ISSN:0306-1353    

DOI:10.1039/AA97202FX001

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

PART I FUNDAMENTALS AND INSTRUMENTATION 1 Light Sources Most of the papers t o be reported relate t o discharge tubes and lasers, though most analytical atomic absorption (AA) spectrometers continue to use hollow-cathode discharge lamps (HCL’s). The most practical light source for atomic fluorescence spectroscopy (AFS) remains the microwave discharge lamp. The greater range of tunable lasers now available may change this situation. However, in view of the elaborate systems needed to drive both microwave lamps and lasers, any means whereby the light output of HCL’s can be increased towards that of microwave lamps would contribute to the wider use of AFS as an analytical tool. 1 . 1 HOLLOW CATHODE LAMPS The effective temperature of the hollow cathode discharge has been investigated by Hannaford (51) and Smith (295).In a study of the Au discharge lamp, Hannaford demonstrated that failure t o consider the hyperfine structure of the spectrum had led to errors in the estimated temperature. Based on the Doppler width of the line, the tempzratures of the 242.8 nm and 267.6 nm lines were calculated to be 720°K and 690 K, respectively, which compare with a temperature of approximately 5000’K estimated by Kreye and Roesler * when no allowance was made for hyperfine s tr u c t ure. A method for constructing a T c HCL by electroplating on to Cu foil and introducing this into the cathode has been reported (40). The operating parameters of such a lamp were investigated and a detection limit for T c of 0.9 mgfI was obtained.IIuman (960) has described an HCL in which a high frequency discharge was used to excite the emission of the atomic vapour produced by cathodic sputtering. Using this type of lamp, AF detection limits of 0.0005 mg/l for Cu and 0.003 mg/l for Au have been achieved. These lamps, however, offer no advantages over conventional HCL’s when uscd in AAS (See ARAAS, 1971, I , p. 2 , refs. 64 and 477). A solid state, modulatable current HCL power supply which can provide up t o 400 V and 40 mA at frequencies up to 1 kHz has been constructed (476). As the spectrum of flame noise has been shown to be inversely proportional to frequency (See ARAAS 197 1, I , p. 15, ref. 638)’ if it were possible to operate HCL’s at frequencies >1 kHz, discrimination against flame noise should be improved.Weide and Parsons? have described the use of current pulsed HCL’s in AFS. They confirmed that the intensity of light emitted during the pulse was much greater than that of the conventional system with no significant increase in the width of self-reversal of the spectral line. In a study of Zn, they obtained a detection limit of 0.003 mg/l; this was almost three orders of magnitude better than the limit they obtained using a conventional HCL system. They attributed this improvement to increased light intensity and the use of a box-car integrator. An alternative mode of use of current pulsed HCL’s has been employed by Lloyd and Lowe (943) to modulate the intensity of the emitted resonance line.In their system the lamp was operated at its normal d.c. level and a high current pulse superimposed. The resonance line was momentarily absorbed by the additional sputtered atomic vapour. By appropriate electronic gating, the output signal of the *Kreye, W. C., and Roesler, F. L., J. Opt. SOC. Am., 1970, 60, 1100. tWcide, J . O., and Parsons, M. L.,Analyt. Lett, 1972,5, 363. 3 Part I: Fundamentals and Instrumentation 4 instrument was the resonance line intensity. in AAS there are certain analyses, e.g., Pb in blood, for which correction for non-specific background absorption is necessary. Procedures for background correction have been reviewed by Kahn and Manning (901). Pairs of lines suitable for correction purposes have been given (1 120) and the possibility of using a multi- element (Cu, Fe and Co) two-discharge lamp to give both increased sensitivity and background correction in AAS has been investigated (788). The developments in HCL's which we have reported are unlikely to have an immediate influence on the development of analytical atomic spectroscopy, their contribution is t o the general fund of knowledge.in most AAS instruments the contributions to instrument noise from lamp instability and from light intensity limitations are comparable (-0.1% s-') and probably close to the best obtainable. In addition to all the major instrument companies, suppliers of HCL's include:- ( 1) Cathodeon Ltd., Nuffield Road, Cambridge, England. (2) Activion Glass Ltd., Halstead, Essex, England.( 3 ) EM1 Electronics Ltd., 243 Blyth Road, Hayes, Middlesex, England. (4) ChemLab Instruments Ltd., 16 Seven Kings Road, Ilford, Essex, England; ( 5 ) Westinghouse Electric Corp., Electronic Tube Division, Elmira, New York, U.S.A. ( 6 ) Micro-Tek Instruments Inc., Baton Rouge, La., U.S.A. (7) Quartzlampen GmbH, Hanau, West Germany. (8) Fivre, Milan, Italy. 1.2 DISCHARGE TUBES The construction (89) and operation of electrodeless discharge lamps (EDL's) continue to be investigated, but with little progress beyond that previously reported (See ARAAS, 1971, I , p. 3). Details have been given (676) for the preparation of electrodeless Hg isotopic lamps having an intensity stability of better than one part in 1000, and in some cases, one part in 10 000.In the preparative stage, particular care is taken to de-gas the tube and associated vacuum system before filling with Ar at 5 torr. A dual element (Cr and Mn) EDL has been used in steel analysis by AFS (192). The effect of lamp temperature was investigated by Browner and co-workers (737) by passing a stream of hot air over the tube. The microwave power was fed to the lamp via an A-type antenna and the temperature of the air-stream varied from 20" t o 500°C. There was a significant improvement in the intensity of the light output and in its stability when compared with that from the conventional resonant cavity. This improvement may be attributed t o separating the function of vapour generation from its excitation, The former is effected by the hot air and the latter by the microwave energy.The effects of antennae and resonant cavities on the stability of the discharge and the cooling of the discharge tube were investigated (141) with a view to determining the isotopic ratio of N-14 to N-15 in agricultural samples by emission spectroscopy . By use of a tuning stub assembly (66) consisting of a length of coaxial line with two branching lines in which the inert conductor is shorted to the outer, the position of the short in each branch being adjustable, the efficiency of coupling the r.f. energy to the EDL was improved. It has been reported that the inclusion of an attenuator between the magnetron and the microwave cavity leads to a decrease in the noise Part I : Fundamentals and Instrumentation 5 associated with the operation of an EDL (30).While this should be particularly true with low boiling point elements, results are disappointing and, in general, amount to a one-third reduction in the noise level and drift rate, but with a doubling of the warm-up time. A circuit for modulating the output of EDL's, by using a transistor to switch a resistor across the magnetron power control, is claimed t o give lower noise levels and high a.c. signals (15-100 W, up t o 20 kHz) (748). Theoretical studies have been made of the mode of operation of EDL's, and of the been frequency elements By implications the studied application (679). of in line The terms of broadening nature a of strong the of ratio the magnetic in electromagnetic AAS of collision field and AFS (-30 frequency determinations field kG t o in to the a ' a '' cylindrical plied Hg of EDL.the more discharge Hadeishi volatile (1 14). and has McLaughlin" produced Zeeman splitting of the 253.7 nm line. This split line was not absorbed by Hg vapour in an absorption cell and could be used t o correct for background scatter and absorption by the sample. A nondispersive Hg analyser based on this system has a detection limit of 4 ng m-' or 5 x lo-' ' g and is manufactured by Scintrex Ltd., Concord, Canada. An alternatjve system has been described (154) in which the 253.7 nm line has been broadened in an Hg lamp and the unabsorbable 'wings' of the line used to provide a means of background correction.An attempt at direct isotope determination by AFS (442) was successful only in the case of Hg. Lamps were made containing pure isotopes and calibration was effected using solutions of known isotopic content. It has been recognised for some time that the Hg resonance line at 184.9 nm would give more than thirty times greater sensitivity in AAS than the forbidden transition at 253.7 nm. Dagnall et a1.(229) used a special long EDLt and a nitrogen-purged system t o investigate the relative sensitivity. They found the line at 184.9 nm was 31.8 times more sensitive than that at 253.7 nm and the noise levels were 1% and 2% of the lamp intensity, respectively. An improvement in the detection limit of up to 64-fold was possible. The behaviour of metal vapour discharge lamps has been studied (145) using a box-car detector to follow the discharge characteristics of the lamps driven on the 60 Hz a.c.supply, The initial afterglow of the self-absorbed Hg 253.7 nm radiation from Hg/Ar discharges is reduced (155) by the addition of 400 torr of Ar. This improvement is due to a reduction in self-absorp tion and increased collisional de-activa tion. 1.3 LASERS Fraser and Winefordner (437) have continued their investigation of the use of lasers in AFS. A pulsed tunable (360-650 nm) dye laser, pumped by a nitrogen laser, was used to examine the relative advantages of various types of fluorescence. They found that resonance fluorescence needed wavelength scan t o measure the scattered light, whereas direct line fluorescence avoided this problem.The signal-to-noise ratio was controlled by scattered light, pulse-to-pulse variation of intensity, and shot noise. Using air/liydrogen, air/acetylene or nitrous oxide/acetylene flames, detection limits ranged from 0.005 mg/l for A1 at 396.1 nm to 0.3 mg/l for Fe at 372.0 nm. Atomic absorption measurements with a tunable organic dye laser are reported to "Hadeishi, T. and McLaughlin, R. D., Science, 1971, 174, 404. TAldous, K. M., Ph.D. thesis, University of London, 1970 6 Part I: Fundamentals and Instrumentation give a sensitivity of 0.002 mg/l for Na vapour (36). Background correction was made by detuning the laser. Several review articles, principally concerned with the mechanism and properties of tunable lasers, have appeared (144,4 16, 538 and 698).Only dye lasers appear to be of interest as light sources in analytical atomic spectroscopy. Such devices can cover the range 116 to 1060 nm, though most are in the range 295 to 650 nm (144). A continuously tunable (304.4 - 327.2 nm) coherent source has been obtained ( 118) by the mixing of ruby and dye laser beams in an ADP crystal. The output power was under 100 W and the spectral band width 0.1 to 0.3 nm. In view of the cost and complexity of tunable lasers it is unlikely that these devices will be used for other than research and special investigations in analytical atomic spectroscopy in the foreseeable future. 1.4 MISCELLANEOUS SOURCES Although little practical use has been made of intense continuum sources in AAS and AFS, it is well established that, when coupled with a high resolution system, they are capable of producing detection limits comparable with those obtained using line sources. The Xe arc has been examined from the point of view of stability (122). It was found that there were three distinct features in the behaviour of these lamps, viz., a slow time-dependent drift, rapid fluctuation, and transition between two or more relatively stable levels. To minimise the effects of these instabilities, the power supply should be well regulated and incorporate a feed-back system. A magnified image reduces the effect of arc wander and new lamps should be ‘burnt-in’ before use. By pulsing a Xe arc (38 1, 383) at currents between 10 and 700 A for up to 1 ms, the light output could be increased by up to 700 times at 200 nm. The lamps withstood several thousand cycles but became unstable after that time. The reproducibility of intensity of lamps fired at 5 s intervals was excellent.

ISSN:0306-1353    

DOI:10.1039/AA9720200001

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

7 Part I: Fundamentals and Instrumentation 2 Excitation Sources and Atomizing Systems 2.1 ARCS AND SPARKS It is an interesting observation that, in relation t o tlie volume of work undertaken daily by direct-reading emission spectrometry, the documentation of this field in the current opcn litcrature is sparse. Whatevcr the reason for this, it is likely that a large quantity of potentially useful information is never made available t o other workers in the area. Fundamental aspects of spark analysis have been treated in two noteworthy papers by Walters (60, 105). In the first (60), experiments were reported for key lines of A1 (358.7, 360.16 and 361.23 nm) sparked in nitrogen against graphite for a variety of discharge current waveforms. The resulting comprehensive data indicated that p mime ter-dependen t excitation indices may require re-ex amination.W alters suggests that the excitation in tile discharge is much more straightforward than previously asstimed arid that tlie manner in whicii the discharge is viewed by tlie spectrometer is critical. He has also discussed (105) at length the formation and growth of a spatially stabilized spark discharge. Fully developed neutral atom and ionic spectra, free from unstable band-rich background radiation, are claimed to be produc’ed. Since the source is free from r.f. interference, it is particularly suitablc for on-line computer ;I p p 1 i c at i on 9. Ultra high speed photographs taken with a Nikon UHF-50OC camera have been used (556) to examine the mechanism of electrode erosion with ii spark discharge in air and argon atmospheres. Use of a low voltage gliding spark has bccn reported (85), its particular advantage being that it permitted analysis of the coati~igon ceramic materials without exciting, the base material.in the conditions of local thermodynamic equilibrium, realizcd iri the plasma of a wall stabilised arc, nieasurcments of C atom concentration were performed by Muller ct id. (1077). By measuring absolute intensities of ninc optically thin lines (136.4, 145.9, 146.3, 147.2, 148.1, 175.1, 176.4, 176.5, 247.8 nm), C atom concentrations frcirn 1.2 x lo1 t o 4.2 x 10’ cm- were obtained. A d.c. UYC plasma burning in opcn-ended water-cooled metal tubes o f different diameters was investigated ( 1079) and the radial temperature distribution determiricd from the ratio of the line intensities of Zn(1) 307.20 nm and Zn(1) 307.59 nm.The concentration of electrons was determined from the line intensities ratio of Mg(1) 285.21 nIn and Mg (11) 279.55 nm. It was shown that the. radial distribution of temperature and electron pressure, as well as the line iritcrisitics of different elements, depended on the tube diameter in which the arc burned. The a.c. intermittent arc was studied by Florian et al. (804, 805, 806, 807) t o determine the factors influencing the excitation processes. The effects of alternating tile polarity of tlie electrodes and the changes of arc temperature with changes in frequency led t o selection of optimum conditions and analytical lines.I n an U.C. UYC plasma, the production of a ‘superintei~sive’ radiation state by depositing the analyte thickly on both electrodes has been described by Paksy ( 310 883). This technique was used to increase the line intensity relative t o background for Fe, Si, Mg, Cu, Zn, Ag and Cd in pure Al. A method has been dcscribed (888,950) for separating the reactions occurring in tlie anode and cathode spaces of a polarized a.c. arc. The products formed in each reaction were measured and conclusions given regarding the axial distribution and intensities of A1 spcctrum h c s . 8 Part I : Fictidurn c7ti tals and Irzstru m cri tatiori The possible use of an a.c. arc in AA analysis was iiivestigated (777) with a stroboscope incorporated into a spectrograph, so tliat tile AA signa! could be recorded during the pause in the current.A study (65) of arc spectral lines for Se and l ' e ill tlie 220-440 iini region found many discrepancies with published data and suggested lines more useful in con- ventional arc analysis than the persistent lines given in most tables. The physical and chemical effects of various two-component gas gtniosplieres on the intensities of A1 spectral lines have been investigated (948 949). A rotating A1 disc was subjected to an arc, using Cu and C counter electrodes in a closed silica cell previously filled with mixtures o f oxygen, argon, nitrogen aiid carbon dioxide i n various combinations and ratios. Harris has investigated ( 106) the effect of extraneous materials in tile point-to-plane mi-arc excitation of Cu-base alloys.He denionstrateti tliat ;i 1 : 1 CiiF/NH4F mixture added to the counter electrode eliminated tlie need to correct for viiriation in the C'II content of t h e alloy and thc same operating method becaine ;ipplic:ible to a wide range of Cu-base alloys. Some multipliascd alloys are particularly difficult simples for a:ialysis by point-to-plane direct-readcr tccliniques. One suci: system, leaded brave,, was invcstigated (741) .using ;I Jarrell-Ash controlled waveform source t o improve tiie precision of the Pb/Cu ratio. Oku and Hirokawa (5.57) used ;I Sliimadzu high precision exitation source to study the relittion between emission iiitcnsities arid alloy phi1 high voltage spark under both .argon and oxygen atmosplicres.First ;I simplc ( system was examined, then an AljSi systcm with a eutectic point, and third 3 complex Ni/AI system liaving intermetallic compounds and ;I eutcctic point. Tlie influence of different phases in iron and steel samples upon the emission results was investigatcd using a 3 mm diameter boron nitride diaphragm to restric,t [lie a ~ a of sample cxamined ( 168). Thc application of an external inhomogeneous magnetic field to :I d.c. xrr h i i s been s:titl (164) to give enlianced line intensity and t o eliminate tlie need f o r t h e use of ;I buffer in determining Bi, Tc, Pb, Sn, Sb, Ag and As in iroii :ind stccl. Furtiier studies ( 113, 558, 945) have dealt witli the effects o f hornogeiienus ni;igrieIic fields on tlie (1.c.are. Various aspects of the well known currier distiilaiior7 technique of d.c. arc analysis have becii iiivestigated. The use of PTFE powder witil tll- formation ot' volatile fluorides (101, 328) promoted tlic vaporization of the !ess volutile elements MII and Fe i l l a GaAs matrix (348) and increased the vaporization rate o E Bc, B, Al, Si, Z r arid M U when these elements were present in oxide form ( 1072). The use of a CS(:: grap!iite mixture by Osumi and Miyake (559, 1080) gave an increased m d constant evaporation rate of the sample. The presence of Cs in tlie arc plasma doubled tile electron density and increased the residence time of the particles in the arc plasma. The line intensity, arc tcmperature, electron density and atomic concentration measurements are given and discussed.Experiments showing the dependence of carrier distillation effects 011 the concentration of the carrier have also been discussed (327, 874). The spectrograpliic charaLteristics of a device for continuously administering powder samples into a d.c. arc have been reported (889). Tlie advtlntages of adding a buffer t o powdered sample blown into an a.c. arc (86') o r a tIiree-pli;ise arc (91) have been discussed. Yurt I . Fundamentals and Instrumentation 9 2.2 P L A S M ~ S Interest has increased in the use of low-power microwave-induced plasmas (2450 MHz, 50-200 W), coupled with emission spectrometers, as detector systems in gas chromatography.In the discharge tube, connected directly t o the gas chromatographic column outlet, organic molecules are fragmented into radicals and atoms which are then excited to emit a spectrum of characteristic bands or lines. This technique* has been used t o determine C, H, N, 0, P, S, CI, Br, I, Hg(399, 578, 642, 663) and metal acetylacetonates ( 5 5 3 ) . Kleinmann (947) and Kawaguchi et al. ( 5 8 3 , 954) have demonstrated the usefulness of a low-power microwave discharge as an atom source for the determination of trace elements in solution, though pronounced inter-element effects were found. Kitagawa and Takeuchi (422) estimated the electron temperature, the degree of ionisation and thc total number of atoms and ions produced when a Mn solution was sprayed into a microwave-excited plasma (2450 MHz, 200 W), and showed that the effects of added alkali and alkaline earth metals were dependent on the ionisation potentials of Mn and the added metals.An emission spectrometer for high-speed qualitative analysis of trace elements in solution has been developed (554) using a 2450 MHz uiiipolar/discharge plasma generated at 400 W power output. Using a high speed galvanometric recorder the spectrum in the range 200-500 nm can be scanned in 10 min. A small W loop filament (1.8 mrn diameter) heated to between 800” and 1200°C in a stream of Ar (300 ml/min) has been used (206) to introduce 2 pl samples, pre-dried un the loop, into an electrodeless discharge (2450 MHz SO W) in a quartz capillary (1.6 mm diameter).Detection limits range from 4 ng Ba t o 0.004 ng Cd. A variety of expressions appears in the literature t o describe the high-power plasma atom sources used in atomic spectroscopy. An IUPAC tentative proposal ( 1098) suggests the use of the term ‘plasma jet’ and discourages the use of the terms ‘plasma flame’ ‘plasma torch’ arid ‘plasma burner’, except where they refer to hot gases produced by combustion. The advantages and disadvantages of the high-power plasma jet (1-40 MHz, >1 kW) as an atom source continue to be discussed (81, 151, 189, 898, 908, 962, 966) but little improvement is apparent in the problems associated with the technique. Greenfield and Smith (5) re-investigated the problems of background continuum and band structure and found that with their equipment (Radyne RD 150 H generator and IJnicam SP900 nebuliser) the background increased with increasing length of the plasma jet?-.Ishida and Kubota ( 5 5 2 ) described the use of an image memory tube to record thc spectrum produced in a plasma jet. Other developments reported include a disc-stabilised plasma jet for solution nebulization (7741, a plasma jet for solid metal samples (673) and a small heated nebulizer for the analysis of inicrolitre quantities of samples (189). Elliott (158) has described a d.c. plasma generator, for spectroscopic analysis, with the anode within a swirl chamber and the cathode, outside the chamber, offset from the axis of the emergent plasma, thereby deflecting ionised gas from the jet ar?d producing a simpler spectrum.Finally, an apparatus and a rrietliod for temperature measurement of heavy particles in an Ar plasma at atmospheric or low * McLean, W.R., Stanton, D.L., and Penketh, GE., paper given at meeting ‘Selectivity in Trace Analysis’, Stirling, June 1972. See.also: Pmc. SOC. analyt. Clietn., 1972,9 296. d-l’ruitt, D., and Kobincon, J.W.. .4rialytic‘a Chim Acta 1970, 49 401. Part I: Fundamentals and Instrumentation 10 pressure, in thermal equilibrium, using attenuation of a monochromatic X-ray beam has been presented (609). On the theoretical side, measurements on the distribution of emission from different parts of a high frequency Ar plasma jet have shown that thc emission cannot be classified solely on the element’s ionisation potential but that the excitation potential for every line must be considered (408).A one-dimensional energy balance equation for steady thermal plasmas has been applied t o the case of an annular plasma contained between coaxial cylinders ( 1 16). The effects of arc configurations, pressure vessels, gas, water pumps and heat exchangers on a steady-state electrical arc discharge generated in Ar at pressures up to 950 atmospheres and currents up t o 200 A have also been studied (470). 2.3 GLOW DISCHARGE LAMPS The glow discharge lamp (GDL) first described by Grimm* has now been developed to a stage at which it must be considered as a competitor to arc/spark excitation for routine spectrographic analysis, Glow discharge source units are commercially available from RSV (8031 Hechendorf, Pilsensee, West Germany) and are reported to be in routine use for quality control in European steel works. The advantages claimed ( 6 3 l ) , viz., linearity of calibration graphs from pg/g to 1 OO%, low background emission, freedom from inter-element interferences and high precision arid accuracy, are most attractive and, if these claims are substantiated by impartial users, one can foresee the GDL becoming the standard source for much emission spectroscopy.Sample size and preparation requirements are basically similar to those for a conventional arc/spark exitation direct-reader system. However, the spectrum produced is somewhat different, with the result that the two forms of source cannot conveniently be used interchangeably in a pre-set direct-reading spectrometer.With a GDL more use is made of the vacuum ultraviolet region; for example 156.1 nm is preferred to 193.1 iim for C because the latter line suffers interference from Ar. Several practical applications of GDL’s have appeared. Belle (742) described his experiences with the use of a GDL, particularly in comparison with a scanning electron microscope, for examination of stainless steel samples which had been exposed t o a Na atmosphere. lager (18) used a GDL with satisfaction for the analysis of 70 quantities of Ag and Cu in Au. He found that calibration curves could be programmed despite a ‘third partner effect’ and that direct determination of all sample components including the matrix element was possible.It has been shown (946) that a GDL is applicable to the analysis of electrically non-conducting material after mixing with a suitable conductor, e.g., graphite or copper powder. RSV’s documentation of their work in this area is comprehensive and is obviously essential reading for anyone contemplating tlie use of this technique.. The capability for the analysis of non-conducting materials has been neatly exploited by Russian workers (102, 103) who deposited electrolytically, on to a thin carbon disc, the metallic impurities in aqueous solutions of organic acids. Bi, Cd, Co, Cu, In, Ni, Pb and Zn were then determined using a GDL set-up. A rather similar approach has been used (184) to determine elements such as Al, Fe, Cu, etc., present in the surface layer of silica samples.In this instance, about 10 mg of substance was removed by leaching with an HF/HNO, mixture and the resulting solution transferred to a graphite hollow cathode for excitation under a He atmosphere. *Grimm, W., Nature, 1967,22,586. 11 Part I: Fundamentals and Instrumentation Further fundamental studies of GDL's have been made. Baumans (210), for example, has examined the sputtering rates of some metals (Al, Cu, Mo, Ni, Ta, W, Zn) and alloys (Cu/Zn, Cu/Ni, Cr/Ni/Fe). He established that there was a linear relationship between the mass sputtered in unit time per unit current and the operating voltage. A tentative evaluation was made of the use of a GDL for thin layer (-1,um) analysis.The possiblity of laser action in the Crookes dark space of a cold cathode discharge was examined (622) using the 441.6 and 537.8 nm Cd lines and 747.8 nni Zn line. The source was operated in a 1-10 torr He atmosphere and :lie metal heated t o obtain vapour. Cd intensity increased with temperature up to 250 C, Zn up t o 36OoC, the decreases at higher temperatures being attributed to quenching. Szilvassy (88 1) reported that the perpendicular application of a magnetic field increased the inhomogeneity of the emission from a hollow cathode operating under both Ar and a He/Ne mixture. The increased intensity at the edge of the cavity also sign if ican t 1 y in creased the sensitivity of analytical detection. Thornton (799) has developed an automatic system with direct concentration read-out for analysis of ten elements in high-temperature alloys using a demountable hollow cathode source.He found that inter-element effects were absent in Si and he therefore mixed the sample with 30% w/w Si as a buffer. A current of 1 A produced a cathode temperature of 2300 C and the emission intensity was then proportional to the boiling point of an element. A hollow cathode source has also been used ( 3 17) for determination of halogens, a He/Ne mixture being used as carrier gas because the excitation energies of the analytical lines used (all in the visible region) were above 12 eV. In this application the stability of emission intensity was such that it was unnecessary to use an internal reference standard.Milazzo and Caroli (696) compared a hollow cathode with spark excitation and found the reproducibility of the hollow cathode t o be much better for the analysis of Cu and brass and slightly better for steel. The AA of Ar and Ne with a demountable HCL as atom reservoir has been described (375), the ionic lines Ar 488.0 nm and Ne 296.7 nm being used. Calibration curves were constructed from 20-loo%, with He as diluent, arid conventional HCL's used as light sources (Ba and Cu lamps for Ar and Ne, respectively). Absorption sensitivities were extremely low, e.g., approximately 1% for 100% Ne, and this work appears to be of academic interest only, Amos and co-workers (565, 705) have re-examined cathodic sputtering as a means of atomisation for AAS* and Showed that it may have applications to the direct analysis of solid samples.Considerably more work will be needed, however, before this could replace conventional flame techniques for routine applications despite the inherent disadvantage of requiring prior dissolution of a sample for flame techniques. Finally, the award of a patent (157) for a modified design of GDL should be noted. 2.4 LASERS Two reviews have appeared, during 1972, covering the use of lasers in spectrochemical analysis (540, 823). Non-metallic samples have been classified into three categories for the purpose of laser microprobe analysis (959), viz., opaque samples with high boiling points (which behave similarly to metals and require no cross-excitation); transparent samples with high boiling points, which behave as those in category one if a high power density is used; and transparent and opaque sarnples.with low boiling points, which require cross-excitation to produce sufficient free atoms for analysis.Some authors have merely described the effect of sampIe matrix on the results obtained by "Gatehouse, B.M., and Walsh, A., Spectrachim. Acta, 1960, 16, 602. 12 Part I : Fun da m e 11 ta Is u n d I n s t r in e I i ti[ t io i i laser microprobe analysis (161, 562, 619, 884) while others have tried to quantity these effects. Ishida and Kubota (204, 207, 561) showed that it was impossible to eliminate matrix effects because the sample erosion was proportional to the heat required to raise the temperature of the sample to the liquidus plus the latent heat of fusion. Felske e t al.(942) critically examined the application of a low power solid-state pulsed laser for the evaporation and excitation of samples and showed the dependence of crater depth on the number of spikes in the laser shot and the melting point and thermal conductivity of the material. Variations in the spot sizes evaporated were made possible by varying the focal length of the laser-focussing lens ( 15 mm focal length produced a 33 pm diameter crater, whilst 27 mm focal length produced a 60 p m diameter crater). A qualitative relationship between the physical properties of the material and the extent and shape of the regions sampled by a single-spike laser pulse was described by Allemand (953), while a freely running laser source was used to show that the inductance and capacitance of a cross-excited spark discharge affected the crater size (944).A stain produced on a sample of Fe around a crater formed by a laser (Carl Zeiss Jena LMA) using cross-excitation with an auxiliary spark discharge was assumed by Yamane (675) to be due to the deposit of C from the electrodes and oxidation of the sample surface heated by the spark plasma. The form and degree of the stain varied with the spark discharge condition. A statistical study (432) of data from the laser microprobe determinations of Ca, Fe and Zn in biological matrices showed that attempts to correct for spectral background signal or variations in laser energy output ( 1 2-17 mJ) produced no imp rovemen t in precision.Other factors concerning the interactions of a laser with a solid target and the resultant plasma produced have been studied. Engelhardt ( 1 15) measured the expansion velocities of the plasmas, produced by a Q-switched ruby laser, from Cu and A1 targets, by using a high-speed framing camera and related them to a theory based on the one-dimensional equations of motion and the conservation of energy. Naka~ima et al. (560) also used a high-speed framing camera, with a maximum photographing speed of 500 000 frames/sec to show that vaporisation of the sample and the laser beam spot were both found to appear in the same photograph. They also showed, using a laser pulse of short duration (20-30 ns), that light originating from the vaporised sample only appeared in the period 10-20 p s after the breakdown and that the longer lasting luminescence may originate from breakdown of atmospheric molecules.The line emission from Mg iom in a plasma, produced by a Q-switched ruby laser, has been studied by Stumpfel and Robitaille (435), who showed that the appearance time of each ion line decreased as the laser power increased. These appearance times rtllowed information to be obtained regarding the overall heating rate and the electron temperature in the heating zone. 2.5 FLAMES 2.5.1 Flame Types A noteworthy development in AA during the last two years has been the d e t e r m i n a t i o n o f n o n - me t a!s at vacuum ultraviolet wavelengths.The n i t r og e n-se p ara t e d n i t ro us o xi d el a ce t y 1 e n e flame is alni o s t c o m p let e 1 y trans p a re 11 t between 180-220 nm and was used for the determination of S ( 2 1 I), I (3h4), P ( 6 3 5 ) , Hg ( 6 3 5 ) , As a:id Se.* Atmospheric absorption was further avoided by passing the * Kirkbright, G.F., and Kanson, t,., Analyr. Chem., 1971, 43, 1238. 13 Reference Detection limi t/mg 1 p~ 1 .o text 635 0.02 25 364 635 21 30 Part I : Fundamentals and Instrumentation light from EDL’s through nitrogen-filled tubes to an evacuated or nitrogen-purged monochromator. Sensitivity (1% absorption)/ mgl Element Wavelength/ nm 2.0 0.05 As H d I ) P I 12 4.8 9 6.5 193.7 184.9 183.0 177.5 180.7 196.0 2.0 211 text S Se ‘Ihe non-availability of HCL’s for I and S may restrict the acceptability of the technique for practical use.Other separated flames have been investigated for AFS. Nitrogen sheathing of an air/acetylene flame for 22 elements resulted (44) in improvements in detection limit by factors of up t o 4 times. Burner configuration, inert-gas and flame sheathing were studied (7 1) for the AF of Sn. Luecke (5 75) has examined nitric oxide/acetylene, nitrious oxide/hydrogen and nitrous oxide/acetylene flames for AA. Although little could be claimed for nitrous oxide/hydrogen, the nitric oxide/acetylene flame* gave the best limits of detection for B , Ba, Sr and Zr. A further study on chemiluminescence and thermal emission spectroscopy in Ar-diluted oxygen/acetylene and nitrous oxide/acetylene flames (697) has been reported.Emission from the secondary inter-conal region of the flame was vicwed, the approach being particularly useful for elements having high excitation energies (e.g., As, Bi, Se) or high monoxide dissociation energies (e.g., Be, Ge, Zr). 2.5.2 Burners and Nebulizers Many papers deal with novel burner and nebulizer designs. U nfortunatcly, genuine innovations are few and it seems that several investigations have tended to go over old ground. An exception appears t o be in the development of ultrasonic nebulizers which have been meticulously evaluated by a few workers (710) over several years. One gains the impression that many of the earlier problems associated with ultrasonic devices have been overcome.AF detection limits (479) for Zn (2.5 x lo-’ mg/l and Cd (1.3 x 1 Op- ’ mg/l) have been obtained using premixed argon/hydrogen/entrained-air flames and Osram lamp excitation. Improvements in AA sensitivity ovcr pneumatic nebulizers are quoted (32) as factors of 13 for Cu, 30 for Fe, 19 for %n arid 1 1 for Pb, while other investigators (441) give improvements of 2 t o 4 times depending on the element. The fact that the enhancement varies with element suggests sorrie improvement i n volatilisation efficiency as well as nebulizer yield. * Slavin, W., Venghiattis, A., and Manning, D.C., Afom. Ahsorp. News/.. 1966. -5, 84. Part I: Fundam en tals and Imtru rri en tat ion 14 Specially designed burners must be used to gain the full benefit from ultrasonic nebulization (32) and allow for air flows of 3 l/min or less.The flame is supported on a series of holes rather than a slot exit port. Stupar and Korosin (312) employed a 3 MHz ultrasonic nebulizer for refractory element determinations using AA with nitrous oxide/acetylene flames and an Ar carrier (2-6 l/min). Sample uptake was 0.37-0.44 ml/min and AA sensitivities (mg/l/%) were: Mg 0.005, Ca 0.008, A1 0.3, Ba 0.15, Ti 1.3 and V 0.5. These sensitivites are about equal to the best obtainable on commercial instruments with pneumatic nebulizers. A pneumatic nebulizer with primary air supply of 7-8 atmospheres has been used (1 37) in conjunction with a narrow capillary through which solution is forced by a pressure of 1-6 atmospheres.Solution uptake rates of 0.6 ml/min gave equal sensitivities to uptake rates of 3-6 ml/min under normal operation. Nebulization efficiencies of 90-93% are claimed and result from the production of smaller droplets having a narrow size distribution. A droplet generator which produces single particles 4-130 prn in diameter, of radius uniform to 1.3% has been made (469) for cloud physics studies. Up to 100 droplets per second can be produced. A Lovelace aerosol particle separator which gives a continuous separation of droplets 0.2 - 5 pm in diameter has been described (382). Comparisons of several flames, burner designs, premix and direct injection nebulizers, heated spray chambers and condenser solvent removing systems have been carried out by Veillon and Murphy (723).Detection limits ranging over 1-2 orders of magnitude were obtained, Cu at 324.7 nm being used as the reference element. Heated spray chambers with solvent removal by condensation have been examined by Soviet workers (784) for absorption, emission and fluorescence and improvements in detection limits over conventional systems observed. Unfortunately no further details are available. Another dcvice for increasing nebulization efficiencies is a heated tube spray chamber used in conjunction with volatile chelates such as acetylacetonates and their fluorinated analogues (2 12). Sensitivities of Cu 0.03, Fe 0.05 and Cr.O.17 mg/l/% were quoted for a Perkin Elmer 303 with Boling burner.Bryand and Hodges (37) haw discussed memory effects in burners when volatile organometallics were nebulued. A sophisticated nebulizer and sample introduction system (72 1) has been constructed which introduces, in turn, sample, standard and blank into the flame and processes the results using a dedicated analogue computer. The sample introduction system is similar t o a multiport sample-loop injector used in gas chromatography and is connected directly t o the capillary of a conventional nebulizer. This approach greatly redfices the effect of experimental parameters such as gas and solution flow rate variations. A chopper system has been designed (965) for modulation of sample aerosol into a flame from two separate nebulizers, one delivering the sample, the other solvent.Flame emission detection limits are 11-25 times better than with continuous nebulization and are accounted for by almost total zlimination of background. Special burner designs include those already described for use with ultrasonic nebulizers (32, 312, 441, 479, '710, 723) as well as a burner for operation at reduced pressure (149). A turbulent flame burner (91 1) has been described with an outer sheath of oxygen to prevent ingress of contaminated ambient air, and a burner for the determination of chloride based on the Beilstein test has been reported by Tomkins and Frank (438). In the latter, chloride was determined indirectly by measuring the Cu present by AA and not by C u r l band emission as in the classical test.The aqueous solution detection limit was 2 mg/l, interference occurring from Na, K, Ca, Cu, 15 Purt I: b’undamentals and Instrumentation phosphate, sulphate and other halides. The studies of burning velocity of premixed flames on commercially available burners reported last year (See ARAAS 197 I , I, p. 14, ref. 6 18) have now been published in journal form (29). An improved burner head for nitrous oxide/acetylene (1084) was reported to give improved sensitivity, freedom from certain interferences and reduced C deposition when compared with the Perkin Elmer burner. In the past, ‘fail-safe’ devices fitted t o gas supplies o n AA spectrophotometers have been found only on expensive, high-performance instruments and not on low-cost instruments, whch might be expected to be operated by less experienced personnel.In 1972 Varian-Techtron and Perkin Elmer added ‘fail-safe’ devices t o their accessory lists and one hopes that other manufacturers will follow suit. 2.5.3. Discrete Sampling Devices A number of sampling devices has appeared recently for use with flames enabling analyses to be carried out on small samples. The readout usually appears as a transient response of which the maximum or the integrated signal may be used for calculation. The well established Dehes sampling cup* has been the subject of several studies (392). Agahigian (380) has improved the stability of the commercially available accessory retailed by Perkin Elmer. A slight rocking motion of the mounting was eliminated by the use of a powerful but inexpensive spring, obtained, it was said, from ‘a nearby hardware store’.A cup holder of lnconel(997) has proved sturdier than the Pt-lr wire originally supplied by Perkin Elmer and is standard on all kits supplied after March 1972. Attempts t o volatilise elements (194) such as Al, Ba, Ca, Ge, Mg, V and W were unsuccessful even in nitrous oxide/acetylene flames; a fused alumina absorption tube and carbon boat were positioned in the red reaction zone. Easily volatilised elements Pb, Ag, Se, Bi and Ga (194) presented few problems, although Hg ( 136) and other elements examined ( 1 94) gave multiple peaks. Direct insertion of Ziquid samples ( 2 ~ 1 ) in the flame on a Pt probe (790) has been used to determine Cd at 228.8 nm in the range 0.01 - 1 mg/l Cd giving a limit of detection of 1 picogram.Injection of a microsample (c.g., 2 pl) into a stream of solvent flowing into a conventional burner/nebulizer assembly has been proposed (1 1) as a technique for the analysis of small samples. Unfortunately there is a penalty in concentration detection limits of 10-1 00 times compared with conventional con t I nu ous nebulization. The direct analysis of‘ solid samples by flame techniques has received very little research effort despite the rather obvious challenge. Govindara et al. (58 1) impregnated a mixture of NaCl and powdered rock samples on to an iron screw rod, which was inserted into an air/acetylene flame.The total recorded response was used to evaluate the Rb content of the samples; a relative standard deviation of 4-574 and complete absence from matrix effects was claimed. A very interesting development in solids analysis has been the preliminary investigation into candoluminescence comiucted at the University of Birmingham (446, 498). Bi, Mn and Sb were incorporated into a matrix of a 1 : l mixture of calcium sulphate and calcium hydroxide. The mixture was pressed into the hexagonal recess of an Allen screw and reproducibly positioned in an air/helium/hydrogen flame. Under these conditions the surface of the sample can be induced t o emit radiation characteristic of the added elements; 5-1000 p of Bi can be determined in this way. * Delves, ILT., Analyst, Lond., 1970, 95, 431.Pb in pencil paint has been monitored rapidly by inserting the edge of a pencil into the AA flame (6 15). Jarrell-Ash (709), Instrumentation Laboratory ( 5 69), Rank Precision Industries and Perkin Elmer (365) have produced inexpensive accessories for the determination of As and Se. The gaseous hydrides are produced by reaction of the sample with acidic stannous chloride and metallic Zn. The evolved hydrides and hydrogen are stored in a reservoir (a rubber balloon in the Perkin Elmer accessory) and at the end of the reaction the gases are swept into an argon/hydrogen diffusion flame. A transient signal is obtained which provides limits of detection lower than either the classical Gutzeit o r Ag diethyldithiocarbamate tests.The use of an electrically heated absorption tube is reported (443) t o give detection limits which are better by a factor of two than an argon/hydrogen diffusion flame. Lichte and Skogerboe (444) have injected a sample (0.1 - 2 ml) containing 1.4 M HC1 and 4% w/w stannous chloride into a tube packed with Zn granules. As was detected down to 3 ng (0.003 mg/l by emission using a microwave excited discharge. Bramen et al. (75 1) determined As (detection limit 1 ng) and Sb (detection limit 0.5 ng) using a d.c. discharge tube for detection by atomic emission. The reducing agent was 25 ml of 1% aqueous sodium borohydride - samples of up t o 10 ml could be accommodated. 2.5.4 Theoretical Studies Several papers have appeared describing J7arne terriperaturc measurements using a two-line AA (33, 84, 350) or AF (616) principle.JII the AA method, the uh absorption is measured at a resonance wavelength and at a wavelength corresponding t o a transition from a thermally populated lower level. The final energy level need not be the same for the two transitions. A continuum source must be used to eliminate variability in the source line width and to simplify equations. The flame temperature is given by: where: E is energy of lower electronic energy levels k is the Boltzmann constant g is the statistical weight of the lower energy level f is the absorption oscillator strength h is the wavelength a is the % absorption at low atomic concentrations. As with the well known two-line atomic emission method, oscillator strengths for the two transitions must be available.However, with the absorption method, calibration for the spectral response of the apparatus is unnecessary, chemical excitation effects are less likely and the light beam may be confined to a well defined geometry. This could be an advantage with inhomogeneous flames where Abel inversion techniques are to be used. In the AF method (616), transitions corresponding to direct-line and t herinally-assist cb direct-li i i v t'iuorexc.ent c wtv-c' c.iio\eri. The degree of thtv-rnczl ionizatiun of metal additives in flameq i? usually Eurt I: Fundamentals and Instrumentation M + X + M+ + e- + X h<+ H 2 0 =+ M O P + H hl + X** + M++ e- + X* 17 calculated assuming the Saha equilibrium." Deviations from the calculated values arise usually from slow recombination of ions or chemi-ionization.Hayhurst and Telford (78) have measured the rates of ionization of Li, Na, K and Cs in oxygen/nitrogen/hydrogen flames using a quadrupole mass spectrometer. As might be expected, the ionization process was shown t o involve flame gas species X : The ionization activation energy, ionization potential and collisional cross-sections were also measured. The same workers have also developed (228) a mass-spectrometric method for the determination of hydrogen atom concentrations in flames based on the equilibrium: (where M = Ca o r Sr). The M+, M O P ion currents are measured. Preist (80) proposed a mechanism of ionization involving electronically excited flame gas species X** or X*: Other kinetic measurements of flame gas reactions and cherni-excitation have been made by Dixon-Lewis and co-workers (664, 665), while AF quenching cross-sections for the Na D doublet (497) and excited Sr atoms (496) have been measured in hydrogen flames.Investigations into the chemistry of metals in flames include the studies of Bulewicz (625) into the fate of Al, Sn and Cr compounds. A1 was shown t o exist mainly as A1(OH)2 rather than AIO, Sn as SnO and Cr as HCr03. Possible interference effects arising from the presence of radical recombination catalysts, e.g., Ca: Sr, Ba, Mn, Mg, Sn and Cr. were proposed. Other workers (79) also support the observation enthalpies that .A1(OH)2 of formation is the main were product derived in (79) oxygen/nitrogen/hydrogen for A10 (-349 * 20 kJlmole flames.and A l Standard (OH), (~ 1005) 2 30 k.J/mole). both components in the vapour phase. Soviet workers ( 2 3 2 ) have studied the ionization and dissociation of Cu and Ca in air/acetylene flames and quote plots of temperature, partial oxygen radical pressure and degree of dissociation of Ca and its oxides with initial scetylene content of the flame. Chemi-excitation and cherni-atomization processes for Na. Ca, Fe and Cu in oxygen/methane have been investigated (337) in a series of low pressure flames. Further studies o n the mechanism of'de-solvation of sample droplets in flames (209) have been made using a discrete droplet generat0r.T Aqueous and organic solvents were investigated with air/acetylene flames; the rate controlling process appeared t o be conduction of heat through a vapour layer t o the droplet surface.Radiative transfer was not considered. A somewhat different apparatus for studying velocities and dimensions of particles in flames ( 8 3 5 ) used two spark units focused on t o a camera along the same optical path through the flame. Discharge of unit one was followed b y discharge of the second unit a set time later, the camepa shutter remaining open throughout. The time interval was monitored by a photocell, the particles in the "Si,mning, L). C., and Capacho-Delgado, L., Analytica Chim. Acta, 1966, 36, 312, iHieftje, G. M., and Malmstadt, H.V., Anal,vf. Chem., 1968, 40, 1860. Part I: Fundamentals and Instrumen tation 18 flame showing up as two sets of black spots. A moving photographic plate was used (840) to follow the atomization processes in the formation of Cu and Mn atoms by AA. A review of the significance of atomic spectral parameters such as line profile, oscillator strength, etc., has been written by L'vov (407). A new method of determining integrated absorption coefficients is included. An accurate approximation (2 17) has been derived for the Voigt line profile for cases where the ratio of Lorentz to Doppler half-widths is not small. The effect of a n electrical field (electrode potential 1500 V) on emission spectra of Ba, Sr and Ca in air/acetylene flames has been investigated (23 1).The field increased the intensity of ionic emission from Ba and Sr; molecular monohydroxide spectra were also studied. The possibility of using AF for the isotopic analysis of H g (442) has been explored using isotopically pure line sources and oxygen/argon/hydrogen flames. Attempts to extend the method t o determine In, Zn and Cu isotopes failed because of hyperfine overlap. The molecular emission spectra of alkaline earth fluorides have'been reported ( 172) in- a variety of flames. It was proposed that the observations offer a possible direct determination of F. Molecular absorption spectra of SO,, (566) and PO (566, 669) have been proposed as methods of analysis for S * and P. The lower limit of detection for S was given as 10 mg/l at 207 nm using a flame in a heated-silica tube, 273 cm in length.A limit of detection of 2 mg/l P at 246.0 rim was quoted, but here a conventional air/acetylene flame was used. Winefordner's theory of signal-to-noise ratios, -f has been used (433, 7 14) to compare experimental and calculated limits of detection in air/acetylene (38 elements) and nitrous oxide/acetylene ( 5 7 elements) flames. Agreement was considered fair if the experimental and calculated values were within a factor of 5 ; results quoted for air/ acetylene were usually much better than this. Several anomalies occurred in the data for the nitrous oxide/acetylene flame-these were ascribed to lack of thermodynamic equilibrium. Shadowgraph studies (467) of air/acetylene flames on two burner configurations have related the flame noise level to the OH radical emission intensity. Signal-to-noise ratios for Cu in absorption and emission have been correlated with flame stoicheiometry and the beneficial effect of nitrogen sheathing confirmed. Alkemade (952) has measured t h e , noise spectra of Na and K emission in oxygen/nitrogen/hydrogen flames over the range 15- 1 O5 Hz.A minimum noise level for both background and alkali metal emission was found at 2-5 kHz; this frequency is recommended for AF measurements. Light scattering by solid particles may cause interference in AA or AF. The graphs of percentage scatter against wavelength were found t o be hyperbolic (739), with the scattered light intensity lowest at long wavelengths.The scatter signal was directly proportional to concentration in air/acetylene, air/hydrogen and nitrogen/hydrogez flames for up to 5% Ca, and the scattered radiation had its minimum intensity at 90 to the incident light. Studies on chemical and physical interference effects continue to be empirical and in many instances the theory provided is unsubstantiated by observation. More rigorous application of theory, however, has enabled Soviet researchers to calculate the degree of interference of nitric (346) and hydrochloric acids (781) on the atomic "F'uwa, K., and Vallee, B. L.,Anelyt. Chem., 1969,41, 188. fWinefordner, J. D., and Vickers, T. J., Analyf. Chem., 1964,36,1947. 19 Part I: Fundamentals and Instrumentation emission of Na in air/propane/butane flames.The mathematical models developed accounted for changes in viscosity, surface tension, vaporisation rate and flame gas composition. Calculated values were within 5% of the experimental. Calculations for correcting solution viscosity effects have keen derived (839) and also partial oxygen radical pressures for flames at 2000-2500 K. The latter information was applied to a theoretical treatment of oxide formation for Ca, Ba, Al, Mg and Zn. The well known interference of Fe on Cr in air/acetylene flames has been ascribed to a formation of chromite FeCr,O, (821) and the releasing action of ammonium chloride was explained by the formation of Cr ammine complexes. The effects of acids on Co emission in air/acetylene flames (55 1) and the influence of ligands on the absorption signal (618) of the same element have been examined.2.4 NON-FLAME CELLS The development of non-flame AA devices is undoubtedly one of the current growth areas, and while their refinement has been a rather slow process,* commercial exploitation has taken the customary five years once the potential of the technique had been appreciated. Moreover, the field is now changing so rapidly that it is very difficult to form an opinion as to the detection limits achievable. Practical considerations such as contamination may well be the limitation for most users, A number of‘ reviews has been published (490, 564, 706, 733, 7 5 9 , 899, 902). 2.6.1 Tube Furnaces and Filaments Today almost all users possess commercially built devices rather than home-made designs.The Perkin Elmer HGA-70 ( 2 8 5 ) , based on a Massmann tube furnace, has been redesigned and is now produced as the HGA-72 in Europe or the HGA-2000 in the United States. The new design consists of an improved cell assembly and a power supply which is current-regulated, featuring continuous and stepwise temperature programming. Varian-Techtron have concentrated on miniature furnace configurations. In addition to their carbon rod and ‘mini-Massmann’ devices they have recently introduced two new atomizers (706, 761). The first is a hollow graphite cylinder (9 mm long by 5.2 mm outside diameter) which is held in the optical beam by spring-loaded graphite rods, pressing against the wall of the cylinder rather than the ends (as in larger Massmann furnaces).This device allows a larger solution sample than the earlier ‘mini-Massmann’ (5 p1 against 2 pl) as well as permitting more light t o pass through because of the larger bore. The second atomizer uses the same work head and is a graphite cup ( 9 mm high by 5.5 mm outside diameter) mounted vertically between two spring-loaded graphite rods. The cup has a transverse hole for the optical beam and is designed to take solid samples (1 - 10 mg) and liquid samples which may swell considerably on ashing. This has also been used as a filter and, by drawing air through it, atmospheric pollutants may be collected (2 14, 76 1). A West carbon rod is now marketed by Shandon Southern Instruments (486), Beckman Instruments retail a Massmann tube furnace, and Instrumentation Laboratory continue with their electrically-teated Ta strip (701, 702, 749).Using a hydrogen purge, Hwang et al. (749) have been able to detect elements Which form refractory oxides (Ca, Al, Si, Ti) by using the Ta strip. Several workers have reported the use of home-built devices, based on accepted designs. West and co-workers (2, 8, 9, 16, 135, 142,452, 455) have studied individual “L’vov, B. V.,Spectrochim. Acta, 1961,17. 761 and 1969,24B, 53. 20 Part I: Fundamentals and Instrumentation elements both in absorption and fluorescence using the graphite rod. They now iavour limited field viewing (9), achieved by optical masking, and reported to reduce interference effects.The use of a rapid response d.c. operational amplifier (9) is also advocated, or alternatively a storage oscilloscope ( 2 ) . A graphite rod atomizer with a hydrogen flame sheath (483, 703) a thick graphite rod capable of taking a 5 0 ~ 1 sample (27), a pyrolytic graphite tube microfurnace (477) and a glass roof over the rod permitting 600 measurements per rod (309) have been described. An electrically- heated ‘in-house’ design of a Ta strip atomizer (134, 563) has been described for use with biological samples; a W filament from a commercial light bulb was used (216) with an Ar sheath to obtain limits of detection for seven elements in 3-1-11 samples. An interesting modification has been described by Soviet workers (766) who replaced the 2.8- mm internal diameter tube in their Massmann furnace with a tube which had an internal diameter of 1.4 mm at the centre and tapered to 2.8 mm at the ends.‘This increased the absorbance signal from 3 pg Pb from 0.01 1 to 0.032, although the reason for the improvement is not immediately obvious. All of the non-flame cells mentioned have been heated by passing a high current at comparatively low voltage through the graphite or Ta cell itself. The use of induction heating has been advocated for some time by Morrison and Talmi* who have constructed a new furnace (439) heated by a 4 MHz supply, principally for atomic emission. The high excitation energies available allow the determination of I at 206.2 nm (612) down to 10 ng by emission.The only similar comparison is with the AA method of L’vov? who used an electrically heated graphite cell and determined down to 1 ng a t 206.2 nm. A similar r.f.-heated ‘well’ has been used for AA w i t i a Unicam SP90 spectrometer (419) and is capable of a maximum temperature of 2670 K. Sealed tubes, heated by an external oven; have been used to determine trace impurities in metals (14)$ and to measure vapour pressures of’ metals (14, 291 ). It would seem that the analytical applications are very limited although these devices could provide a homogeneous, well-defined, atomic population of great value in theore tical studies. The analytical potential of non-flame cells appears so great that very few theoretical studies have been conducted (cf.flame AF)! Furthermore, it would appear that a closer examination of materials science would help in the, development of non-flame atomizers. A paper by POCO Graphite lnc. (700) suggests that several novel and useful forms of graphite have been developed to meet specific requirements, These products behave very differently from conventional graphites and frequently represent extremes of possibilities. The theory of non-flame cells at present is Oilly poorly developed (here we are unfamiliar with much of the Soviet contribution in this area) and few physical measurements have been made. Reeves et al. (703), have measured the temperature profile of a hydrogen diffusion flame used to sheath a graphite i o d atomizer. The rod is situated, in this design, in a cooled flame zone (ca.SO0 K) which is oxygen deficient. The importance of the flame is clearly in the chemical, rather than thermal, environment it provides near the rod. Measurements of the temperature of graphite rod and atomic vapour (350) produced in a Varian ‘mini-Massmann’ furnace indicate *Morrison, G. H., and Talrni, Y . , AnalyI. C’hcrn., 1970, 42, 809. 1 1 ‘vov. B . V., and Khart\y/ov, A. D., J.arialyf. C‘hcrw. IiSSR, I O h Y , ’4, h i 6 $See also Kapperport, E. J . , et al., Rev. scicnt. Instrum., 1970. 41, 1168. 21 t’u r t I : h’u TI da 111 en ta Is aii d 111 s tru tn en tu t io ti that the atomic vapour is always cooler as a result of the open design of the cell. The two-line AA method (33, 84) was used to measure the vapour temperature.West (733) has presented a thorough and controversial paper on the carbon filament atomizer. Because of the rapid heating obtainable, more dense atomic populations are produced than with tube atomizers and molecular absorption is said to be less serious because of the short duration of the atomic pulse. West’s observation (733) of atomic phosplioresceizce of Cd at 326.1 nm excited by 228.8 nm radiation in nitrogen is extremely interesting. Unfortunately n o details of lifetimes or quantum yields are available. The maximum possible atomic population obtainable in a graphite tube furnace is easily calculated by a formula derived by Winefordner: * I1 = 6 x lo2* vcp/v Where: n is the maximum atomic concentration/atoms cm-3 C‘ is the concentration of element sought in solution/moles 1 - ’ V is the volume of solution added to the cell/cm3 p is the degree of atomization v is the inner volume of the cell/cm3 L’vov (See A K A A S , 1971, I , ref.360) considered the production of the atomic vapour as an instantaneous process followed by an exponential decay with time. De Galan (624) has developed a more refined treatment by assuming that the production a n d removal of vapour were described by different exponential functions of time. The derived results were then used in a theoretical comparison with conventional flame A A . 2.6.2 Cells for Mercury Determination Last year a brief historical summary of Hg determination by flameless AA was given (See A R A A S , 1971, I , 20) together with an admittedly irncplete table of commercially available instruments.We reiterate that nearly all manufacturers of AA instruments retail accessory kits for the determination of Hg and that electrothermal devices such as tube furnaces and filaments (Section 2.6.1) have been used for this purpose. Few significant developments have been reported in the determination of Hg by flameless AA and even fewer developments have occurred in its instrumentation. Hg vapour is generated from the sample in a number of ways. Reduction with stannous chloride/sulphuric acid is most usual, although incorporation of Cd salts ( 2 5 ) also reduces organomercurials including methylmercury. Apparatus designs for this include an improved reduction apparatus (507) and an jnexpensive cell (6 14). Mineral samples are determined by heating in air o r oxygen in a furnace followed by collection of the evolved Hg by absorption in permanganate solution (2) or by amalgamation on wires of Ag (21, 31, 456) or Au (62, 278, 281).Hg vapour is released from the wires by heating. Subsequent detection by AF (31,457, 734) is more sensitive than by AA even if modest light sources are used; small interferences by background absorption will be less than with ar, absorption system. The uses of microwave excited plasmas (2 13) and flow-through discharge lamps for Hg detection are also under investigation- the reader is referred to Section 2.2 in Part I and to Part IT of this volume. *Wini.l‘ortincr. J .I).. Atomic Absorption Spcctroscopy, Shefficlrl Sympo4um Plenary Lectures, Ed. Dagnall, R. M., and Klrkbright, G. F., Butterworths (London), 1970, p.36. Part I: Fundamentals and Instrumentation 22 Hg vapour detection using the resonance wavelength at 184.9 nm has been studied by Robinson et al. (858) using a quartz absorption cell and also by Kirkbright et al, (635) using a flame. The use of vacuum U.V. wavelengths would seem to offer very little cost-benefit advantage in this very sensitive determination. Windham (215) has described a simple device for correction of molecular absorption interference in the determination of Hg. After measurement of the absorption, the gas stream was deverted through a glass wool plug containing palladium (11) chloride which removed Hg from the stream; any remaining absorption was due to background.If the Hg detector is coupled with a gas chromatograph (642) or a temperature programmed evaporation device (278), the separation of individual Hg compounds is possible. Finally, two instruments for Hg determination have been reported and are worth comment. Shandon Southern (457) have incorporated their AF Hg accessory into a non-dispersive instrument using a solar blind photomultiplier-a detection limit of 3 ng/l of air was claimed. Scintrex Ltd. (222 Snidercroft Road, Concord, Ont., Canada, and Lincoln House, 296-302 High Holborn, London, WC 1V 7JJ) have introduced four Hg analysis instruments: the HGL-3 a laboratory instrument, the HGG-3 a portable instrument for geochemical prospecting, the HGP-2 a mobile or fixed station pollution monitor and the HGM-2 high sensitivity instrument which also measures total U.V. absorption, These systems are unique in that they employ Zeeman modulation and a double-beam layout with two detectors. A discharge lamp is situated between the poles of an electromagnet: When the field is off, normal AA plus background absorption occurs. When the field is turned on, the lamp spectral line is split into two components, both of which are outside the absorption profile of the sample vapour, so that the true AA is suppressed. The magnetic field is modulated at an unspecified frequency. A beam splitter, situated before the absorption cell, directs a small part of the radiation to a reference detector to compensate for lamp drift. Detection limits range from 4 - 10 ng m-3 IIg or 50 - 100 pg Hg. 2.7 OTHER EXCITATION AND ATOMIZING SYSTEMS As was found last year, several interesting papers have been noted which do not fit into the classification of Sections 2.1 to 2.6, Some of these (275, 626, 694), however, are concerned with the kinetics of reactions involving radiation in the vacuum U.V. region which falls outside the terms of reference of this report, These are therefore considered no further. D’Silva aBd Fassel (750) have shown that Hg 253.7 nm atomic emission can be excited by X-ray irradiation of Hg vapour in an atmosphere of 1% nitrogen in argon. Detection limits were expected t o be of the same order as those obtainable by AA following a similar cold vapour aeration procedure for sample treatment. Morgan and co-workers (769) have used time-resolved spectroscopy to observe the emission from the explosion of a thin A1 wire. Examination of the range 230-700 nm for up to 500 ys showed initially a strong continuum overlaid with broadened lines from neutral and singly ionised Al, followed sequentially by disappearance of singly ionised lines, cessation of almost all atomic emissions with cessation of current through the wire, strong emission of A10 bands and their persistence for sevet;al hundred microseconds. The initial excitatioz temperature was estimated as 10 000 (: and the subsequent gas temperature as 3000 C .

ISSN:0306-1353    

DOI:10.1039/AA9720200007

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

23 3 . Optics Part I: Fu n da m en ta Is an d Ins t r u m en ta t io n 3.1 LIGHT GATHERING AND BEAM MANIPULATION A versatile optical bench suitable for use in the development of spectroscopic instruments has been described (434). The base consists of a large steel plate, ‘/z to 1 inch thick, supported on rubber stoppers to reduce vibration. The optical components are mounted on quick-release magnetic mounts, the equipment can be rapidly rearranged and there is no tendency to creep even with vibrating parts such as mechanical choppers. Alignments are made by means of a low power (% mW) He-Ne laser and a series of height gauges. An extensive literature review on fibre optics (935) has been published but in atomic spectroscopy little use has been made of these devices.A single application (121) describes the use of a water cooled fibre optic probe to examine the spectra of emitting and absorbing gas layers in flames. The probe was connected to the entrance slit of the spectrograph. To increase the effective aperture of a spectrometer, Girard (414) has replaced the entrance and exit slits of the instrument by windows bearing a spiral pattern. One window was vibrated or rotated to modulate the optical beam while a tuned amplifier was used to select the desired wavelength, Some improvement in the light collection efficiency of an Ebert monochromator with an exit slit of 60 mm was achieved (448) by placing a plane mirror tunnel inside the instrument to reflect the exit beam toward the centre of the slit, thereby reducing its effective length to 25 mm, which matched the dimensions of the photocathode of the detector.Attempts t o increase the sensitivity in AAS by multiple passes through the atomic vapour continue to be reported (1 104), but generally prove to be of limited value. To obtain precise compensation for non-specific absorption in AAS by means of a deuterium corrector the alignment of the latter is critical. A revised procedure for its alignment in Perkin-Elmer instruments has been described (509). 3.2 WAVELENGTH SELECTION 3.2.1 Dispersive Systems In the production of dispersive elements the greatest interest has centred on holographic gratings. At a meeting of the Spectroscopy Group of the Tnstitute of Physics (London), the production and properties of holographic gratings were discussed (628, 629, 630, 446 and 744).A holographic diffraction grating is produced by making two plain coherent light waves interfere with one another on a photographic recording emulsion. This grating can have a high dispersive power with spatial frequencies exceeding those which can be ruled by the best machjnes. The spectra produced are free from ghost images, and have better resolution with less stray light than those from conventional ruled gratings. So far it has not been possible to produce completely satisfactory ‘blazed’ gratings. The gratings are best used at near normal incidence to avoid anomalies arising from the beams travelling over the grating at grazing incidence. Concave holographic gratings present a most interesting possibility as they could be designed to reduce the effect of astigmatism and spherical aberration which, in turn, should lead to high quality spectra at large aperture.Two papers (82, 627) have reviewed recent developmments in the production of conventional ruled diffraction gratings while prism and grating monochromators have 24 Part I : Fundamentals and Instrumentation been reviewed by Backlians (535). Reduction of the stray light of a Czerny-Turner monochromator was obtained by incorporating a second diffraction grating to produce additional dispersion ( 5 25). Cresser et al. (720) have continued to demonstrate the advantages of an echelle monochromator in emission flame spectrometry and AAS with a continuum source. They found that spectral interferences in flame emission were only slightly greater than in AAS using a line source, and that sensitivities and detection limits in AA using a continuum were only a little worse than those obtained with a line source.The resolution of their instrument (Spectraspan model 101) varied from 370 000 t o 930 000 with a dispersion of 0.1 1 to 0.042 nm/mm at 500 and 200 nm,respectively. This instrument was suitable for the study of spectral line profiles and gave results comparable with those obtained by a Fabry-Perot interferometer. A procedure for the wavelength calibration of an echelle spectrometer has been described ( 146). The accuracy obtained was of the order of 10- nm provided the temperature was constant to within k 0.005°C.Repetitive wavelength scanning has been combined with a continuous wavelength scan (7 11) and applied t o qualitative flame emission spectroscopy, The technique, which has been reported previously (See ARASS, 197 1, I,p.23, ref. 64 l ) , is claimed to be particularly useful when the analytical line lies on the shoulder of a band because a derivative spectrogram is generated. Simple wavelength scanning has also been used (22) as a means of multi-element analysis by flame atomic emission spectroscopy. The scan rate was 450 nm/min and the output osciilographic recorder had a rise time of 0.01 s. Detection limits were slightly worse than those obtained by conventional means but could be improved by using a slower scan rate.Pruett (1069) has described a photon-counting wavelength-scanning instrument which minimises the scan time while maintaining resolution, precision of the light intensity measurement and capability t o accommodate a wide range of light intensities. A stepping motor is used to drive the wavelength scan while the dwell time at each wavelength setting is determined by the time taken t o accumulate the number of photo-electric pulses necessary t o give the predetermined statistical accuracy. High-speed narrow-waveband scanning has been used t o measure line and background intensities alternately (964). Elser and Winefordner (42) combined mechanical chopping of the light from a continuum source with narrow-band wavelength scanning to isolate the absorption signals in AA.Detection limits of the order of 0.1 mg/l and analytical curves linear over three orders of magnitude were obtained. 3.2.2 Interferometers I n an attempt t o improve the finesse and contrast of Fabry-Perot interferometers, Roychoudhuri and Hercher (450) used a multi-pass system in which the light beam passed through different areas of the same pair of plates a number of times, A major advantage of this system over a single pass is the increased flexibility afforded by the choice of both reflectants and number of passes, thus finesse and contrast can be adjusted independently. Piezo-electric ceramic elements were used for the final alignment. A high resolution piezoelectrically scanned Fabry-Perot interferometer has been used (491), to study the degree of self-absorption in HCL’s, vapour discharge lamps, FDL’s and flames.A time-averager computer was used to scan atomic line profiles in absorption. Air bearings have been used t o support the moving plate of an interferometer (449) to give a scan rate of 0.25 fringe per second. The linear motion was provided by a precision screw driven by a stepping motor and linked to the mirror Purt I : Furidamen fals and lnstrurnentation 25 system via a steel ball resting on a glass plate. By combining a Fabry-Perot interferometer with an echelle monochromator and a pin-hole entrance aperture, Wyller and Fay ( 159) produced a system in which the resolution was variable from 0.5 to 0.0005 nm over the spectral range 350 t o 1300 nm.This system had a 10-fold gain in .light intensity, with very low stray light, over conve:itional grating systems of comparable resolution. The transmittance element of a Michelson interferometer was eliminated by using a diffraction grating as the beam splitter (417). This instrument, which could be used into the far u.v., was used to generate the Fourier transforms of spectra by an oscillatory drive of one of the interferometer mirrors. A variable narrow-band optical filter can be obtained using the chromatic aberration of a Fresnel zone plate lens (418). This system has a Large aperture and a resolution of approximately 3.6 x l o 5 at 600 nm. A similar principle has previously been employed in a spectrograph (See ARASS, 1971,1, p.23, ref.696). A six-channel interference-filter spectrum analyser with a resolution of 0.6 nm and a spectral range of 5.0 nm, was used to study laser light scattered by a plasma (833). 3.2.3 Non-Dispersive Systems Non-dispersive systems for the selection of analytical spectral lines have the advantage of simplicity and greater light gathering power over other arrangements. The application of these techniques t o AAS have been discussed by Walsh (542). A non-dispersive system developed by Vickers (45) for AFS used a chlorine optical filter and a solar blind photomultiplier to limit the spectral range of the instrument. The flame emission below 280 nm from air/hydrogen and oxygen/argon/hydrogen flames was negligible. Using these flames the detection limit for Hg was 1 mg/l and for As 0.3 mg/l. A new type of U.V. band pass filter has been suggested by Olson and Rigney (472). These proposed filters are alkaline metals deposited on t o a substrate such as Supersil silica. The low wavelength cut-off is determined by the filter substrate and the high wavelength cut-off by the aIkaline metal (Na 204.8 nm, K 284.1 nm and Rb 313.7 nm). Alloys of these metals will produce intermediate cut-off points. Ahmed and Ley have described (30 1) a three-stage tunable bi-refringent filter using variable phase shifters of ammonium di-hydrogen phosphate. A voltage of 5 25 kv was used to scan the spectral range 250 nm t o 700 nm. The bandwidths were 41 nm at 350 nm and 80 nm at 700 nm. At 500 nm the transmission loss was 3’3% of the incident radiation.

ISSN:0306-1353    

DOI:10.1039/AA9720200023

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

Part I: Fundamentals and Imtrumentation 26 4 Detector Systems Almost all papers relating t o photo-detectors are based on direct reading systems. The characteristics of the photographic plate have been studied in depth for many years and consequently current work is inevitably, in part at least, a repeat of earlier studies, e.g., the effect of storage on the k-transformation constant (218) and the effect of development conditions on the blackening distribution of a uniformally exposed plate (956). The quantum efficiencies of sever'al photographic emulsions and of a 1P21 photomultiplier have been compared (1071). At 633 nm the former ranged from 4 x 10- to 1.1 x TO- events/photon. 4.1 PHOTOELECTRIC DEVICES 4.1.1 Photomultipliers The effective quantum efficiency of an end-window photomultiplier can be improved by multiple internal reflections within the envelope (See ARAAS, 197 1, I, p.25, refs. 430 and 431). This technique has been used by Prydz and Kolberg (156) who deposited a thin light-tight layer of Ag on the cathode window, optical coupling being provided via a lucite light pipe skew-mounted against the window. The coating was held at 30 V with respect t o the cathode which was earthed. This arrangement eliminated the build-up of electrostatic charge on the window and 'frosting' when the tube was cooled. Significant differences were found in the response sensitivity across the slats of a windowless venetian blind photomultiplier (775). The addition of a mesh in front of the photo-cathode improved the homogeneity and sensitivity of the response.The spectral sensitivity of S10 photocathodes used in RCA 6217 and 5819 photomultipliers has been examined (148) on cooling to -200 A-TS0 !goJ f 400-650 nm was used. On cooling to -17OoC the response of the 6217 photomultiplier increased by 1.5 times, while that of the 5819 photomultiplier fell to 10% of its original value. At constant temperature the sensitivities of both tubes continued t o change slowly with time. The dark current of a 1P28 photomultiplier was reduced 500-fold on cooling with liquid nitrogen in a simple housing (834). Using the well-established technique of biasing the envelope of the photomultiplier tube to near cathode potential, Davies (123) has reported that the dark current of uncooled photomultiplier tubes (EM1 978 lA, RCA 93 lA, Fairchild 7860) could be reduced by as much as three orders of mqgnitude and that in some instances an improvement comparable with that obtained by cooling the photocathode was produced.Rossetto and Mauzerall (471) gated a 1P21 photomultiplier by applying pulses to the first, second and sixth dynodes. Using this system they were able to observe weak light pulses which followed pulses 10' times greater in intensity by switching on the photomultiplier for the duration of the second weaker pulse only. The gating pulses were 10-500 ns with a rise time of 9 ns. Pulsed operation has also been used to reduce fatigue effects (1070). The same paper identified the sources of non-linearity in photomultiplier response as space charges between cathode and first dynode and between last dynode and anode as well as dynode voltage changes due to dynode current drawn from the resistor chain".For use in multichannel optical measurement a commercial instrument incorpor- ating a video processor is now available (1 20). *Land, P.L., Rev. Scient. Instrum., 1971,42, 420. 27 Yurt I: Fundamentals and instrumentation 4.1.2 Solid State Detectors These devices have not yet reached the state of development where their sensitive area and response is comparable with that achieved using conventional photo-tubes particularly in view of their relatively large dark currents. Where long-term stability is required over a period of an hour or more it is claimed that photo-diodes are more stable than photomultipliers (475), but for the observation of rapid events photomultipliers are more suitable.The noise and frequency response of a Si photo-diode/operational amplifier combination has been investigated ( 4 15). When the diode was used photovoltaically the dark noise was random but when used current-wise with a pure resistive load the latter determined the frequency response and noise in the system. A development of considerable interest (793) is an n-n ZnSe-Si isotype heterojunction photo-diode which can discriminate between simultaneous light signals of different wavelengths in the range 500-1 100 nm by proper adjustment of the bias voltage. 4.2 SIGNAL PROCESSING 4.2.1 Photon Counting While papers on this subject continue to appear, their contents tend to confirm the earlier findings of the advantages and limitations of the technique (See ARAAS, 197 1 , 2 , p. 26).Some of the earlier work on photon counting has been reviewed by Malmstadt, Franklin and Horlick (436). The principal advantages are in the measurement of low light intensity and in the production of digital signals. Ingle and Crouch (34) have found that, compared with direct current measuring techniques, in spectrometric work the signal-to-noise ratio in photon counting was only 5-25% higher when the measuremefit was shot-noise limited and if source or background flicker noise predominated, then the signal-to-noise ratios were the same. They also noted (43) that, at high light levels, pulse overlap could cause non-linearity.Afterpulsing due to the generation of ions has been shown t o produce no significant decrease in the signal-to-noise ratio in photon counting (1 60). Dagnall and co-workers (1, 227, 680) have studied the application of photon counting systems to analytical atomic spectroscopy. They described a simplified system (227) in which l,he dark current of the photomultiplier was reduded by a factor of 30 on cooling to-40 C. Using photomultipliers under high gain conditions the pulse amplitude generated was adequate to drive the counting circuit without further amplification. As photon counting is a digital process, drift and gain variation in the photomultiplier tube, which affect pulse amplitude, have no effect on the number of events provided there is no pulse height descriniination.Further, the systems are capable of fast response which may be an advantage when studying transient events such as those from non-flame atomizers. Aldous and Bailey (730) have described the application of a photon-counting AA and AF spectrometer to the determination of trace metals in liquids using discrete sampling systems. 4.2.2 Signal-to-Noise Ratio Heiftje has reviewed the use of instrumental techniques for signal-to-noise enhance- ment (474). The topics considered include modulation, filtration, blocking amplifiers, signal averaging, box-car integration and correlation techniques. From the point of view of improving signal-to-noise, the noise power density spectrum is a fundamental parameter which determines whether modulation techniques effect any significant Part I : Fundamentals and Itistrumcrztcrtioia 28 imFrovement in the signal-to-noise ratio.This parameter has been investigated in terms of spectrometers in general by O'Haver and Thomas ( 7 2 7 ) . They confirmed that the photo-current shot-noise exhibited a sqvare-root dependence on the total photo-count and on the bandwidth of the system. The relationship between noise level and the detection limit in spectrochemical analysis has been discussed by Haisch ( 176). Three limiting conditions are considered (1) a constant noise level; (2) a noise level determined by a photon flux; (3) a noise produced by statistical effects in the photomultiplier.Morris ( 1097) has considered the noise equations for single-beam and double-beam AA spectrometers. The noise level of double-beam instruments is shown to be at least 1.4 times that of single-beam instruments, but the former is claimed to be necessary for high-precision AA to overcome lamp drift. 4.2.3 Miscellaneous Signal Processing Two systems have been described for the measurement of samples of low absorbance, One is an analogue integrator (12) for use with the Perkin-Elmer 303 AA spectrometer. The other is a general purpose system designed for use with an absorption spectrometer provided with double-beam optics (45 1). It is claimed that the latter system will measure transmittance values approaching 100%) with an accuracy of k 0.001% and that it can handle a 10 000-1 variation in input light intensity. Electronic circuits for use with single (1 1 10) and double-beam ( 1 107) AA instruments and for generating reference signals for use with phase sensitive detectors ( 7 7 3 ) have also been described. A mechanical light chopper with a % duty cycle for use with lock-in amplifiers is claimed to give an output which is independent of the waveform sf the input signal (798). For the modulation of light beams at frequencies up to 50 kHz a piezo-optical birefringent modulator* is now available from Hinds International Inc., Portland, Ore., U.S.A. The modulation amplitude may be varied from f 0.005 wave to * 0.5 wave retardation at 1 ,urn. The device can be used over the spectral range 180-2600 nm. "Kemp, J.C., J. Opt. SOC. Am., 1969,5Y, 950.

ISSN:0306-1353    

DOI:10.1039/AA9720200026

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

29 5 Data Processing Part I: Fundamentals and Instrumentation 5.1 EMISSION SPECTROSCOPY An IBM 360 computer has been used (109) to calculate the Fourier transformations and deconvolutions in a technique for resolution enhancement of line emission spectra. A data processing system was constructed (550) t o meet the requirement for the pre-processing prior t o large-scale computer analyses, giving signal-to-noise ratio improvement, resolution enhancement by deconvolution and the decomposition of overlapping bands. Results were given for the Ar emission spectrum, and overlapped absorption bands of ammonia. A programme for the analysis of variance has been worked out for Minsk-22 and M-220 computers (104) and has been used on various examples of emission spectrographic, AA and flame emission methods of analysis.The method of approximation by polynomials was discussed for use in the optimiLation of computer solving of emission spectrographic problems (778) with examples in Cu/Ni and Cu/Zn alloys, and the calculation of the analytical curves and limits of detection by the computer ( 3 3 1, 89 1) gave a uniform comparison of spectrochemical methods, Virtually all the mathematical methods used to assess spectra on photographic plates quantitatively are outlined, with the pros and cons of each method discussed, in a book by Torok and Zimmer (936). The authors have developed a system which would appear to offer an improvement, and the work may be a worthwhile contribution t o the art. Computer methods for the extraction of analytical information from any spectrogram obtained from both prism and grating instruments (896) and connected to automatic spectrophotometers (868) have made possible totally automatic evaluation of spectra, and offer possibilities for the calculation of intensity and elimination of the matrix effect.Although an all-electronic digital read-out system for microphotometry, with the scanned output automatically stopped at the point of minimum transmittance corresponding to line peak is not new, one having been shown by Mostyn and co-workers at the Physics Exhibition, Manchester, U.K., 1965, the system described by Mason (73)is claimed t o offer advantages of high precision, reliability, speed and case of operation.The combination of a single standard with a graphical method for calibration in the spectrographic determination of Cd in high purity Zn (772) was successful. The semi-quantitative analysis of low-alloy steels, stainless steels and Cu alloys, using the NBS table of spectral line intensities (205) requiring no standard sample, was studied. A critical evaluation (567) has been made of the optical performance of some coinmercially available microphotometers; calibration curves were plotted by the two-step method described in ASTM E 1 16-59T. The sharp spectral lines excited in an Fe HCL were photographed to obtain the test line images using a 3.4 m plane grating spectrograph. 5.2 ABSORPTION SPECTROSCOPY There is a wide variety of systems available for data processing in AA of which the most familiar are probably the direct concentration read-out devices retailed by all manufacturers of high-performance spectrometers. Teleprinter outputs are supplied by 30 Part I: Fundamentals arid Instrumen tatiori some manufacturers to enable print-out of sample number, concentration and formatted analytical reports.These last are discussed in some detail (391) relative to the Pye Unicam SP 1900 and a general programme for an off-line computer has been described (724). A simple interface system (384) for a Cary 14 solution spectro- photometer and a Hewlett Packard 2114A computer may be applicable to AA instruments. Several circuits for conversion of logarithmic to linear scales have been given (536). Incorporation of analogue measurement systems in a servo loop to control the sample input of automatic AA spectrometers has been used in two instruments (72 1, 731). Control of gas supply and wavelength setting is included in one of these (731).

ISSN:0306-1353    

DOI:10.1039/AA9720200029

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

Part 1: Fundamentals and Instrumentation 31 6 Complete Instruments Table A lists commercially available emission spectrometers and spectrographs. This was compiled with the assistance of the manufacturers, who responded almost quantitatively to our questionnaires. We believe this information t o be up-to-date, accurate and comprehensive. ’Table B similarly tabulates current instruments for AAS and again we are happy t o acknowledge the help of the- manufacturers in its compilation. To the best of our knowledge it is correct as a t February 1973 and we note that several product lines have been discontinued since 197 1. A similar survey of non-Soviet instruments has been carried out by users in the USSR (603). It is interesting that the correlation of data on commercial instruments warranted the production of a book for that specific purpose ( 1 123).6.1 FLAME-BASED INSTRUMENTS Price (940) has given an authoritative review of advances in AAS instrumentation during the last decade and identified the main features as the introduction of improved light sources, the nitrous oxide/acetylene flame and non-flame atomizers. Winefordner (760) has discussed the advantages .and disadvantages of AA and emission flame spectrometry. He pointed out that both sequential and simultaneous AFS were ideally suited t o multi-element trace analysis. Barnett and Kahn (46, 738) have described their assessment of AF using a modified Perkin-Elmer 403 and have concluded that ‘the present situation for analytical AF seems dim but the future may yet be bright’.Malmstadt and Cordos (731, 732, 900) have described an automated AF spectrometer for the rapid, sequential determination of up t o eight elements. Standard HCL’s were used and optimum conditions for each element pre-set. Interchange between elements required about 3s. Another group (762, 763, 984) has built an automated instrument specially for detection of the transient signals obtained from discrete sampling or non-flame atomizers in AAS or AFS. This comprised a photomultiplier detector with photon-counting electronics, the counter being inter- faced to a mini-computer for programmable control of instrumental parameters. Precision was considerably improved over that obtainable using standard peak-height measurement techniques.Last year (See ARAAS, 1971, 1, p.35) ten papers dealing with non-dispersive instruments for AF were discussed. The advantages of wide acceptance angles and short path lengths were reflected in impressive detection limits when compared with monochromator instruments. Regrettably we have received little further news of these developments and also note that the non-dispersive Techicon AFS-6 instrument, designed for simultaneous multi-element analysis, is no longer available. West (728) has extended the use of the AFS-6 t o include the determination of B , P, S and halogens by molecular emission spectrometry. Vickers et al. (736) have used a filter of C1 gas between the flame and detector to eliminate OH background emission.This filter has a transmittance of 80% at 240 nm, >40% over 190-260 nm and 0% over 280-390 nm. The Soviet Spektr 2 (290) is a filter AA spectrometer which also features automatic sample delivery and periodic baseline adjustment. A serri-automated chemical laboratory for teaching and research in landscape geochemistry has been described (767). This was based upon a Perkin-Elmer 403 AA spectrometer and routinely provided over 6000 element determinations per month. Replacement of a standard Technicon flame photometer by an IL. 143 in a system for W E 3 Irncrh Suonlier Angstrom Inc., P.O. Box 248. Belleville, U.S.A. IAllLr A Tvnr D.R. D.R. D.R. D.R. D.R. U.R. D.R. AsQuanto- meter 33000 Mich. 48111. Applied Research Labs.Ltd.. Wineate Road. LutGn, Bedfordshire, England. Typr of source 3.0m (HV a x . spark 1 HV or LV a.c. arc DC arc LV undirectional Arc. triggered 1.5 m As model 22-101 1.5 m As model 22-101 0.75 m As model 22-101 Various, low voltage, high voltage, multi- source (HVS, LV, d.c. arc). repetition (400 60 50 16 68 16 D.R. 60 (114 lines) D.R. 111 (36 lines) C'OMME KCIALI Y AVAILABLI EMISSION SPbCT KOMtTE RS spectrum monitor. Various excitation stands and read-out options. I capacitor discharge Optical interlock As model 22-101. All wavelengths in vacuum. In-focus wavelength scanning + 3 nm from each Reciprocal dispersion/ Wavelength Focal No. of Channrls nm uer mm ranwlnm 21-30 0.278 21(r518 0.397 211t634 0.556 160-440 0.556 178-320 0.556 190-770 21&350 I70407 17&407 Model 22-101 V-70 V-71 A-70 A-71 Quanto- vac 80 Quanto- meter 80 Quanto- meter 290008 Quanto- vac 29500B Quanto- pact 30,000 Quanto- meter 33000 Quanto- vac 33000 Annlications All ferrous and non- ferrous applications.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. All ferrous and non- ferrous alloys. powders including slags, sinten, ores. rocks, ceramics, soils. etc. Solutions. oils. etc. As Ouantovac 80 but excl;ding determination of C, S and P. As Quantometer 80 Ferrous including low- and high-alloy steels, irons.Particularly suited to non-ferrous materials, e.g.. Al. Mg, Cu, Zn and white metals. Slags, powders, solutions including oils. As Quantometer 80. As Quantovac 80. 1 Surcial frnturrs receiver slit. 0-75 m As model 22~~101 Mobile and operable in hostile environments. All wavelengths in vacuum. Various excitation stands and read-out options. As model V-70 As model A-70 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. Typewriter, teletype, digital computer options.Air or argon excitation stands. Typewriter. teletype and digital computer options. Automated sequential analysis. As Quantometer 33000. 2 -i 5 L-. 3 2 3 2 2 rb 0.556 0.556 0-46 D.R. 60 (84 lines) 0.695 or 0.35 190-610 D.R. 48 (90 lines) 0.35or0 175 190 520 0 46 or 0.23 190 630 0 56 or 0.28 19&705 0 695 or 0 35 19&840 174-235 0 4 6 D.R. 15 (36 lines) 1.388 or 0.695 200 800 0.695 or 0.35 2W4UO (64 lines) 0695 or 0.35 IY(r610 (8 references) 0.46 1 Om 1.0 m As Quantovac 80. 1.5 m As Quantovac 80. 1.0 m Low voltage (50 sparks s-I). High sparks s-l). 0.75 m Low voltage. high voltage. I .O m As Quantovac 80. 1.0 m A5 Quantovac 80. 3 a 5 a -+ $ 2.rb 3 2. I’hot. I 5 or 0 75 370 740 Arc or spark. Built-in order sorter - stl- I I’hot. 1 0 Arc or spark. Built-in order sorter General spectrographic analysis. General spectrographic k 3 5‘ GW-l Bard Atomic Inc., SB-I 175 Middlesex Tuinpike. Bcdford. Mass. 91730, Phot. 0 8 o r 0 4 rn analysis. General spectrographic U.S.A. Arc or spark modular or RE-I analysis. Dual gratings for simultaneous photo- graphy of two spectral regions. Phrrt. 0 8 or 0.4 As GW-l Transient, weak or complex sources. General spectrographic analysis. 0 6 or 0 . 3 I 0 111 Arc or spark modular. ;i 6 Ferrous metals (except S) using C 193.1 nm, P 214.9 nm in 2nd order. 5 Non-ferrous metals, oils. High speed (fil5 5) gratings for rapid examination of transient and/or weak sources or complex spectra Optional echelle grating for U12.1 aperture. Compact, low-cost direct reader with minimum air-condition- ing requirements.Logarithmic read-out. Manual master monitor to check slit alingment. Easy interchange of gratings. Easy conversion to a direct 1 5 m I 5 ni 2 n ni 2 0 ni 3 0 m As GW-I 3 0 m As GW-I 2 0 m 2 0 m As Spectroniet 1000 As Speclromet 11. All 450-750 I H S 2400 1x5 2400 210 590 0.55 o r 0.28 180-2250 180 2250 190-432 1Y0-863 173-432 - 30 60 60 GWR-I SPCCtrolllel I000 GX- I GX-3 Spectromet I1 Speclrovac I 1 D.n. Phot. D.n. 16 per head (normally 3 heads per unit) D.R.1 j . K . si c: 2 R a 2 -+ 3 W W General spectrographic analysis. reader (GX-3). 0-3Y Normal direct rcading capabilities as well as photographic (see GXI). Automatic optical servo monitor continuously maintains correct slit alignment. Logarithmic read-out. Precise electronic setting of slits. 0.294 0.59 All direct reader applications above 190 nm. As Spectromet 1000. Automatic optical servo monitor continuously maintains correct slit alignment. Logarithmic read-out. Manual master monitor to check slit alignment. Temperature compen- sated fixed focal length. Dual stands for argon and air available. 0.29 All direct reader applications including C, P and S. photomultipliers in vacuum.TABLE A COMMERCIALLY AVAILABLE EMISSION SPECTROMETERS - continued Type Phot. Phot. Phot. Type of source 66-716). or 0.75 m As above. Up to 30 0.56 or 0.28 200-800 or 1.5 m As above. 3.4m As.above. 0.5 m Special features 20 inch camera. Choice of 3 gratings. Nitrogen purging extends range 10 175 nm. Optional accessories permit use as direct reader or scanning spectrometer. Comnact. direct reader. Air-conditioning not required. Choice o f 2 gratings. Scanning optional. Easy interchange to photo- graphic (70-310) and scanning version (70-320). I spectrometers, some (Various scanning 7 with vacuum capabilities. Reciprocal dispersion/ Wavelength Focd length Nu. uf Channels nm per mm rangelnm 1.1 or 0.54 210-780 - 1.0 to 0.24 180-3000 depending 180-1500 upon grating 180-750 1.5 m Various available in “Varisource” 3.4 m unit including spark, low- and high-voltage a s .arc, d.c. a&. 2OD-6000 0-75 m Also versatile 200-6OOU 2W6000 4.4 to I .I 3.2 to 0.8 1.6 to 0.4 1.0 m “controlled wave. form excitation 2-0 m source”, (model 168-500 0.54 1.8 or 0.9 1.2 Applications Wadsworth spectrograph. General spectrographic analysis. General spectrographic analysis. Versatile instrument particularly suitable for measuring transient spectra. Most metallurgical analyses. All direct reader applications above I90 nm. All direct reader applications above 190 nm. Suitable for spectro- 1 scopic investigations rather than,for analytical applications.Liquids and solutions. Liquids and solutions. Supplier Jarrell- Ash Division, Fisher Scientific Co., 590 Lincoln Street, Waltham. Mass. 02154, U.S.A. V. A. Howe & England. Jobin-Yvon, 1 rue du Canal, 91160 Longjumeau. France. Model 78-000 70-310 75-150 Co. Lid., 88 Peterborough Road, London S.W.6. Model 750 D.R. Model 1500 D.R. Atom- counter 70-3 I4 82-000 78460 (78-480) 78490 25420 78-120 82400 (82-410) VARAF DELTA - - Up to 30 30-1 - - 1 t 190-400 0.34or0.11 200-510or 190-250 As 70-310 As 70-310 3.28 ’ 200-9M: 200-800 195-770 D.R. Scan. Scan. Scan. Scan. Scan. Scan. Scan.Scan. 1.0 m 0.465 m Flame. . Ohm Flame. I J Czerny-Turner mono- chromator, adjustable bandwidth 0.024 nm. 9 models with various burner and read-out options. Czerny-Turner mono- chromator, adjustable bandwidth 0.024 nm. Automatic wavelength scanning device. Pneumatic nebuliser, laminar flow burner. Ultrasonic nebuliser optional. Reciprocal dihpersionl Wavelength Focal rangelnm lrngth Model Type of source 170-430 or 1.5 m Low voltage at 50, PV 8300 190-407 oils. Supplier M.B.L.E., 80 Rue des Deux Gares, Bruxelles 7, Belgium. 190-700 PV 8200 (SM 150) No. of channel^ nrn per rnm Up to 90 Up to 60 - Type D.R. D.R. Phot. D.R. Optica S.A.S., Via Gargano 21, 20139 Milano, Italy.0.55 0.55 0.69--0.36 200-800 0.69 or 0.36 220-420 0.37 100, 200, 300, 400, 500 Hz. 1.5 m Low voltage source or source with d.c. arc up to 40 amps t- HV spark at 20 kV. 1.2 m All conventional types available. 1.2 m LV triggered arc and spark. HV spark, a.c. and d.c. arc. 1.5 m LV triggered arc and spark. HV spark. 1.0 m Controlled and non- controlled HV spark, a x . arc. 1.2 m LV triggered arc and spark. B5 ESA I ESA 3 Special feature5 Can be equipped with two spark stands. Can connect source unit providing low voltage conditions I d.c. arc up to 40 amps + HV spark at 20 KV. - Stigmatic instrument General purpose. with rotating Ebert grating. Double spark stand both General DurDose metal- in air and inert atmos- I urgical inaiysis, e.g.Al, Pb, Zn, Fe, Cu alloys. Wear metals in phere. Rotrode for solutions. Air-vacuum instrument with all exit slits accessible from outside for adjustment. Many analytical programmes can be arranged in parallel for easy inter- change. Computer facilities available. Scanning monochro- mator with one channel for analytical line and another channel for reference using reflected beam principle. Combined vacuum mono- and polychromator. All excitable elements accessible with scanning system. Applicationlr Steels, iron, non- ferrous metals, slags, Solids, liquids and powders. Non-ferrous metals and oils. oils, etc. Complex analyses involving many spectral lines.Metallurgical works. All material excitable with same source parameter. Routine analysis (including C, S and P) of iron and steel. Non-ferrous alloys. 16 B5C 165-440 B7 v 93 - 9 D.R. Scan. D.R. 0.41 0.36 20&500 160-500 (40 nm as polychro- mator) w wl TAULI? 4 COMMERCIALLY AVAILABLE EMISSION SPECTKOMGTEKS ~ continued - ,./ length ~~~~~ - ~ 177-310 0.546 0.25-2.0 or 174.0-447.7 236.5-607.4 0,741 not stated Model Rank Precision E 1000 Industries Ltd., Polyvac Analvtical DiviGon, Westwood Polyvac E 600 Polyvac Industrial Estate, Ramsgate Road, E 900 Margate, Kent, England. E77718 25 36 - Type of source Solid state electronics. Quartz plate for atmos- thyratran-controlled pheric pressure com- a.c.arc. or low voltage spark, a.c. or d.c. below 200 nm. Direct arc, continuous and reading attachment available. As above. Direct reading attachment available. As above. As above. h'o. of Channels nm per rnm rangelnm Reciprocal dispersion/ Wavelength Focal 1.5 m Various. I J9.6-864.3 0.293- 1 I55 60 0.6 m Condensed arc or hiah-reuetition condensed arc. 0.75 m Various. 1.57 m D.C. arc, HV spark Adjustable slit. Spectral condensed arc. 1.5 m Various. 0.53 m HV spark, con- densed arc, d.c. arc, 3.5 m Glou discharge Paschen-Runge lamp, high, medium mounting specially designed for range intermittent. 2.0 m As above. 1.5 rn Direct reading attachment available.1 - 5 m Glow discharge lamp Novel source claimed to General quantitative {others available). exhibit no background, no matrix effects, linear calibrations for all 1.0 r i i 1.7 ni 1.0 m Not stated. Not stated. 1.0 m Modular unit, low and high voltage spark. Applicurion\ Ferrous and non-ferrous alloys. Geological samples. Wear metals Low and high alloy and tool steels, irons, slags. alloys. Wear metals in oils. Flash photolysis. Examination of line profiles. Non-ferrous metals. Soils. Addiitves and wear metals in oils. General analysis. General analysis. General analysis. General analysis. analysis. Type D.R. D.R. D.R. Phot. D.R. Phot. Phot. D.R. Phot. Phot.D.R. Supplier Siemens Ltd., Great West Great West Road, House, Brentford, Middlesex, England. Shimadzu GE-100 Seisakusho Ltd., GE-170 14-5 Uchikanda GYM-100 I-chome, Chiyoda-ku, Tokyo 101, Japan. Special features Dual spark stands. Computer-controlled instrument. Dual gratings give six systems. in oil. Solid-state electronics with or without computdr control. Curved entrance and Ferrous and non-ferrous exit slits. Solid state electronics or computer controlled. Air or vacuum. Analysis of high-purity length 067 m of which specimens having com- 0.24 m can be selected for plex spectra. Deter- a given exposure. mination of trace element concentrations. Routine qualitative and quantitative analysis.Czerny-Turner mono- chromator. pensation. elements 0-loox,. Not stated. Not stated. Not stated. 191-800 200- I200 0.26-0.97 12 .- 'r 2 2 5 1! g 5 2 a 2 L=. + 1 Y E 549 Medium Quartz R.S.V. GmbH.. SPN 3.5 8031 Hechendorf Pilsensee, West Germany. SPN 2.0 Phot. SPN I5 Phot. SPh' 1.0 ANALY- MAT 200-600 150-490 170-410 0.5-10 0.24 or 0.16 Not stated 0.42 or 0.28 Not stated 1.1 or 0.56 Not stated 1.68 or 0.84 Not stated 0 54 or 0.31 200-650 or 1.66 or 0.83 200.- 0.48 0.46 - - 200- 84 - - 24 - 2 5 -.. =. -+ Tyw ofsourcr lrngrh ? “rl 2 5 E: 2 a ;sl Speciui frururr Double spectrometer Double spectrometer. Vacuum instrument, single or double pass.Supplirr Spectrametrics Spectra- Inc., 2nd Avenue NW Park, Burlington, Mass. 01803, U.S.A. Spex Industries I&., P.O. Box 798, Metuchen, N.J. 08840, U.S.A. Glen Creston, The Red House, 37 The Broadway Stanmore, Middlesex, England. H A 7 . .. -. 4nl. . - _. VEB Carl Zeiss DSA-240 Jena. 69 Jena, Carl-Zeiss Str. I , German Democratic Republic. Carl Zeiss Jena Ltd., 93/97 New Cavendish Street. PGS-2 London W 1 A 2AR, England. Reciprocal dispersion/ Wuvelengfh Focal No. of Chonnrls nm per mm rangelnm 190-900 0.04 1 0.75 m Plasma jet or flame. High dispersion echelle spectrometer. Plasma 0.75 m Plasma jet. 190-900 0.04 10 (inter- changeable casettes) 180-1 500 0.75 m 0.85 m 0.55 0.44 0 8 175-1280 IlCb1500 0.5 m 0.75 m 0.75 m 1.0 m 175-1500 175-1500 180-1500 210-550 Up to 56 - 210-550 - Modrl span 101 Spectra- span 210 1402 1404 I500 l702/ 1704 1802 Q-24 T-vpr Photo.or D.R. Phot. or D.R. Phot. Scan. Phot. Phot. Phot. D.R. Phot. Phot. Applications be less susceptible to chemical and spectral interferences than a flame and lo have a greater dynamic working Atomic analysis of metallic and some non- jet can analyse gases, metallic (CI, B, C, P) liquids and some solids. elements. Claimed to Qualitative capabilities in photographic mode. Background correction included. AA and emission mode with AF capabilities. range. As above.Multi-channel analysis with sensitivity adequate for trace level determina- tion of many elements. Suitable for research applications rather than production control. As above. As above. As above. As above analysis. 2 w 4 1.0 or 0.5 0.8 - Direct reading accessory available. Up to 11 lines analysable All steels and non- in single, automatic ferrous alloys. Bearing metals and solders. Mineral oils and aqueous solutions. scan. Choice of reference line. Built-in automatic temperature and pressure compensa- tiun. Digital signal averaging, display and recording. Rapid change of analysis programme by programme store. Automatic expansion of measuring range. Stigmatic depiction. General spectrographic Dispersion doubled by analysis.Also double passage of light. examination of line Predisperser for order profiles, hyperfine sorting and isolation. structure, etc. Gratings interchangeable. Automatic transport of plate holder. Full range of accessories General spectrographic available. 1.1 0.8 0.78 0.74 or 0.37 200-2800 0.76 I.0m 0.54 m Spark 2.075 ni Arc or spark. 0.54 m Arc or spark. w 00 TABLE B COMMERCIALLY AVAILABLE INSTRUMENTS FOR ATOMIC ABSORPTION SPECTROSCOPY Supplier Description Identification Model 485 Double or single beam, single- and triple-pass optics, grating 1200 lines/mm, blazed at 250 nm, resolution 0.2 nm, automatic filter selection, 50 x scale expansion, meter display, Beckman Instruments, 2500 Harbor Boulevard, Fullerton, Calif.92634, U.S.A. Model 495 As model 485, 100 x scale expansion, digital display. Bausch and Lomb, 142 Linden Avenue, Spectronic AC2-20 Double grating monochromator, 3-speed wavelength drive accessory, manual stray light and 2nd order filter selection. Use also as emission or UV solution spectrophotometer. Rochester, N.Y. 14625, U.S.A. FMD 3 Carl Zeiss, 7082 Oberkochen, Wur t temburg, West Germany. Single beam; range 193-852 nm: grating monochromator, resolution 0.05 nm; 4 lamp turret with 2 stabilised power supplies; digital read-out; curve correction; auto zero; optional automatic calibration and background compensation. EEL 140 EEL 240 Evans Electroselenium Ltd., Church Lane, Brain tree, Essex CM7 5SL. Single-beam, 0.25 m modified Ebert-Fastie monochromator, grating 11 80 lines/mm, dispersion 3.5 nm/mm, non-linear dial read-out, single lamp turret. Single-beam, 0.25 m modified Ebert-Fastie monochromator, grating 11 80 lines/mm, dispersion 3.5 nm/mm, f/& aperture, 4-lamp turret, integration, meter read-out.Modular system Moduiar AA, flame emission, UV spectrophotometer with double or single beam optics and various amplifier/detector systems. Heath/ Schlum berger, Benton Harbor, Mich. 49022, U.S.A. Heath (Gloucester) Ltd., Bristol Road, Gloucester GL2 6Ell. Hilger and Watts See Rank Precision Industries. Model 208 Hitachi Ltd., Nissei Sangyo Co. Ltd. 15-12 Nishi-Shimbashi, 2-Chome, Minato-Ku, Tokv 0.Japan. ? 2 rrc ? rL s" 3 P 3 2; g R -+ ro 2= Single beam, Czerny-Turner monochromator, grating 1440 lines/mm, dispersion 3 1.8 nm/mm, 3 lamp turret, 20 x scale expansion, direct concentration read-out, 2 wavelength drive. -+ -4. e 3 11- 3 5 3 instrumentation Laboratory Inc., 11 3 Hartwell Avenue, Lexington, Mass. 02173, U.S.A. Double beam, dual-channel, 0.33 m I bcrt monochromator, grating 1200 li n t's/nini, tii spe r FI o n 2.5 n ni/ mm , ap c r t u re t 19. 6-1 amp t 11 rrc t , background correction, automatic gas control, dual digital display, auto-7cr0, auto-calibratr, integration, wavelength drive. Uses interference filters as sccond monochromator. Same as IL 353 but single channel, single digital display.Instrumentation Laboratory (UK) Ltd., Station House, Stamford New Road, Altrincham, Cheshire, England. IL 2 5 3 Dial A tom I1 A tomsorb Model 82-500 (hlaximum Versatility ) Model 82-81 0 VARAI: DICLTA iLlodcl6000 Single beam, 0.25 m Czerny-Turner monochromator, grating 11 80 lines/mm, dispersion 3.3 nm/mm, aperture f/7.5, 2-larnp turret, 10 x scale cxpansion,metcr read-out. Single beam, 0.25 m 1:bert monochromator, grating 1 I80 lines/mm, Jarrcll-A sh/ Fisher, 590 Lincoln Street, Waltham, Mass. 021 54, U.S.A. W. A. IIowe & Co. Ltd. 88 Petcrborough Road, London S.W.6. aperture f/3.6, up to 6-lamp turret, 20 w scale expansion, meter read-ou t. Single beam, 0.5 m Lbert monochromator, gating 11 80 lines/mm, various options available on gratings, slits and wavelength drive, multi-pass optics, 6-lamp turret, 20 x scalc cxpansion.Double beam, dual channel, two 0.4 m J-brrt monochromators, gratings 11 80 lines/mm, dispersion 2.1 nmlmm, 5 speed wavelength scan, integration, au to-zero. digital display. Jobin-Yvon, 1 rue du Canal, 91 160 Longjumcau, France. Single bcam, 0.465 m Czerny-Turner monochromator, grating 1220 lines/mm, dispersion 1.8 nmlmm, turbulent or laminar flow burners, recorder or digital display. Single beam, 0.6 m Czcrny-Turner monochromator, grating 1200 lines/mm, dispersion 1.2 nmimm, automatic wavelength scan, optional ultrasonic nebuliscr, recorder or digital display. Optica, Via Gargano 2 1, Single beam, 0.35 m kbert monochromator, automatic filter insertion, pre- focused water-cooled hollow-cathode lamps, continuous regulation of flame temperature, 50 x wale cxpansion, auto-concentration, integration, digital display.201 39 Milan, I talv. ~ continued - U P TABLE B COMMERCIALLY AVAILABLE INSTRUMENTS FOR ATOMIC ABSORPTION SPECTROSCOPY Model 103 Model 107 Single bcam, 0.27 m Littrow monochromator, all mirror optics, grating 1800 lincs/nim, dispcrsion 1.6 nrnlmm, single lamp fitting, autozero, auto tlame ignition, flame emission, 50 x scale expansion, integration, meter display. As Model 103 with electronic digital display. Perkin-Elmer Corp., Norwalk, Conn. 06854, U.S.A. Perkin-Elmer Ltd., Post Office Lane, Beaconsfield, Model 300 Bucks.Single beam, 0.4 m Czerny-Turner monochromator, UV grating 2880 lines/rnm, dispersion I .O nm/mm, \’IS grating 1800 lines/mm, dispersion 1.6 nrn/mm, changeover 420 nm, automatic filter insertion, single lamp fitting, automatic 3-lamp turret optional, autozero, auto-concentration, curve corrector, 40 x scale expansion, automatic flame ignition, flame emission, gas flow interlocks, D2 background corrector, electronic digital display. HP9 1QG. Perkin-Elmer & Co., CmbH, D 7 7 7 Weberling en, West Germany. Pye Unicam Ltd., York Street, Cambridge CB 1 2PX. Rank I’recision Industries Ltd., Analytical Division, Westwood Industrial l:statc, Kmsgatc Koad, Margatc, Kent. Model 300s Model 305 Model 306 Model 403 As Model 300 with analogue meter read-out.Double beam, 0.4 in Czerny-Turner monochromator, UV grating 2880 lines/mm, dispersion 0.6 nmlmtn, VIS grating 1440 lines/mm, dispersion 1.3 nm/mm, spectral bandpass 0.03 nm - 7.0 nm switch selcctahle, single lamp fitting, autozero, auto-concentration, curve corrector, 200 x scale expansion, built-in flame emission, automatic flame ignition, direct meter and digital null read-out, optional D2 background corrector. As Model 305 hut with electronic digital read-out, auto flame ignition/ extinction, signal integration. As Model 306 but with computcd signal averaging, built-in flame sensor and automatic lamp current adjust, intrrfacc lor read-out to teletypewriter. SP 9 0 Series 2 Si’ 1900 Single beam, Littrow monochromator, 303 reat alurninised silica prism, dispersion 3 nrnlrnrn a t 200 nm, 6 nrn/znm at 250 nm and 32 nm/mm at 400 nm, 3-turret accessory, 10 x scale expansion, mctcr display.Single or double bcam, IIbcrt monochromator, prating 1 BOO lines/mm, dispersion 2.2 nm/nim, 20 x scalc expansion. 10 x scalc contraction, auto-zero, integration, 6-lamp turrct, digital display. As SP 1900 but with single lamp fitting. SY 1950 Atoms.pek Mk 3 Singlc heam, 60’ silica prism monochromator, dispersion 1.7 nm/mm at 200 nm, 44.6 nm/mm at 500 nm, 6-lamp turrct, meter display, scale expansion, digital read-out accessory tor concentration with curve correction and in tcgra ti on. v > 2 z 3 2 2 2 A c k c-r 3 ; u \ k 2 - - Y c1 b 7 c f.b - 5. .4-3000 Singic beam, 0.25 m Czcrny-Turner monochromator, grating 630 lines/mm. diywnion 6.0 ninlmm, 4-lamp turrct, 10 \ scalr cupanyinn. mcter display. Shandon Southern Inqtrurncnt5 Ltd. I rirnlcj Road, Camberley, Surrey, lngland. A-3300 Single beam, 0.25 rn CLerny-Turner monochromator, grating 630 lines/mm, dispersion 6.0 nmlmm, 4-lamp turret, auto-zero, emission/absorp tion/fluorescence modes, 5 and 10 second integration in all modes, 25 Y scale expansion, wave- length drive, meter display, DVM output socket. Shandon Southern Instruments 1 n c., 515 Broad Street, Sewickley. Pa. 15143, U.S.A. AA-6 10 Single beam, Czerny-Turner grating monochromator, two lamp turret, 10 x scale expansion, meter display, wavelength drive.Shimadzu-Seisakusho Ltd., 14-5 Uchikanda 1-chome, Chiyoda-Ku, Tokyo 101, Japan. Spectraspan 101 Single beam, compact 0.75 m Llchelle grating spectromcter, dispersion 0.04 nm/mm a t 200 nm, laminar flow burner or plasma jet source. Primarily intended for emission but can be used for AA with a Xe continuous source and photoelectric or photographic detection. Spectrametrics Inc., 2nd Avenue, NW Park, Burlington, Mass. 01803, U.S.A. Spectraspan 210 As Spectraspan 101 but with 10-channel capability. Interchangeable cassettes allow selection of alternative 10-line combinations. Model 1000 Model 1100 Single beam, 0.25 m Czerny-Turner monochromator, grating 1276 lines/mm, dispersion 3.3 nm/mm, aperture f/8, 4-lamp turret, flame emission, 10 x scale expansion, auto-zero, meter display.As Model 1000 but uith 3 and 10 second integration, peak signal retrieval, curvature correction, auto concentration, 50 x scale expansion. Varian Techtron Pty. Ltd., 679 Springvale Road, Springvale, Vic. 3 17 1, Australia. Varian Associates Ltd., Russell House, Molesey Road, Walton-on-Th ames, Surrey, England. As Model 1100 but with digital display. Model 1200 AA-120 Single beam, 0.25 m Ebert monochromator, grating 1276 lines/mm, dispergion 3.3 ninimm, aperture f/lO, 4-lamp lurret, 10 x scale expansion, 2-speed wavelength drive, auto-zero, meter display, A A-5 A A-6 Single beam, 0.5 m libert monochromator, grating 6 3 8 linesimm, dispersion 3.3 nm/mm, aperture f/10,4-lamp turret, 10 s scale expansion, auto-zero, rnetcr display, modular construction.Modular construction similar to AA-5 but with integration, peak height Varian Aerograph Inc., 2700 Mitchell Drive, Walnut Creek, Calif. 94598. U.S.A. retrieval, optional digital or meter display. 42 Part I : Fundamentals and Instrumen tation the continuous measurement of Na and K in serum has been described ( 5 17). Four reagent kits and two flame photometers have been evaluated (514) for these same determinations, the NIL 4-7000 being assessed favourably in comparison with a Coleman Jr Model 2 1. Coleman, however, have subsequently (707) introduced an up-dated version, designated Model 5 1 , of this instrument. An interesting patent for a flame photometer has been noted (150).In this, the sample is impregnated on a porous, combustible support and fed on a non-combustible tape into a small flame. This is claimed to result in less contamination, better control, greater precision and less heat. Flame photometric detection of P and S after gas chromatographic separation continues to be widely used. An improved detector has been designed (708) which uses a pre-mixed oxygenlhydrogen flame and is reported to compare favourably with a diffusion flame detector with respect t o detectability, specificity and linearity. An F-specific gas chromatographic detector has been described (526) in which a stream of metallic Ca in Ar was brought into a small oxygenlacetylene flame.In the presence of F, CaF emission at 529.9 nm was observed and this was measured after passage through an interference filter (527). The Pb content of various anti-knock Pb-alkyls has also been measured (76) by a combined gas chromatographic-flame photometric method, using an oxygen/hydrogen flame and measurement at 405.8 nm, It has been shown (243) that NO, NOz and other N-compounds can be detccted by flame emission by viewing the luminescence of the 680-760 nrn band system in a hydrogen-rich flame. 6.2 SPECIALISED EMISSION INSTRUMENTS In the chemical and other industries in the U.K., there is increased awareness of the importance of ensuring that plant construction materials are of the specified type of alloy. Duff (465) has described the use of a ‘Metascop’ portable spectroscope for in-situ testing of such materials. An extensive analytical scheme has been developed for identification of alloys based upon estimation, by visual inspection of spectra, of their Fe, Mn, Cr, Ni, Mo, Ti, Cu content. Attachment of a modified Polaroid camera provided a permanent record of the spectrum and also permitted estimation of elements such as Al, Nb and Si whose emission lines are around 390 nm. It is probable that there will be increasing emphasis on this type of ‘insurance’ analysis as it is realised that serious industrial accidents, with consequent loss of production, can be caused by use of incorrect materials. Three papers (94, 96, 779) in the Russian literature have provided technical data on Russian spectrographs and spectrometers. The last of these is particularly noteworthy as it compares the parameters of the DFS-1 Om, DFS-3 1, MFS-3, DFS-36 and DFS-4 1 direct-reading instruments with those of the ARL Quantovac 3 1000 and Hilger Polyvac E-600 widely used in Western countries. Olson (120) has described the SSR Instruments New Optical Multi-channel Analyser, which incorporates a video processor with provision for electronic chopping t o minimise stray light and dark current. A spectroscopic isotope-line shift method has been used (880) for the determin- ation of waterJheavy water ratio. A special twin HCL was devised as the light source. This was cooled by liquid air and the spectrum examined by means of a Fabry-Perot interferometer. The H 486.133 nm line was used. An N-analyser (Model NO1-5, marketed by CZ Scientific Instruments Ltd., has been noted which is based upon Part I : Fundamentals and Instrumentation emission spectrometric isotope analysis. Spectral separation of the U.V. band heads due to 14N, and I 4 N 1 ' N molecules is sufficient to permit quantitative assessment of the isotope ratio in a given sample. 43

ISSN:0306-1353    

DOI:10.1039/AA9720200031

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

44 Part I: Fundamentals and Instrumen tation 7 Anciltary Equipment 7.1 STANDARDS This Section contains four tabulations of suppliers of standard materials in everyday use by analytical atomic spectroscopists. None of these lists can be claimed t o be fully comprehensive but they should be helpful t o the analyst who experiences difficulty in obtaining any of these items. The accumulation of an adequate range of standards for calibration of a direct-reading emission spectrometer can be a lengthy and costly process, particularly if a wide range of materials is t o be analysed. Instrument manufacturers should perhaps be encouraged t o provide libraries of standards for loan, hire or sale t o new workers in the field. To help t o alleviate the preseiit situation in which it is necessary either to prepare standards oneself - no mean task for metal alloys-or to beg, borrow or steal from friends who have collections, Table C contains all suppliers of metal and alloy standards known to the compilers of this volume.Unfortunately we cannot comment from personal experience on the quality, e.g., homogeneity, of material from all of these suppliers. Gegus (220) has tested the micro-homogeneity of pig-iron samples. Tables have also been provided (405) of recommended or proposed values for maior, minor and trace elements of ten French geochemical standards. Difficulties arise a t times with supplies of graphite electrodes needed for emission spectrographic analysis. Table D lists suppliers of these. (Many types) (Cu-base) TABLE C - SPECTROGRAPHIC STANDARDS 2 3 1.BNFMRA,Euston Street, London N.W. 1 , U.K. (Cu-base) BCIRA, Bordesley Hall, Alvechurch, Birmingham, U.K. ( C as t I r o ns) Bureau of Analysed Samples Ltd., Ncwham Hall, Newby, Middlesbrough, Teeside, U.K. 4 G. I. Willan Ltd., Sheffield Works, Catcliffe, Rotherham, Yorks., U.K. (Fe-base) 5 Henry Wiggin & Co. Ltd., Holmer Road, Hereford, Herefordshire, U.K. (Ni-base) Tyseley Metals Ltd., Kings Road, Birmingham 1 1, U.K. 6 7 (‘Specpure’ metals) 8 International Alloys Ltd., Bicester Road, Aylesbury, Bucks., U.K. (A I-base) 9 Morris P, Kirk & Son Inc., 2700 South Indiana Street, Los Angelcs, Calif. 90023, (Al-, Sn-, Pb-, Zn-base) 10 Apex Smelting Co., 6700 Grant Avenue, Cleveland, Ohio 44 105, U.S.A.Johnson-Matthey Chemicals Ltd., 74 Hatton Garden, London EC 1P 1 AE. U.S.A. Red House, 37 The Broadway, Stanmore, Middlesex, U.K.) (Al-, Zn-, Mg-base) 11 Spex Industries Inc., Box 798, Metuchen, N.J. 08840, U.S.A. (Glen Creston, The (Many types) 12 Alcoa Research Labs., New Kensington, Pa., U.S.A. (Imperial Aluminium Co. Ltd., Impalco House, P.O. Box 15, Droitwich, Worcs., U.K.), (Al-base) 13 Office of Standard Reference Materials, NBS, Washington, D.C., 20234, U.S.A. (Many types) 14 Zinc et Allioges, 34 Rue Collange, 92-Dervallois-Perret, Boite Postale N o 182, (Z n-base) France. Pechiney, 23 Rue Balzac, Paris 8e, France. 1.5 (Al-base) 45 (Fe-,Al-base) Part I: Fundamentals and Instrumentatior? 16 Bundesanstalt ftir Materialpriifung (BAM), 1 Berlin 45, Unter den Eichen 87, Germany.17 CKD Praha, Na Harfe 7, Praha 9, Vysocany, Czechoslovakia. TABLE D - SPECTROGRAPHIC GRAPHITE ELECTRODES 1 Johnson Matthey Chemicals Ltd., 74 Hatton Garden, London EC 1P 1 AE, 1J.K. 2 La Carbone (GB) Ltd., Portslade, Sussex, U.K. 3 Ultra Carbon Corp., P.O. Box 747, Bay City, Mich. 48706, U.S.A. (Heydon & Son Ltd., Spectrum House, Alderton Crescent, London N.W.4, U.K.) 4 Spex Industries Inc., 3880 Park Avenue, Metuchen, N.J. 08841, U.S.A. (Glen Creston, The Red House, 37 The Broadway, Stanmore, Middlesex, U.K.) 5 National Carbon Co., Union Carbide Corp., 270 Park Avenue, New York 17, N.Y., U.S.A. (ARL Ltd., Wingate Road, Luton, Beds., U.K.) 6 Poco Graphite Inc., P.O.Box 1524, Garland, Texas 75040, U.S.A. 7 Baird-Atomic Inc., 33 University Road, Cambridge, Mass. 021 38, U.S.A. (Baird-Atomic Ltd., 42 Station Lane, Hornchurch, Essex, U.K.) 8 Met Bay Inc., P.O. Box 610, Bay City, Mich. 48706, U.S.A. 9 General Graphites Inc., First & Monroe Street, Bay City, Mich. 48706, U.S.A. 10 Ringsdorff-Werke GMBH, Bonn, Bad Godesberg, West Germany (Mining & Chemical Products Ltd., 70/76 Alcester Road South, Birmingham 14, U.K.) TABLE E - STANDARD METAL SOLUTIONS (MS) & 'REAGENTS FOR AAS' (R) 1 Johnson Matthey Chemicals Ltd., 74 Hatton Garden, London FC'lP IAE, U.K. 8 6 7 5 Koch-Light Laboratories Ltd., Colnbrook, Bucks., U . K . May & Baker Ltd., Dagenham, Essex, RMlO 7XS, U.K. V.A.Howe & Co. Ltd., 88 Peterborough Road, London S.W.6, U.K. Hartman-Leddon c'o. Inc., 60th & Woodland Avenue, Philadelphia, Pa. 19 143, U.S.A. M S J . T. Baker Chemical Co., Phillipsburg, N.J. 08865, U.S.A. 4 3 K 2 BDH Chemicals Ltd., Poole, Dorset, BH 12 4NN, U.K. M S , R Hopkin & Williams Ltd., P.O. Box 1, Romford, Essex, RM 1 LHA, U.K. MS, R Fisons Scientific Apparatus Ltd., Bishop Meadow Road, Loughborough, Leks., LE11 ORG, U.K. ills, R R R MS MS, R 10 Bio-Kad Laboratories, 32nd & Griffin Avenue, Richmond, Calif. 94804, U.S.A. M S nzs IJ3.A. 9 1 1 Barnes Engineering, 30 Commerce Road, Stamford, Conn. 06902, U.S.A. 12 Aldrich Chemical Co., 940 West Street, Paul Avenue, Milwaukee, WK 53233, R 13 Fastman Organic Chemicals, bastman Kodak Co., Rochester, N.Y.14650, U.S.A. R R R 14 K & K Laboratories, 121 Express Street, Plainview, N.Y. 11803, U.S.A. Alfa Inorganics, 8 Congress Street, Beverly, Mass. 01915, 1J.S.A. 16 Instrumentation Laboratory Inc., 1 13 Harwell Avenue, Lexington, Mash. U.S.A. 15 46 Part I: Fundamentals and Instrumentation M S 17 Spex Industries Inc., Box 798, Metuchen, N.J. 08840, U.S.A. MS 18 Nitine h e . , 45 S Jefferson Road, Wippary, N.J. 07981, U.S.A. (and of Shannon, Ireland) MS 19 E. Merck AG, 61 Darmstadt, West Germany. R 20 Carlo Erba, Divisione Chimica Industriale, Via C Imbonati 24, 20 159 Milano, Italy MS TABLE F - ORGANOMETALLICS 1 Messrs. Burt & Harvey Ltd., Brettenham House, Lancaster Place, Strand, London W.C.2, U.K.2 Durham Raw Materials Ltd., 1-4 Great Tower Street, London EC3R 5AB, [J.K. (Nuodex Ltd., Birtley, Co. Durham, U.K.) 3 BDH Chemicals Ltd., Poole, Dorset, BH12 4NN, U.K. 4 Hopkin & Williams Ltd., P.O. Box 1, Romford, Essex, RM 1 IHA, U.K. 5 Office of Standard Reference Materials, NBS, Washington, D.C. 20234, U.S.A. 6 AIfa Inorganics, 8 Congress Street, Beverly, Mass., 01915, U.S.A. 7 Eastman Organic Chemicals, Eastman Kodak Co.,. Rochester, N.Y. 14569, U.S.A. 8 Angstrom Inc., 2454 West 38th Street, Chicago, Ill. 60632, U.S.A. 9 J. T. Baker Chemical Co., Phillipsburg, N.J., U.S.A. 10 ContinentalOilCo.,P.O. Box 1267, Ponca City, Okla. 74601, U.S.A. (“Conostan” series) 11 ROC/RIC, 11686 Sheldon Street, Sun Valley, Calif.91352, 1J.S.A. (Kodak Ltd., Kirkby, Liverpool, U.K.) 12 National Spectrographic Labs. Inc., 19500 South Miles Road, Cleveland, Ohio 44128, U.S.A. 13 Baird-Atomic Inc., 33 University Road, Cambridge, Mass. 02138, U.S.A. 14 E. Merck AG, 61 Darmstadt, Germany. (Anderman & Co. Ltd., Central Avenue, East Molesey, Surrey KY 8 OQZ, U.K.) 15 Carlo Erba, Divisione Chimica Industriale, Via C Imbonati 24, 20159 Milano, Italy. In these cost-conscious days, analysts may prefer to purchase ready-made standard solutions rather than t o use valuable time in preparing their own standards. There are many suppliers of such soh tions and of reagents, e.g. spectroscopic buffers, complexing agents and organic solvents, specifically intended for use in flamc spectroscopy and characterised by the low levels of potentially important contaminants.These are listed in Table E. There is little doubt that when trace amounts of elements are t o be determined it is false economy to purchase other than the highest available purity of all the necessary reagents. Methods of analysis of such materials have been described by Johnson and Watson (390). Table F lists commercially available standards of organo-me tallic materials for which there is an increasing need in both emission and absorption spectroscopy. 7.2 DOCUMENTATION A film (458) describing the historical evolution of AAS from 1953 to 1970 is now Part I: Fundamentals and Instrumentation 47 being shown at symposia and can be recommended for the balanced view it gives of the subject.The AAS bibliography produced regularly by Slavin has been up-dated (355, 377) and the ASTM proposed recommended practices for AAS have been summarised (304). A Perkin-Elmer group (357) has described the direct measurement of limits of detection for a comprehensive list of elements using both flame and graphite furnace sampling. Results were tabulated. A tabulation of 2513 atomic emission lines by element and by wavelength for elements from Li t o Co in the wavelength range below 200 nm has been compiled by Williams (274). 7.3 MISCELLANEOUS Potential sources of contamination in trace analysis have been critically examined by Scott and Ure (464, 1096). Common laboratory materials, viz., glass, rubber, polythene, etc., can cause problems, and it is frequently desirable to take precautions to minimise laboratory dust. Duff (933) and Slickers (957) have reported on the benefits of inserting a rare-gas purifier into the argon supply line to a direct-reading emission spectrometer. The former used a conventional gas chromatographic system while the latter employed active Cu and a molecular filter which reduced the oxygen and water concentrations t o <lpg/g. Rutgers (147) has provided data on the temperature and spectral radiance of standard light sources, viz., tungsten strip lamp, anode of a carbon arc, high pressure xenon arc, which should be valuable when inter-laboratory comparisons of absolute emission sensitivities are attempted. Finally, the sampler described earlier by Raiteri (See ARAAS, 197 1, I , p. 43, ref. 178) has been reported in a more accessible journal (832). This device permits transfer of “IOpg of a large metallurgical product into the electrodes for spectrographic analysis, thereby avoiding the need to transport heavy components.

ISSN:0306-1353    

DOI:10.1039/AA9720200044

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

PART I I METHODOLOGY Introduction In Part 11, the term Methodology covers all aspects of the application of the techniques and instrumentation of atomic absorption, emission and fluorescence spcctroscopy t o chemical analysis. The format adopted for Volume 1 has been retained, with the subject matter treated under the two principal headings of (1) General Techniques, covering the preparation of samples and standards and the interpretation of the experimental data, and (2) Applications, where specific methods of analysis are reviewed and tabulated. The classification of the range of applications into eight main fields of analysis also follows the pattern of Volume 1. Some dupiication of entries may be found in instances where a method is relevant to more than one section. 49

ISSN:0306-1353    

DOI:10.1039/AA9720200048

出版商:RSC    

年代:1972

数据来源: RSC

摘要:

ELEMENT EXPLANATION OF THE TABLES Each of the Applications sections, 2.1 to 2.8, is accompanied by a table which summarises the principal analytical features of the references from which thc corresponding section is compiled. These Applications Tables form a convenient source of information for analysts interested in particular elements, matrices, sample treatments or atomization systems. In many cases, sufficient detail is given for the analytical procedure t o be followed; absence of such detail usually means that the information was not directly available t o the compiler of the table and the original reference should be consulted. The key to the tables is given below. h nm MATRIX CONCENTRATION TECH. ANALYTE SAMPLE TREATMENT A brief indication is given of the sample pre-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 analyscd. The concentration range or level of the element in the original matrix, expressed as % or pg/g for solids and mg/l or ,ug/ml for liquids. The atomic spectroscopy technique is indicated by A (absorption.), E (emission) or F (fluorescence). The form of the sample, as presented to the instrument, i s indicated by S (solid), L (liquid) or G (gas o r vapour). ‘d.1.’ = detection limit in the analyte. required t o 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 t o the main Reference Section, which gives the title of the paper and the name(s) of the author(s), with address. 49

ISSN:0306-1353    

DOI:10.1039/AA9720200049

出版商:RSC    

年代:1972

数据来源: RSC