ISSN:0306-1353    

DOI:10.1039/AA97606FX001

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

2 Excitation Sources and Atomizing Systems 2.1 ARCS, SPARKS, AND LASERS Optical emission spectroscopy i s one of the oldest, and yet still a widely used method of analysis. Some of the more modern excitation techniques for emission have been revicwed in a critical and comprehensive manner by Laqua (785, 1341), and the ‘Analytical Chemistry’ biennial review contains many references to excitation sources (906). 2.1.1 Fundamentals Papers on fundamental aspccts have been relatively few. The axial-rzdiation density distribution in the arc plasma of elements of different ionisation potentials has been dcter- mined (1331). In the presence of additives such as K salts there i s a redistribution of the atoms along the discharge axis (997). Vukanovic et al. (808) have reported further work on the measurement of the transport velocities of elements such as Na, Li, Ca, and Sr by firing ‘liquid bullets’ into a d.c.arc plasma. A photographic method using monochromatic radiation emitted by the arc has been proposed by Dittrich et al. (822), known as ‘equi- d-nsitometry’, for determining the movement of ions in a d.c. arc. Avni and Tnnor (1097) used floating probes, i.e.tungsten wires passing at a velocity of about 2ms-1 through a d c. arc, to measure electron temperature, electric field strength, and electron density. Studies of a d.c. arc constricted in a metal tube of diameter 10 mm have shown that excitation conditions, and consequently spectral line intensities, are altered relative to a free-burning arc (826). Petrovic et al. (1455) contrasted the radial distribution of tempera- ture, electron density, and particles in a free-burning arc in air with that of an arc burning in a water-cooled metal tube.Vukanovic et al. (1454) have developed several types of d.c. arc sources, burning in a horizontal graphite tube, for the determination of extremely low concentrations of elements in powdered samples. A mass spectrometer was used (1459) to analyse products formed in an arc constricted in a tube and in a free-burning arc.Voltage fluctuations at different time intervals during an arc burning between electrodes of differing shapes have been recorded (1456); three distinct burning periods were observed. Several workers have made theoretical studies aimed at the optimisation of the spcctro- chemical processes.Fegas et al. (1493) have considered the criteria influencing the choice of exposure in a d.c. arc procedure. Plsko (859) discussed the influence of various para- meters on ‘detection power’, using the photographic recording technique, and Ustimets (969) assessed absolute limits of detection for Cr in solution under conditions of fluctuating analyte signals.A major source of error is the calibration plot, and in a detailed examina- tion of curve fitting, Goode and Northington (68) observed that, in comparison with linear-linear plots, accuracy is much improved by presenting data in a log-log format. Matherny (998) has also studied the optimisation of the calibration line. Other references of interest - Correlation between theoretical and experimental analytical curves: 1530. Optical diagnostics in an arc plasma: 1517.Statistical procedure for parameter optimisation in spcctrochemical analysis: 757. 2.1.2 Buffers and Matrix Effects Salts of K and Na are most commonly used as buffers, normally in combination with graphite. High-temperature reactions taking place in the arc plasma in the presence of NaCl and NaF have been studied (514), and Zadgorska et al. (139) observed that different K salts affect the rate of evaporation of elements from the surface of the electrode to different 6Part I : Fundamentals and Instrumentation 7 extents, due t o the effect of the anion on the electrode temperature.Sokolowska (829) described experiments in controlled atmospheres in the presence of NaCl which also showed that evaporation rates are dependent on the gas mixture used.Improved detection limits have been claimed in the analysis of Zr (828), and refractory metals and water, using NaCl buffer mixtures (110). Kubova and Plsko (496, 827) investigated evaporation rate in a d.c. arc as a function of crater depth, and provided further experimental confirmation that, as the crater deepens, fractional distillation is suppressed, enabling the simultaneous determination of both volatile and involatile elements without total burning of the sample.Karpel et al. (1668) have studied the vaporization of elements from a mixture of GaAs with graphite powdcr and from graphite powder alone, using radioactive tracers and by the measurement of emitted spectral intensities.Thermochemieal reactions of Ti compounds in a graphite matrix have also been reported (1308). Tetraethylammonium iodide has been shown to give a better separation of the volatile impurities present in refractory metals than when using graphite: the reagent stabilises arc temperatures, a necessary condition for its successful application as a spectroscopic buffer (810).Matrix effects due to changes in the excitation conditions of the plasma can be signifi- cantly reduced by using constant arc temperatures (800). A system is proposed where the temperature of the arc is maintained by means of a regulated voltage. Matherny (1 310, 1315) has discussed the possibility of quantitative classification of matrix effects on the basis of experimentally determined calibration functions.Results are given which were obtained using samples of silicate and carbonate minerals, and ores. Papers presented at a Symposium on Matrix Effect in Atomic Spectroscopy in Poland in October 1975 provide a useful review of the subject (1346), emphasising the role of gaseous atmospheres (1340, 1342, 1343) and the physical and chemical aspects of the analysis of non-conducting materials (1 345).Gordon (317) has shown that matrix vaporization effects can be minimised in a d.c. arc in an Ar atmosphere by coating the carbon electrode with an AgCl buffer. The sample solution or suspension is pipetted onto this coating to form a thin layer, which attains the maximum temperature of the anode spot during the discharge. Malicki (828) also found that the use of an AgCl additive combined with an Ar/N, atmosphere gave very efficient vaporization in the determination of impurities in Zr.Spectral interference effects have been investigated, using rotating graphite disc elec- trodes. Price and MacNaughton (181) used an Fe matrix solution as the basis of a universal method for iron, steel, ferro-alloys, and water.Moselhy et al. (1041) considered the effects of Fe on As, B, Cu, Hg, Sb, and T1; A1 on Pb; Mg on TI; As on Cd; and V on A1 in the analysis of environmental materials, using an Ebert-type 3.4 m spectrometer. Other references of interest - Applications of graphite powder standards: 437. Interacting chemical systems in the spark discharge: 942. Precision measurements using graphite-powder sample admixtures: 1248. 2.1.3 Excitation Techniques The theory and practice of laser microprobe analysis has been reviewed by van Deijck (1241).In the search for rapid methods, McGillivray (108) has used an approach based upon the measurement of spectrogram line-image widths resolved by a prism spectrograph Line widths recorded on Polaroid prints were measured, using a micrometer, and related to measure- ments made using a reference sample and multi-step filters.Using this technique, the compo- sition of small samples (down to 10-8 g) can be quantitatively determined in 10-15 minutes.8 Analytical Atomic Spectroscopy Vyzhelevskii et al. (1373) cmployed a laser dcvicc producing 0.3 J pulses at a repctirior; rate of 0.5-1 .O Hz for the analysis of brass, and Buravlev et al.(966) examined the effect of m:tallurgical structure on thc laser microanalysis of cast iron. Spectroscopic observations mad: on a plasma produced by a dye laser at the surface of garnet crystals have provided information on the mechanism of plasma formation and spectral line characteristics (609). The possibilities of AAS with laser atomization were studied by considering the effect of gas-phase reactions on the optical characteristics of the laser plume (608).Kircheim e f al. (1398, 1464) uszd single crystals of metals and alloys as specimens for laser microprobe analysis in ordrr to study the cffect of orientation on the evaporation process. They noted that the line density in photographically recorded spectrograms was strongly dependcnt 'uycn crystal orientation; as a rule, the line density increases with the density of atoms within a crystal plane.Exploding conductors are still a relatively novel excitation sourcc for ES. The limita- tions of el-ctrodeposition on a silver wire conductor as a sample-introduction technique have been circumvented by the direct dispensing of p1 volumes of the sample on a strip of aluminium foil (91 9, 949, 1108).Exploding-conductor excitation can produce analytical curves which are linear over three or four concentration decades, with photographically determined limits in the low and sub-nanogram rangc, but the high continuum bachgroiind and n:imJrous interferences present significant problems (1 109). A substantial improvement can, however.bc achieved using thin films as substrates with a 7.5 pF capacitative discharge, cbarged to 1.5-3 0 kV. Pavlovic et al. (728, 825) have continued their studies on the influence of an external rotating magnetic field on the analytical properties of the d.c. arc. Only in the case of cathode excitation w x e noticeable changes in intensitics noted. Possible explanations of tb.: enhancements are given in t e r m of altercd evaporation and mass-transport processs The effect of magnetic fields and additives on the determination of trace elements has also been d-scribed (602, 1005).A method has been described for solution analysis in which the solution is introduced through a hollow metallic or carbon electrode into the analytical gap by means of an electrically driven syringe (1466).Analytical possib es have been demonstrated by the analysis of water (1466) and lubricating oils (1460). The application of conventional, rotating graphite-disc elcctrodes has been mentioned in section 2.1.2 (181, 1041). Coraor and Barnes ( I 110) have further developed the liquid-layer spark-sampling technique, in which a thin layer of water or solutton is applied to the surface of a plane sample electrode.The technique has becn shown to increase sensitivity, reduce matrix effects, and provide increased sample rcmoval and melting; these effects are exemplified by the analysis of Pb inclusions in a leaded brass. Hoyt et al. (59) have described a capillary arc for use in com- bination with an aerosol generator remote-sampling device.Aerosol particles (1 pm and smaller) are injected into the arc, which is used purely for dissociation and excitation, thus allowing discharge parameters to be set for optimum conditions of excitation. Improved linearity of calibration curves i s claimed, enabling a wide range of steel types to be analysed using a single working curve. Hobson and Ambrose (509) have now published a description of their remote d c.arc excitation unit, incorporating a fibre-optic light guide. The apparatus is used primarily for the inspection analysis of steel tubes. Until rccently, such techniques were restricted to direct-vision spectroscopes, which, although portable, were entirely dcpcndent upon the judgement of the operator. In this new device, a copper electrode contained in a simple press-down holder is applied to the sample; a d.c.arc provides sufficient light for a light guide several metres long. The apparatus is based upon a multi-channel direct-reading spectrometer and is used to determine such elements as Mn,Part I: Fundamentals and Instrumentation Cr, Ma, and V either in a quantitative mode or on an ‘accept-reject’ basis. Increasing the repetition rate of spark discharges i s a well-established technique for reducing the time taken in the ES analysis of steels, A new sOilrce unit with a repetition rate of 600 discharges s-1 (67), and controlled by the same dedicated computer as that used to control the spectrometer, has been described.It has been suggested that a higher repeti- tion rate may affect analytical performance.Kusnetsova et al. (1519) obtained best results by using a high-voltage spark with a frcquency of 400 discharges s-1, but they noted that the increasc in the total number of spark discharges during the preliminary arcing, and not thc increasc in frequency, was the main factor in improving precision. Imamura et al. (109) have developed a new measuring method of spark discharge which cmploys a multi-channel pulse-height analyser to measure integrated photocurrent for each discharge during the predetermincd sparking period.Improved source units, together with refinements in the selection of spectral lincs and the stabilization of spectrometers, now make it possible to obtain accuracy comparable with that of X-ray fluorescence; this is illustrated by data for the analysis of complex tool steels (73). A comparison between d.c.arc ES, AAS, and XRF for the analysis of rock samples showed no major disparitics; in gencral, ES data were adequate but most prone to crror at low concentrations (318). Boumans et al. (814) compared alternative d.c. arc pro- cedures incorporating fusion for trace analysis in geological materials; they found that elements of low and high volatility gave better accuracy without fusion, whereas elements of medium volatility gave best results when fused.Further studies of the miniature spark discharge with a photoelectric time-rcsolved dctection system have been published by Glass and Crouch (1 12). The highly reproducible spark is formed by the discharge of a coaxial capacitor through a stream of gas flowing at right angles to an optical window. This recent work has been aimed at determining those source parameters which are important for its use as an analytical tool.Information is presented on sample-introduction systems, carrier gases, power supplies, spark capacitor designs, and carrier gases. The discharge has also been investigated for possible use as a detector for gas chromatography, and is capable of detecting picogram quantities of C, P, N, H, S, B, and many other elements (76). 9 Other references of interest - Effect of pores in sample electrodes: 298. Temporal scanning of arc and low-voltage spark spectra: 1515. Vacuum cup electrodes: 500. 2.2 PLASMAS The past year has seen a very considerable growth in the literature pertaining to the application of plasma sources in optical emission spectrochemical analysis.The principal sources of current interest are the h.f. inductively coupled plasma (ICP) and the microwave- induced plasma (MIP); some new work has been described concerning capacitatively coupled microwave plasmas (CMP) and d.c. arc plasmas. 2.2.1 R.f. Inductively Coupled Plasmas Morc than 70 papers concerned with ICP sources have been published in the literature or prcscnted at Confcrcnccs during 1976; these have been dcvoted to all aspects of the opera- tion and application of thesc sources, and include numerous review papers by well-known workers who have pioneered the use of LCPs in optical emission spectrometry ( I , 157, 722, 954, 958, 984, 1040, 1086 1225, 1388).The commercial availability of a number of plasma sources and completc TCP spcctrometer systems for simultaneous multi-element analysis has10 Analytical Atomic Spec troscopj led to rapid expansion of the applications literaturc, although much of this at prcsent originates from the laboratories of the instrument manufacturers. Reports havc appeared of the round-table discussion on ICP sources held at the XVIIlth C.S.I.at Grenoble, France, 1975 (1322) and of the ICP meeting held at Noordwijk, Holland, in June 1976 (1224). The readers’ attention is again directed towards the existence of the ICP Nendetter (see ARAAS, Vol. 5, p. S), which contains much valuablc material and includes correspond- ence. abstracts, reports of meetings, and original papers.Fuizdarizerital Studies. Studies concerned with the nature and characteristics of thc ICP discharge have continued. Fasscl and co-workers (1 38, 1067) have made mcasurements of spatially resolved radial excitation temperatures and radial electron number dcnsity distribu- tions (n,) cxperienccd by analyte species 15 to 25 mm aboce the load coil of a toroidal argon PCP and related these to the analytical performance of the plasmas.The low dcpend- ence of 12, and temperature distributions on thc addition of easily ionized elements suggests that changes in the total composition of the sample do not affect radial excitation tcmpera- tures or the degree of ionization of analytcs in a dominant manner. Other workers (301) haw utilized the relative radiances of Hp, N,, and Hs lines from H, added to a 9 MHz ICP discharge to determine temperature profiles; after Abel inversion and corrcction for self- absorption, the considerable disagreement between the values obtained using Hp : H,.HD : Ha, and H, : 13, intensity ratios led the authors to conclude that thermal equilibrium did not prevail in the source. Dc Galan (1065), in a rcview of excitation temocrntures and interferences in the TCP, has reported temperaturc values ranging from 3000 to 9000 K in 3 medium-powered ICP, depending on the species and the cxcitation process considcrcd: simple explanations of interferences in the ICP on the basis of thermal equilibrium arc thus not possible.This review also cautions against the comparison of optimized TCP systems with unbuffered microwave plasmas when buffering has been shown to produce improve- ment in performance (747).Barnes and Nikdel (314, 1234) have refined and cxtended earlier work conccined with computer simulation of the ICP discharge for spcctrochemical analysis. A modified mathematical model, originally described by Miller and Ayen (J.Appl.Phys., 1969, 40, 5260), 11’15 been used to predict radial and axial temperature profiles in an ICP hit11 N, support gas.This model incorporates an updated description of three concentric gas-flow patterns commonly encountered in spectrochemical applications; the results obtained parallel those of similar simulations for Ar discharges. Chandra and Barnes (1068) have reported n further extension of this model to include consideration of more complex gas-flow patterns.Barnes and co-workers (314) havc also made a detailed comparative study, by computer simulation, of the temperature and velocity profiles, and particle decomposition, for ICP discharges in Ar and N,. Although the Ar discharge temperatures are found to be higher and the linear velocity of the central gas stream lower than for N,!, the N, plasma is shown to bc poten- tially a better medium for the decomposition of A1,0, particles.The N, plasma is predicted to be stable above 3.7 kW and the Ar plasma ;ibove 0.6 kW, which accords with practice Genna and Barnes (1069) have reported preliminary results for experimental temperature and velocity profiles in an JCP; radial pressure profiles wcre determined using a Pitot tube, capacitative pressure transducer, and microcomputer system, and were related to vclocity profiles using a corrected Bernoulli equation.The excitation temperature profiles were measured spectroscopically. Jarosz and Mermet (1094) have reported thc observation of new lrncs of S produccd on introduction of 13,s into an argon ICP. Thc authors postulate that metastable levels of ionized Ar play an important role through charge transfer to an excited state (cf. Penning ionization); thus Ar+”+S-+Ar+S+.Energies of these mctastable levels arc ca. 32 eVPart I : Fundamentals and Instrumentation 11 above the ground state. This could explain the excitation of lines up to 30 eV and the absence of observed Ar(II) lines from 35eV. It is suggested that the overpopulation of metastable levels of Ar(1) and Ar(I1) compared to higher levels is a good assumption to explain the departures from overcxcitation in the ICP.Fisher (823) has measured rotational excitation and Doppler temperatures in a 'brush' discharge at 27 MHz in air. Instrurnenlatiorz. Allemand and Barnes (71) have discussed the electrical parameters affecting the discharge initiation processes in ICP torches; the same authors have also reported the initiation and operational characteristics of various plasma-torch configurations employed during the development of an unattended computer-operated ICP system (69).Defarrari (1025) has pointed out the advantages to be gained from the use of an automatic tuning system with ICP sources operated from crystal-controlled generators; with this type of system the influence of impedance changes during source operation is corrected, and more reproducible analytical results may be attained. Lichte and Koirtyohann (1016) have described a demountable torch assembly for ICP work which is easily constructed and repaired and which results in plasmas of similar stability and symmetry to those operating on all-silica one-piece torches.Ohls and Krefta (1012) have published the design of an adjustable torch which allows optimum positioning of the three silica tubes with respect to each other. Olson, Haas, and Fassel (1087) have described a new ultrasonic nebulizer and its application to emission spectrometry with an ICP source; used in conjunction with aerosol desolvation, this device permits an improvement by an order of magnitude or more in detection limits, Results comparable to or better than those reported by Boumans and de Boer (Spectrochim.Acta, 1975, MB, 309) with similar facilities have been obtained. The nebulizer js chemically inert, reliable, and provides rapid and convenient sample-change Lichte and Koirtyohann (1070) have reviewed the analytical advantages and disadvan- tages of horizontal operation of an ICP source, where the axial channel of the toroidal plasma is viewed 'end-on' rathsr than in the usual direction, in which it is vertical and at 90" to the optical axis of the spectrometer.The advantages claimed for this arrangcment arc increased analyte line intensity, reduced background, reduced molecular band intensity, improved simultaneous multi-element analysis performance, and decreased influence of gas flow-rates.'The disadvantages are stated as decreased dynamic range and increased observed interfercncc effects. A gas 'cutoff' system has bccn used to remove reabsorbing atoms from the optical path and cxtend the dynamic range with this mode of operation. Kiortyohaiin and co-workers (951) have also described the use of a minicomputer-controlled linear photodiode array, optical system, and stigmatic monochromator to collect atomic emission data with both high positional and high spectral resolution from an ICP.Scheeline and Walters (921) have discussed problems associated with obtaining spatially resolved informa- tion from plasma discharges; phenomena which distort measurements are considered in terms of the Abel invcrsion technique for elucidation of radial emission line profiles.Montaser and Fassel (918) have studied the use of an ICP discharge as an atomization c ~ l l for AFS. Using thermostatted EDL sources, the detection limits by AFS for Cd, Zn, and Hg wcrc supxior to thosc observed in emission with the same apparatus. Denton and Wind5or (75) have used an ICP as a detector for gas chromatography; C, H, and S were dctccted in the atomic state in the lower part of the discharge.Molecular band spectra from the tail-flarnt: werc used to detect 0, and N,. Allemand and Benzie (1023) have described an optical arrangsmcnt which allows the12 Analytical Atomic Speciroxvpy opcration of an ICP or arc source with an Ebert spectrograph, so that either source may bc employed on thc same optical bar without thc need for readjustment before use.The design and performance characteristics of commercially availablc multi-channcl ICP spectrometer systems have been reported in a number of publications (468, 677, 1026, 1034). Newland and Mostyn (1014) and Robin and co-workers (542) have similarly described their instruments.The former is a 36 MHz (2.2 kW) system equipped uith ultrasonic nebulization facilities, whilst the latter is a 40 MRz system, which also has ultrasonic nebulization, and which has been employed for both atomic emission and absorption studies. The use of a two-line pseudo-double-beam system in absorption studies with this source was found to improve results. lrzterfermce Effects.A considerable amount of data has been published conccrncd with interference effects obscrved in analytical work with ICP sources, Most of these data relate to the phenomena resulting from nebulization effects and the optical Characteristics of the spectrometcrs employed. Very little information is available from which the effect of concomitant elements on analyte vaporization, dissociation, and ionization can be assessed unambiguously. The nebulization effects observed with acid solutions and pneumatic nebulization have been examined (537).Boumans and de Boer (824) have examincd inter- ferences observed from high concentrations of metal chlorides with and without ultrasonic nebulization and dcsolvation and under compromise conditions suitable for simultancous multi-element analysis.Greater interference was obscrvcd using ultrasonic nebulization and desolvation. The use of buffers was found useful for major and minor elcinent detcrmina- tions, but not for trace determinations when thc salt content is high Robin (1066) has also discussed the possible interference effects observed with ICP sources.The observation of spectral interferenccs in work with the ICP has becn widespread in studies where either single-channel or multi-channel spectrometers have been cmploqcd. Larson et al. (724) have shown that the stray-light problem, which results in changes in thc background signal in the presence of strongly emitting concomitant elemcnts, may arise from ‘Rowland ghosts’ and scattered radiation.General recommendations were given for minimizing stray-light problems via the use of holographically ruled gratings, interference filters in front of cxit slits, double monochromators, or solar-blind photomultipliers. In another paper from the same laboratory (1090) a simultaneous multi-element wavelength- profiling system has been described for diagnosis of stray-light and spectral-line interferencc effects.In this system, for each of a series of small incremental changes in the angle of incidence of the source radiation on the spectrometer grating, integrated intensity measurc- ments are performed simultaneously for all of the analyte lines of interest. Stray-light problems in ICP spectrometer systems have also been discussed by Passel (IOSS), Allemand (1021), and Dahlquist and co-workers (1091).Boumans (812) has reported a formula which can be used to indicate when correction for spectral interference i5 either negligible or exceeds by far the additional statistical uncertainty it introduces. Ward (1093) has discussed the use of solvent-extraction techniques to minimize interference from strongly emitting matrix elements and the application of an automatic background-correction system utilizing a repetitive wavelength-scanning device.Wohlers (1089) has dctailed a study of stray-light reduction in an TCP spectrometer system; additional baffling within the spectrometer, the use of interference filters, and solar-blind photomultipliers resulted in improved tolermec to largc amounts of Ca in the determination of a wide range of elements; the application of a refractor plate device in the polychromator to provide small wavclength displacement at the exit slits and to facilitate automatic background correction wds, however, the most successful.Stelmack (1 022) has reported metal/dielectric (MDM) filters for the isolation ofPart I: Fundamerztds and Instrumentation 13 a single spectral band, with no transmission window at other wavelengths; this may bc valu- able in assisting the reduction of stray light in spectrochemical analysis with ICP sources.Dynamic background-correction systems for optical emission spectrometry have been described by Zeeman (789) and by Skogerboe and co-workers (730). The former has adopted a sequential discrete wavelength-displacement system whereby the background is measured at 0.1 nm intervals on either side of the analyte wavelength.The latter system utilizes a refractor plate mounted on a tuning fork, which crosses the light path behind the entrance slit at 100 Hz, causing a square-wave spectral shift such that the exit slit transmits, alter- nately, the analytical line plus background and the background alone.Applications. A considerable body of interesting data on applications has been published in which ICP discharges have been employed for single- and multi-element determinations of minor and trace elements in a variety of matrices. Reports have been given of the com- missioning and operation of ICP spectrometers in analytical laboratories for routine analysis where high sample throughput is necessary (72, 1377).Abercrombie and Silvester (129) have commented on the degradation of dctection limits for trace element determina- tions in real samples compared to pure aqueous solutions, and have evaluated techniques for circumventing this degradation. General descriptions of methodology and performance of ICP spectrometers applied to samples of biological, environmental, and industrial importance, such as animal and plant tissues, serum, hair, foodstuffs, lake-, river-, and effluent-water, lubricating oils, polymers, and resins have been presented (72, 81, 91, 178, 1013, 1015, 1034, 1091, 1389).Specific application of optical emission spectrometry with the ICP source to the simultaneous determination of As, Se, Ge, and Sn in liver, orchard lcavcs, and coal via a procedure in which these elements are first separated as their volatile bromides has been described (60).Newland and Mostyn (1027) have described in detail the analysis of Ni-base alloys with the ICP source. Further reference to the above applications may be found in Part IT of this volume. 2.2.2 Microwave-excited Plasmas: It is evident from the literature of 1976 that interest in the application of plasmas operated at microwave frequencies has increased considerably.Of the two types of microwave plasma, the capacitatively coupled microwave plasma (CMP) and the electrodeless microwave- induced plasma (MIP), the latter is currently attracting most attention. Skogerboe and Coleman (907) have reviewed the application of microwave plasmas to spectrochemical analysis.They discussed plasma propagation and coupling, plasma charac- teristics, analytical applications, and detection limits. The same workers (732) have described the application of an atmospheric-pressure argon MIP to the determination 01 trace elements in acid digests of food samples; the liquid samples are introduced into the plasma via a nebulizer of low volume that is equipped with a desolvation unit.Kawaguchi and co-workers (295) have reported a low-power microwave plasma source for trace elcnicnt analysis, solution samples being introduced via an ultrasonic nebulizer and desolvation unit. They found a reduction in background and enhancement of analyte spectral line intensities on addition of 1 mgml-1 of KCl to samples.The introduction of samples into a similar atmospheric-pressure plasma, using vaporisation from a tungsten filament heated by dis- charge of a condenser of high capacitance, has been described in a paper from the same laboratory (300); again, the addition of KC1 to samples is reported to improve analytical performance, and detection limits in the range 1 0 4 . 3 ng ml-1 arc quoted for Cu, Zn, As, Cd, and Pb with 2 p1 samples. Fricke, Rose, and Caruso (12) have employed tantalum strip vaporization into an atmospheric argon MIP for the determination of toxic trace elements14 Analytical Atomic Spectroscopy in foodstuffs after their separation by solvent extraction of acid digests with APDC/DDC into chloroform. The tantalum strip was programmed in an evaporation, ash, and atomize cycle similar to that employed in AAS with electrothermal atomizers.In a report from the same laboratory (558), the use of an oscillating-mirror rapid-scanning spectrometer as a detector for simultaneous multi-element analysis with the MIP i s described. In each of the above MTP systems an electrodeless discharge i s maintained in Ar at atmospheric pressure within, or at the exit of, a silica tube of internal diameter in the range 1-2.5mm and supported at 2450 MHz from a generator at powers up to 100-200 W.Electron temperature, electric field strength, and ion- and electron-density measurements have been reported at low pressures (133-1333Nm-2) for an argon MTP unseeded and seeded with Cs (1 112). The influence of flow rate and pressure on electron temperatures and concentrations and on spectral line intensities for various monatomic and polyatomic gases has been investigated for low-pressure MIP sources (1 41).Vinyl chloride has becn deter- mined in air by direct injection into a low-pressure helium MIP and measurerncnt of the Cl(I1) line at 479.45 nm (737); a detection limit of 390 ppm vinyl chloride was recorded, due to the presence of air, which quenches the Cl(I1) emission.The effect of doping gases (primarily 0,) on the microwave emissive spectrometric gas chromatographic detector has also been evaluated (119). Several papers reporting the applications of CMP sources have appeared. Govindaraju and co-workers (747) described the performance of an automated 600 W single-electrode CMP spectrometer and its application to the analysis of silicate rocks and minerals.Other workers (39) have described the determination of Mn in iron and steel with a commercial u.h.f. plasma torch of this type; linear calibration was obtained at 279.48 nm in the range 0.05-50 pg ml-1 after dissolution of samples in HCl/HNO,; substantial interference effects were observed.Nakashima and Sasaki (548) have reported the application of a similar commercial torch system to the determination of Fe, Mn, and Mg in high-purity Mo after dissolution and separation by precipitation with 8-hydroxyquinoline. Problems of inter- element effects with the CMP appear to originate from ionization effects and the failure of nebulized samples to penetrate the plasma discharge and vaporize effcctively.Larson and Fassel (595) have compared inter-element effects encountered with a microwavc CMP and a r.f. ICP; the ICP was shown to be superior in these experiments. Hanamura and Barnes (1 189) employed a source operating on N, to study Hg speciation; samples of tuna fish werc heated in a controlled temperature programme in an electric furnace and the vaporized sample was passed into the plasma.The use of a sealed microwave-excited EDL source, containing the sample and H, filler gas, for direct analysis by measurement of spectral line intensity has becn described (1239). BiI, was added as a buffer, to improve the reproducibility of excitation, and Ge as a ‘getter’, to suppress interfering band spectra.Zn, Cd, In, Pb, and T1 were determined with a detection limit of 1 pgml-1. For the determination of Te and Se, KI was used as both a buffer and getter, The technique was illustrated with practical applications, e.g. the deter- mination of Te in human blood. 2.2.3 D.c. Arc Plasmas A number of reports of the application of d.c. arc plasmas in Ar have been published during the past year.Vickers and Keirs (955) have reviewed the advantages and disadvan- tages of the application of d.c. plasmas for simultaneous multi-element trace analysis. Some applications 01 a commercially availam d.c. arc piasma-echeiie specrrometer system nave been reviewed by Skogerboe (957); interference effects in this source have been studied by Sands (1098).Specific applications of the same device to the determination of P in Cu andPart I: Fundamentals and Instrumeiitation 15 in Ni plating bath solutions have been rcported (1099). An evaluation of the application of this system as a detector for gas chromatography has been reported by Lloyd et al., (F.A.C.S.S. Conference, Philadelphia, 1976, paper 341). Svoboda and Kleinman (862) have described thcir recent studies, using a d.c.arc plasma in Ar/He as an excitation source. Engelsht and co-workers have reported a double-stream d.c. arc plasma (968) and discussed the application of this source to the determination of the light rare-earth elements in solutions and powders of Y,O, (1095). Korolev and co-workers (1665) have reported a ncw design of d.c. arc plasma which operates at 1.5 kW and 1.5 1 min-1 of Ar; detection limits achieved by photographic plate detection were in the range 0.01-5 ng ml-' for a wide range of clemcnts, using desolvated aerosol sample introduction. 2.3 GLOW DISCHARGE 'The glow-discharge source is by now well known, and offers several advantages over other forms of excitation, e.g. high precision and accuracy, and, in many cases, linear calibration.For industrial application the technique is perhaps rather less convenient to operate than arc or spark excitation, but its relative frcedom from matrix effects and the wide dynamic range of calibration could provide serious competition to both ES and X-ray fluorescence methods. The method can be applied to the analysis of solid metals, compressed metals, non-metallics, powders, and (most recently) to the analysis of liquids, following the develop- ment of porous electrodes (1247, 1457).Mavrodineanu (955) has studied the fundamental properties of the glow discharge in comparison with those of the hollow cathode. Other studies of the profile of the spectral lines cxcited by the discharge have shown that line widths vary from approximatcly 0.001 to 0.004 nm, and depend strongly upon the element, current, and concentration (1 114).Theoretical line profiles have been calculated and compared with experimental findings. There was considerable self-absorption and reversal in the lamp at higher currents and conccntrations (145). Emission lines were sufficiently narrow to allow the advantageous use of resonance detection; Butler et al.(136) designed a cathodic sputtering cell for this purpose, and reported linear calibration curves for the determination of Cu in A1 and Ag in Au from 1 pg g-1 to, respectively, 20% and 5%. A Grimm-type lamp in which the gas is evacuated from around the cathode has been modified by introducing secondary electrodes to provide a supplementary boost discharge; this rcsults in increased emission of sputtered metals (817).Layman (1111) has observed a glow dischargc at atmospheric pressure between small tungsten electrodes with a very narrow gap; initial measurements suggest that it may prove to be a practical source for atomic spectroscopy. Butler and Kruger (788) obtained an RSD of 0.0013 in the determination of Cu in brass, while that for alloying elements at the 5% level was 0.014.02; the time required for a full analysis was 1 minute.Non-conducting silicate powders were analysed by Czalcow (836) by mixing the powder with BaO and metallic Cu; synthetic standards made from spectrochemical buffer and oxides were used. Glow discharges may be used as ion-sputtering sources for the analysis of surfaces.Waitlevertch and Hurwitz (105) have described the determination of the depth profile of Mn in steel plate, and the distribution of Al, Fc, Pb, and Zn in galvanised sheet steel. The technique can be used to analyse material from layers a few nm thick. The statc of the art of hollow-cathode excitation has been reviewed in two well- documented studies (1290, 1361). one of which also deals with applications to the emissioii spectrography of gases, metals, and non-metals (1290).A demountable source has been16 Analyticdl Atomic Spectroscopy described which employs a conventional hollow-cathode discharge to produce an atomic vapour by cathodic sputtering, and a secondary discharge to act as a primary source of excitation (70). The analysis of leaded brass for Mn, Ni, and Cd has been achieved, using a discharge tube with a carbon hollow cathode fed from a stabilised d.c.supply; optimum excitation conditions have been determined (1006). A new discharge source designed for plane metal samples has been reported by Bjorkbom for use with the IDES system (92). The hollow-cathode discharge chamber is formed when the sample is pressed against a cathode plate of pure graphite with a conical aperture.The discharge is formed in the volume defined by the sample and the aperture. Advantages claimed are very low self-absorption due to the low pressure of the source, narrow spectral line profiles, high sensitivity down to sub-ppm levels, and wide-range linearity. Other references of interest - Analysis of Au and Pt using a GDL: 1007.Measurement of electron temperature in a hollow-cathode discharge: 1453. 2.4 FLAMES 2.4.1 Theoretical Studies The number of theoretical flame studies of direct relevance to analytical spectroscopy appears to be waning. This no doubt reflects the established position of flame methods and the diversification of research effort into the newer forms of atom cell. L’VOV, Katskov, Kruglikova, and Polzik (144) have discussed the current status and the problems associated with absolute analysis by flame AAS.This paper deals with what may be regarded as the ultimate expression of the flame AAS technique, and gives both research and review information on subjects such as: aerosol distribution in flames; the theory and experimental evidence for the lateral distribution of analytes in flames, and its relevance to the enhancement of sensitivity in the presence of excess matrix; the diffusion transport of vapours and its effect on analyte distribution; the dissociation equilibria of monoxides; the formation of stable Li and Sn carbides; and the evaluation of line profiles. Some of the points dealt with in this work are discussed in other papers by the same authors.For example, the hypothesis that the formation of Sn carbide (750) is responsible for thc anomal- ous behaviour of Sn in C-containing atomizers i s supported by experimental evidence showing the behaviour of Sn in a variety of C- and non-C-containing atom cells. Also, the Walsh theory of AAS has been refined and generalised (751) to include the effects of ultimate line width, asymmetry, shift, and hyperfine structure of the source line on the absorbance value.The droplet generator developed by Hieftje and Malmstadt (Anal. Chem., 1968, 40, 1860) continues to find application in elucidating analyte transport in flames. Bastiaans (1182) has described the use of a droplet generator coupled with a high-fidelity optical system in order to obtain accurate spatial resolution of light emitted, absorbed, or fluoresced from atom reservoirs. The spatial information and time-dependent distribution of the analyte, obtained from individual droplets, are used to evaluate the factors governing atomization efficiency.Sacks and Joshi (1115) have used a modified droplet generator and scanning optical system to study analyte processes in a N20/C,H, flame. The system was applied to studying the transport processes of various compounds of Mn and Ca.A mathematical model of atom production based on a stochastic approach in which the fate of individual droplets is treated as a stepwise process followed by statistical summing to calculate the overall spatial distribution has been proposed by Li (941).The various optical methods for spatial temperature-mapping of flames, e.g. shadowgraph, Schlieren, and the Toepler colour-stripe filter, have been discussed by Pungor and co-workers (101 1).Part I : Fundamentals and Instrumentation 17 An interesting paper which discusses the amplitude and phase relationships between multiplicative noise signals originating in N,O /C,H, flames has been presented by Mossotti (553).The frequency-dependent coherence map of the flame is discussed in relation to spatial co-ordinates, flame species, and flame stoicheiometry. The application of coherence- mapping cross-correlation techniques to effect the measurement of physicochemical quanti- ties in local regions of the flame, without recourse to Abel inversion, i s described.Ingle (739) has reported experimental data on the precision of Cu determination by AAS which supports previously derived theoretical cquations (Anal. Chern., 1974, 46, 2161). In particu- lar, it was shown that the precision at low absorbances i s limited by variations in flame transmission, whereas at higher absorbances it is limited by variations in the absorbance of the analyte.Other references of interest - Effect of particle vaporization and atomic diffusion on analyte distribution in flames: 1184. Laser diagnostics in turbulent combustion flows: 71 1. 2.4.2 General Studies Although virtually every combination of combustible gases has been proposed for analytical spectrometry, useful activity continues in this area, and recent work reports studies on the He/O,/C,H, and the Ar/O,/C,FI, flames.Replacement of the N, in the familiar air/C,H, flame with monatomic gases having lower specific heats and higher thermal conductivities leads to higher flame temperatures, better heat-transfer properties, and a less quenching environment. Winefordner and Johnson (581) reported improved AFS results with the Ar/O,/C,H, flame, showing that variation of the Ar/O, ratio enables a wide range of flame conditions to be realised, with resulting gains in analytical performance.Hieftje and Satur- day (1179) found similar gains for the He/O,/C,H, flame, and reported a four-fold improvc- mcnt in atomization efficiency over the conventional air/C,H, flame. Fuwa and co-workers (140, 442) have studied the formation of stable In monohalides in the presence of halogeno-acids in both air/C,H, and N,O/C,H, flames.Thc moleculir spectra of the individual compounds are shown, and their use both in the determination of In and in the elucidation of interferences is discussed. In a separate paper (1514), the same authors have reported on the photodissociation spectra of Na halides in the air/C,H, flame.A new interference mechanism, analyte reduction to the metal or carbide in the solid phase, has been proposed by Rubcska (861, 868, 924). Using a Meker burner, from which lateral diffusion effects are absent, it was demonstrated that refractory oxides such as A1,0,, ReO, and TiO, can enhance the sensitivity of those elements such as Mo and V which are known to form stable oxides and carbides in reducing flames.It is suggested that the solid Mo and V compete successfully with the concomitant elements for oxygen, and are therefore vaporized in the MO form, which is more volatile than the corresponding metal or carbide form. The mechanism can both enhance and inhibit atom formation, and it competes with thermal vaporization in determining local free atom concentrations.Cresser and Hargitt (6) have attributed the pH-dependent interference observed in the AAS determination of Cr to the state of the Cr0,-/HCrO,- equilibrium. The effect was found to be greatest in thc air/H, flame, less severe in air/C,H,, and not observable in the N,O/C,H, flame. The effect of the oxidation state of Cr was also studied by Britske and Savel’eva (1667).Interferences caused by ligands in the AAS determination of elements in the first transition series have been reported by Fuwa and co-workers (292). As might be expected, interferences occurred in the low-temperature regions of the flame, and were related to the strength of the ligand; CN, for example, caused severe signal depression. An enhancement was found when Cr was18 Analytical Atomic Spectroscopy dctermincd as trisbipyridylCr(II1) in a fuel-rich air /C‘,H, flame.Marszn and Hall (102) have described a variation on the ‘standardised matrix’ procedure in which all saniples are prepared in 2% VjV HC1 containing 4000pgml-1 CsCl in order to minimise analytical trrors. Atom formation and the interferences in both flames and carbon furnaces have been rebiewcd by Ottaway (356).Interference mechanisms and thcir treatment have also bccii discussed by Taylor (474). Hieftje (1073) has described various techniques for interfacing flame spectrometers to computers. Perhaps the most interesting device discussed was a digital gas-flow control consisting of 8 parallel flow pipes, each successive pipe handling twice the flow of thr previous onc.With stop valves on each pipe, the binary weighting allows flow resolution of 1 part in 256 (8 bits) to be achieved. Other references of interest - AFS analysis in air/C,H, flames: 627. Atomization of Mo in air/C,H, flames: 738. tntcrference efiects of ammonium halides in air/C,H, flames: 247. Interference of InCl band spectra on thc determination of Au: 1300. Interferences in flame AAS at high analyte concentrations: 35 Interferences in the dctermination of Mo: 173.Organic solvent cnhanccincnts in premixed O,/H, flames: 319 Production of organic liquid-fuel flamcs for AAS: 335. Tcchniques for structure analysis in low-pressure flames: 899. 2.4.3 Devicer for Sample Introduction Pneumatic nebulization continues to be one of the less well understood aspects of flame spectrometry.There have been no reports of original designs, although modifications to existing typcs are still being reported. Heinemann (1328) has made a comprehensive theoretical study of the factors affecting nebulizer performance. The effects of geometry, capillary diamcter, oscillations in the pressure of support gas, viscosity of the solutions, and variation of temperature were considered.Both Fuller (485) and Miller and Edwards (486) have employed branched-uptake capillary nebulizers for the simultaneous addition of standards, buffers, and releasing agents to the analyte solution. Hieftje and Savage (1 137) have used the decrease in surface tension promoted by high electric fields to produce decrease in droplet size from a pneumatic nebulizer-Beckman burner combination.No analytical results were reported, because the high voltage required (6 kV) could not be sustained within the flame, The effect of nebulizer parametcrs on the enhancement of sensitivity by organic solvents has been studied by Pungor and co-workers (820). Aspiration rates and efficiency of nebulization were measured and correlated with absorbance vahes It was found that for elements behaving normally in the flame, e.g. Pb, Pc, and Ag, the increase in sensitivity was wholly attributable to nebulization paramcters.Skogcrboe and Freeland (1 175) have discussed the importance of aerosol transport phenomena in relation to the efficient design of transport systems for analytical spectrometry. Little has appeared in the literature on this subject, and it is to be hoped that it will attract more attention i n the future.Thermal vaporization or samples from furnaces prior to introduction into a flame ha:$ been employed by several authors. A graphite tube furnace coupled to an air/H, name has been iised for the dctermination of P as POI€ (107, 477). The N, flow through the lurnace was found to be the most critical experimental parameter.Strong interferences were encountered from Ca and Fe(1II) ion, although quantitative results were obtained from a standard matrix of 30 gg ml-’ Ca and 300 pg ml-1 EDTA. Graphite furnaces have alsoPart I: Fundamentals and Instrumentation 19 becn coupled to conventional flames, i.e. air/C,H, and N,O/C,H, (125, 398, 764), with r-salting decreases in the critical nature of the drying, ashing, and atomization steps and improved performance with respect to interferences.A tantalum-filament vaporizer (741) has also becn used in this mode, and applied to the determination of Cu and Fe in serum samples. A carbon-fibre atomizer has been described by Prudnikov (1527). The thread was used for the direct atomization of metal chlorides for AAS and as a pre-atomizer for FES.The maximum temperature employed was 1500 "C, which is rather low for a general- purpose atoniimr. Stephens and co-workers (1288) have now published a paper on their silica-tube atom-trap dcvicc (see ARAAS, Vol. 5, ref. 907). The papcr describes sensitivities, trapping efficiencies, and atomization kinetics of the system A numb.:r o€ papers have been presented describing the use of flame AAS and AES 3s detectors for high-speed liquid chromatography (248, 560, 569, 584, 928, 933, 11 88, 1367).Manahnn and co-workers (1 188) have reviewed the current position, and point out that, whilst the technique yields high specificity compared with conventional methods, the sensitivity is not sufficient €or many applications.They suggest that the recent developments in plasma spectrometry might result in improved sensitivity if the problems associated with using organic solvents can be overcome. Detectors for gas chromatography have been described, Sb and Ca being determined as their volatile chlorides (246). This papcr also describes a novel atomizer consisting of a graphite rod which can be heated to 2000 " C and which is contained in a silica tube through which the column effluent is passed; the ?ample and the carrier gas are introduced at the centre of the rod.The fluorinated ketones of Cr have been a popular subject for GC/AAS (560, 928); tetra-alkyl-Pb compounds have also been determined (1056). Pungor et d. (749) used a laser microprobc for sampling, and reported results for Fe, Na, Cd.Cu, and Zn in various alloys. Optimisation of conditions for emission measurements during pulscd microprobe vaporization into a flame has been studied by Prudnikov and Shapkina (371). An inleresting approach to the sampling of metals for AAS has been reported by Ghiglione et (11. (315). These authors used a water-immersed spark, operating at 3-6 A for cn. 90 s. to produce a colloidal suspension of the metal, which can then be nebulized in the normal way. The colloids appear to be stable and reproducible, and it i s found that the weight dispersed is proportional to the square of the current multiplied by the time. Results for Al, Mg, Mn, and Cu alloys showed good agreement (&5%) with solution work. Rigois and Levy (5) have reported preliminary attempts to determine Si by the formation of SiCI,, generated when 200-700 p g of Si compounds are pyrolysed in CI,/N, at 1100 "C. The vapour is transferred to the flame by a flowing N, stream.Relcher, Townshend, e i ul. have continued their studies on MECA (20, 47, 93, 94, 158, 355, 404, 531, 1124, 1125, 1353) and candolumincscence (160, 183, 708, 749, 1122, 1123), to extend both the range and applicability of the techniques.For example, the hydride- gcneration method (93) has been employed as a means of introducing As, Sn, and Sb into the cavity. MECA has recently been applied to the determination of sulphate in high-purity waters (912). Other r-fercnces of interest - Iiydride generation for AFS with continuum sources: 88. Laminar-flow burner with separate aerosol injection: 81 8.Organic / aqueous stream-combination device: 576. Microsampling cup techniques: 11 3, 134, 1304, 1306, 1658. Pulsed ultrasonic nebulization for analysis of small volumes: 819. Pyrolytic graphite coating of carbon rods, using a carbon-rich flame: 974.20 Analytical Atomic Spectroscopy 2.5 ELECTROTHERMAL ATOMIZERS The increasing maturity of electrothermal atomization techniques has been reflzcted by the scarcity of reports of novcl systems and the emphasis on more basic studies of atomizdion and interference phenomena.There has, however, been one report of a graphite micro- furnace (2X 1SX 10 mm) for the ultramicro-determination of Ca in plasma filtrates (1643) Using 7 nl samples, concentrations of Ca of 50 pg ml-1 were determined under p x l i a l vacuum and with vaporization at 3000 K.L’vov (1141) has discussed problems involved in absolute methods of AA analysis using electrothermal atomization, and ‘rules of thumb’ for method development, based 011 theoretical and practical considerations, have been proposed (1 132). Papers have continued to appear concerning the comparative advantages of peak-height versus peak-area measurc- nient (1 140, 1337).Reviews of electrothermal atomization have generally concentrated upon discussion of practical aspects (375, 385, 872, 986, 1199, 1200, 1242, 1618). A round-table discussion of background measurement and comparison with interferences in other atomic spectroscopic techniques has also been reported (1 320). It was agreed in this discussion that the most satisfactory approach was to minimise the background signal, so that the need for background correction was reduced.There has been an increase in the number of papers dealing with more fundamental aspects of electrothermal atomization and interference mechanisms, which suggests a useful increase in understanding of both these aspects. Fuller has further demonstrated the usefulness of kinetic theories of graphite furnace atomization, and applied these to identify- ing the nature of interferences and to minimizing such effects (152, 720, 1279).The release of Pb atoms from heated graphite into a H, atmosphere has been studied by Torsi and Tcssari (746), and the results compared favorably with the authors’ own kinetic model (ARAAS, 1974, 4, ref. 1326). Frech and Cedergren (51) have considered interferences in the dctermination of Pb in steel, using an interesting theoretical approach; all reasonable products from the reactions of C , H, 0, N, S, Fe, Pb, C1, and Ar were considered by a computer program based on high-temperature equilibrium calculations. The results indicated that a sufficiently high H, concentration in the furnace at 900K should remove Cl,, and thus avoid the formation of volatile Pb halides.This was confirmed by a later practical investigation (52). It was shown that, in a graphite tube, enough HL is generated from the water to remove all the Cl,, but for a mini-furnace, or a rod, HL must bc addzd. Sturgeon and Chakrabarti, using an approach combining thermodynamics and kinetics, postulated three major atomization mechanisms: thermal dissociation of the oxide or halide, or carbon reduction (930, 1142, 1338, 1339, 1347).The transient signal observed in graphite furnace atomization is the resultant of the competing processes of vaporization and removal. Attention in the past h u usually been focusscd upon the vaporization or atomization step, but now de Galan (1151) has studied the diffusion and convection eRects responsible for the removal of atoms.The results showed that the rate of removal is comparable with the rate of entry, and this, combined with the inadequate response time of commercial detection systems, results i n less than optimum sensitivity. Slow instrument response leads to non-linear calibration curves and increases the problems of background absorption.Other workers (1064) have attempted a comprehensive study of Pb atomization by means of microscopic observation of deposits, thermogravimetry, and oscilloscopic recording of signals. Hircq (1 589) has used an ion microprobe and X-ray techniques to investigate the distribution and thermal trans- formation of Cu and UO,(NO,), deposits on graphite filaments.The size of the deposit appears to vary with sample volume rather than with concentration. UO,(NO,), was shown to decompose through oxide phases to a carbide at 2400 “C; the carbide was not removedPart I : Fundameritals arid Iristrutrientatiori 21 from the filament but was converted into the nitratc by the action of HNO.,, thus indicating a mechanism for the build-up of U on the filament and for the poor rcproducibility of signals in uranium matrices.Charge-transfer spectra of alkali-metal halides have been observed in different types of graphite atomizcrs; background matrix absorption was attributed to these species rather than to scattering (53). The thcrmal decomposition of transition-metal salts has been studied spectroscopically, and absorption spectra of metal halides have been observed (544). NO and SO, bands were also observcd, which suggests that nitrates and sulphntcs arc convcrtcd into oxides during the ashing stage.A specially designed Mo tube atomizer with two small indentations 3 mm apart has also been used to dinerentiate bctwecn vapour-phasc and other interferences (706).The analyte (Mg) is placcd in one indentation and the intcrferent in the other: Cr induced a broadened peak both when mixed with Mg and when separated, suggesting a true vapour-phase interference. Thc two-absorption-line intensity-ratio method has been used by several workers to measure clectronic tcmperaturcs in graphite atomizers. Adsms et al. (538) observed the black-body emission at 600 nm to evaluate heating rates in a commercial graphite furnace, and confirmed the results with an optical pyromcter.The tcmperaturc-time profiles were then compared with the clectronic excitation temperature-timc profiles obtained by ohserv- ing the Ga (403.29/417.20 nm) and In (410.18/451.13 nm) lincs. The experimental cuncs wcrc similar to those dcrived by Cresser et a/.(see ARAAS, 1974, 4, ref. 720) from a simple theoretical treatment. The atomic vapour tcmperatures were 15% lower than the wall tem- peraturcs. Sturgeon et al. (1 285) have measured graphite wall temperature-time curves, using an optical pyrometer with a storage oscilloscope. The results obtained were compared with vapour tcmperaturcs mcasured using the two-line method. By using Fe, Sn.In, and Ga, a range of temperatures was covcred in a commercial furnace and a mini-furnacc. The maximum temperature attained by the atoms was found to be 600-1400 K less than that of the walls, and the maximum tcinperaturc varied with the volatility of the elements. An H2 flame appears to increase atomic tcmperaturcs by approximately 100 K at 1900 I< when using the mini-furnace.Using a two-channel double monochromator and thc melting point of Au as a calibration standard, Tsujino et al. (543) measured temperatures up t o 3200 "C in a graphite tube, using the two-wapclength method. The results obtained were confirmed, using the melting points of Pb. Al, Mo, and Ti. The importance of background correction in elcctrothermal atomization is by now well known, and a ncw system has been presented (1217).The accuracy, precision, and limitations of background-correction systems have been described (1 129, I 1 39), and Zander and O'Haver have discussed the methods available and their application (1376). A brief tutorial repicur has been made of the Zeeman effect (l226), explaining its use for back- ground correction by splitting the beam into its components, using a magnetic field. Koizurni et al.(598) have reported the use of R source split into Zeeman components by a magnetic ficld with signal separation by means of a rotating linear polarker; this apparatus was suc- ecssfully applied to thc determination of Pb, Cd, and Zn in blood, livcr, and urine samples. The same authors havc described thc application of the Zeeman effect for background correction to a wide variety of flame and flamelcss determinations (816).This paper contains a useful account of the theory and of the apparatus used. Both EDL sources and a parallel-plate electrode source, cxcitcd by both d c. and h.f., wcrc used. In an interesting variation, Dawson et a/. (28) applied a magnetic ficld to the vapour produced from a graphite rod atomizer.When light from an HCL was passed through thc atoms, only the component parallcl to the magnetic field was absorbed by thcm, and hence, using a rotating pokarizer to scparalc the two polarizations, a background-correction s! 5tem was22 Analytical Atomic. Spcrftoscopy achieved. The sqstem was successfully tested for Ag, Cd, Cr, Cu, Mg, Mn, and Pb.and was applied to the determination of Cu in blood. The use of a dual-chamber furnace in which the atomic vapour diffuses from an outer tube into an inner carbon tube was used (742) for Pb determinations in bovine liver, in conjunction with a source split into Zeeman com- ponents. This technique was found to separate the analytc from matrix vapours, and hencc to decrease background absorption and improve reproducibility.There has been an increase in the number of papers describing automated procedures, particularly automatic sample introduction, for the improvement of the precision and spccd of analysis (1325, 1554). The performance of a commercial auto-sampler for furnace work has been extensively reported (502, 705, 803, 1134, 1144), a relay timer for automatic control of some parameters has been described (327), and a graphite furnace system i n which a microprocessor-controlled spectrometer was wed to optiniise measurement of p c ~ h height (133).For miniature graphite furnaces, a solid-state power controller (312) and a power supply in which the analytical cycle may be interrupted (931) have been described. The advantages of feedback control of furnace temperaturc, using a resistance thermomctcr, have been described (1198), and a very inexpcnsivc temperature monitor based on a photo- voltaic cell has been reported (305).A miniature furnace with an axial slit has been used as a rcsonancc detector, the sample beins atomized from a West filament. Detection limits in the ng range arc reported, but the stability of the signal is, at present, poor (361).The increasing interest in metal speciation studies has been reflected in several papers i n which clectrothermal atomization has been used for this purpose. Ultra-filtration separa- tion followed by furnace atomization (1 191) and differential thermal vaporization (1 190) have been employed. Scvcral workers have utilized gas chromatographic separations followed by detzction using heated T-shaped tube atomizers made of either carbon (227, 1185) or silica (1187, 1259).Other rcfercnces of intcrcst - Air ashing: 186. Microfiche on ‘Flamelcss AA’: 1583 Novel apparatur: 622. 2.5.1 Graphite Hod Device5 An autoinated two-dimensional positioner and saniplcr, controlled by a minicomputer. has bccn employed to position the atomic vapour cell for studies of the spatial distribution of spxies above a filament atomizer (130, 940, 1152).A semi-automatic s3mpler has been described which enables strict control of the rod atmosphere during sampling (597). Pyrolyiic graphite coating of a ‘rod-in-flame’ type atomizer, using a ben7cnc-eariched air /C,H, flame, has been reportcd (974, and significant advantages have been noted. 2.5.2 Graphite Miniature Fiirnaces A new graphite miniature furnace In which the fixed resistance of the circuit IS matched to t h ~ variable resistance of the graphite components, thus allowing grcater control ovtr the tenipcrature profile, has been reported (61, 470, 695), and applied to the analysis of liquid mmplcs that have a high solid content (154).An interesting rewlt of thtsc studies i s that optimum heating rates were found to be 300 to 500 ‘C s-1, c g. optimum signals werc obtained for vanadium by heating, from a ‘hold’ tempcr.iturc: of 2400 “C, a t a ride 01 400 O C s-1. to a final temperature of 3000 OC.Pat1 I : Fundurneritals arid Tnstrurncntatinrz 13 2.5.3 Graphite Tuhc Furnace3 rube furnacec continue t o be apparently the most widely used electrothermal atoinizers Massn~ann has had an opportunity to suwey the prcsent possibilities and limitations (1 47, 716), as has Slavin, emphasising, in this case, the latest advances (1032).Kahn (1538) has compared the use of flames and graphite atomizers. The number of reports of carbon furnace atomic emission has incrcased, but its use has not yet bccome popular.Ottaway et al. have discussed their work (148, 1127) and extended it t o S more elements, viz. Sr, Rb, Cs, Yb, Eu, Dy, Ho, and Er (717). Poor detection limits for volatile elements by crnission, attributed to the rapid diffusion of atoms from the centrc to the cooler ends of the furnace, wcrc considerably improved by reducing the tube thick- ness at the ends (513).The technique has bcen applied t o thc determination of minor constituents OC steels (1143) and Li in Cu (460). Matousek and Srnythe have also studied Li emission in furnace atomization (47 1). The application of wavelength modulation to remove the intense furnace continuum emission has been demonstrated (1 130); the samc paper also suggested ways of overcoming other limitations of the technique in the analy5is of a variety of en1 ironmental matrices. There has been conniderable interest in the study of matrix interferences in graphite tube AA determinations.The suppression of Pb absorption signals by alhaline-earth m-tals was climinatcd in a limited range of water samples by adding ascorbic acid t o the solution (23). It was suggested that this gives a.‘molecular mixture’ of carbon a r d sample during atomization; ascorbic acid was also shown to enhance Cu and Ga signals. Andersson (483) h2s used La to overcome SO, interference in Pb determinations. Ottaway (1 162, 1210) and Matousek (1 161) have stressLd the importance of avoiding Cl media; thc latter presented evidence indicating vapour-phasc interference The effect of the condition of the graphite tube on the extent of the interference by MgCl, or7 the detcrmination of Pb tias been reported (986) The complexity of some interfercnce effects for Cu and Mn has been stressed (573), c.g 50 pg ml-1 of MgCl, gives an cnhancernent of Mn AA of 60%.whereas 100Opgml-1 gives a depression of 80%. Na, in the presence of SO,, has been reported as the most serious intcrfercncc in the determination of As (XIS), and Mg is reported as enhanc- ing the sensitivity and controlling the interference.In a study of thc effects of acids on d-terminations of 71, Fuller (42) showed that HNO, and H,SO, had slight cffects. but those of HCl and HCIO, wcrc qubstantial. The addition of H,SO, minimiscd these latter intei- ferences, and also that of NaCl.Matrix interferences in Si, Ge. Sn, and Pb determinations have bccn rcduccd by the use of a tantalum or tantalum carbidc lining (443). Ottaway and Shaw (359) studied the absorption and emission of Ba atoms and ions, and ccrncludxl that the lower tempcraturz of the furnace compared with that of the N,O/C*EI., flame mean‘, that ionization is not likcly to be a serious problcm in furnace absorption or emission studies.Metal chloride band spectra (1 128) and other absorpton spectra (149) h a w been obscrvcd to cause background light losses Various solid-sampling systems hdve bccn describcd. Techniques habe included the direct introduction into the furnace of alloy chips (56) and micro-boats (131) in which the sample may be charred over small flame prior t o insertion (132).Samples rnixcd directly with graphite powder (1131) or metals deposited electrolytically from solution on to .I graphite rod. which was thcn ground to a fine powder (529), have also becn used Other worhcrs have directly atomized metals from graphite electrodes in a graphitc tube furnwc after controlled-potential electrolytic d~position (1490). Scmi-conductor films Iiave bcen ‘inalysed 1:lycr by layer, using ionic ctching; the sputtered produck w r c collr:cted 011 a graphite cylinder, which was placed in a furnace (256).24 Analytical Atoniic Spcctroscop) Leisz et al.(135) used electron microscopy to study the effects of miscible organic solvents on the distribution of samples in furnace tubes. It was speculated that the lowcred surface tension in the presence of organic solvents might lead to distribution of the sample over a wider surface area and to deeper penetration into the graphite.A method for the determination of Ga following extraction of Ga cupferrate into butanol or chloroform has bccn described (540); chloroform gave poor sensitivity compared to butanol. The dcterrnina- tion of Cd, Cu, and Pb in river- or sea-water, following an APDC/MIBM solvent extraction. has been reported (228).Excellent sensitivity was obtained. Once again, the advantages of pyrolytic graphite coatings have been strcssed (174, 343, 1155, 1196, 1349). Woodriff et 01. have described recent improvements to, and uses of. their conslant-teinperaturc furnace (729, 1153, 1163). Welz et al. have reported a ripid pyromctric system for graphite furnace temperature measurement and control (1 164), and other modifi- cations for this type of furnacc have also been proposed (494, 883).Capacitative heating of a furnacc has been investigated (1 133). Under conditions which cause dielectric breakdown of the surrounding gas, the furnace functions both as an atom cell and an excitation source.At lowcr electric field strengths the furnacc acts as an atom cell only. The curvature of calibration graphs in electrothermal work has for long led to con- troversy. Conditions for obtaining linear Pb calibration curvcs have been optimiscd (524). and thc use of (Zo/Z-l) instead of absorbancz its the ordinate has been advocatcd. An alternative approach, for Cd, has been to pressurise the furnace to achieve a 6-fold incrcesc in linearity at 6 atmospheres of Ar (1165).An indirect calibration method, using two furnaces at differing temperatures in the optical path, has been described (672) and used to detcrmine the slope of the calibration curve for Cu at 324.8 nm (in Ar at 2 atmosphcres). A small graphite tube atomizer has been described for the dcterrnination of S in aqueous samples at the 180.7, 182.0, and 182.6 nm resonance lines (1289). The detection limit observed was 2 ng; the characteristic concentration (for 1% absorption) at 180.7 nm was 0.42 ng.Sulphates, thiocyanatc, and thiourea were studied, and in the casc of sulphatcs thc production of atomic S was attributed to the decomposition of SO,. At one time it was expected that electrothermal atomizers would greatly increase thc application of AFS, but this has not yet proved to be the case.A new graphite furnace for non-dispersive AFS has been described (44), and its application to Cd, Zn, and Pb appears promising. 2.5.4 Metal Filaments: and Furnaces Ohta et al. (533) have studied the atomization of various Pb salts in a molybdenum tube atomizer.The dcgrec and temperature of atomization were shown to depend on the salt used, organic salts being the most favourable and the iodide thc least. l h e same worker? have compared metal strip and tube atomizcrs for Ga determinations (549); thc tube gave better sensitivity because of the greater residence time. An improved tungsten filament atomizer for absorption work has bccn described (580).togcthcr with the appropriate measuring circuitry. Detection limits in the rangc 10-12 g for Ag, Ca, Cu, and Mg were obtained using pure solutions. A copper wire atomizer for Hg determinations (Hg is amalgamated at the copper surface) has also been rcportcd (274); see Brandcnburgcr, H., and Bader, N., At.Ahsorpt.Ncwsl., 1967, 6 , 101 2.6 MISCELLANEOUS ATOMIZERS Thc use of hydridc generation for the dctcrmination of those Indtals which form cocalent hydrides with sodium borohydride is now an established technique.Early workers usxiPart I: Fundamentals and Instrumentation 25 flames to dccompose the hydrides, and modifications to such systems are still being proposed (1 121, 572), but, increasingly, heated tubes arc being used to accomplish this decomposition (481).An automated system based on a Technicon ‘Auto-Analyzer’ and a heated quartz cell has been reportcd (2, 521). Interferences in the hydride-generation technique are now being more frequently investigated. As(V) is convcrtcd into As(II1) before arsine evolution; this reaction is depcndent oil pH, and this equilibrium can be used to distinguish between the species (507).Two useful reports consider practical ways of optimizing arsine generation (507, 592). Inorganic interferences arc to some extent dependent on the order of addition of reagents, and polyethylene and polystyrcne sample cups have been shown to depress the analytical signal (740). In the analysis of metal sulphides for Sb, As, Bi, Se, and Te, it was found that Cu and Ni but, not many other metals, cause severe suppression; they were r-moved beforc analysis (801).A detailed study has been made of interferenc-s in the dctcrmination of scbcn hydride-forming metals in steels; the presence of Fe appeared to be bmeficial (526). EDTA is reported to overcome interfcrences in the determination of Bi in Ni-base alloys (51 1). The cold-vapour technique for Hg determination is very widely used. Koirtyohann and Khabii hakc shown that the reduced Hg follows a defined partition function when equili- brated between a gaseous and aqueous phase; this is affected by H,SO, content and tempera- ture (574). Premature reduction losses and interferences by Pt, Au, and Ag, where the reduced metal or precipitated chloride impeded proper partitioning of the Hg vapour, werc also observed. Rooney has used sodium borohydride as the reducing agent (512), and reports intcrfcrences from those elements which amalgamate with Hg, viz. Au, Pt, Pd, Rh, and Ru. An interesting detector for Hg vapour has been reported (1404) in which the source intensity IS automatically adjusted according to the Hg concentration; very good long-term stability is thus obtained. A copper amalgamator (539) has been used to provide a rapid separation of Hg in environmental samples. Wet and dry oxidation procedures can be avoided by using h.f. induction heating of samples, followed by absorption in acid permanganate and reduction (541). Cresscr ha5 dcscribed (1351) a cold-vapour method for N, determination, using the molecular absorption of NH, at 201 nm in a specially designed cell in the path of an AA spectrometer. The induction furnace of Headridge appears to offer great sensitivity for the analysis of volatile components in metals (156), but calibration samples are difficult to obtain. The same consid:rations must apply to hollow-cathode discharge atom cells; work has been rcportcd in this field, using solutions evaporated in the cathode and samples machined into cathodes (97, 559). A device which allows the simultaneous AA and AE analysis of a sample ashed in a dielectric cuvette (c.g. of quartz) has been patented (1621). The thermally atomimd sample i s excited by an h.f. inert-gas discharge. A 12 kV, 50 Hz spark has been used to sample a flat surface; the particles are then swept into conventional flamcs (25), and determinations are made by AAS or AFS. As the flame caused incomplete vaporization, an ICP was also used, with improved sensitivity, but less freedom from interference. A high-voltagc d.c. arc between two tungsten micro- clectrodes, one of which holds a microlitre sample, (the ‘microarc’) has been employed as an atom cell for AAS and AES (115). Lasers have, in the past, been proposed as suitable atomization systems, and there are again reports of their use. A 400 ns flash-lamp-pumped dye laser microprobe has been used to atomize metal alloy samples; the atoms were then excited to saturated fluorescence by a similar laser (952). Laser absorption and fluorescence techniques have also been used to optirnise conditions for laser microprobe dye-laser-saturated AFS (557). Laser atomization26 Analytical Atontic Spectroscopy has also been applied to the AAS analysis of 9 elements in biological materials (637) and Cd in surface areas of 0.008mm2 on V, Pb, and U oxides (1326, 1534). The vapours pro- duced by a laser microprobe have been transferred, using the venturi effect, to a conven- tional nebulizer and flame for AAS analysis (749). A patent has been granted for a novel flameless atomizer for the AA analysis of water samples (1616). Trace elements are atomized from samples of about 100 pl by focussed i.r. radiation. Other references of interest - Hg determination: 349, 476, 525, 1049, 1352. Hydride generation: Patent: 1644.

ISSN:0306-1353    

DOI:10.1039/AA9760600006

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

3 optics 3.1 BEAM MANIPULATION The continuing revival of interest in emission methods has sustained the interest in wave- length-modulation techniques. Epstein et d. (106, 316) have minimised the problems of black-body radiation and scatter in carbon furnace atomic emission spectrometry by repeti- tive wavelcngth scanning using a vibrating mirror, and for Al, Na, and Ba they obtained improved detection limits, comparable with those obtained by absorption methods. The technique has also been applied by other workers to multi-element microwave-induced plasma emission spectrometry (558).Koirtyohann et al. (1 138) have demonstrated the advantages of using a stepped, square-wave modulation rather than a sine wave for this type of background correction, as the longer dwell-time at the wavelength of interest leads to a much larger useful signal, They also note that the quartz vibrating plate must be of high quality and the spectrometer refocussed if high resolution is to be maintained.A simple system based on scanning the monochromator exit slit, using a telegraph relay, has been proposed (307). Another inexpensive system based on oscillating interference filters, where the bandpass maximum changes with the angle of incidence, giving rise to a modul- ated analyte signal at twice the frequency of oscillation, has been demonstrated (589) for the flame emission determination of Ca (detection limit 10ngml-I).Visser et al. (1282) found that stepped rotation of a quartz refractor plate gave accurate background correction for atomic emission measurements, even when the sample matrix changes.Wavelength modulation has also been applied to AAS, using a continuum source with a high-resolution echelle spectrometer (98, 596); most source noise was suppressed, and results comparable with those obtained using line sources are reported. The continuing interest in the use of the Zeeman effect for background correction in electrothermal atomization has already been noted (see Section 2.5).Koizumi et al. (1172) have reported the use of the Zeeman effect for background correction applied to both the light source and to the sample. The interest in transform techniqucs has not been maintained. The application of a Michelson interferometer system to Fourier transform atomic spectrochemical measurements has been reported (948). The same authors have discussed Fourier domain interpolation as a solution to the apparent oversampling necessary with discrete-array sensors (926).Swift et al. (712) have described the use of binary optical encoding masks for multiplexing in multi- element analysis as an alternative to Hadamard spectrometry. We still await the comprehensive application of optical fibres to spectroscopy, but meanwhile some of the problems have been discussed (1235).Factors governing the illumina- tion of a spectrometer system have been considered from a physical standpoint (815, 821). An optical system to be used in conjunction with a spectrograph to produce stigmatic images for the study of spatially resolved events in a spark gap has been described (1136).For the study of time-resolved spectroscopy, a film holder-exit slit movement for a fixed-grating V.U.V. spectrograph has been designed (1228). A phosphor and a flexible light-guide system is used to couple the output radiation to an external photomultiplier. Other references of interest - Background correction: 1300. Collimating lens for furnace: 324. Wavelength modulation: 1229. 2728 Analytical Atomic Spectroscopy 3.2 WAVELENGTH SELECTION 3.2.1 Dispersive Systems The history of the development of ruled gratings and wavelength tables has been reviewed by American authors (78), who, appropriately, in their bicentennial year, trace the story back to a Philadelphian, Rittenhouse, who constructed gratings three decades before Fraun- hofer.The interest in gratings still continues, and the construction and evaluation of echelle gratings has been described (1194). Keliher and Wohlers (583) have reviewed the uses of these high-resolution gratings in analytical spectrometry; they would seem to offer advan- tages in ES, AAS with a continuum source, and multi-element analysis. Advantage has been taken of the increasing feasibility of computer control of mono- chromators, although this application of computers is not as popular as in thc field of data acquisition. The greatest interest is in emission and fluorescence studies.A micro- computer-controlled dual-wavelength spectrometer has been used for automatic background or scatter correction in atomic emission and atomic fluorescence, respectively (1075).This correction is made at a wavelength close to the analytical line; a computer-controlled refractor-plate background corrector has been described elsewhere (577). In the latter case the computer also controls wavelength selection and slit width. A goniometer control unit. for automatic wavelength changes, and a minicomputer, for data processing and interference correction, have been reported (360) and proposed for use with an ICP.Other reference of interest - Bearing for grating mounts: 911. 3.2.2 Interferometers Fabry-Perot interferometers have been used in the study of spectral line profiles in rcccnt years. Such an instrument has now been used by Behm and Dobele (601) to examine the emission from a dye laser that produces pulses 2ns wide.. Spectral line shifts arising from the Doppler effect, and smaller than the line width, have been measured at the samc Institute (208).The flow velocity of an Ar plasma jet was measurcd, using these shifts on thc 415.8 nm line, and found to be 300-600 m s-l. The use of a scanning Fabry-Perot interferometer for rapid sequential multi-element analysis by AES has now been reported (735).An air/H, flame was used, but only moderate detection limits were obtained; the main advantage of the system appears to be the ability to perform analysis in the ms time-scale. A Michelson interferometer in which the stationary mirror has been rcplaced with a rotating grating has been used as a selectively modulated interferometric dispersive spectrometer (SEMIDS) for atomic spectrometry (564, 910).The oscillating mirror modulates only those wavelengths diffracted perpendicular to the mirror- detector optical axis. High light throughput with good resolution is thus obtained. SEMIDS has initially been applied to flame emission measurements of Sr and CU. 3.2.3 Non-dispersive Systems The slow development of non-dispersive fluorescence systems has been disappointing. Larkins (1 156) has continued his work on non-dispersive multi-element flame fluorescence and has described an instrument comprising a separated flame, line sources, lens, and solar- blind photomultipliers, with provision for scatter correction. An instrument for the simul- taneous non-dispersive determination of Na and K or Li by FES and Ca and Mg by AAS has also been reported (1 324). The instrument incorporated a multi-element HCL and interference filters; a minicomputer was used to control data handling and sample presenta- tion.

ISSN:0306-1353    

DOI:10.1039/AA9760600027

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

4 Detector Systems Interest in the application of solid-state, vidicon, and TV-type detectors for spectrochemical analysis has continued. A number of publications have been devoted to the application of the principle of selective modulation in detection systems for analytical atomic spectroscopy, and reports have been made of the development and performance of flame and electro- thermal atomizers as resonance detectors. 4.1 SELECTIVE MODULATION AND RESONANCE DETECTORS Myers and Huchital (1 11) have reported the performance characteristics of pulsed modula- tors for the selective detection of atomic radiation, (see B. J. Russell and A. Walsh, Spectrochirn. Actn, 1959, 15, 883; R. M. Lowe, ibid., 1969, 24B, 191). An electronic multi- channel analyser was used to store and process repetitive signals to separate modular absorption from emission by the sputtered atoms.A Cu-selective modulator was used to detect resonance emission at 324.7 nm excited by a r.f. ICP source. Cochran and Hieftje ( I 181) have described a selective spectral-line-modulation technique for flame AAS with a continuum Xe arc source; it employs a second ‘source modulator flame’, into which discrete droplets of a solution of the analyte element are periodically introduced from a droplet generator, to absorb source radiation selectively at the line(s) of interest, i.e.produce selective modulation. The technique permits more effective use of a continuum source for AAS and facilitates correction for background absorption; further data are required before the advantages can be assessed in relation to the cost and complexity of the apparatus.Walsh (197) has further discussed the applications of a flame resonance spectrometer (see ARAAS, Vol. 5, p. 24) in which a separated flame resonance detector containing the analyte element is employed, rather than a monochromator, for AAS; some of the advan- tages and problems with this type of system were described. Van Loon and Radziuk (480, 1166) have presented the results of studies in which such a detector was employed for AAS, using either a premixed flame or a graphite furnace for sample atomization.In general, detection limits obtained have been inferior to those attainable by conventional AAS, but advantages lie in the ease with which multi-element determinations may be undertaken.Dawson et al. (361) have described the use of a mini-Massmann-type electrothermal atomizer, loaded with mg amounts of the analyte element of interest, to produce a stable cloud of atomic vapour for use as a resonance detector; the preliminary study was limited to evalua- tion of the system for the determination of Cd, Cu, and Pb by electrothermal AAS. 4.2 SOLID-STATE DETECTORS Work continues to be published on the use of photodiode arrays for spectrochemical measurements. Horlick (308) has reviewed the state of the art for these devices.Current limitations of these systems are cost, limited sensitivity, limited dynamic range, and wave- length coverage. An advantage of photodiode arrays is their lack of any tendency to ‘bloom’ (cf. vidicon systems), and they show promise for u.v.-visible spectrophotometry and for simultaneous multi-element analysis.Yates and Kuwana (902) have evaluated the use of a self-contained detector consisting of a linear array of diodes (Reticon), where the unit replaced the exit slit and photomultipliers of an existing commercial rapid-scanning spectrometer. Danahy and Kaiser (83) have reported on the causes of long-term instability observed with photodiode arrays and have described an ultra-stable silicon photodiode suitable for use for low-level light detection in the v.u.v., u.v., and visible regions of the spectrum.Bubert er al. (1078) have evaluated the use of linear silicon photodiode arrays 2930 Analytical Atomic Spectroscopy with parallel data output as radiation detectors in optical emission spectrometry and cited an example of their use in the determination of Cu in A1 by spark emission spectrometry: the calibration was linear between 6X 10-3 and 2% wjw.Talmi and Davidson (66) have undertaken an extensive study in which they compared the performance of commercially available cnlid-state imaging devices (SSID). A 500-element linear-array charge-coupled device (CCD), L4-element silicon photodiode (SPD) linear array (Reticon), and a 188 X 244-element ,array charge-injection device (CID) were compared.Signal: noise ratios (SNR) in the 200-500 nm region were inferior for each of these devices compared with an IP28 photomultipler. The overall performance of the SPD array was better than that of the other two imagers; its linear dynamic range was 200, it had useful response down to 200 nm, and its operation was much simpler and more flexible. The sensitivity of the SSID devices was enhanced by optical interfacing to a micro-channel plate (MCP) image intensifier; such intensified SSIDs showed a SNR gain of up to 100 over the unintensified devices. 4.3 VACUUM PHOTOTUBES Morrison and co-workers have described a computerized vidicon flame spectrometer and its application to the simultaneous determination of four serum electrolytes (585).Data acquisition, transfer, storage, and manipulation are greatly improved by the computer facility. The same group (1077) has evaluated the performance of u.v.-sensitised and silicon intensified target (SIT) vidicon tubes for atomic spectrochemical analysis, and pointed out that the use of a vidicon detector for simultaneous multi-element analysis by flame emission spectrometry facilitates rapid, accurate, and efficient application of the internal standardiza- tion technique (578).A computer-controlled spectrometer, using a silicon vidicon multi- channel detector, has been developed to examine the operating characteristics of these imaging devices as spectrometric detectors (591); the system is reported to have a signal scan SNR of 220, which is extended to lo4 with signal averaging.Winefordner and co- workers (726) have described the application of a SIT image detector to AES and AFS, using a Xe arc source and air/C,H, and N,0/C,H2 flames; detection limits obtained were greatly inferior to those obtained with the same system when using a photomultiplier detector. Cooke, Pardue, and Santini (734) have now published a description of their vidicon derivative spectrometer in which electronic wavelength modulation is achieved by superimposing a low-amplitude periodic waveform on the horizontal sweep signal of a modified vidicon detector system.A lock-in amplifier referenced to the frequcncy and proper phase of the wavelength modulation waveform generates a signal which is propor- tional to the first derivative of the optical spectrum dispersed onto the vidicon detector.Brehm (128) has described a commercially available high-speed video-spectrum analyscr which overcomes some of the problems of limited dynamic range of vidicon systems, where bright features of a spectrum tend to obliterate those which are weaker; the analyser described is microprocessor-controlled, and during the data-retrieval stage the system dynamically alters the scan rate of the vidicon, so that intense features are read more frequently than weaker ones, Weaker features can therefore be integrated up from the background without the stronger ones saturating or ‘blooming’. Garden and Aldous (1080) have further described the design of their versatile multi- element spectrometer for atomic spectroscopy, using an image dissector. This type of detector has a number of advantages over photodiode and vidicon systems, and i s particu- larly suitable when good temporal resolution of transient signals is required.Part I : Fundamentals and Insfrumen fation 4.4 SIGNAL PROCESSING Horlick and Betty (552) have discussed the application of discrete time analogue signal- processing devices to the processing of noisy signals.With these new devices an analogue signal can be sequentially sampled and each sample stored as an analogue level in the device and sequentially read out; in some devices, parallel read-out of the storage cells is possible. Thus these devices can be regarded as analogue shift registers or delay lines with serial electrical input and serial or parallel (tapped) electrical outputs. Chapman and Williams (1233) have described an analogue-to-digital converter which may be used to transform a multi-channel analyser into an analogue signal averager suitable for analysis of analogue signals in the frequency range d.c.to 40 kHz and signal amplitudes in the mV to V range; this converter effectively transforms an analyser from a specialized digital instrument into a general-purpose laboratory tool. Hamm and Zeeman (304) have reported a useful, inexpensive digitizer operating on a 5 V supply for operation with photomultiplier currents. Signal averaging with a magnetic tape loop and boxcar integrator has been reported (1231); in this analogue technique an amplitude-averaged waveform is recovered from a noisy synchronous signal of short duration.The system reported is advantageous where the changing or non-stationary nature of the events prevents signal averaging by slow scanning, and when equipment for ‘real-time’ analysis is not available.Lawton, Bolden, and Shaw (1230) have described a 512-channel photon counter with a resolution of 10 ns per channel. The input of photon count data is made via a de-randomizer and shift register implemented in emitter-coupled logic circuitry. The instrument was specifically designed to measure fluorescence excited in a weakly ionized plasma by a pulsed, tunable dye laser. Chapman (74) has described the development of a procedure for correction of photometric drift for a direct-reading spectrometer. 31 Other references of interest - Characteristics of modern photographic materials for spectroscopy: 51 6. Photoionization detectors for V.U.V. spectrometry: 7. Review of detection systems for ES: 985. Review of digital techniques in laboratory automation: 1232. TV read-out system for u.v. /visible spectrometry: 1076. Use of Polaroid film with spectrographs: 462. Vidicon detectors in ES: 1364. 1240.

ISSN:0306-1353    

DOI:10.1039/AA9760600029

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

5 Data Processing 5.1 EMISSION SPECTROSCOPY The use of mini-computers in direct-reading ES is well established, both for data process- ing and more recently for the operational control of the spectrometer. Individual labora- tones have found that the increased productivity and analytical sophistication far outweigh the additional cost (946). Lack of understanding of the short-term behaviour of spectro- graphic events may, however, cause difficulties when real-time computer control is attempted.Hieftje (945) has examined recent developments in transducers, techniques, and basic knowledge of spectrographic processes, with a view to their incorporation into new, fully computer-controlled instruments. A new class of integrated circuits now becoming available has exceptional potential for performing real-time, analog data-processing operations (552). Advanced software concepts have been described for the optimisation of experimental parameters in AE and ICP techniques (946).Off-line computing has been used to calibrate spectral data from a synchronous photon counter by comparison with a simultaneously recorded reference spectrum (31 3). Programs have also been described to obtain a linear equation for the photographic emulsion calibra- tion curve, with subsequent calculation of element ratios and concentrations (31 l), and for the identification of different elements present in the sample (1465). 5.2 ABSORPTION SPECTROSCOPY Programmable calculators have proved to be a useful aid for data manipulation in AAS; the calculator can perform such functions as signal averaging, curvature correction, and peak integration (99, 696, 1033).Harrison et al. (1071) have described a particularly com- prehensive system incorporating error analysis, data filing and retrieval, and report genera- tion. Significant improvement in analytical performance is claimed by Kahn and Stux (1033) in electrothermal AAS, where the calculator can be used to provide simultaneous read-out of both peak area and peak height.Improvements in performance can be confidently expected for the microprocesser-based AA spectrometers now becoming available. As an example, Fernandez and Kerber (694) claim constant precision at all absorbances between 0.05 and 1 with RSDs from 0.002 to 0.004. Rapid and efficient handling of data has been achieved by interfacing a minicomputer to a vidicon, optical multi-channel analyser for application to the analysis of blood.The interface described allows for both manual and computer control of the vidicon data output for maximum analytical flexibility (585). Signals generated in many AAS techniques are transient, with widely different profiles. Measurement of integrated peak areas improves precision and linearity (721). There are, however, practical difficulties in setting the limits for integration to accommodate changing profiles due to changes in matrix or concentration. It is suggested that a computer algorithm might be used to provide a decision-making mechanism in such situations (744). Calibration of multi-channel analyser for ES and AFS: 368. General data-processing system: 431. Spectral deconvolution: 923. Other references of interest - 32

ISSN:0306-1353    

DOI:10.1039/AA9760600032

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

- Tables A, B, and C contain a comprehensive summary of instruments for analytical atomic spectroscopy, and once again we are indebted to instrument makers for their collaboration in providing details of their instruments. The tables present information available at January, 1977. The trend in instrument design is toward the incorporation of safety devices, automatic controls, and data processing by computer.Very few new emission spectrometers have appeared on the market this year. Several companies, however, now offer ICP units as standard options. MBLE have introduced the PV 8250 spectrometer system with built-in source unit and read-out by printer, teletype, or computer. Baird-Atomic have entered the field of AA and introduced a new single-beam instru- ment, the A5100, which has automatic background correction and curve correction; in addition, the former Shandon Southern range are marketed.Pye Unicam have upgraded their SP 191 model with simultaneous background-correction facility (SP 192) and intro- duced the SP 2900. Rank Hilger have a new instrument, Atomspek H1551, with built-in background-correction and flame-emission facilities.Varian Techtron have introduced a new model, the AA 175. From Perkin-Elmer there is a range of 3 new instruments, all with microprocessor facilities, and with provision for burner-head safety interlock. Boden- seewerk Perkin-Elmer announce three new instruments, two with double-grating mono- chromators and two with a microcomputer. An interesting instrument promises to be the Hitachi 170-70 AA spectrometer, which has Polarised Zeeman Effect flameless background correction over the complete wavelength range.An automatic sampler, AS1, for use with electrothermal atomizers has been announced from Bodenseewerk Perkin-Elmer. This allows sampling of up to 30 samples, singly or up to 9 times each, and it is claimed to give signifi- cantly improved precision.Some of the unsolved problems of instrumentation in atomic spectroscopy have been identified by Silvester (1 350) as: (i) nebulizer and stray light associated with simultaneous ICP spectrometry; (ii) over-correction by automatic background correctors in AAS; and (iii) speciation and lack of reproducibility in electrothermal AAS. In a recent review of techniques for multi-element analysis, Boumans (1271) concludes that the low-power ICP provides the best balance between cost and performance, but its ease of operation and reliability when in ordinary routine use have yet to be proved.Walsh (467, 1273) has suggested that special-purpose instruments dedicated to a given analysis could be made by combining atomic spectral lamps with sputtering chambers or flames and resonance monochromators, to produce non-dispersive atomic spectrometers suitable for the analysis of solutions and solids by AAS, AFS, or AES.Flexible systems for use in emission, absorption, or fluorescence spectroscopy can be achieved by the use of spectrometers with either vidicon (104, 591) or image-dissector (309) detectors. These systems incorporate computer control to facilitate the rapid selection of a large number of spectral lines.Using the vidicon, the read-out beam can be inhibited, so as to increase the target’s integration time and thus enhance weak signals. 6.1 EMISSION INSTRUMENTS Whilst there seems to be a continuing trend toward the reduction of analysis time in, for example, the steel industry, by locating the spectrometer close to the source of the sample (65), potential users should perhaps question whether the increased expenditure for this end is appropriate in view of the time spent in other stages of the operation. 3334 Analytical Atomic Spectroscopy As the development of new sources, such as the GDL and the ICP, has extended the range of analysis to many more materials, more stringent requirements have been placed on the spectrometer.New instruments incorporating dual holographic gratings which are now appearing on the market (792) may go some way toward meeting these needs. The high resolution and compactness of the echelle spectrometer make it an attractive instrument for analytical atomic spectroscopy, but to date the limited demand for such spectrometers has kept their price rather high.The historical development and the use of the echelle spectro- meter with the d.c. Ar plasma have been reviewed by Cox (1072). Scanning spectrometers have been described for use with plasmas. One (360), incorporat- ing a 0.5 m Ebert monochromator, uses a stepping scan with a movement of 0.003 nm per step, and it achieves an RSD of 0.01 in the measurement of the intensity of an Hg line: another (1028) employs an oscillating galvanometer mirror to give a scan of ca. 100 nm in 10ms. This system was used for the simultaneous measurement of Mg, Mn, Cd, and Bi in an Ar microwave plasma; an RSD of 0.02 was obtained. A slcwed-scan monochromator with rotatable quartz plate for spectral line modulation has been used i n an emission’ fluorescence flame photometer (577).Simple flame photometers continue to perform a valuable function in clinical labora- tories, and in a wide variety of applications, such as the separation of mixed steels by their Mn content (629). For some elements, the range of these simple instruments may be extended by the incorporation of integrating facilities (418). A new type of filter flame photometer (589) employs wavelength modulation, produced by oscillation of an interference filter, to minimise the effect of background emission; the detection limit for Ca was 10 ng ml-1, and at higher concentrations the RSD was 0.004.A four-channel emission/ absorption interference-filter flame photometer for the simultaneous determination of Na, K, Ca, and Mg has been reported (1324). The instrument was used for the analysis of serum, and achieved an RSD less than 0.01. By means of automatic sample dilution and curve correction by computer, a throughput of 120 samples per hour was achieved. 6.2 ABSORPTION INSTRUMENTS AA equipment available in the Netherlands has been reviewed by Hcndrikx-Jongerius and de Galan (1227). The performance of four commercial AA spectrometers was compared for the measurement of 95pgml-1 of Cu in the presence of <90mgml-l Zn (1555).The RSDs ranged from 0.015 to 0.028; double-beam operation was found to be advantageous. Walsh (784) has reviewed recent work at C.S.I.R.O. Australia, on cathodic sputtering in relation to sputtering cells, high-intensity lamps, demountable boosted-output HCLs, and Grimm-type discharge lamps.In spite of promising results, there is, as yet, no widespread use of such systems other than the Grimm lamp. It has been suggested (1334) that the direction of future developments in AA may be in the use of the flame resonance detector and the incorporation of AA detection with chromatographic separation. The question of absolute analysis by flame AAS was examined by L’vov (144).The difficulties in making absolute measurements based on this analytical technique are discussed in detail. In practice, AA is primarily a single-element technique, but efforts continue to be made to develop a multi-element capability. The two principal practical difficulties are the combination of light from several spectral line sources into one optical beam without loss of a major portion of the original intensity and the simultaneous measurement of several spectral lines.A more fundamental difficulty is the limited dynamic range (ca. 2 orders of magnitude), which may not accommodate the concentration ranges of the elements in the sample. The use of an image dissector tube for spectral analysis has been described by Aldous (126, 1169) and applied to the simultaneous determination of Pb, Zn, Ca, and Cd inPart I : Fundamentals and Instrumentation 35 biological specimens, using electrothermal atomization AA.By means of an electro- mechanical programming system, a flame AA spectrometer was used (1570) for automatic sequential multi-element analysis. The concentrations of Co, Cr, Cu: Fe, and Ni at which the RSD was 0.07 were found to be 100, 6, 30. 55, and 100 pg ml-1, respectively, and the characteristic concentrations (1 % absorption) were 2.0, 0.1, 0.4, 1.1, and 1.2 pg ml-1, respectively. A simple system for the simultaneous measurement of Ca and Mg in sub-nl samples of biological material, using electrothermal atomization and interference filters, has been reported by Antonetti and Grosso (185, 1634).Multi-element capability for Ni, Ca, and Ge, with improved precision and with the elimination of systematic errors, has been achieved (774), using a flame as the atomization cell, in combination with a d.c. arc for emission measurements and an HCL plus a continuum source to provide background correction for AAS. A non-dispersive AA instrument comprising an EDL, flame atomizer, photoresistor detector, and a narrow-band amplifier (756) was found to be possible if the analyte element was separated from its matrix by sorption on, and elution from, an ion- exchange resin, The detection limits using a double-beam non-dispersive AA spectrometer for the measurement of Ag, Bi, Cs, Cu, K, In, Na, and Rb were found (1372) to be comparable with those from conventional single-beam instruments.Other references of interest - AA instrument with microprocessor control: 1195. Double-beam AA instruments: 1052, 1213, 1536. Double-beam AA instruments with background correction: 1655. Double-beam AA instruments with electrothermal atomization: 1581, Zeeman-effect AA spectrometer: 1544. 6.3 FLUORESCENCE INSTRUMENTS In comparison with thermal excitation techniques, the advantages of atomic fluorescence ought to be the generation of an emission signal against a low background and the ability to modulate that signal to provide additional discrimination.When thesc advantages are utilised, the minimum resolution required of the spectrometer should be less than that necessary for an equivalent emission system employing arcs, sparks, or plasmas.In practice, when the analyte element is either isolated or in a simple matrix, non-dispersive AF becomes feasible. Advantage has been taken (1188) of the separation afforded by chroma- tography to achieve non-dispersive AF determination of Mn. There continues to be an interest in developing instruments to exploit the potential of atomic fluorescence methods, though information on the application of these devices to real analytical situations is sparse.For the foreseeable future, the me of AF appears likely to be restricted to a few elements, e.g. Cd, Hg, and Zn, where the technique is particularly sensitive. The state of the art of practical developments in AF analysis has been reviewed by West (718). Temperature-controlled EDLs are the most useful sources for excitation.The feasibility of time-resolved multi-element determinations of simple mixtures by control of the heating rate of an electrothermal atomizer was reported. Instruments using a continuum as their primary light source with either an air/C,H, flame (770) or a graphite filament atomizer (1584) have been described. Two techniques have been reported for overcoming the problem of scattered radiation.In one system (86, 1168) two light sources are used; a line source to excite the fluorescence radiation and a continuum to provide a measure of the scattered signal. The light sources are modulated at the same frequency, but are out of phase. Initially the channels are balanced to give zero output, and any disturbance of that balance will be due to fluorescence radiation.The other system (402) uses a single light source but two different wavelengths, one being that of theTable A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS Applied Quanto- Research meter 20 Laboratories Ltd., Wingate Road, Luton. Beds.. England Quanto- vac 28 Quanto- vac 28C Quantc- vac 80 Quanto- meter 80 Ouanto- meter 290008 Quantc- meter 33000 Quanto- vac 33000 Quanto- meter 33000 LA O.A. 137 D.R. D.A. D.R. D.R. W.R. D.R. D.R. D.R. D.R. D.A. 60 1.388 or 0.695 (20 lines) 0.695 or 0.35 60 0.70 or 0.35 As Quanto- 0.70 or 0.35 (28 lines) vac 28 96 0.46 (60 lines) As Quanto- 0.695 or 0.35 vac 00 60 0.35 or 0.175 (48 lines) 0.46 or 0.23 0.56 or 0.28 0.695 or 0.35 (64 lines) 0.695 or 0.35 (8 reference) As Quanto- 0.46 meter 33000 As Quanto- 0.695 or 0.35 meter 33000 48 0.46 200--800 0.75 m 2 o w M ) 175-500 0.5 m 175-500 0.5 m 17-07 1.0m 19~l-610 1.0m 190-520 1.5 m 190-630 196705 190-840 190-610 1.0m 17-07 1.0 m 190-610 1.0m 185-410 1.0m Low voltage, high voltage As Ouantovac 80 As Quantovac 28 Air or argon excitation stands; typewriter and digital computer options As Quantovac 80, but no air-conditioner Complete computer control; teletype or visual display output; off-line computer links Various; low Typewriter, teletype.and voltage, high digital computer options; voltage, multi- source IHVS. LV, options: second stand single or dual stand d.c. arc) As Quantovac 80 As Quantovac 80 As Quantovac 80 As Quantovac 80 H.f. plasma ICP can be argon or air; built-in instrument air-conditioning As Quantovac 80 Typewriter. teletype, and digital computer options; argon and/or air stands ivailable Automated sequential analysis; computer options also available AS Quantometer 33000; computer options also available Automatic loading of up to 24 samples P.p.b. analysis; computer options also available; direct solids nebulizer can be fitted Particularly suited to non-ferrous.e.g. Al. Mg, Cu. Zn, and white metals, slags. powders. solutions, including oils As Quantovac 80. but limited to 28 elements AS Quantovac 28 All ferrous and non- ferrous alloys, powders including slags, sinters, ores. rocks, ceramics, soils, etc.; solutions, oils, etc. As Quantovac 80. but excluding determination of C, S, and P As Quantometer 80 AS Quantometer 80 As Quantovac 80 Solutions Solutions of many materials; ferrous/non- ferrous, slags, clinical and pollution control applications 36Baird-Atomic SB-1 Phot.- 1.5 or 0.75 1.5 m 1.5 m 1.0 m 1.0 m 2.0 m 2.0 m Arc or spark Built-in order sorter General specirographic analysis 2 370-740 450-750 21 0-490 173-767 190.432 190-863 173-432 420-970 21-65 Inc., 125 Middlesex Turnpike, SH-1 Bedford, Mass. 01730, Spectro- U.S.A. met 1000 General spectrographic '1 Ferrous metals (except 5 determination of S) ,. P 214.9 nm in 2nd order; non-ferrous analysis *rl using C 193.1 nm. 2 2 metals,oils 5 Ferrous and 5 non-ferrous metals, including C. S. and P 2 D- 2 Gl T Phot. - 1 .o D.R. 30 0.6 or 0.3 Arc or spark Arc or spark; modular Built-in order sorter Compact, low-cost direct reader with minimum air-conditioning requirements; manual master monitor to check slit alignment Compact, low-cost direct reader with minimum air-conditioning requirements; logarithmic read-out; manual master monitor to check slit alignment; dual stand option Automatic optical servo monitor continuously maintains correct slit alignment; logarithmic read-out; manual master monitor to check slit alignment; temperature- compensated fixed focal length; dual stands for argon and air available As Spectromet I I ; all photomultipliers in vacuum Warner Drive, Springwood Industrial Estate, Rayne Road, Spectro- Braintree. vac-1000 Essex CM7 7YL.England D.R. 30 0.6 or 0.3 Arc or spark; modular As Spectromet 1000 Spectro- met II D.R. 60 0.294 0.59 All direct-reader 2 applications above -..n, 190 nm c -_ D.R. 60 0.29 As Spectromet 1000 All direct-reader applications, including C. P. and S Spectro- vac I1 _- Jarrell-Ash 78-090 Div., Fisher Scientific Co.. 590 Lincoln St., 70-310 Waltham. Mass. 02154. U.S.A. 75-150 - 1.5 m Various available Wadsworth spectrograph; General spectrographic in 'Varisource 20 inch camera analysis unit. including spark, low- and 20 inch camera General sDectroaraohic Phot.- 1.1 or 0.54 Phot. - 1.0 or 0.24, depending upon grating 180-3000 3.4 m I 80-1 500 180-750 high-voltage ' d.c. arcs. . - analysis 1 ~ I S O versatile controlled wave-excitation Choice of 3 gratings; Versatile instrument. source'. Plasma nitrogen purging extends particularly suitable range to 175 nm. for measuring transient optional accessories spectra permit use as direct reader or scanning spectrometer As above, except Computer conlrolled Most metallurgical trolled peak current .electronic con- analyses Phot.- 4.4 to 1.1 1.6 to 0.4 3.2 to 0.8 200-6000 0.75 m 1.0 m or 2.0 m 168--500 0.75 m 168--500 0.75 m 96-750 96-785 (continued ) D.R. Up to50 0.54 D.R. Up to 50 0.54Tablc A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS --continued Supplier Model Type channels No.of ~~~~~~~~~ nm per mm Wavelength range/nm length Focal Type of source Special features Applications (continued) 1500 70-314 96-975 84-405 82-410 82-41 5 82-000 75-150 D.R. D.R. D.R. Scan. Scan. Scan Scan. Scan Up to 60 30 Up to 50 - - - - - 0.56 or 0.28 200--800 or I 10.34 or 0.17 200-510 or 190--400 i 1 . 5 m As above 19k250 As 70-310 As 70-310 3.4 m As above 0.54 168-500 0.75 m ICP 200-900 0.25 m Supplied 3.3 by user 1.6 and 3.3 200-900 0.25 m Tungsten Depends on Depends on 0.25 m As above Deuterium grating grating selected As above As above 0.5 m Supplied As above As above 0.75 m, As above by user 1.0 m 2.0 m Choice of 2 gratings Easy interchange to photographic (70-31 0) version Computer controlled; 1 variable channel All direct-reader applications above 190 nm As for Model 1500 All solutions Various scanning 1 spectrometers Suitable for spectro- scopic investigations rather than for analytical applications b Y Labtest 310 D.R 60 0.56 19&900 1.5 m ‘Transource’ Wavelenath in first order: Ferrous and non-ferrous Equipment Co.high-voltage- CRT; telgtype printer or ’ alloys 11828 La Grange lriggered computer readout systems; Ave., V-25 D.R. 40 0.67 170-550 1.0 m discharge. Low- dual air/inert gas and As above Los Angeles. voltage-triggered solution excitation stand Calif. 90025, 2100 D.R. 30 0.46 188-455 1.0 m d.c. arc. As above U.S.A. 71 D.R. 74 0.52 170-900 2.0 m ICP source for solution analysis General purposes Kontron GmbH ICP D.R 8051 Echina be olasma- 0.8-1.6 mm 185-700 0.5 to ICP - l m Munchen, - spec Oskar-von- System 3 Miller-Str. 1 System 4 Up to 30 023-0.46 mm 187-455 1 m ICP - West Germany General purpose General purposeGlen Creston b NW9 OHL. ? 16 Carlisle Rd. England London 5 2 M.B.L.E., Rue des Deux-Gares 80, Brussels. Belgium 8-1 070. Philips Analytical Dept., Pye Unicam Ltd.. York Street, Cambridge, CB1 2PX.England Philips D.R. 60 0.55 or 0.46 PV 8300 (80 lines) Vacuum Philips D.R. 20 0.46 PV 8350 Vacuum Philips D.R. 60 0.55 or 0.28 PV 8210 (50 lines) Air Philips D.R. 40 0.695 or 0.35 PV 8250' Air 0.83 or 0.42 170-430 1.5 m Triggered capacitor discharge; Monoalternance discharges up to 500 Hz; d.c.arc; intermittent d.c. arc 177410 1 m As for PV 8300 190-700 1.5 m As for PV 8300 'plus ICP 190-610 1 m As for PV 8210 190-780 Optional dual air/argon excitation stand; readout by printer.teletype, or digital computer systems Integrated spectrometer system including source and readout options as for PV 8300 - Staels. iron, non- Q. ferrous metals, and non-conductive powders: 2 air stand for oils, d.c. arc, etc. 1 & !? Steels. iron, non-ferrous 9 metals, non-conductive powders 3 & 5 3 -+ Wavelength range covered All direct-reader in 1st order; remote- analyses above 190 nm.'. controlled roving detector; particularly non-ferrous 2 dxternal excitation; rotrode metals, solutions, oils, and inert atmosphere and non-conductive facilities; readout as powders PV 8300 Integrated spectrometer As for PV 8210 system with built-in source and readout options as for PV 8300 Rank Hilger.El000 D.R. 60 0.29%1.155 159.6-864.3 1.5 m Various, including Dual spark stands; Ferrous and non-ferrous Westwood Polyvac high-repetition computer-controlled alloys: geological Industrial condensed arc instrument; dual gratings samples; wear metals Estate, give 7 systems in oil Margate, Kent, E952 D.R. 36 0.546 or 0.741 174.0-447.7 0.75 m As El000 Curved entrance and Ferrous and non-ferrous CT9 4JL.exit slits; solid-state alloys: wear metals England electronics or computer in oils controlled; air or vacuum Monospek D.R. Single 0.66-15.7 200-22000 1.0 m As selected Curved or straight Scanning monochromator D-400' entrance and exit slits; of particular use for scanning wavelength can monitoring and be read to 0.01 nm from examination of plasma digital counter; wavelength sources. accuracy 2 0.1 nm with 1200 line per mm grating *New equipment since publication of Volume 5 W \3Tablc A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS-cconfiriued -.- Supplier Spectrametrics Inc.. 204 Andover St.. Andover. Mass. 01810, U.S.A. Techmation Ltd.. 58 Edaware Way, Edgware. M idd lesex, HA8 8JP. England Spex Industries Inc., 3880 Park Ave.Metuchen. N.J. 08840. U.S.A. Glen Creston 16 Carlisle Rd London NW9 OHL. England Model AE 2 ~ ~~~ D.R.10 ES 9 RS 1 Spectra- span Ill __ . 1870 1702 1704 1802 Phot., D.R. D.R. Phot. D.R. Phot., D.R. Scan. Phot. Phot. Phot - Spectroscandta IDES D.R. AB, 2080 SF-21660 Nagu Finland Reciprocal channels nm npr mm range/nm No. Of dispersion/ Wavelength EGA Type of source Special features 1 20 (inter- changeable cassettes) Applications ...... r__ ...... 1 (variable wavelength) 20 (inter- changeable cassettes) 0.06 190-900 0.75 m Plasma jet 0.06 190-900 Plasma jet 0.06 Optimised AE system using a high-dispersion, hig h-energy-throug hput ec he I le spectrometer and a high-temperature plasma jet excitation source 190-900 0.75 m Plasma jet, flame, Built-in compute1 or arc stand 190-900 0.75 m Plasma jet, flame 0.06 or arc stand 0.06 190-900 0.75 m d.c argon Optimised A€ system plasma using high-dispersion high-energy-throughput echelle grating spectrometer and a high-temperature plasma jet excitation source; built-in micro-processor; most spectral and matrix effects are eliminated Routine analysis Routine quantitative multi-element analysis Qualitative and semi- quantitative analysis; spectroscopic research Routine sequential; quantitative analysis and multi-element analysis ... . 100 (300 lines) - Multi-purpose unit Routine analysis 1.6 175-1280 0.5 rn 1.1 175-1500 0.75 m - - 0.8 17C1500 1.0 m - - Research Research 0.8 180-1500 1.0 m - Direct-read ing Routine analysis accessory available 0.16 at 200 200-800 0.5 m Hollow cathode Channels not preselected, Ferrous and non-ferrous 0.32 at 400 discharge, changeable at any time: metals, slags, powders. 0.52 at 650 plasma, d.c.arc channel minimum spacing ores, geological 0.2 nm; wavelength specimens, trace 0.63 at 800 elements in metal, high accuracy 0.001 nm; plane samples accuracy at low and high concentrations 0.8 to 0.1 cm; CRT, lineprinter.or teletype readouts; digital computer as standard ~~~~ ~~ ~ ~ ~~ -Oplica S.A.S., 65t Phot. - 0.694.36 200--8W 1.2 m Via Gargano 21, 20139 Milano, Italy 65Ct D.R. 16 0.69 or 0.36 220--420 1.2 m 67Vt D.R. 93 0.37 165-440 1.5 rn ESAlt Scan. - 0.41 zca--500 1.0 m ESA3t D.R. 9 + Scan ESA4t Scan. - 0.36 160-500 1.2 m (40 nrn as POIY- chromator) 0.41 165-500 1.0 m All conventional types available LV-triggered arc and spark; HV spark, a.c.and d.c. arc LV-triggered arc and spark; HV spark Controlled and non-controlled HV spark; a x . arc LV-triggered arc and spark LV-triggered arc. HV spark; a.c. arc Stigmatic instrument with rotating Ebert grating Double spark stand both in air and inert atomsphere; Rotrode for solutions Air-vacuum instrument with all 92 exit slits accessible from outside for adjustment; many analytical programmes can be arranged in parallel for easy interchange; computer facilities available Scanning monochromator with one channel for analytical line and another channel for reference, using reflected- beam principle.Combined vacuum mono- and poly-chromater; all excitable elements accessible with scanning system Scanning vacuum monochromater with one channel for analytical line and another channel for reterence; facilities for analysing two elements simultaneously -c s General purpose; 7 metallurgical analysis. 5 e.g. Al, Pb. Zn, Fe. cu alloys; wear metals in oils, etc. 2 3, Complex analyses involving many spectral 3 lines n General purpose Ic a r t P Metallurgical work; all 4 material excitable with same source parameter f 5. 3 Routine analvsis (including C:S, and P) of iron and steel; non-ferrous alloys Metallurgical work; analysis of ferrous and non-ferrous alloys RSV- SPN 3.5t Phot.. 30 0.14-0.48 200-1000 3.5 rn Glow-discharge Paschen-Runge mounting General analysis Prazisions- D.R. lamp. high- specially designed for mehgerate medium-, or low-range below 200 nm; GmbH., voltage spark, direct-reading attachment E031 Hechendorf a x .or d.c. arc, available Pilsensee. continuous and West Germany inlermittent (continued) tNo up-todate information was available for these instruments when this table was compiled. 5Tablc A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS- coniiriued P EJ Reciprocal nm Der mm range/nm Supplier Yodel Type chN.3;${s dispersion/ kGA Type of source Special features Applications ~~ ~ 0.24-0.84 0.37-1 1 As above (continued) Siemens Ltd., Great West House, Great West Road, Brentford.Middlesex. England SPN 2.0t Phot., D.R. 30 15 - - 40 200-1000 2.0m As above As above SPN l . 5 t Phot.. D.R. 200-1000 1.5 rn As above As above As above SPN l . O t Phot. 0.561.7 0.4-1.7 200-1000 1.0m As above 300-1300 1.0 m As above As above As above As above As above SPN 1.0t Phot.(vat) Analymat D.R. I-airt 0.31 or 0.54 2CC650 1.5 m Glow-discharge lamp (o:hers available) Exhibits no background; no matrix effects; linear calibration for all elements 0--100% As above As above Analymat D.R. Il-vact 40 40 250 0.31 or 0.54 15G-490 1.5 m As above AS above bAnalyrnat D.R.II I-vact 0.42 or 0.5 110-530 1.0 m AS above As above As above Analymat D.R. IVt 0.22 2 x 2.5 m spectrum length 2x 2.0 m As above 200-600 As above As above Analymat D.R. V t 250 250 - - As above 2x 2.0 m As above 150-600 2x 2.0 m As above 120-430 200-630 2.0 m As above As above As above As above As above As above As above b As above Analymat D.R. Analymeter Scan Analymeter Scan.V l t 11 Ilt As above 0.16 - 4, x .... n As above B 0.16 0.16 150-630 2.0 m As above As above Analymeter Scan. lllt 110--630 2.0m As above As above k General purpose B 2 GCT-1007 Phot., (Czerny- D.R. Turner) GE-170t Phot. (Ebert) 0.83 200-850 1.0 m Modular-source DCA ACA LVS, LVA 3 (1200 grooves (10" camera) - 0.46 200-1200 1.7 m HVS (max.) /mm) /mm) 1200 grooves ( 1 0 camera) S,DCA High speed Shimadzu Seisakusho Ltd.. 1 Nishinokyo- Kuwabaracho, Nakagyo-ku. Kyoto. Japan 2 General purpose C h vGEW-170t Phot., 55 0.43 200-1200 1.7 m (Ebert) D.R. (max.) (1200 grooves (20” camera) /mm 1 GQM-75i D.R. 35 0.52 190-430 0.75 m HVS. LVS (max.) (2400 giooves & 510.5 DCA /mn) 589.0. SG-400 . . 518.3 GVM-100t D.R. 60 0.46 173-410 1.0m HVS. LVS (max.) SG-400 3 kinds of readout General purpose electronics are available 1 .built-in computer 2. digital, with linsarizer 3. pen recorder Solid, liquid, powder, metal VEB Carl Zeiss PGS-2t Phot. - 0.74 or 0.37 200-2800 2.075 m Arc or spark Automatic expansion oi General specirographic analysis; also examina- tion of line profiles, .rena, 69 Jena. Carl-Zeiss Str. 1 . dispersion doubled by hyperfine structure, etc.German double passage of light; measuring range; stigmatic depiction; - -. . . . -. . Democratic Republic Carl Zeiss Scientific pre-disperser for order sorting and isolation; gratings interchangeable; automatic transport of plate holder - -. -. . . . . - Instruments Ltd., Q-24t Phot. - 0.76 210-550 0.54 m Arc or spark Full range of General spectrographic PO Box 43.accessories available analysis 2 Elstree Way, Borehamwood, Herts. WD6 1NH England tNo up-to-date information was available for these instruments when this table was compiled. 43P Table B COMMERCIALLY AVAILABLE ATOMIC ABSORPTION SPECTROMETERS P Supplier Single/ Grating Reciprocal Model double Monochromator lines dispersion/ Rero'ution Wave'emh %%'; Other features beam per mm nm per mm lnm range/nm expansion Baird-Atomic Ltd..A5100* Warner Drive, Springwood Industrial Estate, Rayne Road, Braintree. Essex CM7 7YL. England A3400 A3600 Single 0.25 m Czerny- 1200 3.0 0.1 18-60 Digital; Automatic background auto zero; integration; curve correction; wavelength scan; flame ignition; gas safety devices; lens optics; emission and fluorescence Turner x 05-40 correction: 4-lamp turret; Single 0.25 m Czerny- 632 6.0 0.2 19G860 Meter or 4-lamp turret; auto zero; Turner digital; x25 curve correction; integration; flame ignition; Wavelength scan: emission and fluorescence Single 0.25 m Czerny- 632 6.0 0.2 790-4360 Meter or Integration; flame Turner digital; x25 ignition; emission and fluorescence Beckman Instruments GmbH, 1233 Double Littrow 1200 2.7 0.2 1 9 k 8 6 0 Meter; x 55 8 Munich 40, Frankfurter Ring 115, West Germany Beckman-RIIC Ltd.. 1236 Double Littrow 1200 2.7 0.2 190-860 Digital; x55 Eastfield Industrial Estate, Glenrothes, Fife, 1248 Double Littrow 1200 2.7 0.2 190-860 Meter; x 10 KY7 4NG. Scotland 0.2 190-860 Digital; X 10 1272 Double Littrow 1200 2.7 GCA/McPherson Instrument, EU 703 Single - 530 Main St.. Acton.Mass. 01720. U.S.A. ~~ ~~ ~ 1180 2 0 0.1 18k1100 Digital .- Hitachi Ltd.. 170-10 Single Littrow 1440 2.25 0.4 190-900 Meter; Nissei Sangyo Co. Ltd.. xo.l-1 15-12 Nishi-Shimbashi. x 1-10 2-Chome, Minato-Ku, Digital; Tokyo, Japan (optional) Sinyle- or triple-pass optics; % T; abs. or concentration readout As model 1233 b Auto zero and calibrate; integration . As model 1248 EL plus curvature correction s. 5 Modular AA; flame emission; various detectors and gratings available; convertible t o s, single- or double-beam 0 c) U.V. spectrometer h 2 .. .- . - 2 Single lamp mounting, ; NZO-air simultaneously C.I exchanged; concentration 2 readout; continuously variable time constant 2170-30 Single Littrow 170-50 Double Littrow 170-70* Double Littrow 1440 2.25 1440 2.25 1440 2.25 Instrumentation 351 Double 0.33 m Ebert 1200 2.5 Laboratory Inc.. 113 Hartwell Av., Lexington, Mass. 02173, U.S.A. Instrumentation Laboratory (UK) Ltd.. Technical Services Div., Edgeley Road Trading Estate, Cheadle Heath, 251 Double 0.33 m Ebert 1200 2.5 Stockport SK3 OXE, England 151 Single As 251 in all other features 0.4 0.1 0.1 - 0.03 0.03 190-900 As 170-10 19&900 As 170-10 19&9M) Meter/Digital option 190-900 Digital; X 50 190-900 Digital; X 50 Concentration readout; time-weighted signal averaging; AA/AE measurement; auto zero; NzO-air simultaneously exchanged Backaround correction: base-iine drift correction; curve corrector; time-weighted signal averaging; auto zero Polarized Zeeman effect; flameless background correction over the complete 190-900 nm wavelength range; background correction to 1.7 abs 4-lamp turret; wavelength drive; full time integration of peak height or peak area; auto calibration; curve correction; background correction; dual grating; push-button operation; zoom lens; full automatic safety gas controls 4-lamp turret; wavelength drive; full time integration; peak height or peak area; off-line calibration; curve correction; background correction; zoom lens; auto gas controls Jarrell-Ash Division, Dial Single 0.25 m Czerny- 1180 3.3 0.02 193-860 Digital Laminar-flow burner; Fisher Scientific Co., Atom Turner integral gas-flow controls; 590 Lincoln Street, Ill auto zero; concentration Waltham, Mass. 02154, calibration; curvature U.S.A. correction; 2-lamp turret 82-810 Double; 0.4 m Ebert 1180 2.08 0.03 190-900 Digital; Laminar-flow burner; dual x 25 curvature correction; 2-lamp turret (continued) channel 'New equipment since publication of Volume 5Table B COMMERCIALLY AVAILABLE ATOMIC ABSORPTION SPECTROMETERS - continucd * cn Single/ Grating Reciprocal Supplier Model double Monochromator lines dispersion/ Resolution Wavelength scale Other features beam oer mm nm oer mm inm range/nm exDansion (continued) 82-850 Double 0.4 m Czerny- 1180 2.08 0.03 190-900 Digital; Computer-controlled Turner parameters Jobin-Yvon, Division d'lnstruments, 16-18 Rue du Canal, 91160 Longjumeau.France New model to be marketed July 1977. No details yet released. Perkin-Elmer Corp.. Main Ave.. Norwalk. Conn. 06856, U.S.A. Perkin-Elmer Ltd., Post Office Lane, Beaconsfield.Bucks. HP9 lQA, England 272* Single 0.27 m Littrow 1800 1.6 0.2 1901860 Digital; High-energy optical X 0 . 0 1 4 0 system; microprocessor- controlled; auto zero; auto concn.; auto curve correction, with 2 standards; peak height; peak area; integration time selectable from 0.5 to 20 s; flame ignition; optional auto NzO switching; optional burner-head safety interlock; optional deuterium arc background correction x 0.01-50 mirror optics; automatic 372' Double 0.27 m Littrow 1800 1.6 0.2 1901860 Digital; As Model 272.but all gain control; optional deuterium arc double- beam background correction 373. Double 0.27 m Littrow 1800 1.6 0.2 1901860 Digital; As Model 372, but auto x0.01-50 NzO switching; burner- b a head safety interlock; o pressure sensing opticnal flame and 2, 460 Double 0.27 m Littrow 1800 1.6 0.2 1901860 Digital; As Model 373.but x0.01-100 auto curve correction, with up to 3 standards; integration time selectable from 0.2 to 2. 60 s B Turner vis. 1440 1.3 400-900 x0.01-100 optional 4-speed 2 2 - 603 Double 0.4 m Czerny- U.V. 2880 0.65 0.03 180.440 Digital As Model 460, but waveleng:h drive; no r) au:omatic gain controlBodenseewerk, 400s Double 0.33 m Czerny- 1800 1.3 0.2 190-860 Meter;x50 Auto zero; flame Perkin-Elmer 8 Co.GmbH, Turner and x 0.2 ignition; integration Postfach 1120, D-7770 Uberlingen, 400 Double 0.33 m Czerny- 1800 1.3 0.2 190-860 Digital; x50 As Model 400s plus auto West Germany Turner and xO.2 concn.; curve correction; BCD outlet 410. Double Double grating 2800/ I / 0.17/ 190-860 Digital As Model 400.but with double-grating monochromator Czerny-Turner 1800 1.6 0.27 420* Double 0.33 m Czerny- 1800 1.3 0.2 190--860 Digital As Model 400, but with keyboard operation; linearisation with up to 3 standards; EIA RS-232C data outlet Turner microcomputer electronic 0.17/ 190-860 Digital As Model 420, but with 1/ 0.27 double-grating Czerny-Turner 1800 1.6 430' Double Double grating 2800/ monochromator Pye Unicam Ltd..SP 191 Single Ebert 1200 3.3 0.2 190-850 Digital; York Street, x 0.1-25 Cambridge CB1 2PX, England SP 192' Single Ebert 1200 3.3 0.2 190-850 Digital; X 0.1-25 SP 2900' Double Ebert SP 1950 Double Ebert 1200 3.3 1800 2.2 0.2 190-4350 Digital; X 0.1-50 0.1 190-4350 Digital; x20 and x 0.1 SP 1900 Double Ebert 1800 2.2 0.1 190-850 Digital; x20 and x 0.1 4-lamp magazine; auto zero; integration; curve correction; emission As SP 191.but simultaneous background- facility added 4-lamp magazine; auto zero; integration; curve correction, with average calibration facility; peak height measurement with timer; peak area; emission; simultaneous background correction as accessory Auto zero and ianition: integration; C U N ~ correction As Model 1950 plus E-lamp turret Rank-Hilger.Atomspek Single Czerny-Turner 1200 2.6 0.1 190-850 Digital &!amp turret: auto zero Westwood Industrial Estate. H 1550 and flame ignition; Ramsgate Road, curve correction; Margate, Kent, CT9 4JL, integration; background England correciion optional (continued) 'New equipment since publication of Volume 5P Tablc B COMMERCIALLY AVAILABLE ATOMIC ABSORPTION SPECTROMETERS - con/inud a: Single/ Grating Reciprocal Supplier Model double Monochromator lines dispersion/ Resolution Resdout; scale Other features beam oer mm nmoermm Inm range/nm exDansion (continued) Atomspe? Single Czerny-Turner 1200 2.6 0.1 190-850 Digital 6-lamp turret; auto zero; H 1551 flame ignition; integration; inbuilt background correction; flame emission Varian Techtron Pty.. 1100 Single 0.25 m Czerny- 1276 2.8 0.2 185-900 Meter/Digital; 4-lamp turret; auto zero; 679 Springvale Road, Turner x 0.3-50 integration; curve Mulgrave, Vic. 3170, correction; peak reader; Australia f/8 aperture; optional automatic gas-box Varian Associates Ltd., Instrument Group, AA175' Single 0.25 m Czerny- 1200 2.8 0.2 185-900 Digital; 4-lamp turret; reflective 28, Manor Road, Turner xO.3-50 optics with quartz Walton on Thames, overcoat; auto zero; Surrey, England integration; curve Varian Instrument Div., optional automatic 611 Hansen Way, gas-box; simultaneous Palo Alto. background corrector, Calif. 94303, U.S.A. and calculator interface AA6 Single; 0.51 m Ebert 638 3.3 0.05 185-1000 Digital; Modular construction; correction; peak reader; dual x0.3-50 auto curve correction; channel f/lO aperture; optional automatic gas-box, simultaneous background corrector, and calculator interface b VEB Carl Zeias Jena, AAS It Single 0.5 m Ebert 1300 1.5 Continuously 190-820 Meter;xlO 4-lamp turret; auto zero: 69 Jena.adjustable single- or triple-pass 3. Carl-Zeiss-Gt. 1. optics: continuously 0 German Democratic Republic adjustable slit B C Z Instruments Ltd..Herts WD h Beckman Instruments, 485t Double Littrow 1200 2.7 0.2 19IL-860 Meter; x 5 0 Single- or triple-pass B 2500 Harbor Boulevard, optics; automatic filter $ 2 Fullerton, Calif. 92634. selection U.S.A. 495t Double Littrow 1200 2.7 0.2 190-4360 Digital; x 100 As model 485 0Corning Ltd., Halstead, EEL 1407 Single 0.25 m modified Ebert-Fastie 1180 3.5 Non-linear Single lamp mount; meter CO9 ZDX, England EEL 240t Single As EEL 140 1180 3.5 - - Meter 4-lamp turret; integration Essex, ~~~~ ~ ~~ ~ ~ ~~ ~~~~ ~ Diano Corporation, Multispect Single Double-grating 1200 1.5 0.2 190-800 Meter; 3-lamp turret with 3 P.O.Box 346 0.25 m modified x 10 stabilized power supplies; 75, Forbes Boulevard, Czerny-Turner 4-way gas control: % T Mansfield, Ma. 02048, abs. or concn. readout U.S.A. Optica S.A.S.. Via Gargano 21, 20139 Milano, Italy 6000t Single 0.35 m Ebert - - - - Digital: x 50 Auto filter insertion; auto concn.: integration; flame temp. regulation: prefocussed water-cooled hollow-cathode lamps available - - - - Seiko Instruments, SAS 721t Single - - Japan SAS 740t Double; - - - - _ _ Microcomputer and line Tokyo, dual printer channel Shimadzu-Seisakusho Ltd., AA-610.3 Single 1 Nishinokyo-Kuwabaracho, Nakagyo-ku, Kyoto 604, Japan AA-620t Single AA-650t Double - 190-900 Meter: x 1 0 Czerny-Turner 1500 1.9 Czerny-Turner 1500 1.9 - 190-900 Meter; x 10 190-900 Digital; x180 - Czerny-Turner 1200 1.9 Wavelength drive; two lamp holders Wavelength drive: two lamp holders; auto ignition: flame monitor; gas pressure monitor Wavelength drive: two lamp holders; auto zero; integration; curvature correction; background correction; peak detector; auto ignition: flame monitor: gas pressure monitor 2.5 0.05 193-300 Digital 4-lamp turret: 2 Carl Zeiss, FMD 2 t Single Ebert 600 7082 Oberkochen, stabilized power supplies; Wurttemberg, curve correction; auto West Germany zero: optional auto calibrate and background correction tNo up-to-date information was available on these instruments when this table was compiled.'New eqtlipment since publication of Volume 5 P Wv, Table C COMMERCIALLY AVAILABLE ELECTROTHERMAL ATOMIZERS 0 Sensltlvlty for 1% abs. (s.)/pg cu Si Supplier Model Type M;ol;-wF Control Detection limit (d.l.)/pg Special features Eaird-Atomic Ltd..A3470 Graphite rod Warner Drive, Springwood Industrial Estate, Rayne Road, Braintree, Essex CM7 7YL, England. 50 Programmable; d.1. 5 drv. ash (2 stages). atomize. Max. temp. over 3000 "C d.1. 60 Fits most AA spectrometers; air cooled; uses mains power; inert-gas shielding; pyrolytic graphite coating for rods in situ; rapid interchange between flame and electrothermal methods.Beckman Instruments GmbH. 1271 8 Munich 40, Frankfurter Ring 115, West Germany Graphite furnace 100 Programmable; d.1. 4 d.1. 10 Water-cooled; inert-gas dry, ash. atomize, (100 p l ) (100 PI) shielding. Safety feature burn off. Max. for failure of water or temp. 3100 "C purge gas, gas stop; fits Beckman and Pye Unicam instruments Instrumentation Laboratory Inc., 113 Hartwell Avenue, Lexington, Mass. 02173, U.S.A. Instrumentation Laboratory (UK) Ltd.. Station House, Stamford New Road, Altringham, Cheshire, England. Jarrell-Ash Division, Fisher Scientific Co., 590 Lincoln Street, Waltham. 555 Graphite furnace 100 Programmable; d.1. 0.8 d.1. 10 Controlled-temperature six stages, furnace, using feedback ramp or step.from a tungsten Max. temp. temperature sensor; true 3500 "C temperature readout; safety interlock system; automatic cell door: automatic cleaning; cell pressurisation; convenient solid-sampling capacity b using microboats 3 % Y \ MTAP Tantalum strip 50 Programmable; d.1. 2 - Fits most AA 2. dry, ash, atomize. (50 PI) spectrometers; inert-gas Max. temp. and hydrogen shielding ' 2400 "C b s Mass. 02154, U.S.A. FLA 100 Graphite furnace 50 Programmable; d.1. 10 d.1. 50 Fils most AA dry, ash, atomize. spectrometers: inert-gas 3 s. 50 s. 50 shielding. but an air ash G' With ramping and flash possible- tl B 5 2 atomization _____.__ S. 8 J. Juniper & Co., 110 Graphite furnace 50 Programmable; s. 30 - Water-cooled; inert-gas $ 7 Potter Street, dry, ash, atomize, (10 p l ) shielding; all programme Harlow, Essex, burn out.Max. stages cover full England. temp. 3500 "C temperature rangeSoectronic Services E-& J Brereton. ' 4 White Rose Way, Garforth, Leeds LS25 2EF, England. Perkin-Elmer Corp., HGA 2100 Graphite furnace 100 Programmable; d.1. 2 d.1. 50 Main Avenue, dry, ash, atomize. d.1. 1 d.1. 10 Norwalk, Max. temp. 1 pyro-coated 1 { pyro-coated Conn. 06856. 2800 "C. tube tube U.S.A. Ramp accessory provides linear-type r a m temoerature Pyro-coating accessory available for in situ preparation of pyro-coated tubes; AS-1 automatic sampler available for pre-use automatic insertion of up to 30 samples into the HGA. with automatic triggering of HGA and instrument read cycle Bodenseewerk HGA 76 Graphite furnace 100 Programmable; d.1. 1 d.1. 10 Fits Perkin-Elmer and water-cooled; inert-gas Perkin-Elmer & Co. GmbH. dry. ash (2). (100 rl) Postfach 1120. atomize. Max. D-7770 Uberlingen, temp. 2700 "C shielding; permits ramp West Germany ashing; gas stop operation; closed system; safety feature for failure of water or purge gas (100 PI) Zeiss AA spectrometers: inciease in all 3 cycles plus auto high temperature at end of programme.AS-1 * Auto sampler for 100 Automatic as with as with Fits all Perkin-Elmer AA graphite furnace sampling of up to HGA-76 HGA-76 spectrometers with HGA 30 samples once or up to 9 times each Pye Unicam Ltd., York Street. Cambridge CB1 2PX, England. SP9-01 Graphite furnace 50 Programmable; s. 44 dry, ash. atomize, tube clean, tube blank, with cancel and delay stages.Max. temp. 3000 "C - Water-cooled; inert-gas shielding; safety feature for failure of water; tube life indicator and remote recorder control for 1, 2, 3, or all phases Rank Hilger. H1975/ Graphite furnace Westwood Industrial Estate, FA256 Ramsgate Road. Margate, Kent CT9 4JL, England. Varian Techtron Pty. Ltd.. CRA 90 Graphite furnace 679 Springvale Road, (graphite tube), Mulgrove, Vic. 3171, threaded Australia. graphite furnace. graphite cup 100 Programmable: s. 50 - Water-cooled; inert-gas dry, ash. wait, shielding; background atomize. Max. correction when fitted to temp. 2600 'C Atomspek H 1550 25 Programmable; 4 80 Fits most AA dry, ash, atomize. (5 PI) (5 PI) spectrometers; water- Max. temp. cooled; inert-gas shielding 3000 "C and hydrogen flame option; automatic ramp- hold atomization; pyrolytic graphite coating on cup and tubes *New equipment since publication of volume 5Table C COMMERClALLY AVAILABLE ELECTROTHERMAL ATOMTZERS-- co,iiiuu( d -- Sensitivity for 1% abr. (s.)/pg cu Si Special features M,”durem;,le Con,rol Detection limit (d.l.)/pg Suppliers Model Type ul N Barnes Engineering Co., Glomaxt Tantalum strip 30 Commerce Road, Stamfcrd. Conn. 06902, U.S.A. 50 Programmable; d.1. 10 - dry, ash, atomize. (50 $1) burn off. Max. temp. 2400 ‘C Fits most AA spectrometers; air-cooled; inert-gas and hydrogen shielding Water-cooled; inert-gas shielding Gptica S.A.S., CAT 6 t Tantalum strip 50 Programmable; d.1. 10 - Via Gargano 21, dry, ash, atomize. (50 01) 20139 Milano, Italy. Shimadzu-Seisakusho Ltd., GFA-Pt Graphite furnace 50 Programmable; d.1. 5 - Current stabilised to 1 Nishinokyo-Kuwabaracho, current stabilised, obtain highly reproducible Nakagyo-ku, Kyoto 604, dry, ash, atomize. results Japan. Max. temp. 3000 “C b b tNo up-to-date information was available for these instruments when this table was compiled.Part I: Fundamentals aiid Imtrumentatioii 53 fluorescence line and the other a nearby but non-fluorescing ljne. The out-of-balance signal from these two channels is the fluorescent signal. In a non-dispersive instrument dedicated to the determination of Hg (1270) the problem of monochromation is overcome by using a low-pressure Hg lamp under conditions such that 96% of the radiant energy occurs at 184.9 and 253.7nm and stray light is minimised by means of diaphragms. More elaborate instruments, incorporating computers, have been reported for combined multi-element AE and AF analysis. In one (612, 642, 1159), a scanning monochromator with either pulsed HCLs or tunable laser and a Ar/H, flame or low power current-regulated arc is used. Up to 25 elements in emission and 8 in fluorescence can be analysed, and the measurement of 5 elements in blood requires approximately 100 s per sample. The other system (62) uses an image-dissector echelle spectrometer, with a 1.6 kW Xe arc lamp as the excitation source and a flame atomizer.

ISSN:0306-1353    

DOI:10.1039/AA9760600033

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

7 Fundamental Studies 7.1 ATOMIC PROCESSES Resonance-line AF lifetimes of the order of a few ps have been meastired with a two- chopper technique (207). For atomic vapours produced electrothermally, Ar was found to be the optimum atmosphere for minimising the quenching of the fluorescence of Ag, Bi, Cd, and Tl, and He for Hg (761). AA techniques were used to determine oscillator strengths in Br, I, (610) and Hg (611).In the former case, earlier work was corrected for the effects of nuclear hyperfine structure; problems associated with the one- and two-source version? of the line absorption method for determining oscillator strength, including excess kinetic energy, poor collimation, and self reversal, were discussed. Spectral lines of light elements cmittcd from space-charge zones in gas discharges are often Doppler-broadened, due to thc presence of groups of fast atoms.The cffect of particle acceleration in hollow-cathode dis- charges on spectrochemical accuracy has been studied by Sternberg (1451). Spectral overlap is an interference which occurs relatively rarely in AAS. A list has been prepared of observed and predicted overlaps (321), and the effect of increasing amounts of In (252.137 nm) on Co absorption at 252.136 nm was studicd.Other workers (325) habc examined the overlap of Cr 290.905nm on 0 s 290.906nm and Tb 285.214nm on Mg 285.213 nm. The use of the two-line method for the measurement of spectroscopic tempxa- tures has been reviewcd by Schmidt (362), and guidelines are proposed for the choice and measurement of suitable spectral lines.A method has been described by Reif et al. (1571) whereby a non-calibrated W filament lamp may be used in spectroscopic temperature measurements. Other references of interest - Development of spectrochemistry: 1509. Electron excitation of V.U.V. lines of T1: 1237. Oscillator strengths of Tm lines: 1236. 7.2 SIGNAL TO NOISE RATIO The theory of SNR may be used to study, characterisc, and optimise the precision of atomic spectrometric techniques. The sources of noise include: emission of the anal) te and background, which produces shot and flicker noise, shot and flicker noise generated by thc absorption light source, analyte and absorption-ccll shot and flicker noise, and dark-current and amplifier read-out noise.These sources of noise have been investigated theoretically (555) and experimentally (100, 937, 947).In the case of flame AA it was concluded that analyte absorption flicker and flame transmission flicker are the dominant noise sources The latter becomes progressively worse as the wavelength of analysis becomes shorter, where the flame transmission decreases. With an intense light source, the effect of source flicker noise may be reduced by double-beam operation (922).It is claimed (554) that the SNR of experimental data can often be improved by an adaptiL c electronic filter. Digital signal processing was used to enhance narrow-band components, such as the spectral line, abovc the broad-band input, such as the noise, while interferences. such as mains-frequency pick-up, may be filtered out and substracted from the total input.When the required signal can be selzctivcly modulated, correlation techniques such as lock-in amplification, boxcar integiation, signal averaging, and digital smoothing may be used (551, 934) to give SNR cnhancement. Progress in these methods of signal cnhancement benefits considerably from ricw integrated-circuit technology.Signal modulation is an important tool in efforts to d e i elop multi-element non-dispersive systems. In trials or a non-dispersive AF system ( I 43) incorporating a C1, filter, a separated air/C,H, flamc, and a solar-blind photomultiplier, the 54Part I : Fundamentals and Instrumentation 55 limit to precision was set by the shot noise due to the analyte and flame background.Though there was no improvement in SNR by modulation, it was concluded that multi- elcmcnt analysis was feasible. Other workers (579) have used a dedicated digital Fourier analyser to study the noise spectra of several analytical radiation sources. They found that a white-noise spectrum characterised the emissions from a W lamp and microwave plasma source. Premixed burners yielded an approximately white spectrum, with a weak inverse frequency component becoming apparent below 100 kHz.A Meker burner exhibited a stronger 1 if component, and both arcs and sparks show strong 1 /f noise below 1 kHz, with a superimposed mains-frequency modulation and a high harmonic content. A pulsed r.f. plasma source gave a dense non-white spectrum, and all burners exhibited some resonance frequencies.Other workers (553) have examined the origin of multiplicative noise in local regions of N,OiC,H, flames. Winefordner and colleagues (142, 903, 943) have carried out a comprehensive theoretical study of the SNR for single-channel methods, sequential and multiplex scanning, and multi-channel methods in optical spectroscopy. They compared four major types of spectrometers: linear scan in sequence, slew scan in sequence, multi- channel simultaneous multi-channel multiplex, using both Fourier- and Hadamard-transform techniques.They conclude that the multi-channel approach is the best, but the scqucntial slew scan is nearly as good for simple spectra in the U.V. and visible region. 7.3 PRECISION, ACCURACY, STANDARDS, AND REFERENCE MATERIALS The abstract of a paper which was to have been presented at the 3rd Federation of Analytical Chemistry and Spectroscopy Societies’ Meeting by the late Professor H.Kaiser, entitled ‘The Concept of Accuracy in Chemical Analysis’, summarises an aspect of spcctro- chemistry in which he was deeply interested and to which he made many contributions. The paper (1081) was to be a comprehensive examination of the concept of accuracy in chemical analysis, under the headings ‘Terminology’, ‘Classification of Errors’, ‘Propagation of Errors’, ‘The Importance of Selectivity and Linearization of Analytical Function with Respect to Accuracy’, and ‘Strategy to Ensure Accuracy’.The point is made that “much effort is spent in vain in intercomparison analyses, because the planning i s not made on the firm base of theory of chemical analytics”.The ‘range’ of an analytical method is a somewhat indeterminate parameter. Butler (791) suggests a definition based on the precision of the method, e.g. that concentration range within which the precision is equal to or better than 3 times the best precision obtainable; an alternative, widely used approach i s based on the linearity of the response curve.The precision of measurements at high concentrations can often be improved by combining zero suppression with scale expansion. The application of this technique to alloy analysis has been reported (845). When dealing with real samples, analysts are often required to perform high-precision analysis on both major and minor constituents in the same sample, but a wide analytical range is often difficult to achieve.A statistical procedure has been developed (1083) for maximising precision and minimising the risk of gross error with procedures based on thc use of a calibration curve. In AA analysis the effects of integration time, precision versus absorbance level, methods of calibration, optimum working ranges, and a utilization of various automatic correction programmes have been examined (101).By using a micro- processor, the speed, convenience, calibration, and data handling were improved, and curvature-correction programs allowed signals of greater than one absorbance unit to be measured without sacrificing precision, Increases in integration time beyond 9 s did not lead to significantly improved precision.It has been pointed out (163) that calibration by mcans of the standard-addition procedure can be achieved by using a single solution to which56 Analytical Atomic Spcciroscopy aliquots of a solution containing the element are added stepwise. The errors arising from changes in the dilution factor due to use of the solution were corrected by a computer program.To maintain the speed and accuracy of AA analysis of solid samples it is recommended that powdered working samples be included in the scheme of analysis, to provide quality control and thereby give early warning of possible faults in the analytical scheme (176). As certain errors are peculiar to individual samples, they cannot be detected by the usual quality-control checks, such as spiked samples, replicas, and standard reference material.Gold has been used as an internal standard for the early detection of analytical errors (561). The advantage of this procedure lies in its direct association with the sample under test, such that if there is an inadequate recovery of the Au, the amounts of other elements of interest were likely to be in error, due to faults in the sample-preparation procedure. Means of calibrating methods for trace analysis have been reviewed (235), and include the use of certified standard samples, independently analysed samples, standard-addition methods, synthetic standards, and the use of fundamental relationships, e.g.The Law of Mass Action and Beer’s Law. Berman (1348) has noted that the linearity of response, which is practically essential when using the standard-addition method, cannot be assumed to be always true, even at low concentrations, The occasional departure from linearity at low concentration has led Ingle and Wilson (925) to suggest that the extrapolated blank rather than the signal measured with a blank solution should be used when calculating detection limits.To overcome the problem inherent in the dry mixing of spectrochemical standards, a solution method for preparing standard for elements in Ga,O, has been described (1085). The preparation and certification of reference materials have been reported from several sources, as follows: precious-metal ores (220), oxides in lithium tetraborate /lanthanum tetraborate flux (469), geochemical materials (793), T1 in Cd (536), trace elements in metals by ion implantation (805), alloy steels (263), ferrous and non-ferrous alloys and minerals, and refractory materials (210), elements in botanical material (1 105), museum samples (1063), and trace elements in water (1386).The valuable work of the National Bureau of Standards, Washington, U.S.A.in the preparation of well-characterised reference material has been reviewed by members of its staff (914, 959, 1103). Tables D, E, F, and G contain information on suppliers of spectrographic standards, spectrographic graphite electrodes, standard metal solutions and reagents for AAS, and organo-metallic compounds. Other references of interest - Evaluation of detection limits in AAS: 1268.Specificity of analytical methods: 1344.Part I : Fundamentals and Instrumentation 57 Table D SPECTROGRAPHIC STANDARDS Suppller Aluminum Company of America, Alcoa Laboratories, Alcoa Center, Pennsylvania 15069, U.S.A. X Apex Smelting Co., 6700 Grant Avenue, Cleveland. Ohio 44105. U.S.A. X X X BNF Metals Technoloav Centre. X X Bundesanstalt fur Materialprufung (BAM), 1 Berlin 45, Unter den Eichen 87, Germany x x x X x x Bureau of Analysed Samples Ltd., Newham Hall, Newby, Middlesbrough, Cleveland TS8 9EA, England x x x x x x X X CKD Research Institute, Na Harfe 7, 190 02 Praha, Czechoslovakia x x X Comite de liaison des Industries de metaux non-ferraux de la Cornmunaute Europeenne, Boulevard de Berlaimont, 1000 Brussels, Belgium G.L. Willan Ltd..X Sheffield Works, Catcliffe, Rotherham S60 5RL x x x South Yorkshire, England X Johnson Matthey Chemicals Ltd., 74 Hatton Garden, ‘Spectromel’Q powders ‘Specpure’@ metals London EClP lAE, England Moore Boundy Hamill Ltd., Station House, Potters Bar, Herts. EN6 IAL, England Materials, National Bureau of Standards, Washington, D.C. 20234, U.S.A. x x x x x x x x x x x x x Office of Standard Reference x x x x x X x x Various other metals, including high-purity metals X Pechiney, 23 Rue Balzac, Paris 8e, France Spex industries Inc., P.O.Box 798, Metuchen, N.J. 08840, U.S.A. x x X X X X (Glen Creston, 16 Carlisle Road, London NW9 OHL, England) Zinc 8. Alliages, 34 Rue Coliange, 92307 Lavallois-Perret, France X58 Artalytical Atomic Spectroscopy Table E SPECTROGRAPHIC GRAPHITE ELECTRODES 1 Baird-Atomic, Inc., 125 Middlesex Turnpike, Bedford, Mass. 01730, U.S.A. 2 Carbon Products Division, Union Carbide Corp., 270 Park Avenue, New York, N.Y. 10017, U S A . (ARL Ltd., Wingate Road, Luton, Beds., England) 3 General Graphites, Inc., First and Monroe Street, Bay City, Mich. 48706, U.S.A. 4 Johnson Matthey Chemicals Ltd., 74 Hatton Garden, London EClP lAE, England 5 Le Carbone (GB) Ltd., Portslade, Sussex, England 6 Le Carbone Lorraine, 45 Rue des Acacias, 75821 Paris, Francc 7 Met-Bay, Inc., 900 Harrison Street, Bay City, Mich. 48706, U.S.A. 8 Poco Graphite, Inc., P.O. Box 2121, Decatur, Texas 76234, U.S.A. 9 Ringsdos-Werke GmbH, 53 Bonn-Bad Godesberg, West Germany (Mining it Chemical Products Ltd., Alperton, Wembley, Middlesex HA0 4PE, England) I0 Spex Industries, Inc., 3880 Park Avenue, Metuchen, NJ 08840, U.S.A.(Glen Creston, 16 Carlisle Road, London NW9 OHL, England) I 1 Ultra Carbon Corp., P.O. Box 747, Bay City, Mich. 48706, U.S.A. (Hepden & Son Ltd., Spectrum House, Alderton Crescent, London NW4, England) Table F STANDARD METAL SOLUTIONS (MS) AND REAGENTS (R) FOR AAS 1 Aldrich Chemical Co., Inc., 940 W.St. Paul Avenue, Milwaukee. Wis. 53233, U.S.A. (R) 2 J. T. Baker Chemical Co., 222 Red School Lane, Phillipsburg, N.J. 08865, U.S.A. (MS, R) 3 Barnes Engineering Co., 30 Commerce Road, Stamford, Conn 06902, U.S.A. (MS) 4 B.D.H. Chemicals Ltd., Poole, Dorset BH12 4NN, England (MS, R) 5 Bio-Rad Laboratories, 32nd and Griffin Avenucs, Richmond, Calif. 94804, U.S.A.(MS) 6 Carlo Erba, Divisione Chimica Industriale, Via C . Imbonati 24, 20159 Milano, Italy (MS) 7 Eastman Organic Chemicals, Eastman Kodak Co., 343 State Street. Rochester, N.Y. 14650, U S A . (R) 8 Fisons Scientific Apparatus Ltd., Bishop Meadow Road, Loughborough, Lcics. LEI 1 ORG, England (MS, R) 9 Harleco, Div. of American Hospital Supply Corp., 60th and Woodland Avenucs, Philadelphia, Pa. 19143, U.S.A. (MS) 10 Hopkin & Williams Ltd., P.O. Box 1, Romford, Essex RMl IHA, England (MS, R) 11 V. A. Howe & Co. Ltd., 88 Peterborough Road, London SW6, England (MS) 12 Instrumentation Laboratory Inc., 113 Hartwell Avenue, Lexington, Mass. 02173, U.S.A. (MS) 13 Johnson Matthey Chemicals Ltd., 74 Hatton Garden, London EClP IAE, England (R) 14 Koch-Light Laboratories Ltd., Colnbrook, Bucks., England (R) (Anderman & Co.Ltd., Battlebudge House, 87-95 Tooley Street, London SEl, England) IS May & Baker Ltd., Dagenham, Essex RMlO 7XS, England (R) I6 E. Merck, D 61 Darmstadt, West Germany (R) 17 Spex Industries Inc., 3880 Park Avenue, Metuchen, N.J. 08480, U.S.A. (MS) 18 Ventron Corp., Alfa Products, 44 Congress Street, Beverly, Mass, 01915, U.S.A.(MS) (Glen Creston, 16 Carlisle Road, London NW9 OHL, England)Part I: Fundamentals and Instrumentation 59 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Table G ORGANOMETALLIC STANDARDS Angstrom Inc., P.O. Box 248, Belleville, Mich. 4811 1, U.S.A. Baird-Atomic Inc., 125 Middlesex Turnpike, Bedford, Mass. 01730, U.S.A. J. T. Baker Chemical Co., 222 Red School Lane, Phillipsburg, N.J. 08865, U.S.A. B.D.H. Chemicals Ltd., Poole, Dorset BH12 4NN, England Messrs Burt and Harvey Ltd., Brettenham House, Lancaster Place, Strand, London WC2, England Carlo Erba, Divisione Chemica Industriale, Via C. Imbonati 24, 20159 Milano, Italy Conostan Div., Continental Oil Co., P.O. Drawer 1267, Ponca City, Okla. 74601, U.S.A. Durham Raw Materials Ltd., 1-4 Great Tower Street, London EC3R 5AB, England Eastman Organic Chemicals, Eastman Kodak Co., 343 State Street, Rochester, N.Y. 14650, U.S.A. Hopkin and Williams Ltd., P.O. Box 1, Romford, Essex RM1 IHA, England E. Merck, D 61 Darmstadt, West Germany Moore Eoundy Hamill Ltd., Station House, Potters Bar, Herts. EM6 IAL, England National Spectrographic Laboratories Inc., 19500 South Miles Road, Cleveland, Ohio 44128, U.S.A. Office of Standard Reference Materials, National Bureau of Standards, Washington, D.C. 20234, U.S.A. Research Organic /Inorganic Chemical Corp., 11686 Sheldon Street, Sun Valley, Calif. 91352, U.S.A. Ventron Corp., Alfa Products, 44 Congress Street, Beverly, Mass. 01915, U.S.A.

ISSN:0306-1353    

DOI:10.1039/AA9760600054

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

PART I1 METHODOLOGYIntroduction In Part 11, the t e r n Methodology covers all aspects of the application of the techniques and instrumentation of AAS, AES, and AFS to chemical analysis. The format adopted for previous volumes has been retained, with the subject matter treated under the two principal headings of (1) General Information, covering new methods, inter-laboratory comparisons, and referee methods, and (2) Applications, where specific methods of analysis are reviewed and tabulated. The classification of the range of applications into nine main fields of analysis also follows the established pattern. Some duplication of entries may be found in instances where a method is relevant to more than one section. 63

ISSN:0306-1353    

DOI:10.1039/AA9760600061

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

EXPLANATION OF THE TABLES Each of the Applications sections, 2.1 to 2.9, is accompanied by a table which summarises 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 accorn- 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 systems.In many cases, sufficient detail is given for the analytical procedure to be followed; absence 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 Xinm MATRIX CONCENTRATION TECH. ANALYTE SAMPLE TREATMENT ATOMIZATION REF. The elements determined are listed in alphabetical order of chemical symbol, except that, for space economy, multi-element applications (5 elements or more) are 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 % or pgg-1 for solids and mgl-1 or pgml-1 for liquids. The atomic spectroscopy technique is indicated by A (absorp- tion), E (emission2 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). A brief indication is given of the sample pre-treatment required to produce the analyte. 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@) of the author@), with address. 63

ISSN:0306-1353    

DOI:10.1039/AA9760600063

出版商:RSC    

年代:1976

数据来源: RSC

摘要:

1 General Information 1.1 NEW METHODS 1.1.1 Introduction This section describes novel methods of analysis which are considered to be of sufficient general interest to merit discussion here rather than in a subsequent section on applications. Inevitably, the number of genuinely new methods appearing in the literature of any one year is limited. However, the criterion of novelty is applied less rigorously to papers where the authors have made a detailed study of experimental parameters of widespread interest. 1.1.2 Sample Preparation and Extraction Techniques Another detailed scheme for the determination of ambient forms of Hg in air (ARAAS, 1975, 5, p.54) has been described (271). The air stream was passed sequentially through a Millipore filter (0.25 pm), which retained the particulate Hg, Ag-wool, which retained Hg vapour, finely cut Au, which removed simple Hg compounds, and finally through a catalytic combustion unit containing Au followed by a Au collector, which retained all other forms of Hg. The Hg was liberated from each section by heating, and determined using cold-vapour AAS.A convenient procedure for the digestion of biological and geological samples prior to analysis for Hg, using H,SO, /HNO, / HCl/ KMnO, / K,S,O,, has been described (36).A particular advantage of this method is that any HgS in the sample is completely recovered. The Analytical Methods Committee of the Chemical Society havc published recommended methods for the wet oxidation of fats, oils, foods, syrups, and fruit concentrates using 50% m/m H,O, and H,SO, (34).It was particularly stressed that under no circumstances should any form of pressurised vessel be used for this type of dissolution. The addition of 10~gml-1 Au as an internal standard in the analysis of sediment samples by AAS has been found (561) to improve the reproducibility of this type of analysis. Various organic solvents have been investigated for the extraction of the vanadium-cupferron complex (288); MIBK, amyl acetate, and di-n-butyl ether were found to give the best sensitivity and precision.A similar type of study (1374) has determined the optimum solvent-extraction conditions for Co, Cr, Mo, and Zn. Cresser and Hargitt (41) have determined both Cr(II1) and Cr(V1) in a solution, using AAS with a N,O/C,H, flame. The method utilised an anion-exchange resin to remove Cr(V1) from the solution.A novel approach of sampling metal alloys has been reported (315). The sample is immersed in water and a spark discharge applied to it: this results in the formation of a stable colloidal dispersion, which can be directly nebulised into a flame (See Section 2.3.2). Some Soviet workers (756) have reported the use of a dispersionless AAS instrument with a photoresistor detector (ARAAS, 1974, 4, ref. 391). The samples were concentrated by passage through a cation-exchange resin, and the subsequent eluates were then nebulised. Detection limits of 1 ng ml-1 or better were claimed for Cd, Cu, Pb, and Zn. Two rapidly operated mechanical devices for withdrawing fixed weights of powdered samples have been described (972).One device gave sample weights of 10-50mg with a typical RSD of 0.03, whilst the other device gave sample weights of 1-5 mg with an RSD of about 0.1. Contamination during the sample-preparation stage is a continually recurring problem in trace analysis that requires careful attention. A simple sub-boiling distillation apparatus constructed from polypropylene has been described (599) and used to prepare ultra-pure 64Part I I : Methodology 65 HCl, HF, and H,O.An interesting point that arose from this study was that polypropylene containers were found to be better than fluorocarbon containers for the storage of various acids. The latter material was a persistent source of trace levels of Cr, Fe, Mn, and Ni. Zief and Horvath (1051) have given a comprehensive review of the contamination problem in trace analysis.Recommendations for constructing a contamination-free laboratory were given, including the choice of particulate air filters, paints, and wall and bench surface materials. Contamination from the analyst can represent a serious source of error. Cosmetics constitute a particular hazard in this respect.One very sobering finding that emerged from this study was that, in order to clean thoroughly a PTFE bottle for ultra-trace analysis, it took 21 days and cost $100. 1.1.3 Indirect Methods The wide range of substances that can be determined indirectly using AAS continues to increase at a surprising rate. Tin(I1) has been determined down to 0.1 ngml-1 simply by adding a small amount of 0.1 E.rg ml-l Hg solution to the analyte and monitoring the liberated Hg, using a conven- tional cold-vapour system (1062).(ARAAS, 1974, 4, ref. 1019). Cyanide has been deter- mined by adding Cu(1) chloride to the neutralised analyte solution and measuring the solubilised copper (51 5). High concentrations of halides and thiocyanate interfere. Tungsten has been determined by adding Pb acetate to the test solution, removing the precipitated Pb tungstate, and determining the excess Pb (275).Al, Cr(VI), Fe(II), Mn, Mo(VI), and V(V) interfered. Huber and Sand (913) have extended their inhibition titration method to determine sulphate down to 0.2 pg ml-1 by titrating the sample (after passage through a cation-exchange column in the H form) with a Ca solution and monitoring the emission at 620 nm in an air/H, flame.The presence of silicate was essential to obtain the sharp-peaked titration curve. It was also possible to determine phosphate and silicate simultaneously (1500). A similar type of technique has been applied to the determination of Al, Ca, La, and Sr (1557). Indirect methods have also been reported for organic bases such as alkaloids (684), mixtures of complexones such as EDTA and closely related compounds (584), benzyl- pencillin (280), fluorine in organic compounds (284), mixtures of halides (556), and condensed polyphosphates.(236). These methods are more fully discussed in Section 2.2. Other methods have been described for A1 (1537), perrhenate (1567), biuret (699), (see Section 2.6.2), and for both anionic (1 303) and cationic surfactants (645) (see Section 2.7). 1.1.4 Nebulization, Vaporization, and Atomization Techniques A high-capacitance condenser (2.2 X 105 pF), discharging through a 0.25 mm W filament, has been used as a means of vaporising samples into a lOOW, 2450MHz microwave-induced plasma (300). The addition of KCl to the samples (final concentration 0.01 M) enhanced the signal and minimised many inter-elemental effects.Detection limits varied from 10 ng ml-1 for As to 0.3 ng ml-1 for Cd. Kirkbright and Adams (1289) have determined sulphur directly, using a graphite tube furnace and nitrogen-purged optics, Sensitivity and detection limit of 0.42 and 2 ng respec- tively were obtained at the 180.7 nm line; the calibration graph was not linear above 0.3 absorbance units.The application of furnace atomic emission spectroscopy measurements from the optical axis of a graphite tube furnace (ARAAS, 1975, 5, p.19) has been reported by Epstein, Rains, and O’Haver (106). The problem of black-body emission from the graphite tube was minimised by using a wavelength-modulation technique. (Snelleman, W.,66 Analytical Atomic Spectroscopy et al., Anal.Chem., 1970, 42, 394). The detection limits obtained for Al, Ba, Cu, Cr, and Na were similar to those obtained using conventional absorption measurements, and although calibration graphs exhibited considerable non-linearity, they were useable over several orders of magnitude. If multi-element detection could be achieved, this technique would appear to have a promising future. 1.2 ANALYTICAL PARAMETERS 1.2.1 Accuracy, Precision, and Detection Limits A comprehensive examination of the concept of accuracy in chemical analysis has been made by Kaiser (1081); see Section 7.3. Ingle and Bower (100, 739) have investigated the factors that limit the precision of flame AAS measurements (ARAAS, 1974, 4, p.22). For Cu, it was concluded that noise sources associated with the lamp (i.e.signal shot noise and source flicker noise), with the detection system (i.e. amplifier-readout noise, dark-current noise), and with the background emission of the flame are not the limiting factors. Noise due to fluctuations in the trans- mission properties of the flame limits the precision at low absorbances (i.e. A<0.05), whilst fluctuations in the absorption properties of the analyte limit reproducibility for moderate and high absorbances.Optimum measurement conditions have been established for the determination of seven noble metals (1330). The use of a microprocessor to achieve auto- matic curve correction enabled flame measurements to be made at high absorbance levels (i.e. A>1.0) without loss of precision (101).It was also shown that the optimum integration period was about 15s. When detection limits are calculated, it is essential to ensure that the calibration graph is linear near to the calculated detection-limit level. Factors affecting the precision of electrothermal atomization using a graphite tube furnace have been considered (131, 1325), and it was concluded that the manual pipetting of the sample into the furnace constituted the largest error.Abercrombie and Silvester (129) have demonstrated that pure solution detection limits quoted for the r.f. ICP often bear no relation to those obtained when using actual samples. The determination of A1 in some seed samples was used as an example. It was also shown that, although the RSD was less than 0.05, the data were inaccurate by 2 6 5 0 % .The inac- curacies could be reduced considerably if a computer-calculated interference correction for stray light was made, 1.2.2 Interferences 1.2.2.1 Flames. A large number of papers continue to appear, and it is now obvious that AAS is much less of an interference-free technique than is generally supposed. Miller (1178) has reported large enhancements (up to 500%) of Na and K absorbance in the presence of I .5% m/V calcium halogenophosphate in the air/C,H, flame.(All solutions contained 1000 pg ml-1 Cs). This is a surprising result, as it is generally assumed that all Na and K species are completely atomised in the air/C,H, flame). Various W compounds resulted in severe suppression of the Na and K signals. Ammonium halides (10-44.5M) have been found to cause chemical interference effects in the determination of a large number of elements in the air/C,H, flame (277) which could be minimised by avoiding the use of fuel-rich flames and taking measurements higher up the flame.A study of the effects of various ligands (EDTA, CN, (COOH), etc.) upon the determination of Co, Cr, Fe, Mn, and Ni in the air/C,H, flame (292) yielded similar conclusions.It is generally assumed that sulphate exerts a negligible interference effect with Co, Mg, and Ni determinations in the air/C,H, flame. However. Cresser and Macleod (35) have shownPart 11: Methodology 67 that for high analyte concentrations (up to 1000 pg mg-'), which require the use of less sensitive resonance lines and/or burner rotation, suppression of the signal can be observed.The degree of suppression is dependent upon the burner-nebulizer systems used, and can be overcome by using standard releasing agents (La or Sr) or simply by diluting the solution. A change in the pH of a Cr(V1) solution was found to affect the Cr response in air/C,H, and air/H, flames (6), but not in the hotter N,O/C,H, flame. Nicholas and Jones (796) have shown that the mutual interference effects of Au, Ir, Pd, Pt, and Ru can be overcome by extracting the noble metals into MIBK, using S-decyl NN'-diphenylisothiouronium sulphate in sulphuric acid media and adding Zn dibenzyl- dithiocarbamate to the MIBK phase.Sodium sulphate has been shown to be an effective releasing agent for W (927).Comprehensive studies of inter-element effects have been reported for the following elements: Eu (283), In (442), Mo (173, 738), Pd (1332), Pt (1521), Ru (1176), Si (1030), Ti (165), V (1314), and Zr (394). The problem of spectral interference has been studied by various workers (321, 325, 1173). The wavelength range of background absorption measurements can be extended well into the visible region if the conventional D, lamp is replaced by a Xe-Hg arc lamp (587); stray light, however, caused problems at wavelengths below 220 nm, and correction could not be made for Hg at 253.7 nm. 1.2.2.2 Electrothermal Atomization. An interesting paper (573) has shown that dry ashing can be a very critical factor in many analyses. The effect of MgCl, on Mn determinations using conventional dry ashing can result in a 60% enhancement at the 50 pg ml-1 Mg level and an 80% depression at the 1000 gg ml-1 Mg level.By automatically increasing the dry- ashing temperature at a rate of 349 "C min-1 and switching the power off immediately the selected final ashing temperature was reached, the large depressions in the signals caused by 1000 pg ml-l Ca and Mg (as Cl) on Cu and Mn were overcome.Andersson (483) has recommended the addition of 0.75% m/V La to overcome the interference caused by sulphates when determining Pb in manure, plant, and fertilizer samples. The addition of 500 pg ml-1 Mg has been found to minimise the interference from Na when determining As (909). Substantial inter-element effects were observed when Ru was determined in the presence of other noble metals (980).The addition of HC1 or HC10, (0.001-1% VjV) resulted in a severe depression of the T1 signal (42); this was attributed to vaporization of Tl(1). As would be expected, HNO, and H,SO, had a negligible effect on the T1 signal. It is often assumed that the presence of small amounts of miscible organic solvents has little effect on electrothermal atomization signals; it has been shown, however, that pronounced effects can be observed (135).These were attributed to a lower surface tension, causing increased spreading and an increased permeability of the solution into the graphite. Regan and Warren (23) have recommended the addition of 1% m/V ascorbic acid to potable water samples in order to eliminate interferences in the determination of Pb.The method was successful on the six samples analysed, but more work is required before it can be applied routinely. The formation of a tantalum carbide layer or the actual insertion of a Ta lining into a graphite tube furnace (443) was found to reduce chemical interferences observed in the determination of Ge, Pb, Si, and Sn. 1.2.2.3 Hydride-generation and Cold-vapour Mercury Techniques.Pierce and Brown (740) have studied inorganic interference effects in conjunction with an automated hydride- generation technique for As and Se. A number of anions and cations were found to cause a suppressive effect. The interferences were also dependent upon the order of addition of reagent; adding HCl prior to NaBH, was recommended, It was also noted that the use of68 Analytical Atomic Spectroscopy polystyrene or polythene sample cups resulted in a loss of both As and Se.Problems have been experienced with the complete wet oxidation of methylated As compounds (e.g. methylarsinic acid and dimethylarsinic acid), However, it has been found possible (939) to reduce these compounds directly, using NaBH,, to the corresponding methylarsine or dimethylarsine.Both of these compounds are of sufficient volatility to be swept directly into the N,/H, diffusion flame. If a colorimetric finish was used, erroneous results were obtained, as the various hydrides give different sensitivities for As. The use of NaBH, in place of SnCI, to liberate Hg has been reported by two groups of workers (512, 765).The main advantage is that Cu, Ag, I, S 2 0 , and CNS cause much less interference than when reduction with SnCI, is used, although the technique is not satisfactory in the presence of Au. Vitkun et al. (767) have shown that this determination can be satisfactorily carried out by using Na,SnO,/NaOH in the presences KCNS and Zn acetate. Problems with low-level Hg determinations have been thoroughly investigated by two groups of workers (574, 1352).The stabilization of the sample solution is of paramount importance, and various methods for this are discussed. 1.2.2.4 Emission Methods. Inter-element effects of r.f. ICPs have been investigated by a number of workers (468, 537, 824, 1065, 1066). De Galan (1065) has commented that the various operating frequencies (5-50 MHz), powers (0.5-10 kW), nebulization systems, and positions of measurement used will obviously affect the observed inter-element effects.This would account for some of the conflicting reports in the literature. Greenfield et al. (537) have shown that, with their high-power system. viscosity is the dominant paramctcr when mineral acids are nebulised, but for organic acids, surface tension and density are also important.The interference effects are caused by the nebulization system, and are not the result of chemical interference within the plasma. Interference effects can be introduced by the desolvation system of the nebulizer unit if the analyte aerosol particles are easily soluble in the condensed solvent, and in some cases it is better not to use a desolvation stage (824, 1066).Spectral interference can be a serious problem (129, 468, 1066) for some determinations with the r.f. ICP, and can be caused by source-generated continua (particu- larly with organic matrices), spectral line interferences, and stray radiation. Matherny (1310, 1315) has examined the possibility of quantitative classification of matrix effects on the basis of determined analytical calibration functions. Experimental results were obtained for silicate and carbonate minerals, ores, and solutions. A high- voltage spark (298) was used to study the effects of pores in sample electrodes (pure metals and alloys) prepared by powder metallurgy. The effect of the pores varied with the discharge current, and it was concluded that the effect of pores could be reduced by increasing the voltage. Matrix vaporization effects in a d.c. arc have been minimised by coating the carbon electrodes with AgCl buffer and pipetting the sample as a solution or suspension onto this thin layer (317). 1.2.3 Standard Reference Materials and Calibration The various problems of analysing museum samples have been stressed (895, 1063), and results for even the major elements in common bronzes can vary between laboratories by as much as 40-50%. The use of NBS Standard Reference Materials is proposed, in order to minimise these variations (895). Various calibration methods for trace analysis have been reviewed (235); certified standard samples, independently analysed samples, standard-addition methods, and synthetic standards were all considered.

ISSN:0306-1353    

DOI:10.1039/AA9760600064

出版商:RSC    

年代:1976

数据来源: RSC