Delivered by Ingenta to:University of South Australia

IP : 128.72.164.140Sat, 30 Jun 2012 00:40:35

RESEARCH

ARTIC

LE

Copyright © 2012 American Scientific PublishersAll rights reservedPrinted in the United States of America

Journal ofNanoscience and Nanotechnology

Vol. 12, 3006–3010, 2012

Sensitive and Selective Detection ofMercury (II) Based on the Aggregation of

Gold Nanoparticles Stabilized by Riboflavin

Dan Xu1�2, Hua Wen Zhao1�∗, Cheng Zhi Huang2, Li Ping Wu1,Wen Dan Pu1�2, Jia Jia Zheng2, and Ying Zuo1�2

1Department of Chemistry, Third Military Medical University, Chongqing 400038, P. R. China2College of Chemistry and Chemical Engineering, MOE Key Laboratory for Luminescence and Real - Time Analysis,

Southwest University, Chongqing 400718, P. R. China

Gold nanoparticles (AuNPs) can be stabilized by riboflavin against tris buffer-induced aggregation.However, in the presence of mercury (II) (Hg2+), riboflavin can be released from the AuNPs surfaceand the riboflavin-Hg2+ complex formed, leading to the aggregation of AuNPs in tris buffer. Theaggregation extent depends on the concentration of Hg2+. Based on the aggregation extent, asimple and sensitive method of determining Hg2+ is developed. The method enables the detectionof Hg2+ over the concentration range of 0.02-0.8 �M, with a detection limit (3� ) of 14 nM, andshows excellent selectivity for Hg2+ over other metal ions (Cu2+, Co2+, Cd2+, Pb2+, Mg2+, Zn2+,Ag+, Ce3+, Ca2+, Al3+, K+). More importantly, the method avoids complicated surface modificationsand tedious separation processes.

Keywords: Mercuric Ion, Riboflavin, Gold Nanoparticle, Tris.

1. INTRODUCTION

Mercury is a widespread pollutant in the environment andexists as three different forms: elemental, inorganic andorganic. Mercury is highly toxic, especially when trans-formed into methyl mercury, which accumulates in thehuman body through the food chain, and thus exerts seri-ous risks for human health.1 It is important to developa selective and sensitive assay to detect Hg2+ in aque-ous solution. Several methods,2–5 such as atomic absorp-tion spectroscopy, fluorescence spectroscopy, inductivelycoupled plasma mass spectroscopy, inductively coupledplasma atomic emission spectroscopy, provide limits ofdetection at parts-per-billion (ppb) level. However, thesemethods require complicated and expensive instrumenta-tion, or the procedures are laborious and time consuming.Furthermore, they lack the procedural simplicity for onlinemonitoring. Therefore, it is necessary to develop a sim-ple, cost-effective and sensitive method for the detectionof Hg2+.

∗Author to whom correspondence should be addressed.

In recent years, gold nanoparticles (AuNPs) havereceived great attention and been widely used in the devel-opment of colorimetric sensing. The well-dispersed 13 nMAuNPs are wine red, whereas aggregated AuNPs appearpurple or blue.6�7 AuNPs as a colorimetric reporter, mainlybecause their unique surface plasmon resonance (SPR)absorption is extremely sensitive to their size, shape, sur-rounding media and inter-particle distances.7–9 Based onthis principle, several colorimetric assays have been devel-oped for the detection of various analyte such as DNA,10�11

protein,12 ions13–15 and small molecules.16–18

Today, many AuNPs-based colorimetric methods forHg2+ detection have been developed. Among these meth-ods, DNA/AuNPs-based colorimetric methods have beenextensively employed in constructing Hg2+ sensors by tak-ing advantage of the specificity and strong coordinationof thymine to Hg2+. Mirkin’s group and Cheng’s groupused mercury-specific-DNA-functionalized gold nanopar-ticles as a colorimetric probe to detect Hg2+.19�20 Li’sgroup reported a colorimetric method to detect Hg2+

based on unmodified AuNPs interact differently withsingle-strand DNA and double-stranded DNA.21 However,these methods need to use DNA, which is rather expen-sive. Meanwhile, functionalization of AuNPs with DNA

3006 J. Nanosci. Nanotechnol. 2012, Vol. 12, No. 4 1533-4880/2012/12/3006/005 doi:10.1166/jnn.2012.5831

Delivered by Ingenta to:University of South Australia

IP : 128.72.164.140Sat, 30 Jun 2012 00:40:35

RESEARCH

ARTIC

LE

Xu et al. Sensitive and Selective Detection of Mercury (II) Based on the Aggregation of AuNPs Stabilized by Riboflavin

is a tedious, complicated and time-consuming process.Another colorimetric methods for Hg2+ detection are basedon Hg2+-induced the aggregation of acid-capped AuNPsor protein-functionalized AuNPs,22�23 these methods aresimple but lack excellent selectivity for Hg2+ and need toadd masking agent to improve the selectivity. Herein, weintroduce a simple, sensitive and selective Hg2+ detectionmethod based on the aggregation of riboflavin stabilizedgold nanoparticles. Riboflavin, commonly called vitaminB2, could be adsorbed on AuNPs surface through theimide groups.24 Inspired by the fact that the imide groupsof riboflavin are similar to the binding site of thymine(T) for Hg2+,19�21�25�26 we envision that the riboflavin-stabilized AuNPs probe can be used to sensitively andselectively detect Hg2+. Indeed, our experimental resultsdemonstrated that the method show high sensitivity forHg2+ and excellent selectivity over other metal ions.

2. EXPERIMENTAL DETAILS

2.1. Reagents

Hydrogen tetrachloroaurate (III) tetrahydrate (HAuCl4 ·4H2O) and riboflavin were purchased from SinopharmChemical Reagent Co, Ltd. (Shanghai, China). Sodium cit-rate was purchased from Kelong Chemical Reagent Plant(Chengdu, China). All reagents were of analytical reagentgrade without further purification. Tris-HCl buffer (1 mM,pH 8.17) was used to control the acidity of the aqueoussolutions. Doubly distilled water was used throughout.

2.2. Apparatus

Absorption spectra were recorded on a TU-1901 UV-VisSpectrophotometer (Beijing Purkinje General Instrument,Beijing, China). A XW-80A vortex mixer (ShanghaiJingke Industry Company, Shanghai, China) was used toblend the mixture. S-25 digital pH meter (Shanghai JingkeIndustry Company, Shanghai, China) was used to detectthe acidity of the solution.

2.3. Preparation of AuNPs

AuNPs with a diameter of ∼13 nM were prepared bycitrate reduction of HAuCl4 · 4H2O.

27 Briefly, 100 mL ofaqueous solution consisting of 4 mL 1% HAuCl4 · 4H2Owas heated under magnetic stirring until boiling, and then2 mL 5% trisodium citrate was added to the solutionquickly, with the color changing from pale yellow to deepred in 3 min. Followed by boiling for another 5 min, thesolution was cooled to room temperature (25 �C) undervigorously magnetic stirring, and put in refrigerator (4 �C)for reserve.According to the Beer’s law, the concentration of the as-

prepared AuNPs was determined to be around 12.55 nMby using the extinction coefficient of 2�7×108 M−1cm−1at520 nM.21

2.4. General Procedures

Into a 4 mL centrifugal tube were added 250 �L of theTris–HCl buffer solution (1 mM, pH, 8.17), 500 �L of10 �M Riboflavin solution, 250 �L of AuNPs solution, anappropriate volume of 5 �M Hg2+, vortex-mixed. Finally,the mixture was diluted to 2.5 mL with distilled water. Theabove mixture was then vortex-mixed thoroughly and beallowed to react for 30 min at room temperature beforeUV-vis measurements. The absorption ratio of A656/A522

was taken for parametric optimization and construction ofa calibration curve.

3. RESULTS AND DISCUSSION

3.1. Spectral Characteristics of the Interaction ofHg2+ and Riboflavin-Stabilized AuNPs

Figure 1 shows UV–vis absorption spectra and the corre-sponding photographic images of the AuNPs. It can beseen that the surface plasma resonance absorption peakof AuNPs locates at 520 nM (curve 1), and the color ofthe solution is red (inset bottle a). With the addition oftris buffer, the absorbance at 520 nM decreases and theabsorption band exhibits a red shift, meanwhile, the colorchanges purple and a new characteristic absorption bandappears at ca. 650 nM (curve 2, inset bottle b). Curve 3shows the absorption spectra of the riboflavin-stabilizedAuNPs in the presence of tris buffer, the surface plasmaresonance absorption peak and the absorption intensity ofAuNPs are similar to that of AuNPs alone. Meanwhile, thecolor of the solution is still red (insert bottle c), indicat-ing that riboflavin can protect AuNPs against tris-inducedaggregation. However, upon the addition of the appropri-ate concentration of Hg2+, the intensity of the absorption

Fig. 1. Absorption spectra of AuNPs in different conditions. Theinserted picture (a–d) shows the color change corresponding to thecurves (1–4). Concentration: AuNPs, 1.255 nM; Riboflavin, 2.0 �M;Hg2+, 0.5 �M; Tris–HCl buffer, 1 mM, pH, 8.17.

J. Nanosci. Nanotechnol. 12, 3006–3010, 2012 3007

Delivered by Ingenta to:University of South Australia

IP : 128.72.164.140Sat, 30 Jun 2012 00:40:35

RESEARCH

ARTIC

LE

Sensitive and Selective Detection of Mercury (II) Based on the Aggregation of AuNPs Stabilized by Riboflavin Xu et al.

band at 522 nM decreases, and a characteristic absorptionband appears at ca. 650 nM (curve 4). Correspondingly,the color of solution changes to blue (insert bottle d).Though the concentration of the AuNPs is in nano molarbut others reagent in micro molar, the variation of con-centration could indeed influence the AuNPs aggregatingextent, resulting in the fluctuating of absorption spectraand intensity.The absorption and color changes described above can

be interpreted as follows. First, AuNPs in our experimentwere prepared by the reduction of HAuCl4 with trisodiumcitrate and the surface of AuNPs contained plenty of cit-rate. The negative charge of the citrate on the AuNPs sur-face provides sufficient electrostatic repulsion to stabilizethe particles, and thus AuNPs are dispersed well and thecolor of the solution is red. Second, tris can displace citratefrom the surface of the AuNPs and lead to the aggrega-tion of AuNPs (Fig. 2(A)). This is attributed to the factthat the amino and hydroxyl groups of tris were easier toattach onto the surface of the AuNPs than the carboxylgroups of citrate. The rest amino or hydroxyl groups of trisinteract with each other through hydrogen bonding inter-action, so that AuNPs aggregated.28 Third, riboflavin canprotect AuNPs against tris-induced aggregation, becausethe imide groups of riboflavin were easier to attach ontothe surface of the AuNPs than amino and hydroxyl groupsof tris, and thus riboflavin could improve the stability ofthe AuNPs against the tris-induced aggregation.24 Finally,upon the addition of Hg2+, AuNPs can aggregate in trisagain, which is because Hg2+could strongly combine withriboflavin to form a coordination complex, and the release

Fig. 2. Schematic illustration of the AuNPs colorimetric assay for Hg2+.

of riboflavin from AuNPs leads to the aggregation again(Fig. 2(B)). A simple and sensitive detection method ofHg2+is developed based on Hg2+-induced aggregation ofriboflavin-stabilized AuNPs in tris, and the sensing mech-anism is displayed in Figure 2.

3.2. Optimization of Experimental Conditions

To further optimize the detection method, we investi-gated the effect of riboflavin concentration and the pH ofthe solution. Figure 3 shows the effect of the riboflavinconcentration on the absorption ratio at A656/A522. Withthe variation of the riboflavin concentration, the A656/A522

values of riboflavin-stabilized AuNPs without Hg2+ areweak and stable, whereas those of AuNPs containingHg2+ present different traits. That is, the A656/A522 valuesgets increase with increasing riboflavin concentration to1.8 �M, and then remains almost invariable from 1.8 �Mto 2.4 �M. Thus, it is preferable to choose 2.0 �M ofriboflavin. Figure 4 shows the pH effect on the aggregationof riboflavin-stabilized AuNPs in the absence and presenceof 0.5 �M Hg2+. In the presence of Hg2+, the A656/A522

values were almost invariable over the range of 7.08–8.17.At pH above 8.17, the A656/A522 values decreased, whichis attributed to the fact that Hg2+ ions may be depositedby OH−. In our experiment, we chose Tris-HCl (pH 8.17)as buffer system.Moreover, we found that, the A656/A522 values achieved

maximum at 1 mM tris, and then decreased with increas-ing the tris concentration, indicating that in the presenceof Hg2+, the aggregation degree of the riboflavin-stabilized

3008 J. Nanosci. Nanotechnol. 12, 3006–3010, 2012

Delivered by Ingenta to:University of South Australia

IP : 128.72.164.140Sat, 30 Jun 2012 00:40:35

RESEARCH

ARTIC

LE

Xu et al. Sensitive and Selective Detection of Mercury (II) Based on the Aggregation of AuNPs Stabilized by Riboflavin

0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8 3.00.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4A

656/

A52

2

criboflavin/µM

without Hg2+

with Hg2+

Fig. 3. Effect of riboflavin concentration on the absorption ratio ofA656/A522 in the absence and presence of 0.5 �M Hg2+. Concentration:AuNPs, 1.255 nM; Tris buffer, 1 mM, pH, 8.17.

AuNPs was also influenced by tris concentration. This isprobably due to the fact that tris with a high concentrationcan interact with riboflavin through hydrogen bond, which,in turn, reduces the interaction between the riboflavin andHg2+, leading to the decrease of the aggregation degree ofthe AuNPs.

3.3. Sensitivity and Selectivity for Hg2+

Under optimal conditions, quantitative analysis of Hg2+

was performed. As shown in Figure 5, with the increase ofHg2+ concentration, the absorption peak at 522 nM kepton decreasing and redshift, and a characteristic absorptionband appeared over the 600–700 nM wavelength range.Meanwhile, the color of the AuNPs gradually changedfrom wine red to purple, and finally to blue.

7.0 7.5 8.0 8.5 9.00.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8

2.0

A65

6/A

522

pH

without Hg2+

with Hg2+

Fig. 4. Effect of pH on the absorption ratio of A656/A522 in the absenceand presence of 0.5 �M Hg2+. Concentration: AuNPs, 1.255 nM;riboflavin, 2.0 �M; Tris buffer, 1 mM.

Fig. 5. UV-vis absorption spectra upon the addition of different con-centration of Hg2+, insert: corresponding color changes. Concentration:AuNPs, 1.255 nM; Riboflavin, 2.0 �M; Tris–HCl buffer, 1 mM, pH,8.17; Hg2+ (�M), 0, 0.02, 0.04, 0.1, 0.2, 0.3, 0.4, 0.5, 0.8, 1.0.

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

A656

/A522

= 0.01952+1.72018 cHg2+

0.0 0.2 0.4 0.6 0.8

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

A65

6/A

522

A65

6/A

522

cHg2+/µM

cHg2+/µM

Fig. 6. Calibration curve of Hg2+ concentration. Inset: Linear depen-dence of A656/A522 on the Hg2+ concentration.

Fig. 7. Histogram image of the selectivity of Hg2+. The inset showsthe corresponding photographic images of the AuNPs. Concentration:AuNPs, 1.255 nM; Riboflavin, 2.0 �M; Tris–HCl buffer, 1 mM, pH, 8.17;Hg2+, 0.5 �M; Cu2+, 20 �M; Cd2+, 12 �M; other metal ions, 60 �M.

J. Nanosci. Nanotechnol. 12, 3006–3010, 2012 3009

Delivered by Ingenta to:University of South Australia

IP : 128.72.164.140Sat, 30 Jun 2012 00:40:35

RESEARCH

ARTIC

LE

Sensitive and Selective Detection of Mercury (II) Based on the Aggregation of AuNPs Stabilized by Riboflavin Xu et al.

Table I. The results of synthetic samples.

Added Hg2+ Determined Recovery (%)Samples concentration (×10−6 M) Main interferes values (×10−6 M) (n= 5) RSD (%) (n= 5)

1 0.30 Pb2+, Mg2+ Ce3+, K+ 0.30 99.32–105.99 2.722 0.50 Co2+, Zn2+ Al3+, Ag+ 0.47 89.08–98.89 4.00

Concentrations: AuNPs, 1.255 nM; Riboflavin, 2.0 �M; Tris–HCl buffer, 1 mM; pH, 8.17; other metal ions, 20 �M.

The absorption ratio of A656/A522 was chosen to con-struct the calibration curves. There is a linear correlationbetween the absorption ratio and Hg2+ concentration overthe range of 0.02–0.8 �M, with the correlation coefficientof 0.9943 (Fig. 6). The calibration equation is A656/A522 =0�01952+ 1�72018 cHg2+, and the detection limit (3�) is14 nM.In order to evaluate the selectivity of this system for

Hg2+, some environmentally relevant metal ions weretested, including 20 �M Cu2+, 12 �M Cd2+ and Co2+,Pb2+, Mg2+, Zn2+, Ag+, Ce3+, Ca2+, Al3+, K+ of 60 �M.As indicated in Figure 7, only Hg2+ can cause a red-to-blue color change and a strong A656/A522 value. This meansthat our assay responds selectively for Hg2+.

3.4. Detection of Real Samples

To test the present method, synthetic samples containingvarious metal ions were determined, and the results arelisted in Table I. It is showed that the recoveries of thesamples are between 89.08% and 105.99%, and the R.S.Dis lower than 4.0%.

4. CONCLUSION

In this contribution, a simple, cost-effective and sensi-tive method for Hg2+ detection has been developed. Thismethod is based on Hg2+ -induced the aggregation ofriboflavin-stabilized AuNPs in tris buffer, due to the coor-dination of Hg2+ and riboflavin. Our proposed method hasseveral advantages. Firstly, this method has a low detec-tion limit of 14 nM, and exhibits excellent selectivity forHg2+. Secondly, this method is unlabeled, and avoids com-plicated surface modifications and tedious separation pro-cesses. Thirdly, it can be developed to detect Hg2+ justwith the naked eye through the simple color change ofAuNPs.

Acknowledgment: All authors herein are grateful forthe support from the Chongqing Natural Science Foun-dation (CSTC, 2007BB0049) and the National NaturalScience Foundation Committee of China.

References and Notes

1. A. Khatoon-Abadi, A. R. Sheikh Hoseini, and B. Khalili, Int. J. FoodSafety, Nutrition and Public Health. 1, 33 (2008).

2. Y. A. Vil’pan, I. L. Grinshtein, A. A. Akatove, and S. Gucer, J. Anal.Chem. 60, 45 (2005).

3. L. Yu and X. Yan, At. Spectrosc. 25, 145 (2004).4. Y. Li, C. Chen, B. Li, J. Sun, J. Wang, Y. Gao, Y. Zhao, and Z. Chai,

J. Anal. At. Spectrom. 21, 94 (2006).5. C. A. Trimble, R. W. Hoenstine, A. B. Highley, J. F. Donoghue, and

P. C. Ragland, Mar. Georesour. Geotechnol. 17, 187 (1999).6. S. He, D. Li, C. Zhu, S. Song, L. Wang, Y. Long, and C. Fan, Chem.

Commun. 40, 4885 (2008).7. N. L Rosi and C. A. Mirkin, Chem. Rev. 105, 1547 (2005).8. T. K. Sau, A. L. Rogach, F. Jäckel, T. A. Klar, and J. Feldmann,

Adv. Mater. 21, 1 (2009).9. J. Wang, L. Wang, X. Liu, Z. Liang, S. Song, W. Li, G. Li, and

C. Fan, Adv. Mater. 19, 3943 (2007).10. K. Sato, K. Hosokawa, and M. Maeda, J. Am. Chem. Soc. 125, 8102

(2003).11. J. Song, Z. Li, Y. Cheng, and C. Liu, Chem. Commun. 46, 5548

(2010).12. Y. Xu, J. Wang, Y. Cao, and G. Li, Analyst. 136, 2044 (2011).13. F. Tan, X. Liu, X. Quan, J. Chen, X. Li, and H. Zhao, Anal. Methods.

3, 343 (2011).14. H. Wei, B. Li, J. Li, S. Dong, and E. Wang, Nanotechnology. 19, 1

(2008).15. X. Xu, J. Wang, K. Jiao, and X. Yang, Biosens. Bioelectron. 24,

3153 (2009).16. L. Li and B. Li, Analyst. 134, 1361 (2009).17. W. J. Qi, D. Wu, J. Ling, and C. Z. Huang, Chem. Commun. 46,

4893 (2010).18. H. Chi, B. Liu, G. Guan, Z. Zhang, and M. Han, Analyst. 135, 1070

(2010).19. J. S. Lee, M. S. Han, and C. A. Mirkin, Angew. Chem. Int. Ed. 46,

4093 (2007).20. J. Wu, L. Li, D. Zhu, P. He, Y. Fang, and G. Cheng, Anal. Chim.

Acta. 694, 115 (2011).21. L. Li, B. Li, Y. Qi, and Y. Jin, Anal. Bioanal. Chem. 393, 2051

(2009).22. C. C. Huang and H. T. Chang, Chem. Commun. 12, 1215 (2007).23. Y. Guo, Z. Wang, W. Qu, H. Shao, and X. Jiang, Biosens. Bio-

electron. 26, 4064 (2011).24. B. Roy, A. Saha, and A. K. Nandi, Analyst. 136, 67 (2011).25. Y. B. Ruan, A. F. Li, J. S. Zhao, J. S. Shen, and Y. B. Jiang, Chem.

Commun. 46, 4938 (2010).26. C. W. Liu, Y. T. Hsieh, C. C. Huang, Z. H. Lin, and H. T. Chang,

Chem. Commun. 19, 2242 (2008).27. G. Frens, Nat. Phys. Sci. 241, 20 (1973).28. X. Chen, Y. Zu, H. Xie, A. M. Kemas, and Z. Gao, Analyst. 136,

1690 (2011).

Received: 31 August 2011. Accepted: 2 December 2011.

3010 J. Nanosci. Nanotechnol. 12, 3006–3010, 2012