ANALYST OCTOBER 1986 VOL. 111 1113 Directly Coupled Chromatography - Atomic Spectroscopy Part 1 Directly Coupled Gas Chromatography = Atomic Spectroscopy A Review Les Ebdon Steve Hill and Robert W. Ward* Department of Environmental Sciences Plymouth Polytechnic Drake Circus Plymouth PL4 8AA UK Summary of Contents 1. introduction 2. Coupled gas chromatography - microwave induced plasma 3. Coupled gas chromatography - inductively coupled plasma 4. Coupled gas chromatography - direct current plasma 5. Coupled gas chromatography - atomic absorption spectrometry 6. Coupled gas chromatography - atomic fluorescence spectrometry 7. Conclusion 8. References Keywords Review; coupled techniques; gas chromatography; atomic spectroscopy; trace metal speciation 1. Introduction The various techniques of analytical atomic spectrometry1 have been widely used in recent years to obtain total element information particularly on trace metal composition.Where-as these techniques are both selective and sensitive offering detection limits in the ng ml-1 range they yield by themselves only information on total concentrations. Currently there is a much increased demand to quantitatively determine the form of trace metals in a wide variety of samples; this is often termed trace metal speciation. Such information may be of vital importance to the toxicologist to indicate likely sources and transport mechanisms of elements in the environment to the clinician and to indicate the history of a sample. Several approaches to trace metal speciation have been suggested,2-4 including electroanalytical techniques,5,6 but one of the more promising approaches is to couple the separatory power of chromatography with the selectivity and sensitivity of atomic spectroscopy.This area until now has only been partially reviewed’-12 and the purpose of this review is to critically appraise coupled gas chromatography (GC) - atomic spectro-scopic approaches. Part 2 will consider coupled liquid chro-matographic procedures. Atomic spectroscopy offer the possibility of selectively detecting a wide range of metals and non-metals. The use of detectors responsive only to selected elements in a multi-component mixture drastically reduces the constraints placed on the chromatography step as only those components in the mixture which contain the element of interest will be detected.Certain requirements for element-specific detectors may be identified. Atomic absorption spectrometry (AAS) is inher-ently the most selective of the atomic spectroscopic techniques due to the “lock and key” mechanism.1 The various plasma emission sources microwave induced plasma (MIP) direct current plasma (DCP) and inductively coupled plasma (ICP), owing to their very high excitation temperatures produce a wealth of emission lines. Thus although not possessing the inherent selectivity of AAS the use of a suitable high-resolution monochromator reduces the possibility of spectral * Present address Plasma-Therm Inc. Route 73 Kresson Industrial Park Kresson NJ 08053 USA. interference and enables inter-element selectivity.This wealth of emission lines produced also permits multi-element detec-tion which normal line source AAS does not offer. Atomic fluorescence spectrometry (AFS) using line source excita-tion in theory offers similar selectivity to AAS coupled with the multi-element capacity of AES but in practice is limited by the availability of suitable line sources. The second prerequisite is that of sensitive detection for a wide range of elements. The most popular methods of generating atoms for AAS are in flames and by electrother-mally heated furnaces. The former usually gives poorer detectability owing to the shorter atomic residence times in the flame and problems of sample introduction. Relatively short useful linear ranges of 1-2 orders of magnitude are typical of absorption techniques.The plasma emission techniques use high temperature excitation sources thus enabling low levels of detection metals and the favourable source geometry often provides long linear working ranges. In atomic fluorescence, provided a suitably intense line source is available low level detection and long linear ranges are available. Long linear working ranges are frequently cited as desirable characteris-tics of chromatographic detectors but the levels of analyte in real samples are often close to the detection limit and sensitivity may rightly be seen as the major problem. Flames or plasmas whether chemical in FAAS and FAFS, or electrical in the ICP DCP and MIP consist of flowing gas streams and are therefore well suited to accept a gaseous analyte.Their continuous mode of operation is also advan-tageous because although the analyte peak is transient it is introduced into a flowing gas stream. Electrothermal atomisers are typically not continuous in their mode of operation and are designed for use with discrete condensed phase samples and thus require modifications before they can accept a flowing gas stream. The relative merits of the various couplings are perhaps best considered together with a review of their various applica-tions. As there are well over 100 references reported much of the information in this review is classified according to the type of detection system used and is presented in tabular form. This review is confined to the area of directly coupled gas chromatography - analytical atomic spectrometry and hence mass spectrometric detection has not been included nor ha 1114 ANALYST OCTOBER 1986 VOL.111 the related methodology of trapping compounds on chromato-graphic material for later thermal release and atomic spectro-scopic detection of the evolved gases.ls-16 2. Coupled Gas Chromatography - Microwave Induced Plasma The microwave induced plasma (MIP) has two basic charac-teristics that can be utilised when coupling to a GC instru-ment. The low gas temperature of the MIP allows small amounts of sample compatible with those of GC solutes to be introduced without extinguishing the plasma. In addition, sample introduction is easily facilitated as the carrier and plasma gases are the same. These advantages have made coupled GC - MIP a popular technique and many applications have been reported (Table 1).The first use of the MIP as an element-selective detector for organic compounds was reported by McCormack et al. in 1965.17 The effluent from a GC was connected directly to the silica tube containing the plasma discharge. Both the more sensitive tapered cavity and the coaxial cavity intended for larger samples were used. Two plasma types were utilised, low-pressure helium and atmospheric argon the latter being favoured owing to the complexity of the associated vacuum systems required when using low-pressure helium plasmas. Bache and Lisk later used an atmospheric argon plasma to determine pesticides in various samples by the selective detection of phosphorus18 and iodine.l9 Using a low-pressure argon plasma the same authors later lowered the detection limit by a further order of magnitude.20 The more energetic reduced pressure helium plasma has been used for the determination of halogens phosphorus and sulphur using atomic lines.21-23 Moye22 found that if a tapered rectangular cavity with a mixed argon - helium carrier was used a lower background emission for chlorine iodine and phosphorus detection in pesticide residues could be obtained. Dagnall et al.24925 used a quarter-wave radial cavity with low-pressure argon or helium plasmas for the determination of sulphur in various compounds. It was found that the most sensitive and specific emission wavelength was not the same for all the compounds examined.In addition thioglycolic acid was found to be very difficult to fragment,24 although a platinum wire in the base of the detector was found to catalyse the fragmentation process.25 Bache and Lisk29 were the first to use the low-pressure helium plasma for the detection of organomercury compounds after extracting the compounds from salmon using the established procedures of We~tOO.27~28 A potential use of the MIP detector for obtaining inter-element ratios has been reported by Dagnall et al. ,31 using two monochromators one set at a carbon line and the other set to monitor a he teroat om. 0 t her workers37,54,59791-93,95 have also used the MIP detector to determine inter-element ratios in an attempt to establish empirical formulae. The commercially available MPD 850 (Applied Chromatography Systems) low-pressure helium plasma system has also been used in this r01e.53,5~791,92 However Dingjan and De Jong76 found that it was necessary to use a reference compound if accurate ratio formulae were to be obtained.If unequivocal inter-element ratios could be determined independent of the sample type, GC - MIP systems would be capable of much identification work currently performed on more expensive GC - mass spectrometers. An oscillating slit mechanism for the deter-mination of hydrogen isotope ratios has been used by Schwarz et al. ,58 but the poor signal to noise ratios obtained gave poor precisions. More recent publications85791-94 have suggested that the use of capillary gas chromatographic columns, computerised data acquisition and peak-area measurements may improve the precision and accuracy attained.The passage of an organic compound through a plasma may result in the formation of carbon deposits on the walls of the quartz capillary absorbing part of the radiation and increasing the background emission.17 This can be prevented by either initiating the plasma after the solvent has passed through the detector,l8 or by adding traces of a scavenging gas. This gas may be either nitrogen,43 0xygen,3~-39 hydrogen39 or air17 added to the plasma gas; however as a result of this the spectral background is considerably increased. The MIP has proved to be popular as a detector for various metal chelates732J6J8>46 and also as a detector for various Hydride-forming elernents.45.52fj3@ Talmi and Bostick45 have determined alkylarsenical acid compounds in pesticides by generating their hydrides prior to GC - MIP analysis.The separation and sequential detection of As Ge Se Sn and Sb hydrides has also been demonstrated using a mixed argon -helium pla~ma.55~63~64. Little difference in detection levels have been found using various forms of microwave plasma by Mulligan et al.,(jl although the Beenakker TMolo cavity was found to be the easiest to operate. This method was used to determine the above hydride-forming elements in whole blood and enriched flour63 and in NBS orchard leaves.63.64 Coupled GC - MIP has also been used for the detection of various metals in volatile organometallic compounds. Lead has been determined as the tetraalkyl species15356 in petrol80JQ and in the atmosphere,74 and as trialkyllead chloride in water samples.77 Mercury as the diphenyl,56 dimethyl and diethyl derivatives74 has been detected using the TMOl0 cavity.Quimby et al.56 used the same cavity to determine manganese as the methylcyclopentadienyltricarbonyl derivative in petrol and as a silicon-specific detector for tetravinylsilane. The coupling of capillary columns with the TMolo cavity has also been demonstrated with great success for metal-specific detection of volatile organometallics.66~77~80~9~~92 This cavity is increasingly being seen as the optimum for GC - MIP studies as it is capable of operating a He plasma at atmospheric pressure. In a study of the pyrolysis of carborane silicone p0lymers,~8 the group at Amherst found that doping the plasma gas with hydrogen inhibited oxide or silicate formation by promoting borohydride formation which increased the populations of atomic boron rather than the ionic states.Hanie et al.70 have also used capillary columns for the determination of halides in pesticides using a helium plasma and a Surfatron cavity.@ Recent developments of coupled GC - MIP systems have largely been based on the development of software for both system control and data handling. One such system described by Eckhoff et al.83 uses a polychromator - microcomputer system to monitor simultaneously four atomic emission wavelengths throughout an entire chromatographic run. The same system has also been used by Hass and Carus094 as an element-specific detector for the gas chromatography of halogenated compounds.Delaney and Warren85 have used a minicomputer to- modify the interface described by Estes et al. ,77 so that in addition to controlling the switching valves it also controls the monochromator wavelength setting and acquires the analytical data that the MPD and FID monitor. The above and other work in coupled GC - MIP systems are summarised in Table 1. 3. Coupled Gas Chromatography - Inductively Coupled Plasma The high capital cost of inductively coupled plasma (ICP) instrumentation together with the high running costs have resulted in its use mainly as a multi-element exictation source for routine analysis. Consequently use of the ICP as a detector for GC has been limited.However it does offer the advantage of withstanding organic solvents more readily than the MIP owing to the higher gas temperature and so may possibly be further utilised in this role in the future. To obtain a sufficient flow-rate to puncture the fire-ball and produce an annular plasma may mean augmenting the eluent flow from the GC column with an auxiliary Ar flow. This may reduce the sensitivity by dilution but failure to form an annular plasma will be more deleterious ANALYST OCTOBER 1986 VOL. 111 1115 Table 1. Coupled gas chromatography - microwave induced plasma optical emission spectrometry Elements and wavelength/ nm Reference c 17 388.3 F, 516.6 251.6 a , 278.8 Br, 298.5 Detector Tapered and co-axial cavities used the former more sensitive, the latter accepted larger samples.10 mm i.d. discharge tube at low pressure Chromatography Sample Comments Solutions of simple and heteroatom- the preferred carrier containing organic gas. At atmospheric compounds pressure Ar was used At low pressure He was as it gave a stable discharge. Dynamic range four orders of magnitude. Detection limits 2 x 10-16-2 x 10-7 s-l 206.2 s, 257.5 Atmospheric pressure Ar 2 ft glass U column Organophosphorus 1 mm i.d. quartz 5mmi.d.,5% SE30 insecticide residues discharge tube in a tapered cavity on 80-100 Chromosorb W. in pure form agricultural Ar = 20-115 ml min- l , T = 160-200 "C and food samples Diazinon Dimethoate, Ethion Parathion and Ronnel determined. Detection limit: 1.4-9.2 pg s-1 p, 253.565 18 Detection of Ionynil and metabolites.Recoveries from 66-108°/0 achieved. Detection limit: 4 x 10-"'gs-'of12 I, 206.2 I2 band 19 As in ref. 18 See ref. 18 Iodinated herbicide residues and metabolites in wheat oats and soil As in ref. 18 except reduced pressure Ar plasma See ref. 18 p, 253.565 20 Diazinon in grapes; see ref. 18 See ref. 18. Achieved increased sensitivity with low pressure discharge. Detection limit 6 x lo-13gs-* of P Detection limits: 9 x 10-12 6 x 10-"gs-1 Br, 478.55 c1, 479.45 I, 533.82 p, 253.57 s, 545.38 21 Reduced pressure helium plasma using tapered cavity. 5-10 mm Hg pressure 6 ft glass column 10% DC-200 on 80-100-mesh pesticides Gas-Chrom Q.Isothermal set at various T = 130-210°C Organic compounds and Ar - He (15 + 85) mixed plasma tapered cavity as longer lifetimes and less background emission obtained 4 ft x 4 in i.d. glass column 5% SE-30 on Gas-Chrom Q. Flow-rate = 27 ml min-l, T = 180"C, TI = 215 "C Pesticide residues of various P- C1- and I-containing compounds Detection limits: 0.07-1 1.5 ng p, 253.57 a , 221.00 I, 206.20 22 Monitored atomic S and C1 lines a , 479 * 45 s, 545.38 Reduced pressure helium plasma 6 ft x 4 in i.d. glass column 10% DC-200 on 100- 120-mesh Gas-Chrom Q Phenol-substituted insecticides in agricultural samples 23 24 Thioglycolic acid difficult to fragment. Detection limit 0.2 ng for CS2 at C=S band head t h radial line cavity Ar or He low-pressure (13-40 mbar) plasma 2.7 m x 6.5 mm i.d.Cu tubing packed with dinonyl phthalate. 1 .O pl injections S compounds CS2, thiophene, thioglycolic acid, DMSO and SO2 s, S 190.0 S 191.5 C=S 257.6 common to all compounds C2,516 See ref. 24 6 x 0.6 mm i.d. Cu tubing packed with either Porapack P or Q S compounds CS2, thiophene dimethyl sulphide and thioglycolic acid Used Pt wire in base of detector to catalyse fragmentation process. Detection limits low ng range S c, Monitored C=S band head at 257.6 and atomic C line at 247.9 2 1116 ANALYST OCTOBER 1986 VOL. 111 Table l-continued Detector Chromatography Sample Low-pressure (5-10 Torr) See ref.20 S - halogen- and P-He plasma. See ref. 20 containing pesticide residues in a wide range of food products Reduced pressure He plasma in a tapered cavity; cf. ref. 21 Atmospheric pressure Ar plasma, 20 cm x 2 mm i d . quartz tube surrounded by 2 h cavity See ref. 30 2 ft X 5/32 in i.d. glass column 60-80-mesh Chromosorb 101. He = 80 cm3 min - 1, T = lOO"C, 6 ft X '/32 in i.d. glass column 20%,0V-17 and OV-1 (1 + 1 m/m) on 80-100-mesh Gas-Chrom Q. T = 152"C, 30 and 70 cm x 6 mm i.d. packed with Porapak S TI = 140 "C. TI = 208 "C 0.7 m x Q in stainless steel Chromosorb 101 Ar plasma. Essentially 10 f t x in i.d. the same as in ref. 17 stainless-steel column 20% Carbowax 20M on Chromosorb P, 60-80 mesh. Ar = 48 cm3 min-1, T = 75 "C Low-pressure He plasma using MPD 850 system, O2 and N2 used as scavengers to prevent C build-up Methylmercury dicyan-diamide phenyl-mercury(I1) acetate, methylmercury dithizonate MeHgCl in salmon Range of C- 0- N- and halogen-containing compounds Range of C- S- and halogen-containing compounds Various organic compounds Comments Westoo extraction procedure27.28 for MeHgCl in salmon.Linear range: 0.1-100 ng for MeHgCl Several cavities examined 2 h preferred because it produced a long (ca. 8 cm) stable discharge with little local overheating. Detection limits: 10-20 pg s-' Use of two monochromators one set to atomic C line the other set to the hetero-atom line. By monitoring emission from both inter-element ratios were obtained.Detection limits: 0.04-4.5 ng s-1 Found selectivity for Hg over various organic compounds, always >lO3. Detection limit 0.3 ng Detection limits: 0.03-3.0 ng s-1 Element and wavelength/ nm Reference Br 26 478.55 c1, 479 * 45 I, 533.82 p , 253.57 s, 545.38 253.7 Hg 7 29 c, Monitored atomic C line at 247.9 c2 band head at 516.5 and band head at C2/CN 385-389 30 c 31 247.9 1, 206.2 s, 182.0 p, 253.5 (21, 259 band Br, 292 band Hg 32 253.7 c , 247.8 H, 486.1 D, 656.2 0, 777.2 N, 746.9 F, 685.6 c1, 479.4 Br, 470.5 I, 516.1 s, 545.4 3 ANALYST OCTOBER 1986 VOL. 111 1117 Table l-continued Detector Chromatography Sample Similar system to ref.30 2 columns both except t h Evenson 0.6 m x 4.8 mm i.d., cavity 70 W forward power 1. Universal B coated with Ar plasma ignited after 10% Apiezon L. elution of solvent 2.0.5% Apiezon Lon glass beads (0.2 mm diameter). Both conditioned for 36 h at 200 "C Acac and tfa chelates of Al Cr Cu Ga Fe, Sc and V f h Evenson cavity used reduced pressure (10 Torr). He plasma generated in a 6cm x 8mmi.d. quartz tube Ar plasma generated in a quartz capillary, 1.6 mm i d . X 25 cm, placed in a tapered rectangular type cavity Reduced pressure He plasma 0.1-1 YO. O2 or N2 added as scavenger l o r 2 c m x l m m i.d. Cu tubing packed , with either Poropak Q or 5A molecular sieve. T = 125 "C, .- 5O-pl injections Stainless-steel tubing, 72cm x 4mmi.d., 0.5% SE-300n glass beads 60-80 mesh.Ar = 150cm3min-1, T = 160 "C, TI = 200-210 "C Comments MIP responded both non-specifically to C or specifically to the metal of interest. Detection limits: 2 x 10-12-2 x 10-11 g s-1 Element and wavelength/ nm Reference A1 34 396.2 0 , 357.9 c u , 324.7 Ga, 294.4 Fe 7 344.1 s c , 361.4 v, 318.4 CO C02 SO2 and N2 in Gas mixtures were c, known amounts of pure N, Detection limits s, air prepared by injecting 247.9 gas into an air-filled flask fitted with a septum. 20-50 p.p.m. 190.0 337.1 nm (N2 band head) Metal acac chelates A CN band was observed Al, chloroform probably due to N2 Be , Failed to chromatograph Cr, dissolved in for all complexes 396.2 impurity in the Ar.234.9 acac chelates of CuII 425.4 FeI" and VIv. Two orders of magnitude for Be and Cr. One order for Al. Detection limits: 0.01-100 ng 3 m X 2.5 mm i d . 10% Apiezon L on 60-80-mesh solutions element detection used to DCMS-treated calculate empirical formula Chromosorb W. Effluent split 1 1 to FID and MIP of magnitude for F. Wide range of organic Multi-non-metallic of organic compounds. Linear range 4 orders Detection limits: 0.03-3.0 ng s-1 c , 247.8 H, 486.1 D, 656.2 F, 685.6 c1, 479.4 Br, 470.5 I, 516.1 s, 545.4 N, 746.9 0, 777.2 Reduced pressure He plasma doped with 1 % 02 4 h Evenson cavity U tube columns Chelates Cr(tfa), Use of MPD as a specific Cr , packed with Chromosorb Cr(acac), Cr(hfa) detector for Cr and as a 357.87 W-HP with 3% OV-101 non-specific detector by loading monitoring the atomic C line.Detection limit 1.5 X 10-11-8.0 x 10-10gs-lofCr 35 36 37 3 1118 ANALYST OCTOBER 1986 VOL. 111 Table l-continued Detector Chromatogaphy Sample Comments 4 h Evenson cavity DB-5 coated fused-silica Organomercury Interface consists of a maintained at 5 Torr. capillary column -selenium and -arsenic DB-5 coated fused-silica 0.75-m Roland Circle (30 m long 0.25 pm film compounds capillary column passed direct reader with thickness) througha 1.6mmi.d. 12 outputs nickel tube connected to the top of the plasma head Tapered cavity system essentially the same as ref. 17.Ar plasma, 35 W forward power Ar plasma (see ref. 4), atmospheric pressure Reduced pressure 1-10 Torr He plasma Atmospheric pressure Ar plasma 30 W forward power see ref. 40 See ref. 40 4 h Evenson cavity. Atmospheric pressure. Ar plasma 70 W forward power 4 ft x 0.5 mm i d . glass column packed with 4% SE-30 on 30-60-mesh Chromosorb G-HP 3 ft column 4% FFAP on 80-100-mesh Gas-Chrom Q. Ar = 90cm'min-1, T = 150"C, TI = 200 "C 3 ft column 1% FFAP on 80-100-mesh carbon beads. Ar = 95 cm3 min- 1, T = 135 "C, TI = 200 "C 2 ft column Chromosorb 101 He = 80cm3min-1, T = 115 "C, 3 ft column 4% FFAP on 80-100-mesh Gas-Chrom Q. Ar = 110-120 cm3 min- 1 , T = 220-240 "C, TI = 135 "C TI = 245-260 "C 6 f t column 5% Carbowax 20M on 80-100-mesh Chromosorb 101.Ar = 100cm3min-1, T = 175 "C, TI = 180 "C 0.9 m PTFE column 3 mm i.d. 10% SE-30 on 70-80-mesh Gas-Chrom Z . Ar = 30-150 ml min-1, T = 180-190°C, TI = 200 "C Se compounds in environmental samples, looked at various NBS materials with good agreement MeHgX in benzene extracts of biological samples and in air CH3HgX in water and air (CH3),Hg in water and air As and Sb in environmental samples Alkylarsenic acids in pesticide and environmental samples MMA and DMA Human blood serum SeIV complexed with Pd to form the volatile piaselenol complex followed by toluene extraction. Detection limit 40 pg of Se 0.1 pg 1-1 for water samples and 15 p.p.b. for solid samples or OH as all eluted simultaneously; see refs.41 and 42 for explanation. Detection limit: 0.5 pg g-1-1 ng g-l X designates C1 Br I Hg 9 253.7 ~ Element and wavelength/ nm Reference H 39 486.1 c, 247.8 N, 746.9 0, 777.2 F, 685.6 p, 253.6 s, 545.4 (21, 479.5 Br, 470.5 As 9 200.3 Se , 204.0 Hg, 365.0 Se , 204 40 As111 and SbIII As, converted into Ph3AsH 228.8 and Ph3 SbH extracted Sb, into diethyl ether 259.8 separated by GC. Detection limits: 20 pg of As 50 pg of Sb As compounds converted As, into hydrides. Detailed 228.8 study of hydride generation and trappings of the evolved arsines. Linear range 0.01-20 p.p.m Detection limit 20 pg as As in water samples Low-temperature ashing Cr , followed by chelation 357.9 with H(tfa) to form Cr(tfa)3 which is extracted into benzene.Linear range 1-10 pg of Cr. Detection limit: 9 x 1043g 41 42 43 44 45 4 ANALYST OCTOBER 1986 VOL. 111 1119 Table l-continued Detector Low-pressure (150 mbar) He plasma cavity type 214L. Inter-element selectivity improved by use of wavelength modulation Atmospheric pressure Ar plasma in quartz capillary, 25 cm X 1.6 mm i.d. Tapered rectangular cavity 50 W forward power Chromatography Sample Element and wavelength/ Comments nm Reference Organic compounds Demonstrated that at Hg 7 47 Hg( Me)CI low pressures 253.65 fragmentation occurs via collisions with atomic He, whereas at high pressures the collisions are with He2. Linear range: 0.02-0.5 ng.Detection limit 5 x lO-*4g Column 45 cm x 3 mm i.d. Trace levels of Cu and glass 0.5% SE-30 on Al in Zn metal 60-80-mesh glass beads. Ar = 80 cm3 min - 1, T = 14OoC, TI = 180°C Cu and A1 extracted as tfa chelates in CC14. Linear up to 60 ng of Cu, 100 ng of Al. Detection limits 0.5 ng of Al, 1 ng of Cu c u 9 324.8 A1 , 396.2 See refs40,43. Ar plasma 3 ft x 5 mm i.d. glass MeHgCl in water samples MeHgCl extracted as Hg 3 5-10 Torr pressure 18 W 6% FFAP on 80-100-mesh quaternary amine adducts. 253.7 forward power Gas-Chrom Q. Detection limit: Ar = 130-150~m~min-~, T = 180-190"C samples 1-2.5 ng I-' for water TI = 200 "C Atmospheric pressure Used exponential diluter to Gas mixtures He plasma using TMolo cavity demonstrate the applicability of MIP for GC detection Demonstrates advantages of atmospheric pressure He plasma and discusses excitation mechanism.Three to four orders of magnitude linear ranges. Detection limits: 2 x 10-11-2 x lo-' moll-' Low-pressure (3-5 Torr) Glass column Mixture of n-alkanes and Dual FID - MPD Ar plasma; see refs 40 and 6 ft x 3.5 mm 4% OV-101 43 on Chromosorb G-HP carboxylic acids specificity of response to 80-100 mesh. TMS derivatives. Linear Ar = 80 ml min-1 TMS derivatives of (5 1 split) to demonstrate range 0.5-150 ng Low-pressure (90 Torr) He Constant sample Various organic plasma observation introduction for compounds 9 mm downstream from centre of discharge, 75 W forward power. 0.25% VIVO as scavenger or 0.4% VIV NZ.h Evenson cavity Model 214L and 4 h coaxial cavity Model 217L optimisation studies Optimised plasma conditions for gas flow-rates observation position microwave power and gas pressure with the 217L cavity up to 10% of power reflected with 217L only 1 YO reflected. Three to four decades except for H where a non-linear response is found. Detection limits: 0.01-0.5 ng s-l c , 193.1 247.9 H, 486.1 c1, 479.5 481 .O Br, 470.5 478.5 I, 516.1 206.2 s, 545.4 Si, 251.6 Br, 470.47 c , 247.86 c1, 479.45 F, 685.6 H, 486.13 I, 516.12 N, 746.88 0, 777.19 s, 545.39 48 49 50 51 5 1120 ANALYST OCTOBER 1986 VOL. 111 Table l-continued Detector See ref. 33. Reduced pressure He plasma Chromatography Sample See ref.33. Reduced pressure He plasma None given Mixed Ar - He plasma, 110 W forward power 0 W reflected Atmospheric He plasma, TMolo cavity 75-80 W, forward power axial viewing Various organic compounds Trace S in MeOH, yellow P in PC13 specific detection of vinylidene and PCBs Polypenco Nylaflow Hydrides generated from pressure tubing 4.7 mm i d . 1 3 and 6 ft lengths. Packed with Chromosorb 102 60-80 mesh from solutions of As, Ge Sb Se and Sn 3 ft X 4 in i.d. 5% OV-17 Diphenylmercury on 100-120-mesh Chromosorb 750. He = 70 cm3 min-1. 3 ft X & in i.d. 3% QF-1 on 100-120 mesh Varaport 30. He = 50 cm3 min-1. 6 ft X & in i d . 6% Carbowax 20M on 100-120-mesh Chromosorb P.He = 50 cm3 min-1. 6 ft x 4 in i d . 2.5% Dexsil300 on 100-120-mesh Chromosorb TEL 2,5-dimethyl-750. He = 50 (311113 min-1 TBP Tetravinylsilane MMT thiophene. Halobenzenes Comments Signals for four elements monitored simultaneously, added by a SYNC signal and stored for later computer analysis, resulting in inter-element ratios; main concern is in data acquisition and processing Using MPD 850 to obtain accurate empirical formulae obtained detection limits comparable to manufacturers’ claims Hydride trapped in liquid N2 then chromatographed. Elements determined sequentially. Linear over 2 orders of magnitude A design for heating the interface between GC and plasma utilising nichrome resistance wire coupled to a variac given.Detection limits 0.49-63 pg S-1 Element and wavelength/ nm Reference c 53 247.8 H, 486.1 D, 656.2 0, 777.2 N, 746.9 F, 685.6 (21, 479.4 Br, 470.5 I, 516.1 s, 545.4 c 54 247.8 H, 486.1 D, 656.2 0, 777.2 F, 685.6 c1, 479.4 Br, 470.5 I, 516.1 p, 253.6 s, 545.4 Ge 55 303.9 As, 193.7 Se 3 196.0 Sn 7 317.5 Sb , 259.8 Hg 7 56 253.7 p, 253.6 Si , 251.6 Mn , 257.6 Pb > 283.3 s, 545.4 c1, 481 .O Br, 470.5 F, 685.6 I, 206. ANALYST OCTOBER 1986 VOL. 111 1121 Table l-contznued Detector Chromatography Sample Element and wavelength/ Comments nm References 3 h cylindrical cavity 1.8 m x 3.1 mm i.d. 3% Tetraalkyllead compounds Samples cold trapped Pb 15 125 W forward power OV-1 on 80-100-mesh in the atmosphere on SE-50 on Chromosorb 405.78 Ar plasma Chromosorb W.P at -80 "C. Removed background correction Ar = 22 cm3 min-1 by freeze-drying and by wavelength modulation T = 80 "C. concentrated in Ti = 130 "C organic solvent. Detection limit 6-40 pg Low-pressure (5 Torr) He and Ar plasmas. Tapered rectangular cavity, 100 W forward power. 0.3% O2 added to plasma gas See ref. 15 Reduced pressure He plasma. See ref. 33 See ref. 33 Beenakker (3 A), Evenson (4 h) and Broida (3 A) cavities were compared with He - Ar or Ar plasmas 100 W forward power Beenakker TMolo cavity viewed axially He plasma Mixed Ar (400 ml min-1) and He (300 cm3 min-1) plasma power.110 W for forward Evenson a h cavity See ref. 63 Stainless steel H in organic compounds He plasma twice as 3 m X 3 mrn i.d. 3% mlm Dexsil300 on 80-100-mesh Chromosorb W AW. 6 m X 3 mm i.d. Squalane on 80-100-mesh Detection limit: Chromosorb W AW sensitive as Ar plasma due to higher energy and therefore more complete fragmentation. 10-11 g s-1 See ref. 15 H isotope ratios in OSM measures organic compounds in alternately 1H and water samples 2H emissions of hydrocarbons. Major disadvantage is high signal to noise ratios PCBs in seal blubber, cleaning fluids in water Applications of MPD 850 in analysis and also empirical formula determinations. Detection limits: 50 pg s-1 range Biological tissues, coal tars pesticides Brief resum6 of the possible uses of the MPD 850 system 2.5 in x 4.7 mm i.d.Standard solutions Semi-automated hydride packed with Chromosorb 102. Served only to reduce rate of sample throughput to give stable plasma generation from stock solution containing As, Ge Sb and Sn. Beenakker cavity proved easiest to operate. Detection limit 1 p,p.b. at 3 o level for all cavities 10 ft x Q in i d . , stainless-steel Tenax GC water HECD. Found that MIP Haloforms in drinking Compared MIP with was preferable as it gave a uniform molar response and selective detection. Detection limit 1 p.p.b. H, 656.28 'H, 656.28 2H, 656.10 57 58 c H p N 59 F C1 Br, I p Se, As Hg Pb C1 Br I, s p Hg As , 234.984 Ge, 303.906 Sb , 259.806 Sn, 3 17.502 a , 481.0 Br, 470.5 I, 206.2 3 ft X 4.7 mm i.d.Whole blood enriched Hydrides trapped on As , tubing packed with leaves (SRM 1571) condensation tube packed Ge , Polypenco Nylaflow flour NBS orchard liquid N2-cooled 193.7 Chromosorb 102,60-80 with glass helices prior to 303.9 separation on GC column. Se , Detection limits 196.0 3-40 ng Sb , 259.8 Sn , 317.5 mesh. T = 23 k 3°C 60 61 62 63 See ref. 63 NBS orchard leaves; Elements except Ge hydride generation determined both sequentially and simultaneously the former giving lower detection limits. Detection limits: 2C600 ng Simultaneous 64 As , 235.0 Se , 196.0 Sb , 259.8 (2nd order) Sn 317.5 (2nd order) (for sequential, see ref.63 1122 ANALYST OCTOBER 1986 VOL. 111 Table l-continued Element and wavelength1 Reference Detector Chromatography Sample Comments nm See ref. 57 See ref. 57 He plasma TMolo cavity viewed axially 12.5 m X 0.2 mm i d . fused-silica WCOT SP 2100 capillary column, TI = 80-116 "C at 4 "C min-1 to 170 "C; 0.1-p1 injections. Column passed to within 5 mm of plasma He plasma TMolo cavity viewed axially. OV-225 SCOT column. He = 450 (31x13 min-1 100 m X 0.25 mm i.d. He = 4 cm3 min-1, T = 40 "C then 4 "C min-1, TI = 210 "C, Tin = 250 "C H emission from Characterisation of H 65 organic compounds emission from atomic H 656.28 in MIP accounts for non-linearity observed Toluene solutions of The low volume of GC volatile organometailic column (cu.80 p1) compounds [CpV(CO),] is ideally compatible with MMT,[Cp2Fe] [Cp2Ni] MIP. Specificity of detec-[CpCo(NO)(CO),] and tion aids identification of [(CH,),CpCo(CO),] the unresolved [Cp,Ni] and [CpCr(NO) (CO),] complexes Friedel - Crafts 35 redistribution catalysed alkyl group products are formed. redistribution Owing to the requirement reaction of to vent the solvent the methylethyl-n-propyl- low MW products that n-butylsilane elute with the solvent are not recorded He atmospheric plasma, using TMolo cavity 2% OV-101 on compounds 85-90 W forward power Glass 1.5 m x 4 mm i.d., 80-100-mesh Chromosorb He = 60 cm3 min-1, T = 238 "C PBB and related W-HP. He plasma in a 30 m capillary column Pesticides Surfatron cavity coated with OV-101 (see ref.69) methylsilicone. He = 5.9 ml min-1, T = 250 "C for pesticides TI = 275 "C, See refs 56 and 62 See refs 56 and 62 See ref. 72 See ref. 72 TMolo cavity He plasma 80 W forward power o2 as scavenger. He = 40-70 cm3 min-1 15.2 m x 0.508 mm i*d. SCOT column packed with finely ground diatomaceous earth on silica support coated with rn-bis(m-phenoxyphenoxy)-benzene and Apiezon L. He = 0.5-8 cm3 min-1, T = 90 "C c , 247.9 Cr, 267.7 c o , 240.7 Ni , 231.6 Mn , 257.6 Si , 251.6 Br Not as sensitive as the ECD but offers element 478.55 selectivity. Detection limit: 1 ng The Surfatron He plasma gives slightly higher detection limits than those obtained with other cavities.Detection limits: 0.5-20 ng Aqueous chlorination In addition to products of humic and fulvic substances significant number of tnhalomethanes a chlorinated phenolic compounds were found c, 247.8 c1, 479.5 481 .O Br, 470.5 I, 206.2 c1, 479.5 Selenium biomethylation (CH3)2Se (CH3)2Se2 Se , products from soil and and (CH3),Se02 found. 196.0 sewage Detection limit 20 pg for (CH3)2Se Hydrocarbons FID proved 50 times more c , Hg 9 (CH3)2Hg and (C2HJ2Hg sensitive than MIP for C 193.1 and (at 193.1 nm). Both had 247.9 (C2H5)Hg. The MIP was 254.3 twice as sensitive as the FID for (CH3)2Hg using Hg-specific detection. Detection limits: 3.8 X 10-12 and 9.1 X 10-12gs-1 the same sensitivity for 66 67 68 70 71 73 7 ANALYST OCTOBER 1986 VOL.111 1123 Table l-continued Detector Chromatography Sample Comments TMolo cavity He plasma f h Evenson low-pressure (40 Torr). TMolo atmospheric pressure. Ar and He plasmas the latter viewed axially Organic compounds, elemental analysis Linear ranges of 3 orders of magnitude for all elements. Detection limits: 2 x 10-11-8 x 10-10 mol 1-1 C4-C7 n-hydrocarbons Atmospheric pressure SP21OO WCOT fused-silica Trialkyllead chlorides He plasma in a TMolo cavity. Background i.d. and OV-101 SCOT samples correction by quartz glass column refractor plate column 12.5 m X 200 ym in spiked tap water 30 m x 350 ym i.d. TMolo cavity. Atmospheric pressure SP2100 fused-silica B compounds from the He plasma viewed axially WCOT capillary column.pyrolysis of Dexsil T = 60-104 "C at 4 "C min-1 0.1 pl injections 100 1 split. Used for boration studies TMolo cavity. Glass 3 m x 3 mm i.d., Atmospheric pressure columns packed with compounds He plasma 75 W forward either 3% OV-17 on power;He = 80cm'min- 80-100-mesh Shimarate W, 10% Carbowax 6000 on 30-60-mesh Shimarate TPA or Poropak Q, 80-100 mesh. TI = 190 "C 12.5 m X 0.2 mm i.d. Detection of volatile series carborane silicone polymer and from boration of diols with n-butylboronic acid Various organic Tin = 190 "C With the aid of a reference compound it is possible to determine ratio formulae, but the results are inadequate for unknown compounds. Detection limitshg s-1: TM010: He Ar C 0.67 0.2 H 0.13 4.7 2 h Evenson: He Ar C 0.44 0.35 H 0.16 0.36 Gas switches interface illustrated which prevents the solvent from extinguishing the plasma.Linear from 10 p.p.b. to 10 p.p.m. Detection limits: 1S30 p.p.b. H2 doping of the plasma inhibits the formation of oxides of silicates, promotes boron hydride formation and the population of B atomic, rather than ionic states Relative sensitivities for C and H in different compounds were not the same. Attributed to incomplete fragmentation in the low-power plasma used. Detection limits: 1.8-39.0 pg s-* Atmospheric pressure 12.5 m Sp2100 fused- Tetraalkyllead compounds Demonstrates advantages He plasma TMolo silica capillary column. in petrol of element-specific cavity. See ref. 77 detection by comparison of Pb and C responses 100 1 split ratio.T = 40-100 "C at 5 "C min-1 0.01 yl sample Element and wavelength/ nm Reference C? 75 193.1 247.9 H, 486.1 c1, 479.5 481 .O Br, 470.5 487.5 1, 516.1 206.2 s, 545.4 c, 247.86 H, 656.28 C,? 576.52 CH, 431.42 Pb 7 405.8 c , 247.9 B, 247.77 H, 656.279 c , 193.091 F, 685.602 c1, 479.454 Br, 470.486 I, 206.238 s, 545.388 Pb 7 283.3 c, 247.86 76 77 78 79 8 1124 ANALYST OCTOBER 1986 VOL. 111 Table l-continued Element and wavelength/ nm Reference F 81 685.6 Detector Atmospheric pressure, He plasma TMolo cavity. 75 W forward power 12 W reflected Chromatography 1 m X 3 mm i.d. glass column 15% DC-200 on 80-100-mesh Uniport B and 3% OV-17 on 8C100-mesh Uniport HP.He = 80 ml min-1 Sample F in urine Comments F extracted with TCMS and converted to TMFS in toluene. Linear over 4 orders of magnitude. Detection limit 7.5 pg s-* H2 doping of He enables plasma to withstand 1-2 ng s-l throughputs of Pb Ge or Sn. Linear over 3 orders of magnitude. Detection limits: 0.71-6.1 pg Ge, 265,l Sn , 284.0 Pb , 283.3 82 Atmospheric pressure, He plasma TMolo cavity. See ref. 77 12.5 m x 200 pm i.d. SP2100 fused-silica WCOT. Terminated within 1-5 mm of cavity wall Redistribution reactions for Ge Sn and Pb alkyls. Pb alkyls in gasolines Beenakker TMolo cavity He as the support gas. 50 W forward and 0-1 W reflected power.Modified Jarrell-Ash 66000 polychromator 6 ft stainless-steel column (4 in o.d., 2 mm i d . ) packed with 10% Apiezon L on 80-100-mesh Chromosorb P AW at 110 "C. 3 ft silanised glass column (4 in o.d., 4 mm i.d.) packed with 2% OV-101 on 8G100-mesh Chromosorb HP at 270 "C Chlorinated pesticides and brominated flame retardants The polychromator -microcomputer system was developed to monitor four emission wavelengths simultaneously . Detection limits nanogram level with precision in order of 5% RSD c, 247.9 a, 479.5 Br, 470.5 83 84 Atmospheric pressure microwave sustained helium plasma with Beenakker TMolo resonant cavity. Effluent split by 3-way valve with 20% going to FID 2 ft x Q in stainless-steel column packed with Porapak QS 80-100 mesh, using 1-ml gas injections.Bentone 34/DC-550 mixed phase on Chromasorb W HP Application to a number of halomethane and monochlorobiphenyl separations Notes on design, optimisation and utilisation of interface. Detection limits C1 20; P 8.8; Fe, 2.5; Br 10; and S 14.0 ng Cl, 481 .O 479.5 p, 213.6 Fe 7 259.94 Br, 478.6 S, 213.6 545.5 Copper Beenakker cavity 2450 MHz microwave generator and McPherson Model 270 scanning UV - visible monochromator. Interface similar to ref. 77 6 ft x 0.125 in i.d. column packed with OV-17 on Chromosorb W HP. Carrier gas He at 28 ml min-1. Column temperature 85 "C (140 "C for derivatives) Technique used in combination with chemical derivatisation of selected compounds in complex samples e.g., trichloroacetyl derivatives of aliphatic amines Microcomputer used to switch valves; can also be used to control monochromator wavelength settings and acquire analytical data c, 247.9 c1, 479.5 Br, 470.5 85 86 Details not given Modification undesirable c1, in quantitative studies as 479.45 results in degradation of Br, detection limits.Interface 478.55 recommended in ref. 62. Most of paper concerned with hardware and software development for control data acquisition etc. System as described in ref. 68. Minor modification by inserting a stainless-steel tube from the column into the plasma containment tube in hope of reducing dead volume Application to halogenated compounds e.g.Cilex BC-26 Reduced pressure He plasma in parallel with either an FID or ECD. Plasma viewed DB-5. Temperature transversely by a multi-channelspectrometer 10 "C min-1 with a helium Two capillary columns of 30 m X 0.25 mm i.d. with a 1.0 pm thick film of programme 70-300 "C at carrier at 1 ml min-1 Characterisation of fluorine-containing metabolites in blood plasma Inlet splitter to divide F, columns. Interaction of c, fluorine species with 495.7 effluent between the two 685.6 quartz tubing gives rise to peak tailing 8 ANALYST OCTOBER 1986 VOL. 111 1125 Table l-continued Detector Spectrospan IIIB Multi-Element Analyser equipped with a three-electrode DCP-Spectrojet I11 and multi-element cassette.Wavelength scan achieved using Spectrametrics DBC-33 system. Series UV monitor at 280 nm Element and wavelength/ Chromatography Sample Comments nm Reference Gel filtration Speciation of Gel filtration separation c u 88 100 x 2.6 cm column protein-bound requires several hours 324.7 S-300. 5-ml sample and intraVeIlOUS infUSiOIl recalibrated every hour. 213.8 packed with Sephacryl CU Fe and Zn in Serum therefore spectrometer Zn , applied to column fluids Detection limits Cu 3.2; Fe 7 Fe 3.9; Zn 9.3 373.4 Atmospheric pressure Capillary column Pyrolysis products of plasma utilising a 11 m x 0.25 mm i.d., Beenakker type TMolo SE-30 fused silica. He at siloxanes cavity.Low-resolution 1 ml min-1. Temperature scanning monochromator with approximately 0.1 nm after first 6 min.resolution Pyrolysis using Model novel linear silarylene programme 4 " c min-1 100 Pyroprobe Unit Atmospheric He 30 m x 0.25 mm i.d. Organoarsenic plasma operated at thin film DB-1 bonded- compounds 30 W forward power phase fused-silica capillary column Atmospheric He 12 m x 0.25 mm i.d. Chlorinated and plasma TMolo SE-30 fused-silica non-chlorinated cavity. Transfer capillary column organics line similar to ref. 56 As in ref. 91 As in ref. 91. Also a 15 m DB 210 fused-silica capillary column packed with 10% Kel F Oil No. 10 on Chromosorb T and a 3 m Teflon column (Q in o.d., 1/16 in i.d.) packed with 25% dibutyl phthalate on Chromosorb W Organosilicon compounds The interface allowed c , venting of column effluent 253.6 containing large amounts p, of solvents that would 247.6 disrupt helium discharge, while passing labile species without loss Quartz tube extended As , from the interface oven 228.8 into the cavity so that c7 contact with anything 247.9 except column eliminated Use of technique in Se, multi-element detection 203.99 and in determining As 7 empirical formula 228.81 Br, 470.49 CI , 479.45 c , 247.86 p, 253.57 1 7 516.12 s, 545.59 Pb , 283.31 Si , 288.16 H, 656.28 ' F, 685.60 89 90 91 Background emission As in ref.91 92 spectra are compared for plasmas contained within alumina boron nitride and quartz discharge tubes h. coaxial cavity SE-52 FSOT (30 m X Various types of Carrier and scavenger c , No background column and two SCOT with 3-12 carbon atoms use.Determination of H7 DEG 20 m 0, operated at 90 W. 0.315 mm i. d. ) capillary oxygenated compounds gases deoxidised before 247.86 correction used columns empirical formulae 486.13 777.19 (32.5 m x 0.22 mm i.d.) and SE-52 (21 m X 0.22 mm i.d.) 9 1126 ANALYST OCTOBER 1986 VOL. 111 Table l-continued Detector Chromatography Sample System similar to ref. 83 with the internally tuned packed with 2% OV-101. halogenated compounds resonant cavity mounted on the GC oven 3 ft X d in i.d. column Flow-rate 25 ml min-1, Column temperature 300 "C Dioxins and other As in ref. 83 6 ft X 4 mm i.d. glass, OV-101 2% on Chromosorb W HP Pyrethroids and dioxins Comments The H line was monitored with a red-sensitive photomultiplier.Data manipulation as in ref. 83, but modified to store chromatographic data Evaluation of laminar flow torch Element and wavelength/ nm Reference c 94 247.9 Br, 470.5 c1, 479.5 H, 656.3 c , 247.9 p, 253.6 Br, 470.5 C1, 479.5 H, 656.3 F, 685.6 95 The first couplings of GC - ICP were made by Windsor and Denton96.97 in Arizona and Sommer and Ohls99JOO in Dortmund. The former group showed the capability of ICP -optical emission spectrometry (OES) for the elemental analysis of organic compounds96 using an all-argon plasma. This capability was then utilised in a GC - ICP coupling97 for simultaneous multi-elemental analysis of organic and organo-metallic compounds.A natural extension of this work was the derivation of empirical formulae. Windsor and Dentongs used carbon hydrogen and halogen ratios to find the empirical formulae of various organic compounds; however whereas the technique provided the ability to analyse for a large number of elemental constituents suitable lines for oxygen and nitrogen were not found. Sommer and Ohls99 used both all-argon and nitrogen-cooled plasmas for the determination of tetraalkyllead compounds in various petrols by monitoring the lead emission. The same workers100 determined nickel and zinc as diethyldithiocarbamates using a nitrogen-cooled plasma. Fry et al.102 investigated a large number of fluorine atom lines for the selective detection of various fluorine-containing organic compounds using off-line correction to remove interference from the solvent emission.Brown and Fry101 monitored near infrared oxygen emissions to enable oxygen-specific detection. The determination of volatile hydrides of arsenic germanium and antimony by GC - ICP using a sequential slew-scanning monochromator~03 demon-strates how the use of chromatography enables rapid multi-element analysis using a monochromator. Table 2 lists applications of GC - ICP. 4. Coupled Gas Chromatography - Direct Current Plasma The direct current plasma (DCP) is essentially a direct current arc struck between two or more electrodes and stabilised by a flow of inert gas. There are relatively few reported couplings of GC with DCP - OES although the group at Amherst have been particularly active.67@J07JOs They found it possible to use argon helium or nitrogen as a carrier gas,lOs although in certain spectral regions interference from cyanogen bands can occur with nitrogen.The use of a sheathing gas heated to prevent sample condensation around the injector nozzle was found to increase the sensitivity.107JOs The DCP is tolerant to a wide range of gas flow-rates gas and solvent types and this clearly aids versatility even if this is sometimes at the expense of sensitivity. This coupled technique has been used as an element-selective detector for the following manganese as the cyclopentadienyltricarbon yl derivativel07; copper chromium, nickel palladium and zinc chelatesgl; iron in ferrocenellO; and various Group IV metals in an interesting study of Friedel -Crafts catalysed alkyl group redistribution reactions.140 Treybig and Ellebrachtlll utilised a vacuum ultraviolet plasma spectrometer for sulphur-specific detection which compares favourably with MIP detection and has the advantage that solvent venting is not required.The applications of GC - DCP are summarised in Table 3. 5. Coupled Gas Chromatography - Atomic Absorption Spectrometry Coupled GC - AAS applications are summarised in Table 4. Generally these can be seen to involve either flame (FAAS) or electrothermal (ETA) atomisation systems. Flame atomisa-tion offers the advantages of continuous operation simplicity and inexpensive instrumentation. Often it is cited that the low nebulisation efficiency of about 10% for solutions is a disadvantage compared with ETA in which the whole sample is atomised.This is unimportant in coupled GC - AAS as the analyte is in the gas phase prior to entry into the atom cell. However FAAS does suffer the disadvantage of higher detection limits owing to the shorter atomic residence times in the flame. In addition to the increased sensitivity it is also claimed that ETA is safer and lends itself to the possibility of unattended operation. The simplest way of interfacing a gas chromatograph with an atomic absorption spectrometer is to pass the column effluent via an interface tube into the nebulisation chamber, where it is swept by the oxidant and fuel gases into the flame. The first reported GC - FAAS coupling by Kolb et al.112 used this method to determine tetraalkyllead compounds in petrol with an air - acetylene flame. This interfacing method has bee ANALYST OCTOBER 1986 VOL. 111 1127 Table 2. Coupled gas chromatography - inductively coupled plasma optical emission spectrometry Element and wavelength/ nm Reference Detector Chromatography Sample Commen ts 6 ft X fi in i.d. packed with Elemental analysis of Used single and All-Ar plasma Br 97 700.57 c , 247.86 (21, 725.67 F, 634.67 H, 656.28 I, 206.16 Si , 251.61 Fe , 371.99 Pb , 217.00 Sn, 284.00 observations made 9 mm 8% Carbowax 1540 on above load coil. 80-100-mesh firebrick Computer-controlled data acquisition system. See ref. 96 various organic compounds multi-channel monochromators.Using the latter monitored C and H channels for TMT, toluene and p-xylene. Detection limits: 0.8 ng- 1 mg depending on the element All-Ar plasma. See refs 96 and 97. Power 0.8 kW; coolant 12 1 min-1; plasma 0.5 1 min-1; sample 0.9 1 min-1; make-up 0.9 1 min-1 Uses both high-power Ar - SP1000. N2 and low-power Ar - Ar N = 30 cm3 min-1. See ref. 96 98 c, 247.86 H, 656.28 I, 206.16 a, 725.67 Si , 212.4 288.1 Halogen-containing hydrocarbons Elemental ratio determinations for each peak. Typically 200 elemental ratio determinations were taken to yield an average figure 99 100 Determined lead in petrols using standard addition, also TML - TEL ratio and C background at 220.35 nm T = 140 "C (Si), T = 150 "C (Pb) ' plasmas Pb , 220.35 10% Carbowax 20M on Chromosorb P 80-100 mesh.Ar = 25 cm3 min-1, T = 100 "C Tin = 100 "C All-Ar plasma 1.75 kW forward power. Used elongated torch, observation zone 5.5 mm above load coil Monitored near-IR oxygen emissions for various gases and organic liquids Studied effect of varying various plasma gas flows on signal and background levels. Detection limit: 650 ng 0, 777.194 101 6 ft X Q in i.d. packed with Amine 200. Ar = 25 cm3 min-1 T = 105 "C sampling loop used Separation of trifluorobenzene and o-fluorotoluene F/C selectivity of 1.0 at 685.602 nm without background correction. By using "off-line" correction, solvent peak disappears F 102 considered lines in the region 350-895 1 ccg 56 All-Ar plasma All-Ar plasma with slew scanning monochromator.1 kW forward power. Observation 15 mm above load coil 3.5 ft x 3 mm i.d. Chromosorb 102 at ambient temperature Sequentially eluting hydrides monitored. Linear over 2-3 orders of magnitude. Detection limits 4 ng of Ge 50 ng of As and Sb Ge , 303.9 As, 278.0 Sn , 317.5 Sb , 287.8 103 Hydrides generated cold trapped and passed through column into plasma Spherisorb C,,-ODs 10 pm 250 x 4.6 mm 120 x Interface via a 10 cm 1/16 in 0.d. PTFE capillary from column to the nebuliser AII-Ar plasma. Data acquisition via chart Alkyllead compounds Pb , 405.78 105 recorder or microcomputer 4.6 mm. LiChrosorb RP-2, as in ref. 104 10 pm 120 X 4.6 mm.Partisil-10 SCX 25 cm. Various mobile phases examined Effluent from column passed directly into the nebuliser y, 328.937 Also 13 additional elements Ar plasma. All observations 15 mm above LC cartridge in earths in geological load coil. 1.1 kW forward conjunction with a Z material power module Radial 8 PSCX 10 pm Radial Pak Yttrium and selected rare Compression separation system 10 1128 ANALYST OCTOBER 1986 VOL. 111 Table 3. Coupled gas chromatography - direct current plasma optical emission spectrometry Detector Prototype Spectraspan I11 d.c. plasma Cchelle spectrometer See ref. 107. Details of heated interface design given. Dual detection with FID used sheathing gas heated to 230 "C to prevent condensation of eluents.Ti = 230 "C For spectrometer and interface see ref. 109. A 3-electrode jet was used rather than a 2-electrode jet. Ar flow-rates: sheathing 1.42-1.65, cathode = 2.0 and anode = 1.3 1 min-1. Current = 7 A voltage = 40-60 V See refs. 67 and 108 Vacuum UV spectrometer with Spectrametrics d.c. plasma Chromatography 6 ft X Q in i.d. stainless-steel 2% Dexsil 300 GC on 100-120-mesh Chromosorb 750. 1 1 split with FID. He = 25 cm3 min-1 T = 130 "C, TI = 160 "C Ti = 170 "C 6 ft X Q in i.d. 3% Dexsil 300 on 100-120-mesh Chromosorb 750. T = 170 "C He = 60 cm3 min-1 T = 220 "C. 6 ft X + in i.d. 2.5% Dexsil 300 GC. T = 230 "C T = 280 "C. 6 ft X Q in i.d. 3.2% Dexsil300 GC on 100-120-mesh Chromosorb 750. T = 190 "C.6 ft X Q in i d . , 10% SE-30 on 60-80-mesh Gas-Chrom S . T = 170 "C 6 ft x Q in i.d. stainless-steel 5% OV-101 on 100-120-mesh Chromosorb 750. He = 40 cm3 min-1 T = from 80 "C at 6 or 8 "C min-l TI = 210 "C Ti = 220 "C. Nickel tubing 1 m x Q in i.d. 3% OV-201 on 100-120-mesh Ultrabond 20M. He = 40 cm3 min-1, T = from 80 "C at 8 "C min-l TI = 210"C Ti = 220 "C Sample MMT in gasoline, standards in isooctane. Eymantrene as internal standard Cr(tfa), Friedel - Crafts catalysed alkyl group redistribution reactions 100 ft x 0.03 in i.d. Ferrocene and stainless-steel OV-101 haloderivatives PLOT column. T = 170 "C 122 cm X 2 mm i.d., Poropak Super Q. 183 cm x 2 mm i d . 3% OV-101 on Chromosorb W HP, 80-100 mesh. N2 = 80 CS2 thiophene 3-methylthiophene, hexanethiol benzenethiol and dimethyl sulphoxide.Detection limit 0.3 ng s-l cm3 min-1 of s Comments Only sample modification required was addition of the internal standard. 3 min analysis time. Upper limit of linear range was 340 ng. Detection limit: 3 ng Sheathing gas around the issuing GC effluent prevented excessive diffusion as the sample travelled into the plasma from the interface tubing. Linear from 2 to 150 ng for Cr. Detection limits: 0.28-320 pg s-1 Redistribution reactions of the following pairs: n-Pr4Sn+Et4Pb; Et4Sn +n-Bu4Ge; n-Pr4Si+n-Bu4Ge; n-Bu4Ge+Et,Pb; Vn4Si+Et4Sn; and Vn4Si+n-Bu4Ge studied. Formation of PbR3CI and SnR3CI by reactions with AIC13 also studied Paper contains many other organome t allic separations; however the detector used is the FID Element and wavelength1 nm Reference Mn 107 279.8 c u , 324.7 Ni , 341.7 Pd , 340.4 c , 247.8 Cr, 267.7 c , 247.8 Si , 251.6 Ge 7 265.1 Sn 7 286.3 Pb , 368.3 Pb , 368.3 Sn, 286.3 Fe 7 372.0 s, 180.7 108 67 110 111 utilised by various workers.113J20J24 Morrow et al.113 used the dinitrogen oxide - acetylene flame for the silicon-specific detection of silylated alcohols and an air - acetylene flame for atomic emission detection of the same species. A similar coupling was used to determine lead in petro17120J24J37 and in the atmosphere.124 Hahn et al.145 used such an arrangement to determine As Ge Se and Sn after hydride generation using a hydrogen diffusion flame.Cokerl" realised that dilution of the sample and the excessive peak broadening caused by passage through the nebulisation chamber could be avoided. He passed the chromatographic effluent into a manifold just below the burner slot and achieved lower detection limits for tetraalkyllead compounds in petrol than the previous coupl-ings. Wolf130J36 used a similar coupling to specifically determine chromium in standard orchard leaves after chela-tion with trifluoroacetylacetone as did Chanlso when investi-gating tetraalkyllead ratios in petrols from various sources. Work in our laboratories15lJ52 has emphasised that in order to enable true trace level determinations by GC - FAAS the residence times of atoms in the flame must be increased.This was achieved using a flame-heated ceramic tube suspended over a flame in various configurations. The system described by Ebdon et al. 151 or variations of it are now used routinely in a number of laboratories,l52-155 particularly for the speciation of alkyllead compounds. The electrothermal devices used in coupled GC - AAS fall into three main categories (i) laboratory-made electrother-mally heated quartz or ceramic tubes; (ii) commercial graphite furnaces; and (iii) commercial cold vapour mercury analysers. This latter atom cell has been used for mercury-specific detection of organomercurials in various samples. Heyl14 passed the effluent from the chromatograph into a continuous wet chemical reduction cell the reduced Hg(0) being swept into the cold vapour absorption cell of a commercial system (MAS 50 Coleman Instruments).Other workers116119 used a flame-ionisation detector flame to atomise the organomercury species which were then passed into the same cell. Dress-man116 used this method to speciate dialkylmercur ANALYST OCTOBER 1986 VOL. 111 1129 Table 4. Coupled gas chromatography - atomic absorption spectrometry Element and wavelength/ nm Reference Detector Chromatography Sample Comments Flame AAS GC effluent passed via a heated tube Apiezon M on and TEL GC - AAS coupling for into the nebulisation Chromosorb R. N2 = 40 element-specific detection. chamber ml min-1 T = 150 "C Linear range 50-700 2 m x 2 mm i.d. 10% Pb alkyls in petrol TML First paper to describe p.p.m.Pb 112 217.0 Flame AES N 2 0 - C2H2, FAAS air - C2H2 flame. column 20% SE-30 on of n-alcohols C1-C7 stainless-steel (0.0345 in Coupling was through the i.d.) heated in excess of nebulisation chamber T,. Linear ranges AAS 4-20 pg AES 3-100 pg. Detection limits AAS, 0.11 pg; AES 0.72 pg 6 ft X 0.25 in i.d. steel 30-60-mesh Chromosorb W. He = 100 ml min-1, T = 130 "C Silylated pyridine solutions Interface tube Si 113 251.6 Hg 114 253.7 Using cold vapour analyser Organomercury Passed GC effluent into a compounds continuous wet chemical reduction vessel; Hg then flushed into cold vapour cell. Linear up to 10 pg. Detection limit 50 ng Glass 6 ft x 0.25 in i.d. Alkylmercury compounds GC effluent passed into a column 5% HIEFF-2AP in fish tissue MeHgCl and quartz tube combustion on Chromosorb W HP EtHgCl furnace (780 "C) prior to 80-100 mesh.N2 = 120 passing into the cold r.1 min-I TI = 200 "C T vapour cell. Linear up to = 170 "C Ti = 200 "C 45 ng. Detection limit: 2.5 x 10-11 g of MeHgCl gives 1% absorption As ref. 114 Hg 115 253.7 See ref. 114 6 ft x 2 mm i.d. glass column 5% DC-200 + 3% compounds in spiked river QFI on 80-100-mesh waters combust the mercury Chromosorb Q. T = 70 "C hold 2 min then 20 "C min-1 to 180 "C Dialkylmercury The effluent was passed through the FID to compounds prior to entry into the cold vapour analyser. Detection limit: 0.1 ng Hg 7 253.7 116 See ref. 116 Dialkylmercury See ref. 60. Linear from compounds Me2Hg 0.05 to 100 ng. Detection Et2Hg n-Pr,Hg n-Bu2Hg limit 0.02 ng for Me2Hg See ref.114 See ref. 114 Hg 7 253.7 117 118 See ref. 116 Dialk ylmercury See ref. 60. Linear from compounds 0.05 to 100 ng for Me2Hg and Et2Hg. Detection limit 0.02 ng for Me2Hg Hg , 253.7 See ref. 114 6 ft X 0.125 in glass column 5% SP2100 + 3% involved in pathways in SP2401 on 80-100-mesh transformations of microorganisms Supelcon AW DCMS. N2 = 20 ml min-1, T = 60 "C hold 2 min then 32 "C min-1 to 180 "C Mercury compounds Study of methylation microorganisms in soils and sediments Hg 7 253.7 119 120 Air - acetylene flame 3 m x 3 mm PTFE tube. N2 = 40 ml min-1 T = 110 "C Pb alkyls in gasoline samples Effluent passed from GC into spray chamber; 5-cm burner. Linear from 0.2 to 40 pg Pb , 217.0 Graphite furnace kept at 10 cm W transfer line 2700 "C with background connected into an enlarged correction OV-210 on Gas-Chrom Q.hole in graphite tube. Detection limit 10 ng of 6 ft X %16 in i.d. on glass column 4% SE-30 + 6% Ar = 50 ml min-1 T = 150 "C 2 . 0 4 injections Pb Pb alkyls in gasoline Pb 121 217. 1130 ANALYST OCTOBER 1986 VOL. 111 Table k o n t i n u e d Detector Electrothermally heated silica tube (60 x 7 mm i d . T = 1000 "C). Furnace gases air = 120 ml min-1; H2 = 120 ml min-1 Air - C2H2 flame AAS using an electrothermally heated silica furnace. See ref. 122 Air - C2H2 flame. All-glass lining for nebulisation chamber used to prevent absorption of organolead on chamber walls Electrothermally heated silica tube.See ref. 122 Electrothermally heated silica tube. See ref. 122 T furnace atomiser (900-1000 "C; 100 x 20 mm i.d.). Flows into atomiser Hz = 1 1 min-1, N = 6 I min-l. Quartz T furnace Chromatography 1.8 m X 6 mm i.d. glass column 3% OV-1 on Chromosorb W 80-100 mesh. T = 40 "C hold 2 min then 15 "C min-1 to 120 "C. TI = 225 "C 3 ft x 3/16 in i.d. steel column 10% Carbowax 20M on 100-120-mesh Porasil C. H2 = 120 ml min-l T = 130 "C. Laboratory-made column heating system. 5-pI injections See ref. 122 1.8 m x 6 mm i.d. glass column 3% OV-1 on 80-100-mesh Chromosorb W. N2 = 65 ml min-l T, = 40 "C hold for 2 min then 5 "C min-1 to 90 "C Column (see ref. 68). N, Element and wavelength/ Sample Comments nm Reference Me2Se and Me2Se2 in synthetic air samples Se , 196.0 Air samples trapped at -80 "C on 3% OV-1 on Chromosorb W and desorbed into the GC at 80 "C.The trap was heated in a commercial "toaster." Linear up to 50 ng. Detection limit 0.1 ng 122 Pb alkyls in gasoline The effluent from the GC Pb 7 111 passes into a manifold just below the burner slot which evenly distributes the effluent along the flame. Linear up to 200 p.p.m. for TML and 1000 p.p.m. for TEL. Detection limit 0.2 p.p.m. 283.3 TML from methylation of Reported that Me,Pb+ Pb 123 Me,Pb+ salts salts were readily 217.0 converted to TML by microorganisms in lake water or nutrient medium Tetraalkyllead compounds The air sample was Pb 124 in the atmosphere and trapped (see ref.66) then 217.0 gasolines passed through a nebulisation chamber into the flame. Detection limit: 80 ng Tetraalkyllead compounds For sample trap and Pb 125 = 70 ml min-1 T = 56°C in the atmosphere for 2 min then 15 "C min-* to 150 "C, TI = 150 "C chromatographic interface 217.0 see ref. 122. Linear up to 200 ng. Detection limit: 0.1 ng 1.8 m x 6 mm i.d. 3% OV-1 on Chromosorb W, 80-100 mesh. Lead see ref. 68. Selenium NZ = 70 ml min-l T = 40 "C for 2 min then 15 "C min-1 to 120 "C, TI = 225 "C. Arsenic 10% OV-1 on Chromosorb W. N = 30 ml min-1, T = 25 "C = T I , Ti = 100 "C. Mercury, 5% DEGS on Chromosorb W. N2 = 80 ml min-1, Ti = 150 "C. Cadmium, N2 = 70 ml min-1, T = 70 "C, TI = Ti = 80 "C Organometallic compounds in liquid or gaseous samples.For gaseous sample trapping method see ref. 122 T = 145 "C TI = 150 "C, Compounds determined were tetraalkylleads, methylseleniums, methylarsines, alkylmercury chlorides and dimethylcadmium. Detection limits 0.1 ng for each element Hg 9 126 253.6 Pb , 217.0 Cd , 228.5 As, 193.7 Se , 196.0 122 cm x 3 mm i.d. A1 Dialkylselenium The laboratory-made Se 7 127 tube 20% polymetaphenyl compounds chromatographic system 196.0 ether on 60-80-mesh Chromosorb W. N2 = 23 ml min-l T = 82 "C, was contained in the quartz T arrangement TI = 180 " ANALYST OCTOBER 1986 VOL. 111 1131 Table k o n t i n u e d Detector See ref. 127 Flame with chromatographic effluent being delivered directly to the burner cavity Electrothermally heated silica furnace see ref.122, or directly coupled through the nebulisation chamber to an air - C2H2 flame see ref. 124 H2 diffusion flame burning in quartz cuvette. H2 = 250 ml min-1 air = 150 ml min-1 See ref. 117 Graphite furnace with pyrolytic or alumina lining or standard graphite tubes, at various temperatures with and without Ar - H2 (90 + 10) flow (20 ml min - 1 ) Graphite furnace at 2000 "C. The furnace was kept at this temperature throughout the chromatographic run Air - C2H2 flame; see ref. 130 Chromatography See ref. 127 2 ft x 3 mm i.d. PTFE tubing 10% SE-30 on Chromosorb W HP, 80-100 mesh. N2 = 65 ml min-1 T = 180 "C 20-1.11 injection See ref.125 6 m stainless-steel column, 16.5% DC-550 on 80-100-mesh Chromosorb W AW DMCS, He = 80 ml min-1 80 cm x 6 mm i.d. glass column 10% Carbowax 20M on Chromosorb W AW. 5-100-yl injections, N2 = 15 ml min-' for Me2Hg 200 ml min-1 for MeHgCl TI = 200 "C Tc = 60 "C for Me2Hg 200 "C for MeHgCl Element and wavelength/ Sample Comrnen ts nm Reference Organoselenium compounds transpired Astraaalus racernosus The transpired compounds Se 128 129 on Chromosorb W in a dry ice-bath and desorbed at 175 "C into the chromatographic column. Detection limits Me2Se = 10 Me2Se2 = 20 and Et,Sez = 20 ng by were trapped on DC-550 196.0 Inorganic Cr in NBS SRM After a H2SO -k H Z ~ Z 1571 orchard leaves as digestion Cr chelated with Cr(tfa) chelates Htfa (0.1 ml) then extracted with hexane (0.5 ml) prior to injection.Linear from 0.5 to 5 p.p.m. of Cr. Detection limit 1 ng Cr Tetraalkyllead compounds For atmospheric sampling Pb , 217.0 in petrol and air samples see ref. 122. Linear up to 200 ng for furnace. Detection limit 0.1 ng for furnace system Reducible As species in The hydrides of the As As , compounds isolated by 193.7 cold trapping passed down a column and into a furnace. Linear up to 50 ng. Detection limit 0.05 ng for ASH, natural waters MezHg MeHgCl Detection limit 10 p.p.b. Hg 7 Hg 253.7 6 ft X Q in i.d. glass Me3As Me& and Me,Se Best detection levels column 5% SP2100 and in Nz. To simulate an achieved using standard 3% SP2401 on atmosphere over a lake graphite tubes with an 80-100-mesh Supelcon system Ar - H2 flow at 1800 "C.AW DMCS. Ar = 30 Linear up to As 320 Se ml min-l T = 40 "C 313 and Sn 363 ng. Ti" = 100 "C Detection limits 5-12 ng PTFE column 8 ft x 4 in i.d. 20% TCP on Chromosorb W. Ar = 30 atmosphere limit 0.1 ng ml min-I T = 100 "C TI Tetraalkyllead compounds TEL undetected in all 10 in gasoline and the air samples. Detection = 125 "C Ti = 100 "C 18 in x 3 mm i.d. PTFE Inorganic Cr in NBS SRM tubing 5% SE-30 on 1571 orchard leaves and Chromosorb P AW SRM 1569 brewers yeast Fe(tfa)3 and Cu(ofhd),. DMCS 80-100 mesh. N2 as chelates also Co Fe = 120 ml min-1 and Cu chelates Detection limits 1.&500 The chelates determined were Co(fod), Fe(fod),, Linear from 0.5 to 8.0 pg. ng Tc = 160 "C TI = 150 "C As , 193.7 Se , 196.0 Sn, 224.6 Pb , 283.3 130 131 132 133 134 135 Cr 136 c o Fe c 1132 ANALYST OCTOBER 1986 VOL.111 Table 4-continued Detector Element and wavelength/ Sample Comments nm Reference Chromatography Both a flame air - C2H2 20% SE-52 on Tetraalkyllead compounds The furnace technique was Pb 137 the effluent introduced Chromosorb W. Ar = 90 in gasoline samples 100 and 75 times more 283.3 through the nebuliser and a graphite furnace at 1300 "C TEL respectively. ml min-1 T = TI = 125 "C Ti = 130 "C sensitive than the flame coupling for TML and Detection limits flame, TML = 17 and TEL = 81 ng; furnace TML = 0.12 and TEL = 1.1 ng anelectiically heated quartz furnace at 620 "C Hg compounds atomised in 10% SP2300 on Alkylmercury compounds Rapid method for Hg , Chromosorb W.N2 = 90 in fish quantitative extraction of 254.0 ml min-1 T = 145 "C, TI = 200 "C Electrothermally heated See ref. 122 silica tube. See refs. 122 and 125 Graphite furnace atomisation at 1700 "C Various atom cells, air - C2H2 flame flame and electrothermally heated quartz tubes, graphite cup and furnaces Graphite furnace atomisation (see ref. 135) at 1500 "C Electrothermally heated silica furnace see ref. 122 Electrothermally heated silica tube furnace see refs 139 and 125 Graphite furnace atomiser organomercury compounds from fish given. Linear up to 120 ng. Detection limit: 3.5 ng Tetraalkyllead compounds Extraction procedures for Pb , 217.0 in water sediment and fish three sample types given.Detection limits water (200 ml) = 0.5 pg 1-1, sediment ( 5 g) = 0.01 pg g-1 fish (2 g) = 0.025 Pg g-l 150 cm X 6 mm i d . glass Tetraalkyllead compounds The Pb compounds from Pb, column 3% OV-101 on in air 70-1 air samples were 283.3 Chromosorb W 80-100 mesh. Ti = 80 "C T = 90 "C then 40 "C min-1 to 200 "C or isothermal at 150 "C trapped at -72 "C on the chromatographic packing. Detection limit 40 pg 150 cm x 6 mm i.d. glass Tetraalkyllead compounds If Ti > 300 "C Pb , column 3% OV-101 on decomposition of lead 283.3 Chromosorb W 80-100 mesh. N2 = 140 ml min-1 T = 50 "C then 40 "C min-1 up to 200 "C compounds occurred and interference from remobilisation by the solvent resulted.Detection limit 30 pg with HGA 2100 furnace 18 in X i in i.d. PTFE column 20% Ucon surface and evaporates. 283.3 Non-Polar on Chromosorb P. Ar = 60 ml min-1, T = 140 "C TI = 150 "C, Ti = 140 "C TEL in sea water Some TEL migrates to the Pb , The majority forms the soluble Et3PbC1. Evidence of further degradation was found. Detection limit: 1 pg rnl-1 See ref. 122 TML in methylation of Found a chemical Pb 9 Pb" salts in aqueous methylation pathway for 217.0 solution converting PbI1 salts into methyl derivatives 138 139 16 140 141 158 See refs. 125 and 139 Tetraalkyllead compounds Samples were analysed for Pb 142 in fish sediment total Pb volatile Pb 283.3 vegetation and water tetraalkyllead and samples hexane-extractable Pb 2.3 m X 6 mm i.d.3% MMT in air samples The air samples were Mn 143 OV-101 on Chromosorb collected (see ref. 84) at 70 279.5 W HP 80-100 mesh. N = 80 ml min-l T = 115 "C, ml min-* for 8 h. Detection limit 0.05 ng m-3 TI = 150 "C Ti = 150 " ANALYST OCTOBER 1986 VOL. 111 1133 Table Aontinued Element and wavelength/ nm Reference Detector Chromatography Same as ref. 142 Sample Comments Determination of total, hexane-extractable, volatile and tetraalkyllead in fish water sediment and vegetation samples. See ref. 142 Graphite furnace atomisation Coupling of chromatograph transfer line to the furnace was via friction-fitted Ta connector (ref. 140). Detection limits 2 p.p.b. of hexane-extractable, 0.5-1.5 p.p.b. of volatile and 0.5 p.p.b.of tetraalkyllead Pb 7 144 283.3 3 ft X 4.7 mm i.d. Polypenco Nylaflow tubing Chromosorb 102. T = 23 "C Determination of As Ge, Se and Sn after hydride generation and cold trapping of hydrides H2 diffusion flame, samples introduced through nebuliser Chromatographic separation allowed manual lamp change and monochromator change between peaks. The overlap of SeH2 and SnH, required their separate detection. Detection limits 60-260 ng As 7 193.7 Ge , 265.2 Se, 196.0 Sn 7 224.6 145 Coupling via 1 m x 0.5 Pb 7 mm i.d. glass tube. Linear up to 50 ng. Detection limits 40 pg of TML and 90 pg of TEL 283.3 Graphite furnace atomisation at 2000 "C. External gas flow of 0.9 1 min-1 Glass column 180 cm X 2 TML and TEL in petrol mm i.d.3% OV-101 on Gas-Chrom Q 100-120 mesh. Ar = 30 ml min-', T = 50 "C then 20 "C min-' up to 150 "C, Ti, = 200 "C 146 Pb 7 283.3 147 Graphite furnace atomisation see ref. 96 Same as ref. 146; samples desorbed from short glass column of 6 1 min-1 chromatographic material at 90 "C into chromatograph Tetraalkyllead compounds in air sampled for 1 h at Pb compounds sampled on to glass beads at -130 "C, then transferred to a short column of chromatographic packing at - 196 "C. Detection limits TML = 0.1 and TEL = 0.3 ng m-3 2 m x 6 mm i.d. glass column 3% SE-30 on Chromosorb G AW DMCS. For R = Me N2 = 16 ml min-1 T = 120 "C. For R = Et N2 = 50 ml min-1 T = 180 "C Tetraalkyltin and alkyltin chlorides (R,SnCl,-,, R = Me and Et) Owing to column rearrangements all four methyltin compounds cannot be examined.Passed column effluent directly to atomiser and also generated hydrides prior to atomisation. Linear up to 400 ng. Detection limits 1 .O ng for Me,Sn 2.0 pg for Me$n if hydride is atomised Electrothermally heated quartz tube Sn , 286.3 148 Graphite furnace atomisation see ref. 96 Same as ref. 146 149 Tetraalkyllead compounds in air (cf. ref. 97) petrol (cf. ref. 96) river and rain water Degradation of TML and Pb , TEL in river water 283.3 investigated. Detection limits TML = 0.2 TEL = 0.5 pg 1-1 Interface line was 4 ft X 0.02 in i.d. stainless steel. Linear up to 400 ng for TML 1400 ng for TEL Pb 7 217.0 Air - C2H2 flame; effluent from chromatograph introduced just below burner slot 10 ft X i) in i d .steel column 20% Carbowax 20M on Chromosorb P. N2 sources = 120 ml min,-' T = 120 "C 2 p1 injected Tetraalkyllead compounds in petrol from a variety of "C TI = 140 "C Ti, = 110 15 1134 ANALYST OCTOBER 1986 VOL. 111 Table A o n t i n u e d Element and w avel e ng t hl Comments nm Reference Chromatography Detector Sample 4 ft glass column 20% OV-3 on Chromosorb W, 80-100 mesh. N2 = 80 ml min-l T = 30 "C for 3 min then 20 "C min-1 up to 110 "C TI = 85 "C, Tin = 76 "C Electrothermally heated silica furnace (see refs 122 and 125) at 850 "C. H2 = 150 ml min-1 Methyltin compounds sampled from the headspace above sediment samples in a methylating environment Headspace sampling (see Sn 156 ref.122). Experiments 224.6 indicated that SnrI was methylated by CH31 but SnIV was not. Detection limit 0.1 Electrothermally heated silica furnace; see refs 122 and 125 For chromatographic conditions see refs. 124, 125 and 126 Methylated derivatives of As Hg Pb and Se Study of the effect of pH on methylation in the 193.7 aquatic environment. Detection limits 0.1 ng of each element 253.6 As, Hg, 157 Pb , 217.0 Se , 196.0 Tetraalkyllead compounds in the atmosphere. Samples taken from rural, urban and gasoline station environments Elevated levels of Pb , tetraalkyllead compounds 283.3 were found around gasoline stations and in areas with heavy traffic.Linear up to 50 ng. Detection limits 40 pg of TML and 90 pg of TEL 159 Graphite furnace atomisation; see refs. 146 and 147 See refs. 146 and 147 Electrothermally heated silica tube see ref. 122 180 cm X 6.4 mm i.d. 3% OV-1 on Chromosorb HP, 80-100 mesh. N2 = 25 ml min-1 T = 70 "C, TI = 150 "C Tetraalkyllead compounds formed in study of methylation pathways in coastal sediments Reported that Pb , bioconversion of Pb" to 217.3 TML unlikely in marine environments 160 Electrothermally heated quartz tube (cf. ref. 125) at 980 "C 8 cm X 3.2 mm i d . stainless-steel column, Porapak Q 80-100 mesh. TML was trapped on column and flushed off with N2 (150 ml min-l) by placing the column in a toaster (cf. ref. 125) at T = 235 "C Methylation was affected Pb, by methyllithium and only 283.3 a 50% conversion was achieved.Linear up to 200 ng. Detection limit 5 ng 167 Determination of inorganic Pb in aqueous samples as tetramethyl derivative formed by methylation of the extracted dithiocarbamate complex Sn , 224.6 162 Electrothermally heated quartz tube (see ref. 125) 1.8 m X 6 mm i.d. glass column 3% OV-1 on Chromosorb W 80-100 mesh. N2 = 65 ml min-', TI = 180 "C T = 90 "C then 20 "C min-1 up to 190 "C Tin = 165 "C Organotin compounds, Me,SnBu+ in water Tin compounds were extracted with a 0.1% tropolone in benzene solution from spiked water samples. Linear up to 33 ng. Detection limit 0.1 ng Flame and a flame-heated ceramic tube 1.5 m X 4 mm i.d.glass column 5% Carbowax 20M on Chromosorb 750, 80-100 mesh. T = TI = Ti = 159-175 "C Tetraalkyllead compounds Pb 9 Various atom cells developed and simplex 283.3 optimised. Detection limit: 17 pg for most sensitive atom cell 151 Mo furnace surrounded by an alumina sleeve heated at 250 K s-' to 2473 K 247 mm X 1.22 mm i.d. Mo column with a wall thickness of 0.81 mm. Carrier gas of either Ar at 44.7 k 2.1 p1 s-1 or Ar + H2 at 35.1 k 0.8 and 13.5 k 0.4 p1 s-l respectively. T = 2093 K Na Cu Mn and Mg in inorganic salts Ar (3.8 1 min-1) and H2 Na 7 provide an air-free c u 7 Mn 9 (1.2 1 min-1) used to atmosphere around tube 16 ANALYST OCTOBER 1986 VOL. 111 1135 Table k o n t i n u e d Detector Modified form of flame AAS system used in ref.151 to permit the use of Perkin-Elmer burners requiring high gas flow-rates Silica furnace consisting of an electrically heated quartz T-tube encased in a shaped firebrick. Assembly mounted in an aluminium cradle positioned within the optical beam of the spectrometer Flame AAS system based on ref. 151 Element and waveleng thl Comments nm Reference Chromatography Sample 1.5 m X 6 mm 0.d. X 2 Ionic alkvllead comDounds Problem of sample Pb 7 153 mm i.d. column packed with 10% OV-101 on Chromosorb W 80-100 mesh. Temperature programme 50-250 "C at 10 "C min-1 in water 1.8 m x 6 mm i.d. glass column packed with 10% OV-101 on 80-100-mesh Supelcoport. He flow-rate: 35 ml min 1-1. Temperature programme up to 250 "C Alkyllead compounds in environmental samples 1 m x 6 mm 0.d.2 mm i.d. glass column in air containing 3% OV-101 on Gas-Chrom Q (100-120 mesh) Tetraalkyllead compounds introduction into the atom 283.3 cell overcome using commercially available open silica cell normally employed with the Perkin-Elmer MHS-10 mercury - hydride system. Detection limits ng 1-1 Furnace operating conditions 900 "C and hydrogen make-up gas at 50 ml min-l. Detection limits about 30 pg with claims of possible improvement by improving the chromatographic efficiency Pb 7 217.0, 283.3 Samples collected using Pb , cryogenic trapping at 283.3 -196 "C then flash-hea ting 164 154 compounds in spiked river waters. Blair et al.119 also used this method in a study of mercury transformations in aquatic environments. Gonzalez and Ross115 used a quartz combus-tion furnace prior to the detector to determine methyl- and ethylmercury chlorides in fish tissues and found better selectivity towards mercury than that exhibited by electron-capture detectors towards the organomercury chloride. The use of an electrothermally heated silica tube as an atom cell for coupled GC - AAS was pioneered by Chau et a1.122 The furnace heated to around 1000 "C with a through flow of air and hydrogen was used with a selenium-specific detector in the determination of dimethyldiselenium and dimethylsele-nium.122 Chau with a number of co-workers then used this coupled technique for numerous environmental applica-tions.122~239126~56,157 This group later developed the same technique for the metal-specific detection of organolead in the atmosphere,124J25 the aquatic environmentl39J42 and for methylation studies of lead,123J58 tin,156 arsenic mercury and selenium.157 Thompson160 utilised a similar atom cell to study methylation pathways in coastal sediments whereas Brueg-gemeyer and Carus0161 used the same. system for the determination of inorganic lead in aquatic samples after methylation of the extracted dithiocarbamate lead complex. Van Loon and Radziuk127-129 developed a silica T-tube for coupled GC - AAS. This inexpensive arrangement had the chromatographic column contained in the long arm of the T and the effluent then passed into the cross-piece atomiser purged with flows of hydrogen and nitrogen.The system was used as a metal-specific detector for organoselenium com-pounds127 and in the study of organoselenium transpiration by Astragalus racernosus.128J29 Bye and Paus138 used an elec-trothermally heated quartz furnace to atomise organomer-curial compounds prior to their detection in an unheated silica cuvette. In a comprehensive study of various tetraalkyl, methyl- and ethyltin chloridqs148 Burns et al. used an electrothermally heated quartz tube as an atomiser. They found that detection limits could bed lowered substantially if the hydrides were generated prior to atomisation. In a comparison of various atom cells for coupled GC - AAS by Radziuk et a1.140 the graphite furnace proved the most sensitive for lead and gave a 50-fold increase in response compared with the early Kolb-type flame coupling.It is clear from the work of Ebdon et a1.151 reported above that by using ceramic tube atomic traps FAAS couplings can be as sensitive if not more so than furnace couplings as the vital parameter in optimisa-tion is the residence time of the atoms in the absorption cell. The first GC coupling to a commercial graphite furnace was rather crudely achieved by Segar.121 The end of a tungsten transfer line was passed through an enlarged hole in the graphite tube so that the effluent impinged on the hot tube wall. Parris et al. 134 considered the effect of using pyrolytically coated alumina-lined and standard graphite tubes at various atomisation temperatures with and without hydrogen (10%) added to the chromatographic effluent.The best detection levels were achieved for As Se and Sn using standard graphite tubes with hydrogen added to the effluent flow and an atomisation temperature of 1800 "C. Robinson et al.135 passed the chromatographic effluent through a graphite electrode into the optical path of a laboratory-made atomiser which was kept at 2000 "C throughout the chromatographic run. This atomiser was used for lead-specific detection of tetraalkyllead compounds in petrol135 and in a study of the degradation of tetraethyllead in sea water.141 Bye et al. 137 found that graphite furnace atomisation was 100 times more sensitive than flame atomisation for the determination of tetramethyllead in petrol. The determination of tetraalkyllead compounds in various matrices had been well researched; for example by Cruz et al.165 in fish water sediment and vegetation samples. The group in Antwerp developed the most sensitive GC -GFAAS coupling for tetraalkyllead compounds146 and used it to determine these compounds in petro1,146J49 the atmo-sphere147J49 and in a preliminary study of their degradation in river water.149 The determination of another anti-knock petrol additive methylcyclopentadienylmanganesetricar 1136 ANALYST OCTOBER 1986 VOL. 111 Table 5. Coupled gas chromatography - atomic fluorescence spectrometry Detector Chromatography Sample Element and wavelength/ Comments nm Reference Circular N2 shielded See Table 4 in ref. 140 Tetraalkyllead compounds FAFS 3 times more Pb 140 circular air - C2H2 flame Electrothermally heated no better than quartz tube furnace electrothermal AAS sensitive than FAAS.Electrothermal AFS was Graphite cup furnace at 1000 “C bonyl in the atmosphere was achieved by Coe et a1.143 at levels down to 0.05 ng m-3 concentrations. Winefordner et a1.163 have demonstrated a novel method of avoiding matrix interference by selective volatilisation using coupled high-temperature (ca. 2093 K) GC - AAS. They used a molyb-denum column - atomiser for the separation of sodium, copper manganese and magnesium ions with an excellent correlation of analytical signals for each metal in both pure and mixed solution. This work opens a new area of application for GC - AAS as prior to this only elements that formed volatile hydrides or chelates in inorganic matrices could be separated.The technique thus offers a possible method for separating interfering concomitants from the analyte prior to atomic spectroscopic analysis. 6. Coupled Gas Chromatography - Atomic Fluorescence Spectrometry To date chromatographic applications of FAFS have utilised only line sources (Table 5). Van Loon166 first suggested the possible use of non-dispersive AFS as a detector for chromato-graphy noting its multi-element capability ability for low level detection and simplicity of usage. Although this latter point is debatable the most likely reason for the dearth of published work using GC - AFS is probably the lack of sufficiently intense stable and simple light sources. Van Loon’s group in Toronto have published the only GC - AFS work,140 in which they used il nitrogen separated circular air - acetylene flame, an inert gas shielded electrothermally heated quartz tube and a modified graphite cup atomiser.In the lead-specific detection of tetraalkyllead compounds flame AFS proved a factor of three more sensitive than FAAS; however no increase in detectability was found using AFS over AAS when the graphite cup or quartz tube atomisers were used. The availability of a commercial AFS instrument should increase the usage of the technique as the advantages of multi-element analysis and sensitive detection make AFS an excellent method for the determination of metals. 7. Conclusion Historically the MIP has proved the most popular excitation source to couple with gas chromatography.This is probably a reflection of the MIP’s ability to monitor certain non-metallic elements in addition to metals and particular mention should be made of the ability of the helium MIP to monitor halogens. Commercially available GC - MIP systems have unfortunately used low-pressure He plasmas and thus have suffered from the attendant problems of vacuum lines and gas transfer from atmospheric pressure in the chromatograph to low pressure in the detector. The availability of the Beenakker TWlo cavity, which allows an atmospheric He plasma to be sustained yields a more satisfactory GC - MIP coupling. All the plasma emission detectors offer a multi-element facility and long linear ranges which makes them attractive as GC detectors.Unfortunately the ICP and to a lesser extent the DCP involve high capital investment and high operating costs so that coupling of these detectors to GC may not prove cost effective to any but the largest laboratories. Atomic absorption detectors although having short work-ing ranges offer adequate sensitivity for trace metal specia-tion work especially if electrothermal atomisation or atom traps are used. It has been shown that the use of simple ceramic tube traps in conventional flames offers low levels of detection. It is therefore not surprising that most laboratories with a trace metal speciation requirement in which the analytes are a thermally stable volatile organometallic species favour coupled GC - AAS systems. At this time the de-mountability of such couplings is an advantage and single element detection aids the elucidation of unequivocal analy-tical information from highly complex samples.The availa-bility of couplings as commercial accessories from instrument suppliers would receive a ready response from many users. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. References Ebdon L. “An Introduction to Atomic Absorption Spectro-scopy,” Heyden London 1982. Florence T. M. Talanta 1982 29 345. de Mora S. J. and Harrison R. M. Water Res. 1983,17,723. Brewer P. G. and Hao W. M. in Jenne E. A. Editor, “Chemical Modelling in Aqueous Systems,” ACS Symposium Series No. 93 American Chemical Society Washington DC, 1979 p. 261.Brezonik P. L. in Rubin A. J. Editor “Aqueous Environ-mental Chemistry of Metals,” Ann Arbor Science Publishers, Ann Arbor MI 1974 pp. 167-191. van den Berg C. M. G. Anal. Proc. 1984 21 359. Fernandez F. J. At. Absorpt. Newsl. 1977 16 33. Van Loon J. C. Can. J. Spectrosc. 1981,26 22A. De Jonghe W. R. A, and Adanis F. C. Talanta 1982 29, 1057. Byckhovskii M. Ya. 1. Anal. Chern. (USSR) 1983 12 38. de Mora S. J. Hewitt C. N. and Harrison R. M. Anal. Proc. 1984 21 415. Harrison R. M. Hewitt C. N. and de Mora S. J. Trends Anal. Chem. 1985,4 8. Braman R. S. and Tompkins M. A. Anal. Chem. 1978,50, 1088. De Jonghe W. R. A. Chakraborti D. and Adams F. C., Anal. Chem. 1981 15 421. Reamer D. C. Zoller W. H. and O’Haver T. C. Anal. Chem. 1978,50 1449. Radziuk B.Thomassen Y. Van Loon J. C. and Chau, Y. K. Anal. Chim. Acta 1979 105 255. McCormack A. J. Tong S. C. and Cooke W. D. Anal. Chem. 1965,37 147. Bache C. A. and Lisk D. J. Anal. Chem. 1965,37 1447. Bache C. A. and Lisk D. J. Anal. Chem. 1966,38 783. Bache C. A. and Lisk D. J. Anal. Chem. 1966 38 1756 ANALYST OCTOBER 1986 VOL. 111 1137 21. Bache C. A. and Lisk D. J. Anal. Chem. 1967 39 786. 22. Moye H. A. Anal. Chem. 1967 39 1441. 23. Bache C. A. and Lisk D. J. J. Gas Chromatogr. 1968 6, 301. 24. Dagnall R. M. Pratt S. J. West T. S . and Deans D. R., Talanta 1969 16 797. 25. Dagnall R. M. Pratt S. J. West T. S. and Deans D. R., Talanta 1970 17 1009. 26. Bache C. A. and Lisk D. J. J. Assoc. Ofl. Anal. Chem., 1967,50 1246. 27. Westoo G. Acta Chem.Scand. 1967 21 1790. 28. Westoo G. Acta Chem. Scand. 1968,22,2277. 29. Bache C. A. and Lisk D. J. Anal. Chem. 1971 43 950. 30. Dagnall R. M. West T. S. and Whitehead P. Anal. Chim. Acta 1972 60 25. 31. Dagnall R. M. West T. S . and Whitehead P. Anal. Chem., 1972,44,2075. 32. Grossman W. E. L. Eng. J. and Tong Y. C. Anal. Chim. Acta 1972 60 447. 33. Lowings B. J. Analusis 1972 1 510. 34. Dagnall R. M. West T. S . and Whitehead P.,Analyst 1973, 98 647. 35. Dagnall R. M. Johnson D. J. and West T. S. Spectrosc. Lett. 1973 6 87. 36. Kawaguchi H. Sakamoto T. and Mizuike A. Talanta 1973, 20 321. 37. McLean W. R. Stanton D. L. and Penketh G. E. Analyst, 1973 98 432. 38. Serravello F. A. and Risby T. H. J. Chromatogr. Sci. 1974, 12 585. 39. Olsen K.B. Sktarew D. S. and Evan J. C. Spectrochim. Acta Part B 1985 40 357. 40. Talmi Y. and Andren A. W. Anal. Chem. 1974 46 2122. 41. Jones D. R. Tung H. C. and Manahan S. E. Anal. Chem., 1976 48 7. 42. Luckow V. and Russel H. A. J . Chromatogr. 1977 138, 381. 43. Talmi Y. Anal. Chim. Acta 1975 74 107. 44. Talmi Y. and Norvell V. E. Anal. Chem. 1975 47 1510. 45. Talmi Y. and Bostick D. T. Anal. Chem. 1975,47,2145. 46. Black M. S. and Sievers R. E. Anal. Chem. 1976,48,1872. 47. Houpt P. M.,Anal. Chim. Acta 1976 86 129. 48. Sakomoto T. Kawaguchi H. and Mizuike A. J. Chroma-togr. 1976 121 383. 49. Talmi Y. and Norvell V. E. Anal. Chim. Acta 1976 85, 203. 50. Beenakker C. I. M. Spectrochim. Acta Part B 1977,32,173. 51. Bostick D. T. and Talmi Y . J. Chromatogr.Sci. 1977 15, 164. 52. Van Dalen J. P. J. de Lezene P. A. and de Galan L. Anal. Chim. Acta 1977 94 1. 53. Bonnekessel J. and Klier M. Anal. Chim. Acta 1978 103, 29. 54. Brenner K. S. J . Chromatogr. 1978 167 365. 55. Fricke F. L. Robbins W. D. and Caruso J. A. J . Assoc. Ofl. Anal. Chem. 1978 61 1118. 56. Quimby B. D. Uden P. C. and Barnes R. M. Anal. Chem., 1978 50,2112. 57. Schwarz F. P. Anal. Chem. 1978,50 1006. 58. Schwarz F. P. Braun W. and Wasik S. P. Anal. Chem., 1978,50 1903. 59. Hobbs J. S. Mannan C. A. and Beeston B. E. P. Am. Lab. 1979 11 43. 60. Hobbs J. Eur. Spectrosc. News 1979 26 40. 61. Mulligan K. J. Hahn M. H. Caruso J. A. andFricke F. L., Anal. Chem. 1979 51 1935. 62. Quimby B. D. Delaney M. F. Uden P. C. and Barnes, R.M. Anal. Chem. 1979 51 875. 63. Robbins W. B. Caruso J. A. and Fricke F. L. Analyst, 1979 104,35. 64. Robbins W. B. and Caruso J. A. J. Chromatogr. Sci. 1979, 17 360. 65. Schwarz F. P. Anal. Chem. 1979 51 875. 66. Estes S. A. Uden P. C. Rauch M. D. and Barnes R. M., J . High Resolut. Chromatogr. Chromatogr. Commun. 1980,3; 471. Estes S. A. Poirier C. A. Uden P. C. and Barnes R. M., J. Chromatogr. 1980 196,265. 67. 68. 69. 70. Mulligan K. J. Caruso J. A. and Fricke F. L. Analyst 1980, 105 1060. Moisan M. Leprince P. Beaudry C. Bloyet E. US Pat., 4049940 1977. Hanie T. Coulombe S. Moisan M. and Hubert J. in Barnes R. M. Editor “Developments in Atomic Plasma Spectrochemical Analysis,” Heyden London 1981 pp. 337-344. Quimby B. D. Delaney M.F. Uden P. C. and Barnes, R. M. Anal. Chem. 1980 52 259. Reamer D. PhD Thesis University of Maryland College Park MD 1975. Reamer D. C. and Zollar W. H. Science 1980 208,500. Wasik S. P. and Schwartz F. P. J . Chromatogr. Sci. 1980, 18 660. Beenakker C. I. M. Boumans P. W. J. M. and Rommers, P. J. GITFach. Lab. 1981 25 1979. Dingjan H. A. and De Jong H. J. Spectrochim. Acta Part B , 1981 36 325. Estes S. A. Uden P. C. and Barnes R. M. Anal. Chem., 1981 53 1336. Sarto L. G. Jr. Estes S. A. Uden P. C. Siggia S . and Barnes R. M. Anal. Lett. 1981 14 205. Tanabe K. Haraguchi H. andFuwa K. Spectrochim. Acta, Part B 1981 36 633. Uden P. C. Anal. Proc. 1981 18 181. Chiba K. Yoshida K. Tanube K. Ozuki M. Haraguchi, H. Winefordner J. D. and Fuwa K. Anal.Chem. 1982,54, 761. Estes S. A. Uden P. C. and Barnes R. M. J. Chromatogr., 1982 239 181. Eckhoff M. A. Ridgway T. H. and Caruso J. A. Anal. Chem. 1983,55 1004. Cerbus C. S. and Gluck S. J. Spectrochim. Acta Part B , 1983 38 387. Delaney M. F. and Vincent Warren F. Spectrochim. Acta, Part B 1983 38 399. Mulligan K. J. Zerezhgi M. and Caruso J. A. Spectrochim. Acta Part B 1983 38 369. Hagen D. F. Belisle D. F. and Marhevka J. S . Spectro-chim. Acta Part B 1983 38 377. Gardiner P. E. and Bratter P. Spectrochim. Acta Part B, 1983 38 427. Riska G. D. Estes S. A. Beyer J. O. and Uden P. C., Spectrochim. Acta Part B 1983,38 407. Limenentani G. B. and Uden P. C. J. Chromatogr. 1985, 325 53. Slatkavitz K. J. Uden P. C. Hoey L. D. and Barnes R. M., J.Chromatogr. 1984 302 277. Slatkavitz K. J. Hoey L. D. Uden P. C. Barnes R. M., Anal. Chem. 1985 57 1846. Zeng K. Ov. Q. Wang G. and Yu W. Spectrochim Acta, Part B 1985 40 357. Haas D. L. and Caruso J. A. Anal. Chem. 1985 57 846. Bruce M. L. and Caruso J. A. Appl. Spectrosc. 1985 39, 942. Windsor D. L. and Denton M. B. Appl. Spectrosc. 1978, 32 366. Windsor D. L. and Denton M. B. J. Chromatogr. Sci. 1979, 17 492. Windsor D. L. and Denton M. B. Anal. Chem. 1979 51, 1116. Sommer D. and Ohls K. Fresenius Z. Anal. Chem. 1979, 295 337. 100. Ohls K. and Sommer D. in Barnes R. M. Editor, “Developments in Atomic Plasma Spectrochemical Analysis,” Heyden London 1981 pp. 321-326. Brown R. M. and Fry R. C. Anal. Chem. 1981 53,532. Fry R. C. Northway S.J. Brown R. M. and Hughes S. K., Anal. Chem. 1980 52 1716. Eckhoff M. A. McCarthy J. P. and Caruso J. A. Anal. Chem. 1982 54 165. McCarthy J. P. Jackson M. E. Ridgway T. H. and Caruso, J. A. Anal. Chem. 1981,53 1512. Ibrahim M. Nisameneepong W. Hass D. L. and Caruso, J. A. Spectrochim. Acta Part B 1985 40 367. Aulis R. Bolton A. Doherty W. Vander Voet A. and Wong P. Spectrochim. Acta Part B 1985 40 377. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 101. 102. 103. 104. 105. 106 ANALYST OCTOBER 1986 VOL. 111 1138 J. Environ. Sci. Health 1979 14A,’65. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117.118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. Uden P. C. Barnes R. M. and DiSanzo F. P. Anal. Chem., 1978 50 852. Lloyd R. J. Barnes R. M. Uden P. C. andElliott W. G., Anal. Chem. 1978,50 2025. Coker D. T. Anal. Chem. 1975,47 386. Uden P. C. Henderson D. E. Disanzo F. P. Lloyd R. J., and Tetu T. J . Chromatogr. 1980 196 403. Treybig D. S . and Ellebracht S. R. Anal. Chem. 1980,52, 1633. Kolb B. Kemmner G. Schleser F. H. and Wiedeking E., Fresensius 2. Anal. Chem. 1966 221 116. Morrow R. W. Dean J. A. Shults W. D. and Guerin, M. R. J. Chromatogr. Sci. 1969 7 572. Hey H. Fresensius 2. Anal. Chem. 1971 256 361. Gonzalez J. G. and Ross R.T. Anal. Lett. 1972 5 683. Dressman R. C. J. Chromatogr. Sci. 1972 10 472. Longbottom J. E. Anal. Chem. 1972 44 1111. Longbottom J. E. and Dressman R. C. Chromatogr. Newsl., 1973 2 17. Blair W. Inverson W. P. and Brinkman F. E . Chemo-sphere 1974 4 167. Katou T. and Nakagawa R. Bull. Inst. Environ. Sci. Technol. 1974 1 19. Segar D. A, Anal. Lett. 1974 7 89. Chau Y. K. Wong P. T. S. and Goulden P. D. Anal. Chem. 1975,47 2279. Wong P. T. S. Chau Y. K. and Luxon P. L. Nature (London) 1975,253 263. Chau Y. K. Wong. P. T. S. and Saitoh H. J . Chromatogr., Sci. 1976 14 162. Chau Y. K. Wong P. T. S . and Goulden P. D. Anal. Chim. Acta 1976,85 421. Chau Y. K. and Wong P. T. S . in Ewing G. W. Editor, “Environmental Analysis,” Academic Press New York 1977, Van Loon J.C. and Radziuk B. Can. J. Spectrosc. 1976,21, 46. Radziuk B. and Van Loon J. C. Sci. Total Environ. 1976,6, 251. Van Loon J. C. Radziuk B. Kahn N. Lichwa J., Fernandez F. J. and Kerber J. D. At. Absorpt. Newsl. 1977, 16 79. Wolf W. R. Anal. Chem. 1976,48 1717. Chau Y. K. Wong P. T. S. and Goulden P. D. in Branica, M. and Konrad Z. Editors “Lead in the Marine Environ-ment,” Pergamon Press Oxford 1980 pp. 72-82. Andreae M. O. Anal. Chem. 1977 49 820. Grimm P. Libs S . Cressely J. and Deluzarche A. Ann. Falsif. Expert. Chim. 1977 70 523. Parris G. E. Blair W. R. and Brinckman F. E. Anal. Chem. 1977,49 378. Robinson J. W. Kiesel E. L. Goodbread J. P. Bliss R. and Marshall R. Anal. Chim. Acta. 1977 92 321. Wolf W. R. J. Chromatogr. 1977 134 159. Bye R. Paus P. E. Solberg R. and Thomassen Y. At. Absorpt. Newsl. 1978 17 131. Bye R. and Paus P. E. Anal. Chim Acta. 1979 107 169. Chau Y. K. Wong P. T. S . Bengert G. A. and Kramar O., Anal. Chem. 1979 51 186. Radziuk B. Thomassen Y. Butler L. R. P. Van Loon, J. C. and Chau Y. K. Anal. Chim. Acta 1979 108 31. Robinson. J. W Kiesel. E. L and Rhodes. I. A. L., pp. 215-225. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. Chau Y. K. Wong P. T. S . Kramar O. Bengert G. A , Cruz R. B. Kinrade J. O. Lye J. and Van Loon J. C. Bull. Environ. Contam. Toxicol. 1980 24 265. Coe M. Cruz R. and Van Loon J. C. Anal. Chim. Acta, 1980 120 171. Van Loon J. C. Anal. Chem. 1979 51 1139A. Hahn M. H. Mulligan K. J. Jackson M. E. and Caruso, J. A. Anal. Chim. Acta 1980 118 115. De Jonghe W. Chakraborti D. and Adams F. Anal. Chim. Acta 1980 115 89. De Jonghe W. R. A. Chakraborti D. and Adams F. C., Anal. Chem. 1980 52 1974. Burns D. T. Glockling F. and Harriott M. Analyst 1981, 106 921. Chakraborti D. Jiange S. G. Surkign P. De Jonghe W., and Adams F. Anal. Proc. 1981 18 347. Chan L. Forensic Sci. Znt. 1981 18 57. Ebdon L. Ward R. W. and Leathard D. A. Analyst 1982, 107 129. Ebdon L. Ward R. W. and Leathard D. A. Anal. Proc., 1982 19 110. Chakraborti D. De Jonghe W. R. A. Van Mol W. E., Van Cleuvenbergen C. and Adams F. C. Anal. Chem., 1984 56 2692. Hewitt C. N. and Harrison R. M. Anal. Chim. Acta 1985, 167 277. Harrison R. M. and Radojevic M. Environ. Technol. Lett., 1985 6 129. Chau Y. K. Wong P. T. S . Dramar O. andBengert G. A., in “Proceedings of the 3rd International Conference on Heavy Metals in the Environment September 1981 Amsterdam,” CEP Consultants Edinburgh 1981 p. 641-643. Baker M. D. Wong P. T. S . Chau Y. K. Mayfield C. I . , and Innis W. E. in “Proceedings of the 3rd International Conference on Heavy Metals in the Environment September 1981 Amsterdam,” CEP Consultants Edinburgh 1981, Ahmad I. Chau Y. K. Wong P. T. S. Carty A. J. and Taylor L. Nature (London) 1980 287 716. De Jonghe W. R. A. Chakraborti D. and Adams F. C. in “Proceedings of the 2nd International Conference on Heavy Metals in the Environment Amsterdam September 1980.” Thompson J. A. J. in Proceedings of the 3rd International Conference on Heavy Metals in the Environment September 1981 Amsterdam,” CEP Consultants Edinburgh 1981, p. 653-656. Brueggemeyer T. M. and Caruso J. A. Anal. Chem. 1982, 54 872. Chau Y. K. Wong P. T. S . and Bengert G. A. Anal. Chem. 1982 54,246. Ohta K. Smith B. W. and Winefordner J. D. Anal. Chem., 1982,54 320. Forsyth D. S. and Marshall W. D. Anal. Chem. 1985 57, 1299. Cruz R. B. Loronso C. George S . Thomassen Y., Kinrade J. D. Lye J. and Van Loon J. C. Spectrochim. Acta Part B 1980 35 775. Van Loon J. C. At. Absorpt. Newsl. 1976 15 72. pp. 645-647. Paper A61100 Received March 25th I986 Accepted May 16th 198