* Corresponding Author

Tel: +86-312-5079618

E-mail: zhangdq@mail.hbu.du.cn

Abstract: A method has been described for the determination of arsenic by hydride generation atomic fluorescence spectrometry (HG-AFS). The experimental conditions influencing the atomic fluorescence intensity and the reduction of arsenic were investigated and optimized. The influence from foreign ions and its elimination were studied. The detection limit and relative standard deviation (RSD) were found to be 87 ng/L and 1.24%, respectively. The proposed method was applied to the determination of arsenic in traditional Chinese Medicines including Niuhuang Qingxin Wan (1,2) and Niuhuang Qingxin Pian produced by different manufactures with a recovery range of 93.6-107.3%. The advantages of the HG-AFS technique include simplicity, speed and less cost.

Key words: determination, HG-AFS, arsenic, Traditional Chinese Medicine

The important of Traditional Chinese Medicinal (TCM) as a natural medicinal material has been gaining increasing recognition worldwide in recent years^[1]. However, the safety, quality, and efficacy of TCM must be critically evaluated before they can be put into clinical trials or placed on the market ^[2]. For its high toxicity and the poison effects to human body, arsenic must be determined at very low concentration levels in various samples including TCM. For the determination of arsenic, many analytical methods had been proposed such as atomic absorption spectrometry (AAS)^[3], inductively coupled plasma-atomic emission spectrometry (ICP-AES)^[4], inductively coupled plasma-mass spectrometry (ICP-MS)^[5] and atomic fluorescence spectrometry (AFS)^[6,7]. Ni et al. reported a method to minimize the phosphate interference in the determination of arsenic with elctrothemal AAS ^[3]. Rhodium in the form of (NH₄)₃RhCl₆ plus citric acid as mixed matrix modifier was proposed to the minimization of phosphate interference in the determination of arsenic in urine samples arsenic at 20 ng ml^-1 level could be determined using the proposed method. The detection limit of arsenic with Pd(NO₃)₂ as modifier was found to be 0.3 ng ml^-1 and the spectral interferences can be avoided using the proposed method. An ICP-mass spectrometry (ICP-MS) method was used for the determination arsenic in TCM drugs under the comparison of different sample digestion methods according to Wang et al.^{[5 ]}

For the determination of arsenic, the hydride generation technique was always applied with the combination with AFS ^[6,7] for its high sensitivity, high selectivity and relative freedom from interference.

The main purpose of this work is to establish a method for the determination of trace arsenic in TCM using AFS combined with hydride generation technique.

2.1. Apparatus

An AFS-230 model atomic fluorescence spectrophotometer (Beijing Haiguang Instrumental Company, Beijing, China) equipped with an arsenic hollow cathode lamp (Beijing Haiguang Instrumental Company, Beijing, China) was employed for the measurements of atomic fluorescence intensity of arsenic. An automatic intermittent hydride generation device was used to generate the arsenic hydride. Argon gas was used as the carrier gas for the transposition of arsenic hydride from the hydride generator to the atomizer.

2.2. Reagents

A 1% KBH₄ (m/v) solution was prepared by dissolving approximate amount of KBH₄ in 0.2% NaOH (m/v) daily just before use without filtration.

Arsenic (III) stock solution (1000 mg/L) was prepared by dissolving 1.3200 g of As₂O₃ in 25 ml of 20% (m/v) KOH solution, which was neutralized with 1% H₂SO₄ and diluted to 1000 ml with 1% H₂SO₄.

A KI solution (120g/L) was obtained by dissolving appropriate of KI in distilled water and 5 ml of this solution was added to the sample, regent blank and standard solutions as a reductant for the generation of hydride of arsenic.

An appropriate of thiourea was dissolved in distilled water to obtain a solution of 100 g/L and 5 ml of this solution was added to the sample, regent blank and standard solutions as sensibilization reagent.

All reagents used were of analytical reagent grade and distilled water was used through the experiment.

Traditional Chinese medicine samples including Niuhuang Qingxin Wan1, Niuhuang Qingxin Wan2 and Niuhuang Qingxin Pian were obtain commercially and used as received.

2.3. Sample digestion

The TCM samples were digested using HNO₃-H₂O₂ under the digestion procedure. A 2.0-2.5g portion of TCM sample was accurately weighed into a 100ml beaker to which 20ml of concentrated HNO₃ and 4 ml of H₂O₂ were added. After gentle shaking, the sample was heated on a hot plate for about 2 hours. After cooling down to room temperature, 2 ml of H₂O₂ was added and the sample was heated again until the sample solution was clear. The HCl, KI and thiourea were added in the residues and the sample solution was transferred to a 50ml volumetric flask and diluted to the mark with water.

2.4. Measurement procedure

The digested sample was mixed with 8 ml of HCl, 5 ml of KI and 5 ml of thiurea, then diluted with distilled water to 50 ml. The obtained sample solution was with 2 mol/L of HCl, 12g/L of KI and 10g/L of thiourea. Samples transported were mixed with a 1% (m/v) KBH₄ solution. The generated arsenic hydride was transferred to the atomizer with argon gas and the atomic fluorescence signal was measured under the optimum conditions. All experiments were carried out using the optimum conditions shown in Table 1 and Table 2.

Table 1. Operating conditions for HGAFS

Table 2. Working program for the intermittent flow reactor

3.1. Effect of high voltage of PMT

The influence of the high voltage of PMT on the atomic fluorescence signal and the stability was investigated. The results show that the atomic fluorescence signal increased with the increase of the high voltage of PMT while the stability decreased. This result indicates that the detection limit of arsenic changed with the increase of high voltage and the best detection limit was obtained with 300 V. Higher high voltage will result in the poor stability and detection limit. Therefore, a 300 V of high voltage was used in the later experiments.

3.2 Effect of atomizer temperature and the atomizer height

The atomizer used is a quartz furnace in a shielding mode, which is consisted of the inner tube and the outer shielding tube, heated by means of externally wrapped resistance wire. Hydrogen and the vapor of arsenic, which originated from the intermittent flow system, are swept into the inner tube by an argon carrier gas and atomized. Around the top of the quartz furnace outlet, there is a resistance wire making the gas mixture ignited at a very low furnace temperature. The influences of the atomizer height on the fluorescence were studied. The result show that the fluorescence signals increased with the increase of the atomizer temperature and up to a maximum value at 400° C. But a further increase of atomizer temperature will result in a lower fluorescence signal. This is because at elevated furnace temperature the gas volume expands, decreasing the atom density and consequently the measured intensity. In this case, the blank only decreased slightly when the atomizer temperature increased，thus atomizer temperature of 400° C was considered as the most appropriate.

Atomizer height is the distance from the top of the furnace atomizer to the point where the atomic fluorescence signal is measured. The effects of atomizer height on the fluorescence signals were studied, The results demonstrated that the atomizer height affects the atomic fluorescence intensity seriously. It was found that the increased atomizer height resulted in the fluorescence signals for arsenic increased until the atomizer height was 12mm. In order to get the optimum signal-to-noise ratio, the atomizer height of 12 mm was considered in the determination of arsenic.

3.3. Effect of lamp current on signal

An increase of the lamp current resulted in an increase of the atomic fluorescence signal and the maximum value for arsenic was observed at a lamp current of 60 mA. Furthermore, the lamp current higher than 60 mA caused the reduction of the atomic fluorescence signal of arsenic and the lifetime of the lamp. Moreover, the background varied slightly with the increased lamp current. Therefore, the lamp current of 60 mA was selected.

3.4. Effect of flow rate of carrier gas and shield gas on signal

High purity argon gas was used as the carrier gas to strip the arsenic hydride from the generator to the atomizer, and the atomic fluorescence signal for arsenic was examined in relation to the flow rate of carrier gas. The flow rate of carrier gas in the range of 100-800 ml/min was tested and it was found that the atomic fluorescence signal decreased with the flow rate of carrier gas both lower and higher than 400 ml/min. So a carrier gas flow rate of 400 ml/min was used to strip arsenic hydride.

The shield gas is often used to prevent extraneous air from entering the flame and this has been found useful to increase the signal and reduce the flame radiation noise. The flow rate of shield gas affects the atomic fluorescence signal and the maximum atomic fluorescence signal was obtained at the flow rate of shield gas of 900 ml/min. Therefore, a 900 ml/min of shield gas flow rate was selected.

3.5. Effect of hydrochloric acid concentration on hydride generation

The HCl was used as the acid medium in the generation of arsenic hydride and the effect of the concentration of HCl on the atomic fluorescence signal of arsenic was studied and the results are shown in Fig. 1. The results in Fig. 1 indicate that the effect of HCl concentration is serious on the atomic fluorescence signal while no significant effect was observed on the background signal. The results also indicate that the interference from interfering ions can be prevented with higher HCl concentration. The HCl concentration of 2 mol/L was selected as the acid medium.

The HCl was also used as acid medium in the carrier solution. The influence from HCl concentration on the atomic fluorescence signal was tested and the results are shown in Fig. 1. As can be seen in Fig.2, the background signal almost remained unchanged with HCl in a concentration rage of 0.5-4 mol/L and the maximum atomic fluorescence signal was observed when 1 mol/L HCl was used as acid medium in the carrier solution. .

Figure.1 Effect of concentration of HCl on the fluorescence of 50ng mL^-1 arsenic.

u : Effect of concentration of HCl of medium on the fluorescence and n : Effect of concentration

hus a 1 mol/L HCl was selected as HCl of carrier solution on the fluorescence

3.6. Effect of potassium tetrahydroborate concentration on hydride generation

In the present method, KBH₄ is used not only as a reductant but also as the hydrogen source, which is necessary to sustain the argon-hydrogen flame. Consequently, in this technique, the concentration of KBH₄ has a large impact on the signal response. Firstly, it affects the hydride generation process and secondly, it influences the argon-hydrogen flame, and this directly affects the atomization process and the flame emission noise. At a higher KBH₄ concentration, the flame become larger and the visual part of the flame increases. However, the noise level caused by flame emission also increased with a higher concentration of KBH₄.

The examination of the effect from concentration of KBH₄ on the atomic fluorescence signal was carried out and the results are depicted in Fig. 2. The results show that, the signal increased with the increase of the concentration of KBH₄ before 1%(m/v) and the signal reached a maximum value while the concentration of NaBH₄ is 1%(m/v). After 1%(m/v), the signal decreased and then descended to a plateau. The results indicate that a 1%(m/v) of NaBH₄ concentration was required for the formation of arsenic hydride, consequently a concentration of NaBH₄ of 1% (m/v) was selected.

Figure 2. Effect of KBH₄ concentration on the atomic fluorescence signal of 50 ng/ml of arsenic. ■: arsenic signal and ♦: background signal.

3.7. Effect of reductant

For the conditions previously selected for AsH₃ generation, the signals obtained form As(III) and As (V) were compared and the regression equations are summarized as following.

I_f=8.1375 C_{As (V)}. + 0.8112 r=0.9993

I_f=16.887 C_{As ()} + 1.336 r=0.9995

It was found that the atomic fluorescence signals obtained from As (V) solutions are about 50% of those obtained from As(III) solutions with the same concentration. Because of that, the effect of previous reduction conditions on the fluorescence intensity of arsenic was investigated in order to obtain as higher as possible signal form As (V).

Potassium iodide is an efficient reductant for the reduction of As (V) to As(III). The experimental results show that a concentration of 12 g/L KI was enough to obtain the same atomic fluorescence signals from solutions of As(III) and As (V) with the same concentration.

3.8. Study of effect of sensitization

For the determination of arsenic, ascorbic, thiourea or mixture ascorbic-thiourea are usually used as sensitization reagents. In order to obtain the highest signal of arsenic, the domino effect of sensitization of ascorbic, thiourea and ascorbic-thiourea were investigated. A concentration of 10 g/L thiourea was considered the best regents for sensitization of signal of arsenic.

3.9. Effect of foreign ions on atomic fluorescence signal

Table 3. Relative signals obtained for As(III) in the presence of foreign ions mixed with and without thiourea

The influence from other metallic ions on the determination of arsenic was investigated. It was reported ^[8] that alkali metals and alkaline metals did not affect the atomic fluorescence signals whereas transition metals, noble metals and metals that can form hydrides influence the atomic fluorescence signals. So the effects from Fe³⁺, Mn²⁺, Zn²⁺, Cr³⁺, Cd²⁺, Co²⁺, Ni²⁺, Ag⁺., Mo²⁺, Cu²⁺, Hg²⁺ and Se⁴⁺ were studied. These ions were added individually to a 100 ml of solution containing 20 ng of As(III). It is found (shown in Table 3) that those ions actually influence the fluorescence signals. The interference was studied with the addition of foreign ions to a 100 ml of solution containing 20ng of As(III) and 1g of thiourea, and the results show that the ions mixed with thiourea did not affect the fluorescence signals. In all the experiments, thiourea was added for avoiding the foreign ions effect the fluorescence signals.

Nitric acid has been reported seriously affect the fluorescence signals which could not be eliminate by ascorbic acid/thiourea ^[9,10], so it is required to exhaust the nitric acid during the digestion procedure.

3.10. Analytical figures of merit

The detection limit (3s ) was found to be 87 ng/L and the relative standard deviation (RSD) is 1.24% at 20 ng/ml level of arsenic.

The regression equation for the calibration curve for standard arsenic (prepared with KI-thiourea previous reduction) was Y=28.78X+1.542 with correlation coefficient of 0.9996, where Y is atomic fluorescence intensity and X is concentration of arsenic(ng/L). The linear range of calibration curve is 0 to 90 ng/ml of arsenic.

3.11. Analysis of TCM samples

The proposed method was applied to the determination of arsenic in TCM samples and the results are summarized in Table 4. Recovery test by spiking arsenic to real TCM samples were carried out and the recoveries are also listed in Table 4. The results show that the recoveries are in the range of 93.6%-107.3%.

4. Conclusion

HGAFS provides an extremely sensitive methodology for determination of arsenic in TCM samples. With KI-thiourea previous reduction, the sensitivity and detection limit has been proved. The advantages of the proposed method include high sensitivity, wide linear range, speed of analysis, ease of use and low cost.

Regulations on maximum levels of metals allowed in foodstuffs were currently under revision [11]. It is necessary to determine arsenic in the TCM drugs because even low concentration of arsenic can cause serious toxic effects to organisms. However, quite often, toxic metals are intentionally included in TCM as part of the active ingredients. Niuhuang Qingxin Wan1, for instance, arsenic in the form of realgar, a crystal from of As₄S₄ or As₂S₂ presented as a part of the drug. Niuhuang Qingxin Wan has been used for many years because of its demonstrated effectiveness in treating deficiency of energy and blood and other related mind disease. But the exact function of realgar in the formulations remains unclear. Chemically, the arsenic in realgar, when mixed with other components in the drug, may well lead to some other chemical forms through chemical reactions. Arsenic is true in many TCM drugs, the effect of arsenic should be paid more attention.

Table 4. Analytical results and recoveries for arsenic from real samples

Reference

1. Wong A.B., Hong Kong Industrialist, 12,33,1997.
2. Chen D.M., Science and Technology Review, 27,1,1998.
3. Ni Z.M., Rao Z. and Li M., Anal. Chim. Acta, 334, 177-182, 1996.
4. Zhang L., Shan X.Q. and Ni Z.M., Fresenius Z. Anal. Chem., 332, 764-768, 1988.
5. Wang X.R., Zhuang Z.X., and Sun D.H., et al.At. Spectrosc., 20, 86-91, 1999.
6. El-Hadri F., Morales-Rubio A. and Guardia M., Talanta, 52, 653-662, 2000.
7. Gallardo M.V., Bohari Y., and. Astruc A, Anal. Chim. Acta, 441, 257-268, 2001.
8. Zhang P., Guangpuxue Yu Guangpufenxi, 9,61,1989.
9. Gondlen P.D., Brooksbank P., Anal. Chem, 46,431,1974.
10. Agemian H., Thomson R., Analyst, 105,902,1980.
11. Coleman M.E., Elder R.S., and P. Basu, JAOAC Int., 75, 615,1992.