JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 I VOL. 6 623 Slurry Sampling and Fluorination-Electrothermal Vaporization Inductively Coupled Plasma Atomic Emission Spectrometry for the Direct Determination of Molybdenum in Food Hu Bin Jiang Zucheng* and Zeng Yun'e Department of Chemistry Wuhan University Wuhan 430072 China A method for the direct determination of Mo in various types of food by fluorination-electrothermal vaporization inductively coupled plasma atomic emission spectrometry has been developed and optimized. Slurry samples were prepared by ultrasonic wave vibration after pre-treatment and direct injection into the graphite furnace using polytetrafluoroethylene as a fluorinating agent. The procedure was applied to the determination of Mo in National Institute of Standards and Technology Standard Reference Materials 1567 Wheat Flour 1568 Rice Flour and 1577 Bovine Liver.The values found were in reasonable agreement with the certified values with a detection limit of 0.7 ng mi-l and an RSD of 3.2% (n-10) at a concentration of 0.1 pg mi% The proposed procedure has been applied successfully to the analysis of various types of food samples; the recovery ranged between 92 and 105%. Keywords Molybdenum determination; fluorination-electrothermal vaporization; polytetra fluoroethylene slurry; inductively coupled plasma atomic emission spectrometry Molybdenum is an essential trace element required by both plants and animals.' The importance of Mo in animal nutrition has been recognized for over 40 years. The antagonistic effects on the metabolism of Cu in ruminant animals has attracted much attention in the past,2 but recent findings have indicated that Mo itself can have important direct effects on the biological processes control- ling growth and reproductive perfo~mance.~ It has also been shown from studies with patients receiving total parented nutrition that Mo is an essential element for man,4 the major source for man of Mo being in food.There is a need therefore for a convenient method of assessing the Mo status in man. Although there are a number of techniques available for the determination of Mo in biological materials recent interest has been focused on the use of electrothermal atomic absorption spectrometry (ETAAS) and inductively coupled plasma atomic emission spectrometry (ICP-AES).However the formation of stable non-volatile carbides of Mo on the graphite surface is a serious drawback to the use of ETAASSd Barium difluoride has been proposed as a chemical modifier for the determination of Mo in serum' and milk.* When using this method the appearance temperature of Mo is lowered the signal is increased and matrix interference effects are partly overcome but the memory and signal tailing effects still remain especially with solid sampling.* Electrothermal vaporization (ETV) has developed into an important tool for trace element analysis in recent years as it combines the advantages of both ETAAS and ICP-AES. Aziz et aL9 have successfully coupled a commercially available graphite furnace (Perkin-Elmer HGA-74) to an ICP torch and determined Cd Pb Mn and Zn in two National Institute of Standards and Technology (NIST) Standard Reference Materials (SRMs) and commercially available serum samples. Matusiewicz and BarneslO de- scribed a method for the determination of Al Li Au Os Pd Pt and Ru in body fluids at therapeutic levels by ICP- AES with introduction of 5 pl samples by ETV.Recent developments of this technique have been reviewed by Ng and Carusol and Matusiewicz. l2 However this approach retains the disadvantage of electrothermal atomization with respect to the formation of ~ _ _ _ ~ ~ _ _ _ _ _ _ ~ ~____ *To whom correspondence should be addressed. refractory carbides by elements such as Ti Zr V Cr Mo W Nb Ta Hf B Si and rare earth elements,13-15 which leads to a decrease and sometimes complete suppression of the analyte signals.Improved methods have been reported that include conversion of the analytes into volatile hal- i d e ~ ~ ~ - ~ ~ or graphite tubes coated with a layer of compacted metal carbides to prevent carbon from reacting with the elements under investigation. l6 A previous paper from this laboratory17 described a method for the direct determina- tion of B in a plant sample by fluorination and ETV-ICP- AES using polytetrafluoroethylene (PTFE) as a fluorinating agent; the detection limit for B was as low as 24 pg. The present study was carried out to develop an accurate precise and rapid method for the determination of Mo in various types of food by slurry sampling fluorination-ETV- ICP-AES using PTFE as a fluorinating agent.Experimental Apparatus The experimental details for the inductively coupled plasma and electrothermal vaporizer have been described previ~usly.~~ A commercial 2723 MHz Ar ICP source (Beijing Broadcast Instrument Factory Beijing China) with a 2 k W plasma generator was interfaced to a WDG 500- 1 A monochromotor (Beijing Second Optics Beijing China). The output of the photomultiplier (R456 Hama- matsu Japan) was amplified and registered on a strip-chart recorder (L23- I04 Sichuan Fourth Meters Shanghai China). Table 1 ETV-ICP-AES operational parameters Wavelength Incident power Carrier gas (Ar) flow rate Coolant gas (Ar) flow rate Auxiliary gas (Ar) flow rate Observation height Entrance slit-width Exit slit-width Drying temperature Ashing temperature Atomization temperature Sample volume 202.030 nm 1.0 kW 0.5 1 min-' 16 1 min-I 0.8 1 min-' 12 mm 25 pm 25 pm 100 "C ramp 15 s hold 15 s 380 "C ramp 10 s hold 20 s 2100 "C 3 s 10 jd624 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL.6 A WF-1 type heating device with a matching graphite furnace (Beijing Second Optics) was used as the electrother- mal vaporization device. The instrument operating condi- tions and wavelength used are given in Table 1. Reagents A stock solution of Mo with a concentration of 1 g 1-l was prepared by dissolving 0.1500 g of MOO (Specpure Shanghai Reagent Factory China) in 10% v/v ammonia solution (Suprapur Shanghai Reagent Factory) and diluting to volume with doubly distilled water. A 60% m/v slurry of PTFE (viscosity 7 x 10-,-1 5 x lo- Pa s) was supplied by the Shanghai Institute of Organic Chemistry.Except for Na2Si03 (analytical-reagent grade Shanghai Reagent Fac- tory) the other reagents (NaC1 KCl CaCl CuCl FeCl AlCl and MgC1,) were all of Specpure grade. Doubly distilled water was used throughout. Preparation of Slurry Samples Samples of pollen and spinach were dried at 105- 1 10 "C for at least 24 h and garlic was shelled and washed with tap water followed by doubly distilled water and air dried. The samples were crushed and 1 .OOOO g of the spinach and garlic and 0.5000 g of the pollen were accurately weighed into acid-washed crucibles. The materials were charred in a muffle furnace at 600 "C for 2 h then quantitatively tranferred into acid-washed PTFE bottles together with agate spheres and 5 ml of a slurry containing varying amounts of PTFE depending on the sample (pollen spinach or garlic) were added accurately.The bottles were agitated on a flask shaker for 1 h to reduce the particle size. The addition of Mo to the slurry samples was achieved by pipetting 0 10 30 and 100 pl of a 1 pg ml-l Mo solution into 0.1 0.5 and 0.5 ml slurries of the pollen spinach and garlic respectively. The mixtures were diluted to 1 ml and dispersed with an ultrasonic wave vibrator for 15 min after which the bottles were shaken vigorously prior to sampling. This results in slurry samples containing 0 10 30 and 100 ng ml-l of added Mo respectively and 6% m/v of PTFE. Powder samples (milk powder 1.0000 g and wheat flour 0.5000 g) were mixed with 2.5 ml of water then heated to 50 "C and stirred.Samples (0.5 ml) of these slurries were taken and diluted to 1 ml with the appropriate amounts of PTFE slurry and various volumes of the Mo standard solutions. The mixtures were dispersed with an ultrasonic wave vibrator for 20 min after which the bottles were shaken vigorously prior to sampling. The resulting slurry samples contained 0 10 30 and 100 pg 1-l of added Mo and 6% m/v of PTFE. Procedure After igniting the plasma the gas flow rate power and viewing height were adjusted to the conditions given in Table 1. The details of the drying ashing and atomization stages of the electrothermal vaporizer are also given. The appropriate wavelength was selected by aspirating a Mo solution. The pneumatic nebulizer was then disconnected from the plasma torch and replaced by the ETV system.A 10 p1 volume of sample was then deposited into the furnace and the drying and ashing steps were initiated. After ashing was complete the injection hole of the graphite furnace was sealed with a graphite rod and the vaporizing sequence was repeated according to the conditions described in Table 1. The desolvated vaporized sample was carried into the plasma by the Ar carrier gas. The transient emission intensity for the Mo line selected was recorded by the strip- chart recorder. A calibration graph was constructed using peak height measurements. Results and Discussion ICP Discharge Parameters The ICP discharge parameters were established using a standard solution of 0.1 pg ml-l of Mo containing 6% m/v PTFE and the signal-to-background ratios were used to take the measurements. The results showed that 1.0 kW power a carrier gas flow rate of 0.5 1 min-l and a 12 mm observation height were the optimum conditions.Optimization of the Graphite Furnace Programme Experiments were carried out to determine the best temper- ature and times for the various drying ashing and atomiza- tion steps. Optimum drying conditions were required to provide a smooth even evaporation of the solvent (water) with no spluttering; a drying temperature of 100 "C was used. Proper choice of the ashing temperature is very impor- tant. When the ashing temperature is lower than the decomposition temperature of any organic compounds present it can lead to interferences from the organic matrix.However when the ashing temperature is higher than the temperature at which the analyte vaporizes decreases in the emission signal of Mo may occur. In order to optimize the ashing and atomization temperatures ashing and atomiza- tion curves were constructed for samples containing 0.1 pg ml-* of added Mo and the optimum concentration of fluorinating agent; the results are shown in Figs. 1 and 2. It was seen that when PTFE was present decreases in the Mo analytical signal occurred at less than 400 "C whereas without PTFE no decrease was found at temperatures of more than 1240 "C. This indicated that on addition of PTFE Mo reacted with it in the graphite furnace and vaporized in the form of the volatile MoF (boiling-point 36 "C). The optimum ashing temperature was 380 "C.It is evident from Fig. 2 that the presence of PTFE greatly influenced the vaporization behaviour of Mo and the vaporization reached a plateau at 1500 "C. This showed that the fluorination reaction between Mo and PTFE was complete. By contrast in the absence of PTFE no plateau was found in the temperature range tested. From Figs. 1 and 2 it can also be seen that with fluorination-vaporiza- tion the Mo emission signal intensity is much more intense than that without the use of PTFE as a fluorinating agent. In this experiment 2100 "C was chosen as the atomization temperature for the determination. Optimization of Amount of Fluorinating Agent Different amounts of PTFE were added to a series of 500 p1 sample suspensions containing 0.2 pg ml-l of Mo. The mixtures were diluted to 1 ml with water and subjected to I A 1 I I I 200 600 1000 1400 i a TemperaturePC Fig.1 Ashing curves for Mo using ETV-ICP-AES for the introduction of 10 pl of slurry sample into the graphite furnace A 0.1 pg ml-I of Mo in 6% m/v PTFE; and B 1 pg ml-* of Mo without PTFEJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL. 6 .- a9 .!6- 0 .- w .- .- U k > 3 - - 625 f A' / z c 6 E 3 - .- P *; .- *.' - a K Fig. 2 Atomization curves for Mo using ETV-ICP-AES for the introduction of 10 pl of slurry sample into the graphite furnace A 0.1 ,ug ml-1 of Mo in 6% m/v PTFE; and B 1.0 pg ml-1 of Mo without PTFE -Y I I 1 I I PTFE concentration (% m/v) Fig. 3 Effect of PTFE concentration on fluorination-vaporiza- tion; 10 pl sample 0.1 p g ml-I Mo the furnace programme. The Mo emission signal was accentuated by PTFE at concentrations up to 5% d v .The maximum emission intensity achieved with this concentra- tion remained constant up to the highest amount studied (10%) (Fig. 3) but the stability of the plasma discharge decreased markedly and actually extinguished at high PTFE concentrations owing to vigorous decomposition of the PTFE in the plasma. The optimum concentration of PTFE chosen was 6%. Matrix Interferences The influences of the matrix on the fluorination-vaporiza- tion of Mo were investigated. The interfering elements were Al Ca Cu Fe K Mg Na and Si with a Mo concentration of 0.1 pg ml-l. It was found that by using PTFE as fluorinating agent amounts of Ca Cu K Mg and Na of up to 5 g 1-1 and A1 of up to 2 g 1-l did not affect the Mo vaporization process.Iron did not interfere with the determination of Mo at a concentration below 1 g 1-l. However when the amount of Fe was in excess of 2 g I-l the emission intensity of Mo was enhanced markedly. This is due to the Fe 202.050 nm spectral interference caused by evaporation of Fe. When the Si content was more than 1 g l-l the Mo signal decreased significantly. This is because a large amount of Si reacts with the PTFE and inhibits the vaporization of Mo. Memory Effects When using the optimum operating conditions no memory effects were apparent for 10 pl of a 10 pg ml-l solution of Mo (Fig. 4). C U 3s U B W I - r Fig. 4 Recorder tracings for 10 pl samples A 0. I pg ml-I of Mo in 6% m/v PTFE; B residual signal of the first firing after vaporizing 10 p1 of a 10 pg ml-1 of Mo in 6% m/v PTFE sample; and C residual signal for the second firing Mo concentrationlpg mr' Fig.5 Calibration graph for Mo obtained at 202.030 nm using 6% m/v PTFE as the fluorinating agent Calibration In order to obtain a calibration graph standard solutions containing 0.0 1- 10 pg ml-l of Mo with 6% m/v PTFE were subjected to the furnace programme. The results are shown in Fig. 5. As can be seen the graph is linear over a concentration range of three orders of magnitude. Detection Limit Precision and Accuracy According to the recommendations of the American Chemical Society Committee of Environmental Improve- ment the detection limit the lowest concentration level that can be determined to be statistically different from a blank is defined as three times the within-batch standard deviation of a single blank determination corresponding to a 99Oh confidence level.The detection limit for Mo with fluorination-ETV-ICP-AES is 0.7 ng ml-l at a concentra- tion of 0.01 pg ml-l but the detection limit for Mo without PTFE is 30 ng ml-l. The detection limit was improved by approximately two orders of magnitude by using fluorina- tion-vaporization. The relative standard deviation (RSD) of this method obtained for ten replicate determinations at a concentration of 0.1 pug ml-l was 3.2%. Table 2 Concentration of Mo in NIST SRMs obtained by fluorination-ETV-ICP-AES Reference material Found value Certified value SRM 1567 Wheat Flour 0.43 k 0.02 0.40 SRM 1568 Rice Flour I .48 -t 0.10 1.60 SRM 1 577 Bovine Liver 3.40 f 0.17 3.50 (PPm) (PPm)626 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 199 1 VOL.6 Table 3 Results for determination of Mo in various types of food obtained by standard additions and slurry sampling fluorination-ETV- ICP-AES. All of the results are the means of triplicate analyses Amount of Mo/pg 1-I Sample Pollen Milk powder Wheat flour Garlic Spinach Content Before Amount After Recovery in sample/ 54.8 50.0 107.5 105 5.48 k0.21 addition added addition (%I C(g g-I 10.8 10.0 20.0 92 0.054 k 0.0 1 42.0 40.0 79.5 94 0.42 k0.06 21.8 20.0 41.2 97 0.22 kO.05 18.0 20.0 36.6 92 0.18 k0.04 To study the accuracy of the method NIST SRMs 1567 Wheat Flour 1568 Rice Flour and 1577 Bovine Liver were used. The results are given in Table 2.The agreement with the certified values was very good. The accuracy of the method was also determined by measuring the recovery of standard additions of Mo to the samples. The recovery ranged from 92 to 105% (Table 3). Sample Analysis The proposed method was applied to various types of food samples for the determination of Mo. The results obtained using the standard additions procedure are shown in Table 3. Conclusions The results of this study show that the use of PTFE as a fluorinating agent not only enhances the sensitivity for the determination of Mo by ETV-ICP-AES but also eliminates memory effects. By using slurry sampling the procedure provides a means for determining M o in solution or solid samples which can reduce sample preparation time de- crease analyte losses due to volatilization and contamina- tion.Furthermore any element capable of forming a fluoride that is more volatile than the original form of the analyte present after drying could potentially benefit from this approach. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 References Underwood E. J. Trace Elements in Human and Animal Nutrition Academic Press New York 1977. Mason T. Toxicology 1986,42 99. Phillippo M. Humphries W. R. Atkinson T. Henderson G. D. and Garthwaite P. H. J. Agric. Sci. 1987 109 321. Abumrad N. N. Schneider A. J. Steel D. and Rogers L. S. Am. J. Clin. Nutr. 1984 34 2551. Muller-Vogt G. Wendl W. and Pfundstein P. Fresenius Z. Anal. Chem. 1983 314 638. Sneddon J. Ottaway J. M. and Rowston W. B. Analyst 1978,103 776. Ericson P. McHalsky M. L. and Saselskis B. At. Spectrosc. 1987 8 101. Wagley D. Schmiedel G. Mainka E. and Ache H. J. At. Spectrosc. 1989 10 106. Aziz A Broekaert J. A. C. and Leis F. Spectrochim. Acta Part B 1982,37 369. Matusiewicz H. and Barnes R. M. Acta Chim. Hung. 1988 125 777. Ng K. C. and Caruso T. A. Appl. Spectrosc. 1985 39 719. Matusiewicz H. J. Anal. At. Spectrom. 1986 1 171. Kirkbright G. F. and Snook R. D. Anal. Chem. 1979 51 1938. Ng. K. C. and Caruso J. A. Analyst 1983 108 476. Huang M. Jiang Z. and Zeng Y. Gaodeng Xuexiao Huaxue Xuebao 1989,5 288. Ng. K. C. and Caruso J. A. Anal. Chim. Acta 1982,143,209. Hu B. Jiang Z. and Zeng Y. Fresenius J. Anal. Chem. 199 1 340,435. Paper 0/058 12B Received December 28th I990 AcceDted June 5th. 1991