JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 69R ATOMIC SPECTROMETRY UPDATE-CLINICAL AND BIOLOGICAL MATERIALS FOODS AND BEVERAGES Simon Branch The Lord Rank Research Centre R. H. M. Research and Engineering Lincoln Road High Wycombe Buckinghamshire HP12 3QR UK Helen M. Crews Ministry of Agriculture Fisheries and Food Food Safety Directorate Food Science Laboratory Colney Lane Norwich NR4 7UQ UK David J. Halls* Trace Element Unit Institute of Biochemistry Royal Infirmary Castle Street Glasgow G4 OSF UK And rew Ta y I or Supra-Regional Assay Service Metals Reference Laboratory Robens Institute of Industrial and Environmental Health and Safety University of Surrey Guildford Surrey GU2 5XH UK Summary of Contents 1 Analysis of Clinical and Biological Materials 1.1.General Reviews 1.2. Sampling and Sample Preparation 1.3. Developments in Multi-element Analysis 1.3.1. Inductively coupled and direct current plasma atomic emission spectrometry 1.3.2. Inductively coupled plasma mass spectrometry and other mass spectrometric techniques 1.3.3. X-ray fluorescence spectrometry 1.3.4. Other multi-element techniques and studies 1.4. Developments in Single-element Techniques 1.5. Reference Materials and Inter-laboratory Trials 1.6. Hair Analysis 1.7. Progress for Individual Elements 1.7.1. Aluminium 1.7.2. Arsenic 1.7.3. Beryllium 1.7.4. Bromine 1.7.5. Cadmium 1.7.6. Calcium 1.7.7. Chromium 1.7.8. Copper 1.7.9. Fluorine 1.7.1 0. Gold 1.7.1 1. Iodine 1.7.12. Iron 1.7.1 3. Lead 1.7.1 4. Lithium 1.7.15. Magnesium 1.7.1 6. Manganese 1.7.1 7.Mercury 1.7.18. Molybdenum 1.7.1 9. Nickel 1.7.20. Phosphorus 1.7.21. Platinum 1.7.22. Potassium 1.7.23. Selenium 1.7.24. Silicon 1.7.25. Strontium 1.7.26. Tellurium 1.7.27. Tin 1.7.28. Uranides 1.7.29. Zinc 1.8. Conclusions Table 1. Summary of Analyses of Clinical and Biological Materials * Review Co-ordinator. to whom correspondence should be addressed.70R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 2 Analysis of Foods and Beverages 2.1. Sample Preparation 2.2. Sample Introduction 2.3. Speciation Studies 2.4. Developments in Methodology for Atomic Absorption Spectrometry 2.5. Developments in Methodology for Plasma Emission Spectrometry 2.6. Developments in Methodology for Inductively Coupled Plasma Mass Spectrometry 2.7. Topical Applications 2.4.1.Flame atomic absorption spectrometry 2.4.2. Electrothermal atomic absorption spectrometry 2.7.1. Aluminium in foods 2.7.2. Reference materials and collaborative trials 2.7.3. Characterization studies Table 2. Summary of Analyses of Foods and Beverages This Update contains reviews on recent developments in the application of atomic spectrometry to the analysis of clinical and biological materials and of foods and beverages. These two reviews cover the references 90/1159-90/4166 and 91/1-91/825 which are listed fully as Atomic Spectrometry Update References in Volumes 5 and 6 of JAAS respectively. A list of references to published papers in abbreviated form appears at the end of this Update. The Tables in an improved format summarise the methods and studies covered in these reviews.Last year’s Update (J. Anal. At. Spectrom. 1990 5,75R) covered developments in the preceding year. 1. ANALYSIS OF CLINICAL AND BIOLOGICAL MATERIALS David J. Halls and Andrew Taylor This review covers recent developments in atomic spectromet- ric techniques for the analysis of clinical and biological materi- als. Table 1 summarizes published and conference papers produced within the review year. Two minor changes have been made to the format of the review. Solid and slurry sampling have been incorporated into sampling and sample preparation. Interest in direct solid Sam- pling appears to have waned. Slurry sampling the more practi- cal alternative is however flourishing and has matured as a technique. Interest in hair analysis remains somewhat surpris- ingly high.It was thought worthwhile for this review to discuss this as a separate section particularly to try to assess whether hair analysis was really providing new information. 1.1. General Reviews Tolg’s lecture entitled ‘Where is analysis of trace elements in biotic mafrices going?’ has now been published (90/1407). This review surveys various atomic spectrometric techniques assessing their performance criteria The practical application of these techniques to the determination of essential and toxic elements in clinical samples was reviewed by Paschal (90/ 2088). The main considerations at each stage of analysis were illustrated with two examples the determination of Pb in blood and TI in urine by ETAAS. Determination of trace elements in marine biological samples by ETAAS was reviewed by Sturgeon (90/2087).The analysis of solids slurries and digested samples was discussed. Slurry sampling overcomes many of the limitations of direct solid sampling and according to Sturgeon it holds great promise for widespread use. Our last Update (90/3500) covered advances in the determi- nation of clinical and biological materials published approxi- mately between September 1988 and September 1989. 1.2. Sampling and Sample Preparation The use of micmwave heating for digestion continues to grow in popularity (90/1291 90/1360 90/2193 90/3327 91/51). Unfortunately many of the papers cover the same ground reflecting the conversion of those researchers to this technique rather than presenting anything new.To reduce blanks Patter- son et al. (90/C 1680) evaluated the use of vapour-phase diges- tion with HNO within a microwave digestion vessel. Dried urine was oxidized by the HNO for subsequent analysis for Cr by isotope dilution MS. Modifications to reduce the internal volume of a Parr bomb were described by Nicholson et al. (90/ 1291) to allow the digestion of micro-amounts (50 mg) of bio- logical samples. Kingston’s group (90/C1762,9 1/C736) devel- oped direct feedback control of a microwave oven with a microcomputer linked to temperature and pressure monitors in the digestion vessel. As a further step an expert system was developed to assist in establishing programmes for dissolution. ‘The objective was to produce a system that was independent of type of oven allowing standardized conditions to be repro- duced readily between laboratories.As an extension of this idea discussions have begun between manufacturers of labora- tory instruments to harmonize instrument settings and condi- tions. Knapp (90/3559) compared the advantages and disadvant- ages of commercial systems for sample decomposition based on pressure digestion with microwave or conventional heating low temperature plasma ashing and combustion in O?. This paper stressed the importance of the materials generally used for vessels and their element contents were tabulated. It was concluded that for most purposes quartz is best. Adaptation of an automated wet-digestion system to handle small sample masses was described by Frank (90/1411) and this was applied to the determination of Se in biological materials (90/141 I 90/ ‘I 4 1 2).S himizu and co-workers demonstrated the usefulness of tetr-amethylammonium hydroxide for solubilizing biological materials for analysis by ETAAS and ICP-AES (90/1410 91/C536) and by ICP-MS (90/1410 91/C537). The advan- tage of the technique is that this digestion avoids losses of volatile elements. Thus B could be determined in biological materials by ICP-AES with ETV (91/C582) and I and Br could be measured by ICP-MS (91/C537 9 1/C583) without loss of these elements in digestion. The probe technique in ETAAS has been adapted by Lucker c p t al. (90/3994) to handle solid samples. Determination of Cd Cu Mn Pb and Zn in various biological matrices was shownJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 71R to be achieved with an accuracy and precision comparable to other solid sampling ETAAS systems. Direct ETAAS analysis of hair and stomach membrane for Cu and Zn was carried out by Heune et al. (90/C2033) using a graphite inner mini-cup coated with lanthanum. The inner cup acted in the same way as the L’vov platform in delaying atomization and reducing interferences. Calibration was with biological CRMs. The Perkin-Elmer ultrasonic agitation system for slurry sampling has been described by Carnrick and co-workers (90/ 3295 90/C3638) with examples of its application to biologi- cal samples. In a further paper (90/3294) the technique was applied to the determination of Mn in five NIST biological SRMs including Bovine Liver. About 10 mg of sample were weighed directly into autosampler cups containing 1 ml of a suspension fluid of dilute HNO and Triton X-100.Determi- nation was made at the less sensitive 403.1 nm line against an aqueous calibration graph. No charring stage was used in the furnace programme. The patentee of the ultrasonic agita- tion system Miller-Ihli has described experience with this system (90/C1793 90/2199 91/C692). Ebdon et al. (91/6) explored the advantages of using an air-ashing stage in the analysis of slurries by ETAAS. Higher ashing temperatures were possible (e.g. for Pb) allowing removal of organic ma- terial and hence reducing background absorbance from smoke. It also reduced chemical interferences. The technique was successfully applied to the determination of Cr Co Mn and Pb in five CRMs including NIES Mussel Chlorella and Human Hair. Holcombe’s team (90/C 179 1,90/C 1829) have been continu- ing their work on pre-concentration of metals by micro- organisms (see last year’s review 90/3500). They have now tried the algal biomass (100 mg) in a column pumping the so- lution of interest through to collect metals.These were later eluted by adjusting the pH. 1.3. Developments in Multi-element Analysis 1.3.1. Inductively coupled and direct current plasma atomic- emission spectrometiy Several workers have reported problems with the direct deter- mination of trace elements and electrolytes in serum by ICP- AES after simple dilution. Melton et al. (90/4032) had prob- lems with fouling of the nebulizer and torch by protein depos- its while Jiang et al.(91/185) found 10-fold dilution gave results consistent with those after sample digestion for Cu Mn and Sr but not for Ba Ca Fe Zn and Mg. In addition simple dilution measures total Fe. Clinical chemists however nor- mally interpret measurements made by colorimetric methods which measure only transferrin-bound and unbound Fe and thus are free from the interference from Fe in haemolysed samples. Melton et al. (90/C293 I 90/4032) overcame these problems by adding a single reagent containing TCA and HCI for protein precipitation hydroxylamine sulphate for reduction of Fe and Y as an internal standard. After centrifugation the supernatant was analysed for Ca Cu Fe K Mg Na P and Zn using a single multi-element standard for calibration. A similar technique using TCA and HNO as the precipitant was de- scribed by Kita et al.(90/3369). Jiang et al. (91/185) resorted to digestion with HN0,-HCIO,. Their claim to include Mn in the range of elements determinable should be viewed with some scepticism as the normal concentrations of Mn in serum (el pg I-I) are below the detection limit for ICP-AES with conventional nebulization. In contrast Bussiere et al. (90/ 3326) reported that they could determine Ca Cu Fe Mg Mn and Zn in amniotic fluid samples by direct nebulization without dilution. Formation of protein deposits on the torch was avoided by the use of a sheath gas device. Standards were prepared in a matrix containing 0.5% m/v of bovine serum albumin and 0.76% m/v NaCI. The system was tested with NIST Human Serum CRM (diluted 1 +4 with dilute HNO,) and showed good accuracy precision and stability.The ques- tion of whether direct determination of elements in serum by ICP-AES is possible is not trivial. If as some of the papers above suggest preliminary treatment is necessary then ICP- AES cannot be considered as a serious competitor in the clini- cal laboratory. Multi-channel analysers using colorimetry and ion-selective electrodes can handle most of the same range of elements with many other additional tests at a fast rate of ana- lysis without preliminary sample treatment. Another possibility for ICP-AES in the clinical laboratory is screening for toxic elements in urine. Lo and Arai (90/1288) developed a procedure for screening for As Be Cd Co Cr Cu Hg Mn Ni Pb and Zn in acidified urine using a simulated urine standard for calibration.Detection limits ranged from 1 pg I-’ (Be) to 50 pg 1-’ (As). Coefficients of variation ranged from 3 to 10% RSD and recoveries from 94 to I 10%. Barnes’ group (90/2064 90/2065) reported the analytical groundwork they carried out for a major study on soil inges- tion by children. Eight marker elements (Al Ba Mn Si Ti V Y and Zr) were determined in food dust soil faeces and urine by ICP-AES; three of these (V Y and Zr) were also measured by ICP-MS. Faecal samples were freeze dried ashed ground to a fine powder and fused with LiBO,. Considerable practical problems were encountered in sampling and processing soiled diapers. In a practical test of the procedure (90/2065) six adults ingested 500 mg of soil per day and the recovery of the elements in faeces was measured.This showed that Al Si and Y were probably the best indicators of soil ingestion. In the context of an EEC programme on ‘Exposure and Health Effects’ Caroli and collaborators (90/C 1898 90/1979 90/2237) established reference values for trace elements in human lung tissue with a view to using the techniques estab- lished for assessing occupational exposure. Samples were col- lected using titanium nitride coated surgical instruments and 48 elements determined by ICP-AES and NAA. Results obtained by these two techniques were compared for eight elements (90/1979). The linear correlation of ICP-AES results (y-axis) against NAA results (x-axis) gave slopes of less than one for seven of these elements.Inadequate precision at normal concentrations was found for the determination of Al and Cd by NAA and V by ICP-AES. Accuracy was poor for Al by ICP-AES (125% recovery) and for Cu by NAA (82.5% recovery). A comparison of results by NAA and ICP-AES was also made for the analysis of human brain by Andrasi et al. (90/C1901). Taking into account the inhomogeneity of the sample they reported a good correlation. Instrumental NAA required no sample preparation minimizing contamination but analysis time was long and precision was relatively poor. The ICP-AES technique was faster with good detection limits and better precision. To handle the high solids content of digested rat bones Hight (90/C1756) introduced samples into the ICP with a Hildebrand grid nebulizer for the determination of Ca Cu Fe K Mg Mn Na P and Zn.The bones were digested in HCIO leading to a final concentration of 10% v/v. Standards con- tained matched concentrations of Ca Mg and P. Li and Shen (90/C2009) determined Al Cr Cu Fe Mn Sr and Zn in ants by ICP-AES after pressure digestion with HN0,-HClO,. To obtain reliable results external contamination had to be removed by a cleaning protocol involving soaking in water detergent solution and then EDTA solution. These workers re- ported that female ants were digested more easily than the males; there has to be a message in there somewhere! Braetter’s group continued their studies of speciation using a combination of HPLC and ICP-AES (90/1414 90/1902). Species containing the elements Ca Cu Fe Mg and Zn were measured in body fluids (90/1902) and in human milk and infant milk formulas (90/1414) and related to the distribution profiles of proteins.In human milk the citrate concentration increased by a factor of 25 after birth increasing the propor- tion of Fe and Zn bound to citrate. They extended their study72R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991. VOL. 6 to Se in serum erythrocytes lysates and human milk (90/ 1902). The changes in speciation of Ca Cu Fe Mg P S and Zn in human milk were followed throughout lactation by Suzuki et al. (91/C531) using gel-filtration HPLC and ICP- AES. They too recorded marked changes in Zn distribution during the first week which settled down in the following two weeks. Changes in the other elements were seen also.Developments in MIP technology could make this an attrac- tive alternative to the ICP as power and gas consumption are less. Itai et al. (91/C526) applied an He MIP to the determina- tion of Cu and Zn in serum and urine. The development of the toroidly shaped Ar MIP has allowed MIP-AES with ETV to reach detection limits up to 10-fold better than previously pos- sible according to Heltai et al. (91/C491). They demonstrated this with the analysis of several biological CRMs. A tungsten-filament vaporizer developed by Buckley and Boss (91/36) was used with a DCP to determine Cu in whole swine blood. A microprocessor controlled the heating of the filament and allowed synchronization of vaporization and capture of the data from the transient signal. Very low detection limits in the range 80-200 fg were possible with this technique. 1.3.2.Inducti\tely coupled plasma mass spectrometiy and other mass spectrometric techniques Progress is continuing to be made in the application of ICP- MS to the determination of trace elements in clinical and bio- logical materials. Many of these relate to the determination of one or two elements and are covered in section 1.7. Interest continues in the application of isotope measurements to identi- fy sources of exposure (see Pb section 1.7.13; and Uranides section 1.7.28) and in the use of stable isotopes for nutritional or clinical studies (see Cu section 1.7.8; Fe section 1.7.12; and Zn section 1.7.29). For stable isotope measurements both ICP-MS and FABMS have been used. The group in Gent in Belgium have continued their work on multi-element determinations in serum by ICP-MS (90/C18 1 1 90/C1900,90/C2001 91/C509).Their preparation involves di- lution of samples 1 + 4 or 1 + 9 with 0.14 mol dm-3 HNO con- taining as an internal standard. Elements measured satisfactorily so far were B Ba Bi Br Cs Cu Co Fe I Li Mo Pb Rb Sr Sn and Zn but Al As Cr Mn and Se were precluded because of polyatomic interferences. The feasability of direct analysis of urine by ICP-MS for screening exposure to toxic elements was studied by Mulligan et al. (91/3). Samples were diluted 1 + 9 with 2% v/v HNO,. At this dilution analyses could be carried out for a whole day with no evidence of appreciable accumulation of solids on the sampler orifice and no degradation of performance. Generally detection limits were below 1 pg I-' in the original urine even when scanning from 2 to 250 u over 2 s and averaging 120 successive scans.Certain elements (As Cr Se Ti and V) were precluded because of isobaric interference. Comparison of results for Cd Hg and Sb with results by alternative AAS tech- niques and with target values of inter-laboratory quality as- sessment samples showed reasonable agreement except for Hg. In the multi-element analysis of tissues by ICP-MS Friel et al. (90/2193) found that the extent of digestion was important. Use of incompletely digested samples led to a loss of sensiti- vity with time; in addition thorough digestion (as shown by a low "C count in the mass spectrum) reduced isobaric interfer- ences. Samples were first digested with HNO in a closed di- gestion vessel using microwave heating and then digested further with HNO,-H,OI on a hot-plate.The digests were ana- lysed by ICP-MS with five internal standards. Analysis of NIST SRMs Bovine Liver and Oyster Tissue and IAEA Animal Muscle CRM for 12-14 elements gave results which were well within the certified range for most elements. Simple pressure digestion with HNO using microwave heating was used by Gelinas et al. (91/C601) for the determination of 27 elements in rat tissues by ICP-MS to establish an 'inorganic fingerprint'. Applications of ICP-MS to marine biology were described by McLaren et &(9 l/C480). The simplicity of interfacing ICP-MS to HPLC for speciation studies was stressed and illus- trated with the determination of As and Sn species.Multi- element analysis of a new lobster hepatopancreas RM (LUTS- 1 ) was also described. Ishil (91/C576) applied ICP-MS to the determination of 13 elements in marine organisms; less inter- ference was found for the elements of higher relative atomic mass. Applications of SIMS to elemental imaging in studies of bones and teeth (90/2652) and in marine biology (90/2572) have been reviewed. The value of the technique was demon- strated in studies of human and rat teeth by Chabala et al. (90/ 2824). Distributions were obtained for Ca CN- and F and in selected specimens additionally for C- CI K Mg Na 0- and PO-. Information on the kinetics of mineralization could be ob- tained. Mony et al. (90/2568) studied distributions of Ca Fe K Li Mg and Na in the tissues of mice treated with Li.Images of Li distribution coincided with those of Na and K. 1.3.3. X-ray fluorescence spectrometry Recent developments in X-ray techniques were covered in a companion Atomic Spectrometry Update (J. Anal. At. Spec- from. l990,5,243R). In the determination of trace elements in serum by TXRF Bethel et al. (90/2147) found serious matrix interferences. To overcome these samples were ashed with a low temperature plasma asher and then the elements separated from the main inorganic matrix by ion exchange. Simultaneous determination of Ca Cu Fe K Pb Se Sn Rb and Zn was possible. Prange et al. (90/3422) digested whole blood and serum samples with HNO and microwave heating for the determination of Ca Cu Fe K P Pb Rb S Se Sr and Zn.After separation of the main salts and Fe then Mn and Ni could additionally be deter- mined. A limited range of elements can be determined without sample preparation as Yap (90/2 183) demonstrated. Samples were diluted 1 + 1 with water containing Y as an internal stan- dard for the determination of Br Cu Fe and Zn. In view of the small volume of sample required it was considered ideal for paediatric studies. Sun et al. (91/188) also achieved analysis without sample pre-treatment for Al Ca CI Fe K Mg Na P S Si and Zn by applying the thin-sample method. A small volume (0.1 ml) of blood was pipetted on to a filter-paper and the spot analysed. Thin-film PIXE was applied by Barbour et al. (90/3433) for the determination of Br in urine. Results ranged from 2 to 18 mg I-'.No significant difference was seen between urinary Br concentrations in urinary stone formers and controls. The variation of trace element concentrations in kidney stones was studied by Galassini et al. (90/2139) using EDXRF and PIXE. The concentrations and distribution of a range of elements (Br Ca Cu CI Fe K P Pb Si Sr and Zn) were presented. The Pb content seemed to be a good indicator of environmental exposure. The Sr and Zn contents of various parts of human teeth were determined by Frank et a f . (90/1854) using EDXRF. The Sr content was higher in enamel than in dentine and increased with age in the enamel but not in dentine. The Zn contents were fairly high in both enamel and dentine with no significant change with age. In a study of hones dug up in archaeological sites Kliment (90/2119) measured Sr:Ca ratios by EDXRF with a ' T d source.Bones from horses and ruminant animals had a high Sr:Ca ratio which was related to the high Sr:Ca ratio in the plants that formed their diet. The ratio was lower in human bones and only small differences were seen between different prehistoric bones.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 73R An experimental method for evaluating matrix c-on-ections in the determination of trace elements in thick biological samples by EDXRF was outlined by Kumar et al. (90/2145). The method was tested on the analysis of NIST biological SRMs in the form of thick pellets. Cercasov and Schreiber (90/ 1409) made their corrections from measurements of the back- ground intensity at two different energy regions (e.g. 5 and 15 keV).From this the effective mass and effective thickness were calculated. Synchrotron radiation XRF (SRXRF) with tunable mono- chromatic excitation gives superior detection limits to PIXE in the analysis of biological samples according to a study by Torok et al. (91/366). This advantage is most pronounced for elements with atomic number greater than 30. Compared with EPMA SRXRF was more sensitive for elements of atomic number above 40. Where X-ray techniques are superior to other analytical techniques in clinical and biological analysis is in their ability to measure the concentration in a very small area and hence to show the distribution of the major and trace elements. This application has been reviewed by Hall (91/314) by Nagy (91/ 256) specifically for cancer research and by Roomans (90/ 1248) for studies of cystic fibrosis.The trace element distribu- tion can help distinguish between malignant and normal tissue as was demonstrated by Drake and Sky-Peck (90/1871). By using ultra-micro EDXRF for measurement elements found to be most useful were Ca Rb and Zn in breast samples Ca Fe and Zn in colon and Cu Fe and Mn in lung samples. Nine- 10- and 1 1 -element discriminant functions resulted in accura- cies of classification of 98% for breast 100% for colon and 100% for lung tissues respectively. Strontium and Br were evaluated by Wroblewski et al. (91/317) as tracers for Ca and CI respectively. The Sr:Ca and Br:Cl ratios in the acinar cells of the submandibular gland of rats were measured by EDXRF microanalysis after the rats had ingested SrCI or NaBr.In a further study from the same Institute (91/315) a comparison was made between the information provided by EDXRF mi- croanalysis NAA and AAS on normal and injured muscle. The localized changes in Na CI and P could be followed in muscle cells by EDXRF and were different from the magni- tude of the changes seen with NAA and AAS which measured the concentrations in the bulk of the white muscle biopsy in- cluding those in extracellular non-muscular components. A similar study of the changes in Ca CI K Na and P in peri- pheral nerve axons following injury was made by LoPanchin et al. (90/1249). The subcellular ion distribution in the intesti- nal mucosa of the mouse was studied by Von Zglinicki and Roomans (91/318) using XRF microanalysis of ultra-thin cryo- sections.1.3.4. Other multi-element techniques and studies Miller-Ihli (90/2 199 90/3584) gave further demonstrations of the ability of simultaneous atomic- absorption spec-trometric analysis with a continuum source (SIMAAC). Analytical results were presented for a range of biological CRMs and samples with flame and electrothermal atomization. Satisfac- tory accuracy and precision was obtained in ETAAS with a compromise heating programme for the graphite tube and plat- form which included charring at 500 "C and atomization at 2700 "C. Examples were given of sample preparation by wet and dry ashing and of slurry and direct liquid sampling (90/ 2 199). A Norwegian study compared trace element concentrations in human tissues from the populations of Bergen and the Faroe Islands (90/1833).Calcium Cd Fe Hg Se and Zn were deter- mined in autopsy samples of liver kidney cortex and medulla heart muscle and spleen by ETAAS and FAAS. High Hg and Se concentrations in the liver and kidney of Faroe Islanders was explained by the high consumption of pilot whales. Automated F l sysrems for the determination of Ca Cu Fe Li Mg and Zn in serum and urine by FAAS and FAES were developed by Araujo and Lima (90/1549). Three separate mani- folds were used differing in the extent of dilution used one for K Li and Na (80 determinations per hour) a second for Ca and Mg (1 80 determinations per hour) and a third for Cu Fe and Zn (240 determinations per hour).For large dilutions part of the stream was split off and pumped to waste and the re- maining part pumped directly into the nebulizer. This removed effects related to different physical properties of sample and standard. Comparison with manual dilution methods showed good agreement but the FI results showed better precision. The Ca K Mg and Na concentrations in human lympho- cytes were measured by Weissberg et al. (91/327) by AAS after a single step separation procedure of the cells from blood on a discontinuous Ficoll-Histopaque gradient. Variation between individuals was less when results were expressed as per mg of protein rather than per unit cell. 1.4. Developments in Single-element Techniques Vanadium(v) as a chemical modifier in ETAAS was thor- oughly evaluated by Tsalev et al.(90/4145). Maximum thermal pre-treatment temperatures were for some elements higher than those obtained with Pd but a mixed modifier of V + Pd was even better showing excellent stabilization for Ag As Bi Cu Ga Ge In Mn P Pb Sb Se Sn Te and T1. The value of the V or V + Pd modifiers was demonstrated by the determination of elements in several CRMs including As in NIES Mussel Se in IAEA Horse Kidney and Sn in NIES Human Hair. The horse kidney was analysed as a slurry whereas the other two were solubilized with tetramethylammo- nium hydroxide. Fast analysis by ETAAS has been studied by Kunwar et al. (91/72) and Slavin et al. (90/2089). Both recognized that the drying time could be reduced by injecting the sample on to a hot tube or platform.Kunwar et al. injected urine RMs or digested biological tissue RMs on to a pyrolytic graphite tube at 120 "C; Slavin et al. reduced the cool down time of the platform so that it was still warm when the sample was inject- ed. A short (2 s) burst of heating at an applied temperature of 700 "C helped to warm up the platform quickly for drying. No charring step was used. To cope with the higher background absorbance Zeeman-effect background correction was used. Application of this technique included the determination of Cd Cu and Pb in NIES Mussel CRM Cd and Cu in NIST Oyster Tissue SRM and Cr in NET Elevated Urine SRM. All of these determinations were carried out on HNO digested samples without a modifier and gave good agreement with the certified values. Kunwar et al.however wished to use a uni- versal modifier (Pd-HJ with a standard pre-treatment tempera- ture of 1000 "C. A short preceding step at 300 "C in the presence of H was found to be important to convert the Pd into the elemental form. This was satisfactory for all elements tried except for Cu which gave low recovery. If H was re- placed by Ar in the 1000°C step then correct recovery was obtained for Cu. Satisfactory results were demonstrated for the determination of Cu Ni and Pb in urine RMs and Cu and Fe in NIST Bovine Liver SRM. The application of laser-enhanced ionization specwxopy in the certification of SRMs in NIST was described by Turk and Kingston (90/C 1765). Prior to determination the alkali and al- kaline earth metals in biological samples were separated by che- lation chromatography with Chelex 100.Determinations described included Mn and Ni in NIST Bovine Serum SRM and Cu Mn Ni and Pb in a proposed NIST Total Human Diet SRM. lndir-ect FAAS methods for the determination of pharmaceu- ticals were described in recent publications. Two of these used ion-pair formation for the determination of alkaloids. Kir and Temizer (90/1993) extracted the ion-pair complex of Ni( SCN),'- with atrophine sulphate or pilocarpine hydrochlo-74R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 ride into an organic phase (IBMK or 1,2-dichloroethane). Nickel was then determined in the organic phase by FAAS. The optimum pH for extraction was 5.2 for atrophine sulphate and 4.0 for pilocarpine hydrochloride. The method was applied to eye drops injection solutions and tablets.The complex Bi1,- was used by Nerin et al. (91/66) for the ion-pair extraction of nemorensine platyphylline senecionine and se- neciphylline into I ,2-dichloroethane. Flame AAS was then used to determine Bi in the extract. Caffeine in analgesics was reacted with phosphotungstic acid to form a precipitate in a method developed by Bazzi et al. (90K1678). The precipitate was re-dissolved and the W measured by FAAS with an N,O- C,H2 flame. Montero et al. (91/12) determined chlordiazepox- ide (Librium) in pharmaceutical preparations by reduction on a column of metallic Cd or Zn in an FI system. Cadmium or Zn ions liberated were detected by FAAS. This method is selec- tive to the N-oxide group in chlordiazepoxide allowing deter- mination in the presence of other 1,4-benzodiazepines.1.5. Reference Materials and Inter-laboratory Trials Roelandts (90/1959) has extended his original compilation of biological RMs (Spectrochim. Acta Part B 1989 44 281) with the publication of a listing of the more recently introduced materials. In the 1989 Update (J. Anal. At. Spectrom. 1989 4 47R) comment was made about the worrying trend to use CRMs as calibrants. Parr (9l/C590) made the same point em- phasizing the limited amount and expense of CRMs and stating ‘it behoves the analyst to use them sparingly and efficiently’. The proper use of RMs in quality control was discussed. The future CRM programme of the National Institute for Environ- mental Studies (NIES) Japan was outlined by Okamato (90/ C1640).Stocks of NIES Human Hair and Mussel are now ex- hausted but new CRMs planned include Fish Tissue (with certified values for total Sn tributyltin and triphenyltin and or- ganomercury compounds) and Marine Organisms (for As com- pounds). This theme of certification for speciated metals runs through the programmes of the National Research Council Canada (NRCC) (90/C1639) and the Community Bureau of Reference (BCR) of the EEC (91/C646) for marine biological RMs. There are now three marine biological tissues certified for trace elements in the NRCC range and one in the BCR range. The NRCC programme has stimulated the development of new more sensitive reliable and faster methods of analysis placing particular emphasis on the use of ETAAS as the analyt- ical technique of choice (90/135 I).Bovine blood lead RMs were prepared by Cox et al. (90/ 3424) of the Centers for Disease Control Atlanta GA USA to cover the range 6.4-54.0 pg per 100 ml. Concentrations were unchanged when stored over 48 months at -20 “C. Measure- ments of Pb were made by ASV two versions of the Delves microcup method ETAAS and solvent extraction-FAAS. The Italian National Institute of Health runs inter-laboratoiy quality assurance schemes for blood Pb (90/2240) and Cd (90/ 2242). They provide the same internal quality-control materi- als to the various laboratories and regularly evaluate perfor- mance by circulating external quality control samples. For Pb the proportion of laboratories with ‘good performance’ rose from 30% in the first phase of the scheme to 50% in the fourth phase.Morisi et al. (90/2241) evaluated whether this assess- ment accurately reflected the performance of these laboratories with real samples. The performance of 1 1 laboratories was compared with the reference laboratory for quality-control samples and for real samples generated in an extensive bio- logical monitoring programme in which 10% of the samples were required to be checked at the reference laboratory. Statis- tical analysis of the data showed that the laboratories per- formed similarly with the two types of sample. Reasons for the wide range of results obtained in the IAEA Intercomparison Surveys for Cr in biological materials have been evaluated critically by Schelenz et al. (90/1206) high- lighting possible sources of error.When the intercomparison was repeated with seven selected laboratories there was a dra- matic drop in the mean reported value for the materials and a three-fold reduction in the standard deviation of the results. 1.6. Hair Analysis In 1986 Taylor (Ann. Clin. Biochem. 1986,23 361) surveyed the problems in the measurement of trace elements in hair with the object of critically assessing its usefulness. It was conclud- ed that data from trace element analysis of hair did not corre- late well with relevant tissue concentrations. The only occasions in which hair data might reflect body status were extreme situations (e.g. high exposure to toxic elements). Later in 1986 a leading article by Dormandy in the Br. Med. J . (1986 293 975) cited Taylor’s article and came out against the rise of commercial laboratories offering hair analysis to members of the public as a trace element profile.Interest in data from hair analysis worldwide has not waned as is shown by the number of papers to be covered for this review. It appears to be particularly popular in China. Perhaps we might invite our readers in China to write to J . Anal. At. Spectrom. to explain why it is considered useful there. For the clinician hair sampling has an attraction in that it avoids the necessity to take a blood sample particularly stressful for children as Wilhelm et al. (90/3435) stated. For the analyst the higher trace element concentrations and the relatively low concentrations of CI K and Na make hair a more attractive material to analyse than blood or urine.The object of this section is to assess the current work on hair ana- lysis particularly to see whether new information about its usefulness has been gathered and to see whether Taylor’s con- clusions need to be changed. The first problem in hair analysis is to establish a washing procedure that removes external contamination without re- moving the element incorporated into the hair in growth. Yang and Zhuang (91/452) soaked hair in acetone washed with H,O then soaked in a neutral detergent washed again and then dried. The lipid matter in hair was found by Attar et al. (91/325) to contain Ca K Mg Na Ni and Sr in relatively high proportion (> 20% of total). Removal of the lipids from hair by detergents or solvents was thought to remove the elements present from environmental exposure leaving the elements present in the fibre as representative of the nutritional status of the patient.In this and other papers (e.g. 91/265) the belief seems to be strong that if only the sampling and sample prepa- ration could be put right then hair analysis would be able to fulfil its true potential. Analysis of hair for the essential elements is fairly straight- forward. Digestion followed by multi-element determination by ICP-AES is a popular route allowing maximum informa- tion to be obtained from one sampling (90/1384 90/2239 90/ C3115 90/4166 91/175 91/452 9l/C539). Selenium requires separate determination e-g. by hydride generation AAS (9 I / 423). Flame AAS allows the determination of elements such as Cu Fe and Zn (90/1268 90/1316 91/265) while ETAAS allows the determination of elements present at lower concen- trations such as Mn (91/C657) and Cr (90/3528).Solid sam- pling ETAAS has even been demonstrated for direct analysis of Cu and Zn in hair (90/C2033). X-ray fluorescence spectro- metry allows simultaneous determination of a range of ele- ments (90/1308,90/2 1 15). Eltayeb and Van Grieken (90/2 1 15) dissolved hair in HCI0,-HNO added Y to the digest and pre- cipitated the metals with APDC. The filtered precipitate was then analysed by XRF for Cu Fe Ni Pb and Zn. For clinical applications it is important that the essential element concentrations in the hair reflect the nutritional status of the patient. In rat studies Dever and Bresee (90/3329) demon- strated by the use of an enriched TU isotope that Cu was incor-JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 75R porated into the hair from the diet. Isotope measurements were by NAA and total Cu measurements by ICP-AES. The ratio TU to total Cu in the hair increased from 0.3 1 on normal diet to 0.53 with T u supplemented feed. Ziu et al. (90/13 16) found that the Zn:Cu ratio in hair was higher in patients with arteriosclerosis than in healthy controls. Unfortunately no further details were available to this reviewer (D.J.H.). Hair Zn concentrations were lower in anorexic children than in healthy children but hair Cu concentrations were similar according to Yuan (9 l/438). Com- parison of concentrations of Cu and Zn in scalp and pubic hair by Wilhelm et al.(91/265) showed some correlation ( r = 0.549 for Cu and r = 0.263 for Zn). Concentrations were higher in scalp hair (the first 2-3 cm closest to the scalp) than in pubic hair. An extensive study to establish reference ranges for children has been reported by Senofonte et al. (90/2239). For toxic elements the need for multi-element determination is less important than in nutritional assessment. Many papers therefore report results or methods for one or two elements. Numakura et al. (90/1384) determined Al in human hair by ICP-AES. Interferences were avoided by extracting Al from the digest with 8-hydroxyquinoline into CHCI,. A mixed Pd- Ni modifier was used by Long and Liu (91/386) in the determi- nation of Hg in hair by ETAAS. Arsenic in a hair digest was extracted by Gu et al.(91/449) as an Sb-Mo-As hetero-poly acid into IBMK for subsequent determination by HGAAS. Hair measurements do seem to have some relevance in assessing differences in environmental exposure of various population groups. Tavares et al. (91/C657) found high hair Mn concentrations (mean 38.7 pg g-') in the population of a village in Bahia Brazil that surrounded an Fe-Mn alloy foundry. No evidence of Mn poisoning was found in the women studied. Another study from the same group (90/1838) looked at hair Pb concentrations in children living near a Pb smelter. Concentrations ranged from 20 to 4933 pg g-I with a geometric mean of 3 13 pg g-I. Parameters found to increase the hair Pb concentration were younger age shorter distance between home and smelter longer residence time dark skin curly hair and being the child of a Pb worker.Girls had higher hair Pb concentrations than boys. Wilhelm et al. (90/3435) compared hair Pb concentrations in children in industrial Duis- burg (Germany) and rural Westphalia finding geometric mean values of 3.7 and 1.9 pg g-I respectively. Summer values were higher than winter values in both regions and boys had higher values than girls. Non-German children had higher values than German children. They concluded that hair Pb measurements were a useful tool for epidemiological studies. In the compari- son of scalp and pubic hair (91/265) members of the same group found some correlation for Pb ( r = 0.576) and Cd ( r = 0.474) with scalp hair measurements being higher than pubic hair concentrations.Better correlations were found between Cd and Pb concentrations in the same material ( r = 0.691 for scalp hair and r = 0.621 for pubic hair). Higher than normal hair Pb concentrations in an occupationally exposed group were reported by Eltayeb and Van Grieken (90/ 21 15). Hair Pb concentrations in the exposed group ranged from 70 to 3700 pg g-' compared with 3 to 17 pg g-' in the non-exposed group. Zhang et al. (91/C539) found that low level Cd exposure from a chinaware factory led to an increas- ing hair Cd concentration with age in children and that in- creased Cd exposure led to a decrease in hair Zn concentration and blood haemoglobin. Higher hair concentrations of Cd Cu Hg and Pb were reported in a violence-prone group of children by Marlowe et al.(90/C3662). Multi-element analysis of hair samples by ICP-AES from patients on haemodialysis showed higher concentrations of P and Zn and lower concentrations of Si than in healthy persons (90/4166). Interestingly. Al concentrations were little different. It would seem that the use of hair analysis for nutritional as- sessment is still not convincing. The weak correlation between the two types of hair scalp and pubic for Cu and Zn is indica- tive of the poor predictive value of these measurements. For epidemiological assessment of environmental exposure to toxic elements hair analysis seems to be useful although it is worth noting that Wilhelm et al. (90/3435) would not use hair data on an individual basis but would prefer a blood Pb meas- urement. A more convincing case can be made for occupation- al exposure.In last year's review (90/3500) for example two papers were discussed both showing that hair Cr concentra- tions were a good indicator of exposure to chromates and that the data correlated well with urine Cr concentrations. In this situation however it may still be difficult to distinguish exter- nal contamination from aerosols in the atmosphere and the element incorporated in growth especially for elements such as As and Hg which bind tenaciously to hair. The distinction may not be too relevant however when risk to exposure is to be assessed as a high air concentration indicates a greater bio- logical risk. In summary there is little reason to modify Taylor's original conclusions about the value of hair analysis.1.7. Progress for Individual Elements 1.7.1. Aluminium Studies investigating the effects of Al are extremely topical par- ticularly those involving elevated environmental levels of this toxic element. In 1988 20 tonnes of AI,(SO,) were accidental- ly released into the public water supply feeding the town of Ca- melford Cornwall UK. Clinical effects were experienced by a number of the inhabitants some of whom have still not recov- ered. Taylor (90/C4036) reported some of the biochemical in- vestigations that were carried out. Some but not all of the serum samples examined had elevated serum Al concentrations. Of two patients given a desferrioxamine test one showed a small rise in urinary Al excretion. Both patients had evidence of a localized deposit of Al in bone biopsies.Electron emission X- ray spectroscopy confirmed the presence of Al although total Al concentrations measured by ETAAS in these and sample bi- opsies from four other individuals (not showing positive stain- ing) were all within the reference range. These findings have recently been published (Eastwood J.B. Levin G.E. Pazianas M. Taylor A.P. Denton J. and Freemont A.J. Lancet 1990 336 462). At the same meeting Candy and Edwardson (90/ C4035) discussed Al as one of the factors contributing to Alz- heimer's disease and reasoned that further study of the localiza- tion of Al in the brains of dialysis patients might provide clues to the role of Al in the disease. Using secondary ion MS A1 was located in the cerebral cortex of the brain of these patients asso- ciated primarily with pyramidal neurones.In a control group the Al accumulation was absent. Delves (91/C739) has been carrying out studies of the bio-availability of Al from certain foods and drinks by measuring small changes in serum and urine Al after ingestion. A pronounced increase was seen in serum Al after ingestion of fruit juice and soya milk. An early paper by Allain and Mauras (Anal. Chem. 1979 51 2098) described the determination of Al in serum by ICP- AES but since then ICP-AES has not been widely adopted for this determination as few have found it possible to attain their detection limit. Arniaud et al. (90/C2976 90/C3 160) have been re-examining the conditions necessary to achieve a satis- factory determination. Using a Jobin-Yvon JY 38 sequential spectrometer they were able to develop a direct method without sample dilution having a detection limit of < I pg I-' and a linear calibration range of over four orders of magnitude.Publication of this work is awaited with interest. Delves (91/ C739) described a method for the determination of Al in serum by ETAAS which involved platform atomization oxygen ashing and modification with (NH,),HPO,. In an attempt to produce a more robust method for the determination of A1 in serum by ETAAS Cunningham et al. (91/C753) ex- amined some of the factors that would reduce between-batch variation. The use of HNO in the diluent gave an ashing profile with a wider plateau region reducing the possibility of76R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS Element A1 A1 A1 A1 A1 Al Al Matrix Hair Blood serum red cells Serum bone Urine plasma Liver Serum Serum A1 Dialysis concentrates A1 Brain A1 Serum urine bone A1 Brain A1 Serum Alkaloids Pharmaceuticals As Biological samples As Biological samples As Urine As Urine serum hair As Biological samples As Urine As Biological samples Technique; atomization; analyte form* AE;ICP;L AA;ETA;L AA;ETA;L AA;ETA;L LMMS;-;- AE;ICP;L AE;ICP;L AA;F NzO-CZH2;L AE;ICP;L MS;-;S AA;ETA,L AA;ETA;L AA;ETA;L AA;-;- AE;ICP;L AA;Hy;L AA;ETA;L AA;ETA;L XRF-;S MS;ICP;L XRF;-;S Sample treatmentlcomments An Al-oxine complex was formed and extracted with CHCl (in Japanese) The samples were diluted 4-fold with 0.05Oh Triton x- 1 00 Serum A1 concentrations were predictive of levels in bone with approximately 7% error Concentrations were measured in specimens from patients after renal transplantation The ultrastructural location of A1 within cells was achieved Instrumental parameters and chemical interferences were carefully and systematically studied Instrumental parameters and chemical interferences were carefully and systematically studied to develop a protocol with aspiration of undiluted samples which gave sub-ppb limits of detection A system for concentration of A1 on an on-line mixed ion-exchange column was presented.A1 at 30- 190 ng ml-I was determined in sample volumes of 1-10 ml environmental A1 in Alzheimer's disease SIMS was used and showed focal accumulation of A1 in pyramidal neurones in the cerebral cortex of patients with chronic renal failure A1 was measured in samples from individuals who had consumed water contaminated with high Al concentrations An acid digestion procedure was carefully evaluated to establish contribution from contamination and the effect of precipitation of non-digested fatty residues A sample diluent containing 1% v/v HNO and Triton X-100 was used.Serum-based and aqueous KN0,-based standards gave reasonably comparable results A1 kaloids were indirectly measured. The alkaloids (atropine pilocarpine) with Ni(SCN),2- were extracted into an organic solvent and the Ni measured The As was extracted into xylene ASH generated and transferred to the ICP for determination at 228.82 nm detected following chromatographic separation.A device for on-line photo-oxidation reaction was described which converted organic species into As"' within 20 s measurement of total As. Solvent extraction protocols for separation of AS]'' or for combined determination of As1" AsV monomethylarsonic acid and dimethylarsenic acid were also developed 76.4s was used to investigate behaviour of As"' and AsV at different stages of the furnace heating programme and the stabilization effects of various modifiers. With solutions containing C1- As"' could only be stabilized in the presence of HN03 Samples were digested the hydride produced and trapped on to a cellulose filter which was taken for WDXRF 15 As species chromatography. Aliquots (1 0 pl) of fractions were dried on to a polypropylene film for EDXRF analysis In an investigation of the possible role of Some As species do not form ASH and cannot be Pd or Pd-ascorbic acid was used for direct HPLC systems were established for the separation of As species were separated by ion-exchange Reference 901 1 3 84 90lC 1797 9012243 9012277 90J2330 90lC29 76 90lC3 160 90/3478 901C4035 901C4036 9 11326 9 1 JC753 9011 993 9011980 90lC30 16 9013 5 5 3 9014002 911228 9 1 JC5 35 911C652JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 77R Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS- continued Element Matrix As Urine Au Blood urine Au Blood B Biological tissues Be Biological samples Bi Pharmaceuticals Br Urine Br Plasma urine Br Biological tissues Ca Epithelial cells Ca Bone Ca Biological samples Ca Urine Caffeine Analgesics Cd Urine Cd Seminal plasma serum Cd Blood serum urine Cd Urine Cd Urine Cd Lung Technique; atomization; analyte form* MS;ICP;L MS;ICP;L MS;ICP;L AE;ICP;L AA;ETA;L AA;F;L PIXEi-;- MS;ICP;L MS;ICP;L XRF;-;- XRFi-i- MS;ICP;- AA;F air-C,H,;L AA;F N,O-C,H,;L AA;F N20-CzHZ;L AA;ETA;L AAi-i- AA;ETA;L AA;ETA;L AA;F air-C,H,;L AE;DCP;L Sample treatment/comments A low flow of N into the Ar prevented the formation of ArCl which overlaps the 7 5 A ~ peak.It was proposed that Ar-N-N-Cl and N-CI species are formed patients were separated using various types of chromatographic columns coupled to the ICP system patients treated with Au were separated by HPLC coupled to the ICP-MS system tetramethylammonium hydroxide and introduced to the ICP from an electrothermal vaporizer A 200 mg sample was prepared in oxalic acid-HC1 passed through a cation-exchange column and the Be eluted with sulphosalicylate at pH 7.The solution was purified by anion-exchange and taken for determination of Be. Concentrations in original samples were at ng g- levels find best sensitivity for Bi differences between subjects with urinary stones and controls internal standard gave results that were in agreement with expected concentrations. Recovery and reproducibility were good Samples were solubilized with tetramethylammonium hydroxide without losses due to volatilization Cells from patients with cystic fibrosis had high Ca concentrations and defective CI- transport Sr:Ca ratios were measured in prehistoric bones for dietary reconstruction Stable isotopes were used to determine human intestinal absorption of Ca from different foods Elimination of the interference caused by was investigated.20 g 1 - I La with the air-C,H flame gave the most accurate results Phosphotungstic acid formed a precipitate with caffeine. The precipitate was dissolved in NH3-NH,CI buffer and the W measured to calculate the amount of caffeine. Caffeine concentrations in final solutions were 2-40 pg ml-i Urine samples were diluted 1 + 1 with 1.5% HNO,. No other modifier was necessary and calibration was with standards prepared in a pooled urine specimen Increased Cd was found in serum and seminal plasma from heavy smokers but there were no effects on spermatozoa. Cd Se and Zn concentrations in seminal plasma did not correlate An FI system with thermospray sample deposition into a heated cuvette was described.The chemical modifier was introduced first and diluted sample with volumes of 10-500 pl were added 1% HN03 1% (NH,),HPO and l0h NH,F were compared as chemical modifiers to eliminate the matrix effect in urine. NH,F was the most satisfactory (in Japanese) column chromatography and a flame atom trapping device Cd content of lung increased with smoking habits. Concentrations returned to normal on cessation of smoking Samples from gold treated rheumatoid arthritis Au metabolites in blood of rheumatoid arthritis Samples were solubilized at alkaline pH with Designs of slotted tube atom traps were evaluated to Values of 2- 18 pg ml- were reported with no Samples diluted 1 + 9 with 1% HNO and Eu as the Sensitivity was enhanced by use of pre-concentration Reference 91x759 90lC 1450 90/2070 9 llC582 9013 3 3 3 901 148 9013433 9014 1 59 911C583 901 1248 9012 1 19 90lC3 9 8 3 9014033 901C 901 901 901C 901 678 339 367 595 877 9011923 901 194578R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 ~ Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS-continued Technique; atomization; Element Matrix analyte form* Cd Blood 3 3 Cd Gdney liver AA;-;L . . _- Cd Urine Cd Serum Cd Pharmaceuticals Cd Blood Cd Tissues Chlordiaz- epoxide Pharmaceuticals CI Epithelial cells Cr Biological samples Cr Urine Cr Tissues blood urine Cr Plasma Cr Urine Cr Lung Cr Urine Cr Serum cu Urine cu Hair stomach membrane cu Blood c u Urine c u Faecesurine serum c u Kidney liver AA;ETA;L AA;ETA;L AA;F air-C2H2;L AA;ETA;L AA;-;- AA;F air-C2H2;L XRFi-i- AAiETAi- AEiICPi- AA;ETA;L AA;-;- AA;ETA;L AA;ETA;L AE;DCP;L AA;ETA;L AA;ETA;L AA;F;L AA;ETA;S MS;ICP;L AA; ETA;L AA;-;L MS;-;L AA;-;L Sample treatmentlcomments See also Pb ref.9012240 Isoforms of metallothionein were separated by HPLC and quantified by measurement of the Cd Cu and Zn Dei:ection limit was 0.08 pg I-' The agents commonly used for thennostabilization of Cd are rarely Cd free. A procedure was described to remove the contamination by injection of an aliquot of Mg-Pd modifier and heating at 1100 "C. This was followed by injection of the sample on to the purified modifier Chllordiazepoxide (Librium) was reduced with Cd or Zn.The reducing columns were coupled on-line to an FI system and an AA spectrometer. Measurement of the Cd or Zn gave quantitative results for the drug in the range of 2.0-25 pg ml-l with an analysis rate of 150 samples h-I in the presence of other 1 ,Cbenzodiazepines chemical modifier containing (NH,),HPO and Mg(NO,),. Blood was diluted with Triton X-100 and the same modifier added polyacrylamide gel electrophoresis. The gel was sliced to give the different fractions and Cd measured in each Aqueous calibration standards were used with a Cd-binding proteins were separated by See Cd ref. 9 111 2 See Ca ref. 9011 248 Different methods used to examine RMs gave large confidence intervals and possible sources of error were discussed digested with HN03-H202 and oxidized with KMnO (in Chinese) No differences in concentrations of Cr and Ni were found in samples of normal and tumour tissues Cr"' was reacted with trifluoroacetylacetone to form a volatile complex which is lost prior to atomization so that the CrVi could be measured 3 nil of urine in a silica tube were heated to dryness HzOz added and the tube placed into a closed microwave vessel with HN03.Digestion in the acid vapour give a residue which was not acidic Cr content increased with age and smoking habits. Concentrations were unaltered after cessation of smoking Methods were adapted to speed up sample analysis rates A mixture of H and Ar was recommended as the purge gas Intra-individual variation in excretion was shown to depend on seasonal changes in food intakes Washed and dried samples were placed into a small lanthanum coated graphite cup within the tube.Atomization was delayed as with L'vov platforms and graphite probes Recovery of added Cr was 89-1 15% in samples See Au ref. 9012070 See Cr ref. 9012089 Diets enriched with 65Cu were fed to volunteers for 90 d. Faecal and food samples were dry ashed at 500 "C and dissolved in 1 mol dm-3 HNO for measurement of total Cu by AAS or the stable isotope by TIMS See Cd ref. 90129 14 Reference 9012242 90129 14 9013426 9014 148 91/12 911250 91lC636 91112 901 1248 901 1206 9011358 901 136 1 9011415 9OlC 1 680 9011 945 9012089 901C2880 9011 905 9OlC2033 9012070 90/2089 9012 8 6 3 90129 1479R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS- continued Element c u cu c u cu c u c u CU cu F Fe Fe Fe Fe Fe Fe Hg Hg Hg Hg Matrix Serum Hair Plasma Faeces Biological samples Blood Hair Serum Biological tissues Liver tisue Serum Blood serum red cells Plasma red cells Biological fluids Biological samples Biological samples Artificial saliva Hair Blood saliva urine hair Hg Pharmaceuticals Hg Biological samples Hg Biological tissues Technique; atomization; analyte form* MS;ICPL AEiDCPi- AA;F air-C,H,;L MS;-;- MS;-;- AE;DCP;L AA;F;L AA;ETA;L Molecular absorption;-;L AA;ETA;L AE;ICP;L AA;ETA;L AA;ETA;L AE;ICP;L MS.-.- AA;cold vapour;L 7 AA;cold vapour;L AA;cold vapour;L AA;cold vapour;L MS;ICP;L AA;cold vapour;L AA;-;- Sample treatmentlcomments species. Cu was thereby separated from Na+ and PO,,- which interfere with the determination of Cu stable isotopes by MS Rats were fed a diet enriched with W u . The isotope in hair was measured by NAA and the total Cu concentration determined by DCP-AES Pulse nebulization of small volumes of undiluted sample avoided possible contamination associated with preparation and permitted analysis of specimens from neonates extrinsically labelled bread to study intestinal absorption Cu and Fe enriched with W u and 54Fe were used to measure absorption in young and elderly men Recent developments with a tungsten filament vaporizer were described.Microprocessor control of the temperature programme and data collection gave improved detection limits with 200 fg reported for Cu of Cu in hair from healthy and sick children (in Chinese) Non-protein bound Cu was separated by a membrane micropartition system.The concentration of this fraction was measured in maternal and cord blood samples 20 p1 of 0.1 mol dm-) AI(NOJ3 were dried inside the graphite furnace. 10 p1 of test solution were placed into the cooled furnace and the AIF formed when the temperature was increased to 2400 "C. Molecular absorption was measured at 227.45 nm (Pt hollow cathode lamp) conditions were determined were used (in Japanese) 0.05Oh Triton X-100 SEC was used to isolate the Cu protein bound T u was used to provide intrinsically and No difference was found between the concentrations Concentrations of Fe in different pathological Analysis was improved when deproteinized samples The samples were diluted 100- or 10000-fold with Samples were diluted with 0.05% Triton X-100 HPLC coupled to an ICP-AES system was used to See Cu ref.90/4 133 0.1 g of sample with 1 ml of HN03 were heated under pressure in a microwave oven. Hg2+ in the digests was reduced concentrated by Au amalgamation and then measured Dissolution of metallic Hg into various solutions was determined after incubation for up to 14 d Hg from moustache samples was heated at 800 "C and collected on to porous Au. The Hg was thermally desorbed into a 20 cm path length light cell (in Japanese) SnClz was added to undigested samples and mixed for 1 min before liberation of the Hg vapour. Results were superior to those from a procedure which used immediate liberation with better recovery greater sensitivity and use of less reducing agent.Calibration against aqueous standards was possible Preparations were diluted with H20 and the Hg species separated by HPLC to determine thiomersal decomposition products digestion within a high-pressure bomb separate ferritin-containing drugs Similar results were given by cold mineralization and - Reference 9013497 9013 329 901C404 1 9014 1 3 1 9014 133 9 1/36 9 11438 911824 9013508 9011 327 9011 395 901C1797 901C2 109 9012244 9014 133 901 1 179 901 1 330 901 1 944 901 1949 901207 1 9013293 901333780R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS -continued Technique; atomization; analyte form* AA;ETA;L Reference 911386 Sample treatmentlcomments Loss of Hg during the pre-atomization steps was reduced with a Pd-Ni modifier (in Chinese) Samples were wet ashed the I precipitated with AgCI and dried as AgNH41 on to a tantalum filament.The I was thermally desorbed in a laser resonance mass spectrometer. Isotope ratios were measured in samples previously equilibrated with 1291 ldentical to I ref. 90/2300 See Br ref. 9014 159 See Br ref. 91/C583 A diluent with Li as internal standard was used. Accuracy measured with RMs was + 5 and +0.4% with aqueous and matrix matched calibration respectively (in Japanese) Cells were harvested by density gradient centrifugation lysed and diluted for analysis Methods were compared. AE was less accurate than AA KH,PO as a modifier produced a large increase in sensitivity with a detection limit of 0.5 pg I - ' (in German) background correction was used.Reference values for serum and urine were 0.16 pmol dm-' and 5.24 pmol d-I respectively :Interferences associated with C1- and NO3- were studied. No interferences were found with a tungsten-coated tube and NH,N03 or (NH,),HPO as modifier (in Chinese) Dilution of samples with 1% HNO was suitable for determination of Li only. For multi-element measurements it was necessary to digest samples with acid Proteins were precipitated from serum with 10% HN03 urine specimens were diluted with 5% NH,NO,. With tantalum-coated pyrolytic graphite tubes the maximum ash temperature was increased to 1200 "C. Background correction was unnecessary. A detection limit of 0.05 pmol 1-' was achieved with 25 p1 of sample centrifigation procedure suspended in H20 and diluted with La buffer relating to precision of analysis and the clinical usefulness of results IVH,N03 was used as chemical modifier and Leucocytes were harvested by a careful multiple This letter commented on previous observations See K ref.91/825 Matrix Hair Element Hg 9012300 MS;ETA;L Oyster I 9012353 9014 1 59 91lC583 901 1 506 Biological samples Plasma urine Biological tissues Serum MS;ETA;L MS;ICP;L MS;ICP;L AE;F;L I I I K AE;F;L 911825 9011 876 9011 878 K Li Li Mononuclear cells Serum erythrocytes AA;F;L AE;F;L AA;ETA;L Serum. urine AA;ETA;L 9013534 Li Serum urine AA;ETA;L 9 11407 Li Blood Li Li Serum Serum urine MS;ICP;L AA;ETA;L 9 1lC799 9 11C8 13 901 1484 Leucocytes AA;F;L Mononuclear blood cells 9014034 .. -- I AA;F;L AA;ETA;L 911825 901 1203 Mg Mn Mononuclear cells Serum erythrocytes Four sample preparation procedures were evaluated. Dilution 1 + 4 with 0.1 mol dm-3 HN03 gave best results with superior sensitivity and accurate values with aqueous calibration. The other methods (digestion freeze-drying and dissolution) required matrix-matched standards to provide accurate results This study confirmed that contamination can derive from the needles used for blood collection Interferences from alkali and alkaline earth elements were removed with a robot system for chelation chromatography autosampler cup containing 5% HN03-0.04% Triton X- 100. Measurements were made at 403.1 nm to reduce sensitivity Platelets and blood cells were separated by discontinuous gradients of colloidal polyvinylpyrrolidone-coated silica Dried powdered sample was placed into an Mn Mn Blood Serum AA;ETA;L LEI;-;- 901 14 13 9OlC 1765 AA;ETA;SI 9013294 Mn Biological samples Blood cells AA;ETA;L 911324 MnJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 199 1 VOL.6 81R Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS-continued Technique; atomization; Element Matrix analyte form* Mn Biological samples AA;ETA;L Mn Blood urine Mn Hair Mo Urine faeces AA;ETA;L AA;ETA;L AA;ETA;L Mo Plasma AA ; ETA; L Mo Biological samples Mo Breast milk Ni Biological tissues Ni Tissues blood urine Ni Serum Ni Pharmaceuticals Ni Serum Ni Urine Ni Urine Ni Urine Ni Blood serum hair Ni Serum urine 0 Urine P Biological samples AA;F N,O-C,H,:L AA;ETA;L AA;ETA;L AA;ETA;L AA;-;- LEI;-;- AA;-;- AA;ETA;L AA;ETA;L AA; ETA;L AA; ETA;L AE;M I P; L MS;ICP;L MS;-;- AA; ETA; L Sample treatmentlcomments Sample preparation instrumentation and matrix interference were investigated.L-Ascorbic acid as a suppressor was recommended - Hair was used to monitor exposure from an industrial source Good recovery and precision results were obtained with methods that required minimal sample preparation HNO3-HC1O4-H2SO4. Excess of acid was boiled off the pH adjusted to 1.6 and the Mo complex with 8-hydroxyquinoline was formed. The complex was extracted into CHCI dried and re- dissolved with Tris buffer at pH 7.95. The sensitivity was 3 ng ml-' with a 1 ml sample Samples were ashed at 500 "C the Mo dissolved in HCl and complexed with ammonium dithiophosphoric acid U,U-diethyl ester.The solution was concentrated by passage through filter-paper covered with activated charcoal. The Mo was desorbed from the carbon with HNO and taken for analysis Milk was analysed without treatment. A modifier with BaF was used and a detection limit of 0.89 pg 1-I was obtained HN03-HC1-H202. Control of contamination was described and recovery of three Ni compounds was found to be 92- 106% Plasma or serum samples were digested with Microwave heating was used to digest samples with See Cr ref. 901 1361 See Mn ref. 9OlC1765 Alkaloids were indirectly measured. The alkaloids (atropine pilocarpine) with Ni(SCN),,- were extracted into an organic solvent and the Ni measured Serum samples from healthy individuals and patients with renal failure were diluted with 1% Triton X- 100-1 mol dm-) HNO,.50 pl were injected into the graphite furnace. Extensive precautions to maintain cleanliness were described and normal concentrations were much lower than has previously been reported 0.14 f 0.09 pg I-' Urine was diluted 1 + 2 with HN03 Detection limit was 0.1 pg 1 - I Urine was diluted 1 + 1 with 2% HN03-0.001% Samples were digested with acid and the Ni2+ Triton X-1 00 reduced by BH,' to NiO reacted with CO to produce the volatile carbonyl which was trapped on Chromosorb. Rapid desorption by heat released the carbonyl into the MIP. The detection limit was 3 pg Good sensitivity is compromised by formation of CaO which interferes with the Ni isotopes. Principal components analysis was successfully used to correct for the presence of Ca in samples spectrometric procedure was used to determine ' 8 0 A molybdenum tube atomizer was used with measurements made at 2 13.6 nm.A high temperature pyrolysis stage eliminated chemical interferences. The detection limit was 2.3 ng ml-I with a characteristic mass of 1.9 ng P A continuous fast flow isotope ratio mass Reference 911414 91lC654 91lC657 9011283 901208 1 9013419 911251 901 1360 901 1 36 1 9OlC 1 76 5 901 1993 9012082 9013290 9013426 9013552 9 1 I248 9 llC666 9012329 9012 19582R JOURNAL. OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991. VOL. 6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS- continued Technique; atomization; analyte form* Sample treatmentlcomments AA; ETA; L Samples digested with H2S04 HN03 and K,S2O8 were extracted with dithiocarbamate in CHC13.The Pb was back-extracted with an Hg" solution. This procedure eliminated interference from CI- of Pb n-butanol (in Chinese) and SO:- and increased the concentration Tissues were homogenized with H20 and treated with See Cd ref. 9OlC1595 Hair samples from 263 children aged 1-9 y and who lived within 900 m of a Pb smelter were analysed Calibration by standard additions was not necessary Dilution with NH,H,PO,-HN0,-Triton X- 100 gave results with 'legally defensible accuracy' for occupational monitoring purposes X-100 was atomized from a L'vov platform after an O2 ash stage Improved results reported by participants in an ongoing inter-laboratory quality assurance programme were noted.The same RMs were used for internal control by all participants. See Pb ref. 9012241 Fresh human blood and control samples gave comparable performance within the furnace Multiple absorption peaks are derived from the various Pb species in urine and modifiers to simplify the analysis were studied. Pd(N03)2-Mg(N03)2 was superior to either agent alone and to NH,H2P0 10 pi placed on to the L'vov platform with 10 pl of modifier (2 mg I-' Pd-2% citric acid-0.01 mol dm-3 HNOJ. O2 ashing was also used Separation of Pb species was achieved by HPLC Quality control data were used to compare methods. ETAAS offered superior precision to that of ASV methods In children concentration differences associated with age sex residential area and occupation of father were found sample preparation were measured to identify the sources of exposure in cases of childhood Pb poisoning Pb concentrations in samples from 200 Saudi schoolboys aged 6-8 y were 6.9 f 3.4 pg dl-' A lanthanum chemical modifier was used (in Chinese) Comparison of I 1 chemical modifiers was reported.The most useful were H3P04 (NH4)H2P0 Pd and Pt (in Chinese) Biological fluids were analysed without treatment. Tissues were heated with HN03-H202 at 85 "C for 1 h (in Chinese) Siimples were taken from patients treated with carboplatin and the Pt species separated by HPLC Pharmacokinetics and protein binding of Pt compounds used to treat cancers were investigated Pu concentrations and isotope ratios were measured in samples from persons living around a nuclear fuels plant Lfow concentrations of Sb"' were extracted with N - p methoxyphenyl-2-furylacrylohydroxamic acid in CHCl to provide detection at 1 x lo-* ppb Blood diluted with NH4H2PO,-Mg(N0J2 and Triton Blood was diluted 1 + 9 with 0.1% Triton X-100 and A critical review of interferences and methods for Pb isotopes in blood and environmental materials See As ref.911228 See Cd ref. 9011367 Element Matrix Pb Biological samples Reference 901 1499 901 547 901C 595 901 838 90/1922 9012088 9 0 x 2 100 9012240 AA;ETA;L Pb Tissues AA;ETA,L AA;-;- Pb Blood serum urine Pb Hair Pb Urine Pb Blood AA;ETA,L AA;ETA;L Pb Blood AA;ETA;L Pb Blood 901224 1 9OlC2958 Pb Blood Pb Urine AA;ETA;L AA;ETA;L AA;ETA;L 9OlC30 19 Pb Blood 901C3235 9013425 Pb Pb Biological samples Blood M S; I CP; L A A; ETA; L Pb Hair 9013435 Pb Pb Blood Blood AA;ETA;L MS;ICP;L 90/3549 9 112 Pb Blood AA;ETA;L 9 11308 Biological samples Blood AA;ETA; L AA;ETA;L 9 11328 9 1 I422 Pb Pb Pt Biological samples AA;ETA;L 901 1 394 Pt Biological fluids AE;ICP;L 90lC 18 10 Pt Pu Plasma urine Tissues AE;ICP;L MS;-;- 9012 2 4 5 9012808 Sb Tissues AA;ETA;L 9011 473 Sb Se Biological samples Seminal plasma serum XRF-;S AA;-;- 9 11228 901 1 367JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991.VOL. 6 83R Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS - continued Technique; atomization; analyte form* AA;ETA;L Element Matrix Se Blood Sample treatmentlcomments Low concentrations of Se were found in samples from Careful digestion with HNO H,SO and HCIO at a sick neonates final temperature of 310 "C in special flasks was necessary to give reliable results A molybdenum coated graphite tube and a L'vov platform were used with Ni(N03) modifier.Concentrations of Se were unaltered by running a marathon (in German) were analysed with a Pd-Cu modifier pregnancy. Concentrations of Se decreased with gestational age After a comprehensive systematic investigation of digestion chemical modification calibration and type of purge gas an accurate method was described that employed microwave digestion with HNO P~(NO,),-CU(NO,)~ msdifier and addition of Triton X-100. Direct and standard additions calibration with Zeeman-effect background correction gave equivalent results for a RM HCIO,. The digests were applied to an FI hydride generation system to allow use of small volumes with rapid sample throughput.Results compared well with those given by ETAAS Blood samples collected and dried on to filter-paper were sent from sites remote from the laboratory. Discs equivalent to 17 pl of blood were digested and taken for analysis Samples were diluted with 0.2% HNO,-O.2% Triton X-1 00. Matrix matched calibration standards were prepared. A chemical modifier with CU(NO~)~ and Mg(N03),.6H20 was used developed. An electrically heated quartz tube in the light path was used to atomize the evolved hydride. An Ar-H flame ignited at the tube outlet and enhanced the atomization process. The design provided a detection limit of 27 pg Samples digested with HNO by microwave heating Samples were collected from 14 to 24 weeks of Samples were digested with HNO H,SO and A hybrid atomizer with two components was See As ref.911228 Tissue samples were digested with HN03 HzS04 and HCIO,. The H,Se was formed with NaBH and transferred to the Ar plasma A mixed Pd-Pt modifier permitted measurement of Se in the presence of Fe and PO,'-. Deuterium- arc background correction was employed Good agreement with results obtained by NAA was observed at concentrations of 100- 12000 pg g- Organosiloxane released from hair by an enzyme digestion procedure was extracted into an organic solvent for the measurement of Si to assess deposition of silicone hair products Organotin species were separated by HPLC coupled to the ICP-MS system for detection Specimens digested with HNO and HCIO were taken for hydride generation Samples were diluted 20- to 50-fold with dilute HN03.Concentrations of 16 and 158 pg I - ' were found in serum and urine of normal subjects Sr and Zn were measured by EDXRF in enamel and dentine of teeth from subjects aged 9-64 y See Ca ref. 90121 19 86Sr was used to study metabolism of Sr. Gender and age related differences in uptake were observed Reference 9011 383 901 1408 9011 55 1 9OlC 175 1 901 193 1 9012057 9012 194 90129 1 3 9014030 9014 1 52 9 11228 91131 1 91lC814 901 I 198 901 138 1 9OlC3 197 91lC524 9011483 9011 854 9012 1 19 91lC538 Se Biological fluids AA;Hy;L Se Plasma AA;ETA;L Se Liver Se Amniotic fluid AA;ETA;L AA;Hy;L Se Liver AA;ETA;L Se Plasma and serum AA;Hy;L Se Blood Se Serum blood AA;Hy;L AA;ETA;L Se Tissues AFHy;L Se Biological samples Se Biological samples XRF-;S AE;ICP;L Se Blood urine milk AA;ETA;L Si Biological samples Si Hair Sn Biological samples Sn Biological tissues MS;ICP;L A€; 1CP;L AA;ETA;L Sr Plasma urine Sr Teeth XRFi-i- Sr Bone Sr Blood urine saliva XRF;-;- MS;ICP;L84R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS-continued Element Matrix Te Urine Th Biological samples T1 Blood serum urine T1 Urine U Biological samples V Biological tissues W Analgesics Zn Platelets lymphocytes granulocytes Zn Seminal plasma serum Zn Teeth Zn Urine Zn Blood Zn - Zn Faeces Zn Kidney liver Zn Biological samples Zn Plasma Zn Pharmaceuticals Zn Urine Zn Hair Various (1 0) Biological tissues Various (5) Nerve cells Various (1 1) Urine Various (7) Biological tissues Various (10) Blood bile Various Biological tissues Various (6) Tissues blood Various (6) Milk Various (8) Serum urine Various Urine Various Biological samples Various Hair Technique; atomization; analyte form* AA;ETA;L MS;ICP;L AA;ETA;L AA;ETA;L MS;ICP;L AA;F N,O-C,H,;L AA;F N,O-C,H,;L AA;ETA;L AA;-;- XRF;-;- AA;FL MS;ICP;L MSi-i- MS;-;- AA;-;L MS;ICP;- AA;F air-C,Hz;L AA;F air-C,H,;L AE;ICP;L M F L AA;ETA;L XRF;-;- AE;ICP;L - AE;ICP;L AA;ETA;L AA;-;- AE;ICP;L AA;F;L AE;FL AA;ETA;L Sample treatmentlcomments Urine was digested with HN03-HC104 and extracted with IBMK.The organic phase was removed for analysis and values of (0.1-1.2 pg I - ' were found Bi and TI were used as internal standards See Cd ref. 9OlC 1 59 5 A method with a detection limit of 0.5 pg 1-' was See Th ref.9011978 An on-line anion exchange resin column was used to concentrate the V. Silica-bonded phases were examined for separation of different V species A method for the indirect determination of caffeine Platelets and cells were isolated from blood and digested with HN03. Matrix effects were small with peak-area measurements but calibration by standard additions was used mentioned See Cd ref. 9011367 See Sr ref. 9011854 Intra-individual variation in excretion was found to See Au ref. 9012070 Intestinal absorption of Zn was measured by stable isotope analysis by FABMS Zinc absorption from enriched cereals was assessed. Concentrations of Zn isotopes were slightly different when determined by FABMS compared with TIMS depend on dietary intake See Cd ref.90129 I4 See Ca ref. 9OlC3983 See Cu ref. 901C4041 See Cd ref. 9 11 12 Excretion of Zn was determined in 1472 female subjects. No effects due to age or smoking were found but regional variations were evident Thr Zn concentrations in hair from sick children were lower that those from healthy children (in Chinese) Homogenized lyophilised tissue was heated with IR radiation and ashed at 450 "C. The ash was dissolved by heating in 2 mol dm-3 HN03 at 95 "C for 5 min (Al Cd Co Cr Cu Fe Mn Ni Pb Zn) Distribution of elements in hydrated and dehydrated cells was investigated (Ca C1 K Na P) Spectrum shift background correction was used. Typical operating conditions were established (As Be Cd Co Cr Cu Hg Mn Ni Pb Zn) A microwave acid digestion procedure was described and evaluated Different diluents were evaluated (in Japanese) This review described how RMs were used to assist the development of improved analytical methods The effects of Pb and Sn on the tissue concentrations of essential trace elements were measured (Ca Cu Fe Pb Sn Zn) Proteins in breast and cows milk were separated by gel filtration to study the binding of metals (Ca Cu Fe Mg Mn Zn) FI analysis was compared with conventional flame .4A and AE methods (Ca Cu Fe K Li Mg Na Zn) shown to be similar to that of NaC1.Background interferences were reduced by addition of HN03 (in Chinese) The background signal associated with urine was See Various ref. 901135 1 Use of RMs was discussed Reference 9013950 9011 978 90lC1595 9012088 9011978 901C1767 90/C 1678 901 1 326 9011 367 9011854 901 1 905 9012070 9012820 901282 1 90129 14 9OlC398 3 90/C404 1 91/12 911268 911438 901 1 180 901 1249 9011288 9011291 901 13 1 5 901 135 1 901 1 368 90114 14 9011 549 9011553 90/C 1639 90/C 1 64085R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS- continued Technique; atomization; Element Matrix analyte form* Sample treatmentkomments Various Tissues AE;ICP;L Tissues were digested with H2S0 HNO HClO and ultrasonic nebulization was used for efficient sample introduction that included a Hildebrand gnd nebulizer. Suppression caused by the high Ca concentrations was compensated for by matrix matched standards (Ca Cu Fe K Mg Mn Na P Zn) Various (9) Bone AE;ICP;L Bone digests were analysed in an optimized system Various Biological samples Various (9) Biological tissues Various (16) Serum Various (6) Human tissues Various (1 6) Human tissues Various (48) Lung Various Biological samples Various (1 3) Serum Various (4) Brain Various (6) Biological samples Various Tissues blood urine Various (8) Lung Various (8) Lung Various (8) Faeces urine Various (8) Faeces urine Various ( 5 ) Hair Various (1 1) Kidney stones MS;ICP;L MS;ICP;L AA;F;L AA;ETA;L XRF-;- AE;ICP;L MS;ICP;L MS;ICP;L AE;ICP;L AE;ICP;L AE;ICP;L AE;ICP;L AE; I C P; L MS;ICP;L AE;ICP;L MS;ICP;L XRF;-;- Programmed computer control was described to give reproducible conditions for microwave-oven digestion heating was used.HClO was avoided to limit the formation of species which interfere with the determination of As Cr and Zn (As Cd Cr Cu Hg Ni Sr Pb Zn) The analysis of a second generation RM was examined.Internal standards and a PTFE introduction device were used to overcome some of the systematic errors High concentrations of Hg and Se in tissues from subjects who lived in the Faroe Islands were attributed to dietary intake of whale meat (Ca Ca Fe Hg Se Zn) Concentrations of some elements were different in certain malignant tissues compared with normal samples Samples were prepared using titanium nitride coated instruments to reduce contamination Values were obtained for a large number of elements for which little previous information was available in RMs An internal standard Il5In was added and the specimen diluted 5-10-fold with 0.14 mol dm-3 HNO Samples were collected from 12 regions of the brain and digested with HN03 in a PTFE pressure vessels (Cu Mg Mn Na) An HPLC system was coupled to the ICP for the measurement of Ca Cu Fe Mg and Zn.Se species were determined in serum breast milk and erythrocyte lysates This compilation of available RMs is an addition to lists produced earlier (see ref. 901140) Results were compared with those obtained by NAA. With careful analytical technique both methods gave similar results (Al Cd Cr Cu Mg Mn V zn) Tissue and HN03 were heated at 400 "C. Residues were dissolved in HN03 and diluted with H 2 0 Samples were frozen lyophilised and ashed at 490 "C. The residues were fused with LiBO and the solid dissolved in 5% HN03.This paper gives methods for details of results see Part 2 ref. 9012065 (Al Ba Mn Si Ti V Y Zr) were compared with the amounts measured in the diets (Al Ba Mn Si Ti V Y Zr) See ref. 90/C539 (in 1990 review) (Cu Fe Ni Pb Zn) evaluate possible factors involved in their formation Digestion in sealed containers with microwave See ref. 9012064. Recoveries of elements in faeces Hair samples were digested with HNO and HClO,. The compositions of kidney stones was measured to Reference 90/C 1650 901C1756 90/C 1762 90lC1763 90/C1811 9011833 901 187 1 90/C 1898 90/C 1899 901C 1900; 90/C200 1 90/C 1 90 1 901 1902 901 1959 9011 978 90/ 1979 9012064 9012065 9012 1 15 9012 13986R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS - continued Technique; atomization; Matrix analyte form* Element Various ( 9 ) Serum XRF;-;S Various ( 4 ) Serum Various (24) Biological samples Various (1 1) Biological samples Various ( 12) Biological samples Various (1 8 ) Hair Various ( 6 ) Heart thyroid red cells Various Teeth bone Various Teeth Various ( 9 ) Biological samples Various Biological samples Various ( 9 ) Serum blood milk urine Various ( 6 ) Amniotic fluid Various Biological samples Various ( 9 ) Serum Various ( 1 3 ) Blood serum Various Clinical and biological Various Biological samples samples foods Various (1 6) Biological samples XRF;-;- MS;ICP;L AA;ETA;L AA;ETA;Sl AA;ETA;L AE;ICP;L AE;ICP;L MSi-i- MS;-;- MS;-;- AA;ETA;S AE;ICP;L AA;ETA;L AE;ICP;L AA;ETA;L AE;ICP;L AE;ICP,L Sample treatmentlcomments components were minimized by pre-treatment with a low-temperature asher prior to TXRF analysis (Ca Cu Fe K Pb Rb Se Sn Zn) (in German) Small amounts of sample were analysed with minimal pre-treatment.See ref. 891201 3 (in 1990 review) (Br Cu Fe Zn) A procedure for sample preparation was specially developed. HN03 digestion with microwave heating under pressure was performed with half the sample the container cooled in a freezer and the remainder of the sample added for a second digestion step. Acid was removed by boiling and digestion completed by additions of HN03 and Interferences caused by organic and inorganic H2OZ Direct analysis of solutions wet and dry ashing and slurry preparation were used for single element and simultaneous multi-element analysis of RMs Ciireful analysis of RMs was described to provide reliable reference values Reference ranges for subjects less than 15 y old were established Frozen tissue samples were placed in an ion microscope for analytical measurement and localization of the elements by SIMS (Ca Fe K Li Mg Na) A review of the applications of SIMS to the analysis of biological hard tissues SIMS was used for high-resolution (50 nm) imaging of elements in teeth The elements in a sample were electrolytically deposited on to a L'vov platform which was placed into the furnace and heated to release the atoms (Ca Cd Cr Li Mg Mn Ni Pb Zn) A rapid method with microwave heating which provided enhanced concentrations was referred to but not described Samples were diluted I + 3 with 0.IW Triton X-100 and common conditions could be used for many measurements only adjustments to the atomization temperatures were necessary (Al As Bi Cd Cr Mn Ni Pb Se) Samples were untreated.Standards contained 0.5% albumin and 0.76% NaCl (Ca Cu Fe Mg Mn Zn) HC1-HN03-HF was presented for the measurement of nearly 25 elements Proteins were precipitated by addition of 4 ml of a mixed acid solution to give final concentrations of 0.28 mol dm-3 HN03 and 5Oh CC1,C02H. Samples were centrifuged and the supernatant taken for analysis. Variable results were obtained when compared with other procedures (Ca Cu Fe K Mg Na P Si Zn) (in Japanese) TXRF was applied to samples digested with HN03 in a microwave oven. Limits of detection were around 0.08-0.2 pg ml-' The 1989- 1990 ASU review A microwave dissolution procedure with 1.'; ml of water were added to 0.5 ml of sample. -- .. 9 AA;F;L Developments in sample decomposition,low Results for several sample types and preparation contamination of reagents etc. were discussed procedures with a SIMAAC spectrometer were given. See also Various ref. 9012199 AA;ETA; L AA;ETA;SI Reference 9012 147 9012183 9012 193 9012 199 9012231 9012 2 3 9 9012568 9012652 9 012 8 24 90K2944 901C2948 9OlC29 83 9013 326 9013327 9013369 9013422 9013500 9013559 9013584JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 87R Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS -continued Technique; atomization; analyte form* Sample treatmentlcomments Reference 901 3 9 9 4 9014032 AA ; ET A; S AE;ICP;L Solid samples were succesfully analysed on a graphite A diluent containing CC13C02H and HCI for protein probe (Cd Cu Mn Pb Zn) (in German) precipitation hydroxylamine sulphate to reduce Fe and Y as internal standard were used.Recoveries of 92-10796 were obtained (Ca Cu Fe K Mg Na P Zn) V and mixed V + Pd modifiers were investigated and found to be efficient thermostabilizers for a wide range of volatile elements in slurried and liquid samples benzothiazolylazo)-6-bromo-2-naphthol (Co Cu Fe Ni Zn) (in Japanese) patients with chronic renal failure were determined (Al Ca Cu Fe Mg P Si Sr Zn) (in Japanese) Urine was diluted 10-fold with 2% vlv HNO aspirated into the ICP and the mass range 2-256 u was scanned every 2 s for 4 min.Detection limits were less than 1 ng ml-I and there was good agreement with results obtained by other methods for Cd Hg and Sb. Interferences limited the usefulness of the procedure in the range mlz = 10-80 Lyophilised samples were mixed with H,O and Zr beads for 1-4 h to form slumes which were injected into the graphite furnace. An air-ashing stage included in the temperature programme reduced the non-specific absorption and allowed the use of higher ash temperatures. Good agreement was obtained with certified values for a series of RMs (Cr Co Mn Pb) Digestion with HN03 and HClO gave good results but 1 O-fold dilution produced anomolous values (Ba Ca Cu Fe Mg Mn Sr Zn) (in Chinese) Rapid accurate precise results were obtained with 0.1 ml of blood dried on to filter-paper (in Chinese) Factors such as sex age hair colour time of year scalp or pubic hair influenced the levels of Cd Cu Pb and Zn centrifugation (Ca Mg Na K) from which the lipid fraction had been extracted.Significant amounts of some elements were removed with the lipid and could be lost if samples are cleaned with detergents 9014 145 Analytes were co-precipitated with 142- 9014 165 Trace element concentrations in hair samples from 9014 166 9113 Lymphocytes were isolated by density gradient Elements were determined in washed hair and in hair Tissues were homogenized with H20 and lyophilised Element Matrix Various ( 5 ) Biological samples Various (8) Serum Various (25) Biological tissues AA;ETA;SI AA;ETA;L Various ( 5 ) Liver Various (9) Hair XRF-;- AE;ICP;L Various Urine MS;ICP;L Various Biological samples AA;ETA;SI 9116 Various (8) Serum AE;ICP;L 911185 Various (1 1) Blood 911188 XRF;-;S Various (4) Hair 911265 Various (4) Lymphocytes Various (1 3) Hair AA;-;L AE;ICP;L 911327 911325 9 11349 911352 Various Biological tissue Various Biological tissue SR with a spectrum of X-rays up to 40 keV was used as a microprobe beam focused on to 10 x 10 pm2 sections described and the importance of quality control was emphasized Changes in trace element concentrations were found in brains from patients with atherosclerosis compared with the controls (Ca Cu Fe K Mn Rb Se Zn) Optimization of instrumental parameters was SRXRF was superior to that of a PIXE technique PIXE;-;- 911353 Various Biological samples Various (8) Brain PIXEi-;- 911354 Various Biological samples Various (20) Hair XRF;-;S PIXE;-;S AE;ICP;L 9 1 I366 9 11452 A procedure to wash and digest hair specimens was described (in Chinese)88R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 Table 1 SUMMARY OF ANALYSES OF CLINICAL AND BIOLOGICAL MATERIALS -continued Element Matrix Various Biological samples Various Biological samples Various (1 3) Serum Various (7) Breast milk Various ( 5 ) Muscle Various (10) Bile Various Biological tissues Various (43) Biological tissues Various Biological samples Various Biological samples Various Biological samples Various ( 5 ) Biological samples Technique; atomization; analyte form* MS;ICP;L AE;MIP;L MS;ICP;L MS;ICP;L AA; ETA; L AE;ICP;L AE;ICP;L AA;ETA;L AA;ETA;L MS;ICP;L AE;ICP;L Sample treatmentlcomments Advantages of ICP-MS in marine analytical chemistry such as interfacing with HPLC multi- element detection sensitivity were discussed Samples were vaporized from a graphite furnace into a toroidally shaped MIP maintained inside a quartz tube.Detection limits were 10-fold lower than are obtained with a conventional MIP Serum samples were diluted 5-10 times with 0.14 mol dm-3 HN03. I151n was included as an internal standard for the analysis of samples from 12 normal subjects Changes in concentration at 2-200 d after birth were monitored in milk skim milk and whey Wet-ashed muscle was analysed. Concentrations were higher in soleus (red) than in extensor digitorum longus (white) muscle (Al Cu Fe Mn Zn) Concentrations in gall bladder and hepatic bile from patients and controls were determined Uses for RMs in the maintenance of analytical quality control were described An ‘elemental fingerprint’ was devised for each of ten tissues A series of RMs developed for marine monitoring programmes were described An approach to controlled microwave digestion was presented with selection of descriptors for each type of sample.The objective was to develop reproducible standard methods Clinical applications of atomic spectrometric analysis were described Tissues were digested the thiocarbazohydrazone complex formed and a 30-fold concentration step achieved by extraction into organic solvent (Cd Co Cu Ni Zn) Reference 9 1lC480 91lC491 9 IlC509 91x531 91x527 91lC528 9 1 lC590 91lC601 9 1 lC646 91lC736 9 1lC739 9 1 lC803 *Hy indicates hydride generation and S L G and S1 signify solid liquid gaseous or slurry sample introduction respectively.Other abbreviations are listed elsewhere. a difference in applied temperature in the ashing stage affect- ing the result. More importantly a greater day-to-day variation was seen in the peak height of standards than for serum quality-control samples. The study showed that the addition of KNO reduced this variability. The final method involved di- luting standards and samples 1 + 2 with a diluent containing 1.5% v/v HNO 10 g I-’ K and 0.05% Triton X-100. The determination of Al in dialysate concentrates presents a formidable challenge in that a low concentration of Al must be determined in a high concentration of inorganic salts.Delves’ group (9 1/C739) overcame this problem by collecting the Al on a chelating ion-exchange microcolumn and eluting with HNO for determination by ETAAS. The determination could be carried out on 1 ml of sample in under 10 min with a detection limit of 1.3 pg I-’. A similar approach built as an F1 system was used by Pereiro Garcia et a / . (90/3478). Detemi- nation was by ICP-AES or by FAAS with an N,O-C,H flame. Of a range of resins investigated Chelex 100 and Amberlite IRA-400 microcolumns were found to be the most suitable. The required sensitivity could be reached by increasing the sample volume; for determination by FAAS at low concentra- tions sample volumes of 25 ml were used. The work of Davenport et al. on urinary Al e-.uc*r*efion fol- lowing renal transplantation covered in last year’s Update (90/ 3500) has now been published (90/2277).Patients formerly on dialysis showed a rise in urinary AI excretion after transplanta- tion increasing to a maximum of 9.7 -t 2.4 pmol per 24 h on che sixth day after operation. Thereafter urinary Al excretion decreased but at the end of the study (14 d) it was still in excess of normal excretion. Measurements were by ETAAS. The value of methods that can localize Al in tissues was demonstrated by Vandeputte et al. (90/2330). Using LMMS it was possible to discover that Al was associated with large cells in the livers of rabbits treated intravenously with aluminium rnaltol. Van Ginkel et al. (91/326) studied ways of improving the determination of Al in brain tissue by ETAAS.Better condi- tions of wet digestion were established and contamination was reduced to a satisfactory low level. The presence of non- digested fatty residues did not affect the accuracy of the method. The final method gave a detection limit of 5 ng g-’ wet inass in tissue a within-batch precision of 4.8% and a between- batch precision of 5.5% RSD. The Al content of human cortex samples ranged from 0.14 to 0.22 pg g-’ wet mass. High pres- sure bomb digestion with 6 mol dm-3 HNO was the technique preferred by Parra and Romero (90/C2 1 10) for preparation of brain and bone samples prior to determination by ETAAS. As expected concentrations of Al found in the brain and bone of patients on dialysis were higher than controls. I .7.2. Arsenic Methods for the determination of As by hydride generation generally involve lengthy sample digestion to break down organic As compounds to inorganic As.The original approachJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL I99 I VOL. 6 89R of Kalman and Attallah (90/C3016) was to apply photo- oxidation in a continuous flow system by passing the sample solution with potassium persulphate in NaOH through 4 m of tubing wound around a low-intensity mercury lamp. This was shown to break down arsenobetaine monomethylarsonic acid dimethylarsinic acid and phenylarsonic acid efficiently. This system when coupled to an anion-exchange column allowed speciation of As in biological and environmental samples. With the aim of optimizing the determination of As in urine serum and hair digests by ETAAS Krivan and Arpadjan (90/4002) used radiotracers to study the stability of As'" and As' in the graphite furnace during the heating programme. Tri- valent arsenic showed less stability than AsV in the absence of a modifier.Nickel or Pd increased the stability but in chloride- containing matrices such as urine the stability was improved by the presence of HNO or H,O,. The technique was applied to the determination of As in hair and the result obtained was in good agreement with that obtained by NAA. In a further publication of recent work Subramanian (90/ 3553) described the difficulty of directly determining As in urine by ETAAS caused by over-correction of the background absor- bance. The problem was overcome by solvent extraction of the chloride complex from the urine in 9 mol dm-3 HCI into toluene in the presence of KI.The As was back-extracted into a solution of Ni in dilute HNO,. This approach determined inorganic and organic forms of As except for arsenobetaine which was not dis- sociated in the process. More information can be obtained by complete speciation of As in urine. Branch et al. (91/C759) carried out speciation by HPLC-ICP-MS of the urine of volun- teers who ate fish-based meals. Isobaric interference from 40Ar"5C1 in the ICP was overcome by introducing a low flow of NZ into the plasma gas. Quality of results was demonstrated by deter- mination of As in a number of CRMs and comparison of total As measurements in urine with the sum of the concentrations of the individual species. Similar studies were carried out by Shibita and Morita (91/C535) again using HPLC-ICP-MS.Three differ- ent chromatographic systems allowed the identification and quantification of 15 arsenic species. One was a gel-permeation column and the other two used ion-pair chromatography for the separation of cationic and anionic species respectively. Arse- nobetaine was the predominant species seen on the samples of urine from normal volunteers. Inductively coupled plasma MS is ideal as adetector for HPLC as direct coupling is straightforward. Favier er al. (91/C652) used EDXRF for their measurements re- quiring the collection of fractions. A small volume (10 @) of the fraction was deposited on to a polypropylene film and evaporated for measurement of As. Separation was carried out on an anion- exchange column.1.7.3. Beryllium A combined ion exchange-ETAAS procedure was developed by Hayashibe et al. (90/3333) for the determination of Be in several biological RMs. Beryllium in the sample digests was collected on a cation-exchange column eluted and further purified by anion exchange followed by determination by ETAAS without a modifier. Application of the method to biological RMs demonstrated the ability to determine Be in the ng g-' region for a sample mass of just 200 mg. 1.7.4. Bromine Allain et al. (90/4159) determined Br in plasma and urine by ICP-MS. Samples were simply diluted 10-fold with a diluent containing Eu as internal standard. The method had a linear range from the detection limit of 52 pg I-' up to 40 mg I-'. For a group of 26 healthy subjects a mean Br concentration of 4.1 f 0.9 mg I-' was found in plasma and urinary Br excretion was 3.6 f 1.7 mg per 24 h.Inductively coupled plasma MS was also used by Cho et al. (91/C537) to determine Br in tissue CRMs following solubilization with tetramethylammonium hydroxide. 1.7.5. Cadmium In assessing occupational exposure of workers to Cd measure- ments of whole blood and urine Cd are normally made. These determinations are not that straightforward as shown by con- siderable variation in quality control schemes. In an evaluation of the performance of 20 laboratories in an Italian quality- control scheme for blood Cd Morisi et al. (90/2242) conclud- ed that both advanced instrumentation and experienced person- nel seemed to be necessary. The value of using oxygen ashing was stressed by Maeda et al.(90/3371) not only in lowering background absorption from the blood matrix but also in re- ducing matrix interference. De Benzo et al. (91/250) found it was possible to use aqueous calibration if a modifier of (NH,),HPO,-Mg(NO,) was used with platform atomization. Samples and standards were diluted with Triton X-I00 solu- tion. Evaluation of results was by peak height. Two studies have shown that by careful choice of atomiza- tion conditions in ETAAS it is largely possible to separate the Cd peak from the background absorption in the determination of Cd in urine. Ong et al. (90/3426) found that with their Varian system a slow heating rate in the atomization stage produced the Cd peak before the peak of background absor- bance.In the study of Dube et al. (90/1339) using Perkin- Elmer instrumentation the peak appeared between two back- ground peaks the first of which could be greatly reduced by the inclusion of ashing stages at 300 and 400°C. Both used pyrolytic graphite tubes and matrix-matched calibration. Ong et al. diluted samples 1 + 9 with 0.01 mol dm-' HN0,4.02% v/v Triton X-100. Accuracy was demonstrated by excellent results in an external quality-control scheme. The object of the study by Dube et al. was to develop a method suitable for measurement of normal levels of Cd in the general population. Consequently a lower dilution (1 + 1) was used and the diluent (1.5% HNO,) helped reduce background absorption; remain- ing background was corrected by a Zeeman-effect system.A detection system of 0.05 pg I-' was achieved. In the determination of Cd in serum by ETAAS Bulska et al. (90/4148) found that a Pd-Mg modifier gave a significant blank value. This was reduced by first injecting the modifier heating the graphite tube to 1 100°C to remove Cd cooling and then pi- petting the sample into the tube. The modifier allowed an ashing temperature of 800 "C to be used. In the final programme air ashing prevented accumulation of carbonaceous residues and a cool step prior to atomization improved the sensitivity by 30%. As part of a study of the relationship between fertility and smoking habits Saaranen et al. (90/1367) determined Cd in seminal plasma and serum of 64 men by ETAAS. Smokers had higher serum Cd concentrations than non- smokers; differences in seminal plasma were only evident if smoking exceeded 20 cigarettes per day.No effect on fertili- ty was demonstrated. 1.7.6. Calcium Nutritional absorption of Ca by humans from milk was studied by Mellon et al. (90/C3983) using a double label stable isotope technique and measurements by FABMS. For the determination of Ca in urine Welch et al. (90/4033) compared two FAAS methods measurement in the air-C,H1 flame with lanthanum as interference suppressor and measure- ment in the NZO-C,H flame with potassium as ionization sup- pressor. Different sample dilutions and concentrations of lanthanum were compared. A 1 + 19 dilution with a high con- centration of lanthanum (20 g I-') and measurement in the air- CZHZ flame gave the most accurate results.1.7.7. Chromium Watkins and Rademeyer (90/C2880) claimed advantages for the use of a mixture of H! in Ar as a purge gas during the char- ring and atomization steps of the determination of Cr in serum90R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 by ETAAS. The build-up of carbonaceous material was pre- vented and depression of the analytical signal by involatile oxides was stated to be reduced. Differentiation of Crl'I from Cr"' in plasma was attempted by Schmitt and Kruse-Jarres (90/ 141 5) using solvent extraction with trifluoroacetylacetone (TFA). However interference was seen from the TFA in the determination of Cr by ETAAS. A stable isotope dilution GC-MS technique for the determi- nation of Cr in urine has been reported by Aggarwal et al.(90/ 2297). Samples were first digested with HN0,-H20,. A vola- tile chelate of Cr was formed by complexing with lithium bis- (trifluoroethy1)dithiocarbamate. This was separated on a fused silica capillary column and the isotope ratios measured by a general purpose mass spectrometer. Accuracy was demonst- rated by the determination of Cr in two urine RMs. Chromium and Ni concentrations in tumour tissue were found by Raithal et al. (90/1361) to be not significantly differ- ent from concentrations in tumour-free tissue. Whole blood and urine Cr and Ni concentrations had been observed to be higher in patients with tumours than in normal persons. Con- tamination of fluids and tissue with Cr and Ni containing in- struments during major surgery was thought to be the cause.Chromium (and Cd) in lung tissue were investigated by Paako er al. (90/1945) as markers of smoking. Mean lung Cr concen- trations were about 4-fold higher in smokers than non- smokers. Concentrations tended not to decline with time after giving up smoking. Measurements were made by ICP-AES. 1.7.8. Copper Although the essentiality of Cu has been known for a long time and measurements of plasma Cu are frequently made to estimate the Cu status of patients these measurements really tell us little about the adequacy of Cu nutrition in humans. The Recommended Dietary Allowance (RDA) for Cu is 2 mg d-' for adults but most dietary studies find a daily intake of nearer 1 mg. Particularly relevant therefore are recent studies on ab- sorption of Cu using the stable isotope T u and measurements by TIMS.Tumland et al. (90/2863) found that percentage Cu absorption tended to increase with decreasing Cu dietary intake. Copper balance could be achieved by the young men acting as volunteers in their study with an intake of 0.8 mg d-' of Cu suggesting that the RDA for Cu is unnecessarily high. An earlier study from the same group (90/4133) com- pared Cu absorption in young and elderly men given a diet containing 3 mg of Cu. No significant difference was seen between the two age groups. High Zn in the diet (16.5 mg) was found not to increase faecal Cu excretion (90/4132) but Cu absorption was higher (48%) than on a low intake of 5.5 mg d-1 of Zn (38%). Most studies use "Cu added extrinsic-ally as a tracer but the question arises of the validity of this ap- proach.Johnson and Lykken (90/4 13 1 ) compared absorption of Cu from whole wheat bread labelled extrinsically with ab- sorption from bread made of wheat labelled in growth. The lack of significant difference in results supported the use of ex- trinsic tracers. The poor predictive value of serum Cu measurements to in- dicate Cu status was mentioned above. Most of the Cu in serum is bound to proteins principally caeruloplasmin and changes in serum Cu generally reflect changes in the protein concentrations rather than a change in Cu status. For example serum Cu rises dramatically in pregnancy. Noubah and Al- Awqati (91/284) used ultrafiltration to assess how the non- protein bound fraction of Cu in serum changed in pregnancy.This fraction was separated using the Amicon micropartition system. Both total Cu and ultrafilterable Cu were measured by ETAAS. The non-protein bound fraction could be measured on 1 ml of serum in less than 2 h with a within-batch and between-batch precision of 5.7 and 9.3% RSD respectively. Little difference was seen in the concentration of this fraction between normal females pregnant females and foetal blood samples (mean values 0.253,0.262 and 0.247 pmol I F respec- tively). In normal females this formed 1.4% of the total Cu in serum. A highly significant correlation was found between matched cord (foetal) and maternal blood samples. Further ap- plications of this technique are awaited with interest. The reverse separation was used by Lyon and Fell (90/ ,3497) to solve a problem in the determination of the h3C~:hsCu ratio in serum.Polyatomic species interfered at mass 63. By SEC with disposable columns containing G25 Sephadex it was possible to separate the high relative molecular mass pro- teins containing caeruloplasmin (and hence most of the Cu) from other inorganic ions such as Na' and Pod3- which gave Tise to the polyatomic interferences. The measured isotope ratio in the protein fraction was 2.082 in good agreement with the ratio found for a simple aqueous solution of Cu (2.1 10). Rocks (90/C4041) described a method for the determination of Cu and Zn in paediatric plasma samples by pulse nebulization FAAS. Small volumes of undiluted plasma were injected into a small inverted funnel attached to the nebulizer tube and the peak height (or area) measured for the transient peak.1.7.9. Fluorine Gomez et al. (90/3508) evaluated various methods for prepar- ing samples for the determination of F in an electrothermal at- omizer by measuring the absorption at 227.5 nm by the AIF species. A solution of 0.01 mol dm-3 A1 and 0.01 mol dm-3 Sr (as modifier) was pipetted on to the graphite tube and dried before the addition of the sample. Urine F could be determined directly after dilution; addition of NH,NO was necessary to help reduce background absorption. For solid samples oxygen flask combustion was the most satisfactory mineralization pro- cedure. The technique offers greater sensitivity for the deter- imination of F than determination by an ion-selective electrode. 1.7.10.Gold Elder and collaborators (90/C 140 90/2070) used ICP-MS as a sensitive detector for the separation of Au drugs and rheir metabolites by HPLC. With a single injection by FI informa- tion on the Au Cu and Zn distribution in plasma from arthritis patients treated with the drug auranofin was obtained by sep- aration on size-exclusion or anion-exchange columns. The former showed the protein-binding of the elements whereas the latter gave information on the low relative molecular mass components separating the different metabolites. I .7. I 1. / d i n e Iodine concentrations in plasma and urine were determined by ,4llain et al. (90/4159) using ICP-MS. Samples were simply diluted 10-fold with a diluent containing Eu. The method had a detection limit of I .6 pg I-' and a linear range up to 400 pg I-'.For 24 healthy individuals the mean plasma I concentration measured was 58 k 12 pg I-' and the mean urinary I excretion !)4 k 97 pg per 24 h. Iodine in tissue CRMs was determined by Cho et al. (9 I /C537) using ICP-MS following solubilization with tetramethylammonium hydroxide. Fassett and Murphy (90/2300) from NIST used isotope dilution RIMS to establish ii value for I in NIST SRM Oyster Tissue of 4.55 pg g-I. Further data for SRMs were obtained by isotope dilution TIMS using LaB for ionization enhancement (90/2353). :I .7.12. /ran Iron in sei-um is generally more conveniently measured in a clinical laboratory by well established colorimetric methods. Atomic spectrometric techniques have few advantages but are regularly explored.Uchida et a!. (90/ 1395) deproteinized serum for measurement of Fe by ICP-AES; results were com- parable to those obtained by a colorimetric technique. AnJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 199 I VOL. 6 91R ETAAS method was developed by Navarro and Romero (90/ C1797 90/C2109). Plasma was diluted 100-fold and whole blood 10 000-fold with a 0.05% Triton X-100 solution prior to determination of Fe. Measurement of liver Fe concentration is the best way of as- sessing the extent of iron-storage diseases. Van Deursen et al. (90/1327) compared histochemical assessment by staining and determination of Fe in liver homogenate by ETAAS. They concluded that chemical determination of the liver Fe content was preferable. Separation by HPLC on a gel-permeation column followed by determination of Fe by ICP-AES allowed La Torre et al.(90/2244) to study ferritin-containing drugs. Comparison of peak profiles with a UV detector and with the ICP-AES detec- tor allowed identification of the eluted proteins. The availability of stable Fe isotopes enables studies of Fe metabolism to be made. Amarisiriwardena and Barnes (90/ C3163) reported their analytical study of measurement of the s6Fe:s7Fe ratio in rat blood by ICP-MS. Samples were first di- gested under high pressure. The precision of measurement of the ratio in the final method was better than 1% RSD. The ab- sorption of Fe from diet was studied in six healthy young men by Turnlund et al. (90/4133) using the stable isotope s4Fe. Measurements were by TIMS. Comparison with similar data from a group of elderly men showed no significant difference in absorption.However Fe absorption differed significantly between individuals whereas Cu absorption did not. 1.7.13. Lead Another critical review by Subramanian on the determination of Ph in blood by ETAAS has been published (90/3549). Pro- cedures to overcome interferences were outlined and the ad- vantages of the STPF concept described. This approach together with O2 ashing was used by Granadilio and Romero and described in two conference presentations. In one (90/ C2100) the modifier was NH,H2P0 + Mg(NO,)? and in the other (90/C30 19) Pd with citric acid. Blood samples were diluted 1 + 9 with 0.1% v/v Triton X-100 solution and the modifier was added separately to the furnace. Oxygen ashing was also examined by Maeda et al.(90/3371) who observed that it resulted in a reduction in background absorbance and matrix interference. Osterloh et al. (90/3245) found that ETAAS gave superior within- and between-batch precision to commercial or modified ASV procedures for the determination of low blood Pb concentrations. Determination of Pb in blood by ICP-AES is unusual. Alvarado et al. (91/65) used ETV to introduce Pb from diluted blood into the ICP. To achieve good results signal integration and careful selection of the integra- tion period were important. No appreciable matrix effect was found. A study from Saudi Arabia (91/308) showed that the mean blood Pb concentration in 200 children was 69 + 34 pg I-' as measured by ETAAS. In a sub-group of children with higher blood Pb concentrations haematological measurements indi- cated an early stage of microcytic anaemia.Further studies of the use of isotopic composition of Pb in establishing sources of exposure were made by Viczian et al. (91/2). Measurements were made by ICP-MS of the four Pb isotopes in the blood of children poisoned with Pb and in their possible environmental sources dust soil and paint. Use of all three isotope ratios was found to increase confidence in assigning sources of exposure but in the cases studied no unique source of exposure emerged and discrimination between the various sources was difficult in practice. Direct determination of Pb in urine by ETAAS is difficult. Parsons et al. (90/C2958) found multiple absorption peaks from Pb in urine suggesting different species.This was reduced by addition of a mixture of Pd and Mg nitrates; the mixture was shown to be better than either salt alone and better than NH,H,PO,. However Mak et al. (90/1922) de- scribed a direct method with the latter as a modifier. Both wall and platform atomization gave satisfactory recoveries but plat- form atomization was preferred as it gave a wider linear range better precision and longer tube life. Speciation of trialkyllead species in biological samples has been achieved by coupling ion-pairing HPLC to ICP-MS in a study by Al-Rashdan et al. (90/C3235). The suitability of hair Pb levels to reflect environmental pollution was demonstrated in a Brazilian study (90/1838). For 263 children living near a Pb smelter hair Pb levels measured by AAS showed a range of 20-4933 pg g-I with a geometric mean of 313 pg g-I.To determine Pb in biological RMs Lo et al. (90/1499) de- veloped a two-step extraction to separate Pb from the matrix. Lead extracted from the digest with APCD-NaDDC into CHCI was back-extracted into an Hg1I solution for determina- tion by ETAAS. 1.7.14. Lithium The determination of natural levels of Li in serum and urine has become a popular subject for study. Bencze et al. (90/ 1878) used a KH2P0 modifier which increased the sensitivity of ETAAS markedly giving a detection limit of 0.5 pg I-'. Ammonium nitrate was the modifier preferred by DUK et al. (90/3534) giving a detection limit of 0.2 pg I-'. Background correction was found to be important even though the wave- length of measurement was high (670.8 nm).A tungsten lamp background correction system was found to give superior sen- sitivity and greater linearity of calibration than a Zeeman- effect system. Mean normal serum Li was found to be 0.16 pmol I-' (1.1 pg I - I ) and the mean 24 h excretion in urine 5.24 pmol (37 pg). Sampson (91/C813) managed to avoid the need for background correction by coating the graphite tubes with tantalum. This allowed the use of a higher charring tem- perature (1200 "C) removing the substances giving rise to the background. Serum samples were deproteinized with 10% v/v HNO and urine samples diluted with 5% m/v NH,NO,. Zhing and Zhang (91/407) studied the interference effect of chloride in the determination of Li. The interference was overcome by NH,NO or (NH,),PO when a tungsten coated graphite tube was used.This was applied to the determination of Li in blood. Determination by ICP-MS was explored by Abou- Shakra and Ward (91/C799). Simple dilution of samples was satisfactory for the determination of Li alone but for multi- element analysis wet digestion was necessary. Serum Li was measured on 150 samples treated in this way giving a range For the determination of the higher Li concentrations in the serum of patients on Li therapy Quinonero et al. (90/1876) compared FAES and FAAS. In general methods using FAES gave lower detection limits but results were generally less ac- curate than FAAS. of 0.3-2.4 pg I-'. 1.7.15. Magnesium Reference values for Mg in Ieucmytes and ei-ythroc-ytes were obtained by Geven et al.(91/825) for children from birth through to adolescence. Discontinuous density gradient cen- trifugation on Ficoll-Hypaque was used to separate the leucoc- ytes from blood and the Mg content of the lysed leucocytes and erythrocytes was measured by FAAS. Results for leucocy- tes were expressed as fmol per cell pmol per g of protein and mmol per g of DNA. No significant correlation was found between Mg in serum and leucocytes nor between Mg in serum and erythrocytes. Leucocyte Mg was largely indepen- dent of age and within the adult range of values. Erythrocyte Mg however was significantly lower in cord blood and during the first month of life compared with later in infancy. A new type of plastic tube developed for a single-step separation of92R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 lymphocytes on a density gradient for Mg determination was described by Weissberg et al. (91/327). To harvest leucocytes from blood for measuring leucocyte Mg the polymer Ficoll is generally added. Frandsen and McNair (90/1484) wished to avoid this addition as it interfered with protein measurement and it precluded the use of the plasma for other measurements. Their method involved three successive centrifugations which succeeded in isolating 8-28% of the leucocytes and 04.025% of the erythrocytes. Magnesium in the cells was measured by FAAS. In blood samples from 98 healthy people a mean (kSD) concentration of Mg of 26.4k3.9 pmol per g of protein in leucocytes was measured. The best way of expressing leucocyte Mg remains controver- sial. In response to a letter by Urdal and Landmark (Clin.Chem. 1989 35 1559) reporting better precision when results were expressed as pmol per g of protein rather than fmol per cell Elin et al. (90/4034) criticized the poor precision with which the workers had measured cell count. [It should be noted that the same reduction in variation is seen in the results of Geven et al. (91/825) for results expressed as pmol per g of protein and in the work of Weissberg et al. (91/327) on lym- phocytes.] Urdal and Landmark had concluded that analytical imprecision reduced the usefulness of leucocyte Mg assay par- ticularly when results were used to detect Mg deficiency in single patients. Elin et al. refuted this statement arguing that the clinical value and application of this Mg test for patient care had still to be defined.However do not hold your breath! This saga has been running longer than the average soap opera. I .7.16. Manganese The Mn content of the cellular components of blood was studied by Milne et al. (91/324). Platelets and blood cells were separated on a discontinuous Percoll gradient and the Mn measured by ETAAS after digestion with HN0,-H,O,. Most of the Mn (about 66%) was in the erythrocytes with about 30% associated with the buffy coat (platelets and leucocytes). It was suggested that the latter might better indicate the Mn status of the patient as they turn over more rapidly than eryth- rocytes. For determination of Mn in serum by ETAAS Hulanicki et al. (90/1203) compared various sample pre-treatment proce- dures.The simplest dilution with 0.1 mol dm- HNO was satisfactory but high recoveries of Mn were seen. This was cured by adding Ca and H,PO to the standards. Stobbaerts and Deelstra (91/474) recommended the use of L-ascorbic acid as a modifier to overcome interferences from KCI and NaCl in biological matrices. Slurry sampling for the determination of Mn in biological samples was described by Jordan et al. (90/3294); more details of this application appear in section 1.2. 1.7.17. Mercury Cold vapour atomic absorption spectrometry (CVAAS) is the most popular technique for the determination of Hg. More workers have made use of collection of Hg on Au (90/1179 90/1944) or Ag (90/34 10) with subsequent release by heating to obtain higher sensitivity.Many of the remaining problems of Hg determination in biological samples relate to the diges- tion procedure in achieving adequate mineralization without incurring losses of Hg. Vermeir et al. (90/1179) developed a pressure digestion technique using PTFE-perfluoroalkoxy (PFA) vessels with microwave heating. About 100 mg of sample were heated with 1 ml of HNO in a domestic micro- wave oven with a three-step programme lasting 20 min. Subse- quent determination was by CVAAS with collection on a gold- coated sand adsorber. The vessels used for digestion did not contribute significantly to the blank value and did not lead to detectable losses of Hg. The procedure was applied to NIST biological SRMs with satisfactory results. De Vargas and Romero (90/3293) compared two digestion procedures a cold mineralization with HN0,-H,S04-HC10,-KMn0 at - 10 "C for a week and a high-pressure digestion with HNO at 130 "C for 2.5 h.Both gave no observable losses of Hg and satisfacto- ry results on NIST SRM Albacore Tuna. For a study of Hg in a large number of shrimp and fish samples from Lake Mara- caibo Venezuela the cold mineralization procedure was preferred as it allowed a greater throughput of samples with less of the analyst's time. The dissolution of metallic Hg in artificial saliva and in so- lutions of its constituents was studied by Takahashi er al. (90/ 1330) using ETAAS. The rate of dissolution in artificial saliva was seven times higher than in saline solution principally because of the presence of glutathione. The group at Nagoya University Japan have continued their theme of high-temperature separations in AAS with a paper on the determination of Hg in rat tissues (90/3337).The separation was carried out on a column atomizer containing activated charcoal heated to 1100 "C. Mercury accumulation in the body has frequently been determined through hair analysis normally head hair. Ko- bayashi et al. (90/1944) however studied moustache hair with a sensitive method based on atomization of Hg from the hair at 800 OC collection on porous Au vaporization and measurement by AAS with a cell of 20 cm path length. They detected a rise of about 1 pg g-' Hg in moustache hair 7 d after ingestion of 100 g of tuna fish. Bushee et al. (90/207 1) developed a procedure to determine thimerosal (sodium ethyl mercurithiosalicylate) an antimicro- bial agent used as a preservative in certain biological products such as vaccines.Separation by LC followed by detection by ICP-MS enabled thimerosal and its possible degradation prod- ucts to be determined. Direct measurement by ICP-MS with FI sample introduction gave the total Hg concentration. No evi- dence for significant degradation was found even on outdated samples but losses of thimerosal were seen in some instances because of absorption by certain types of rubber stopper. 1.7.18. Molybdenum Although Mo is known to be an essential element there is still very little known about dietary requirements and its metabo- lism. Lack of convenient analytical methods has slowed up progress with this element.This review year it is clear that a number of workers have been directing their efforts at solving some of these problems. The sensitivity of ETAAS is barely adequate for the determination of Mo in biological fluids but in addition there are serious matrix interferences. Bermejo- Barrera et al. (91/251) overcame these interferences in the de- termination of Mo in infant formula and in human milk with the addition of BaF as modifier. The detection limit was 0.9 pg I-'; levels in human milk ranged from 2.3 to 8.4 pg I-'. An alternative approach is to solvent extract the Mo which offers the additional advantage of increasing the sensitivity. Morrice et al. (90/2081) extracted Mo from digested plasma with 8- hydroxyquinoline into CHCI,. After evaporating the extract to dryness the residue was re-dissolved in a smaller volume of buffer to give a 2-fold increase in sensitivity.Accuracy was demonstrated by satisfactory recovery of added Mo and a result for NIST SRM Bovine Serum which agreed with the certified value. The method was applied to plasma from sheep and cattle but as these workers noted extra sensitivity would be required for human plasma. Pre-concentration of Mo on to activated carbon after chelation with the ammonium salt of the dithiophosphoric acid 0,O-diethyl ester was described by Monte and Curtius (90/3479). After a two stage acid treatment to recover Mo from the carbon it was measured by ETAAS with Mg(NO,) as modifier. Enrichment factors of 30 were re- alized although the final ETAAS determination was still subject to interferences from sulphate and phosphate from theJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 199 1 VOL.6 93R decomposition of the complexing agent. The method was applied to the determination of Mo in NIST SRMs Oyster Tissue and Bovine Liver with satisfactory results. Pinna et al. (90/1283) described methods for the determination of Mo in diet urine and faeces by ETAAS with minimal sample pre- treatment. Diet and faeces were homogenized dry ashed and the residue dissolved in dilute HNO,. Acidified urine samples were measured directly. Using peak area evaluation recover- ies were 92.3 106.3 and 100.2% for urine faecal and diet digests respectively. Background correction was found to be unnecessary but extended firing of the tube was required after atomization to reduce memory effects.1.7.19. Nickel Well worth detailed study is an excellent paper by Nixon et al. (90/2082) on the determination of Ni in serum by ETAAS. They reinforced the need to use extra vigilance in eliminating all possible contamination and to use techniques to increase the sensitivity so that very low concentrations can be meas- ured. A clean room was used for sample preparation and ana- lysis and all materials coming into contact with the sample (including the syringe) were acid washed. Samples were diluted 3 + 1 with I % m/m Triton X-100-0.001 mol dm-2 HNO and 50 p1 were injected on to the surface of a pyrolytic graphite coated graphite tube. After drying and ashing a further 50 pl of prepared sample were injected and the full furnace programme initiated.Background after ashing at I600 "C was acceptably low (ca. 0.1 A) and corrected with a Zeeman-effect system. The large amount of sample on the tube enabled a very low detection limit (0.06 pg I-') to be reached. Calibration was by matrix-matched standards and ac- curacy was demonstrated by analysis of serum RMs including the Second Generation Serum RM of Versieck. Normal con- centrations in the serum of 38 volunteers was found to be 0.14 f 0.09 pg I-' which is about 4-fold lower than has been found in studies by other workers. Patients on haemodialysis showed higher serum Ni concentrations (mean f SD 6.38 k 3.36 pg I-' n = 27 patients). Three methods published this review year for the direct de- termination of Ni in urine by ETAAS have many similarities.All used pyrolytic graphite coated graphite tubes; Paschal and Bailey (90/3552) commented that the use of a platform gave lower sensitivity. Diluents were mixtures of HNO and Triton X-100 but the dilution ratio and the concentrations varied 1 + 1 with 2.0% v/v HNO 0.001% v/v Triton X-100 (90/ 3552); 1 + 2 with 0.6% v/v HNO (90/3290); and 1 + 1 with 0.02% v/v Triton X-100 0.01 mol dm-' HNO (90/3426). Ong et al. (90/3426) achieved a better detection limit of 0.01 pg 1-' by the use of triplicate injection of 20 pl aliquots on to the tube drying and ashing after each injection. White and Boran (90/3290) who used the higher dilution quoted a de- tection limit of 2.1 pg I-' whereas Paschal and Bailey (90/ 3552) with a 1 + 1 dilution reported 1.1 pg I-'.All methods were designed for assessing occupational exposure; the lower detection limit only becomes of value in assessing concentra- tions in normal unexposed individuals which Ong et al. (90/ 3426) reported as 0.9-8.2 pg I-' in 20 persons with a mean of 3.1 pg I-'. Aqueous calibration was used by Paschal and Bailey (90/3552) however the other two used matrix- matched calibration but unfortunately provided no evidence that aqueous calibration would not work. The demonstration by these three papers all from laboratories very much in- volved in routine occupational monitoring that direct ETAAS methods for this determination do really work will hopefully put an end to the earlier solvent extraction procedures advo- cated for this determination. Determination of Ni in serum and urine by ICP-MS suffers from isobaric interference from oxides of Ca at all Ni isotope masses.Vaughan and Templeton (9 1/C666) explored the possi- bility of using mathematical correction with principal compo- nent analysis and target vector rotation to isolate the signal due to Ni in the presence of high concentrations of Ca. Correction of interference from 300 mg I-' of Ca was shown to be possible. Benson et al. (90/1360) determined Ni in tissues by ETAAS after digestion with HN0,-HCI0,-H,O using microwave heating. Recovery of added Ni was acceptable for lung liver and lymph node but kidney showed lower recovery (85%) when NiSO was added. As a result of an observation that patients with turnours had increased Cr and Ni concentrations in serum and urine Raithel et al.(90/1361) wished to discover whether there was an accumulation of these elements in tumour tissue. Their measurements by AAS however showed that there was no significant difference in the concentration of those elements in tumour and tumour-free tissue. They concluded that the high concentrations of Cr and Ni seen were due to contamination of tissues and body fluids during major surgical procedures. 1.7.20. Phosphorus A molybdenum tube atomizer was used by Ohta et al. (90/ 2195) for the determination of P in biological samples by ETAAS at the 213.6 nm non-resonance line. A study of inter- ferences showed that by including a pyrolysis step at 1030 K interferences were all removed except for Al at higher concen- trations. Results obtained by analysing NIST biological SRMs (including Bovine Liver) after pressure digestion with HN0,- H,O were in good agreement with certified values.1.7.2 1 . Platinum As Caroli et al. (90/C1810 90/2245) pointed out it is the 'free' Pt-complex (i.e. bound to species of low relative mole- cular mass) that exerts a cytotoxic effect when Pt-based drugs are used. Therefore they used a combination of HPLC and ICP-AES to measure the 'free'component in the plasma of pa- tients given the drug carboplatin by infusion. The group at Velindre Hospital Cardiff UK have continued their studies of the kinetics of cisplatin. On patients with malignant melanoma (90/C4037) they followed the pharmacokinetics of Pt by ETAAS measurements when the limb was isolated surgically and perfused with cisplatin.Relatively low concentrations were measured in the main systemic circulation and as a con- sequence the normal side-effects were absent. In their other study (90/C4038) the kinetics of Pt elimination were followed on patients put on multiple infusions of cisplatin every 3 weeks. At the end of the 6 h infusion the mean (k SD) plasma Pt concentration was 2.6 f 0.4 mg I-' which then declined tri- phasically until after the end of the first 3 week period it was 0.32 k 0.11 mg 1-'. They concluded that Pt would accumulate in patients when infusions were given at 3 week intervals. In order to study the same accumulation of cisplatin in rats Li et al. (90/1394) measured Pt in tissues. Samples were digested with HN0,-H,O at 85 "C for 1 h and Pt was determined in the digests by ETAAS.Delves (91/C739) used ICP-MS to measure Pt in the blood of patients treated with a low energy catheter ablation technique for correcting heart defects. Platinum micro- electrodes are used in this technique and they could corrode in use. The method developed measured Pt down to 0.3 pg I-' in blood. 1.7.22. Potassium In assessing the accuracy of determination of K in serum by FAES with Li as an internal standard Umemoto ef al. (90/ 1506) found that with simple K standard solutions for calibra- tion results obtained on CRMs were about 5% higher than certified values. By using one CRM to calibrate the instru- ment deviation was reduced to 0.4%. The final recommenda- tion that the CRM be used to calibrate the spectrometer is not one that these reviewers would support.Reference materials94R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 are for proof of accuracy in method development and for quality control not for Cali brat ion. Reference values for K in mononuclear cells and erythrocy- tes in children were established by Geven et al. (9 l/825). After separation of the cells or erythrocytes they were lysed and the K concentration of the lysate measured by FAES. Concentra- tions of K in erythrocytes and mononuclear cells showed no change with age (0-17 years). Erythrocyte K showed a small but significant correlation with erythrocyte Mg concentration. A single-step procedure for the isolation of lymphocytes was described by Weissberg et al. (91/327). A new type of plastic tube was used with a fixed polystyrene disc below which Ficoll-Histopaque solution could be inserted.After centrifuga- tion the cells were collected from the boundary between plasma and Ficoll-Histopaque. After washing the cells were dissolved in HNO and K determined by FAAS. When results were expressed as fmol per cell pregnant women had significantly higher values than controls; this difference however was no longer apparent when results were calculated as fmol per mg of protein. 1.7.23. Selenium For Se there has been a steady development in modifiers for ETAAS automation of hydride generation techniques and application of the techniques to establish reference ranges for plasma Se and clinical conditions in which Se depletion could be demonstrated. For ETAAS Sampson (91/C814) reported the use of a mixed Pt-Pd modifier which had the high sensitivity of the Pd modifier with the beneficial effect of the Pt in removing the over-correction effects of phosphate seen with a deuterium-arc background correction system.Determination of Se in a number of matrices (including urine) was claimed to be pos- sible. For their method for Se in whole blood and serum McMaster et al. (90/4030) chose a Cu-Mg(NOJ2 modifier Zeeman-effect background correction and stabilized tempera- ture platform furnace (STPF) conditions. Plasma samples were diluted 1 + 3 and whole blood 1 + 12 with 0.2% v/v HN0,- 0.2% v/v Triton X-100. Calibration was by standard additions. A thorough evaluation of the problems of determining Se in tissues by ETAAS has been carried out by Martin and Wil- liams (90/C1751,90/2057).Using NIST SRM Bovine Liver as a model they investigated simple sample dissolution proce- dures and modifiers to allow determination by direct calibra- tion using STPF conditions and Zeeman-effect background correction. Samples were weighed into calibrated glass flasks and allowed to pre-digest with HNO ( 1 ml per 0.1 g) over- night. The flasks were then either heated on a hot-plate for 45 min at temperatures up to 150 "C or heated in a microwave oven in a 12 min cycle of alternate heating and standing. Both techniques were effective. Their objective in choice of a modifier was to establish conditions in which the appearance time and peak shape were identical for aqueous standards and sample digests. With a modifier of Pd or Pd + Mg(NO,) and the use of H2-Ar as a purge gas the peaks did not coincide; they did with Cu + Mg(NO,) as a modifier but peak shape was different and analytical recovery was low (75-85%). A mixture of Pd (600 mg I-I) and Cu (500 mg I-' ) with H2-Ar as a purge gas was found to be the most effective but for com- plete recovery Triton X-100 also needed to be present.Absence of matrix interferences was confirmed by comparing results by standard additions and direct calibration. An average recovery of 98.5 f 7.2% (n=38) was obtained for a range of sample masses from 0.005 to 0.650 g. Hydr-ide generation techniques for determining Se are now becoming more automated. Flow injection with continuous hydride generation was described by McLaughlin et al.(90/ 2194) for the determination of Se in serum after wet digestion. With an injection volume of 330 pl the detection limit was 1.2 pg I-'. Comparison with the ETAAS method of McMaster et al. (90/4030) showed good agreement in results. Application of the Perkin-Elmer FI system (FIAS-200) to the determina- tion of Se in biological fluids by HGAAS was described by Guo et al. (91/48). Much of the time taken in the analysis of biological fluids by hydride generation techniques is taken up by digestion. Galgan and Frank (90/1412) automated the whole process. Automated wet digestion with HN0,-HCIO was followed by continuous HGAAS with FI. The latter part had a capacity of producing 120 measurements in 1 h. Contin- uous flow hydride generation was applied by Tracy and Moller (91/311) to determine Se in biological materials by ICP-AES.No gas-liquid separator was necessary as this function was carried out by the standard nebulizer. The detection limit was 0.4 pg I-' corresponding to 4 pg kg-I in a 1 g sample. All the preceding systems used commercially available AAS or ICP spectrometers for quantification. The apparatus for Se determi- nation can be considerably simplified by using non-dispersive AFS as shown by D'Ulivo et al. (90/4152). As an atomizer they used an electrically heated vertical quartz tube. Light from an EDL was focused above the tube and the fluorescent intensity was measured by a solar-blind PMT without a mono- chromator. It was found important with the batch hydride gen- eration system used that the H2 evolved in the process should be spontaneously ignited at the mouth of the tube.Without the presence of a flame the sensitivity was 80-200-fold lower. The method was applied to the marine RMs DORM-I and DOLT-1. The detection limit was 27 pg and the calibration was rectilinear up to 300 ng of Se. Many of the papers this review year are less concerned with the methodology of determining blood or plasma Se but more with understanding what the measurements mean. Morisi et al. (90/2238) reported reference ranges for serum Se related to age and sex for 4201 Italian adults and 1217 children. From Northern Ireland McMaster et al. (90/4030) reported ranges for blood and plasma Se in 100 adult subjects. As part of a study of whole blood Se concentrations in various countries Lombeck et al.(90/2913) developed a blood-spotting tech- nique for collecting samples. Blood was spotted on a filter- paper and dried. Circles of 28 mm diameter (corresponding to 17.0 p1 of sample) were cut from the paper for analysis by wet digestion and HGAAS. The application of this screening test in Greece and China showed significantly lower blood Se con- centrations in Greek mothers after birth than in Chinese mothers. Greek infants showed a dramatic decrease in blood Se concentrations within the first 4 months of life. Sampson and Gilbertson (90/C4039) found a similar decrease in British infants from a concentration of 0.40 f 0.12 pmol I-' at birth to 0.16 k 0.07 pmol I-' at 3 weeks. Infants on total parenteral nu- trition (TPN) showed a decrease to concentrations less than 0.05 pmol I-' (i.e.the detection limit of the determination by ETAAS) as a result of the absence of Se in the TPN regimen. Low birth weight infants seem particularly at risk from Se deficiency according to a study by Lockitch et al. (90/1383). Plasma Se and glutathione peroxidase concentrations taken within 3 d of birth were lower in low birth weight infants than in full-term infants. Many of these end up being fed by TPN and in these plasma Se concentrations again were below the detection limit of the ETAAS method. A decrease in the Se concentration in amniotic fluid with the progress of pregnancy was found by Karunanithy et al. (90/1931). In their study of amniotic fluid samples from 1 1 1 healthy pregnant women by using HGAAS a highly significant negative correlation was found between gestational age and Se concentration.Marathon runners may be interested to know that running a marathon does not lead to any significant changes in plasma Se (901 1551). 1.7.24. Silicon Silicones on hair- from cosmetics were measured by Gooch and Kohl (90/1381). The method they developed involved en-JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 95R zymatic digestion of the hair extraction of the organosilanes released with an organic solvent and determination of the Si in the extract by AAS. 1.7.25. Strontium Strontium is rarely measured in plasma and urine. Leeuwen- kamp et al. (90/1483) however wished to carry out Sr absorp- tion tests which had been proposed as a non-radioactive alternative to the normal technique of assessing Ca absorption with a radioactive Ca isotope.Determination of Sr was made by ETAAS on a pyrolytic graphite tube after dilution of plasma 20-fold and urine 50-fold with 0.2% v/v HNO,. Matrix effects were absent at this dilution. Endogenous concentrations of Sr in plasma and urine were found to be 16 f 8 pg I-' (mean f SD) and 158 k 26 pg I-' respectively. As part of a programme to assess the uptake of radioactive elements by the human body Dalgano et al. (91/C538) studied the uptake of the stable isotope 86Sr in volunteers. By using ICP-MS methods were developed for the measurement of Sr isotope ratios in blood urine and saliva samples. 1.7.26. Tellurium For monitoring of occupational exposure to Te Kobayashi and Imaizumi (90/3950) developed a sensitive method for deter- mining Te in urine by solvent extraction and ETAAS.Samples were concentrated and digested with HN0,-HCIO,. The Te was extracted into IBMK and the organic phase analysed. Concentrations found ranged from <O. 1 to 1.2 pg I-'. 1.7.27. Tin For the determination of Sn in tissues Yokoi et al. (91/C524) used continuous hydride generation with ICP-AES. The detec- tion limit was 0.03 pg I-' and calibration was linear up to 10 pg I-'. Samples were digested in HN0,-HCIO and analysed using standard additions calibration to compensate for matrix interference. 1.7.28. Uranides Inductively coupled plasma MS is particularly good at deter- mining the elements of high atomic number. Isobaric interfer- ences are absent but matrix interferences occur which can be overcome by internal standardization.Igarashi et al. (90/1978) particularly addressed themselves to the problem of finding the most suitable internal standard for the determination of Th and U in biological materials. The better internal standards were those similar in mass; 209Bi was the best with 205Tl in second place. A number of biological RMs were digested with HN0,- H,Oz. A further treatment step with HF and HNO removed Si and organic matter respectively. The results obtained for Th and U in the digested samples were in good agreement with certified or literature values. Popplewell et al. (90/2808) determined Pu isotope ratios and concentrations in the tissues of people who lived near a nuclear fuels plant. The results strongly suggested that the Pu in the tissues had come from aerial discharges at the plant.1.7.29. Zinc The availability of mass spectrometric techniques to measure stable Zn isotopes has made possible studies on the absorption of Zn. Determination of Zn stable isotopes by FABMS was re- viewed by Miller et al. (90/2820) outlining the history of the technique its interferences and future development. Eagles et al. (90/282 1 ) compared FABMS with TIMS for the determina- tion of the apparent Zn absorption from cereals. Human sub- jects were fed enriched 67Zn with a bran-based cereal and faecal samples collected over subsequent days until all the un- absorbed isotope had been collected. Sample preparation in- cluded an anion-exchange separation. The mean apparent Zn absorption was 20.I % when measured by FABMS and 18.6% by TIMS. A further study (90/C3983) compared these two techniques with ICP-MS. A micro-method for the determination of Zn in plasma by pulse-nebulization FAAS was developed by Rocks (90/ C4041). Small volumes of undiluted plasma were injected into a small conical funnel attached to the capillary tube feeding the nebulizer and the peak height or area of the resulting tran- sient signal measured. Methods for the determination of Zn in platelets lympho- cytes and granulocytes were described by Wang et al. (90/ 1326). Separation was carried out on a discontinuous gradient of Ficoll-Hypaque. Cells were digested with HNO and Zn measured by ETAAS using standard additions calibration. For normal subjects they obtained mean values of 50.9 pg per 1O1O cells for lymphocytes 45.9 pg per IO'O cells for granulocytes and 3.3 pg per 1O1O cells for platelets. Two studies focus on urinary Zn excretion of females.Wat- anabe er al. (91/268) studied female Japanese farmers finding no effect of age or smoking but variations amongst the eight regions of Japan studied. A significant positive correlation was found with urinary Cd concentrations. Measurements were by ICP-AES. Iyengar (90/1905) measured by FAAS daily urine Zn excretion of three females over a year. No seasonal influence was found but it was noted that a vegetarian diet led to about half the Zn excretion found with a meat-based diet. 1.8. Conclusions In last year's Update (90/3500) comment was made that ICP- MS originally hailed as a multi-element technique was being more frequently used for the determination of one or two ele- ments for speciation studies and for measuring isotopic com- position for nutritional studies and for identifying sources of environmental exposure.This trend has continued. Multi- element studies are still being made (e.g. 90/2193 91/3) but they show many of the practical difficulties of using ICP-MS as a multi-element technique. Compared with ICP-AES very successful as a multi-element technique ICP-MS is much more sensitive but is subject to more interference. A thorough evaluation of accuracy for each element is required. Multi- element analysis of tissues by ICP-AES has opened the possi- bility of obtaining an extensive bank of data (90/1979 90/ 2237). The value of ICP-MS can be considered as its ability to measure those elements that are below the sensitivity of ICP- AES.Experience with single-element techniques already has taught us that at low concentrations each element may have its own particular problems with stability contamination and interferences. Thus more reliable progress can be made for some of these difficult elements by a more thorough investiga- tion of just one or two elements. While it is not intended to denigrate the development of ICP-MS as a multi-element tech- nique it is obvious that many workers in this field have already considered the advantages of sensitivity ability to measure isotopes separately and ease of coupling to HPLC as more significant and have exploited these possibilities. Analytical measurement techniques have continued to develop but one must not lose sight of the fact that in clinical and biological analysis although a result may be accurate it may not be any use unless it tells us something significant about a human or animal.Many measurements in general use have to be interpreted with a considerable degree of caution. Serum Cu concentrations as discussed in section 1.7.8 are a good example. Serum Cu concentrations are high in pregnancy and low in patients with Wilson's disease (paradoxic-ally a condition of storage of excess of Cu). The measurements reflect changes in caeruloplasmin a Cu-containing protein which is high in pregnancy and very low in Wilson's disease.96R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991. VOL. 6 Developments to measure the fraction of Cu not bound to cae- ruloplasmin are therefore welcome (9 1/284).Speciation has also been useful in identifying toxic and non-toxic forms of As in urine (section 1.7.2) and in measuring the free Pt in serum that apparently exerts a cytotoxic effect when patients are given Pt-based drugs. The search for alternative more reliable means of assessing the nutritional status of patients and the extent of exposure to toxic elements is important and should continue. It has however generated bandwagons which seem to get nowhere and are difficult to stop. Hair analysis for nutri- tional assessment is one example and was discussed in section 1.6. The other more recent example which seems to be heading the same way is the measurement of elements such as Mg and Zn in blood components particularly leucocytes.2. ANALYSIS OF FOODS AND BEVERAGES Simon Branch and Helen M. Crews A summary of the review is given in Table 2. In comparison with last year's review (90/3500) there appear to be more papers dealing with the analysis of all types of milk but fewer reporting the use of ICP-MS. Several papers deal with charac- terization studies usually for legislative purposes using a combination of multi-element measurements with statistical analyses. 2.1. Sample Preparation Two papers reported the analysis of samples after minimal prior treatment. Vegetable oils (0.25-0.50 g) were diluted to 100 g with THF (tetrahydrofuran 99.75-99.50 g) and trace amounts of Fe determined by ETAAS (90/1400). Bermejo- Barrera et al.(90/1478) measured Cs in mineral and thermal waters by ETAAS having mixed the sample solution with 0.2% HNO,. The detection limit was 1.85 pg I-' and the pres- ence of potential interferents at levels higher than that of the analyte was tolerated. The same workers measured Rb in waters (90/1507) by FAAS. Samples were treated with 0.1% m/v KCl solution to suppress the effect of Rb' ionization. No effect was observed from ten potential interferents present at levels greater than those normally found in waters. Other workers reported the use of extraction and cmcentra- tion procedures to facilitate the determination of various ele- ments. In order to measure Cd Co Cu Pb and Zn in beers by FAAS Ybanez et al. (90/1365) combined ashing with liquid- liquid extraction using APDC and IBMK.The method can also be used for lemonade and for Co Cu and Pb in white wine. Pre-concentration by extraction chromatography was used to determine trace amounts of Cd Pb and Zn in drinking water by FAAS (90/C2034). The characteristics of the column were described. It was found that the presence of 19 coexistent ions did not interfere with the method. Recoveries were 96- 103% with a precision better than 5%. Dabeka (90/3550) used ETAAS to measure Cd Co Ni and Pb in infant formulas and evaporated milk following HN0,-HClO digestion and co- precipitation with APDC. The same group (90/C1766) has re- ported the simultaneous determination of Al Cu Mn Mo and Sn in infant formulas and evaporated milks by ETAAS. Co- precipitation was not necessary but 20% citric acid was added prior to digestion with HNO to prevent Sn loss from blanks.Palladium was used as a chemical modifier for Al and Sn. Cervera et al. (90/1284) have reported the use of APDC and IBMK to complex and extract Cd Cu Pb and Zn from food brines after dry ashing. The organic extract was measured by FAAS and detection limits in brine were Cd 0.8; Cu 3; Pb 15; and Zn 15 ng g-'. Pre-concentration for specific elements was described in the following papers. Trace amounts of Ni in waters were deter- mined by ETAAS (90/243) with tungsten ribbon atomization after pre-concentration on activated carbon impregnated with cyclohexane- 1,2-dionedioxime and by FAAS (90/740) after solvent extraction with 2,4,6-tri-2-pyridyl- 1,3,5-triazine. The detection limit for the former technique was 0.3 ppb and for the latter 0.5 ppb of Ni.A 400-fold concentration was ob- tained in the determination of P in natural waters by Kubota et al. (90/742). Trace amounts of P were collected on a mem- brane filter as an ion pair of molybdophosphate and dodecyltri- methylammonium. After dissolving both filter and ion pair in DMF (2.5 ml) an aliquot (10 pl) of the DMF was measured by ETAAS using a zirconium treated graphite tube. Kimura et al. (90/3367) separated and pre-concentrated Mo"' from tap water river water and sea-water using activated carbon as col- lector. The effects of coexisting salts and the presence of EDTA on the adsorption capacity of the carbon were investigated. Real samples (200 ml) were concentrated to 1 ml by desorption of the MeV' from the activated carbon by 0.1 mol dm-3NaOH solution.Measurement was by ETAAS and the detection limit was 0.001 1 ng ml-I. The use of masking agents with FI pre-concentration was found to improve the determination of Pb in tap water by FAAS (90/35 12). The paper described the development and use of a buffer solution containing suitable masking agents (triethanolamine thiourea fluoride acetylacetone or cyanide) for the suppression of interferences. Analyses by FAAS using this method were in good agreement with results obtained by ETAAS. Lead was determined in vegetable tissues (90/1463) by ETAAS and (NH,)?HPO ( 1 % m/v) was used as chemical modifier in preference to La(NO,) following HN0,-HCIO di- gestion in an open beaker. In contrast Pb in potato samples was measured after dry ashing at 650 "C with H,SO followed by co-precipitation to concentrate the analyte (91/C754).To achieve accurate and precise results it was necessary to freeze dry the samples to facilitate grinding to a fine powder and to add MgO in addition to H2S0 as an ashing aid. The limit of determination was 5 pg kg-' dry mass. Freeze drying was also used as part of a method for the de- termination of 25 elements in milk (90/C3639). A combination of freeze drying with low-temperature ashing in an O2 plasma and subsequent acid digestion was found to be more promising and less cumbersome than conventional or bomb acid diges- tion. A combination of measurement techniques AAS ICP- AES and FI-AAS was used. Martinez et al. (91/C602) were able to screen preliminarily for 30 elements in milks down to ppb levels using ICP-MS after sample treatment in a Pan- microwave digestion bomb.A typical procedure was 1 ml of milk plus 1.25 ml of HNO plus 0.25 ml of H,02 heated for 2 min at medium power in a conventional microwave oven. Digests were cooled for 15 min at room temperature prior to measurement. Other workers (90/3472) reported the use of pressurized microwave vessels to digest food samples (0.2 g) which were subsequently measured for Br Ca Cu Fe Mg Mn P Sr and Zn by ICP-AES. Good recoveries (90-103.8%) and RSDs (0.5-9.65%) were reported. Closed PTFE vessels were used with microwave oven digestion for Ca Cu Fe Mg Mn and Zn in foods (90/1913). Knapp (90/3559) reviewed mechanical methods of sample decomposition for trace and ultra-trace analyses.Various aspects were discussed including the reduction of systematic errors (by for example reducing contamination) implementa- tion of the method mechanization of the decomposition process and reduction of decomposition time. The automation of some stages of sample preparation was found to have ad- vantages (90/C1765) in that it reduced the workload for the analyst and assured low blanks. A laboratory robot performedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 97R an open column separation procedure using Chelex 100 chelat- ing resin. This was necessary in order to remove alkali and al- kaline earth elements which could produce a variety of interferences when using laser-enhanced ionization (LEI) spectroscopy for the determination of Cu Mn Ni and Pb in some NIST RMs.Finally Kingston et al. (91/C736) discussed automated microwave sample preparation. The concept of automation as a potential way of standardizing laboratory methods as opposed to just being a way of controlling equip- ment was proposed. 2.2. Sample Introduction The effectiveness of solid and slurry sampling was reviewed by Welz and Sperling (90/C1683). Whilst direct solid sam- pling can minimize the risk of contamination and analyte loss the automation of such a procedure is difficult. Slurries have the advantage that they can be handled in the same manner as solutions and can therefore be more easily modified and cali- brated. The analytical signals and background absorption from food slurries were found to be similar to those from digested food samples.However interferences and chemical effects were similar to those observed for solid samples. Slurry sam- pling was recommended for refractory samples such as coal and ash with high analyte concentrations but not for samples which could be brought easily into solution. Milk powder was measured by both direct solid sampling and by slurry forma- tion (90/3288). The elements Cd Cr Cs Cu Mo Pb Se and Zn were determined by ETAAS. The solids technique was less suited for the refractory elements especially Mo than slurry analysis which was suitable for all elements. Accuracy and precision was checked using RM IAEA-11 standard milk powder. For slurry analysis accuracy and precision were 94- 11 1% and 5-20% respectively and for the solids method 91- 105% and 9-25% respectively.Powdered rice (1 + 2 mixture with EDTA) was analysed for Cu by direct solid micro-sampling in a graphite cup with Zeeman-effect corrected AAS (90/3529). Results were linear to 0.8 ng of Cu. The method was considered suitable for the determination of Cu in wheat corn and other cereals. The ad- vantages of incorporating an air-ashing stage into the ETAAS analysis of slurries were reported by Ebdon et al. (91/6). Re- duction of interferences from non-specific absorption and a significant decrease in the interference of MgCI on Pb deter- minations were found for the measurement of Co Cr Mn and Pb in five RMs. Burguera et al. (90/C3594) were able to measure Ca K Mg and Na in tropical fruits using FAAS.Ground fruit was mixed with Triton X-100 solution to form a slurry. Evaluation of the method with RMs showed it to be comparable in accuracy and precision to a traditional acid di- gestion procedure. Flow injection (30 pl samples) was used to determine total and free Ca in milk (91/155). Interference in the measurement of free Ca by spectrophotometry (at 530 nm) was removed by dialysis which also separated the free and total Ca. The latter was determined by FAAS at 422.7 nm. Interference from P043- in the measurement of total Ca was overcome by using an N20-C,H flame with suppression of ionization with K'. Flow injection was also used in a comparison of methods for the determination of Hg by cold vapour AAS (90/C2055). Two FI systems (one without amalgamation sample volumes from 0.2 to 1.0 ml; the other with amalgamation on Au-Pt gauze sample volumes up to 6.0 ml) were compared with a batch system also with and without collection of Hg by amal- gamation on Au-Pt gauze.The advantage of the batch systems was that the large sample size (maximum 50 ml) gave the best sensitivity and detection limit near 1 ng I-' when amalga- mation was used. However reagent consumption was high sample throughput low and automation difficult. Additionally interferences from Cu and Ni were pronounced. In contrast FI had the highest sampling frequency lowest reagent consump- tion and was easy to automate. It was least susceptible to heavy metal interferences but low sample volumes permitted only Hg concentrations greater than 0.5 pg I-' to be determined routinely. The use of continuous flow (CF) mode FI with amalgamation gave a 10-fold increase in sensitivity so that de- tection limits were 0.1 pg 1-I.Other parameters such as sam- pling frequency depended upon the sample size used and were between the values obtained for batch and non-CF-FI. Low detection limits and low blank values were obtained for the determination of Pb in food slurries by hydride generation AAS (90/35 10). When three oxidant media were evaluated K2Cr,0,-lactic acid was the most satisfactory for determining Pb in fish and vegetables. Whilst Pb in vegetables was accu- rately determined using an (NH,),S,O,-HNO medium it was only semiquantitative for Pb in fish. The final medium H,O,- HNO was unsatisfactory due to the decomposition of H,O by the organic matter in the samples.2.3. Speciation Studies The problems with and prospects for metal speciation studies in foods in particular the relative merits of in vitro HPLC- ICP-MS techniques and computer modelling approaches were discussed (9 1/C578). Chromatographic techniques can be an essential part of speciation studies and all of the following papers incorporate either LC or GC methods. The Holak method which uses LC-ETAAS for the determination of methylmercury in seafood was collaboratively tested (90/ 3429). Methylmercury was isolated from the blended samples (unspiked swordfish lobster and tuna; spiked lobster and tuna) by CHCI elution from diatomaceous earth-HC1 and then extracted into a small volume of 0.1 mol dm-3 Na2S,0,. An aliquot was injected onto the Zorbax column and eluted with MeOH-0.05 mol dm-3 NH40Ac containing 0.1% C,H,SH.A specially designed interface between a Zorbax ODS column and ETAAS was used to generate Hg vapour. The accuracy of the technique when compared with reference values was 94.4-99.6%. The method was adopted by the AOAC as an official first action. Jiang ef al. (90/1209) de- scribed a rapid method for the speciation of Hg in fish using capillary GC interfaced with AAS. Organomercurials in a benzene extract of frozen fish samples (0.5 g) were separated by GC using a WCOT glass column of OV-17. The eluate was passed to a pyrolyser at 700 "C and Hg detected by AAS. Re- coveries were 95% and the limit of detection for methylmercu- ry was 0.04 ppm on a 0.5 g sample.Gas chromatography- AAS was also used to analyse fish samples for organolead and organotin (9 1 /C685). Samples were enzymatically hydrolysed and extracted with diphenylthiocarbamate (ionic species) or hexane (for tetraalkyl species). Methyl or butyl Grignard rea- gents were used to form derivatives of the ionic alkyllead and - tin compounds. A similar derivatization procedure was used to determine ionic alkyllead in potable water by GC-AAS (90/ 3546). Species containing Ca Cu Fe Mg Mn and Zn in human milk and milk formulas were characterized using gel filtration chromatography with ICP-AES (90/1414). The distribution profiles for Cu Fe and Zn binding with various proteins (caseins albumins and lactalbumins) were different in human milk and milk formulas. The behaviour of Cu and Zn with citrate in human milk differed before and after birth.Citrate concentration increased by a factor of 25 after birth on the first day of lactation. Size-exclusion chromatography coupled to ICP-MS was used to conduct preliminary studies on the speci- ation of A1 in tea (90/C36 18). The estimated relative molecular mass of the Al species varied with pH and with the nature of in \itro enzyme treatment. For example at pH 6.6 A1 in a tea in- fusion was associated with estimated relative molecular masses of 1 10oO and 6200 u. In contrast after in riti-o gastric treatment at pH 2.5 a single peak corresponding to a relative moleclular mass of less then 2000 u was obtained. Favier et al.98R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 Table 2 SUMMARY OF ANALYSES OF FOODS AND BEVERAGES Element Matrix A1 Canned drinks A1 Al A1 Beer Tea Tea Al Foods A1 Foods As Vegetables herbage As Tomato products As Rapeseed refined and hydrogenated oil As Foods As Drinking water As Food additives As Mussel tissue fish B Peach leaves drinking water B Wine Ca Soft drinks Ca Skimmed milk calcium enriched milk water-cress soup Ca Milk Cd Foods Technique; atomization; analyte form* AA;-;L AA;ETA;L NMR;-;L AE;-;L MS;ICP;L AA;ETA;L AE;ICP;L NAA;-;L AA;ETA;L XRF-;S AA;Hy;L AA;ETA;L A F; H y ; L AA;FL AA;ETA;L XRF;-;- AA;ETA;L AE;F;L AA;F;L MS;FAB;L AA;F,N,O-C*H2;L AA;ETA;- Sample treatmentlcomments Lacquered A1 cans were filled with de-ionized water and various buffers and stored at 20°C. The solutions were analysed by AAS at intervals over 60 d and the leaching of Al from the cans examined Platform atomization peak-area integration deuterium-arc background correction chemical modification and the 396.2 nm line combined were found to be suitable conditions for reliable determination of Al in beer attack in a micro-Kjeldahl flask.5% (vlv) HNO was used to dilute the digests and tea infusions prior to NMR analysis Al species in tea infusions were separated using size- exclusion chromatography. The molecular nature of A1 species was investigated following simulated gastro-intestinal digestion of tea Three techniques were compared for the analysis of food RMs. ETAAS gave the most reliable and consistent results Matrix interferences in the determination of Al in liquid foods were overcome by using L'vov platform (NH&HPO chemical modification and in situ O2 ashing Samples ( 1 .O g) were analysed by EDXRF following co-precipitation.Results for grasses broccoli and spring cabbage agreed with those obtained by NAA As in crushed Spanish tomatoes was determined by dry ashing. The ash was dissolved in HCI. Following hydride generation atomization was in a flame heated quartz cell Inorganic As and triphenylarsine were determined by STPF-ETAAS using Pd and Ni as chemical modifiers (in Chinese) Foods were digested in HN03 dissolved in HCI reduced and As and Sb measured by hydride generation non-dispersive AFS (in Chinese) TO separate As from interferences an Sb-Mo-As heteropoly acid was formed and extracted with IBMK and KBH,-NaCl (in Chinese) A half cylinder tantalum-coated graphite furnace was applied to determining As in food grade H,PO,.The determination was performed in 8-1Ooh H2 in Ar as the H2 was found to prolong furnace lifetime (in Chinese) Off-line EDXRF was used to determine As species in eluate from an anion-exchange column A method using Ca-Mg as chemical modifier was described (in Chinese) Samples were evaporated and Me3B0 generated from the solid by the action of H,SO and MeOH The composition and effect on teeth of eight of the most popular UK soft drinks were examined. AAS was used to investigate demineralization of dental enamel FABMS was applied to determining Ca absorption using a double label stable isotope technique Tea leaf digests were prepared by HN0,-HCIO Total Ca in milk was determined by AAS (422.7 nm) and free Ca by spectrophotometry (530 nm). The two forms were separated by a dialyser which also eliminated interferences in the determination of free Ca This report related the application of a commercial probe system to Cd determination in a range of food RMs.Chemical modifiers were not used Reference 9011 297 901191 1 9014 16 1 9OlC36 18 9 1 /C694 91x739 901 166 9013292 9 1 I405 9 1 I446 9 1 I449 9 11450 91lC652 901345 1 91/91 9011 924 9OlC3 9 8 3 9 90/( 1155 369499R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 Table 2 SUMMARY OF ANALYSES OF FOODS AND BEVERAGES-continued Element Matrix Technique; atomization; analyte form* Cd Vegetables drinking water AA;ETA;L Cr Milk powder c s Waters c u c u c u c u c u F Fe Fe Hg Food Rice Bread Food Edible oils fats Foods Vegetable oils Edible oils fats Fish Hg Wine Hg Vegetables Hg Seafood Hg Waters I Oyster tissue Mn Flour AA ; ETA ;- AEiICPi- AA;ETA;L AA;ETA;S MS;-;- Molecular absorption;-;L AA;ETA;L AA;quartz furnace;L AA;cold vapour;L AA;-;L AA;cold vapour;L MS;-;S AA;ETA;L Sample treatmentlcomments Environmental pollution in three different areas of Spain was monitored by determining the levels of Cd Pb and Sb in a range of produce and water The relative merits of a number of techniques for Cr determination in IAEA RMs were critically assessed Cs was determined in mineral and thermal waters using the standard additions technique.Sample solution (20 ml) was mixed with 0.2Oh HN03 introduced into the graphite furnace and determined at 852.1 nm.The detection limit was 1.85 pg I-' Diet studies using stable isotopes indicated that Cu absorption is strongly dependent on dietary Cu levels Powdered rice was mixed with EDTA and placed in a graphite cup for ETAAS. The method was also proposed for wheat corn and other cereals (in Chinese) The results of experiments involving the intrinsic and extrinsic labelling of bread were used to support the use of extrinsic Cu tracers to assess Cu absorption Stable isotope analysis was used to determine the effect of two dietary Zn levels on faecal Cu levels and Cu absorption Following a collaborative study the described method was adopted as an official first action by IUPAC-AOAC Molecular absorption in a graphite furnace was used to measure F as AlF3 at 227.45 nm.A variety of mineralization techniques for preparing liquid and solid samples were evaluated (0.25-0.50 g) which were diluted to 100 g with THF (in Chinese) ETAAS was used to determine Fe in oil samples See Cu ref. 911309 Coupled pyrolysis capillary GC-AAS at 253.6 nm was used to speciate Hg in fish. Frozen fish (0.5 g) was homogenized with CuSO and H20 mixed with acidic KBr solution and extracted with benzene. An aliquot of the dried organic phase was injected Samples were digested by HN0,-H2Cr0,-H2S04 and Hg2+ then reduced by Sn2+. The metallic Hg was swept into the quartz furnace by Ar H,S04. Results for RMs by the proposed method compared favourably with those found using the AOAC digestion unspiked seafood using AAS coupled to HPLC via a novel vapour generation interface. The complete method which included CHC13-Na2S20 sample extraction was adopted as an AOAC official first action Three methods were tested a batch system with and without collection by amalgamation on-line FI without amalgamation and on-line FI with amalgamation precipitation with AgCI.After drying as silver ammonium iodide on a tantalum filament I was thermally desorbed in a resonance ionization mass spectrometer Digestion was performed by HN03-HC104 attack. Mn was measured at 279.5 nrn using platform ETAAS (in Chinese) Samples (0.5 g) were digested using K2Cr20 in dilute Methylmercury was determined in spiked and Samples were wet ashed and I separated by co- 901 901 Reference 9 11229 206 478 9012863 9013529 9014 1 3 1 9014 132 9 1 I309 9013508 901 1 400 9 1 I309 9011 209 90/ 1 509 9012 1 13 9013429 901c2005 9012300 9 11396IOOR JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL.6 Table 2 SUMMARY OF ANALYSES OF FOODS AND BEVERAGES-continued Mo Tap water river water and sea-water Pb Pb Pb Pb Pb Pb Rb Sb Tap water Potable water Drinking water Vegetables drinking water Fish Potatoes Mineral and thermal water Water plant and animal tissues Sb Foods AA;-;L AA;F;L AA;ETA;L AA;-;L AA;ETA;L AA;FL AA;ETA;L AFHy ;L Mo Milk AA;ETA;L Mo Infant formula human milk AA;ETA;L Ni Natural waters AA;ETA;L Ni Edible oils fats P Milk Pb Vegetables Pb Wine Pb Food AAiETAi- AA;ETA;L AA;ETA;L AA;Hy;L AA;Hy;Sl Technique; atomization; Matrix analyte form* Sample treatmentlcomments Element Mo Foods AA;ETA;L Freeze-dried samples were ashed for 24 h at 500 "C a few drops of 6 mol dm-) HN03 added and after a further 5 h ashing the ash was dissolved in 0.5 mol dm-3 HN03.The chosen wavelength was 313.3 nm Mo ions were collected from water samples and pre- concentrated using activated carbon. The Mo was desorbed by NaOH prior to analysis by ETAAS. The detection limit was 0.001 1 pg 1 - I (in Japanese) Milk samples with octyl alcohol were introduced directly into the furnace. The relative merits of HN03 BaF or neither as chemical modifier were investigated !;ample plus BaF chemical modifier were injected directly into the graphite tube. The limit of detection was 0.89 pg 1-' Activated carbon impregnated with cyclohexane-l,2- dionedioxime was used to pre-concentrate Ni prior to ETAAS.Detection limit 0.3 pg 1-' (in Japanese) AA;ETA;L See Cu ref. 911309 P in milk was determined by ETAAS with Zeeman- effect background correction and an STPF. A chemical modifier of La(N03)> was used HN03-HClO was applied to vegetable samples. The chemical modifier was (NH,),HPO,. Pb recoveries from the samples were between 100 and 101% (NH4)2S203 and NaBH,. Atomization occurred in a flame-heated silica tube Three different oxidizing media for generating lead hydride from food slurries were compared (i) H202-HN03; (ii) (NH,),S203-HN0,; and (iii) K2Cr207-lactic acid. Method (iii) was found to offer the lowest detection limits The determination of Pb in spiked tap water by FI- FAAS was used to demonstrate the efficiency of masking agents in overcoming transition metal interferences during pre-concentration of Pb.The analyses were performed at 283.3 nm Grignard derivatization and coupled GC-AAS was applied to environmental samples. Alkylleads were present as ultra-trace contaminants Data obtained using the Delves cup FAAS method were found to agree well with that obtained by ETAAS A three step wet digestion procedure using Lead hydride was generated from wine using AA;F air-C2H2;L A speciation method involving NaDDC extraction See Cd ref. 911229 F'b and Sn in fish caught in the Great Lakes were speciated by coupled GC-AAS. Grignard reagents were used to derivatize the analytes prior to A method involving freeze drying grinding ashing GC-AAS and concentration by co-precipitation was described for the low level determination of Pb in potatoes.Levels were found to be typically 8 pg kg-I dry mass The conditions for FAAS determination of Rb in water were described Following extractive separation with N-(4- methoxyphenyl)-2-furylacrylohydroxamic acid Sblll and SbV were determined by ETAAS using Cu as chemical modifier. The chosen wavelength was 217.6 nm See As ref. 911446 Reference 901 1283 9013367 9013507 911251 90/243 9 1 /309 9013296 901 1463 901 1479 9013510 9013512 9013546 901C3017 9 11229 911C685 9 1/C754 9011 507 901 1473 9 11446JOURNAL OF ANALYTlCAL ATOMIC SPECTROMETRY APRIL I99 I VOL. 6 l O l R Table 2 SUMMARY OF ANALYSES OF FOODS AND BEVERAGES-continued Technique; atomization; analyte form* Element Matrix Sample treatmentfcomments Reference 9 1 I229 901 166 9011278 9011 550 9012057 9012247 901245 1 9014 152 91lC685 9011417 9012820 901282 1 9012862 9013 52 7 9OlC3983 901 1284 9011 365 9011 398 9011 399 9011 4 14 9011 839 901 19 12 9011 9 13 9011959 Sb Vegetables drinking water AA;H y;L See Cd ref.91/229 Se Vegetables herbage XRF-;S See As ref. 90/ I66 Se Infant milk formula AA;ETA;L Prior to ETAAS analysis samples were digested with H2S04-HN0 ( 1 + 1). Pd was used as chemical modifier. The wavelength used was 196 nm (in French) Se Milk AA;ETA;L Zeeman-effect corrected ETAAS with an STPF was used to analyse milk following dilution. Results were compared with those obtained by fluorimetry following digestion and extraction Samples were microwave digested with HN03 and then analysed by Zeeman-effect ETAAS using Pd(N03) mixed with Cu(NO,) as chemical modifier Se Bovine liver AA;ETA;L Se Food AE;ICP;L As part of a programme to monitor the intake of Se by the Italian population a continuous flow HG ICP-AES method for Se determination was developed separated by HG and SeH absorbed in HNO on which isotope dilution TIMS was performed tube-flame atomizer for the HG-non-dispersive AF determination of Se was constructed.Results for RMs agreed well with certified values. An absolute detection limit of 27 pg of Se was quoted Food samples were digested (HN03-HC10,) Se A combined electrothermally heated quartz See Pb ref. 911C685 ETAAS measurement of foods found the general concentration to be t 7 ng g-' wet mass A review with seven references of the determination of Zn absorption by stable isotope FABMS FABMS and TIMS were compared for evaluation of human Zn absorption from cereals Zn homeostasis in humans was investigated by monitoring Zn by TIMS A separation and pre-concentration technique using 8-sulphydrylquinoline-polyurethane foam was used to determine Zn at the pg 1-' level in water (in Chinese) investigate the effects of different food processing techniques on Zn absorption The sample treatment was ashing HN03 digestion and extraction with APDC and IBMK (Cd Cu Pb Zn) (in Spanish) A method for the simultaneous determination of five elements in bottled and canned beer was described (Cd Co Cu Pb Zn) The results for seven different elements were used to evaluate two sample pre-treatment procedures (Ca Fe K Mg Mn Na Zn) (in Japanese) Low-temperature ashing was found to give similar values to acid digestion when determining minerals in food (Ca Fe K Mg Na) (in Japanese) Coupled gel filtration chromatography-ICP was used to characterize the species containing six different elements (Ca Cu Fe Mg Mn Zn) The correlation between elemental levels in greenhouse soils and vegetables was investigated The level of 15 elements in Chinese teas was measured (in Chinese) Samples (0.2 g) were microwave digested in HN03-H20 prior to ICP-AES analysis (Ca Cu Fe Mg Mn Zn) (in Chinese) Three mass spectrometric methods were used to The paper lists additions to available biological RMs Animal tissue MS;-;L Se Se Reference materials AF;Hy ;L Sn Te Fish Foods AA;-;- AAiETAi- Zn Food Zn Cereal MS;-;- Zn Food MS;-;- Zn Water AA;F;L Zn Foods Various (4) Food brines MS;ICP TI FABi- AA;F;L Various ( 5 ) Beer AA;F;L Various ( 7 ) Canned fruit tomatoes AA;-;L vegetable juices Various (5) Food AA;-;L Various ( 6 ) Human milk milk formulas AE;ICP;L Various (8) Vegetables Various (1 5 ) Chinese teas Various (6) Food AA;-;L AE;ICP;- A E; IC P; L Various Food animal and plant tissues .. -- ,102R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991. VOL. 6 Table 2 SUMMARY OF ANALYSES OF FOODS AND BEVERAGE,S-continued Technique; atomization; analyte form* AA;ETA;L Reference 9012089 9012 169 9012 1 93 901283 1 901 328 8 90/3300 9013427 901355 1 9013584 90/40 1 1 9014 145 9OlC 1 640 90/C 1 683 901C1765 90K1766 90/C2 104 9 O/C2 9 2 9 Element Matrix Various Reference materials Sample treatmentkomments effect ETAAS analysis of a range of RMs in less than 1 min per sample.An STPF was used for the determination trace elements from fresh water (Bi Cd Cu Fe Mn Ni Pb Zn) beaker-HN0,-H,O digestion allowed multi- element analysis of three RMs The determination of stable isotopes by an elemental analyser and MS were reviewed and several applications reported (in Italian) elements in milk were reported (Cd Cr Cs Cu Mo Pb Se Zn) A review of recent publications concerned with analytical topics 22 laboratories participated in a collaborative study to develop mineralization procedures for toxic elements in food The dietary intake of infants aged 0- 12 months was estimated for four elements (Cd Co Ni Pb) Analytical results for a range of biological RMs analysed by SIMAAC were reported €:I-FAAS or FI-FAES using a stirred dilution chamber was applied to water analysis (Ca K Mg Na) Vanadium was evaluated as a modifier.The maximum pre-treatment temperatures of 25 elements were recorded A review of the first 10 years of the NIES CRM programme. Candidate CRMs especially for organometallics were discussed l’he merits of solid and slurry sampling in ETAAS were critically appraised Chelation chromatography separated alkali and alkaline earths from sample matrices prior to trace element determination by LEI spectroscopy (Cu Mn Ni Pb) of five elements was described. Citric acid was used to prevent loss of Sn from blanks (Al Cu Mn Mo Sn) Trace-element profiles and multivariate analysis allowed the country of origin of agricultural products to be identified (B Ba Cu K Mg Mn P Rb Zn) Ion-pairing HPLC-DCP-AES in combination with FTIR were applied to trace metal speciation of foods l h e country of origin of nuts was determined using discriminant analysis.Several statistical approaches were evaluated (Ba Ca Cu Fe K Mg Mn P zn) The value of chemometric methods in understanding complex data relationships was described The philosophy and experimental detail of a programme to establish reference values for milk were outlined Pollution from a coal fired power plant was evaluated by monitoring trace element levels in transplanted fresh water mussels Ground fruit was suspended in Triton X-100 prior to aspiration. Results were comparable to those obtained by wet-acid digestion (Ca K Mg Na) Milk analysis by freeze drying followed by low- temperature ashing in an oxygen plasma and subsequent acid digestion was described The omission of the pyrolysis step allows Zeeman- Functional filter-papers were used to pre-concentrate A combination of microwave-HNO with open Various methods for determining essential and toxic An ETAAS method for simultaneous determination Various (8) Fresh water XRF-;S Various (24) Animal muscle liver oyster MS;ICP;L Various Alcoholic beverages food MS;-;- Various (8) Milk milk powder AA;ETA;L S S1 Various (28) Food beverages Various (7) Food Various (4) Evaporated milk infant formulas Various (1 6) Biological materials AA;ETA;- AA;F NZO-C?Hl ETA;L AE;F;L AA;F;L AA,ETA;L S1 Various (4) Water Various (25) Biological and environmental reference materials Reference materials Various Various Various (4) Various ( 5 ) Various (9) Various Various (9) .. -9-9- Foods AA;ETA;S S1 LEI;-;- Apple leaves diets peach leaves Evaporated milk infant formulas AA;ETA;L Orange juice nuts AEiICPi- AA;-;L Foods AE;DCP;L Nuts AE;ICP;L AA;-;L 17 18 59 90K3 901C3 901C3 Various (1 5 ) Orange juice Various (1 3) Human milk AE;ICP;L AE;ICP;L Various (26) Fresh water mussels AA;-;L PIXE;-;- 90K3360 9OlC3594 90/C3639 Various (4) Tropical fruits AA;F;SI -;-;L Various (25) MilkJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 103R Table 2 SUMMARY OF ANALYSES OF FOODS AND BEVERAGES-continued Element Matrix Technique; atomization; analyte form* Sample treatment1comments Various ( 10) Wine Various Chinese tea Various (21) Human diets Various (10) Wheat flour Various Food Various ( 1 5 ) Food Various (30) Milk Various (7) Foods Various (1 1) Mussel tissue Various Reference materials AE;ICP;L XRF;-;- AA;-;- MS;ICP;L AEiICPi- AA;-;L -;-;L Various (4) Tomato leaves mussel AA;ETA;SI Air ashing was found to reduce interferences from non-specific absorption significantly and allow an extended range of ashing temperatures (Co Cr Mn Pb) allowed successful analysis of biological CRMs.The total heating cycle lasted 45 s (Cu Fe Ni Pb) grape type and origin of wines to be assigned (Al Cd Co Cr Cu Fe Mn Pb V Zn) The role of XRF in certification exercises for SRMs was discussed. Results obtained by XRF were in good agreement with those obtained by other analytical techniques Dietary composition for three income categories in India were assessed by NAA and AAS.Adequate amounts of essential trace elements were found in all categories HN03-HC104 ( 5 + 1) and then analysed by ICP- AES (Al Ba Ca Cu Fe Mg Mn Ni Y Zn) An in vitro gastric simulation was used to digest foods prior to metal speciation by coupled SEC-ICP- MS. This approach was compared with a computer model developed to predict major metal species present in the gut and Drug Administration dietary survey cow to the consumer was monitored by ICP-MS. Characteristic relationships between specific elements and fat and/or protein were found The results of a collaborative study of heavy metals hydrocarbons and PCBs in German food were presented Progress in certification of RMs for marine monitoring was reported A prototype expert system for computer control of microwave digestions was described.The system was able to adapt to the level of automation in different instrument configurations Three different dissolution procedures for determining heavy metals in 13 different samples were described. ICP-MS with N2 addition was used for the final analysis (As Cd Hg Pb Sn U) Various (4) Bovine liver milk powder AA;ETA;L Hot-injection and Pd-H2 chemical modification ICP-AES analysis of ten trace elements allowed the AA;-;L NAA;-;- AE;ICP;L Wheat flour was open vessel digested with MS;ICP;L This presentation gave an overview of the US Food The complete inorganic composition of milk from the Various (6) Vegetarian and vegan food supplements MS;ICP;L Reference 9116 91172 911105 9 11206 9 11304 911331 911C578 911C581 9 1 /C602 iC643 lC646 lC736 9 IlCS06 *Hy indicates hydride generation and S L G and S1 signify solid liquid gaseous or slurry sample introduction respectively.Other abbreviations are listed elsewhere. (91/C652) described the determination and speciation of As using ion-exchange chromatography with EDXRF. They con- cluded that the capability of the method to recover different forms of As (AsOd3- AsOz- monomethylarsonic acid and di- methylarsinic acid) was sufficient to allow routine determina- tion of As species from a mixed solution down to 10 ng ml-I. 2.4. Developments in Methodology for Atomic Absorption Spectrometry 2.4.1 . Flame atomic absorption spec~tr.ometry Sanz et al. (90/1479) reported the use of aflame heated silica tube in the determination of Pb in wine by hydride generation AAS.Lead was generated directly from the wine matrix with (NHJ2S203 as the oxidizing agent and NaBH as the reducing agent. The lead hydride was atomized in a flame heated silica tube and the Pb concentrations in wine determined in this way agreed with those obtained using three other methods. Flow in- jection FAAS with a dilution chamber was used to determine Ca K Mg and Na in water (90/4011). The dilution chamber made it possible to carry out the calibration using the dilution profile of a single concentration standard for each element. 2.4.2. Elecwothermal utomic- absorption specti-ometr-y Both FAAS and ETAAS were used (90/3584) for multi- element analyses of biological materials with a prototype simultaneous multi-element AAS (SIMAAC) system which in- cludes a continuum source and an khelle polychromator.Data were reported for a variety of RMs including ETAAS results for NIST SRM 1568 Rice Flour a freeze-dried urine andI04R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 plasma protein solutions and FAAS results for a diet RM and other food RMs. Compromise atomization conditions did not significantly limit accuracy or precision. Several papers (90/1278 90/1283 90/1550 90/2057 90/ 3507,90/4145 91/251) reported the use of chemical modifiers with ETAAS. Selenium was determined in milk samples using Pd [O.I% solution in 2% HCI (90/1278) or Pd in HNO and Mg(NO,) (90/1550)] as modifier and in bovine liver following microwave digestion [Pd(NO,),-Cd(NO,) mixed modifier (90/2057)] to prevent loss of volatile Se.Barium difluoride was used as a chemical modifier in the determination of Mo in untreated infant formulas (90/3507 9 1/25 1). The BaF- was thought to produce preferential formation of Mo fluorides hence reducing carry-over between injections. The detection limit for this method was 0.89 pg I-' of Mo. Pinna et al. (90/ 1283) measured Mo in human samples and diets by ETAAS following ashing of samples in a muffle furnace and dissolu- tion of the ash in 0.5 mol dm-3 HNO,. No chemical modifier was added and the results suggested to these workers that the technique would be suitable for use in balance studies in humans. The potential of Vv as a chemical modifier in ETAAS was evaluated by Tsalev et al.(90/4145). The V was in the form NH,VO and the determination of 25 elements was investiga- ted. For volatile analytes a mixed V-Pd modifier provided thermal stability. Both slurried and digested RMs were ana- lysed. It was concluded that Vv has good potential as a chemi- cal modifier particularly for As Cd P Pb Pd se Sn and TI. The modifier was compatible with alkaline sample solutions samples solubilized in tetramethylammonium hydroxide (TMAH) and slurries. The mixed V-Pd was very efficient in the thermal stabilization of Ag As Bi Cu Ga Ge In Mn P Pb Sb Se Sn Te and TI. Chemical modifiers were omitted for some determinations employing stahilized temperature platform furnace (STPF) methods. Slavin et al. (90/2089) analysed RMs for several ele- ments including As Cd Cr Cu Ni and Pb to test the feasibi- lity of fast STPF methods omitting both the pyrolysis step and in most instances chemical modification but using Zeeman- effect correction for large background signals.Highly mineral- ized waters were measured for As Cd Pb and Se using an STPF method with mixed Pd-Mg(NO,)? as the chemical modifier with Zceman-effect background correction (91/7 I). The use of a pre-atomization cool-down step to 20 "C reduced background absorption whilst increasing the Mg(NO,) im- proved the peak shape. However interferences still remained for waters containing Na,SO and MgSO,. Zhou et al. (91/405) found that addition of Pd to edible oils being measured for in- organic As and triphenylarsine by STPF-ETAAS gave better recoveries than addition of Ni.A tantalum-coated tube was used with ETAAS to determine As in food-grade H,PO used as a food additive (91/450). Addition of H (8-10% in Ar) pro- longed the tube life. Low levels of Cd were determined in some food RMs by probe atomization without chemical modification (90/C3694). An open quartz tube method was de- scribed for determining Hg in wine by ETAAS (90/1509). An electrothermally heated quartz tube was combined with a flame atomizer for hydride generation non-dispersive AFS measurement of Se in RMs including fish liver and muscle tissues (9014152). 2.5. Developments in Methodology for Plasma Emission Spectrometry Kardos et al. (90/2247) reported the measurement of Se in some Hungarian and Italian foods using hydride generation with ICP-AES.Samples were digested using a laboratory mi- crowave system and the over-all procedure was suitable for a routine and reliable Se assay. An overview of the determina- tion of trace metals in foods by DCP-AES was given by DeMenna and Jacobs (90/C2929). The robustness of the tech- nique was highlighted and the use of ion-pairing HPLC with DCP-AES to remove alkali and alkaline earth salts from complex food matrices was reported. Safronova et al. (91/220) described a direct method for the simultaneous determination of 20 elements in waters using ICP-AES with a water-cooled torch. In a routine method for multi-element analysis of wheat flour by ICP-AES (9 1/33 I) coefficients for spectral interfer- ences by Ca Fe and Mg in the determination of Ba Cu Ni Y and Zn were determined and applied.2.6. Developments in Methodology for Inductively Coupled Plasma Mass Spectrometry Very few papers have appeared in this review year which deal with ICP-MS for food and beverage analyses. Friel et al. (90/ 2193) analysed three RMs following a cornhination of micro- wave dissolution with open beaker digestion using HNO and H,O,. An on-line five element internal standard was used to correct for chemical and instrumental drift effects. Foulkes et al. (91/C806) compared three dissolution procedures for the determination of As Cd Hg Pb Sn and U in 13 samples of sea vegetables and derived 'food supplements'. Microwave di- gestion bombs gave better recoveries and repeatibility than open tube reflux methods. In order to reduce polyatomic inter- ferences in As determinations it was necessary to use N addi- tion to the carrier gas.2.7. Topical Applications 2.7.1. Aluminium in foods There is still considerable interest in all aspects of A1 in foods and its relation to dietary uptake and subsequent hioavailabili- ry. The possible leaching of Al ions from drink containers was studied by Sugden and Sweet (90/1297) and the possible toxi- cological hazard discussed. It was concluded that leaching of Al from drink cans does not present a significant risk to people with healthy kidneys. Beer was analysed by various ETAAS methods in order to develop a reliable and accurate method for Al in this beverage (90/1911). Both normal off the wall and platform atomization employing either a direct concentration mode or standard additions resulted in inconsistent but always high recoveries for added Al to beer.By using peak area rather than peak height integration incorporating chemical modification and deuterium-arc background correction in con- junction with platform atomization at 396.2 nm consistent re- coveries of 99-100% were obtained. The method was verified by using standard additions. The determination of Al in tea has been mentioned above (section 2.3). In addition NMR (90/ 4161) has been used to quantify Al in tea. The results agreed with those determined by AES. Delves (91/C739) presented data from bioavailability studies in which small increases in the serum and urine Al of volunteers were detected following consumption of certain food/drinks. Matrix interferences in the determination of Al in most liquid foods serum and urine were reported to be overcome by using L'vov platform ETAAS and chemical modification with (NH,)H,PO and with in situ 0 ashing.Nine food RMs were used to compare instru- mental methods for the determination of trace amounts of Al (91/C694). Samples were subjected to dissolution by micro- wave heating for ETAAS and ICP-AES measurements and were irradiated and counted directly as solids for instrumental NAA. The last technique was subject to interference by P and Si whilst ICP-AES was too insensitive. The most consistent and reliable Al results were produced by ETAAS. 2.7.2. Reference materials and c*ollahorati\fe trials Five conference reports reviewed aspects of the cei-tificarion and production of RMs (90/C643 90/C646 90/C1640 90/ C 1641 90/C3 159).The presentations covered the broadJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 105R aspects of certification procedures and problems as well as dealing with specific materials e.8.. human milk (90/C3 159). A list of additional biological RMs has been compiled (90/ 1959). Other published papers dealt with a collaborative study of a method for Cu Fe and Ni in oils and fats by ETAAS (91/ 309) and an inter-laboratory study of a method for toxic ele- ments in food products (90/3427). The former paper from the American Food and Drug Administration described an IUPAC-AOAC method whilst the latter gave results which were used to develop a national standard for the USSR. 2.7.3. Characterization studies Several workers have reported data used to characterize or ‘finger-print’ some foodstuffs.This type of data is multi- element or deals with isotopic abundances. For example Fav- retto et al. (90/1180) used principal components analysis to classify polluted and unpolluted mussels having determined ten element concentrations by ETAAS. Fruit juices have been investigated with regard to the differences in method of pro- cessing (90/1935) and in the country of origin (90/1935 90/ C2 104 90/C3 1 18). Spectrochemical analysis of orange juice has been studied for many decades (90/C3 118) and the con- centrations of many major and minor elements are well known. Similarly analysis of some 50 inorganic components of wine by ICP-AES allowed grape type and the origin of wines to be defined and assisted in helping to assign a vintage year (91/105).The country of origin of edible nuts and orange juice was determined following microwave diges- tion and measurement (primarily by ICP-AES) using trace element profiles and multivariate data analysis (90/C2 104 90/C3 1 17). LOCATION OF REFERENCES The full list of references cited in this Update have been published as follows 90/C1-90/C539 J . Anal. At. Spectrom. 1990,5( I) 56R-74R. 90/C540-90/1158 J. Anal. At. Spectrom. 1990,5(2) 131R-150R. 90/1159-90/C1460 J. Anal. At. Spectrom. 1990,5(4) 167R-178R. 90/1461-90/2277 J . Anal. At. Spectrom. 1990,5(5) 215R-242R. 90/2278-90/359 I J . Anal. At. Spectrom. 1990,5(7) 278R-32 1 R. 90/359 1-90/4 179 J . Anal. At. Spectrom. 1990,5(8) 36 1 R-378R.91/1-91/825 J. Anal. At. Spectrom. 1991,6( I) 41R48R. Abbreviated forms of the literature references quoted (excluding those to Conference Proceedings) are given on the following pages for the convenience of the readers. The full references names and addresses of the authors and details of the Conference presentations can be found in the appropriate issues of JAAS cited above. Abbreviated List of References Cited in Update 901148. Anal. Proc. 1989 26 258. 90/166. Analyst 1989 114 919. 901243. Bunseki Kagaku 1989 38 120. 90/740. Bunseki Kagaku 1989 38 252. 90/742. Bunseki Kagaku 1989 38 322. 9111179. Anal. Chim. Acta 1989 220 257. 90/1180. Anal. Chim. Acta 1989 220 135. 9011198. Anal. Chem. 1989 61 1834. 90/1203. Anal. Lett. 1989 22 1341. 90/1206.Fresenius Z. Anal. Chem. 1989 333 33. 90/1209. Fresenius Z. Anal. Chem. 1989,334 27. 90/1248. Microheam Anal. 1989 24 23. 90/1249. Microheam Anal. 1989 24 26. 9011278. Guangpuxue Yu Guangpu Fenxi 1988 8(5) 54. 9011283. J . Micr-onutr. Anal. 1989 5 127. 90/1284. Rev. Agroquim. Technol. Aliment. 1989 29 117. 9011288. Am. lnd. Hyg. Assoc. J. 1989 50 245. 90/1291. Clin. Chem. (Winston-Salem N.C.) 1989 35 488. 90/1297. Pharm. Acta Helv. 1989 64 130. 90/1308. Hejishu 1989 12 243. 90/1315. Biryo Kinzoku Taisha 1989 (17) 39. 9011316. Yingyang Xuehao 1989 11 42. 90/1326. J. Micronutr. Anal. 1989 5 181. 90/1327. J. Clin. Chem. Clin. Biochem. 1989 27 345. 90/1330. Dent. Muter. 1989 5 256. 9011339. Analyst 1989 114 1249. 90/1351. Emiron. Sc-i. 1988 34 (Chem. Prof. Environ 1987) 333.90/1358. Diandu Yu Huanhao 1989 9(2) 24. 90/1360. J . Appl. To-vicol. 1989 9 2 19. 9011361. J. Environ. Pathol. Toxicol Oncol. 1989 9 115. 9011365. J . Inst. Brew. 1989 95 257. 9011367. Andrologia 1989 21 140. 90/1368. Biol. Trace Elem. Res. 1989,20 233. 90/1381. J . Soc. Cosmet. Chem. 1988 39 383. 9011383. J . Pediatr. (St. Louis) 1989 114 865. 9011384. lgaku to Seihutsugaku 1989 118 263. 90/1394. Yaowu Fenxi Zazhi 1989 9(2) 82. 9011395. Rinsho Kensa 1989 33 221. 90/1398. Chiha-ken Eisie Kenkyusho Kenkyu Hokoku 1988 ( 12) 6 1. 90/1399. Osaka-furitsu Koshu Eisei Kenkyusho Kenkyu Hokoku Shokuhin Eisei Hen 1988 19 17. 90/1400. Zhongguo Kexue Jishu Dame Xuehao 1989 19 141. 9011407. Trace Elem. Anal. Chem. Med. Biol. Proc. Int. Workshop 5th 1988. 9011408. Trace Elem.Anal. Chem. Med. Biol. Proc. Int. Workshop Sth 1988 25. 9011409. Trace Elem. Anal. Chem. Med. Biol. Proc. Int. Workshop 5th 1988 48. 9011410. Trace Elem. Anal. Chem. Med. Biol. Proc. Int. Workshop 5th 1988. 72. 90/1411. Trace Elem. Anal. Chem. Med. Biol. Proc. lnt. Workshop 5th. 1988 78. 90/1412. Trace Elem. Anal. Chem. Mad. Biol. Proc. lnt. Workshop 5th 1988 84. 90/1413. Trace Elem. Anal. Chem. Med. Biol.. Proc. lnt. Workshop 5th 1988 90. 9011414. Trace Elem. Anal. Chem. Med. Biol. Proc. Int. Workshop Sth 1988 145. 9011415. Trace Elem. Anal. Chem. Med. Biol. Proc. Int. Workshop 5th 1988 167. 90/1417. Trace Elem. Anal. Chem. Med. Biol. Proc.. lnt. Workshop 5th 1988 412. 90/1463. Analyst 1989 114 1397. 90/1473. Anal. Lett. 1989 22 237. 90/1478. Microchem.J. 1989 40 103. 90/1479. Micr-ochem. J. 1989 40 115. 9011483. Clin. Chem. (Winston-Salem N . C.) 1989 35 19 I 1. 90/1484. Clin. Chem. ( Winston-Salem N . C.). 1989 35 1986 90/1499. Anal. Chim. Acta 1989 224 139. 9011506. Bunseki Kagaku 1989 38 T. 140. 9011507. Anal. Lett. 1989 22 1601. 9011509. At. Spectrosc-. 1989 10 85. 9011547. Huanjing Kexue 1989 10(4) 49. 90/1549. J. Trace Elem. Electrolytes Health Dis. 1989 3(2) 97. 90/1550. Neth. Milk Diary J. 1989 43 185. 9011551. Beitr. Gerichtl. Med. 1989 47 97. 90/1553. Huaxue Xuehao 1989 47 804. 90/1833. Sci. Total Environ. 1989 84 25. 90/1838. Int. J. Environ. Anal. Chem. 1989 36 221. 90/1839. J. Sci. Food106R JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 Agric. 1989 49 307. 9011854. Arch. Oral Biol.1989 34 593. 9011871. Cancer Res. 1989 49 4210. 9011876. Ann. Chim. (Rome) 1989 79 31 1. 9011877. Sangyo lgaku 1989 31 22. 9011878. Aerztl. Lab. 1989 35(5) 102. 9011902. Trace Elem. Anal. Chem. Med. Biol. Proc. Int. Workshop 5th 1988 119. 9011905. Trace Elem. Med. 1989 6 47. 9011911. J. Am. Soe. Brew. Chem. 1989 47 68. 9011912. Huanjing Huaxue 1989 8 80. 9011913. Yingyang Xuebao 1989 11. 65. 9011922. Med. Lab. Sci. 1989 46 272. 9011923. Arch. T o x i d . Suppl. 1989 13 (Biol. Monk Exposure Response Subeell Level Toxic Subst.) 124. 9011924. Br. J . Nutr. 1989 62 451. 9011931. Gynecol. Obstet. Invest. 1989 27 148. 9011935. Sci. Aliments 1989,9 147. 9011944. Nippon Eisieguku Zasshi 1988 43 979. 9011945. Environ. Res. 1989 49 197. 9011949. Proc. ICMR Semin. 1988 8(Proc.Asia-Pac. Symp. Environ. Occup. Toxicol. 1987) 259. 9011959. Spectrochim. Acta Part B 1989 44 985. 9011978. J . Anal. At. Spectrom. 1989 4 571. 9011979. J . Anal. At. Spectrom. 1989 4 577. 9011980. J . Anal. At. Spectrom. 1989 4 581. 9011993. J. Anal. At. Spectrom 1989 4 657. 9012057. J. Anal. At. Spectrom. 1989 4,691. 9012064. J . Anal. At. Spectrom. 1989,4 727. 9012065. J . Anal. At. Spectrom. 1989 4 737. 9012070. J. Anal. At. Spectrom. 1989 4 767. 9012071. J. Anal. At. Spectrom. 1989 4 773. 9012081. Analyst 1989 114 1667. 9012082. Analyst 1989 114 167 1 . 9012087. Spectrochim. Acta Part B 1989 44 1209. 9012088. Spectrochim. Acta Part B 1989 44 1229. 9012089. Spectrochim. Acta Part B 1989 44 1237. 9012113. Analyst 1990 115 173. 9012115. J . Radioanal. Nucl. Chem.Articles 1989 131 331. 9012119. J . Radioanal. Nucl. Chem. 1989 137 265. 9012139. Nucl. Instrum. Methods Phys. Res. Sect. B 1989 43 556. 9012145. X-ray Spectrom. 1989 18 207. 9012147. Fresenius 2. Anal. Chem. 1989 335 885. 9012169. Fresnius 2. Anal. Chem. 1989 334 558. 9012183. Appl. Spectrosc. 1988 42 1250. 9012193. Analyst 1990 115 269. 9012194. Analyst 1990 115 275. 9012195. Analyst 1990 115 279. 90/2199. Spectrochirn. Acta Park B 1989 44 122 I . 9012237. Ann. 1st. Super. Sanita 1989 25 379. 9012238. Ann. 1st. Super. Sanita 1989 25 393. 9012239. Ann. 1st. Super. Sanita 1989 25 385. 9012240. Ann. 1st. Super. Sanita 1989 25 405. 9012241. Ann. 1st. Super. Sanita 1989 25,417. 9012242. Ann. 1st. Super. Sanita 1989 25 449. 9012243. Ann. 1st. Super. Sanita 1989 25 457.9012244. Ann. 1st. Super. Sanita 1989 25 481. 9012245. Ann. 1st. Super. Sanita 1989 25 487. 9012247. Ann. 1st. Super. Sanita 1989 25 565. 9012277. Ann. Clin. Biochem. 1990 27 25. 9012297. Anal. Chem. 1990 62 1 1 1. 9012300. Anal. Chem. 1990 62 386. 9012329. Biomed. EnLiron. Mass. Spectrom. 1989 18 543. 9012330. Biomed. Enliron. Mass Spectrom. 1989 18 598. 90/2353. Stand. Technol. 1989 94 215. 9012451. Fresenius 2. Anal. Chem. 1989 335 751. 9012568. Second. ion Mass Spectrom. Proc-. Int. Conf. 6th 1987 877. 9012572. Second. Ion Mass Specstrom. Proc. lnt. Conf. 6th 1987 921. 9012652. Second. ion Mass Spectrom. Proc. lnt. Conf. 6th 1987 865. 9012808. Radiar. Prot. Dosim. 1989 26 313. 9012820. Spec. Puh1.-R. Soc. Chem. 1989 72 (Nutr. Availability Chem.Biol. Aspects) 38. 9012821. Spec. Pub1.-R. Soc-. Chem. 1989 72 (Nutr. Availability Chem. Biol. Aspects) 35. 9012824. Swed. Dent. J. 1988 12 201. 9012831. Technol. Chim. 1988 8 138. 9012862. Spec. Puhl.-R. Soc. Chem. 1989 72(2) Nutr. Chem. Biol. Aspects 213. 9012863. Am. J . Clin. Nutr. 1989 49 870. 90/2913. J . Trace Elem. Electrolytes Health Dis. 1989 3 175. 9012914. J . Chromatogi.. 1989 482 87. 9013245. Miki-ochim. Acta 1989 2 185. 9013288. At. Spectrosc.. 1989 10(4) 106. 9013290. A f . Spectrosc. 1989 10(5) 141. 9013292. At. Spectrosc. 1989 10(5) 154. 9013293. At. Spectrosc. 1989 10(5) 160. 9013294. At. Spwrinw. 1989 10(6) 165. 9013295. At. Spec-tI-osc. 1989 10(6) 170. 9013296. At. Spectr-osc. 1989 10(6) 183. 9013300. Appl. Spec.tr-osc*. 1989 43 7 19. 9013326. Anal. Chim. Acta 1989 224 73. 9013327. Anal. Chim. Acta 1989 225 159. 9013329. Anal. Chim. Acta 1989 225 185. 9013333. Anal. Sci. 1989,5,53 1. 9013337. Anal. Sci. 1989 5 765. 9013367. Bunseki Kagaku 1989 38 529. 9013369. Bunseki Kagaku 1989 38 547. 9013371. Bunseki Kagaku 1989 38 734. 9013410. Fresenius Z. Anal. Chem. 1989 334 740. 9013422. Fresenius Z. Anal. Chem. 1989 335 914. 9013424. J . Anal. Toxicol. 1989 13 204. 9013425 J . Anal. Toxicol. 1990 14 8. 9013426. J . Anal. Toxicol. 1990 14 29. 9013427. J . Assoc. Of. Anal. Chem. 1989 72 286. 9013429. J. Assor. 08. Anal. Chem. 1989 72 926. 9013433. J. Radioanal. Nucl. Chem 1989 137 327. 9013435. J. Trace Elem. Electrolytes Health Dis. 1989 3 165. 9013451. Fenxi Huaxue 1989 17 456. 9013472. Fenxi Shiyanshi 1989 8(5) 33. 9013478. J. Anal. At. Spectrom. 1990 5 15. 9013479. J . Anal. At. Spectrom. 1990 5 21. 9013497. J . Anal. At. Spectrom. 1990 5 135. 9013500. J . Anal. At. Spectmm. 1990 5 75R. 9013507. Analyst 1990 115 549. 9013508. Analyst 1990 115. 553. 9013510. Analyst 1990 115 563. 9013512. Analyst 1990 115 57 1. 9013527. Yejin Fenxi 1989 9(3) 45. 9013528. Lihna Jianyan Huaxue Fence 1988 24 (4) 229. 9013529. Lihua Jianyan Huaxue Fence 1988 24(6) 322. 9013534. Anal. Biochem. 1989 182 245. 9013546. Sci. Total Environ. 1989 84 249. 9013549. Sci. Total Environ. 1989 89 237. 9013550. Sci. Total Environ. 1989 89 271. 9013551. Sci. Total Environ. 1989 89 279. 9013552. Sci. Total Environ. 1989 89 305. 9013553. Sci. Total Environ. 1989 89 31 1. 9013559. Anal. Proc. 1990 27 112. 90/3584. Talanta 1990 37 119. 9013950. Anal. Sci. 1990 6 83. 9013994. Fresenius 2. Anal. Chem. 1989 335 176. 9014002. Fresenius 2. Anal. Chem. 1989 335 743. 9014011. Fresenius 2. Anal. Chem. 1989 335 975. 9014030. Clin. Chem. (Winston-Salem N.C.) 1990 36 21 1. 9014032. CIin. Chem. 1990 36 247. 9014033. Clin. Chem. 1990 36 351. 90/4034. Clin. Chem. 1990 36 821. 9014131. J . Agric. Food Chem. 1988 36 537. 9014132. Bid. Trace Elem. Res. 1988 17 31. 9014133. Nutr. Res. (N.Y.) 1989 8 333. 9014145. J . Anal. At. Spectrom. 1990 5 189. 9014148. J . Anal. At. Spectrom. 1990 5 203. 9014152. J . Anal. At. Specrrom. 1990 5 225. 9014159. Analyst 1990 115 813. 9014161. Analyst 1990 115 823. 9014165. Bunseki Kagaku 1989 38 468. 9014166. Bunseki Kagaku 1989 38 553. 9112. J. Anal. At. Spectrom. 1990 5 293. 9113. J. Anal. At. Spectrom. 1990 5 301. 9116 J. Anal. At. Spetwom. 1990 5 321. 91112. Analyst 1990 115 943. 91136 Appl. Spectrosc. 1990 44 505. 91148. At. Spectrosc. 1990 11( l) 24. 91151. Spectroscopy (Eugene Oregon) 1989 4(7) 16. 91/65. Anal. Lett. 1989 22 2975. 91/66. Anal. Lett. 1989 22 3041. 91/71. Talanta 1990 37 545. 91172. Talanta 1990 37 555. 91/91. Microchem. J. 1990 41 164. 911105. Mikrochim. Ac-fa 1989,3 291. 911155. Analyst 1990 115 605. 911175. Fenxi Huaxue 1989 17 847. 911185. Fenxi Huaxue 1989 17 1045. 911188. Fenxi Huaxue 1990 18 84. 911206. X-Ray Spec-from. 1990 19 85. 91/220. Zh. Anal. Khim. 1990 45 757. 91/228. Fresenius J. Anal. Chem. 1990 336 226. 911229. Fresenius J . Anal. Chem. 1990 336 303. 911248. Anal. Chim. Acra 1990 231 265. 911250. Anal. Chim. Acta 1990 231 283. 911251. Anal. Chim. Acta 1990 231 321. 911256. Scanning Mic*rosc*. 1989 3,473. 911265. Sci. Total Environ. 1990 92 199. 911268. Sci. Total En\iron. 1990 94 169. 911284. Acta Chim. Hung. 1989 126 325. 911304. lnt. J . En\viron. Anal. Chem. 1990 38 565 911308. J . Environ. Sci. Health Part A 1989 24 943. 911309. J. Assoc-. Off. Anal. Chem. 1990 73 320. 911311. J . Assoc. Off. Anal. Chem. 1990 73 404. 911314. Scanning Mic*r-osc*. 1989 3 46 1. 911315. Scanning Mii*i.osc*. 1989 3 467. 91/317. Scanning Mi(-rosc.. 1989 3 861. 911318. Sc*annincq Micmw.. 1989 3 483. 911324. CIin. Chem. (Winston-Salem N.C.) 1990 36 450. 911325. Clin. Chem. (Winston-Salem N.C.). 1990 36 477. 911326. CIin. Chem. (Winston-Salem N.C.) 1990 36 658. 911327. Clin. Chim. Acta 1990 187 28 1. 911328. Fenxi Ceshi Tongbao 1989 8,JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY APRIL 1991 VOL. 6 107R 56. 911331. Fenxi Ceshi Tonghao 1989 8 57. 911349. Nucl. lnstrum. Methods Phys. Res. Sect. B 1989 40 630. 911352. Nucl. Instrum. Methods Phys. Res. Sect. B 1990 45 527. 911353. Nucl. lnstrum. Methods Phys. Res. Sect. B 1990 49 183. 911354. Nucl. Instrum. Methods Phys. Res. Sect. B 1990 49 191. 911366. Nucl. Instrum. Methods Phys. Res.. Sect. A 1989 282 499. 911386. Guangpuxue Yu Guangpu Fenxi 1989 9 60. 911396. Guangpuxue Yu Guangpu Fenxi 1989,9(3) 74.911405. Guangpuxue Yu Guangpu Fenxi 1989 9(4) 65. 911407. Guangpuxue Yu Guangpu Fenxi 1989,9(6) 54. 911422. Lihua Jianyan Huaxue Fence 1989 25 21. 911423. Lihua Jianyan Huaxue Fence 1989 25 33. 911438. Lihua Jianyan Huaxue Fence 1989 25 168. 911446. Lihua Jianyan Huaxue Fence 1989 25 27 1. 91/49 Lihua Jianyan Huaxue Fence 1989 25 294. 911450. Lihua Jianyan Huaxue Fence 1989 25 341. 911452. Lihua Jianyan Huaxue Fence 1989 25 361. 911474. Bull. Soc. Chim. Belg. 1989 98 513. 911824. Clin. Chem. 1990 36 860. 911825. Clin. Chem. 1990,36 1323.