31 ANALYST, JANUARY 1991, VOL. 116 - Simultaneous Determination of Silver and Copper by Flame Atomic Absorption Spectrometry With Alternate Irradiation by Two Hollow Cathode Lamps LS2 . Osamu Sakurada, Shunitz Tanaka and Mitsuhiko Taga" Department of Chemistry, Faculty of Science, Hokkaido University, Sapporo 060, Japan Tei ji Ka kiza ki Laboratory of Chemistry, Hokkaido University of Education at Iwamizawa, lwamizawa 068, Japan i PI0 ' CPU The simultaneous flame atomic absorption spectrometric determination of silver and copper is described. The resonance lines of both silver (328.07 nm) and copper (327.40 nm) were introduced within the same bandpass of the monochromator. Therefore, it was possible to measure the absorption signals of silver and copper simultaneously, when the silver and copper hollow cathode lamps were alternately irradiated. Acquisition of the absorption data was synchronized with the irradiations using a computer.For the introduction of the sample into the flame, a discrete nebulization method was investigated in order to minimize the sample volume required and the analysis time. Keywords: Simultaneous atomic absorption spectrometry; flame atomization; discrete nebulization; silver; copper N D Atomic absorption spectrometry (AAS) is an excellent analytical method in terms of selectivity and sensitivity. The selectivity of AAS is attributed to the use of an analyte-specific resonance line being emitted from the radiation source, a hollow cathode lamp (HCL). However, this indicates that for simultaneous multi-element determination by AAS, it is necessary to prepare multi-channel radiation sources and optical systems for each element.Simultaneous multi-element determination by AAS has been attempted by many workersl.2 and the instrumentation is now commercially available.3 Harnly et al. proposed a multi-element AAS system consisting of a continuum light source such as an Xe arc lamp, a high-resolution echelle polychromator and a computerized high-speed data system. Nakamura and Kubotas also reported an instrument for multi-element AAS consisting of a specific multi-element HCL, a single detection channel with one photomultiplier tube (PMT) and a time-divided high-speed data acquisition system. However, these instruments are very complicated and expensive to use for routine analyses. On the other hand, the spectral interferences that arise in AAS from the overlap of the absorption lines of the analytes consequently lead to large experimental errors .6 Therefore, a graphite furnace AAS method has been developed for the simple simultaneous determination of copper and silver by the use of their neighbouring resonance lines.' This method is based on the difference in the appearance temperatures of the analytes in the graphite furnace.Silver is atomized at a lower temperature than copper. The mixed radiation of the resonance lines of silver (328.07 nm) and copper (327.40 nm) is introduced into the graphite furnace atomizer simultaneously. By measurement of the peak heights of silver and copper on a chromatographic absorption-time profile, the simultaneous determination of silver and copper is achieved.In this work, the simultaneous determination of silver and copper by flame AAS was investigated using the neighbouring resonance lines. In flame AAS, the difference in the appear- ance temperature, used in the graphite furnace method, cannot be expected. Therefore, alternate irradiation of the sample with the silver and copper HCLs was attempted. By utilizing a computerized high-speed data acquisition system to * To whom correspondence should be addressed. collect the absorption signals alternately, the absorption signal for each element can be obtained with one measurement. A discrete nebulization methods for the introduction of the sample into the flame was also investigated in order to minimize the sample volume required and the analysis time.The proposed simultaneous method is simpler and less expensive than other methods. The method was applied to the determination of silver and copper in commercially available silver brazing filler metals used for welding. Experimental Apparatus A Hitachi Model 170-50 atomic absorption spectrometer was used, with a pre-mix burner for the air-acetylene flame. In order to introduce the mixed radiation of the resonance lines of silver and copper into the flame, the optical system for deuterium background correction was used as the secondary radiation source. As shown in Fig. 1, the silver and copper HCLs were placed in the positions of the HCL and the deuterium lamp (in the normal mode), respectively. The light beams from the silver and copper HCLs were pulsed at 50 Hz alternately, as shown in Fig.2. The light beams were made spatially coincident using the half-mirror and subsequently passed through the flame atomizer into the spectrometer. The HCL2 - 1 HCLl M PMT32 I HCL2 Cu HCL ANALYST, JANUARY 1991, VOL. 116 HCL1 . Ag HCL spectral bandpass of the monochromator was set at 1.1 nm to detect both the silver and copper resonance lines. The voltage analogues of the logarithmic converted circuit of the PMT (Hamamatsu R456) were converted into digital data by a 12 bit analogue to digital converter ( N D ) [Contec AD12-16A(98)]. Timing of the A/D cycle was obtained using software-controlled radiation pulses from the HCLs, with a parallel input-output module [Contec (PI0-48W( 98)].The vertical scale in Figs. 3-6 is shown in volts (V) and is comparable to absorbance. The peak area integrated the peak height with the time (s). A Model PC-9801UV2 personal computer (Nippon Electric) was used for data acquisition. Software to perform the data acquisition and analysis was written in BASIC and incorporated a Machine Code sub- routine for rapid data acquisition.9 The discrete nebulization method was used to introduce the sample solution into the flame. The device for this was assembled from a miniature polytetrafluoroethylene funnel connected to the nebulizer capillary. The sample solution was injected into the funnel and nebulized in the flame. The computer-controlled data acquisition was triggered by the detection of the sample passing through the nebulizer capillary by means of a photocoupler (Omron Model EE-SV3) registering the change in the light transmission. The absorbance-time signals were digitally stored in the memory of the computer.The stored time, namely, the time required for the sample volume, was pre-set, normally 4 s for a 100 p1 injection. The system for the detection of the injection was a modification of the system proposed by Goto et al.lO Micropipettes (Eppendorf 4700 and 4710) were used for sample injection. 1 Reagents Standard solutions of silver and copper were prepared by dissolving silver nitrate (analytical-reagent grade, Wako Pure Chemicals) in lmol dm-3 HN03 and by dissolving copper metal (99.999% pure, Mitsuwa Pure Chemicals) in concentrated nitric acid and diluting with water to a final concentration of 1 mol dm-3 HN03.Other reagents were of analytical-reagent grade. Doubly distilled water was used throughout. - Results and Discussion Flame Conditions For the simultaneous atomic absorption spectrometric determination of silver and copper, it is important to determine the optimum flame conditions for the two elements. In order to do this, response surface graphs11312 were constructed from the results of the atomic absorption spectrometric determination of silver and copper plotted as burner height versus acetylene pressure. The burner height values were indicated by the scale reading on the AAS apparatus. In the distribution pattern of silver and copper shown in Fig. 3, the maximum absorbance peaks are located at Off (1.1 4ms 4ms Fig.2 HCLs and ( b ) signal sampling (a) Timing chart of emission signals from the silver and copper a flame height of approximately 2.5 and an acetylene pressure of 0.25 kg cm-2. In subsequent experiments, the following experimental conditions were chosen as the optimum flame conditions; air flow pressure, 1.6 kg cm-2; acetylene flow pressure, 0.25 kg cm-2; and burner height, 2.5 (arbitrary units). Absorbance Signals of Copper and Silver Obtained by Alternate Irradiation of the Sample Silver and copper are simultaneously atomized and hence there is usually no time difference in the appearance of the signal, in conventional AAS. Therefore, the different absorp- tion signals cannot be distinguished when the mixed resonance lines of silver and copper pass through the flame at the same I I 6 3l L 0 a L .- Y 3 a L 0 0.2 0.3 0.4 0.5 0.6 0.7 P C ~ H J ~ S cm-* 2 0 0.2 0.3 0.4 0.5 0.6 0.7 P C 2 H J b cm-2 Fig.3 Response surfaces of silver and copper obtained by continu- ous nebulization. Sample solution of (a) silver ( 5 ppm); and (b) copper (5 ppm). Burner height value was indicated by the scale reading on the AAS apparatusANALYST, JANUARY 1991, VOL. 116 33 time. However, if the silver and copper HCLs alternately irradiate the sample and acquisition of the absorption data is synchronized with the irradiation, it is possible to distinguish between the absorption signals of silver and copper. The two HCLs are pulsed at 50 Hz alternately as shown in Fig. 2 and both radiations pass through the flame. By utilizing a computerized high-speed data acquisition system to collect both of the absorption signals alternately, each element could be determined with a single measurement.The absorbance- time profiles for silver and copper, obtained simultaneously using the instrument previously mentioned, are shown in Fig. 4(a). A sample solution of silver (5 ppm) and copper ( 5 ppm) *I 0 0 m Time Fig. 4 Absorption profiles of silver (top) and copper (bottom). (a) Continuous-flow nebulization method. (b) Discrete nebulization method. Sample injection volume was 100 pl. The numbers on the profiles refer to sample composition and correspond to the concentration of the metal in ppm z Y z 2 a 1 0 1 0 50 100 150 Injection vol u rne/pl Fig. 5 Effect of injection volume on peak height and peak area of (a) silver and ( b ) copper obtained with 5 ppm silver and 5 ppm copper mixed solutions in 0.1 mol dm-3 HN03 was injected by using a conventional method.The absorbance-time profiles obtained in the mixed solution were compatible with the signals obtained in each single element solution. Mutual interference due to the coexistence of silver and copper was not observed. Conse- quently, it is clear that the simultaneous determination of silver and copper can be performed with the alternate irradiation method. In conventional flame AAS, the continuous flow injection method is popular for sample injection. However, it has the disadvantage that a large sample volume is required. In order to minimize the sample volume required and the analysis time, a discrete nebulization method was investigated.The signal peak shape obtained with this method is shown in Fig. 4(b). Effect of Injection Volume The dependence of the peak height and the peak area on various injection volumes of the mixed solution, containing 5 ppm of Ag and 5 ppm of Cu, is shown in Fig. 5. The peak height increases with the volume of sample solution injected up to about 100 p1. Thereafter, the limit of the peak height is the same as the height obtained with continuous aspiration. 6 3 s E 0 r 0) .- Y m n 2 1 0 0 1 2 3 4 5 6 Concentration (ppm) Fig. 6 Calibration graphs for (a) silver and ( b ) copper obtained with a constant volume (100 pl) of mixed solutions of silver and copper Table 1 Results for the determination of silver and copper in silver brazing filler metals. Figures in parentheses represent the relative standard deviations (YO) of five analyses.The proposed method of simultaneous flame AAS used a 100 p1 injection volume; flame AAS was the conventional single-element flame method using continuous-flow nebulization. Reference values are as reported by the manufacturer Content (YO m/m) Proposed method Peak Peak Sample Element height area Low melt* Ag 43.9 44.1 (0.5) (2.4) Cu 15.6 15.2 (2.7) (1.8) SampleNo. 318t Ag 25.7 25.2 c u 34.4 33.7 (2.3) (1.3) (2.1) (2.1) * Obtained from Nippon Yushi. t Obtained from Kinzokuyouzai. Reference Flame value AAS (%m/m) 45.1 44-46 15.3 14-16 25.7 5 2 5 34.8 =35 (1.1) (2.5) (1.4) (2.5)34 ANALYST, JANUARY 1991, VOL. 116 The reproducibility of the signal is improved by increasing the injection volume.With an injection volume of greater than 100 pl, the relative standard deviation for ten determinations of peak height is less than 1%. The peak-area value shows a linear relationship to the injection volume. Calibration Graph The calibration graphs for silver and copper, as the absorbance-time profiles, are shown in Fig. 6. The peak height and peak area gave straight lines for silver and copper. As a constant injection volume was used, there was a linear relationship between the peak-area value and the concentration. Application The proposed simultaneous AAS method using discrete nebulization was applied to the determination of silver and copper in two types of silver brazing filler metal alloys that consisted of silver, copper, zinc and cadmium.Samples were dissolved in nitric acid. The results shown in Table 1 were obtained by measuring the peak height and peak area and are in agreement with the reference values reported by the manufacturers. In addition, the proposed method was vali- dated by conventional single-element flame AAS using continuous-flow nebulization. In applying the Student’s t-test to the two methods, there was no significant statistical difference in the results obtained using the two methods. The simultaneous determination of silver and copper by flame AAS was also performed using the neighbouring resonance line. The method requires only a simple modifi- cation to the normal AAS system. The proposed method can be employed in order to minimize both the sample volume and the analysis time. 1 2 3 4 5 6 7 8 9 10 11 12 References Busch, K. W., and Morrison, G. H., Anal. Chem.. 1973. 45, 712A. Harnly, J . M., Anal. Chem., 1986, 58, 933A. Okumoto, T., Tsukada, M., Tobe, H., Kitagawa, M., Yone- tani, A., and Sawakabu, H., The Hitachi Scientific Instrument News, 1987,30, 2787. Harnly. J. M., O’Haver, T. C., Golden, B., and Wolf, W. R., Anal. Chem., 1979, 51, 2007. Nakamura, S., and Kubota, M., Analyst, 1990, 115. 283. Vajda, F., Anal. Chim. Acta, 1981, 128, 31. Taga, M., Tanaka, S., and Sakurada, O., Bunseki Kagaku, 1989, 38,403. Cresser, M. S., Prog. Anal. At. Spectrosc., 1981, 4, 219. Kakizaki. T., Hikma, S., and Hasebe, K., Anal. Sci., 1989, 5, 781. Goto, K., Uchida. T., and Iida, C., Rev. Sci. Instrum., 1983,54, 291. Mossholder, N. V., Fassel, V. A., and Kniseley, R. N., Anal. Chem., 1973, 45, 1614. Fujiwara, K., Haraguchi, H., and Fuwa, K., Anal. Chem., 1975, 47, 743. Paper 0103357J Received July 24th, 1990 Accepted September 11 th, 1990