ISSN:0003-2654    

DOI:10.1039/AN98409FX037

出版商:RSC    

年代:1984

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98409BX039

出版商:RSC    

年代:1984

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN98409BP077

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST OCTOBER 1984 VOL. 109 1229 Spark-source Mass Spectrometry Recent Developments and Applications A Review Jeffrey R. Bacon and Allan M. Ure Macaula y Institute for Soil Research Craigiebuckler Aberdeen AB9 2QJ UK Summary of Contents 1 Introduction 2 The vacuum spark discharge 2.1 The vacuum spark model 2.2 Ion formation 2.3 Ion production and sensitivity 2.4 Negative ions 3 Analytical procedures 3.1 Specific procedures 3.2 3.3 Sample preparation 3.4 Pre-concentration 3.5 Solutions 3.6 Electrode preparation 3.7 Sparking procedures 3.8 Photoplate evaluation 3.9 Spectral interferences 3.1 0 Sensitivity Isotope dilution spark-source mass spectrometry 4 Instrumentation 4.1 General 4.2 Ion source 4.3 Analyser 4.4 Ion detectors 4.5 Data processing 5 Applications 5.1 Metals 5.2 Non-metallic high-purity materials 5.3 Other manufactured materials 5.4 Environmental materials 5.5 Fuels 5.6 Geological and related materials 5.7 Biological materials 5.8 Miscellaneous materials 5.9 Standard reference materials 6 Concluding remarks 7 References Keywords Review; spark-source mass spectrometry 1.Introduction About 50 years ago Dempsterl introduced the idea of spark-source mass spectrometry (SSMS) as an analytical technique for elemental analysis and it is now some 25 years since the first commercial instruments became available. After a period of rapid development and application in the 1960s activity has declined and the technique is now at a critical point in its lifespan. There are several reasons for this.Firstly SSMS has generally suffered from the misconception that it was only a qualitative or at the most semi-quantitative technique. This probably arose because most of the original applications were in metals or materials research that needed quick-quantitative analyses. Further the instrumentation is complex and expensive and most commercial instruments in service are now becoming out-dated. There is virtually no active support from instrument manufacturers and at present there is only one instrument available commercially. It would appear that all development work is currently undertaken in users’ laboratories. That SSMS is still a very useful technique for certain applications is endorsed by the development work being undertaken and by the number of reported applications.The whole of the Periodic Table can be covered in one analysi 1230 ANALYST OCTOBER 1984 VOL. 109 with little or no knowledge of the sample type or element concentrations. All elements exhibit relatively uniform sensi-tivity for all sample types with detection limits of 0.1 p.p.m. or less in the solid sample. About 25 mg of sample are required for analysis but this can be reduced by using tipped electrodes. Considerable efforts have been made to achieve quantitative analyses and coupled with techniques such as isotope dilution a precision of 5% can be attained. The combination of multi-element analysis with little if any sample pre-treatment and the ability to achieve accuracy within an order of magnitude without calibration are particular strengths of the technique for certain applications.The major disadvan-tages of the technique are the generally poor precision which in most applications cannot approach that of other techniques, long analysis time complex instrumentation and expense. For the technique to survive and flourish it is essential that any major developments become available commercially and that simpler and cheaper instruments are produced. This review is intended to cover all developments and applications reported since 1972. For earlier work the reader is directed to the book edited by Ahearn2 published in 1972, while another recommended review of the principles and applications of SSMS is that by Deines.3 A detailed review of mass spectrometry of solids up to 1970 including a biblio-graphy was given by Honig.4 Various general and more limited reviews have appeared, including some in less common languages.These include wider reviews of inorganic mass spectrometry and solids analysis,'-H over-all reviews of the SSMS technique and instrumentation9-18 and more specifically the multi-element isotope dilution SSMS technique,lg descriptions of "recent" developments and applications ,2(&28 a discussion of systematic and statistical errors,29 a theoretical discussion of plasma formation30 and reviews of the application of SSMS for layer-by-layer analysis31 and the analysis of metals.32 34 geological samples,35 ?7 coal and other fuelsq3S 3y environmen-tal sarnples,@43 biological materials,4+45 forensic samples46 and liquids.47 2.The Vacuum Spark Discharge In the SSMS technique an electrical discharge is produced between two electrodes and in the process material from one or both of the electrodes is atomised and ionised. Because this process of ion production shows approximately equal sensitiv-ity for all elements it is possible to use the technique for semi-quantitative analysis without regard to the actual changes in sensitivity that occur. For qaantitative analysis, however the ionisation processes need to be understood and controllable. One of the major factors in the poor precision of the technique is the irreproducible production of ions during the discharge. Considerable research has been made into understanding the fundamental processes involved and, although the mechanisms of ion formation are still not completely understood real advances have allowed changes in instrumentation to be proposed.There are two kinds of discharge commonly used-the pulsed r.f. spark and the triggered low-voltage arc discharge. As most instruments use the former it follows that most research has been made into the r.f. spark source in which a pulsed 1-MHz r.f. voltage of up to 100 kV (peak to peak) is fed to the electrodes in vacuum. Various combinations of pulse length and repetition rate can be chosen. There have been three different approaches to understand-ing the discharge-study of the voltage breakdown and plasma formation itself study of the ion species formed during the discharge and study of the effects of spark parameters on ion production and sensitivity.2.1 The Vacuum Spark Model A detailed description of the physics of the electrical discharge is given by Franzen in Chapter 2 of reference 2. Since then the research of Chupakhin and co-workers4*58 has led to a model for the origin and development of the breakdown of the vacuum gap. Initially the breakdown was considered to consist of three distinct stages,48-52 but later a fourth pre-breakdown stage in which little ionisation occurs was added.58 The latter is considered to be relatively unimportant. The three principal stages can be summarised as follows. 1. Initiation of breakdown from the appearance in vacuum of the current carrier to completion of formation of the discharge channel. The beams of auto-electrons emitted from points located on the cathode surface bombard the anode and discharge energy in the surface layer.The ion current increases irreversibly and a plasma cloud is formed in less than 10 ns. 2. Spark stage a transitionary stage with a high-voltage discharge. The current increases at first and then dec-reases. The major ionisation process is electron bombard-ment of the anode. 3. Arc stage a low-voltage discharge and quasi-stationary state. Here ionisation occurs by sputtering i.e. ion bombardment of the cathode. By modifying the spark circuit it has been found possible to cut off the spark discharge at any time and thus study the three principal stages inde~endently.~H In the first stage the popula-tion of multiply charged ions was high whereas that of multi-atomic ions (known also as polyatomic cluster or molecular ions) was low.Compared with the later stages, relatively little of the electrodes was consumed. In the second stage the intensity of singly charged ions and multi-atomic species increased and electrode consumption was higher. Formation of ions in the final stage was determined by the properties of the element and condition of discharge i.e. it was variable and could affect the sensitivity. Consequently it was suggested that if the ion current during the arc stage was gated out the precision was improved sensitivity increased and the intensity of multi-atomic species decreased. The discharge model has also been considered in terms of the processes involved and the energy consumed.-5-7-55 57 In this model the three stages were as follows (a) atomisation, (b) ionisation and (c) plasma dispersal followed by ion detection .(a) In the atomisation stage a weakly ionised plasma was formed and only lo/" of the energy flux was used for atomising the sample .5h There were two mechanisms for atomisation: firstly explosive which was short and unselective and secondly by evaporation which was a longer process and was selective according to properties such as boiling-point and vapour pressure. In a short discharge of maximum power the explosive mechanism predominated i. e. all elements ionised with equal sensitivity.53 (b) During the ionisation stage. with an energy discharge of 1 x 10-4 - 5 x 10-3 J material passing into the electrode gap was completely ionised.(c) In the plasma dispersal stage intense recombination of ions occurred. This has been studied55 using a computer model in which the degree of ionisation of particles escaping the plasma depended on their characteristic recombination path length. If this was greater than the initial plasma radius all ions would escape before recombining. With a discharge energy in the range 10-4 - 10-3 J all ions with three or more positive charges would recombine but doubly and singly charged ions would not. If after dispersal all the multiply charged ions were converted in this way to doubly and singly charged ions the ratio of doubly charged to singly charged ions would be similar for all elements. Atomisation and ionisation depend mainly on the power (W) liberated in a discharge and the duration of the discharge ( t ) .These are not controllable in commercial instruments an ANALYST OCTOBER 1984 VOL. 109 1231 consequently the sensitivity is not equal for all elements. It is suggested55 that a new spark source is needed in which it is possible to control W and t. With a short powerful discharge the explosive atomisation phase predominates and all ele-ments are atomised and consequently ionised with equal sensitivity. A similar model of the spark discharge has been proposed following the use of a time-resolved chopping device to study 5-ps periods of the spark discharge.59- 60 In this the high-energy high-temperature portion of the spark pulse showed the formation of a plasma of electrons and ions together with neutral atoms.In the first 1-5 ps this non-equilibrium plasma produced ions of all energy states but higher energy states predominated. After a few microseconds the voltage and temperature dropped and a more stable lower energy plasma existed in which the singly charged ions predominated. Whereas Chupakhin et al.48 recommended that this final arc stage should be gated out to achieve better reproducibility, Franklin and Dean59.hO believed the spark stage should be gated out so as to accept only the more stable lower energy arc stage (30-90 ps of a 100-ps pulse). The latter strategy would effectively produce a triggered, low-voltage arc source as used in the Varian-MAT instrument. The advantages of the latter type of source were summarised by Radermacher and Beske61 as a much lower energy distribution a steady-state discharge with constant current, the production of ions of a higher ionisation state and a constant ratio of multiply to singly charged ions.Demortier and CO-workers62-64 used X-ray spectroscopy of the hard X-rays produced in the r.f. spark discharge to study the processes of electron emission that gave rise to the ions. The spectra showed that the r.f. electron current was produced by field emission and was very low the breakdown was initiated at sharp points on the electrode surface and actual breakdown involved thermoelectronic emission at high voltages (10-15 kV). This indicated that at the beginning of the breakdown highly multiply charged ions were produced at the electrode surface under the impact of energetic electrons.Shelpakova et ~ 1 . 6 5 have also demonstrated that admission of material to the plasma of the spark discharge is by an explosive mechanism in which atomisation of matrix and trace elements was equally probable and that the predominant mode of ionisation was electron impact ( i . e . in the spark stage of the discharge). The dispersal of ions has been studied for various sparking procedures by fitting a detector on a ring placed around the electrodes at different positions within the source.66 The material dispersed anisotropically and the distribution in space depended on the polarity of discharge and the impedance of the discharge circuit but not on variable spark parameters such as pulse rate and length.Muheim67-71 has studied the ionisation processes involved in the spark to obtain information on electronic and lattice structure in the matrix He concluded that “the intense r.f. spark electron bombardment on to the impact electrode probably gives rise to a positive space charge build-up under the isotropic and instantaneous explosive action of which at least part of the atoms are being smoothly disassembled, preserving largely the microscopic electronic structure of the real solid.”67 The resulting gaseous spark plasma represented an almost perfect picture of the actual solid state. From measurement of the intensity of multiply charged ions@ or multi-atomic ions70 it was possible to derive information on electronic structures and valence. Similar studies have been reported by Joyes et af.72 2.2 Ion Formation Three types of ion are produced during the spark discharge: multiply charged singly charged and multi-atomic ions.Time-resolved studies have allowed the formation of the ion types to be timed. Use of beam-chopping devices showed that ion formation was not uniform throughout the pulse”) and that the intensities of the multiply charged ions were at a maximum at 5-10 ps and then decreased to reach a minimum at 20 ps, whereas the singly charged ion intensities did not reach a maximum until 20-30 ps after which ion production was relatively stable.59,60 From measured and theoretical times of flight for ions produced by metal samples it was shown that doubly charged ions and ions from volatile elements arrived at the detector about 250 ps before singly charged ions.73.74 The evidence from the sparking of two dissimilar copper electrodes was that all these ions were produced in the vapour phase.Measurement of the formation times for multi-atomic ions showed that multi-atomic carbon ions appeared about 300-500 ns after singly charged ions and that other multi-atomic ions also appeared at a later but not so well defined time.27 All of the preceding evidence therefore indicates that the first ions to be formed are multiply charged ions that form the singly charged ions at a later time. The multi-atomic ions are the last to appear. There is however conflicting evidence on whether the multi-atomic species are ejected as such from the electrodes or are formed by recombination processes in the vapour phase.All the earliest evidence suggested that they were ejected as molecules from the electrodes. The fragmen-tation patterns of metal carboxylate salt mixtures were studied and found to be consistent with rapid decomposition of polymeric radical ions with the same composition as the salt .75,76 Recombination mechanisms charge exchange and solid - solid reactions were considered to have minimum effect. The study of spectra produced by CaF2 suggested that two groups of molecular ions were present which represented two different mechanisms of formation.77 One group was formed by evaporation from the electrodes and ionisation by high-energy particles whereas the other mechanism involved sputtering and/or thermal ionisation processes.In both mechanisms the multi-atomic species were released as such from the electrodes. Most recent evidence has shown however that multi-atomic ions are formed in the vapour phase predominantly by recombination processes. The simple fact that electrodes formed by mixing a metal oxide (MO) and carbon (C) will produce plentiful CO+ but little MO+ was stated by Stefani78 as evidence for this. Analysis of alkali metal halides showed that M2X+ ions were formed by a gas-phase mechani~rn.~~ From the analysis of A1,0,. species produced by alumina mixed with salts containing enriched 1 x 0 it was concluded that A10 and A120. and probably other A1,0, species were formed by single-atom reactions in the spark plasma.xO.8l Ramendik et al.x2 used electrodes made from erbium or its oxide to show that the dominant mechanism for the formation of multi-atomic ions was association.Evidence from the analysis of metals is less definite, however and the conclusions drawn are contradictory. Multi-atomic ions formed by copper - aluminium alloys showed abundant mixed Cu,Al species that could only be formed by recombination yet the authors83 argued that large aggregates such as AP+ could not be formed by the same mechanisms and suggested a mechanism in which multi-atomic ions are formed at an interface that separated the plasma and the solid crystalline phases. A complete rearrange-ment of atoms took place and the large aggregates could be formed and ejected. However the analysis of copper - gold alloys led the same authors84 to conclude that “plasma reactions are at most a minor cause of the appearance of polyatomic ions.Molecular ions can be produced by volatilis-ation of neutral complexes followed by the ionisation in plasma or through sputtering processes.” It would appear from these studies that no single process can account for all the multi-atomic ions observed in spectra. The predominant mechanisms in most instances however, will involve recombination processes as many of the mol-ecular species formed are chemically unstable and could not exist as such in the electrode material 1232 ANALYST OCTOBER 1984 VOL. 109 Various workers have studied the spectra of relatively pure materials to determine the ionic species which are formed and the relative abundance of each type.Metal samples studied include Cu - A1 alloys,83,85,86 Cu - Au alloys84 and W meta1.87 In a study of AgBr crystals the ratios of Br2+ AgBr+ and Ag2+ to Ag+ were related to the mean positron lifetime.88 A correction for the fraction of ions produced in the +2 and +3 ionisation states has been proposed.89 In a study of the multi-atomic species formed in the analysis of CaF2 two groups of multi-atomic ions were proposed those whose intensity was related to the bulk composition and those which were not.77 The ions produced in the spectra of various rare earth matrices have been studied and the authorgo remarked that “high-intensity molecular ions due to formation of almost any imaginable combination of individual atom with atoms of the pelleting medium are possible.” The ratio of rare earth oxide to rare earth metal varied from 0.005 (for Eu and Tm) to 0.2 (for Y and La).The existence of A102 species which is disputed on the evidence of other techniques was shown in the analysis of alumina and the relative abundance of the A1,0 species were measured.80 The multi-atomic species formed by GazO3 have also been studied.91 The formation of the B,C, and Si,C, species has been investigated by sparking electrodes of oxides mixed with graphite.81 In the analysis of frozen drops of liquid samples the ratio MO/M was measured for various elements and those for the lanthanoid elements were in the range 0.5-1.92 The analysis of SF6 gas using high-pressure SSMS with a quadrupole analyser has shown the formation of a large number of multi-atomic species.93.94 The intensities of multi-atomic ions95 and multiply charged ions96 have been measured for various elements to derive information on atomic structure.It is generally assumed that multiply charged multi-atomic ions are either not formed or are negligible. The existence of these species has been demonstrated but their intensities in most analyses will be too weak to be significant. The existence of Si32+ Si42+ and Si62+ was shown in the analysis of silicon samples97 and the ratios of CaF+ to CaF2+ 77 and Tho+ to Th02+ intensities98 have been measured as 1000 1 and 3000 1 respectively. The ion-beam profile for Cu“+ ions was relatively homogeneous in singly charged ions but became increasingly less so as n became greater.99 Electrostatic repulsion forces resulted in beam expansion the effect of which was greater for larger n .Other experiments on ion beam composition have shown the ion beam to be homogeneous over its full width, however.100 The energy distribution curves have been measured for a number of elements in standard steels and were found to be similar for all elements with the single maximum shifting slightly to lower energy with increased mlz.101 The width at half-maximum value was about 800 V for all elements. In the counter probe technique a counter electrode of a pure conducting material is sparked against the sample electrode. The contribution of the counter electrode to the ion popu-lation has been measured and found to vary considerably according to the element used and the spark gap width.102 In the frozen drop method the counter electrode contribution varied from 0.2% for W to 66% for A1.103 No deterioration of absolute sensitivity was found as the width of the counter electrode was increased,104 but a so-called “edge effect” resulted from material sputtered from the surface of the groove formed.105 2.3 Ion Production and Sensitivity A number of investigations have been made into the effect of various source parameters on ion production.Increasing pulse voltage in the range 38-49 kV did not affect results in the analysis of frozen solutions106 but caused transmission to decrease in the analysis of semiconductors.~~7 The ion intensity of element lines was relatively independent of spark voltage but the intensity of multi-atomic species decreased with increasing spark voltage the optimum operating value of which was 40 kV.108 Similarly the intensity of multi-atomic species was found to decrease markedly with increasing spark breakdown voltage.109 Changes in the discharge circuit parameters such as added capacitance have been found to change the relative populations of doubly charged ions and counter electrode ions considerably.110~111 As the distance of the electrodes from the slit increased so did the ratio of singly to doubly charged ions,112 while the ion intensity decreased steeply for distances between 2 and 6 cm, then levelled out to the optimum position at 8 mm from the slit. The relative ion populations changed significantly between 9 and 12 mm.113 The effect of changing the electrode position in the other two planes was much less marked, however. 106.113 It is generally considered that maintaining a constant spark gap is a critical factor in achieving good precision but whereas the population of doubly charged ions was found to increase significantly with the spark gap,l14 the effect on the analytical results was considered to be small.113 The shape of electrodes has also been considered for the counter probe technique and maximum transmission was obtained when the sample elec-trode was planar and the thin pointed counter electrode was angled at 45” to it.1°7 The production of CO+ C+ and O+ ions was found to increase as the source pressure increased.115 The ratio of singly to doubly charged ions was found to be dependent on the beam deflector setting but independent of element concentration.112 Some ions are formed not in the ion source but in the analyser section through a charge-transfer process Mn+ -+ Mp+. Measurement of the shift in spectra of the position of these ions from their expected position gave evidence of a privileged location for the charge exchange and suggested that a collision-induced process is questionable. 116,117 Relative sensitivitiy factors (RSFs) which are discussed in Section 3.10 are usually introduced in quantitative analysis to allow for the small but significant differences in elemental sensitivities from element to element and in different sample materials. In general they are determined from the analysis of standard materials in which the true element concentrations are known and RSFs can be considered as a correction factor, equal to (experimentally determined concentration)/(true concentration).Several workers have determined RSFs to study the effect of changing various source parameters. From the measurement of RSFs in steel standards Van Hoye and co-workers118J19 concluded that changes in spark pulse repetition rate and length and accelerating voltage do not significantly affect the precision of analysis and the same conclusion has been reached from studying the effects of photoplate emulsion sensitivity and electron multiplier sensi-tivity on RSFs.101 In the analysis of aluminium and copper, however the same authors found that RSFs did change with sparking conditions120 and in particular elements with lower melting- and boiling-points than the internal standard showed increased sensitivity with more energetic sparking.121 Gross changes in RSFs between various matrices were attributed to non-homogeneity of samples and spark effects,5’ and RSFs for some elements are stated to vary with changing spark parameters.122 More specifically the spark parameters found to have the greatest effect on RSFs were spark gap,123 spark gap and spark voltage,124 spark gap and spark position relative to the exit slit,12s the breakdown voltage and spark pulse frequency126 and breakdown voltage.65 This last parameter is only controllable in commercial instruments indirectly by changing the spark voltage and spark gap.Other uncontrollable parameters are discharge duration and power, which were also found to be critical in maintaining constant sensitivity.55 In the analysis of silver halides variations i ANALYST OCTOBER 1984. VOL. 109 1233 spark gap spark voltage and electrode position affected the RSFs for C1 and Br but not that for 1.127 It can be seen that changes in most spark parameters have been found to affect sensitivity in some samples and the only acceptable procedure is to maintain all parameters as constant as feasible during analysis. 2.4 Negative Ions The formation of negative ions in the spark has been studied comprehensively by Kishi for a large number of ele-ments.12g131 Whereas negative ions were formed from the halogens and Group VIB elements none were seen for the rare gases alkaline earths Zn Cd Hg Sc Ti Mn V Ta Re or the lanthanoid elements.Some negative multi-atomic ions were also observed. The negative ions observed in the spectra of SF6 gas include F- F2- SF5- and SFh- .93 Relative sensitivity factors have been determined for 16 negative ions.132 3. Analytical Procedures The analysis of samples by SSMS involves a sequence of procedures sample preparation electrode preparation, sparking procedure ion detection and calculation of results. Each of these steps has been studied in detail by various workers to optimise the operating procedures for different types of sample and to achieve easier more precise and more accurate analysis. General descriptions of the principles and operation of the spark source mass spectrometer have been given,13>136 but most procedures are concerned with a particular type of sample.3.1 Specific Procedures Metal samples are usually sparked directly and there have been no reported developments in the general procedures for metal analysis. A procedure for the single exposure analysis of 244Cm and 252Cf has been described in which small amounts of sample are mounted on gold electrodes.137 Steel samples have been dissolved in acid evaporated to dryness and mixed with pure graphite.138 Procedures for the analysis of rare earth matrices have been described in full both for the metals and for their compounds.9~ Methods for the analysis of other pure oxide materials have been described. 13!~141 Whereas the analysis of metals is relatively simple and the procedures have become standardised the analysis of non-conducting materials is more complex and generally requires the addition of a conducting powder to the sample.As a consequence a number of different procedures have been developed. This is especially so for geological materials to which the technique has been widely applied. A review of methods for geochemical and extraterrestrial samples has been given.142 Comprehensive descriptions of procedures have been presented by a number of workers. In the methods of Taylor and Gorton143J@ and Hintenberger145 the sample was mixed with graphite as conducting material and in that of Ni~holls13~122J~6 the sample was fused to improve homogeneity and then mixed with graphite. The counter probe technique has been used by Chupakhin and co-workers,147J4* with the probe electrode sparking against a thin disc of the sample and by Ure and Bacon,149J50 with an aluminium counter electrode sparking against an electrode pelletted from a mixture of the sample with aluminium powder as the conducting material.For the analysis of lunar samples graphite powder has been used.151 A procedure has also been described for the analysis of geochemical materials using the low-voltage arc discharge. 152 The analysis of biological materials is more difficult than that of geological samples in view of the problems of sample size sample preparation low concentration levels of trace elements and the higher concentrations of alkali and alkaline earth elements.A number of procedures for SSMS analysis of biological samples have been described however usually with special consideration of the sample preparation stage. 112.1~~-160 One procedure involved analysis of the lyophilised sample directly without ashing but it was only applicable to the determination of the elements of high mass. 161 Whereas most procedures described have used photoplate detection electrical detection systems have also been employed for the analysis of non-conducting powdered samples. 162,163 A review of the application of SSMS to environmental samples has been given by Cornides164 and the use of a single spark method has been described for the analysis of thin conducting films. 16s Gases have been analysed by SSMS. In one procedure UF6 was dissolved in water which was then mixed with conducting powder and freeze-dried.166 Arsine and monogermane gases, however have been analysed by admitting them directly into the source chamber where they were decomposed on thin electrically heated tungsten electrodes which were sub-sequently sparked. 167 3.2 Isotope Dilution Spark-source Mass Spectrometry In this technique spikes of the elements to be determined but of different isotopic composition are added to the sample and from the altered isotope ratios the concentrations of elements can be calculated. It has the advantages of being very sensitive precise accurate and not dependent on quantitative chemical pre-treatment. Although thermal ionisation is more precise SSMS can be used in this mode if the elements are involatile or have a high ionisation potential and a multi-element analysis with a minimum pre-treatment is required.Reviews of isotope dilution SSMS (ID-SSMS) have con-sidered both “wet” and “dry” techniques as well as extensions of the met hod 19339,168 In the “wet” technique the sample and spiked isotopes are allowed to equilibrate in solution. Such a technique has been used with electrodeposition of a number of elements on to gold wire cathodes for the analysis of a wide range of sample types 16% 171 Procedures have also been described for sample solution equilibration with the spike solution followed by drying on to electrode powder.172-174 The “wet” technique has limitations imposed by the impurity of acids used problems of losses during the drying phase and difficulties in complete dissolution of complex materials.A “dry” technique has been developed in which graphite powder is spiked with a solution of enriched isotopes, dried and mixed with the powdered sample.175 Although no true equilibration between sample and spike isotopes occur, the technique is a compromise between the improved pre-cision of ID-SSMS and the minimal pre-treatment required for direct SSMS. Similar procedures have been described by other workers for the analysis of geochemical samples.173J76J77 The use of electrical detection systems is particularly suitable for ID-SSMS and one study showed peak switching to be the preferred approach .I78 A double isotope dilution technique has been used to determine Cu in fish samples and sea water.179 Copper enriched in 65Cu was used as a tracer and copper enriched in 63Cu as the SSMS spike.Sulphur has been determined in steels by dissolution spiking with sulphur enriched in 34S precipitat-ing the sulphur as BaS04 and mixing with graphite.1g0 A so-called isotope dilution method has been used for the direct analysis of gases in metals.18131*2 The sample was sparked against an electrode that contained certain elements enriched isotopically. If the sparking was controlled correctly, material from both electrodes was mixed in the plasma and isotopic equilibration was achieved 1234 ANALYST OCTOBER 1984 VOL. 109 An extension of the technique used elements enriched with stable isotopes as multiple internal standards. 1x3 This was intended particularly for monoisotopic elements that cannot be determined by ID-SSMS.3.3 Sample Preparation Any sample treatment required before analysis especially of biological materials is usually described as part of the over-all method but a few studies have paid particular attention to this part of the procedure. Contamination is a problem in all trace element analyses but especially so in the analysis of biological materials. Possible sources of contamination have been investigated152 and ashing procedures discussed. 157,158,160 Oxidative acid digestion and high- and low-temperature dry ashing procedures are widely used each having its own advantages and disadvantages. A procedure has been des-cribed for the acid digestion of powdered biological material in a digestion bomb followed by rotary evaporation on to graphite.184 Methods for the homogenisation of powdered geological samples have been investigated185J86 and sampling procedures for rare earth matrices metallic and as oxide have been described.m 3.4 Pre-concentration Although a major advantage of SSMS is its ability to analyse samples with a minimum of pre-treatment for certain applications it is necessary to pre-concentrate the analyte in the sample relative to the matrix thus improving the detection limit. The consequent reduction in the concentration of the matrix elements minimises multi-atomic ion interferences produced by them. The matrix in which the elements are concentrated can often be made constant irrespective of the original sample matrix and problems of sensitivity changes with matrix are consequently reduced.The homogenity of the new matrix can also be improved. The different techniques that have been used are illustrated by the following examples. Evaporation Fused silica was dissolved in HF and evaporated to give a concentration factor of about 150.187 Co -p recip ita tion Ammonia has been used to precipitate the lanthanoid elements from fused silicate materials188 and to precipitate uranium from waters using iron as a carrier.189 A number of elements were co-precipitated from soil extracts by quinolin-8-01 tannic acid and thionalide (pH 2) using aluminium as carrier. 190 Selenium and tellurium were determined in copper standards by co-precipitation with gold (added as carrier) from solution by hypophosphorous acid.191 Co-crystallisation Metals dissolved in geothermal water samples have been co-crystallised with 1-(2-pyridylazo)-2-naphthol (PAN) reagent .I92 Ion-exchange chromatography Cation-exchange chromatography has been developed for the separation of the lanthanoid elements as a group from rock matrices193--'95 but a comparison of this technique and mixed-solvent anion-exchange chromatography for the same purpose found the latter to be a preferable t e ~ h n i q u e . l ~ ~ J ~ ~ Adsorption Transition metals in solution have been chelated by quinolin-8-01 and subsequently adsorbed on to activated carbon. 198 Cementation In this technique a solution was passed through a small column of aluminium powder and a number of elements were collected by the method of spontaneous electrochemical displacement and deposition known as cementation.190.199 This method is particularly suitable for the SSMS procedure150 using aluminium powder electrodes as no further sample preparation other than drying and pelletting is required. Electrodeposition A number of elements can be electrodeposited on gold wire cathodes and the technique has been used with ID-Solvent extraction Elements can be extracted from iodide solution with 4-methylpentan-2-one,201JQ back-extracted and evaporated on to alumina. Similarly the platinum metals have been extracted with N,N-hexamethylene-N'-phenylthiourea and concentrated on Cu0.203 The platinum metals have also been concentrated using the fire assay technique and wet chemical methods but no details were given.204 Disadvantages of pre-concentration and other pre-treatments are the added risk of contamination and the extended treatment time.This is compounded in some methods by the use of organic reagents which thereby introduces an ashing step in the procedure. Losses resulting from digestion procedures or ashing can be minimised by using the isotope dilution technique but this restricts the numbers of elements that can be determined. SSMS. 170,171,200 3.5 Solutions As SSMS requires a solid sample electrode direct analysis of solutions is not possible. In the above pre-concentration procedures evaporation coprecipitation deposition and other methods are used to transfer elements from a liquid to a solid phase.Such methods can also be used for the SSMS analysis of solutions whereby elements are concentrated in the process. A brief review of methods for analysis of solutions has been given.j7 In the simplest procedure described the solution was dried on to the conducting material used for the analysis of acidic solutions of nuclear reactor fuels.205 High-purity water and acids were analysed by evaporation to a small volume and transferring the drop to the end of silver or graphite electrodes which were sparked against silver wire counter electrodes.206 Metal chloride and nitrate solutions were analysed by applying them to the end faces of graphite electrodes that had been impregnated with a benzene solution of polythene to fix the sample in a thin surface film.2o7 An alternative procedure was to freeze the solution and spark against a counter electrode.This method has been investigated106 and applied to the analysis of Antarctic ice.92 For some samples it was necessary to deposit a thin gold film on the surface of the frozen sample to make it conducting.208 Layers of semiconductor material have been analysed by removing with acid freezing the solution and sparking against tantalum wire.209.210 The freezing method has also been used for the analysis of high-purity tin which was melted frozen rapidly and sparked against tungsten wire. 103 Aluminosilicate catalyst has been analysed by dispersing in water freezing, covering with a gold film and sparking against a gold counter electrode but the sensitivity and the precision proved inferior to those obtained by analysing a pressed powder electrode.211 3.6 Electrode Preparation The preparation of metallic samples is simple because after cleaning they can be sparked directly or if in powder form, pressed into electrodes and sparked ANALYST OCTOBER 1984 VOL.109 1235 Procedures for non-conducting materials are more numer-ous and a review of such methods has been given.212 For solid pieces an auxiliary electrode can be placed inside around or close to the sample. In the analysis of magnesium oxide gold wire was wound around the sample and sparked against a gold wire counter electrode213; pure quartz was analysed similarly with gold wire inserted into the sample214 and monocrystals have been analysed by wrapping in purified aluminium foil.8’ Alternatively the sample can be crushed or ground to a fine powder and analysed as for powdered samples a procedure that was also used for quartz samples.214 Boron fibres have been analysed by deposition on tungsten wire215 and alumina, beryllia and boron have been analysed following vacuum deposition of copper or aluminium on the sample.2’6 In general there are two techniques for the analysis of non-conducting powders.In the first the sample is made conducting by mixing with a conducting powder and pressed into two electrodes usually cylindrical which are sparked together. The most commonly used conducting material is graphite but aluminium silver and gold powders have also found applications. The choice of conducting powder can be dictated by the application.Several investigations have been made into the choice of conducting material. Gold powder was chosen because analysis was more sensitive than with graphite and drops of solution diffused deeper into the electrode,217 because it presents fewer interference problems than graphite alumi-nium or silver197 and because the sample was not diluted atomically so much as with graphite.218 In particular applica-tions such as the determination of silver as its sulphide gold powder has also been used.219 Silver wire was used as the cathode in a pre-concentration procedure using electrodeposi-tion.171JmJ20 Silver powder was chosen instead of gold or graphite for the analysis of uranium hexafluoride because of impurities in the gold,166 and was preferred to graphite because of faster and more sensitive analysis even though the silver gave problems of inhomogeneity.221 Other examples of the use of silver powder include the analysis of calcium fluoride77 and of objets d’arts.222 In the analysis of cinnabar, gold was consumed too quickly and detection limits and blanks were lower for graphite than for silver.223 Similarly graphite was chosen instead of gold or silver for the analysis of fibre-optic glass because the analysis was easier and the cost lower.224 It is for these reasons and the much higher purity of graphite that most workers use graphite as the Conducting material except in special situations in which interference effects and other considerations dictate otherwise.The addition of In and Re as internal standards to graphite by a slurry technique has been described.225 Other materials that have been used are copper,226 which gives a cleaner spectrum than graphite under more energetic sparking conditions and aluminium,150 which is used in a hybrid technique with an aluminium wire counter electrode.The second technique for non-conducting powdered materials is the counter probe technique which can be used for analysing thin sample layers without addition of conduct-ing materials and for localised micro-probe analysis. This technique has been developed by Chupakhin and co-workers,5’,211,227-230 who have used a number of different materials for the counter electrode including rhenium ,227 aluminium,228.229 niobium51 and tantalum.51.211,230 The choice of counter electrode depends on the application and need not necessarily be that which gives the least contribution to the ions formed.In microprobe analysis for example the depth resolution was improved if less sample as distinct from counter probe material was consumed and for this application aluminium was used as :his contributed more than the other 11 materials tested .228 For most applications. however the counter electrode consumption needs to be kept to a minimum and for this reason tantalum is widely used. The counter electrode should also be the cathode with a unipolar discharge to minimise its consumption.126 Tantalum has also been used for the analysis of alumina beryllia and glass,214 of ionic crystals (in preference to tungsten)231 and of semiconductor materials (in preference to aluminium).110 For the analysis of high-purity tin by the frozen drop method however tungsten counter electrodes were chosen in preference to aluminium and tantalum.103 Pure gold counter electrodes have been used as a probe for the localised analysis of vanadium metal232 and silver counter electrodes for the analysis of high-purity water and acids deposited on silver or graphite powder.206 3.7 Sparking Procedures From the investigations of the effect of changing spark and other conditions on sensitivity ion production precision and other parameters it is possible to suggest optimum operating conditions. These conditions will depend largely on the type of sample being analysed and the purpose of the analysis.A fundamental principle must be that precision will be improved if all source conditions are held as constant as possible,233 but the sensitivity will only be maximum under certain defined conditions. Spark gap width Several studies have found changes in sensitivity some considerable with changes in spark gap width. 114,123,125,132.234.23s A constant gap width is therefore desirable. At narrow gaps the ion energy distributions were broader236 and the maximum shifted to higher energies,l32 but ion intensities were higher for wider g a ~ s . 2 3 ~ As the gap was increased the production of multiply charged ions and of multi-atomic ions increa~ed.6338~.132 The electrode tempera-ture increased with increasing gap width and this could affect the analysis of volatile elements especially mercury.238 Generally however gap widths of 50 pm,233 40-50 pm239 and 0.034.1 mm56 have been recommended to improve precision, to reduce the counter electrode contribution and to optimise the discharge energy respectively.Electrode - slit distance As the electrode - slit distance was increased the resolution increased,9”132 the absolute intensities decrea~ed,7~3~~ the relative intensities or sensitivity ~hanged99.125.132.2~0 and the ion energy distribution became more asymmetric with the maximum shifting to a slightly higher value.236 This parameter should therefore be kept constant expecially if external standards are used when both sample and reference elec-trodes must be sparked at the same distance from the slit.112Js6 Further with an increased electrode - slit distance the consequent beam expansion resulted in a reduced number of multiply charged ions being sampled99 and also the intensity of multi-atomic ions decreased.xsJ32 Electrode position Although generally not thought to be an important parameter, it has been shown that off-axis sparking affected accuracy and so electrodes should be accurately aligned in front of the extraction slit .I14 Electrode size and shape These can affect the analytical sensitivityQ5J41 and should be chosen so that masking of the slit is not significant.114 Planar electrodes resulted in increased ion transmission.107 Electrode vibration In addition to continually changing the spark gap this also led to transfer of electrode material and the production of surface irregularities and should if possible be av0ided.11~ Extraction hole diameter As this was reduced the beam intensity was also redu~ed.7 1236 ANALYST OCTOBER 1984 VOL.109 Spark duty cycle Whereas some studies showed that pulse length and repetition frequency have little or no effect on analy~is,l~~~ll9.13'.2~5.~~2 others showed an effect. As the duty cycle increased the sparking became more violent and hotter and could affect the intensities of some elements considerably,122.237 especially for volatile elements238 and those with boiling- and melting-points considerably lower than those of the matrix.120 The pulse width and repetition frequency had no effect on the ion energy distribution.236 These parameters should be kept constant during analysis and generally the recommended values are low238 with for example a pulse length of 50 ps.243 For metal analysis the precision was improved if the repetition frequency was fixed at 1000 or 3000 Hz,242 and for rock analysis maximum sensitivity was obtained with a duty cycle of 3% .Z41 Spark voltage There is evidence that this does not affect the analytical precision and sensitivityl18,242,243 except at values greater than 40 kV.It is suggested therefore that the optimum spark voltage should be the lowest that will maintain regular sparking.243 However a change in accuracy was found for oxygen in iron and steel237 and definite changes in relative intensities have also been observed.24" Multi-atomic ion intensities increased with increasing spark voltage.85 The ion energy distribution changed,236 but was dependent on whether the electrode at the accelerating voltage acted as anode or cathode the energy spread being narrower in the former instance .74 The transmission efficiency dropped with increased voltage.107 Breakdown voltage This is a function of spark voltage gap width source pressure and electrode shape. For accurate analysis it should be kept constant within fairly narrow limits during all stages of a spark discharge .*44 Discharge energy If this is regulated by placing an additional capacitance in the electrode circuit,230 the analytical precision is improved.59q5s Optimum values for the discharge are a total energy flux of 1W-10~ W cm-7 discharge duration 15-100 ns and inter-electrode distance 0.03-0.1 mm.57 Accelerating voltage In the analysis of metals this was found to have no11932 or only a possiblellx effect on precision.For the AEI MS702 instrument the width of spectral peaks (at half-height) have a minimum at an accelerating voltage of 20.3 kV so a value of 20 kV is recommended.223 Using the gold probe technique €or analysis of high-purity compounds an accelerating voltage of 24 kV consumed far less material than one of 16 kV produced darker lines on the photoplate and produced a higher number of multiply charged ions.245 Source pressure Pressures greater than 5 X 10-7 Torr have been found to affect the analysis of carbon nitrogen and oxygen in metals.239 In the low-voltage discharge source high currents and low graphite concentration resulted in transfer of material which caused the spark discharge to become erratic and finally to be extinguished.152 In the analysis of thin layers by the counter probe technique shallow craters are usually required. The best counter electrode to achieve this was aluminium229 and the spark energy should be reduced by reducing gap width and modifying the spark circuit.58 The depth of the layer removed increased with increasing breakdown voltage and surface roughness but decreased with a wider counter electrode .246 There is some disagreement over the ideal counter electrode dimensions ranging from a point with zero thickness,l'7 to 3.4-4.5 mm wide x 0.05-0.08 mm thick,lIO 3.5 mm wide246 and up to 12 mm wide.104 In a study of silicon no clear relationship has been found between exposure and sparking area thickness of layer removed and relative consumptions of sample and counter electrode.247 It is clear that the choice of counter electrode shape is dictated by the lateral and depth resolution required.Ion beam choppers improve the analytical precision by increasing the consumption of material for short exposures,242 but can also be used to accept only ions produced during required parts of the spark discharge cycle. The arc discharge stage can be excluded by accepting only the initial part of the breakdown5O151 or the arc stage alone may be accepted for 30-90 s of the discharge.60 Both were claimed to increase the analytical precision. To reduce the intensity of the multiply charged ions the electrodes can be cooled.87 For the analysis of thin layers of metal on cylindrical rods the rods should be rotated during sparking to increase the layer volume available for consump-tion.248 3.8 Photoplate Evaluation The photoplate is still the most widely used means of detection and some studies have been made to improve photoplate development. The most commonly used photoplate Ilford Q2 has been compared with Kodak SWR and Ionomet JM plates.249 The Ionomet and Kodak plates have considerably higher sensitivity than Ilford Q2 plates but the Kodak plates have a reduced dynamic range. The Ilford Q2 plates have, however a much lower background level and better line definition. The response of different types of ions has been studied.25O Monatomic ions produce darker lines than multi-atomic ions of the same mass and energy and for ions of the same mass the more compact ions produce darker images.The relationships between ion density ion energy ion mass and the line area have been investigated.251 The use of an internal developer gives more intense lines on the photoplate103 and the background fog hgs been reduced by using a high-contrast phenidone developer252 or a bleaching agent.IO3 The latter was not entirely successful as the line intensities were weakened and the photoplate had less contrast. Modified development conditions have been des-cribed .243 The methods used for evaluating photoplates have been reviewed.l(),132.'64,25~,254 Three photoplate calibration methods have been compared in terms of reproducibility, speed and simplicity for manual processing.255 The method of Kai and M i k P is the simplest and fastest but the precision is low whereas the method of Hull257 gives the highest precision but is tedious.A computer study of calibration methods258 found the Churchill two-line method259 and Mattauch and Ewald method260 to be equally acceptable but the latter was recommended. New or modified photoplate calibration methods have been proposed by various workers. In the method of Taylor261 isotope intensity ratios measured on one exposure were used to obtain an intensity versus density relationship using a Seidel function. The intensities at a chosen and fixed density reading were used to calculate element concentration.The calibration curve of Bouvy and Gauneau214 used the equation D = {[log 1 + aE(1 + abE)]/[l +aE(l + abE)lO-c]} where D is optical density E is exposure and a b and c are constants. Fergason and Young262 used a mathematical procedure to calculate the number of ions that produced a particular line and included a background correction. In the analysis of thorium oxide Childs263 used a method in which the logarithm of peak area was plotted against the logarithm o ANALYST OCTOBER 1984 VOL. 109 1237 exposure for each charge state from Th+ to Th7+. Then the peak area equivalent to a 1 nCi exposure was plotted against ionisation potential for each state and three methods of element concentration determination were used depending on the concentration level.The method used by Pearton264 was basically the Schuy and Franzen expansion265 of the Hull method which introduced a grain distribution function. Calculated relative exposure levels Q values were intro-duced by Pilate and Adams266 and their use and other factors in photoplate calibration have been discussed.267 A completely different approach the so-called parabola method has been developed by Radermacher and Beske for the low-voltage discharge source.61.2h~271 A plot of the logarithm of ion intensity against ionisation state was found to lie on a parabola that was specific for each element and was influenced by the matrix. From these a mean ion charge number (Zi) was ca!culated for each element and the matrix, and the aperture of each element parabola was altered to make the element Zi equal to that of the matrix Zi.The intensities of impurity lines were compared at the matrix Zi for concentration determination. This method requires the matrix spectral line intensities to be measured and so very short exposures must be taken. The precision of photoplate densitometry and calibration has been studied"J72 and found to improve if a correction for line width variation was niade233J73 or integrated line areas used.272~~~~ Three different methods of measuring line area have been studied.274 The evaluation of the same photoplate by eight different laboratories using well defined conditions showed a considerable range of analytical results.275 There was no advantage in using the defined conditions in place of the laboratories' own methods.Calibration graphs have been found to change close to the matrix lines but the reasons are not clear.276 A study of the Ilford Q2 photoplate response to ion mass showed that response decreases proportionally to M0.4 and not Mo.5 as is widely used.273 There is rarely any problem of the identification of elemental lines but a method has been proposed that compares standard and sample spectra on the same plate.277 Another method for assigning mass used for the spectra of organic ccmpounds used the mass and position of calibration lines over small segments of the photoplate.278 3.9 Spectral Interferences Before any line intensities can be used in concentration calculations they must be seen to be clear of superimposed interferences or corrected using other lines present in the spectrum.The problems of overlapping element lines are usually not great as all elements (except indium) have a unique isotope. The presence of multiply charged ions is seldom a hindrance for they often fall at fractions of masses and in the spark source are not so intense as the singly charged ions. The interfering species that give rise to most problems are the multi-atomic ions which often but not always arise from the matrix element. Oxides carbides and halides are often present and the combinations possible seem endless and are not restricted to normally stable chemical species. In the spectra of high-purity materials the interferences tend to be specific on particular elements of interest whereas in complex matrices the problems of interferences are much more widespread.In many applications the occurrence of interferences is not discussed and this could account for some so-called matrix effects. One study claimed that "spectral interferences would be very slight or non-existent using mass spectrometry for the analysis of a complex mixture such as coal and coal ash."279 However the reports in which interferences have been identified or discussed are listed in Table 1 and these indicate that interferences do indeed exist for all types of matrices. Table 1. Reports in which interferences have been identified and/or discussed Matrix General . . . . . . . . Metals . . . . . . . . . . Pure materials: Pure chemicals and crystals Semiconductors . . . . Rareearthmatrices .. . . Pure acids . . . . . . . . Azides . . . . . . . . Glasses . . . . . . . . Rocks . . . . . . . . Silicate matrices: Coal . . . . . . . . . . Platinurnmetalores . . . . Environmental materials: Airparticulates . . . . . . Water . . . . . . . . Sewagesludge . . . . . . General . . . . . . . . Plants . . . . . . . . Animaltissue . . . . . . Humantissue . . . . . . Negativeions . . . . . . Biological materials: References 13,23,122,170,274,280-282 98,103,120,135,191,283-287 77,89,108,141,166,223,288,289 1 15,290-292 90 293 294 97,113,187.214,224,295-297 132,143,144,149,150,174,175, 39,160 204 185,196,197,199.241.298-303 164.304 305 288 9,157 45,158,288 45,154,158 159,306.307 130,131 Although they are rarely a major problem and "spectral complications must not be over emphasised ,"lo2 they should never be ignored or underestimated and chemical pre-concentration may be necessary to remove them.15oJ87J903199 There is a danger however that chemical pre-treatment may introduce new interferences as in the dissolution of rock samples.174 The complexity of multi-atomic species formed is best illustrated by the study of complex materials such as rocks1"J and biological materials. 45.158.159.2X8 Combinations of calcium and .aluminium with oxygen or carbon seem to form particularly easily and can spread beyond the high-mass elements. 143,144,288 The lanthanoid elements are of consider-able interest particularly in geochemical samples and the interferences in this region have been well stud-oxide and carbides to complex species such as CaA10C?,,1j3 and BaCx0,,Nz.z94 Mathematical procedures have been pro-posed for correcting these interferences both on individual lanthanoid element~'~3~175~298 and on the group as a whole.l49~,"()2 A computer program has been used to identify multi-atomic species of the general formula AkB[Cn,Dn where A is the conducting matrix B the cation C the complex former ( e .g . Si) and D the anionVzx2 ~~~Y~J,14~,1~4.14'~,~7~,1~h.1'17.2~4,29X.300.302 and range from simple 3.10 Sensitivity For quantitative analysis the calculated results are usually corrected using sensitivity factors that are determined by the analysis of standards with the same or similar matrix as the sample and in which the element concentrations are known.Values for the correction factors have been presented by numerous workers and comparisons have been made between these studies. Such comparisons are difficult however as the definitions of the correction factors are not standardised and details of the exact means by which they have been calculated are often not reported. Given this even the terminology is not consistent as two terms (relative sensitivity factor and relative sensitivity coefficent) are frequently used and appear at times to be interchangeable. The relative sensitivity factor (RSF) can be defined as RSF = fsftfp wheref%,f andf, are individual factors for the source transmission and photoplate relative sensitivities,308 i.e. the RSF includes all the discriminatio 1238 ANALYST OCTOBER 1984 VOL.109 effects involved and if no allowance for these effects has been made in the calculations the correction factor is the RSF. The relative sensitivity coefficient (RSC or S,) should be the termf only i.e. the relative ease of formation of ions but is generally accepted to be fq ft so that if photoplate or detector discrimination effects are included in the calcula-tions the correction factor is the RSC.1jS In some instances attempts have been made to allow for discrimination in ion formation before the calculation of RSCs. 150 The basic formula for sensitivity factors as used by most workers is (measured concentration)/(true concentra-tion),gg.3(@ but in a few instances the inverse of this relation-ship has been used.l5*.2().5.’9h.31() When an internal standard is used in the calculation the RSF or RSC is defined as [(CJC,) measuredl(C,lC,,) trueIz where C and C, are the concentra-tions of the unknown element and internal standard respec-tively and t refers to the matrix.Another equation used for the calculation of senstivity factors is E I A 1 C i = C s . - .- . - . -E I A RSC‘ where C is concentration E is the exposure needed for a defined blackening I is the isotopic abundance A is the relative atomic mass and the subscripts s and i refer to the internal standard and unknown element ~ respectively. 1 1 ~ 1 5 ~ This equation has been modified in some studies by raising the mass term to the power 0.8 and making corrections for line width,312 by leaving this term out of the relationship139.2vs or by replacing it with isotope intensity ratios.313 Graphical methods have also been used to determine RSFs or RSCs.86.264 Reports that give calculated sensitivity coefficients are listed in Table 2 and include comprehensive studies of sensitivity ranging from one element in one matrix to a large number of elements in several matrices. The matrix does not always refer to the type of sample being analysed. as it can be altered in the sample preparation stage as for example with water samples dried on to a graphite powder. The variation of sensitivity factors with matrix is still a matter of debate and the evidence presented is contradictory. One major study concluded “that the RSC of an individual element does not change significantly from matrix to mat-In the analysis of the platinum group metals.the RSFs did not vary from one platinum metal to another.j‘h.3’8 Other studies, riX,”132.3?4 and this is supported by other studies 172.30‘).33h.3-17 Table 2. Reported sensitivity coefficicnts Matrix References Detailedstudies . . . . . . 101,118,119. 132.286.30s. Metals . . . . . . . . . . 59.65.84.86.87. 101. 11&120,123. 3 13-3 18 132,138,139.152.174,196. 221. 226.233,235.237.264.285.286. 308,310.311.314.316.317, 3 19-33 1 Pure materials: Graphite . . . . 123,152.194.196.205.207.221. Chemicals . . . . . 55,90,127,139,188.202.217,223, 272,314,318,332-335 288,289,308.309.312,3 336-340 Semiconductors . . . . 115,341-344 Glass . . . . . . . 224,296.308 Frozen solutions .. . . 55.92.339 Rocks and related materials . . 17 132,143.145 148 150. Biological materials . . . . 112 114. 132. 153 154 158 Negative ions . . . . . . 129,132.350 298,314,324.345-357 307,314.324.348.349 3.315. 88,203. 161. however found that “for most elements relative sensitivity factors differ significantly with the type of matrix””‘“ or that there were large changes with matrix.5g.13y It has been suggested that if all the charged states of the measured element are used then the RSFs would approach ~ n i t y . 3 3 ~ The absolute sensitivities of the individual elements in any one matrix are different and attempts have been made to correlate these differences with physical and chemical proper-ties of the elements. Whereas two studies have found no dependence of RSCs on the chemical forms of the ele-ment 132.325 another found large differences dependent on whether the elements were in the oxide or fluoride The chemical form could have an influence in certain samples, for example in the analysis of thorium in rock matrices which, it was suggested was dependent on the titanium content and the decomposition of ThTi06.34h Most studies show that the RSCs are independent of element concentration I 1% 1 7 0 3 7 2 2 ’ 2 8 f i but some changes in RCSs have been found at low concentrations.”() The vaporisation terms that were best correlated with RSCs were the melting-point,2-75 boiling-point 143.31 1 vapour pressure,s5.3()X.32() heat of sublimation139 and both melting-point and heat of sublimation.316 Similarly, the correlation of RCSs with ionisation terms has been found to be best for ionisation p0tential”~123,1”.2Z”X or ionisation c r o s s - s e c t i o n ~ ~ ~ ~ 7 2 ~ 3 ~ ~ ~ ~ ~ ~ 2 whereas one study concluded that neither was satisfactory.92 Certain equations based on these physical properties have been proposed for calculating theoretical sensitivity factors and comparisons have been made of experimental and theoretical factors to determine the most satisfactory equa-tion.Several workers132.315,317.32~ have found closest agree-ment with the equation of Goshgarian and Jensen3”: where CR is the covalent radius AH is the heat of sublimation at 298 K and‘ IP is the first ionisation potential. Some agreement has also been found’32.315 with the equation of Socha and Masumoto352: where 8 is the temperature at which the vapour pressure is 10-8 Torr and c is an empirical constant.Other equations have been proposed by Billon-33? RSC,tY = and by Itoh and Sata”7: 4. Instrumentation 4.1 General All the commercial SSMS instruments are now of dated design and only the JEOL instrument is still available. Consequently, a number of improvements in design have been suggested for improving the precision and ease of analysis and to implement improved analytical procedures. Descriptions have been given of a JEOL instrument that had been modified with an interchangeable Knudsen cell source,353 of the Thomson -CSF TSN 212 instrumentlo and of the AEI MS702R instru-ment and the operating procedure for it.135 Because of their age it is becoming necessary to upgrade the older instruments.An improved cold-trap filling system334 and new solid-state power supplies354.355 have been designed for the AEI MS7/70 ANALYST OCTOBER 1984 VOL. 109 1239 instrument and a complete upgrading of a CEC 21-110 instrument included new electronics pumping systems source unit and focusing system.-356 A totally different instrument using a time-of-flight mass analyser was designed for the analysis of micro-particles but the instrument was only applicable to this specific purpose and had a very poor resolution.357 Another specialised application was the high-pressure analysis of gases using a spark-source instrument combined with a quadrupole analyser.93.244 There is a need for the development of new simpler instruments or as stated in one report for “improved r.f.spark ion sources combined with relatively uncomplicated rationally working and inex-pensive analysers.”3”9 4.2 Ion Source The spark source unit has been redesigned to make it interchangeable with other types of ion source such as the Knudsen ce11,353 ion microprobe source ,358-3(50 hollow-cathode gas discharge source361J62 and the laser source.363 An AEI MS7 instrument has been redesigned with a low-voltage d.c. arc discharge source.364 For the specialised application to the analysis of radioactive materials the source unit must be redesigned for remote handling of samples by means of a glove-box.365 For y-emit-ting materials all loading of samples sparking and even repairs must be carried out remotely with the aid of closed-circuit television ,366 In some applications such as the determination of gases in semiconductor materials the ion source pressure must be considerably reduced most commonly by fitting a cryogenic pump using either liquid nitrogen367 or liquid helium.l l s ~ ~ 1 ~ 6 5 ~ ~ ~ ~ ~ 6 Y To remove hydrogen from the system it is necessary to retain the diffusion pump341 or cover some of the cooling fins with activated carbon.369 A cryosorption pump has been designed using activated charcoal cooled with liquid nitrogen and fitted with a heater to desorb the gases.370 The combination of a mechanical cryopump and high-speed diffusion pump has been fitted to a modified source hous-ing,37lJ72 and stainless-steel grids were added at a later date to reduce the desorption of gases from the source ~aIls.373.37~ A source unit has been modified to accommodate a 1200 1 s-l oil diffusion pump cold-trap baffle and a gate ~alve.3~5 The time required for sample loading limits the possibility of reducing the analysis time.A probe-type sample changer has been used for the analysis of a large number of saxples for one element mercury.376 An alternative approach has been to design multi-sample holders for two pairs of electrodes,114-377 six samples103 or up to twelve samples.245 A system for rotating vertically mounted cylindrical elec-trodes has been described for improving analytical pre-cision.378 Automatic scanning systems havz been designed for the analysis of semiconductor surfaces using a counter electrode traversing over a rotating disc379 and for layer-by-layer analysis using a probe scanning in two mutually perpendicular directions.380 Two methods have been used for the analysis of gases.In one the gas is heated and introduced into the spark gap through a small longitudinal hole drilled through one of the metal electrodes.381JX2 In the other method the source is filled with the gas under pressure and sparked between two tungsten electrodes .93 Various modifications have been made to the spark circuit electronics to improve sparking procedures. The output stages of the r.f. generator have been modified by the addition of a high-voltage kenotron and capacitor to make the discharge unipolar.48,126,231,38~38S With the same aim modifications have been described that allow only the portion of the beam that was formed with one electrode acting either as cathode or anode to be transmitted.386JX7 Using both systems the electrodes connected to the accelerating voltage should act as anode.The second system has been used with a spark circuit modified by the addition of resistors to give a self-triggered damped di~charge.73~388-389 Other modifications to the spark circuit have been used to control the spark discharge. Cut-off capacitors in the dis-charge circuit reduce the duration of the spark dischargel48J3~ and a new spark circuit discharge generator has been designed although no details were given to control the power and duration.57 Ballast elements have been added to the circuit to reduce the discharge intensity and to decrease the sampling depth in layer analysis,110,384 as well as to extend the lifetime of frozen drops.209 Ballast capacitance has also been added to reduce the effects of stray capacitances.11’ An alternative modification provided single pulses for the analysis of thin films.165 The energy distribution of the ion beam has been controlled with a new method of commutation of the electrodes,”) for which no details were given and by an ion-beam compression system .391 Another device used to stabilise the matrix line intensities irrespective of spark gap width has been reported but again without details.111 Ion-beam chopping devices have been designed to improve precision by consuming more material than is actually detected or by accepting the ion beam from only a part of the total discharge.The devices can be synchronous in which only a pre-selected time interval from each pulse is accep-ted48.50,60,383,392,393 or asynchronous in which only a certain number of whole pulses are accepted.392 Alternatively the portions of the discharges accepted were totally rand0m.s~),3~~ Devices have been designed to control the spark gap width in order to improve precision. The AEI Autospark unit394 has been modified to make it possible to select the mean electrode gap by controlling the r.f. voltage developed between the electrode~.114.2”~395 In a similar device a constant peak r.f. voltage was maintained.396 Circuits have been given for measuring the breakdown voltage and current385J97 and this breakdown voltage has been used to keep the spark gap constant.1~9,244 The electrode position can also be controlled by keeping the ion-illumination angle constant. A device has been designed- for keeping this and the gap width constant by intercepting the ion beam with two plates on either side of the beam and measuring the ratio of the signals.39x.399 A system for the self-vibration of electrodes used the attractive force experienced by the electrodes during the breakdown and the resilience of an elastic holder.247 4.3 Analyser All the commercial instruments have a double-focusing analyser with Mattauch - Herzog geometry and such is the soundness of the design that there has been very little modification to the analyser section.Tirne-0f-flight93~~~)~) and quadrupoleg3 analysers have been used in special applications but the resolution was very poor. The quadrupole analyser has been investigated for SSMS and if the energy band width could be reduced the combination might be ~iable.4()1-~02 New magnet power supplies have been designed for the AEI MS702.243 Electrical detection requires accurate control of the magnet and systems have been described for the fitting of a Hall probe403 and a temperature compensation network for a Hall probe.404 The alignment of the ion beam on the optical axis276 and determination of the position of the image p1ane4O5 have been described. An internal shield of nickel-coated soft iron has been placed between the source and analyser to reduce the fringe magnetic field.406 Secondary electrons ejected from the ion collector by impinging ions cause non-reproducibility of the monitor reading and the ion collector design has been modified to reduce this.407 A thin metallic shield has been fastened to the photoplate cassette in the position of the matrix lines to reduce background fogging caused by secondary emission but wit 1240 ANALYST OCTOBER 1984.VOL. 109 the loss of some analysis lines.103.408,4"9 The size of the &-slit has been reduced by welding thin strips of tantalum across the slit in order to increase the resolving power.410 4.4 Ion Detectors Although still widely used there has been very little develop-ment of the photographic plate detector. A simple method has been described to check that the plates lie in the focal plane.411 A comparison of three detection systems found peak scanning to be the fastest and peak switching to be the most precise but that photoplates still had the advantage of simultaneous detection.218 Most development work has been on electrical detection systems.The system developed for the JEOL instrument has been described412 and used to determine the precision and accuracy of the method for the analysis of steels.119 In the peak scanning mode precision was 18% whereas in the peak switching mode it was 10%. The peak scanning mode has been digitised to allow the data to be stored on magnetic tape and a signal averager used to sum the spectra.413 One electrical detection system has been designed using peak scanning with a slow scanning speed.321 Changes to the AEI electrical detection system included a new digital integrator system a tapered collector slit modified amplifier circuits and the use of magnetic peak switching with a Hall probe in place of electrostatic peak switching.114 A multi-channel analyser has been used to record repetitive multiple scans in the scanning mode and to ~ignal-average.3"~3775.414 Peak switching systems have been designed with either electrostatic peak switching240 or magnetic peak switching.349 The latter is considered more precise but takes longer. A digital ratio circuit for the peak switching mode has been described.396 Automation of the electrical detection system has been described using an autoprogrammer that reads instructions from a paper tape415; the electrostatic peak switching mode was used.On-line computer-controlled electrical detection systems using both scanning and magnetic peak switching modes have been designed for the Nuclidel62.416 and JEOL instruments.3") Automation of the AEI system in the peak scanning417 and peak switching418 modes has also been described. The computers control the mass spectrometer by setting the magnet current and electrostatic analyser voltage, and simultaneously acquire and reduce data. A circuit has been described for the automatic attenuation of the signals to an electrometer so that very weak and very strong ion currents can be measured.419 For the measurement of very small ion currents a detector has been used that monitors the rate at which charge accumulates across a calibrated capacitor .420 The major disadvantage of these electrical systems is that they are single-channel systems and the ion currents are measured sequentially and not simultaneously.A simul-taneous ion beam collection has been described for isotope ratio work in which a Faraday collector with a slit was placed in the focal plane.421 One isotope passed through the slit to a separate detector while the remaining isotopes fall on the Faraday collector. A multiple collector system each movable relative to the others has been described but no applications of the technique have been recorded.422 Attempts have been made to develop new electron - optical ion detector systems. In one system a channel electron multiplier array (CEMA) was used with three optical readout systems to assess their viability.423-425 Twelve CEMAs would be required to replace a photographic plate and there would be considerable problems with the data readout.Resolution was poor and the response was non-linear with the high peak ion currents driving the CEMA into saturation. It was concluded that the CEMA was not suitable for a pulsed source and that the optical detectors were not sensitive enough. An alternative concept used three channel plates in a chevron array and self-scanning silicon photodiode arra~s.426~427 Applications of this design have not been reported and there would be problems of cost and dynamic range. The need for the development of new simultaneous electrical detection systems is still present and the possible approaches that have been suggested28 are channel electron multiplier charge-coupled devices electrostatic build-up on insulated plates and Fourier transform mass spectrometry.4.5 Data Processing Manual processing of data for the photoplate is slow and tedious and much effort has been made to automate to various degrees the densitometry and calculation of results. This is highlighted by the number of computer-controlled systems (Table 3) that have been reported for the aquisition and subsequent processing of data from photoplates. An early detailed description of an on-line computer-controlled mic-rodensitometer and data processing system also critically assessed methods of calibrating and reading photographic plates.434 Different systems vary in their sophistication from small interactive microcomputers to the large totally auto-matic systems that only require initial setting up of the densitometer and the input of basic data.There are two basic design concepts that differ as to whether the computer is instructed which lines to measure or whether it detects all peaks and subsequently identifies them. The latter system needs much larger data storage. Other differences lie in the definition of peaks the measurement of peak height width or area the photoplate calibration method used and the degree of automation of the densitometer. The trend is towards a two-step operation with a small computer used on-line for data acquisition and data reduction ,and then a larger computer, either time-shared or off-line used for data evaluation.Some systems use one dedicated computer for both steps. Such systems are however more expensive and often permit few operator decisions and these disadvantages have dictated the design of a simpler system with no on-line computer.439 The problems of interferences in spectra do not always seem to be considered but in one system the program identified interfer-ing multi-atomic species.*** The modification required to automate a manually con-trolled densitometer have been described,446.449 including in one report a system for maintaining the photographic plates in the focal plane.44'1 Programs for the on-line computer control of a microphotometer have been given451 and another system punches reduced data from photographic plates on to paper tapes .452 Simpler generally earlier systems have been described in which data from photographic plates are fed to the computer through the keyboard,261 punched cards453 or punched tapes.20.435 In one system the densitometer is lined up manually on peaks to be measured and the digital output can be fed directly to the computer.4s4 Desk-top computers455 and programmable calculators with printers456 have also been used for the data processing routines.Systems have also been described for the processing of data from the electrical detection 5. Applications The power of SSMS is demonstrated by the very wide range of sample types for which analyses have been reported. These are but the tip of the iceberg for the number of unreported Table 3.Reports giving details of computer systems for data acquisition and processing from photographic plates References 20,61 264 266 268 269 274 277 282 428-45 ANALYST. OCTOBER 1984 . VOL . 109 1241 analyses will be much greater . The analyses range from the determination of the concentration of one element to a complete survey of the whole Periodic Table with detection limits given for those elements not detected . 5.1 Metals The analysis of metals (Table 4) is generally for quality control. to check that products are up to specification. that purification methods are achieving their object and that the concentrations of certain elements are not above their critical level . The procedures are usually straightforward as metal sam-ples can be sparked directly without pre.treatment .Some metals. however. do require special methods if they are liquid. for example sodium,4**.489 or melt at a low temperature. for example tin.103 In the latter instance the frozen drop technique can be used for analysis.103.408.470 The main problem with metals is that internal standards cannot be introduced into the sample and usually a matrix isotope is used for this purpose . Ion implantation has been used to prepare standards.329.461.511 One method of introducing an internal standard has been to spark a sample electrode with a reference electrode containing a known amount of internal standard and allowing the internal standard to distribute over the surface of both electrodes.460 5.2 Non-metallic High-purity Materials This group of compounds (Table 5 ) is considered here to include non-metallic elements and simple compounds.a large number of which are semiconductor materials . As these materials are non-conducting the analysis often requires mixing of the sample with a conducting material such as graphite. the purity of which must also be determined . These analyses. as for those of metals. are generally made in order to ascertain whether impurity element concentrations are below the tolerated or specified levels . Recent reviews on the analysis of semiconductor materials have underlined the importance of these impurity levels.478-518 The electrical properties of 28 GaAs crystals manufactured by eight different companies by two techniques have been correlated with their trace element ~oncentration.52~ A similar study related free electron or free hole concentrations to dopant concentrations in epitaxial layer and bulk crystal GaAs.528 The trace element concentrations are critical in materials that are used in other applications .Germanium single crystals are used as radiation detectors and the concentrations of Cu. Ni and 0 are critical.517 Doped AgCl crystals are used as detectors of nuclear particles and the purity of synthetic crystals prepared by different methods has been deter-mined.lO8.537 Fused silica used in the manufacture of optical waveguide glasses187 and semiconductor-grade silica used in the manufacture of solar cells3443522 have been studied . Lead chalcogenide lasers have been applied to gas spectroscopy and pollution monitoring and their performance depends largely on the quality of the PbS substrate crysta1.546 Other examples of analysis for which quality control is important are mono-crystals (e.g.PbMo04") . oxide insulators ( e . g . MgO and A1203213) and pigments (e.g. . The materials analysed are not necessarily in the form of solids . Germanium and arsenic hydrides (gases) have been analysed following decomposition in the ion source on to tantalum wire electrodes.167 Liquid samples can be analysed by conversion into solid compounds. 338 by freezing208 or by evaporation on to metal wires2933540 or po~ders.203~206~544 The purity of waters and acids has been shown to be dependent on the materials from which beakers and other containers are manufactured.2O6The contamination that results from grinding graphite in different containers has also been investigated .157 ~~ Table 4 . Reported analyses of metal samples Metal General . . Aluminium . . Barium . . Bismuth . . Cadmium . . Calcium . . Cobalt . . . . Copper . . Iron . . . . Gallium . . Gold . . . . Hafnium . . Indium . . Iridium . . Lead . . . . Molybdenum Nickel . . . . Niobium . . Platinum . . Ruthenium . . Silver . . . . Sodium . . Tantalum . . Thorium . . Tin . . . . Titanium . . Tungsten . . Uranium . . Vanadium . . Yttrium . . Zinc . . . . Zirconium . . Lanthanoids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transuranium elements Alloys: Steels . . . . . . Copperalloys . . . . Nickel alloys . . . . Zirconiumalloys . . Otheralloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References 233.45 9-461 10.89.120.135.287.317.329.367. 466 408 408 466 181 51.58.135.178.191.201.239.243. 286.310.322.330.331.377.418.462.467. 468 461-465 118.119.181.237.239.321. 469 292.47 0-472 89.224.239.284.295.296.47 3-477 478 292.408. 479 480 243.248.285.292.334. 481 181.370.482. 483 135.191.239.484. 485 478 . 486 284. 487 135 224.295. 327 488 . 489 453.462.478. 486 98. 490 11.103.321. 487 28.181. 491 87.371. 372 292. 32 1.474.479. 492 232.478.493. 494 495 292 276.478. 496 90.283.495.49 7-499 137.217.365.366.500- 505 32.114.125.138.173.174.182.202. 219.220.237.239.243.316.320. 321.349.377.482.496.50 6-508 83.85.222.239.286.418.462. 482 320.334.509. 510 396.496. 511 22.384.487.495.499. 506 Table 5 . Reported analyses of non-metallic high-purity materials Graphite. carbon . . . . . . 157.224.295.296.308.51 2-514 Boron . . . .. . . . . . 215.216.515. 516 Semiconductor materials: Material References Antimony . . . . . . 292 Arsenic . . . . . . . . 495 Germanium . . . . . . 10.115.290.292.341.379.380.478. Silicon . . . . . . . . 104.115.209.210.290.322. 541.51 7-519 341-344.370.379.380.391.462. 478.479.52 0-522 Tellurium . . . . . . . . 384. 523 Gallium arsenide . . . . 104.210.246.290.291.342.368. 370.379.380.463.478.518.520. 524-528c Others . . . . . . . . 115.245.290.291.369.490.528a. 528d Simple compounds: Oxides . . . . . . . . 90.91.139-141.187.213.216.218. 226.263.296.313.337.340.487. 529-536 Halides . . . . . . . . 77.88.108.127.208.231.338. Lanthanoid compounds . . 67.90.218.226.499.510.535.539. Pure water and acids . . . . 202.206.293. 544 Others . .. . . . . . 10.58.89.167.294.490.545. 546 537-541 542. 54 1242 ANALYST OCTOBER 1984 VOL. 109 5.3 Other Manufactured Materials This group (Table 6) is taken to include more complex manufactured materials not included in Sections 5.1 and 5.2, and also some specific applications that include samples from all three groups. Quality control is important in the manufac-ture of special glasses such as those used in fibre optics.187,296,549 Attenuation of the signal along the fibre is critically dependent on the concentration of certain trace elements. Such analyses have not always been s u c c e ~ s f u 1 ~ ~ ~ however probably owing to inhomogeneous distribution of the elements and an alternative approach has been to use a pre-concentration procedure that gives a more homogeneous matrix and allows the determination of elements at the parts per billion level.187 Some special glasses are treated with lanthanoid compounds whose concentration levels are impor-t an t.297 Trace levels of certain elements cause discoloration in TiO2 pigments and SSMS is an ideal technique for the analysis of such insoluble and refractory materials.531 A number of facial cosmetics have been analysed and the elements Pb As Se and Cr were found to be above acceptable concentrations in some of them.557 Catalysts used in a coal liquefaction process were found to be contaminated by trace elements associated with the coa1.558 A brief review of the analysis of radioactive materials has been given.562 Such materials associated with nuclear reac-tors include 233U02 used to make fuel,'40 acid solutions of irradiated uranium - plutonium fuel rods,205-366 uranium and plutonium oxides,530,536 insoluble residue left from fission dross (mainly silicon)561 and waste sludges and supernatants from the processing of fission products.5~~~~6() The analysis of glasses for forensic purposes has been discussed46.~~3,555 and generally involves correlating two glass samples one from the scene of the crime and one from the suspect.An over-all elemental analysis can differentiate glasses that are indistinguishable by physical methods. The types of glass studied include vehicle headlamp and auxiliary lamp glass,548 window glas~550-552,5~~ and container gla~s.55~ Similar studies have been made of bullet lead285-481 and copper wires.468 Human liver samples have been analysed to assess the feasibility of using such analyses in forensic toxicology.307 Archaeological samples have been analysed to identify their origin but such analyses have achieved only limited success.A number of pigments used in Indian paintings have been identified,222 but it was not possible to distinguish between the red and grey shards used in Iranian ceramics.556 Analysis of copper artifacts from North America showed differences in samples from the same site and so was of little use,467 whereas the analysis of Peruvian copper artifacts showed changes in manufacturing methods in different eras.3'0 5.4 Environmental Materials With the growing concern for the quality of the environment there has arisen a need for multi-element analysis of environ-mental samples ranging from air particulates and waters to sewage sludges.Analyses that have used SSMS are listed in Table 7. A general scheme for the assessment of environmental samples especially in relation to pollution included SSMS analysis for 73 elements.56"5@ Air pollution analyses have been made of urban air samples,*3,188,327,~6~~~66 laboratory and external air157 and of samples taken within or in the vicinity of various industrial plants (see Table 7). Fly ashes are generally produced by coal-fired generating stations and by municipal in~inerators.5~3 The analyses of respirable dust and coal dust of respirable size were particularly comprehensive.574 Air-borne samples were usually collected on filters made of nitro~ellulose23~~~5 (which requires ashing) silver mem-brane3g7 (which can be sparked directly) glass-fibre304 or millipore filters.160,578 Wastes studied include the bottom ash from coal burning plant~56~V573 and from refuse incinerators,573 chars and tars from coal conversion plants,579 sludges from nuclear power plants559.5@ and sewage sludges. l9",288,580 Industrial effluents from condensers scrubbers and chillers and liquid wastes from processing plant~577.58~,597 have been studied. Most water samples have been analysed by evaporation on to the electrode powder which is usually graphite but freeze-drying,587 filtration on to Ag powder,582 cementation on to A1 powder,l99 collection on activated charcoal,l98 electrochem-ical deposition ,171 ion exchange594 and solvent extrac-ti0n192~5~4 have all been used.5.5 Fuels The analysis of fuels (Table 8) has usually been for the purpose of monitoring possible pollution effects arising from the use of such fuels. Coal samples are usually ashed before analysis but it has been shown that it is possible to obtain a comprehensive analysis with unashed samples.574-598,600 Petrol could be a major source of environmental Pb S Cr Ni Cd Mn and v.*5 Table 6. Reported analysis of other materials and specific applications Material Glass . . . . . . . . . . Ceramics . . . . . . . . Pigments . . . . . . . . Cosmetics . . . . . . . . Catalysts . . . . . . . . Applications: Radioactive materials . . Forensicsamples . . . . Archaeological samples .References 46,113,187,188,216,224,295-297, 487,534,547-555 337,529,556 222,531 557 211,558 140,205,217,365,366,490,530, 46,113,285,307,468,481,548,550, 222,310,467,556 533,536,559-562 552-555 ~~ Table 7. Reported analyses of environmental materials Material General . . . . . . . . Air particulates: Fresh air . . . . . . . . Fly ash . . . . . . . . Coalminedust . . . . . . Other industrial plants . . Wastesandsludges . . . Waters: General . . . . . . . . Purifiedwaters . . . . . . Taplintake waters . . . . Lake and river waters . . Geothermal waters . . . . Seawaters . . . . . . Industrialeffluents . . . . References 563,564 23,157,164,188,304,327,565,566 22,171,188,279,567-573 22,574,575 22,160,171,347,571,576-579 190,288,559,560,567,573, 579-581 189,196,582,583 175,333,335 157,198,199,333,573,584 23,198,199,305,332,333,577, 585-590 192,586,590-592 544.593-597 171,573,577,579,331,584,596, 597 Table 8.Reported analyses of fuels Material References Coal . . . . . . . . . . 22,160,170,171,183,279,558,566. 567,570,571.573-576,579, 598-604 Fuel oil . . . . . . . . 567,570 Petrol . . . . . . . . . . 172,335,570.605 Refuse . . . . . . . . 573 Nuclear fuels . . . . . . 140,205.60 ANALYST OCTOBER 1984 VOL. 109 1243 5.6 Geological and Related Materials This is the largest group of complex matrices that have been analysed by SSMS (Table 9). For geochemical research the technique has proved itself valuable for giving over-all multi-element analyses and is probably the only technique that can determine the whole lanthanoid group at trace levels, simultaneously.Such analyses allow conclusions to be made on the history of samples and consequently on the formation of the geo-sphere.Lanthanoid elements are immobile in most metamor-phic conditions but mobility has been associated with uranium mineralisation.634 There was little relative fractionation of the lanthanoid elements during the sedimentary process but the lanthanoid element pattern changed from Archean to post-Archean rocks.303.636 Sharp increases in thorium and uranium were also associated with the Archean - Proterozoic boun-dary.637 The analysis of lanthanoid elements has been used to discuss the origin of plutonic rocks from Nova Scotia.614 Investigations into the origin of volcanic lavas and rocks have, in a number of instances used SSMS analyses for lanthanoids and other trace elements.617.6lX.h2~633.64(~644 Samples from the Oklo fossil reactor have been analysed to study migration of uranium fission products.622-625 Considering the large number of analyses of rock samples by SSMS it is perhaps surprising that relatively few analyses of similar soil samples have been reported. Whereas rocks are of geochemical interest soils are also the first link in the food chain. Trace element deficiencies occurring in plants or animals can be correlated with low soil contents or low availability to plants and toxicity effects can occur where soil trace element contents are high. There has been only one comprehensive study of trace element contents in soil using SSMS,150-1997656 but the technique has also been used to determine the lanthanoid contents of soils.613,654 The analysis of soil extracts indicates the availability of some trace elements to plants.150J90 Soil extracts have also been analysed to detect the presence of bromine- and fluorine-containing herbi-ci de s .657 In the analysis of meteorites the wide element coverage of SSMS is particularly useful and a large number of samples have been analysed either comprehensively or for a number of trace elements.Similarly materials returned to earth by the Luna 16 and 20 and Apollo 11 12,14,15 and 16 moon landing missions have been analysed. Comprehensive analyses can be made with the relatively small amount of material available.Analysis of Apollo 11 samples showed that organogenic elements are present in small amounts677 and that the trace element composition was different from that of Apollo 12 samples.679 The results for Apollo 12 samples suggested that the moon interior was heterogeneous on a small scale that lunar material was heated to high temperatures before accretion and that tektites were not of lunar origin.681 5.7 Biological Materials Of all the groups of materials perhaps the greatest analytical challenge is given by samples of biological origin (Table 10). The matrices can be complex and widely varying from sample to sample. The relatively high levels of alkali and alkaline earth metals halogens and phosphorus give rise to a large number of complex interfering species that can make analysis difficult.The concentrations of most trace elements are much lower than in for example rock matrices and coupled with the necessity for ashing and possible losses are much more sensitive to problems of blank levels and contamination. The problems of contamination157 and sample preparation688 have been discussed in detail. The analyses of plant materials range from the determina-tion of one element for example boron in radish leaves by isotope dilution,69’ to comprehensive analyses for example of normal and deficient wheat.697 Lanthanoid element contents have been determined in water lily,hsX lichens and m0sses,1~~,694 cIubmoss,6Y3 horsetail,6” dwarf shrubs and trees694 and ferns6s4 and compared with those for the soil or rocks on which the plants grew.The contents of other trace elements in lichens and mosses348 and spruce samples690 have also been reported. Other plants analysed include r i ~ e ~ 5 ~ @ 8 apple691 and pear696 samples. Usually photoplate detection has been used for the analysis of biological samples but electrical detection in the scanning mode has been used for some plant analyses.692-693 The analyses of marine samples have generally been for the detection of water pollution and include the determination of mercury in fish meal700 and copper in plaice.179 Multi-element analyses have been reported for musse1,23.7()1 oyste1-54~ and trout158 samples. Mammalian samples analysed include liver ,108,154,158.170 kidney 158 whole blood ,702 serum 108 and urine.108 A detailed study has been made of the yttrium and lanthanoid element contents of various organs from rats fed on different diets.154 Heavy metals can in some instances be determined in unashed samples by using a high sparking voltage to minimise the production of organic molecular ions above mlz 192.108 The accumulation of Cd Pb and Zn by earthworms has also been studied.655 Analyses of human materials are numerous and most internal organs and body fluids have been analysed. A particularly comprehensive survey by Hamilton et al. 688 gives contents for 28-53 elements in 10 different organs and blood. Most analyses attempted to correlate trace element contents with specific diseases. In a study on alcoholism the yttrium and lanthanoid element contents were determined in a large number of various internal organs from alcoholics and non-alcoholics.154 Grossly contaminated miner’s lung has been compared with normal tissue22J60 and hilar lymph nodes have also been analysed in a large survey.’s9 Patients with advanced carcinoma were found to have increased copper ~~ ~~ Table 9. Reported analyses of geological and related materials Material Terrestrial : Standardrocks . . . . Rocksamples . . . . Minerals . . . . . Soils . . . . . . Soilextracts . . . . Lake and river sediments Meteorites . . . . Extra-terrestrial: Lunarmaterials . . References . . 18.50 122. 143 144 148. 150. 152, 199,212,299,302,303,330.346. 163,173,177,178,188,193-197. 433,607-616 . . 18. 163,2O3.230,300,345,346.604, 608,611.614,616-644 .57,77.160,204,212,214,223,226, 230,289,310,346,358,359,477, 604,610,619,626,627,629,638, 645-653 654-656 . . 149,150,188,190,199,302,613, . . 150,190,657 . . 23,585,658 . . 23,142,144,145,173,174.176,185, . . 57,132,147,148,151,185,241,261, 186,241,299,659-676 299,473,661,677-687 Table 10. Reported analyses of biological materials Material References General . . . . . . . . 45,157,158,688,689 Plants . . . . . . . . . . 153,158 169 170,184,288,348, Marine . . . . . . . . 23,158,179,544,700,701 Animal . . . . . . . . 108 114 154,158,170,655,702 Human . . . . . . . . 22,23,112,154-156,159,160,306, Other samples . . . . . . 160,183,712-715 377 451. 654 658 69&699 307,314,688,703-71 1244 ANALYST OCTOBER 1984 VOL.109 content and lower zinc content in their blood serum.23 Skin samples have been analy~ed709,~l~ but no significant difference could be detected between mycetoma grains and normal dermis.710 Hair and fingernail samples have been analysed for forensic applications~~2~~55~156~~~~~~05 and accumulations of Hg , As and Pb in hair and As in fingernails have been found.704 Similarly the use of liver analysis for forensic toxicology has been investigated.307 A large number of dental enamel samples from 17 American States have been analysed to establish the normal range for trace element concentra-t i o n ~ . ~ ~ ~ The presence of sulphur in the biochemical Slow Reacting Substance was first indicated by SSMS.7” Other samples analysed include diet160 and food sam-p l e ~ ~ ~ ~ raw and refined sugars,712,713 honey714 and the enzyme Subtilisin carlsberg.715 5.8 Miscellaneous Materials Other analyses have been of wear-metals in lubricating 0ils,23 polyheavy water,716 wide explosives72Y4 some polymers and plasti~s,~~0~717 bromine- and fluorine-containing herbicides extracted from soil657 and Trizma base used in subtilisin enzyme assay.715 Gaseous materials analysed by SSMS include uranium hexafluoride166.381.718 and silicon hexafluoride .939 Although principally an elemental analytical technique, SSMS has been used to yield structural information from the fragmentation pattern of non-volatile metal carboxylates .75376 5.9 Standard Reference Materials A large number of certified reference materials (CRMs) have been analysed using SSMS and the list of reported analyses (Table 11) contains the more common ones for which analyses have been reported by two or more SSMS laboratories.Table 11. Reported analyses of selected standard reference materials USGS: BCR-1 AGV- 1 w-1 . . G-1 . , G-2 . . GSP-1 PCC-1 Japanese: JB-1 . . JG-1 NIM: D . . G . . L . . N . . P . . s . . NBS: Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SRM 1632 (coal) SRM 1633 (fly ash) . . . . . . . . . . . . . . . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . SRM 1571 (orchard leaves) SRM 1577 (bovine liver) . . SRM 444 (stainless steel) . . SRM 461 (low-alloy steel) . . SRM685w(gold) . . . . SRM1213(zircalloy) . . . . References 18,143,144,148,149,152,194,195, 197,212,241,299,302,330, 610-615,628 148,150,194,195,212,616 122,145,148,152,163,173,176, 193,194,212,433,608,611,615, 672 145,163,212,616 150,152,188,194,195,199,212, 302,608 188,194,195,212,612 194,195,212 196,197,612,616 196,199,612 609,615 194196,609,615 609,615 196,609,615 609,615 609,615 170,171,183,279,566,571,602 171,188,279,566,571 158,169,183,184,332,377,658, 695 158,170 173,377 243,349,506 239,284 276,396 6. Concluding Remarks From the number of reported analyses that have been listed in this review and from the fact that some laboratories are investing in the updating of power supplies vacuum systems, detectors etc.it is evident that the technique is still attractive for certain applications. This is especially so when a multi-element survey analysis is required for which the accuracy of the technique is sufficient (generally 15-3070 with calibration or within an order of magnitude without). If SSMS is to be used for analysis its cost and long analysis time have to be justified. Comparisons of the technique with various other instrumental techniques have been sample types. Factors considered include the number of detectable elements whether analysis is simultaneous detec-tion limits accuracy precision analysis time matrix effects, charge-up and field problems resolution (depth and lateral), beam-induced chemical changes and on-stream capability.The general conclusion can be summarised by those given in a review of twelve instrumental techniques used for water analyses721; SSMS is the most specific and the most com-prehensive multi-element technique but the accuracy about 30% is very poor. Similarly in the analysis of sea water SSMS had the best sensitivity and provided multi-element data but was also expensive and sl0w.5~~ SSMS is considered to be a “sensitive survey method” for the analysis of surfaces and thin films.521 The strengths of SSMS are its multi-element nature, reasonably uniform sensitivity and its high absolute sensitiv-ity.Its detection limit has been compared with those for a number of other technique~.9.~”~5.l34,199,5~,580.723 In a com-parison of nine techniques SSMS detects the greatest number of elements to the lowest detection limit,9 and for solid samples it gives the lowest detection limit.28,lW It is superior to neutron activation analysis in that the detection limits by SSMS are constant for all elements.45-544 Two other types of ion source have been found to have the same advantages as the spark source but fewer of the disadvantages. These are the laser source724726 and the hollow-cathode ion source ,361,482 but neither has yet found wide applications. Recent commercial developments have included the coupling of an inductively coupled plasma source with a quadrupole mass spectrometer for the rapid analysis of solutions727 and the use of a d.c.plasma discharge source for the analysis of solid materials.728 The expense of and limited demand for SSMS instruments will probably preclude the development of a new generation of instruments and even simpler instruments would still not be cheap. It is more likely that users of SSMS instruments will modify and update present instruments to improve perfor-mance. Probably fewer than 200 instruments have been produced and the number still in use is decreasing each year. The technique is at the point where it will fall into decline unless instrumental developments become significant enough for increased use to be considered. madel8.309.4Y1.521,566,592,595,h8Y.719-722 for the whole range of 1.2. 3. 4. 5. 6. 7. 7. References Dempster A. J. Proc. Am. Philos. SOC. 1935,75,755. Ahearn A. J. Editor “Trace Analysis by Mass Spec-trometry,” Academic Press New York 1972. Deines P. Pa. State Univ. Coll. Earth Miner. Sci. Exp. Sta. Circ. No. 78 1970. Honig R. E. in Ogata K. and Hayakawa T. Editors, “Recent Developments in Mass Spectroscopy ,” University Park Press Baltimore 1970 pp. 116-149. Elser R. C. in Winefordner J. D. Editor “Trace Analysis: Spectroscopic Methods for Elements,” Wiley New York, 1976 pp. 383-417. Liebl H. Mikrochim. Actu 1978 1 241. Demortier G. Rev. Quest. Sci. 1979 150 509 ANALYST OCTOBER 1984 VOL. 109 1245 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47, 48. 49. Stefani R. Adv. Mass Spectrom. 1978 7A 729. Morrison G. H. in Ogata K. and Hayakawa T. Editors. “Recent Developments in Mass Spectroscopy,” University Park Press Baltimore 1970 pp. 158-168. Vastel J . and Valeriano P. Rev. Tech. Thomson - CSF (France) 1971 3 489. Bukowski J . A, and Patryas Z . Pr. h’auk. Inst. Chem. Nieorg. Metal. Pierwiastkow Rzadkich Politech. Wroclaw., 1973 18 29. Morrison G. H. Bull. SOC. Chim. Belg. 1975 84 581. Nicholls G. D. and Wood M. in Nicol A. Editor, “Physiochemical Methods of Mineral Analysis,” Plenum New York 1975 pp. 195-229. Cuna C Ioanoviciu D. and Cuna S . Stud. Cercet.Fiz., 1976 28 481. Wang T. S. Fen Hsi Hua Hsueh 1978 6 387. Gauneau. M. Spectra 2000 1979 7 No. 53 55. Dietze H. J. ZtF. Mitt. 1980. 28 39. Beske H. E. Gijbels R. Hurrle A. and Jochum K. P., Fresenius Z. Anal. Chem. 1981 309 329. Carter J . A. Franklin J. C. and Donohue D. L. ACS Symp. Ser. 1978 No. 70 299. Cornu A. Proc. 2nd Con$ Mass Spectrom. tspra Italy Sept. 1971 Euratom Luxembourg 1972 pp. 317-341. Rigby L. J. Report No. IR/R445/1971/155 Standard Tele-communication Laboratories Harlow Essex 1971 17 pp. Brown R. Jacobs M. L. and Taylor H. E. Am. Lab. 1972, 4 29. Bingham R. A. and Vossen P. G . T. Lab. Pract. 1975,24, 233. “L’Analyse Elementaire des Solides par Spectrometrie de Masse,” collection of articles Spectra 2000 1979.7 No. 5 1,29; No.53 45; No. 54 61. Stefani. R. Trends Anal. Chem. 1981 1 84. Bacon J. R. and Ure A. M. Anal. Proc. 1981 18 356. Adams F. Philos. Trans. R. SOC. London Ser. A 1982,305, 509. Ure A. M. Trends Anal. Chem. 1982 1 314. Cornides I. and Gal T. Kem. Kozl. 1976. 46 41. Ramendik G. I. Priroda (Moscow) 1977 26. Derzhiev V. I. Liebich V. Mai H. and Ramendik G. I., Wiss. Ber. Akad. Wiss. DDR Zentralinst. Festkorperphys. Werkstofforsch. 1979 17 259. du Bois E. A and Wrigley K. B . Proc. Chem. Conf. B t h , British Steel Corp. Sheffield 1972 pp. 97-102. Wrigley K. B. and Dubois E. A. Report No. MG/CC/569/ 72 British Steel Corp. London 1972 13 pp. Tunney A. A. Report No. GS/TECH/479/1/75/C British Steel Corp. Sheffield. 1976 28 pp. Herzog L. F. Int. J. Mass Spectrom.ton Phys. 1970,4,253. Mapper D. and Smales A. A. in Wainerdi R. E. Editor, “Analytical Chemistry in Space,” Pergamon Press Oxford 1970 pp. 209-275. Cornides I. Miner. Slovaca 1973 5 489. Carter J. A. Walker R. L. and Sites J. R. in Babu S. P., Editor “Trace elements in Fuel,” American Chemical Society, Washington DC 1975 pp. 74-83. Carter J . A. Donohue D. L. Franklin J. C. and Walker, R. L. in Karr Jr. C. Editor “Analytical Methods for Coal and Coal Products,” Volume 1 Academic Press New York 1978, pp. 403-420. Roboz J . in Ciaccio L. L. Editor “Water and Water Pollution Handbook,’’ Volume 4 Marcel Dekker New York, Alford A. Biomed. Mass Spectrom. 1975 2 229. Alford A. Biomed. Mass Spectrom. 1977 4 1. Alford A. Biomed. Mass Specrrom. 1978 5 259.Lehmann W. D. and Schulten H.-R. Agnew. Chem. Int. Ed. Engl. 1978 17 222. Morrison G. M. in “Elemental Analysis of Biological Materials,” Technical Reports Series No. 197 International Atomic Energy Agency Vienna 1980 Chapter 10. Butterworth A. German B. Morgans D. and Scaplehorn, A. Int. Microform J . Legal Med. 1972 8 card 5 10 pp. Leyder F. Spectra2000 1979 7 No. 53,64. Chupakhin M. S . Ramendik G. I. Derzhiev V. I. Tatsii, Yu. G. and Potapov M. A. Sov. Phys. Dokl. 1973,18,430. Ramendik G. I. Derzhiev V. I . Gushchin V. N. and Chupakhin M. S . Sov. Phys. Dokl 1975 20 289. 1973 pp. 1617-1661. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76.77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88 * 89. 90. 91. 92. Ramendik G. I . Chupakhin M. S . and Potapov M. A. J. Anal. Chem. USSR 1975,30,564. Ramendik G. I. Derzhiev V. I. Chupakhin M. S. Liebich, V. and Mai H. J. Anal. Chem. USSR 1975 30 1609. Ramendik G. I. and Derzhiev V. I. J . Anal. Chem. USSR, 1978 32 1197. Ramendik G. I . and Derzhiev V. I. J. Anal. Chem. USSR, 1978 32 1204. Derzhiev V. I. and Ramendik G . I. Sou. Phys. Tech. Phys., 1978 23 186. Ramendik G. I. and Derzhiev V. I . J. Anal. Chem. USSR, 1979 34 647. Derzhiev V. I. Liebich V. Mai H. and Ramendik G. I., Wiss. Ber. Akad. Wiss. DDR Zentralinst. Festkorperphys. Werkstofforsch. 1979 17 265. Ramendik G . I. Adv. Mass Spectrom. 1980 8 408. Derzhiev V.I Ramendik G. I. Liebich V. and Mai H., lnt. J. Mass Spectrom. lon. Phys. 1980 32 345. Franklin J. C. Thesis Tennessee University Knoxville TN, 1971 121 pp. Franklin J . C. and Dean J. A. Anal. Lett. 1971,4 857. Radermacher L. and Beske H. E. Spectrochim. Acta Part B 1979 34 105. Demortier G. and Lamberts L. Znt. J. Mass Spectrom. Ion Phys. 1979,31 135. Demortier G. Lamberts L. and Mathieu G. Adv. Mass Spectrom. 1980 8B 1934. Demortier G. Spectra2000 1980 8 No. 55 50. Shelpakova I. R. Saprykin A. I. Gerasimov V. A. and Yudelevich I. G. J . Anal. Chem. USSR 1980,35 413. Gerasimov V. A. Shelpakova I. R. and Rudaya N. S . J. Anal. Chem. USSR 1979,34 13. Muheim J. T. Proc. Rare Earth Res. Conf. loth Volume 1, National Technical Information Service Washington DC, 1973 pp.208-217. Muheim J. T. Helv. Phys. Acta 1976 49 739. Muheim J. T. Rare Earths Mod. Sci. Technol. 1980 2 429. Muheim J. T. Spectra2000 1980 8 No. 55 53. Muheim J . T. Spectra2000 1980.8. No. 55 54. Joyes P. Leleyter M. and Hoareau A. J . Phys. (Paris) Lett. 1979 40 131. Berthod J. Cornu A. and Stefani R. C.R. Acad. Sci. Ser. B . 1977; 285 195. Berthod J. Andreani A. M and Stefani R. Int. J . Mass Spectrom. Ion Phys. 1978 27 305. Hass J . R. Thesis University of North Carolina Chapel Hill, NC 1972. 222 pp. Buck R. P. and Hass J . R. Anal. Chem. 1973.45 2208. Luck J. Moller. P . and Szacki W. Int. J. Muss Spectrom. Ion Phys. 1974 13 25. Stefani R. J . Phys. (Paris) Colloq. 1977 19. Rechsteiner C. E. Youngless T. L. Bursey M. M.and Buck R. P., Int. J . Mass Spectrom. Ion Phys. 1978 28 401. Cornides I. and Gal T. High Temp. Sci. 1978 10 171. Cornides I . JEOL News 1980 16A 20. Ramendik G. I . Kryuchkova 0. I. Derzhiev V. I . , Stroganov N. S . and Strel’nikova E. B . Dokl. Phys. Chem., 1979 245 336. Mathieu G. and Lamberts L. Int. J. Mass Spectrom. Zon Phys. 1979,31 125. Mathieu G. Lamberts L. Thomas M. and Demortier G., Int. J. Mass Spectrom. Ion Phys. 1981 38 35. Reding A. M. Labar C. Mathieu G. and Lamberts L. lnt. J . Mass Spectrom. Ion Phys. 1978 26 421. Mathieu G. and Lamberts L. Bull. SOC. Chim. Belg. 1979, 88. 21. Itoh N. and Sata T. Report No. 6 Res. Lab. Eng. Mater., Tokyo Institute of Technology Tokyo 1981 pp. 131-138. Muheim J. T. and Surbeck H. Helv. Phys. Acta 1975,48, 446.Lasne C. Spectra2000 1979 7 No. 53 59. Conzemius R. J . in Gschneidner K. A. Jr. and Eyring L., Editors “Handbook on the Physics and Chemistry of Rare Earths,” North-Holland Amsterdam 1979 Chapter 37C. Cornides I. and Gal. T. High Temp. Sci. 1981 14 71. Chupakhin M. S. Kryuchkova 0. I. and Sulimova S. I . J. Anal. Chem. USSR 1974 28 1504 1246 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. Frees L. C. Christophorou L. G. Ellis H. W. and Sauers, I. Abstr. 28th Annu. Conf. Mass Spectrom. Allied Topics New York May 1980 pp.197-198. Frees L. C. Sauers I. Christophorou L. G. and Ellis, H. W. Abstr. 29thAnnu. Conf. Mass Spectrom. Allied Topics, Minneapolis May 1981 p. 145. Muheim J. T. Muter. Res. Bull. 1972 7 1417. Muheim J. T. Helv. Phys. Acta 1977 50 584. Vidal G. Int. J . Mass Spectrom. Ion Phys. 1972173 10 204. Vidal G. Galmard P. and Lanusse P. Int. J. Mass Spectrom. Ion Phys. 1969,2,405. Wadlin W. H. Thesis University of Virginia Charlottesville, VA 1969 110 pp. Boulin Y. Spectra2000 1980 8 No. 55 44. Van Hoye E. Gijbels R. and Adams F. Anal. Chim. Acta, 1980 115 239. Bedrinov V. P . Ukrainskii Yu. M. and Chupakhin M. S . J . Anal. Chem. USSR 1972 27 1729. Shabanova L. N. Shelpakova I. R. and Yudelivich I. G J . Anal. Chem. USSR 1978 33 885. Bedrinov V.P. Belousov V. I. and Khrornova N. V. J . Anal. Chem. USSR 1974,28,2132. Bedrinov V. P. and Belousov V. I. J . Anal. Chem. USSR, 1976 31 1701. Chupakhin M. S . Kryuchkova 0. I. and Semenov A. D. J. Anal. Chem. USSR. 1972,27 1724. Chupakhin M. S . and Ramendik G. I. J. Anal. Chem. USSR 1972,26 1241. Fitchett A. W. Thesis University of North Carolina Chapel Hill NC 1973 137 pp. Colby B. N. and Morrison G. H. Anal. Chem. 1972 44, 1263. Gerasimov V. A. Saprykin A. I. Shelpakova I. R. and Yudelevich I. G. J . Anal. Chem. USSR 1978 33 998. Pustovit A. N. and Sikharulidze G. G. Mikrochim. Acta, 1981,11 219. Clemena G. G. Thesis University of Virginia Charlottes-ville VA 1971 165 pp. German B. and Scaplehorn A. W. J. Forensic Sci. SOC., 1972 12 367.Magee C. W. Thesis University of Virginia Charlottesville, VA 1973,256 pp. Clegg J. B. and Millett E. J . Philips Tech. Rev. 1974 34, 344. Stefani R. Adv. Mass Specfrom. 1980. 8B. 1859. Stefani R. Inr. J. Mass Speclrom. Ion Phys. 1980. 34. 3 17. Van Hoye E. Gijbels K. and Adams F. Talunta 1076 23. 369. Van Hoye E Adams F. and Gijbels R . . Tulanta 1976. 23, 789. Van Hoye E. Adams F. and Gijbels R Talanta. 1978 25, 73. Van Hoye E . Adams. F. and Gijbels R. Znt. J. Mass Spectrom. Ion Phys. 1979 30 75. Nicholls G. D. Proc. Br. Ceram. Soc. 1970 April 85. Magee C. W. and Harrison W. W. Anal. Chem. 1973 45, 852. Oda S . Aihara H. and Kamada H. Bunseki Kagaku 1973, 22 757. Yanagihara K . Sato S. Oda S . and Kamada H . Anal. Chim. Acta 1978 98 307.Ramendik G. I. Chupakhin M. S. Tatsii Yu. G. and Derzhiev V. I. J. Anal. Chem. USSR 1974 29 202. Aoki M. Nishiyama Y. and Aoyama T. Abstr. Pittsburgh Conf. Anal. Chem. Appl. Spectrosc. Cleveland OH 3-7 March 1975 No. 172. Makita T. Kishi H and Kodera K. Shitsuryo Bunseki, 1973 21,293. Kishi H. Bull. Chem. SOC. Jpn. 1981 54 703. Kishi H. Bull. Chem. SOC. Jpn. 1981 54. 1999. Kishi H. Bull. Chem. SOC. Jpn. 1981 54 2005. Jaworski J. F. Thesis Cornell University Ithaca NY 1974, 123 pp. Chupakhin M. S . Ocherki Sovrem. Geokhim. Anal. Khim., 1972 591. Morrison G. H. Ann. N . Y . Acad. Sci. 1972 199 162. Pearton D. C. G. and Sobiecki A. Report. No. 1984, National Institute for Metallurgy Randburg South Africa, 1978 22 pp. 136. 137. 138. 139.140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. ANALYST OCTOBER 1984 VOL. 109 Beske H. E. Hurrle A. and Jochum K. P. Fresenius Z . Anal. Chem. 1981 309 258. Chastagner P. and Tiffany B. Int. J. Mass Spectrom. Ion Phys. 1972 9 325. Saito M. and Sudo E . Trans. Natl. Res. Inst. Met. 1977 19, 178. Kai J. Ogata Y. Ohi K. and Watanabe M. in Ogata K., and Hayakawa T. Editors “Recent Developments in Mass Spectroscopy,” University Park Press Baltimore 1970 pp. 314-318. Franklin J. C. Landau L. Donohue D. L. and Carter J. A. Anal. Lett. 1978 11 347. Sobiecki A. Report No. 2014 National Institute for Metal-lurgy Randburg South Africa.1979 7 pp. Jochum K. P. and Seufert M. Geol. Runhch. 1980,69,997. Taylor S. R. and Gorton M. P. Geochim. Cosmochim. Acta, 1977 41,1375. Taylor S. R. in Gschneidner K. A. Jr. and Eyring L., Editors “Handbook on the Physics and Chemistry of Rare Earths,” North-Holland Amsterdam 1979 pp. 359-376. Hintenberger H. Fortschr. Mineral. 1977 54 141. Nicholls G. D. Inst. Min. Metall. Trans. Sect. B 1971. 80, 299. Chupakhin M. S. and Polyakov A. L. J . Anal. Chem. USSR, 1972 27,460. Ramendik G. I. J. Anal. Chem. USSR 1977 32 1578. Ure A. M. and Bacon J. R. Proc. Anal. Diw. Chem. SOC., 1976 13 124. IJre A. M. and Bacon J. R . Analyst 1978 103 807. Dietze H. J . Freyer K. Geisler M. Haase. G. Hartmann, G.Maul E. and Panse H. Gerlands Beitr. Geophys. 1976. Jackson P. F. S . and Strasheim A . Analyst 1974. 99 26. Erametsa O. and Yliruokanen I. Suom. Kem. B 1971.44, 121. Sihvonen M.-L. Ann. Acad. Sci. Fennicae SerA2,1972,168. Harrison W. W. Ryan M. A. Copper L. D. and Clemena, G. C. in Mertz W. and Cornatzer W. E. Editors “Newer Trace Elements in Nutrition,” Marcel Dekker New York, 1971 Chapter 17. Harrison W. W. and Clemena G. G. Anal. Chem. 1972,44, 940. Hamilton E. I. Minski M. J and Cleary. J . J. Sci. Total Environ. 1972 1 1. Wong S. F. Thesis Cornell University Ithaca NY 1973, Brown. R. and Taylor H. E. Report No. PB-274206. National and Technical Information Services Washington, DC. 1975. 55 pp. Jacobs M. L. Report No. 75-186 US Department of Health, Education and Welfare Washington DC 1975 91 pp.Fitchett A. W. Buck R. P. and Mushak P. Anal. Chem., 1974 46 710. Morrison G. H. and Colby B. N . Anal. Chem. 1972 44, 1206. Nieminen K. and Yliruokanen I . Bull. Ceol. SOC. Finl. 1974, 46 167. Cornides I. in Malissa H. and Robinson J. W. Editors, “Analysis of Airborne Particles by Physical Methods,” CRC Press West Palm Beach 1978 pp. 73-91. Liebich V. and Mai H. Adv. Mass Spectrom. 1974 6 655. Journoux J. P. Teulieres P. and Boyer R. in Karr C Jr., “Proceedings of Symposium on Analytical Methods in Nuclear Fuel Cycle Vienna Austria 1971,” IAEA Vienna 1972 pp. Gaivoronskii P. E. Larin N. V. and Chernikovich M. P. J. Anal. Chem. USSR 1978,33 766. Cesario. J. Spectra 2000 1979 7 No. 54 68. Alvarez R.Anal. Chim. Acta 1974 73 33. Paulsen P. J. in Mavrodineanu R. Editor “Procedures Used at the National Bureau of Standards to Determine Selected Trace Elements in Biological and Botanical Materials,” National Bureau of Standards Washington DC. 1977 pp. 33-48. Koppenaal D. W. Lett R. G. Brown F. R. and Manahan, S. E. Anal. Chem. 1980 52 44. Franklin J. C. Carter J. A. Donohue D. L. and Stelzner, R. W. Abstr. 23rd Annu. Conf. Mass Spectrom. Allied Topics, Houston Texas May 1975 355. 85 497. 83 PP. 3 15-328 ANALYST OCTOBER 1984 VOL. 109 1247 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203.204. 205. 206. 207. 208. 209. 210. Jochum K. P. Seufert M. and Knab H.-J. Fresenius Z . Anal. Chem. 1981 309 285. Jochurn K. P. Seufert M. and Best S. Fresenius Z . Anal. Chem. 1981,309 308. Jaworski J. F. and Morrison G. H. Anal. Chem. 1975 47, 1173. Knab H.-J. and Hintenberger H. Anal. Chem. 1980 52, 390. Dietze H. J. Isotopenpraxis 1979 15 46. Dietze H. J. and Opauszky I. Isotopenpraxis 1979,15,309. Harvey B. R. Anal. Chem. 1978,50 1866. Saito M. and Sudo E. Bunseki Kagaku 1981,30 281. Glavin G. G. Kopetskii Ch. V. Nikolaev R. K. and Sikharulidze G. G. J . Less Common Met. 1972 26 1. Sikharulidze G. G. and Glavin G. G. Metody Opred. Issled. Sostoyaniya Gazov Met. Vses Konf. 3rd 1973 1 98. Donohue D. L. Carter J. A. and Franklin J.C. Anal. Lett., 1977 10 371. Verbueken A. Michiels E. and Van Grieken R. Fresenius 2. Anal. Chem. 1981 309 300. Rothenberg A. M. Thesis Cornell University Ithaca NY, 1971 101 pp. Morrison G. H. and Rothenberg A. M. Anal. Chem. 1972, 44 515. Tong S. S. C. Su Y.-S. and Williams J. P. Anal. Chim. Acta 1976 84 327. Sugimae A. Bunseki Kagaku 1978 27 169. Callis E. L. Report No. NBL-262 New Brunswick Labora-tory New Brunswick NJ 1972 pp. 52-55. Welch K. H. and Ure A. M Anal. Proc. 1980 17 8. Paulsen P. J. Dunstan L. P. and Diaz M. G. Abstr. 25th Annu. Conf. Mass Spectrom. Applied Topics Washington, DC 1977 pp. 703-704. Blommaert W. Vandelannoote R. Van’t Dack L Gijbels, R. and Van Grieken R. J . Radioanal. Chem. 1980,57,383. Strasheim A. Strelow F. E.W. and Jackson P. F. S. Proc. X V I I Colloq. Spectrosc. Int. Florence 1973 1 306. Strelow F. W. E. and Jackson P. F. S. Anal. Chem. 1974, 46 1481. Strelow F. W. E. Victor A. H. and Weinert C. H. S. W., Geostand. Newsl. 1978 2 49. Van Puymbroeck J. and Gijbels R. Bull. SOC. Chim. Relg., 1978 87 803. Van Puymbroeck J . and Gijbels R. Fresenius 2. Anal. Chem. 1981 309 312. Vanderborght B. N. and Van Grieken R. E. Talanta 1980, 27 417. Ure A. M. in “Proceedings of the VIth Czechoslovakian Spectroscopic Conference September 1-5 1980 Nitra,” Cze-choslovak Spectroscopic Society Prague 1981 pp. 61-76. Koppenaal D. W. Lett R. G. Brown F. R. and Manahan, S . E. Abstr. 30th Pittsburgh Conf. Anal. Chem. Appl. Spectrosc. Cleveland OH 5-9 March 1979 No. 437. Zolotov Yu.A. Analyst 1978 103. 56. Zolotov Yu. A. Shakhova N. V. Kryuchkova 0. I . , Gronskaya S. I. Spivakov B. Ya. Ramendik G. I. and Gushchin V. N. J . Anal. Chem. USSR 1978,33 983. Petrukhin 0. M. Zolotov Yu. A. Shevchenko V. N., Kryuchkova 0. I . Gronskaya S. I. Gushchin V. N., Ramendik G. I. Dunina V. V. and Rukhadze E. G. J . Anal. Chem. USSR 1979,34 256. Mastrangelo F. Matteucci E. and Restino G. Mem. Sci. Geol. 1979 33 205. Franklin J. C. Landau L. and Carter J. A. Radiochem. Radioanal. Lett. 1979 41 217. Mykytiuk A. Russell D. S. and Boyko V. Anal. Chem., 1976,48 1462. Gladskoi V. M. Ivanova G. A. Kuzovlev 1. A. and Kuz’min N. M. J . Anal. Chem. USSR 1971,26,975. Chupakhin M. S . and Kryuchkova 0. I. Khim. Svostva Soedin. Redkozemel. Elem.Dokl. Vses. Soveshch. Fiz. -Khim. Primen. Redkozemel. Elem. Ikh. Soedin. Splavov 6th 1969, 1973 82. Yudelevich I. G. Shelpakova I. R. Shabanova L. N., Gerasimov V. A. and Korda T. M. J . Anal. Chem. USSR, 1976 31 474. Yudelevich I. G. and Shelpakova I. P. Mikrochim. Acta, 211. 212. 213. 214. 215. 216. Chupakhin M. S. Kogan L. O. and Polyakova A. A. J. Anal. Chem. USSR 1974,29,881. Andreani A. M. Brun J. C. Mer J. P. Fillot A. and Stefani R. Methodes Phys. Anal. 1971 7 258. Nakano M. Yamaguchi G. and Saito K. J. Ceram. SOC. Jpn. 1971 79. 72. Bouvy G. and Gauneau M. Analusis 1971 1 15. Viala J. C. Thevenot F. and Duclot J. C. Analusis 1972,1, 485. Orzhonikidze K. G. Melashvili V. A. Sikharulidze G. G., and Glinskikh V. M. in Ogata K. and Hayakawa T., Editors “Recent Developments in Mass Spectroscopy,” Uni-versity Park Press Baltimore 1970 pp.362-366. 217. Carter J. A. Cameron A. E. and Dean J. A. Anal. Lett., 1971 4 337. 218. Bingham R. A. and Vossen P. G. T. Int. J . Mass Spectrom. Ion Phys. 1974 14 259. 219. Paulsen P. J. and Burke R. W. Abstr. 29th Annu. Conf. Mass Spectrom. Allied Topics Minneapolis MN I981 p. 235. 220. Paulsen P. J . Alvarez R. and Mueller C. W. Appl. Spectrosc. 1976 30 42. 221. Kannapin L. Luck J. Moller P. and Szacki W. Fresenius 2. Anal. Chem. 1974 271 1. 222. Cairns T. and Johnson B. B. in Perkins A. J. Grove E. L., Kaeble E. F. and Wetermeyer J. E. Editors “Developments in Applied Spectroscopy,” Volume 10 Plenum Press New York 1972 pp. 3-14. Gauneau M.and Bouvy G. Analusis 1973 2,314. Ikeda Y. Umayahara A. Kubota E. Aoyama T. and Watanabe E. Abstr. 20th Annu. Conf. Mass Spectrom. Allied Topics Dallas Texas 1972 pp. 142-145. Jaworski J. F. Burdo R. A. and Morrison G. H. Anal. Chem. 1974 46 805. Bulaceanu R. and Dumitrescu P. Rev. Chim. (Bucharest), 1976 27 606. Chupakhin M. S. Ramendik G. I. and Kryuchkova 0. I . J . Anal. Chem. USSR 1969 24,249. Chupakhin M. S . Ramendik G. I. and Kryuchkova 0. I. . Anal. Chem. USSR 1969,24 773. Chupakhin M. S. and Ramendik G. I. in Ogata K. and Hayakawa T. Editors “Recent Developments in Mass Spectroscopy,” University Park Press Baltimore 1970 pp. 319-322. 230. Ramendik G. I . Derzhiev V. I. and Vasyuta Yu. V. J . Anal. Chem. USSR 1979,34 1016. 231. Babushkina T.V. Sikharulidze G. G. and Nikolaev R. K., J . Anal. Chem. USSR 1975 30. 142. 232. Schmidt F. A . Conzemius R. J. Carlson. 0. N . and Svec, H. J. Anal. Chem. 1974 46. 810. 233. Konishi F. Shirsuryo Bunseki 1971 19 284. 234. Haemers J. J . Phys. E Sci. Instr. 1974 7 294. 235. Ito M. Sato S . and Yanagihara K. Anal. Chim. Acta 1980, 120 217. 236. Oda S. JEOL News 1973 IIA 15. 237. Oda S . Ohashi Z . Furuya K. and Kamada H. Talanta, 1972 19. 779. 238. Harrison W. W. and Donohue D. L. Anal. Chem. 1973,45, 1741. 239. Konishi F. Int. J . Mass Spectrom. Ion Phys. 1972,9 33. 240. Colby B. N. Thesis Cornell University Ithaca NY 1971 79 PP. 241. Kashuba A. T. Jr. Thesis Cornell University Ithaca NY, 1970 104 pp. 242. Ito M. and Yanagihara K. Denki Seiko 1972 43 171.243. Pearton D. C. and Sobiecki A. Report No. 2075 National Institute of Metallurgy Randburg South Africa 1980 15 pp. 244. Fursov V. Z. and Chupakhin M. S . J . Anal. Chem. USSR, 1975 29 1449. 245. Socha A. J. Masumoto E. M. and Willardson R. K., Report No. AD-728219 Bell and Howell Research Center, Pasadena CA 1970 96 pp. 246. Dorokhov A. N. Shelpakova I. R. and Yudelevich I. G. J . Anal. Chem. USSR 1975,30 672. 247. Gladshoi V. M. Kuznetsov Yu. N. Babenkov L. M., Belousov V. I. and Kuzovlen I. A. Ind. Lab. 1975,40,1796. 248. Vidal G. Galmard P. and Lanusse P. Int. J . Mass Spectrom. Ion Phys. 1969 2 373. 249. Masters J. I. Report No. NSF/RA-780507 Ionomet Co., 223. 224. 225. 226. 227. 228. 229. 1978 I 547. Waban MA 1978,-36 pp 1248 ANALYST OCTOBER 1984 VOL.109 250. Vouros P. Desiderio D. M. Leferink J. G. Odiorne T. J., and McCloskey J. A. Znt. J. Mass Spectrom. Zon Phys. 1972, 10 133. Grais K. I. and Grais G. S. Indian J. PureAppl. Phys. 1971, 9 127. Dale L. S. Anal. Chem. 1976 48,2040. Blain C. Spectra2000 1979 7 No. 53 47. Labar Ch. and Lamberts L. Spectra 2000 1979 7 No. 51, 46. Ordzhonikidze K. G. Samadashvili 0. A. Glinskikh Y. M., Yur’eva 0. K. and Karpenko I. K. J. Anal. Chem. USSR, 1973 27,2100. 256. Kai J. and Miki M. Mitsubishi Denki Lab. Rep. 1964 5 , 175. 257. Hull C. W. Abstr. Tenth Mass Spectrom. Confer. New Orleans 1966 p. 404. 258. Belousov V. I. Bedrinov V. P. Gladskoi V. M. and Kuznetsov Yu. N. Znd. Lab. 1974,40 535. 259. Churchill J. R. Znd.Eng. Chem. Anal. E d . 1944 16 653. 260. Mattauch J. and Ewald W. Naturwissenschuften 1943 41, 487. 261. Taylor S. R. Geochim. Cosmochim. Acta 1971 35 1187. 262. Fergason L. A. and Young R. C. Appl. Spectrosc. 1972,26, 620. 263. Childs B. G. Report No. PR-CMa-23 Chalk River Nuclear Laboratories Chalk River Ontario 1973 pp. 79-81. 264. Pearton D. C. G. Report No. 2062 National Institute for Metallurgy Randburg South Africa 1980 21 pp. 265. Schuy K. D. and Franzen J. Fresenius Z. Anal. Chem. 1967, 225,260. 266. Pilate A. and Adams F. Anal. Chim. Acta 1980 122 57. 267. ‘Verbueken A. and Van Grieken R. Bull. SOC. Chim. Belg., 1981 90 331. 268. Radermacher L. and Beske H. E. Report No. Jul-1325, Kernforschungsanlage Julich Julich 1976 82 pp. 269. Radermacher L.and Beske H. E. Report No. Jul-Spez-7, Kernforschungsanlage Julich Julich 1978 44 pp. 270. Radermacher L. and Beske H. E. Adv. Mass Spectrom., 1978 7A 545. 271. Radermacher L. and Beske H. E. Fresenius Z. Anal. Chem. 1981,309,319. 272. Labar C. and Lamberts L. Analusis 1979 7 175. 273. Konishi F. and Nakamura N. Shitsuryo Bunseki 1971 19, 213. 274. Burdo R. A. Thesis Cornell University Ithaca NY 1973, 155 pp. 275. Ailloud P. and Allegre J. Spectra 2000. 1977 5 No. 35 16. 276. Andreani A. M. Blanchard B. Brun J. C. Mermoud J. P., Pellet J. and Stefani R. Adv. Mass Spectrom. 1971 5 534. 277. Owens E. B. Rev. Sci. Znstrum. 1970 41 636. 278. Weber J. H. Earnshaw D. G. Severin D. W. K. and Decora A. W. Report No. RI 7938 National Technical Information Service Washington DC 1974 14 pp.Guidoboni R. J. in Karr C. Jr. Editor “Analytical Methods for Coal and Coal Products,” Academic Press New York, 1978 pp. 421-434. 280. Jackson P. F. S. and Strasheim A. Proc. XVZZZ Colloq. Spectrosc. Int. Grenoble 1975 1 183. 281. Fuller C. W. Proc. SOC. Anal. Chem. 1974 11 176. 282. Burstenfinder J. Luck J. and Szacki W. Fresenius 2. Anal. Chem. 1981. 309 325. 283. Jordan R. G. and Jones D. W. J. Less-Common Met. 1973, 31 125. 284. Beske H. E. Eurisotop Off. Znf. Bookl., 1974 No. 72,657. 285. Haney M. A. and Gallagher J. F. Anal. Chem. 1975,47,62. 286. Van Hoye E. Gijbels R. and Adams F. Talanta 1977,24, 625. 287. Sobiecki A. Report No. 1985 National Institute for Metal-lurgy Randburg South Africa 1978 9 pp. 288.Bacon J. R. and Ure A. M. Adv. Mass Spectrom. 1980,8, 362. 289. Luck J. Moller P. and Szacki W. Fresenius Z. Anal. Chem. 1973 267 186. 290. Andreani A. M. Brun J. C. Mkrmound J. P. and Stefani, R. Methodes Phys. Anal. 1971 7 269. 291. Gauneau M. Rupert A. Minier M. Regreny O. and Coquille R. Anal. Chim. Acta 1982 135 193. 292. Marugaiyan P. Pure Appl. Chem. 1982 54 835. 251. 252. 253. 254. 255. 279. 293. 294. 295. 296. 297. 298. 299. 300. 301. 302. 303. 304. 305. 306. 307. 308. 309. 310. 311. 312. 313. 314. 315. 316. 317. 318. 319. 320. 321. 322. 323. 324. 325 326. 327. 328. 329. 330. 331. 332. Kuehner E. C. Alvarez R. Paulsen P. J. and Murphy T. J. Anal. Chem. 1972,44 2050.Ribaudo C. Semel S. and Blosser E. R. Technical Report 4014 Picatinny Arsenal Dover NJ 1970 43 pp. Ikeda Y. and Umayahara A. JEOL News 1973. 11A 22. Gaunaeu M. Analusis 1975,3 368. Bulaceanu R. and Dumitrescu P. Rev. Chim. (Bucharest), 1976 27 707. Roaldset E. Tulanta 1970 17 593. Morrison G. H. Kashuba A. R. and Rothenberg A. M., Adv. Mass Spectrom. 1971 5 520. Nance W. B. andTaylor S. R. Geochim. Cosmochim. Actu, 1976,40 1539. Ure A. M. and Bacon J. R. Adv. Mass Spectrom. 1978,7A, 549. Bacon J . R. and Ure A. M. Anal. Chim. Ada 1979 105, 163. McLennan S. M. and Taylor S. R. Chem. Geol. 1980 29, 333. Daughtrey E. H. and Bostick J. K. Abstr. 28th Annu. Conf. Mass Spectrom. Allied Topics New York May 1980 pp. 433-434. Wahlgren M. A. Edgington D.N. Rawlings F. F. and Rawls J. L. in “Proceeding of 15th Conference on Great Lakes Research,” International Association for Great Lakes Research Ann Arbor MI 1972 pp. 298-305. Gregory N. L. Anal. Chem. 1972 44 231. Locke J. Boase D. R. and Smalldon K. W. Anal. Chim. Acta 1979 104 233. Sasamoto T. Itoh Y. Hara H. and Sata T. Bull. Tokyo Znst. Technol. 1975 126 91. Ehrlich G. Stahlberg U. and Scholze H. Spectrochirn. Acta Part B 1982 37 45. Friedman A. M. and Lerner J. Adv. Chem. Ser. 1978 No. 171 70. Opauszky I. and Nyary I. Report No. KFKI-1981-09, Hungarian Academy of Sciences Budapest 1981 12 pp. Sasamoto T. Hara H. and Sata T. Shitsuryo Bunseki 1976, 24,121. Haaland J. and Stijfhoorn D. E. in Michelsen 0. B., Editor “Analysis and Application of Rare Earth Materials,” Universitetsforlaget Oslo 1973 pp.153-164. Hamilton E. I. and Minski M. J. Znt. J. Mass Spectrom. Ion Phys. 1972 10 77. Conzemius R. J. and Svec H. J. Talanta 1973 20 575. Itoh M. and Yanagihara K. Bunseki Kagaku 1973,22 10. Van Hoye E. Adams. F. and Gijbels R Talanta 1979,26, 285. Vanderborght B. and Van Grieken R. Talanta 1979 26, 461. Billon J. P. Report No. CEA-R-3883 Commissariat a 1’Energie Atomique Paris 1970 59 pp. Bratton W. D. and Wood C. H. Appl. Spectrosc. 1970,24, 509. Blanchard B. Methodes Phys. Anal. 1971 7 280. Beske H. E. Frerichs G. and Melchers F. G. Fresenius 2. Anal. Chem. 1973 267 99. Billon J. P. Adv. Mass Spectrom. 1974 6 649. Jaworski J. F. and Morrison G. H. Anal. Chem. 1974,46, 2080.Billon J. P. Proc. XVIZI Colloq. Spectrosc. Int. Grenoble, 1975 3 819. Jackson P. F. S. Strasheim A. and Dixon K. Procs XVZZZ Colloq. Spectrosc. Znt. Grenoble 1975 1 164. Oda S. Kubo H. and Kamada H. Bunseki Kagaku 1975, 24 66. Jackson P. F. S. Strasheim A. and Dixon K. “Extended Abstracts International Symposium on the Analytical Che-mistry of Exploration Mining and Processing Materials 1976,” CSIR Pretoria South Africa 1977 pp. 553-556. Matus L. Opauszky I. Nyary I. and Pasztor E. Adv. Mass Spectrom. 1980 8B 1962. Dale L. S. Liepa I. Rendell P. S. and Whittem R. N., Anal. Chem. 1981,53 2288. Winkler P. Fresenius Z. Anal. Chem. 1981 309 266. Wahlgren M. A. Edgington D. N. and Rawlings F. F. in Vogt J. R. Editor “Nuclear Methods in Environmental Research,” Proceedings American Nuclear Society Univer-sity of Missouri Columbia MO 1971 pp.97-103 ANALYST OCTOBER 1984 VOL. 109 333. 334. 335. 336. 337. 338. 339. 340. 341. 342. 343. 344. 345. 346. 347. 348. 349. 350. 351. 352. 353. 354. 355. 356. 357. 358. 359. 360. 361. 362. 363. 364. 365. 366. 367. 368. 369. Hamilton E. I. and Minski M. J. Environ. Lett. 1972,3,53. Crocker I. H. and Knight C. H. Report No. PR-CMa-23, Atomic Energy of Canada Ltd. Chalk River Ontario 1973, pp. 45-48. Carter J. A. Donohue D. L. Franklin J. C. and Stelzner, R. W. Proc. 9th Annu. ConJ Trace Substances in Environ. Health Colombia MO 1975 303. Griffith D. A. Conzemius R. J. and Svec H. J. Talanta, 1971 18 665.Leipold M. H. in Kirkendale G. A. Editor “Analytical Methods for Materials Investigation,” Gordon and Breach, New York 1971 pp. 145-162. Howard 0. H. Anal. Chem. 1972 44. 1482. Kryuchkova 0. I. Derzhiev V. I. Ramendik G. I., Stroganova N. S . and Strel’nikova E. B. Sov. Phys. Dokl., 1979 24 924. Kudryavtsev V. N. and Vlasov V. N. USSR Pat. 1979 No. 673 903. Clegg J. B. Gale I. G. and Millett E. J . Analyst 1973,98, 69. Clegg J . B. and Millett E. J . Proc. SOC. Anal. Chem. 1974, 11 52. Hurrle A. and Dietl J. Proc. Electrochem. SOC. 1980, Hurrle A. and Dietl J. Fresenius Z . Anal. Chem. 1981,309, 277. Bulaceanu R. and Dumitrescu P. Rev. Chim. (Bucharest), 1976 27 894. Luck J. Szacki W. and Pluhar E. Fresenius 2. Anal. Chem. 1981,309 281. Pilate A.and Adams F. Fresenius Z . Anal. Chem. 1981, 309 295. Erametsa O. and Yliruokanen I. Suom. Kern. 1971 44B. 372. Magee C. W. and Harrison W. W. Anal. Chem. 1974 46, 474. Makita T. Kodera K. and Kishi H. in Ogata K. and Hayakawa T. Editors “Recent Developments in Mass Spectroscopy,” University Park Press Baltimore 1970 pp. 32!9-333. Goshgarian B. B . and Jensen A. V. Abstr. I2thAnnu. Conf. Mass Spectrom. Allied Topics Montreal 1964 p. 350. Socha A. J. and Masumoto E. M. Abstr. 15th Annu. Conf. Mass Spectrom. Allied Topics Denver 1967 p. 179. Watanabe E. and Naito M. in Ogata K. and Hayakawa. T. Editors “Recent Developments in Mass Spectroscopy.” University Park Press Baltimore 1970 pp. 249-256. Dabbs M. D. G. Lab. Pract. 1978 27 560. Pearton D.C. G. and Hulse N. D Report No. 2040, National Institute for Metallurgy Randburg South Africa, 1980 15 pp. Herzog L. F. Eskew T. J. Marshall D. J. and Underwood, K. H. Abstr. 29th Annu. ConJ Mass Spectrom. Allied Topics, Minneapolis 1981 pp. 98-99. Auer S. and Berg 0. E. Rev. Sci. Instrum. 1975,46 1530. Banner A. E. Bateman R. H . Halliday J. S . and Willdig, E. Abstr. 20th Annu. Conf Mass Spectrom. Allied Topics, Dallas 1972 pp. 15C155. Banner A. E. and Stimpson B. P. Vacuum. 1974 24 51 1. Blattner R. J. Baker J. E. and Evans C. A. Anal. Chem., 1974,46 2171. Harrison W. W. and Magee C. W. Anal. Chem. 1974 46, 461. Mattson W. A. Thesis University of Virginia Charlottes-ville VA 1975 201 pp. Bingham R. A. and Salter P. L. Anal. Chem. 1976 48, 1735.Franzen J. in Ogata K. and Hayakawa T. Editors “Recent Developments in Mass Spectroscopy,” University Park Press, Baltimore 1970 pp. 296-299. Billon J. P. Analusis 1973 2 137. Warmack R. J. Landau L. Christie W. H. and Carter, J. A. Report No. CONF 811025-21 Oak Ridge National Laboratory Oak Ridge 1981 9 pp. Degreve F. Adv. Mass Spectrom. 1971 5 529. Andreani A. M. Spectra 2000 1979 7 No 54 61. Jansen J. A. J. and Witmer A. W. Fresenius 2. Anal. Chem. 1981 309 262. 80-85 106. 370. 371. 372. 373. 374. 375. 376. 377 * 378. 379. 380. 381. 382. 383. 384. 385. 386. 387. 388. 389. 390. 391. 392. 393. 394. 395. 396. 397. 398. 399. 400, 401. 402. 403. 404. 405. 406. 407 408. 409.410. 411. 412. 413. 1249 Glinskikh V. M. Melashvili V. A. Ordzhonikidze K. G., and Samadashvili. 0. A. J . Anal. Chem. USSR 1974,29,89. Dugger D. and Oblas D. Abstr. 22nd Annu. Conf. Mass Spec from. Allied Topics Philadelphia 1974 pp. 4 14-4 16. Dugger D. L. and Oblas D. Talanta 1977 24 447. Dugger D. L. and Oblas D. W. Abstr. 24th Annu. Conf. Mass Spectrom. Allied Topics Sun Diego 1976 pp. 682-684. Dugger D. L. and Oblas D. W. J . Vac. Sci. Technol. 1980, 17 826. Harrison W. W. and Mattson W. A. Anal. Chem. 1974.46, 1979. Carter J. A. and Sites J. R. Anal. Lett. 1971 4 351. Magee C. W. Donohue D. L. and Harrison. W. W. Anal. Chem. 1972 44 2413. Haemers J. J . Phys. E 1978 11 897. Clegg J. B. and Millett E. J . Acta Electron. 1975 18 27. Ramendik G.I. Tatsii Yu. G. and Chupakhin M. S . J . Anal. Chem. USSR 1973 28 653. Callis E. L. Report No. NBL-265 New Brunswick Labora-tory New Brunswick IVJ 1972 pp. 53-54. Callis E. L. Report No. NBL-267 New Brunswick Labora-tory New brunswick NJ 1973 p. 93. Chupakin M. S . Usp. Anal. Khim. 1974 44. Potapov M. A. Chupakhin M. S. Shtanov V. I. and Zlomanov V. P. J . Anal. Chem. USSR 1978 33 820. Gerasimov V. A. Chupakhin M. S . and Shelpakova. I . R., J. Anal. Chem. USSR 1980,35 138. Berthod J. Ger. Offen. 2719243 1977. Berthod J. Br. Pat. 1576453. 1980. Berthod J . Alexandre B and Stefani R. Int. J . Mass Spectrom. Ion Phys. 1973 10. 478. Berthod J. Adv. Mass Spectrom. 1974,6 421. Zelenin A. E. Pustovit A. N. and Sikharulidze G. G. Sov. Phys. Tech.Phys. 1980 25 1374. Saprykin A. I. Shelpakova I. R. andyudelevich I. G. Izv. Sib. Otd. Akad. Nauk SSR Ser. Khim. Nauk 1979 6 123. Schuy K. D. and Frazen J. US Pat. 3809896 1974. Warmack R. J. and Futrell T. L. Anal. Lett. 1980 13.965. Bingham R. A. Powers P. and Wolstenholme W. A. in Ogata K . and Hayakawa T. Editors “Recent Developments in Mass Spectroscopy,” University Park Press Baltimore, 1970 pp. 339-350. Magee C. W. and Harrison W. W. Anal. Chem. 1973,45, 220. . Eskew T. J . and Smith A. J . Abstr. 20th Annu. Conf. Mass Spectrom. Allied Topics Dallas 1972 pp. 194-199. Gerasimov V. A. and Saprykin. A. I. Instrum. Exp. Tech., 1977 20 809. Conzemius. R. J. and Svec. H. J. Talanta 1973 20 477. Conzernius R. US Pat. 3 91 1 272 1975. Auer S . Report No.N74-31920 Goddard Space Flight Centre Greenbelt MD 1974 12 pp. Fujii T. and Fuwa K. Shitsuryo Bunseki,. 1980 28 151. Fujii T. andFuwa K. Int. J . MassSpectrom. Ion Phys. 1980, 33 325. Degreve F. Figaret R. Le Goux J . J. and Calavrias L. Int. J . Mass Spectrom. Ion Phys. 1974 14 183. Vozenilek E. F . Anal. Chem. 1976 48 1654. Berthod. J. and Stefani R. Int. J. Mass Spectrom. Ion Phys., 1976 22 121. Andreani A. M. Berthod J. Mercadier D. and Mermoud, J. P. Analusis 1979 7 172. Gal’burg V. A. Glavin G. G. and Zelenin A. E. J . Anal. Chem. USSR 1975,29 1223. Shabanova L. N. Shelpakova I. R. and Yudelevich I. G., Izv. Sib. Otd. Akad. Nauk SSSR Ser. Khim. Nauk 1979 6, 126. Fuller C. W. and Whitehead J. Analyst 1971 96 779. Jacobs M. L. Sweeney S.L. and Kerfoot E. G. Abstr. 24th Annu. Conf Mass Spectrom. Allied Topics Sun Diego 1976 p. 504. Brouwer G. Meeuwsen Th. P. M. and Witmer A. W. Int. J. Mass Spectrom. Ion Phys. 1973 12 397. Arai M. Naito M. Takagi T. and Takeuchi M. Abstr. 21st Annu. Conf. Mass Spectrom. Allied Topics Sun Francisco, De Ceuninck F. Van Hoye E. and Adams F. Proc. Anal. Div. Chem. SOC. 1978 15 104. 1973 pp. 507-51 1 1250 ANALYST OCTOBER 1984. VOL. 109 414. 415. 416. 417. Mattson W. A. and Harrison W. W. Anal. Chem. 1975,47, 967. Conzemius R. J. Rhinehart W. A. andSvec H. J. Talanta, 1972 19 1147. Morrison G. H. Colby B. N. and Roth J. R. Anal. Chem., 1972 44 1203. Giltrap J. W. Mapper D. and Farren J. Report No. AERE R-8866 UK Atomic Energy Research Establishment Harwell, 1978,20 pp.Giltrap J. W. Mapper D. and Farren J. Report No. AERE R-9583 UK Atomic Energy Research Establishment Harwell, 1979 15 pp. 419. Conzemius R. J. and Sleege G. A Talanta 1976 23 556. 420. Herzog L. F. Abstr. 20th Annu. Conf. Mass Spectrom. Allied Topics. Dallas 1972 pp. 295-305. 421. Conzemius R. J. and Svec H. J . Talanta 1974 21 171. 422. Powers P. US Pat. 3 522428 1970. 423. Donohue D. L. Thesis University of Tennessee Knoxville, TN 1979 124 pp. 424. Donohue D. L. Carter J . A and Mamantov G. Int. J. Mass Spectrom. Ion Phys. 1980 33 45. 425. Donohue D. L. Carter J. A. and Mamantov G. Int. J . Mass Spectrom. Ion Phys. 1980 35 243. 426. Radermacher L. Beske H. E. Holzbrecher H. and Maeckelburg D Int. J . Mass Spectrom. Ion Phys.1976 20, 333. 427. Radermacher L. and Beske H. E . Adv. Mass Spectrom., 1980 8A. 378. 428. Franzen J. Schonfeld W and Stuwer D. Adv. Mass Spectrom. 1971 5 . 322. 429. Frazer J. W. Carlson L. R. Kray A. M. Bertoglio M. R., and Perone S. P. Anal. Chem. 1971 43 1479. 430. Hunt M. H. Nikula J. V. Posen H and Vickers V. E., Abstr. 20th Annu. Conf. Mass Spectrom. Allied Topics Dallas, 1972 pp. 409-413. 431. Matsumoto R. Suzuki R. Sato K. and Yamaguchi N., Shitsuryo Bunseki 1972 20 63. 432. Frisch M. A. and Reuter W. Anal. Chem. 1973,45 1889. 433. Burdo R. A Roth J. R. and Morrison G. H Anal. Chem., 1974,46 701. 434. Millett G. J Morice J. A. and Clegg J. B Int. J. Mass Spectrom. Ion Phys. 1974 13 1 . 435. Ordzhonikidze K. G. Orlova L. A. Samadashvili 0.A., and Yur'eva 0. K. J. Anal. Chem. USSR 1974 29,684. 436. Woodward W. S . Ridgway T. H. and Reilley. C. N., Analyst 1974 99 838. 437. Stelzner. R. W. Sites J. R. McKown. H. S . . Abstr. 169th National ACS Meeting Philadelphia 1975 No. ANAL72. 438. Van Hoye E Adams F . and Gijbels. R. Bull. SOC. Chim. Belg. 1975 84 595. 439. Conzemius R. J. Adduci D. J Foss G. O. and Svec. H. J., Anal. Chem. 1976,48 1647. 440. Dale L. S . and Whittem R. N. Appl. Spectrosc. 1976 30, 469. 441. Naudin G. Report No. CEA-N-1825. Commissariat ?I I'Ener-gie Atomique Paris 1976 26 pp. 442. Stiiwer D. Int. J . Mass Spectrom. Ion Phys. 1976. 20. 387. 443. Nakamura N. Kai Y. and Konishi F. Shitsuryo Bunseki. 1977 25 67. 444. Hoyland J. R. and Alexander C . A. Abstr. 26th Annu.Conf. Mass Spectrom. Allied Topics St. Louis 1978. pp. 647-649. 445. Vanderborght B. and Van Grieken. R. Anal. Chim. Acta, 1978 103 223. 446. Desantis R. R. Lett R. G. and Plankey. F. Abstr. 31st Pittsburgh Conf Anal. Chem. Appl. Spectrosc. Atlantic City, 2980 No. P15. 447. Knab H. J. Z . Naturforsch Teil A . 1980 35 18. 448. Udalov V. I. and Gaivoronskii P. E. J. Anal. Chem. USSR, 1980,35 1203. 449. Naudin G. Report No. CEA-CONF-5773. CEA Centre d'Etudes de Bruyeres-le-Chatel Montrouge France 1981. Pearton D. and Heron C. Report No. 2117 National Institute for Metallurgy Randburg South Africa. 1981. 8 pp. Warashina M. Mori Y. Shishido T. Kato. Y. Kondo M., Nishihara T. Fujii M. Yamashita T. and Kato. M. Seikatsu Eisei 1978 22 145. 452. Babushkina.T. V. Glavin G. G. Zelenin A. E. and Kostenko. T. A. J. Anal. Chem. USSR 1977. 32. 856. 418. 16 PP. 450. 451. 453. Chupakhin M. S. and Khusainova B. N. J. Anal. Chem. USSR 1971 26 971. 454. Woolston J. R. Abstr. 2lst Annu. Conf. Mass Spectrom. Allied Topics Sun Francisco 1973. pp. 514-517. 455. Viczian M. Kada T. and Rethati R . Publ. Hung. Min. Inst. 1976 No. 19 219. 456. Daughtrey E. H. and Bostick J. K. Abstr. 26th Annu. Conf. Mass Spectrom. Allied Topics St. Louis 1978 pp. 623-625. 457. Bingham R. A. Brown R and Powers P . Adv. Mass Spectrom. 1971 5 326. 458. Marushia P . and Delmore J. Abstr. 29th Annu. Conf. Mass Spectrom. Allied Topics Minneapolis 1981 pp. 462463. 459. Ailloud P. Spectra2000 1979 7 No. 53 50. 460. Gaivoronskii P.E and Chernikovich M. P J. Anal. Chem. USSR 1979 34 1429. 461. Beske H. E. Frerichs G. Jahn M. D. and Melchers F. G., Proc. Int. Conf. Anal. Non-Met. Met. 1980 1981 107. 462. Beske H. E. Fresenius. Z . Anal. Chem. 1971 252 103. 463. Bedrinov V. P. Ukrainskii Yu. M. Fursov. V. Z . and Chupakhin M. S . J . Anal. Chem. USSR 1972. 27 535. 464. Beske H. E. Welter J. M. Frerichs. G. and Melchers F. G., Fresenius Z . Anal. Chem. 1981 309 269. 465. Matus L. Opauszky I . Nyary I. and Pasztor E. Fresenius 2. Anal. Chem. 1981 309 316. 466. Evers J . Weiss A. Kaldis. E. and Muheim J. J. Less-Common Met. 1973 30 83. 467. Crocker I. H. and Knight C. H. Report No. PR-CMa-24, Atomic Energy of Canada Ltd. Chalk River. Ontario 1973, pp. 48-50. Locke J . Boase D. R. and Smalldon.K. W. Med. Sci. Law, 1980 20. 175. Burke R. W. Paulsen. P. J. Maienthal E. J. and Lambert, G. M Talanta 1982 29 809. Shabanova L. N. Shelpakova I. P. Gerasimov V. A. and Yudelevich I. G. Izv. Sib. Otd. NaukSSSR. Ser. Khim. Nauk, 1977 14 55. Kurthy J. Acta Chim. Acad. Sci. Hung. 1978 96 209. Nyary I. and Opauszky I. Fresenius Z . Anal. Chem 1981, 309 274. Socha A. J. Updegrove W. S. and Oro J. Abstr. 18th Annu. Conf. Mass Spectrom. Allied Topics Sun Francisco, 474. Golightly D. W. Methodes Phys. Anal 1971 7 275. 475. Strasheim A. and Jackson P. F. S . J. S. Afr. Inst. Min. Metall. 1971 72 121. 476. Vasile M. J. and Malrn D. L Anal. Chem 1972 44 650. 477. Krendelev F. P. Teplov S. N Karmanov N. S . . and Yablokova S. V. Geol. Rudn. Mestorozhd.1980 22 89. 478. Karpov Yu. A and Alimarin I. P. J. Anal. Chem. USSR. 1979 34 1085. 479. Mahalingham. T. R Murugaiyan. P. Sonni K. S . and Venkateswarlu C. Report No. BARC-797 Bhabha Atomic Research Centre Bombay 1975 7 pp. Callis E. L. Report No. NBL-272 New Brunswick Labora-tories New Brunswick NJ 1974 pp. 62-65. Haney M. A. and Gallagher J. F. J. Forensic Sci. 1975 20, 484. Colby B. N . and Evans C. A. Jr. Anal. Chem. 1974 46, 1236. Maciejewska A and Substyk D. Pr. Nauk. Inst. Chem. Nieorg. Metal. Pierwiastko w Rzadkich Politech. Wroclaw. , 1975 32 67. 484. Graham M. J. and Caplan D. J. Electrochem. Soc. 1973. 120 769. 485. Pauwels J Report No. EUR 6304 Commission of the European Communities Brussels 1979 14 pp. 486. Surowiec J Pr. Nauk.Inst. Chem. Nieorg. Metal. Pierwiast-kow Rzadkich Politech. Wroclaw. 1975 32. 103. 487. Fowkes A. J Br. Ceram. Res. Assoc. Spec. Publ 1975 No. 86. 82. 488. Berkey E. and Hickam. W. M Adv. Mass Spectrom 1971, 5 543. 489. Berkey E. Sweeney G. G. and Hickam W. M. Trans. Am. Nucl. Soc. 1971 14 616. 490. Crocker I. H. Armstrong S. V . andTaylor J. G. V. Report No. PR-CMa-40 Atomic Energy of Canada Ltd Chalk River, Ontario 1977 p. 48. Bunus F. Dumitrescu P. Rodeanu. T and Spiridon. S . J . Radional. Chem. 1978 42. 463. 468. 469. 470. 471. 472. 473. 1970 pp. B222-B226. 480. 481. 482. 483. 491 ANALYST OCTOBER 1984. VOL. 109 125 1 492. 493. 494. 495. 496. 497. 498. 499. 500. 501. 502. 503. 504. 505. 506. 507.508. 509. 510. 511. 512. 513. 514. 515. 516. 517. 518. 519. 520. 521. 522. 523. 524. 323. 526. 527. Northrup C. J. M. Jr. J . Phys. Chem. 1975 79 726. Conzemius R. J. and Svec H. J . Abstr. 22nd Annu. Conf. Mass Spectrom. Allied Topics Philadelphia 1974 pp. 420-423. Van Deventer E. H Renner T. A. Pelto R. H. and Maroni V. A. J . Nucl. Muter. 1977 64 241. Taylor J. B. Calvert L. D. Despault J. G. Gabe E. J. and Murray J . J. J . Less-Common Met. 1974 37 217. Beske H. E. Frerichs G. Jahn M. D. and Melchers F. G., Mikrochim. Acta Suppl. 1979 8 493. Wrigley K. B. Proc. SOC. Anal. Chem. 1972 9 298. Spedding F. H. Sanden B. and Beaudry B. J. J . Less-Common Met. 1973 31 1. Taylor J. B. Calvert L. D. Utsunomiya. T. Wang Y.and Despault J. G. J . Less-Common Met. 1978 57 39. Buffereau M. Vierney J . De La Graviere M. and Pichotin, B . in Karr C. Jr. Editor “Analytical Methods in the Nuclear Fuel Cycle,” IAEA Vienna 1972 pp. 77-99. Chastagner P. and Tiffany B. Adv. Mass Spectrom. 1971,5, 539. Ganivet. M. Genty C Huart. A . and Pichotin B. in Karr C. Jr. Editor “Analytical Methods in the Nuclear Fuel Cycle,” IAEA Vienna 1972 pp. 101-113. Billon J . P. Bull. Inf. Sci. Tech. Commis. Energ. At. (Fr.), 1975 200 93. Billon J . P. Buffereau M. and Deniaud S. Report No. CEA-N-1790 Commissariat il I’Energie Atomique Paris 18 PP. Billon J. P. Buffereau M. and Deniaud S . Proc. XVIll Colloq. Spectrosc. lnt. Grenoble 1975 3 820. Guidoboni R. J. and Evans C. A. Jr Anal. Chem. 1972, 44 2027.Sato K. Uzuki Y. Suzuki R. and Yamaguchi N. Tetsu To Hagane 1972 58 1495. Van Craen M. Verlinden J Gijbels R and Adams F., Talanta 1982 29,773. Stefani R. Andreani A. M. Berthod J. and Eloy J . F., Abstr. 22nd Annu. Conf. Mass Spectrom. Allied Topics, Philadelphia 1974 pp. 425-427. Crocker I . H . Armstrong S. V. andTaylor J. G. V. Report No. AECL-5851 Atomic Energy of Canada Ltd. Chalk River, Ontario 1977 pp. 37-39. Christie W. H Carter J. A. Eby R. E. and Landau L., Anal. Lett. 1979 12A 1123. Vanderborght B. and Van Grieken. R Anal. Chim. Acta, 1977 89. 399. Siu D. P. and Immorlica A. A. Jr. Report No. ERC 41013,3RF Rockwell International. Thousand Oaks CA, 1980 54 pp. Jansen J . A. J and Witmer A. W. Fresenius Z. Anal. Chern. 1981 309 305.Cueilleron. J . Thevenot F. and Viala. J. C. in Tavadze, F. N. Editor “Boron. Production Structure and Properties,” Izdatel’stvo Nauka Moscow 1974 pp. 27-32. Thevenot F. and Viala J. C. Analusis 1975 3 76. Crocker I. H. and Wray L. W. Can. Spectrosc. 1971,16,77. Gauneau M. Spectra2000 1979 7 No. 53 52. Saprykin A. I. Shelpakova 1. R. Yudelevich I. G. and Kanter Yu. O. Izv. Sib. Otd. Akad. Nauk SSR Ser. Khim. Nauk 1979. 4 100. Huber A. M. and Moulin M. J . Radioanal. Chrm. 1972,12, 75. Honig R. E. Thin Solid Films 1976. 31. 89. Lindstrom R. M. Fleming R. F. and Paulsen P. J Nut. Bur. Stand. (US) Spec. Publ. 1980 No. 572 61. Gaivoronskii P. E. Larin N . V and Chernikovich M. P Poluch. Anal. Veshchestv. Osoboi Chist. (5th Dokl. Vses. Konf. Moscow USSR 2976) 1978 165.Look D. C. J . Appl. Phys. 1977 48 5141. Dorokhov. A . N. Shelpakova I. R. Yudelevich I. G. and Christanova S. T. Izv. Sib. Otd. Akad. NaukSSR Ser. Khim. Nauk. 1977 61. Brozel M. R. Clegg. J. B and Newmann. R. C J . Phys. D , 1978 11 1331. Huber A. M. Morillot G. Merenda. P. and Linh N. T. in Goldanskii V. I. Gomer R Schafer. F. P and Toennies. J . P. Edifors “Proceedings of the 2nd International Conference on Secondary Ion Mass Spectrometry (SIMS II),” Springer-Verlag Berlin 1979 pp. 91-94. 528. 528a 528b 528c. 528d. 529. 530. 531. 532. 533. 534. 535. 536. 537. 538. 539. 540. 541, 542. 543. 544. 545. 546. 547. 548. 549 * 550. 551. 552. 553. 554. 555. 556. 557. 558. 559.560. 561. 562. Clegg J . B. Grainger F. and Gale I. G. J . Muter. Sci. 1980, 15 747. Blackmore G. W. Clegg J. Hislop. J. and Mullin J. J. Electron. Muter. 1976. 5 401. Martin G. M. Jacob G. Hallais J. Grainger F. Roberts, J . A. Clegg J. B. Blood P. and Poiblaud G. J. Phys. C., 1982 15 1841. Clegg J. B. in Makram-Ebeid S . and Tuck B Editors, “Semi-insulating 111-V Materials,” Shiva Publishing Nantwich, 1982. p. 80. Clegg J. B. Mullin J . Timmins K Blackmore G. Everett, G. and Snook R. J. Electron. Mater. 1983 12 879. Tong S. S. C. and Williams J. P J . A m . Ceram. SOC. 1970, 53 58. KO R. Report No. HEDL-SA-392. Hanford Engineering Development Laboratory Richland WA 1972. 9 pp. Beveridge R. L. Appl. Spectrosc. 1973 27 271. Gladskoi V.M. and Belousov V. I. Tr. Khim. Khim. Tekhnol. 1973 110. Billon J. P. Buffereau M. Sanson C. and Vienney J . , Report No. CEA-N-1788 CEA Centre d’Etudes de Bruyers-le-Chatel Montrouge France 1975 15 pp. Williams J. P. Su Y.-S. and Wise W. M. Microchim. Acta, 1977 11 527. Allegre J. Spectra2000 1979 7 No. 53 61. Alkire G. J . Report No. HEDL-6798 Hanford Engineering Development Laboratory Richland WA 1981 6 pp. Fitchett A. W. and Buck R. P. Anal. Chem. 1973,45,1027. Kaushik S. B. Gupta S . R. Chatterjee S . N. and Laroia, K. K. Indian J . Pure Appl. Phys. 1974 12 389. Spedding F. H. Beaudry B. J . Henderson D. C. and Moorman J. J . Chem. Phys. 1974 60 1578. Barnes I. L. Moore L. J Machlan L. A . Murphy T. J., and Shields W. R. J . Res. Natl. Bur. Stand.Sect. A 1975,79, 727. Dietze H. J . and Zahn H. ZIF. Mitt. 1982 46 60. Kowalczyk J. Lukas W and Tomasik Z. Pr. Nauk. Inst. Chem. Nieorg. Metal. Pierwiastkow Rzadkich Politech. Wroclaw. 1975 32 17. Crocker I. H. Schruder J. A. Bokwa S . R. and Hilde-brandt B. W. Report No. AECL-5851 Atomic Energy of Canada Ltd. Chalk River Ontario 1977 pp. 39-40. Russell,.D. S . Chem. Can. 1979 31 13. Garner E. L. Murphy T. J Gramlich. J. W. Paulsen. P. J . , and Barnes I. L. J . Res. Natl. Bur. Stand. Sect. A 1075 79, 713. Maier. H. Daniel D. R . Herkert R. and Luck J . J . Muter. Sci. 1978 13. 297. Bergholz J . Luck. J . Moller P. and Szacki W Fresenius Z . Anal. Chem. 1974 269. 121, Butterworth A. German B. Morgans. D. and Scaplehorn, A. J. Forensic Sci. SOC.1974 14 41. Fuller C. W. Himsworth G. and Whitehead J. Analysf, 1971 96 177. Dabbs M. D. G. German B. Pearson E. F and Scaple-horn A. W. J . Forensic Sci. SOC. 1973 13 281. Ikeda K. Bunseki Kiki 1973 11 456. Drayton J. V. Chem. Br 1975 11 439. Haney M. A J . Forensic Sci. 1977 22 535. German B. Morgans D. Butterworth A . and Scaplehorn, A. J . Forensic Sci. Soc. 1978 18 113. Locke J . Boase D. and Smalldon K. W. J. Forensic Sci. SOC. 1978 18 123. Stefani. R. Berthod J. Pellet J. Bourguillot R. Cornu. A . , and Bouchez R. ~ Report No. CEA-CONF-2406. Commis-sariat a 1’Energie Atomique Saclay 1973 16 pp. Jacobs M. L. Sweeney S. L. and Rogowski M. F. Abstr. 27th Pittsburg Conf. Anal. Chem. Appl. Spectrosc. Cleveland, OH 1976 No. 220. Sharkey A. G. Jr.in Freudenthal R. I. and Jones P. W., Editors. “Aromatic Hydrocarbons Chemistry Metabolism and Carcinogenesis,” Raven Press New York 1976 pp. Kelley J . A Radiochem. Radioanal. Lett. 1973 14 95. Gray L. W. and Ondrejoin R. A. Report No. DP-MS-74-45, E. I . Du Pont de Nemours Aiken SC 1974,22 pp. Bodnar L. Z. Report No. NBL-277 New Brunswick Labora-tories New Brunswick NJ 1976 pp. 14-19. Naudin G. Spectra 2000. 1979 7 No. 54 65. 341-347 1252 563. 564. 565. 566. 567. 568. 569. 570. 571. 572. 573. 574. 575. 576. 577. 578. 579. 580. 581. 582. 583. 584. 585. 586. 587, 588. 589. 590. 591. 592. 593. 594. 595. 596. 597. 598. Briden F. Dorsey J. and Johnson L. Znt. J . Environ. Anal. Chem. 1981 9 189.Briden F. and Lewis D. Znt. J. Environ. Anal. Chem. 1981, 9 249. Brown R. andvossen P. G. T. Anal. Chem. 1970,42,1820. Daughtrey E. H. and Bostick J. K. Abstr. 27th Annu. Con5 Mass Spectrom. Allied Topics Seattle 1979 pp. 41041 1. Carter J. A. Walker R. L. and Sites J. R. Am. Chem. SOC., Div. Fuel. Chem. Prepr. 1973 18 78. Davison R. L. Natusch D. F. S. Wallace J. R. and Evans, C. A. Jr. Environ. Sci. Technol. 1974 8 1107. Natusch D. F. S. Wallace J. R. and Evans C. A. Jr., Science 1974 183 202. Von Lehrnden D. J. Jungers R. H. and Lee E. R. Jr., Anal. Chem. 1974 46,239. Daughtrey E. H. Jr. Abstr. 24th Annu. Conf. Mass Spectrom. Allied Topics Sun Diego 1976 pp. 63tA40. Linton R. W. Loh A. Natusch D. F. S. Evans C. A. Jr., and Williams P. Science 1976 191 852.Trischan G. M. Abstr. 27th Annu. Conf. Mass Spectrom. Allied Topics Seattle 1979 pp. 406-409. Kessler T. Sharkey A. G. and Friedel R. A. Report No. USBM-TPR-42 US Bureau of Mines Pittsburgh 1971 15 pp. Sharkey A. G. Jr. Kessler T. and Friedel. R. A A h . Chem. Ser. 1975 No. 141 48. Bolton N. E. Carter J. A. Emery J. F. Feldman C., Fulkerson W. Hulett L. D. and Lyon W. S. Am. Chem. SOC. Div. Fuel Chem. Prepr. 1973 18 114. Taylor C. E. Report No. EPA/660/2-74-078 Environmental Protection Agency Athens GA 1974 41 pp. Brown R. Taylor H. E. and Mesich F. Abstr. Pittsburgh Conf. Anal. Chem. Appl. Spectrosc. Cleveland OH 1975, No. 1. Gasior S. J. Lett R. G. Strakey J. P. and Haynes W. P., Am. Chem. SOC. Div. Fuel Chem. Prepr. 1978 23 88. Ure A.M. Anal. Proc. 1980 17 409. Taylor C. E. and Taylor W. J. Report No. EPA-66012-74-001 Environmental Protection Agency Athens GA 1974, Moody J . R. Paulsen P. J. Rains T. C. and Rook H. I. in Lafleur P. D. Editor “Accuracy in Trace Analysis Sampling, Sample Handing Analysis,” Volume 1 National Bureau of Standards Gaithersburg MD 1976 pp. 267-273. Paulsen P. J. and Moody J. R. Abstr. 24th Annu. Conf. Mass Spectrom. Allied Topics Sun Diego 1976 pp. 501-503. Ribaudo C. Campbell C. Bulusu S . Fisco W. Chen T., Kramer H. and Castorina T. Report No. ARLCD-TR-81028 US Army Armament Research and Development Command Dover NJ 1981 61 pp. Brown R. and Vossen P. G. T. in Westley B. Editor, “Identification and Measurement of Environmental Pollu-tants,” National Research Council of Canada Ottawa 1971, Klose M.Fresenius Z . Anal. Chem. 1971 254 195. Crocker I. H. and Merritt W. F. Water Res. 1972 6 285. Wahlgren M. A. Edgington D. N. and Rawlings F. F., Report No. ANL-7860 (Pt. 3) Argonne National Laboratory, Argonne IL 1972 pp. 55-63. Crocker I. H. Am. SOC. Test. Muter. Spec. Tech. Publ. 1975, STP573 71. Vandelannote R. Blomrnaert W. Gijbels R. and Van Grieken R. Fresenius 2. Anal. Chem. 1981,309,291. Klose M. Fresenius 2. Anal. Chem. 1971,254 7. Watson J. C. Report No. PLN-MA-572 Batelle Pacific Northwest Laboratories Richland WA 1978,406 pp. Sturgeon R. E. Berman S. S. Desaulniers A. and Russell, D. S. Anal. Chem 1979,51,2364. Mykytiuk A. P. Russell D. S. and Sturgeon R. E. Anal. Chem. 1980,52 1281. Sturgeon R.E. Berman S. S. Desaulniers J. A. H., Mykytiuk A. P. McLaren J. W. and Russell D. S. Anal. Chem. 1980,52 1585. Walters D. B. and Russell R. B. Abstr. 24th Southeastern Regional ACS Meeting Birmingham AL 1972 No. 92. Pellizzari k. D. Bursey J. T. Smith D. J. Castillo N. P., and Willis S. L. Abstr. 26th Annu. Conf. Mass Spectrom. Allied Topics St. Louis 1978 pp. 42-45. Kessler T. Sharkey A. G. and Friedel R. A. Report No. PB-204151 US Bureau of Mines Pittsburgh 1971 19 pp. 30 PP. pp. 427-431. 599. 600. 601. 602. 603. 604. 605. 606. 607. 608. 609. 610. 611. 612. 613. 614. 615. 616. 617. 618. 619. 620. 621. 622. 623. 624. 625. 626. 627, 628. 629. 630. 631. 632. 633. 634. 635. 636.637. 638, 639. 640, 641. ANALYST OCTOBER 1984 VOL. 109 Guidoboni R. J. Anal. Chem. 1973,45 1275. Kessler T. Sharkey A. G. Jr. and Friedel R. A. Report No. RI-7714 US Bureau of Mines Pittsburgh 1973 8 pp. Pollock E. N. Am. Chem. SOC. Div. Fuel Chem. Prepr., 1973 18 92. Daughtrey E. H. Jr. and Bostrick J. K. Jr. Abstr. 25th Annu. Conf. Mass Spectrom. Allied Topics Wahsington DC, 1977 pp. 700-702. Lyon W. S . Lindberg S. E. Emery J. F. Carter J. A., Ferguson N. M. Van Hook R. I. and Raridon R. J., “Nuclear Activation Techniques in the Life Sciences,” IAEA, Vienna 1978 pp. 615-625. Kronberg B. I. Brown J. R. Fyfe W. S. Peirce M. and Winder C. G. Fuel 1981 60 59. Jungers R. H . Carter J. A. Clay D. A. and Bumgarner, J. E. Report No. CONF-760401-4 Oak Ridge National Laboratory Oak Ridge TN 1976,22 pp.Alkire G. J. Anderson H. J. Delvin W. L. Hartcorn, L. A. Matsumoto W. Y. and Moore R. L. in Karr C. Jr., Editor “Analytical Methods in the Nuclear Fuel Cycle,” IAEA Vienna 1972 pp. 295-313. Christie 0. H. J. Talanta 1975 22 1048. Howarth R. and Rankin P. C. Chem. Geol. 1975 15 239. Jackson P. F. S. and Strelow F. W. E. Chem. Geol. 1975, 15 303. Maksimovic Z . and Roaldset E. in “Travaux du Comite International pour 1’Etude des Bauxites de 1’Alumine et de I’Aluminium,” Academie Yougoslave des Sciences et des Arts, Zagreb 1976 No. 13 pp. 199-220. Roaldset E. Geochim. Cosmochim. Acta 1975,39 455. Rankin P. C. Chishitsugaku Zasshi 1976 82 215. Rankin P. C. and Childs C. W. Chem. Geol. 1976 18 55.De Albuquerque C. A. R. Geochim. Cosmochim. Acta 1977, 41 1. Jackson P. F. S. and Strasheim A. “Extended Abstracts, International Symposium on the Analytical Chemistry of Exploration Mining and Processing Materials 1976,“ CSIR, Pretoria South Africa 1977 pp. 606-610. Rankin P. C. N. 2. Dep. Sci. Znd. Res. Bull. 1977,218 110. Taylor S. R. Capp A. C. Graham A. L. and Blake D. H., Contrib. Mineral. Petrol. 1969 23 1. Jakes P. and Gill J. Earth Planet. Sci. Lett. 1970 9 17. Roaldset E. and Rosenqvist I. Th. Lithos 1971 4 417. Roaldset E. Lithos 1973 6 349. Cook E. W. Chem. Geol. 1973 11 321. Neuilly M. Bussac J. Frejacques G. N. Vendryes G. and Yvon J. C. R. Acad. Sci. Ser. D 1972 275 1847. Hagemann R. Nief G. and Roth E . Bull. Znf. Sci. Technol. Paris 1974 193 65.Frejacques C. Blain C. Devillers R. and Ruffenach J. C., “The Oklo Phenomenon,” IAEA Vienna 1975 pp. 509-524. Walton R. D. Jr. and Cowan G. A. “The Oklo Phenom-enon,” IAEA Vienna 1975 pp. 499-507. Roaldset E. and Christie 0. H. J. Schweiz. Mineral. Petrogr. Mitt. 1975 55 191. Neiva A. M. R. Chem. Geol. 1975 16 153. Glikson A. Y. Geochim. Cosmochim. Acta 1976,40 1261. Kyle P. R. and Rankin P. C. Geochim. Cosmochim. Acta, 1976,40 1497. Peccerillo A. andTaylor S. R. Earth Planet. Sci. Lett. 1976, 32 121. Peccerillo A. and Taylor S. R. Contrib. Mineral. Petrol., 1976,58 63. Gorton M. P. Geochim. Cosmochim. Acta 1977,41 1257. Taylor S. R. and Hallberg J. A. Geochim. Cosmochim. Acta 1977,41 1125. McLennan S. M. and Taylor S. R. Nature (London) 1979, 282 247.Taylor S. R. and McLennan S. M. Geochim. Cosmochim. Acta 1979,43 1551. Bavinton 0. A. and Taylor S. R. Geochim. Cosmochim. Acta 1980 44 639. McLennan S. M. and Taylor S. R. Nature (London) 1980, 285 621. Neiva A. M. R. Chem. Geol. 1980 29 49. Jenner G. A. Chem. Geol. 1981 33 307. Bhatia M. R. and Taylor S. R. Chem. Geol. 1981,33 115. Irving A. J. and Price R. C. Geochim. Cosmochim. Acta, 1981,45 1309 ANALYST OCTOBER 1984 VOL. 109 642. 643. 644. 645. 646. 647. 648. 649. 650. 651. 652. 653. 654. 655. 656. 657. 658. 659. 660. 661. 662. 663. 664. 665. 666. 667, 668. 669. 670. 671. 672. 673. 674. 675. 676. 677. Taylor H. E. and Lichte F. E. Geophys. Res. Lett. 1980,7, 949.McCulloch M. T. and Perfit M. R. Earth Planet. Sci. Lett., 1981 56 167. Whitford D. J. White W. M. and Jezek P. A. Geochim. Cosmochim. Acta 1981,45 989. Cornu A. and Brun J. C. Dosage Elem. Etat Traces Roches Autres Subst. Miner. Natur. Actes Colloq. 1968 1970 215. Hunt M. H. Abstr. 18th Annu. Conf. Mass Spectrom. Allied Topics Sun Francisco 1970 pp. 24&244. Kopp 0. C. and Statts P. A.,J. Phys. Chem. Solids 1970,31, 2469. Nalbantoglu M. Dosage Elem. Etat Traces Roches Autres Subst. Miner. Natur. Actes Colloq. 1968 1970 223. Agiorgitis G. Panagos A. and Marien R. Chem. Erde, 1976 35 199. Viczian M. Foldt. Korl. 1978 108 47. Kronberg B. I. Couston J. F. Stilianidi Filho B. Fyfe, W. S. Nash R. A, and Sugden D. Econ. Geol. 1979 74, 1869. Ramendik G.I. and Lisitsyn A. E. Geochem. Inter. 1979, 117. Cesbron F. and Drin N. Report No. CEA-R-5118 Centre d’Etudes Nucleaires de Fontenay-aux-Roses France 1981,110 PP. Erametsa O. and Haukka M. Suom. Kem. 1970 B43 189. Van Hook R. I. Bull. Environ. Contam. Toxicol. 1974 12, 509. Ure A. M. Bacon J. R. Berrow M. L. and Watt J. J., Geoderma 1979 22 1. Tong S. C. Gutenmann W. H. St. John L. E. and Lisk, D. J. Anal. Chem. 1972 44 1069. Ure A. M. and Bacon J. R. Geochim. Cosmochim. Acta, 1978 42 651. Hintenberger H. and Berghof W. in Ogata K. and Hayakawa T. Editors “Recent Developments in Mass Spectroscopy,” University Park Press Baltimore 1970 pp. 657-667. Mason B. and Graham A. L. Smithson. Contrib. Earth Sci., 1970 3 1. Smales A. A. Mapper D.Webb M. S. W. Webster R. K., and Wilson J. D. Science 1970 167 509. Morrison G. H. Nadkarni R. A. Potter N. M. Rothenberg, A. M. and Wong S. F. Radiochem. Radioanal. Lett. 1972, 11 251. Allen R. O. Jr. and Mason B. Geochim. Cosmochim. Acta, 1973 37 1435. Hintenberger H. Jochum K. P. and Seufert M. Earth Planet. Sci. Lett. 1973 20 391. Hintenberger H. Jochum K. P. and Seufert M. in Kiesl, W. and Malissa H. Jr. Editors “Analyse extraterrestrichen Materials,” Springer-Verlag Vienna 1974 pp. 125-145. Jochum K. P. Seufert M. and Hintenberger H. Meteoritics, 1974 9 355. Jacques A. L. Lowenstein P. L. Green D. H. Kiss E., Nance W. B. Taylor S. R. and Ware N. G. Meteoritics, 1975 10 289. Shima M. Jochum K. P. Sighinolfi G. P. and Hinten-berger H. Meteoritics 1974 9 199.Jochum K. P. Hintenberger H. and Buchwald V. F., Meteoritics 1975 10 419. Mason B. Nelen J. A. Muir P. and Taylor S. R., Meteoritics 1975 11 21. Hintenberger H. Jochum K. P. Braun O. Christ P. and Martin W. Earth Planet. Sci. Lett. 1978 40 187. Knab H. J. and Hintenberger H. Meteoritics 1978 13,522. Li Z-H. Liang Z-C. Zhou H. Li Y .-G. Zhao D.-Z. and Zhang M.-L. Ti Ch’iu Hua Hsueh 1979 249. Jochum K. P. Grais K. I. and Hintenberger H. Meteoritics, 1980 15 31. Jochum K. P. Seufert M. and Bergmann F. Z. Natur-forsch. Teil A 1980 35 57. Knab H. J. Geochim. Cosmochim. Acta 1981,45 1563. Oro J. Updegrove W. S. Gilbert J. McReynolds J., Gil-av E. Ibanez J. Zlatkis A. Flory D. A. Levy R. L., and Wolf C. J. in Levinson A. A. Editor “Proceedings of the Apollo 11 Lunar Science Conference,” Volume 2 Per-gamon Press Oxford 1970 pp.1901-1920. 678. 679. 680. 681. 682. 683. 684. 685. 686. 687. 688. 689. 690. 691, 692. 693, 694. 69s. 696. 697. 698. 699. 700. 701. 702. 703. 704. 705. 706. 1253 Bouchet M. Kaplan G. Voudon A. and Beroletti M.-J. in Levinson A. A. Editor “Proceedings of the 2nd Lunar Science Conference,” Volume 2 MIT Press Cambridge MA, Morrison G. H. Gerard J. T. Potter N. M. Gangadharam, E. V. Rothenberg A. M. and Burdo R. A. in Levinson, A. A. Editor “Proceedings of the 2nd Lunar Science Conference,” Volume 2 MIT Press Cambridge MA 1971, pp. 1169-1185. Smales A. A. Mapper D. Webb M. S. W. Webster R. K., Wilson J. D. and Hislop J.S. in Levinson A. A. Editor, “Proceedings of the 2nd Lunar Science Conference ,” Volume 2 MIT Press Cambridge MA 1971 pp. 1253-1258. Taylor S. R. Rudowski R. Muir P. Graham A. and Kaye, M. in Levinson A. A. Editor “Proceedings of the 2nd Lunar Science Conference,’’ Volume 2 MIT Press Cambridge MA, 1971 pp. 1083-1089. Strasheim A. Jackson P. F. S. Coetzee J. H. J. Strelow, F. W. E. Wybenga F. T. Gricus A. J. Kokot M. L. and Scott R. H. in Heymann D. Editor “Proceedings of the 3rd Lunar Science Conference,” Volume 2 MIT Press Cam-bridge MA 1972 pp. 1337-1342. Taylor S. R. Kaye M. Muir P. Nance W. Rudowski R., and Ware N. in Heymann D. Editor “Proceedings of the 3rd Lunar Science Conference,’’ Volume 2 MIT Press, Cambridge MA 1972 pp. 1231-1249. Taylor S.R. Gorton M. P. Muir P. Nance W. B., Rudowski R. and Ware N. Geochim. Cosmochim. Acta, 1973,37 2665. Taylor S. R. Gorton M. P. Muir P. Nance W. Rudowski, R. and Ware N. “Proceeedings of the 4th Lunar Science Conference,” Volume 2 Pergamon Press New York 1973 pp. 1445-1459. Hubbard N. J. Ramendik G. I. Gronskaya S. I. Gubina I. Ya. and Gushchin V. N. Geokhimiya 1975 803. Hubbard N. J. Vinogradov A. P. Ramendik G. I. and Chupakhin M. S. “The Soviet American Conference on Cosmochemistry of the Moon and Planets,” Part 1 NASA Special Publication SP-370 1977 pp. 259-262. Hamilton E. I. Minski M. J. and Cleary J. J. Sci. Total Environ. 1972/73 1 341. Morrison G. H. ICP Inf. Newsl. 1977 3 165. Erametsa E. Erametsa O. and Yliruokanen I. Suom. Kem. 1972 B45 311.Cornides I. Pais I. and Viczian M. in Proc. X V I I Colloq. Spectrosc. Int. Florence 1973 1 301. Erametsa O. Haarala A.-R. and Yliruokanen I. Suom. Kem. 1973 B46 234. Erametsa O. Viinanen R. and Yliruokanen I. Suom. Kem. 1973 B46 355. Yliruokanen I. Bull. Geol. SOC. Finl. 1975 47 71. Yliruokanen I. Ann. Acad. Sci. Fenn. Ser. A2 1975,176,28 PP. Chamel A, Andreani A. M. and Mercadier D. Commun. Soil Sci. Plant Anal. 1979 10 1311. Leonhardt J. W. Dahn E. Dietze H. J. Freyer K. Geisler, M. Hartmann G. Jung K. and Schelhorn H. “Nuclear Activation Techniques in the Life Sciences,” IAEA Vienna, Ramaswamy K. Perumal R. and Pandurangan S. V. Appl. Spectrosc. 1979 33,516. Chamel A. R. Andreani A. M. and Eloy J. F. Plant. Physiol. 1981,67 457. Ball D.F. Barber M. and Vossen P. G. T. Sci. Total Environ. 1973 2 101. Ball D. F. Barber M. and Vossen P. G. T. Sci. Total Environ. 1975 4 193. Ball D. F. Barber M. and Vossen P. G. T. Biomed. Mass Spectrom. 1974 1 365. Erametsa O. and Sihvonen M. L. Ann. Med. Exp. Biol. Fenn. 1971,49 35. Harrison W. W. Clemena G. G. and Magee C. W. J. Assoc. Off. Anal. Chem. 1971 54 929. Harrison W. W. and Clemena G. G. Clin. Chim. Acta 1972, 36,485. Losee F. Cutress T. W. and Brown R. in Hemphill D. S . , Editor “Trace Substances in Environmental Health,” Univer-sity of Missouri Columbia MO 1973 pp. 19-24. 1971 pp. 1247-1252. 1979 pp. 91-102 1254 ANALYST OCTOBER 1984 VOL. 109 707. 708. 709. 710. 711. 712. 713. 714. 715. 716. 717. 718. Gooddy W.Williams T. R. and Nicholas D. Brain 1974, 97 327. Gooddy W. Hamilton E. I . . and Williams T. R. Brain, 1975 98 6.5. Christie 0. H. J. Dinh-Nguyen. N. Vincent J. Hellgren. L and Pimlott W. J . Invest. Dermatol. 1976 67 587. Findlay G. H. and Vismer H. F. Br. J . Dermatol 1977,97, 497. Murphy R. C. Abstr. 28th Annu. Conf. Mass Spectrom. Allied Topics New York 1980 p. 208. Hamilton E. I . and Minski M. J . Sci. Total Environ., 1972173 I 375. Pommez P. and Clarke M. A. Report No. ARS-S-51 US Agricultural Research Service Southern Region. New Orleans LA 1975 pp. 40-46. Tong S. S. G. Morse R. A. Bache C. A. and Lisk. D. J . , Arch. Environ. Health 197.5 30 329. Locke J . Carpenter R. andosselton M. D Med. Sci. Law. 1981 21 123. Davis R. E.in Shirley. D. A, Editor “Electron Spectro-scopy,” North Holland Amsterdam 1972 pp. 903-908. Aoki M. and Kurosaki K . Kobunshi Kagaku 1972,29.442. Callis E . L . Report No. NBL-272 New Brunswick Labora-tories New Brunswick NJ 1974 p. 61. 719. 720. 721. 722. 723. 724. 725. 726. 727. 728. Cristy S . S . and Annis B. K. Report No. Y-Da-5554 Oak Ridge Y-12 Plant Oak Ridge TN 1973 27 pp. Braier H. A. and Eppolito J. in Yen T. F . Editor “Role of Trace Metals in Petroleum,’‘ Ann Arbor Science Publishers, Ann Arbor MI 1975 pp. 65-87. Baudin. G J . Radioanal. Chem. 1977 37. 119. Morrison G. H. in Lafleur P. D . Editor “Accuracy in Trace Analysis Sampling. Sample Handling Analysis,” Volume 1, National Bureau of Standards Gaithersburg MD 1976 pp.Kane P. F. Chem. Technol. 1971 532. Bingham R. A. and Salter P. L. Int. J . Mass Spectrom. Ion Phys 1976 21 133. Conzemius R. J. and Svec R. J. Abstr. 24th Annu. Conf. Mass Spectrom. Allied Topics San Diego 1976. pp. 323-326. Opauszky I. Pure Appl. Chem. 1982 54 879. Date. A. R. and Gray A. L. Analyst 1981 106 1255. Cantle. J. E. Hall E. F. Shaw C. J. andTurner P. J . Int. J . Muss Spectrom. Ion Phys. 1983 46 1 1 . 65-77. Paper A4149 Received February lst 1984 Accepted May 14th 198

ISSN:0003-2654    

DOI:10.1039/AN9840901229

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST OCTOBER 1984 VOL. 109 1255 Evaluation of an Electrothermal Atomisation Procedure for the Determination of Lead in Potable Water John Webster" Department of Water and Drainage Lothian Regional Council 6 Cockburn Street Edinburgh, EHl INZ UK and Andrew Wood Napier College Edinburgh UK An electrothermal atomisation procedure using a lanthanum-impregnated tube was investigated for its suitability for the routine determination of lead in potable water. The precision and bias were found to be acceptable but to ensure adoption of the procedure for routine use a detailed comparison was made between the results and those obtained using two other common methods involving concentration followed by atomic-absorption spectrophotometry. The first involved concentration by evaporation and the second extraction of the ammonium tetramethylenedithiocarbamate complex into isobutyl methyl ketone.A statistical comparison of data from all three procedures was made. Keywords Lead determination; potable water; atomic-absorption spectrophotometry; electrothermal atomisation; lanthanum-treated tube Electrothermal atomisation procedures are widely used in many fields for the direct determination of trace metals in a variety of materials. The determination of lead in water has been the subject of many papers but has not been widely accepted because of reports of serious matrix interference effects; various reagents have been suggested for eliminating or reducing such effects or for improving the precision.'.* In addition L'VOV~ has recommended the use of a modified tube design to ensure isothermal heating of the sample at atomisation and to give increased sensitivity and improved precision of analysis a technique adopted by Sturgeon et ~ 1 .~ for the analysis of marine sediments. More recently it was reported that either the addition of lanthanum to samples or impregnation of the tube overcame interferences in the determination of lead in a variety of potable waters and met the requirements recommended by the Water Research Centre for analytical methods by having a total standard deviation of not more than 1.5 pg 1-1 or 5% of the concentration and a bias of not more than 5 pg 1 - 1 or 10% of the concentration whichever was the greater.5 It was decided to investigate this procedure using an impregnated tube6 as part of the work to be unaertaken by a student from Napier College Edinburgh on a 6-month work experience secondment to the Lothian Region's Water and Drainage Department.The background to the project lay in the need to analyse comparatively large numbers of household tap waters as part of a programme to identify areas of high lead content (>100 pg 1-1) within the City of Edinburgh originating from the use of lead tanks or piping within older properties. The water supplied to the City comes from upland sources and is relatively unpolluted and of low hardness. In some areas the raw water may be coloured by humic acids but this is much reduced by treatment. During the course of the investigation a range of raw and treated waters from within the Region were used for tests and the chemical characteristics of four of these, selected to cover a range of hardness and organic carbon content are given in Table 1.Most of the work was carried out on samples originating from Alnwickhill and Fairmilehead as these supplies provide the bulk of the water to the City, particularly in the areas where elevated lead levels would be expected. *. To whom correspondence should be addressed. The statistical performance of the electrothermal atomis-ation procedure was evaluated after optimising the instrumen-tal conditions and the data produced by this procedure were compared with lead levels measured routinely in the labora-tory using flame AAS after concentration by evaporation. In addition a number of samples were analysed in replicate by extraction of the ammonium tetramethylenedithiocarbamate (ammonium pyrrolidinedithiocarbamate; APDC) complex into a solvent again followed by flame AAS.Experimental Reagents Nitric acid 70% m/V. Aristar grade (BDH Chemicals). Lanthanum nitrate La( N 0 3 ) 3 . Analytical-reagent grade. Apparatus All glassware was soaked in 10% nitric acid for 24 h before use. Measurements were made using a Perkin-Elmer 272 atomic-absorption spectrophotometer with a Perkin-Elmer HGA 2200 graphite furnace accessory which allowed the use of pre-set drying ashing and atomising temparature cycles but with no ramping. Background correction was made by deuterium arc and a lead hollow-cathode lamp (IL) was used at 283.3 nm. Signals were recorded on a 10-mV recorder as peak heights.Twenty-microlitre aliquots of sample and standard were used for all measurements delivered by a Perkin-Elmer auto-sampler into pyrolytically coated graphite tubes (Perkin-Elmer) held in the light path by end-clamps which also served as electrical contacts. The tubes were impregnated with lanthanum as described below. Disposable acid-soaked 4-ml polystyrene cups were used with the autosampling accessory. Electrothermal Atomisation A tube was soaked in saturated lanthanum nitrate solution overnight wiped dry then subjected to a series of heating runs in the range 110-500 "C wiping any crystals from the outside of the tube each time with a tissue. Finally a programmed run with atomisation at 2200 "C was made at least twice or until no signal was observed during the run.It was important to follow this procedure with a new tube i 1256 ANALYST OCTOBER 1984. VOL. 109 ~ ~~ Table 1. Summary of major ion concentrations in waters used during tests. All results in milligrams per litre except where stated Property Conductivity/pS cm- . . Totalsolids . . . . . . Hardness (CaC03) . . . . Chloride(C1) . . . . . . Sodium(Na) . . . . . . Potassium(K) . . . . . . Magnesium(Mg) . . . . Total organic carbon (C) . . Alnwickhill 60-80 35-50 8-1 1 5 0.5 2-4 2-3 95- 120 Fairmilehead 7 w 1 0 40-60 3 w 5 5-8 4 0.3 1-2.5 2-4 Pateshill 100-130 85- 105 45-70 6-12 5 0.6 8-13 2-3.5 Harburn head 250-280 150-180 120-150 7-9 6 1.3 3 5 4 .8 1-3 Table 2. Recovery data for lead by the APDC procedure Results and Discussion Sample Lead added/pg 1- I Lead found*/pg 1- 1 Performance Characteristics of the Electrothermal Atomis-ation Procedure Distilledwater . . . . 25.0 28.7 75.0 100.0 50 50 50 Alnwickhillsupply . . . 0 Fairmilehead supply . . 0 Castle Moffat (raw) . . 0 * Duplicate measurements. 70.0 98.7 1 52 5 52 1 45.5 order to avoid short-circuiting of the power or physical damage. The conditions of analysis were optimised according to the procedures outlined in the manufacturer’s manual using a sample of raw water comparatively high in organic carbon (1 1 mg 1-*) and spiked with 50 pg 1-1 of lead. The optimised conditions for subsequent analysis were as follows: Drying temperature .. . . 110 “C; time . . . . 50 s Charring temperature . . . . 500 “C; time . . . . 30 s Atomisation temperature . . 2200 “C; time . . . . 5 s The argon flow used to purge the tube during a run was set at 30 units on the meter (300 ml min-1) with a flow stop time of 3 s at atomisation. All samples and standards were acidified with 10 ml of concentrated nitric acid per litre of sample. Concentration by Evaporation Procedure A 200-ml volume of acidified sample was transferred into a beaker then concentrated to about 20 ml on a hot-plate before transfer into a 25-ml calibrated flask and dilution to the mark giving an 8-fold concentration. The lead level was measured at 283.3 nm by flame AAS on a second instrument (IL 151) with background correction from a deuterium hollow-cathode lamp.Extraction of APDC Complex Procedure The method employed was essentially that described in “Methods for the Examination of Water and Associated Materials.”7 One modification was introduced to avoid having to aspirate isobutyl methyl ketone into the AAS instrument. Lead was back-extracted from the solvent into 25 ml of 50% nitric acid then determined by flame AAS. Excellent recoveries were obtained for standards treated in this way and for spiked water supply samples as can be seen from Table 2. The APDC method was used as an independent comparison to test if results from the furnace gave statistically acceptable values for lead. Fig. 1 shows a calibration graph corrected for the blank due to trace amounts of lead in the nitric acid.The blank was equal to about 1-2 pg 1-1 in the samples. No scale expansion was employed during the runs; typically a 100 pg 1-1 lead standard gave a peak height of about 120 mm which was equivalent to an absorbance of 0.5. The calibration in this instance included five points in the 1-10 yg 1-1 range and the linearity is excellent (correlation coefficient = 1.0) with the line passing through the origin. Some performance characteristics are given below. Substance Type of sample Calibration graph . . Linear to at least 100 pg 1-1 Standard deviation (within batch) determined . . Lead . . Tested on raw and treated water . . The following are pooled estimates: 15pgl-1 1.54pgl-1 13 D.F.* 39 2.02 45 D.F. 126 5.81 28 D.F.Estimated at 3.2 pg 1-1 (17 D.F.) . . Typically less than 5 pg 1-1 or 10% of Criterion ofdetection Bias . . . . the concentration (see Table 3) In practice the calibration was defined in each run using six standards in the range 1C100 pg 1-1 inclusive with the result that any between-batch random errors caused by calibration were minimised. The within-batch standard deviation was calculated by pooling estimates of standard deviation obtained from 15 runs over a period of 2 months to give the values reported above. The criterion of detection was calculated from replicate analyses of a blank solution consisting of acidified distilled water as used in the preparation of standards. The value of 3.2 yg 1-1 was calculated from 2.33SW for the blank.Tests for Bias Bias was tested from the difference between the “true” result as measured by the standard additions method and that obtained from direct analysis using a lanthanum-impregnated tube. Confidence limits were calculated on the result obtained by direct analysis and used with the difference between the results from the two procedures to see if this exceeded 10% of the concentration found by standard addition or 5 1-18 1-1, whichever was the greater. Table 3 presents some of the results obtained. Only the result from supply sample B was unacceptable and this was associated with a higher than usual standard deviation attributed at the time to a tube nearing the end of its useful * Degrees of freedom 1257 ANALYST OCTOBER 1984 VOL. 109 Table 3.Test for bias by standard additions Sample Raw Fairmilehead (spiked) . i Tap water A . . . . . . . . B . . . . . . . . c . . . . . . . . D . . . . . . . . Raw Alnwickhill (spiked) . . . . * With 90% confidence limits. “True” result from standard additions 81 .O 96.0 133 61 40 91 Direct electrothermal analysis 80.8 94.5 134.5 54.2 36.8 88.8 Difference* 0.17 k 0.62 1.5 k 0.45 1.5 k 0.69 6.8 z t 1.21 3.2 k 0.34 2.2 k 0.62 Max. allowable difference 8.1 9.6 13.3 6.1 5.0 9.1 120 I- I 1 00 2 80 ET) W .-$ 6 0 W 0 40 20 20 40 60 80 100 [Leadl/pg I-’ Fig. 1. Lead calibration graph life. Experience indicated that a treated tube could be used for about 200 cycles before deterioration became evident by loss of precision.Examination of the tube suggested that the trouble was due to the build-up of deposits within the tube, perhaps resulting in occlusion of the sample and subsequent difficulty at the atomisation stage. In general however the results obtained by direct determi-nation are not significantly different from those by standard additions and with one exception meet the requirements of a bias of less than 10% of the concentration or 5 pg 1-1. In view of the acceptable results obtained from these tests a detailed comparison was made of results obtained by the concentration by evaporation and flame AAS procedure used routinely by the laboratory and direct electrothermal analysis. Comparison with Results from Concentration by Evaporation -Flame AAS During the project measurements were made on many of the samples routinely received by the laboratory so that a comparison could be made between the two procedures.Over a period of 2 months 86 pairs of results were obtained in this way. The differences between the lead values found by each procedure were statistically compared on the assumption that both procedures should yield the same result. The actual differences should therefore in theory be normally distributed with a mean of zero and a test made as to whether the mean difference is significant. When all 86 results were considered and the calculated value for the t-statistic was compared with the tabulated value, there was a significant difference at the 95% confidence level (t = 5.9 compared with 1.96).However inspection of the data showed that early runs gave poorer agreement than later runs, possibly owing to greater experience with the electrothermal atomisation technique and increased care on the part of analysts in general. To test this later sets of results for two runs about a week apart were compared in the same way. The Table 4. Comparison of results of concentration by evaporation - AAS and electrothermal atomisation procedures Concen tration/pg 1 - 1 Tap water sample AAS A . . . . . . 32 B . . . . . . 134 c . . . . . . 97 D . . . . . . 5 E . . . . . . 93 F . . . . . . 23 G . . . . . . 5 H . . . . . . 10 I . . . . . . 32 J . . . . . . 32 K . . . . . . 51 L . . . . . . 46 M . . . . . . 42 N .. . . . . 139 0 . . . . . . 87 Number of samples (N) = 15 ZD= -15 Z 0 2 = 427 . (ZD)2=225 Electrothermal atomisation Difference (D) 40 131 89 13 86 26 11 7 31 32 53 38 36 136 84 +8 -3 -8 +8 -7 +3 +6 -3 -1 0 +2 -8 -6 -3 -3 (2DIN)- 0 t = = 0.71 S/ <N data contained 33 pairs and on this occasion gave a calculated f-statistic of 1.82 compared with the tabulated value of 2.01, confirming that more recent measurements by the two procedures were not statistically different. Table 4 shows a complete batch of results obtained by the two procedures from a routine survey of tap waters from various sources within the Region. The agreement is generally good and a t-test showed no statistically significant difference (t = 0.71 compared with 2.14 for 14 degrees of freedom).The most obvious discrepancy occurs for very low levels of lead in the 5-15 pg 1-1 range which were easily measured by the electrothermal atomisation procedure but were at the limit of detection for the flame AAS procedure even after concentra-tion of the samples. Closer examination of the latter procedure indicated a high noise level on the recorder trace (2-3%) with resulting uncertainty in drawing a base line. Background correction on the instrument (an IL 151) was poor owing to the condition of the deuterium lamp which required high currents when in use with a resulting increase in noise. The detection limit of the flame AAS procedure was estimated at 0.07 mg 1-1 so that the concentrations in many samples even after 8-fold concen-tration were too close to this for adequately precise results to be obtained 1258 ANALYST OCTOBER 1984 VOL.109 Table 5. Comparison of APDC and electrothermal atomisation results (vg 1- I ) using the test for non-homogeneous standard deviations Electrothermal APDC method atomisation method Standard Standard Mean deviation Mean deviation Tan 0* = G A ) (SA) ( i E ) ( S E ) SF E VSiA* + si; Sample Alnwickhill (spiked) . . . . Fair milehead (spiked) . . Tap water A B C D E F G H I 31.0 31 .O 45.0 158.2 47.2 164.8 37.2 130.0 87.2 76.7 95.8 5.33 3.52 2.19 6.50 12.07 24.02 4.60 6.68 8.87 6.41 5.88 35.3 29.1 51.0 134.5 54.2 172.3 36.8 122.2 82.5 80.8 94.5 1.93 0.93 4.05 0.84 1.47 1.37 0.41 0.98 0.55 0.75 0.55 2.77 3.79 0.54 8.50 8.98 19.3 12.33 8.33 17.71 8.52 10.7 1.85 1.26 3.19 8.10 1.28 0.70 0.18 2.33 1.18 1.58 0.55 * The values of tan 8 represent angles ranging from 70 to 87”.Critical values at the 5% level of significance and for 6 degrees of freedom are 60” = 2.435 75” = 2,440 and 90” = 2.447. rhe value of IXA- XFI/vSiA2+ S,F2 should therefore be less 2.4 for no statistically significant diff-erence between the means for the two procedures. Concentration by Extraction of APDC Complex Followed by AAS The modified APDC procedure described earlier was used on a number of samples over the period of the investigation.Each sample was analysed in replicate (usually five or six times) by both the APDC and electrothermal atomisation procedures. The mean values were then compared to see if the difference was statistically significant. Table 5 gives the results for 11 sets of results. Table 5 shows very high standard deviations for the APDC procedure but confirms the excellent precision of the electrothermal atomisation method. The marked difference in standard deviation invalidated the use of t-testing so the means were compared by a test for non-homogeneous variances.8 The ratio of the standard deviations for the two sample means (tan 6) was used to determine the significance of the difference between the means as summarised in Table 5 . Only one result was highly significant on this criterion with the APDC method giving a mean of 158.2 compared with 134.5 by the electrothermal atomisation procedure.The standard additions technique when used with the electrothermal atomisation procedure on this sample gave 133 pg 1-l so it seems likely that the APDC data were less accurate in this instance. In general however the direct electrothermal atomisation procedure gave results comparable to those obtained by the APDC method but with a much superior precision in many instances by a factor of 10 although this could probably be improved with more experience of the APDC procedure. Conclusions The work reported confirms that the use of lanthanum either to impregnate the tube as in this instance or added to solutions prior to the atomisation as reported by Bertenshaw et al.,5 can eliminate matrix interferences in the determina-tion of lead in potable waters.The precision and bias were acceptable and excellent recoveries were obtained from a range of raw and treated waters. The results obtained by direct electrothermal atomisation showed no statistically significant difference from those obtained by flame AAS after concentration. In addition, generally acceptable agreement was observed between the mean values obtained by replicate analysis by the electrother-mal atomisation and an APDC - AAS procedure. Waters from the Lothian Region are low in dissolved salts compared with supplies from many parts of the UK and may therefore be expected to suffer less from matrix effects. Such effects were nevertheless present when analysis was attempted using an untreated tube (e.g. only 72% recovery for 50 pg 1-l added to Pateshill water) and lanthanum treatment effectively improved the recovery figures giving a simple and rapid procedure for determining low levels of lead. The authors express their thanks to Miss Alison White and staff of the Water Services Laboratory for analytical assis-tance. Thanks are also due to Mr K. Richards Director of the Water and Drainage Department for permission to publish this paper. 1. 2. 3. 4. 5 . 6. 7. 8. References Regan J. G. T. and Warren J. Analyst 1978 103. 447. Hodges D. J. Analyst 1977 102 66. L’vov B. V. Spectrochim. Acta Part B 1978,33 153. Sturgeon R. E. Desaulniers J . A. H. Berman S. S . and Russell D. S. Anal. Chirn. Acta 1982 134 283. Bertenshaw M. P. Gelsthorpe D. and Wheatstone K. C., Analyst 1981 106 23. Thompson K. C. and Reynolds R. J. “Atomic Absorption, Fluorescence and Flame Emission Spectrophotometry,” Charles Griffin London 1978 p. 231. “Lead in Potable Waters by Atomic Absorption Spectropho-tometry,” Methods for the Examination of Waters and Associated Materials HM Stationery Office London 1976. Kennedy T. B . . and Neville A. M. “Basic Statistical Methods for Engineers and Scientists,” Second Edition Harper Interna-tional New York 1976 p. 213. Paper A4185 Received February 29th 1984 Accepted May 4th 198

ISSN:0003-2654    

DOI:10.1039/AN9840901255

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST OCTOBER 1984 VOL. 109 1259 Collaborative Study of a Graphite-furnace Atomic-absorption Screening Method for the Determination of Lead in Infant Formulas* Robert W. Dabeka Food Research Division Food Directorate Health Protection Branch Health and Welfare Canada Ottawa, Canada KIA OL2 Blind duplicates of three infant formulas and one evaporated milk were analysed by nine laboratories. Lead levels in the samples ranged from 29 to 200 ng 9-1. Only one of the laboratories obtained good agreement with reference values for all samples; two laboratories were outliers. Analysis of the data indicated that the remaining laboratories had instrumentation problems the main cause of which was inadequate simul-taneous background correction. The mean over-all sample reproducibility (relative standard deviation; RSD) was 66.8% for the data as submitted and 18.7% after adjustment for background correction.The mean precision between blind duplicates was 38.9% (RSD) for the submitted data and 6.6% after adjustment and rejection of statistical outliers. The results imply that for matrices and elements for which moderate (0.2 absorbance unit) non-specific background can be expected most laboratories with conventional background correctors will have difficulty in obtaining accurate results using graphite-furnace atomic-absorption spectrometry. Keywords Lead determination; graphite-furnace atomic-absorption spectrometry; infant formulas; collaborative study A study conducted by the International Atomic Energy Agency for the inter-laboratory determination of trace ele-ments in a single lot of powdered milk' revealed that the world status of methodology for lead in milk is very poor.Twelve laboratories reported lead values for the sample ranging from 0.017 to 246 pg g-1 a range of over four orders of magnitude. The toxicological importance of lead has prompted several countries to promulgate regulations specifying the maximum allowable level of lead in canned milks and infant formulas. As a result many samples require accurate and precise analyses for lead. As compliance methods are usually time consuming and expensive for regulatory laboratories rapid screening methods which can considerably reduce the number of unnecessary compliance analyses provide an inexpensive means of assessing the approximate levels of lead in formulas, whilst simultaneously increasing the number of samples that can be analysed.A collaborative study of one such graphite-furnace atomic-absorption spectrometric method,2 which has the advantages of using a single unwashed polystyrene test-tube for each sample and inexpensive reagents is described in this paper. The purpose of the collaborative study was to evaluate the performance of the method in different laboratories. This paper analyses the results of the study detailing the instrumental difficulties encountered when graphite-furnace atomic-absorption spectrometric methods progress from one laboratory to many. Experimental Design of Study Four samples were chosen one evaporated milk [2% butter fat (B.F.)] one milk-free formula (concentrated liquid) and two milk-base formulas (ready-to-use).The lead levels in all samples were background levels ranging from 29 to 200 ng g-1 and no samples were spiked. All samples had been purchased commercially and were in lead-soldered cans. Some had been stored in the laboratory for several years, explaining the high lead levels. * Collaborators G. R. Cutris G. G. Durany F. Estiandan D. A. Gooden J.-P. Hanchay L. J . Hart C. D. Howes P. A. Keller M. J. Lingenfelser L. H. Marion R. A. Moffitt L. I. Rivera K. K. Rousey T. L. Shannon M. K. Tikkanen N. G. Webb and R. W. Woods The lead levels in the formulas were determined using a reference method3 and the rapid screening method being tested. The latter method was also used to test sample homogeneity which was about +5% of the confirmed lead levels for all samples for the sample sizes sent to collaborators.Samples (blind duplicates randomly numbered) and instruc-tions were sent to 14 collaborators. Samples were sent in their digestion tubes. No practice sample was included in the study. In addition each collaborator received empty unwashed test-tubes for preparation of standards and reagent blanks a bottle of modification solution (0.5% citric acid 0.5% hydrogen perQxide and 0.25% ammonium dihydrogen phos-phate) two washed pipettes and two lead stock standard solutions (0.2 and 0.4 yg ml-1) for preparation of working standards. The stock standards had been prepared from US National Bureau of Standards certified lead nitrate and were in 1% nitric acid.Collaborators had to supply their own concentrated nitric acid (about 30 ml in total) for digestions and skim milk for standards. It was requested that the skim milk be purchased locally from a retail outlet and the concentrated nitric acid contain less than 10 ng g-1 of lead. Procedure To 2 ml of ready-to-use or 1 ml of concentrated-liquid formula was added 1 ml of concentrated nitric acid. The test-tubes were tightly capped and heated at 58-63 "C for 6-16 h in a water-bath. For reagent blanks two test-tubes containing 1 ml of water and 1 ml of concentrated nitric acid were digested as above. For standards and standard blanks seven test-tubes containing 2 ml of skim milk were digested as above. After digestion four of the skirn-milk digests were spiked with two lead stock standard solutions to yield final lead concentrations of 10 20 30 and 40 ng ml-1.All sample standard and blank digests were then diluted to 10 ml with modification solution, and the tubes were capped with their original caps and shaken briefly. Determinations were made using graphite-furnace atomic-absorption spectrometry. The instrument conditions reques-ted were as follows aliquot volume 10 pl for the HGA-400, HGA-500 HGA-2200 IL-455 and Hitachi 170-70 20 pl for the HGA-2100 and HGA-2000 and 2 or 5 pl for the CRA-90 and CRA-63 instruments; drying temperatures adjusted S ANALYST OCTOBER 1984 VOL. 109 1260 Table 1. Sample description and collaborator results (ng g-l) as submitted Laboratory No. and valuest Sample description Evaporated milk Conc.liquid Ready-to-use Ready- to-use (2% B.F.) soya base formula milk base formula A milk base formula B O$ 1 57 48.1 50.9 200 246 244 53 66.8 59.6 29 27.4 31.7 lA§ 44.2 42.3 199 184 48.9 48.9 23.9 26.8 2 62.0 59.3 139 140 39.4 37.0 8.9 8.9 3 44.4 31.5 168 187 42.3 28.8 10.6 10.6 4 25 .O 3.7 236.8 62.5 65.9 20.2 22.1 -5 29.4 59.1 85.0 101.7 22.5 20.1 11.9 18.6 6 265 167 28 1 308 93 100 78 75 7 286.8 36.3 194.0 189.5 88.3 83.5 17.9 16.2 8 34.4 59.9 440.3 296.6 79.4 44.2 4.7 44.2 * Numbers in parentheses were submitted to collaborators. t Blind duplicate values for each sample.$ Reference value from this laboratory. § Values obtained by laboratory 1 using peak-area measurements. These values were not included in the statistical evaluation. 9 105 109 282 277 64 79 9 34 that 2- 5 10- and 20-pl samples dry in about 5,8,12 and 20 s, respectively; ashing temperature and time 650 "C for 20 s; recorder mode continuous to monitor background noise during the atomisation cycle; background correction simul-taneous; gas-interrupt mode if available used during the atomisation step; graphite tubes uncoated if available; measurements peak height; and other parameters manufac-turers' specifications. Three replicate determinations were made for each solu-tion. The measurement sequence was fixed in the following general pattern 20 ng ml-1 standard - all blanks - all stan-dards - all samples (any order) - all standards.Results and Discussion Raw Data Nine laboratories participated in the study. The results as submitted are listed in Table 1. Only laboratory 1 which analysed the samples using both peak-height and peak-area measurements obtained good agreement with reference values for all samples (Table 1). Most of the other laboratories obtained values for some samples that appeared reasonably acceptable; however the results were generally poor. Laboratory 4 reported that sample 3 was broken during shipping and that losses were suspected during the digestion for samples 1 and 2. The value for the lost sample was set at 237 ng g-l that obtained for the blind duplicate and was used in all subsequent statistical manipulations.Identification of Problems The raw data indicated that most of the collaborators experienced problems with the method. Two possible reasons for this were considered contamination and instrumentation. As the method was simple and there was little chance of contamination and as most collaborators obtained good agreement between blind duplicates contamination was excluded as the prime factor influencing the quality of results. Instrumentation was investigated as the second reason for the poor results and subsequent examination of the method using a new atomic-absorption spectrometer (Varian Model 775-ABQ) in our laboratory revealed that inadequate simul-taneous background correction could cause accuracy prob-lems.The non-specific background at the wavelength of 283.3 nm was about 0.13 absorbance unit using an ashing tempera-ture of 650 "C. At the 217.0-nm line the background was a factor of 2 greater and using the Varian instrument at this wavelength could not be adequately corrected. This problem was not detected when the method was first developed because for the spectrometer used (Perkin-Elmer Model 403) only the wavelength of 283.3 nm recommended by the manufacturer was used and extensive care was always taken to assure proper lamp alignment for background correction. Table 2. Adjusted values (ng g-'). A constant was added to all the reported solution concentrations so that a mean value of 29 ng g- * was obtained for samples 7 and 8 Sample No. 1 2 3 4 5 6 7 8 Ref.ralue* 57 57 200 200 53 53 29 29 1 47 50 245 243 66 59 27 31 2 101 98 198 178 60 57 29 29 Laboratoryt 3 4 6 67 19§ 75 204 251 189 223 251 217 61 70 45 47 74 53 29 28 30 29 30 28 80 408 1731 7 3107 59 218 212 iooy 957 30 28 9$ 95 98 208 203 55 65 227l 367 * Composite using rapid screening method and extraction method.3 t Results from laboratories 5 and 8 excluded owing to poor $ Data based on linear least-squares fit for standards as well as Q Laboratory reported possibility of losses during digestion. 7 Statistical outliers at 95% confidence level (Dixon or Cochran instrument performance. adjustment. test). On examining the recorder tracings submitted by six of the nine laboratories it became obvious that inadequate back-ground correction revealed by the presence of negative peaks or positive peaks or shoulders that occurred during atomisation and could not be attributed to lead could indeed be one of the factors affecting the quality of the results.Assisting in the evaluation was the observation of differences between the time - absorbance profiles of blanks for standards containing skim milk and reagents without milk. Adjustment of Data for Background Correction The possible effects of background on the data was numeri-cally evaluated by assuming that for any one laboratory uncorrected background absorption was identical for all sample solutions regardless of tube age and that any uncorrected background absorption did not affect the calibra-tion slope.If these approximations are correct and if background is the main factor causing the poor results it should be possible to improve the results by adjusting for each collaborator the reported solution concentrations by a constant factor so that the mean value for one of the samples equals the reference value. The sample chosen as a reference for this adjustment was ready-to-use milk-base formula B and its value was set at 29 ng 8-1. For laboratory 9 a calibration problem was detected and prior to the background adjust-ment a linear least-squares treatment of the calibration graph was made. In spite of the crudeness of the approximations made the results of the adjustments (Table 2) showed a significan ANALYST OCTOBER 1984 VOL.109 1261 improvement for most of the laboratories supporting the hypothesis that incomplete background correction was the major cause of the poor inter-laboratory performance of the met hod. Laboratory 5 appeared to have a general accuracy problem, which revealed itself by erratic changes in sensitivity for some standards run before and after samples. The sensitivity increased 33% for the 10 ng ml-1 standard whereas for the 40 ng ml-1 standard it decreased by 8.7%. Based on linear least-squares fits for the two groups of standards (10 20 30 and 40 ng ml-1 in each group) the slopes of the calibration graphs before and after sample determinations changed by 22% and the intercept with the x-axis changed by 0.079 absorbance unit equivalent to a blank uncertainty range of about 6.3 ng ml-1.Based on the author's experience with the CRA-63 and CRA-90 atomisers the problem was probably due to unstable electrode - graphite tube contact inapprop-riate drying temperature or too great an aliquot volume (5 pl). Although 2-5 pl was the recommended aliquot volume for the furnace recent studies showed that 2 pl is the maximum aliquot volume that can be reliably used with the furnace.4 For laboratory 8 the instrument sensitivity was low and, after submission of results a weak lead electrodeless dis-charge lamp was reported. This may have caused the poor precision between blind replicates [mean 55% (RSD)]. Owing to the above difficulties laboratories 5 and 8 were treated as outliers.Statistical Evaluation of Data About 6 months after this collaborative study was run our laboratory organised a workshop in which seven analysts from five Field Laboratories of the Health Protection Branch, Health and Welfare Canada participated. One of the methods studied was the rapid screening method for lead in infant formulas and the analysts were given the same samples and standards that were used in this collaborative study. Each analyst performed the sample analysis independently, although solutions were analysed using the same well aligned instrument. Results for the workshop were good and no procedural homogeneity or contamination problems were encountered.5 The results are summarised in Table 3 together with the collaborative study data.For the adjusted data base (Table 2) six values were individual outliers at the 95% level of significance (Dixon and Cochran tests) and losses were suspected for two other samples. These values encompassed 14% of the data. In comparison for the workshop results only one value was a statistical outlier.5 The mean levels for all collaborators agreed reasonably well between workshop and collaborative study results (Table 3); the major difference between the two studies was that the precision for the latter was much poorer. The mean RSD was 3.8% for the workshop and 38.9% for the submitted values for the collaborative study. Adjustment of data and rejection of outliers and samples suspected of losses reduced the latter value to 6.6%. The mean over-all sample reproducibility (RSD) was 7.0% for the workshop and 66.8% for the collaborative study.The latter was reduced to 18.7% (range 11.1-30.2%) when the data were adjusted. Instrument Performance The collaborative study results demonstrated that accurate application of graphite-furnace atomisers is difficult to achieve on an inter-laboratory basis. As a fairly representative population of spectrometers atomisers and autosamplers were used (Table 4) the problems did not appear to be limited to any specific model or design of spectrometer or atomiser. A discussion of instrument performance diagnosed using preci-sion and background correction parameters is given to illustrate the nature of the problems encountered and the steps that can be taken to overcome them.Table 3. Statistical evaluation of data Parameter Mean levelshg g- . . . . . . . , Standard deviation between replicateshg g-1 . . RSD between replicates YO . . . . Reproducibility (standard deviation)/ng g-1 . . Reproducibility (RSD) % . . . . . . . . * Data include 1 outlier out of 56 values in total.5 Sample No. . . 1 2 3 4 1 2 3 4 . . 1 2 3 4 1-4 1 2 3 4 . . 1 2 3 4 1-4 . . . . Data base Works hop values" 60.0 215.2 57.6 29.3 1.8 7.0 2.5 1.3 3.0 3.3 4.3 4.6 3.8 3.6 14.6 2.6 3.1 6.1 6.8 4.5 10.5 7.0 Submitted values (all labs.) 82.0 225.2 59.8 25.0 64.4 35.0 10.0 11.2 78.4 15.6 16.7 44.9 38.9 80.7 87.9 21.9 21.6 98.4 39.0 42.4 87.5 66.8 Adjusted valuest 77.0 217.1 59.3 29.O t 5.0 10.7 6.0 1.5 6.2 4.9 10.1 5.3 6.6 24.0 23.9 8.9 1.1 30.2 11.0 15.0 3.7 18.78 t Values adjusted according to Table 2. Laboratories 5 and 8 not included. Replicate pairs containing a statistical outlier and samples for $ As a result of adjustment values meaningless for comparative purposes. § Value does not include sample 4. which losses were suspected were not included in the calculations 1262 ANALYST OCTOBER 1984. VOL. 109 Table 4. Instrumentation and performance characteristics Adjust- Analyst Instrument Mean RSD ment Stan- experience Aliquot precision between made dard with GFAAS volume/ yl (mean % RSD) duplicates % for blank- (months) AA Observed back- reagent Lab.spectro- Fur- Auto- Recom- h/ Stan- Sub- Adjust- non-specific ground/ blank/ Lead No. meter nace sampler Used mended nm dards Sample mitted ed absorbance ngml-1 ng ml- I General detn. 1 IL-951 IL-551 IL-254 2 PE-503 HGA- AS-1 2100 3 PE- HGA- AS-40 5000 500 400 5 Varian CRA- ASD-53 4 PE-603 HGA- AS-1 475 90 6 PE-306 HGA- AS-1 2100 305B 2100 5000 500 7 PE- HGA- AS-1 8 PE- HGA- None 9 PE-306 HGA- None 10 Varian HGA- AS-1 2100 775 400 20 10 20 20 10 10 10 10 5 2-5 20 20 20 20 20 10 20 20 10 10 283.3 283.3 283.3 217.0 217.0 283.3 283.3 217.0 283.3 283.3 2.8 5.2 5.8 4.0 1.6 11.4 2.0 2.3 9.1 8.9 14.6 6.6 6.6 17.7 28.8 3.1 9.7 12.6 24.8 -2.3 2.6 11.6 6.2 2.2 1.5 30.6 2.4 9.8 12.1 55.1 -3.8 14.9 25.2 3.7 1.3 2.4 3.8 -0.6 Negative peaks -2.9 Positive peaks 7.8 Positive peaks - 1.6 Negative peaks -Positive peaks 16.1 Positive peaks 3.9 -- -4.5 0.0 --0.3 1.1 -1.8 -0.5 2.0 5.6 0.2 -1.7 4.3 -6-12 6-12 (6 ( 6 6-12 6-12 >12 >12 6-12 6-12 >12 >12 >12 >12 >12 >12 <6 (6 - -For the lead levels determined in the collaborative study all the instruments used when properly optimised should yield an instrumental precision better than 5% (RSD).Failure to achieve this is indicative of problems. For collaborators, instrument precision for samples [range 1.5-17.7% (RSD)] and standards [range 1.&9.8% (RSD)] varied widely. The mean precision for the standards [5.3% (RSD)] was better than for the samples [9.7% (RSD)] and both were high in comparison with the respective workshop values of 1.3 and 2.4% (RSD).While laboratories 1 6 and 7 obtained good precisions for both samples and standards the precision for laboratories 3 5 and 8 was generally poor (Table 4). The causes of poor general precision are traceable to inadequate background correction too large an aliquot volume and instrument noise (caused by faulty lamps. electronic noise and poor optical throughput). For laboratories 2. 4 and 9 the precision for standards was reasonable [range 1 .&6.6% (RSD)] however for samples the RSD was an average of 4.6 times higher. This effect is indicative of problems with furnace parameters (drying temperature aliquot volume) and to a smaller extent with background correction deficiencies.The ranges of aliquot volumes recommended by manufac-turers for their furnaces usually include aliquot volumes too large to be handled accurately on a routine basis by any but highly skilled analysts. Large aliquot volumes can cause precision and accuracy problems owing to distribution effects within the furnace4 (and for this method significant back-ground absorbance). As no publications have appeared on this topic less experienced analysts tend to use aliquot volumes that are unsuitable for their furnaces in order to improve the detection limits. Although the author attempted to overcome this potential problem by specifying aliquot volumes for each model of furnace used his recommendations of 20 pl for the HGA-2100 and 5 1.11 for the CRA-63 and CRA-90 were in retrospect too large for the furnaces.This may have had a significant impact on the collaborative study results because the above furnaces were used by half of the collaborators. Regarding simultaneous background correction the adjust-ments required for correction of inaccuracy due to non-specific absorbance varied from -2.9 to 16.1 ng ml-*, indicative of both positive and negative background effects (Table 4). The submitted recorder tracings were ineffective in determining the sign of the background effect. While the effects of wavelength and aliquot volume on background correction were discussed proper lamp alignment is just as important a parameter. For example laboratory 3 using an appropriate wavelength and aliquot volume still required a background adjustment of 7.8 ng ml-l equivalent to a 78 ng g-1 error for a 1-g sample.The most probable reason for this was improper alignment of the deuterium and lead lamps, the beams of which should be coincidental with respect to size and shape as they pass through the furnace. Because for the specific spectrometer model the deuterium lamp must be aligned by the manufacturer and the instrument allows little flexibility with respect to lead lamp alignment owing to the type of lamp mount the analyst can do little to ensure proper background correction. All other spectrometers used by collaborators do allow analysts to make proper lamp align-ment. Conclusions The collaborative study of the method was a failure primarily owing to inadequate simultaneous background correction on the instruments used.This was substantiated by (a) significant improvement of the data adjusted for background and (b) by the comparatively good results obtained by seven analysts using the same samples and standards in a workshop situation where only one well aligned instrument was used. The importance of the results however extends beyond this particular method. The non-specific background for other elements and matrices often exceeds that for lead in milk. For instance the background for cadmium at 228.8 nm is roughly twice that for lead at 283.3 nm for most of the food matrices studied in our laboratory and even after dilution factors of 10, has been observed to exceed 1 absorbance unit.Thus it can be predicted that no graphite-furnace cadmium method applic-able to all foods is likely to achieve good inter-laboratory accuracy unless more powerful background-correction systems such as Zeeman or Smith - Heifje are used or the method includes the separation of cadmium from the bulk of the inorganic matrix. The implications of this for future collaborative studies are multiple relating both to the choice of methods and to verification of the adequacy of background correction. For situations when background correction is applied bu ANALYST OCTOBER 1984 VOL. 109 1263 not tested for adequacy organisers of collaborative studies would be wise to choose methods for which the maximum non-specific background will not exceed roughly 0.05 absor-bance unit a level significantly smaller than that which can be conveniently corrected with well optimised instrumentation.This approach would ensure wide applicability of the method without the analyst spending excessive time optimising the background corrector of the spectrometer. Once a method has been chosen samples giving the highest background levels that can be expected for the method should be included in the study. For collaborative study of methods so convenient that meticulous instrument adjustment for compensation of moderate (0.1-0.5 absorbance unit) background is warranted, it is necessary to include as part of the method a method for ensuring adequacy of background correction. Unfortunately, there is currently no published graphite-furnace atomic-absorption method available whereby the analyst can guaran-tee that non-specific background is being corrected by the spectrometer.Approximations can be made however and the simplest way to do this would be to include again as part of the method analysis of a practice sample. Such a practice sample to be valid and useful would be unlike those commonly suggested for collaborative studies for several reasons. Firstly it would have to be accessible to all analysts wishing to use the method i . e . it would have to be a standard reference material such as those supplied by the US National Bureau of Standards. Secondly the level of analyte in the sample would have to be qertified and below or near the detection limit of the method.Thirdly the level of back-ground for the sample after digestion etc. would have to exceed that of the “worst” sample to which the method would be applied. Fourthly the matrix composition of the test sample would have roughly to approximate that of real samples to avoid unexpected interferences and to match the source of the background. Using the method of interest the sample is analysed twice once unspiked and once spiked to give a signal roughly 10 times the detection limit of instrument-ation. If the results for both the spiked and unspiked sample correspond to the true levels it can be assumed that the instrument as aligned can adequately correct for the back-ground of unknown samples. Gratitude is expressed to the following for their participation in the collaborative study G. G. Durany and M. J. Lingen-felser Hunt-Wesson Foods Fullerton CA; L. H. Marion and J.-P. Hanchay Health Protection Branch Montreal Quebec; N. G. Webb Florida Dept. of Agriculture Tallahassee FL; R. A. Moffitt F. Estiandan and L. I. Rivera Carnation Research Laboratory Van Nuys CA; P. A. Keller Louisiana State University Experimental Station Boston Rouge LA; C. D. Howes G. R. Curtis and D. A. Gooden Lorna Linda Foods Mt. Vernon OH; M. W. Tikkanen State of New Mexico Scientific Laboratory Division Albuquerque NM; L. J. Hart and K. K. Rousey Mead Johnson & Co., Evansville IN; and R. W. Woods and T. L. Shannon US Department of Agriculture Athens GA. 1. 2. 3. 4. 5 . References Dybczynski R. Veglia A. and Suschny O. “Report on the Intercomparison Run A-1 1 for the Determination of Inorganic Constituents of Milk Powder,” IAEA/RL/68 International Atomic Energy Agency Vienna 1980. Dabeka R. W . “Rapid-screening Determination of Lead in Canned Milks and Infant Formulas Using Graphite-furnace Atomic-absorption Spectrometry,” Procedure LPFC-114, Health Protection Branch Laboratory Ottawa Canada 1980. Dabeka R. W . Anal. Chem. 1979 52 902. Dabeka R. W. “Drying Temperature Studies Using Graphite-furnace Atomic-absorption Spectrometry,” 8th Annual Meet-ing Federation of Analytical Chemistry and Spectroscopy Societies Philadelphia PA Sept. 20-25 1981. Dabeka R. W. J . Assoc. Off. Anal. Chem. 1982 65 1005. Paper A41103 Received March 13th 1984 Accepted May 3rd 198

ISSN:0003-2654    

DOI:10.1039/AN9840901259

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST OCTOBER 1984 VOL. 109 1265 Low-pressure Evaporation Concentration Prior to Discrete Nebulisation Flame Spectroscopic Analysis Fiona A. Robertson Anthony C. Edwards and Malcolm S. Cresser University of Aberdeen Department of Soil Science Meston Walk Old Aberdeen AB9 2UE UK The use of low-pressure evaporation in a vacuum desiccator for small-volume sample solutions is described. The procedure utilises hydrophobic plastic AutoAnalyzer cups without heating which reduces deposition on the container walls the final volume is assessed by weighing and therefore no dilution is required no reagents are necessary and the system is closed to atmospheric contamination during the evaporation stage. Keywords Pre-concentration; evaporation; atomic-absorption spectroscopy; atomic-emission spectroscopy Numerous pre-concentration techniques have been employed in analytical atomic spectroscopy over the years.In flame spectroscopy where improved detectability is essential sol-vent extraction is most widely used.' However solvent extraction is not without problems for trace analysis. Substan-tial amounts of solvent and reagents of adequate purity are required and the solvents used are invariably flammable. Moreover a considerable input of skilled operator time is necessary unless an expensive automated system is available. Cation exchange is generally regarded as a less popular alternative to solvent extraction but purified reagents are still required and manual sample processing is still very time consuming. Also samples are unavoidably diluted upon elution unless the resin is dissolved or analysed as a solid.Concentration by evaporation has never been very popular. At elevated temperatures evaporation in glass vessels often leads to loss of sample by deposition or absorption on to the glass walls above the liquid surface. Evaporation is slow if the samples are covered but open samples are prone to contami-nation from the atmosphere. A slow air bleed through the samples at reduced pressure speeds up the evaporation but again may introduce contamination. This paper describes the use of low-pressure evaporation in a vacuum desiccator for small-volume sample solutions. Hydrophobic plastic AutoAnalyzer sample cups are used without heating to reduce deposition on the container walls. The final volume is assessed by weighing so no dilution is necessary.No reagents are required and the system is closed to atmospheric contamination during the evaporation stage. Experimental Apparatus A 240-mm i.d. vacuum desiccator was used with a sample holder made from a 182 x 106 mm rubber block containing five rows of ten 16-mm diameter holes to hold 50 Technicon AutoAnalyzer standard sample cups. When more uniform evaporation was required a 160-mm diameter disc sample cup holder was used. It was made from expanded polystyrene, with 24 equally spaced holes drilled out their centres being 64 mm from the disc centre Each desiccator was fitted with a sintered-glass filter in the extraction - air inlet line to prevent dust entrainment on releasing the vacuum.Procedure Pipette 1.5 ml of each sample or standard into a sample cup, and weigh the cup to +1 mg. Load the samples into a sample tray and place the tray in a vacuum desiccator over granular calcium chloride. Grease the lid thoroughly with vacuum grease close and evacuate using a water pump until there is no change in the sound of the water pump when the desiccator tap is closed. This generally takes 10-15 min. Close the tap and leave to stand until the required degree of evaporation is observed. For 24 samples in a normal sized vacuum desiccator, 10-fold concentration is achieved in 10-24 h depending on the efficiency of the water pump and the number of times the calcium chloride has already been used. Allow air to re-enter slowly via a filter. Weigh the sample cups and contents to 21 mg and calculate the mass of water evaporated and hence the volume of solution remaining.Determine the element of interest by flame atomic-absorption or -emission spectroscopy using the discrete nebulisation technique ,2 reading the absorbance or emission intensity peak from a suitable chart recorder. Use matrix-matched standards allowing for the concentration of matrix components if necessary or run suitable standards and blanks through the evaporation procedure. Calculate the concentra-tion factors based on the volumes of solution remaining and hence the amount of analyte in the original sample solutions. Results and Discussion Choice of Desiccant The net removal of water from the sample surface and hence the rate of concentration.depends on the rate at which water molecules escape from the surface of the sample solution the rate of diffusion of water molecules to the desiccant surface and the rate of absorption of water molecules by the desiccant. At normal atmospheric pressure five desiccants were studied viz. granular calcium chloride phosphorus pentox-ide concentrated sulphuric acid anhydrous calcium sulphate (freshly ignited) and silica gel. Four sample cups each containing 1.5 ml of distilled water were weighed and placed in each of a series of 75-mm i.d. desiccators containing the five desiccants. The sample cups and contents were re-weighed after 18 23 40 46 and 64 h with a view to comparing the trapping efficiencies of the five desiccants and to seeing if their efficiency deteriorated as they became wetter.The efficiency of the desiccators decreased in the order sulphuric acid (most efficient) > calcium chloride > phos-phorus pentoxide = silica gel > calcium sulphate. Calcium chloride was used in subsequent work because it was only marginally inferior to sulphuric acid and is safer to use under vacuum. For calcium chloride and sulphuric acid the desic-cant efficiency remained constant as the desiccants became wetter. For the other three desiccants the absorption effi-ciency decreased with increasing time and wetness. Without evacuation the samples lost only about 500 yl each over 60 h. The evaporation rates were greatly accelerated at low pressure indicating that the evaporation of water was limited by the rate at which water molecules escaped from the sample solution surface and/or their rate of diffusion to the desiccant, as well as the trapping efficiency of the desiccant 1266 ANALYST OCTOBER 1984 VOL.109 0.1 pg ml-1 M I Time 4 I Time -b Fig. 2. Chart recordings for ( a ) the direct determination of copper in dilute hydrochloric acid extracts of five plant ash samples; and ( b ) for the same solutions after evaporation pre-concentration Fig. 1. Volumes (PI) remaining after evaporation from 1.5-ml sub-samples. ( a ) SO water samples after 8 h ; ( b ) 19 dilute hydrochloric acid extracts of plant ash after 8 h. Values obtained depend on pump efficiency and prior use of desiccant Evaporation Rates at Reduced Pressure Using a rectangular block containing five rows of ten sample cups the rate of evaporation was very uneven as shown in Fig.l(a). For 1 S-ml samples the evaporation was greatest at the block corners and lowest at the centre of the block. If a more uniform evaporation rate is required a symmetrical, circular pattern of sample cups is necessary. A 162-mm disc containing 24 equally spaced cups around its circumference gave a uniform 10-fold concentration over 10-24 h depending on pump efficiency and desiccant age. The desiccator should not be left permanently connected to the water pump once a suitable low pressure has been attained because the water vapour pressure in the desiccator does not then fall below the saturation vapour pressure at the operating temperature of the pump and the evaporation rate is thus reduced.It should be noted that if the disc sample holder is not completely filled the evaporation rate is no longer uniform from sample to sample. The pattern shown in Fig. l(b) for 0.06 M hydrochloric acid solutions from ashed plant samples for a batch of 19 samples is typical. Examples of Applications The technique has been applied successfully to the atomic-absorption determination of iron and manganese in natural fresh water samples. For iron contamination problems were encountered' initially. These were traced to minute but significant inputs of dust on releasing the vacuum. Using the rectangular sample block contamination was greatest at the centre of the block immediately under the desiccator tap and lowest but still a problem.at the edges. The effect was eliminated by placing a fine glass sinter in the air input line and slowing down the rate of vacuum release. Reduced-pressure evaporation has also been applied to the determination of copper in 0.06 M hydrochloric acid solutions from 6 M acid extracts of ashed plant samples. Typical results with and without the evaporation concentration step are shown in Fig. 2. Copper was determined by atomic-absorption spectroscopy using a Baird A3400 spectrometer and an air -acetylene flame. The improved detectability and precision are immediately apparent. A 10-fold concentration of ten repli-cate 1.5-ml portions of a single hydrochloric acid digest gave a result of 0.080 pg ml-1 for copper in the ash extract with a relative standard deviation of 1.9%.Conclusions Evaporation concentration in a vacuum desiccator provides a simple procedure for up to 10-fold pre-concentration of analyte elements prior to determination by discrete nebulisa-tion flame atomic-absorption or -emission spectroscopy. Because the system is closed and addition of reagents is generally unnecessary contamination is minimal. Because no heating is used inexpensive hydrophobic plastic sample cups may be employed and solid deposition on to the evaporation vessel walls which in our experience may be a problem with heated glass systems appears to be negligible. The conical shape of the bottom of the Technicon sample cups is ideal for discrete nebulisation of 100-200-p1 samples ANALYST OCTOBER 1984 VOL. 109 1267 Although in this work the technique has been used prior to flame atomic-absorption spectroscopy it is also perfectly well suited to other techniques that may be successfully completed using small volumes of sample solution. If concentration factors greater than 10-fold are required a second 1.5-ml portion of sample may be added after the first evaporation and the evaporation repeated to achieve 20-fold concentration. The use of mass change to measure final volume is very precise and eliminates the need to dilute to a fixed. small volume. The authors are indebted to George Wilson and Jim Wallace for constructing the sample holders and to the Department of the Environment and Natural Environmental Research Coun-cil for financial support. References 1. 2. Cresser M. S . “Solvent Extraction in Flame Spectroscopic Analysis,” Butterworths London 1978. Fry R. C. Northway S. J. and Denton M. B . Anal. Chem., 1978 50. 1719. Paper A411 I7 Received March 23rd 1984 Accepted May Ilth 198

ISSN:0003-2654    

DOI:10.1039/AN9840901265

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, OCTOBER 1984, VOL. 109 1269 Combined Reagent Purification and Sample Dissolution (CORPAD) Applied to the Trace Analysis of Silicon, Silica and Quartz V. J. Phelan and R. J. W. Powell GEC Research Laboratories, Hirst Research Centre, Wemble y, Middlesex, HA9 7PP, UK A simple and reliable method is described for the determination of impurities in very high-purity single-crystal silicon and in siliceous materials of importance to the electronics industry. Reagent purification and sample dissolution take place simultaneously within a closed PTFE vessel; the analyses are completed by carbon furnace atomic-absorption spectrometry. The very low reagent blank levels enable aluminium, chromium, copper and iron t o be determined at levels as low as 5 ng g-1. Keywords: Silicon, silica and quartz analysis; reagent purification - sample dissolution; atomic-absorption spectrometry; trace metal determination The relevant desirable properties of some materials used in high-technology industries are seriously affected by trace metal impurities.The concentration above which an impurity has a deleterious effect on the behaviour of the material is often below the microgram per gram level. Three examples of such materials are given in Table 1 with indications of the impurity elements of interest, the range of interest and the properties affected by the impurity. The detection and determination of metals at the required level can be achieved by a variety of techniques such as mass spectrometry, neutron activation and carbon furnace atomic- absorption spectrometry (CFAAS).The first two techniques require complex and costly equipment that is not widely available. Mass spectrometry also requires standard samples of known concentration for calibration purposes. For iron determination, difficulty may be experienced because of the confusion between Si2+ and Fe+, both of which have ions with a mass number of 56. CFAAS provides adequate sensitivity for the elements of interest and in the work described here it has been the method of choice for the provision of quantitative results. It can be operated using either solid samples or solutions. The use of solid samples avoids the problem of impurities in the dissolving acid. However, it suffers from the necessity of weighing repeatedly and accurately small amounts of sample and the difficulty of calibrating the met5od.1 With siliceous materials there would also be the problem of releasing the impurities from a relatively large amount of refractory material in the furnace.In view of the above problems associated with the direct examination of solid samples, attention has been concentrated on solution techniques. For silicon or silica samples this approach has a particular advantage in that attack with hydrofluoric acid leads to the formation of silicon tetra- fluoride, which is volatile and conveniently leaves the impuri- ties in an acidic solution. Matrix effects are thus largely eliminated. The dissolution of samples does, in general, have a major drawback in that impurities in the dissolving acid contribute a blank to the determination. One method of achieving low reagent blank values is the separate purification of reagents by sub-boiling distillation followed by conventional dissolution. The use of this method has been reported by Stewart and Newton.2 However, we did not adopt this approach for several reasons: a platinum still was not available to us; the construction of a sub-boiling dissolution unit from PTFE would be at least as complex as the making of our equipment; there are numerous difficulties in handling and storing high-purity acids once they have been prepared3; and further clean space is required in which to carry out the dissolution procedure.The principle made use of in the procedures discussed here is that of isopiestic distillation in which vapour from concen- trated volatile acids is absorbed in high-purity water surround- ing the sample.The concentrated acid and sample are in separate vessels inside one sealed container. The transferred acid vapour effects the dissolution but does not contain the impurities present in the concentrated acid as these do not transfer with' the vapour. Experimental Apparatus The analyses reported here were carried out using a PTFE vessel based on a design by Wooley.4 The design was modified so that relatively large samples could be accommodated with minimum handling (i.e. , crushing, grinding and transferring of solutions). Figs. 1 and 2 show the vessel and the four sample cups that it contains. Features of the modified design are as follows. (i) For "as-received" samples up to 1 g can be accommo- dated.The relatively large volume of the sample cups allows each sample to be made up to 10 ml in situ. (ii) All four sample cups may be readily removed simul- taneously. (iii) Chamfers on the top edges of the cups reduce the risk of condensed acid running into them. Table 1. Effect of impurities on three materials used in high-technology industries Impurities of Range of interest/ Material Property affected relevance ngg-' Siliconsinglecrystal . . . . . . Electrical breakdown Fe 1-100 of devices High-purity synthetic Optical fibres quartzcrystal . . . . . . Radiation hardness A1 ,Fe 5- 100 (high-purity silica) . . . . Optical transmission Cu ,Cr,Ni ,Fe 5- 1001270 ANALYST. OCTOBER 1984, VOL. 109 Sample Preparation The silicon samples are received as slices of approximately 50 mm diameter (other common sizes are 75, 100 and 125 mm).These are cleaved into strips prior to cleaning. Samples for analysis as-received are washed with ethanol and then rinsed in de-ionised water. If bulk analysis is required the silicon is heavily etched using a mixture of hydrofluoric and nitric acids (acids of the highest purity available are used). The samples are thoroughly rinsed in de-ionised water before dissolution. This procedure was found to be unsatisfactory for bulk iron analysis. A commonly recognised cleaning procedure (RCA clean) for removing surface contamination employs a mixture of hydrochloric acid, hydrogen peroxide and de-ionised water.6.7 Cleaning is followed by thorough rinsing of the sample with de-ionised water.75 75 5 Quartz The quartz samples are received as blocks and are washed with ethanol prior to a light etch in hydrofluoric acid. This etch is followed by cleaning with hydrochloric acid, hydrogen per- oxide and water as described for silicon preparation. The sample is finally rinsed thoroughly with de-ionised water. Fig. 1. PTFE vessel and lid. All dimensions in millimetres 0 Hole tapped and rod threaded OBA Fig. 2. PTFE cradle and cup. All dimensions in millimetres (iv) Shoulders on the outside of the cups ensure that they do not lean against the wall of the vessel. (v) The four-cup design allows the dissolution of up to three samples and one blank or more commonly two samples and two blanks. The atomic-absorption measurements were obtained using a Perkin-Elmer Model 280 spectrometer fitted with an HGA 500 graphite furnace.The samples were introduced using an AS-40 autosampler. The spectrometer readout was recorded on a Perkin-Elmer Model 56 chart recorder. Standard conditions for each element were listed in the users' manual.5 The graphite tubes used were all supplied by Perkin-Elmer and no problem with residual blanks was encountered in any of the analyses reported. Optical Fibres These samples are received as collapsed tubes. The inner core is removed by trepanning and etched in hydrofluoric acid to remove any remaining outer tube and trepanning residues. The core is then cleaned in a mixture of hydrochloric acid, hydrogen peroxide and water and finally the sample is rinsed thoroughly in de-ionised water.The optical fibre, silicon or quartz crystal samples are weighed, cleaned and placed in the sample cups. The acids used for dissolution are placed in the bottom of the large vessel (1 + 1 V/V hydrofluoric acid - nitric acid for silicon; hydrofluoric acid for quartz and silica. Aristar grade acids are used). The four sample cups are then placed in the rack in the top of the vessel. The apparatus is sealed with the PTFE lid and the whole is clamped together between brass plates and placed in an oven at approximately 110 "C. The assembly is usually left in an oven overnight. When the vessel is removed from the oven it is spun about its axis on a turntable at approximately 30 rev min-1 for a few minutes to encourage any condensation on the roof of the vessel to run back into the bulk acid.The pot is cooled, the clamps are released and the pot is transferred to a clean environment before opening. Although silicon tetra- fluoride has been generated there is no build-up of pressure and the lid has to be gently prised open. There is some advantage in cooling below room temperature when silicon has been dissolved as this reduces the evolution of oxides of nitrogen from the bulk solution. All of the samples examined are siliceous and therefore most of the silicon is removed as a vapour during dissolution leaving a liquid residue containing only the fluorides and/or nitrates of the impurities associated with the sample. It was found that leaving the samples to dissolve for a longer time resulted in a larger liquid residue in the sample cups.To facilitate manipulation of the residues and remove any residual fluorides, which could cause interference problems in subsequent furnace analysis,g 50 pl of Aristar sulphuric acid are added to each cup. The cups are then heated at approximately 100 "C on a hot-plate. This leaves a residue of 50 p1 containing all the impurities associated with the samples. The liquid is then made up to volume (typically 1 ml) using a micropipette to dispense the de-ionised water (Elgastat Spectrum ROl), The samples are subsequently analysed against standards matched in acid concentration. The analyses of standards and samples were carried out in triplicate. The standard deviation (6) of results on 30-p.1ANALYST. OCTOBER 1984, VOL. 109 1271 Table 2. Impurities sought in single-crystal silicon.Sample mass. ca. 1 g; initial sample volume of 5% H2S04. 1 ml Impurityhg g- I Sample Fe Cr Ni Cu Pb As received from manufacturer 200* 7 10 15 20 Afterfirst HF-HN0,etch . . 40 40 10 10 AftersecondHF-HN0,etch . . 40 5 3 10 RCAcleanon2 . . . . . . 17 After heavy HF - HN03 etch followed by RCA clean on as-receivedsilicon . . . . 5 Blankassuming 1 gofsample . . 3 10 3 10 5 * Sample diluted 100-fold for Fe analysis. Table 3. Impurities sought in quartz crystals. Sample mass, ca. 1 g; initial sample volume of 5% H2S04, 1 ml Impurityhg g ~ I Quartz sample Fe Al Commercially available quartz (1) . . . . . . 100* 9000* Commercially available quartz (2) . . . . . . SOO* 11 000* Commercially available quartz (3) . . . . . . 100* 9000* Pure quartz A .. . . . . . . . . . . 50 30 Pure quartz B . . . . . . . . . . . . 5 10 Blanks assuming 1 g of sample . . . . . . 5 S *Samples diluted 100-fold for analysis. samples of a 0.01 p.p.m. iron solution was 0.0005 (based on ten replicate determinations). Using the same procedure the standard deviation for the analysis of replicate samples was similar. It has been possible to maintain the blank levels at 3-8 ng of iron per sample. A similar method of sample preparation has also been used to provide matrix-free samples for spark-source mass spec- trometry (SSMS). In this method the residue, or a portion of the residue, is dried on to graphite (Ringsdorff RW-A/Total), which is pelleted prior to analysis. SSMS is a semi-quantitative method of trace analysis which, in these applications.is limited by the large number of complex ions of different mass produced by silicon isotopes in combination with carbon and oxygen. Using residues from the present method of dissolu- tion, where silicon has been removed, allows more elements to be detected at lower levels than before. However, the sulphuric acid treatment has to be omitted from the procedure to prevent the formation of complex ions based on sulphur and oxygen. In the studies reported here, SSMS results have been used to show which elements should be measured quantitatively by atomic-absorption spectrometry. Results and Discussion Analysis of Silicon A high iron content is expected with the as-received slice owing to the unavoidable presence of surface impurities.This is well known in the semiconductor industry and it is standard practice in device technology to oxidise the silicon (approxi- mately 100 nm) and remove this layer by dip-etching in hydrofluoric acid. The initial results from the analysis of the silicon crystal slices confirmed the high iron content of as-received slices. In Table 4. Impurities sought in optical fibres. Sample size, 1 g; initial sample volume of 5% H2S04. 1 ml Impurityhg g-’ Sample Cr Ni Cu Fe FibreB118:end 1 . . . . . . 8 50 130* 70 end2 . . . . . . 7 SO* 140* 10 FibreB120 . . . . . . . . 5 10 10 140 Fibre B111 . . , . . , . . 10 5 160* 60 Commercial silica (used for preform tubing) . . 2 10 10 240 Blanks assuming 1 g of sample . . 2 5 2 10 * Samples diluted 10-fold for analysis. order to determine a bulk iron content for the silicon the samples were pre-etched in a mixture of hydrofluoric and nitric acids to remove the silicon surface.Although this procedure did lead to lower levels of iron being detected they were still higher than expected, possibly because iron can be plated back on to the fresh silicon surface. After heavy etching followed by an RCA clean the iron level was reduced to 5 ng g-1, a level close to but distinguishable from the current blank level. The results in Table 2 show that there is a high chromium content obtained after the first etch. As this was higher than both the “as-received” chromium content and the “subse- quently cleaned” chromium content it was thought that it was most likely to have been due to some particulate contamina- tion introduced during the cutting - polishing procedures.This is supported by other “as-received’’ analyses where samples from the same slice can give widely differing iron contents.9 Analysis of Quartz The results (Table 3) from the analysis of the two high-purity quartz crystals (A and B) showed that the iron and aluminium contents of these crystals are considerably lower than those found in commercially available material. With respect to aluminium these crystals are purer by nearly three orders of magnitude. This is of great significance as aluminium is known to be the most important impurity limiting the so-called “radiation hardness” of quartz crystals. 10 Analysis of Silica Optical Fibres The analysis of silica optical fibres is a requirement arising from the necessity for extremely low levels of potential colouring elements, to minimise attenuation of the optical signal transmitted along fibres that may be several kilometres in length.The results obtained (Table 4) have been of value in locating the presence of such impurities and in evaluating the degree of success of attempts to achieve lower impurity levels at various stages in the development of fibre-producing processes. Conclusion In this work the results obtained indicate that the immediate requirements in the determination of concentration of metal- lic impurities in high-purity silicon and silica have been met. With the quartz crystals and silica optical fibres, the limits of detection now attained are adequate for the foreseeable future.However, for silicon, advances in technology will probably require higher purities than are currently available and hence will generate a need for analytical methods with even lower limits of detection.1272 ANALYST, OCTOBER 1984, VOL. 109 Employing the dissolution process described, the detection limit is set by the blank value of the sulphuric acid used to remove residual fluoride. It is planned that improvement be sought by several different means. These include use of high-purity sulphuric acid, a larger sample size (up to 5 g could be used) with the same amount of sulphuric acid, or use a smaller volume of acid. With the success obtained in application of the isopiestic distillation combined with dissolution as a means of lowering blank levels on siliceous samples, it is intended to pursue similar techniques for other materials.Semiconductor tech- nology has generated a demand for high-purity materials such as gallium arsenide, indium phosphide and cadmium mercury telluride and there is thus a need for means of analysis. Whereas with these materials there will not be the advantage of loss of matrix effects from volatilisation of silicon tetra- fluoride, the ability to effect solution with very low reagent blanks will enable quantitative analysis to be carried out at lower levels than is easily possible at present. It is expected that the results of investigations with such materials will be the subject of a further paper. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Fuller, C . W., Anal. Chirn. Acta, 1972, 62, 261. Stewart, D. A., and Newton, D. C., Analyst, 1983, 108, 145. Moody, J. R., and Beary, E. S., Tulantu, 1982, 29, 1003. Wooley, J. F., Analyst, 1975, 100. 896. “Perkin-Elmer Users Manual for HGA 500 Graphite Fur- nace,” Perkin-Elmer, Norwalk, CT, 1978. Kern, W., and Puotinen, D. A., RCA Rev., 1970, 31, 187. Kern, W., RCA Rev., 1970, 31, 234. L’vov, B. V., Spectrochim. Actu, Part B , 1978, 33, 153. Ward, P. J . , J . Electrochem. Soc., 1982, 129, 2573. Martin, J . J . , Doherty. S. P., Halliburton, L. E., Marks, M., Koumvakalis, N . , Sibley, W. A.. Brown, R. N., and Arming- ton, A , , “Procccdings of the 33rd Annual Symposium on Frequency Control, 1979,” US Army Electronics Command, Fort Monmouth, N J , 1979, pp. 134147 (copies available from the Electronic Industries Association. 2001 Eye Street, N.W., Washington, DC, 2006. Paper A4178 Received February 23rd, I984 Accepted May 9th, 1984

ISSN:0003-2654    

DOI:10.1039/AN9840901269

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, OCTOBER 1984, VOL. 109 1273 Determination of the Flow-point of Lubricating Greases and Petroleum Waxes Including Petrolatum Dhoaib Al-Sammerrai, Hazim Al-Najjar and Wejdan D. Selim Petroleum Research Centre, P. 0. Box 10039, Jadiriyah, Baghdad, Iraq The flow-point has been determined for lubricating greases, petroleum waxes, including petrolatum, and protectives and the results obtained have been compared with the corresponding drop-points. The flow-point can be used as a characterisation parameter in the identification of these products. Keywords: Flow-point determination; lubricating greases; petroleum waxes It is important to be able to identify lubricating greases, petroleum waxes, including petrolatum, and protectives for the purpose of classifying them and quality control.Methods of identification of these products include the determination of their drop- and melting-points, which can be defined as the temperature at which the product passes from a solid or semi-solid to a liquid state under the conditions of the test. Many standardised methods are available for such determinations. 1-3 In the work described in this paper, the flow-point, which is usually defined as the temperature at which a plug of sample slides down to a complete halt from one leg into the bend of a specimen U-tube ,4 has been determined for lubricating greases, petroleum waxes, including petrolatum, and a protective. The results obtained have been compared with the corresponding drop-points, measured according to standard procedures. It was observed that the flow-point was of a slightly higher value than the drop-point and that it can be used as a characterisation parameter in the identification and classifica- tion of these products.Experimental The instrument used to obtain the flow-points is the Buchi Model 510 melting-point determination apparatus. The instru- ment consists essentially of two parts, the control unit and the apparatus. The control unit contains all operating elements and the electronic temperature controller with a control range of 0-300 "C. Control is effected by a proportional-action con troller. The U-shaped oil container, mounted on top of the control unit, is capable of accepting the specimen tubes. A specially designed stirrer ensures an intense circulation and mixing of the silicone oil used.This produces uniform temperature distribution around the specimen table and the thermometer. The glareless lighting and adjustable magnifying glass permits observation of the specimen with both eyes. For the temperature rise during the determinations, tem- perature programmes with heating rates of 0.2,0.5,1,2,3 and 5 "C min-1 are available. A thermometer (0-360 "C) is used for measuring the temperatures (see Fig. 1). The specimen glass U-tubes supplied with the Biichi 510 have the dimensions shown in Fig. 2. The specimen glass U-tubes were filled with lubricating greases by inserting one leg of the tube about 10 mm into the product to be tested and pushing the plug of product downwards with a filling wire until it is about 10 mm above the bend of the specimen tube.It was found from experience that repeated insertions are required in order to ensure that a sufficient amount of sample (ca. 15-20 mg) had been pushed into the specimen tube. It is also advisable to place the tube with the pushed-out plug of product into a refrigerator for a 3 - . 4 1 I I 1 I . I I J BUCHI 510 * . Irt Y , -- Fig. 1. Flow-point determination apparatus. 1 , Main switch; 2, switch for lighting and cooling fan; 3, selector switches for tempera- ture programme; 4, set-point potentiometer; 5 , oil container; 6, specimen table; 7, adjustable magnifying glass; 8, lighting; 9, heating coil; 10, stirrer motor; 11, holder for spare capillary tubes; 12, thermometer; 13, cooling fan; and 14, expansion vessel short time so that the plug can then be pushed downwards much easier and to avoid the formation of air bubbles.This cooling procedure does not cause any changes in the structure of the grease, as nowadays most lubricating greases are for multi-purpose use and can withstand very low temperatures. Petroleum waxes, including petrolatum, and protectives are added to the specimen tubes by inserting one leg of the tube about 10 mm into a hot melt of the product (approximately 10 "C above its drop-point). The U-tube is then heated on a hot-plate, tilted slowly until the product flows downwards to about 10 mm above the bend of the U-tube and the product is then allowed to solidify on cooling. If the flow-point is not known, an orientation determination is carried out by turning the set-point potentiometer to increase the oil temperature rapidly until the sample flows and the approximate temperature is immediately recorded by the thermometer.The set-point potentiometer is then adjusted to1274 ANALYST, OCTOBER 1984, VOL. 109 a temperature 10-15 "C below the approximate reading. The silicone oil is allowed to cool to the set temperature and a new specimen is placed in the determination apparatus. The desired temperature rise is then achieved by pressing the appropriate programme button. The U-tubes can be cleaned and used for more than one determination by flushing out the plug of grease sample with a strong fine jet of light petroleum followed by a jet of soapy water and finally acetone. As for wax samples, the U-tubes are heated on a hot-plate and inverted allowing the wax melt to flow out, the tubes are cleaned with a hot solvent (toluene), then acetone and dried. Results and Applications Flow-point determinations were all conducted by the same operator at a recommended5 temperature programme of 0.5 "C min-1, starting 10-15 "C below the approximate reading.No trial determination was run as the drop- and melting-points of all the samples were known. Five lubricating greases prepared from mineral-based oils and different thickeners were chosen for our investigation. The data shown in Table 1 are for a set of ten flow-point determinations for each sample. Means and standard devia- tions are presented. Betonite grease exhibited no flow and began to decompose at a temperature of about 250 "C, i.e., degradation of the oil liquid phase.The data shown in Table 2 are for a set of ten flow-point determinations on three petroleum-based waxes, a petrolatum and a protective grease. Means and standard deviations are presented. The results in Tables 1 and 2 show that the repeatability for samples with a wide range of drop- and melting-points is good. Fig. 2. millimetres ( a ) Specimen U-tube; and ( b ) filling wire. All dimensions in ~ ~~ Table 1. Flow-points for lubricating greases at a temperature programme of 0.5 "C min-I Flow-point */"C Lithium (198 "C) (no drop-point) hydroxystearate Betonite Calcium complex (290 "C) 292.5 292 292 292.5 292.5 293 292.5 292.5 292.5 293 292.5 Sodium (144 "C) 145.5 145 145 145 145 145.5 145 145 145.5 145 145.1 Determination No. 1 2 3 4 5 6 7 8 9 10 Mean .. . . . Standard deviation . . . Calcium (98 "C) 99 99 99 98.5 99 99 99 99 99 98.5 98.9 200 200 200 200.5 199.5 199.5 199.5 199.5 200 199.5 199.8 Decomp. 253 Decomp. 253 Decomp. 250 Decomp. 250 Decomp. 252 Decomp. 253 Decomp. 252 Decomp, 252 Decomp. 252 Decomp. 253 0.33 0.32 0.2 0.23 * The results in parentheses are the drop-points of the grease samples measured according to ASTM standard methods. Table 2. Flow-points for petroleum waxes, petrolatum and a protective grease at a temperature programme of 0.5 "C min-l Flow-point */"C Determination Microcrystalline Microcrystalline .'a-affin Petrolatum Protective No. wax (A) (73 "C) wax (B) (71 "C) wax (64 "C) (59 "C) (63 "C) 1 2 3 4 5 6 7 8 9 10 Mean . . Standard deviation 73.5 73.5 73.5 74 73.5 74 74 74 73.5 73.5 73.7 71.5 71.5 71.5 71.5 71.5 72 71.5 72 72 71.5 71.6 64.5 64.5 64.5 65 64.5 64.5 64.5 65 64.5 64.5 64.6 60 59.5 59.5 59.5 59.5 59.5 60 59.5 59.5 59.5 59.6 63.5 63.5 63.5 63.5 63.5 64 64 63.5 63.5 63.5 63.6 0.24 0.23 0.2 0.2 0.2 * The results in parentheses are melting-points for the samples measured according to ASTM standard methods.ANALYST, OCTOBER 1984, VOL.109 1275 The slight increase in the flow-point, over the drop- and melting-points is due to the time it takes the melt to flow from one leg of the specimen U-tube into the bend, and this will vary according to the type of product to be tested and the viscosity of its melt. Conclusions It has been shown that the flow-point of a product is slightly higher than its drop- or melting-point and that it can be used for the identification of lubricating greases and petroleum waxes, including petrolatum.However, the method is limited to a temperature of 300 “C when using silicone oil as the heating medium. A flow-point determination is a less time- consuming procedure than a drop- or melting-point determi- nation and requires only a minute amount of sample. For a lubricating grease exhibiting no drop-point, it is possible to record the temperature of decomposition of the product when using a flow-point determination apparatus. References 1. “Standard Method for Measurement of Dropping Point of Lubricating Greases over Wide Temperature Range, Designa- tion D-2265, 1982 Book of ASTM Standards,” American Society for Testing and Materials. Philadelphia, 1982. “Standard Method for Measurement of Dropping Point of Lubricating Greases, Designation D-566, 1982 Book of ASTM Standards,” American Society for Testing and Materials, Philadelphia, 1982. “Standard Method for Measurement of Drop Melting Point of Petroleum Wax including Petrolatum, Designation D-127,1982 Book of ASTM Standards.” American Society for Testing and Materials, Philadelphia, 1982. “Instruction Manual for Buchi 510 Melting Point Determina- tor,” Buchi Laboratoriums-Technik AG, Flawil, Switzerland. 2. 3. 4. 5 . Al-Sammerrai, D., unpublished work. Paper A41129 Received March 29th, 1984 Accepted May loth, 1984

ISSN:0003-2654    

DOI:10.1039/AN9840901273

出版商:RSC    

年代:1984

数据来源: RSC

摘要:

ANALYST, OCTOBER 1984, VOL. 109 1277 Analytical Uses of Charge-transfer Complexes: Determination of Pure and Dosage Forms of Piperazine" U. Muralikrishna Department of Chemistry, Andhra University, Postgraduate Extension Centre, Nuzvid 521 201, India and Mannam Krishnamurthy and N. Sorneswara Rao Analytical Chemistry Laboratories, School of Chemistry, Andhra University, Waltair 530 003, India The charge-transfer interaction of piperazine with benzoquinone and its halo derivatives was used for the simple spectrophotometric determination of piperazine in chloroform medium. The sensitivity of the method and the time stability of the charge-transfer complexes were found to depend on the electron affinity of the quinones. The drug-base in different dosage forms was extracted into chloroform after alkali treatment and the piperazine was assayed.The method is sensitive and rapid and the base tolerates a wide variation of reagent concentration. The results compare well with the official BP method. Keywords: Charge-transfer complexes; piperazine determination; benzoquinone; spectrophotometry; tablets Piperazine, a pyrazine derivative, is one of the most potent drugs and is used as an anthelmintic for the treatment of threadworms and roundworms in humans and animals. Sodium tetraphenylborate , I ammonium reineckate,l-3 bro- mothymol blue,4 Folin - Ciocalteu reagent,5 sodium 1,2- naphthoquinone-4-sulphonate ,677 2,5-dichloro- 1,4-naphtho- quinone ,8 1,4-benzoquinone ,9-11 dichloro- 1,4-benzo- quinone12 and 3,5-dichlorobenzoquinone chlorimidel3 have been proposed as reagents for the determination of piperazine and its formulations. However, these methods are either time consuming or indirect or suffer from the disadvantage of maintaining a narrow range of pH.The molecular interactions between electron donors and acceptors are generally associated with the formation of intensely coloured charge-transfer complexes,l4 which absorb radiation in the visible region. The photometric methods based on these interactions are usually simple and conve- nient,lj because of the rapid formation of the complexes. Piperazine is a good n-electron donor and will form charge- transfer complexes16~17 with organic x-acids such as benzo- quinones. These types of complexes are used in the determina- tion of the components of the complexes. 17-18 Belal et al.19 and Rizk et ~ 1 . 2 " reported the determination of piperazine using tetrachloro-l,4-benzoquinone as reagent, but in dioxane and butanol as solvents. Dioxane21 and alcohols22 are known to interact with quinones and their utility for quinones is stated to be non-ideal.23 In continuation of earlier work on the analytical uses of charge-transfer complexes,17,18.24-26 we have now examined the utility of benzoquinone and its halo derivatives as reagents for the determination of piperazine. Materials and Methods Apparatus A Carl Zeiss Spekol spectrophotometer was used for absor- bance measurements with matched, stoppered cells of 1-cm path length. Reagents A saturated solution of each quinone in chloroform was prepared from the purified samples. Piperazine (Fluka, puriss * Presented at the National Symposium on Absorption Spectrometry, Bhabha Atomic Research Centre, Bombay, India, February, 1984.grade sample) was recrystallised twice from cyclohexane solution and a 3 x M stock solution was prepared in chloroform. Procedure for Colour Development Place an aliquot, ranging from 0.3 to 4.0 ml, of piperazine in a 10-ml calibrated flask, add 2 ml of the reagent and dilute to the mark with chloroform. Measure the absorbance of the solution at the wavelength of maximum charge transfer (ACT) after the appropriate time indicated in Table 1. Calculate the unknowns from Beer's law graphs. The results are given in Table 1. Procedure for the Assay of Salts and Dosage Forms of Piperazine Salts and tablets Weigh an amount of the salt or powdered tablet containing 25 mg of piperazine into a suitable vessel and add 15-20 ml of distilled water.Shake, filter, wash with two 5-ml aliquots of water and combine the filtrate and washings. Add 10 ml of 10% sodium hydroxide solution. Extract the drug base with three 30-ml volumes of chloroform, transfer the combined extracts into a 100-ml calibrated flask and dilute to the mark with chloroform. Take an aliquot and develop the colour as described under Procedure for Colour Development. Solutions and syrups Dilute a volume of piperazine hydrate solution or syrup containing 25 mg of piperazine to 25-30 ml with water. Add the alkali and extract the drug base as described under Salts and tablets. Develop the colour as described under Procedure for Colour Development.The results of the assay of salts and dosage forms of piperazine using tetrachloro- and tetrabromo-l,4- benzoquinones are compared with the official method27 in Table 2. Results and Discussion Piperazine is a fully saturated molecule and the electrons that are responsible for the charge-transfer transition are the non-bonded electrons on the nitrogen atom. Benzoquinones are well known x-acceptors and hence the nature of the excitation in these interactions is an n-n* transition. The absorption bands of the complex are well separated from those of either of the components.1278 ANALYST, OCTOBER 1984, VOL. 109 Table 1. Analytical data for the determination of piperazine Sample No. Reagent 1 1,4-Benzoquinone . .. . . . 2 Monochloro-1,4-benzoquinone . . 3 2,5-Dichloro- 1,4-benzoquinone 4 2,6-Dichloro-l,4-benzoquinone 5 2,3,5,6-Tetrachloro-l,4- . . . . benzoquinone . . . . . . 6 2,3,5,6-Tetrabromo- 1,4- . . . . benzoquinone . . . . . . 7 3,4,5,6-Tetrachloro-l,2- . . . . benzoquinone . . . . . . * Values from reference 30. Electron affinity* 0.77 1 .oo 1.15 1.20 1.37 1.37 1.55 k I - 1 nm 506 527 538 546 575 576 588 Time to attain maximum absorbance 3 h l h 30 min 20 min 15 min 10 min 5 min Beer's law range1 pgml-1 13-120 11-1 10 10-100 10-90 7-70 7-70 6 6 0 Sandell's sensitivity1 pg cm-2 0.128 0.108 0.098 0.092 0.078 0.076 0.064 Time stability/ h 3 12 20 24 36 36 48 0.8 0.7 0.6 0.5 a C (II e n a g 0.4 0.3 0.2 0.1 0 \ \ \\ \ I 17 I I I 460 490 520 550 580 610 640 Wavelengthlnm Fig.1. Absorption spectra of charge-transfer complexes of piperazine ( 5 x in Table 1). The broken lines represent the absorption of benzoquinones (saturated solutions diluted five-fold) M) with benzoquinones (numbers refer to compounds indicated The continuous variations28 and the molar ratio29 methods show that the composition of the complexes is 1 : 1, except that for tetrachloro-1,2-benzoquinone, where a ratio of 1 : 2 (acceptor to donor) is observed. This indicates that only a single nitrogen is involved in the formation of charge-transfer complexes, although piperazine is a twin-site donor. From Table 1 it can be seen that the reaction is rapid with tetrahalo-l,4-benzoquinones. No reagent blank is necessary in measuring the absorbance of the complexes (cf., Fig.1). The complexes were found to be stable. The results also indicate that the method is sensitive to microgram amounts of piperazine. The time stability and the sensitivity were found to have a direct relationship with the electron affinity of the quinones (Fig. 2). The results for the variation of reagent concentration (Table 2) indicate that 1 x 10-3 to 4 x 10-2 M of tetrachloro- or tetrabromo-l,4-benzoquinone will not affect the determina- tion of piperazine. The higher concentrations of the reagent may, on the other hand, be useful for rapidly reaching equilibrium, thus minimising the time required to attain the maximum absorbance at ACT. The assay of dosage solutions involves extraction of the drug base into chloroform. A single extraction is capable of extracting 93.8% of piperazine and thus three batch extrac- tions are sufficient, instead of the five reported earlier.20 Piperidine and N-methyl- and N-(aminoethy1)piperazines interfere at all concentrations, but pyridine, pyrazine and 2,5-dimethylpiperazine (up to a 5-fold excess), triethanol-ANALYST, OCTOBER 1984, VOL.109 1279 Table 2. Assay of salts and dosage forms of piperazine using three different methods Recovery,* o/o of label claim Tetrachloro- Official 1,4-benzo- Piperazine product method27 quinone (C.V.) Hexahydrate . . Hydrate solution Phosphate BPC Phosphate tablets CitrateBP . . Citrate tablets T1 Citrate tablets T2 Citrate syrup S, Citrate syrup S2 Di hydrochloride AdipateBP . . Adipate tablets . . . . . . 99.5 . . . . . . 100.4 . .. . . . 99.7 . . . . . . 101.2 . . . . . . 99.6 . . . . . . 100.8 . . . . . . 99.3 . . . . . . 101.7 . . . . . . 98.2 . . . . . . 99.0 , . . . . . 98.9 . . . . . . 99.3 99.8 100.2 99.3 101.8 99.7 100.2 99.8 102.1 97.8 99.0 98.4 99.0 (0.39) (0.29) (1.07) (0.65) (0.57) (0.40) (0.40) (0.76) (0.58) (0.39) (0.80) (0.48) * The coefficient of variation (C.V.) values are derived from three determinations for each product. Tetrabromo- 1,4-benzo- quinone (C.V.) 100.4 (0.65) 99.3 (0.83) 102.0 (0.65) 100.0 (0.65) 100.4 (0.56) 99.2 (0.57) 102.1 (0.65) 98.6 (0.29) 98.6 (0.29) 98.8 (1.13) 99.6 (0.57) 99.4 (0.28) 6o F 50 30 P) E F 20 0.14 1 I N I 1 0.12 CI) 5. > .z 0.08 - 8 0.06 - > 0.10 - 4- .- .- c 0.04 ‘ I 1 I I 0.5 1 .o 1.5 Electron aff inity/eV Fig. 2. Correlation of the sensitivity of the method and the time stability of the charge-transfer complexes with the electron affinity of the acceptors (numbers refer to compounds indicated in Table 1) amine (up to a 0.3-fold excess) and ethylenediamine (up to a 0.2-fold excess) do not interfere.The assay results on dosage forms of piperazine using the present method compare well with those obtained using the official BP method. Further, the results (coefficients of variation in Table 2) indicate a good precision for the method. 4. Das Gupta, V., Am. J . Hosp. Pharm., 1976, 33,283. 5. Rao, G . R., Kanjilal, G . , and Mohan, K. R . , Analyst, 1978, 103, 993. 6. Hanana, S., and Tang, A., J . Pharm. Sci., 1973, 62, 2027. 7. Dessouky, Y. M., and Ismaiel, S. A , , Analyst, 1974, 99, 482. 8.Abou-Ouf, A. S . , Taha, A . , and Saidhom. M. B.. J . Pharm. Sci., 1973, 62, 1700. 9. Cavett, J. W., and Heotis, J. P., J . Assoc. Off. Agric. Chem., 1958, 41, 323. 10. Parlmulter, S. H., J . Assoc. Off. Agric. Chem., 1958, 41, 506. 11. Loucks, M. F., and Nauer, L., J . Assoc. Off. Anal. Chem., 1967, 50, 268. 12. Wachsmuth, H., and Koeckhoven, L. V . , J . Pharm. Belg., 1962, 17, 220. 13. Baggi, T. R., Surinder, N. M., and Rao, G. R., J . Assoc. Off. Anal. Chem., 1974, 57, 1144. 14., Mulliken, R . S., and Person, W. B., “Molecular Complexes,” Wiley-Interscience, New York, 1969. 15. Townshend, A . , Proc. SOC. Anal. Chem., 1973, 10, 39; 1976, 13, 64. 16. Muralikrishna, U., Seshasayi, Y. V. S . K., and Krishnamurthy, M., J . fndian Chem. SOC., 1983, 60, 447. 17.Muralikrishna, U . . and Krishnamurthy. M., Indian J. Chem., 1982, 21A. 1018. 18. Muralikrishna, U., and Rao, N . S . , Indian 1. Chem., 1978, 16A. 993. 19. Belal, S . , Elsayed, M. A., Abdel-Hamid, M. E., and Abdine, H., J . Pharm. Sci., 1981, 70, 127. 20. Rizk, M. S., Walash, M. I., and Ibrahim, F. A., Analyst, 1981, 106, 1163. 21. Kobashi, H., Tomioka, Y., and Morita, T., Bull. Chem. SOC. Jpn., 1979, 52, 1568. 22. Kuboyama, A., Bull. Chem. SOC. Jpn., 1960,33, 1027. 23. 24. 25. 26. Thomson, R . H., “Naturally Occurring Quinones,” Academic Press, London, 1971, p. 44. Muralikrishna, U., Rao, N. S . , and Ramanadham, G. V., Curr. Sci., 1975, 44, 534. Muralikrishna, U., Rao, N. S . , and Krishnamurthy. M., Zndian J . Pharm. Sci., 1983, 45, 28. Muralikrishna, U., and Krishnamurthy, M., Microchem. J . , in the press. “British Pharmacopoeia,” Volume 1, Pharmaceutical Press, London, 1980, p. 352. Job, P., Ann. Chim. (Paris), 1928, 9, 113. Yoe, J . H., and Jones. A. L . , fnd. Eng. Chem., Anal. Ed., 1944. 16, 111. Briegleb, G., Angew. Chem., Znt. Ed. Engl., 1964, 3, 617. The authors thank the Council of Scientific and Industrial Research, New Delhi, for the award of a Senior Research Fellowship to M.K. 27. 28. 29. 30. References 1. 2. 3 . Masse, M., Pharm. Acta Helv., 1958, 33, 80. Pankratz, R. E., 1. Pharm. Sci., 1961, 50, 175. Shirsat, P. D., Indian J . Pharm., 1975, 37. 101. Paper A31403 Received November 11th) I983 Accepted May 4th, 1984

ISSN:0003-2654    

DOI:10.1039/AN9840901277

出版商:RSC    

年代:1984

数据来源: RSC