|
11. |
Analyte volatilization procedure for the determination of low concentrations of iodine by inductively coupled plasma atomic emission spectrometry. Invited lecture |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 159-165
Taketoshi Nakahara,
Preview
|
PDF (759KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 159 Analyte Volatilization Procedure for the Determination of Low Concentrations of Iodine by Inductively Coupled Plasma Atomic Emission Spectrometry* Invited Lecture Taketoshi Nakahara and Toshio Mori Department of Applied Chemistry University of Osaka Prefecture Sakai Osaka 593 Japan A simple method is described for the generation of a continuous flow of volatile iodine by the oxidation of aqueous iodide for the determination of low concentrations of iodine by inductively coupled plasma (ICP) atomic emission spectrometry in the normal ultraviolet and vacuum ultraviolet (VUV) regions of the spectrum. For measuring spectral lines in the VUV region the monochromator and the enclosed external optical path between the ICP source and the entrance slit of the monochromator were both purged with nitrogen to minimize light absorption by atmospheric oxygen.The iodine atom emission lines at 178.28 183.04 and 206.1 6 nm were selected as the analytical lines of interest. Of the various oxidation reactions investigated an oxidizing solution of 5.0 mmol I-’ of sodium nitrite in 8.0 mol I-’ sulfuric acid was found to be the most appropriate for the generation of elemental iodine. The gaseous iodine is separated from the solution in a simple gas-liquid separator and swept into the argon stream of an ICP for analysis. The best attainable detection limits (30 criterion) for iodine at 178.28 183.04 and 206.16 nm were found to be 0.39 0.55 and 2.1 ng ml -’ respectively. Typical calibration graphs obtained under the optimized experimental conditions are rectilinear over approximately four orders of magnitude of concentration.The present method has successfully been applied to the determination of total iodine (i.e. iodide + iodate) in several brine samples. Keywords Inductively coupled plasma atomic emission spectrometry; iodine determination; analyte volatiliz- ation; gas-phase sample introduction; brine sample In natural environments iodine is known to be widely distrib- uted. However the amount of iodine contained in natural environmental materials such as natural water rocks and vegetables is often at trace concentration levels,’-4 and sufficient sensitivity is required for the determination of iodine at such levels. Inductively coupled plasma atomic emission spectrometry (ICP-AES) is the current method of choice for the determi- nation of over 70 elements.However the direct determination of non-metals such as iodine has to date been quite limited.5 A major reason for the difficulties encountered is that the sensitive resonance lines for most non-metals lie in the vacuum ultraviolet (VUV) and/or near-infrared spectral regions making their determination by AES difficult. Also the high excitation energies of non-metals cause excited-state popu- lations to be low. Recently for example Nakahara and Wasa6 reported the direct determination of iodine in aqueous solu- tions with an argon ICP where typical detection limits for iodine were 0.092 0.088 and 0.559 pg ml-’ at 178.28 183.04 and 206.16 nm respectively.In addition the same workers described a prior-oxidation procedure6 which improved the detection limits by a factor of approximately 40 at both 183.04 and 206.16 nm. One of the most important features in any sample introduc- tion system is transport efficiency. For normal pneumatic nebulizers this efficiency is l-2%.7 If the analyte is present in solution in a volatile form analyte transport efficiency could approach 100% and hence the detection limit would be improved by a corresponding factor. However most non-metal microwave induced plasma (MIP) AES has been commonly performed with a gas-phase sample introduction mode for a relatively low power (<200 W) MIP. This gas-phase sample introduction mode includes indirect determination of iodine based on the cold-vapour method for mercury,8 electrothermal vaporization ( ETV)9 and analyte volatilization On the other hand little work has been reported on the * Presented at the XXVIII Colloquium Spectroscopicurn Internationale (CSI) York UK June 29-July 4 1993.application of this analyte volatilization method in ICP- AES.13,14 Thus in this study the on-line generation of volatile iodine by the oxidation of aqueous iodide was examined for argon ICP-AES. Iodine emission lines were observed in the UV and VUV spectral regions. A nitrogen-purged optical system was used for the observation of the VUV spectra.” The optimiz- ation and analytical performance of the method and the application of the proposed technique to the determination of total iodine (ie. iodide +iodate) in several brine samples are described below.Experimental Instrumentation and Apparatus A Nippon Jarrell-Ash Model ICAP-SOSM inductively coupled plasma emission spectrometer was used in combination with a gas-phase sample introduction system. Specifications of the instrumentation used in this work are given in Table 1 and the complete analytical system is schematically represented in Fig. 1. A detailed description of the gas-liquid separator made of borosilicate glass is given in Fig. 2. The internal mono- chromator path and the enclosed external optical path between the ICP source and the monochromator were both purged with nitrogen to reduce light absorption by atmospheric oxygen. The detailed schematic diagram of this simple nitrogen- purged optical system has been described previously by Nakahara.” With the use of this system the VUV spectrum between 170 and 190 nm could be observed. Reagents All the reagents used in this study were of analytical-reagent grade.High-purity water (Milli-Q water) was obtained by passing distilled water through a Milli-Q ion-exchange and membrane-filtering system (Millipore). A standard stock solution of iodine (1000pgml-’) was prepared daily by dissolving potassium iodide in Milli-Q water.160 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 1 Experimental instrumentation Component Description Generator (r,f.) Crystal-controlled type; 27.12 MHz; automatic power control; automatic impedance matching network; maximum output power 2.0 kW All quartz; Fassel type Argon for coolant plasma and carrier gases Ebert 0.5 m mounting; 1180 lines mm-' grating blazed at 240 nm; reciprocal linear dispersion (first order) 1.6 nm mm-' Plasma source focussed as 1 1 image onto an entrance-slit with a 60 mm focal length Suprasil lens D.c.amplification/lO s integration; digital voltmeter Rikadenki Model R-21 (1.0 V full scale) Plasma torch Gas for ICP Nebulizer Pneumatic cross-flow type Spectrometer Optics Photomultiplier Hamamatsu Photonics R-106UH Signal measurement Chart recorder Peristaltic pump Tokyo Rikakikai MP-3 Y Y Y F C A B Fig. 1 Schematic diagram of the ICP-AES system with gas-phase sample introduction for the continuous-flow determination of iodine A sample solution; B oxidizing solution; C peristaltic pump; D gas-liquid separator; E liquid waste; F carrier gas; G elemental iodine +carrier gas; H coolant gas; I plasma gas; J gas-flow controller; K argon tank; L plasma torch; M ICP; N nebulizer chamber; 0 monochromator; P photomultiplier; Q high voltage supply; R amplifier; S digital integrator; T chart recorder; U tuning and coupling; V r.f.power generator; W lens; X light cell; and Y nitrogen purge gas Sample Oxidizing solution solution ___._I) ICP Waste 5 cm Fig. 2 Gas-liquid separator for continuous-flow generation of elemental iodine From this standard solution working standard solutions were prepared immediately before use by serial dilution. Oxidizing solutions sodium nitrite potassium persulfate hydrogen peroxide potassium permanganate potassium dichromate and potassium perbromate were prepared acidified with sulfuric acid and used for the continuous-flow generation of elemental iodine from iodide.Argon (99.99% pure) served in a dual role i.e. as a plasma- sustaining gas for the ICP and as a carrier gas for introduction into the ICP of the iodine vapour that was generated. Nitrogen (99.99% pure) was used as a purge gas for measuring the VUV lines of iodine. Procedure A previously acidified oxidizing solution and a sample solution containing iodine as iodide were conveyed continuously into the iodine generator (Fig. 2). The chemically evolved iodine vapour was separated from the reaction solution and swept into the plasma by a continuous flow of argon carrier gas while the resulting solution was drained to waste. The iodine emission lines in the UV and VUV spectral regions were monitored.All emission intensities when stabilized were inte- grated at least in triplicate over 10 s. For off-line background correction at the same wavelength as the analyte the back- ground emission (which included any emission from the reagent blank) integrated over the same period of time (10s) was subtracted from all the emission data. Under the standard operating conditions the iodine emission signal begins to rise from the baseline ,within 20s of the sample (or standard) solution being pumped in and it reaches a plateau after about 30 s. Furthermore the tailing edge of the signal returns to the baseline well within 20 s of the sample (or standard) solutions being exchanged for Milli-Q water. Six methods for chemical vapour generation of iodine were investigated; the oxidation reactions along with the standard oxidation-reduction potentia1sl6 are shown in Table 2. The influence of the six oxidizing agents shown in Table 2 at 0.005-1000 mmol l-.' and acidities of 1.0-8.0 moll-' in sulfuric acid on the iodine emission intensity at 178.28 183.04 and 206.16 nm was briefly examined.These parameters in addition to other experimental conditions were optimized by a univari- ate search technique with the use of 0.2 1.0 and 5.0 pg ml-' iodide solutions at 178.28 183.04 and 206.16 nm respectively unless otherwise stated. Results and Discussion Optimization of Experimental Parameters In a preliminary study the VUV emission spectrum in the 175-186nm range was obtained from an iodine solution of 0.2 pg ml-' witlh the use of the nitrogen-purged optical system developed previou~ly'~ (see Fig.3). For comparison the normal UV spectrum at 205-207 nm was measured under the same conditions and is depicted in Fig. 3 also. In an attempt to obtain a maximum line-to-background intensity ratio In/Ib (where I is the background-corrected net analyte intensity and Ib is the background emission intensity) Table 2 Iodine generation reactions with various oxidizing agents ~~~~~ Oxidizing agent Iodine generation reaction K2SzOs (+ 2.123)* H,O (+ 1.776) KMnO (+ 1.507) KBrO (+ 1.423) K,Cr20 (+ 1.232) S20s2- + 21- + 2Hf = I 2 + 2HSO,- H,O + 21- + 2Hf = I + 2H20 2Mn0,- + 101- + 16H+ = 2Br03- + 101- + 12H' = Cr,'07'- +6I- + 14H' = 51,+ 2Md' + 8H20 51 + Br + 6H20 31 + 2Cr3+ + 7H,O NaNO (+ 0.938) 2N02- + 21- +4Ht =I,+ 2N0 + 2H,O *Standard oxidation-reduction potentials ( V versus a normal hydro- gen electrode) for each oxidizing agent. Standard oxidation-reduction potential of iodine ( I + 2e- = 21-) is + 0.536 V.All potential values were taken from ref. 16JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 161 C Baseline bA 207 205 185 180 175 Wavelengthhm Fig.3 ICP emission spectral regions of 175-185 and 205-207 nm obtained for a 0.2 pg ml- ' iodide solution by continuous-flow iodine generation. A S I 180.73; B S I 182.04; C S I 182.62; D I I 206.16; E I I 184.44; F I I 183.04; G I I 179.91; and H I I 178.28 nm for iodine various operating parameters were examined and optimized individually while the other parameters were kept at their optimum values.The parameters investigated were r.f. power analytical wavelength observation position in the ICP source argon flow rates of coolant plasma and sample gases and nitrogen purge gas flow rates. The selection of the most sensitive lines for iodine determi- nations was achieved by the comparison of major iodine lines shown in Fig. 3. The I IJIb and background equivalent concentration (BEC) obtained for major lines of iodine after gas-phase sample introduction or conventional solution nebul- ization are given in Table 3. The BEC is defined as the concentration of the analyte expressed in pgml-' of iodine that yields a net intensity signal equal to the intensity of the background. As a consequence three major iodine lines I I 178.28 I I 183.04 and I I 206.16 nm were the analytical lines selected and the optimized operating conditions at these three wavelengths are summarized in Table 4.The ICP discharge was stable even at lower r.f. power (<l.OkW) probably because the perturbation of the plasma by excessive water vapour aerosol could be avoided in the present gas-phase sample introduction method. Unknown emission lines marked with A B and C in Fig. 3 were found to be S I 180.73 S I 182.04 and S I 182.62 nm Table 3 Emission characteristics of major iodine (I I) lines Table 4 iodine Optimized operating conditions for the determination of Parameter I 178.28 I 183.04 1206.14 R.f. forward power/W 0.9 0.9 1.2 Carrier gas flow rate/l min-' 1.1 1.2 1.2 Coolant gas flow rate/l min-' 17.0 18.0 18.0 Plasma gas flow rate/l min - ' 0.3 0.3 0.8 Observation height/nm 17.0 17.0 18.0 Slit-height/mm 5.0 5.0 2.0 Purge gas flow rates/l min-' Light cell 6.0 6.0 0 Monochromator 6.0 6.0 0 Sample solution flow rate/ml min-' 18.0 18.0 18.0 Oxidizing solution flow rate/ml min-' 9.0 9.0 9.0 (above load coil) respectively.These emission lines arose as a result of a small amount of sulfuric acid vapour being conveyed together with the generated elemental iodine. This was confirmed by the fact that no appreciable emission lines of sulfur were observed when conventional solution nebulization of a standard solution of much higher iodine concentration (ie. 25.0 pg ml-' of iodide) was carried out. Analyte Volatilization Reactions For sodium nitrite the concentration dependencies of the oxidizing agent and sulfuric acid on background-corrected emission intensity and line-to-background intensity ratio at the I 1 183.04nm line are shown in Fig.4.The optimum concentration and acidity of each of the six oxidizing solutions tested (see Table 5) were selected taking into consideration the emission characteristics of the iodine lines as well as the lifetime of the tube conveying the acidic solution. It can be seen from Table 5 that except for sodium nitrite the optimized concen- trations for the oxidants are to some extent different depending upon the iodine 178.28 183.04 and 206.16 nm lines; the reason for these differences is not known at present. Furthermore the use of these oxidizing solutions did not appear to produce any significant cumulative memory effects.Analytical Calibration Graphs and Detection Limits Under the optimized experimental conditions listed in Tables 4 and 5 double logarithmic analytical calibration graphs were obtained for iodine at I I 178.28 183.04 and 206.16 nm by the six analyte volatilization reactions. All analytical working curves showed relatively large dynamic ranges whereas those achieved with potassium dichroma te potassium permanganate and potassium perbromate gave somewhat shorter dynamic ranges. Typical calibration curves obtained with the other three oxidants H202 NaNO and K2S,0 are shown in Fig. 5. Each point on the plots was determined from at least five replicate measurements. The lowest point of each graph gener- Iodine generation with 0.2 pg ml-' of iodine Wavelength/nm In* zn/lbf BEC/pg ml-' 178.28 1 .oo 1 .oo 0.01 179.91 0.26 0.16 0.07 183.04 0.88 0.27 0.04 184.44 0.12 0.03 0.34 206.16 0.58 0.04 0.29 Solution nebulization with 25.0 pg ml-' of iodine In$ In/Ib§ BEC/pg ml ~ 0.74 1 .oo 2.4 0.26 0.21 11.4 1 .oo 0.32 7.6 0.17 0.05 53.2 0.67 0.05 45.5 *In is relative to the intensity of the 178.28 nm line. tlJlb is relative to the intensity ratio for the 178.28 nm line.$1 is relative to the intensity of the 183.04 nm line. $ I J I is relative to the intensity ratio for the 178.28 nm line.162 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 20 -= 10 0 ( C - B A I I I I I I I 20 ,= -= 10 -. 0 I I 1 1 1 1 0.1 0.5 1 5 10 50 ~NaNO,l/mmol I-' Fig.4 Dependence of (a) I and (b) ZJIb on the concentration of oxidizing agent (NaNO,) and sulfuric acid using the I 1 183.04nm line.Concentration of H,S04 A 3.0; B 5.0; and C 8.0 moll-'. Table 5 Optimized concentration and acidity of oxidizing solutions Oxidizing agent acidified with H,S04 H,O,/mmol 1-' H,SO,/mol 1 -' NaNO,/mmoll-' HzS04/mol 1- ' K,SzO,/mmol 1 - H,S04/mol 1- ' KBrO,/mmoll-' H,SO,/moll- ' K,Cr,O,/mmoll- ' H,SO,/mol 1-' KMnO,/mmoll-' H,SO,/mol I-' Wavelength/nm 178.28 10 8 5 8 5 8 10 1 1 8 0.01 3 183.04 100 8 5 8 10 8 50 1 10 8 0.1 8 206.1 6 50 8 5 8 50 8 5 3 5 8 0.1 8 ally corresponds to the determinable concentration limit of the analyte. The lowest concentration range was limited mostly by spectral background noise and by little or no significant reagent blank. Moreover it was considered that the upper concentration limit would be set by the chemistry of iodine generation and transport efficiency rather than the ICP detec- tion system.The difference in linear dynamic range between the iodine 178.28 183.04 and 206.16 nm lines can be attributed to the differences in the nature of these three lines. Relatively poor linearity at I I 178.28 nm has also been pointed out in the study with MIP-AES in this laboratory.'' The calibration graphs obtained at these three wavelengths show good linearity with correlation coefficients of better than 0.999 as illustrated in Fig. 5. Detection limits for iodine were extrapolated from the linear calibration graphs (Fig. 5 ) and are defined as the con- centration of analyte that would produce a net signal (Le. background-corrected line intensity) equal to three times the standard deviation of the background emission intensity in accordance with IUP'AC recommendation^.'^ The detection limits for the iodine lines studied using the three iodine generation methods are summarized in Table 6 where the results achieved in the present work are compared with those \obtained by the conventional solution nebulization method.6 'The detection limits were improved by two or three orders of magnitude. As a result in this study the best attainable limit of detection and the smallest BEC for iodine were achieved by using the iodine 178.28 nm line in combination with iodine generation by use of hydrogen peroxide as an oxiaant. In addition the best attainable limits of detection achieved here are approximately an order of magnitude in concentration better than that already reported in the 1iterat~re.l~ The instrumental precisions expressed as relative standard deviations (RSDs) were obtained at the three wavelengths and by the three methods for iodine generation with the use of different concentrations of iodine as iodide and are shown in Table 7.In conclusion taking the optimized oxidant concentrations (Table 5 ) and some analytical figures of merit into consider- ation an oxidizing solution of 5.0 mmol 1-' of sodium nitrite in 8.0 moll-' of sulfuric acid was found to be the most suitable of the different oxidation reactions for the generation of elemental iodine and was used subsequently. In addition to an oxidizing solution the use of the I 1 178.28 nm line was discarded for the remainder of this work because of its smaller linear dynamic range (Fig.5 ) . Interference Study Using the iodine lines at 183.04 and 206.16 nm with continu- ous-flow gas-phase sample introduction a study of various concomitant ions at concentrations 1 10 100 and 1000-fold greater than iodine (0.5 pg ml-' of iodide) was carried out. Several foreign species (i.e. cations and anions) at different concentrations were selected taking the constituents of brines into account. The results of the effect of some foreign ions interfering with the emission intensity of iodide at 183.04 and 206.16 nm by the proposed procedure are shown in Tables 8 and 9. Interference is considered to have occurred when the change in emission intensity is greater than f 3 % from that for iodine alone.The following ions did not interfere SO4,- NO,- C1- F- NH4+ and C204*-. And the following elements did not interfere Al As B Ba Be Bi Ca Cd Cs Cu Ge In K La Li Mg Mn Mo Na Ni P Pb Rb Sb Se Sr Th T1 W Y and Zr. Relative intensity is defined in this context as the ratio of the iodine emission intensity obtained in the presence of the foreign ion to that obtained when no foreign ion was present in the analyte solution. The presence of C032- and HC03- in particular caused an increased background due to copious amounts of simul- taneously evolved carbon dioxide which in the ICP source produced CO molecular band emission'' observed at 206.77 and 184.15 nm in the vicinity of the 206.16 and 183.04 nm lines of iodine respectively. As shown in Table 8 a slight enhance- ment in emission intensity of iodine lines in the presence of iodate was produced by oxidation of iodide with iodate as follows; 1 0 3 - +51-+6H+=312+3H20 whereas the presence of bromate and cyanide ions caused a depressing interference according to the following reactions respectively 2Bt-0,- + I = Br + 210,- 2CN- +12=21CN+2e- In addition the great decrease in iodine emission intensity in the presence of Ag Au Hg Pd and Pt ions as shown in Table 9 can be attributed to the formation of their stable and insoluble iodides.Furthermore there was little or no significant difference in the extent of interferences between the 183.04 and 206.16 nm iodine lines.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 163 (a) 108 - ( b ) ( C) A .- c A 10' 103 10 l o I 10 102 lo3 10" 1 10 lo2 lo3 l o A lo5 lo6 10 lo2 lo3 10" 105 106 Iodine concentrationhg ml-' Fig.5 Calibration graphs for iodine with various oxidizing agents A hydrogen peroxide; B sodium nitrite; and C potassium persulfate. (a) I I 178.28; (b) I I 183.04; and (c) I I 206.06 nm Table 6 Detection limits for iodine Wavelength/nm Oxidizing agent I 178.28 NaN02 H202 None* I 183.04 NaN02 H202 None* I 206.16 NaNO K2S208 K2S208 H202 K2S,O* None* Detection limit/ng ml- ' 0.392 0.34 1 0.820 0.553 0.676 1.18 2.12 3.10 11.1 92 88 559 BEC/pg ml-' 0.0 18 0.008 0.015 4.34 0.034 0.028 0.033 7.82 0.337 0.274 0.582 37.9 *Direct solution nebulization (ref. 6 ) . Table 7 Reproducibility* of iodine emission intensity measurements Iodine concentration/pg ml-l I 1 178.28 nm I I 183.04 nm 11 206.16 nm Oxidizing agent 0.1 1 .o 10 0.1 1 .o 10 0.1 1 .o 10 NaNO 0.9 3 0.92 0.51 1.74 1.21 1.12 2.82 1.94 1.82 H202 2.37 1.63 1.61 2.03 1.36 1.56 2.29 1.45 1.23 K23208 1.35 0.94 0.78 1.41 1.22 1.01 2.65 1.07 1.03 *Expressed as relative standard deviation (%); n = 10 Table 8 Effect of foreign ions on iodine emission intensity Species IO - Br- BrO,- ClO,- co,2- HC03- S2 - Compound added Na2S0 KIO KBr KBrO NaC10 Na2C0 NaHC0 Na2S Amount added/ pg ml-' 50 0.5 50 0.5 5 50 50 5 Relative intensity* I1 183.04 nm 103 107 100 99 99 97 102 101 I 1 206.16 nm 103 105 100 97 99 101 97 99 *Relative to 100 for the emission intensity of iodine (0.5 pg ml-l) alone.Applications iodine (iodide and iodate) in brine iodate must be reduced to - - Pre-reduction of iodate to iodide iodide prior to use of the present gas-phase sample introduction method.After optimization of the concentration of ascorbic It is well known that iodine is present in natural waters both as iodide and iodate." Therefore for the determination of total acid as a pre-redu~tant,'~ 0.003% m/v ascorbic acid was added to the sample solution to reduce iodate to iodide.164 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 9 Effect of foreign elements on iodine emission intensity Element Au Ce c o Cr Fe Ga Pd Pt Sn Te Ti V Zn Ag Hg Compound added 4 0 0 3 HAuC1 * 4H20 Ce2(S04)3' 8H20 Metal in HCI Feel * 6H20 Metal in HCI PdCI H2PtCl Metal in HC1 Na,Te03 Metal in HCI Metal in HCI KzCrzO7 NH4VOj Amount added/ pg ml-' 0.5 0.5 50 50 50 5 50 0.5 0.5 0.5 50 50 50 5 50 Relative intensity* I I 183.04 nm 0 31 98 99 101 95 99 0 0 31 100 102 97 98 101 I I 206.16 nm 1 47 98 100 100 99 100 0 0 35 100 100 98 99 101 *Relative to 100 for the emission intensity of iodine (0.5 pg m1-I) alone.Table 10 Determination of iodine (pg m1-l) in brines This work* Standard additions method Calibration graph method Sample 183.04 nm 206.16 nm 183.04 nm 206.16 nm Reference method A 90.8 -t 2.0 87.6 f 2.2 90.9 1.3 91.0k 1.9 90.5t 92.4$ B 70.1 k 2.6 68.7 & 3.5 71.4 & 2.2 70.1 f3.4 70.6t 69.61 C 105.9 f 2.4 104.7 f 3.2 106.0 k 2.7 107.3 2.9 106.5t 107.31 *Mean value f standard deviation (n = 5). ?Determined by ICP-AES with prior-oxidation (ref. 6). $Determined by atmospheric pressure helium MIP-AES with continuous-flow gas-phase sample introduction (ref.10). Determination of total iodine in brines The accuracy and precision of the present continuous-flow gas-phase sample introduction method for ICP-AES were established by analysing several samples of natural water. To this end the procedure described below was followed for the determination of total (iodide and iodate) concentrations in brines taken at Mobara City Chiba Prefecture Japan. The presence of some matrix components would be expected to interfere in the determination of iodine by the proposed method depending upon whether or not the gas-phase sample introduction procedure and the off-peak background correc- tion techniques were used. Recovery studies therefore were made by adding known amounts of iodine as iodide or iodate to diluted samples and applying the recommended procedure with 5 mmoll-' of sodium nitrite in 8.0 moll-' sulfuric acid.Mean recoveries of added iodine were in the range 97.1-104.6% depending on the brine sample being measured. Therefore both the calibration graph method and the method of standard additions were employed for the determination of iodine. A sample was filtered through a No. 2 filter-paper (Toyo Roshi) and the filtrate was analysed for total iodine after spiking with iodine in the range 1-10 pg m1-l. A dilution factor of 50 was used for the analysis of brine samples. The results obtained by the present method for several samples of brines are presented in Table 10. At the same time the total iodine in the samples was determined by ICP-AES with the prior-oxidation procedure6 and quantified by atmospheric- pressure helium MIP-AES with continuous-flow gas-phase sample introduction." It should be noted that there is little or no appreciable difference in the iodine concentration between the 183.04 and 206.16 nm iodine lines and that the results obtained by the present method are in good agreement with those obtained by other techniques.Conclusions A simple analyte volatilization procedure for ICP-AES has been developed and the results presented which allows the continuous-flow determination of iodine at low concentrations. As well as can be established this on-line system with a simple gas-liquid separator for the generation of elemental iodine by oxidation of aqueous iodide has been proposed for the first time and the best attainable limits of detection obtained in this work were approximately an order of magnitude better than those reported in the 1iterat~re.l~ In conclusion the proposed method has been successfully applied to the determination of total iodine (Le.iodide plus iodate) in samples of brines giving RSDs of 1.4-5.1%. The authors thank DP. I. Makino of Asahi Glass for supplying the brine samples investigated in this work. References 1 Niazi S. B. and Mozammi M. Anal. Chim. Acta 1991 252 115. 2 Koh T. Ono M. and Makino I. Analyst 1988 113 945. 3 Ebihara M. Saito N. Akaiwa H. and Tomura K. Anal. Sci. 1992 8 183. 4 Cox R. J. Pickford C. J. and Thompson M. J. Anal. At. Spectrom. 1992 7 635. 5 McGregor D. A. Cull K. B. Gehlhausen J. M. Viscomi A. S. Wu M. Zhang L. and Carnahan J . W. Anal. Chem. 1988 60 1089A. 6 Nakahara T. and Wasa T. Appl. Spectrosc. 1987 41 1238. 7 Browner R. F. and Boorn A. W. Anal. Chem. 1984 56 786A. 8 Nakahara T. and Wasa T. Microchem. J. 1990 41 148. 9 Barnett N. W. and Kirkbright G. F. J. Anal. At. Spectrom. 1986 1 337.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 165 10 Nakahara T. Yamada S. and Wasa T. Appl. Spectrosc. 1990 44 1673. 11 Quintero Ortega M. C. Cotrino Bautista J. Saez M. Menendez Garcia A. Sanchez Uria J. E. and Sanz Medel A Spectrochim. Acta Part B 1992 47 79. 12 Dolores Calzada M. Carmen Quintero M. Gamero A. and Gallego M. Anal. Chem. 1992 64 1374. 13 Dolan S. P. Sinex S. A. Capar S. G. Montaser A. and Clifford R. H. Anal. Chem. 1991 63 2539. 14 Cave M. R. and Green K. A. J. Anal. At. Spectrom. 1989,4 223. 15 Nakahara T. Spectrochim. Acta Part B 1985 40 293. 16 CRC Handbook of Chemistry and Physics ed. Weast R. C. CRC Press Boca Raton 66th edn. 1985 p. D-151. 17 IUPAC Nomenclature Symbols Units and Their Usage in Spectrochemical Analysis Revision 1975 Part 11 Spectrochim. Acta Part B 1978 33 248. 18 Gaydon A. G. The Spectroscopy of Flames 2nd edn. Chapman and Hall London 1974 p. 351. 19 Nakayama E. Kimoto T. and Okazaki S. Anal. Chem. 1985 57 1157. Paper 3/048140 Received August 1 I 1993 Accepted August 31 1993
ISSN:0267-9477
DOI:10.1039/JA9940900159
出版商:RSC
年代:1994
数据来源: RSC
|
12. |
Detection of surface aggregates of trace amounts of copper and silver on graphite using secondary ion mass spectrometry at elevated temperatures. Invited lecture |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 167-170
Jason G. Jackson,
Preview
|
PDF (511KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 167 Detection of Surface Aggregates of Trace Amounts of Copper and Silver on Graphite Using Secondary Ion Mass Spectrometry at Elevated Temperatures" Invited Lecture Jason G. Jackson Rodney W. Fonesca and James A. Holcombet Department of Chemistry and Biochemistry University of Texas at Austin Austin TX 78712 USA Static secondary ion mass spectrometry is used to investigate the extent of dispersion of copper and silver on a pyrolytic graphite coated graphite surface. The appearance of secondary fragment ions consisting of more than one atom of the metal analyte is used as an indication that the sample is present on the surface as aggregates. This signal for the polymeric metal ion is compared with the signal for the monatomic fragment ion for various concentrations.Concentration studies were conducted for both metals and copper was found to exist as dispersed copper species at lower concentrations while at higher concentrations evidence suggests surface aggregates or microdroplets. Silver was found to exist as aggregates for the range of concentrations studied. Keywords Secondary ion mass spectrometry; copper; silver; graphite; surface aggregation The signal in electrothermal atomic absorption spectrometry (ETAAS) is dependent in part on the initial generation of atoms from the surface. If a kinetic model is considered for the release then the activation energy E surface coverage 0 and order of release n are the governing parameters. It has been suggested that the order of release may give some information on the morphology of the desorbing species on the surface during the desorption process.'-4 In vacuum desorp- tion studies Redhead' suggested that the order of release could be deduced from the general shift in the peaks when the surface coverage was altered.McNally and Holcombe2 applied this theory to vaporization processes with a graphite furnace in ETAAS applications. More recently Rojas and Olivares6 and Yan et al.7 presented a means of mathematically treating the absorbance data to permit quantitative determination of a value for n. However all these techniques provide information only at the temperature where desorption occurs and represent indirect evidence of the analyte morphology on the surface. Surface techniques have been used to study metal clusters directly on graphite although some require atomic level smoothness for their success (e.g.scanning tunnelling microsc~py*~~) and others require relatively high surface cover- ages for detection (e.g. Auger spectroscopy" and electron spectroscopy for chemical applications). Unfortunately in ETAAS extremely low analyte concentrations located on microscopically roughened surfaces are of interest It may be possible to overcome both of these limitations using secondary ion mass spectrometry to study trace metals deposited on pyrolytic graphite coated graphite surfaces used in ETAAS. In this study temperature programmed static secondary ion mass spectrometry (TPS-SIMS) was evaluated as a means of differentiating between clusters of analyte and dispersed adsorbed analyte on the surface.Specifically the extent of dispersion of Cu and Ag was investigated using TPS-SIMS both at room temperature and during a thermal ramp to determine if any morphological changes occurred preceding vaporization. Experimental Apparatus and Solutions The TPS-SIMS system has been described previously.12 The primary Ar' ion beam was operated at 1 nA and 2.3 keV * Presented at the XXVIII Colloquium Spectroscopicurn Inter- t To whom correspondence should be addressed. nationale (CSI) York UK June 29-July 4 1993. using a spot size of approximately 100p.m. The beam was rastered across a surface area of approximately 2 x 2 mm. This low primary ion flux and beam energy implies that only the outermost layers of the sample surface are probed and that there is minimal surface perturbation over the analysis time.This ensures operation under 'static' conditions. In most studies solution samples containing the dissolved analyte salt were deposited on a pyrolytic graphite coated graphite platform (10 x 5 x 1 mm Stackpole/UltraCarbon Bay City MI USA). The platform was radiatively heated by a tantalum-strip heater located below the platform. The heating rate was maintained by a proportionally controlled heating circuit with feedback provided from a thermocouple placed within a hole (0.6 mm diameter x 2 mm long) drilled into the graphite platform. All data collection was performed by software written in ASYST in this laboratory running on an 80486 based PC interfaced to a Keithly Series 500 data acquisition and control system.High-purity Ag wire was dissolved in HNO and diluted to prepare a stock solution of 1000 mg 1-1 Ag in 0.5% HNO,. A stock solution of copper nitrate was prepared by dissolving ACS reagent grade CU(NO,)~.~+ H 2 0 in distilled de-ionized water. The working solutions were prepared daily by dilutions of the stock solution. Procedure A 2 p.1 aliquot of either the Cu or Ag solution was placed on the graphite platform located on the translational rod12 and dried at approximately 60°C. The platform was then trans- ferred through two differentially pumped stages into the main vacuum chamber. The base pressure of the main chamber was less than 8 x lop9 Torr (1 Torr = 133.322 Pa) and was approxi- mately 3 x Torr with the SIMS Ar+ gun active.Once the platform was accurately positioned under the mass spec- trometer the temperature of the platform was ramped from ambient temperature to approximately 773 K at a rate of 1 K s-'. Results and Discussion The SIMS technique does not necessarily sputter 'stoichio- metric units' from the surface a fact which is often the source of the difficulty in spectral interpretation. However when clusters are removed they originate in the immediate vicinity of the primary-ion collision with the surface. As a consequence,168 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 t a) if condensed-phase microparticles microcrystals or micro- droplets of metal or metal containing compounds are located on the surface one would expect to see secondary ions containing more than a single metal atom.To verify this postulate a simple sheet of Cu was used in place of the graphite platform and served as the primary Ar+ beam target subject to static-SIMS analysis. Fig. l(a) shows the static SIMS spectrum for the higher masses of the pure Cu sheet at 298 K. Owing to the close proximity of Cu atoms on the surface peaks for Cu2+ and Cu3+ are evident. Oxide species (e.g. Cu20f) are present as a result of an expected oxide layer on the Cu. As the temperature increases the oxide peak decreases and Cu2+ and C U ~ ( O H ) ~ + dominate [Fig. l(b)]. At 751 K the Cu30+ increases in inten- sity relative to that observed at lower temperatures [Fig. l(c)]. The distinct appearance of polyatomic Cu ion species from these closely spaced Cu atoms of the Cu sheet suggests that such spectral features may be a characteristic of any micro- scopic clusters of a metal on a surface.In contrast dispersed metal species should exhibit a low probability of encountering two or more closely positioned metal atoms at the site of the collision of the primary-ion beam with the surface. As a result 1 cu,oi cuz+ CuJOH),' 100 120 140 160 180 200 mlz Fig. 1 and (c) 747 K S-SIMS spectra of a copper platform at (a) 298 (b) 429 only monomeric metal ions should be detected. Thus this information may be useful in investigating the surface mor- phology of metals on a graphite surface. It can also be seen in 'Fig. 1 that as temperature is increased there is a general decrease in the various oxygen containing species this could Ibe due to the changes in the degree of surface oxidation or some other change in surface species.Copper solution samples resulting in deposited Cu masses ranging from 2-2 000 ng were dried onto a pyrolytic graphite coated graphite plaiform. The ratio of the intensity of lCu,O+:Cu+ at room temperature was used as an indication of the extent of aggregation on the surface. As shown in Fig. 2 the ratio is relatively constant for masses above 40ng but dropped significantly for the 2 and 20ng sample loads. The Cu20f signal was used in place of another polymeric Cu ion because of its greater sensitivity. The error bars on Fig. 2 are due to both variations between different experiments and variations in the baseline. The former variation affects the higher masses 40-2000 ng and since for 2 and 20 ng there is no observable signal the variation in the baseline is dominant.If the Cu20+ signal were to retain the same ratio with respect to the Cu' signal it should be easily detected above the noise level even at 2 ng of deposited Cu. Thus its decrease suggests that the sample is present as dispersed Cu atoms or Cu containing molecules at the lower concentrations. Figs. 3 and 4 show the contrasting spectra for 40 and 20 ng sample loads respectively. The signal-to-noise ratio for the aggregates (e.g. Cu2O+ and Cu2+) is large with a 40ng sample but for the ;!Ong sample the polyatomic ion signals are barely detectable above the noise even though there is only a factor of two decrease in the amount of Cu that has been 0.25 0.20 .- 0 0.15 F 3 0.10 u .- v) C 0.05 - 0 -0.05 f 1 10 1 x 102 1 ~ 1 0 3 1 ~ 1 0 4 Mass/ n g Fig.2 Ratio of Cu,O+:Cu+ measured at room temperature for a 2 pl sample of a copper nitrate solution deposited on pyrolytic graphite coated graphite at increasing concentrations I4O I 120 - - I v) C v) 100 - e 80 - s m 6 0 - -. c > v) 5 40 20 .- - 4- C - - cu 0 50 60 70 80 90 100 110 120 130 140 150 m/z Fig.3 nitrate on a pyrolytic graphite coated graphite platform S-SIMS spectra of 40ng of Cu deposited as the aqueousJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 169 - - - - - 3 0 ~ I 30 I 25 ," 20 4- 8 m 15 g > 4- .- l o 2 w - 5 - - 0 25 I m f 20 3 0 15 r -. Y > 'v 10 Q) C .+d - 5 50 60 70 80 90 100 110 120 130 140 150 m/z Fig.4 S-SIMS spectra of 20ng of Cu deposited as the aqueous nitrate on a pyrolytic graphite coated graphite platform deposited.While uncertainty in the analyte position after initial droplet drying makes absolute quantitative comparisons from one sample to the next very difficult the relative signal magnitudes within the same sample of different m/z values should be relatively accurate. These results indicate that Cu is located on the graphite surface as dispersed adsorbed species and not as microdroplets islands or any other aggregated form at solution concen- trations below 0.02 ppm Cu (40 ng). This is consistent with the first order of release and dispersed atoms as suggested by McNally and Holcombe2 and is supported by the work of Lynch et d3 The results also suggest that the low concen- trations of Cu are dispersed on the surface immediately after the sample is dried. For an Ag metal sheet the SIMS spectrum also shows the presence of poly atomic species.The dominant polymeric frag- ment observed for the Ag sheet was Ag,'. Silver samples with masses between 2 and 2000 ng were also deposited as 2 p1 sample volumes and dried onto a pyrolytic graphite coated graphite platform. Fig. 5 shows that unlike Cu the ratio of the intensities of Ag,+:Ag+ after drying is constant over the entire concentration range studied. This implies that the Ag on the surface is present as microdroplets or some other form of aggregates within the concentration range studied. The error bars represent primarily variations between different experi- ments and as in Fig.2 the error bars at lower sample masses are due to variations in the baseline. Using the alignment of the ETAAS peaks and the general variation of E Fonseca et concluded that Ag existed as aggregates above 0.1 ng and probably was present as dispersed atoms for less than 0.07 ng of Ag. Eloi et a1.,14 however found no evidence to suggest a surface bound species of Ag as they had seen for Cd 0'1° I Ma sstng Fig. 5 Ratio of Ag2+:Ag+ measured at room temperature for a 2 pl sample of a silver nitrate solution deposited on pyrolytic graphite coated graphite at increasing concentrations and Pb when using Rutherford backscattering spectroscopy as a probe. Instead they suggested that most of the sample was distributed in the bulk of the graphite platform. Various Cu and Ag ion signals were monitored continuously to provide data for TPS-SIMS.In this technique the tempera- ture is ramped to the desired final temperature (uiz. 773 K) then linearly cooled at the same rate. Multiple ion monitoring was used to record the temperature-dependent intensities for several masses. The resulting plots of signal intensity versus temperature can provide information on the reversibility of processes within the temperature range studied. Fig. 6 shows the TPS-SIMS plot for different Cu containing ions with a 400 ng load. It is interesting to note that the signal decreases rapidly at approximately 550 K. As the final tempera- ture which is below the atomization temperature in vacuum is reached and the sample begins to cool the signal remains below detectability.Data collection is stopped when heat loss from the platform is slower than the desired cooling rate. When the sample finally cools to room temperature the signal remains at baseline. This indicates that the Cu is irreversibly removed from the surface (e.g. sub-surface migration) or transformed into another chemical form which is less sensitive to positive ion SIMS detection. At the lower concentrations where the samples exist as dispersed atoms a different trend is seen in the TPS-SIMS scans. Fig. 7 shows that at lower concentrations there is a slow decrease in the signal intensity of the same masses shown in Fig. 6 but without a peak appearing near 550 K. Again little signal is detected in either case for these ions above 600 K. By monitoring evolved gases from a microgram sample load during a thermal ramp it is evident that the signal decrease at 550 K corresponds to the final decomposition of the basic 30 35 .- 25 v) 20 00 "0 15 ? -.4- r .- g lo a 4- - 5 0 300 400 500 600 700 Tern perat u re/K Fig.6 TPS-SIMS scan of 400ng of Cu deposited as the aqueous nitrate on a pyrolytic graphite coated graphite platform 160 140 7 120 m v) 100 13 5 80 > 2 60 40 20 .w .- a c - 300 400 500 600 700 Temperature/K Fig. 7 TPS-SIMS scan of 2 ng of Cu deposited as the aqueous nitrate on a pyrolytic graphite coated graphite platform170 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 dispersed molecules on the surface. Thus static-SIMS can be used to give indications of the extent of metal dispersion on a substrate.Data for Cu show that at sample amounts below 20 ng Cu is dispersed and probably exists as dispersed atoms or copper containing molecules. This adsorption to the surface occurs during desolvation (ie. the drying cycle). At masses greater than 40 ng polymetallic fragment ions are observed thereby indicating that Cu exists as microdroplets or some other form of surface aggregates at room temperature. Silver however shows evidence of aggregates throughout the concen- tration range studied. This work was supported in part by the National Science Foundation grant No. CHE9020591. - 5 - 300 400 500 600 700 Temperature/K Fig.8 Positive and negative TPS-SIMS scans of 2000ng of Cu deposited on a pyrolytic graphite coated graphite platform as the hydrated nitrate copper nitrate salt to CuO(s).It is possible that the CuO+ SIMS signal shown in Fig. 8 originates from fragmentation of the basic nitrate crystal on the surface by the primary-ion beam. This is supported by the negative SIMS ion spectra for CuO-. Fig. 8 shows the signal for CuO- increasing at tempera- tures where the CuO' signal decreases. This change near 550 K is located at the temperature expected for the conversion of the basic nitrate salt to C U O ( S ) . ~ ~ ~ ~ This is reasonable since surface CuO would be more likely to form a negative ion during sputtering and hence an increased CuO - signal. The peak for the CuO- signal centered around 330 K corresponds to the temperature of dehydration of the hydrated copper nitrate reported by Taylor et As noted in Fig.6 at the sample masses where there is evidence for surface aggregates the TPS-SIMS scan shows a sharp drop corresponding to the decomposition of basic nitrate salt this sharp decrease in signal is absent for masses where the sample appears to be dispersed. This indicates that either the dispersed sample exists as dispersed atoms possibly bound to the surface through a Cu-0-surface bond or that for the dispersed samples the bulk characteristics of the basic nitrate decomposition are lost. Conclusion Metals which exist as microdroplets give rise to sputtered ions containing multiple metal atoms. The probability of sputtering polymetallic secondary ions is low when the analyte exists as 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 References Arthur J. R. and Cho A. Y. Surface Sci. 1973 36 641. McNally J. and Holcombe J. A. Anal. Chem. 1987 59 1105. Lynch S. Sturgeon R. E. Luong V. T. and Littlejohn D. J. Anal. At. Spectrom. 1990 5 311. Fonseca R. W. McNally J. and Holcombe J. A. Spectrochim. Acta Part B 1993 48 79. Redhead P. A. Vucuum 1962 12 203. Rojas D. and Olivares W. Spectrochim. Acta Part B 1992 47 387. Yan X.-p. Ni Z.-m. Yang X.-t. and Hong G.-q. Spectrochim. Acta Part B 1992 48 605. Ganz J. Sattler K. and Clarke J. 1989 Surface Sci 219 33. Schleicher B. Jung Th. Hug H. and Burtscher H. 2. Phys. D. 1991 19 327. Mason M. G. Phys. Rev. 1983 B27 748. Eigelhoff W. F. Jr. and Tibbets G. G. Phys. Rev. 1979 B19 5028. Hassell D. C. Majidi V. and Holcombe J. A. J. Anal. At. Spectrom. 1991 6 105. Fonseca R. W. McNally J. and Holcombe J. A Spectrochim. Acta Part B 1993 48 79. Eloi C. Robertson J. D. and Majidi V. J. Anal. At. Spectrom. 1993 8 217. Ghose J. and Kanungo A. J. Thermal Anal. 1981 20 459. Taylor T. J. Dollimore D. and Gamlen G. A. Thermochim. Acta 1986 103 333. Paper 3105841 G Received September 28 1993 Accepted December 14 1993
ISSN:0267-9477
DOI:10.1039/JA9940900167
出版商:RSC
年代:1994
数据来源: RSC
|
13. |
Occurrence and reduction of noise in inductively coupled plasma mass spectrometry for enhanced precision in isotope ratio measurement |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 171-176
Ian S. Begley,
Preview
|
PDF (861KB)
|
|
摘要:
JOURNAL O F ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 171 Occurrence and Reduction of Noise in Inductively Coupled Plasma Mass Spectrometry for Enhanced Precision in Isotope Ratio Measurement* Ian S. Begley and Barry L. Sharpt Department o f Chemistry Lo ughbo ro ugh University o f Tech n o logy Lo ughb orough L eices f ershire UK LEI I 3TU The limitations imposed upon measurement precision by instrumental noise sources present within the inductively coupled plasma mass spectrometer used have been established using noise spectral analysis. Based upon the spectral information gained a methodology for the sequential measurement of isotopes has been developed to minimize the deleterious influence of non-random instrumental noise. Efficient use of the noise reduction techniques available within the peak-jumping mode have been found to remove the majority of the instrumental noise associated with sample introduction and excitation facilitatin a measurement precision of approximately 0.05% relative standard deviation with respect to the lo7Ag "Ag isotope ratio.The precision of isotope ratios has been found to be limited by inaccuracies associated with the operation of the quadrupole mass analyser and the statistical error arising from the random arrival of ions at the detector. Keywords Inductively coupled plasma mass spectrometry; noise spectral analysis; isotope ratio measure- ment; instrumental parameters The use of inductively coupled plasma mass spectrometry (ICP-MS) in isotopic ratio studies has been somewhat restric- ted by its failure to offer the precision and accuracy attainable by more established techniques such as thermal ionization mass spectrometry (TIMS).The precision achievable by ICP-MS in the range 0.1-1.0% relative standard deviation (RSD) for isotope ratios is generally regarded as being primar- ily limited by instrumental in~tability.'-~ Using TIMS a pre- cision of less than 0.005°/~ RSD for isotopic ratios can be achieved by use of acquisition periods of up to 24 h. A number of studies have utilized noise spectral analysis or high-speed photography to gain an understanding of the frequency charac- teristics and origin of instrumental noise occurring in ICP-MS.4-8 Noise spectral analysis has shown the noise charac- teristics of ICP-MS instruments to be essentially similar to those observed in inductively coupled plasma atomic emission spectrometry (ICP-AES).4*s The majority of llfnoise occurring in ICP-AES has been shown to arise from nebulization and vaporization proces~es.~ Several studies have addressed the optimization of instru- mental parameters for accurate and precise measurement of isotopic rati0s.'-~9~ Ting and Janghorbani3 found that many of the trends observed were the result of reducing the counting statistic through increasing the analyte sensitivity of the ICP-MS instrument.Under optimized conditions the precision of measured isotopic ratios has at best been 2-3 times that imposed by the counting statistics. Furuta6 studied the influ- ence of dwell time per channel upon measurement precision of lead isotope ratios in the mass scanning mode to gain improvement via elimination of low-frequency noise. The detri- mental influence of random llf and discrete frequency noise upon the precision of isotope ratios required to be quantified prior to adoption of a suitable strategy for noise reduction.In this study the acquisition parameters available in the peak- jumping mode have been optimized based upon noise spectral information and the nature of the residual noise investigated. Experimental Instrumentation The ICP used in this study employed a 27 MHz crystal- controlled supply ( Plasma-Therm Kresson NJ USA Model * Presented at the XXVIII Colloquium Spectroscopicum Internationale (CSI) York UK June 29-July 4 1993. 7 To whom correspondence should be addressed. HFP 2500F) with an automatic impedance-matching network.The ICP was operated at 1300 W and the outer intermediate and carrier gas flow rates were 12.0 0.0 and 0.75 1 min-' respectively. A glass concentric nebulizer (J. E. Meinhard Santa Cruz CA USA Model TR-30-CZ) peristaltic pump (Gilson Villiers Le Bel France Minipuls 2) and water-cooled spray chamber (VG Elemental Winsford Cheshire UK) were used for sample introduction. The plasma source was centred about the sampling orifice which was located approximately 12 mm above the load coil. The quadrupole mass analyser was a VG Micromass (Altrincham Cheshire UK) Model 12-12S fitted within the vacuum system of the second UK ICP-MS instrument built on behalf of the British Geological Survey as described previously." The mass resolution was set to 0.65 u defined as the peak width at 5% of the peak height.Data acquisition was undertaken in the peak-jumping mode with three acqui- sition points each separated by 0.05 u being allocated to each mass peak. Noise Spectral Analysis The collection of data was undertaken using the operating conditions utilized in isotopic ratio measurement following an interval of at least 1 h after light-up to allow the instrument to equilibrate. The ion current from the electron multiplier (Galileo Sturbridge MA USA Model 4780V) was amplified using an operational amplifier current follower of in-house design and low pass filtered to comply with the Nyquist sampling theorem. A sample and hold amplifier (Metrabyte Taunton MA USA SSH-4) and 12-bit analogue to digital (A/D) converter (Metrabyte DAS20) were used to digitize the analogue signal for processing by an IBM compatible personal computer.A series of ASYST 3.1 (Keithley Rochester NY USA) programs were used for control of data acquisition and calculation of noise power spectra. For the collection of noise-power spectra in the range from 0 to 5 Hz the sampling rate was 20 Hz and the -3 dB point of the in-house low pass filter having a - 12 dB per octave roll-off was set at 10 Hz. For the collection of the 0-400 Hz spectra the sampling rate was 1000 Hz and the -3 dB point of the in-house low pass filter having a - 30 dB per octave roll-off was 400 Hz. In all cases 1024 data points were acquired in each data set resulting in a frequency resolution of 0.020 Hz for the 0-5 Hz spectra,172 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 and 0.98 Hz for the 0-400 Hz spectra. Noise amplitude spectra were calculated by Fourier transformation taking the sum of the squares of the real and imaginary components of the transformed data and signal averaging in the frequency domain. For the 0-5 Hz spectra eight sets of data were averaged and 24 sets of data were averaged for the 0-400 Hz spectra. Procedure for the Determination of Isotopic Ratios There are two different modes of data collection available in ICP-MS instruments mass scanning and peak jumping. It is the peak-jumping mode of operation that has been investigated within this study. In the peak-jumping mode the centres of the mass peaks are sequentially transmitted by 'hopping' between selected isotopes.The acquisition method used in the peak-jumping mode is determined by a number of software definable parameters. The parameter set utilized was as follows acquisition points per peak (three); mass step between points (0.05 u); dwell time per channel (varied); settle time between peaks (varied); number of sweeps (varied); and number of repeat integrations (lo). The analyte concentrations utilized in isotope ratio analysis were 100 or 500 ng ml-' for silver and 1000 ng ml-I for lead. Reagents Stock solutions of silver and lead were of analytical-reagent grade. De-ionized water for the dilution of the stock solutions was obtained from a laboratory-reagent grade water system (Liquipure Bicester Oxfordshire UK) generally operated at 18 Ma.All solutions were prepared so as to contain 1% v/v HNO of ultrapure reagent grade and stored in high-density poly(propy1ene) bottles. Results and Discussion Noise Spectral Analysis A noise-power spectrum obtained over the frequency range 0-5 Hz for ion-current monitoring of the 40Ar40Ar+ signal at a count rate of 2 x lo6 counts s-' is shown in Fig. l(a). A l/f noise component present below approximately 1 Hz was observed similar to that found in ICP-AES when using a glass concentric nebulizer." Discrete noise components were observed at 2.46 and 4.92Hz the latter being a harmonic of the former for a solution uptake rate of 0.73 ml min-l. These discrete noise components were absent from noise-power spectra obtained for a dry plasma. Variation of the solution uptake rate was found to cause a shift in the frequency at which these noise components occurred as had been observed by Goudzwaard and de Loos-Vollebregt.12 The frequency of the fundamental noise peak was observed to be equivalent to the rate at which the individual rollers on the pump head of the peristaltic pump squeezed the pump tubing.A noise-power spectrum for 40Ar40Ar + in the frequency range 0-400 Hz is given in Fig. l(b). The audio-frequency (a.f.) noise at 324 Hz is a consequence of a phenomenon occurring within the ICP discharge. The origin of this phenomenon has been described as being derived from the rotation of the discharge as a consequence of non-linear flow of the plasma gas," and also the passage of vortex rings down the central axis of the discharge owing to the flow of plasma gas into the surrounding static air.13 The frequency of the a.f.noise peak was observed to vary with plasma operating conditions between 200 and 350 Hz. The influence of incident power gas flow rates and observation height upon the frequency of the a.f. noise in the instrument used has been found to follow the trends previously r e p ~ r t e d . ~ . ~ The discrete noise components occurring at 100 150 200 250 and 300 Hz are all considered to be related to the discrete noise at 50Hz owing to pick-up from a.c. power lines. Goudzwaard and de Loos-Vollebregt12 - 40 - 45 - 50 - 55 - 60 - 65 - 70 -75 m P 1 4- 'ijj -80 (a) 5 0 1 2 3 4 5 e c -35 - 40 - 45 i -50 ~ I I -55 4 I I 1 -60 -\If - 65 I I 0 100 200 300 400 FrequencyIHz - 70 Fig.1 Noise-power spectra for 40Ar40Ar+ ion current from the ICP-MS instrument for introduction of de-ionized water under stan- dard operating conditions (a) frequency range 0-5 Hz; and (b) fre- quency range 0-400 Hz found that in ICP-AES the noise intensity of components occurring at 100 and 200 Hz were proportional to the square of the analyte concentration suggesting that these are associ- ated with signal modulation within the plasma. Noise-power spectra obtained for the 208Pb+ signal from a solution contain- ing 1000 ng ml-' of lead were consistent with those shown in Fig. 1 for 40Ar40Ar+. Furuta et aL5 considered similar obser- vations to imply that noise peaks are the result of a modulation of plasma ionization conditions at the sampling orifice.The contributions of the various noise components observedJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994. VOL. 9 173 Table 1 Contribution of type of noise to overall instability Noise type Percentage of total instability Random* Frequency dependent llf 40 40 20 * For data collected over a period of 24 s at a count rate of approximately 2 x 106 counts s ~ '. in noise-power spectra can be estimated by calculating their relative contributions to the over-all instability of the ion- current signal. The RSD of a measured signal can be approxi- mated from the mean noise level indB ( N ) by use of the equation RSD = 10"" ( 1 ) A white noise level of between -60 and -65 dB was observed for an ion current equivalent to approximately 2 x lo6 counts s-l [Fig.l(b)]. Using eqn. (l) the white noise rep- resents an RSD of 0.06-0.1% of which approximately 0.07% can be accounted for by the counting statistic. The counting statistic can therefore be considered the primary white noise source within our instrument. This limits the best possible precision attainable by isotope ratio analysis for use of a data acquisition method which removes the detrimental influence of non-random noise. Typically the precision observed for isotope ratio measurement by ICP-MS is 2-3 times that due to the counting statistic. Digital signal processing was utilized to determine the improvement in precision that would be attained by removing the influence of specific frequencies. This was accomplished by computing the Fourier transform of the signal and multiplying this by the desired digital filter.The inverse Fourier transform was then computed to give the filtered signal. These calculations were carried out for a number of data sets acquired under conditions identical with those used in the collection of noise-power spectra over the frequency range from 0 to 400 Hz. The improvements in precision suggest that for the ICP-MS instrument used the contributions of the various noise types to the over-all instability were as given in Table 1. Isotope Ratioing Noise spectral analysis has shown low-frequency noise to make a major contribution to the over-all instability of the ion- current signal. If the mass peaks of interest are swept rapidly such that the time separating the measurement of isotopes is much smaller than the period of low-frequency noise then the signal intensity can be considered unchanged and the isotopic ratio unaffected by low-frequency noise.It is the period of time that elapses between the start of measurement of the first isotope and the end of measurement of the second isotope within the ratio referred to herein as the elapse time which determines the range of noise frequencies removed by ratioing. The elapse time for an acquisition procedure involving a single isotopic ratio is given by Elapse time = 2 x td x np + t (2) where td is the dwell time per channel np is the number of points per peak and t is the settle time. The use of three points per peak was adopted as this allowed monitoring of the peak profile for the measured isotopes. For noise components with periods longer than twice the elapse time isotope ratioing will be beneficial in reducing noise by acting as a substitute for simultaneous measurement of the isotopes as in a multi- detector system.The improvement in measurement precision for the lo7Ag '"Ag isotope ratio observed upon decreasing the elapse time via the dwell time is shown in Fig. 2. The elapse time was reduced by decreasing the dwell time per channel of which there were three per isotope from 81.92 to O'I5 ~ B v 4 A v 0 10 20 30 40 50 60 70 80 90 Dwel I ti m e/ms Fig. 2 Effect of dwell time on RSD for isotopic ratio measurement A lo7Ag:'O9Ag; and B counting statistic. The settle time between peaks was 10ms 5.12 ms. The integration time per isotope was held constant at 49.152 s by increasing the number of sweeps.By retaining a fixed integration time the total ion count and thus the counting statistic remained fairly constant. Reducing the elapse time from approximately 500 to 41 ms was beneficial to measure- ment precision as the detrimental influences of llf noise and peristaltic pump induced noise [Fig. 1 (a)] were eliminated. Further reduction in dwell time was observed to have no appreciable influence upon measurement precision possibly because the increase in sweep rate attained by reducing the dwell time below 5.12 ms was relatively small while utilizing a settle time of 10 ms. It was therefore considered necessary to determine whether reduction of elapse time by reducing the settle time would benefit the precision. The influence of reduction in elapse time by reducing the settle time upon the measurement precision for the lo7Ag lo9Ag isotope ratio is shown in Fig.3. For a dwell time of 2.56 ms reducing the settle time from 10 to 2 ms resulted in a marginal improvement in measurement precision. Variation of the settle time within this range was observed to have no significant influence upon the accuracy of the lo7Ag '"Ag isotope ratio. However reduc- ing the settle time to below 0.5 ms was found to cause a measurable change in the lo7Ag lo9Ag isotope ratio suggesting the quadrupole mass analyser had insufficient time to stabilize following 'jumping'. Signal Averaging Signal averaging could be beneficial in enhancement of the signal-to-noise ratio (S/N) by reduction of any of the noise 0.15 0.10 - 8 n - v U 0.05 I t I I I 0 2 4 6 8 1 0 Settle time/ms 2 Fig.3 Effect of settle time on RSD for isotopic ratio measurement A "'Ag '09Ag; and B counting statistic.The dwell time per channel was 2.56 ms174 JOURNAL OF ANA types referred to above. However to clarify the strategy adopted herein signal averaging will be discussed solely in terms of its effect in the reduction of non-random noise while the reduction of random noise will be considered in terms of the signal integration period. Signal averaging of non-random noise is operational at two separate levels during data collec- tion these are (i) the summation of sweeps in the multi- channel averager; (ii) the low pass filtering and anti-aliasing (preventing under-sampled high frequency components appear- ing at spurious low frequencies) influence of the dwell time.The summation of sweeps is an effective means of noise reduction as isotope peak profiles are added coherently in the multi-channel averager while noise components are reduced by smoothing. The frequency of interest in the summation of sweeps is that derived from the time taken between replicate measurements of the same isotope referred to as the cycle time. For an analysis routine in which two isotopes are monitored to give a single isotope ratio the cycle time is equivalent to the sum of the elapse time and the settling time required to return from the second to first isotope. The frequency domain representation of the accumulation of sweeps is similar to that given by a comb filter whose teeth are centred at the averaging frequency and its harrn0ni~s.I~ The width of the bandpass at these frequencies is given by Frequency bandwidth = 0.886/nS x t (3) where n is the number of sweeps and t is the cycle time. Noise frequencies outside these bandpass regions are reduced by the accumulation of sweeps.While the cycle time is some- what fixed by the elapse time required for isotope ratioing the number of sweeps can be increased to gain a decrease in the bandwidth of the teeth. For example with a typical cycle time of 81 ms and 25 sweeps a tooth bandwidth of 0.44 Hz is obtained however for 500 sweeps the tooth bandwidth is reduced to 0.022 Hz. Increasing the number of sweeps accumu- lated therefore improves noise reduction at frequencies in close proximity to the averaging frequency or its harmonics.It is of course necessary to select a cycle time that does not correspond with a prominent noise component such as that at 50 Hz owing to pick-up from a.c. power lines. Dwell time acts as a low pass and anti-aliasing filter having a high frequency cut-off at the frequency given by twice the inverse of the dwell time. Noise at frequencies above that associated with the dwell time per channel (49 Hz for a dwell time of 20.48 ms) undergoes reduction within each channel. The dwell time is crucial in the reduction of non-random noise by both isotope ratioing and signal averaging. Decreasing the dwell time is beneficial in reduction of low frequency noise by isotope ratioing while increasing dwell time is beneficial in reduction of high frequency noise by signal averaging.Thus it is necessary to find a balance for dwell time to gain maximum benefit in the precision of isotopic ratios. As the instability caused by llfnoise is generally more detrimental to precision in ICP-MS than high frequency noise such as a.f. noise it was considered best to optimize dwell time for maximum effective- ness in isotope ratioing. Signal Integration Period In addition to the influence accumulation of sweeps has upon non-random noise the accumulation of sweeps is beneficial in reducing the uncertainty owing to the counting statistics. If the majority of non-random noise was removed by the com- bined efforts of isotope ratioing and signal averaging the counting statistic would most probably be precision limiting. Assuming the counting statistic is limiting the S/N should improve in proportion to the square root of the signal integration period.12 The improvement in precision of the Io7Ag lo9Ag isotope ratio observed upon increasing the signal integration period for an elapse time of approximately 71 ms (dwell time= 10.24 ms) is shown as a log-log plot in Fig.4(a). ,YTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 I I 100 1 0.1 Log (integration 1 timels) 10 100 - $ OY 100 1 10 100 1000 Log (concentrationhg ml '1 Fig. 4 Influence of non-random noise on signal-to-noise ratio (SIN) for isotopic ratio measurement. Variation in (a) isotope integration time for a solution containing 100ngml-l of silver and (b) silver concentration for an integration time of 49.152 s.A lo7Ag lo9Ag; and B counting statistic The line representing the counting statistic has a gradient of 0.5 as is characteristic of random noise. The S/N of the isotope ratio is shown to benefit from the improvement in the S/N of the counting statistics suggesting that the counting statistic is the limiting noise type. The improvement in precision of the Io7Ag Io9Ag isotope ratio observed upon increasing the analyte concentration for an elapse time of approximately 133 ms (dwell time = 20.48 ms) is shown as a log-log plot in Fig. 4(b). Comparison of the rise in S/N ratio of the Io7Ag lo9Ag isotope ratio when increasing the analyte concentration and increasing the integration period shows that the full benefit of increased integrated count is not realised as there is some divergence from the counting statistic when the integration period is increased.This would suggest that although the counting statistic continues to be the limiting noise an additional noise type has been introduced which is increasing in relative intensity with integration period. It is suggested that the most probable explanation is that an increase in the signal inte- gration period causes the significance of instrumental drift to rise owing to prolonging of the time taken in acquisition of the ten integrations constituting a determination. That is to say the RSD for ten integrations is influenced by change in the observed isotope ratio between integrations arising from drift occurring incoherently in both isotopes. Hence while llf noise at frequencies above that associated with the time taken in acquisition of a single integration (2 min) is minimized drift occurring within the 20 min period required to undertake ten such integrations is detrimental to measurement precision.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 175 Improved Precision The variability in measurement precision of the Io7Ag Io9Ag isotope ratio for consecutive determinations over a 1 h 40 min period are shown in Fig. 5(a). The operating parameters util- ized were dwell time per channel = 10.24 ms; settle time = 10 ms; and number of sweeps= 1600. It was observed that although the error owing to the counting statistics remained unchanged the precision fluctuated about a mean of approxi- mately 0.05% RSD suggesting the magnitude of noise varied with time. Replication of this experiment upon a VG PlasmaQuad (VG Elemental) fitted with a high-extraction interface has shown that results similar to those given in Fig.5(a) can be obtained. However as a consequence of the higher transportational efficiency the reduced limitation imposed by the counting statistic for a specific analyte concen- tration allowed a precision approaching 0.05% RSD to be obtained at 100 ng ml-I as opposed to 500 ng ml-l. 0.07 0.06 0.05 2 0.04 0 $ 0.03 - 0.02 0.01 0 1 2 3 4 5 6 Determination I >. c .- 2 99 a C 4- .- a P c .- $ .t 98 97 A I 100.4 - &? - 0 100.3 5 a 0 .- c .- 4- 100.2 .; .- C 0 I- Y - 100.1 .; 3 100 99.9 1 96 ' ' 99.8 0 10 20 30 40 50 Integration Fig. 5 Variation of silver isotope intensities and isotopic ratio over a period of 1.66 h for a solution containing 500 ng ml-' of silver.(a) Five consecutive determinations for A lo7Ag '09Ag; and B counting statistic; and (b) 50 consecutive 2min integrations for . lo7Ag; + lo9Ag; and A '07Ag lo9Ag Remaining Noise The data for each of the integrations giving rise to the determinations shown in Fig. 5(a) acquired at 2 min intervals are shown in Fig. 5(b). For both the Io7Ag and lo9Ag isotopes a long-term drift upon which short-term fluctuations were superimposed was observed. The long-term drift in isotope intensities is reflected in the Io7Ag lo9Ag isotope ratio but in reverse. For example a rise in the measured ratio from integration 0 to 25 and at the same time a decrease in the isotope intensities is shown in Fig.5(b). In addition the isotope ratio appears to deviate from the mean value in instances for which there is notable differential change in the intensities of the two isotopes as for integrations 3 and 39. The results for a study of the change in the 206Pb and 207Pb intensities and 206Pb '07Pb isotope ratio over a 4 h period (2 min per integration) are shown in Fig. 6. The operating parameters utilized were dwell time per channel = 10.24 ms; settle time = 2 ms; and number of sweeps = 1600. The isotope intensities appear to have undergone significant long-term drift both positively and negatively during the 4 h period. The influence of this drift is prominent in the isotope ratio causing the precision of the 206Pb 207Pb isotope ratio averaged for the 12 determinations (one determination being the average of ten integrations) to rise to about 0.13% RSD approximately twice the mean of the measurement precision for individual determinations.The drift associated with the ICP ion source arises from either a change in energy transfer from the plasma to the sample or variation in the efficiency of the sample production and transportation." As these should influence both isotopes coherently the direct influence of drift associated with the ICP ion source upon the isotope ratio can as shown above be rendered insignificant. However Figs. 5(b) and 6 imply that drift over a time scale of tens of minutes or intermittent noise unequally influencing the isotopes of interest is preventing the measurement precision of isotope ratios from being limited only by the counting statistic. It is therefore necessary to I A I I I I 1 1 80 ' 99.6 0 20 40 60 80 100 120 Integration Fig.6 Variation of a '''Pb; + 207Pb isotope intensities; and A '06Pb '07Pb isotopic ratio over 120 integrations during approxi- mately 4 h176 JOURNAL OF ANA identify the possible sources of this noise.There are two likely sources of instability associated with the mass spectrometer that have been identified. These arise from uncertainties associ- ated with the mass bias and mass calibration of the ICP-MS instrument. The mass bias the deviation of the measured isotope ratio from the actual value occurs because the sensitivity of the instrument varies with mass owing to differences in ion trans- mission. Generally in ICP-MS instruments the mass bias takes the form of a loss of transmission at either end of the mass range (from lithium to uranium).Correction of mass bias within samples can be undertaken by normalizing the data to a constant accepted ratio. It is common to use a reference standard in determination of the mass bias however it is necessary to assume that the measured bias remains constant during sample analysis. Russ and Bazan' have observed sig- nificant change in mass bias for their ICP-MS instrument (VG PlasmaQuad) over several hours. The stability of the mass calibration [the relationship between measured mass represented by the digital to analogue (D/A) converter channel number and actual mass] is generally limited by the constancy of the ratio of the d.c.voltage to the radiofrequency (r.f.) voltage applied to the quadrupole rod pairs." For instance a change of the order of 0.05% in the r.f. voltage amplitude for 100~1 would cause a shift of about 0.05 u. The daily shift in mass calibration for quadrupole mass spectrometers is typically less than _+ 0.02 u however ambient temperature change could cause additional shift." As the resolution of the quadrupole mass analyser is inversely pro- portional to the ratio of the d.c. voltage to the r.f. voltage any shift in mass calibration caused by a change in this ratio would also be reflected in the peak shape. In combination mass shift and resolution change may cause drift which occurs incoher- ently between isotopes. The nature of both mass bias and mass calibration is such that the significance of the associated instabilities would rise from middle mass (silver) to high mass (lead).In an attempt to quantify the noise arising from instability of mass bias and mass calibration a number of consecutive determinations were undertaken for a fixed m/z ratio at the centre of the '''Pb isotope peak using the acquisition pro- cedure utilized in measurement of the "'Pb '07Pb isotope ratio. By monitoring 206Pb in both isotope channels without 'jumping' the quadrupole mass analyser noise coherent to both isotopes was passed into the isotope intensities and their ratio as before. However incoherent noise arising from the mass spectrometer would have been eliminated. The accuracy C0.9999 +0.0002 (n = 5 ) ] and precision (0.032-0.041 % RSD) for the measured isotope ratio was found to be limited only by the counting statistic (0.034-0.037% RSD).The mean difference between the measured precision and the counting statistic is only 0.0()3%. Therefore it is suggested that the mass spectrometer and/or its associated ion optics are the source of non-random noise observed in the isotope ratio measurements reported above. ,Y 'ICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Conclusions This study has shown that educated selection of acquisition parameters can give rise to a practical precision of about 0.05% RSD for the Io7Ag '"Ag isotope ratio. Rapid isotope ratioing was successful in eliminating low frequency noise associated with plasma and sample introduction processes. The benefits of signal averaging were to ( i ) reduce the influence of noise occurring at frequencies that did not correspond with the bandpass frequencies of the equivalent comb filter; and ( i i ) reduce noise occurring at frequencies above twice that associated with the dwell time per channel.Having reduced the influence of non-random noise improved precision was obtained by reducing the counting statistic via increased inte- gration time. Under optimized conditions instabilities associ- ated with the mass bias and mass calibration were found to impose limitations upon the accuracy and precision of isotope ratio measurements. Various methods of data processing mass spectra to minimize the uncertainties arising from the mass bias and mass calibration are currently undergoing assessment. The authors thank the British Geological Survey for the loan of the ICP-MS instrument VG Elemental for supporting this work and the Science and Engineering Research Council for the provision of a studentship to I.S.B. References 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Russ G. P. and Bazan J. M. Spectrochim. Acta Part B 1987 42 49. Longerich H. P. Fryer B. J. and Strong D. F. Spectrochim. Acta Part B 1987 42 39. Ting B. T. G. and Janghorbani M. J. Anal. At. Spectrom. 1988 3 325. Crain J. S. Houk R. S. and Eckels D. E. Anal. Chem. 1989 61 606. Furuta N. Monnig C. A. Yang P. and Heiftje G. M. Spectrochim. Acta Part B 1989 44 649. Furuta N. J. Anal. At. Spectrom. 1991 6 199. Gray A. L. J. Anal. At. Spectrom. 1992 7 1151. Koperdraad R. H.Sc. Thesis State University of Gent 1990. Hobbs P. Spillane D. E. M. Snook R. D. and Thorne A. P. J. Anal. At. Spectrom. 1988 3 543. Date A. and Hutchison D. J. Anal. At. Spectrom. 1987 2 269. Belchamber R. M. and Horlick G. Spectrochim. Acta Part B 1982 37 17. Goudzwaard M. P. and de Loos-Vollebregt M. T. C. Spectrochim. Acta Part B 1990 45 887. Winge R. K. Eckels D. E. DeKalb E. L. and Fassel V. A. J. Anal. At. Spectrom. 1988 3 849. Heiftje G. M. Anal. Chem. 1972 44 69A. Carre M. Poussel E. and Mermet J.-M. J. Anal. At. Spectrom. 1992 7 791. Bley W. G. Vacuum 1988 38 103. Paper 3/045881 Received July 30 1993 Accepted October 6 1993
ISSN:0267-9477
DOI:10.1039/JA9940900171
出版商:RSC
年代:1994
数据来源: RSC
|
14. |
Spectral interferences encountered in the analysis of biological materials by inductively coupled plasma mass spectrometry |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 177-185
Hans Vanhoe,
Preview
|
PDF (1248KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 177 Spectral I Materials nterferences Encountered in the Analysis of Biological by lnductively Coupled Plasma Mass Spectrometry* Hans Vanhoe Jan Goossens Luc Moens and Richard Dams Laboratory of Analytical Chemistry University of Ghent Institute for Nuclear Sciences Proeftuinstraat 86 B-9000 Ghent Belgium In order to evaluate the potential of inductively coupled plasma mass spectrometry for the analysis of biological materials a systematic study on the occurrence of spectral interferences was carried out. All polyatomic ions originating from the matrix elements Ca CI P K C Na and S that give rise to spectral overlap with the analyte elements were examined systematically. For these matrix elements which are representative of most biological materials a list of all the polyatomic ions which can be seen to interfere up to a matrix element concentration of 1 g I-' is given.In addition for each spectral interference the corresponding apparent analyte concentration was calculated for different matrix concentrations. In this way an interference factor (IF) could be calculated for each spectral overlap so that the extent and importance of each potential spectral interference can be estimated. This factor is defined as follows IF = 1 O6 x apparent concentration (analyte element)/concentration (matrix element). A study on the formation and short- and long-term stability of polyatomic ions was also carried out to evaluate the use of some simple correction procedures to overcome spectral overlaps.Finally a review of the correction procedures applied to biological materials is given together with some examples. Keywords lnductively coupled plasma mass spectrometry; spectral interference; biological material Although for trace element analysis inductively coupled plasma mass spectrometry (ICP-MS) has several advantages over other techniques such as a quasi-simultaneous multi-element capability and excellent detection limits in a large number of applications it suffers from both spectroscopic and non- spectroscopic interferences. The latter group of interferences can be overcome by internal standardization standard additions isotope dilution or sample preparation (e.g. separ- ation) techniques. This subject has recently been reviewed by Evans and Giglio.' Spectroscopic interferences on the other hand are caused by atomic or polyatomic ions that have the same nominal mass as the analyte of interest.Such interferences can be divided into two categories depending on the origin of the interference. Firstly there are isobaric interferences caused by the overlap of isotopes of different elements. They are easy to predict and can be overcome by use of alternative isotopes of the element of interest. However when an element is introduced into a plasma the monoatomic singly charged analyte ion (M') is not the only species observed in the final mass spectrum. Elements with a low second ionization energy will be partly doubly ionized (M2+) whereas for elements that have a high MO bond strength oxide (MO') and hydroxide ions (MOH') are formed.The intensities of these species can be reduced to 1-2% by adjusting instrumental settings for MO' M+ even for the most refractory elements. In addition to these species polyatomic ions originating from the plasma (Ar) from the matrix (0 H and often C N P S and C1) and from the air surrounding the plasma (C N and 0) exist. There are probably three places where these polyatomic ions are formed in the plasma itself (especially for refractory elements that form oxide ions) in the boundary layer that is formed in the vicinity of the sampling cone surface and finally in the expansion region. Vaughan and Horlick2 have given an over- view of all spectral interferences due to the formation of metal monoxide ions (MO') hydroxide ions (MOH') and doubly charged ions (M2+).When analysing biological materials the number of potential spectral overlaps is dramatically increased by the acids (HC1 HClO and H2S04) used for the digestion of the sample and/or by the large variety of matrix elements present. Information * Presented at the XXVlII Colloquium Spectroscopicurn Inter- nationale (CSI) York UK June 29-July 4 1993. on these spectral interferences is therefore necessary to obtain accurate results. Tan and Horlick3 have extensively described the background mass spectra for water 5% solutions of HNO HC1 and H,SO,. Since no quantitative data were given this paper does not allow an estimation of to what extent the polyatomic species mentioned can contribute to the analyte signal intensity. In addition when biological materials are analysed other matrix elements such as C and Na also give rise to spectral overlap. Mulligan et aL4 have reported potential interferences on trace metals in the analysis of urine.Both qualitative and quantitative data were given. Evans and Giglio' gave an overview of all spectroscopic interferences (including those for biological materials) which have been described in the literature. In the present work the polyatomic ions related to the following matrix elements which are representative of most biological materials were examined Ca C1 P K C Na and S. Since both qualitative and quantitative results were obtained in most circumstances it is possible to estimate the magnitude and hence the importance of the spectral overlap when biological samples with a known composition are ana- lysed.In addition the formation and stability of some poly- atomic ions were studied in more detail so that the use of some simple correction procedures to overcome spectral over- laps could be evaluated. Experimental Instrumentation The ICP mass spectrometer used is a VG PlasmaQuad PQ1 (VG Elemental) equipped with a Gilson Minipuls-2 peristaltic pump a Meinhard concentric glass nebulizer (Type TR-30-A3) a Scott-type double pass spray chamber with surrounding liquid jacket and a Fassel-type torch. Details of the operating conditions are summarized in Table 1. Reagents and Standards High-purity water was obtained with a Millipore Milli-Q water purification system (resistivity of 18 Mi2 cm). Concentrated HN03 ( 14 moll- ') was purified by sub-boiling distillation in a quartz still using analytical-reagent grade HNO (Union Chemique Belge) as feedstock.Pre-cleaned polyethylene Cali- brated flasks and glass pipettes were used throughout.178 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 1 Operating conditions for ICP-MS Stage Parameter Plasma R.f. power Forward Reflected Gas flow Plasma Intermediate Aerosol Peristaltic pump Nebulizer Spray chamber Ion sampling Sampling cone Skimmer cone Sampling depth Vacuum Expansion stage Intermediate stage Analyser stage 1.35 kW <10 w Condition 13 1 min-' 11 min-' 0.725 1 min-' Minipuls 2 (Gilson) pumped at 0.9 ml min- Meinhard Tr-30-A3 concentric glass nebuli Double-pass Scott type water-cooled (10 "C) Nickel 1.0 mm orifice Nickel 0.75 mm orifice 10 mm (from load coil) 2.4 mbar* 1.0 x lop4 mbar 4.0 x mbar - 1 zer * 1 bar= 1 x lo5 Pa.For qualitative and quantitative study of the spectral inter- ferences for each matrix element investigated (Ca C1 P K C Na and S) ten solutions were prepared with concentrations of 0.05 0.1 0.5 1 5 10 50 100 500 and lOOOmgl-'. Each solution also contained In as an internal standard (concen- tration of 100 pg 1-'). Stock solutions with a concentration of 10 g 1-' of the matrix element were prepared by dissolving an appropriate amount of the solid salt or by diluting a concen- trated solution Ca(N03),.4H20 (UCB pro analysi) sub- boiled HC1 (10 mol 1-') H,P04 (Merck pro analysi) KNO (UCB pro analysi) acetic acid (Merck Suprapur) NaNO (Carlo Erba pro analysi) and sub-boiled H2S04 (18 moll- ').All solutions were prepared with 0.14moll-' HNO as the solvent and the diluent. In order to calculate the apparent analyte concentrations four multi-element standard solutions with a concentration of 100 pg 1-' for each element were used (external calibration) the first solution contained V Cr Mn Fe Co Ni Cu Zn and In (internal standard); the second contained Ti Mo and In (internal standard); the third contained Ge As Se and In (internal standard); and the fourth Ca Sc Ga Br and In (internal standard). All these solutions were prepared from commercially available 1 g 1-l atomic absorption spectrometry standard solutions and stored in polyethylene flasks. The blank solution used consisted of 100 pg 1-l In (internal standard) in 0.14 moll-' HNO,.Use of an Internal Standard In order to correct for the matrix-induced ion signal fluctu- ations and instrumental drift an internal standard was used. Although Vanhaecke et aL5 suggested that an internal standard with a mass close to that of the analyte is to be preferred for all elements In of mass 115 u was used [1'51n(95.7%)] because other internal standards (with a mass below 80u) could give rise to additional spectral interferences. Measuring Procedure Optimization of the ion signal Before each measuring procedure the aerosol gas flow rate and the ion lens settings which are generally accepted to have a strong effect on the signal intensity were optimized in order to give maximum signal intensity for '%+ . The optimum aerosol gas flow rate was found to fluctuate around the value of 0.725 1 min-' mentioned in Table 1.Other parameters including r.f. power intermediate and plasma gas flow rates and temperature of the spray chamber were not changed during the experiments described in the present work. Measuring conditions The qualitative and quantitative data for the spectral inter- ferences were obtained using the scanning conditions summar- ized in Table 2. The range between m/z 42 and 126 was scanned 100 times with a dwell time of 320 ps per channel. In this way one measurement required approximately 1 min. Five replicate measurements were made for each solution. In order to reduce memory effects the following measuring sequence was used firstly a blank was measured then the ten matrix element solutions with increasing concentration and finally the four standard solutions were measured.After meas- uring each solution the sample introduction system was rinsed for 2 min with 0.14 mol 1-' HNO,. Calculations For each solution the signal (peak area integrated over an m/z value of 0.8 around the peak maximum) of each nuclide of interest was normalized to the signal of 'IsIn (internal standard). The mean and standard deviation of the five resulting normalized. signals of each solution were calculated. External calibration was used to calculate the apparent concen- tration of the analyte elements. In addition for each spectral interference an interference factor (IF) was calculated as follows apparent concentration (analyte element) concentration (matrix element) IF= lo6 - with the matrix element and the analyte element concentrations expressed in the same mass per volume unit.In this way it is possible to estimate the extent and importance of each spec- tral overlap. Results and Discussion Spectral Interferences Qualitative and Quantitative Data The interference factors were calculated based on the results that were obtained for the range of matrix element solutions containing various concentrations. For each matrix element concentration for a given spectral interference an interference factor was calculated based on the formula given under Calculations. A mean of all interference factors obtained was then calculated together with a standard deviation. Normally all results were taken into account except for those which have a relatively high uncertainty [percentage relative standard deviation (RSD) above 20%].In cases where only one or two results are available an indicative value is given. The relatively small uncertainty in the interference factors obtained for most of the spectral interferences indicates a good and in most cases linear correlation between the matrix element concentration and the apparent analyte concentration. It is an exception for the uncertainty obtained to be high whereby the spectra1 overlap at low matrix element concentrations is underestimated. Table 2 Scanning conditions used to obtain qualitative and quantitat- ive data for spectral interferences Mass range No. of channels Dwell time per channel No. of sweeps Measuring time 42-126 u 2048 320 ps 100 = 1 minJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 179 Calcium An overview of the spectral interferences arising from Ca is given in Table 3. In addition to spectral interferences due to the formation of polyatomic ions some Ca isotopes also give rise to an isobaric overlap 46Ca (0.0040/,) coincides with 46Ti (8.0%) and 48Ca (0.19'%0) with 48Ti (73.8%). The interference factors for 46Ti and 48Ti are respectively 543 and 4010. These overlaps reduce the result for determination of Ti in a Ca-rich matrix. Some conclusions can be made from Table 3. Firstly the formation of CaO' and CaOH' is dominant. Since Ca has six stable isotopes with mass numbers between 40 and 48 the corresponding spectral interferences are spread out between m/z 56 and 65.The Ca polyatomic ions mainly interfere with the determination of Fe. The main Fe isotope 56Fe (91.7%) as well as 57Fe (2.20%) are subject to interference to a large extent interference factors of 650 and 3710 were observed respectively. The problems experienced with calcium oxide and hydroxide interferences in the analysis of human serum and urine have been given by Vanhoe et aL6 and Vaughan and Templeton7 respectively. Other elements subject to interference are Ni (58Ni and 60Ni) and to a lesser extent Co ("Co) and Zn (64Zn). Secondly calculations show that the ratios CaOH+ CaO+ and CaO+:Ca+ are 0.1 and 7.4 x respectively. The contribution of Ca160H+ to the total signal (Ca160H+ and Cai70+) was found to be more than 99%. Thirdly the formation of minimum amounts of 40Cai60160 + was only observed at high Ca concentrations (500 mg 1-' and above).Finally it should be mentioned that neither ArCa' nor Ca2+ could be observed. The detection of such species is however hampered by the presence of a background peak (Ar2+). Chlorine The spectral interferences arising from C1 are summarized in Table 4. The important formation of C10' and ClOH' should first be noted. Since C1 has two isotopes (mass numbers of 35 and 37 respectively) these polyatomic species can be found in the mass spectrum between m/z 51 and 54. The occurrence of such polyatomic ions substantially hampers the determination of two elements in a Cl matrix V and Cr. Both "V (99.7%) which is almost monoisotopic and ',Cr (9.5%) suffer from interferences from 35C1160 + and 37Cli60+ respectively. The high interference factors found 29 and 88 respectively indicate the importance of these interferences.In addition they are observed at a C1 concentration of only 5 mg 1-'. The combi- nation of Cl with less abundant isotopes of 0 170 and l80 contributes to a lesser extent to an increase in the blank signal. These combinations only play a significant role at Cl concen- trations of 500mg1-' or higher. The ratio CIOHf:CIOf is about 3.2 x whereas the contribution of C1160H+ to the total signal (Cli60H+ and Cl170+) was established to be Secondly the formation of 35Cli60160+ was observed at a C1 concentration of 1 g 1- '. The size of this spectral interference is however very small analogous to that of CaO,'. Thirdly at C1 concentrations of 500 mg 1-' and higher a signal significantly different from the background signal was observed at m/z 49. This is attributed to the formation of 35C114N+ [spectral overlap with 49Ti (5.5%)] as suggested by Tan and H ~ r l i c k .~ The importance of this interference is limited because several other isotopes of Ti are available and because the size of the signal is very small. The ratio for ClN+ C10+ is about 2.5 x which is initially in contradiction with the ratio of the bond strengths (333.9 kJ mol-' 272 kJ mol-' = 1.23).8 However the concentration of N f and 0' in the plasma has to be taken into account when such polyatomic ions are formed. Gray' reported a concentration of 3 x 1014 cm-3 for 0' and only 2.5 x 10" cm-3 for N + (for 0.14 mol I-' HNO,).Fourthly an important CI interference problem is caused by the formation of ArCl' (ArCl' CIOf = 8.0 x This mainly hampers the determination of As in that 75As (100%) (monoisotopic) suffers from an interference from 40Ar35C1 +. In addition such polyatomic ions are observed at C1 concen- trations of 5-10mg1-' and higher. For the determination of Se the interference of 77Se (7.6%) with 40Ar37C1f can be %99%. Table 3 Calcium spectral interferences. Stable isotopes of Ca 40Ca (96.6%) 42Ca (0.65%) 43Ca (0.14%) 44Ca (2.08%) 46Ca (0.004%) and 48Ca (0.19%). The most important polyatomic ions are printed in italics. Values given are apparent analyte concentration (Fg 1-') &standard deviation of the mean n = 5 Ca concentration/mg I-' Analyte 56Fe (91.7%) 57Fe (2.20%) 58Ni (68.3%) 60Ni (26.1 Yo) 64Zn (48.6%) 72Ge (27.4%) 59c0 (100%) 5 10 50 100 500 <2.2 40f14 177f13 229f 15 372 f 22 - < 15 137f4 314f 11 1861 f 92 - <0.30 0.39f0.12 0.78k0.17 3.7 & 0.5 - - - < 0.04 0.24 & 0.09 - - <0.10 0.74+0.44 11.35 1.0 - < 0.47 - - - <0.10 0.35 2 0.15 - - - 1000 560 f 30 4270 f 140 7.3 & 0.6 1.1 &0.1 28.5 k 1.1 0.84 & 0.18 1.5f0.1 I F 650f 130 3710f 570 7.5 f 0.3 1 26$4 0.8 1.5 Table 4 Chlorine spectral interferences.Stable isotopes of C1 35Cl (75.8%) and 37Cl (24.2%). The most important polyatomic ions are printed in italics. Values given are apparent analyte concentration (pg 1-') fstandard deviation of the mean n=5 Analyte 49Ti (5.5%) 52Cr (83.8%) 53Cr (9.5%) 54Fe (5.8%) 55Mn (100%) 67Zn (4.1 %) 70Ge (20.5%) 75As (100%) 77Se (7.6%) 51v (99.7%) Interferen t 35C1'4N 35C1160 37C1'4N 35C1160H 35C1170 37C1160H 37C1170 3 7 ~ 1 1 6 0 3 7 ~ ~ 0 35~1160160 35C135C1 40Ar35 CI 40Ar37C1 C1 concentration/mg 1-' 1 5 - - < 0.07 0.17 & 0.02 < 0.26 0.40 f 0.04 - - - - - - < 0.08 0.14 f 0.03 - - 10 50 - <0.71 0.31 k0.02 1.2f0.1 < 0.03 0.96 f 0.04 4.2 f 0.2 - - - 0.19 0.02 0.76f 0.14 < 1.3 3.8 f0.5 100 1.4f.0.2 2.8k0.1 0.11 f.0.04 7.8 f 0.5 < 5.5 < 0.02 <0.10 1.6 f 0.2 7.6 f 2.5 - 500 7.5 5 0.4 15.5f.0.5 0.5f0.1 44.5 f 0.6 6.6 f 2.5 0.06 f 0.02 < 0.43 0.17 f 0.02 7.6 k 0.2 39.9 f 1.7 ~ 1000 15.0 + 0.7 33.8 5 2.6 1.2 * 0.1 99f 11 16.1 +4.7 0.07 & 0.03 3.0 + 0.4 0.30 + 0.09 16.0k 1.1 86+ 14 ~ I F 14.7 f 0.6 29k4 1.1 f O .1 8 8 2 9 16 0.07 3 0.3 15.6 f 0.5 7 9 f 5180 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 overcome by the use of other isotopes. Some of these (76Se 78Se and are however subject to interference by Ar,'. Problems encountered with the determination of As and Se have been reported for serum,1s12 ~ r i n e ~ ? ~ '-19 protein solu- t i o n ~ ' ~ and other biological materials such as bovine l i ~ e r . ' ~ ~ ~ Finally it can be noted that at very high Cl concentrations (> 500 mg 1-') the formation of Cl,' was observed. Phosphorus The spectral interferences arising from P are presented in Table 5. Firstly the formation of PO' and POH' is important. Since P is monoisotopic ("P) these polyatomic ions are situated at between m/z 47 and 49 interfering with three isotopes of Ti.These spectral interferences however do not hamper the determination of Ti in a P matrix because 46Ti (8.O%) which is free from interferences by P can be used. The ratio of POH' :PO' is about 2.8 x In addition the contribution of 31P180' to the total signal at m/z 49 is about 80%. The remaining part is due to the formation of 31P160H2+. A ratio for POH2' PO' of 5.1 x Secondly there is the important formation of PO2' at P concentrations of 5 mg 1-' and higher. This polyatomic ion influences the determination of Cu and to a lesser extent that of Zn. A ratio for PO,' :PO' of 8.3 x lo- was observed. Calculations show that the signal at m/z 64 can be attributed fully to the formation of 31P160160H' and not to that of 31P160170f (about 0.2%).The formation of a tetraatomic ion was also observed at m/z 79 i.e. PO,'. The intensity of the signal however is very small (PO,' PO' = 3.6 x Thirdly at high P concentrations (> 50 mg 1-I) the forma- tion of ArP+ was observed. This polyatomic ion influences the determination of Ga but its importance is rather limited ( I F = 3.5 and ArP' PO' = 5.5 x lop3). Mulligan et ~ 1 . ~ have reported on the spectral interferences of P (PO PO3 and H2P03) in the analysis of urine. was found. Potassium The spectral interferences arising from K are given in Table 6. Although K has three stable isotopes 39K (93.3%) 40K (0.01%) and 41K (6.7%) a significant spectral interference due to KOf was only observed at m/z 55. The formation of KO' is rather small in comparison with other XO' polyatomic ions owing to a low XO bond strength:8 276 kJ mol-' in comparison with 596.6 kJ mol-' for PO.In addition the detection limits for Fe at m/z 56 and 57 are relatively high 2.2 and 15 pg1-' respectively so that the formation of KOH' was not observed at all. At K concentrations of lOOmgl-' and higher the formation of ArK' was noticed at m/z 79 and 81 (spectral interferences with the two isotopes of Br) analogous to that of ArC1' and of ArP'. In addition to peaks due to the expected formation of KO+ and ArK' the mass spectrum of K also contained some other peaks. These are presented in Fig. 1 which gives the spectrum between m/z 90 and 100 for a 1 g I-' solution of K. As can be seen peaks are found at m/z 94 95 96 and 97 Taking into account the relative intensities of the peaks they can only be attributed to the formation of K20' and of ArKO'.Although these polyatomic ions have not been described in the literature other combinations seem most unlikely. A similar formation of Na20' and of ArNaO' was observed with solutions of Na (see later). Finally it should be pointed out that an experimental ratio for 39K39K'60+ 40Ar39K160' of 0.22 is in contrast with a ratio for 39K160f 40Ar39K' of about 1. As has been described by Lam and Horlick,20 such polyatomic ions are probably formed at the interface by collision of monatomic and/or polyatomic ions. In this way K,O' can be formed by collision of K with KO (the formation of K2+ was not observed the detection of such a species is however hampered by the presence of Ar2') whereas collisions of Ar with KO and of ArK with 0 can lead to the formation of ArKOf.Mulligan et aL4 have reported spectral interferences of K (KO and ArK) in the analysis of urine. Carbon The spectral interferences arising from C are summarized in Table 7. Two groups of spectral interferences were observed. The first group is due to the formation of polyatomic ions consisting of C and 0. Although not given in Table 7 CO' and COH' are predominantly formed at m/z values between 28 and 32. These interfere with all isotopes of Si. Besides CO' CO,' is also formed (at m/z 44 and 45) and to a lesser extent CO,' (m/z 60). The following ratios were observed 0.14 and 8.1 x for CO,' CO' and CO,' CO' respectively. In addition it can be calculated that the contribution of 12C160160H' to the total signal at m/z 45 is about 75% and that the ratio for C02H+ :CO,+ is 6.1 x The second group is due to the formation of polyatomic ions consisting of Ar and C. These polyatomic ions mainly affect the determination of Cr because two isotopes of Cr ',Cr (83.8%) and 53Cr (9.5%) are subject to interference by 40Ar12C+ and 40Ar1'3C' respectively at C concentrations of 100 mg 1-' and higher.The spectral overlap is however small ( I F values of 2.4 and 2.1 respectively). Nevertheless problems experienced with the determination of Cr in plasma protein solutions have been reported by Lyon et a[.' Carbon spectral interferences are much more severe when organic solutions are measured. An overview of the polyatomic ions obtained has been given by Hutton.21 Sodium The spectral interferences arising from Na are summarized in Table 8.The vicinity of the intense peak of 40Ar' hampered the detection of 23Na160f and this ion was therefore not investigated further. Firstly in contrast with other matrix elements such as P and C the formation of NaO,' was not observed probably because of the relatively low bond strength? 256.5 kJ rriol-' for NaO 596.6 kJ mol-I for PO and 1076.5 kJ mol-' for CO. Secondly there is the important formation of Na,' at Na concentrations of 50 mg 1-' and higher. The signal at m/z 47 can be attributed to the formation of Na,H+. A ratio for Na,H' Na,' of 9.8 x lo- was observed. Thirdly there is the remarkable formation of 23Na23Na160 ' with a ratio for NazO+ Na2' of 4.5. Fourthly the formation of 40Ar23Na+ was observed which affects the determination of Cu.A ratio for A.rNa' Na20' of about 1 was noted. Finally at m/z '79 the formation of 40Ar23Na160+ was observed. The intensity is however relatively small (ArNaO' ArNa' =; 5.8 x lo- and ArNaO+ NaNaO' = 4.9 x Lyon and Fell2 studied the spectral interferences of Na20 and of ArNa on Cu in serum in detail. It should also be noted that for the calculation of the interference factors for Na,' and Na,Of the square of the Na concentration must be used [IF = apparent analyte concentration x 106/(Na con- ~entration)~] and that for Na,H' and ArNa+ no relationship could be found. Sulfur The spectral interferences arising from S are given in Table 9. The formation of SN' mainly of mass 46 was not observed although this species has been suggested by Tan and H ~ r l i c k .~ It is clear from Table 9 that the combination of S with 0 gives rise to most of the spectral interferences. A first group of interferences is due to the formation of SOf and SOH'. Since S has four stable isotopes with mass number between 32 and 36 these polyatomic ions can be found between m/z 48 andTable 5 f standard deviation of the mean n = 5 Phosphorus spectral interferences. Stable isotope of phosphorus 31P (100%). The most important polyatomic ions are printed in italics. Values given are apparent analyte concentration (pg 1-I) P concentration/mg I-' Anal yte 47Ti (7.3%) 48Ti (73.8%) 49Ti (5.5%) 63Cu (69.2%) 64Zn (48.6%) Interferent 31~160 3 1 ~ 1 6 0 ~ 3 1 ~ 1 7 0 31~180 3 i p l 6 0 ~ ~ 31p160160 3 1 p 1 6 0 1 6 0 ~ 31p160170 31p160180 2 40Ar31 P 31p160160160 0.1 0.5 1 5 - - < 0.04 0.13f0.02 - - < 0.02 0.59 f 0.02 <0.61 4.3 f 0.5 8.2 0.4 41.9 f 0.3 - - - - 10 78.2 f 1.8 0.24 f 0.01 <0.35 0.90 & 0.04 50 366 & 8 0.99 k 0.05 0.86f0.16 4.1 f0.2 100 659 & 24 1.7 & 0.1 1.9f0.2 7.5 f 0.5 500 3060 f 160 7.7 f 0.5 8.0k 1.4 30.8 f 0.4 1000 6410 f 380 14.9 f 1.2 17.0f 1.5 59.6k2.5 I F 6850f 700 17f2 17_+ 1 65f8 0.46 f 0.04 < 0.04 - - 1.4f0.2 < 0.01 0.17f0.02 < 0.3 1 2.8 k 0.3 0.13 f 0.03 0.38 & 0.06 0.93 f 0.12 11.0f0.4 0.94 f 0.17 1.7k0.2 4.9 _+ 0.4 23.4f1.0 1.73 f 0.04 3.4 f 0.3 9.7 f0.3 2523 1.6 k0.3 3.5 * 0.2 9.6 f 0.3 65Cu (30.8%) 71Ga (39.9%) 79Br (50.7%) Table 9 concentration (pg 1-I) fstandard deviation of the mean n=5 Sulfur spectral interferences.Stable isotopes of S 32S (95.0%) 33S (0.75%) 34S (4.2%) and 36S (0.02%). The most important polyatomic ions are printed in italics. Values given are apparent analyte 0.05 0.1 0.5 1 0.27k0.07 0.44k0.08 0.98f0.05 1.5f0.1 - < 1.7 - - < 0.43 4.1 f0.3 - - 5 5.7f0.1 3.2 f 0.3 5.0 f 0.3 10 10.8 * 0.1 4.5 f 1.0 8.7 & 1.9 7.8 f 0.2 1.8fO.1 - 50 50.2 _+ 0.7 10.3 _+ 0.6 33.5 & 2.2 100 90.2 f 1.6 15.5 k0.6 57.6 f 1.7 < 0.02 61.2 f 0.4 8.6 0.4 500 445 f 9 1000 842 _+ 8 144f3 511f8 0.30 f 0.08 500f 5 64.8 f 2.5 I F 980f 120 150+6 580 k 70 0.3 6302 110 80f 15 Analyte 48Ti (73.8%) 49Ti (5.5%) 50Ti (5.4%) 52Cr (83.8%) 64Zn (48.6%) "Zn (27.9%) 6 5 C ~ (30.8%) F W W "P 75.9 k 1.4 272 f 2 0.14f0.03 275 f 8 36.3 f 0.5 0.33f0.08 0.7010.15 1.4f0.3 2.1f0.2 - - - < 0.09 4.5 f 0.4 1.2f0.1 34.2 f 0.8 4.9 T 0.3 1.3 f 0.2 - 1.7 f 0.3 < 3.5 - 3.9 & 0.2 5.5 f 1.5 < 0.69 < 0.5 1 <0.36 < 1.6 6.8 f 0.2 6.7 f 0.9 2.0 _+ 0.6 0.69 f 0.05 1.6 f 0.3 5.3 f 1.0 24.7 * 0.7 17.6 f 3.2 2.5f0.1 1.7 & 0.4 8.1 f0.2 9.8f4.0 44.0 _+ 0.4 23.4 _+ 2.0 3.1 f0.4 2.1 k0.2 15.4 f 0.3 12.6 +_ 2.7 60f 16 29f8 4.1 f 1.3 3fl 15.9 f 0.4 16+5 67Zn (4.1 YO) 68Zn (18.8%) 72Ge (27.4%) 'lBr (49.3%) 82Se (9.2%)182 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 Table 6 are printed in italics. Values given are apparent analyte concentration (pg 1 - I ) f standard deviation of the mean n = 5 Potassium spectral interferences. Stable isotopes of K 39K (93.30/b) 40K (0.01%) and 41K (6.7%).The most important polyatomic ions K concentration/mg 1-' Anal yte "Mn (100%) 79Br (50.7%) 81Br (49.3%) 95Mo (15.9%) 96Mo (16.7%) 9 4 ~ 0 (9.3%) 9 7 ~ 0 (9.6%) In terferen t 40Ar39K 40Ar41K 3 9 ~ 1 6 0 3 9 ~ 3 9 ~ 1 6 0 4 0 ~ ~ 3 9 ~ 1 6 0 3 9 ~ 4 1 ~ 1 6 0 4 0 ~ ~ 4 1 ~ 1 6 0 10 50 100 < 0.08 0.1 1 + 0.02 0.29 f 0.04 - < 0.45 0.57 f 0.13 - < 1.3 - - <0.11 - < 0.04 0.17 f 0.05 - - < 0.05 - < 0.02 0.08 f 0.03 - 500 2.9 & 0.2 17.2 f 4.1 3.2 f 0.5 1.30f0.13 9.0 f 1.6 0.32 f 0.08 2.4 f 0.4 1000 5.7 * 1.1 59518 9.0f 1.0 4.0 rt 1.2 38 k 14 0.99 f 0.34 9.5 & 3.5 I F 5.8f0.1 47k17 9 4 38 1 10 500 I 400 - - a C I u a a 300 - L v) 200 - +- =I 0 u 100 - 90 91 92 93 94 95 96 97 98 99 100 m/z Fig. 1 Mass spectrum between 90 and 100 m/z of a 1 g 1-1 K solution 53.This causes a dramatic spectral overlap with three isotopes of Ti including the most abundant isotope 48Ti (73.8%). The interference factors vary from 150 to 980. The important formation of SO' can be explained by the relatively high bond strength' (521.7 kJ mol-I) which is comparable to that of P (596.6 kJ mol-l). A ratio for SOH' :SO' of 5.6 x was found. A second group contains the polyatomic ions SO,' and S02H' which can be found at between m/z 64 and 69. They mainly hamper the determination of Zn because almost all Zn isotopes are subject to interference. In addition the intensity of SO,' is remarkable a ratio for SO2' SO' of 0.31 was noted. This is higher than for similar XO,':XO+ ratios. Examples are ratios of 8.3 x for PO2+ PO' and of 0.14 for CO,' CO'.Although Tan and Horlick suggested that in addition to SO,' S2' can be formed it can be seen from the present data that a significant contribution due to the formation of such species is dubious. Up to a S concentration of 1 g I-' a linear response between the S concentration and the signal at m/z 64 (64Zn s9Co) was obtained indicating that SO2' is formed almost exclusively. A third group of polyatomic ions contains SO,' and SO,Hf. The most important polyatomic ion of this group namely 32S160160160' could not be detected because of an intense spectral overlap with 40Ar40Ar'. Less intense peaks were seen at m/z 81 and 82. For SO,' :SO+ a value of 1.9 x lop2 was found. Finally at S concentrations of 100 mg 1-1 and higher the formation of ArS' was also observed. The intensity of this signal is however small (ArS' SO' = 8.8 x Spectral interferences due to S (interfering with Cu Zn and Br) have been reported for human ~ e r u m ~ ~ and urine.4 Formation and Stability of Polyatomic Ions A study of the formation and short- and long-term stability of polyatomic ions was carried out in order to evaluate the use of some simple correction procedures to overcome spectral overlaps.The signals for s9Co+ (as a reference) CIO' SOz' and ArCl' m/z 51 64 and 75 respectively as a function of time Table7 Carbon spectral interferences. Stable isotopes of C "C (98.8%') and I3C (1.1Y0). The most important polyatomic ions are printed in italics. Values given are apparent analyte concentration (pg 1-l) f standard deviation of the mean n = 5 C concentration/mg I-' Analyte 44Ca (2.08%) 48Ti (73.8%) 49Ti (5.5%) "Cr (83.8%) 53Cr (9.5%) 6oNi (26.1 %) 45sc (100%) Interferent 12~160160 1 2 ~ 1 6 0 1 6 0 ~ 1 3 ~ 1 6 0 1 6 0 36Ar1 ' C 36 Ar 13C 40Ar12C 40Ar'3C 1 2 ~ 1 6 ~ 1 6 ~ 1 6 ~ 10 50 100 500 < 14 43.8 +:!.7 96.6-+_ 5.8 501 f 19 < 0.06 0.18 + 0.01 0.29 2 0.07 1.0 _+ 0.1 - - < 0.04 0.10f0.04 - - <0.35 1.7f0.5 < 0.02 0.13 +0.01 0.28 f 0.09 1.1 f0.2 - <0.1.5 0.66 0.07 1.2f0.3 - <0.10 - - 1000 985 k 39 2.0k0.1 0.19 & 0.02 2.4 0.3 2.2 * 0.2 1.7k0.3 0.90 k 0.15 I F 957 f 56 2.6 k 0.8 0.2 3 2.4 f 0.3 2.1 f 0.5 0.9 Table 8 Sodium spectral interferences. Stable isotope of Na 23Na (100%).The most important polyatomic ions are printed in italics. Values given are analyte concentration (pg 1-I) &standard deviation of the mean n = 5 Na concentration/mg 1-' Analyte Interferen t 10 50 100 500 1000 I F 46Ti (8.0%) 23 Nuz3 Nu <0.13 1.1 k0.3 6.7 f 2.3 189 k 31 589f81 610f 140" 47Ti (7.3%) 23 Nuz3 Nu H - <0.10 0.89k0.17 3.6k0.1 4.9 f 0.5 62Ni (3.59%) 23Na23Nu160 < 1.1 1.8 f 1.1 18.2 f 7.0 609 k 80 1890f270 2050 k 340* 2.7 f 1.1 92+ 12 281 + 30 63Cu (69.2%) 40Ar23Nu <0.16 ?3r- (50.7%) 40Ar23Na160 - - < 0.45 1.7 k0.4 5.9 +_ 1.6 6 - * (Concentration matrix element)' x x interference factor ( I F ) = apparent analyte concentration.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 183 (the first 2 min after the start of aspiration of the corresponding solution) are given in Fig. 2. A first significant signal is obtained for both 5 9 C ~ f and the polyatomic ions after 40-50s.The signals due to the polyatomic ions however are less stable and fluctuate to a larger extent the RSD calculated between the first and the second minute amounted to 0.86% for 59C0+ 1.1% for ClO' 3.1% for SO,' and 1.9% for ArCl'. The behaviour described can be extrapolated to almost all poly- atomic ions (the exceptions will be described further). The larger fluctuation on the signals from polyatomic ions indicates that temperature fluctuations in the plasma changing sampling conditions etc. exert a different influence on the formation of monatomic and polyatomic ions. The behaviour of some polyatomic ions is different from that described in Fig. 2. This can be seen in Fig.3 where the signals for 59C0+ (as a reference) Na,' Na,O' and ArNa+ (m/z 46 62 and 63 respectively) as a function of time (the first 2 min after the start of aspiration of the corresponding solution) are presented. In contrast to most of the polyatomic ions no constant and stable signal is obtained for polyatomic ions of Na after the observation of the first significant signal. In addition at that moment the signal increases almost linearly for at least 1 min. It is remarkable that this behaviour was only observed for polyatomic ions containing Na or K (similar observations were made for ArK+ at m/z 79). Both matrix elements have a low ionization energy 5.14 and 4.34eV respectively. The behaviour of ArNa+ (m/z 63) has been confirmed by Lyon and They proposed the condensation of Na on the surface of the interface with an accumulation of deposit on the surface as the run proceeds as the origin of the established behaviour.The ArNa' is subsequently produced and transported from the surface to the supersonic jet at the interface where it is detected by the mass spectrometer. The A 0 60 120 Timels Signal intensity for A 59C0; B ArC1' (m/z 75) (x 5); C SO2+ (mjz 64) (x 0.1); and D C10' (m/z 51) as a function of time (the first 2 min after the start of aspiration of the corresponding solution) 0 30 60 90 120 Time/s Fig. 3 Signal intensity for A s9C0; B Na2' (m/z 46); C ArNa' (m/z 63); and D Na20+ (m/z 62) as a function of time (the first 2 min after the start of aspiration of the corresponding solution) interface as the origin of species such as ArX+ has also been suggested by Lam and Horlick.,' A similar explanation can be given for the formation of Na,' Na,O+ and ArK'.A more or less constant signal for Na,' Na,O' and ArNa+ (m/z 46 62 and 63 respectively) is obtained after an aspiration time of about 10 min as shown in Fig. 4. For the long-term experiment (several hours) a stable and reproducible signal was obtained for ClO' SO,' and ArC1+ (m/z 51 64 and 75 respectively) whereas the signals of Na,' Na,O+ and ArNa' (m/z 46 62 and 63 respectively) were subjected to short- and long-term fluctuations. Reduction of and/or Correction for Spectral Interferences In order to reduce or to correct for some of the spectral interferences encountered in the analysis of biological materials several procedures have been proposed.Some of these are given here and illustrated. In several cases an isotope subject to less or no interference can be used. An example is the determination of Zn in human serum via the 68Zn (18.8%) isotope ( I F 4.1) which has less interference and not via the 64Zn (48.6%) isotope ( I F 630). Since the concentration of S in human blood plasma24 is about 1 81-' it can be calculated from Table 9 that the following apparent Zn concentrations are expected 630 60 29 and 4 pg I-' when 64Zn 66Zn 67Zn and 68Zn respectively are used for the Zn determinations. Comparison of these values with the Zn concentration in human serum2' (about 1 mg l-') shows that the spectral overlap for 68Zn is only about 0.4% and thus can be neglected. This was confirmed by the analysis of a reference serum reported by Vandecasteele et a1.26 For spectral overlap with polyatomic ions containing H N 0 and/or Ar (present in the plasma) a correction can usually be made by the use of a blank solution containing the solvent.However one should be aware of the fact that the extent to which species such as Ar,' and ArO' are formed is dependent on the total matrix composition of the sample solution.'' An example was given by Vandecasteele et al.27 who accurately determined Br in human serum. Both bromine isotopes 79Br and "Br were corrected for the Ar,H + overlap by subtraction of the blank. Vanhoe et aL6 evaluated the use of a so-called simulated blank solution to correct for spectral interferences owing to the serum matrix elements Na S and Ca. This solution contained the same amount of the interfering matrix elements as the serum solution that was analysed.Accurate results were obtained for Fe Co Cu and Zn. They concluded that accurate and precise results can only be obtained when the correction factor is limited to less than 10% of the total signal. A similar method was used by Friel et a[.,* to correct for Ca and C1 spectral interferences during the analysis of several biological reference materials. The use of correction formulae based on the isotopic abun- 25 n a I I I I 0 4 8 12 16 20 Time/m in Fig. 4 Signal intensity for A ArNa' (m/z 63); B Na,' (m/z 46); and C Na20+ (m/z 62) as a function of time (19 min)184 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 dances of the interfering matrix element is another method used to correct for spectral overlap.Spectral interferences arising from C1 can be corrected for in this way; "V and 53Cr are subject to interference from 35C1160+ and 37C1160+ respectively (Table 4). The ratio of these two interferences is comparable to the ratio of the relative abundances of 3sCl and 37Cl (75.8% 24.2% = 3.132). So the contribution of 35C1160f at m/z 51 (in order to determine V) can in the absence of Cr be calculated from the 37C1160+ value at m/z 53. This correc- tion method has been applied to lobster hepatopancreas by Ridout et al.,29 whereas Park et determined Cr in human serum [National Institute for Standards and Technology (NJST) Standard Reference Material (SRM) 909 Freeze-Dried Human Serum] neglecting the presence of V in the sample.The interference of 40Ar35C1+ on 75As can be corrected for in an analogous way. The 40Ar35C1+ and 40Ar37Clf give an interference with 75As and 77Se respectively (Table 4). This method was also applied successfully to lobster hepatopancreas by Ridout et a1.29 Another illustration of this correction method is given in Table 10. Selenium was determined in a reference serum. Since the Cl concentration in human blood plasma24 is about 3.55 g l-' it can be deduced from Table 4 that the spectral interference of 40Ar37C1+ on 77Se gives an apparent Se concentration of about 280 pg I-' (compared with a serum Se c~ncentration~~ of about 100 pg 1-l). A correction for this overlap was made in two ways. Before the analysis of 8-fold diluted serum samples the ratios of 40Ar35C1+ 40Ar37Cl+ and 35C1160+ 40Ar37C1+ were determined experimentally with 1 % HCl.The ratios found were 2.960 and 18.90 respectively. In this way the mass fractionation introduced by the ICP mass spectrometer was taken into account. These ratios did not differ significantly from those obtained by spiking the diluted serum samples with an appropriate amount of HCl and were used to correct for the 40Ar37C1+ interference on 77Se. Both results agree well with the certified value. The errors introduced by the presence of As and V in the serum solution could be calculated from the certified concentrations and were 1 and less than 0.1% respectively and thus could be neglected. Another method of correcting for spectral interferences is based on the use of multivariate correction methods.A four- component principal component analysis technique has been applied by Vaughan and Templeton7 and by Xu et ~ 1 . ~ ~ to separate Ca Na and K interferents from the Ni data in order to determine Ni accurately in urine and serum. In addition to the use of blank solutions or simple equations a more extensive sample preparation procedure can also be applied. Separation of the analyte from the interfering matrix element is the most popular technique. Lyon and co-workers10*22 used size-exclusion chromatography (gel fil- tration) to de-salt the protein solution to be analysed in order to determine Se in serum. This method offers the potential of separating the proteins from the inorganic salt fraction of a serum sample.Sheppard and c o - w o r k e r ~ ~ " ~ ~ used ion chroma- tography to eliminate the interference due to ArC1+ on the determination of As in urine whereas Heitkemper et ~ 1 . ' ~ and Beauchemin et a1.33 coupled high-performance liquid chroma- tography with ICP-MS for the speciation of As compounds in urine and a dogfish muscle reference material respectively. At the same time they avoided the ArCl+ interference on As. Table 10 Results for Se for the 'second-generation' biological reference material Freeze-Dried Human Serum Se as "Se Correction with 40Ar35C1 Correction with 35Cl'60 Certified value3' Se concentration/pg I-' 95.1 (1.7)* 100.3 (4.5)t 95.5 f 4.61 * Starrdard deviation n=4. 7 Standard deviation n = 5. 95% confidence limits. Goossens and C O - W O I T ~ ~ ~ S .~ ~ ' ~ ~ used an anion-exchange resin column (Dowex-1) to separate Cl and S from the trace metals in serum and urine. In this way they were able to determine Cu Zn As and Se in serum and As and Se in urine. Plantz et al.34 used an on-line sample treatment method to separate the analyte elements (V Cr Ni Co and Cu) from alkali and alkaline earth elements and anions. The separation consisted of complexation of the metals mentioned with the reagent bis(carboxymethy1) dithiocarbamate. The proposed method was applied to the analysis of a urine standard. Janghorbani et al.35 developed a chemical separation of Br from S based on distillation from acidified solutions to eliminate the spectral overlap with S03Ht. Correct Br isotope ratios could be obtained.Serfass et Amarasiriwardena et and Patterson et used several extraction procedures to separate Zn from the matrix. In this way they were able to determine the Zn content and the Zn isotope ratios accurately in a number of biological materials such as blood plasma faeces and urine. An on-line anodic stripping voltammetry system was coupled with ICP-MS by Pretty et ~ 1 . ~ ' to avoid the Na spectral interferences allowing the determination of Cu in urine. The use of a different sample introduction method is often a good alternative. The most promising technique is electro- thermal vaporization (ETV). Since the sample is introduced into the plasma in the absence of any accompanying solvent the levels of certain polyatomic ions can be significantly reduced.Whittaker et aL4' showed the possibility of the accurate determination of isotope ratios for Fe in serum by ETV-ICP-MS. In addition the use of an ashing step and/or the addition of a chemical modifier can eliminate some matrix elements. Carey et showed the selective elimination of C1 by the addition of NH40H (with the formation of volatile NH4C1). Another possibility is the use of hydride generation (HG) to separate C1 from As and Se to determine both elements in biological materials. Ting et al.42 and Buckley et ~ 1 . ~ ~ have reported the successful determination of Se in several biological materials by HG-ICP-MS. More recently mixed-gas plasmas have been investigated to reduce or eliminate some spectral interferences. With the addition of N to the aerosol or the plasma gas Branch et ~ 1 .l ~ and Wang et a1.18 were able to determine As in urine accurately. Hill et ~ 1 . ' ~ have determined V As and Se in several biological materials by the addition of methane to the nebulizer gas An alternative method is the addition of an organic solvent to the sample. Goossens et d." reported on the accurate determi- nation of As and Se in serum and urine by the addition of 4% ethanol to the sample in combination with careful adjustment of the aerosol gas flow rate. Finally the most effective way to overcome spectral inter- ferences is of course the use of a mass spectrometer with sufficient resolution to resolve peaks that have similar m/z values. Bradshaw et ~ 1 . ~ ~ reported a double focusing magnetic sector mass spectrometer that was able to achieve a resolution up to 10000.So far no applications with biological materials have been reported. References Evans E. H. and Giglio J. J. J. Anal. At. Spectrom. 1993 8 1. Vaughan M. A. and Horlick G. Appl. Spectrosc. 1986 40 434. Tan S. H. and Horlick G. Appl. Spectrosc. 1986 40 445. Mulligan K. J. Davidson T. M. and Caruso J. A. J . Anal. At. Spectrom. 1990 5 301. Vanhaecke F. Vanhoe H. Dams R. and Vandecasteele C. Talanta 1992 39 737. Vanhoe H. Vandecasteele C. Versieck J. and Dams R. Anal. Chem. 1989 61 1851. Vaughan M. A. and Templeton D. M. Appl. Spectrosc. 1990 44 1685. Handbook of Chemistry and Physics ed. Weast R. C. Astle M. J. and Beyer W. H. CRC Press Boca Raton 64th edn. 1984.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 185 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Gray A. L. Spectrochim. Acta Part B 1986 41 151. Lyon T. D. B. Fell G. S. Hutton R. C. and Eaton A. N. J. Anal. At. Spectrom. 1988 3 601. Goossens J. Vanhaecke F. Moens L. and Dams R. Anal. Chim. Acta 1993 280 137. Goossens J. Moens L. and Dams R. J. Anal. At. Spectrom. 1993 8 921. Lyon T. D. B. Fell G. S. Hutton R. C. and Eaton A. N. J. Anal. At. Spectrom. 1988 3 265. Heitkemper D. Creed J. Caruso J. and Fricke F. L. J. Anal. At. Spectrom. 1989 4 279. Sheppard B. S. Shen W.4 Caruso J. A. Heitkemper D. T. and Fricke F. L. J. Anal. At. Spectrom. 1990 5 431. Branch S. Ebdon L. Ford M. Foulkes M. and O’Neill P. J. Anal. At. Spectrom. 1991 6 151. Sheppard B. S. Caruso J. A. Heitkernper D.T. and Wolnik K. A. Analyst 1992 117 971. Wang J. Evans E. H. and Caruso J. A. J. Anal. At. Spectrom. 1992 7 929. Hill S. J. Ford M. J. and Ebdon L. J . Anal. At. Spectrom. 1992 7 1157. Lam J. W. H. and Horlick G. Spectrochim. Acta Part B 1990 45 1327. Hutton R. C. J. Anal. At. Spectrom. 1986 1 259. Lyon T. D. B. and Fell G. S. J. Anal. At. Spectrom. 1990 5 135. Goossens J. and Dams R. J. Anal. At. Spectrom. 1992 7 1167. International Commission on Radiological Protection Report of the Task Group on Reference Man ICRP Publication 23 Pergamon Press Oxford 1975. Versieck J. and Cornelis R. Trace Elements in Human Plasma or Serum CRC Press Boca Raton 1989. Vandecasteele C. Vanhoe H. and Dams R. J Anal. At. Spectrom. 1993 8 781. Vandecasteele C. Vanhoe H. Dams R. and Versieck J. Anal. Lett. 1990 23 1827. Friel J. K. Skinner C. S. Jackson S. E. and Longerich H. P. Analyst 1990 115 269. 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 Ridout P. S. Jones H. R. and Williams J. G. Analyst 1988 113 1383. Park C. J. Smith D. C. and Vanloon J. C. Trace Elements Med. 1990 7 103. Versieck J. Vanballenberghe L. De Kesel A. Hoste J. Wallaeys B. Vandenhaute J. Baeck N. Steyaert N. Byrne A. R. and Sunderman F. W. And. Chim. Acta 1988 204 63. Xu S. X. Stuhne-Sekalec L. and Templeton D. M. J. Anal. At. Spectrorn. 1993 8 445. Beauchemin D. Siu K. W. M. McLaren J. W. and Berman S. S. J. Anal. At. Spectrom. 1989 4 285. Plantz M. R. Fritz J. S. Smith F. G. and Houk R. S. Anal. Chem. 1989 61 149. Janghorbani M. Davis T. A. and Ting B. T. G. Analyst 1988 113 405. Serfass R. E. Thompson J. J. and Houk R. S. Anal. Chim. Acta 1986 188 73. Amarasiriwardena C. J. Krushevska A. Foner H. Argentine M. D. and Barnes R. M. J. Anal. At. Spectrom. 1992 7 915. Patterson K. Y. Veillon C. Moser-Veillon P. B. and Wallace G. F. Anal. Chim. Acta 1992 258 317. Pretty J. R. Blubaugh E. A. Evans E. H. Caruso J. A. and Davidson T. M. J. Anal. At. Spectrom. 1992 7 1131. Whittaker P. G. Lind T. Williams J. G. and Gray A. L. Analyst 1989 114 675. Carey J. M. Evans E. H. Caruso J. A. and Shen W.-l. Spectrochim. Acta Part B 1991 46 1711. Ting B. T. G. Mooers C. S. and Janghorbani M. Analyst 1989 114 667. Buckley W. T. Budac J. J. Godfrey D. V. and Koenig K. M. Anal. Chem. 1992 64 724. Bradshaw N. Hall E. F. and Sanderson N. E. J. Anal. At. Spectrom. 1989 4 801. Paper 3/043 15 K Received July 21 1993 Accepted October 15 1993
ISSN:0267-9477
DOI:10.1039/JA9940900177
出版商:RSC
年代:1994
数据来源: RSC
|
15. |
Application of inductively coupled plasma mass spectrometry to the certification of reference materials from the Community Bureau of Reference |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 187-191
Luc Moens,
Preview
|
PDF (623KB)
|
|
摘要:
JOURNAL O F ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 187 Application of Inductively Coupled Plasma Mass Spectrometry to the Certification of Reference Materials from the Community Bureau of Reference* Luc Moens Hans Vanhoe Frank Vanhaecke Jan Goossens Michael Campbell and Richard Dams Ghent University Laboratory of Analytical Chemistry Proeftuinstraat 86 B-9000 Ghent Belgium Inductively coupled plasma mass spectrometry (ICP-MS) is widely used for applied analysis. The application of a method to the certification of reference materials is a severe test of its accuracy and precision. We have been using ICP-MS for the certification of Community Bureau of Reference (BCR) reference materials since 1988. This paper indicates that while the method proved to be potentially accurate and precise accuracy and precision cannot be taken for granted and that each analysis requires a thorough study in order to obtain results that meet BCR standards.Under these conditions however ICP-MS can compete with well established reference techniques such as neutron activation analysis. Keywords lnductively coupled plasma mass spectrometry; certification; certified reference materials Since 1983 when instruments for inductively coupled plasma mass spectrometry (ICP-MS) became commercially available the method has been rapidly accepted in both research and application laboratories. The ICP-MS technique is widely used for routine analysis and research is currently being carried out to improve the analytical performance and to extend the field of application of the method.' Most of this research is however concerned with methodological and technical novel- ties new sample introduction methods are being designed and techniques are being developed for the reduction or elimination of spectral and non-spectral interferences.2 The accuracy of the analytical results obtained seems to be taken for granted.The application of ICP-MS for the certification of reference mate- rials is a means of testing this assumption. From the 1970s onwards work has been carried out in this laboratory on the certification of trace element concentrations in the certified reference materials (CRMs) issued by the Community Bureau of Reference (BCR) of the Commission of the European Communities. Considering the 3 1 environmental and food reference materials certified in the period 1982-1990 this laboratory contributed to the certification of about 67% of all certified trace element^.^ Most of the data were obtained with neutron activation analysis (NAA).3 From 1987 onwards when the equipment for ICP-MS became available it was tested and later used for certification work.The contribution of ICP-MS applied in this laboratory to BCR certification work between 1990 and 1993 is shown in Table 1. Except for CRM 090 (doped Ti) all the materials are of biological origin or were certified for their environmental importance. The ICP-MS method was used to test the homo- geneity of the distribution of B in CRM 090 (doped Ti) and was also used in three intercomparison studies organized by BCR. Two of these studies concerned the determination of Pb in different types of wine; in a first inter-comparison only European laboratories took part whereas in a second results of European specialist laboratories were compared with data from laboratories in America and Canada.In a third study trace elements were determined in estuarine water from the river Tejo (Portugal). The analysis and certification of the reference materials mentioned above showed that ICP-MS is sufficiently accurate and precise for BCR certification work. With some exceptions the results obtained with ICP-MS were accepted for the certification on the basis of both their accuracy and precision. This general conclusion is illustrated in Fig. 1 which shows the results that were finally used for the certification of the * Paper presented at the XXVIII Colloquium Spectroscopicum Internationale (CSI) York UK June 29-July 4 1993.concentration of Na in BCR CRM 399 (Fre~hwater).~ Results obtained using ICP-MS are in good agreement with those from well established atomic absorption and emission methods and from instrumental NAA. Moreover the ICP-MS value coincides with the certified value and the precision compares favourably with the precision obtained with other techniques. In this paper the problems associated with the use of ICP-MS in BCR certification work are discussed and methods for overcoming these problems are presented. It is shown that in order to obtain the highest accuracy and precision a detailed study of the possible sources of error must be carried out.Experimental All measurements were carried out with a VG PQI ICP-MS instrument (Fisons VG Elemental Winsford UK) equipped with a Fassel torch a Gilson Minipuls-2 peristaltic pump a Meinhard type Tr-30-A3 concentric glass nebulizer and a double-pass Scott-type spray chamber with surrounding liquid jacket the temperature of which was controlled with a recircul- ating refrigerating-heating system. Sampling cones ( 1.0 mm orifice) and skimmers (0.75mm orifice) were made of nickel. Typical operation conditions are summarized in Table 2. Solutions were prepared with Milli-Q water (Millipore Bedford MA USA) and use was made of reagents of the highest available purity. The nitric acid that was used in most r 0 341 T ICP I NAA T L g 2 6 1 0 Method 0 Fig. 1. Determination of Na in BCR CRM 399 (Freshwater).Concentrations and 95% confidence intervals obtained with different methods (pg g-') and used in the final certification188 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 1 Contribution of ICP-MS at Ghent University Belgium to the certification of trace element concentrations in BCR reference materials CRM 101 090 303 304 398 399 403 414 422 142R 143R 145R 063R Product Spruce needles” Titanium” Human serum13 Human serum13 Freshwater (low ~ o n t e n t ) ~ Freshwater (high ~ o n t e n t ) ~ Sea water14 Plankton’’ Cod muscle16 Light sandy soil Light sandy soilt Sewage sludge-amended Sewage sludge-amended7 Sewage sludge Sewage sludget Skimmed Milk Powder Analyte Mg B Nb W Li Mg Li Mg Na Mg Al C1 Ca Mn Na Mg Al CI Ca Mn Fe Mo Mn Cu Zn Cd Pb Mn Cu Cd Pb Mn Cu Pb Mn Cu Zn Cd Pb Mn Cu Zn Cd Hg Pb Mn Cu Zn Cd Pb Mn Cu Zn Cd Hg Pb I Hg Pb Year 1990 1992 1992 1992 1992 1992 1992 1993 1993 * * * * * * * - - - - - - - *Final report in preparation.?Aqua regia soluble fraction. Table 2 Typical operation conditions of the VG PQI spectrometer Plasma r.f. power/” Forward Reflected Plasma Intermediate Nebulizer Gas flow rates/l min-’ Sample uptake rate/ml min-’ Spray chamber temperature/”C Ion sampling depth Vacuum/hPa Expansion stage Intermediate stage Analyser stage 1350 < 5 13.5 0.5-1 0.725-0.8 0.9 10 10 mm (from load coil) 2.4 10-4 4.8 x experiments was purified in the laboratory by sub-boiling distillation. Standards were prepared by dissolving high-purity metals or compounds with known stoichiometry.For isotope dilution studies isotopically enriched certified reference mate- rials were obtained from Isoflex (Consett County Durham UK) for Hg (Hg metal enriched in 201Hg) and from the National Institute of Standards and Technology (NIST Washington USA) for Pb [Pb metal enriched in 206Pb stan- dard reference material (SRM)983 Lead Radiogenic Isotopic]. The certified isotopic composition of spikes was always checked against an independent reference material or against a material with natural abundance. Results and Discussion Choice of Internal Standard It is common practice to add an internal standard to sample standard and blank solutions to correct for matrix-induced signal suppression or enhancement for signal drift and for instrumental instability.It is also known that in order to obtain accurate results the internal standard should closely match the analyte in terms of mass number and ionization potential. For the VG PQI mass matching turned out to be much more important than ionization potential matching. However mass matching between internal standard and analyte does not necessarily guarantee that the internal stan- dard will appropriately correct for matrix effects. In the BCR intercomparison study of the determination of Pb in wine the wines were analysed without any sample pre-treatment except for a 10-fold dilution with 1% nitric acid.’ Fig. 2 shows the I I I I I 0 0.5 1 .o 1.5 2.0 2.5 3.0 Ethanol concentration (%) 1.00 - Fig. 2. Determination of Pb in wine showing signal enhancement in aqueous ethanol solutions relative to 1% nitric acid solutions for different potential internal standards A ’‘’TI; B 184W; C 209Bi; and D I9’Pt.Data normalized to the enhancement for 208Pb enhancement of the signals of four potential internal standards when going from a 1% nitric acid solution to solutions containing in addition an increasing amount of ethanol. The enhancements shown have been normalized to the enhance- ment obtained for 208Pb. Similar matrix effects due to ethanol can be expected for the diluted wine samples. Only for 205Tl does the signal enhancement equal the enhancement observed for 208Pb within the experimental uncertainty. Bismuth-209 which is often used as an internal standard for Pb determi- nations was found to be unsuitable in the experimental arrangement used in this work since its use leads to an error of more than 10%.The observed effect cannot be attributed to a difference in the ionization potential since the latter is 7.4 eV for Pb which is closer to the 7.3 eV of Bi than to the 6.1 eV of T1. Similar observations were made for medium mass ranges.6 For certification work it is therefore necessary to test different internal standards to make sure that suppression or enhancement of the analyte signal is correctly compensated for by the internal standard. Both the accuracy and the precision are improved by matching the mass of the analyte and the internal standard. This is demonstrated in Fig. 3 where the precision of the Y signal is compared with the precision of the ratio of the same signal to the signal of the internal standard.Results have been plotted for different internal standards as a function of the mass of the internal standard. It is clear from Fig. 3 that whenJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 189 12 m 1 10 - 8 n m c 6 4 2 I 1 I I 0 50 100 150 200 250 Mass number of the internal standard Fig. 3. Relative standard deviations (YO) for the signal of @ "Y and a signal ratio ("Y :internal standard) as a function of the mass number of the internal standard. Results of a long-term stability experiment (mass scanning mode) the difference between the mass of the internal standard and the mass of the analyte (Y) is large the precision of the ratio can be even worse than the precision of the Y signal itself.The optimum precision was obtained when the masses were similar. Unlike the conclusions of other researcher~,~ the precision obtained with the equipment used in this work does not depend upon the similarity between the first ionization poten- tials of the analyte and the internal standard. Correction for Spectral Interferences Spectral interferences are an important problem and a major limitation of quadrupole ICP-MS. To get accurate analytical results spectral interferences must be identified and eliminated or corrected for. Different methods can be applied for this purpose. For the determination of Mo in BCR CRM 403 (Sea Water) three techniques were applied. Although Mo has several stable isotopes none of these was free of interference when measured in the sea-water matrix as can be seen in Table 3.The isotopes at m/z 92 and 94 suffer from isobaric interferences from Zr. The other isotopes suffer from interference from polyatomic ions. Because the contribution of BrOH + and BrO+ species to the peak at m/z 89 is less than 5% "Mo was selected for the certification.8 First matrix matched blanks were prepared. In the 10-fold diluted sea-water used in this analysis a Br concentration of 6.75 mg 1-1 was found. Blanks with the same Br content were prepared and the signal intensity at m/z 98 was subtracted from the signal intensity at the same mass in the sea-water spectrum. A concentration of 10.21 pg 1-1 (standard deviation 0.42 pg 1-') was found. Next the inter- ference was corrected for by calculation. A mathematical correction based on the intensity of the 95Mo peak and on the isotopic abundances of the Mo isotopes and the experimentally determined 81BrOH+:79Br0+ ratio was used.8 A value of 10.99 pg 1-1 (standard deviation 0.52 pg 1-I) was found.Finally the Br interference was eliminated by removing the Br Table3 Major spectral interferences on the Mo isotopes in a sea-water matrix ~~~ Abundance (%) 14.8 9.1 15.9 16.7 9.5 24.4 9.6 Interferences "Zr + 94Zr+ KKO+ "BrO' ArKO+ "Zr' 79BrOH+ "BrO' NaKCl+ *'BrOH+ 81Br170+ *4sr0 + from the solution with an anion-exchange resin prior to the measurement. The 98Mo peak could then be measured free from interferences. The resulting concentration was 10.83 pg 1-1 (standard deviation 0.52 pg 1-l). Within the exper- imental uncertainties the three results compared well with one another and a mean value of 10.72 pg 1-1 (standard deviation 0.52 pg 1-l) was reported to BCR.This value is in agreement with the finally certified concentration of 10.1 pgl-' with a 95% confidence interval of L2.0 pg 1-I. For most matrices the presence and nature of the polyatomic interfering species is predictable and interferences can be identified based on the known or measured matrix composi- tion. Appropriate methods for interference correction can then be applied. Occasionally however polyatomic species were observed that sometimes could not be identified. For the determination of about 3 pg I-' of Cu in BCR estuarine water only the 65Cu isotope was measured since the 63Cu isotope was strongly interfered with by ArNa'.The Cu peak at m/z 65 is also expected to suffer interference mainly by polyatomic species containing S or P. These interferences were fully eliminated by removing S and P from the solution by anion exchange. Nevertheless results were too high by a factor of 2 and were rejected for certification. To investigate this problem further the samples were analysed with high-resolution ICP-MS. A new technology developed by Finnigan MAT (Bremen Germany) and used in a new high-resolution ICP-MS instrument soon to be available on the market was applied.' The resulting spectrum obtained for 1 pg1-I Cu is shown in Fig. 4 and reveals the presence of an interfering species at m/z 64.982 that until now could not be identified. Evident possibil- ities such as 24Mg,160H+ or Na2180H+ could be ruled out because the corresponding oxides could not be observed in the spectrum.The interfering peak may be due to an organic polyatomic species. Standardization Standardization is of course an important topic in certification work. Calibration procedures must be accurate and traceable which means that in principle one should always start by weighing an appropriate amount of an element standard or a compound with known stoichiometry. Three calibration methods can be used in ICP-MS. In many cases external calibration is sufficient on condition that an appropriate internal standard is chosen. Standard additions was used when matrix induced signal suppression or enhancement was expected or turned out to be insufficiently corrected for by the internal standard.Isotope dilution was applied only exception- 160 r I40 1 ; 120 rn rn 3 100 5 80 2 60 40 20 0 v) 0. cu fi \ ? I 64.869 64.919 64.969 65.019 Masslu Fig. 4. Spectral interference in the determination of Cu in BCR CRM Estuarine Water. High resolution ICP-MS spectrum of a 1 ppb solution at a resolution of 2200 [MIAM (M is the mass of a species AM is the difference in mass between two neighbouring species) 10% valley definition]'190 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 ally. In this laboratory it is common practice to test both external calibration and standard additions in BCR certifi- cation work. If after a first set of experiments both methods yield the same result with a comparable precision the less time consuming external calibration method is then used for further analysis. For the determination of Mg in BCR human serum (CRMs 303 and 304) external calibration did not yield sufficiently precise results.Beryllium had to be used as an internal standard because other elements appropriate for internal standardization were present in the serum samples. Because the relative difference in mass between Be and Mg is large the precision of the results was poor. To check the accuracy of the standardization an NIST SRM 909 Human Serum was also analysed. Values of 28.6 [standard deviation (n= 5)= 1.2 mg 1-'1 and 28.7 mg 1-' [standard deviation (n = 6) = 0.38 mg 1-'1 were found with exter- nal calibration and standard additions respectively. Both results are in agreement with the range of 28.7-30.6 mg I-' specified by NIST.The exceptionally large uncertainty on the external calibration results was also observed for the NIST material and the better precision of standard additions is obvious. Both methods are accurate. For the certification of the BCR material standard additions was used. In the determination of Cd in BCR CRM 143R Sewage Sludge-amended external calibration resulted in much lower values (63.6 pg g-'; standard deviation 0.9 pg g-') than stan- dard additions (71.5 pg g-'; standard deviation 1.2 pg g-'). This was not to be expected since the internal standard In in this case did not reveal any signal suppression or enhance- ment and since for other elements determined in the same material agreement between results from both standardization methods was excellent.In this case standard additions was used and results later turned out to be in excellent agreement with those of other participants and with the finally certified value (71.8 pg g-'; standard deviation 1.2 pg g-'). Isotope dilution is an expensive and time consuming stan- dardization method but it can be useful when sample pre- treatment causes loss of analyte material. It then corrects appropriately- for incomplete recovery but obviously not for contamination. Therefore blanks are still necessary and unless isotope dilution is applied to the blanks also the use of an internal standard is still necessary. In practice isotope dilution has been applied to the determination of Pb in BCR CRM 414 (Plankton) BCR CRM 422 (Cod Muscle) and wine and to the determination of Hg in BCR CRM 422 (Cod Muscle).Results of the latter analysis" are presented in Fig. 5. Parallel to the c 0) 0) 700 1 - (0 2 600 .- a C $ 500 G 8 $! 400 In m u 300 0 .- 4- 2 4- 6 200 0 C 0 0 ICP-MS CVAFS ' . Certified X I € isotope dilution 2 Standard additions T T 3- External calibration 1 _L Isotope dilution 1 t External calibration 2 I 1 Method Fig.5. Determination of Hg in BCR CRM 422 (Cod Muscle). Concentrations and 95% confidence intervals (ng g-'). For Isotope dilution 1 microwave digestion was carried out immediately after spike addition; and for Isotope dilution 2 overnight room temperature digestion was carried out prior to microwave digestion ICP-MS analyses Hg was also determined with cold vapour atomic fluorescence spectrometry. For ICP-MS analyses about 0.1 g of material was digested in nitric acid in a closed vessel in a microwave oven.The recovery of Hg was greater than 96%. When using standard additions results were higher than with external calibration for which repeatability was poor as is shown in Fig. 5 for two series of analyses. The data obtained with standard additions were confirmed by the cold vapour atomic fluorescence values. Isotope dilution yielded different results depending on the details of the digestion procedure. When the Teflon bomb with the sample and the nitric acid was immediately heated in the microwave oven the results obtained were too low. When the mixture was allowed to equilibrate overnight prior to microwave digestion the results obtained compared well with those of standard additions and atomic fluorescence.A possible explanation is that organomer- cury compounds survive the microwave digestion and that the ICP-MS signal per unit mass of Hg is lower for organomercury than for inorganic Hg ions. The overnight equilibration in this case enabled the spiked mercury to be incorporated into the organic species. An overnight equilibration step was also included for the standard additions. Conclusion In general it can be concluded that BCR has principally accepted ICP-MS for certification work and that ICP-MS substantially contributes to BCR certifications in the field of biological and environmental materials. The ICP-MS technique was shown to yield accurate results only if matrix effects and spectral interferences were rigorously studied and dealt with.Accuracy and precision therefore cannot be taken for granted. Whereas in the period from 1982 'to 1990 ICP-MS contrib- uted to less than 1% of the certification work carried out in this laboratory on biological and environmental samples its contribution since then has risen to about 60%. Each certifi- cation required substantial scientific research leading to data that was usually acceptable for certification. The accuracy and precision of the results obtained by ICP-MS are comparable to or better than those of reference methods such as neutron activation analysis. 1 2 3 4 5 6 7 8 9 10 11 12 13 References Jarvis K. E. Gray A. L. and Houk R. S. Handbook of Inductively Coupled Plasma Mass Spectrometry Blackie Glasgow and London and Chapman and Hall New York 1992 p.1. Houk R. S. Shum S. C. K. and Wiederin D. R. Anal. Chim. Acta 199 250 61. Cornelis R. Dyg S. Griepink B. and Dams R. Fresenius' J. Anal. Chem. 1990 338 414. Quevauviller P. Vercoutere K. Bousfield D. and Griepink B. Report EUR 14062 EN ECSC-EEC-EAEC Brussels Luxembourg Goossens J. De Smaele T. Moens L. and Dams R. Fresenius' J. Anal. Chem. 1993 347 119. Goossens J. Vanhaecke F. Moens L. and Dams R. Anal. Chim. Acta 1993 280 137. Thomson J. J. and Houk R. S. Appl. Spectrosc. 1987 41 801. Vanhaecke F. Goossens J. Dams R. and Vdndecasteele C. Talanta 1993 40 975. Schmidt G. Finnigan MAT Bremen Germany personal communication. Campbell M. J. Vermeir G. Dams R. and Quevauviller P. J. Anal. At. Spectrom. 1992 7 61 7. Griepink B. Maier E. A. and Muntau H. Report EUR 12680 EN ECSC-EEC-EAEC Brussels Luxembourg 1990 pp. 1-82. Vandendriessche. S. Griepink B. and Marchandise H. Report EUR 14056 EN ECSC-EEC-EAEC Brussels Luxembourg 1992 De Leenheer A. P. Yeoman W. B. and Colinet E. S. Report EUR 10627 EN ECSC-EEC-EAEC Brussels Luxembourg 1992 1992 PP. 1-64. pp. 1-55. pp. 1-64.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 191 14 Quevauviller P. Kramer K. J. H. Vercoutere K. and and Griepink B. Report EUR 14557 EN ECSC-EEC-EAEC Griepink B. Report EUR 14061 EN ECSC-EEC-EAEC Brussels Luxembourg 1992 pp. 1-53. 15 Quevauviller P. Vercoutere K. Muntau H. and Griepink B. Report EUR 14558 EN ECSC-EEC-EAEC Brussels Luxembourg 1993 pp. 1-71. 16 Quevauviller P. Imbert J. L. Wagstaffe P. J. Kramer G. N. Brussels Luxembourg 1993 pp. 1-64. Paper 3/04699K Received August 4 1993 Accepted December 6 1993
ISSN:0267-9477
DOI:10.1039/JA9940900187
出版商:RSC
年代:1994
数据来源: RSC
|
16. |
Speciation analysis of chromium by inductively coupled plasma mass spectrometry with hydraulic high pressure nebulization |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 193-198
Norbert Jakubowski,
Preview
|
PDF (800KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 193 Speciation Analysis of Chromium by lnductively Coupled Plasma Mass Spectrometry With Hydraulic High Pressure Nebulization* Norbert Jakubowski Brigitte Jepkens Dietmar Stuewer and Harald Berndt lnstitut fur Spektrochemie und angewandte Spektroskopie Postfach 10 13 52 0-440 13 Dortmund Germany The speciation analysis of Cr was studied using hydraulic high pressure nebulization in combination with inductively coupled plasma mass spectrometry. Ion-pair chromatography with tetrabutylammonium acetate as the ion-pairing reagent and 25% methanol in the eluent was applied for the separation of Cr"' and Cr"'. Interferences from carbon could be reduced by the addition of oxygen to the aerosol gas. With careful optimization of the operating conditions a procedure has been established that results in detection limits of down to 1 ng ml-' for the Cr species.Keywords lnductively coupled plasma mass spectrometry; nebulization; speciation; chromium Speciation analysis is becoming of greater interest mainly being focused on selected elements the species of which appear in environmental samples and are known or at least suspected to be toxic. Among these elements Cr is obviously of particular interest as it is widely distributed in the environment because of many industrial applications e.g. in galvanization and the steel industry. It exists in several species of which Cr"' is considered to be essential whereas CrV' is thought to be strongly toxic owing to its high oxidation potential and the ease with which it penetrates biological membranes.This is the reason why speciation analysis of Cr in particular in drinking water has already been investigated extensively by application of different techniques such as precipita- tion adsorption solvent extraction volatilization and chromatography. Not only does speciation analysis require selectivity but also high sensitivity because in real samples the elements in ques- tion must be determined at ultratrace levels. Inductively coupled plasma atomic emission spectrometry (ICP-AES) and mass spectrometry (ICP-MS) together provide these require- ments,' and in particular ICP-MS excels by offering not only extremely low detection limits but also real multi-element capabilities and isotope information the latter enabling the application of isotope dilution techniques which are unsur- passed for accuracy of quantificati~n.~,~ Considerable work has already been devoted to the perform- ance enhancement of ICP-MS for speciation analysis with respect to the detection power and also to the ease of operation.Detection limits could be improved by preconcentrati~n~.~ or by application of new sample introduction techniques such as electrothermal vaporization thermospray (TS),6 direct injec- tion nebulization ( ultrasonic nebulization ( USN),9 hydraulic high pressure nebulization ( HHPN)1° and direct coupling of gas chromatography to ICP-MS.ll Most profitable as regards the ease of operation is an analytical procedure that enables the determination of the species in question to be made in a single step.The analysis time can be reduced considerably by application of on-line techniques which have recently been reviewed by Sperling et aL4 for speciation analysis of Cr. In particular high-performance liquid chromatography (HPLC) on-line separation which combines high chromato- graphic resolution with a short analysis time is favourable. Procedures for speciation analysis by plasma source mass spectrometry have been developed for several important elements.12 However it is somewhat strange that so far there are only a few results available for the speciation analysis of Cr by plasma source mass ~pectrometry.'~ With organic liquids plasma source spectrometry is generally impeded by inter- * Presented in part at the XXVIII Colloquium Spectroscopicum Internationale (CSI) York UK June 29-July 4 1993.ferences owing to the presence of carbon and in particular this is the case for Cr because the main isotope 52Cr+ is obscured by the abundant molecule 40Ar'2C+. Clogging of the sampling orifice is a further problem that arises as a conse- quence of the high carbon content. Addition of oxygen to the aerosol gas could be considered as a means of reducing any problems due to interferences from carbides and clogging. In a previous in~estigation,'~ it has been shown that HHPN as a sample introduction technique in ICP-MS simply com- bined with an effective method of desolvation can increase the sensitivity for the majority of elements in comparison with conventional pneumatic nebulization. In principle it is extremely well suited for direct coupling with HPLC.Thus it was clear that an investigation should be carried out of the capabilities of HHPN for the speciation analysis of Cr because with respect to the above mentioned requirements the tech- nique seems particularly appropriate for this type of appli- cation. For chromatographic separation of the Cr"' and CrV' species ion-pair chromatography with the reagent tetrabu- tylammonium acetate (TBAA) was utilized according to a procedure which was developed by Syty et aE.15 for on-line separation of Cr species and that was later modified by Berndt and co-workers for atomic absorption spectrometric (AAS)16 and ICP-AES dete~tion.'~ Experimental A schematic diagram of the whole experimental arrangement is shown in Fig.1; operating details are compiled in Table 1 t 6 -7 Ar II "V 5a C ' 4 7 I 5b - 7 3 1 / \ Fig. 1 Schematic diagram of experimental arrangement 1 HPLC pump; 2 sample loop; 3 chromatographic column; 4 HHPN nebuliz- ation chamber with impact bead; 5 desolvation system with heating (5a) and two-stage cooling (5b and 5c); 6 ICP torch; and 7 drain194 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Table 1 Compilation of operating conditions HH PN system- Sample uptake rate Sample loop volume Pressure Nozzle diameter Desolvation system- Heating temperature Cooling temperature ICP-MS system- Power Reflected power Aerosol gas flow rate Intermediate gas flow rate Outer gas flow rate Oxygen flow rate Sampling distance Data acquisition system- Dwell time per data point Total acquisition time 1.0 ml min-' 0.1 ml 9.0 MPa 15 pm 120°C First stage 0°C Second stage -20°C 1500 W 25 W 0.75 1 min-' 0.5 1 min-' 15 1 min-' 0.06 1 min - ' 10 mm 327.68 ms ~ 3 0 0 s including the standard parameters derived from optimization experiments.The dual piston HPLC pump (Knauer Berlin Germany) enables flow control and is specifically for elemental analysis equipped with an inert Ti pumping head. The sample is introduced by a metal-free syringe to the system via an inert sample introduction valve (Knauer) connected to a 100 pl loop made from poly ether ether ketone ( 5 cm x 4.6 mm i.d.) (PEEK). The chromatographic column is filled with particles with a diameter of 5 pm (Eurospher 100 CI8 Knauer). All these components as well as the subsequent HHPN system with the desolvation unit and an inlet for addition of oxygen are operated outside the cabinet of the commercial ICP-MS system (PlasmaQuad PQ2 Turbo Plus Fisons Winsford UK).The design and function of the HHPN unit has been described in more detail e1~ewhere.l~ Specific details for these investigations are a nozzle with a diameter of 15 pm made from Pt-Ir in a nozzle holder made from Ti and fixed in the poly( tetrafluoroethylene) (PTFE) ground plate of the spray chamber an impact bead used as a droplet converter and a spray chamber designed in this laboratory. A Ti filter with a pore diameter of 3 pm protects the nozzle from clogging. It should be mentioned that HHPN is a functional part of the HPLC system; the high pressure of the pump is not only used to overcome the pressure drop of the column but also for nebulization so that a pressure of 9MPa is required with respect to the nozzle diameter and a sample uptake rate of 1 ml min-'.For reasons of transportation only an aerosol gas with an optimized flow rate of 0.75 1 min-' is added tangen- tially to the flow in the PTFE holder. The spray chamber is coupled to a quarz tube (0.8cm i.d. x 30 cm) which is heated by a temperature controlled heating device to about 120 "C. Cooling was performed in two stages. The first uses a glycol-water mixture to achieve a temperature of 0 "C. The second is made up by a water-cooled Peltier element (Type CP1 4-127-45L AMS Thermotech Miinchen Germany) providing a temperature of about - 20 "C. The whole HHPN system including the desolvation equip- ment as described is now commercially available in an inte- grated unit.The solutions draining from the spray chamber and the desolvation system are removed by use of a peristaltic pump. After desolvation oxygen is added to the gas flow which is then fed to the ICP torch uia a flexible tubing with a length of about 1 m. A calibrated flow controller (Type 246 MKS Miinchen Germany) is used to stabilize the flow of oxygen. Stock solutions of 0.1 moll-' TBAA were prepared by dissolution of the solid reagent (Fluka Chemie Neu Ulm Germany) in doubly distilled water. These were used to add the ion-pairing reagent to the sample and acetic acid for adjustment of the pH value to within an operating range of 3.0-3.3. It should be mentioned that 1 x 10-4mo11-1 of ammonium acetate solution (Merck Darmstadt Germany) was always added to the sample which is necessary to improve ion-pairing reactions in samples with a high salt load.Although this is not the case in drinking water samples it was also used here for investigations of different water samples. In the experiments for optimizing performance of the procedure doubly distilled water was used for blank measurements and a solution with a concentration of lOOngml-' of both Cr species was used to obtain the analytical signals. A tap-water sample from this Institute and a mineral-water sample were used in a test analysis. Standards for Cr"' and CrV' as Cr042- were prepared from two stock solutions (Merck) with a total Cr concentration of 1000 mg 1-' each. Diluted standards for calibration or standard additions were prepared daily because of the limited stability particularly of CrV'.18 For optimization of ion intensities in the organic eluent solution a standard solution was prepared containing In at a concentration of 20ngml-I and 25% methanol.A mixture of methanol (Fluka Chemie) doubly distilled water and ammonium acetate (1 x moll-') was used as the eluent the pH value of which was adjusted to within the range 3.0-3.3 by addition of acetic acid (96% Suprapur Merck). Ion-monitoring profiles were recorded at m/z 50 and 53 using the peak-jumping mode with a dwell time of 327.68 ms per data point. The total scan time was limited to about 300 s. Scanning was started just before injection of the sample. The peak area of the analytical signal was used as the intensity value.For correction of the blank value the intensity value of a blank solution was determined in the same way as for the analytical signals. No blank value correction was applied in the optimization experiments. Results and Discussion Optimization In the case of organic solutions independent optimization of instrumental parameters is necessary because standard working conditions differ considerably from those obtained for non- organic solutions. In comparison with the direct determination of Cr from standards diluted in doubly distilled water the detection limits are higher in the case of an organic solution mainly owing to interferences. On injecting the sample via a sample loop Cr"' is not retained by the column and is measured in the more or less aqueous environment of the sample whereas the retained species Cr"" is measured in the organic environment of the eluent mixture. In order to obtain the best detection limits for the more hazardous species optimization was always carried out with the organic eluent containing In as an internal standard.Because optimization of transient signals is difficult a 5 ml loop was used instead of the 100 pl loop so that injection of the organic standard resulted in a signal that remained constant for more than 5 min. Considering the analytical signals of the two Cr species it should be mentioned that it was always possible to achieve identical sensitivity for both species by careful tuning but with respect to the effort required this was not always considered a primary aim in the general optimization experiments.In contrast to previous e~perience,'~ desolvation required a two-stage cooling procedure in the present investigation to remove the methanol-water eluent effectively. In the first stage more than 95% of the water was removed by condensation while the methanol was mainly removed in the second stage. This two-step operation was indispensible because otherwise the aerosol tubing with a diameter of 8 mm would be clogged by ice formation after about 15 min of operation. For operation of the ICP when nebulizing organic solutions the power had to be increased from the usual value of 1.35 toJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRYy MARCH 1994 VOL. 9 195 1.5 kW. Even with the addition of oxygen the reflected power could not be tuned to less than 25 W.This increases with increasing methanol concentration in the eluent and also with increasing oxygen addition. A very high methanol content also leads to unsatisfactory stability. Improvements in the matching unit with respect to applications with organic solutions should provide a challenge for a manufacturer. Problems due to Carbon Several drawbacks to the operation of ICP-MS instrumen- tation can arise from the presence of an organic solvent in the aerosol spectral and non-spectral interferences clogging of the sampling orifice and the skimmer and a loss of sensitivities for certain elements. Utilization of HHPN instead of pneumatic nebulization can fully balance this loss for most elements by the gain in intensity achieved with this technique but only with effective desolvation.The desolvation system with two- stage cooling was therefore an essential pre-requisite of this work. The efficiency of operation is high enough to reduce the organic load to the ICP drastically about 94% can be con- densed from the aerosol before introduction to the ICP torch.20 In order to avoid the sampler and skimmer from becoming clogged addition of oxygen is known to be helpful. Normally this is associated with the disadvantage of the appearance of greater amounts of oxide ions in particular those of Ar. However in the case of Cr this is not a severe limitation. Although expected no increase in erosion of the skimmer or sampler material was observed. Furthermore the addition of oxygen has proved to be an effective means of reducing interferences.In general addition of oxygen leads to reduction of interferences induced by carbon by up to an order of magnitude with the only exception being m/z 56 for which the signal is increased owing to ArO'. Concerning the Cr isotopes almost no reduction of the interfer- ing molecules appears for the main isotope at m/z 52 owing to the high signal of Arc' whereas a significant reduction is observed for the minor isotopes at m/z 50 and 53 which can now be considered for quantitative evaluation. For the two isotopes "Cr and 53Cr considered for quantitat- ive evaluation the influence of oxygen addition has been studied in more detail. The ion-monitoring profiles of the two isotopes for a blank measurement with varying amounts of oxygen added are shown in Fig.2(a). The blank value for 53Crf is about one order of magnitude higher than for "Cr'. From this result higher oxygen addition seems generally to be preferable but this does not hold true when the analytical signals are taken into consideration. This is demonstrated in Fig. 2(b) which shows measurements as before but now with both species of Cr at a concentration of 100 ng ml-' each. For both species a pronounced maximum is observed at a flow rate of about 60 ml min-'. According to its higher natural abundance 53Cr' exhibits the higher intensity but despite this analytical interest was mainly focused on "Cr + because of its lower blank value. Chromatographic Working Conditions For improvement in the optimum chromatographic operating conditions not only was the addition of the ion-pairing agent to the eluent considered as Syty et a/.' did but alternatively addition of TBAA to the sample as has been done previously by Posta et a1.,16 was also investigated.The sample was injected by switching the sample loop into the high-pressure flow. No additional mixing loops were necessary. The results of these optimization experiments are shown in Fig. 3(a) and (b). In the measurements represented in Fig. 3(a) the concentration of TBAA in the eluent was kept fixed at a comparatively high value while the concentration in the sample was varied. Without TBAA in the sample the highest signal intensity was obtained for both species and they are clearly separated. 0.80 4- .- C 3 >.0.60 c g m 0 0.40 2 Y m 2 0.20 1 I I I I I 0 20 40 60 80 100 Oxygen flow rate/ml min-' Fig.2 Measurements for optimization of the oxygen flow rate. (a) Blank value at A 53 and B 50 m/z. (b) Analyte signals from Cr"' and CrV1 100 ng ml-' each A 53Cr"1; B 53CrV1; C s°Cr"'; D 50CrV1 .- -2 .g 8 - C al C Y - 6 4 2 0 100 200 Time/s 300 Fig. 3 Measurements for optimization of TBAA content. (a) A in the absence of TBAA; B 5 x lop3; and C 5 x mol 1-1 TBAA in the samples and a constant concentration of 3 x lop3 moll-' in the eluent. (b) A in the absence of TBAA; B 5 x lop3; and C 5 x moll-' TBAA in the eluent and a constant concentration of 3 x lov5 moll-' in the sample196 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Increasing the concentration of TBAA in the sample results in a loss of sensitivity with almost no change in the retention time For the measurements represented in Fig. 3(b) the TBAA concentration in the eluent was kept fixed at the low value of 3 x moll-' while the concentration of TBAA in the sample was varied.A minimum concentration of about 5 x mol 1-1 is obviously necessary to achieve satisfactory separation with highest sensitivity for both species; this was therefore included in the final procedure. The preceding investigation was performed using a mixture of methanol and water as the eluent which was known to result in good chromatographic resolution. However variation of the concentration of methanol in the eluent was of course also taken into account for optimization of the operating conditions.From the results represented in Fig. 4 it can be seen that the methanol content influences both the retention time and the signal intensity. A concentration of 25% methanol was chosen as an acceptable value in order to achieve high sensitivity for both species and well separated signals. It is one of the peculiarities of HHPN that the sample uptake rate can be varied over a broader range than is usually accessible with pneumatic nebulization. For the HHPN nozzle chosen a range of between 0.5 and 3 ml min-' can be covered. The results of the corresponding measurements are shown in Fig. 5. The maximum intensity of both species remains constant for sample uptake rates of from 0.8 up to 1.6mlmin-'. A value of 1 ml min-' was chosen as being suitable. It should be mentioned that complete elution of CrV1 after injection will take more than 2 min.This is much longer than expected from the sample uptake rate. The reason is a pro- nounced dispersion effect caused by the high dead volume of the nebulizer and the desolvation system. Nevertheless chrom- atographic working conditions could be chosen here so as to resolve both species. In general however it might be possible 9 1 1 I I 0 100 200 300 400 Time/s Fig. 4 Optimization of methanol concentration in the eluent A 10; B 20; C 30; D 40; and E 50% Time/s Fig.5 C 1.2; and D 1.6 ml min-' Optimization of the eluent liquid flow rate A 0.8; B 1.0 to reduce the analysis time significantly by application of a dispersionless sample introduction system. Analytical Measurements The first measurement using the chosen parameters was a reproducibility test with seven repeat injections of a mixture of both Cr species at a concentration of 100 ng ml-' each.The reproducibility of the retention time is very good. Fluctuations in the signals as can be seen from Figs. 3 and 4 are mainly caused by the greater instability of the plasma owing to the high carbon and oxygen load and can only partially be ascribed to pulsations of the pump. Nevertheless the relative standard deviation (RSD) of the integrated signals amounts to 2.9% for both Cr species which is nearly the same as is obtained for the analysis of aqueous solution. Finally a calibration procedure was performed using four solutions with concentrations of up to 100 ng ml-' of both Cr species.The recorded single ion monitoring profiles for 50Cr + are shown in Fig. 6. With blank value correction the cali- bration procedure provides satisfactory linearity with detection limits (3s) of 0.6 ng ml-l for Cr"' and 1.8 ng ml-' for CrV1. The CrV1 shows a slightly lower sensitivity which is attributed to a certain depression of the signal by the organic environ- ment. The calibration curve is linear over more than two orders of magnitude down to the ng ml- ' region. On evaluating 53Cr the detection limits are worse by a factor of about four. Recovery and stability of the Cr"' were checked by injection of this species alone. A signal was seen to appear without being retained for the Cr"' but this did not exceed 6% of the peak area of the final CrV' signal.This signal must be attributed to a partial reduction of CrV' to Cr"' for the chosen working conditions of pH 3. Therefore all investigations were performed with CrV1 solutions which were prepared immediately before use. As a preliminary application the whole procedure was applied to measurements of the local tap water and of a mineral water with quantification by standard additions for both species. For the mineral water the single ion monitoring profiles of 'OCr+ that were recorded are presented in Fig. 7. Again doubly distilled water was used for blank measurement. A depression of the CrV' signal by 50-70% is observed for both samples which is stronger for the mineral water which has the higher salt content. However the determination of Cr"' is not affected.Actual investigations indicate that compen- sation for the signal depression could be possible due to the presence of a higher amount of acetic acid and TBAA in the sample. Nevertheless detection limits for the determination of CrV1 in real samples are in the low ng ml-' region as can be 7.50 D C . . . . . . . . . . . . . . .. . L . .- . . . . - . . .~ . . . . . . . . 0 100 200 300 Tirne/s Fig. 6 Calibration with varying concentrations of both Cr"' and Cr" ion-monitoring profiles at 50 m/z A blank; B 10; C 50; and D 100 ng ml-'JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 197 1.50 I I 0 100 200 300 Time/s Fig.7 Test analysis of a mineral water sample showing recorded single ion monitoring profiles of 'OCr+ with standard additions of both Cr"' and Cr" A blank; B sample; C sample + 5 ng m1-I; and D sample + 10 ng ml- ' estimated from these measurements. No signal for Cr"' is observed for the tap-water sample whereas for the mineral water a Cr"' signal corresponding to almost 4 ng ml-' appears.In both samples a third species occurs in the chromatogram after the CrV' signal the identity of which cannot be confirmed so far. It should be mentioned that owing to the higher detection limit this species was not observed when applying ICP-AES instead of ICP-MS. In conclusion the results demon- strate promising possibilities for the on-line determination of Cr"' and CrV1 in drinking water. Comparative Assessment From the work published so far no procedure is known for the speciation analysis of Cr by ICP-MS which enables determination of Cr"' and CrV' in only one step.For the determination of Cr"' alone Roehl and Alforq~e'~ have reported detection limits of about 1 ng ml-I for an application with ICP-MS but nearly identical detection limits were obtained by colorimetry. Procedures that were primarily developed for TCP-AES can often be adapted to ICP-MS with only minor modification so that ICP-AES procedures for the determination of CrV' will also be considered here for comparative assessment of the results obtained. With any procedure developed for the deter- mination of Cr"' Cr"' can always be determined by measure- ment of the total Cr content and subtraction of the CrV' value. The relevant data available from the literature are compiled in Table 2. In comparison with AES the detection limits obtained in the present work offer a certain improvement by a factor of about four with the additional advantage of reduced analysis time but for evaluation of these results one should be aware that the determination of Cr is in some way a worst case scenario because the interferences introduced by carbon have the strongest disturbing influence for this particular element.From this point of view the work of Cox et (see Table2) could be a promising alternative also for ICP-MS. However for future work the advantage of the described procedure could be for pre-concentration of Cr" as has been shown by Posta et a1.,16 to be additionally applied. Preliminary results show that detection limits for CrV' can be improved by more than one order of magnitude by an increase in the volume of the sample loop.Further improvement of the analysis time by automation using a commercial HPLC auto- sampler is also possible. Conclusion Hydraulic high pressure nebulization as a functional part of an HPLC system is a powerful means for providing an interface-free coupling of HPLC to ICP-MS with the additional advantage of improved sensitivity. This has been demonstrated in the present work by development of a pro- cedure for the speciation analysis of Cr by application of ion- pair chromatography in the determination of Cr"' and CrV' by ICP-MS in one step. Addition of oxygen to the aerosol gas and effective desolvation were necessary prerequisites in order to apply ICP-MS as a highly selective and sensitive detection technique.Ion-pair chromatography has been applied to separ- ation of the species so that the procedure described here can be used for speciation analysis of other elements with only minor modification. Detection limits down to 1 ng ml-' were achieved with the main limitation being interferences. An even better analytical advantage can be attained from this promising technique by addition of a pre-concentration step which could be useful for improving the detection limits significantly and to extend the applicability to species of other elements with environmental significance. For a full evaluation of the capabili- ties of the technique the work must also be verified by high mass resolution which is the subject of future studies. This work was supported financially by the Bundesministerium fur Forschung und Technologie and by the Ministerium fur Wissenschaft und Forschung des Landes Nordrhein-Westfalen. Table 2 Detection limits (DL) for different ICP applications of ICP for speciation analysis of Cr Technique* HPLC-DIN-ICP-AES DIN-ICP-AES HPLC-USN-ICP-AES HHPN-ICP-AES TS-HPLC-ICP- AES FI-ICP- AES IC-ICP-MS HPLC-HHPN-ICP-MS Crvl Cr"' DL/ng ml-' 20 20 12 12 8 4 4 2 2 1.4 0.2 1 0.6 1.8 Time/min Reference 8 LaFreniere et al.ref. 7 8 LaFreniere et al. ref. 7 4 1-2 Wang and Jiang ref. 9 Berndt and Luo ref. 17 3 Roychowdhury and Koropchak ref. 6 1 Cox et al. ref. 21 8 Roehl and Alforque ref. 13 1-2 This work CrV1 *IC = ion chromatography; FI =flow injection. t Not specified.198 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 1 2 3 4 5 6 7 8 9 10 11 12 References Broekaert J. A. C. in Metal Speciation in the Environment ed. Broekaert J. A. C. Gucer S. and Adams F. Nato AS1 Series G Ecological Sciences Springer Verlag Berlin 1990 vol. 23 p. 213. Heumann K. G. in Metal Speciation in the Environment ed. Broekaert J. A. C. Gucer S. and Adams F. Nato AS1 Series G Ecological Sciences Springer Verlag Berlin 1990 vol. 23 p. 153. Okamoto K. Spectrochim. Acta Part B 1991 46 1615. Sperling M. Xu S. and Welz B. Anal. Chem. 1992 64 3101. Syty A. Christensen R. G. and Rains T. C. At. Spectrosc. 1986 7 89. Roychowdhury S. B. and Koropchak J. A. Anal. Chem. 1990 62 484. LaFreniere K. E. Fassel V. A. and Eckels D. E. Anal. Chem. 1987 59 879. Shum S. C. K. Neddersen R. and Houk R. S. Analyst 1992 117 577. Wang S.-R. and Jiang S.-J. J. Chin. Chem. SOC. (Taipei) 1991 38 327. Berndt H. Fresenius’ Z . Anal. Chem. 1988 331 321. Houk R. S. and Jiang S . J. J. Chromatogr. Libr. 1991 47 101. Vela N. P. Olson L. K. and Caruso J. A. Anal. Chern. 1993 65 585A. 13 14 15 16 17 18 19 20 21 Roehl R. and Alforque M. M. At. Spectrosc. 1990 11 210. Jakubowski N. Feldmann I. Stuewer D. and Berndt H. Spectrochim. Acta Part B 1992 47 119. Syty A. Christensen R. G. and Rains T. C. J. Anal. At. Spectrom. 1988 3 193. Posta J. Berndt H. Luo S . K. and Schaldach G. Anal. Chem. 1993 65 2590. Berndt H. and Luo S . K. J. Anal. At. Spectrom. submitted for publication. Griepink B. in Metal Speciation in the Environment ed. Broekaert J. A. C. Giicer S. and Adams F. Nato AS1 Series G Ecological Sciences Springer Verlag Berlin 1990 vol. 23 p. 361. Jakubowski N. Feldmann I. and Stuewer D. Spectrochim. Acta Part B 1992 47 107. Luo S. K. and Berndt H. Spectrochim. Acta Part B submitted for publication. Cox A. G. Cook I. G. and McLeod C . W. Analyst 1985 110 332. Paper 3/04551 J Received July 27 1993 Accepted October 5 1993
ISSN:0267-9477
DOI:10.1039/JA9940900193
出版商:RSC
年代:1994
数据来源: RSC
|
17. |
Matrix separation by chelation to prepare biological materials for isotopic zinc analysis by inductively coupled plasma mass spectrometry |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 199-204
Steven F. Durrant,
Preview
|
PDF (789KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 199 Matrix Separation by Chelation to Prepare Biological Materials for Isotopic Zinc Analysis by Inductively Coupled Plasma Mass Spectrometry* Steven F. Durrant Antoaneta Krushevska Dula Amarasiriwardenat Mark D. Argentine Sabine Romon-Guesnier and Ramon M. Barnes University of Massachusetts Department of Chemistry Lederle Graduate Research Center Towers Amherst MA 0 1003-0035 USA Following an evaluation of three chelating resins [Chelex-1 00 poly(dithi0carbamate) (PDTC) and carboxyme- thylated poly(ethy1eneimine)-poly(methylenepolypheny1ene) isocyanate (CPPI)] a procedure was established with the last of these for the separation of Zn from biological matrix elements prior to 70Zn:68Zn isotopic analysis by inductively coupled plasma mass spectrometry (ICP-MS).The method was verified by establishing Zn recoveries and by determining its effectiveness in removing CI and Na from buffered test solutions. Calcium Na and Zn concentration data were determined by inductively coupled plasma atomic emission spectrometry. Chlorine was measured by electrothermal vaporization ICP-MS. The efficacy of the technique was demonstrated by the determination of zinc isotope ratios in bovine milk and human urine. Results compared favourably with those obtained using a previously established extraction procedure. Keywords Zinc isotope ratios; inductively coupled plasma mass spectrometry; matrix separation; bovine milk; human urine Inductively coupled plasma mass spectrometry (ICP-MS) is a widely accepted reliable method for elemental and isotopic determinations in diverse The multi-elemental capability of the technique is particularly useful for environ- mental samples and ICP-MS accounted for 1% of the total number of the analyses reported between 1986 and 1989.4 Both ICP-MS and isotopic'-'' analyses of biological materials provide unique information.Biological applications have recently been reviewed.''*12 The advantages of ICP-MS have been do~umented;'-~ how- ever a number of difficulties are encountered in the analysis of practical samples. Spectral interferences from oxides doubly- charged or polyatomic species of elements present in the sample and reagents13-" and non-spectroscopic interferences from solutions containing relatively high (about 0.2% m/m) dissolved solids typically observed as a suppression of analyte responses,'&'* hamper convenient analyses.A number of strategies have been developed to minimize spectral interferences. These include mathematical correction," addition of a molecular gas such as nitrogen to the argon plasma to modify the plasma chemistry,20 and separation of the analytes from interfering elements prior to analysis.21 The last method is also effective in reducing or eliminating non- spectroscopic interferences. Separation techniques for spectro- chemical analysis including detection by ICP-MS have been reviewed by Horvath et aLZ2 Resins have long been used to separate analytes from matrix elements prior to elemental analysis. For example Kingston et used a Chelex-100 resin to separate Cd Co Cu Fe Mn Ni Pb and Zn from alkali and alkaline earth elements in sea-water prior to their determination by electrothermal atomic absorption spectrometry. Resin separation procedures have been automated.For example Wang and Barnes24 used two chelating resins to preconcentrate Cu and Zn for the analysis of water samples by flow injection inductively coupled plasma atomic emission spectrometry (ICP-AES). A previous study described an acid digestion of biological samples with microwave heating." Urine red blood cells and plasma were prepared by extraction with ammonium * Presented at the XXVIII Colloquium Spectroscopicum t Present address Hampshire College School of Natural Science Internationale (CSI) York UK June 29-July 4 1993. Amherst MA 01002 USA.pyrrolidin-1-yldithioformate (APDC) and carbon tetrachloride (CCl,) for the measurement of 70Zn 68Zn isotope ratios by ICP-MS. In the present report an equally effective but quicker and less labour intensive method was sought. The motivation for studying 70Zn 68Zn ratios lies in their use in zinc bioavail- ability studies of children and pregnant ~ o m e n . ~ ' * ~ ~ Three chelating resins were tested for their ability to separate biological matrix components for subsequent Zn isotope ratio measurements using ICP-MS. Owing to their contribution to spectroscopic and non-spectroscopic interferences C1 and Na were the principal target elements; the former as 35Cl is coincident with 70Zn+ while the latter if present at a suffic- iently high concentration causes signal suppression.The three chelating resins investigated were Chelex-100 poly(dithi0carbamate) (PDTC) and carboxymethylated poly (ethyleneimine) - poly (methylenepolypheny1ene)isocyanate (CPPI). Preliminary studies involved the recovery of Zn from buffered solutions in the pH range 4.8-10.2. Based on these studies and further experiments to measure the effectiveness of the resins in removing Na C1 and Ca from test solutions the CPPI resin was selected for testing with real samples of bovine milk and human urine. In urine Cl and Na are present typically at > 3000 pg ml - and > 1500 pg ml- ' respectively;26 thus Zn in urine analysis is a challenging test sample. Owing to the lower Cl and Na concentrations expected in bovine milk (1000 and 500 pg ml-' re~pectively),~~ milk analysis is an example of a relatively straight-forward measurement.Total Zn recoveries determined by ICP-AES and a pre- established APDC-CC14 extraction are compared with those obtained using the new separation technique employing the CPPI resin. Zinc (70Zn 68Zn) isotope ratio measurements obtained after sample preparation by the two methods were also obtained. Experimental Instrumentation ICP-AES The Ca Na and Zn concentrations of the test solutions were determined by ICP-AES using a sequential spectrometer with two monochromators and internal reference channels (Perkin- Elmer Plasma 11 Norwalk CT USA). Operating conditions are given in Table 1. The nebulizer flow rate and viewing200 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 Table 1 Instrumental operating parameters for ICP-AES Parameter Generator frequency/MHz Forward r.f. plasma power/kW Reflected power/W Argon gas flow rates/] min-' Outer Intermediate Nebulizer Observation height/mm Read delay/s Integration time/ms Replicates Analytical wavelengths/nm 27.12 1.0 <5 Value 15.0 1 .o 0.7 11 20 100 3 Ca I1 393.366 Co I1 228.616 Cu I 324.754 Mn IT 257.610 Na I 589.592 Sc I1 424.683 Y I1 371.030 Zn 1213.856 height were optimized for Zn I emission signal-to-background ratio. Other parameters are default conditions recommended by the manufacturer. Final solutions after extraction or resin separation were 2mol 1-' in nitric acid and their analysis was possible with acid-matched standards. However the acid concentration of the initially digested milk or urine varies with the nature and amount of the reagents used and their evaporation rates during digestion. Thus matching of the sample and standard acid concentrations is impractical. However differences in viscosity between unmatched solutions can lead to deterioration in analytical precision and accuracy.Therefore Myers-Tracy signal compensation was used for the determinations in the digested material with 50 pg ml-' of Sc added as a reference element. 28 ICP-MS Zinc isotope ratios were determined by ICP-MS (SCIEX Elan Model 250 Thornhill Ontario Canada). The mass spec- trometer settings and plasma conditions were optimized with a 200 ng ml-' solution of Zn prior to isotope ratio measure- ment. The instrument operating conditions and data collection parameters for the isotope ratio measurements are listed in Table 2.As only small sample volumes were available after the isotope ratio and elemental measurements chlorine was deter- Table 2 Instrumental operating parameters for zinc isotope ratio ("Zn 68Zn) determinations by ICP-MS Parameter Generator frequency/MHz Forward r.f. plasma power/kW Reflected power/W Distance/mm Load coil-sampler orifice Torch injector tip-sampler orifice Torch (Sciex 'long')/mm Spray chamber Nebulizer Argon plasma gas flow rate/ min-' Outer Intermediate Nebulizer Resolution Measuring mode Measurements per peak Measurement time/s Dwell time/ms Cycle time/s Replicates per integration Solution uptake rate/ml min-' 27.12 1.2 <5 Value 27 6 123 Scott double-pass Perkin Elmer cross-flow 11 1.4 1 u at 10% Peak hop 1 1 .ooo 20 0.85 6 0.6 1.05- 1.15 mined by electrothermal vaporization (ETV)-ICP-MS.For these determinations the central channel flow rate was optim- ized by maximizing the response at m/z 35 from a 150 pg ml-' C1 solution. The analysis conditions are given in Table 3. Reagents Sub-boiled nitric acid and isothermal ammonia solution were prepared. Ammonium acetate buffers (0.1 mol I-') at pH values in the range 4.8-10.2 were prepared from glacial acetic acid and ammonium acetate (Fisher Scientific Fairlawn NJ USA). The buffer was cleaned by passing it through a CPPI column. Distilled de-ionized water (DDW; NanoPure Sybron/Barnstead Boston MA USA) was used throughout to make-up solutions. For the ICP-AES determinations standard solutions of Zn Ca and Na were prepared by dissolving Zn metal (Johnson Matthey Ward Hill MA USA) CaCO and NaCl respect- ively in sub-boiled HNO and diluting to the appropriate volume with DDW.For the determination of C1 by ETV- ICP-MS standard solutions were prepared in the range 3-1 500 pg ml - ' from NaCl (Fisher Scientific Fairlawn NJ USA) dissolved in DDW. Cobalt at 0.2 pg ml-' was used as an internal reference. Columns Columns of Chelex-100 PDTC and CPPI were prepared. The Chelex- 100 was a commercial material (Biorad Laboratories Richmond CA USA). Preparation of the PDTC and CPPI resins has been described by Barnes29 and Horvath and Barnes,,' respectively. Resins were soaked in 2mol 1-1 nitric acid (made from sub-boiled nitric acid) for a few hours and then rinsed with DDW prior to use.Each resin (200-400mg) was packed in plastic funnels (capacity about 5 ml) fitted with glass wool retaining plugs. Typically natural flow rates were about 0.25 ml min-'. Table 3 Instrumental operating parameters for the determination of chlorine by ETV-ICP-MS Parameter Argon plasma gas flow rate/] min-' Outer Intermediate Central channel Forward r.f. plasma power/kW Reflected powerp Measurement parameters Resolution Dwell time/ms Replicates Replicates/reading Measuring mode m/z Sample volume/jd Electrothermal vaporization furnace Start/"C FinalPC 20 700 700 700 700 1100 1100 1100 1100 1200 1200 1200 1200 2400 2400 2400 2400 20 Ramp Sequence- Value 11+0.1 o2 1.4 1.25 1.2 t 5 High 20 300 (1 point/peak) 1 Peak hop 35 10 Perkin-Elmer HGA-400 Time/s 10 30 5 10 5 10 2 6 5JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 201 Digestion Procedures Urine A bulk urine sample (24 h collection) was digested in a microwave oven (CEM MDS-8 1 CEM Corporation Matthews NC USA) according to a procedure given pre- viously.l0 The resulting final volume after disgestion was half of the original representing approximately a doubling of the elemental concentrations compared with the original sample. Scandium was added to give 50 p ml-' in the final solution. Milk A 10ml volume of commercial pasteurized bovine milk (Vitamin D milk grade A homogenized Maple Hill Farms Bloomfield CT USA) was prepared in an ashing furnace (Fisher Isotemp Model 497).1° A few millilitres of sub-boiled nitric acid together with 10 ml of 250 pg ml-' of Sc solution were added to the residue and the final solution made up to 50 ml with DDW.Procedure Resins The ability of the three resins to extract metals from buffered solutions (0.1 mol 1-l ammonium acetate) was studied. The columns were prepared by washing with 2mol 1-l nitric acid followed by 5 ml of buffer. A 5 ml volume of the buffered test solution was then added. The standard buffered solutions contained Zn Cu Co Mn and Y at 0.5vgml-'. The pH values studied were 4.8 6.7 8.9 9.4 and 10.2. Elution was with 5 ml of 2 mol 1-l nitric acid. The separation of Na and C1 from Zn was also investigated. Ammonium acetate buffer (5 ml) at pH 5.5 containing 5000 pg ml-' of Na 200 pg ml-' of Ca and 0.5 pg ml-' of Zn was passed through the columns and eluted.Additional wash- ing with 5 ml of buffer preceded the elution step. CPPI resin with milk and urine Extractions of the urine and milk digests were performed in a class-100 clean room according to the already established procedure." For urine 15 ml aliquots of the digested material were extracted. The final volume was 5 ml (2 mol I-' HN03). For milk 1 ml aliquots were extracted and the final volume was 5 ml (2 mol 1-l HNO,). Aliquots of the digested urine (15 ml) were evaporated to dryness in pre-leached quartz beakers. The residue was taken up in sub-boiled nitric acid neutralized with isothermal ammonia solution and buffered with ammonium acetate buffer (final buffer concentration about 0.1 mol l-' pH 5.5). The resulting solution was passed through a CPPI column and eluted with 5 ml of 2 mol 1-' nitric acid into a vial (12 ml number 6133 Spex Industries Metuchen NJ USA) prior to analysis by ICP-MS.A 3-fold preconcentration resulted com- pared with the digested material. An identical procedure was employed for milk except that the initial aliquots were 1 ml. Only 1 ml of digested milk was required since it contains relatively high Zn concentrations (typically 1 pg ml - I). Results and Discussion ICP-AES Optimization Optimization of the Zn determination by ICP-AES has been described re~ently.~' The Zn 1213.856 nm line exhibited greater signal-to-background ratios than the Zn I1 202.548 nm emis- sion. Consequently the optimal conditions were established with the Zn I line based on the 3a detection limit (DL) and the background equivalent concentration (BEC).The DL and BEC for Zn at various nebulizer flow rates and measuring heights are given in Table4. The lowest DL and BEC values occur at a nebulizer flow rate of 0.7 1 min-'. This flow also corresponds to the best correlation between the intensities of the Zn 1213.856 nm and the Sc I1 424.683 nm line as a function of viewing height. The Sc I1 line was used for signal compensation. No significant difference exists in the BEC for viewing heights between 10 and 15 mm above the induction coil but the DL decreases with decreasing viewing height to a minimum at 7 mm. Signal-to-noise ratios are high at viewing heights above 15 mm or below 10 mm. The optimal viewing height was 11 mm at which the best peak precision was also found.At a nebulizer flow rate of 0.7 1 min-' and different viewing heights signal precision was tested with solutions containing 20 pg ml-' of Zn 50 pg ml-I of Sc and 2 10 and 20% v/v of nitric acid. The relative signals for Zn and Sc were estimated against solutions in 2% nitric acid with and without Myers-Tracy compensation As a check the Sc I1 424.683 nm emission was measured as an analytical line normalized to itself as measured by the second monochromator. The results are illustrated in Table 5. Investigation of the correlation between the Sc I1 424.683 nm line with atomic and ionic lines at different nebulizer flow rates and viewing heights has been presented recently.32 The best correlation of Zn I 213.586 nm with the Sc reference line over a broad range of viewing heights occurred at a nebulizer flow rate of 0.71 min-'.When the nitric acid concentration is in the range 1-20% v/v 10% v/v nitric acid can be used to prepare the standard solutions. The Myers-Tracy compensation also improves the precision by a Table 4 Detection limit (30 measured in pg m1-l) and background equivalent concentration (BEC) for Zn I 213.856 nm (n= 10) Nebulizer flow rate) min-' 0.7 1 .o 1.3 Viewing height/mm DL BEC DL BEC DL BEC 20 18 15 14 13 12 11 10 9 8 7 6 0.019 0.01 8 0.0099 0.0094 0.0078 0.0077 0.0070 0.0065 0.0065 0.0061 0.0056 0.0066 0.10 0.027 0.14 0.093 0.44 0.10 0.073 0.01 5 0.088 0.060 0.33 0.077 0.074 0.076 0.073 0.080 0.0 10 0.073 0.044 0.27 0.083 0.087 0.084 0.1 1 - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -202 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL.9 130 110 - E 90- 2 > 2 7 0 - a 50 30 Table 5 nebulizer flow rate of 0.7 1 min-' Percentage relative signal for Zn and Sc with (MT) and without Myers-Tracy (NMT) signal compensation (n= 10 RSD = 2%) at a Acid concentration/% HN03 - - I I I I I ; 6 7 8 9 10 10 20 Zn s c Zn s c Viewing height/mm MT NMT MT NMT MT NMT MT NMT 20 96 88 99 89 93 81 99 83 18 97 87 99 89 94 82 99 83 15 98 82 99 83 96 81 100 83 14 98 93 99 94 99 93 99 93 93 100 93 96 94 100 96 13 97 12 97 90 100 96 95 91 99 94 11 98 86 99 87 96 82 99 84 10 98 91 99 91 96 91 99 94 9 98 84 99 85 95 73 99 76 8 98 78 99 74 95 73 99 85 7 98 80 99 79 96 77 99 82 factor of 2-5.32 Frequent recalibration is unnecessary because the responses exhibit good long-term stability. Fig.1 Recovery of elements following exchange on Chelex-100 A Mn; B Zn; C Cu; D Y; and E Co. Original solutions buffered in 0.1 mol 1-' ammonium acetate buffer. All elements originally at 0.5 pg ml-' 120 1 t 1. 100 - A 80 - 2 ? = c 60- 40 - T A 5 6 7 8 9 10 PH Fig.2 Recovery of elements following exchange on CPPI A Mn; B Zn; C Cu; D Y; and E Co. Original solutions buffered in 0.1 mol 1-' ammonium acetate buffer. All elements originally at 0.5 pg ml-' Matrix Removal Using a test solution containing Ca Na and Zn in 0.1 mol 1-l ammonium acetate buffer (pH about 5.5) the effectiveness of Chelex- 100 and CPPI for matrix separation was examined. Calcium is present at relatively high concentrations in bovine milk (about 1200pgml-1)33 and may compete with Zn for binding sites. Calcium and Na also cause analytical signal suppression with ICP-MS.The concentrations of Na and Ca in the eluent from the resin columns are shown in Table 6. The CPPI resin is very effective at removing Na. Although some Ca is captured and eluted from the resin no indication exists that this interferes with the exchange of Zn under these conditions. Recoveries of the latter are always about 90%. Although present at several thousand pg ml-' in the starting solution C1 concentrations in the eluents were negligible. In spite of the relatively poor ionization of C1 in the ICP (=0.9%0) a detection limit of <lOpgrnl-' was obtained based on the response using a 10 p1 sample. Other chlorine species (i.e.37Cl 51C10 and 75ArC1) were measured but 35Cl ion was the most sensitive. This result is expected because the 35Cl isotope is the most abundant (75.8%) and suffers no significant interference. Chlorine was evolved above 1200 "C; therefore data acqui- sition began as soon as this temperature was reached and lasted 30 s. The response peak typically observed at about 12JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 203 Table 6 Concentrations (pg m1-I) in the eluent following separation; original solution 5000 pg ml-' of Na and 200 pg ml-' of Ca in 0.1 mol 1-' ammonium acetate buffer at pH 5.5 Eluent Na Ca c1 ~ ~~~ Chelex-100- Column 1 62 182 15 Column 2 332 161 19 Column 3 40 176 <lo* Column 1 1.9 93 < 10 Column 2 0.4 112 < 10 Column 3 3.3 157 < 10 CPPI- *DL= 30 of blank measured in pg m1-l.and 23 s was integrated for calibration purposes. The cali- bration exhibited good linearity (r2 = 0.988) in the range 10-1500 pg ml-'. Based upon the successful removal of Na and Cl from synthetic solutions the effectiveness of the CPPI separation with milk and urine samples was examined. Chelex- 100 appears less capable and suffers from considerable volume changes depending on the ambient pH. The 70Zn 68Zn ratio was determined in digested milk from which the matrix had been extracted by the established extrac- tion procedure and digested milk from which the matrix had been extracted by the chelating method using the CPPI resin. The data obtained are given in Table 7. The isotope ratios determined in the untreated milk digest differ significantly from those determined after matrix separ- ation.This is expected because the former contains both C1 and Na which give rise to the interference 70Cl; and matrix suppression respectively. Thus the measured 70Zn 68Zn ratio may be elevated or depressed depending on the concentrations of Cl and Na. However no statistically significant difference in the isotope ratios is obtained for the two treatment pro- cedures. Precision of the isotope ratios of the treated milks are a little poorer than those of the untreated milk. This may be due in part to the dilution of the treated samples and the resulting lower signal count rate. Excellent Zn recovery (99%) is obtained with the resin separation method. Concentrations of elements present in the test samples (and related to the determination of Zn but not reported in Table 7 include Ca Na and C1.The concentration of Ca in the digested milk was about 260 pgml-'. After column separation the concentration was reduced to <20 pg ml-'. Sodium concen- tration was reduced from 60pgml-' in the digest to < 1 pg ml-' following column separation. Chlorine concen- trations were also low (about 20 p ml-') in the final solutions. For urine the Zn isotope ratios determined after extraction or matrix separation are not significantly different (Table 7). However the expected difference between treated and untreated samples is significant for the same reasons given above for the milk analysis. The precision of the Zn isotope ratios are better for the treated urine samples than for the untreated digest which might be due in part to the preconcen- tration of Zn that accompanies the urine treatment.A Zn recovery of 76% was obtained with the resin procedure. Typically Ca concentrations in the digested urine were about 200 pg ml-l. These were reduced to < 15 pg ml-' fol- lowing column extraction despite the 3-fold concentration produced by the separation. Similarly Na concentrations were reduced from 3800pgml-' in the untreated digest to < 1 pg ml-' in the solution subjected to column matrix separ- ation. Chlorine concentrations in the final treated solutions were typically < 15 pg ml-'. The lower concentrations of interferent elements and the higher Zn concentrations found in milk make the determi- nation of Zn in milk straightforward compared with that of urine.Moreover the digestion of milk is simple and rapid while the digestion of urine requires more manipulations and thus introduces a greater probability of elemental loss or contamination. Precipitation also can occur in the urine digest and requires the addition of distilled de-ionized water and re-heating of the sample. These factors perhaps explain the relatively low (76%) Zn recovery for urine with the resin procedure. Conclusions Although care must be taken with the handling of the digested material both to minimize contamination and to ensure adequate redissolution of the residue following drying of the digested material the proposed resin separation procedure is simpler and less labour intensive than the extraction methods previously Saturation of the CPPI resin would lead to low Zn recovery and sufficient resin should be placed in the column to avoid this.A similar limitation exists in the conventional CCl extraction. For example recovery of only approximately 50% was obtained in preliminary experiments with bovine milk and the conventional extraction resulting from too large an aliquot volume used for the extraction. Any error depends on the final Zn Concentration obtained. A poor recovery implies poor sensitivity and precision. However in principle the isotope ratio measurement should not be affected if the Zn concentration remains significantly above the limit of determination. In practice measuring the Zn concentration in the digested material is prudent.This allows identification of samples with high Zn content and permits determination of Zn recovery from the separation procedures. Thus an alternative sample treatment for matrix separation of biological materials has been established for subsequent Zn isotope ratio measurements by ICP-MS. The procedure is simpler than the conventional extraction procedures.10734 Moreover the procedures established earlier"*34 involve the extraction of Zn into a solution containing CCl and either APDC or diethylammonium diethydithiocarbamate (DDDC) respectively. Subsequently the Zn is extracted into nitric acid. As CC14 must not be transferred into the final solution careful and time-consuming sample handling is required. This intrinsic disadvantage of working with chlorine-containing reagents is not shared by the resin extraction method.The next stage in development is to employ an automated Table 7 70Zn 68Zn isotope ratios determined in biological samples by ICP-MS; analysis of six replicates Sample Ratio Standard deviation Zn recovery (%) Bovine Milk Solution- Digest 0.03478 0.00002 Conventional extraction 0.03742 0.00045 Column separation 0.03706 0.00076 Digest 0.1 1035 0.00395 Conventional extraction 0.03921 0.001 10 Column separation 0.04117 0.000 3 8 Human Urine Solution- 89 99 - 94 76204 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 system for the column separation. Similar procedures for both off- and on-line preconcentration with ICP-AES have been described.354’ A semi-automated procedure was applied to preconcentrate and separate trace metals from high-salt matrices prior to analysis by ICP-MS.21 On-line approaches reduce the time of analysis and the risk of sample contamination.We thank C. Amarasiriwardena for valuable discussions of the digestion procedures and ICP-MS analyses and P. Kandola for technical assistance. This research was supported by the ICP Information Newsletter and some equipment was provided by The Perkin-Elmer Corporation. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 References Applications of Inductively Coupled Plasma Mass Spectrometry eds. Date A. R. and Gray A. L. 1988 Blackie Glasgow. Hieftje G. M. and Vickers G. H. Anal. Chim. Acta 1989 216 1. Jarvis K. E. Gray A. L. and Houk R. S. Handbook of Inductively Coupled Plasma Mass Spectrometry 1992 Blackie Glasgow.Braun T. and Zsindely S. TrAC Trends Anal. Chem. 1992 11 267. Ward N. I. Abou-Shakra F. R. and Durrant S . F. Biol. Trace Elem. Rex 1990 26 177. Durrant S . F. Analyst 1992 117 1585. Amarasiriwardena C. J. Gercken B. Argentine M. D. and Barnes R. M. J. Anal. At. Spectrom. 1990 5 457. Ting B. T. G. Mooers C. S. and Janghorbani M. Analyst 1989 114 667. Cary E. E. Wood R. J. and Schartz R. J. Micronutr. Anal. 1990 8 13. Amarasiriwardena C. J. Krushevska A. Foner H. Argentine M. D. and Barnes R. M. J. Anal. At. Spectrom. 1992 7 915. Ward N. I. Abou-Shakra F. R. Durrant S. F. Thompson J. Havercroft J. M. and Yadegarian L. in Monography Proceedings Round Tables and Discussions of the 7th Int. Symp. on Trace Elements in Man and Animals (TEMA-7) ed.Momcilovic B. IMI Edinburgh Scotland 1991 pp. 33. Durrant S . F. TrAC Trends Anal. Chem. 1992 11 68. Zhu G. and Browner R. F. Appl. Spectrosc. 1987 41 349. Date A. R. Cheung Y. Y. and Stuart M . E. Spectrochim. Acta Part B 1987 42 3. Gray A. L. and Williams J . G. J. Anal. At. Spectrom. 1987 2 599. Gregoire D. C. Spectrochim. Acta Part B 1987 42 895. 17 18 19 20 21 22 23 24 25 Tan S. H. and Horlick G. J. Anal. At. Spectrom. 1987 2 745. Kim Y.-S. Kawaguchi H. Tanaka T. and Mizuike A. Spectrochim. Acta Part B 1990 45 333. Lyon T. D. B. Fell G. S. Hutton R. C. and Eaton A. N. J. Anal. At. Spectrom. 1988 3 265. Branch S. Ebdon L. Ford M. Foulkes M. and O’Neill P. J. Anal. At. Spectrom. 1991 6 151. Heitmar E. M. Hinners T. A. Rowan J. T. and Riviello J.M. Anal. Chem. 1990 62 857. Horvath Z. Lasztity A. and Barnes R. M. Spectrochim. Acta Reu. 1991 14 45. Kingston H. M. Barnes I. L. Brady T. J. Rains T. C. and Champ M. A. Anal. Chem. 1978 50 2064. Wang X. and Barnes R. M . J. Anal. At. Spectrom. 1989,4 509. Egan C. B. Smith F. G. Houk R. S. and Serfass R. E. Am. J.-Clin. Nutr. 1991 53 547. Ward. N. 1.. J. Micronutr. Anal. 1986. 2 211. 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 Rose,’ M. S The Foundations of Nutrition 3rd edn Macmillan New York 1938 p. 391. Myers S. and Tracy D. Spectrochim. Acta Part B 1983,38 1227. Barnes R. M. Biol. Trace Elem. Res. 1984 6 93. Horvath Z. and Barnes R. M. Anal. Chem. 1986 58 1352. Krushevska A. Barnes R. M. Amarasiriwardena C. J. Foner H. and Martines L. .I. Anal. At. Spectrom. 1992 7 851. Krushevska A. Barnes R. M. and Martines L. 1992 Winter Conference on Plasma Spectrochemistry Paper TP29 San Diego California January 6-11 1992. Emmett S . E. J. Anal. At. Spectrom. 1988 3 1145. Patterson K. Y. Veillon C. Moser-Veillon P. B. and Wallace G. F. Anal. Chim. Acta 1992 258 317. Barnes R. M. and Genna J . G. Anal. Chem. 1979 51 1065. Barnes R. M. and Lu S.-j. Can. J. Spectrosc. 1983 28 139. Knapp G. Muller K. Strunz M. and Wegscheider W. J . Anal. At. Spectrom. 1987 2 611. Prakash N. Csanady G. Michaelis M. R. A. and Knapp G. Mikrochim. Acta 1989 111 257. Schramel P. Xu L.-Q. Knapp G. and Michaelis M. Mikrochim. Acta 1992 106 191. Schramel P. Xu L.-Q. Knapp G. and Michaelis M. Fresenius’ J. Anal. Chem. 1993 345 600. Israel Y. Krushevska A. P. Foner H. Martines L. J. and Barnes R. M. J. Anal. At. Spectrom. 1933 8 467. Paper 3/0391 OB Received July 6 1993 Accepted September 1 1993
ISSN:0267-9477
DOI:10.1039/JA9940900199
出版商:RSC
年代:1994
数据来源: RSC
|
18. |
Determination of calcium, potassium, magnesium, iron, copper and zinc in maternal milk by inductively coupled plasma atomic emission spectrometry |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 205-207
Nereida Carrión,
Preview
|
PDF (441KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 205 Determination of Calcium Potassium Magnesium Iron Copper and Zinc in Maternal Milk by Inductively Coupled Plasma Atomic Emission Spectrometry* Nereida Carrion Ana Itriago Miguel Murillo Elias Eljuri and Albert0 Fernandez Centro de Quimica Analitica Escuela de Quimica Facultad de Ciencias Universidad Central de Venezuela P. 0. Box 4 71 02 Caracas 704 7 -A Venezuela A simple and rapid method was developed for the determination of Ca P Mg Fe Cu and Zn in maternal milk samples by inductively coupled plasma atomic emission spectrometry. Milk samples were emulsified with ethoxy nonylphenol to 0.03% m/v. The emulsified samples were diluted 10-fold with nitric acid and to 1 O/O v/v. Aqueous solutions with the same amount of emulsifier and acid were used as calibration standards.The accuracy of the proposed method was assessed using the National Institute of Standards and Technology Standard Reference Material 1549 Non-Fat Milk Powder. No significant difference at the 95% confidence level was observed. The precision of the method when applied to real samples was in the range 0.3-2% expressed as %RSD with no internal standard. Maternal milk samples from 50 women residing in Caracas Venezuela collected at three different lactation stages (on the third seventh and twenty-first days postpartum) were analysed. Keywords lnductively coupled plasma atomic emission spectrometry; human milk; trace element determination Human milk has long been regarded as the optimal source of essential nutrients for the young infant if the maternal diet is nutritionally adequate and a sufficient amount is consumed.Accurate data on the concentration of trace elements in human milk throughout lactation are important both for formulating nutritional requirements and for obtaining baseline levels leading to an understanding of the physiology of milk secretion.’ For this reason sensitive rapid and precise methods for the determination of trace elements in human milk are necessary. Inductively coupled plasma mass spectrometry (ICP-MS) and ICP atomic emission spectrometry (AES) have been used in the past for the determination of trace elements in Both techniques have analytical characteristics that are appli- cable in the routine determination of elements in milk i.e.low detection limits long linear working ranges and the capability of simultaneous multi-element determinations. Different sample treatments such as dry and wet digestion and decomposition with microwave heating in open and closed vessels have been used. Their efficiency is based on the time required and the completeness of the decomposition. Emmett2 reported trace metal determination in liquid and powdered milk by ICP-MS. Samples were digested by a wet acid pro- cedure and the results obtained compared well with the reference values. The decomposition method was lengthy. Dry ashing3-’ and wet acid digestio1-1~9~ procedures were also used for the determination of trace metals in liquid and powdered milk by ICP-AES. Human milk is a complex colloidal system that can be difficult to dissolve completely.Krushevska et a1.* carried out a comparison of several dry ashing and wet digestion pro- cedures for the determination of Zn in milk samples using ICP-AES. They reported that the wet dissolution procedure with a hot-plate is time-consuming and prone to contamination due to the large amounts of reagent required. The use of microwave systems increased the speed of sample dissolution compared with classical digestion but did not completely eliminate the requirement for chemical reagents. High-pressure digestion with a high-temperature programme destroys the organic carbon almost completely with only HN03. Zinc losses were observed at high temperatures when a dry ashing pro- * Presented at the XXVITI Colloquium Spectroscopicurn Inter- nationale (CSI) York UK June 29-July 4 1993.cedure was used without an ashing aid especially in the presence of chlorides. Direct dilution of sample is simple it can be automated and is less time-consuming than alternative procedures such those mentioned above. Durrant and Ward’ determined 18 elements in milk samples by ICP-MS. The samples were simply diluted to 2% v/v in aqueous acidic solution but the accuracy of the method is not very clear. Emmett2 in an attempt to directly analyse milk samples diluted with water reported a poor accuracy that was attributed to the fatty nature of the milk. Coni et aL5 also tried direct aspiration of the liquid milk samples into an ICP torch. They obtained an analytical signal lower (as much as one third) than those obtained by aspiration of the digested milk sample.This was associated with the rather large average droplets size characteristic of the un-treated milk and the incomplete atomization of components during residence in the plasma. They suggested that this method of determination is impracticable due to the difference in viscosity and surface tension values between real samples and aqueous calibration solutions. It has been reported1@13 when using flame atomic absorption spectrometry that the addition of surfactants to aqueous sample solutions improves the analyte sensitivity. It reduces the average droplet size in aerosols produced by pneumatic nebulizers due to a depression of the surface tension. Nevertheless Bertagnolli et a1.14 have reported that the use of surfactants in ICP does not improve sensitivity. They have mentioned that only the sample transport process is better thereby allowing a more stable plasma when using high sample uptake rates. One of the useful uses of surfactants in atomic spectrometry has been the emulsification of oil samples in water for the direct introduction of emulsions in both and ICP1s22 atomizers.The finely divided particles are uniformly dispersed in the water phase and the sample behaves like an aqueous solution. In this way the use of emulsions without prior destruction of the organic matrix has become an alternative method for sample introduction in atomic spectrometry. The aim of this investigation was to develop a direct rapid and simple method for the determination of Ca P Mg Fe Cu and Zn in human milk by ICP-AES by direct aspiration of emulsified samples.The accuracy was assessed by analys- ing the National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 1549 Non-Fat Milk Powder.206 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Experimental Apparatus A Jobin-Yvon Model JY24 inductively coupled plasma spec- trometer was used. Experimental parameters are presented in Table 1. Reagents and Materials All reagents were of the highest purity available. Milli-Q de- ionized water (Millipore) was used throughout the work. Nitric acid was Merck Suprapure grade. Ethoxy nonylphenol was obtained from Etoxil of Venezuela. Standard solutions were prepared from 1000 pg ml-' stock solutions (BDH) of each element.A 100 pg ml-' phosphorus standard stock solution was obtained by dissolving K2HP04 (BDH) in high-purity water. Cleaning of the plastic- and glass-ware All laboratory glass-ware and polystyrene plastic-ware con- tainers were acid-washed overnight in 3% v/v nitric acid and rinsed repeatedly with de-ionized water before drying at 60 "C. Once dried all plastic ware was heat-sealed in clean plastic bags to prevent further contamination. Sam p 1 e s A total of 50 lactating mothers provided milk samples for this investigation. All were residents of the city of Caracas (Venezuela). Maternal milk samples were collected at three Table 1 Instrumental parameters ICP atomic emission spectrometer Nebulizer Spray chamber Forward power/W Sample delivery Coolant flow rate/l min-' Nebulizer flow rate/l min-' Sample flow rate/l min-' Observation height Analytical lines/nm Jobin Yvon Model JY-24 Meinhard C-type Scott - type 1000 Peristaltic pump (Sorvall type 49061) 12 0.7 0.5 15 mm above load coil Ca I1 393.66 Cu I1 324.754 Fe I1 239.562 Mg I1 279.553 P 1213.618 Zn I1 213.856 Table 2 Analysis of SRM materials by direct ICP-AES methods Element Concentration Certified Found Zn P g g-' 46.1 f 2.2 45.0 f 0.7 c u Pi2 g-' 0.7 f 0.1 ND* Fe Pg g-' 1.78 _+ 0.02 1.72 f 0.12 Ca YO m/m 1.30+0.05 1.29 f 0.03 Mg YO m/m 0.120 f 0.003 0.1 18 k 0.005 P % m/m 1.06 f 0.02 1.05 k 0.03 * ND =Not detected.different lactation stages (on the third seventh and twenty- first days postpartum). Samples were collected in the early morning into previously prepared acid-washed 20 ml con- tainers from INDEMEDICA.Ethoxy nonylphenol aqueous solution (25% m/v) was added as an emulsifying agent (final concentration 0.30% m/v). The samples were stored at - 15 "C prior to analysis. Procedure The maternal emulsified milk samples were diluted 10-fold with nitric acid to 1% v/v (final concentration 0.030% m/v ethoxy nonylphenol and 1 % v/v nitric acid). Aqueous standard solutions for calibration purposes were prepared using identical acid and ethoxy nonylphenol concentrations as for the samples. The milk reference material (SRM 1549 Non-Fat Milk Powder) was reconstituted by suspending 2 g of the dried material in 25 ml of 0.03% ethoxy nonylphenol solution and then homogenized by ultrasonic agitation. The iron concen- tration was determined in this solution.For the determination of Zn however this solution was diluted 8-fold and 40-fold for Ca Mg and P. No agglomeration of solid particles was observed under these conditions. Results and Discussion Optimization of Analytical Conditions Under ideal conditions freezing is the best long-term preser- vation method. However during freezing some irreversible processes may take place because of denaturation of proteins and re-distribution of elements due to rupture of cell walls by ice crystals. An oily layer on the top of the liquid milk is produced which cannot be re-dispersed by mechanical pro- cedures such as stirring shaking or ultrasonic treatment. Ethoxy nonylphenol is a non-ionic surfactant that is com- pletely soluble in water.It is not affected by the presence of Ca Mg and Fe salts.23 This surfactant has been successfully used by Murillo et al. to emulsify lubricating oils2' and crude samples for the determination of metals by ICP-AES. The concentration of surfactant was evaluated to obtain the most stable milk emulsion. For this the chosen criterion was the re-dispersion of the oil layer into the whole milk sample so as to produce a homogeneous and stable emulsion. The homogenization of the emulsion was visually evaluated. The ethoxy nonylphenol concentration was varied from 0 to 0.53 % m/v. The optimum emulsifier concentration selected was 0.3%; concentration greater than 0.3% m/v did not improve the emulsion stability. At this optimum concentration stable emulsions are formed which can be kept under refriger- ation for a long time.Reconstruction of such samples is carried out by manual shaking. The emulsified sample was diluted 10-fold with nitric acid prior to nebulization (final concen- tration 0.030% m/v ethoxy nonylphenol and 1% v/v nitric acid). Reproducibility of the analyte signal was checked using samples that had not been frozen. In order to obtain representative portions of the sample it was better to add the emulsifier before sampling. By adding ethoxy nonylphenol to the samples several limitations associ- ated with milk fat separation during sample storage under frozen conditions could be avoided. Also ethoxy nonyphenol enabled direct sample introduction without blockage of the Table 3 Concentration ranges of elements in maternal milk collected 3 7 and 21 days postpartum ~ Concentration/pg ml- ' Day Ca Mg Ft c u Zn P 3 107-388 20.8-47.6 0.22-0.74 0.30-0.78 2.6-1 1.6 35-169 136-278 7 192-393 20.2-39 0.22-0.57 0.32-0.72 2.0-5.75 21 144-330 19.8-32.5 0.22- 0.5 5 0.32-0.65 1.01-4.3 105-234JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 207 nebulizer.The introduction of emulsions has improved trans- port and atomization efficiency of milk sample in ICP-AES. This may be associated with the reduction of the average size of the droplets.24 The technique also overcomes the problems associated with viscosity and surface tension differences between real samples and aqueous calibration solution^.^ Detection Limit The detection limits were calculated based on the Kayser definition25 with a signal-to-noise ratio of 3.The detection limits obtained were 0.05 0.0089 0.0079 0.023 0.023 and 0.055 pg ml-' for Zn Cu Fe Ca Mg and P respectively. Analysis of an SRM by the Proposed Method To assess the accuracy of the proposed method the NIST SRM 1549 Non-Fat Milk powder was analysed. The results obtained are presented in Table 2. The significance testz6 indicated no significant difference at the 95% confidence level between the observed and certified values for the elements determined. Analysis of Maternal Milk Samples The proposed method was applied to the determination of Ca Cu Fe Mg P and Zn in 50 maternal milk samples. The concentration ranges found are presented in Table 3. The precision was found to be satisfactory (between 0.3 and 2%).It can be seen that element concentration is higher in the early period of lactation namely colostrum (third day) and in transitory milk (seventh day postpartum). Similar results have been found by other worker^.^^.^^ Conclusions The method described enables the rapid and direct determi- nation of Ca Mg Fe Cu Zn and P in whole maternal milk and similar materials by ICP-AES. The addition of ethoxy nonylphenol to milk samples bypasses several limitations associated with milk fat separation during storage of frozen samples. This surfactant also facilitates direct sample introduction without nebulizer blockage. Aqueous standard solutions prepared under identical acid and surfactant concentrations as the samples were adequate for calibration purposes and complete recovery was obtained.For real samples reproducibility of the proposed method (as %RSD) varied from 0.3 to 2%. Good accuracy was obtained with no internal standard. This work was supported in part by Consejo de Desarrollo Cientifico y Humanistic0 of Universidad Central de Venezuela (Research Grant 03.12.2136.89). The authors gratefully acknowledge the financial support that has made this work possible. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 References Barnes L. A. Mauer A. M. Anderson A. S. Dallman P. R. Forbes E. B. Nichols B. L. Roy C. Smith N. J. Walker W. A. and Winick M. J. Pediatr. (St. Louis) 1978 2 591. Emmett S. E. J. Anal. At. Spectrom. 1988 3 1145. Feely R. M. Eitenmiller R. R. Jones J. B. and Barnhart H. J. Pediatr. Gastroenterol.Nutr. 1983 2 262. Feely R. M. Eitenmiller R. R. Jones J. B. and Barnhart H. J. Clin. Nutr. 1983 37 443. Coni E. Stacchini A. Caroli S. and Falconieri P. J. Anal. At. Spectrom. 1990 5 581. Suzuki K. T. Tamagawa H. Hirano S. Kobayashi E. Takahashi K. and Shimojo N. Biol. Trace Elem. Res. 1991 28 109. Li J. Yoshinaga J. Suzuki T. Abe M. and Morita M. J. Nutr. Sci. Vitaminol. 1990 36 65. Krushevska A. Barnes R. M. Amarasiriwaradena C. J. Foner H. and Martines L. J. Anal. At. Spectrom. 1992 7 851. Durrant S. F. and Ward N. I. J. Micronutrient Anal. 1989,5 111. Kodama M. and Miyagawa S. Anal. Chem. 1980 52 2358. Yan Z-y. and Zhang W. J. Anal. At. Spectrom. 1989 4 797. Mora J. Canals A. and Hernandis V. J. Anal. At. Spectrom. 1991 6 139. Ruiz A. I. Canals A.and Hernandis V. J. Anal. At. Spectrom. 1993 8 109. Bertagnolli J. A. Neylan D. and Hammargren D. D. At. Spectrosc. 1993 14 1. Berenguer V. and Hernhndez J. Quim. Anal. 1977 31 81. Berenguer V. Guinon J. L. and De la Guardia M. Anal. Chem. 1979 294,416. Polo-Diez L. Hernandez-Mendez J. and Pedraz-Penalva F. Analyst 1980 105 37. De la Guardia M. Salvador A. and Berenguer V. Analusis 1980 8 488. Lord C. Anal. Chem. 1991 63 1594. Borszeki J. Knapp G. Halmos P. and Bartha L. Mikrochim Acta 1992 108 157. Murillo M. Gonzalez A. Ramirez A. and Guillkn N. At. Spectrosc. in the press. Murillo M. and Chirinos J. paper presented at the XXVIII Colloquium Spectroscopicurn Internationale (CSI) York UK June 29-July 4 1993. Martinez A. A. Introduccidn a la Quimica de SuperJcies y Coloides Alhambra S.A. Madrid 1977. Becher P. Emulsions Theory and Practice Reinhold New York 2nd edn. 1965 ch. 6 and 7. Liteanu C. and Rica I. Statistical Theory and Methodology of Trace Analysis Wiley New York 1980 p. 255. Miller J. C. and Miller J. N. Statisticsfor Analytical Chemistry Wiley New York 1985 pp. 52 and 82. Casey C. E. Hambidge K. M. and Neville M. Am. J. Clin. Nutr. 1985 41 1193. Casey C. E. Neville M. and Hambidge K. M. Am. J. Clin. Nutr. 1989 49 773. Paper 3/03888B Received July 6 1993 Accepted December 14 1993
ISSN:0267-9477
DOI:10.1039/JA9940900205
出版商:RSC
年代:1994
数据来源: RSC
|
19. |
Determination of extractable organic chlorine by electrothermal vaporization inductively coupled plasma mass spectrometry |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 209-211
Pentti K. G. Manninen,
Preview
|
PDF (393KB)
|
|
摘要:
209 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 Determination of Extractable Organic Chlorine by Electrothermal Vaporization Inductively Coupled Plasma Mass Spectrometry* Pentti K. G. Manninen VTT Reacfor Laboratory P.O. Box 200 Fin42151 Espoo Finland Extractable organic chlorine (EOCI) is a sum parameter used to describe the organic chlorine concentration in solid or liquid samples that can produce injurious effects on health. In this work a new method is presented for the determination EOCI. Inductively coupled plasma mass spectrometry (ICP-MS) is not especially suitable for the determination of chlorine owing to the high ionization energy of chlorine. However when electrothermal vaporization (ETV) is compared with ICP-MS the sensitivity is high enough to facilitate sum parameter analysis from limited sample volumes.Ethyl acetate has been used as an extraction solvent. Chlorine was determined by collecting data during heating by monitoring mlz 35 using the single-ion monitoring mode. The detection limit for EOCl by ETV-ICP-MS is approximately 10 ng. Keywords Extractable organic chlorine; electrothermal vaporization ; inductively coupled plasma mass spectrometry Extractable organic chlorine ( EOCl) the lipophilic fraction of total organic chlorine (TOCl) can be characterized as that part of organically bound chlorine which can be separated from a sample matrix by using an organic water-immiscible solvent extraction procedure. In recent years several methods have been developed for the determination of EOCI especially for environmental research.lP4 Mainly because most of the known organic chlorine compounds such as chlorinated biphenyls ( PCBs) dibenzodioxins and phenols are generally considered as harmful or dangerous substances.Most of these methods use a microcoulometer or neutron activation analysis (NAA) as the detection method. Typical solvents are light petroleum (b.p. 40-60 "C) disopropyl ether pentane hexane cyclohexane and ethyl acetate. In this work a new detection method for EOCl is presented. Inductively coupled plasma mass spectrometry (ICP-MS) is not especially suitable for the determination of chlorine owing to the high ionization energy of chlorine. The detection limit for chlorine is > 10 pg 1-' compared with detection limits of < 1-10 ng 1-' for most other elements.However when electro- thermal vaporization (ETV) which was first introduced by Gray and Date,' is combined with ICP-MS the detection capability for the absolute amount of chlorine from limited sample volumes is increased. Sample introduction by ETV has several advantages over solution nebulization:6 for example a reduction in matrix and spectral interferences higher analyte transport efficiency because all of the vaporized sample is transported to the plasma small sample volume requirement and most importantly here the ability to analyse organic solvents directly. Recently Richner and Wunderli7 used the ETV-ICP-MS technique for determining polychlorinated biphenyls (PCBs) in oil. They have shown that it is possible to separate organic and inorganic chlorine by modifying the heating profile and even separate PCB isomers to some extent. Experiment a1 Equipment A Fisons PlasmaQuad PQ 11+ with a Fisons ElectroThermal Vaporizer ETV Mk I11 were used in this work.The operating conditions are listed in Table 1. The pyrolytic graphite coated graphite furnace tubes were from Ringsdorf-Werke The graphite furnace and the plasma torch were connected with clear nylon tubing (80 x 0.4 cm). Mass calibration and response * Presented at the XXVTII Colloquium Spectroscopicum Internationale (CSI) York UK June 29-July 4 1993. Table 1 Operating conditions Parameter R.f. power/W Outer gas flow rate1 min-' Intermediate gas flow rate/l min- Nebulizer gas flow rate/l min-' Dwell time per channel/ms Acquisition time/s Single-ion monitoring mode m/z Heating programme Time/s 0-60 60-65 65-80 80-90 Value 14 1 350- 1450 0.9 0.5-0.6 50 60 35 Temperature/"C 20- 100 100- 1000 1000-2600 2600 calibration were carried out by introducing a 100 pg I-' solu- tion (Be Co In La Pb Bi U) by the conventional solution nebulizing system consisting of a Gilson Minipuls 3 peristaltic pump V-groove De Galan nebulizer a Scott-type quartz spray chamber and a Fassel-type quartz torch.After tuning the nylon tube was connected between the torch and the graphite furnace and the torch position was re-checked by injecting a 10 mg 1-' solution of Hg into the furnace and monitoring the '"Hg signal without heating. Usually only a slight adjustment of the position of the torch was needed. Portions (10-50 pl) of sample solution were injected manually into the graphite tube and the dosing hole was closed with a carbon rod before the heating sequence was started.The data were collected by monitoring m/z 35 using the single-ion monitoring (SIM) mode during heating. The typical heating profile and acquiring time are presented in Fig. 1. Data were also collected at least once in every working period by scanning the mass range m/z 34-36 to check that the mass calibration which must be very precise in the SIM mode was stable. The EOCl results obtained by ETV-ICP-MS were compared with those obtained by NAA which was used for chlorine determination during the extraction procedure. Samples were irradiated in the Triga Mark I1 reactor in the 4 x 10'' n cm-2 s-' neutron flux using sealed polyethene cap- sules which were transferred to the reactor by a pneumatic transfer system.Chlorine was analysed by measuring a photo- peak at 1642 keV using a Ge(Li) detector and a Canberra MCA 40 multichannel analyser. A more detailed description of the method is published210 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 400 ,K! 300 8 3 0 5 200 or 100 1000 2600 20 4000 3000 2000 1000 0 Heating te m pe ra t u re/" C Fig. 1 Signal from (a) an ethyl acetate blank (b.p. 77 "C) and (b) a TCA standard (8 mg l-' b.p. 198 "C). Injection volume 20 p1 Reagents Analytical-reagent grade ethyl acetate (Baker) was used. Organic chlorine compounds which were used during the development and testing of the method were trichloroacetic acid (TCA Merck) pentachlorophenol (PCP EPA) 0- chlorobenzoic acid (CBzA Merck) and p-chloroanisole (CAn Fluka).Purified water was produced by the Millipore R04+Milli-Q system and nitric acid was pro analysi grade (Merck). Standard solutions of TCA (3 6 9 and 12mg1-I) were prepared in ethyl acetate. Procedure For solid samples the procedure was as follows. Sample ( NN 1 g) was extracted with 20ml of ethyl acetate under reflux for 30min. The extract was cooled in a cold water-bath and cleared by centrifugation. Liquid samples (5-20 ml) were extracted with 20 ml of ethyl acetate. All extracts were washed with 20-50 ml of purified water to remove inorganic chloride. The extracts were then stored in a freezer (-18°C) where water was separated from ethyl acetate as small ice pearls.The true amount of extract can be determined after filtration of the ice pearls. The extracts can be injected into the ETV- ICP-MS without any further treatment. The injected volumes varied from 10-5Opl depending on the type of sample and experiment. The data were collected during the heating period as shown in Fig. 1. The peak area was used for calculations. After every third or fifth sample 50 pl of 1% nitric acid were injected into the graphite tube and the heating profile was run. This procedure was necessary for the removal of carbon deposits which result in a reduction of the signal intensity from the sample cone. For NAA 2-3 ml of extract were put into the polyethene capsules which were sealed in plastic bags. These bags were put into liquid nitrogen before irradiation.After irradiation for 1 min the extracts were melted in a hot water-bath and 1 ml of extract was transferred to a clean capsule for measurement. Results and Discussion Ethyl acetate was selected as the organic solvent because of its high extraction power and applicability especially for sediment sarnple~.~*'~ The mean extraction efficiencies for the organic chlorine compounds tested were 91% for TCA 87% for CBzA and 107% for PCP. Another advantage of ethyl acetate is the very low chlorine background which facilitates its use without any pre-distillation or other purification procedure. Several heating profiles were tested and the slowly rising temperature programme was chosen for the determination of chlorine. With this programme the organically bound chlorine was separated very cleanly from the somewhat noisy signal coming from the vaporization step where the temperature is about 2600°C.Ethyl acetate was gently removed by heating the tube to 100°C over a period of 1 min. The temperature was then raised to 1000°C in 5 s. During this step all the organic chlorine compounds tested were vaporized and trans- ferred to the plasma where atomization and ionization of the element occurs. The organic load from the solvent and the sample did not noticeably affect the energy consumption of the plasma. Some carbon deposits from the solvent were found to build up especially on the sampler cone after 3-5 sample injections. Injection of 50 pl of 5% nitric acid efficiently removed the black deposits and therefore this procedure was repeated after every 3 or 5 samples.Problems were also associated with carbon deposits inside the connection tube between the ETV and the plasma torch. These deposits decreased after carefully readjustment of the temperature Cali- bration of the instrument. In Fig. l(a) the signal obtained from an ethyl acetate blank is shown. In Fig. 1 (a) and (b) the x-axis presents the data- acquisition time during the heating cycle. The temperatures reached are labelled on these axes rather than time in order to give a better view of what is happening during the heating. A small peak with a net area of about 900 counts is observed a few seconds after the start of the heating step from 100 to 1000 "C. This corresponds to a time of 62-63 s. A more noisy signal pattern is observed when the furnace temperature reaches its highest values.During this step the furnace was cleaned and a signal from inorganic chloride solutions (NaCl) was obtained. Hence if an inorganic chloride is present in an injected sample it is well separated from organic chlorine and does not disturb the determinations of EOC1. This has also been reported by Richner and W~nderli.~ In Fig. l(b) there is a clear peak from TCA (8 mg 1-' injection volume 20 pl). The limit of detection (LOD) for EOCl by ETV-ICP-MS is greatly dependent on the tuning of the instrument. A well- known problem is how to get a stable and sufficiently long signal for tuning the lens system and for centralizing the torch position when one is working with the ETV. Owing to these difficulties the chlorine response normally varied in the range 105-106 counts pg-'.The calibration line was linear (r=0.99) the slope was 12900 and the intercept was -3600. In this experiment the ethyl acetate blank value was 14 100 & 2500 counts. Using the formula LOD = 3sb an LOD of 10 ng was obtained for chlorine. This is an acceptably good value despite the fact that the repeatability is not very good. The relative standard deviation (RSD) for TCA standards is normally about 5% but for real extracts it usually varies from 5 to 30%. The organic chlorine compounds were then extracted into ethyl acetate and the EOCl concentrations were deter- mined by NAA p-chloroanisole (EOC134 mg 1-') pentachlor-JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 21 1 Pent ac h I o ro p h e n o I Trichloroethene pChloroanisol 0-C h I o ro benzo ic acic Organic chlorine compound Fig.2 Integrated peak areas divided by chlorine content of some organic chlorine extracts. The true absolute amount of chlorine is determined by NAA. The response to chlorine was of the same order of magnitude for all of these compounds ophenol (EOCl 8 mg l-') trichloroethene (EOC1 6 mg 1-I) and o-chlorobenzoic acid (EOC1 3 mg 1-'). The integrated counts pg-' of C1 have been determined by ETV-ICP-MS (Fig. 2). Each determination was repeated 3-5 times. The results show that a moderately good response and repeatability can be achieved using ETV-ICP-MS for the determination of chlorine with these compounds which each represent a differ- ent stage of chlorination.The method was also applied in the analysis of a real sediment sample from the Gulf of Bothnia (EOCl 570 pg 8-I dry mass by NAA). The ETV-IPC-MS result was slightly higher (165000f38000 RSD 23% n = 5 ) than for the test compounds indicating some matrix effect. These results are just tentative and the existence and nature of the matrix effects have not been investigated thoroughly. Conclusion Although the method discussed here is not a standard pro- cedure the possibility of its use in the determination of EOCl is clearly demonstrated. The technique is sensitive enough compared with conventional microcoulometer or NAA tech- niques and is element selective. These experiments have been carried out without any auxiliary gases or extra valve systems so they should be easily transported and repeated with any similar facilities.Future investigations are needed to assess possible matrix interferences and to improve repeatability. In this experiment it is thought that the poor repeatability and occasional loss of sensitivity could at least partly be attributed to a malfunction of the furnace instrument. References 1 2 3 4 5 6 7 8 9 10 Gron C. Vatten 1988 44 205. Wegmann R. C. C. 'Determination of Organic Halogens' Proceedings of the Second European Symposium on Analysis of Organic Micropollutants in Water Killarney Ireland November Griffin H. C. Copeland R. A. Epstein P. Harger R. and Wade D. R. Radiochem. Radioanal. Lett. 1981 50 67. Riggin R. M. Lucas S . V. Jungclaus G. A. and Billets S. J. Test. Eval. 1984 12 91. Gray A. L. and Date A. R. Analyst 1983 108 1033. Gregoire D. C. Lamoureux M. Chakrabarti C. L. Al- Maawali S. and Byrne J. p. J. Anal. At. Spectrom. 1992 7 579. Richner P. and Wunderli S. J. Anal. At. Spectrom. 1993 8 45. Manninen P. Ann. Acad. Sci. Fenn. Ser. AZZ. No. 225 1990. Manninen P. K. G. and Hasanen E. J. Radioanal. Nucl. Chem. 1993,167 353. Makinen I. Poutanen E.-L. and Manninen P. The Science of the Total Environment Elsevier Amsterdam in the press. 17-19 1981 pp. 249-263. Paper 3103891 B Received July 6 1993 Accepted October 7 1993
ISSN:0267-9477
DOI:10.1039/JA9940900209
出版商:RSC
年代:1994
数据来源: RSC
|
20. |
Simulation of the vaporization process in inductively coupled plasma atomic emission spectrometry with a modified model using the Monte Carlo technique |
|
Journal of Analytical Atomic Spectrometry,
Volume 9,
Issue 3,
1994,
Page 213-216
Hu Yanping,
Preview
|
PDF (512KB)
|
|
摘要:
JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 213 Simulation of the Vaporization Process in Inductively Coupled Plasma Atomic Emission Spectrometry With a Modified Model Using the Monte Carlo Technique* Hu Yanping and Zhang Zhanxiat Department of Chemistry Zhongshan University Guangzhou 5 10275 People's Republic of China Zheng Jianguo Guangdong Import and Export Commodity Inspection Bureau Guangzhou People's Republic of China An improved Monte Carlo program based on the velocity model and two-temperature model for the simulation of the aerosol vaporization process in an inductively coupled plasma (ICP) is presented. The influence of carrier gas flow rate Q on the useful mass transport rate W and the influence of Q on W at a given observation height above the load coil are studied.The variation of the evaporated and unevaporated droplet numbers with Q at a known observation height in the axial channel of the ICP is also presented. The results confirm experimental observations that carrier gas flow rate and observation height are the critical parameters in controlling W and hence in large part the observed emission signal. As expected the quality of the aerosol input is also shown to be an important factor. Keywords Monte Carlo simulation; vaporization process; inductively coupled plasma atomic emission spectrometry In previous work,' it has been shown that the average size and range of droplet sizes are functions of experimental parameters including nebulizer design liquid flow rate and gas flow rate. The nebulization process in an inductively coupled plasma (ICP) is also discussed and the relationship between the carrier gas flow rate Qg and the total mass transport rate wet investigated.According to Cresser and Browneq2 the value of KO is the sum of the excess mass transport rate We and the useful mass transport rate W and only the latter is related to the analyte emission signal. Thus W is an important parameter for the study of vaporization processes in the ICP. Generally a value for W is not easy to obtain experimentally. In this work the simulation of vaporization processes in the ICP with an emphasis on the value of W by using the Monte Carlo technique is described. Two models namely the two- temperature model and the velocity model are proposed for the simulation of the vaporization process of the aerosol droplets in the ICP.The two-temperature model is based on the fact that the gas temperature differs from the electron temperature and that they both vary with axial and radial position in the ICP whereas the velocity model is based on the fact that the moving velocity of the aerosol droplets differs from the gas velocity and that the former varies with the droplet diameter in the course of vaporization. Mathematical Models Aerosol Droplet Diameter According to Cresser and Browner,2 the variation of the aerosol droplet diameter d with vaporization time t and temperature can be expressed as follows d 3 = do3 - 48D,~P,M~t(p,R7')-~ exp(-MCIM,) (1) where do is the initial diameter of the aerosol droplet; D is the diffusion coefficient of the solvent vapour; G is the surface tension of the solvent; P is the saturated vapour pressure of the solvent; M is the relative molecular mass of the solvent; R is the gas constant; T i s the absolute temperature of the system; C is the concentration ( O h m/m) of the solution; M is the relative molecular mass of the solute; pa is the solvent * Presented at the XXVIII Colloquium Spectroscopicurn Internationale (CSI) York UK June 29-July 4 1993.t To whom correspondence should be addressed. density and t is the vaporization time. The equation is based on the assumption that the liquid and vapour phases of the solvent are in equilibrium. It is well known that the ICP discharge is spatially inhomo- geneous and is therefore characterized by concentration and temperature gradients that give rise to the transport of species by convection and diffusion.The parameters D P and t~ are the temperature dependent parameters. They can be derived from the following expre~sion:~ where MA and MB are the relative molecular masses of water and argon respectively if water is used as the solvent and Ar as the carrier gas P is the atmospheric pressure (atm) and T is the temperature of the system. Assuming that the solvent vaporizes rapidly in the high temperature environment of the ICP and that the vapour is an ideal gas the saturated vapour pressure of the solvent P can be obtained from the following equation P = P1 T/& (3) where 7'' is the temperature at a known saturated vapour pressure e.g. if 7'' = 653 K then PI is the saturated vapour pressure of the solvent at 653 K.Owing to the high temperature environment of the plasma the vaporization process of the aerosol in the ICP can be considered as a super-heated vaporization process. Thus the surface tension of the solvent can be expressed as f01lows:~ o V * = k ( T - T - d ) (4) where T is the critical temperature and V is the molar volume. For water as the solvent the constant d = 6.0 k = 2.2 erg per "C ( 1 erg z 4.36 x J). By substitution of eqns. (2) (3) and (4) into eqn. (l) the variation of the aerosol droplet diameter d with vaporization time t and temperature can be obtained. Two-temperature Model Since the temperature distribution in the ICP is inhomo- geneous it is necessary to establish an expression for the calculation of the temperature at different locations in the ICP.The atom-ion energy balance expression for the two-214 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 temperature model5 is given as follows where pVC,dT/aZ is the energy contributed by the atom-ion; 6 is the fraction of energy loss when a collision occurs between the electron and atom or ion. For Ar this value is 2.7 x lo-'; v is the collision frequency; and lai is the heat conduction coefficient of the atom-ion. For Ar gas this value is 2.7 x lo-'. The electron number density n can be derived from the expression where n is the atom number density m is the electron mass g is the statistical weight of the electron gi and g are the statistical weight of ions and atoms respectively and Ei is the ionization energy.The atom number density n can be derived from the partial pressure law (7) The electron collision frequency is determined by the follow- ing equation v = V,(nm$ern + na$ea + ni$ei) ( 8 ) where $, $ and $ei are the collisional cross-section of molecule atom and ion respectively and the electron velocity V = (8kT,/zm)* ( 9 ) Eqn. (5) is only an approximate equation in which the ambipolar diffusion is neglected and the ionization energy and dissociation energy in the transport process are not considered. The significant point in this two-temperature model is that the gas temperature of the atom-ion is determined by the given T or by the experimentally determined T,. The atom-ion gas temperature is not easy to obtain but to date an enormous amount of electron temperature data have been obtained experimentally with reasonable accuracy.With this two-tem- perature model T data can be used to calculate the atom-ion gas temperature. Assuming the values of T and pValong the central channel and the values of l v n and C along the cross-section are average constants the variation of gas tem- perature with vertical and radial displacement can be derived from the following expression where 0 < r < R for the central channel 0 < 2 < 1 for the channel length; a=pVC,; b=$(kvbn,); C=41/T2; R is the radial distance and Z is the axial distance. Velocity The velocity of an aerosol droplet introduced into the ICP is different from that of the gas. This difference should be taken into consideration in the Monte Carlo simulation.Assuming that drag and gravitation6 are the only factors that influence the moving velocity of an aerosol droplet the momentum equation for every aerosol droplet moving up vertically to the ICP torch is as follows Boundary condition when t = 0 Up = Uo,Uo = Q,/zr12vT/G; and T = 350 K. For Ar gas the molar expansion coefficient of the gas v is 1.0. This equation is based on the fact that the aerosol droplet has free diffusion velocity only in the radial direction and that the relationship between the droplet velocity and the gas velocity in the radial direction is neglected. The Up is the aerosol droplet moving velocity in the axial direction at Z = 0; To is the environmental temperature; Q is the carrier gas flow rate; rl is the internal radius; pa is the aerosol droplet density; p is the carrier gas density; d is the diameter of the aerosol droplet; g is the gravitation constant; Cd is the drag coefficient and U is the gas velocity within the plasma torch defined approximately as rzL - G where Q is the gas flow rate in the intermediate tube and r2 is the internal radius of the intermediate tube.The moving velocity of the incompletely evaporated droplet is assumed to be Up and that of the completely evaporated droplet U which is actually the gas velocity within the plasma torch. Basis of the Monte Carlo Simulation As mentioned previously,' a normal random number has been used to describe the aerosol droplet size distribution at every moment thus the droplet size introduced into the ICP also has a normal distribution.If the aerosol droplet size distri- bution is within the [O,d,] interval the normal random number R within the [0 dp] range could be selected to describe the size of the aerosol droplet. Since the plasma torch is cylindrical in structure only the axial and radial variations are considered. The aerosol droplet will diffuse freely in the radial direction and at the same time will also move along with the gas in the Z direction. Thus within a small time interval At the spatial displacement of an aerosol droplet can be described as follows dZ = R,,(2DTAt)* + U,dt (13) (14) loAf d R = Rg2( 2DTAt)* where Dz and dR are the axial and radial movement respect- ively R and R are the random Gaussian numbers distrib- uted about 0 with a standard deviation of 1.0 respectively and DT is the temperature dependent diffusion coefficient.If the aerosol droplets are vaporized completely eqn. (13) can be expressed as follows dZ = Rgl(2DTt)' + UAt (15) Owing to evaporation the diameter of each of the random aerosol droplets will become smaller and smaller as it moves randomly. Accordingly its moving velocity will also vary. The simulation time interval At is 0.01 ms. For each At interval the temperature is different but within this very short period the temperature can be assumed to be constant. In other words the temperature in the whole process varies in a step-wise manner. By using eqn. (lo) the temperature in the axial and radial movement can be determined. Experimental Instrumentation The dimensions of the ICP torch structure are shown in Fig.1 and the operating conditions are given in Table 1. The dimen- sions of the concentric nebulizer nozzles provided by Canals et aL7 are used in this simulation work. They are tabulated in Table 2. The analyte solution used is 1000 pg ml-I of Mn(N0,)2.JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY MARCH 1994 VOL. 9 215 z=o r1 1 Fig. 1 Schematic diagram of ICP torch structure internal radius of outer tube R = 1.88 cm; internal radius of intermediate tube r2 = 1.25 cm; external radius r3 = 1.50 cm; and injection tube radius ri = 0.12 cm Table 1 Operating conditions for the ICP torch Parameter Value Plasma gas flow rate (Ql)/l rnin-l 12.0 1 .o 0.39-1.1 5-25 Intermediate gas flow rate (Q?)/l min-' Carrier gas flow rate (Qg)/l min-' Observation height above load coil/mm Table 2 Dimensions of the nebulizer nozzles7 Concentric nebulizer Parameter 1 2 3 Inner diameter of the inner sample Outer diameter of the inner sample Inner diameter of the outer gas Cross-section area of the gas Cross-section area of the liquid tube/mm 0.318 0.508 0.424 t u be/mm 0.510 0.700 0.538 tube/mm 0.580 0.730 0.574 outlet/mm2 0.060 0.034 0.031 outlet/mm2 0.079 0.203 0.141 Recess of the inner tube/mm +0.07 -0.06 -0.06 The Monte Carlo program is written in Fortran 77 and run on an M340 computer.It has a memory of 12 megabytes and the hard disk has a memory of 4300 megabytes. The CPU capacity is 2 500 000 computations per second. Simulation Procedure The simulation was carried out as follows. (i) Generate the normal random number R of the aerosol droplet size distribution.(ii) At temperature (when i = 0 Tis the initial tempera- ture) use eqn. (1) to obtain the aerosol droplet diameter d i after vaporization within the At interval. (iii) Derive the velocity of the aerosol droplet from the velocity equation. (iu) Substitute the parameters obtained into eqns. (13)-( 15) respectively to obtain the axial and radial movement of the aerosol particle. (u) Using eqn. (lo) the accumulation Ed value and EdR value calculate the plasma gas temperature T+l at the next interval At. (vi) Repeat procedures (ii)-(v). (vii) When the Ed value is equal to or greater than the required observation height or the Ed value is greater or equal to the boundary radius R accumulate the evaporated droplet numbers for d < 0.01 pm droplets with d > 0.01 pm are considered to be unevaporated.Then return to procedure (i). Repeat all the procedures until the evaluation for all the simulation particles (in this work 100 000 particles) is complete. Calculation of Useful Mass Transport Rate The useful mass transport rate W can be calculated by using the following equation w u = nz/N x w o t ( 16) where Kot is the total mass transport rate N is the total simulation particle number and n2 is the number of evaporated droplets in the central channel. Results and Discussion Influence of Qg on W The effect of Q on W at a height of 10mm above the load coil (ALC) for three different concentric nebulizers are depicted in Fig. 2. A peak value of W for all three nebulizers is observed at Qg=0.841min-'.Beyond this value the W decreases rapidly and at Q = 1.1 1 min-l it decreases almost to zero. This phenomenon is fairly different from that observed with Fat. This shows the importance of selecting the optimum Q,. At the optimum value of Q not only will the diameter of the droplet be sufficiently small for rapid vaporization but also the droplet will have sufficient residence time in the ICP for vaporization to occur. Variation of W With ALC Fig. 3 shows that the W values decrease with an increase in height ALC for all the Q studied. Thus the height ALC should be selected carefully to obtain the appropriate value of W,. This figure further confirms that a carrier gas flow rate of 0.84 lmin-I gives the highest W at all heights ALC for nebulizer 1; nebulizers 2 and 3 have also been evaluated and were found to exhibit the same phenomena.Influence of Qg on Droplet Numbers The influence of Q on the maximum diameter of the aerosol droplet and the number of unevaporated and evaporated 3.00 + I I I I 0.30 0.50 0.70 0.90 1.10 Q Fig. 2 Influence of carrier gas flow rate Qg on useful mass transport rate W at 10mm observation height A nebulizer 1; B nebulizer 2; and C nebulizer 3216 JOURNAL OF ANALYTICAL ATOMIC' SPECTROMETRY MARCH 1994 VOL. 9 Table 3 (Evap.) droplets in the central channel of the ICP Influence of Qg on the maximum diameter (dmax) of the aerosol droplet and the number of unevaporated (Unevap.) and evaporated Carrier gas flow ratejl min-' 0.39 0.52 0.65 0.84 Nebulizer Parameter No. 1 dmaxlilm Evap.Unevap. Evap.7 ('10) Evap. Unevap. Evap. (YO) Evap. Unevap. Evap. (YO) No. 2 dmax l P No. 3 d m a x l P 10* 23.7 1329 171 88.6 20.1 1117 238 82.4 16.9 1100 44 96.2 15* 1966 30 97.3 - - 966 57 94.4 - 875 8 99.0 10 15 20.5 1405 1094 74 14 95.1 98.7 15.8 - 1210 976 55 0 95.6 100 14.4 - 1149 924 14 2 98.9 99.8 - 10 15 18.2 - 1451 1120 30 0 97.9 100 12.4 .- 1189 958 1 0 100 100 11.3 - 1125 916 0 0 100 100 10 16.2 1384 180 88.5 9.8 1155 71 94.2 7.5 1971 119 90.0 15 1067 137 - 88.6 - 957 50 95.0 - 90 1 81 91.8 1.10 10 12.7 150 12404 1.19 7.5 130 10885 1.2 7.0 99 13415 0.7 15 104 12065 - 0.9 - 95 10595 0.9 - 72 13085 0.5 * Height ALC in mm. Evap. Evap. + Unevap. Evap. (%) = x 100. 3.50 3.00 .- 2.50 I v) 0 5 2.00 7 1.50 1 .oo m I 0 3' 0.50 0 5 10 15 20 25 30 ALC/mm Fig.3 Useful mass transport rate W as a function of height ALC at the central channel of the ICP for nebulizer 1 for flow rates of A 0.39; B 0.52; C 0.65; D 0.85; and E 1.10 (1 min-') droplets are shown in Table 3. The following phenomena are observed. (i) For all the nebulizers studied the maximum diameter of the droplets introduced into the ICP decreases with an increase in Qg. The data also show that nebulizer 3 generates smaller diameter aerosol droplets than the other two nebulizers. Evidently this Monte Carlo simulation is helpful for assessing the quality of the nebulizer. (ii) The data show that even at 10-15mm ALC a certain number of unevaporated droplets still remain in the ICP and the numbers vary with the value of Qg. However the number of unevaporated droplets decrease as the height ALC increases from 10 to 15 mm at the central channel.(iii) The percentage of evaporated droplets is closely related to the maximum diameter of the droplet. The smaller the diameter the greater is the percentage. Thus the nebulizer that generates droplets with smaller maximum diameter eg. nebulizer 3 will give a higher percentage of evaporated droplets. Conclusion The proposed Monte Carlo simulation program is valid for examining the effects of Qg on the maximum diameter of the aerosol droplets and the number of unevaporated and evapor- ated droplets in the central channel and can also be used to demonstrate the effects of Qg and height ALC parameters on W in the ICP. The program is useful for selecting optimum operating parameters and nebulizers. The spatial distribution of the analyte particle can also be obtained using the proposed program; reports on this will be given in a future paper. Though W is closely related to the analyte emission inten- sity it is however not directly related to the emission intensity of the analyte. Further investigation of the atomization ioniz- ation and excitation of the analyte useful mass should be considered. This work was supported by the Natural Science Foundation of China. References Hu Y. Zheng J. and Zhang Z. Guangpuxue Yu Guangpu Fenxi submitted for publication. Cresser M. S. and Browner R. F. Spectrochim. Acta Part B 1980 35 73. Shi J. Chemical Engineering Handbook Chemical Industry Publisher Beijing 1989 ch. 10 p. 83. Tan M. and Huan W. Surface Physical Chemistry China Construction Industry Publisher Beijing 1985 p. 24. Dresvin S. V. Low Temperature Plasma Physics and Technique Publisher of the National Committee of Soviet Minister CCCP on Utilization of Atomic Energy Moscow 1972. Boulos M. Pure Appl. Chem. 1985 9 1321. Canals A. Hernandis V. and Browner R. F. Spectrochim. Acta Part B 1990 45 591. Paper 31039540 Received July 7 1993 Accepted November 1 I 1993
ISSN:0267-9477
DOI:10.1039/JA9940900213
出版商:RSC
年代:1994
数据来源: RSC
|
|