ISSN:0306-1353    

DOI:10.1039/AA97404FX001

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Part I : Furidarneitfals arid Instrurneiitatioit 5 2 Excitation Sources and Atomizing Systems 2.1 ARCS AND SPARKS Lishanskii (686) studied the emission intensity of Cu through the gas cloud generated by a low voltage spark, an a s . arc, a low-voltage spark with an arc tie and bipolar combined discharges. The intensities of the arc and low-voltage spark discharges were inhomogeneous along the discharge axis.By using the combined discharge forms it was claimed that a higher emission intensity could be obtained. A d.c. arc source has been describcd which consists of a horizontal arc burning in a graphite tubc surrounded by an inhomogeneous magnetic field (558). The temperature distribution between the clectrodes and the tube wall was measured and detection limits for Tn and Pb reported.The effects of easily ionised elements on the d.c. arc temperature have been examined undcr various conditions (63, 552, 571, 1230). The general conclusion seems to be that, although the arc temperature is lowered, detection limits can be improved by several orders of magnitude. Decker (560) has compared the performances of three c1.c. arc sources - a manually controlled source, a constant-current source and a constant-power sourcc - for the control of variations in arc temperature, electron pressure and anode temperature. He showed that there was little difference between the systems for well buffered samples, but that the constant-power source gave the greatest control with poorly buffered samples.In a study of the spcctrum produced by a d.c.arc between aluminium and carbon clectrodes (636), it was dcmonstrated that the intensity of the A1 305.01 nm line was greater when the aluminium electrode was the anode but the intensities of the emission lines from the impurities were greater when the aluminium electrode was made the cathode. Two important papers by Nickcl and Mazurkiewicz (1201, 1202; see also ARAAS, 1973, 3, 7, ref. 819) have described the effects of the discharge gas (Ar, He and Ar/O,) on the processes of evaporation and excitation in a carbon arc. High-speed photography and radioactive nuclides (to determine material distribution in the electrodes and in the discharge cell after burning) were used to study macroscopic particle movement in the process of evaporation. Results for various Fe compounds showed that changes in emission intensity wcrc mainly due to variations in the b.p. of the various compounds and to the differences of cvaporation rates.Avni and Goldbart (52, 56) have measured the rate of volatilization of free particles of U, Th and Zr matrices from the anode cavity into the discharge zone of a d.c. arc in two ways, (a) indirectly, by determining the residue in the anode cavity, and ( b ) by passing an aluminium wire through the plasma at a constant velocity and determining the concentration of elements coated on the wire.The total particle concentration of U, calculated by the aluminium wire method, was compared with values obtained by absolute intensity measure- ments of ion and atom spectral lines to obtain an estimate of the molecular concentration of U in the d.c.arc plasma. A study of electrode erosion (1140) in low-current arcs claimed that metal droplets (0.05-0*5 prn diametcr) wcrc cjccted from onc of the electrodes, by the rccoil forcc that follows the abrupt cessation of ion or electron bombardrncnt, each time bombardment ceased. Kaiser (1452) reviewed the use of sparks as spectrochemical light sources using a historical approach. Several points were made, but of particular importance (not only to the area of spark emission analysis!) was the comment that a wealth of information in the older literature is oftcn overlooked; this leads, therefore, to frequent repetition of previous work.A study of the effect of spark-source parameters on spectral-line intensities and excitation states has been made (568).From the results obtained excitation conditions could be pre- dicted and, by control of source conditions, it was possible to separate spectra of close ionization states cithcr by total-pulse or time-resolved observations. Studies of the changcs6 Part I : Fundamentals and Znstrumentation in emission spectra of medium-voltage sparks with variations in electrode polarity and the number of ignitions in unit time were made by Florian (1506).He also described an evaluation factor, F, dependent on the important parameters calculated from scatter diagrams, which allowed an easy choice of optimum line pairs (1348). Optimisation of emission spectrographic techniques by isolating selected regions in the spark gap has been studied (461), and methods for viewing the spark in order to reduce background and line interferences were suggested.Strasheim and Blum (885) have examined the phenomena taking place on the surface of A1 containing 5% Cu, using high-speed photography, while a single medium-voltage spark discharged under atmospheric pressure. The formation of cathode spots was found to bc dependent on the condition of the sample-electrode surface and changed during the occur- ence of one single discharge.The work showed that a number of cathode spots formed even on a polished surface. The application of a liquid layer to a solid sample-electrode surface during sparking was shown to improve analytical results (111). In addition, a standard in the solution layer could be used as an internal referencc instead of the conventional matrix element in the solid sample.The influence of O,, present in Ar or which leaks into the spark chamber from the atmosphere, on oxide formation in the spark has been described previously. Slickers (68) has considered occluded oxygen and oxides in the sample as a third possible source of 0 2 . This problem has been overcome by directing a stream of Ar on to the electrode surface to remove any 0, formed during the pre-spark period.Other references of interest - Arcs - Applied magnetic fields: 438, 489, 490, 886 Electron-density calculations: 61, 1109 Electrode-temperature measurement : 1224 Vapour-liquid equilibria /analytical curve : 900 Laser/spark: 1171 Electrodes: 1434, 1515 Effects of gases: 10188, 1223 Effect of electrode temperature: 553 Sparks - 2.2 PLASMAS Further studies have been made in all of the major fields of research utilising plasmas as spectrochemical atomization and excitation sources.It is particularly noticeable that more and more authors are at last starting to recognise (and emphasise) the potential advantages of plasmas, especially for multi-element analysis (1 12, 1234, 124'1, 1255, 1264, 1265, 1266, 1424, 1552).Thus, for example, the wide working range available has been shown to permit the determination of 6 elements in A1 alloys at a single dilution (1264). Critical comparisons with flame spectrometric analysis often serve to underline the merits of plasma emission techniques, particularly with respect to useful concentration range and (though perhaps to a lesser extent) detectability (1 21 7, 1264, 1552); comparative inter- ference studies also appear to favour plasma atomization (1447). Moreover, plasmas of only moderate power, (i.e., 2 kW or less) appear to be perfectly adequate for a wide range of sample types.It is perhaps appropriate to point out here that Koirtyohann and Wallace have com- mented that inexperienced workers may initially be faced with several time-consuming (but not insuperable) problems when entering the field of induction coupled plasmas for the first time (1471), e.g., melting of the quartz jet owing to improperly formed discharges, difficulties in electrically earthing the system, the problem of noise rejection by the electronics and the intolerance of the discharge for air contamination in the Ar.They have discussed their solutions to these problems in the hope that others can avoid repeating their mistakes.Purt I : Furldameritals and Instrumentation 7 The use of simple pneumatic nebulizers is increasing in favour for the introduction of liquid samples into plasmas (112, 1195, 1217); de-solvation is no longer always regarded as mandatory.Fassel and co-workers (575) have described the construction of a purpose-built pneumatic nebulizer with an uptake rate of 3 ml m i d at an Ar flow rate of only 0.8 1 m i d . The same group has also combined a simple tantalum filament atomizer with one of their inductively coupled plasmas and shown that plasma emission analysis can compare very favourably with AA and A F for the analysis of microlitre amounts of sample (118).An analysis rate of 20-30 samples per hour was still possible. Much attention has been focussed upon higher-powered, inductively-coupled r.f. Ar plasmas, and the fundamental properties of this type of system have been studied quite extensively. Parameters investigated include the axial and radial distribution of temperatures (1 195, 1283, 14418), electron and element concentrations (1 195, 1283) and the origin of the background radiation cmitted, analyte residence time and self-absorption (1 264).Barnes and Schleicher (1450) have discussed the experimental verification of computer models oE an inductively coupled plasma using several techniques for flow visualisation and gas velocity determinations.An interesting report (64) has been published of the spectra and properties of Ar and Ar/CH4 plasmas containing S, P and halogens. The authors commented that the poor detectability of these elements confined any potential application of the system to gas analysis. They also noted that the presence of atomic C resulting from CH, introduction was supcrfluous as far as the atomization of refractory oxide-forming elements was con- cerned, as might be expected for relatively high-powered plasmas.However, other authors have commented (524) that the incorporation of C3H8 or organic solvents into Ar plasmas is a distinct advantage for the analysis of such elements. Dreher and Frank (22'5) have described the use of a low-powered induction coupled plasma (250 W, 27-5 MHz) for emission spectrometry in the vacuum ultraviolet. Although the detection limits reported for As and I might be regarded as analytically useful, sample changeover was a somewhat lengthy procedure and the tantalum foil used to introduce the sample has a very limited life.Merchant and Veillon (738) have reported a study of the use of a low-current, d c .arc pZusma jet as an atomization and excitation source. They observed an unexpectedly rapid drop in temperature (5000 K) in the flame immediately above the cathode region, over a distancc of about 2 mm. Optimum detection limits were obtained in this low-temperature region, but only because of the low-dispersion nionochromator employed. The authors commented that the superior detection limits obtained with similar sources by othcr authors must be attributable to thc greater resolution of the spcctromctcrs used.Low-power microivave plasmas have been studied on both a theoretical and a practical basis. Busch and Vickers have reported (14) the results of an extensive study of the funda- mental properties of this type of excitation source. Other references of interest - Reviews: 1231, 1234, 1553 Sealed sample tubes: 577, 1070, 1232, 1235 Isotopc analysis: 872 Temperature measurement: 127 1, 1275 Line-broadening in plasmas: 1168 D.C.arc plasma jet: 788, 1459, 1474 600 W, 2450 MHz, N, plasma: 1241 2.3 GLOW DlSCHARGE LAMPS Although glow discharge lamps (GDLs) have advantages compared with arcs and sparks, (see ARAAS, 1972, 2, 10, ref. 631), as with any analytical technique, problems exist. It was known that dischargc cratcrs could give an impression of selective erosion, although the range and simplicity of the calibration curves did not seem to support this. Papers arc,8 Part I : Fundamentals and Instrumentatioft however, now appearing that report examinations of circumstances in which changes in linearity and emission rates are said to occur.For example, Jager and Blum (562) used a scanning electron microscope to examine cross-sections of the burn spots produced by a GDL on both brass and Au samples. The examination was claimed to show that Pb inclusions in both types of sample were eroded to a lesser degree than the sample matrix. Further work (1040) on Au samples led to the use of a &element interference correction programme for Au, Ag, Cu and Pb by an on-line computer coupled to a 2 m spectrometer, using a commercial GDL.Explanations given for the non-linear calibration curves werc: changes in the sputtering rate, caused mainly by variations in the Cu content of the samples, and self-absorption, which was observed when resonance lines were used.A comparatively simple relationship of the emission intensity to the gas vapour pressure and the source power used in the GDL was reaffirmed by work with mild steel and Cu/Ni alloys reported in a paper by Hirokawa (550). Broekaert (1237) investigated the dependence of detection limits, with an HCL, on the pressure of Ar or He discharge gases, modes of operation, current, and sample introduction technique for the analysis of metals in aqueous solutions.Other references of interest - Hollow-cathode lamps: 53, 565, 1002, 12318, 1337 2.4 LASERS An improved Q-switched ruby laser microprobe for multi-element analysis by emission spectrometry has been described (1281). Re-designed components in the laser cavity led to greater stability, and operating convenience was improved by using a binocular viewing microscope (incorporating a camera and safety device to protect the viewer), a direct-reading laser-energy monitor and a pulse indicator.Various modifications to the optics of the system were also described. The degree of erosion by a laser pulse was studied as a function of m.p. for various pure metals and alloys (746). Erosion of pure metals showed a maximum for Pb with a rapid fall off for elements with increasing m.p.; for binary alloys the erosion, as expected, was strongly dependent on alloy composition.A method has been described (6311) for measuring the divergence angle of a lascr beam (used to evaporate substances from selected areas) based upon the measurement of the interference patterns produced behind a ground-glass plate when the beam was projected via an objective on to the glass plate.Other references of interest - AA with automatic background correction using a laser atomizer: 1006 Laser-excited emission from aerosol particles : 1270 Plasma production by laser irradiation of droplets: 1277 Laser/spark atomization: 1135 2.5 FLAMES 2.5.1 Flame Types Little original work has been reported on the application of new flame types to analytical spectrometry. Detection limits obtained for a number of elements (883) using an O2-shic1ded air/CzH2 flame for FES (see ARAAS, 1973, 3, 11) have now been reported and compared with those attainable with the N20/CzHz flame.Similar values were obtained for many elements with both systems studied, although the NzO/C2Hz flame gave higher sensitivity for refractory- oxide-forming elements.The influence of flame-gas thermal conductivity on atom formation in flame spectro- metry has been demonstrated (650) from a comparison of the rate of de-solvation of water droplets in a He/Oa/CzHz flame with a previously determined rate in an air/C2Hz flame. A close correlation between de-solvation rate and thermal conductivity was observed.With aPart I : Fundamentals and Jnstrumcritatioii 9 conventional air /C2H2 flame the emission intensity from sample solutions increased when air was replaced by a He/Oz mixture of comparable 0, content. Enhanced emission was attributed to reduction in droplet size and an increase in de-solvation and vaporization efficiency. Small diff crences in particle vaporization rates between the two flames are important only in the determination of refractory-oxide-forming elements.A fibre-optic system was used in this study to improve the spatial resolution in emission-profile measure- ments. ,4n excellent experimental and theoretical evaluation of the NzO /C2H2 flame as an atomization cell has been described (66). The formation of atoms from metal solution acrosols introduced into this flame was examined by ( a ) infcrence from well known reactions and equilibria prevailing in cooler flames, ( b ) calculations using a thermodynamic flame model, and (c) experimental observations of free atom densities in the flame as a function of flame stoicheiometry.The calculated partial pressures of the major natural flame species and some of the spectroscopically observed minor species were presented as a function of the oxidant/fuel flow ratio.The predicted relative number densities of Na, Mg, Cu, Fe, Li, Be, Al, W, Ti and Si have been compared with measured frec atom absorbances in an Ar-shielded flame. These comparisons were completed for various heights of observa- tion in order to detect any time delays in the attainment of equilibrium.The data reported show that (a) the degree of metal atomization in the flame can be adequately described by the equilibrium state, (b) in general, when solute vaporization is complete there exists a value of p at which atomization is complete for metals that form monoxides with dissocia- tion energies less than 6.5 eV, and (c) certain metals may form carbon-containing compounds in the interconal zone.Winefordner and co-workers (1463) have studied electrical effects in air /C2Hz flames by the use of wire probes inserted into the flame to allow measurement of the macroscopic electric field strength. The variation in electric field as a function of air/fuel ratio and addition of metallic elements was studied.A double floating probe was obtained by supply- ing an external voltage to the wires in the flame; from current-voltage plots the electron density, temperature and ion density were derived. The effects of different air/fuel ratios and added metallic elements were also determined on these variables. The use of low rise-velocity high-temperature flames for the emission spectrometric analysis of individual particles of micrometre and sub-micrometre diameter has been described (1328).The design of a burner system and calibration procedurzs are repoited with results presented for 02/C2N2 and N,0/CzH2 flames. Legg (9SO) has analysed burner gases, after the addition of trichloroacetic acid to solutions nebulised into flames, during AA procedures. The combustion products were shown to contain CHCl,, C1, and phosgene at concentrations of <0*2, 40 and 2 pg 1-I, respectively.It was suggested that the levels of phosgene may be responsible for symptoms of headache and lachrymation, unless an efficient extraction system is employed, when trichloroacetic acid is used as a precipitant for protein. Interest in the application of flame systems as chromatographic detectors continues.Systems which depend on the use of hydrogen flames for detection of S and P compounds separated by GLC remain of interest (172, 173', 417). A fluorine-specific detector, in which the intensity of SrF band emission at 580 nm is monitored in a turbulent Hz/Oz flame, has been described (1016). Herrmann (1032) has compared the use of a flame emission method of this type with a technique based on AA for the detection of F compounds after gas chromatography.A GLC detector consisting of a flameless atomizer and AA spectrometer has been used in the analysis of commercial gasoline, containing alkyl-lead compounds. and of fluorinated acetylacetonates of chromium (832). The applicability of AA as a detection technique for gel chromatography has been discussed (2); the effluent from a Sephadex G-15 column was introduced directly into the nebulizer of an AA spectrometer.The system was shown to provide sensitive determination of metal ions. An AA detector for the liquid-liquid chromatographic separation of chelating10 Part I: Furidameritals mid liistrunzentatioiz agents such as EDTA and NTA has also been reported (5).The solution containing the chelating agent was passed over a column of Cu-form chelating ion-exchange resin; a quantity of Cu, proportional to the quantity of chelating agent present, was stripped from the column and the resulting solution passed to the air/C.H, flame of an AA spectrometer. 2.5.2 Burners and Nebulizers A study of the application to AA of total-consumption burners with external control of the solution uptake ratc, using a piston pump, has been reported (557).The shape of calibration curves and the efficiency of thc apparatus was discussed in terms of solution uptakc ratc. It is claimed that these burners can give satisfactory results even when sample flow rates are cxternally adjusted to values higher or lower than those obtained undcr free aspiration conditions; the accessible concentration range is thus extended by more thnn onc order of magnitude.Denton and Swartz (827) have described an improved ultrasonic nebulizer for the generation of high-density aerosol dispersions. Conversion rates greater than 1 ml min-’ were obtained and generation of various droplet sizes was achieved through overtonc opera- tion and simplified transducer substitution.Studies comparing this system with a standard pneumatic nebulizcr, using a premixed 02/EI, flame, showed a significant improverncnt in detection limits. Hieftje and Bastiaans (837) have repmted a furthcr application of an isolated-dropkt s:implc-introduction device to the investigation of the vaporization behaviour of \ arious Ca salts in a laminar flow air/CzH: flame.The rate and extent of vaporization were determined indirectly by mcasuring the emission intensity of vaporized atoms as a function of flarnc hcight. The results indicated that mass transfcr was thc controlling mechanism in tfi: vaporization oi small particles of calcium chloride. Cershman and Glushanok (626) have reported a pneumatic sampling device for powders in which the samples are placed in a vertical storage tubc coupled to a 50 Hz vibrator.The sample was fed into the air inlet nozzle at a rate which was reproduciblc to within 8-1056. A patent has been issued (171) for a burner designed to minimize turbulent flow, which causes concsntrated sample solutions to form precipitates and block the burner slot.This consists of a slot burncr in which the slot tapers at the base in an arc of 2 3 mm radius. A flashback-resistant burncr for the N20/C2H, flame has been described (13); i n this design a single 50 mm row of 50 holes, 0-57 mm in diameter, is employed rather than a single slot. Optimum nebulization and residence times arc achieved at an oxidant flow rate of 9 1 min-l with analytical performance claimed to be comparable to that of a slot burner.Suddcndorf and Denton (97) have reportcd a specially designed one-hok insert burncr used for the study of burning parameters of premixed 0,/H2 and OI/C2H2 flames. Gas mix- tur:s can be preheated and premixed; a pop-off valve was also incorporated to rclcase pressur,: in flashback studies. The effects of port diameter, oxidant/fucl ratio, total gas flow and burner temperature on flame stability have been studied.The results provided design criteria and operational data for these gas mixtures; it is claimed that flashback can be eliminated by applying thesc criteria. 2,5.3 Diwrete Sampling Devices A number of reports have appcarcd dcscribing Ihe extension and consolidation of carlicr work with devices of this type, l)d?vssn 2 n d w-workei-s too ntimcrous to mention.(1 129) ha\)e now describcd i n dctail their work on clxtrolytic extraction combined with flame AA reported in part in earlier communications (see ARAAS, 1973, 3, 12); in this technique metals arc deposited from aqueous solution on an iridium wire and then atomized into the optical path of an AA Spectrometer using an air/H? flame and a 10 cm Nimonic tube.This method is claimed as ‘feasible' for quantitative analysis, but as only a fraction (<1%) of any element is deposited from thc qolution, the sensitivity is no bcttcr than with direct flame methods.Part I : Fundamentals and Iizstrumeritation I I Studies of molecular emission cavity analysis (MECA) (see ARAAS, 1973, 3, 13) by Belcher, Townshend and co-workers have continued.Papers which describe the determina- tion of halides ($01, 929), Se and Te (718), I3 (1042) and As and Sb (1050) have been published. For B, As and Sb, volatile methyl borate, AsHz and SbH3 were generated and passed into the cavity, where band emission was monitored. Commercial instrumentation is now available for this technique, which should assist its acceptance for analysis.A flame atomic emission technique which employs vaporization from an electrically heated tantalum filament, with atomization and excitation in an air/& or air/CzHz flame, has been reported (1454). Detection limits €or Li in aqueous solution were 0-009 ng (0.0018 pg ml-*; 5 pl sample) and 04tH ng (0.006 pg ml-I; 5 p1 sample) in the air/M2 and air/C2Hr flames, respectively.A modification of the Delves cup technique, in which a 4 mrn diameter hole was punched into the base of the cup, has been shown to permit higher tcrnperaturcs to be achieved and to allow the volatilization of Cu from aqueous aild diluted serum samples (1243). The samples are retaincd in a ridge in the base of the cup; this device may permit extension of the utility of the technique to the determination of a number of other elcments. The same group of workers has described the use of a N20/C2H2 flamle for work with a microsampling cup assembly (i430).The increased heating rate and higher terminal temperature result in more complete sample volatilization and atomization. 2.5.4 Theoretical Studies Again only those theoretical flame studies of possible interest in analyticai spectroscopy are reported, Reif, Fassel and Kniseley (1 5) have reviewed the basic theoretical principles concerned with temperature nxasurenients in flames by spectroscopic line-reversal, emission-a bsorp- tion, slope and two-line technique.A systematic examination of the validity of temperature data obtained by spectroscopic techniques is reported. Valucs were found to depend on the method employnd, the energy of the quantum states involved in the transition(s) producing the line@) studied, the particular temperature gradient in the flame and the conccntsabion distribution of the thermometric species.In a second papcr (563) the same authors have calculated apparent temperatures for several theoretical niodels representing typical labora- tory flames and shown that diiferences as large as 800 K may arise between the measured values and the average or weighted average temperatures.Moreover, temperatures measured by the same techniques and with the same thermornetric species, but using different spectral lines, may differ by as much as 200 K even if no other errors were involved (such as uncertainties in values taken for transition probabilities).The authors rightly state, therefore, that physical interprctations which go beyond these uncertainties flatter the actual experi- mental results and are not justified. Techniques for measuring temperatures and free atom concentrations in cylindrical flames of air/CsHz, aia/H2 and air/C3Hs have been described (525).Measurcments 011 integrated signals were made along lines in the cross-sectional plane and perpendicular to the direction of observation. Optical densities and emission line intensities were calculated as a function of flame raditls using a computer prograrnrnc. Zizak (1 107) has discussed the applicability of the Abel inversion tcchnique to calculatc the local conditions in axially symmetric flames.Where this technique cannot be used, a numerical invcrsion programme to integrate the radiative transfer equation by successive approximations was proposed. Studies concerned with the ionization of alkaline earth additives in H, / O2 / Nz flames have been described (1 137, 1 138). Only positive ions were observed; principally MOH+ and M+ involved in the equilibrium M+ + HzO MOW+ -t H.In a mass spectrometric study of ion concentrations at various points in flarnes the rates of production and recombination of MOH+ ions by the processes M + OH ;t MOH+ 4- c and MO -t H - MOH* -t e- were measured. The rate constants for both processes at different tempera- tures were determined. A12 Part I : Fundamentals arid IrrstrLiurzetitation Bulewicz, Padley and Smith (615) have examined the visible and ultraviolet spectra of reduced pressure Oz/C2Nz/Hz flames to obtain evidence on the ground state concentrations of species in the flame gases.The 02/CzN, flame is difficult t o stabilise (especially when rich) owing to its low burning velocity; it is thus experimentally more convenient to study the Oz/CzNz/Hz flame.Vibrational and rotational temperatures were measured for CN and Cz species and found to be in the range 5500-6508 K. The electronic (line-reversal) tempera- tures of Na in the flames were normal away from the reaction zone; concentrations of CN, at least in 02-rich flames, were many orders of magnitude above thermal while in contrast neither C, nor CH could be detected, even in the most fuel-rich flames (Le., the concentra- tions of these species were considerably below equilibrium values).The CN radical has been used as an indicator species to characterise the reducing properties of the N 2 0 /C2Hz flame (932). Quantitative relationships have been given between thc occurrcncc of CN in the flame and flame atomization efficiency; these have been used to determine experimental values of atomization efficiency and to correlate the relative emission responses of different radicals (NH, CN, CII and C,).The work of Rubeska, concerned with the determination of relative atomization efficiencies of Sn in air/CzHz, air/Hz and Ar/H, diffusion flames from peak absorption measurements, has been published (1 197).The results indicate that the Ar /Hz diffusion flame, despite its low temperature (1130 K) is the most efficient for Sn atomization, which was explained by assuming that free Sn atoms are formed through reactions of tin oxide with free hydrogen atoms. Some indirect supporting evidence for this hypothesis was given. Other references of interest - Flames: 315, 482, 596, 1331 Flame-in-tube burner: 49 Efficiency of atomization and sample introduction: 734 Nebulization: 482, 18'87, 1213, 1511, 1513, 1514 Spark-discharge vaporization: 1 155 Graphite atomizer in flame: 209, 684, 865, 1 182, 15 10 MECA: 1476 Screw feed: 1484 Flame detector for 14: 627 Radiative dis-equilibrium in flames: 11 01 Velocity and temperaturc measurement in oscillatory diffusion flames: 1 I39 Coherence in flame-emission noise signals: 1453 2.6 ELECTROTHERMAL ATOMIZERS This section deals with new practical and tlicoretical developments in electrothermal atomiz- ation; it does not include methodology papers using previously described systems.There has becn a decline during 1974 in the numbcr of references concerning the development of new electrothermal atomization systems.The emphasis has been on various instrumental and procedural modifications of existing systems and an improved theoretical understanding of atomization processes and interelenient eflects. The advantages of continuous formation of a pyrolytic coating during the operation of heated graphite devices has becn described previously (ARAAS, 1973, 3, 15) and further work using CH, (342, 1486) and C3H8 (670) has been reported.A novel device for the precise measurement of microlitre sample volumes for use with tungsten or platinum filaments has been described (121). The filament is heated to a temperature just below the boiling point of the solvent; on adding the solvent a rapid drop in the filament temperature occurs which results in a sharp decrease in the resistance of the device.When all the solvent has evaporated, the filament rapidly returns to the original temperature resulting in a sharp increase in the resistance. Thus the time between the voltage changes (caused by the variations in resistance) is directly proportional to the addedPart I : Fundamentals m d lristrumcritatiori 13 sample volume. This technique should also be viable for other electrothermal devices having a low heat capacity and could prove useful for automated systems.Non-specific background absorption is more likely to be observed with electrothermal atomizers than with flamr: techniques and it is geiierally agreed (1216) that before manual or automatic background correction is applied the background absorption should be niinimised by instrumcntal or mcthodology modification (e.g., correct sample dissolution, careful collimation of the light bcam and judicious setting of the dry ash and atomization times and voltage settings).Dry ashing in the presence of oxygen has been suggested as a means of reducing background absorption (1 11 3) and was found to minimise condensation with a graphite furnace and permit the use of a much lower dry ashing temperature (760 K instead of 1370 I<) €or the determination of Cu in vegetable oils (1532).Woodriff (954) has compiled a review comparing the advantages and disadvantages (with respect to accuracy, sensitivity, convenience, measurcment procedure and matrix effccts) for the L’Vov, Woodrill, Massmann and mini-Massmann furnaces. Matrix effects were found to be less serious with long tubes than for tubes with openings near the centre (e.g., sample entry port or vent holes).It was claimed that, when peak heights are measured, the relative sensitivity (k, solution concentration) is proportional to the tube length and, when peak areas are measured, the relative sensitivity is proportional to the third power of the tube length. 2.6.1 Graphite Rod Devices Cresser and Mullins (’720) have used a simple theoretical approach to calculate temperature and power consumption versus time curves for graphite and tungstcn filaments at various voltage settings. For very small diameter filaments it was postulated that a capacitor dis- charge would give a very rapid temperature rise (2500 K in a iraction of a millisecond).This should result in high sensitivity and very reproducible atomization conditions. A theorctical model which described the time dependence of the number of atoms in a point, a rectangular and a circular space above a graphite rod atomizer, assuming linear thermal perturbation, has been reported (1 326). The relative influences of the parameters controlling the rclease and transport processes of the atoms were considered.Aggclt and Sprott (1045, 1386) have determined the minimal atomization (appearance) temperature for a number of elements for both graphite rod and tantalum strip atomizers. Unfortunately these temperatures were defined as the ‘lowest temperature at which atomic absorption signals were observed’; a more precise definition could have been used.The free energies for the dissociation of the oxide to the metal and for the carbon reduction of the oxides to the metal were calculated at the appropriate appearance temperaturcs. The following conclusions were reached: (a) €or elements such as Ag and Au the oxides dissociate prior to atomization, ( b ) for elemcnts such as Al, Ca and Mg atomization occurs directly from the oxide, (c) graphite acts as a reducing agent towards the oxides of Co, Fe, Ni and Sn where the appearance temperature is lower for graphite than for tantalum, (d) for elements such as Bi, Cr, Cu and Pb the graphite does not act as a reducing agent, despite a favourablc free cnergy for the graphite reduction reaction at the appearance temperature - this was attributed to a kinetic effect.The graphite rod appcarance tempera- tures for Al, Cd, Cu and Sn however did not agree (300-500 K difference) with those reported by Ottaway and Campbell (1072) using a graphite tube furnace (see Section 2.6.3). Those for Ag, Ca, Cr, Mg, Mn, Ni, Pb and Zn agreed within 120 K. Thcse results indicate that experimental results obtained using a given atomizer cannot always be extrapolated to other systems.An intcr-element study (10485) on the binary combinations of Cu, Mg and Ni was also undertaken. Interferences originated both on the graphite rod (independent of aperture height) and in the vapour phase immediately above the rod (dependent on aperture height). Interference caused by condensation from the vapour phase was found t o be dependent on the volatility of the interferent, the mutual affinity of the elements and the14 Part I : Fundamentals arid Iiistrumentatiori tendency of an interfercnt to recombine with traces of oxygen and condensc as the oxide rather than the metal.Various new types of electrothermal graphite atomizers have been described. Hwang (920, 1455) has described a system which can accept tungsten or graphite in various con- figurations and allows the system to be pressurised.Another system (653, 1489) utilised a braided graphite filament (2 mm 0.d.) with temperatures up to 2900 K attainable with 350 W power. The porous nature of the braid gave a furnace-type environment during the various hcating steps. Na (254), P and S (721) have been dctermined by flame emission from the entrained air/N2/I12 flame formed above a graphite filament, but the detection limits were not very good. 2.62 Graphite Miniature Furnaces A fundamental study of the original mini-Massmann graphite rod (ARAAS, 1971, 1, 19) has been reported (1483) with the importance of a rapid rcsponsc amplification system strcssed. The influence exerted by the shcathing gas (Ar, He and Nz) could be explained in terms of diffusion phenomena rather than the variation in the heat transfer coefficients.The work of L’Vov (ARAAS, 1971, 1, ref. 360) on the characterisation of transient absorption pulses has been extended (91 8) to the cascs of a mini-Massmann tiibe and a con\ientioiial Massman tube furnace using the characteristic times of ‘appearance’, ‘peak height’, ‘atoniiza- tion’ and ‘rcsidcnce’ of the atomic species.Lining a miniature graphite tube with tantalum foil (572) was found to improve the sensitivity and precision for the determination of As and Se in effluents. A lower atomization temperature could bc used which not only decreased the background absorption but allowed better resolution of the atomic absorption peak from the preceding non-specific absorption peak.However, the foil had to be replaced aftcr 20 to 30 sample injections. Gold-plated miniature graphite tubes havc been used for the determination of Hg in solution (103) and Hg in air (65, 100). Winefordner et al. (1338) havc described R vitreous carbon furnace through which pneumatically nebulized samples were passed. AF mcasurements were per- i‘ormed above the furnace.A considcrable improvement was obtained by using a pulse nebulization technique (1486) and adding CH, to the purge gas. Furthcr work (1530) has shown that a remarkable gain in precision can be obtained by instrumentally controlled sample introduction (A RA AS, 1972, 2, ref. 385). 2.6.3 Graphite Tube Furnaces The study of atomization mechanisms in graphite tube furnaces (see ARAAS, 1973, 3, 17) has been continued.Ottaway and Campbell (1072) have related the lowest atomization (appearance) temperature of a substantial population of atoms in a graphite tube furnace with the temperature at which the metal oxidc reduction by graphite, MO(s) + C(s) + CO(g) + M(g) : AGIO, is therinodynamically feasible, ie., AGIO =I 0. It should bc stressed that these appearance tcmperaturcs agreed with those observed from both a graphite rod and a tantalum filament (10415) for a number of elcmcnts (Ag, Cay Cr, Mg, Mn, Ni, Pb and Zn) for which different atomization mechanisms were proposed in some cases (see Section 2.6.1). Substantial losses were found to occur at temperatures considerably below the appearance temperatures for a number of elements (1072) during prolonged dry ashing (up to 300 s).At 1400 K the losses, for the sulphate salts, decreased in the order Cu> Ni >Co >F,: (see also AKAAS, 1973, 3, 17). This was attributed to the free energy charigc (GG,”) for the reaction MO(s) + C(s) -+ CO(g) + M(s) becoming ncgativc at a lower temperature than that at which AG,” becomes negzttivc.Thc metal was assumed to bc more volatile than the corrcsponding oxide. Fuller (1094, 1131) has used a kinetic approach to describe the atomization of Cu at fixed atomization temperatures betwecn 1720 and 2220 K in a graphite tube furnace. Tt was shown that a slow first order reaction involving reduction of CuLO by graphite followed byPart I : Furiclameiztals arid Iizstrumerzlntion 1s the rapid vaporization of the resulting Cu is the most probable reaction mechanism.The theoretically calculated time versus absorbance profiles agreed with the experimentally determined results. \%'hen a tantalum-lined graphite tube was used the Cu absorption peak height increased and the peak width decreased compared to the graphite tubc. A faster rate of reduction of the CLLO on the tantalum surface was tentatively suggested as a contributory factor.Peak area nicasurcments have been found (1058) to be of greater value than peak height measurements, with the advantages of the former being a greater linear range of the calibration graph, better reproducibility and no change in signal due to a change in the atomization kinetics, caused by variations in matrices.However, an important consideration is that peak area read-out, in the absence of automatic background correction, cannot discriminate between an atomic absorption peak and background absorption peaks; even with automatic background correction a decrease in the signal-to-noise ratio is often ciicountered. The direct analysis of solid samples has given some encouraging results but calibration rcmains a diiificult problem.Lundgrcn et ul. (39, 576, 1325) have extended their earlier work on temperature controlled heating (ARAAS, 1973, 3, ref. 1513) using a very r:qid rate of increase of fixrnacc temperature up to a pre-sclectcd atomization temperature. Using this system with a L'Vov type furnace (576) it was possible to determine Pb and Zn directly in biological tissx (up to 200 pg dry weight).L'Vov (1509) has described a novel approach which should rniniinise thc problem of heterogeneity in this type. of work and also minimisc background absorption effects. He uscd two concentric graphitc tubes (31 and 7 mm i.d.) with the solid samplc (UP to 0.1 g) placed between the two tubes so that, on heating, the atomic vaponrs diffused in to thc inncr tube where the analytical measurements were made.h two-part furnace atomizer for the determination of relatively volatile elements has been described (1336). The sampf,: (solid or liquid) is vaporized and oxidiscd in a separate chamber prior to being trr,nsported into a heated graphite tubc. Langmyhr ct al. (1039, 1048, 1060) have extended their earlier work (ARAAS, 1973, 3, ref.GQ2) using an r.f. heated graphite tube for the direct determination of Ag, Cd, Ga, In, Pb, TI and Zn in various inorganic materials. The samplz (0.5-10 mg), finely ground (53 pm), was placed in the middle ol: the graphite tube by means of a small tantalum scoop; lor Ag, Pb and Zn much better sensitivity was observed if the sample was mixed with an qua1 amount of graphite powder.Liquid standards could not be uscd for calibration and in most cases standard solid reference samples wcrc uscd. ltitckground absorption from the sample matrix still CBUSCS inany analytical problems when using clcctrothernial atomization; this oftcn rmults in a severely degraded detection limit compnrcd to th: purc soltation detcction limit. A method of minimising this problcm (399) was to pass thc inert shield gas through the ends of the graphite tube and allow it to cxit rhrough the sample cntry hole in the middle of the tube.Using this technique for thc dzterminr-ition of Cu in sea water the background absorption was reduced 15 times (see also the introduction to Section 2.6, above). A graphite tube with reduccd wall thickness sections cithcr side of the central sample cntry port (730) was used for the dctermination of Pb in waters and found to give good scmitivity ( 5 X 10-" g) and rcyrodvcibility (an RSD of 0.04 at the 0-1 pg ml-' Pb kvel over a period of 4 weeks, with several tubes, was obtained).However, results for Co (and other lcss volatilc elements) showed little advantage over conventional flames and indicate that the inaxirnum operating tcmperaturc of the tube is somewhat limited.Kirkbright et al. (784) have deterniincd I using a N,-purged graphite tube in conjunction with a microwave excited lamp and a vacuum monochromator. The main problem was severe background absorption caused by common salts (e.g., NaCI, KCI, Na2S04), which could not be tolerated above a 100-fold weight cxccss.Howcver, with NaN03 this excess could be increased to 1000-fold.16 Part I : Fuiidcimeritals arid Imtrumentntion Electron-impact excitation of the atoms generated in a King-type furnace has been used to determine B, Cd, Hg and Mg (54). The detection limits (except for B) were somewhat poorcr than those obtained using much simpler electrothermal or cold-vapour techniques. 2.6.4 Metal Filament Devices There have been relatively few developments in this area. Hwang (920, 1455) has described a new tungsten-type atomizer. A molybdenum filamcnt has been shown (1550) to give better sensitivity, longer life, better reproducibility and less inter-element effects than a tantalum filament for Co, Cu and Ni. The difference in behaviour was attributed to the more inert surface of molybdenum.The presence of a residual oxide film on the tantalum filament resulted in a sharp increase in the degree of interference effects observed in the determina- tion of Cu. This effect was not observed with molybdenum. However, a very important drawback of molybdenum is that nitrate ion at concentrations above 0.011 M reacts with the filamcnt, causing permanent damage.Elcctro-deposition on to a tungsten wire has bcen used for the determination of Cd in sea water (1036, 1046) in order to overcome the background absorption problems that can be experienced without prior separation. Further work (ARAAS, 1973, 3, 19) on the technique of electro-deposition on to silver wire (69) followed by explosive destruction of the wire using a large current pulse has been used for thc AA determination of Cd, Ni, Hg and Pb.It was also found possible to replace thc silver wire with aluminium foil (1477) and therefore evaporate the sample solution directly onto thc foil. The AA determination of I at 183.0 nm using a platinum loop has been reported (882). The detection limit (18 pg ml-‘) and the severe background absorption effects experienced limit the usefulness of the technique.Work by Cresser and Mullins (720) and Aggett and Sprott (1045, 1386) involving graphite rods as well as metal filaments has been described in Section 2.6.1. Other references of intcrcst - Background correction: 1210 Chelating agents: 401, 1062, 11 14 Design modifications: 312, 526, 542, 1276, 1446, 1458 Design patents: 170, 408, 409, 410, 412, 456, 467, 468, 469, 815, 868 Hydride generation : 978 Interferences: 1 I17 Non-dispersive systems: 378, 379 Optical pyrometers : 1 104 Reduced-pressure operation : 38 Solid sampling: 684, 757, 1457 Theory: 448,720, 11 15 2.7 OTHER EXCITATION AND ATOMIZING SYSTEMS The most interesting developments in this diverse field are the AA measurements in the liquid phase that have been reported by West et al.(931, 1527). The reaction [Mn+.aq] f ne- -+ [Mo.aq] was used to generate free solvated ‘atomic’ species in thc solution phase adjacent to the surface of a platinum or mercury electrode. The absorption bands of the solvated species (which were related to the line spectra of the gas-phase atoms in flames) had half-widths of 2C-100 nm, which would tend to limit severely the selectivity of the technique. Typical molar absorptivities of 20-40,000 were observed. There have bcen few new developments in the cold-vapour Hg technique. When N2, containing Hg vapour in a closed cell, was irradiated by an intense Hg lamp a rapid decrease in Hg absorption was observed (6), although the Hg could be regenerated by heating the ccll to 343 K. It has been shown (836) that the cold-vapour fluorescence tech- nique has considerable advantages over the corresponding absorption technique with respect to detection limit and background absorption.Part I : Fundameritals aricl Iiistrumcntatioii 17 The determination of As and other elements using Izydride generation techniques is rapidly becoming an accepted method of analysis. A silica tube (17 cm long, 8 mm i.d.) heated by an air/CzHz flame has been used (1128) for the atomization of hydrides, generated lrom a sodium borohydride solution and passed directly into the tube by a flow of N,. Good detection limits were obtained €or As, Bi, Sb, Se, Sn and Tc; results were also obtained for Pb by this technique. An auxiliary transverse inert-gas flow across both ends of the heated silica tube helped to maintain the inert atmosphere within the tube and ensured that negligible background absorption was observed for As and Se. Automated systems for the determination of As in atmospheric particulate matter (338) and waters (1528) have been dcscribcd using mechanical samplers, proportioning pumps and gas-liquid separators. A non-dispersive AF technique using a solar-blind photomultiplier has been used to determinc As (1049). Walsh et al. (53) have continued their studies of cathodic sputtering and have determined Cry Cu, Mn, Ni and Si in various Fc-base alloys. With AF, multi-element analysis; can readily be pcrformed by this technique. The work on shock-wave atomization and excitation (ARAAS, 1973, 3, 19) has now been published (1516). Other references of interest - Cathodic sputtering: 722 Hydride generation: 404, 405, 851, 1049, 1321 Hg: 392, 1047, 1163, 1517 Non-dispersive: 1049 Patents (Hg): 1163 Theoretical kinetic studies of atomic species: 140, 141, 448, 1144, 130.2, 1303

ISSN:0306-1353    

DOI:10.1039/AA9740400005

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

18 Purt I : Furidurnen fals arid Instrumeiitatiori 3 Optics 3.1 OPTICAL COMPONENTS The use of AF to determine radiochemically produced ""a at very low atomic densities has been described (28 1). The very significant improvement in detection limit obtained (compared with that reported by earlier workers) was attributed to the use of carefully designed, horn-shaped light traps, similar to those encountered in Raman spectrometry, which greatly reduce stray light rcflcctions of the exciting radiation. 3.2 BEAM MANIPULATION A patent has been granted for the use of the Zeeman effect to split resonance lines into pairs of non-absorbing lines as a means of providing radiation ideally suited for making background corrections (472) (see also ARAAS, 1973, 3, 4). Either a modulated cxternal magnetic field or a constant transvcrse magnetic field can be used to produce perpendicular polarization in the two split lines and parallel polarization in the unsplit resonance line; separation is then achicved by a rigid polarization filter or by alternately-operating polarization filters, respectively. Winefordner and co-workers (654) havc used double modulation in AF to eliminate the scatter problem when using a 900 W Xe arc.Wavelength modulation of the spectral image over a small range was achieved with a refractor plate (see ARAAS, 1973, 3, 20). 3.3 WAVELENGTH SELECTION 3.3.1 Dispersive Systems The nicchanism for the production of high-resolution transmission gratings by laser-induced refractive-index changes in peroxidized, polymerized methyl methacrylate has been studied (220).The authors concluded that the mechanism depends upon the photo-polymerization of the unreacted monomer, which depends upon the amount of peroxidized unreacted monomer. As this latter parameter is difficult to control, it was suggestcd that irreproduci- biljty will remain a problem in the production of these gratings. A geometric theory of ruled and holographic gratings has bcen discusscd (823), with particular referencc to the design of aberration-corrected gratings.The performance of these concave holographic gratings was much improved compared with that of mechanically ruled concave gratings (825). Thc advantages of using holographic gratings, with correctcd mountings, to produce spectral images with reduccd aberration in the vacuum ultraviolet have also been discussed (3106).Michels (407) has studied the importancc of the change of blaze wavelength as a function of the position on the surface of a concave grating from both a theoretical and a practical basis. The application of cchellc monochromators has bcen extended in plasma eniission analysis (1266), flame AA analysis using a continuum source (789) and for studying spectral- line profiles at wavclengths below 300 nm (1467).Winefordncr and co-workers (1462) have describcd an AF system for multi-element analysis, using a continuum excitation sourcc with a synchronous photon-counting detection system; the monochromator was driven by a stepping motor under cornputcr control, so that it was possible to slew rapidly from the emitted line of onc element to that of the next. 3.3.2 Interferometers Capacitance micrornetcrs, which have bcen used €or soinc timc to dctcct wry small mechanical displaccments, have now been used (1 273) in a servo-controlled Fabry-Perot interferometer to minimisc the effects of thermal and mechanical drift by achieving long- term stability of the ctalon spacing. A method has beer: described (1274) for determining the instrumental function associated with a Fabry-Pwot ctalon in a pressure-scanned systcm in which a sccond etalon is used to provide transmission maxima that are small comparcd with tlic instrumental function bcing determined.Part I : Fundamentals arid Instrumeiitatiori 19 The design of a field-compensated interference spectrometer for the region 400-1 000 nm has been described (1279).The resolving power of the instrument does not depend upon the area of the field stop and therefore has a much larger flux-gathering power from a suitably extended source. It is particularly useful for studying weak, extended sources at reasonable resolution. 3.3.3 Non-dispersive Systems Interest in thc construction o€ intcrfcrencc filters for use at wavelengths below 300 nm continues to grow, a point which has bcen strcsscd in a brief, but intcresting, review by Davidson (968).There has been a particular tendency towards the preparation of interfcr- ence filters for use at low wavelengths, with narrowcr spectral band-passes than those reported previously (see ARAAS, 1973, 3, 22). Malhcrbc (406) has described the use of LaF3 or NdF, and MgF, coatings in the construction of dichroic mirrors and interference filters.Filters may be made with a transmittancc of 85% and a band-width of 2-5 nm centred at wavelengths as low as 200 nm. Previously the use of multi-dielectric multilayer filters has extended through the visiblc only to about 250 nm, the lower wavelength being limited by the problem of finding materials capable of evaporation that are transparent in the ultra- violet.Spillcr (282) has described the construction of an aluminium-MgFz interference filter with a 25% peak transmission and a bandwidth of about 4 nm at wavelengths around 200 nm. Title and co-workers have reported that the wavelength drift of narrow-band ZnS and cryolite filters depends critically upon their thermal and radiation history, and upon their operational temperature (1124), 1125).Baking at 373 K €or 50 days stabilized peak wavc- lengths to 0.1 nm provided the filter was storcd at below 311 K and was not illuminatcd at wavelengths below 500 nm. A 10 mm cell containing 500 g 1-I of nickel sulphate hexahydrate, which transmits below 300 nm, has been used to remove Pb emission at 405.8 nm from the excitation-source spectrum to eradicate the problem of scatter in the determination of blood Pb by measure- ment of Pb direct-line fiuoresccncc at this wavelength (1059). Other references of interest - Background correction: 1006, 1287 Beam manipulation: 41 1, 468, 890, 891 Ca line profiles: 884 Camera focal length and dctcction limit: 231 Gratings: 279, 824 Logarithmic filter construction: 47 Monochromators: 217, 333 Non-dispersivc systcms: 243, 391, 996, 1000, 1 153 Rcsolution: 640 Windowlcss lamp design: 1219

ISSN:0306-1353    

DOI:10.1039/AA9740400018

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

20 4 Detector Systems The photomultiplier rcmains the most sensitive and widely used radiation detector. Its superiority was demonstrated (1284) in a comparison between photographic and photo- electric rccording as used in the laser microprobe analysis of low-alloy and stainless steels. The precision was better by a factor of 2 or 3 and detection limits by up to 2 orders of magnitude. In attempts to overcome its principal limitation, i.e., lack of multiplex capability, studies of the suitability of vidicon tubes and photodiode arrays for multi-element analysis have been carried out.The general subject of multi-channel methods has been reviewed by Codding (1451) and Horlick (1466) has compared the merits of photodiode arrays and Fourier transform spectrometers for atomic spectral measurements.The detection limits that can be achieved with these devices arc not as low as thosc from photomultiplicrs owing to their inferior signal-to-dark-current characteristics. With the widesprcad use of light-source modulation techniques and transient-response atomizers, the mcasuremcnt of fast optical signals is important if the best analytical results are to be obtained. The available technology and the difficulties encountered in the measurc- mcnt of such signals have been reviewed by Lyttle (91 1, 1341). 4.1 VACUUM PHOTOTUBES Developments relating to photomultipliers have been of a minor nature and are extensions or modifications of well established practice. A simple system (580) using circulating cooled, dry gas, to lower the temperature of the photocathode of an end-window photomultiplier to 108 K, reduced the dark current of an S1 spectral-response cathode by a factor of lo4.Pulsing the supply voltage to photocathodes or dynodes is often used in the study of transient phenomena and weak modulated signals. A transistor switch has been deviscd (897) €or the first dynode which has a rise time of 2 ps and fall time of 3 ps.An end-window photomultiplier has been directly coupled by means of a simple rubber gasket to a vacuum spectrograph (898). The constant need to extend the wavelength rangc of the response of photocathodes has been assisted by a technique for developing Ag-O-Cs photocathodes which are sensitive to visible and near infrared radiation up to 1250 nm (1282). Imaging tubes have been used for some time in exploratory studies of their suitability for multi-element analysis. The limited spectral response of the generally available vidicon tube renders it unsuitable for use in the mid and far ultraviolet; recent work has therefore been directed at overcoming this limitation.The output of a Bendix Bx 8025-4522 image converter tube (spectral rcsponse: 120-3150 nm) was viewed by an SEC vidicon system (216).The modulation transfer function of the system fell to 10% at a spatial frequency of 8 lines per mm and 160 nm. The converter used a MgFz input window and a CsTe photo- cathode to give a virtually solar-blind response. An alternative device is the ‘ranicon’ tube (resistive anode image converter) (1212) in which the incoming photon releases an electron in a micro-channel electron multiplier.The charge generated is focussed on a large-arca resistive anode plate with pick-up electrodes on its edges. Each event is located either by the ratio of the charges collected at the edges or by the signals’ rise times. Silicon diode array vidicon cameras with enhanced ultraviolet response (200-1 100 nm) have been used for multi-clement FES and AA (11, 109, 649, 651, 736, 1333, 1435, 1464, 1465).The vidicon tube is mounted in place of the exit slit of the monochromator and views a spectral range corresponding to about 1 cm of the dispersed spectrum (20-100 nm). Line pairs with a scparation of 0.05 nm can bc resolved. Detection limits by emission methods are at least 1 order of magnitude worse than thosc generally obtainable in single element analysis owing to the impossibility of obtaining optimum atomization and excitation condi- tions for all elcments at the same time.Scan rates up to 250 Hz have been crnployed and an analytical rangc of 4 orders of magnitude can be obtained.Purr I : Fundanzcritals arid Instruiricrituiioli 21 4.2 SOLID STATE DETECTORS Boumans et al.(55) have examined various configurations of planar silicon photodiode arrays as detection devices for multi-channel emission spectrometry. They obtained a spectral resolution equivalent to that 01 a system using an exit slit and photomultiplier, The response was linear over 3 orders of magnitude above noise level. Comparisons were made with various types of photomultipliers, but the latter were found to be superior by a factor of 100 to 500 in the signal-to-noise ratio at low light levels in the wavelength range 250-550 nm.Codding and Horlick used an array with 1008 photodiodes per inch (583, 840). The device, which could integrate spectral information for 10 ms to 4 s, was linked to a computer and used for d.c. arc studies. The same authors (584) carricd out further experiments on thc application of the diode array to multi-element AA.In view of the relatively strong signals encountered in AA analysis compared with emission, the problem of diode dark-current is likely to be less serious. A technique for computing the geometric design of the photo- sensitive detectors in a linear array to yield the highest signal-to-noise ratio at the spatial rcsolution limit of the array ha3 becn developed by Eliot (1272). 4.3 THE PHOTOGRAPHIC PLATE 'The subjcct of photographic-type imaging proccsscs has been dealt with in a book by Dainty and Shaw; of interest to thc spectrographcr will be the sections on the photographic proccss, detective quantum efficiency and microdensitometry. The declining use of the photographic plate for the recording of spcctra has led one major manufacturer to discontinue production of most emulsions on a glass base.A working party has been set up by the Spectroscopy Group of the U.K. Institute of Physics to measure the essential spectrographic character- istics of these emulsions and to compare them with readily available alternatives. A simple scmi-automated system for measuring the density of photographic plates has been described (48).Alniatov et al. (1185) used the photographic plate for the simultaneous detection of several trace elements by AA. Thc sample (10-30 mg) was atomized for 2-4 s at 2300 K, with an exposure of 0.01 to 1 s depending upon the lamp used. The sensitivity was the same as for photoelectric recording, e.g., lo-'% for in. 4.4 SIGNAL PROCESSING Two devices have been described for the proccssing of the transient signal produced by the cool vapour method for Hg: one (1211) incorporated an integrator and the other (785) a digital voltmeter with a display indicating the maximum and minimum readings. A generally applicable analog circuit for producing a continuous average of an input signal over a prc-set time period (0.5-50 s) has been devised by Balslev and Hougs (1280).This type of circuit can be used with a strip chart recorder, and by making the time period equal to the expected duration of the transient signal, e.g., emission or absorption pcak, the maximum of the recorder response will correspond to the area under the emission or absorption curvc. A photon-counting system (6134 has been used in an integrating flame photometer for thc micro-analytical determination of Na and K in biological samples with an RSD, at the g of Na level, of 0-05.Other uses of photon counting have included locating the centrc of gravity of a spectral line whcn observed by a Fabry-Perot interferometer (988) and the measurement of exponential decay curves (895); in general, photon counting is of value only when weak light signals arc to be measured. A double-clipped digital correlator has been described (894); it works in real time at sampling rates up to 16 MHz for measuring the cor- relation of weak light beams such as the auto-correlation of light scattercd by a medium. The signal-to-noise ratio whcn using a single detector in either transform multiplex or scanning mode for the detection of optical images has been examined theoretically for photon noisc liniited situations (221, 1123). Both studies concluded that, whcn the principal sourcc of noisc was in the initial signal, scanning was superior except for simple images consisting of a few well-defined featurcs against a low-intensity background, for which the transform tcchniquc had the advantage.

ISSN:0306-1353    

DOI:10.1039/AA9740400020

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

22 Part I : Futidatneritals and Irtstrumentatiori 5 Data Processing 5.1 EMISSION SPECTROSCOPY A computer program for least squares regression analysis using background intensity valucs on either side of the wavelength of interest was used by Risova and Plsko (232) to provide a reliable basis for determining the limit of detection. Relative fluctuations of the back- ground signal were said by Friedhoff er al.(1285) to be constant throughout large ranges of a spectrum, so that measurements at only a few wavelengths were sufficient to establish the overall noise level in trace analysis. Using backgroundmeasurements at 320.0 nm and 203-9 nm for trace analysis in A Cu matrix, the detection limits of 16 elements were between 0.8 and 52 pg g-', Background measurements in low-alloy steel and cast iron were also discussed.A high-speed analog-to-digital converter, which permitted accurate measurement of sub- nanoampere currents over a very wide dynamic range, has been described (1330); the converter was applied to AF and AE analysis. Other references of interest - Calibration: 71, 873, 1028, 1118 Background correction: 8% 5.2 ABSORPTION SPECTROSCOPY Various computer programs for evaluating data obtained from AA spectrometers have been described (914, 955, 966, 1297).One (955) estimated the sample concentration, standard deviation, coefficient of variation and the 2a and 3a limits of confidence for the detcrmina- tion of up to 11 elements in a variety of samples. Roos (67) has now published his extended treatment of error functions in AAS. The major error component for most determinations was the noise asgociated with the dynamic nature of the flame. Other references of interest - Control charts: 1397 Errors: 709 Ifiterfacing: 979 On-stream analysis: 732

ISSN:0306-1353    

DOI:10.1039/AA9740400022

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

Part I : Fundamentals and Instrumentation 23 6 Complete Instruments The availability of commercial instruments has been reviewed again this year. Many new or modified emission spectrometers, atomic absorption spectrometers and electrothermal atomizers have been introduced for 1975, and these are included in Tables A, B and C, respectively, on pages 25 to 37. The tables represent information available at February, 1975. 6.1 ARC/SPARK EMISSION INSTRUMENTS This year has seen the introduction of plasma excitation sources from Applied Research Laboratories and Jarrell-Ash to add to that already available from Spectrametrics. This undoubtedly reflects the growing interest in plasmas as sources for multi-element determina- tions. Dreher and Frank (224) have described the design and construction of an inexpensive vacuum spectrograph.The instrument uses a Czerny-Turner type monochromator and a grating ruled with 1180 lines per mm to produce a reciprocal dispersion of 0.8 nm mm-l. With photographic detection, down to 130 nm can be recorded. 6.2 FLAME-BASED INSTRUMENTS Jongerius and de Galan (1233) have published a comprehensive survey of atomic absorption spectrometers and flameless atomizer accessories available in The Netherlands.The authors include in their survey the current prices for each instrument; this information is not included in the volumes of ARAAS as it is felt that these prices are of limited use because they vary in different countries and they become so rapidly outdated. Conversion of a Varian Techtron model AA5 single-beam spectrometer into a dual-beam instrument to permit simultaneous background correction has been achieved by the use of an off-axis continuum light source and a half-coated mirror (92).The authors give full electronic details. A multi-channel flame spectrometer for the simultaneous determination of K, Na and Ca in 0.15 ml of sample has been developed, based on the Russian ISP-51 spectrograph and a platinum wire spiral for sample introduction (1 175).Three exit slits from the monochromator are used to measure the emission signal for each element simultaneously. The construction of a single-beam flame spectrometer for AA, AE and AF has been described (390), based on a Zeiss Jena SPM-2 monochromator. An article (336), which could interest anyone who has to maintain his own AA spectro- meters, describes various aspects of instrument maintenance, and presents a table indicating symptoms associated with breakdowns of specific sections of an instrument.Anacon Instruments have introduced a MECA flame analyser to utilise the technique of molecular emission cavity analysis. At this stage in the development of the technique the instrumentation would seem to be rather highly priced for the limited range of applications so far published.Table A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS No.of {~~~~~~~ Wavelength Focal Special features Applications Model Type channels nm Der mm range/nm length Type Of source Supplier Angstrom Inc., 22-101 D.R. 60 0.278 210-430 3-0 m HV a.c. spark HV Optical interlock All ferrous and non- P.O.Box 248, 0-397 210-518 3.0 m or LV a.c. arc; LV spectrum monitor. ferrous applications Belleville, 0.556 210-634 3.0 m undirectional d.c. Various excitation stands Mich. 48111, U.S.A. acitor discharge arc, triggered cap- and read-out options V-70 D.R. 50 V-71 D.R. 22 0.556 160-440 1 * 5 m As model 22-1 01 0 * 556 178-320 0.75 m As model 22-101 A-70 D.R. 68 0.556 190-770 1 - 5 m As model 22-101 A-71 D.R. 22 0.556 210-350 0-75 m As model 22-101 As model 22-101. All wavelengths in vacuum. In-focus wavelength scanning 3 nm from each receiver slit Mobile and operable in hostile environments. All wavelengths in vacuum. Various excitation stands and read-out options As model V-70 As model V-70 Ferrous and non-ferrous metals, oils, soils and geological specimens As model V-70 As model V-70 but including determination of C, S and P As model A-70 Applied Research Laboratories, Ltd., Wingate Road, Luton, Beds., England (Entry continues on next page) Quanto- D.R. 20 1 -388 or 0.695 200-800 meter 20 (36 lines) 0.695 or 0.35 200400 Quanto- D.R. 28 0.70 or 0.35 175-500 vac 28 (49 lines) Quanto- D.R. 28 0-70 or 0.35 175-500 vac 28C Quanto- D.R. 6.0 0.46 170-407 vac 80 (84 lines) Quanto- D.R. 60 0.695 or 0.35 190-610 meter 80 (84 lines) Quanto- D.R. 48 0.35 or 0.175 190-520 meter (90 lines) 0-46 or 0.23 190-630 29000B 0.56 or 0.28 190-705 0.695 or 0.35 190-840 0.75 m Low voltage, high voltage 0.5 m As Quantovac 80 0.5 m As Quantovac 28 1.0 m Various, low volt- age, high voltage, multi-source (HVS, LV, d.c. arc) 1 * O m As Quantovac 80 1.5 m As Quantovac 80 Air or argon excitation stands.Typewriter and digital computer options As Quantovac 80 but no air-condi t ioner Complete computer control, teletype or visual display output. Off-line computer links Typewriter, teletype and digital computer options. Single or dual stand options. Second stand can be argon or air. Built-in instrument air-conditioning As Quantovac 80 Typewriter, teletype and digital computer options.Argon and/or air stands available 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 80Table A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS - corztirzued No. of :$:%::) Wavelength Focal Special features Applications Ap p I ied Quanto- D.R. (64 lines) 0.695 or 0.35 190-610 1.0 m As Quantovac 80 Automated sequential As Quantometer 80 Research meter (8 reference) analysis Laboratories 33000 Ltd.Mode' Type channels nm per mm range/nm length Type Of Supplier (continued) Quanto- D.R. As Quanto- 0.46 170-407 1.0 m As Quantovac 80 As Quantometer 33000 As Quantovac 80 vac meter 33000 33600 meter meter 33000 to 24 samples Quanto- D.R. As Quanto- 0.695 or 0-35 190-610 1.0 m H.F. plasma Automatic loading of up Solutions 33000CA Q.A. 137 D.R. 48 0.46 185410 1.0 m R.F. plasma P.p.b. analysis Solutions Baird Atomic SB-1 Phot. Inc., 125 Middlesex TurnDike, SH-1 Phot. Bedford,. Mass. 01730, U.S.A. GW-1 Phot. GWR-1 Phot. Spectto- D.R. met 1000 - 1 a5 or 0.75 370-740 1.5 m Arc or spark Built-in order sorter General spectrographic - 1.0 450-750 1.5 m Arc or spark Built-in order sorter General spectrographic - 0 - 8 or 0 - 4 185-2400 2.0 m Arc or spark, Dual gratings for simul- General spectrographic analysis analysis modular or RE-1 taneous photography of analysis two spectral regions - 0.8 or 0.4 185-2400 2.0 m As GW-1 High speed (f/15.5) Transient, weak or "cl gratings for rapid exam- complex sources.Q and/or weak sources of analysis ? i? air-conditioning require- using C 793.1 nm, 3 alignment 9 s" ination of transient General spectrographic - 31 complex spectra.Optional echelle grating for f/12 aperture 5 3 ments. Logarithmic P 214.9 nm in 2nd s read-out. Manual master order. Non-ferrous 6 monitor t o check slit Easy interchange of General spectrographic gratings. Easy conversion analysis t o a direct reader (GX-3) Automatic optical servo monitor continuously capabilities as well as 2 maintains correct slit photographic - E alignment.Logarithmic (see GX-1) read-out. Precise elec- 30 0.6 or 0-3 210-590 1.0 m Arc or spark, Compact, low-cost direct Ferrous metals (except modular reader with minimum determination of S) metals, oils a 2 Normal direct reading $ -. tronic setting of slits g GX-1 Phot. - 0.55 or 0.28 180-2250 3.0 m As GW-1 GX-3 D.R. 16per head 0-39 180-2250 3-Om AsGW-1 (normally 3 heads per unit)Spectro- D.R.met I I in 'Varisource unit including spark, low and high-voltage a.c. arcs, d.c. arcs. p s o versatile controlled wave-. form excitation source'. Plasma 180-1 500 180-750 or 2.0 m 60 2 analysis analysis 20 inch camera General spectrographic 2 3 Choice of 3 gratings. Versatile instrument %* 5 Nitrogen purging extends particularly suitable range to 175 nm.for measuring transient 3 Optional accessories 0-294 190-432 2.0 m As Spectromet Automatic optical servo 0.59 190-863 1000 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 Spectro- D.R. 60 0.29 173-432 2.0 m As Spectromet As Spectromet II. All vac II 1000 photomultipliers in vacuum 3 All direct-reader applications above 190 nm z 2 '? 5 2 $ 0" F A l l direct-reader applications including C, P and S b Jarrell-Ash Div., Fisher Scientific Co., 590 Lincoln St., Walt ham, Mass. 02154, U 3 . A V. A. Howe & Co. Ltd., 88 Peterborough Road, London S.W.6, England 78-000 70-31 0 75-1 50 90-750 90-785 1500 Atorn- counter 70-31 4 82-000 (78-489) 78-490 25-020 78-420 78-460 82-400 ( 82-41 0) Phot.Phot. Phot. D.R. D.R. D.R. D.R. Scan. Scan. Scan. Scan. Scan. Scan. 1.1 or 0-54 1.0 to 0.24 depending upon grating Up to 50 Up to 50 Up to 30 30 + 4.4 to 1 *1 3.2 to 0-8 1 - 6 to 0.4 0.54 0.54 0*56 Or o.28 ( 0.34 or 0.17 As 70-310 Depends upon 1 grating i selected 3-28 permit use as direct I reader or scanning I spectrometer 168-500 0-75 rn As above, except Computer controlled 163-500 0.75 m trolled peak 200-800 or 1.5 m As above Choice of 2 gratings 200-510 or j 19CL250 As 70-310 3.4 m As above Scanning optional. Easy 1 electronic con- current 190-400 1 interchanae to Dhoto- spectra Most metallurgical analyses All direct-reader applications above 190 nm All direct-reader applications above graphic (70-310) and 190 nm scanning version (70-320) Various scanning spectro- Suitable for spectro- meters, some with scopic investigations vacuum capabilities rather than for analytical applications 1.0 m N 200-goo 0-25 rnTable A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS - continued No.of ~ ~ ~ ~ ~ $ , Wavelength Focal Special features Model lype channels nm per mm range/nm length Type Of Source Supplier Applications Jobi n-Yvon, VARAF Scan.- 1-8 or 0.9 200-800 0.465 m Flame Czerny-Turner mono- Liquids and solutions 1 Rue du Canal, chromator, adjustable 91 160 bandwidth 0.02-4 nm. Longjumeau, Nine models with various France burner and read-out options chromator, adjustable bandwidth 0.024 nm.Automatic wavelength scanning device. Pneumatic nebuliser, laminar flow burner. Ultrasonic nebuliser optional DELTA Scan. - 1.2 195-770 0.6 m Flame Czerny-Turner mono- Liquids and solutions 190-900 1 - 5 m 'Transource' Ferrous and non-ferrous Labtest ' 310 D.R. 60 0.56 Equipment Co., alloys 40 0.67 170-550 1.0 m As above 11828 La Grange v-25 D.R. Ave . , Los Angeles, Calif. 90025, U.S.A. As above 2100 D.R. 30 0-46 188455 1.0 m As above As above 71 D.R. 74 0-52 170-900 2-Cm As above General purpose M.B.L.E., PV 8300 D.R. 60 0.55 or 0.46 170-430 1-5 m Triggered capac- Optional dual air/argon Steels, iron, non-ferrous 80 Rue des Vacuum (80 lines) itor discharge. excitation stand. Printer metals, oils, non- Deux Gares, High voltage spayk.teletype or computer conducting powders 1070 Bruxelles, Monoalternance read-out systems Belgium triggered capac- itor discharge, d.c.arc, inter- mittent d.c. arc PV8210 D.R. Up to60 0.55 190-700 i - 5 m AS PV8300 Air Wavelength range Solids, liquids and covered in 1st order. powders. Non-ferrous Remote controlled roving metals and oils detector. External excitation. Rotrode and inert atmosphere faci tities Optica S.A.S., 85 Phot.- 0.69-0-36 200-800 1 -2 m All conventional Stigmatic instrument General purpose Via Gargano 21, types available with rotating Ebert 20139 Milano, grating Italy B5C D.R. 16 0.69 or 0.36 220-420 1-2 m LV triggered arc Double spark stand both General purpose. and spark. HV in air and inert Metallurgical analysis, spark, a.c. and atmmtphere. Rotrode e.g., Al, Pb, Zn, Fe, d.c.arc for solutions Cu alloys. Wear metals in oils etc.B7V D.R. 93 ESAl Scan. - ESA3 D.R. 9 ESA4 Scan. - g 4 .? 2 a" 0.37 165440 1.5 rn LV triggered arc Air-vacuum instrument Complex analyses and spark. HV with all exit slits involving many spectral spark accessible from outside lines for adjustment. Many analytical programmes can be arranged in parallel for easy inter- change.Computer facilities available 0.41 200-500 1 .O m Controlled and Scanning monochromator Metallurgical work. All non-controlled HV with one channel for material excitable with spark, a x . arc analytical line and same source parameter another channel for reference using reflected beam principle and polychromator. All system 0.36 160-500 1.2 m LV triggered arc Combined vacuum mono- Routine analysis (40 nm as and spark of (including iron and C, steel.S and P) POlY- excitable elements chromator) accessible with scanning Non-ferrous alloys 0.41 165-500 1.0 m LV triggered arc. Scanning vacuum mono- Metallurgical work. HV spark, a.c. arc chromator with one Analysis of ferrous and channel for analytical non-ferrous alloys line and another channel for reference.Facilities for analysing two elements simultaneously Ferrous and non-ferrous Rank Precision El000 D.R. 60 0.293-1.155 159-6864-3 1.5 m Various, including Dual spark stands. Industries Ltd., Polyvac high repetition Computer-controlled alloys. Geological Analytical Div., condensed arc instrument. Dual grat- samples. Wear metals Westwood ings give 6 systems in oil Industrial Estate, E952 D,R.Ramsgate Road, Polyvac Margate, Kent CT9 4JL, England 36 0.546 or 174.0-447.7 0.75 m As El000 Curved entrance and Ferrous and non-ferrous 0-741 236.5-607'4 exit slits. Solid state alloys. Wear metals electronics or computer in oils controlled. Air or vacuum €742/3 Phot. - Not stated 191-800 1-57 m D.C. arc, HV Adjustable slit. Spectral Analysis of high-purity Large spark condensed length 0.67 m, of which specimens having Quartz/ arc 0-24 m can be selected complex spectra.Glass for a given exposure Determinations of trace element concentrations E777/8 Phot. - 0-26-0-97 200-1200 1 a 5 m Various Czerny-Turner mono- Routine qualitative and chromator quantitative analysis. Flash photolysis. Examination of line profiles E549 D.R. 72 0.5-10 200-600 0.53 m HV spark, con- Solid state electronics. Non-ferrous metals. densed arc, d.c. Quartz plate for Soils. Additives and controlled a.c. compensation arc Medium Quartz arc, thyratron- atmospheric pressure wear metals in oils N Ww 0 Table A COMMERCIALLY AVAlLABLE EMISSION SPECTROMETERS - continued Reciprocal No* Of dispersion/ Wave'ength Type of source Special features Supplier Model Type channels nm per mm range/nm Appllcations R.S.V.GmbH., SPN 3.5 Phot. 30 0.14-0.48 200-1000 3.5 m Glow discharge Paschen-Runge mounting General analysis 8031 Hechendorf D.R. lamp, high, specially designed for Pilsensee, medium or low range below 200 nm. West Germany voltage spark, a.c. Direct reading attach- or d.c. arc, ment available Siemens Ltd., continuous and Great West intermittent West House, Road.Great SPN 2.0 P_hO_t. 30 0.24-0.84 200-1000 2.0 m As above As above General analysis Brentford, D.H. Midd lesex, SPN 1.5 Phot. 15 0-37-1 $1 200-1000 1.5 m As above Direct reading General analysis England D.R. attachment available SPN 1.0 Phot. SPV 1.0 Phot. Analymat D.R. ivac) Lair Analymat D.R. I I -vac Analymat D.R.I I I-vac Analy- D.R. meter ~~ ~ S hi madzu GE-100 Phot. Seisakusho Ltd., 14-5 Uchikanda GE-170 Phot. - 0 56-1 .7 200-1 000 - 0.4-1 .7 300-1 300 40 0.31 or 0.54 200-650 40 0-31 or 0.54 150-490 40 0.42 or 0.5 110-500 Scan. 0-12 200-800 - 1.66 or 0.83 200- - 0.48 200- 1.0 m As above 1-Om Asabove 7 . 5 m Glow discharge lamp (others available) 1.5m Asabove 1.0 m As above 2-Om Asabove Direct reading attachment available As above Exhibits no background: no matrix effects.Linear calibration for all elements 0-1 00% As above As above As above ~~~~ 7 .O m Not stated Not stated 1.7 m Not stated Not stated General analysis General analysis General analysis As above As above As above *FI .? 9 -b - - 2 3 cs, 1-chome, Chiyoda-ku, GVM-100 D.R. 24 0.46 170-410 1.0 m Modular unit, Not stated Tokyo 101, low and high -_ 3 - s Japan voltage spark i;; Inc., D.R.using a high dispersion 2 Spectrametrics AE 2 Phot. 1 0.06 190-900 0.75 m Plasma jet Optimised AE system , Routine analysis i3 204 Andover St., high energy throughput 190-900 0.75 m Plasma jet i- echelle spectrometer 3 Routine quantitative a Andover, D.R.10 D.R. 20 0-06 Mass. 01810, (inter- and a high temperature multi-element analysis U.S.A.changeable i~;: jet excitation 2 HA8 8JP, wavelength) s. cassettes) Techmation ES 9 Phot. - 0.06 190-900 0.75 in Plasma jet, flame Built-in computer 3 Ltd., or arc stand 58 Edgware Qualitative and semi- Way, Edgware, RS 1 D.R. 1 0.06 190-900 0.75 m Plasma jet, flame quantitative analysis. Middlesex (variable or arc stand Spectroscopic research ' England xRoutine analysis Soex Industries 1870 Scan.- 1.6 175-1280 0.5m - Multi-purpose unit lric., 3880 Park Ave., Metuchen, N.J. 08840. 1702 1704 Fho?. Phot. U.S.A. 1802 Phot. Glen Creston, The Red House, 37 The Broadway, Stanmore, Middlesex HA7 4DL. England 1 - 1 0.8 175-1500 0-75 m - 175-1500 1-Om - 180-1500 1.0 m -- - 0.8 Research Research Direct reading Routine analysis accessory available VEB Carl Zeiss Jena, 69 Jena, Carl-Zeiss Str. 1, German Democratic Republic Carl Zeiss Jena Ltd., 93/97 New Cavendish St., London W1A 2AR, England. DSA-240 D.R. Up to 56 0.78 210-550 0-54 m Spark Up to 11 lines analysable in single, automatic scan. Choice of reference line. Bui It-in automatic temperature and pressure compensation. Digital signal averaging, display and recording.Rapid change of analysis pro- gramme by programme store measuring range. Stigmatic depiction. Dispersion doubled by double passage of light. Pre-disperser for order sorting and isolation. Gratings interchangeable. Automatic transport of plate holder accessories available PGSQ Phot. - 0.74 or 0.37 200-2800 2.075 rn Arc or spark Automatic expansion of (2-24 Phot.- 0.76 210-550 0.54 m Arc or spark Full range of A l l steels and non- ferrous alloys.Bearing metals and solders. Mineral oils and aqueous solutions General spectrographic analysis. Also examina- tion of line profiles, hyperfine structure etc. General spectrographic analysisw p3 Table B COMMERCIALLY AVAILABLE ATOMIC ABSORPTION SPECTROMETERS - Single/ Grating Reciprocal Supplier Model double Monochromator lines dispergion/ Resolution Wavelength Read-out; Other features beam per mm nm per mm nm range/nm e x ~ ~ ~ ~ o n Beckman Instruments, 485 Double Littrow 1200 2.7 0.2 190-860 Meter; x 50 Single and triple pass 2500 Harbor Boulevard, optics; automatic filter Fullerton, Calif. 92634, selection U.S.A. 495 Double Littrow 1200 2.7 0.2 190-860 Digital; x 100 As model 485 Beckman Instrument GmbH, 1233 Double Littrow 1200 2 - 7 0.2 19C860 Meter; x 55 Single and triple-pass 8 Munich 45, Frankfurter Ring 11 5, West Germany optics; YO T, abs.or concn. read-out 1236 Double Littrow 1200 2.7 0.2 190-860 Digital: x 55 As model 1233 Beckman RllC Ltd., Eastfield Trading Estate, 1248 Double Littrow 1200 2.7 0-2 190-860 Meter; x 55 Auto zero and calibrate Glenrothes, Fife, KY7 4NG, Scotland 1272 Double Littrow 1200 2.7 0-2 190-860 Digital; x 55 As model 1248 Carl Zeiss, FMD 3 Single 600 0.05 193-900 Digital 4-lamp turret, 2 7082 Oberkochen, stabilized power supplies.Wuerttemberg, Curve correction, auto West Germany zero, optional auto calibrate and background correction ki Corn i ng-E EL, EEL 140 Single 0-25 m modified 1180 3-5 Non-linear Single-lamp turret ? St.Andrews Works, E bert-Fastie meter Halstead, Essex, 9 England EEL 240 Single As EEL 140 1180 3.5 Meter 4-1amp turret; integration 5 c) Diano Corporation, Multi- Single Double grating 1200 1.5 0.2 190-800 Meter: x 10 3-lamp turret with 3 75 Forbes Boulevard, spec 0-25 m modified stabilized power supplies; Mansfield, Mass. 02048, 2-20 Czerny-Turner 4-way gas control; % T, U.S.A. abs.or concn. read-out 2 b 5 GCA/McPherson Instrument, EU 703 Single 1180 2.0 0.1 180-1000 Digital Modular; various R 530 Main St., Acton. detectors and gratings rr 2 Mass. 01720, U.S.A. 5 2 Hitachi Ltd., 208 Single Czerny-Turner 1440 1.8 Nissei Sangyo Co. Ltd., 15-12 Nishi-Shimbashi, 2-Chome, Minato-Ku, Tokyo, Japan x 20 3-lamp turret: wavelength drive.Concn. read-out 5, 2.Instrumentation 453 Double; A. 0.33 m Ebert A. 1200 Laboratory inc., dual 113 Hartwell Avenue, beam B. 0.16 m Ebert B. 1200 Lexington, Mass. 02173, U.S.A. 1 nstrumentation Laboratory (U.K.) Ltd., 353 Double; A. 0-33 m Ebert A. 1200 Station House, dual B. Interference - Stamford New Road, beam filters Altrincham, Cheshire, England 251 Double 0.33 m Ebert 1200 151 Single 0.33 m Ebert 1200 A. 2.5 A. 0.03 A. 190-900 Digital; x 250 6-lamp turret, wavelength drive; integration, auto B. 5.0 B. 0.2 B, 190-900 zero and calibrate. Abs. or concn. read-out. Curve correction, auto back- ground correction and internal standard mode A. 2.5 A. 0.03 A. 190-900 Digital; x 250As model 453 - B. 2.5- B. 248-766 6.0 2 - 5 0.03 190-900 Digital; X 50 Integration; abs.or concn. read-out; auto zero, curve correction 2.5 0.03 190-900 Digital; X 50 As model 251 Jarrell-Ash/Fisher, 590 Lincoln Street, Waltham, Mass. 02154, U.S.A. V. A. Howe and Co. Ltd., 88 Peterborough Road, London, S.W.6, England Dial Atom I I 650 Atom- sorb 82-500 82-81 0 82-81 1 82-850 Single 0.25 m Czerny- 1180 3.3 Turner 190-900 Meter; x 10 2-lamp turret; f/7.5 aperture Single 0.25 m Czerny- 1180 3.3 Turner 190-900 Digital; x 10 Single 0.25 m Ebert 1180 Meter; x 20 Single 0-5 m Ebert 1180 x 20 Double; Two 0-4 m 1180 2.1 0.03 dual Ebert channel Double; Two 0.4 m 1180 2.1 0.03 dual Ebert channel Double 0.4 m Ebert, 2 gratings Digital Digital As Dial Atom II 6-lamp turret: f/3.6 Various optics available Wavelength scan; integration; auto zero As 82-810, plus computer- controlled read-out Computer-controlled parameters and measurements Jobin-Yvon, VARAF Single 0-465 m 1220 1.8 1 Rue du Canal, Czerny-Turner 91160 Longjumeau, France DELTA Single 0.6m Czerny- 1220 1.2 Turner Digital Digital Wavelength scan Ootica.Via Gargano 20139 Milan, 21, Italy 6000 Single 0.35 m Ebert Digital; x 50 Auto filter insertion, auto concn., integration.Flame temp. regulation. Pre- focussed water-cooled hollow-cathode lamps availableTable B COMMERCIALLY AVAILABLE ATOMIC ABSORPTION SPECTROMETERS - continued Single/ Grating Reciprocal Supplier Model double Monochromalor lines dispersion/ Resolution Re:%?; Other features beam per mm nm per mm nm range/nm exmnsion Perkin-Elmer Corp., 103 Norwalk, Conn. 06856, U.S.A.Perkin-Elmer Ltd., 107 Post Office Lane, Beaconsfield, 360 Bucks. HP9 lQA, England 370 3056 306 503 Single 0.27 m Littrow Single 0.27 m Littrow Double 0-27 m Littrow Double 0.27 m Littrow Double 0.4 m Czerny- Turner Double 0-4 m Czerny- Turner Double 0-4 m Czerny- Turner 1800 1800 1800 1800 U.V. 2880 vis. 1440 U.V. 3880 vis. 1440 U.V. 2880 vis. 1440 1.6 0.2 190-860 Meter; x 50 1.6 0.2 190-860 Digital; x 50 1.6 0.2 190-860 Meter; x 50 and x 0.5 1.6 0.2 190-860 Digital: x 50 0.65 0-03 180-440 Meter; x 100 1.3 400-900 and x 0.5 All mirror optics: integration: auto zero and flame ignition As Model 103 All mirror optics; auto zero, gain control, flame ignition and optional nitrous oxide switching; integration; curve correction; peak reader As Model 360 All-mirror optics: auto zero and flame ignition: peak reader; integration: curve correction 0-65 0.03 180-440 Digital; x 100 As Model 3058 plus 1.3 400-900 and x 0.1 auto concn. 0.65 0.03 180-440 Digital; x 100 As Model 306 plus signal 1.3 400-900 and x 0.1 averaging: auto nitrous oxide switching; flame, pressure and burner head sensors Perkin-Elmer & Co. GmbH, 400s Double 0.33 m Czerny- 1800 1.3 0.2 190-860 Meter; x 50 Auto zero and flame Postfach 1120, Turner and X 0.2 ignition; integration 7770 Ueberlingen, West Germany 400 Double 0.33 m Czerny- 1800 1 - 3 0.2 190-860 Digital; x 50 As Model 400s plus auto Turner and X 0.2 concn.; curve correction ~ ~~ Pye Unicam Ltd., SP 191 Single Ebert 1200 3,3 e-2 190-850 Digital; x 25 4-lamp turret; auto zero York Strset, and x 0.1 and ignition; integration; Cambridge, CB1 2PX, curve correction England - SP 190 Single Ebert 1200 3.3 0.2 190-850 Digital; X 25 As Model 391 without and X 0.1 emission capability SP 1950 Double Ebert 1800 2.2 0.1 19s850 Digital; x 20 Auto zero and ignition; and x 0.1 integration: curve correction SP 1900 Double Ebert 1800 2.2 0.1 190-850 Digital; x 20 As Model 1950 plus and x 0.1 6-lamp turretRank Precison Industries Ltd., ATOM- Single Silica prism - 1 * 7 at Analytical Division, SPEK 200 nm Westwood Industrial Estate, H 1170 44-6 at Ramasate Road. 500 nm Meter 6-lamp turret Margate, Kent, CT9 4JL, England ATOM- Single Czerny-Turner 1200 2.6 0.1 190-850 Digital 6-lamp turret; auto zero SPEK and flame ignition; H 1550 curve correction; intearation Seiko Instruments, SAS 721 Single Tokyo, Japan SAS 740 Couble; dual channel Microcomputer and line printer ~~~ Shandon Southern A3600 Single 0.25 m Czerny- 632 6.0 0.2 190-860 Meter; x 25 Integration Instruments Ltd., Turner Frimley Road, Camberley, Surrey, GU16 5ET, England A3400 Single 0.25 m Czerny- 632 6.0 0.2 190-860 Meter; x 25 4-lamp turret; auto zero Shandon Labortechnik correction; integration GmbH, wavelength scan Frankfurt/Main 50, West Germany Turner and flame ignition; curve Shimadzo-Seisakusho Ltd., AA610 Single Czerny-Turner Meter; x 10 2-lamp turret; 14-5 Uchikana 1-chome, wavelength drive Chi yod a-Ku, AA610S Tokyo 101, Japan AA600 Varian Techtron Pty.Ltd., 1100/ Single 0.25 m Czerny- 1276 2.8 0.2 185-900 Meter/ 679 Springvale Road, 1200 Turner North SDrinsvale, Vic., Digital; x 50 Australia 3171 Varian Associates Ltd., 1150/ Single; 0.25 m Czerny- 1276 Russell House, 1250 dual Turner Molesey Road, channel Walton on Thames, Surrey, England AA6 Single; 0.51 m Ebert 638 Varian Instrument Div., channel 611 Hansen Way, Palo Alto, Calif. 94303, U.S.A. dual 2.8 3-3 0-2 185-900 Meter/ Digital; x 50 0.05 185-1000 Meter or digital; X 0.3-x 50 4-lamp turret; auto zero: integration; curve correction, peak reader; f/8 aperture As Models 1100/1200 plus simultaneous background correction Modular construction; 4-lamp turret; auto zero and correction; calibrate; integration; curve peak reader; f/10 aperture VEB Carl Zeiss Jena, AAS 1 Single 1300 1 - 5 190-820 Meter; x 10 4-lamp turret; auto zero. 69 Jena, Carl-Zeiss-Str. 1, optics German Democratic Republic Carl Zeiss Jena Ltd., 93/97 New Cavendish St., London W1A 2AR, England Single or triple pass ww Table C COMMERCIALLY AVAILABLE ELECTROTHERMAL ATOMIZERS m Sensltivity for 1% abs. (S) c u Si Special features ","dS;E"),"1" Control unit Detection limlt (d.1.) Supplier Model Type Barnes Engineering Co., Glomax Tantalum strip 50 Programmable; d.1. 10-11 g N.S. Fits most AA spectro- 30 Commerce Road, dry, ash, atomize, (50 & I ) meters. Air-cooled: inert- Stamford, burn off. Max. gas and hydrogen Conn. 06902, U.S.A. temp. 2400°C shielding Beckman Instruments GmbH, 1268 8 Munich 45, Frankfurter Ring 115, Graphite furnace West Germany - I nstrumentation Laboratory Inc., lL355 113 Hartwell Avenue, Lexington, Mass. 02173, U.S.A. Tantalum strip iL455 Tungsten and graphite furnace 100 Programmable; d.1. 4 x 10-12 g d.1. 10-11 g dry, ash, atomize, (100 & I ) (100 fill burn off. Max. temp. 3100°C 50 Programmable; d.1. 10-11 g d.1. 10-8 g dry, atomize. Max. (25 p l ) (25 P I ) temp. 2700°C 80 Programmable; S. 2 x 10-11 g S. g six stages. Max. temp. 3500°C Water-cooled; inert-gas and hydrogen shielding. Safety feature for failure of water or purge gas Fits most AA spectro- meters.Air-cooled; inert- gas and hydrogen shielding. Maintains constant current during atom isat ion Safety interlock system. Automatic furnace cleaning ~~ Jarrell-Ash Fisher. MTA-2 Tantalum strip 50 Programmable; d.1. 2 x 10-12 g N.S. Fits most AA spectro- Waltham, Max. temp. 2400°C hydrogen shielding Mass. 02154, U.S.A. 590 Lincoln Street, dry, ash, atomize. (50 p I ) meters. Inert-gas and b 5 FLA 10 Graphite furnace 50 Programmable; d.1. 2 X g N.S. Fits most AA spectro- 7 2 S. & J. Juniper & Co., 110 Graphite furnace 50 Programmable: S. 3 X g N.S. Water-cooled; inert-gas & stages cover full ? temp. 3500°C temperature range - 3 Via Gargano 21, dry, ash, atomize (50 p l ) shielding 9 dry, ash, atomize.(5 pI) meters. Inert-gas shielding Max. temp. 2800°C but an air ash possible shielding. All programme $ - 7 Potter Street dry, ash, atomize, (10 p l ) Harlow, Essex, burn out. Max. England Optica S.A.S., CAT 6 Tantalum strip 50 Programmable; d.1. 10-rl g N.S. Water-cooled; inert-gas 20139 Milan, Italy Postfach 1120, 7770 Ueberlingen, atomize. Max. Water-cooled; inert-gas - 3 - L 3 k Perkin-Elmer & Co.GmbH, HGA 74 Graphite furnace 100 Programmable; d.1. 2 X g d.1. 5 x 10-11 g Fits Perkin-Elmer and - % West Germany temp. 2700°C shielding. Permits ramp ? dry, ash (21, (100 4) (100 PI) Zeiss AA spectrometers. 5 2 ashing, gas stop opera- Perkin-Elmer Corp., tion. Closed system. Norwalk, Safety feature for failure Conn. 06856, U.S.A. of water or purge gas w -.HGA 2100 Graphite furnace 100 Programmable; d.1. 2 x 10-12 g d.1. 5 x 10-11 g Water-cooled, inert-gas 5 dry, ash, atomize. (100 P I ) (100 PI) shielding. Gas-stop and * - Max. temp. 2800°C reduced gas flow opera- Y tion. Closed system. LI Constant current operation .. and direct temp. "r! calibration. Safety feature for failure of water or p\ purge gas a 2 Pye Unicam Ltd., Graphite furnace 100 Programmable; d.1. 4 x 10-12 g d.1. g Water-cooled; inert-gas .-r York Street, dry, ash, atomize, (100 P I ) (100 PI) and hydrogen shielding, $ England temp. 3100°C of water 0; purge gas 2 * - Cambridge CBI 2PX, burn off. Max. Safety feature for failure a Rank Precision Industries Ltd., H1975/ Graphite furnace 100 Programmable; S. 5 x 10-11 g N.S. Water-cooled; inert-gas 5 Analytical Division, FA256 dry, ash, wait, shielding 2 s 9 Shandon Southern A3470 Graphite rod 25 Programmable; d.1. 5 x 10-!2 g d.1. 6 x 10-11 g Fits selected AA spectro- 8 $ Westwood Industrial Estate, atomize. Max. Ramsgate Road, temp. 2600°C Margate, Kent, CT9 4JL, England -+ -. Instruments Ltd., dry, ash (21, ( 5 PI) (5 P I ) meters. Air-cooled; inert- Frimley Road, atomize. Max. gas shielding. Pyrolytic Camberley, Surrey, GU16 5ET. temp. 3000°C graphite coating for rods England available in situ. Flame- less accessory can be left in instrument during flame measurements Shimadzu-Seisakusho Ltd., GFA-1 Graphite furnace 14-5 Uchikana 1-Chome, Chiyoda-Ku, Tokyo 101, Japan Varian Techtron Pty. Ltd., 63 Graphite furnace 5 Programmable; d.1. 4 x 10-12 g d.1. 8 x 10-1' g Fits AA spectrometers 679 Springvale Road, dry, ash, atomize. (5 PI) (5 PI) with optical aperture f/8 North Springvale, Graphite cup 25 Max. temp. 3000°C or less. Water-cooled: Vic. 3171, Australia inert-gas and hydrogen shielding. Pyrolytic graphite coating on cups and tubes

ISSN:0306-1353    

DOI:10.1039/AA9740400023

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

24 7 Ancillary Information 7.1 STANDARDS The information on standards supplied in ARAAS, 1973, 3, 43-45 is still valid, but in addition to this a comprehensive list of spectrochemical standards has been compiled by M.B.L.E. (1534). Two new standard rcference materials have been prepared by the National Bureau of Standards. Standard reference material No. 1579 is a powdered Pb-based paint intended for the calibration of methods used in the determination of Pb in old paint samples (531).Standard mercury solutions at the 1.5 pg ml-l and 1.2 ng ml-' levels were prepared by acidifying with nitric acid and stabilising with trace amounts of Au (1403). Four standard geochemical samples prepared by the Association Nationale de la Recherche Technique have been analysed by 99 laboratories and recommended values proposed for the major and minor elements (1 520).An interesting technique has been described which permits generation of very low levels of Hg vapour for dynamic testing of air monitoring instruments (1402). Five types of plastics, after impregnation with dimethylmercury, could bc maintained in a controlled temperature air-flow to regenerate the dimethylmercury at a predictable rate.A simple pyrolyzer also permitted the dimethylmercury to be converted to elemental Hg. One of the greatest problems involved in thc determination of impurities in solid samples below 0-1 p.p.m. is the availability of sufficiently pure acids to dissolve the materials. Work is currently being considered by the Commission of the European Community to analyse comprehensively commercially available high purity acids.The ultimate intention is to secure the production of higher purity acids, an objective that can be achieved only by the development of universally acceptable methods of analysis at these extremely low impurity levels. 7.2 DOCUMENTATION An extremely useful listing of books, journals, abstracts, conference proceedings and wave- length tables, of interest to workers in the field of optical emission spectrometry, has been compiled by M.B.L.E. (1013) and, for workers in the field of AAS, Perkin-Elmer have continued to publish their listing of relevant references (268, 773). Reviews published during the year are listed elsewhere in this book, but one worthy of mention here is the critical review, by Browner, of AFS as an analytical technique (1130).

ISSN:0306-1353    

DOI:10.1039/AA9740400024

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

PART II METHODOLOGYI nt rod uct ion In Part 11, the term 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 Techniques, 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. 41

ISSN:0306-1353    

DOI:10.1039/AA9740400039

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

EXII'LANATION 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 accomp- anying text discusscs only the more noteworthy contributions. These Applications Tables form a convenient source of information for analysts interested in particular elements, matrices, sample trcatments or atomization systems.In many cascs, 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 h/nm &fAl''RIX CONCENTRATION TECH. ANALY'rE SAMPIA.G 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, expresscd as 76 or pg g-' for solids and mg 1-* or pg ml-' for liquids. Thc atomic spectroscopy technique is indicated by A (absorp- tion), E (emission) or F (fluorescence). The form of the sample, as presented to the instrument, is indicatcd by S (solid), L (liquid) or G (gas or vapour). 'd.1.' = detection limit in the analyte. A brief indication is given of the sample pretreatment 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(s) of the author(s), with address. 41

ISSN:0306-1353    

DOI:10.1039/AA9740400041

出版商:RSC    

年代:1974

数据来源: RSC

摘要:

42 Part II: Methodology 1 General Techniques 1.1 NEW METHODS 1.1.1 Introduction With many techniques, particularly those as old as emission spectroscopy, very few methods can genuinely be described as new. The methods discussed in this section contain dcvelop- ments in methodology, via the use of better materials, electronic and electromechanical devices, that may enable techniques previously restricted to specialist rcsearch laboratories to become applicable to routine analysis.In can often be difficult, when evaluating new analytical techniques, to decide on the best ‘over-all’ procedure. A statistical method of ‘steep slopc’ analysis has been suggested by workers in the University of Lenin in Tashkent (753) as a solution to this problem. 1 .I .2 Samplie Preparation Thc collection of trace elcments on specifically designed chelating resins has been used for the selcctive concentration of low levels of Ag and transition metals prior to analysis (800).Another pre-concentration reagent was an ultrasonically prepared Hg / H,O emulsion, which was used to collect Ag, Au and Pd (442, 443). The solubility of many metals in Hg might possibly extend this tcchnique to other systems.The relatively high solubility of MIBK in water can bc a disadvantagc: Etcrson and Parker (1538) reported that 3-hcptanone is one-fifth as soluble as MTBK in water, whcrcas the solubilities of metal chelates in the solvents are similar. The usc of 3-heptanone should result in greater extraction cfficiency. However, this ‘new’ solvent has a lower nebulization efficiency than MIBK. 6.1.3 Emission The d.c. arc, one of the most sensitive of spectrocheniical excitation devices, is particularly useful for multi-element analyses. Vukanovic et al. (558) have reported the development of a novel excitation system in which a horizontal arc burning in a graphite tubc was sub- jected to an inhomogeneous magnetic field. The detection limits €or Pb and In were 400 pg and 600 pg, respectively.Karyakin et al. (240) also varied the magnetic field around a normal 5 to 10 A d.c. arc operation. Although this work has bcen carried out on solutions, it could provoke some fruitful invcstigation on solid samples. Barnes has continued his work on the liquid-layer-on-solid-sample spark technique (111). This method was investigated for the determination of Si in A1 in the range 0.08 to 1076, 2nd for Ni, Fe and Cr in steels.It was found that onc analytical curve could replace several that were previously necessary to cover the range. The results of future investigations should be well worth watching. Drehcr and Frank (225) used an iitductively coupled r.f. plussma to determine As and I in the vacuum ultraviolet.Although the method was somewhat tedious, these clcments arc difilcult to determinc by alternative techniques. 1.1.4 Absorption Prudnikov (209) obtained sensitivity increases OC up to 10-fold in FAAS by combining pulsed electrothermal vaporization into the flame with a ‘IT-shaped quartz absorption tube. Modifications of the graphite rube technique have bcen rcportcd (261, 526).Rcnshaw (261) used a tantalum foil liner lo prevent Ba combining with the graphite to form involatilc: Ba carbidcs. Kuzovlev and co-workers (526) prevented carbide formation by filling the cuvette with an organic Ta or Nb compound, rapidly heating it (2-3 s ) to 675-725 K and maintaining this temperature for 1-2 min. The cuvettc wall becamc coated with a bright film, 100-150 ym thick, of the Nb or Ta metal.The lifc of the cuvette did not appear to be reduced provided the device was not heated above 2750 K.Part II: Methodology 43 A reduction in the losses of Hg from cups and tubes was achieved by previously plating them with gold (103). Up to 90% of Hg was lost when this modification was not made. Once again, the limitation in use is that the cup must be kept below the melting point of the coating element, in this case Au (1336 K). 1.1.5 Fluorescence Non-dispersive AF has been used for the determination of Bi (378) and for Ag, Cd, Zn, Bi and T1 (399). In the latter report selective volatilization from a carbon rod atomizer produced a tirne-resolved spectrum of fluorescence signals from mixtures of up to 3 of the above elements. 1 2 INTER-LABORATORY COMPARISONS AND REFElREE METHODS 1.2.1 Introduction This section reports methods of analysis that have been submitted for evaluation to various organisations in all parts of the world, e.g., the U.S. National Bureau of Standards, the British Standards Institution and the newly emerging European Standards Organisations. It is not intended to judge the value of the various organisations or inter-laboratory groups, but to report and comment on their findings.There are some generally accepted guidelines for workers participating in collaborative studies. Firstly, meticulous care should be taken over calibration of flasks, burettes and pipettes, and in the cleaning of all apparatus to be used for the analyses, as well as with the drying and weighing of the solid materials to be converted into standard solutions.Grinding, mixing and sampling of solid materials are potcntial sources of error and great care is necessary during these stages, Secondly, it is important that the operating parameters of the instrument should be accurately recorded, particularly if ‘slight modifications’ have been made to a standard item supplied by the manufacturer, since it will not be possible to assess the relative merits of different techniques unless full experimental details are known.Finally, all of the results or measurements obtained should be recorded. If it is not possible to say before or during the execution of one of the experiments that a given result will be wrong, then all the replicate deterniinations should be included in the calculation of the average result. 1.2.2 Inter-laboratory Comparisons The results of a 4Glaboratory study on the chemical analysis of airborne particulates were presented by workers at C.C.R. Euratom (1414). The homogeneity of the samples was checked by 3 independent techniques and found to be less than the analytical error for 40 of thc 56 elements determined. The techniques used by the participating laboratories included AAS, AES, neutron activation analysis, X-ray fluorescence spectroscopy and various wet-chemical methods.Initial evaluations of the results indicated good agreement for Al, Fe, Zn, Ca, Pb, C1, S, Si, Ti and Mn. While no definite conclusions could be drawn for the other elements, some systematic errors were apparent.The International Atomic Energy Agency (Vienna) distributed simulated air filters (filter papers ‘spiked’ with 13 elements), and aqueous solutions ‘spiked’ with 15 elements to a number of laboratories for analysis (1413). In the air filter study 22 laboratories returned 922 results obtained by 6 different methods. For 8 elements the nunibcr of results obtained by neutron activation and AAS was nearly the same, so permitting a comparison of these techniques, which were found to have a similar precision and accuracy.In the study on water analysis, the results from 40 laboratories, mostly using AAS without pre-concentration, were disappointing. A study carried out by the N.B.S. (1403) on the determination of low levels of Hg (1.2 pg 1-‘ and 1.5 mg 1-I) in water showed that it was necessary to add gold solution and HNOB to stabilise the solutions.A report from the American Watcr Works Association (184) showed that the cleanliness of glassware and the care of water samples containing44 Purl I / : Methodology suspended matter were critical in Hg analysis. In this study, carried out by 7 laboratories, AAS was found to be better than either the dithizone or crystal violet spectrophotometric methods.Other inter-laboratory studies on water analysis were carried out in Canada (858), Germany (682) and New Zealand (504). The first 2 studies compared results obtained by direct and by pre-concentration methods. The British Steel Corporation has set up a working party to study AA analysis of steels (637). It was concluded that the wide variations in the magnitude of the matrix effect precluded the use of standard correction procedures for the determination of Cr and Al.A B.S.T. method for the AAS determination of Zn in aluminium and aluininium alloys (579) was shown to agree well with 1 3 . 0 . method No. 2637. The American Society of Brewing Chemists carricd out inter-laboratory studies on the analysis of beer for Ca and Mg (194) and Fc and Cu (346).Although AAS determination of Ca agreed with the EDTA method, it was concluded that AAS determinations of Ca and Mg could not be included in the A.S.B.C. Methods of Analysis. This organisation also concluded that further work was needed to improve precision in the determinations of Fe and Cu in beer. The first ‘European Intercomparison Programme’ on the analysis of Pb in bZood in 1972 produced poor results.A second study has just been completed in which 80 laboratories participated and the analyses wcrc extended to include Cd and Hg, and both blood and urine samples were analysed (691). Tntra- and inter-laboratory variabilities wcre rcported. A similar study on Pb in blood and Pb and Cd in urine was undertaken by 33 laboratories in Germany (506). Most of the laboratories had an error (RSD) of between 0.07 and 0.15, but 30% of the laboratories had errors exceeding 0.20.The National Environment Research Centre has conductcd an extensive inter-laboratory study (119) on the comparison of ES, AAS, neutron activation analysis, spark source mass spcctrometry, X-ray fluorescence spectroscopy and anodic stripping voltammetry for the determination of 28 elements in matcrials such as coal ash, fly-ash, fuel oil and gasoline.The most reproducible results were obtained for Si, Ca, K, Sr, Fe, Cr and Ni, but poor results were obtained for Sn, Li, Ba, Ag and Na. The authors expressed the need for standard reference materials with certified trace-element contents. 1.2.3 Referee Methods and Reference Materials A referee method for serum Ca determinations (Clin. Chem., 1973, 19, 1208) has been evaluated over a 1-year period (465, 1245) and the accuracy of the method confirmed. However, the method was not considered suitable €or routine use because of thc largc sample volumes and long analysis time required. The method was considered suitable as a reference procedure against which the accuracy of routine methods may be judged. Standards for the analysis o€ oil samples have been prepared by mixing neutral Zn, Sn, Cu and Ni sulphonates and slightly alkaline B sulphonates (816). An N.B.S. publication (527) rcported on the stability of organo-metallic solutions for oil analysis. An N.B.S. standard reference paint sample with a Pb content of 11-87 t 044% Pb has been prepared (531). This sample has been analysed by AAS and differential cathode-ray polarography. K O . Standards have been prepared for the determination of major and minor elements in rocks (1 520), A1 in cryolites (1 5 5 9 , and Na in KOH (1 554).

ISSN:0306-1353    

DOI:10.1039/AA9740400042

出版商:RSC    

年代:1974

数据来源: RSC