Materials Transactions, Vol. 51, No. 12 (2010) pp. 2215 to 2219 #2010 The Japan Institute of Metals Degradation Process of Graphite Furnace Estimated from the Atomic Gas Temperature of Iron in Graphite Furnace Atomic Absorption Spectrometry Tetsuya Ashino 1; *, Haruki Shimabukuro 2 and Kazuaki Wagatsuma 1 1 Institute for Materials Research, Tohoku University, Sendai 980-8577, Japan 2 Ryukyu-Cement Co. Ltd., Urasoe 901-2023, Japan GF-AAS has been employed for determining several trace impurities in metallic materials. The analytical precision in the continuous analysis tends to worsen due to thermal degradation of the graphite furnace at the atomization stage. Changes of the atomization condition along with increasing the number of measurements were investigated by monitoring the temporal variation in the atomic gas temperature of iron. In a working solution containing low-concentration of acid, the inner wall of the graphite furnace was gradually damaged and then the efficiency for the atomizing process was reduced. In a working solution containing sub-mol/l order of sulfuric acid, thinning or cracking of the graphite furnace progressed more rapidly and the function of it was greatly deteriorated. [doi:10.2320/matertrans.m2010180] (Received May 21, 2010; Accepted September 6, 2010; Published October 20, 2010) Keywords: graphite furnace atomic absorption spectrometry, atomization furnace, degradation process, atomic gas temperature of iron, temporal variation 1. Introduction Electrothermal atomic absorption spectrometry (ET- AAS) 1) is extensively employed for the determination of trace amounts of metallic elements, because several elements, such as Bi, 2) Cd, 3) Pb, 2,4) Se, 5,6) Sn 7) and Sb, 8) can be detected with better sensitivity compared to the other analytical methods. These metals, in some cases, cause a serious degradation of the quality of manufactured metallic materials. 9,10) Furthermore, metallic elements such as Cd and Pb have a harmful effect against the environment around the human beings. 11,12) Accordingly, trace analysis of these elements has been required and ET-AAS can be thus employed as an effective analytical method not only for environmental assessment but for industrial and medical uses. Regarding the analytical procedure in ET-AAS, a sample solution is injected into an atomization furnace. Then, this furnace, which is commonly made of graphite (graphite furnace, GF), 1,13 17) is electro-thermally heated with a suitable temperature program for each analyte element to be pre-determined for generating an atomic gas to measure the atomic absorption. In continuous analysis by GF-AAS, the analytical precision becomes worsening due to thermal degradation of the atomization furnace; as the result, it is generally lower compared with the other analysis method, such as inductively coupled plasma-optical emission spectrometry (ICP-OES) or inductively coupled plasma-mass spectrometry (ICP-MS). The authors consider that the influence of changes of the graphite furnace on the atomization process should be investigated to understand the reason for this phenomenon. During heating and cooling of the graphite furnace repeatedly, it is difficult to keep a good and stable condition of the graphite furnace for precise analysis and its function as the atomizer deteriorates as the number of atomizing increases. In order to estimate a temperature of atoms in gaseous phase (atomic gas temperature) in a graphite furnace directly, *Corresponding author, E-mail: ayustet@imr.tohoku.ac.jp the authors have reported a monitoring system by using a two-line method. 18 20) In the developed method, iron is selected as a probe element. The atomic gas temperature of iron is observed by using two iron atomic lines assigned to the 3d 6 4s-3d 5 4s4p transition. Each atomic absorbance is monitored and then converted into a characteristic temperature under the assumption of the Bolzmann distribution. In the present work, the authors investigated effects of the number of measurements on both the lifetime of the furnace and the atomization condition in the repeated operation in GF-AAS. Evaluating the atomic gas temperature in the atomization process yields useful information on the state of the graphite furnace and thus the evaluation procedure of analyte species. 2. Experimental 2.1 Experimental apparatus A Z-9000 (Hitachi Corp., Japan) simultaneous multielement atomic absorption spectrometer with an autosampler system was employed. As the light source, a hollow-cathode lamp (Hamamatsu Photonics Corp., Japan) was used. Table 1 lists the operating conditions for the measurement by the Z-9000. The graphite furnace employed was a graphite tube without a platform. The method to control the temperature of the furnace was based on measurements with a radiative thermometer installed in the instrument, by monitoring the radiation from the furnace; therefore, the temperature of the furnace wall was directly estimated. Two hollow cathode lamps were installed to monitor two different wavelengths of iron atomic lines simultaneously. The absorbance of iron was then measured under the conditions listed in Table 1. The absorbance values for each spectral line were recorded on a personal computer through an AD converter at an interval of 0.02 s. Finally, the temporal variations in the gas temperature were directly estimated. In the employed graphite furnace, its electrical resistance may be different individually or changed as the number of
2216 T. Ashino, H. Shimabukuro and K. Wagatsuma Table 1 Instrument and operating condition. Instrument Hitachi Z-9000 multi-element atomic absorption spectrometer Background correction dc-polarized Zeeman effect Light source Hollow cathode lamp Lamp current (ma) 5.0 Wavelength (nm) 372.0 and 373.7 Ar gas flow rate (Lmin 1 ) 0.2 (without atomizing stage) Injection volume (ml) 20 Temperature, T/K Time, t/s Drying 353 423 30 Charring-1 423 1373 30 Charring-2 1373 30 Atomizing 2573, 2973 or 3073 10 Cleaning 3273 10 Table 2 Concentration of iron (III), c/mgl 1 Sample solutions employed in the present work. Acidity Concentration of acid, c/moll 1 Atomizing temperature, T/K 1 1.2 HNO 3 0.001 3073 2 1.2 H 2 SO 4 0.2 2973 3 1.2 H 2 SO 4 0.2 2573 analyzing increases. In the employed apparatus, it is difficult to monitor a variation in the resistance of the furnace with a heating cycle. However, the temperature of the graphite furnace is controlled by monitoring not the electrical current through the furnace but the radiation from the furnace wall. It is thus considered that the graphite furnace is reproducibly heated under the same process of heating even when different furnaces are individually employed or the condition of the furnace is gradually changed with the heating cycle. 2.2 Reagents and procedure For preparing an iron stock solution (1.0 g/l), 0.10 g of metallic iron (Mairon HP, Toho Zinc Co. Ltd., Japan) was dissolved in 10 ml of 7 mol/l of nitric acid and exactly diluted to 100 ml with water. Table 2 lists iron sample solutions for measuring the atomic gas temperature of iron in the present work. The concentration range of iron, having a linear correlation between the concentration and the absorbance, was confirmed. The concentration of the sample solution in the present work was determined to be the highest in this range. 18,19) Working solutions were each prepared by diluting from the stock solution before use. The nitric acid and the sulfuric acid employed were an electronics industry grade (Wako Pure Chemical Industries Ltd., Japan), and deionized water (Yamato Scientific Co. Ltd., Japan) was used for all preparations of the sample solutions. The experiment procedure is as follows: An as-received graphite furnace was installed in the ET-AAS instrument. Then, the sample solution listed in Table 2 was injected into the graphite furnace and the temporal variation in the atomic gas temperature of iron was monitored. The procedure described above was repeatedly conducted until the graphite furnace was destroyed. In the 0.2 mol/l of H 2 SO 4 solution, continuous monitoring of the gas temperature was once stopped at 200th measurement and was restarted after leaving for 12 h. 2.3 Two-line method The principle of this method has already been described elsewhere. 18,20) It is based on the difference in the number density between two energy levels, which can be determined by a characteristic temperature under thermodynamic equilibrium. A graphite furnace is considered to be in local thermodynamic equilibrium (LTE), because the kinetic energy of particles in the furnace atmosphere can be averaged through their collisions occurring in the ambient pressure condition actively. A general formula for the two-line method is as follows: lnðr 1 =R 2 ÞþlnðF 2 =F 1 Þ¼ðE 2 E 1 Þ=kT ð1þ where k and E are the Boltzmann constant and the excitation energy of each energy level. T is an equilibrium temperature which should coincide with the atomic gas temperature in the graphite furnace under LTE condition. The intensity ratio of the incident and absorbed radiation, R ¼ I abs =I 0, where I abs and I 0 mean the intensity of the light source after and before an absorption, is measured for a spectral line related to each energy level. Then, a term expressed as F ¼ 2 ga, where, g and A respectively mean the wavelength, the statical weight and the Einstein coefficient for the spontaneous emission of a spectral line, is calculated using literature data for each analytical line. Therefore, the atomic gas temperature in the graphite furnace can be estimated from the absorbance ratio between two spectral lines whose lower energy levels are different. Moreover, the variations in the atomic gas temperature can be determined not only by the temperature distribution, but by the spatial distribution of iron atoms through their diffusion in the atomization stage. On the other word, it reflects that the atomic gas generates at a center (high-temperature area) and moves to an edge (low-temperature area) of the furnace. 3. Results and Discussions 3.1 Variation in atomic absorbance and maximum atomic gas temperature in continuous analysis by GF-AAS It is commonly recognized that the conditions of the graphite furnace deteriorate as the number of atomizing increases. In order to evaluate the change of the furnace in continuous operation in GF-AAS, the authors investigated the variation in the atomic absorbance and the maximum atomic gas temperature of iron in the atomization stage. Figure 1 shows the relationship between the number of measurements and the atomic absorbance of an iron line in the solutions listed in Table 2. In the 0.001 mol/l of HNO 3 solution, the atomic absorbance observed was almost constant from the beginning of measurement to 460th measurement. After then, it was fallen and the graphite furnace was destroyed by 475-fold atomizing. However, the variance of the atomic absorbance in the 0.2 mol/l of H 2 SO 4
Degradation Process of Graphite Furnace Estimated from the Atomic Gas Temperature of Iron 2217 Fig. 1 Relationship between number of measurements and atomic absorbance of iron. Fig. 3 Comparison of variation in atomic gas temperature of iron in several sample solution. Fig. 2 Relationship between number of measurements and maximum atomic gas temperature of iron in atomizing stage. solution was larger compared with that in the 0.001 mol/l of HNO 3 solution and the graphite furnace was destroyed at 236-fold atomizing when the atomizing temperature was 2973 K or at 290-fold atomizing when it was 2573 K. Figure 2 shows the relationship between the number of measurements and the maximum atomic gas temperature at the atomization stage. In both solutions, the atomic gas temperature was observed to be 2000 K in measurements just after the beginning even when employing different atomizing temperatures were pre-set in the instrument, and it was gradually elevated as the number of measurements increased up to 70. However, the variations in the atomic gas temperature were different between both solutions or among the atomizing temperatures. In the 0.001 mol/l of HNO 3 solution, the atomic gas temperature was almost constant from 70th to 240th measurements. Next, it was again elevated up to 280th atomizing and almost constant between 280th and 370th measurements. After then, the atomic gas temperature was fallen until destruction of the graphite furnace. On the other hand, the atomic gas temperature in the 0.2 mol/l of H 2 SO 4 solution was largely changed during repetition of the atomization and its variance is larger compared with that in the 0.001 mol/l of HNO 3 solution. This phenomenon was especially predominant when the atomizing temperature was 2973 K. Moreover, the discontinuity in the atomic gas temperatures was observed before and after 200th measurements because the measurement was stopped for 12 h. The reason for these phenomena is considered to be the following. In the beginning of the continuous measurement, the inner wall of the graphite furnace becomes better condition for generating the atomic gas to improve the efficiency of the atomizing process while the number of measurement is up to 70. In the 0.001 mol/l of HNO 3 solution, the graphite furnace seems to be moderately changed due to the low-concentration of acid. In the 0.2 mol/l of H 2 SO 4 solution, the inner wall of the graphite furnace would be damaged by sulfuric acid containing in the solution and its lifetime is shorter compared with that of the 0.001 mol/l of HNO 3 solution. Therefore, the performance of the graphite furnace greatly deteriorates during heating and cooling repeatedly, and the degradation occurs more rapidly when the atomizing temperature pre-set in the instrument is higher. 3.2 Temporal variation in atomic gas temperature in continuous analysis by GF-AAS In order to evaluate the state of the atomic gas in the repeated atomizing operation, the authors monitored temporal variations in the atomic gas temperature of iron at the atomizing stage. Figure 3 shows variations in the atomic gas temperature for each solution listed in Table 2 during the atomizing time between 0 and 2 s. In the 0.2 mol/l of H 2 SO 4 solution, the atomic gas temperature reached the maximum more rapidly compared with that in 0.001 mol/l of HNO 3 solution. It is considered that a chemical form of charred species and its atomizing rate are different between the sample solutions having different compositions of acids. Figure 4 shows temporal variations in the atomic gas temperature at several different numbers of measurements in the continuous measurement. In both solutions, the generation of the atomic gas started earlier and the gas temperature reached the maximum more rapidly, as the number of measurements increased.
2218 T. Ashino, H. Shimabukuro and K. Wagatsuma Fig. 4 Temporal variation in atomic gas temperature of iron at several different numbers of measurements. (1) 0.001 mol/l HNO 3, 3073 K; (2) 0.2 mol/l H 2 SO 4, 2973 K; (3) 0.2 mol/l H 2 SO 4, 2573 K. Fig. 5 Temporal variation in atomic gas temperature of iron at different numbers of measurement just before distribution of graphite furnace. (1) 0.001 mol/l HNO 3, 3073 K; (2) 0.2 mol/l H 2 SO 4, 2973 K; (3) 0.2 mol/l H 2 SO 4, 2573 K. Figure 5 shows temporal variations in the atomic gas temperature at several atomization stages just before destroying the graphite furnace. In the 0.001 mol/l of HNO 3 solution, the gas temperature became lower than the programmed atomizing temperature and their large variations were observed. On the other hand, the rising rate of the gas temperature decreased and the maximum atomic gas temperature was fallen as the number of measurements increased in the 0.2 mol/l of H 2 SO 4 solution. Moreover, the atomic gas temperature was less elevated at the last measurement just before the destruction. The reason for these phenomena is considered to be the following. Along with increasing the number of atomizing, the furnace wall is eroded by heating and cooling repeatedly and thus the thickness becomes smaller. Because the wall temperature is elevated more rapidly due to the decreased wall thickness, the atomizing of analytes by conductive heat from the wall could be caused more quickly. However, differences in the deterioration process of the graphite furnace were observed between both solutions. In the 0.001 mol/l of HNO 3 solution, the inner wall of the graphite furnace was gradually changed not to be suitable for generating the atomic gas with a good efficiency for the atomizing process as the number of measurements increase before destruction. In the 0.2 mol/l of H 2 SO 4 solution, the graphite furnace was suddenly destroyed because thinning or cracking of the graphite furnace would progress more rapidly. These effects would be derived from a difference in the reactivity of the acids employed.
Degradation Process of Graphite Furnace Estimated from the Atomic Gas Temperature of Iron 2219 4. Conclusion The effects of the number of atomizing in repeatedly operation of the graphite furnace were investigated by monitoring the temporal variation in the atomic gas temperature of iron. When an as-received cuvette begins being heated, the wall of the graphite furnace was fast turned to be a better condition for generating the atomic gas with better efficiency for the atomizing process as the number of measurements increased. However, the state of the graphite furnace deteriorated as the number of atomizing further increased. In the solution containing low-concentration of acid, the inner wall of the graphite furnace was gradually damaged and then the efficiency of atomization was worse. In the solution containing sub-mol/l order of sulfuric acid, thinning or cracking of the graphite furnace progressed more rapidly and the function of it greatly deteriorated. For sensitively and precisely analyzing elements by GF- AAS, several numbers of atomizing for pre-measurement are needed before analyzing actual samples. For keeping better and stable conditions and long lifetime of the graphite furnace, the concentration of acids in the sample solution should be minimized and also the atomizing temperature should be pre-set as low as possible. REFERENCES 1) H. Massman: Spectrochim. Acta 23B (1968) 215 226. 2) O. Kujirai, T. Kobayashi and E. Sudo: Trans. JIM 18 (1977) 841 845. 3) T. Ashino and K. Hirokawa: Anal. Sci. 11 (1995) 703 706. 4) S. Hasegawa, T. Kobayashi and R. Hasegawa: Mater. Trans., JIM 41 (2000) 841 845. 5) T. Ashino, K. Takada and K. Hirokawa: Anal. Chim. Acta 297 (1994) 443 451. 6) T. Ashino and K. Takada: Anal. Chim. Acta 312 (1995) 157 163. 7) K. Takada, T. Ashino, T. Itagaki, Y. Morimoto, K. Wagatsuma and K. Abiko: Mater. Trans. 43 (2002) 105 110. 8) T. Ashino and K. Takada: J. Anal. Atom. Spectrom. 11 (1996) 577 583. 9) A. Toi, J. Sato and T. Kanero: Tetsu-To-Hagané 83 (1997) 850 855. 10) A. Toi and J. Sato: Tetsu-To-Hagané 84 (1998) 534 539. 11) S. Kuyacak and M. Sahoo: Can. Metall. Quart. 35 (1996) 478 484. 12) X. Chan, A. Hu, M. Li and D. Mao: J. Alloy. Compd. 460 (2008) 478 484. 13) B. V. L vov: Spectrochim. Acta 24B (1969) 53 56. 14) K. Kuga and K. Tsujii: Bunseki Kagaku 27 (1978) 441 443. 15) B. V. L vov: Spectrochim. Acta 33B (1978) 153 193. 16) H. Koizumi, K. Yasuda and K. Katayama: Anal. Chem. 49 (1977) 1106 1112. 17) K. Takada: Talanta 32 (1985) 921 925. 18) H. Shimabukuro, T. Ashino and K. Wagatsuma: Anal. Sci. 24 (2008) 1165 1170. 19) T. Ashino, H. Shimabukuro, S. Morimoto and K. Wagatsuma: Anal. Sci. 25 (2009) 1261 1264. 20) S. Morimoto, T. Ashino and K. Wagatsuma: Anal. Sci. 26 (2010) 809 813.