 389. Relative oscillator strengths, Ti : 428. Review, early work, cyanogen flame AES and long path AAS, Cd and Zn : 1260. Significant improvement in sensitivity by parameter optimization : 1599. Other references of interest - 1.3.2 Atomization and Interference Studies Investigations of free-atom fractions and atomization processes have been reported for different elements in the air/C,H, flame in which the role of oxygen in restricting atomization was discussed (C1566). In a discussion of the 01,/C,lM,/He flame, the importance of the He dilution in enhancing the free-atom fraction for the elements Ba, Ca, Cr, Cu, Fe, and Sr was emphasized (16683.The authors concluded that the atomization capabilities of this flame were intermediate between those of air/C,H, and N,O /C,H, and that its background cmission was weaker than that of the latter.A suitable burner was also described. Spatial variation of OH radical concentration and its effect on atomization have been discussed ((373). In a study of atom trapping (ARAAS, 1979, 9, Ref. 1464?, a role for the sputtering action of the scouring flame gases in atomization has been indicated that may have relevance to atomization mechanisms in ETA (508).Additional references on the preceding topic - 379, 4891, 523), 994. An authoritative review of interference effecls in flame spectroscopy has been published by Rubeska (1304). He has adopted the difficult but rewarding approach of presenting the basic processes, on a theoretical level, which give rise to the confusing observations of interferences.Voinovitch et al. (1 247) have adopted a comprehensive and novel approach lo the study of interactions between elements in AAS, by relating them to their electronic configuration as reflected in the Periodic Table. Some 34 elements were studied pair-by-pair with element ratios varying from 0.01 to 100.The authors were able to assign interferences to four categories according to their positions in the Periodic Table. Further additions to the already voluminous literature on interferences on specific elements have been made. These include a detailed theoretical and experimental study of the effect of pH on Zn in FAAS (133\7). Diepression at pM 2 7 was attributed to the formation of particles of Zn'(OH),, which are partly lost in the nebulizer spray chamber, while at low pH coprecipitation on A1 and Fe hydroxides occurred.Losses by the first mechanism could, it was suggested, be a measure of nebulizer efficiency since it was strongly dependent on aerosol droplet size. Other investigations have been made into the elimination of sulphate interference on Ca by the presence of carbon black (0.008%) in the solution (316) and into the mechanism of sulphate and phosphate interference on Ca (68'1).The vaporization interference mechanism proposed entails a non-linear relationship between absorbance and concentration that could result in a 10-20% error in analyses where a simple standard addition procedure is used.The interferences of Ca and Si on Fe determination in an air/C,H, flame were found to be minimized in lean flames and by the presence of 2% m / V NH,C10, (1078). Optimal operational parameters have been obtained for the mini- mization of CaOrH and SrOH interference in the determination of Ba and Li, respectively, by FAAS in an air/C,H, flame (1054, 15819). Interactions between Cr and Fe have been attributed to the formation, by mutual catalytic oxidation, of refractory oxides of theA tomizaltion and Excitatiort 51 elements (489).The selectivity of hi&-resolution d.c. plasma OES has been compared, in terms of spectral and chemical interference, with FAAS for the analysis of samples such as sea-water, serum and orange juice (1 106).The authors conclude, perhaps a little optimistic- ally, that the d.c. plasma technique is almost ideal in these respects, whereas sample pre- treatment is required for FAAS. Additional reference on the preceding topic - 379. The effect of cmnplexing @gents such as EDTA in enhancing Ba in FAAS ((28'2) and of KCN in eliminating interference by anions and other complexing agents in the determination of C o (960) have been examined.The influence of MIBK and o-xylene on the atomization of Co, Cu, Fe, Mn, Ni and Zn complexed with 1 -phenyl-3-methyl-4-bcnzoyl-5-pyrazolone has been studied (356). Two studies of the enhancement by nitrogenous compounds of A1 AA in an air/C,H, flame have been presented; the most effective compounds were tertiary amines (518). The enhancement in an 0,-sheathed air/C,H, flame was attributed to an increased formation of Al(OH),, to a reduced and more uniform aerosol particle size, and to a more reducible Al(OH), clotlet (530).A comprehensive series of studies of the molecular absorption spectra in the air/C,H, flame of salts of the elements Al, Be, Ce, Cr, Fe, Tn, La, Mn, Mo, Pr, Si, Sn, T1, V, and W have becn reported and thei,r role in FAAS interference evaluated (1352, 1364, 1365).Other references of interest - Discussion of the lateral diffusion interference effect : 1716. Effect of organic solvents on flame geometry : 252. 1.3.3 Devices for Sample Introduction Recent investigations of the fundamentals of nebulizer operation have significantly advanced our knowledge of the principles involved.There is general agreement on the need for an appreciation of the basic aerodynamics of nebulizer design (11) and of the importance of solvent vaporization effects in controlling the analyte signal (C1370). I .3.3.1 Nebulizers. New techniques for estimating aerosol droplet size and distribution include the use of cascade impactors (I 187, C1463). The saturated vapour pressure increases as droplct size decreases but this can be compensated by increasing the salt content of the solution. By using this principle it has becn demoinstrated that evaporation from droplets as small as 0.51pm is negligible at high enough salt concentrations (11%). Laser forward scattering methods have also been used to study droplet size and distribution in a new design of nebulizer (829).The cascade impactor technique has been used to study the influence of nebulizer component parts such as impact beads and paddles, of auxiliary air flow-rate and of nebulizer uptake rates (8156). Not surprisingly, the advantages of impact devices, viz, fewer large droplets and an increase in large droplet fragmentation, have been confirmed. It was also shown that as sample uptake rate is increased mass transport reaches a peak or a plateau.The less than expected effect of temperature on nebulizer efficiency has been attributed, in one study, to the fact that the increase in viscosity produced by a lowered temperature is partly compensated for by a concomitant reduction in droplet size (473). Adsorption losses on nebulizer surfaces have been shown to be important in unacidi- fied solutions at concentrations below 1 pgml-1 but were eliminated by the presence of 0.05-0.2M HC1 in solution.The relationship between the amount adsorbed and the solution concentration was derived (C1548). Additional reference on preceding topic - (21563. New nebulizer designs includc one using the centrifugal principle (279), others which are demountable ((283, C1156) and one making use of a stepped throat to increase sensitivity52 Analytical Atomic Spectroscopy (C1465). A new design has interchangeable flow spoiler and impact bead and while the bead improved limits of detection in an air/C,H, flame it gave a poorer RSD in the N,O/C,H, flame (108'9).Currently available nebulizers have been surveyed in terms of design, efficiency and droplet size and distribution (996). 2 .3.3.2 Microsampling Devices. Comparisons of discrete sample nebulization with conven- tional nebulization have attributed the higher nebulization efficiency of the former to changes in the aerosol droplet size distribution (1250; C1457). This benefit has bccn applied to the determination of 10 elements in several standard reference materials, including NBS SRM 1577 (Bovine Liver), using a micro-digestion (2mg sample) procedure (505).In combination with long-tube AAS using an air/H, flame the technique was claimed to pro- vide detection limits, for a range of elements, which approach those of ETA methods and with supcrior speed and precision (48'3). Discrete sample nebulization has been shown to offer a solution to the problems involved in nebulizing chlorinated hydrocarbons into flame atomizers by continuous methods (9 1 1).An automatcd instrument has been described which incorporated continuous flow injection analysis and, using continuum source FAAS, was able to determine up to 16 elements simultaneously at a rate of 60 samples h-1 in samplc volumes of 25-300 pl (C74).Flow injection analysis has been applied to the determination of Zn in whole blood, plasma and to NBS wheat and rice flour (SRM 1567 and 1568) digests with good results in sampk volumes of 20-50 pl (C77). The design of a manifold for flow injection AAS has also been described (897). 2.3.3.3 Somple Introduction by VolatiCizatioiz. Development of the flame microsampling (Delves) cup technique has involvedl automatic sampling and programmable analogue integration (1516) and this has been applied to the determination of Pb in air particulates by direct analysis of the cellulose filter discs used to collect them (1517).Good agreement with XRF and conventional FAAS was obtained and a precision of 2 6-15'% quoted, with cycle times of 33s. The allied technique using a platinum loop device has achieved detection limits 20-500 times superior to those obtained by discrete sample nebulization methods for a variety of volatile elements '(C 1602).The electrically heated chamber- or capsule-in-flame method, which uses a mixture of the powdered sample with graphite, has undeservedly received little attention outside the USSR but has been the subject of continued research and application there.hplications for this method were drawn from a theoretical consideration of the geometrical form and size of the atom source in carbon-filament and chamber type atomizers (I 327). Applications included the determination of Li and Rb in silicate ores (1001) and Co and' Fe in biological materials (224). An electrically heated tungsten filament has been used to volatilize elements into an O,/H, flatme for determination by AES.A precision of 4 5% was obtained and only 3 pl of sample solution was required, but a more sophisticated background correction system appeared to be desirable (180). Arc-volatilization into a flame for AAS has bcen applied to the determination of Cu, Fe, K, Mn and Pb in different matrices such as glass, rock, oil and metal alloys (1670).The vapours from a carbon cup furnace, drawn into the dual flame of a Cu or In sensitized Van der Smissen burner, have been used to determine total chloride by measuring the Cu or In chloride band emission (C1155). 1.3.3.4 Direct Analysis of Fluid Suspensions. Despite thc obvious attractions of direct nebulization of emulsions, suspensions and slurries into a flame for AAS or AES analysis, this technique is still limited by practical difficulties.A few reports, however, are of interest.Atomization ond Excitation 53 A detailed investigation of the analysis of soil suspensions concluded that the atomization efficiency was, understandably, dependent on soil concentration, particle size and mineral composition.However, by using a standard suspension of a soil, typical of the samples analyzed, results for Cu, Fe, Mg and Mn accurate to about 15% were obtained with considerable saving in sample preparation time (C1605). Aluminium has been extracted with oxine into MIBK, emulsified with water, Tergitol XDI and ethanol, and nebulized into a N,O/C,H, flame. A detection limit of 0.07mg 1-1 with an upper limit of 3’mgl-1 was achieved (1608).Water emulsions of petrol have been used for the determination of Pb in various alkyl forms with no interference effects and good agreement with a standard method (173). Additional reference on the preceding topic- 12. 1.3.3.5 Chromatographic Detectors. The use of spectroscopic, element-specific detectors for chromatography is now firmly established and molecular band emission methods for P have been further developed for modern liquid chromatographs with their reduced flow rates ((333) and for HPLC where the use of an air/H, flame in “reversed” mode has been described (939).A detailed investigation of FAFS as an HPLC detector reported on per- formance and compared it with that of conventional U.V.spectrophotometric detectors (C1482). In a comparison of a FAAS gas chromatographic detector with a katharometer and a flame-ionization detector, the FAAS detector was found to be superior for the deter- mination of Ga and In organometallic compounds (1029). Additional references on the preceding topic - 1030, 1031, 1.3.4 Atomic Fluorescence Spectrometry The expanding interest in AFS is reflected in the number of papers and reviews presented this year although there have been more discussions of its potential than of its analytical application to date.The history and potential of AFS has been reviewed with emphasis on recent advances in high radiance Xe arc continuum source AFS and predictions of further advances in multi-element analysis (1421). A comparison has been made of the different types of AFS and d.c.plasma and ICP emission techniques (C28). A report of three different instrumental approaches to the application of AFS in clinical analysis (C201) gave some practical examples of the determination of Cd, Cu and Mg in urine and pointed to developments in combining AFS with hydride generation for the determination of Se (see also Section 1.5). Interesting and novel applications have been made of the inductively coupled plasma in AFS for scattering correction and for the determination of, for example, Fe in fly ash by ICP emission using air/H, flame fluorescence as a resonance detector (1186).AFS with an ICP as a continuum source has been comparedl with line source (EDL) AFS, in an air/C,H, flame, for 20 elements.Although ICP-AFS exhibited the advantages of simultaneous multi- element analysis, its detection limits were inferior to conventional AFS by factors of 2 to 10 Several new developments in sample introduction to AFS atomizers have appeared this year. In a design for a miniature N,O/H, burner, provision was made for mounting a graphite cup in the burner tube from which sample was volatilized by electrical heating and the vapour carried by the H, fuel into the flame.Improved sensitivity, compared with direct nebulization, was claimed and reduced molecular interference anticipated ((32’). Gaseous S compounds, namely H,S and SO,, have been used for the determination of S by AFS in a N,/H, flame using a Xe arc continuum source (C51). An apparatus has been described for the determination of W in which the sample was chlorinated by SOCl, at 553 K in a sealed quartz Carius tube.The WCl, formed was volatilized into a quartz reactor where it was (9m15).54 Arialyticnl Atomic Spectroscopy reduced at 1100 K by a H,/He mixture prior to AFS determination. Rock, plant and water samples were analysed with a detection limit of 0.8 ng and an analytical range of 34-900 ng (1 339).Other references of interest - AFS Pb monitor for smoke stacks : 1072. AFS theory: 305. Eight channel non-dispersive AES instrument : C1178. Tndirect determination of P by AFS : 491. On-line AFS monitor, Mg in boiler feed water : 506. Study of air/H, flame : 306. 1.3.5 Applications of Lasers Interest in the analytical applications of laser technology has continued at a high levcl although commercial exploitation of the novel ideas and measurements in this field still appears to be a long way in the future.Several authors have presented reviews or review lectures in which different techniques have been compared, and possibilities for the future have been assessed. Falk (351) presented a comparison of methods based on lascr-excited atomic sourccs for quantitative analysis, whereas Winefordner (C 1420) widened his discus- sion and compared atomic and molecular spectrometic methods.A review of laser-enhanced ionization in flames has appeared (12951, and applications of single atom detection by rcsonance ionization spectroscopy have also been discussed (C6731, although the author was principally concerned with applications of interest to physicists. I .3.5.1 Atomic Fluorescerice Spectrometry.The measurement of pulsed dye laser excited AF signals in an TCP, reported in a conference paper last year (ARAAS, 1979,9, Ref. 1373), was said to be poorer than either flame AFS or ICP-OES (4'79). The reasons given were the greater background and noise of the TCP compared to the N,-separated air/C,H, flame, and to a small extent the lower light throughput of the optical system required for the operation of the laser with the TCP.Epstein et ccl. (855, 1126) have applied laser-excited AFS to the determination of Ni and Sn (855) and Fe (1126) in several NBS SRMs. An N,-separated air/C,H, flame was used and detection limits of 0.5, 3 and 0.2 ng ml-1 were reported for the above-named clements, respectively.Serious spectral interference from Fe on Ni excited at 300.3nm was observed, but was removed by reducing the laser bandwidth to 0.002nm (855). For Fe, the Stokes direct-line AF transition, 296.7 nm excitation / 373.5 nm emission, was used (1 126). The theory of spectrophotometric detectors suitable for lascr AFS determinations has been reviewed (38).Laser-excited fluorescence methods for the measurement of flume temperatures have received increased interest. Bradshaw et al. (1 594) described the theoretical basis of five methods, one based on the linearity between the AF signal and the laser spectral irradiance and three based on saturation. The laser-induced molecular fluorescence of OH was also used to measure flame temperatures (620).The method was based on matching the observed spectrum to a numerical model in which the detailed balance temperature is a parameter. Saturated laser-excited fluorescence spectroscopic methods of temperature measurement based on atoms and flame species haw been described in an NRS special publication (342). Results on thc observation that the intensity ratio of the two AF lines for the yellow Na doublet departs systematically from 2 in a wide range of H, and C,H, based flames, which was reported last year (ARAAS, 19'79, 9, Ref. 1348), have now appeared in the literaturc (271). Polarization of sodium atomic fluorescence at 589.0' am produced by lascr- induced dissociation of Na, has been reported (1 646). Near-rcsonant Rayleigh scattering inAtomization and Excitation 5s a flame in the region of the Na resonance line has been observed and shown to be in agree- ment with theoretical predictions (230).Saturated laser fluorescence measurements of Li and Na in fuel rich O,/H,/N, flames at 17OQ-2200 K have indicated significant depletion of the excited state atom populations by reactions with flame molecules such as H,O and H, to form metal hydroxides or hydrides (1 302).A steady state redistribution occurs between atomic and molecular forms of the element within the initial fraction of the microsecond laser pulse. Use of nanosecond pulse-length lasers is required to circumvent these chemical relaxation reactions. Frueholz et al. (1593) have described a new method that gave an improved SNR for CW laser-induced AF detection of Na in an air/C,H, flame.The method is called harmonic saturated spectroscopy and takes advantage of the non-linear response of an optically saturated atomic transition to reject background interferences. The dependence of the harmonic signal on intensity, excitation waveform and depth of modulation was reported. Other references of interest - Laser fluorescence spectroscopy of U in a pulsed hollow cathode : 348.Laser fluorescence study of reactions of Ba with HCl and HBr : 1647. 1.3.5.2 Moleculcrr Fluorescence Spectrometry. Two interesting papers in this field appeared during 1980. The molecular fluorescence spectra of polynuclear species have been detected by laser excitation of premixed air /paraffin and air/gasoline flames (848).Maximum emis- sion was observed at low flame heights (<2 mm). Continuous-wave dye-laser induced fluorescence emission and thermal emission spectra of YO molecules were measured simul- tanously in a O,/H, /Ar flame at 2430 K (1214). The differences in the spectra were used to show that collisional de-excitation of an excited vibronic level takes place by vibrational relaxation, decay to the electronic ground state and by intermultiplet transfer in order of increasing probability. I .3.5.3 Loser-enhmced Zonizaticun Spectrometry (LEI).Apart from the review (1295) men- tioned earlier, most of the reports concerning LET that appeared during 1980 have been in the form of abstracts of conference papers and full details are not readily available.It is clear however that authors working in this field have turned their attention, from the pro- cesses leading to enhanced ionization (see ARAAS, 1979, 9, Section 1.3.4.3), to the mechan- ism of signal or ion collection at the electrodes, in particular in an attempt to diminish or remove interference effects. Travis et al. (C1441) described as experimental technique used to devise a vertical profile of the laser-produced iojn flux at the collecting electrode.The position and width of this “ion image”, as a function of applied voltage, have been correlated with theoretical models in order to study the relative importance of the flame velocity, diffusion, electrostatic repulsion, and field-induced mobility in determining ion. trajectories. Experiments indicated 100% efficient ion collection.Znterferences in LEI due to thermal ionization of low ionization potential matrix elements (see ARAAS, 1979, 9, Section 1.3.4.3), which cause the formation of a positive ion sheath at the cathode and signal suppression, have been overcome by a novel electrode arrangement ((21443). Normally the burner head i s used as the anode and a pair of cathodes for ion collection are mounted just outside the flame on opposite sides.In the new arrangement, a single water-cooled cathode was placed directly in the flame, just above the laser beam and 1 cm from the burner head anode. Water cooling prevented electrode deterioration, and no contamination or memory effects were observed. Tolerance to a matrix containing Na was improved from 300 to SOOOpgml-1.Another group from the University of Arkansas have reported extensive investigations of electrodes for LEI and56 A naly ticd A t omic Spectroscopy have suggested a number of methods of reducing interference effects from low ionization potential matrix elements ((2554, C578). Their modifications included the use of plates rather than rods as cathodes, and the correct optimization of the plate area and the position and diarnetcr of the excifing laser beam in relation to the plate cathodes (C578).They also investigated the use of high voltage electrodes to remove interfering species before LEI dctection, and the effect of using a pulsed applied voltage, of varying pulse width and varying synchronization with the laser pulse, to this collector electrode ((3543.Surface studies of LEI electrodes were reported which indicated that acid solutions leached analyte off the previously electroplated cathodes (C567). Two papers have described the mechanism of shifts in spectral line profiles in LEI spectrometry (C553i, C555). Piepmeier et al. (CS55) have defined a model that predicts the wavelength shift of the spectral line absorption profile of a transition approaching laser saturation.Both papers indicate that the wavelength chosen for maximum LEI signal will depend on how close the transition is to being saturated. Two step or stepwise laser-enhanced! ionization of sodium in an O,/H,/Ar flame has been achieved using two co-linear laser beams (C1439). The Na LEI signal was enhanced by a factor of 150 compared to normal LEI (see also ARAAS, 1979, 9, Ref. 1987). Stepwise photo-ionization of Cs and Li in an air/C,H, flame has also been described with detection limits of 1 X lo.-* and 3X 10-5 pg ml-1, respectively (1498, 1530). The same authors have reported the detection of 1 fg of Na using a graphite cup atomizer (892). The atomic vapour releascd at an atomization temperature of 1900 O C was exposed to radiation from a single Nd3+ laser and two (nitrogen-laser pumped) dye lasers operated at 1064, 590.01 and 568.8 nm, respectively.The resulting ions were detected using a tungsten wire electrode placed over the graphite cup and operated at 600V. A SNR of 201000 was reported for 101pg of Na. Optogalvanic detection has been proposed as a method for the direct measurement of the intensity of resonance radiation (0514).Thus the optogalvanic signal from the atom cell acts as a combined resonance monochromator and photocell. Implications with respect to alternative designs of atomic absorption instruments were considered by the author. Ultraviolet photodissociation of simple metal salts has been shown to be a convenient means of producing laser action on the resonance lines of many metal atooms (C1467).Such lasers are said to be simple and cheap to construct, involving a heated sealed-off sample cell and a fast rare-gas monohalide flash lamp for a pump source. Each laser was locked to the resonance wavelength of the metal used with a bandwidth adjustable to a limited extent by varying the dissociating-pump wavelength and the buffer-gas pressure in the sample cell.The use of these lasers for the detection of two elements by LEI was reported. Other references of interest - Tnfrared optogalvanic spectroscopy of He : C6'74. Optogalvanic spectroscopy in hollow cathode discharges : 1591, 1597. 1.3.5.4 Other Studies. Novel applications of lasers in spectroscopy are reported every year and appear to increase as the sophistication and flexibility of laser technology increase. Although many of the papers in this area are likely to be of interest only to theoretical spectroscopists, a number have genuine analytical interest and others may have potential analytical application that is hard to discern at present. The lifetimes of the excited states of atoms in an air/C,H, flame have been measured using a novel cross-correlation technique (Cl440). With fast response detectors, comparisons in time of the AE signal with the pulse train of the excitation laser, allowed the measurement of excited state lifetimes of several elements capable of absorbing either the fundamental or second harmonic of rhodamine 6G. The optical delay could be scanned with a 3 ps time resolution.Atomization and ExciWon 57 Fluorescence radiation generated from two atoms during a collision process has been observed (1 595). The emission, termed radicrtfve collisionai fluorescence, was at a wavelength corresponding to the sum of the energies of the thermally excited electronic states of the colliding species, and was generated from mixtures of Ba and T1 in a tube at 1700 "C. The fluorescence phoron burst method was used to detect individual atoms of Na in a helium buffer-gas with a SNR of 5.1 (1590~. The motion of the atoms across tbe laser beam could also be monitored by the technique developed by these authors. Other references of interest - Observations of Ca2+ and Ba2+ in atomic vapours excited by laser beams : 622. Optically induced magnetization of alkali vapours using non-resonant laser radiation : 1592. Spin-orbit relaxation of excited Pb atoms in O,/H,/N, flames : C825 (see also ARAAS, 1979, 9, Ref. 396).

ISSN:0306-1353    

DOI:10.1039/AA9801000048

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

Atomization and Excitatton 57 1.4 ELECTROTHERMAL ATOMIZERS The past year has seen continued intense activity in the field of atomizer improvements. There now appears to be a general realization that isothermal heating i s necessary for minimum interferences and maximum sensitivity. Towards this end, the graphite platform previously proposed by L’vov (ARAAS, 1976, 6, Ref. 11411, has been adopted by many workers and its utility is gaining widespread acceptance.What is however surprising is the ad hoc use of various matrix modifiers. It is to be hoped that a more rational use of these aids will develop in the future. 1.4.1 Atomizer Design Atomic spectroscopic methods have often been criticized in that they are not absolute methods of analysis. The concept of an absolute rnerhod in ETA, originally expounded by L’VOV (see ARAAS, 1978, 8, Ref. 1392) is without doubt an attractive one. Chakrabarti and his co-workers (C612, 1057, 1070, C1167, C1458) may have achieved significant progress towards this goal. The capacitively heated anisotropic graphite atomizer reported previously (see ARAAS, 1979, 9, 25) is now claimed to be “almosty’ the ultimate in ETA.The authors state that it is free from all interferences, since all of the sample is atomized. The method requires only single-point calibration, since using appropriate proportionality factors, absorbance is directly related to analyte mass. The papers on this work are of great interest since they may well be describing what is close to the ultimate in electrothermal atomizer design.Such extravagant claims, however, coming in the wake of reality about the ICP, make the publication of an independent appraisal of ETA heated by capacitive discharge an eagerly awaited event. Much work on attempts to achieve a closer approximation to atomization under isothermal conditions is still in evidence. Woodriff and Nichols (C114) have further modificd their constant temperature furnace to give greater temperature uniformity (both spatial and temporal).To their “T” shaped atomizer, these workers have added two end posts which conduct current from the electrical connections at the base of the furnace assembly. This modification was claimed to reduce temperature variations significantly along the 26-cm length of the atomizer, giving less than a 30K temperature gradient. The more isothermal environment was claimed to allow interference free atomization, at often as little as 300 K above the analyte appearance temperature. Similarly, Slavin et al.(509, C605) modified a commercial ETA of the Massmann design, by contouring the atomizer tube; in this case the temperature gradient was reduced to 100 K. No significant improvements in sensitivity were recorded by the authors.An interesting modification to the mini-Massmann type atomizer was described by Lawson and Woodriff (1261). By replacing the conventional58 Analytical Atomic Spectroscopy centre supports for this furnace by two-pronged supports that held the atomizer tube at both ends, the current was made to flow across the atomizer ends thus heating the centre by conduction only. This resulted in an initial temperature gradient of 700~900K between the centre and the ends.Interferences on Pb were decreased, presumably due to a reduction in condensation at the tube ends and a hotter vapour-phase environment for atomization. Use of the L’vov platform (see ARAAS, 1976, 6, Ref. 1141) continues to be advocated by many workers and Slavin and Manning (1256) have reviewed the “state of the art” regarding this device.The advantages to be gained from the use of this simple modification are now well documented and it has been shown to be a genuinely practical way of reducing molecular interferences. One unforeseen advantage of platform atomization was reported by Ottaway and co-workers (1233, C1724) who found it most effective when measuring AE signals from a graphite furnace.The platform system was used with an echelle spectrometer- based instrument, custom built specifically for emission work. Compared to previous work, improved detection limits were reported for 14 out of the IS elements studied. Another device designed by L‘vov, intended for sampling into a constant thermal environment, was the tungsten loop (see ARAAS, 19178, 8, Ref. 1027). This sampling technique has not reached the degree of development of the platform, and Manning and Slavin (C104) described problems associated with tungsten carbide formation resulting in melting of the probe. When they replaced this by a flat graphite rod or probe, higher working tempera- tures could be used, although analytical precision was poor.The authors attributed this to the variable porosity of the graphite substrate. A modified approach using a similar tungsten-probe sampling system was described by Cedergren et al. (C1170). With this devicc the pyrolysis products, formed during vaporization, were mixed with a gas buffer then passed through two equilibrium zones. Since the residence times of the atoms in this system were of the order of seconds, there appears to be a higher probability of these atoms attaining a state closer to thermal equilibrium.High temperature equilibrium calcula- tions were used to determine the ideal composition of the gas atmosphere. The study of atomization surfaces is of interest now, as workers become more aware of the effects of surface interactions on atomization.Almeida and Seitz (C96) have described a technique for the preparation of carbide-coated tubes. Graphite tubes were soaked in solutions of Mo, Ti or W chlorides under slight negative pressure. The chlorides were then hydrolysed prior to firing the tube at high temperature to form the carbide. Titanium and W carbide coated tubes proved the most successful; tubes coated with Mo tended to deform and chip.Verwolf et crl. (C100) have advocated the use of vitreous carbon rather than pyrolytic graphite as an atomization substrate. They have studied the physical and electrical properties of vitreous carbon and concluded that only minor design changes would be required to present atomizers to allow the use of this alternative.Another interesting surface modification was decribed by Bath (CllZ), i.e., Cu was electroplated on to a standard graphite tube as an aid to the atomization of As. Two-fold enhancements of the As signal were recorded compared to standard tubes, but no enhancements were found when Cu was added to the sample solution. Papers on the use of metals as crtumizers were less prevalent this year and only a few workers continue to investigate this alternative.Sychra and co-workers ((21549) have further improved their W tube atomizer to facilitate controlled heating at rates of up to 20 K ms-I. Bezak and Vasiliev (1 109) have employed a Ta foil boat (Donega and Burgess, Anal. Chem., 1970, 42, 1521) atomizer operated within a vacuum chamber. Low pressure operation was found to be detrimental to sensitivity.Other references of interest - Sample introduction : 1496. Temperature measurement : 1684.Atomization and Excitation 5 9 1.4.2 Atomization and Excitation L’vov and Pelieva have continued their invcstigation into the mechatiism of Qiomizution involving the formation of gaseous monocyanides (sce ARAAS, 1979, 9, 26). They have calculated the atomization energies for gaseous molecules of alkali-metal monocyanidcs (895), and Ba, Ca and Sr monocyanides (1342).In thc former case, the values calculated agreed well with tabulated dissociation energies of monofluoride, whcrcas in the latter work, the existence of the metal cyanide molecules was confirmed by identification o€ their emission and absorption molecular spectra.In further work (1 6885), a statistical thermo- dynamic approach was incorporated using a rigid-rotor / harmonic oscillator method to calculate the frce-energy function. Equilibrium constants were calculatcd from integral absorbance values and hence dissociation energies for first row transition-metal mono- cyanidcs determined. The author concluded that the gaseous monocyanidcs were morc stablc than the chloridks.In a parallel study, L’vov and co-workers (1574) calculated the heat of formation of C, molecules in graphite furnaces at 3000-3400K. They compared two techniques, involving emission and absorption at the edge of the Swan band at 516.5 nm and determined values of 891 -+ 37 and 841 3- 3i7 kJ mol-1. Using crror analysis, the absorption method was shown to be inferior in accuracy to the emission technique.Kitagawa et al. (472) have calculated degrees of atomization for all atoms cxtrapolated from expcrirnental measurements of CuCl molecular emission and Cu atomic emission. The degree of atomiza- tion was found to be not purely a function of ternpcrature and could be doubled by the introduction of a small percentage of H,.Several methods md models have been proposed by numerous workers as an aid to the understanding of atomizaifion processes. One of the most cornprehcnsive was that advocated by Frech and co-workers (C1171, C1460) who employed high temperature equilibrium calculations. This approach has been used to advantage to predict the optimum conditions required for the atomization of Si (478) and P (1704).Silicon atoms were formed from SiO(g) or SiC(s) depending on temperature. Phosphorus was shown to be formcd via i t s oxide species PO. and POI,, at temperatures below 1800 K, whereas above this, an equilibrium bctween P, and P was predicted. In this study, the molecular spectra WCR measured to verify the theoretical predictions. Such calculations have been shown to provide reliable results, provided that equilibrium can actually be achieved, i.e., atomizers arc operated under isothcrmal conditions.In commercial non-isothermal atomizers, however, the kinetics of the processes must bc considered. Akman et al. (1220) described a simplified mathematical analysis of atom formation in such a commercial atomizer, using first order kinetics. A kinetic approach was also used by L’vov and his group (1531) to investigate the atomization mechanism for group IB and VITIB elements in both graphite and Ta-lined commercial atomizers.The activation energy was shown to be dependent on the nature of the surfacc but independent of the condition of the surface. Tn elegant studies on analyte ionizaition, using microwave attenuation methods, Sturgeon et al.(C609, C1150, 1378, C1459) determined the electron concentrations in both graphite and Ta atomizers. They obtained values of 2.5X 1011 to 7.8X 1012 cm-3 in the temperature range 2470~3000K in graphite atomizers and 1.1X1012 t o 8.7X1022 cm-3 in the range 2413-2743 K for Ta. Ionization, it was concluded, was negligible for elements of ionization potential greater than 4.2 eV in Ta atomizers and 4.5 eV in graphite atomizers.Ishibachi and Kibuchi (520) have studied di#usiorz e#ects for atoms in various sheath gases in furnace atornizcrs. Using Al, Coi and Cu as test clements, the effect of Ar, He, N, and mixtures of these with H, were studied. It was found that the gas flow-rate was directly related to the logarithm of the absorption signal and that, overall, Ar provided the best sensitivity.Falk and Thann (1229) have continued their investigations into the expansion60 Analytical Atomic Spectroscopy processes occurring in the vapour cloud above open atomizers and rcsolvcd that these processes could not be fully explained in terms of diffusion alone. L’vov and Polzic (1 326) calculated thc atom distribution in open atomizcrs with horizontally and vertically extended atom sources.A simple method was developed to predict the analytical signal, based on diffusion effects in two directions. The effect of hyperfine structure on Zecman splitting and line crossing, in afomic Faraday spectroscopy was described by Kitagawa et al. (2). A theoretical approach was developed that closely agreed with experimental observations. This described the dependence of transmitted intensity on the magnetic field strength for line crossings betwecn two Zccman components.Atomic Faraday spectroscopy was concluded t o bc comparable in sensitivity to conventional AAS. Results of investigations into calibration graph curvature in carbon furnace AES werc reported by Littlejohn and Ottaway (1121).Self absorption was found to be the most common cause and was shown to be independent of temperature. Self reversal however was only significant at concentrations greater than 1000 times the dctcction limit and then onl), for light elements. Curvature in calibration graphs for refractory elcmcnts was shown to b: due primarily t o atomization effects. 1.4.3 Interferences Without doubt, the most important contribution to the allcviation of interferences has bccn the trend toward attaining a more isothermal heating environment in ETA.In this context there have been many papers expounding the general utility of graphite platforms (C1475. C1477, C1168, C1169). Slavin and Manning (C113) found improvements in the dctcrmina- tions of Al, Cd, Mn and Tl in the presence of halide matrices.Sukhoveeva and co-workers (1 328) fabricated platforms from discs cut from grapkite electrodes and reported reductions of interferences for Ga and In in chloride media. Better results were however obtained when thc discs were pre-treated with solutions of 1% m / V ascorbic acid. Tantalum-coated platforms were fabricated in batches by Koirtyohann et al.(C103) and used in Ta-coated tubes, prepared according to the procedure of Zatka (AHAAS, 1978, 8, Ref. 950). Thew platforms were capable of accepting up to 50 pl of solution. In tests using low, medium and high volatility elements in synthetic and natural water matrices; improved performance. close to pure aqueous standards, was achieved. As in the previous rcfcrencc (1328), the use of a matrix modifier, in this case 0.5% NH,H,PO,, reduced all matrix effects to lcss than 10%.The conclusion to be drawn from studies such as these i s that the attainmcnt of 21 more constant thermal environment during atomization is necessary in order to minimize interference. Despite the fact that there appears to bc a “witches brew” approach to the use of matrix modification for the reduction of intcrferenccs in ETA, the effcctivcness of the technique is reflected in the number of applications in which it has appeared. Itsuki and Ibeda (529)’found the addition of La to be advantageous in the rcmoval of interferences occurring from the presence of various cations and organic acids during the atomization of Sn.Various combinations of matrix modifiers and atomization surfaces were invcstiga tcd by Sefflova and Komarek (1527) in order to find an optimum for the dctcrmination of Pb in waters and particulate matter.A combination of atomization from a Ta carbide co,itcd surface in the presence of 1% m / V La solution was found to give minimal intcrfcrcnces from phosphate, sulphate and chloride. Soluble organic acids have been previously reported to reduce interference effects in sea-watcr analysis, and Hydcs (991) confirmed that the use of 1% m/V ascorbic acid was advantageous in removing such interference effects on Ca, Cu and Mn.Goss ((37) reported that thc use of NaBH, removed interfcrencc effects in the determination of Pb in drinking water, and that this was a superior modifier to ascorbicA tomization md Excitation 61 acid or La.Matsusaki et al. (481) removed chloride interfcrences on Cr by the addition of an EDTA salt, and Cooksey and co-workers (C101) have advocated the addition of 0, to the purge gas during ashing, as an effective matrix modifier. Several important interference studies havc becn reported concerning the effect of the utomizer surface. Erspamer and Niemezyk (102) concluded that the physical and chemical properties of the atomizer surface play a key role in determining the degree to which thc matrix effects the atomization mechanism.Veillon et al. (1519) have uscd a radio-tracer technique with S*Cr to demonstrate how that element was irreversably retained by graphite furnace tubes and that thc degree of retention varied with matrix and tube condition. Penetration of analyte solution into the graphite tube during the drying stage was cited by Dabeka (C1163, C1404) as being detrimental to analytical accuracy.In this work, the author proposed the theory that previously reported matrix effects are in fact related to problems associated with the drying temperature. The authors however failed to mention the effect of drying rate.Miller and Sgontz (C559) havc noted that poor precision may result from fluctuation in the gas flow between the charring and atomization steps. They proposed the inclusion of n flow stabilization cycle to improve precision. Other references of interest - Double peaks : C98, C611, C1149, C1461. Halide interfcrences : 189, 1264, 1664. Isothermal sampling : 429, 902.Ionization : 521. Response times : 1275. 1.4.4 Advances in Methodology One of the main areas of interest this year has been the interfacing of chromatographic systems, in particular HPLC systems to conventional electrothermal atomizers. The problems of interfacing HPLC or an ion-exchange column to an electrothermal atomizer were discussed by Chau et al. (C1179) who demonstrated the applicability of such techniques to speciation studies of alkyl compounds of As, Pb, Se and Sn.Arsenical pesticide residues and some of their metabolites were similarly determined by Woolson and Ahronson (1 622) by coupling an HPLC to the automatic sampler accessory of an electrothermal atomizer. Such systems have been shown to be simple to automate, as was demonstrated by Brinckman (1618) who again determined various alkyl As compounds in soils and drinking water.Various HPLC techniques such as size exclusion and reverse bonded phase were used by the same group (984) to study Si and Sn macromolecules. Tetra-alkyl Pb compounds were separated by conventional GC by D e Jonghe et al. (486). Thesc authors described an alternative way of interfacing the GC to an electrothermal atomizer, whereby the eluate from the column was run directly into thc inncr gas purge stream of the graphite furnace, which was maintained at 2000 OC.Some advances in dealing with problems associated with the atomization of biolagical fluids have been evident. Frigicri and Trucco (1200) concluded that the only effective back- ground correction system for such applications was that based on the inverse Zeeman effcct.Similar conclusions werc reached by Magyar and Vonmont (1206) who found 0, source background correction inadequate for the determination of Hg in urinc. An alternativc to con- ventional AA analysis of biological fluids was proposed by Ottaway et al. (C213, C1725,1736). These workers reported improved detcction limits when graphite furnace AE was used in an instrument employing an echelle spectrometer with wavelength modulation Por background correction.An interesting indirect method for the determination of F- in urinc and blood62 Analytical Atomic Spectroscopy was described by Fuwa and co-workers (1318). The tube was pre-treated with A1 and thc molecular absorption of AIF was used as a measure of the F- concentration. Results were compared with ion-selective electrode techniques and agreement was found for F in urine but not in blood, where the authors concluded that some F was protein bound. Carbide-coated tubes are gaining in popularity. L'vov et al. determined platinum group metals (383) down to 0.1 ng and Y in steels (821) down to 0.001% using Ta-coated tubes. The necessity for such tubes when determining refractory metals was emphasized by Hanamura (C93), who determined Zr using a Zeeman AA system. Metal carbide tubes haw also been used to advantage by Harrison et al, (C569) who determined organometallics of groups ZVA, VA and VIA. The modified surface was claimed to decrease analyte carbide formation and reduce molecular volatilization. Other references of interest - Automated analysis : 87, C1410. Marine samples : 600, 601, 1659, 1705. Solid sampling : C8127, 883, 12415, C1462, C1561.

ISSN:0306-1353    

DOI:10.1039/AA9801000057

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

62 Analytical Atomic Spectroscopy 1.5 VAPOUR GENERATION 1.5.1 Hydfide Generation Thc vaporization of volatile hydrides continues to be a popular method for the determina- tion of a limited group of elements. Developments in 1980 included some novel atomizers, new information on interferences and the speciation of several elements, and some new detection systems. Them also appears to be increasing interest in thc determination of Pb using this method.The introduction of hydrides into the ICP and d.c. arc plasmas continucs to gcnzrate interest (see Section 1.2.1.3) but new flame methods arc also being introduced. A fuel-rich O,/H, flame burning in the inlet part of a T-shaped quartz tube has bcen used to atomize SeH, (1201). It was shown that atomization is brought about by free radicals gcncrated in the flame.A more traditional flame-heated silicon T-tube atomizer has also been dcscribcd (499, which was claimed to give improved sensitivities compared with earlicr versions (sec ARAAS, 1974, 4, Ref. 1128). Branched capitlary aspiration has bcen proposcd as a means of separately introducing the acidic sample and alkaline borohydridc solution (951).Thc hydride is then evolved inside the premixing chamber of the atomic absorption instrument. An analogous procedure was developed in which the two solutions were supplied directly to the mixing chamber using two pneumatic nebulizers (C84). A patcnt for a commercial clectrically heated atomic absorption atomization cell has also been reported (280). Interest continues to be shown in the developmcnt of rzort-dispersive AFS as a means of measuring As (174) and !3e (443, 846, 1281 and 16733 after generation of their volatile hydrides.An Ar/H,/entrained air flame i s usually preferred. Atomization in a hot inert gas stream has, however, also been suggested (1673) and the reported detection limit of 20 pg appears to be lower than that obtained in flames.Some authors have described the optimization of reduction conditions for the determination of tead (C73, C121, 186, 466, C1152 and 1353) and Sn (Cl21 and C1152) by hydride generation and AAS. Some authors confirmed the necessity for the addition of tartaric acid and potassium dichromate, as well as control of pH in Pb determinations (C121, 466). Others preferred to carry out a prior separation by co-precipitation of Pb with manganese dioxide t o overcome the interference of Cu and Ni in drinking water samples from a nickel mining and smelting area (C73, 186).Atomization m d Excitotion 63 A fiefd method in which a hydride generation cell and a portable Ar cylinder were used to release the hydrides for collection on HgCl, or AgNO, impregnated filters has been described (1226).Samples were desorbed in the laboratory and analysed using a slotted quartz-tube flame cell. The sequential determination of As, Ge, Se and Sn by hydride generation, separation by gas-solid chromatography and sequential AAS detection has also been proposed (1280). Since the light source and wavelength must be altered between each element the procedure appears complex but all four elements were determined in 3min per sample.There appears to be continued interest in automation of hydride generation and the associated analysis procedures, and automatic systems have been described for As and Se (C1479), Bi in rucks (912), Pb in drinking water (C73) and soluble As (1082). Considerable efforts are being made to detect different species of hydride forming elements, in some cases by modifying the chemical treatment and in others by incorporating separation methods. Thus Se(1V) and Se(V1) were differentiated (C1151 and 1711), and the separate determinations of Sb(II1) and Sb(V) (1079), Te(N) and Te(VI) (CllSl), as well as As(III), As(V), methylarsine, dimethylarsine, trimethylarsine and phenylarsine (l08Zx C1603) wero all described this year.New releasing agents for minimizing interferences have also been proposed : thiosemicarbazide and 1 ,lo-phenanthroline for Bi (91 2), and pyrogallol, thiosemicarbazide, oxine, cupferron, lanthanum nitrate and alizarin fluorine blue for AS ( 1 0 82). Two informative papers on mechanisms have been presented, the first dealing with adsorptive properties of the materials of construction of hydride generation cells (C120). With radioactive 75Se, significant adsorption of SeH, on Teflon, polypropylene and glass reaction vessels was demonstrated.Teflon vessels appeared to saturate with hydride and treatment of glass vessels with silicones significantly reduced adsorptive losses. Brown et al. (C&l) have reported further studies on the use of BH,CN- (safety notc : see Section 3.1.1.4) as the reducing agent in place of BH,-.The role of metals in the catalytic production of volatile oxides of nitrogen, which can cause interference at other points of the apparatus. such as the drying agent or cold trap, was also described. 1.5.2 Mercury Determination Once again, numerous papers have been published on cold-vapour methods for the deter- mination of Hg.There are however very few new ideas and1 most authors are merely con- cerned with the application of commercially available systems, the latter reflecting the wide- spread interest in Hg in environmental studies. In terms of techniques, additional information has become available on methods intro- duced in recent years, such as non-dispersive AFS (35, 899, 1084) and non-dispcrsive AAS (534, 1518).Atomic fluorescence defection limits of 0.01 pg were reported following collec- tion of Hg on gold wool (899) and 4 pg using collection on a gold amalgam and analysis in a “windowless” cell shielded with Ar (35). A simple non-dispersive AJ? spectrometer has been constructed that yielded a detection limit of 0.045 pg 1-1 (1084).Using nun-dispersive AAS, and the 184.9nm Hg resonance line, an improved detection limit of 0.01 ng has been reported (1518i). The apparatus was similar to that described in ARAAS, 1979, 9, 30. A fiefd method involving collection of Hg on a gold sampling tube has been described (1521). Subsequent analysis was carried out using a He d.c. plasma emission spectrometer system (253.7 nm line) and a detection limit of 0.5 ppb was reported for a 20 ml sample.The ICP measurement of Hg has also been reported (C7). An important paper described the development of a Zeeman atomic absorption system for the continuous measurement of Hg in a gas stream in the presence of high concentra- tions of organics, smoke, oil, and mist (426). The Hg emission line (T component) and a64 Analytical Atomic Spectroscopy reference line (a component) were generated by a single Hg discharge lamp operated in a 15 kG magnetic field.The instrument was used to measure Hg in oil shale gases (see also ARAAS, 1972, 2, 5). Several groups have reported developments in Hg speciation, particularly for distinguish- ing between organic and inorganic Hg (864, 935, 1610, 1710).Goalden et al. (1710) released inorganic Hg with EDTA and alkaline NH,OH, inorganic plus aryl Hg with EDTA and SnCl, and all forms of Hg with CdCl, and SnC1,. Mizunuma et al. (864) used NaBH, in NaOH for inorganic Hg and NaBH, in the presence of iron(III) solution at pH 2-3 for total Hg. Other references of interest - Construction of an automatic atomizer : 16201.Disproportionation of acidic solutions of Hg(I) under gas sparging : 917. Problems in using magnesium perchlorate as a desiccant : 170'1. 1.5.3 Methods Based on Molecular Absorption One very interesting application of the Zeemm effect that has been reported in the past year allows the determination of gas molecules to be carried out using atomic line sources (795, 1522). In these two papers, Koizumi et al. used an electromagnet to tune one of the Zeeman components of an atomic emission line to coincide exactly with a single rotational-vibrational line of a molecule such as NO. The molecular absorption was corrected for background absorption using the other Zeeman component, which is shifted away from the molecular absorption line. Thus down to 50 ppb of NO was detected using the 114Cd 214.438 nm line, the measurements being made at near room temperature. The Zn line at 21 3.856 nm was preferred for measurements at higher temperatures such as those existing in flames (1522). Molecules determined to date include NO, NO,, I,, S,, SO, and formaldehyde (795).

ISSN:0306-1353    

DOI:10.1039/AA9801000062

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

CHAPTER 2 lnstrumentatlon 2.1 LIGHT SOURCES Increasing interest is now being shown in the development of HCLs and other lamps using cathodic sputtering. The HCL has always been the most favoured light source for AAS, but workers seeking higher intensity sources, particularly for AFS, have usually concentrated on developing EDLs. An improved cathodic sputtering boosted discharge lump recently developed by Sullivan and Van Loon (see ARAAS, 1979, 9, Refs. 490, 1559) was, however, reported to improve the sensitivity and linear range for FAAS and to provide better detec- tion limits for FAFS. These demountable lamps are now being manufactured under licence (C126) and the cathode, in the form of a sod or powder pressed into a support medium, could easily be fabricated by the user, Assembly and disassembly are rapid and warm-up times are short.The authors obtained such high intensities from these lamps that they suggested it might be necessary to attenuate the radiation to avoid overexposing the PMT. These lamps could be a most important development, allowing better performance than has previously been obtained with detectors such as vidicons and photodiode arrays, which have generally poor sensitivity compared with PMTs; they could at last lead to increased popularity for AFS.Gough and Meldrum (944) used such a source for non-dispersive MS with a cathodic sputtering atom cell. Three separate power supplies were required for the lamp, one operated at approximately 30 mA and '700 V to sputter atoms from the cathode, a filament heater run at 5 A and 2 V to provide a source of electrons, and finally a supply operated at 600mA and 40V connected between the anode of the lamp and the filament to cxcite further the sputtered atoms.Detection limits were approximately ten times better than for absorption measurements. A demountable water-cooled boosted discharge hollow-cathode lamp has been described (32), where the boosted discharge passed along the direction of light emission and coaxially through the hollow cathode (see ARAAS, 19718, 8, Ref. 137). Emission intensities for Al, Cu, Mo, Ti and V were generally 10-20 times higher than with standard commercial lamps, while linc broadening was said to be minimal. T h i s coaxial lamp was, however, more complicated and expensive than are conventional lamps.Performance improvements were claimed for a pulsed Cu HCL run with 1Olps r.f. bursts superimposed on the pulses (920;). There was no enhancement of intensity over the pulsed current mode, but improved uniformity of line profile and intensity led to a better SNR for AAS. The use of high frequency HCLs (910) for determining Cd, Pb and Zn yielded FAAS sensitivities 1.5 times superior to those obtained with conventional lamps.Detection limits for non-dispersive ETA-AFS were 1 X 10-13 g (Cd), 3X 10-11 g (Pb) and 2X 10-13 g (Zn). A study of line profiles from an As HCL with Ne fill gas showed some spectral interferences on the 193.7nm line by two ion lines (1244). This would produce pronounced curvature in AA calibration graphs. Measurements of the Ca 422.7nm line profile in a HCL demonstrated the linc narrowing associated with selective modulation ( I 241). The fundamental characteristics of HCL discharges were studied using a plasma model based on the Corona discharge (C125).This was said to be a better model for predicting the electron density and the temperatures of the atomic species and electrons than the more traditional thermal equilibrium m o d 4 which considers only collisional processes and neglects radiative energy losses.Additional references on the preceding topic - 623, 81 5, 847, C1406, C15.56, 1578, 1581.66 Analyticd Atomic Spectroscopy The controiied temperature-gradient l m p (see ARAAS, 1979, 9, Refs. 799, 1678) is especially suited to volatile elements that make less successful HCLs.A commercial version (C127’) can bc mounted in conventional spectromcters and provides higher intensity than EDLs with narrower line widths. The lamps have been produced for a number of elements including As, Cd, P, Pb and Se. An interesting new source, which appears to have potential for analytical spectroscopy, has been described. The heat-pipe arc lamp (1596) consisted of a disc-shaped oven, contain- ing Ar at 1 Torr, made from two circular stainless-steel plates with a window in the edge of the disc.This allowed a large solid angle of a Li discharge to be viewed. The application of a d.c. voltage across the plates maintained a glow discharge at currents of 10-5 to A, and an arc discharge at currents greater than 0.1 A (an unstable changing discharge occurred at intermediate currents).The Li 61 0.4 nm doublet (0.0 13 nm separation) could be resolved using a Fabry-Perot interferometer, indicating that minimal linc broadening occurred in the lamp. The use of an inductively coupled plasma as an excitation source for AFS was reported previously (see ARAAS, 19179, 9, Ref. 1851). Its use to determine Cd and Zn in fly-ash using an air/H, flame has been described (1 186).The applications of lasers are reported in Scctions 1.1.2 and 1.3.4. Other references of interest - Mercury emission in a low-pressure ring discharge : C 1564. Microwave power measurements for EDLs: 1225. R.f. generator for EDLs : 810. Rejuvenating commercial EDLs : 1289. 2.2 OPTICS 2.2.1 Background Correction Zeeman-eflect AAS can provide a more effective method of background correction than the use of a D, arc lamp with conventional AAS.This was shown in a detailed study by Fernandez et crl. (940) who examined elements in several matrices including Pb in 1% NaCl and Se in Ca,(POI,),. A Zeeman-effect AAS instrument, able to correct background absorp- tion up to 2 absorbance units, has been described (1231). An extensive critical review (274) of background and scatter correction in AAS and AFS included consideration of deuterium source background correction, Zeeman-effect AAS and wavelength modulation using a continuum source.Additional references on the preceding topic - C618, 1107. 2.2.2. Optical Systems A new generation of ICP spectrometers using slew-scan programmable rnonochroniators was prcviously reported (see ARAAS, 1979, 9, 40).A double monochromator instrument, after slewing to the analytical line acquires data over a 0.3 nm range centred around the linc (C737, (21437); this allows %point background correction at each wavelength. Monitoring an Ar line before each run compensates for wavelength inaccuracy. A programmable monochromator was used with a SIT vidicon detector for FAES (27’6) and ICP-OES (844).This detector had the advantage that less accurate wavelength positioning w a s required compared with the USC of slit/PMT systems (see also Section 2.3). Additional reference on the preceding topic - 3%.Instrumentation 67 Ottaway et al. (1233) combined an echelle rnoizochrornator with an electrothermal atomizer for carbon furnace atomic emission.High resolution coupled with background correction using square-wave Wavelength modulation contributed to improved detection limits for 14 out of the 15 elements studied. An echelle monochromator has now been made to scan (C564) by the addition of microcomputer controlled stepper motors to both prism and grating drives. The very high resolution of the echelle permitted the spectral range 200-600 nm to be scanned and recorded in wavelength increments of only 0.0001 nm.It was reported that slew-scanning was also possible with this system. The performance characteristics of imaging detectors (silicon target vidicon and ID) with an echelle mono- chromator, for simultaneous multi-element AAS and AES, have been evaluated (459).New advances in holographic gratings for use in direct reading spectrometers have been described (C 1427). The astigmatism in a conventional Rowland circle configuration was substantially reduced to permit high quality imaging. Also, a new design of flat field grating permitted a change in spectral range over the flat field by simply rotating the grating; no further optics were required.A modified polychromator had 3 toroidal mirrors mounted 011 a turret so that each could be positioned in front of the grating (34). The system, when operated at grazing incidence, was stigmatic in one dimension in the V.U.V. An automatic drift control on a direct-reading polychromator was activated by an auto-sampler (C75). As the sample probe moved from the wash to the sample solution, a switch was activated to drop a Hg lamp in front of the entrance slit for 151 s.Signals from 2 detected Hg lines were fed to a null-balance motor which turned a correction plate to compensate for any drift. The use of 2 poly- chromators in a single system doubles the available space for the exit channels. It also allows the effective placement of channels on alternate polychromators when the very close proximity of 2 spectral lines would preclude their use on the same focal curve.A number of these systems have been described (C44, C1389), including one used for end-on viewing of an ICP, which was claimed to allow more than 200 PMTs to be mounted (C735), and one permitting measurement on up to 256 wavelengths (C672). In neither case, however, was the maximum number of channels used.A 3-channel optical system for simulronaous AESIAAS has been described by Ure et al. (1634) (see also Section 2.5.2). New typcs of U.V. optical filters have been described (ClSl), and the authors speculated that high resolution band-pass filters in the future could have band-widths as narrow as 0.001 nm. Other rcfcrcnces of iriterebt - An explanation of the photoelectric effect : 1246.New Eagle-type monochromator mounting : 621. Quartz-coated optical components in spectrometers : 33. Wavelength scanning and mirror alignmcnt of a multipass Fabry-Perot interfero- meter : 158'6. 2.3 DETECTOK SYSTEMS Further reports on the usc of rupid-scartizing photoelectric detectors for both AES. and AAS have been published, and in most cases detection limits were still inferior to those obtained using PMTs. Codding et al.(1717) have, however, constructed a photodiode array spectro- meter for FAAS and obtained better detection limits than previously reported for this type of system. Typical values were 0.001, 0.004 and 0.069 ppm for Mg (285.2 rim), Zn (21 3.9 nm) and Cu (324.7 nm), respectively. They speculated that present technology should permit a spectrometer to bc designed that was not detector-noise limited.when, presumably.68 Analytical A t omic Spectroscopy detection limits could be improved further. Boumans previously evaluated linear diode arrays with separate outputs (i.e., not self-scanning) for detection with ICP-OES (see ARAAS, 1979, 9, Ref. 863). Further studies (C734) have shown this system also to be detector-noise limited, although at wavelengths above 500 nm SNR performance approached that using PMTs.This improved SNR of SSIDs at high wavelengths was exploited by Fry et al. (C1040) who found an ICP diode array spectrometer to be suitable for the near i.r. simultaneous multi-element determination of non-metals such as Br, C, Cl, F, H, N and 0 in organic compounds.A system in which analyte flicker was the dominant noise sourcc has been described (1316). An ICP was used with a 1024-element photodiode array spectro- meter and the authors suggested that the sample delivery and gas-flow systems of the ICP were responsible for the high flicker noise. Additional references on the preceding topic - C70, C146, 517, C545, C588, C616, C 1 472. A silicon intensified target vidicon detector has been used with a programmabk monochromator for multi-element FAES (276).The determination of 22 elements in a N20/C2H2 flame gave detection limits that were poorer in the u.v., although similar in the visible region, to PMT detection. Thc same authors used the detector for ICP-OES with a 1 m monochromator (844).In all cases detection limits were poor compared with a direct readcr. Also, the linear range was smaller with the SIT system. Vidicon mcasurcments reported in the literature have almost always been for steady-state signals, target lag being the biggest problem if transient signals are measured. A SIT vidicon has been evaluated for use with transient non-repetitive radiation sources and besides lag, additional adverse factors were reported (C1426).A nearly 5-fold increase in background noise occurred with single-scan operation, while electron beam defocusing was experienced with time-gated operation. It was found that non-linearity due to lag could be reduced by making scveral target scans, but at the expense of background noise. A silicon target vidicon and a n image dissector have been used with an echellz spectrometer for simultaneous multi-element AAS and AES determinations (459).A KBr-coated photocell permitted non-dispersive cold-vapour AAS of Hg at 185.0 nm with high sensitivity (534). The high quantum yield of this detcctor in the v.u.v. led to the achievement of a characteristic concentration of 0.26 ng and a detection limit of 0.043 ng using a 3001 mm path-length absorption cell.Other reference of interest - Use of a GDL with a resonance detector : 1208. 2.4 DATA PROCESSING 2.4.1 Emission Reports describing instruments that use slew-scanning programmable monochromators for ICP-OES are increasing (see Section 2.5.1). Their versatility for rapid determinations at different wavelengths under optimized conditions depends heavily on the sophistication of the computer control.Consequently, as thc instruments develop the major emphasis will probably bc on softwarc improvements. A graphics utility softwarc package has been developed for use with the Perkin-Elmer ICP 5000 (C1383). Disk storage of the complete spectra enables their recall and comparison with spectra of pure standards.This allows spectral characteristics oE the sample matrix to be examined when spectral interferenccs will be readily apparent. The Instrumentation Laboratory Plasma 100 scans a spectral range of over '700 nm in only 13 s, and an important function of the microcomputer control i s toZnstrumen tation 69 allow accurate and rapid wavelength selection. A search routine (C741) finds the wavc- length of maximum emission after the monochromator has been slcwcd t o a position closc to the line.Other features of this instrument have been described (C830), including back- ground correction and routines for correcting interferences. Thc software system developed t o operate the Applied Research Laboratories rapid-scanning sequential instruincnt also includes a routine for pcak-seeking and hence accurate wavclcngth sclcction (C 132).The programs for instrument control, data acquisition and data calculation arc written in a high level language, ARL Extended BASIC. Floyd er al. ( 1 5 1 9 , in dscribing a computer controlled scanning monochromator system, claimed to have dcvclopcd a supcrior pcak- seeking routine that is used in the Applied Research Laboratorizs instrumcnt.After approximate peak location, the instrument step-scans over a range of 0.3 nni in 0.01 nm increments. The exact position i s then located by fitting a Gaussian-shaped profile to thc intensity measurements so obtained. Other software features includc storage in mcmory of the analytical wavelengths, pre-sclcction of analytical and background wavelcngths, and a routine for calculating detection limits in order t o monitor instrument performance.A method of data proccssing that was claimcd to improve precision by 2040% (1514) involved scanning the spectral line in 90 increments over a wavelength range of 0.1 8 nm. After accumulating 416 scans any irregularities in the pcak profile were smoothed out by fitting a Gaussian curve t o the data points by the least squares method.Peak intcnsity was then calculated from thc equation of the curve. Additional references on thc preceding topic - C66, C131. A minicomputer linked t o a 12-channel direct reading emission spcctromctcr (C75) besides performing the usual control, signal processing and data proccssing functions, also controlled a sample changer.A procedure for thc automatic compensation of spcctromctcr drift, which was performed between sample changes, was also included (scc Section 2.2.2). Rapid proccssing of spcctrographic rcsul ts by means of a compu tcrizcd rnicrophoro- meter has been described (81 6). A partitioned mcmory permitted the simultaneous acquisi- tion of transmittance from the microphotometer.Results for 64 elcments were calculated from storcd calibration data and outputted only 5min after developing the plate. Additional references on the preceding topic - C1544, C1554, (21577. Horlick and Hall (C616, C48) dcscribed a method for processing Fourier transform interferograms. A time and software saving alternative t o Fourier transformation involved generation of a mask interferogram containing the species of interest, followcd by cross- correla t i on.Other references of interest - Abel inversions on ICP spatial emission profiles : C70, C588. Review of data acquisition and control proccdurcs for TCP-OES : (259.1. Software correction methods for TCP-OES : C31. 2.4.2 Absorption The speed and efficiency of routine analysis by AAS can be greatly enhanced by external conzpicter control of the instrument, and an increasing number of applications are appearing in the literature. When a Perkin-Elmer model 5000 AA spcctromcter was interfaced t o a 32K computer, better baseline and slope correction features than those built into the instrument were achieved (C128).The major time-saving feature of the computer was its ability to produce results corrected for the blank and dilution factors.The system enablcd 1000 results/day t o be obtained, which was a 3-fold increase over conventional manual AAS. Enrione and Stransky (C130) also linked the same spectrometer t o a cmputcr, which prc- pared sample lists including standards and quality control sarnplcs. I l i e microproccssor in the AA spectrometer computed concentrations and then transmitted tlicm to the computer70 Aizwlytical A tomic Spectroscopy where they were stored with the sample numbers.At the end of the day, all data wcrc transmitted to a central computer for further processing along with data from othcr labora- tory instruments. Two Pye-Unicam SP9 AA spectrometers with automatic samplers werc coilpled to a 48K microcomputer and operated simultaneously (C160).Data wcrc storcd and filed on disks for later statistical manipulation. The system was applied to the detcrrnina- tions of several trace metals in plants and it was hoped in this way to carry out 300Cd individual metal determinations in one growing season. Other instniments that h a w been interfacxl with a computer were an Tnstrumentatjon Laboratory model 351 with ETA (C29) 2nd a Varian AA6 (C836).Routh (C1411) suggested that the coupled computer may also prow wcful in interactive method development by editing instrumental parameters. cxamiriing raw data and evaluating statistical results. A’dditional referenccs on the preceding topic - 377, 792. 0’Haver and co-workers described some recent advances in computer software for coiztiizuurrz source rclzelle IvrrveZerzgfh-I.tzoduCated AAS (C1408, C14233.T h e minicomputer generatcd the wavelength modulation waveform, measured and stored intensity data o n up to 16 channels, and performed various signal and data processing functions. A particular zrdvantagc for multi-elemcnt continuum source AAS was the processing of absorption profile shapc data to give a greatly improved dynamic range.Thc importance of rupfd elcctrorzic rasportse in ETA-AAS has been stressed (1275). It was reported that the more subtle effects of limiting low-pass time constants in the sjgnal processing clectronics wcre to afiect pcak areas as well as pcak heights. This could lead to large errors whcn matrix effects unpredictably change the shape of transient signals.Whiteside et aE. (1255) reviewed the state-of-the-art of signal and data procming tcch- niqucs. Tncluded were methods of signal averaging, the importance of precision calculations, correct calibration procedures and curve correction methods. Other references of interest -- Analogue-digital convertor for double-beam instrument : 1 343. Compensation of additive interferences : 1576. Correction procedure for nebulization effects : 19m2.Statistical evaluation of interfercnces : 1628. Two-dimensional simplex optimization : 1672. 225 COMPLETE INSTRUMENTS 2.5.1 Emission Instruments Interest in sequeiitial-scanrziiig inductively coupled plasma spectrometerJ cquippcd with programmable monochromators i s continuing to grow. Floyd et al.(1 515) dcscribcd a system that they claimed was more sophisticated than any reported previously. Thc 0.5m doublc monochromator instrument was equipped with computer-controlled wavelength scanning at speeds of 500 nm min-1 and 10 nm min-1, and had sophisticated software routines (sce Scction 2.4.1). They presented results from determinations of up to 26 elcmcnts in CRMs (17181).An instrument, operating in BASIC, with a programmable double mono- chromator had automatic background correction and data were displayed on a video scrccn aftcr being subjccted to a 60 Hz noise-rejection routine (C737, C1437). Additional references on the preceding topic - C10, C60, C64, C596, (2737, C1398, C1437, 1730. The more sophisticated dirocl-reuding spectrometers may be cquippcd with two poly- chromators, thus allowing more space for the placement of exit slits/PMTs.kccl-rtc ei rrl. (C44) described an TCP system that enabled 61 clements to bc determined simultaneously i nInstrumentation 71 geological samples. A refractor plate was used for background correction. A dual-grating instrument with the first grating used for U.V. wavelengths and the second for higher wave- lengths up to the i.r.region was claimed to be capable of providing u13 to 256 wavelength channels with a resolution of 0.02nm (C672). Performance with an TCP was said to bc similar to that of other direct readers. Additional references on the preceding topic - C 1 375, C 1 389, 154 1. A d.c. plasma vacuum ultraviolet emission spectrometer (0.3m focal length) had a useful working wavelength range down to 160 nm (C736). Problems of light absorption were overcome by using a Suprasil I1 focusing lens, coating the mirrors with MgF, and purging the system with Ar.A d.c. plasma echefte spectrometer has been available commercially for several years, and two recent papers have described its modification to incorporate an ICP sourcc.McLaren et al. ((21367, C1402) interfaced a minicomputer to the system and time-niultiplsxcd the currents from 20 FMTs to a single amplifier. Gram et al. (CS3) used fibrc-optic inter- facing of 32 channels from an echelle spectrometer to a high-speed readout containing separate PMTs and an analogue to digital convertor. This was linked to a minicomputer that controlled each PMT voltage, the source position, and an electronic shutter and refractor plate for wavelength position control and background correction.Echelle mono- chromators have a lower light throughput than single prism or grating instruments and the use of fibre-optics would inevitably lead to further light losses. Unfortunatcly, sensitivity data were not given for this system. Ottaway et 01. (123131, C1275) have built an ETA echelle spectrometer for carbon furnace atomic emission spectrometry.Improved detection limits compared with previous work were attributed partly to the use of a graphite platform for isothermal atomization and partly to the high-resolution optical system with background correction using square-wave wavelength modulation. Additional reference on the preceding topic - 1737.Other reference of interest - Spectrometer with visual detection far multi-element steel analysis: 1 345. 2.5.2 Absorption Instruments The multi-element continuum source echelle wavelength-modulated A A spectrometer developed by O’Haver and ccv-workers (see ARAAS, 1979, 9, 41) has been further refined (C74, C88). Used with an ETA and auto-sampler, the start of the atomize stage triggered data acquisition to measure both peak height and area.Up to 1001 measurements could be made in this way without operator interference. In the FAAS mode the instrument could operate with standard nebulization or by using non-segmented flow injection. The latter permitted discrete microvolumes of liquid sample to be inserted into a continuously flowing solvent stream, which was pumped directly to the nebulizer.Data acquisition was triggered by the valve shaft, which shunted the sample loop into the mainflow to the nebulizer. U p to 16 elements per sample were determined at the rate of 60 samples per hour. Additional references on the preceding topic - C604, C6 18. A dedicated multi-element % m e spectrometer constructed by Ure et al.(1634) had 3 monachromators arranged so that 3 elements could bc determined simultaneously in the same flame. For soil extracts and plant-ash solutions the instrument determined Ca and Mg by FAAS and Na by FAES using a N,O/air/C,H, flame. A unique twin-nebulizer system was arranged so that K (to suppress ionization of Na) and Sr or La (to suppress interference by PO,3- on Ca) were both introduced concurrently with the analyte.A chopper arrange- ment allowed calculation of the absorbance with flame background emission cancelled out.72 Analytical Atomic Spectroscopy An RSD of 0.025-0.031 and good accuracy were demonstrated. Sequential multi-elcment analysis from 6 flowing sample streams has been achieved with an automated ETA-AA spectrometer (C87).Each stream had a by-pass flowing continuously through a sampling cup, and discrete samples were pipetted from the cups. Additional references on the preceding topic - C94, 251, 458. Dcscriptions of a number of Zeeman-eflect AA spectrometers have been pitblishcd (940, 1204, 1231). A new Varian AA spectrometer was claimed to provide significant performance and data handling irnprovcments (C86, C560, C834).These include a pneumatic ncbulizer with modiiicd spray chamber design, real time data acquisition and display of transients, and various calibration routines ixluding 5-standard calibration and standard additions. Com- putcd data arc fcd into a high-speed printer. A simple high-sensitivity non-dispersive A A spectrometer for the cold-vapour dctxmina- tion of Hg (534) used a KBr-coated photocell detector. Its high quantum yield enabled thc 185.0 nm line to bc used.Several ncw Russian instruments have been described (435, 4691, 976, 1575). Other references of interest - Backgr0ur.d and dsvelopment of the Pye-Unicam SP9 : 1729. History of Perkin-Elmer instruments : 1733. Review of instruments on the Norwegian market : 1727.Simultaneous multi-clement flame spectrometer : 945. 2.5.3 Fluorescence Instruments The new high-intensity light sources developed by Sullivan and Van Loon (scc Section 2.1) have been uscd in a non-dispersive AFS spectrometer for the direct determination of metals by cathodic sputtering (944). A pseudo-double-beam system enabled the light source intensity to be monitored. Jn order to eliminate most of the emission noise the sputtering cell was modulated and this gave detection limits an order of magnitude better than had previously been obtained by this technique.A non-dispersive dedicated instrument for cold-vapour Mg determinations has been described (1084). A detection limit of 0.045 pg I-' was claimed without the need for a solar blind PMT. Other references of interest - Cold-vapour Hg spectrometer : 899.On-line instrument for determining Mg in boiler feed water : 5%. Spectrometer for determining Pb in stack gases : 81 3. 2.6 NEW COMMERCIAL INSTRUMENTS 2.6.1 Emission Spectrometers Jarrell-Ash have added three instruments to their range : the model 82-487 with photo- graphic detection; the 82-484, a scanning monochromator instrument; and the 82-000 scanning instrument, which can be interchanged to photographic detection.Ncw from bird-Atomic is the Spectromobilc MS2, a 24-channel direct rcadcr. The Spectramctrics d.c. plasma echelle spectrometers have bcen revised. The Spectraspan IIIB has interchange- able casscttes to provide either single- or multi-element capabilities. Additional micro- computer control i s available in the form of a data station that is easily interfaced to this and any existing Spectraspan instrument.A refractor plate background correction system is also available. The lower cost Spectrospan IV is a sequential instrument.Instrumentation 73 2.6.2 Absorption Spectrometers The sophistication and versatility of the Ferkin-Elmer model 5000 AA spectrometer has been further extended with the introduction of the Zeeman 5000.me Zeeman module incorporates a modified furnace with an applied magnetic field of 9 kGauss during the atomization cycle. Instrumentation Laboratory have introduced the IL357. This is the single-beam 157 with additional software and the RS232C data output used in their more sophisticated instruments. A single-beam video instrument, the JL45 1, has been introduced, and the IL457 has replaced thc IL257.A more sophisticated version of the Pye-Unicam SP9 computer i s their PU9090 Data Graphics system. Varian have renumbered their 275 and 475 spectrometers; they are now the 1275 and 1475, respectively. The modifications are the incorporation of a new nebulizer/spray chamber assembly and some changes in the interface to external data handling systems (see also Section 2.5.2.).Table 2.6A -COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS Supplier Reciprocal Focal nm oer mm m Model dispersion/ ~ ~ ~ ~ $ ~ ~ h length/ Type of Source Speclal features Applied Research Quantometer Laboratories Ltd., 3 4 m c Wingate Road, Luton, Beds., England Applied Research Quantometer Laboratories Ltd., 34ooOD En Vallaire CH-1024, Ecu b lens/Lausan ne, Switzerland Applied Research Quantometer Laboratories Ltd., B34000C 9545 Wentworth Street, P.O.Box 129, California. U.S.A. Societe Franqaise Quantotest d’lnstruments 360(#3 Controlee d ‘Ana lyses, B.P. No. 3, F78320, Le Mesnil, St. Denis, France Low voltage spark High voltage spark Full computer control t o provide direct concentration print out; full rapge of options including dual floppy discs, VDUs, fast printers, remote terminals and computer links em.; twin Stand facility including Ar, air, hollow cathode, rotrode, plasma, glow discharge etc.; 48 channels 0.465 170-407 0.520 193456 0.347 or 0.694 190-610 ; :: j- or d.c.arc 0.930 or 0.465 190-820 or 0.310 As 34OOOC As34000C As 34OOOC As 34UOOC As34000C As 34OOOC 0.70 200-400 Low voltage spark and/or high voltage spark and/or d.c.arc Full computer control to provide direct concentration print out; optional local and remote printers; Ar or air excitation stands: 48 channels As 3400OC As 34OOOC; twin stand facility including Ar, air, hollow cathode, cotrode, glow discharge, efc.; allows for expansion to include a large number of elements and offers comprehensive computer options to handle multiple and complex alloy programmes: 60 channels 0.3 Low voltage Small transportable Quantometer with Go No-Go inspection type electronics: 10 channels Baird Corporation, Spectromet 0.6 or 0.3 210-590 1.0 Arc or spark; Compact, low-cost direct reader with minimum air 125 Middlesex loo0 mod u I ar conditioning requirements: manual master monitor to Turnpike, Bedford, MA 01730, U.S.A.Spectrovac 0.6 or 0.3 173-767 1 - 0 Arc or spark; As Spectromet. 1000: logarithmic read-out; check slit alignment; 30 channels 1000 modular dual stand option; 30 channels Baird-Atomic Ltd. Warner Drive, Spr i ngwood Industrial Estate, , Spectromet I1 0-294 19-32 2.0 As Spectromet 1000 Automatic optical servo monitor continuously 0.59 190-863 maintains slit alingment; logarithmic read out; manual master monitor to check slit alignment; dual stand option; 60 channels 60 channels Braintree, Essex CM7 7YL, England Spectrovac I I 0.29 173432 2.0 As Spectromet 1000 As Spectromet I I ; all photomultiplers in vacuum; Spectromobile 0.3 or 0.6 200-600 1.0 Low voltage Mobile spectrometer for metal sorting or checking: MS2 * bstep d.c.arc 24 channels b * New equipment since publication of Volume 9Jarrel-Ash Division, 78.090 1 -1 or 0.54 426970 Fisher Scientific Co., 21 12485 590 Lincoln Street, 70-310 1 - 0 O[ 0.2 180-3000 Waltham, MA 02254, depending on 180-1500 U.S.A. grating 180-750 96-750 0.54 168-500 96-785 0.54 168-500 1500 0.56 or 0.28 200-&(30 or 0.34 or 0.17 2Oo-510 or 190-400 190-250 1 a 5 3.4 Various available in Wadsworth Spectrograph; 20 in camera: choice of varisource unit three gratings; N2 purging extends range to 175 nm; Includes spark, low optional accessories permit use as direct reader or and high voltage d.c.scanning spectrophotometer arcs. Also versatile controlled wave excitation source 0.75 As above except Direct reader; computer controlled electronically 0.75 controlled peak current 1.5 As above Choice of two gratings Labtest Equipment Co. 310 0.56 190-900 1.5 'Transource' high Wavelength in first order; CRT; teletype printer or Ltd., computer read out systems: dual air/inert gas and 11828 La Grange V25 0.46 170-450 1.0 discharge. Low solution excitation stand; V25 vacuum for C and S in Avenue, voltage triggered ferrous materials Los Angeles, CA 90025, 2100 0.46 188-455 1.0 d.c. arc; ICP source U.S.A.for solution analysis voltage triggered 71 0.52 170-900 2.0 M.B.L.E., Rue des Deux Gares 81070, Brussels, Belgium PV8350 0.46 177-410 1 . O Monoalternance arc, glow discharge, hollow cathode, or ICP 80, Vacuum 500 Hz spark, d.c. - - Philips Analytical Department, PV8250 0.69 190-61 5 1 * 0 As for PV8350 Pve-Unicam Ltd., Air 0.59 190-531 York Street, Cambridae CB1 2PX.England - PV8210 0.55 190-7ff5 1.5 AS for PV8350 Air Integrated spectrometer system with optional dual air/Ar excitation stand; choice of programmable calculator and computer configurations with dual cassettes or floppy discs; rapid printer; VDU extension options As for PV8350 0.92 190-820 0.46 190-410 As for PV8350 Rank-H i lger Ltd., Westwood, Margate, Kent CT9 4JL, England E10W 0-293-1.155 156-880 1.5 Various, including Solid state electronics; microprocessor control high repetition condensed arc, ICP, G DL available.Dual gratings give 12 standard systems to select optimum dispersion and wavelength coverage. Special grating if required; dual spark stands P olyvac E960 0.546 or 0.741 174.0-447.7 0-75 As ElOOO Curved entrance and exit slits; microprocessor control available; air or inert gas discharge stands -1Table 2.6A - COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS- continued Reciprocal Focal nm per mm rn Supplier Model dispersion/ ~~~~~~~h fength/ Type of Source Special features Spectrametrics Inc., AEZ 0.06 1 9 w m 0.75 Plasmajet Optimized AE system using a high dispersion, 204 Andover Street, high-energy, throughput echelle spectrophotometer Andover, MA 01810, DRlO 0.06 190-800 0.75 Plasmajet and a high temperature plasmajet excitation source.U.S.A. Techmation Ltd., ES9 0.06 190-800 0.75 Plasmajet, flame Built in computer 5% Edgware Way, Edgware, Middlesex HA8 8JP, England FtS1 0.06 190-800 0.75 AsES9 One channel, variable wavelength or arc stand VEB Carl Zeiss Jena, 69 Jena, Carl-Zeiss Str. 1 German Democratic Republic Carl Zeiss Scientific Instruments Ltd., P.O. Box 43, 2 Elstree Way, Boreham Wood, Herts WD6 lNH, England PGSP 0.74 or 0.37 ZCH3-280 2.075 Arc or spark Atl3s for spectra evalution; wide choice of precision diffraction gratings; high resolving power; dispersion doubling or multiplying as required; automatic transport of cassette, wavelength scale for quick orientation of the user within the spectra; wide range of accessories available including LMA 10 laser-microspectral analyser b bTable 2.6B- COMMERCIALLY AVAILABLE PLASMA SPECTROMETERS Supplier ~~ Generator Reciprocal Focal Model dispersion/ length/ Output Operating nm per mm m power/ frequency kW MHz Special features Applied Research Quantometer 0.930 or 0.465 1 .O 2 27-12 Czerny-Turner Spectrometer.Full computer control to provide direct Laboratories Ltd.* 340OO/ICP or 0.310 concentration print-out; full range of options including dual floppy discs, VDU, fast printers, remote terminals and computer links, etc. Concentric flow nebulizer. Automated scanning of grating for wavelength range studies, profile facilities and electronics providing qualitative analysis on chosen spectrum lines. Full computer control Quantometer 0-80 1.0 2 27.12 35000C 0.60 0-50 0.40 As for 34000 Baird Corporation * Plasma 0-66 1.0 2 - 5 27.12 One metre polychromator with 120 exit slits in a rigid focal curve.Data acquisition is controlled by a Tektronix 4052 graphic computing Spectromet 0.33 system.Optional integral scanning monochromator 0.22 Instrumentation Plasma 100 - Laboratory 1nc.t 2.5 27.12 Microcomputer controlled scanning double monochromator for sequential multi-element analysis; instructions for programming appear on video display with single keystroke operation; emission profiles of analytical line appear on video screen for selection of wavelength for background correction, for study of inter-element effects and observation of spectral interferences; all circuitry for r.f.power generation, monochromator optics, and microcomputer are built into the instrument. Ebert (double) monochromator, crossflow neb u I izer Jarrell-Ash Division, 96.975 0.54 0.75 2.0 27.12 Computer control; variable channel; concentration print-out Co.Ltd.* 96.988 0.54 0.75 - - Computer control; N +1 channel scanning attachment, spectrum Fisher Scientific shifter attachment for automatic background correction; special K and L i channels; data management system Jobin-Yvon, J Y 37P - 0 - 6 Plasmatherm Czwny-Turner monochromator, large aperature monochromator (f 5.4 grating size 120x 140 mm); manual or computer controlled; constant - 1.0 1-5 27-12 time integration or in ratio mode; choice of concentric glass Division nebulizers, adjustable concentric nebulizers in zirconium, or d'lnstruments, J Y 38P 91160 Longjumeau, 5 - ultrasonic nebulizer 16-18 Rue du Canal, France - 2.5 * Address as in Table 2.6A i- Address as in Table 2.6CTable 2.6B -COMMERCIALLY AVAILABLE PLASMA SPECTROMETERS- corzrirrued Generator Reciprocal Focal Supplier Model dispersion/ length/ Output Operating nm per mm m power/ frequency kW M Hz Special features EDT Research, JY48P 14 Trading Estate Road.London NWlO 7JU, England Kontron GmbH, ICP 1000/2o0O Postfach 8057, Eching bei Munchen, ASS 80 Oskar-von-M ul ler 0.39 1.0 Durr-Jobin-Yvon Air or vacuum spectrometer; 86 positions of photomultipliers, fully 0.46 2.2 56 automatic read-out computer option; choice of two standard gratings 4 - (or specials i f necessary) 1800 grooves/mm (wavelength range - - 180-590, reciprocal dispersion 0.55 nm/mm), or 2500 grooves/mm 0.55 0.82 (wavelength range 1+15, reciprocal dispersion 0.39 nm/mm), scanning entrance slit under computer control for identification of interfering spectral lines, background correction, and for analysis of elements not installed in the programme, Paschen-Runge monochromator; choice of nebulizers as for JY38P measurement of grating position by angle encoder; software for ICP concentric glass nebulizer 0.6 0.6 1.5 to 27- 12 Czerny-Turner monochroma!or; computer control; direct 7 - analysis and graphic d!splay on video screen; printer/plotter; Str. 1, West Germany ICP 1000/2000 0.5 0.75 1.5 to 27.12 Paschen-Runge monochromator; computer control; video display of results; printer plotter; concentric glass nebulizer - ES 750 7 Labtest Equipment Plasmascan - 0.35 2 27-12 Czerny-Turner monochromator; microprocessor control; enclosed co. * 700 sample pumping system; computer readout system; crossflow, MBLE * Philips 0.55 or 0.28 1.5 2 50 Paschen-Runge spectrometer; wavelength range covered i n 1st order; concentric glass or ultrasonic nebulizer remote controlled roving detector; readout by printer, teletype or digital computer systems, crossflow nebulizer options, spectrometer and nebulizer as for PV8210 PV8250 Air Philips 0.69 or 0.35 1.0 2 50 Integrated spectrometer system with built-in source and readout PV8250 Air 0.59 or 0.35 0.92 or 0.46 0.46 or 0.23 -.Philips 0.46 1.0 2 50 Integrated spectrometer system including source and readout G' PV8350 op!ions, spectrometer and nebulizer as for PV8210 % vacuum $ 3 - Rank-Hilger Lid.* ElOOO 0.293-1.155 1 - 5 - 27.12 Paschen-Runge spectrometer dual gratings give 12 standard systems 3 h polyvac to select optimum dispersion and wavelength coverage; special ?' grating if required; dual spark stands; solid-state electronics; microprocessor control available; crossflow or concentric glass nebulizer 27.12 Paschen-Runge spectrometer, curved entrance and exit slits; E960 0.5460.741 0.75 - Y microprocessor control available; crossflow or concentric glass nebulizer !? - 8 U * Address as in Table 2.6A t Address as in Table 2.6C ?Perkin-Elmer ICP 5000 U.V. 0.65 0 . 4 2.5 27.12 Completely automated sequential ICP system can analyse up to 20 Corp oration . t elements in an operator selectable multi-element programme; analytical parameters including wavelength selection, background increment selection and signal handling are programmable and storable via standard microcomputer: optical path purgeable permitting analysis t o 175 nm; optional HGA5M) furnace and gas-control systems permit instrument t o operate as completely automated ICP, flame AA and furnace AA system; ICP system retrofitable to existing model 5000 AA spectrophotometers vis. 1.3 Spectrametrics 1nc.h: Spectraspan 0.06 at 200 mm 0.75 - d.c. Optimized AE system using a high-dispersion, high-energy lllB 0.24 at 800 rnm throllghput echelle spectrometer: high temperature d.c.plasma source. Computer controlled with background correction facility. Modular and expandable simultaneous operation in presence of complex matrix solutions Spectraspan As IllB As I l l B As I l l B As l l l B Single channel module, sequential operation; high sensitivity even IV Spex Industries Inc., 1870 1-6 0.5 Plasma-Therm Crerny-Turner spectrograph/monochromator Metuchen, NJ 08840, 1702 1 . 1 0.75 - 27.12 Slew-scan monochromator 3880 Park Avenue, U.S.A. Glen Cpeston 1704 0.8 1 . 0 - - As 1702 16 Dalston Gardens, 1269 0.65 1.26 - - High resolution slew-scan monochromator Instruments Ltd., Stanmore, Middx. HA7 lDA, England * Address as in Table 2.6A t Address as in Table 2.6CTable 2.6C-COMMERCIALLY AVAILABLE ATOMIC ABSORPTION SPECTROMETERS Model beam) Supplier (single/double Resontion/ Type of Data Output Special features Baird-Atomic Ltd., A51 00 0.1 Bit parallel Automatic background correction by D2 HCL; 4-lamp turret; auto zero: integration; Warner Drive, (single) curve correction; wavelength scan: flame ignition: gas safety devices: lens optics: Springwood emission and fluorescence: optional microprocessor control for up to 8 standards Industrial Estate, with re-slope facility: illuminated status indication and date/trme clock Rayne Road, Braintree, Essex CM7 7YL England BCD (TTL levels) Baird Corporation, Alpha 1* 125 M idd lesex (single) Turnpike, Bedford, MA 01730, Alpha 2* U.S.A.Alpha 3* Alpha 4* Alpha computer systems 0.1 AS 5100 Single-lamp turret: fail-safe gas safety;digital concentration readout As Alpha 1 plus 4-position lamp turret As Alpha 1 plus automatic 0 2 HCL background correction and 2-speed wavelength scanning As Alpha 3 plus a 4-position lamp turret Colour or monochrome video display, unlimited curve and report storage on floppy disc; printer option GBC Scientific Equipment Pty. 7/63 Park Drive, GBC 0-5 IEEE-488 Ltd.,t SB 900 dandenong, ’ (single) Victoria 3175. Australia EDT Research, 14 Trading Estate, London NWlO 7LU, England 2-lamp supply; optional background correction: hydride generation and calculator available; length 700 mm, width 200 rnm, height 225 mm Hitachi Ltd., Nissei Sangyo Co.Ltd Mori 17th Building, 26-5 Toranomon, 1-chome, Minato-Ku, Tokyo, Japan Nissei Sangyo Instruments Inc., 392 Potorero Avenue, Sunnyvale, CA 94086, U.S.A. Nissei Sangyo GmbH, 4 Dusseldorf, Am Wehrhahn 41, West Germany 170-1 0 0.4 - (single) (single) 170-30 0.4 - 170-50 0.1 - 17C70 0.1 - (double) (double) Single lamp mounting, NzO/air simultaneously exchange: concentration readout: continuously variable time constant Concentration read-out; time weighted signal averaging: AAS/AFS measurement; auto zero; NzO/air simultaneously exchanged Base-line dritt correction; curve corrector: time weighted signal averaging; auto zero Polarized Zeeman effect background corrector to 1 - 7 absorbance units * New equipment since publication of Volume 9 t No up to date information suppliedlnstrumentat ion Laboratory Inc., 68 Jonspin Road, Wilmington MA 01887, U.S.A.Instrumentation Labora:ory (CIK) Ltd., Kelvin Close, Birchwood Science Park, Warrington. Cheshire, England 157 (single) 357::: (single) 451 * (single) 457* (double) 551 (double ) 991 (double; dual channel) 0.04 - Microcomputer controlled; calibration curve linearized using two standards. Fully automated gas box is standard feature; optional Ds arc background correction, optional 4-lamp turret and wavelength scan 0.C4 RS 232C As for IL 457 0.M RS 232C As for IL 551.CRT video readout: DZ arc background correction 0.04 RS 2326 Microcomputer controlled: calibration curve linearized using up to 5 standards; provides full statistics on results: fully automated gas box is standard feature; optional D? arc background correction; 4-lamp turret: wavelength scan and alphanumeric printer Microcomputer controlled: calibration curve linearized using up t o 5 standards: memory will store up to 10 calibration curves simultaneously: VDU displays standard conciitions for each element, the working curve and will show transient signals; fully automated fail-safe gas box is standard feature: optional background correction, 4-lamp turret, wavelength scan and alphanumeric printer Microcomputer controlled; calibration curve linearized in both channels using up to 5 standards: CRT video readout, will display 2 elements simultaneously A, B, A/B or A-B; internal standard and non-absorbing line background correction: VDU displays standard signals: automatic gas box is standard feature; optional 4-lamp turret; wavelength scan and built-in alphanumeric printer 0.04 RS 232C 0.04 RS 2326 Perkin-Elmer Corporation, Main Avenue, Norwalk. CA 06856.U.S.A. Perkin-Elmer Ltd., Post Office Lane, Beaconsfield, Bucks. HP9 1QA England Bodenseewerk. 2280 0.2 (single) 2380 0.2 (double) 4000 0.3 (double) Perkin-E!mer & Co. GmbH, Postfach 1120, D7770 Uherlingen West Germany 5000 0.3 (double) * New equipment since publication of Volume 9 EIA-RS 2326 High energy optical system: microprocessor controlled: auto-zero: auto concn.; auto curve with up t o 3 standards: peak height: peak area: integration time selectable from 0.2 to 60 s; statistics: flame ignition optional: auto N2O switching and burner head safety interlocks: optional flame and pressure sensing by microcomputer burner control. 0 2 arc background correction optional As model 2280 but all mirror optics: automatic gain control; auto NzO switching: burner head safety interlocks: optional 02 arc background correction with automatic intensity control Semi-automated sequential AA system: automatic gain control: instrument can analyse up to 6 elements with little operator participation; analytical parameters including standardization and signal readout can be entered and stored internally: digital stepper motor wavelength selection: flame ignition, auto N20 switchover, burner head interlocks: optional flame and pressure sensing by microcomputer, burner control: optional double-bezm background correction for all U.V.and visible wavelengths with automatic intensity control: lamp turret available Completely automated sequential AA system: instrument can analyse up t o 6 elements with minimal operator participation: all analytical parameters including lamp current, wavelength selection, resolution, gas flows, standardization and signal readout can be entered and stored using magnetic cards: optional double- beam background for all U.V.and visible wavelengths: when used in conjunction with HGA 500 it will provide sequential analysis for up to 6 elements with the same analytical ease as flame: when used with ICP-emission accessory it will provide sequential multi-element analysis for up to 20 elements with operator selectable background parameters EIA-RS 232C 2 way EIA-RS 2326 2 way EIA-RS 232C oc Ibo w Table 2.6C - COMMERCIALLY AVA1LABLE ATOMIC ABSORPTION SPECTROMETERS- coiztiirued Model beam) Supplier (single/dauble ResoAzion/ Type of Data Output Special features Zeeman:> 0.3 2 way (double) furnace AA Flame operation both with and without background correction as for model 5000.The instrument can be equipped simultaneausly for Zeeman-corrected graphite 5000 EIA-RS 2326: Pye-Unicam Ltd., SP 2900 0.2 BCD York Street, Cambridge CB1 2PX, England ( dou b fe ) SP 9 (single) SP 9 Computer PV 9090 0.2 - RS 232c 4-lamp magazine.auto zero' 2-standard curve correction, peak height peak area measurement; eiission; o-io mV fast analogue output (b-1 A), 10 m\i output for integrated or peak height reading. Fully microprocessor controlled data centre available, processing by programmable calculator. 3 types of curve correction; statistics; standard additions and other data handling; full calculator ability retained 8 modules available with combinations of 4-lamp turret; auto-gas control module with full safety interlocks; scale expansion; 2-standard curvature correction; burner interlock; output for SP 9 Computer and PV 9090 Data Graphics System and 0-10 mV (G-1 A) analogue output as standard Microprocessor data processing and control of flame automattc system for SP 9 and SP 2900.Curve correction with 5 standards in fixed and variable ratios. Peak height and/or peak area, full statistics, running mean, error warnings, built-in self-test routines. Integration and peak read times 0-1-100 s As SP 9 Computer, plus with PV 9095, video display of flame and furnace cookbook calibration curves and transient peak profiles with automatic scaling printout of furnace parameters Ran k-H i lger Ltd., Atomspek 0.2 - Westwood, M arg ate, H 1580 Kent CT9 4JL, Englanc; (single) 6-lamp turret; auto zero and flame ignition; curve correction; integration programmable calculator.Printer available !& k Shimadru-Seisakusho AA 625 0.2 - Quantitative flame emission; flameless capacity, flow lines for air, C?H2 and N?O 14-5 Uchikanda, 1 Chome, Chiyoda-Xu, AA 630 Tokyo 101, Japan (single) Ltd., (single) Road, England Ltd.,t (single) Ei 5 3 V.A. Hoive & Co. AA 640 0-2 - Au:omatic background correction; quantitative and qualitative flame emission; 3 88 Peterborough integration % London SW6 3EP, ? 2 0.2 - Quafit;tative/qualitative flame emission; ftarneless capacity, flow lines for air, C-H.' and N2O; flame monitor; gas pressure monitor; wavelength drive flameless capacity, flow lines for air, C2Hz and N20; flame monitor: wavelength drive, b 3 * New equipment since publication of Volume 9 t NO up to date information suppliedVarian-Techtron Pty. Ltd., 679/701 Springvale Road, Mulgrave, Victoria 3170, Australia Varian Associaies Ltd., 28 Manor Road, Walton-on-Thames Surrev KT12 2QF.England Palo Alto, CA 94: U.S.A. AA 1275* 0.02 IEEE-488 Two-lamp turret; overcoated reflective optics; automatic gas control system; (single) compatible with samplers, printer, hydride and furnace atomization systems. Intel 8080 with 10K ROM provides signal processing; background correction, absorbance conversion, integration; 3 standard curve fitting; peak height; peak area measurement; lamp-current control. D? arc background correction; new integrated high sensitivity atomization system RS 232C and parallel BCD AA 1475* (double) AA 875 (double) 0.2 IEEE-488 As for AA 1275 RS 232C and parallel BCD 0.05 lEEE-488 and duplex RS 232C Computer compatibls via two-way RS 23X for real-time signal acquisition, comprehensive report generation and instrument control; new integrated, high sensitivity atomization system.Four-lamp turret; compatible with desktop computer, printer, samplers, hydride and furnace systems. Intel 8080 with 15K ROM provides double beam background correction; absorbance conversion; integration; 5-standard curve-fitting; peak height and peak area measurement; statistics; self test and error detect ion VEB Carl Zeiss Jena, AASl 69, Jena, (single) Carl Zeiss Str. 1 AAS1 N German Democratic (single) equipment) Republic ( NzO Carl Zeiss Scientific Instruments Ltd., P.O.Box 43, 2 Elstree Way, Boreham Wood, Herts. WD6 I N H , England - 160 mV 4-lamp turret; single or triple pass optics; auto zero; titanium burner heads; flow lines for air, C~HZ, NeO; gas pressure monitor: gas flow monitor: burner head safety interlock: automatic flame ignition (600 ohms) for potentiometric recorder or absorbance converter.TECl printer or computer; VlDTEC1; signal output 775 mV (5 kohms) for linear recording or absorbance * New equipment since publication of Volume 9T k t b l C 3.6D-COlMMERCIALLY AVAILABLE ELECTROTHERMAL ATOMlZEKS AND AUTOSAMPLERS Supplier Model Control unit Ramp rate range Special features Baird Corporation ::: A 3470 Graphire Rod Programmable, dry, - ash 12 staaes).atom'ize; niax.'temp. 3500 "C F,its most AA spectrophotometers; air cooled; uses mains power; inert gas shielding; pyrolytic graphite coating for rods in situ' rapid interchange between flame and electrothermal methods' Instrumentation 655 Laboratory Inc.*: Graphite Programmable, 6 - furnace stages; ramp or step for auto zeroing and auto Calibrating the spectrop ho!ometer 254 Autosampler Digi:al timers for - for flame sample deposiiion, furnace trigger circuitry for operation auto zeroing and auto calibrating the spectrometer Perkin-Elmer HGA 400 Graphite Corporation * furnace Bodenseewerk, Perltin-Elmer Co., GmbH.A 540 Auto sampler HGA 500 Graphite furnace True on temperature read-out; LED display; safety interlock system; automatic cell door; automatic cleaning, cell pressurization; convenient solid sampling capaciry using micro boats Flame/furnace autosampling technique with auto calibration (FASTAC); employs an aerosol deposition technique of introducing aerosol into cuvette, which is at elevated temperature; the sample volume, which evaporates on contact with graphite surface, is controlled by length of time sample is sprayed into furnace, allowing operator to control sensitivity by varying deposition time 1-999 s Microprocesser unit From 2000 "C High-speed temperature accessory permits rapid heating to provides up to 8 steps s-1 'to 999 s temperature between 800 and 3000 "C for for optimal atomization of controlled+heating; be!ween any temp., ramp dme, two temps.hold time, gas and other furnace and spec?rophotometer control functions are programmed by direct keyboard entry; digital displays provide read-out of temp., time and prog. status Microprocessor unit As for provides up to9 steps of controlled heaiing; temp., ramp time, gas and other furnace and spectrop hotomeier control functions are programmed for each step by direct keyboard entry; digital displays provide read-out of temp., time and prog.HGA 400 00 P Automatic insertion of up to 35 samples and blank and 35 standards into the furnace; will also perform automatic method of additions; automatic matrix modification; recalibration; automatic triggering of furnace and instrument read cycle for unattended operation $ 3 2. 0 a_ - Furnace control programmes for up to 6 different elements may be stored in 6 programme memories; programme parameters for more than 6 elements can be stored on magnetic cards and recalled at the touch of one button; the optical temperature sensor and digital gas flow control for 2 different purge gases add to the versatility of the furnace programme; when used in combination with the A 540 microcomputer furnace auto sampler and the model 5000 AA up to 35 samples, blank and 3 standards, may be analysed for up to six elements each without operator attention $ 3 5- 2 d a 2 * Address as in Table 2.6Astatus; up to 6 complete programmes can be stored and recalled at the touch of 1 key -Unicam Ltd.* PV 9095 Graphite Microprocessor 18 ramp ra?es, Microprocessor selection and control of all functions including video furnace control of six phases 9 linear built-in auto sampler controls, video display of parameters and furnace and ramps to 3000°C 2-2000 "C s-1 status. Non-volatile storage of 10 programmes. Gas stop and Voltage or temp. and recorder control on all phases. Video display of peak shapes, control, no adjustment 9 exponential calibration and cookbook conditions when used with PV 9090 Data of photodiode sensor Graphics System; fits all current Pye-Unicam spectrophotometers necessary SP 9 furnace SP 9 furnace autosampler Graphite furnace 4-phases, each programmable to 2-2000 "C s-1 indication. Fits all Pye-Unicam spectrophotometers 3000 " C . Voltage or temperature control, no adjustment of photodiode sensor 9 ramp rates Digital parameter selection comprehensive status and fault Automatic sampling of - 38 samples and two wash positions. Selectable number of readings and volume for each sample: positive identification of blanks, standards and samples Rank-Hilger Ltd.* H 1475 Graphite Programmable: dry, - furnace ash, wait atomize; max. temp. 2600 "C Water cooled, inert gas shielding Shimadzu-Seisakusho GFAZ Graphite Programmable, dry, - $< t furnace ash, atomize: max. temp. 3000 " C Current stabilized to obtain reproducible results Var i an-Tec htron CRA 90 Graphite Programmable, dry, - Pty. Ltd.* furnace ash, atomize; max. (Graphite tube,temp. 3000 "C threaded graphite tube, graphite cup) Fits most AA spectrophotometers; water cooled; inert gas shielding and hydrogen flame option; automatic ramp hold atomization; pyrolytic graphite coating on cup and tubes * Address as in Table 2.6A t No up to date information supplied

ISSN:0306-1353    

DOI:10.1039/AA9801000065

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

CHAPTER 3 Methodology 3.1 NEW METHODS This Section describes novel methods of analysis that arc considered to be oE sufficient general interest to merit discussion here as well as in the appropriate section on specific applications. The Section also includes consideration of papers by workers who have made a detailed study of experimental parameters of widespread relevance. 3.1.1 Sample Preparation Techniques 3.1.1.1 Sample Drying, Washing, Solubilization and Unusual Dissolution Procedures.With the advent of rapid instrumental analysis techniques, the sample preparation and dissolution steps are probably the most time-consuming stages of many analyses. The drying of large numbers of sample aliquots in a conventional oven can be a lengthy process. Koh (1719) described the use of a microwave oven for rapid drying of homogenized samples of liver and carp fillct and also finely chopped lucerne.Drying of 10 g samples could take as little as 15 min, but if 15 samples were simultaneously dried, an increased drying time of 45 min was required. Weight losses wcre similar to those obtained using a conventional oven and no significant losses of Cd, Cu, Fe, Mn, Pb and Zn were observed.It should be noted howevcr that this technique might not be suitable for all types of samples. Thc absence of standardized washing procedures prior to the mialysis of human hair often causes irrcproducible results, poor comparability between different laboratories and prevents the establishment of reliable baseline results for unexposed subjects.In an attempt to overcome some of those problems, Arbige and six co-workers (C79) evaluated a number of washing agents such as dilute HNO,, sodium dodecyl sulphate (SDS) and SDS+EDTA at a range of pH values from 4-10 (see also ARAAS, 1979, 9, Ref. 487). The direct analysis of hydrophobic liquids and greases, after dilution with a suitable solvcnt system that is miscible with small amounts of aqueous inorganic master standard solutions, considerably simplifies the analysis procedure. It avoids the necessity of prcparing multi-clement standards from “oil soluble” compounds or using a suitable digestion tech- nique (see also ARAAS, 1979,9, 60).Hon et al. (1085) used isobutyric acid to dilute lubricat- ing oil samples and incorporated 1 ml of an appropriate aqueous calibration standard per 100ml.The final nebulized solutions could contain up to 5% m/V oil and satisfactory results for Ba, Ca, 0.1, Fe and Zn were obtained. The disadvantage of the method was that calibration by standard addition was necessary for most elements and that the N,O/C,H, flame could not bc used because of the relative insolubility of K or Na ionization suppres- sants in the isobutyric acid.The same group has similarly determined Fe in vegetable oils by diluting the oil using either propionic acid or MIBK/ethanol (8 : 3) and incorporating a small amount of an Fe aqueous standard (476). The determination of Zn in ointments (C625) and Pb in petrol (173), after solubilization with non-ionic emulsifiers, has been dcscribed. An ultrasonic agitator was used to produce the emulsions. The determination of l e d in petrol by a conventional procedure involving reaction with iodine (ARAAS, 1975, 5, 57) has been automated (C129).A continuous flow auto-analyser was used to dilute the sample with MTBK, react the alkyl lzad compounds with iodine and stabilize the response of the different alkyl lead-iodine compounds by complexation with Aliquat 336.An analysis rate of 30 samples h-1 was achieved.88 Analytical Atomic Spectroscopy A novcl method for the determination of ahmiria in wid on high-purity aluminium has been described by Matsumoto et al. (492). The sample was reacted with phcnol to convert the A1 into its phenoxide. The Al,O, was then filtered off, fused with KHSO, and the A1 subsequently determined by FAAS.Gale ((2875) has developed a very rapid anodic dissolution method for the determination of acid soluble A1 in steel. The steel sample was made into an anode and together with a graphite rod cathode was immersed in 25% V / V HCl. A current of 4 A was passed through the solution for a fixed time (typically 2min). This procedure resulted in dissolution of a constant amount of steel.Other rzferences of interest - Determination of dimethylpolysiloxanes in fats and oils : 445. Determination of total, organic solvcnt extractable, volatile and tetralkyl lead : 1242. Separation of Sb as SbT, : C1562. 3.1 .I .2 Sample Digestion, Oxidation and Fusion Procedures. The large number of digestion, oxidation and fusion procedures described in the literature each year shews JIO signs of abating.The newcomer must find it increasingly dificult to select the optimum method for any particular application. Rooney (C873) has outlined the advantages of incorporating perchloric acid in the wet digestion of a wide variety of organic materials (e.g., grass, meat, oil, etc.). While it could not be claimed that the use of this reagent is free from all hazards, it was contended that most of the dangers have been greatly exaggerated.Digestions incor- porating this reagent were claimed to be much more rapid than are most alternative tech- niques. (The reader is specifically referred to the Report of Analytical Methods Committee on “Notes on Perchloric Acid and its Handling in Analytical Work,” Analyst, 1959, 84, 214). Losses of chromium during wet digestions incorporating HClO, have frequently been attributed to the volatilization of CrO,Cl,.Pickett and co-workers (1 188), using radio- chemical techniques, have conclusively shown that this did not occur when yeast grown in a medium containing “Cr was digested with T-T,SO,/HNO,/HClO,. Any losses must be attributed to adsorption. T h i s confirms earlier work reported in ARAAS, 1978, 8, Ref. 534. Additional reference on the preceding topic - 1623. The optimum sample preparation technique for the aizalysis of sewage sludges for safe disposal to land is a contentious subject. Important criteria are that the selected method as well as being accurate, should be rapid, safe and suitable for the analysis of large batches of samples by junior personnel.Various methods have been recommendcd, such as HNO,/H,O, digestion (C71), acid digestion in a PTFE bomb (C71) and ultrahomogeniza- tion followed by direct ETA analysis (676) (ARAAS, 1979, 9, 153). Artiola-Fortuny and Fuller (540) compared six techniques and rccommendcd dry ashing or Na,CO, fusion. A much simpler technique has been described (1076). This involved refluxing 0.5 g of the dried sludges with HNO, in 50 ml calibrated glass tubes for 15 min with subsequent dilution to the 50ml mark with distilled water.Satisfactory results for Cd, Cr, Cu, Ni, Pb and Zn were obtained and this method was considered adequate for the purpose of screening sludge before disposal to the land. Pressure digestion vessels can be used to wet ash most plant and animal tissue, including bone, simply using a HNO, digestion (C214, 245).An inexpensive pressure digestion system (C49) utilized tightly capped polypropylene bottles in an autoclave at 121 “C (20 psig). By using this system it was possible to dissolve rapidly a wide range of alloys in 8M €€NO, and 6M HC1.Methodology 89 Digestions with phosphoric acid have rarely been used in the past because of the severe chcmical interference effects subsequently encountered in flame analysis. These effects, however, are not significant with the ICP.Thus many ores and rocks were successfully analyscd after digestion in H,PO, and dilution to a final 20% m/V H,PO, ((21464). The resulting polyphosphatcs can hold most elcments, including thc alkaline earths and silicon, in solution.Thc increased viscosity, however, reduccd the TCP nebulizer uptake ratc and conscqucntly degraded thc dAcction limits by 2-10 times. This drawback was countcr- balanced by the simplicity of thc digestion technique. Fusion techniques using LiBO, are well establishcd but are rather time consuming. Botto (6, 704) dzscribed an automatic fusion dcvicz that simultaneously fused six ash samples with LiBO, and cast the resulting flux mixtures into dilute HC1.The entire opcra- tion required about 15 min for six samples. An ingenious method of fusing micro-samples has been dcscribcd by Julietti (C162). A small bead of the flux is electrically hcated in a loop of Pt wire using a variable power supply; the powdercd sample i s then picked up from a polished stainless-steel surface with the bead in a tacky condition and the temperature then raised until fusion is complete. 3.1.1.3 Sample Preparation and Reduction Procedures for Mercury Anclysis. 'The rate of publication on this topic shows no signs of dzclining (see also ARAAS, 1979, 9, 61).Sodium borohydride has been used to reduce Hg(II), methyl mercury and phenyl mercury species to Hg') at the place of sampling, followed by 2dsorption of the liberated Hg on Au foil, which was thcn transported back to the laboratory (1521).This procedure minimized problems of sample preservation (see also ARAAS, 1979, 9, 59). Mizunuma et ol. (864) used NaBH, for the selective determination of Hg species. At pH I 1 only inorganic Hg is reduced, while at a pH of 2.5, in the prescncc of 1000 pg ml-1 Fe(III), the addition of 1% m/V NaBH, solution results i n total Hg reduction.Goulden and Anthony (1710) developcd an automated method for thc cold-vapour determination of various forms of Hg; EDTA/NH,OH was used to determinc the inorganic Hg, EDTA/SnCI, for inorganici-aryl Hg and CdCl,/SnCl, /NaOH for total Hg. Another speciation mcthod for foodstuffs and biological materials utilized homogenization with ethanolic KOH.The lipids were then removed with heptanc, NH,OH was used to liberate inorganic Hg and CuSO,/SnCI, was used to determinc total Hg (1610). Additional references on the preceding topic - 1301, 1620. Various digestion reagents have bcen recommended for total Hg determination V,O,/HNO, /H,S'O, for animal feedstuffs (1034), V,05 /HNO, with heating in an autoclave to 150 "C for biological materials (175), V,O,/HNO,/H,SO, at 140 OC for rice and soil (386), HNO, /H,O, or HNO3/I-Z,SO, /HClO, for marine organisms (896), HNO, /HCIO, / HCIO, for fish (299) and HNO,/HCl (9 : 1) at 50 OC for sediments and soils (1681). Additional references on the preceding topic - 393, 798, 905, 934.Tong and Leow (1 524) have directly determined total Hg in undigested biological samples. The samples (2.5g) were homogenized and shaken with 12.5 ml 30% m/V NaOH/O.G% m/V NaC1.The reduction step was carried out using SnCl,/CdCI, in 15% m / V NaOH. A novel method of determining Hg at high sensitivity using a conventional air/C,H, flame/nebulizer system has bcen developed by Palliere and Gernez (951). The acidic sample solution was fed to one branch and NaBH, solution to the other branch of a branchcd capillary uptake system.The HgO vapour formed in the spray chamber was then carried directly into the flame (see also Section 3.1.1.4). Iodine monochloride has been used t o absorb Hg from stack gases. The resulting solutions were nebulized directly into a d.c.arc plasma (1525). Thc method was 10 timcs90 Analytical A toniic Spectroscopy faster than the cold-vapour method, but could only be used for solution concentrations greater than 0.5 ,ug ml-1. Preconcentration of Hg by various techniques is still being reported, for example, ion- exchange using either acid digestion of the resin (679) or elution from the resin with thiourea/HCl (487), adsorption 011 to activated carbon (516, 1503) or adsorption on to sulfhydryl cotton fibre (935). Replacing AAS with the more sensitive AFS detection (see ARAAS, 1975, 5, Ref. 633) would avoid the necessity for preconcentration in most cascs.Other references of interest - ETA analysis for Hg : 488, 1191. 3.1.1.4 Sample Preparcrtion and Reduction Procedures for Hydride Generation Techniques.Considerable effort i s now dircctcd towards speciation studies of environmental samples. Hinners (1083) has shown that it is virtually impossible to determine selcctivcly As(III), As(V), mono and di-methyl As species in a samplc by dircct reduction unless the various hydrides are first physically separated. Separation by volatilization has been used (1603) to determine selectively inorganic As, mcthyl As, di-methyl As, tri-methyl As and phenyl As species.Gas-solid chromatography, after hydride gcneration, has been used in speciation studics of As, Ge, Se and Sn (1280), whereas ion-exchange, prior to hydride generation, has been used for As speciation (547). Nakashima (1079) scleetively determined Sb(IT1) by liberating SbH, with 0.35 M HC1 in the prcsence of 4 mg ml-1 Zr(1V).Under thcsc condi- tions Sb(V) was not reduced by NaBH,. For total Sb the reduction was carried out in 2M HCl/ 1 yo m / V KT. Melcher and Welz (1 323) showcd that it was essential to use Sb(V) standards when determining Sb in steel samplcs after dissolution in aqua regia. Rate of analysis mid inter-element effects are probably the main limitations of the hydride generation technique and the search for nicthods of overcoming these continues.Automation can overcome the former but interelemcnt effects still cause severe problems. Vijan and Leung (1711), using an automated technique, found that the interelement effects of Cu, Fe(II1) and Ni on Se(1V) are considerably minimized if the samplc contains 6-8M HCl.Various methods have been specified (1082) for minimization of the interference effects of Sb(TII), Au(III), Pt(IV), Fe(TTT), F- and S2- in the automatic hydridc gcneration of total As without destruction of methyl arsenic species before the hydride generation step. Agemian and Thompson (1074) determined total As and Se using an automated system and complete recovery from a wide range of organic As and Se compounds was obtained.The use of NaBH,CN in place of NaBH, minimized the deleterious effect of large amounts of Ni on the determination of As and Se (C81). It was also shown that volatilc boranes are evolved during hydride evolution with NaBH,. It should be noted that NaBH,CN liberates HCN on contact with acids and appropriate safety measures must be taken. Additional references on the preceding topic - 1501, 1538, C1546.The determination of lead by hydride methods is particularly prone to intereleinent effects (ARAAS, 1977, 7 , Ref. 295) and for natural waters this has been overcome by initial co-precipitation of the Pb on to MnO, (C73, 186). This procedure cffcctively eliminated Cu and Ni interference. Jin and Taga (1353) used K,S,O,/NaBH,l to gencratc PbH, and collected it in a rcservoir before passing it into an air/H,/N, flame.The method was successfully used for the detcrmination of Pb in steel and natural water samples. Additional references on the preceding topic -C121, 466, 1607. Severe losses of H,Se on to various materials has been observed by Reamer and Veilloii (C120) using radiochemical detection techniques. Even PTFE cxhibited significant H,Sc absorption.If all the glassware was treated with dimcthyl dichlorosilanc, the adsorption decreased to a mean level of 3-5%. Watling and Watling (1226) have generated As, Sb and Se hydrides at the site of sampling using NaBH,. The hydrides were absorbed on to HgCI,Methodology 91 or AgNO, impregnated pads, which were then transported back to the laboratory for analysis.Direct nebulizarion of acidified solutions and a NaBH,l solution through a branched capillary nebulizer has been used (941) to directly generate hydrides in the spray chamber prior to flame atomization (see also Section 3.1.1.3). A similar technique where two separate nebulizers wcre used has also been reported (C84). The acidified sample was passed through one nebulizer and a 5% in/V NaBH, solution was passed through the other.An approximately 10-fold improvement in detection limit compared with conventional nebuliza- tion of the acidified sample was observed. Other references of interest - AsH,/SbH, in battery gases : C76. Hydride generation ICP-OES : C84. S, S2-, SO2-,, SO”-, speciation using MECA : 998.Se determination using MECA : 471. 3.1.2 Precoricentration Techniques Papers describing new preconccntration procedures continue to proliferate. Problcms are envisaged with translating many of these to the analysis of large numbers of samplcs. Often these procedures require precise attention to detail, which is difficult to achieve for prolonged periods in a routine laboratory.Also, in many cases it is difficult to discern any marked advantages over previously published methods. Techniques that can be automated or partially automated have much to recommend them. 3.1.2.1 Solvent Extrcrction. Multi-element extraction of natural waters with APDC / MTBK/ xylene at pH 3.8 was used to determine As, Cd, Co, Cr, Cu, Mn, Ni and Zn (539). The organic extract was evaporated to dryness and the residue dissolved in a small amount of acid.A 50-fold concentration was achieved. 4-Benzoyl-3-methyl-1 -phenyl-5-pyrazolone has been used to extract Co, Cu, Mn, Ni and Pb from solar evaporation ponds (1679). The addition of sodium tartrate and extraction at pH 11.5 has been recomrncnded for minimizing Fe interference when determining Ag, Cd, Cu, Ni and Pb in natural waters using a DDC/ MIBK extraction (1699). When solvent extraction methods are directly applied to natural water samples without any pretreatment, the endogenous analyte is not always quantitively extracted.Further work has again demonstrated this often overlooked fact (1690). Very poor recovery of naturally occurring trace metals in estuary water was found when using a direct APDC/CCI, extraction.A HNO,/H,SO, pre-digestion was required to ensure quantitative recovery (see also ARAAS, 1979, 9, Ref. 1467). Back extraction of Freon TF or CHCI, extracts into dilute acid has again been recommended for overcoming instability of some metal chelates in “wet” organic solvents (I 192, C1770) (see also ARAAS, 1978, 8, 64). The solvent extraction of chromium has attracted considerable attention.T’he conven- tional APDC/MIBK system can bc prone to serious errors unless considerable care is exercized in controlling both the extraction pH and the amount of KMnO, used to oxidize the Cr(II1) to Cr(V1). The following extraction systems have been recommended for Cr : 5% m / Y methyltricaprylammonium chloride / MIBK (1 188), thiosemicarbazide/MIBK (1 703), trioctylamine/MIBK in the presence of HC1 and NH,Cl (1665), directly into MIBK (CI 555) and tri-benzylamine /CHCl, followed by back extraction into NH,0H/H202 (1492).Solvent extraction of thaflium species is rarely reported; however Aihara and Kiboku (525) have cxtracted T1 into MTBK from a solution containing potassium xanthate and ammonium tartrate at pH 8. The technique was compared with conventional APDC and DDC extractions and has also been adapted for Bi.Selective determination of inorganic TI92 A nal y tical A tomic Spectroscopy in rat urine has been achieved (982) by allowing the urine to react with Br, water, which converts inorganic T1 to TIBr, which was directly extracted into MIBK. The (CH,),Tl species remaining in the aqueous phase were then extracted with NaDDC/MIBK.Selective noble metal solvent extraction systems have proved very useful when determin- ing low levels of the noble metals in complex matrices (see ARAAS, 1979, 9, 129). Tsuka- hara and Tanaka (197) developed a method that allows 2 pg Ag to be separated from 1 g Zn and 5 g Ni using tri-n-octylammonium bromide dissolved in n-butyl acetate to extract the Ag from an HBr solution.The method has also been adapted for determining trace levels of Pd (lSI1). Russian workers (1019) have used o-isopropyl N-methyl thiocarbamine /toluene to extract Ag from geological samples, whereas others (10 18) used triphenylphosphine /C,H, t a extract Ag from chlolride solutions of Pt. Haddon and Pantony (185) directly extracted Au from ores and products by suspending the ground sample in 0.1M HCI and shaking for S h with amyl acetate containing a small amount of Br,.The method was specific for Au and none of the other 25 elements tested was extracted in significant amounts. Further work using petroleum sulphides to extract Au has been reported (1 103) (see also ARAAS, 1974, 9, 129). Seknium(ZV) has been extracted from a wide range of samples with 4-nitro-o- phenylenediamine /toluene (443) and APDC / MIBK (1 489).In the latter case a 1 00-fold excess of Cu and 5% m/V EDTA were added to minimize the extraction of other elements and to enhance the Se sensitivity when using ETA analysis. Other references of interest - Determination by solvent extraction of -A1 : 1608. -Cd, Pb : 1680.-CO: 1637. -CU : 1500. -Hg : 1333. -V : 1363. -Ni : 1362. 3.1.2.2 Zon-exchange Methods. Strelow (1491) separated Fe from Al, Ca, Cu, Mg, Mn and Ni in 8'M HCl using AGl-X4, a 4'y0 cross-linked anion-exchange resin. Tiron (disodium 1,2-dihydroxybenzene-3,5-disulphonic acid), retained on Amberlyst A-26 macroporous anion exchanger, has been used to precolncentrate a wide range of trace elements (190).Brajter et al. (188) have studied the optimization of conditions for the separation of Pt and Rh using cation exchange resins. Additional reference on the preceding topic - 794. Barnes ei al. (C1394) have continued their work on preconcentration using polydithio- carbarnare chelating resins followed by nitric acid dissolution of the resin (see ARAAS, 1979, 9, 63).More than 50 elements including Th, U, noble metals and rare earths could be preconcentrated without simultaneous preconcentration of Ca, K, Mg, and Na. By careful selection of pH it was possible to diffcrentiate between Cr(1IT) and (VI), Fe(I1) and (III), U p ) and (Vl) and Ce(I1I) and (IV). Additional reference on the preceding topic - 898. Knapp et al. (C209) have adsorbed Cd, Co, Cu, Ni, Pb, Se and Zn on to an APDC- loaded Chromosorb Supelco W / H P reversed-phase column.The column was dried by an Ar stream and directly connected to either an AA burner or an ICP torch. The column was then inductively heated resulting in rapid volatilization of the metal chelatzs into the Ar stream. Additional reference on the preceding topic - (C575).Methodology 93 3.1.2.3 Adsorption Concentration and Other Techniques. Calhoun and Hurley (1 31 9) added 0.5 g Zn dust to 11 of the sample (i.e., a pure solution) to concentrate Cu.The solid residue was collected and dissolved in a small amount of 4M HNO, and Cu wxs determined by FAAS. It is doubtful whether this method would be satisfactory for samples that con- tained organically bound forms of Cu unless pre-digestion with acid was used.Hanagan and Skogerboe (C9) preconcentrated a number of elements after reduction and precipitation using NaBH4. Selective adsorption on to a suitable substrate is a useful preconcentration technique. Iridium, Pd, Pt, Rh and Ru have been co-precipitated on to CuS and 2-mecaptobenzothiazo1 (997); silver has been selectively adsorbed onto dithizone particles from an acidic solution containing a large amount of Pb (468}; charcoal and tannic acid have been used to concen- trate Mo, Sn and W from natural waters (1263); spheron thiol has been used to adsorb As, Bi and Sb (898), and APDC was used to adsorb Bi, Cd, Co, Fe, Tn, Ni, Pb, TI and Zn selectively from an aluminium matrix (81 8).Other references of interest - Determination of complexation capacity using copper phosphate : 1692, 1693. Freeze drying preconcentration of natural waters : 485.Methods for multi-element preconcentration from pure lead : 1197. 3.1.3 Indirect Methods It would be of considerable interest to know how many laboratories routinely used any of the large number of indirect methods cited in this Section over the last few years; probably only a few (see ARAAS, 1979, 9, 64).Two methods for the determination of sulphate after BaSOl, precipitation with excess Ba have been reported (842, 928). Tn the latter method, instead of simply determining the excess Ba, it was precipitated as BaCrO, and the excess Cr determined by AAS. The indirect determination of phosphate after formation of molybdophosphate species still continues to be reported (see ARAAS, 1978, 8, 65) using both flames (367, 1293) and ETA (491, C631).It is dificult to appreciate the advantages of many of the published indirect AAS methods over existing colorimetric methods, many of which can be readily automated. Two methods for the determination of cyanide have been reported, the first (240) utilized the formation of Ni(CN),2- followed by precipitation of the excess Ni as sulphidc and subsequent AAS dctermination of the residual Ni(CN),Z-, and the other involved the formation of Hg(CN),, which was not readily reduced to HgO (249).A cyanide characteristic concentration of 0.4 ng ml-1 was achieved by this latter method. Halide ions have been determined by monitoring the molecular absorption of various metal halides in a graphite-tube furnace; the halides used were AlF (1318), T1T (498), InBr and TlBr (497).A full investigation into the use of Al, Ga and Tn for the determination of bromide, chloride and fluoride was carried out by Fuwa et al. (1 252) (see also AHAAS, 1979, 9, 65). Methods for the determination of chloride by precipitation of AgCl and determina- tion of the excess Ag continue to be advocated (806, 1032).The determination of detergents by indirect AAS methods appears to have potential considering the alternative methods that are available. Matsueda and Morimoto (1 361) determined linear alkyl benzosulphonates (LAS) in natural waters. Ion pairs were formed between LAS and the thiourea-Cu(1) complex and these were then extracted into MIBK.Two methods for the determination of non-ionic detergents have been proposed; one was based on precipitation of the complexes formed with molybdophosphoric acid (1644) and94 Analytical Atomic Spectroscopy subsequent FAAS determination of Mo, whereas the other relied on precipitation with modified Dragendorff’s reagent followed by determination of the Bi in the precipitate.Ingenuity is the hallmark of indirect methods and this year there has been reported (292) a method for determining Cd with a filter flame photometer. The samplc solution was evaporated to dryness with 0.001N H,SO, and treated with excess LiI solution. The Li,CdI, formcd was then extracted into MIBK and thc Li determined by FAES. The detecti.on limit was 0.4 pg ml-f Cd.Indirect AAS methods have also been proposed for : biacetyl (863), B (313), 1,2-diols (C628, 1526), histidine (337), proteins (339) and sulphide (1075). 3.1.4 Nebulization, Vaporization and Atomization The weakest link in most ICP systems is the nebulizer. A grcat deal of effort is therefore being directed towards the design of nebulizers that arc robust, not readily blocked by fine particulates and will tolerate appreciable (> 1 % m / V ) dissolved solids concentrations.Various improved permanently aligned, cross-flow systems have been described that incor- porate : a fixed jewel orifice for the nebulizing gas and a fixed precious metal capillary tube for the sample (C61, C751); an all-glass nebulizer (C15, 1523); a Pt/Ir nebulizcr (C63); a nebulizer that can be rapidly removed from the spray chamber if a blockage should occur (C62).An HF resistant concentric ncbulizer, spray chamber and torch has been describcd by Goulter (C 1392) and applied to alloy analysis. The precision of most ZCP nebulizers is of the order of 0.01-0.02 KSD. Anderson et al. (C749) described a high-pressure (200 psi& cross-flow nebulizer that gave typical RSDs of 0.0025-0.005.The main factors limiting the precision were variations in the coolant and nebulizing gas flow-rates caused by temperature fluctuations. Mass-flow controllcrs mini- mized these variations. Some very remarkable results were claimed for Zr, a mean RSD of 0.00056 being observed over a period of twelve days. Babington nebulizers continue to bc investigated (C61, C751, C752); suspended matter up to 1 mm diameter can be tolerated and a wide range of sample types nebulized (see also ARAAS, 1979, 9, 65).It is unlikely however that suspended matter with a diameter significantly in excess of 10 pm would be fully atomized, even assuming that it is transported into the flame or plasma. The direct rzebulizartioiz of slurries continues to show significant advantages with respect to sample preparation time (see ARAAS, 1979, 9, 65).Stupar and Ajlec, using an ultrasonic agitator, ((21605) have prepared 0.5-2% m/V soil suspensions in 25% V / V isopropanol. The relative efficiency of atomization of Cu, Fe, Mg and Mn was investigated as a function of suspension concentration, type of soil, particle size and flame temperature.It was concluded that, with suitable calibration procedures and a 100 mm path length 30% N,O/70% air/C,H, flame, it was possible to obtain satisfactory results for the above-mentioned elements. The typical accuracy was to within t 15% with a time saving of - 70% compared with normal sample preparation techniques. Additional references on the preceding topic - C12, 635.Taylor and Floyd ((258) have evaluated an ultrasonic nebulization aiid desolvatioiz system for the ICP analysis of drinking waters and obtained “lowcst quantitatively deter- minable concentrations” for Ag, As, Ba, Cdc, Cr, Pb and Se that were below the EPA maximum contamination levels (see also ARAAS, 1977, 7 , Ref. 1691). Commercial manufacturers are still devoting time to the development of FA AS nebulizers, New nebulizer designs with significantly improved detection limits have been reported (C83, CS29, C833, C1465).Although it is not normally feasible to nebulize chlorinated solvents continuously, pulse nebulization of 50 pl aliquots of CC1, and CHCI,, solutions after extraction of trace metal APDC complexes has been reported (91 1).Methodology 95 The possibility of direct analysis of orgarzic species by ETA, using molecular absorp- tion, has been further explored by using an instrument with a cathode ray tube display ((2115) (see also ARAAS, 1979, 9, 29).Vijan (1320) determined low levcls of Ni using nickel carbonyl generation. The sample was spotted on to a filter-paper and reduced to elemental Ni using NaBH,.The Ni was then allowed to react with CO at room temperature and the resulting Ni(CO), was passed to an electrically heated silica tube. The method was extremely sensitive and a characteristic concentration of 19 pg ml-1 was observed for a 1 ml sample volume. Two methods for improving atomization eficiency in air/C,H, flames have bcen reported. The addition of 0.08% m / V carbon powder increased the Ca response and removed chemical interference of sulphate ion (316).The presence of tertiary amines allowed the direct determination of Al in the air/C,H, flame (518), although the detection limit was somewhat worse than that usually obtained in the N,OiC,H, flame. Rigin et al. (1 528) determincd zirconium by atomic fluorescence spectrometry after formation of volatile ZrCI, or ZrF, and, subsequent reduction by Na vapour at 2000 "C.There has been a large increasc in the number of papers directed towards the ICP deter- mination of elements that are considered to be difficult to cxcitc and/or detect. Considerable further work on the vacuum ultra-violet wavelength region I: as bcen reported, thc following elements having been detcrmined by ICP : Hg, P and S (C6139, C1397), B (C1396), C, P and S (1043), S (170, C7158) and B, Hg, I, P, and S (C658, C831).Dcnton et at. (C136) described a vacuum spcctrometcr detection systcm that responded to radiation down to 120 nm and could consequently detect Br, C, Cl, N, 0 and S with good sensitivity, thus allowing total 0, and N, determinations that arc very difficult to achieve by any other technique.North- way et al. utilized non-resonance visible transitions from high-energy states to detect F at 685.6 nm (1 31 I), N at 821.67 nm (857) and 0 at 779.2 nm (858). For 0 and N dctermina- tions a power in excess of 2 kW was requircd to minimize chemical interference effects. element effects in graphite tube ETA dctcrminalions of a wide range of clemcnts in various samples (C103, 1256, 1328, C1475).It appears to have considerable potential for the direct atomization of solid samples (1056, 1237) and Ottaway et al. (C1724) extcndcd its use to the emission ETA analysis of clinical samples using an cchelle monochroniator and wavelength modulation. Rapid heating of ETA devices minimized the inter-element effects nornially obscrved for Cd and Pb in a NaCl matrix (1249).The graphite-tube atomizer was heated at a rate of 2000 "C s-1. Chakrabarti et 01. have continued their work on capacitive hearing o f ETA devices (see ARAAS, 1979, 9, 25). Significantly improved sensitivity and minimal i.nter- fcrence effects were observed (1057). It was claimed that as all samples were completely atomized, it was possible to analysc samples by an "absolute method", because the sensitivity was independent of the matrix.This also has the advantage that minimal sample preparation is needed and no calibration graphs are required (C600, C606, C612, 1070, C1167, C1458). Copper-plated graphite tubes for ETA determination of As : C112. Effects of impact beads and mixer paddles and auxiliary oxidant flow on FAAS : 856.Flow injcction FAAS analysis : C77. Pulse nebulization/ICP : 1055. The L'vov platform is becoming an increasingly popular niethomd of reducing inter- * Other references of interest -96 Analytical Atomic Spectroscopy 3.2 DETECTION LIMITS, PRECISION AND ACCURACY Two significant factors affecting the precision of major component determinations by FAAS are ( i ) the very high dilution factors required to bring thc working solutions within the linear working range and (ii) small changes in flame stoicheiometry necessitating frequent re-calibration.The use of an internal standard, where a dual channel instrument is available, is becoming morc widesprcad judging by the incrcased number of papers published this year. A good example of the use of an internal standard to overcome the above problems is the dctcrmination of Ca in cement.Sotera et nl. (C85, (21052, C1372, C1409) have shown that the addition of an accurately measurcd amount of Sr at the beginning of the samplc- preparation stage considerably improved the precision as well as the speed of the detcrmina- tion. Satisfactory results were obtained for NBS and Portland Ccment Association ccments using a N,O/H, flame. Tsujino et 01.(519, 527) found significant zllcviations in the viscosity effects attributable to the presence of large amount of cthaiiol and glucose when analysing wine for Cd, Cr, Cu, Fe, Mn, Pb and Zn by FAAS after the addition of Au as internal standard. Anderson et al. (C749) have described the irnprovcments in precision for ICP-AES obtained when using a fixed high-pressure cross-flow nebulizer dcsigiied to opcratc at 200psig Ar pressure.By using thc high pressure nebulizer together with an internal standard of 1 pg ml-1 Sc, an overall RSD of 0.005 was claimed (see also Section 3.1.4). As a result of proposals made at the Second ICP Conference in Noordwijk aan Zee in 1978 an international detection limits committce was set up and a dctcction limits programme published (C23).The purpose of thc programme i s to arrive at cxperimcntal data on detection limits attained with various XCPs under “rcal sample” analysis conditions. The detailed programme document gave a number of niles for the experimental determination of detection limits in order that unambiguous comparisons of complete XCP analysis systems could be made as well as assessment of the separate roles played by the main components of the system, i.e., the source, optics and detector read-out device.The preparation of properly matched synthetic matrix calibration solutions is oftcn very difficult and the method of standard additions is a widely used alternative. Hall et al. (1081) (see also ARAAS, 1972, 2, Ref. 371) described a simple direct conccntration read-out mcthod for samples containing different analyte concentrations but the same matrix composition; the standard additions need only to be made to one sample. The instrument settings are adjusted so that direct positive read-out of analyte concentrations is obtaincd for subsequent samples. Thc applicability of the method was checked by using as cxamplcs the dcpressivc effect of Fe(TI1) on Cr(II1) in an air/C,H, flame and the enhancing effect of Ba on Ca also in an air/C,H, flame.Timm et al. (1017) reviewcd the physical and mathematical implications ol thc mcthod of standrrrd additions, thereby forming the basis of a routine procedure for planning, performing and evaluating analytical data with particular reference to AAS.Methods for improving the limited dyizamic range of the atomic ubsorptioiz technique have been the subject of much study. Epsteiii and Winefordner (194) asscssed the technique of horizontal burner rotation with particular emphasis on the analytical precision and linearity of response. Analytical data on the detcrmination of Zn in urban particulate matter using an air/C,H, flame were presented to illustrate that an increasc in dynamic range in the upper region of the working curve had not bccn achicved at the expense of precision.Limbeck et al. (376) claimed a significant extension of the AA working range using a rational-method calibration algorithm, which allowed synthesis of a mathematical curve that closely matched the curvature of all types of calibration curves at high absorbance values.For Ni, at 232.0 nm, an extension of the working range to 50 pg ml-1 was found to beMethodology 97 possible. The routine application of these techniques might possibly result in carryover problems for samples with widely different concentrations. Other references of interest - Non linearity of working curves : C123.Concentration range covered by ICP-AES’ calibration curves : C767. 3.3 STANDARDS AND STANDARDIZATION 3.3.1 Reference Materials A major International Symposium entitled “The Production and Use of Reference Materials” organized by the Bundesanstalt fur Materialprufung was held in Berlin in November 1979. Some 40 papers were presented covering all aspects of reference materials under the broad headings : general aspects; metallic reference materials; inorganic reference materials; reference materials with physical and physico-chemical properties; reference materials with physical and technological properties; nuclear and radioactive reference materials; clinical and organic reference materials.Proceedings of the symposium are available from BAM, 1 Berlin 45, Unter den Eichen 87, Germany. Rossi (1 540) described the preparation and certification of an oxide silicate-base copper ore RM carried out under the auspices of the European Community Reference Bureau.Thc samplc was analysed by 14 laboratories and techniques used were : electrogravimetry, potentiometry, iodimetry, polaroigraphy, AAS and isotope dilution MS. 3.3.2 Standardization Studies The primary aim of round robin interlaboratory analysis of RMs i s to establish accurate determinand levcls in the materials.Diversification of analytical techniques i s highly desirable as a means of eliminating analytical bias and there is considerable value in the methodological data that emerges from such collaborative exercises. Steger and Faye (1 87) summarized the information on methods used for the analysis of 26 reference material ores for 27 elcments produced by the Canadian Centre for Mineral and Energy Technology.A critical survey of a number of trace element analytical techniques used by life scientists has been published (1583). Techniques included were FAAS, ETA, ICP-AES, MS and XRF and the criteria used for comparison included sensitivity, specificity, accuracy and precision, single versus multi-element capability, analysis time and general applicability to analytical problems.The paper cited 203 references. A concluding report (168) of the exhaustive intcrlaboratory comparison studies carried out by the South African National Institute for Water Research has been published. Two interesting points to emerge for those interested in AAS were first, that a considerable variety in the type and quality of interference suppressants used in FAAS was apparent, and secondly, that NaBH, was more effective as a hydride generating agent than was zinc slurry.The data from 16 laboratories were found to be acceptable €or final certification in a study involving 46 determinands. The results of two national studies of the Ca, Cd, Co, Cr, Cu, Mg, Ni, Pb and Zn contents of raw and potable water in Canada (536) have shown that none of these were found at hazardous concentrations.It was found that direct ETA-AAS compared favourably with the APDC extraction method except for Pb and Cd. Cadmium was the only metal to be subject to interference from suspended particulate matter.One of the problems involved in interlaboratory method studies lies in persuading every participant to follow exactly the prescribed methodology. A study of the AAS deter-98 A naly tical A torn ic Spectroscopy mination of Sn in foods, organized by the National Food Process Association, Washington DC (253), yielded results from eight laboratories, only six of which followed the prescribed procedure that involved digestion of the sample in HNO,/H,SO, and addition of methanol to enhance the absorption signal before aspiration into a N,O/C,H, flame.The method was adopted as an “interim official first action”. A cohborative study of the analysis of foods for As, Cd, Cu, Pb, Se and Zn using pressure digestion with HNO, has been madc (579).Arsenic and Se were determined by AAS with hydride generation and Zn by FAAS. Thc remaining elements were measured by anodic stripping voltammetry. The collaborating laboratories achieved 89-108% element recoveries with RSDs of 0.16 and 0.20. An AAS method for the analysis of work place atmospheric aerosols (924) has been tested by circulating to sixteen laboratories sample aerosols, having particulate diameters of 0.1-10pm, that had been collected on filters.Thcse were analysed for Cd, Co, Cr, Ni and Pb at three concentrations levels. The overall percentage recoveries and standard deviations for the metals were, respectively, 100.8 t 9.9, 97.6 2 13.9, 96.6 t 10.8, 98.6 2 10.3 and 98.7 f 12.2. In a similar study of the AAS Reference Method for Pb (805), the US Environ- mental Protcction Agency established that results would be expected to differ by less than * 17% for within laboratory measurements and by -t 28% for between laboratory measurc- ments.Notice of the designation of three “manual equivalent” methods for the determina- tion of Pb in suspended particulate matter has been given by the US Environmental Protec- tion Agency (1050).They are : (i) determination by FAAS following ultrasonic extraction with hot HNO, /HCI; (ii) dctermination by ETA-AAS; (iii) determination by ICP-AES. The first method is identical to the EPA Reference Method for Pb except for the extraction procedure, which is more stringent. The second and third methods use either of thc two extraction procedures specified in the Reference Method or the ultrasonic procedure.A reference method for the determination of K in serum has been established by the US National Measurement Laboratory (453). The accuracy of the FAES method was established by comparison of the values obtained by 12 laboratories against those obtained by a definitive analytical method based on isotope dilution MS. Seven serum pools with K concentrations ranging between 1.3 and 7.3 meq 1-1 were analysed.The South African National Institute for Metallurgy has been active in the development of laboratory methods for the analysis (inductively coupled plasma OES) of Mn-bearing materials, phosphate rocks and H,SO, leach liquors (650, 1038). Thc methods can be used to determine a number of elements from a single sample solution and in each case employ LSc as an internal standard.Methodology TABLES 3.38.1 -3.3A.5 : REFERENCE MATERIALS Table 3.3A.1 FERROUS METALS AND ALLOYS 99 Supplier Finely divided form Solid form Amt fur Standardisicrung und Warenpriifung (ASMW), 102 Berlin, Wallstrasse 16, D.D.R.Bundesanstalt fur Materialpriifung (BAM), 1 Berlin 45, Unter den Eichen 87, Germany Bureau of Analysed Samples Ltd., Newham Hall, Newby, Middlesbrough, Cleveland TS8 9EA, England Bureau National de Metrologie (B.N.M.), 21 rue Casimir Perier, 75007- Paris, France Ccntro Nacional de Lnvestigaciones Me talurgicas, Cuidad Univcrsitaria, Madrid 3, Spain Gosstandart of the USSR, 9 Leninsky Prospckt, 11704, Moscow, U.S.S.R.Tnstitut. dc Rccherches dc la Sidkrurgic Francaise, B.P. 129, 78104-Saint Germain en Lays, France Iron & Steel Institute of Japan, .lapa n MBH Analytical Limited, Station House, Potters Bar, Herts. EN6 IAL, England Metalimpex, POB 330, H-1376 Budapest, Hungary National Bureau of Standards, Ofice. of Standard Reference Materials, Washmgton, DC 20234, U.S.A. Spex Industries Inc., 3880 Park Avenue, Metuchen, NJ 08840, U.S.A. Unalloyed & alloyed steels, cast irons, slags, fcrro alloys Unalloyed & alloyed steels, slags, cast irons, ferro alloys High purity irons, unalloyed & alloyed steels, slags, cast irons, ferro alloys Unalloyed & alloyed steels, cast irons High purity irons Unalloyed & alloyed steels Unalloyed & alloyed steels, cast irons Unalloyed RL alloyed steels, ferro alloys, steels cast irons, slags Unalloyed & alloyed Unalloyed & alloyed steels, cast irons steels Unalloyed & alloyed Unalloycd & alloyed steels, cast irons Unalloyed & alloyed steels steels, cast irons Unalloyed & alloyed Unalloyed & alloyed steels, cast irons, ferro alloys Unalloyed & alloyed steels, cast irons Unalloyed & alloyed steels, cast irons100 A naly tical A tom ic Spectroscopy Table 3.3A. 1 FERROUS METALS AND ALLOYS- continued Supplier Finely divided form Solid form Swedish Institute for Metal Research, Drottning Kristinas vag 48, S-11428 Stockholm, slags Sweden South African Bureau of Standards, Private Bag X191, Pretoria, Transvaal 0001, South Africa Unalloyed & alloyed stecls, ferro-alloys, Ferro alloys Table 3.3A.2 NON-FERROUS METALS AND ALLOYS ~~ Supplier Finely divided form Solid form Aluminium Company of America, Alcon Technical Center, Alcon Center, PA 15069, U S A .Aluminium Pechiney, 23 bis, rue Balzac, 75360 Paris Cedex 08, France Amt fur Standardisierung und Warenpriifung (ASMW), 102 Berlin, Wallstrasse 16, D.D.R. British Aluminium Co. Ltd., Chalf ont Park, Gerrards Cross, Bucks. SL9 OQB, England Bundesanstalt fur Materialprufung (BAM), 1, Berlin 45, Unter den Eichen 87, Germany Bureau of Analysed Samples Ltd., Newham Hall, Newby, Middlesbrough, Cleveland TS81 9EA, England BNF Metals Technology Centre, Grove Laboratories, Denchworth Road, Wantage, Oxfordshire, England A1 base High-purity metals, Al, Mg base Sn, Al, Mg base Al, Cu A1 base Cu, Ni, Al, Mg base High-purity metals, High-purity metals, Al, Mg, Cu, Ni, Sn, Pb base Al, Ch, Ni base Cu baseMethodology 101 Table 3.3A.2 NON-FERROUS METALS AND ALLOYS- continued __ Supplier Finely divided form Solid form Canada Centre for Mineral and Energy Technology, c / o Coordinator, CANMET, 555 Booth Street, Ottawa, Ontario K1A OGI, Canada Commissariat a l'Energie Atomiquc, (C.E.A.) Cristal Tec, B.P.no. 85 Centre de tri, 38041 --renoble Cedex, France Centre Technique des Industries de la Fondcrit; {C.T.I.F.), 44, Avenue de la Division Leclerc, 92310- Svres, France Centre Technique du Zinc, 34, rue Collange, 92300- Levallols Perret, France Gosstandart of the USSR, 9 Leninsky Prospekt, 1 1704 Moscow, U.S.S.R. Inco Europe Limited, European Research and Development Centre, Commercial Development Department, Birmingham B16 OAJ, England Japan Brass Makers' Association, Japan Japan Light Metal Association, Japan Light Metal Smelters Association, Japan Johnson Matthey Chemicals Ltd., Orchard Road, Ro ys t on, Herts.SG8 SHE, England MBH Analytical Limited, Station House, Potters Bar, Herts. EN6 IAL, England Mercure Industrie, 13, rue Saulnier, 92800- Puteaux, France Cu base Cu base Al, Mg base Cu base High-purity metals, Zn base Cu base Ni base Cu base Al, Mg base Al, Mg base High-purity metals High-purity metals Al, Cu, Ni, Zn, Co base High-purity metals102 Aitalytical Atomic Spectroscopy Table 3.3A.2 NON-FERROUS METALS AND ALLOYS- contirtuecl Supplier Finely divided form Solid form ~~ Metalimpex, A1 base POB 330, H-1376 Budapest, Hungary National Bureau of Standards, High-purity metals, OfIice of Standard Referencc Materials, Al, Co, Cu, Ni, Pb, Washington, DC 20234, U.S.A.Mg, Sn, Pb, Ti, Zn, Zr base Planet-Wattohm, High-purity metals 053104a Roche de Rame, France Spcx Industries Inc., 3880 Park Avenue, Metuchen, NJ 08840, U.S.A. Al, Cu, Pb, Ni, Ti, Zn, Zr basc Cu, Pb, Sn base Tablc 3.3A.3' GEOLOGICAL MATERIALS Supplier Finely divided form ~~ Amt fiir Standardisierung und Warenpriifung (ASMW), 102 Berlin, Wallstrasse 16, D.D.R. Bundesanstalt fur Materialprufung (BAM), 1 Berlin 45, Unter den Eichen 87, Germany Bureau of Analysed Samples Ltd., Newham Hall, Newby, Middlesbrough, Clcvcland TS8I 9EA, England Canada Centre for Mineral and Energy Tcchnology, c / o Coordinator, CANMET, 555 Booth Street, Ottawa, Ontario KlA OGI, Canada Centre National de la Recherche Scientifique, Centre de Recherche Petrographiques et Geochimiqucs (C.N.R.S./C.R.P.G.), 14, rue Notre Dame des Pauvres, Case Officielle No. 1, 54500 Vandocuvre-lez-Nancy, France Mn, Cr, Sn ores Fe ores Fe, Mn, Cr, Al, ores fluorspar, sillimanite, Na & K feldspar, magnesite, dolomite, limestone Sb, Co-Mo, Au, Fe, Mo ores syenite, gabbro, ultramafic rocks Ores, rocksMethodology Table 3.3A. 3 ,GEOLOGICAL MATERTALS- continued 103 Supplier Finely divided form Commission of European Communities, Community Bureau of Refcrencc (BCR}, 200 Rue de la Loi, B-1049 Brussels, Belgium Geological Survey of Japan, Japan Gosstandart of the USSR, 9 Leninsky Prospekt, 1 1704> MOSCOW, U.S.S.R. Tnternational Atomic Energy Agency, Analytical Quality Control Services, Laboratory Seibersdorf, PO Box 590, A-101 1 Vienna, Austria Junta de Energia Nuclear, Chidad Universitaria, Madrid-3, Spain L.R.M., 13.P. 3013, 54000 Nancy Ccdex, France National Bureau of Standards, Office of Standard Reference Materials, Washington, DC 20234, U.S.A. National Chemical Laboratory for Industry, lapan South African Bureau of Standards, Private Bag X191, Pretoria, Transvaal 000 1, South Africa Zn, Sn, Cu ores Rocks Fe ores U ores Lignite Rocks Fe, Al, Cu, Mo, Li, Zn, W ores, fluorspar, Na & K feldspar, clays Rocks Rocks, Fc, Cr ores Table 3.3A.4 GLASSES, CERAMICS AND) REFRACTORIES Supplier Finely divided form Bureau of Analysed Samples Ltd., Newham Hall, Newby, Middlesbrough, Cleveland TS.8 9EA, England Centre d’Etudes et de Recherches de L’industrie des Liants Hydrauliques, 23, rue de Cronstadt, 7501 5- Paris, France Silica brick, firebrick, rnagnesite-chrome Portland cement, zircon, high purity silica Cement104 Analytical A tomic Spectroscopy Table 3.3A.4 GLASSES, CERAMICS AND REFRACTORIES- coizlirzued Supplier Finely divided fom Centre National de la Recherche Scientifique, Centre de Recherche Petrographiques et Geochimiques (C.N.R.S./C.R.P.G.), 15, rue Notre D a m e des Fauvres, Case Officielle No. 1, 54500 Vandoeuvre-lez-Nancy, France Federation Europeenne des Fabricants de Produits Refractaircs (P.R.E.), 44, rue Copernic, 75016 Paris, France L.R.M., B.P. 3013, 54000' Nancy Cedex, France National Bureau of Standards, Office of Standard Reference Materials, Washington, DC 20234, U.S.A.National Chemical Laboratory for Industry, Japan Society of Glass Tcchnology, 20 Hallam Gate Road, Shefield S10 SBT, England Glasses (2 available) Refractory materials Refractory materials Lead/ barium, opal, high and low boron, soda limc glasses, silica, aluminosilicate and chrome refractories, Portland cements Sodalime silica, silica, high silicic acid - high boric acid glass Glasses (3 available)Methodology Table 3.3A.5 ENVTRONMENTAL MATERIALS 105 Supplier Finely divided form Bureau of Analysed Samples Ltd., Newham Hall, Newby, Middlesbrough, Cleveland TS8 9EA, England Tnstitut de Recherches de la Sidhrgic Francaisc, B.P. 129, 78104-Saint Germain en Laye, France National Bureau of Standards, Office of Standard Reference Materials, Washington, DC 20234, U.S.A.Furnace dust (LD) Furnace dust (electric) Orchard lcavcs, bovine liver, rivcr sediment, urban particulatc matter, coal, fly ash, wheat flour, rice flour, yeast, tomato lcavcs, pine needles, spinach, oyster tissue Table 3.3B SUPPLIERS OF SPECTROGRAPHIC GRAPHITE ELECTRODES 1 2 Baird Corporation Inc., 125 Middlescx Turnpikc, Bcdford, MA 01730, U.S.A.Carbon Products Division, Union Carbide Corp., 270 Park Avcnuc, New York, NY 10017, U.S.A. (ARL Ltd., Wingatc Road, Luton, Bcds., England) 3 Labtest Equipmcnt Co., 1182& La Grange Avenue, Los Angela, CA 90025, U.S.A. 4 Johnson Matthey Chemicals Ltd., Orchard Road, Royston, Herts. SG8 5HE, England. 5 Le Carbone (GB) Ltd., Portslade, Sussex, England 6 Le Carbone Lorraine, 37-41 Rue Jcan-Jaurcs, 9223 1 Gennevilliers, Francc 7 Jarrell-Ash, 590 Lincoln Strect, Waltham, MA 02154, U.S.A. 8 Zebac Inc., P.O. Box 345, Bcvca, OH 44017, U.S.A. 9 Ringsdorffe-Wcrke GmbH, 53 Bonn-Bad Godcsberg, West Germany (Mining & Chemical Products Ltd., Alpcrton, Wembley, Middlesex HA0 4PE, England) Spex Industries Inc., 3880 Park Avenue, Metuchcn, NJ 08840, U.S.A. (Glen Crcston, 16 Dalston Gardens, Stanmorc, Middlcsex HA7 lDA, England) Ultra Carbon Corp., P.O.Box 747, Bay City, MI 48706, U.S.A. (Hcyden & Son Ltd., Spcctrum House, hldcrton Crescent, London NW4, England) 10 I 1106 Annlytical Atonzic Spectroscopy 1 2 7 8 9 10 11 12 13 14 15 16 17 18 Table 3.3C SUPPLIERS OF STANDARD METAL SOLUTIONS (MS) AND REAGENTS (R) FOR AAS Aldrich Chemical Co.Inc., 940 W. St. Paul Avcnue, Milwaukee, WI 53233, U.S.A. (R) J. T. Baker Chemical Co., 222 Red School Lane, Phillipsburg, NJ 08865, U.S.A. (MS, R) Barnes Engincering Co., 30 Commerce Road, Stamford, CO 06902, U.S.A. (MS) BDlH Chemicals Ltd., Poole, Dorset BH12 4NN, England (MS, R) Bio-Rad Laboratories, 2200 Wright Avcnue, Richmond, CA 948041, U.S.A. (MS) Carlo Erba, Divisione Chimica Industriale, Via C. Imbonati 24, 20159 Milano, Italy (MS) Eastman Organic Chemicals, Eastman Kodak Co., 343 State Street, Rochestcr, NY 14650, U S A . (R) Fisons Scientific Apparatus Ltd., Bishop Meadow Road, Loughborough, Leics. LEll ORG, England (MS, R) Harleco, Div. of Amcrican Hospital Supply Corp., 60th and Woodland Avenues, Philadelphia, PA 19143, U.S.A. (MS) Hopkin & Williams Ltd., P.O. Box 1, Romford, Esscx RMl lI-IA, England (MS, R) V. A. Howc & Co. Ltd., 85 Pctcrborough Road, London SW6 3EP, England (MS) Instrumentation Laboratory Inc., I 1 3 Hartwell Avcnue, Lexington, MA 02173, U.S.A. (MS) (R) Johnson Matthey Chemicals Ltd., Orchard Road, Royston, Hcrts. SG8 WE, England Koch-Light Laboratories Ltd., Colnbrook, Bucks., England (R) (Anderman & Co. Ltd., Central Avcnuc, East Molesey, Surrey KT8 OQZ, England) May & Baker Ltd., Dagcnham, Esscx RMlO 7XS, England (R) E. Merck, D 61 Darmstadt, West Germany (R) Spcx Industries Inc., 3880 Park Avenue, Metuchcn, NJ 08840, U.S.A. (MS) ALFA Division, Vcntron Corp., 152 Andovcr Street, Danvcrs, MA 01923, U.S.4. (MS) (Glcn Creston, 16 Dalston Gardens, Stanmore, Middlcsex AM7 lDA, England)Methodology 107 Tablc 3.3D SUPPLTERS OF ORGANO'METALLIC STANDARDS I 2 3 4 5 6 7 8 9 1 0 11 12 13 14 15 16 1 7 Angstrom Inc., P.O. Box 248, Bcllevillc, MT 481 11, U.S.A. naird Corporation Inc., 125 Middlescx Turnpike, Bedford, MA 01730, 1J.S.A. J. T. Baker Chcmical Co., 222 Red School Lanc, Phillipsburg, NJ 08865, U S A . BDH Chemicals Ltd., Poolc, Dorsct I33211 2 4NN, England Burt and Harvey Ltd., Brcttcnhani House, Lancaster Place, Strand, London WC2, England Carlo Erba, Divisionc Chimica Industrialc, Via C . Iinbonati 21, 20159 Milano, Ttaly Conostan Div., Continental Oil Co., P.O. Drawer 1267, Ponca City, OK 74601, U.S.A. Durham Raw Mzterials Ltd., 1-4 Great Tower Street, London EC3R 5A13, England Eastman Organic Chemicals, Eastman Kodak Co., 343 State Street, Rochcstcr, NY 14650, U.S.A. Hopkin and Williams Ltd., Y.O. nox 1, Romford, Esscx RM1 l I M , England E. Mcrck, D 61, Darrnstadt, West Germany MI3H Analytical Ltd., Station House, Potters Bar, Ilerts. EN6 1 AL, England Division of Chcmical Standards, National Physical Laboratory, Teddington, Middlesex 'l'Wl1 OLW, England National Spectrographic Laboratories Inc., 19SOO South Miles Road, Clcvcland, OH 44128, U.S.A. National Bureau of Standards, Office of Standard Reference Materials, Washington, DC 20234, U.S.A. Kcscarch Organic/Tnorganic Chcmical Corp., 11 686 Sheldon Strcct, Sun Valley, CA 91 352, U.S.A. ALFA Division, Ventron Corp., 152 Andover Strcct, Danvers, MA 01923, U.S.A. (Glen Crcston, 16 Dalston Gardens, Stanmorc, Middlescx HA7 1 DA, England)

ISSN:0306-1353    

DOI:10.1039/AA9801000087

出版商:RSC    

年代:1980

数据来源: RSC

摘要:

CHAPTER 4 Applications EXPLANATION OF THE TABLES Each of the Applications Sections, 4.1 to 4.9, i s accompanied by a Table which summarizes the principal analytical features of the references from which the corresponding Section is compiled. All relevant references are included in the appropriate Table, while the accom- panying text discusses only the more noteworthy contributions. These Applications Tables form a convenient source of information for analysts interested in particular elements, matrices, sample treatments, or atomization systcms. In many cases, sufficient detail is given for the analytical procedure to be followed; absencc of such detail usually means that the information was not directly available to the compiler of the table, and the original reference should be consulted.The key to the tables is given below. ELEMENT h/nm MATRIX CONCENTRATION TECH. ANALYTE The elements determined are listed in alphabetical order of chemical symbol, except that, for space economy, multi-element applications (5 elements or more) arc given at the end of some tables. The wavelength, in nanometres, at which the analysis was performed. An indication, necessarily brief, of the material analysed.The concentration range or level of the element in the original matrix, expressed as 96 or pgg-1, for solids and mgl-1 or pgml-1 for liquids. The atomic spectroscopy technique i s indicated by A (absorp- tion), E (emission), or F (fluorescence). The form of the sample, as presented to the instrument, is indicated by S (solid), L (liquid), or G (gas or vapour). SAMPLE TREATMENT A brief indication i s given of the sample pretreatment required to produce the analyte. ATOMIZATION REF. The atomization process is indicated by the abbreviations A (arc), S (spark), F (flame), or P (plasma), usually with some additional descriptive detail. The number refers to the main Reference section, which gives the title of the paper and the name(s) of the author(s), with address. The prefix ‘C’ indicates a paper presented at a meeting. 10.9

ISSN:0306-1353    

DOI:10.1039/AA9801000109

出版商:RSC    

年代:1980

数据来源: RSC