ANALYST OCTOBER 1986 VOL. 111 1139 Use of a Matrix Modifier and L‘vov Platform in the Determination of Copper in Pooled Human Saliva by Electrothermal Atomic Absorption Spectrometry lnes Game,* Leonard0 Balabanoff Rita Valdebenito and Luz Vivaldi Facultad de Ciencias Departamento de Quimica Universidad de Concepcion Casilla 3-C Concepcion, Chile The L’vov platform was used with a common matrix modifier HN03 and NH4N03 to eliminate matrix interferences present in saliva samples during the determination of copper by furnace atomic absorption spectrometry. The NaCl matrix effects were evaluated from a graph of the absorbances versus charring temperatures using a copper solution of 20 pg I-’ with NaCl added. The results obtained by atomisation of the standard either by tube or by L‘vov platform indicate that severe interferences are observed if the modifier is absent.Precision was improved by using a combination of platform matrix modifier and standard additions techniques. The analysis of ten samples of pooled saliva gave an average of 16.7 _+ 1.6 pg I-’ of Cu and a recovery of 99 k 1.6%. The sensitivity was increased from 0.1 ng (common tube) to <0.1 ng (L‘vov platform tube). Keywords Copper determination; pooled saliva; a tomic absorption spectrometry; electrothermal a tom isa -tion; L‘vov platform matrix modifier effects During the last decade worldwide interest in trace elements has stimulated a number of studies of their concentrations in biological fluids and human tissues in order to establish their natural concentrations and to detect illnesses occupational diseases and possible toxic effects.One of the most useful methods for the determination of these trace elements is atomic absorption spectrometry with electrothermal atomisation (ETA-AAS) However the appli-cation of ETA-AAS to the determination of trace elements in biological samples is complicated by the high content of sodium chloride. This compound interferes with the determi-nations producing broad band absorption and also lowers the absorption peak of the analyte. In order to reduce interference from sodium chloride matrix modification can be performed using matrix modifiers. The following reagents have been proposed for this purpose ammonium nitrate I sodium peroxide,2 ascorbic acid,’ carbon dioxide ,4 formic acid5 and thiourea .6 This study was initiated in order to improve the ETA-AAS method for the determination of trace amounts of copper in samples with a sodium chloride matrix.A sample of human saliva was chosen. This fluid contains up to 0.10% m/V of NaCl. In order to reduce or eliminate matrix interferences we propose the use of nitric acid and ammonium nitrate as matrix modifiers. We used a combination of a graphite tube and a L’vov furnace platform (which has been proposed for improv-ing the ETA-AAS methodcs) and compared the results with those obtained using the common graphite tube. Initially we analysed sodium chloride solutions spiked with copper (synthetic solutions) and then analysed samples of pooled human saliva for copper. Experimental A Perkin-Elmer Model 503 atomic absorption spectrometer, equipped with an HGA 2100 graphite furnace and a graphite tube with a L’vov platform was used.The light source was a copper element hollow-cathode lamp intensitron and sample solutions were injected into the graphite tube by an auto-sampler (Perkin-Elmer AS-1). * To whom correspondence should be addressed. Materials and Reagents The glass containers were cleaned by soaking for 48 h in a 1 + 1 mixture of 10% V/V H2S04 and 10% V/V HN03, followed by several rinses with doubly distilled water. Drying was carried out at room temperature in a plastic-covered hood. Flasks and polyethylene containers were filled with 50% V/V HCl for 48 h. All reagents were of Merck Suprapur quality and high-purity distilled water was used for the preparation of all solutions.The furnace was purged with argon (purity 99.998% by volume). Standard solutions A 100 mg 1-1 copper standard solution was diluted such that 20 1.11 contained 0.4 0.8 1.2 and 2.0 ng of copper in each of four solutions. Analysis of Synthetic Solutions Working solutions were prepared according to the compo-sition summarised in Table 1. The synthetic solutions were used in the study of the maximum charring temperature with the L’vov platform. The results obtained are plotted in Fig. 1. Table 1. Composition of synthetic samples used for the study of absorbance vs. maximum charring temperature. Each synthetic sample always contains 0.4 ng of Cu (20 pl of 20 pg 1-’ solution). The concentrations of NaCl HN03 and NH4N03 are 1.5,189 (3 M) and 40 g 1-1 respectively Composition of the matrix * Symbols as used in Fig.1 1140 ANALYST OCTOBER 1986 VOL. 111 0.040 I I 500 1000 C h a r r i ng tern pe rat u re/"C Fig. 1. Absorbance versus charring temperature using L'vov plat-form and synthetic solutions. Symbols and matrix composition as in 500 1000 Charring temperaturePC Fig. 2. Effect of charrin temperature on absorbance for saliva pool sample with (A) HNO, ?B) HN03 + NH4N03 (L'vov platform) and (C) HNO + NH4N03 (common tube) Determination of Copper in Saliva Saliva collection method Samples of saliva were obtained in our laboratory from voluntary donors by a spitting method.9 No stimulation was used. The collected samples were mixed and homogenised by stirring and stored at 4 "C in polyethylene containers.Sample treatment and procedure Slow but substantial precipitation of protein occurred during storage at 4 "C. Consequently the total pooled sample is again homogenised by shaking for 5 min shortly before removing aliquots for analysis. Treat 5 ml of saliva* with 2 ml of concentrated nitric acid. Add 1 ml of ammonium nitrate (40 g 1-I) digest for 3 h at 80 "C and dilute to 10 ml with doubly distilled water. Using the autosampler inject 20-p1 aliquots of the sample solution ( a ) into a graphite L'vov platform tube; and ( b ) into a graphite tube without a platform. Repeat the injection five4imes for each selected charring temperature. The results are plotted in Fig. 2. * This work is part of a project that includes the analysis of saliva for a series of elements i e .Ca Mg Na and K by flame AAS P043- by spectrophotometric analysis with flow injection F- with an ion-selective electrode and Cu Mn and Zn with an HGA graphite furnace. Hence we use 5-ml aliquots of saliva because we try to analyse for all the above elements in the same (pooled) sample. If one were to analyse only for Cu the saliva sample volume could be reduced. Table 2. Instrument conditions used for the determination of copper in pooled saliva samples. Instrument parameters wavelength 325.4 nm; spectral band width 0.7 nm; lamp current 20 mA; argon flow-rate 20 ml min-1 Procedure Temperature/"C Time/s Charring . . . . . . . . 500 30-40* Drying . . . . . . . . 100-200* 40-SO* Atomisation .. . . . . 2700 8-8 * * L'vov platform. ~ Table 3. Determination of copper in pooled saliva Common tube L'vov platform tube Modifier Cu k s.d. R* k s.d. Cu f s.d. R* +_ s.d., p.p.b. Y O p.p.b. YO ACT 13.8 5 1.1 84.6 k 2.8 ACT 14.3 f 1.6 91.0 k 1.6 HN03 CCfl0.0 k 0.1 CCS14.6 I! 0.1 ACT 13.7 k 1.0 84.5 k 4.4 ACT 16.7 k 1.6 99.0 k 1.6 HN03 + NH4N0, CCf 8.2 _t 3.6 CCf 15.2 k 0.1 * R% Recovery. Each value is the mean of five determinations t AC Standard addition mean value for 10 samples of pooled f CC Calibration graph mean value for 10 samples of pooled with two additions. saliva. saliva. Procedure for measurements A temperature of 500°C was chosen for the ashing stage. Copper was determined in 5-ml aliquots of saliva that had received two standard additions of 0.4 and 0.8 ng per 20 pl and that had been treated with concentrated nitric acid and ammonium nitrate as described above.The blanks were determined by the standard additions method.10 The temperature and time of drying and atomisation of the saliva sample were determined in the usual way." The instrument parameters and conditions are given in Table 2 and the analytical results in Table 3. The recoveries of the two copper additions of 0.4 and 0.8 ng to 2 0 4 aliquots of sample solution are also given in Table 3. Results and Discussion The experimental results presented in Fig. 2 allowed us to choose 500°C as an adequate charring temperature. If the temperature is approximately 1000 "C there is diminished absorbance probably owing to the loss of part of the free atomic population of copper in the vaporisation stage.A temperature of 2700 "C was chosen for atomisation. From Fig. 1 it may be concluded that the absorption signal of the standard copper solutions with sodium chloride is lower than that of the standard copper solution without sodium chloride at different charring temperatures even when using the L'vov platform. A loss of sensitivity is observed if nitric acid is added to a copper solution that contains sodium chloride. Nevertheless the nitric acid is required to destroy the organic matter present in the saliva.12 Fig. 2 showing the maximum absorbance vs. charring temperature for saliva with nitric acid and with nitric acid and ammonium nitrate, respectively under L'vov platform analysis leads to the conclusion that the ammonium salt greatly increases the sensitivity.From Table 3 showing the results for copper in saliva it may be concluded that a sample treated with acid shows a low recovery of this element even though the L'vov platform is used. This conclusion is consistent with the graphs in Fig 2 ANALYST OCTOBER 1986 VOL. 111 The mean value found for copper with standard additions and nitric acid and ammonium nitrate treatment in a graphite tube is 13.7 k 1.0 p.p.b. with an 84.5 k 4.4% recovery. With the L’vov platform the mean value of copper rises to 16.7 k 1.6 p.p.b. and an optimum recovery of 99.0 k 1.6%. The results for the determination of copper in saliva derived from the calibration graph are lower than those obtained using the standard additions method.The combination of matrix modifier L’vov platform and standard additions allows one to obtain reproducible results. This is an indication that the L’vov platform by itself does not eliminate the sodium chloride interference. The absolute sensitivity of the method was calculated as the amount of copper in a sample that produced a 1% absorption or 0.0044 absorbance unit. This was found to be <O. 1 ng with the L’vov platform tube and 0.1 ng with the graphite tube. For artifical standarised samples of copper sodium chloride and nitric acid Figs. 1 and 2 show that treatment with ammonium nitrate substantially improves the sensitivity. Finally it can be concluded that for the determination of copper in human saliva it is convenient to add ammonium nitrate with nitric acid to digest the sample and to analyse for copper in a graphite tube with a L’vov platform using the method of standard additions.This research is part of a project (20.13.17) financed by the Direccion de Investigacion of the University of Concepcion, Chile. 1141 The authors thank Prof. P. C. Minning and Prof. M. Zamudio for their critical reading of the manuscipt. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Ediger R. Peterson G . and Kerber J. At. Absorpt. Newsl., 1974 13 61. Churella D. J. and Copeland T. R. Anal. Chem. 1978,50, 309. Hides D. Anal. Chem. 1980 52 959. Tominaga M. and Umezaki Y . Anal. Chim. Acta 1983,148, 285. Susuki M. Ohta K. and Yamakita T. Anal. Chem. 1981, 53,9. Bezur L. Marshall Y. Ottaway J. M. and Fakhrul-Aldeen, R. Analyst 1983 108 553. Hindeberger E. J. Kaiser M. L. and Koirtyohann S. R. At. Spectrosc. 1981 2 1. Harezov I. and Ivanova E. Fresenius 2. Anal. Chem. 1983, 315 26. Navazesh M. Christensen M. J . Dent. Res. 1982,61 1158. Camrnan K. Fresenius 2. Anal. Chem. 1982 312 515. “Analytical Methods for Atomic Absorption Spectroscopy Using the HGA Graphite Furnace,” Perkin-Elmer Norwalk, CT 1977 p. 4-1. Game I . Balabanoff L. Valdebenito R . and Vivaldi L., Bol. Soc. Chil. Quim. 1982 27 340. Paper A51455 Received December 18th 1985 Accepted April 3rd 198