Học viên Trần Ngọc Thanh MỤc lụC

All atomic absorption measurements were performed via AAnalyst 600 (Perkin Elmer, Norwalk, CT USA) including dual wavelength spectrometer integrated with longitudinal Zeeman-effect background correction, graphite furnace with cooling system, autosampler, tin hollow cathode lamp. WinAA software was used for instrument controlling, data handle and processing. The performance of instrument was daily checked as suggestion of the manufactory guideline. In brief, GF-AAS was daily checked by analysis of copper solution in nitric acid. The peak are of copper have been met critical specifications. Data evaluation was carryout through either dedicated software or Microsoft excel and Microcal Original version 7.0. Instrumental operating parameters of AAS apparatus are showed in Table 1. For occupational sampling, low volume air sampling system model LV20P (Shibata, Japan) was used. Pre-weighted cellulose ester membrane filter (37 mm i.d. and 0.8 µm pore size, Millipore, Singapore) was used for collection air-bone particle samples.

All volumetric plastic and glassware were calibrated before using according to ISO guideline. All laboratoryware were decontaminated by immersion in 5% HNO₃ overnight and rinsed with tap water. Afterwards, the materials was rinsed with deionized water and then immersed in 10% HNO₃ at least 48h, copiously rinsed with ultrapure water; collected from a Milli-Q system (Millipore, Bedford, USA) and dried at 60°C before use, avoiding any contact with dust or metallic surface. All laboratoryware were checked by laboratoryware-blank measurement.

All the reagents used were of analytical reagent grade. A tin standard solution in nitric acid was purchased from Merck (Singapore). Aqueous working solutions were daily prepared by dilution of stock solution in 0.2 % nitric acid solution. All steps of dilution of standards and samples have been carried in the laminar fume hood in order to avoid contamination from environment.

Table 1. Instrumental operating parameters of GF-AAS apparatus for tin analysis

Operating conditions
Element	Sn
Wavelength	286.3 nm
Slit	0.7 nm
HCL current	24 mA (80% Imax)
Background correction	Zeeman effect based (longitudinal)
Measurement mode	Integrated absorbance (peak area)
Graphite furnace operation
Graphite furnace	THGA graphite tubes with end caps and integrated platform (Perkin-Elmer number)
Sheath and purge gas	Argon (99.999%)
Sample volume	20 µL samples (standards)
Chemical modifier	3 µg Mg(NO3)2 + 5 µg Pd(NO3)2
Dry temperature	110 - 130 °C with two stages (60 seconds in total)
Pyrolysis temperature	1000 °C (for 30 seconds)
Atomization temperature	2200 °C (for 4 seconds)
Cleaning temperature	2450 °C (for 3 seconds)

2.4 Sampling and storage

Samples have been collected from welding working zones according to occupational sampling guideline. In general, sample was collected form working sites via high volume air samples utilizing cellulose ester membrance filter (37 mm i.d. and 0.8 µm pore size) with sampling flow-rate according to guideline for occupational sampling. The air-bone particle samples on the pre-weighted cellulose ester membrance filters were kept on the desiccator. The weight of particles in occupation samples was measured by difference between pre-weighted filter before and after sample collection.

Air-bone particle samples were treated by wet digestion. In brief, air-bone particle samples was transferred into Kjeldahl flask, acidified by concentrated nitric acid. The samples was heated on the sand bath until the solution inside the Kjeldahl flask turned to clear. The samples was cooled to room temperature, and transferred to 25 mL volumetric flask and filled until mark by dilute nitric acid. Sample solutions was analysed as rapid as possible or stored at -4 °C for maximum three days as mentioned in the stability experiment part. Concentration of tin in the real samples was analysed via matrix-matched external calibration curve.

3. 
   Results and discussion
   

From analytical spectrochemistry point of view, tin is hard element for GF-AAS measurement. In addition, low concentration level of tin in the real samples is required high sensitivity and selectivity method. The use of chemical modifiers becomes an essential part of graphite furnace atomic absorption spectrometry with the introduction of the stabilized temperature platform furnace. Chemical modifier plays an important role in GF-AAS measurement. They act as increasing pyrolysis and atomization temperatures therefore sample matrix will be eliminated. The right selection of chemical modifier is not only improvement of selectivity also enhance of sensitivity. There are two kinds of chemical modifiers: permanent and temporary modifier. For tin analysis, the temporary chemical modifier is the most popular because of easy handle. In addition, the repeatability of measurement utilizing with temporary modifier is better than that using permanent modifier. The most chemical modifiers for tin analysis are Mg(NO₃)₂, Pd(NO₃)₂, (NH₄)₂HPO₄ or mixture of these compounds. In this study, Mg(NO₃)₂, Pd(NO₃)₂ and mixture of both reagents were used as temporary chemical modifiers. The mass ratio of single and mixture of chemical modifiers to analyte have been investigated (data not shown). The integrated absorbance and peak shape of atomization steps have been used for assessment. We observed that the mixture of 5 µg Pd(NO₃)₂ and 3 µg Mg(NO₃)₂ was worked as the best chemical modifier regarding to both intensity and peak shape. Indeed, the integrated absorbance and peak shape are also depended on the pyrolysis and atomization temperatures and sample matrices as well. Therefore this mixture was selected for the further investigated experiment and for real sample analysis.

Pyrolysis temperature and atomization temperature are critical parameters in the atomic absorbance measurement. Both temperature should be optimized for a given element and sample matrix also in order to achieve the highest sensitivity. These temperatures are also depended on the kinds of chemical modifiers.

For optimization of pyrolysis and atomization temperature, the studies was performed by varying of parameter investigated while all other parameters were fixed as values showed on the Table 1. The lower pyrolysis temperature and atomization temperature, the lower sensitivity and selectivity. In contrast, higher pyrolysis temperature and atomization temperature, the higher intensity and selectivity. However, the higher temperature may be losing sample during pyrolysis and atomization steps. As a result, the lower sensitivity should be achieved. The pyrolysis and atomization temperature curves were showed in the Figure 1. As can be seen in the Figure 1, the optimum temperatures for pyrolysis and atomization were 1000 °C and 2200 °C, respectively. These temperatures were the best compromise between maximum intensity and minimum background signal. In addition, peak profile was resolved within only 3 seconds. These temperatures have been checked by employing with spiked matrix- matched solutions. In addition, peak profile of atomization process has also been investigated as an alternative parameter for both temperature selection.

In order to evaluate the developed method in term of sensitivity, linearity and limit of detection, six standard solutions with concentration from 10 to 200 ng mL^-1 were measured in triplicate. Calibration function was calculated by linear regression of integrated absorbance as a function of concentration of tin. Regression equation and correlation coefficient square were Abs = 0.000757*C_Sn(ng mL^-1) - 0.00169 and 0.999, respectively. Linearity of measurement has been achieved in two order of magnitude of tin concentration. Systematic error was checked through comparison with another linear regression equation that was fit through zero value. It was no significant difference between two regression equations (data not shown). Limit of detection (LOD) and limit of quantification (LOQ) were calculated as the concentration corresponding to the three- and ten-times standard deviation of blank measurements (peak height) according to the guidance for industry bioanalytical methods validation [1]. LOD and LOQ of the developed method were 4.35 and 14.53 ng mL^-1 corresponding to 0.92 and 3.04 µg m^-3 for occupational samples, respectively. The LOD and LOQ of the developed method were far from the maximum level of inorganic tin in occupational samples according to USA-OSHA and WHO.

For the intra-day and inter-day stability test, five 25 ng mL-1 tin solutions were freshly prepared in dilute nitric acid. These solutions were rapidly analysed independent after preparation. The standard solutions then was stored at 4°C and frequently measured after 24h, 48h and 72h. Standard deviation of inter-day and intraday were 0.90 % and 0.65% (data not shown), respectively. It could be concluded that the developed method is highly repeatability and stability over three days. In addition, air bone samples can be stored at 4°C for three days after sample preparation.

According to Bulska et al. measurement of tin via GF-AAS is mainly interfered by chloride and sulphate in the matrix samples[18], [19]. However, concentration level of chloride and sulphate in the occupational samples that collected from welding working site is quite low. Therefore, we have focused on the investigation of interference of heavy metals on measurement of tin via GF-AAS. We observed that the analytical signal of tin is no significantly difference between without and present of other heavy metals at fifty times concentration levels than that of tin (data not shown). For comparison, concentration of other heavy metals in the real samples was semi-quantitative via ICP-QMS.

Air-bone samples were collected and treated as described in the Experiment section. Quantification of tin in the samples was carried out via matrix matched external calibration curve and the obtained concentration are listed in Table 2.

Table 2. Concentration ± standard deviation of tin in the air-bone samples analysed by our GF-AAS. Samples are coded as company names and sampling location

According to the guide for validation of bioanalytical methods, the method accuracy can be assessed by either using certificated reference materials or inter-comparison with other independent methods. Due to the fact that the lack of certificated reference materials for occupational samples, the validation of the developed method have been performed by comparison with inductively coupled plasma quadrupole mass spectrometry (ICP-MS), that method is required by USA-OSHA. ICP-MS is used as ‘golden method’ for metals analysis, in general. This method also is used as recommended method for inter-laboratory comparison[9].

Figure 2. Data comparison of concentration of tin in several air-bone particle samples via ICP-QMS (yellow) and our developed GF-AAS (green) measurements

3. 
   Conclusion
   

Ngoc-Thanh Tran is kindly acknowledged for financial support from Vietnam General Confederation of Labor via project No 502/QĐ-TLĐ

[1] E. a. Ostrakhovitch, Chapter 56, Tin in Handbook on the Toxicology of Metals. 2015.

[2] M. E. Weeks, J. J. Chisholm, S. Francisco, D. P. H. Laxen, and T. C. Hutchinson, “Germanium, Tin and Lead,” Chem. Elem., pp. 367–405, 1997.

[3] E. Shams, H. Abdollahi, M. Yekehtaz, and R. Hajian, “H-point standard addition method in the analysis by differential pulse anodic stripping voltammetry Simultaneous determination of lead and tin.,” Talanta, vol. 63, no. 2, pp. 359–64, May 2004.

[4] L. Průša, J. Dědina, and J. Kratzer, “Ultratrace determination of tin by hydride generation in-atomizer trapping atomic absorption spectrometry,” Anal. Chim. Acta, vol. 804, pp. 50–58, Dec. 2013.

[5] R. Moretto Galazzi and M. A. Z. Arruda, “Optimization of a hydride generation metallic furnace atomic absorption spectrometry (HG-MF-AAS) method for tin determination: analytical and morphological parameters of a metallic atomizer.,” Talanta, vol. 117, pp. 543–8, Dec. 2013.

[6] S. V. De Azevedo, F. R. Moreira, and R. C. Campos, “Direct determination of tin in whole blood and urine by GF AAS,” Clin. Biochem., vol. 46, no. 1–2, pp. 123–127, 2013.

[7] S. Rončević, A. Benutić, I. Nemet, and B. Gabelica, “Tin content determination in canned fruits and vegetables by hydride generation inductively coupled plasma optical emission spectrometry.,” Int. J. Anal. Chem., vol. 2012, p. 376381, Jan. 2012.

[8] D. L. Anderson and W. C. Cunningham, “Nondestructive determination of lead, cadmium, tin, antimony, and barium in ceramic glazes by radioisotope X-ray fluorescence spectrometry.,” J. AOAC Int., vol. 79, no. 5, pp. 1141–57, Jan. .

[9] M. a White, “A comparison of inductively coupled plasma mass spectrometry with electrothermal atomic absorption spectrophotometry for the determination of trace elements in blood and urine from non occupationally exposed populations.,” J. Trace Elem. Med. Biol., vol. 13, no. 1–2, pp. 93–101, 1999.

[10] F. Daverat, L. Lanceleur, C. Pécheyran, M. Eon, J. Dublon, M. Pierre, J. Schäfer, M. Baudrimont, and S. Renault, “Accumulation of Mn, Co, Zn, Rb, Cd, Sn, Ba, Sr, and Pb in the otoliths and tissues of eel (Anguilla anguilla) following long-term exposure in an estuarine environment.,” Sci. Total Environ., vol. 437, pp. 323–30, Oct. 2012.

[11] J. D. Herner, P. G. Green, and M. J. Kleeman, “Measuring the trace elemental composition of size-resolved airborne particles.,” Environ. Sci. Technol., vol. 40, no. 6, pp. 1925–33, Mar. 2006.

[12] D. R. Smith and M. Nordberg, General Chemistry, Sampling, Analytical Methods, and Speciation∗, Fourth Edi. Elsevier, 2015.

[13] R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid, and J. F. Tyson, “Atomic spectrometry updates. Review of advances in elemental speciation,” J. Anal. At. Spectrom., vol. 29, no. 7, p. 1158, Jun. 2014.

[14] J. Cavalheiro, H. Preud’homme, D. Amouroux, E. Tessier, and M. Monperrus, “Comparison between GC-MS and GC-ICPMS using isotope dilution for the simultaneous monitoring of inorganic and methyl mercury, butyl and phenyl tin compounds in biological tissues.,” Anal. Bioanal. Chem., vol. 406, no. 4, pp. 1253–8, Mar. 2014.

[15] “Occupational health guideline for inorganic tin compopunds.pdf.” pp. 1–6, 2013.

[16] “Chemical Sampling Information | Tin, inorganic compounds (except oxides) (as Sn).” [Online]. Available: https://www.osha.gov/dts/chemicalsampling/data/CH_271800.html. [Accessed: 10-Aug-2015].

[17] A. Boehncke, G. Koennecker, I. Mangelsdorf, and A. Wibbertmann, “Concise international chemical assessment document 6,” IPCS Concise Int. Chem. Assess. Doc., no. 6, pp. 1–37, 1999.

[18] E. Bulska, “Modifiers in Graphite Furnace Atomic Absorption Spectrometry – Mechanisms,” At. Spectrosc.

[19] B. Welz, G. Schlemmer, and J. R. Mudakavi, “Palladium nitrate?magnesium nitrate modifier for electrothermal atomic absorption spectrometry. Part 5. Performance for the determination of 21 elements,” J. Anal. At. Spectrom., vol. 7, no. 8, p. 1257, 1992.

[20] Eurachem, The Fitness for Purpose of Analytical Methods. 1998.

[21] L. Yang, Z. Mester, and R. E. Sturgeon, “Species-specific isotope dilution-based calibration for trace element speciation and its combined uncertainty evaluation: determination of tributyltin in sediment by HPLC-ICPMS.,” Anal. Chem., vol. 74, no. 13, pp. 2968–76, Jul. 2002.

[22] M. Uveges, L. Abrankó, and P. Fodor, “Optimization of GC-ICPMS system parameters for the determination of butyltin compounds in Hungarian freshwater origin sediment and mussel samples.,” Talanta, vol. 73, no. 3, pp. 490–7, Sep. 2007.



Corresponding author: Dinh-Binh Chu

Email: binh.chudinh@hust.edu.vn