Translate this page into:
Separation and preconcentration of trace amounts of gold from water samples prior to determination by flame atomic absorption spectrometry
⁎Corresponding author at: Department of Chemistry, Payame Noor University of Kerman, Kerman, Iran. Tel./fax: +98 341 3321492. bahramy.2010@yahoo.com (Habibe Bahrami)
-
Received: ,
Accepted: ,
This article was originally published by Elsevier and was migrated to Scientific Scholar after the change of Publisher.
Peer review under responsibility of King Saud University.
Abstract
A preconcentration/separation procedure is presented for the solid phase extraction of trace gold(III) as its rubeanic acid (dithiooxamide) chelate on silica gel, prior to determination by flame atomic absorption spectrometry. The influences of analytical parameters including pH of the aqueous solution, the amount of the sorbent, time of the complex formation, ligand amount, flow rates of sample and elution solutions and the type, concentration and volume of elution solution on the quantitative recoveries of Au(III) were investigated. At pH 3.5, the maximum sorption capacity of Au3+ was 7.5 mg g−1, by column method. The linearity was maintained in the concentration range of 1.0–3.4 × 104 ng mL−1 for gold in the original solution. The preconcentration factor of 100 and relative standard deviation of ±1.7% were obtained, under optimum conditions. The limit of detection (LOD) was calculated as 0.80 ng mL−1, based on 3σbl/m (n = 8) in the original solutions. The proposed method was successfully applied to the determination trace amounts of gold in the water samples.
Keywords
Gold
Preconcentration
Rubeanic acid
Silica gel
1 Introduction
Gold belongs to the group of elements which occur on the earth in very low natural contents. Its concentration is about 4 ng g−1 in basic rocks and 1 ng g−1 in soils. The values of 0.05 and 0.2 ng mL−1 were found in seawater and river water, respectively (Medved et al., 2004). Its products are extensively used in various areas such as the petrochemical industry, medicine, electronics and nuclear power industries. The significance of developing accurate and dependable analytic procedures for gold analysis is related to its increasing presence in the environment and to a growing interest in the elucidation of its role in living organisms and the impact on human health (Underwood, 1977; Tsalev and Zaprianov, 1985). In spite of improvements in the sensitivity of analytical systems, preconcentration and separation of analytes from the sample matrices are still necessary to obtain reliable results.
Solid phase extraction (SPE) technique has become increasingly popular when compared with other techniques such as liquid–liquid extraction, because of its advantages of high enrichment factor, high recovery, rapid phase separation, low cost, low consumption of organic solvents and the ability of combination with different detection techniques in the form of on-line or off-line mode.
The choice of sorbent is a key point in SPE, because it can control the analytical parameters such as selectivity, affinity and capacity. Preparation of new material for selective solid phase extraction of analytes is an important trend of solid phase extraction (Pool, 2003). Reagents can be modified on organic or inorganic support as solid phase extractants. Silica gel presents the advantages of no swelling, fast kinetics, mechanical, thermal and chemical stability under various conditions (Jal et al., 2004). Therefore; it is a widely used support for various solid-phase extractants. Recently, some chelating reagents have been modified on silica gel as solid-phase extractants for separation/preconcentration of some metal ions.
Aniline formaldehyde condensate coated silica gel, was used separately, for preconcentration of Cr(VI) and Cu(II) in aqueous solutions (Kumar et al., 2007a,b). Titanium dioxide immobilized on silica gel was used to enrich and separate gold prior to flame atomic absorption spectrometry (Liu and Liang, 2007). Silica gel modified with murexide was applied to preconcentration of uranium (VI) ions from water samples (Sadeghi and Sheikhzadeh, 2009). Silica gel–polyethylene glycol was used as a sorbent for solid phase extraction of cobalt and nickel (Pourreza et al., 2010). Cd, Cu, Hg and Pb in environmental and biological samples were determined with silica–mercaptopropyltrimethoxysilane (Huang and Hu, 2008). Silica gel modified with sulfanilamide was prepared for extraction of some metal ions (Zou et al., 2009) and silica gel modified with alizarin violet for extraction of lead (Fan et al., 2008).
So far, several sorbents have been used for solid phase extraction of Au(III) (Shamspur and Mostafavi, 2009; Wu et al., 2004; Hu et al., 2005; Yin et al., 2009; Morales and Toral, 2007; Hassanien and Sherbini, 2006; Pohl and Prusisz, 2005) . They are produced by immobilization of complexing or chelating reagents on different solid surface through physical loading or chemical bonding.
Rubeanic acid (dithiooxamide) has been used as organic chelating agent for preconcentration of metal ions (Ghaedi et al., 2007; Celik et al., 2010; Dolak et al., 2009; Soylak and Erdogan, 2006; Soylak and Tuzen, 2006), but there is no report on preconcentration of this agent with Au(III) ions on silica gel. Therefore, this work is devoted to the examination of rubeanic acid as a complexing agent with Au(III), and the feasibility of silica gel as a solid phase extractor was investigated. The results showed that the gold ions were easily retained in the form of metal–rubeanic acid complex on silica gel from acidic aqueous solution. The proposed method was successfully applied to the analysis of various water samples.
2 Experimental
2.1 Apparatus and reagents
An atomic absorption spectrometer model Sens AA (Dandenong, Australia, (http://www.gbcaustralia.com) equipped with deuterium lamp background corrector was used for determination of gold in air–acetylene flame. The instrumental settings of the spectrometer were as follows: wavelength, 242.8 nm; slit width of 1.0 nm; lamp current, 4 mA; acetylene flow 1.5 L min−1 and air flow 3.5 L min−1. Funnels-tipped glass tube (5 × 100 mm) equipped with stopcock was used as column for the preconcentration purposes. A 691 Metrohm pH meter (Herisau, Switzerland, http://www.metrom.com) was employed for pH measurements.
All reagents were of analytical grade. Deionized and distilled water were used in all experiments. The stock solution 100.0 ppm of Au(III) was prepared by dissolving 0.0154 g AuCl3 (Merck, Darmstadt, Germany) in 100.0 mL of 0.5 mol L−1 HCl. The standard working solutions were diluted daily prior to use. A 0.1% (w/v) solution of rubeanic acid (Merck) was prepared by dissolving 0.10 g of this reagent in 100.0 mL ethanol. Buffer solution was prepared from 0.1 mol L−1 acetic acid and 0.1 mol L−1 sodium acetate (Aldrich, Milwaukee, USA) for pH 3.5.
2.2 Silica gel and column preparation
Silica gel (0.063–0.2 mm or 70–230 mesh ASTM) (Merck) was used as the supporting material. A small amount of glass wool was placed in the end of column (5 × 100 mm) equipped to stopcock, to prevent loss of the sorbent during sample loading. Then, the column was packed with 100 mg of the silica gel. The bed height of the silica gel in the column was approximately 4 mm. It was preconditioned by passing buffer solution prior to use.
2.3 General procedure
The performance of the column preconcentration method was investigated with the synthetic sample solution before its application to the real samples. A total of 30 mL synthetic sample solution containing 20 μg of Au was taken and 0.5 mL 0.1% rubeanic acid was added. The pH of the solution was adjusted to the optimum value determined experimentally (pH ∼3.5) with acetic acid/acetate buffer. The resulting solution was passed through the preconditioned column by a flow rate adjusted to the optimum value determined experimentally (ca. 3 mL min−1). The retained gold ions were then eluted from the solid phase with 5.0 mL 0.5 mol L−1 thiourea and then 5.0 mL 1 mol L−1 HNO3 solution. This solution was aspirated into an air–acetylene flame for the determination of Au by FAAS.
3 Results and discussion
The optimum sorption and desorption properties of silica gel for Au were found by using the column technique. Because quantitative recovery was not obtained for gold, without using a chelating agent in our preliminary experiments rubeanic acid was used as the chelating agent and the effect of some analytical parameters such as pH, amount of reagent, amount of adsorbent, type and volume of elution solution, flow rate of sample solution, and volume of sample solution on the recovery of the gold ions have been investigated.
3.1 Effect of pH
The pH of the sample solution plays an important role in retention of metal ions. Thus, the effect of pH on the recovery of Au(III) was examined. The pH value of the sample solution was adjusted in a range 1–10 with HNO3 or NaOH. The obtained solutions were passed through the column at a flow rate 3 mL min−1. The metal ions were then eluted by an appropriate eluent and determined by FAAS. As can be seen in Fig. 1, quantitative recovery (>95%) was found in the pH range of 2–4 for gold. pH∼3.5 was selected for further studies and was adjusted with acid/acetate buffer solution.
Effect of pH on the recovery of gold after preconcentration with the proposed method. Conditions: Au, 20.0 μg; flow rate, 3 mL min−1; sorbent, 100 mg; elution solution, 5.0 mL of 0.5 mol L−1 thiourea and then 5.0 mL of 1 mol L−1 HNO3. Instrumental settings: were the same as Table 1.
3.2 Ligand amounts
Prior to adsorption of trace heavy metal ions on a solid phase for preconcentration, generally metal ions were converted into a suitable form including metal chelates or metal inorganic complexes. Because of this point, rubeanic acid was selected as chelating agent for gold ions. Different volumes of 0.1% of rubeanic acid were added to the solutions containing 20 μg Au(III). Then, these solutions were passed through the silica gel column. Quantitative recoveries were obtained for gold ions in the 0.5 and 1 mL rubeanic acid solution. Because of insufficient ligand amounts in the solutions, the recovery of analyte was not quantitative when less than 0.5 mL rubeanic acid solution was added. Above 1.0 mL of rubeanic acid solution, recovery was below 95%, due to competition on the adsorption between rubeanic acid metal chelate and excess rubeanic acid in the solution. In further works, 0.5 mL rubeanic acid solution was added to the solution.
3.3 Effect of amount of the sorbent
The amount of sorbent is another important parameter that affects the recovery. A quantitative retention is not obtained when the amount of sorbent is less. On the other hand, an excess amount of sorbent prevents the elution of the retained chelates by a small amount of eluent quantitatively. For this purpose, different amounts of the sorbent (5–100 mg) were examined. The results showed that quantitative recoveries of the metal ions were obtained when the sorbent quantity was greater than 30 mg. With 100 mg of the sorbent, the highest recovery was obtained. Therefore, 100 mg was chosen for further experiments.
3.4 Effect of the type and volume of elution solution
Desorption of the retained Au(III) from the column was tested using various eluting agents such as nitric acid, hydrochloric acid, sodium thiosulfate, sodium sulfite, potassium thiocyanate, EDTA and thiourea. Since the complex of Au(III) with rubeanic acid is stable, most of them could not be eluted Au(III) from the column, completely. As can be seen from Table 1, the maximum recovery has been obtained by using 5.0 mL of 0.5 mol L−1 thiourea and then 5.0 mL of 1 mol L−1 HNO3, separately.
| Solution | Recovery (%)a |
|---|---|
| 5.0 mL of 3.0 mol L−1 HCl | 11.1 ± 1.8 |
| 5.0 mL of 1.0 mol L−1 Na2SO3 | 15.2 ± 2.0 |
| 5.0 mL of 1.0 mol L−1 Na2S2O3 | 6.1 ± 1.2 |
| 5.0 mL of 0.1 mol L−1 EDTA | 4.4 ± 0.81 |
| 5.0 mL of 0.1 mol L−1 KSCN | 17.8 ± 4.3 |
| 5.0 mL of 0.5 mol L−1 thiourea in 3.0 mol L−1 HCl | 8.3 ± 1.0 |
| 5.0 mL of 3.0 mol L−1 HCl and then 5.0 mL 0.5 mol L−1 thiourea | 65.4 ± 1.3 |
| 5.0 mL of 3.0 mol L−1HNO3 and then 5.0 mL 0.5 mol L−1 thiourea | 82.1 ± 1.1 |
| 5.0 mL of 3.0 mol L−1 H2SO4 and then 5.0 mL 0.5 mol L−1 thiourea | 23.7 ± 0.4 |
| 5.0 mL of 0.5 mol L−1 thiourea and then 5.0 mL of 1.0 mol L−1 HNO3 | 99.2 ± 1.7 |
| 6.0 mL of 0.5 mol L−1 KSCN 4.0 mL of 3.0 mol L−1 HNO3 | 49.7 ± 0.9 |
| 5.0 mL of 3.0 mol L−1 HCl and then 5.0 mL 1.0 mol L−1 thiourea | 57.8 ± 1.1 |
| 3.0 mL 0.5 mol L−1 thiourea and then 2.0 mL 3.0 mol L−1 HCl | 30.3 ± 0.6 |
Conditions: Au, 20.0 μg; pH ∼3.5; buffer, 2 mL; flow rate, 3 mL min−1; sorbent, 100 mg.
Instrumental settings: current of Au hollow cathode lamp 4 mA; absorption line Au 242.8 nm; slit width 1.0 nm; acetylene flow 1.5 L min−1 and air flow 3.0 L min−1.
3.5 Effect of flow rate of sample and eluent solution
The retention of an element on a sorbent also depends on the flow rate of the sample solution. Thus, the effect of flow rate of the sample and elution solution on the retention and recovery of gold ions were investigated under optimum conditions. The solution containing Au(III) was passed through the column with the flow rates adjusted in a range 0.1–4 mL min−1. It was observed that, at flow rates greater than 3 mL min−1, there was a decrease in the recovery of Au. The reason for this decrease is probably insufficient contact of the metal ions and the sorbent to reach equilibrium. Therefore, a flow rate of 3 mL min−1 was selected for subsequent experiments.
For desorption of gold ions, flow rate was varied between 0.5 and 4 mL min−1. The flow rate of 3 mL min−1 was adequate for desorption of the analyte. Therefore, a flow rate of 3 mL min−1 was used in both steps, sorption and desorption, and was adjusted with a stopcock in end of the column.
3.6 Effect of complex formation time
In order to investigate the effect of time on the Au(III)–rubeanic acid complex formation, extraction was carried out for a series of solutions containing 20 μg Au(III) and 0.5 mL rubeanic acid at the adjusted pH, in the time range 0–20 min. The results showed that the time has no effect on the extraction efficiency of gold and the complex formation between gold and rubeanic acid was very fast.
3.7 Effect of the sample volume (breakthrough volume)
The measurement of breakthrough volume is important in solid phase extraction because breakthrough volume represents the sample volume that can be preconcentrated without loss of analyte during elution of the sample. The volume of the first aqueous phase, containing a fixed amount of analyte (20 μg Au), was varied in the range of 100–1200 mL under the optimum conditions, keeping other variables constant, and was passed through a column for preconcentration (Fig. 2). It was observed that adsorption was constant up to 1000 mL. At higher sample volumes, the recoveries decreased gradually with increasing volume of sample solution. Since the final elution volume was 10.0 mL, preconcentration factor of 100 was obtained for gold.
Breakthrough volume curve of gold. Conditions: Au, 20.0 μg; pH ∼3.5; flow rate, 3 mL min−1; sorbent, 100 mg; elution solution, 5.0 mL of 0.5 mol L−1 thiourea and then 5.0 mL of 1 mol L−1 HNO3. Instrumental settings: were the same as Table 1.
3.8 Effect of coexisting ions
The effect of common coexisting ions in natural water samples on the recovery of gold was studied. In these experiments, 5 mL of the solutions containing 20 μg Au and various amounts of interfering ions were treated according to the proposed method. The tolerance limit was set as the concentration of the ion required to cause ±3% error. The results obtained are given in Table 2. Among the cations and anions examined most could be tolerated up to milligram levels. Thus, this method is selective and can be used for determination of gold in the water samples.
| Salt or ion | Tolerance limit (mg) |
|---|---|
| Na2CO3, NaCl | 500 |
| NaHCO3, Na2SO4 | 500 |
| (NH4)2SO4 | 500 |
| KBrO3, CH3COONa | 500 |
| Trisodium citrate | 500 |
| KI | 450 |
| Sodium oxalate | 50 |
| NaBr | 50 |
| Pb2+ | 25 |
| Zn2+ | 50 |
| Ni2+ | 3.0 |
| Cd2+ | 3.0 |
| Pd2+ | 1.0 |
| Fe3+ | 20 |
| Co2+ | 0.3 |
| Cr3+ | 0.3 |
Tolerance limit is defined as concentration of the ion required to cause ±3% error.
Conditions: Au, 20.0 μg; pH ∼3.5; buffer, 2 mL; flow rate, 3 mL min−1; sorbent, 100 mg.
Elution solution, 5.0 mL 0.5 mol L−1 thiourea and then 5.0 mL 1 mol L−1 HNO3.
Instrumental settings: were the same as Table 1.
3.9 Sorption capacity of the sorbent
To determine the amount of analyte retained on the column, that is the maximum quantity of the gold, which can be sorbed by silica gel, several solutions were made by difference concentrations and introduced into the column. The outlet solutions were collected and the presence of the analyte was tested in each of them by FAAS. When Au was detected in the outlet solution, the test was stopped and the sorption capacity was calculated. The sorption capacity was found to be 7.5 mg for 1.0 g sorbent.
3.10 Analytical performance
Calibration curve for the determination of gold was prepared according to the proposed procedure under the optimum conditions. The linearity was maintained in the concentration range of 0.05–34 μg mL−1 in the final solution or 1.0–34 × 104 ng mL−1 in the original solution. The equation of the line is A = 0.0263C + 0.0015 with the regression coefficient 0.9994 where A is the absorbance and C is concentration of the metal ion (μg mL−1).
The limit of detection (LOD) of the present work was calculated after application of the preconcentration procedure to blank solutions. LOD of gold based on three times standard deviations of the blank on a sample volume 1000 mL was 0.80 ng mL−1. Eight replicate determinations of a mixture of 2.0 μg mL−1 of gold in the final solution, gives a mean absorbance of 0.0582 with relative standard deviation ±1.7%.
3.11 Real sample analysis
To assess the capability of the method for real samples with different matrices containing varying amounts of diverse ions, the method was applied to the separation, preconcentration and determination of gold from water samples. According to the results, the concentration of gold in analyzed water samples was below the LOD of the method. The suitability of the proposed method for the analysis of natural water samples was checked by spiking samples with 5.0 and 10.0 mL of 2.0 μg mL−1 gold. Good recoveries were obtained for all analyzed samples (Table 3).
| Sample | Gold (ng mL−1) | Recovery (%) | |
|---|---|---|---|
| Spiked | Founda | ||
| River waterb | 0.0 | ND | – |
| 20.0 | 19.92 ± 0.40 | 99.6 | |
| 40.0 | 40.10 ± 0.71 | 100.2 | |
| Spring waterb | 0.0 | ND | – |
| 20.0 | 20.32 ± 0.38 | 101.6 | |
| 40.0 | 39.80 ± 0.81 | 99.5 | |
| River waterc | 0.0 | ND | – |
| 20.0 | 20.12 ± 0.36 | 100.6 | |
| 40.0 | 40.08 ± 0.68 | 100.2 | |
| Tap waterd | 0.0 | ND | – |
| 20.0 | 19.69 ± 0.41 | 98.5 | |
| 40.0 | 38.90 ± 0.62 | 97.2 | |
| Distilled water | 0.0 | ND | – |
| 20.0 | 20.20 ± 0.35 | 101.0 | |
| 40.0 | 40.30 ± 0.78 | 100.7 | |
In order to evaluate the accuracy of the procedure, recovery experiments were also carried out with spiked water samples because of certified reference materials for gold was not available. The results are given in Table 3. The recovery percentages of gold ions were evaluated and the results showed that the real sample matrixes did not affect the recovery of the gold.
4 Conclusions
The results of the present investigations in this paper demonstrate the usability of silica gel with rubeanic acid for selective preconcentration of Au(III) from water samples and determination by atomic absorption spectrometry. The accuracy of the results was verified by analyzing the spike water samples. A comparison of the proposed system with some recent studied procedures is given in Table 4. The analytical performance is not significantly different to those achieved by other methods described in the literatures. Detection limit is comparable to those presented by other methods.
| Sorbent | Detection technique | PFa | LOD (μg L−1) | RSD% | Sorption capacity (mg g−1) | Reference |
|---|---|---|---|---|---|---|
| Chelating resin | FAAS | 200 | 16.6 | <6 | 12.3 | Senturk et al. (2007) |
| MWCNTS | FAAS | 75 | 0.15 | 3.1 | 14.8 | Liang et al. (2008) |
| Coprecipitation | FAAS | 75 | 1.3 | 1–10 | – | |
| Silica gel/ nanometer TiO2 | FAAS | 50 | 0.21 | 1.8 | 3.56 | Liu and Liang (2007) |
| Dowex/M 4195 | FAAS | 31 | 1.61 | <5 | 8.1 | Tuzen et al. (2008) |
| Coprecipitation | ICP-MS | 30 | 0.3 | <2.8 | – | Aydin and Soylak (2007) |
| Nano clay | FAAS | 105 | 0.1 | 2.3 | 3.9 | Afzali et al. (2010) |
| Silica gel/rubeanic acid | FAAS | 100 | 0.8 | 1.7 | 7.5 | This work |
Due to relatively high preconcentration factor, trace gold ions at ng mL−1 level can be determined by this proposed method. The use of organic solvents in the proposed method is eliminated. There is no time for complex formation between gold and rubeanic acid before the preconcentration procedure to obtain quantities recovery. The good precision, relatively high sorption capacity and high tolerance to interferences from matrix ions are other advantages.
References
- J. Hazard. Mater.. 2010;81:957.
- Talanta. 2007;73:134.
- J. Hazard. Mater.. 2010;174:556.
- Asian J. Chem.. 2009;21:165.
- Talanta. 2008;74:1020.
- J. Hazard. Mater.. 2007;147:226.
- Talanta. 2006;68:1550.
- J. Chromatogr. A. 2005;1094:77.
- Spectrochim. Acta B. 2008;63:437.
- Talanta. 2004;62:1005.
- Sep. Purif. Technol.. 2007;57:47.
- J. Hazard. Mater.. 2007;143:24.
- Spectrochim. Acta B. 2008;63:714.
- Anal. Chim. Acta. 2007;604:114.
- Anal. Bioanal. Chem.. 2004;379:60.
- Miner. Eng.. 2007;20:802.
- Micochim. Acta. 2005;150:159.
- Trends Anal. Chem.. 2003;22:362.
- Talanta. 2010;81:773.
- J. Hazard. Mater.. 2009;163:861.
- J. Hazard. Mater.. 2007;149:317.
- J. Hazard. Mater.. 2009;168:1548.
- J. Hazard. Mater.. 2006;137:1035.
- J. Hazard. Mater.. 2006;138:195.
- Atomic Absorption Spectrometry in Occupational and Environmental Health Practice. Boca Raton, Florida: CRC Press; 1985.
- J. Hazard. Mater.. 2008;56:591.
- Trace Elements in Human and Animal Nutrition. New York: Academic Press; 1977.
- Talanta. 2004;63:585.
- J. Hazard. Mater.. 2009;169:228.
- Intern. J. Environ. Anal. Chem.. 2009;89:1043.