Facet dependent binding and etching: Ultra-sensitive colorimetric visualization of blood uric acid by unmodified silver nanoprisms

Biosensors and Bioelectronics 59 (2014) 227–232

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Facet dependent binding and etching: Ultra-sensitive colorimetric visualization of blood uric acid by unmodified silver nanoprisms Kanghui Tan, Guang Yang, Huide Chen, Pengfei Shen, Yucheng Huang n, Yunsheng Xia n Key Laboratory of Functional Molecular Solids, Ministry of Education, Center for Nano Science and Technology, College of Chemistry and Materials Science, Anhui Normal University, Wuhu 241000, China

art ic l e i nf o

a b s t r a c t

Article history: Received 4 January 2014 Received in revised form 22 March 2014 Accepted 24 March 2014 Available online 3 April 2014

By combination of experiments and density functional theory calculations, we present a simple but effective “facet dependent binding and etching” strategy for non-enzymatic and non-aggregated colorimetric sensing of blood uric acid (UA), using unmodified Ag nanoprisms as the signal readout. In the absence of UA, the triangular Ag nanoprisms are etched alongside (1 1 0) facets by H2O2 and form round nanodiscs, and a more than 160 nm surface plasmon resonance (SPR) blue shift is observed. Because of special affinity between UA and side facets of the Ag nanoprisms, pre-added UA can well protect the Ag nanoprisms from etching. Such protection effect can be used for well quantifying UA in the range of 10–3000 nM, based on the inverse proportion of the SPR blue shift with the added analyte. Due to very thin plate morphology (5 nm) and facet dependent binding/etching effects of the Ag nanoprisms, the sensing system has ultrahigh sensitivity. The detection limit is only 10 nM, which is about 2 to 4 orders of magnitude lower than that of previous colorimetric sensing systems. In addition to accurate quantitation, the proposed strategy can conveniently discriminate the patient of hyperuricemia from normal person by naked eyes. So, the present simple, low-cost and visualized UA chemosensor has great potential in the applications for point-of-care diagnostics. & 2014 Elsevier B.V. All rights reserved.

Keywords: Colorimetric visualization Blood uric acid Silver nanoprisms Facet-dependent properties

1. Introduction Hyperuricemia is the level of uric acid (UA) in blood that is abnormally high. Hyperuricemia causes several diseases like gout, arthritis, Lesch–Nyhan syndrome, and it is also associated with high risk of type 2 diabetes, hypertension, stroke and cardiovascular diseases (Bos et al., 2006; Dehghan et al., 2008; Kannan and John, 2009). Because of rapid increasing trend (In the United States alone, the prevalence of gout among adults in 2007–2008 was 3.9% (8.3 million individuals), which was only 0.275% in the 1980s (Schumacher, 1988; Zhu et al., 2011), hyperuricemia has been considered one of the biggest public health threats following hypertension, diabetes and hyperlipidemia (Becker and Jolly, 2006). For preventing from hyperuricemia and the related diseases, it is important to frequent monitoring and tight control of blood UA level. To date, various UA assay systems have been designed based on different physicochemical principles, for example, molecular spectroscopy, electrochemistry, high performance liquid chromatography, capillary electrophoresis, etc (Bera et al.,

n

Corresponding authors. Tel.: þ 86 553 3869303; fax: þ 86 553 3869303. E-mail addresses: [email protected] (Y. Huang), [email protected] (Y. Xia). http://dx.doi.org/10.1016/j.bios.2014.03.048 0956-5663/& 2014 Elsevier B.V. All rights reserved.

2011; Huang et al., 2008; Martinez-Pérez et al., 2003; Wang et al., 2000; Zhou et al., 2013). Among them, colorimetry has attracted special interest because of its simplicity and practicality (Kim et al., 2011). Traditionally, serum UA has been colorimetrically detected by the reduction of phosphotungstic acid (Kageyama, 1971). Unfortunately, in addition to tedious deproteinization processes, the coexisting reducing substances, such as ascorbic acid, strongly interfere the detection. As an effective alternative, enzyme based sensing strategy can exclusively detect UA without deproteinization processes (Chen et al., 2010). However, enzyme is costly and delicate (Pollak et al., 1980). So, it is urgent to explore reliable, simple, low-cost and robust colorimetric chemosensors for UA assay. Noble metal (e.g., Au and Ag) nanoparticles (NPs) are attractively colorimetric reporters because of their excellent optical properties (Stewart et al., 2008; Xie et al., 2012; Zhou et al., 2008). One of the most interesting features is their distance dependent surface plasmon resonance (SPR) absorption. For example, as 15 nm spherical Au NPs aggregate, the initial 520 nm SPR peak decreased and a new absorption appeared at longer wavelength (600–800 nm). At the same time, the solution color turns from wine to blue. To date, such aggregation/deaggregation modulation based sensing mode has been well accepted and widely used for myriad analyte targets with the assistance of

228

K. Tan et al. / Biosensors and Bioelectronics 59 (2014) 227–232

versatile surface modification chemistry. However, the suspension of the aggregates is not stable due to the increased particle size and reduced surface repelling force, which tends to precipitate and leads to the color of the suspension diminishes or even disappears (Zhou et al., 2008); furthermore, the aggregation can also be caused by non-specific effects, especially for some complex bio-samples (Aslan et al., 2004). Obviously, these factors would prevent accurate analyte target determination. Recently, anisotropic noble metal NPs have attracted considerable attention ranging from controllable synthesis to application exploration (Huang and Lin, 2012). Because of strong shape dependent SPR wavelength, anisotropic Au (or Ag) NPs are promising to be used in the design of wavelength-variations sensing platforms based on their specific reactions with analyte targets. To date, the approaches of etching (Chen et al., 2013; He et al., 2012; Liu et al., 2013; Xia et al., 2013), etching protection (Malile and Chen, 2013), and epitaxial growth (Rodríguez-Lorenzo et al., 2012) of anisotropic NPs have been employed for different target sensing. Such sensing mode may facilitate to overcome above problems, because the signal transduction is dependent on the particle morphology change instead of aggregation/deaggregation transformation. On the other hand, anisotropic NPs possess well-defined facet-dependent physicochemical properties (including photocatalytic, catalytic, electrical, and molecular adsorption properties, and so on) (Huang and Lin, 2012). The facet-dependent behaviors have been used for shape controlled synthesis and various catalytic applications (Huang and Lin, 2012). Therefore, a conceptual demonstration of a sensing platform, in which analyte targets can be colorimetrically visualized by combination of shapedependent SPR wavelength and facet-dependent properties of anisotropic NPs, will be significant for both fundamental research and applications. Such platform taking full advantage of optical, surface and structural properties of anisotropic noble metal NPs, would bring us new application fields and better sensing performances. Herein, we propose a “facet dependent binding and etching” strategy for non-enzymatic and non-aggregated colorimetric sensing of blood UA with unmodified Ag nanoprisms as the signal readout. The assay principle is different from previous study and shown in Scheme 1: In the absence of UA, the triangular Ag nanoprisms are well etched along side (1 1 0) facets by H2O2 and form round nanodiscs, and a more than 160 nm SPR blue shift is observed. The added UA molecules just selectively bind to the side facets, which well protects the Ag nanoprisms from etching. Furthermore, the SPR blue shift of the Ag nanoplates is inversely proportional to the added analyte, which can be used for UA quantification ranging from 10 to 3000 nM. The detection limit is as low as 10 nM. In addition to strongly shape dependent SPR wavelength, the present ultrahigh sensitivity results from three reasons: (1) H2O2 etching proceeds along to the Ag nanoprism side faces; (2) UA preferentially binds to side (1 1 0) facets of the Ag nanoprisms; (3) the very thin plate morphology of the Ag

Scheme 1. Schematic illustration of “facet dependent binding and etching” strategy for colorimetric sensing of UA. .

nanoprisms leads to their very small proportion of side area. Impressively, the proposed method can conveniently distinguish hyperuricemia patient from normal person by naked eyes. The whole processes, from Ag nanoprism synthesis to UA assay, need not any sophisticated equipments, expensive reagents, timeconsuming particle surface modification, tedious sample pretreatments, as well as high technical demands. So, the proposed visualized, low-cost and simple UA chemosensor has great potential in the applications for point-of-care diagnostics.

2. Materials and methods 2.1. Reagents and materials Sodium citrate tribasic dehydrate (99%) was purchased from Shanghai Lingfeng Chemical Reagent Co., Ltd. Poly (vinylpyrrolidone) (PVP, Mw E10, 000), uric acid (UA, 99%), glucose (99%), dopamine, glutathione, urea, hypoxanthine, adenine, xanthine and uricase were obtained from Sigma-Aldrich. Horse radish peroxidase (HRP) was purchased from Aladdin. Sodium borohydride (NaBH4, 96%), Silver nitrate (AgNO3, 99.8%), Hydrogen peroxide (H2O2, 30 wt%), ascorbic acid (99%), various amino acids, tetramethylbenzidine (TMB, 99%) and other routine chemicals were acquired from Sinopharm. Serum samples were donated from Wuhu Hospital of TCM. All solutions were prepared with ultrapure water.

2.2. Apparatus A Hitachi-U-3010 spectrometer was used to record the UV– visible spectra. Characterizations of scanning electron microscopy (SEM) were carried out on Hitachi S-4800 under the accelerating voltage of 10 kV. The samples for SEM measurements were prepared by the deposition of one drop of dilute aqueous dispersion on a silicon substrate, and the solvent was removed by evaporation in air. Characterizations of transmission electron microscopy (TEM) were carried out on Tecnai G2 F20 S-Twin transmission electron microscope under the accelerating voltage of 200 kV. The samples for TEM measurements were prepared by the deposition of one drop of aqueous dispersion on a copper grid coated with thin films of carbon, and the solvent was removed by evaporation in air. To valid the accuracy of the proposed method, the Olympus AU 400 autoanalyzer was used for UA assay.

2.3. Procedures for UA sensing The Ag nanoprisms were synthesized by the procedure reported previously (Zhang et al., 2011) (The details for synthesis process were presented in Supplementary materials). To a series 5 mL calibrated test tubes, 360 μL of crude Ag nanoprisms and 20 μL of UA standard solutions with different concentrations (or pre-diluted serum samples) were diluted to 2 mL with ultrapure water; then the mixed solutions were transferred separately into a 1 cm quartz cuvette. After 15 min, 40 μL of 2 mM H2O2 were rapidly added. For real sample assay, all the serum samples were pre-diluted 5 times by 0.01 M phosphate buffer solution (pH 7.4). All measurements were performed at room temperature. For interference experiments, various amino acids, glutathione, urea, hypoxanthine, adenine, xanthine, ascorbic acid, dopamine, glucose and their mixture were pre-mixed with UA, which were then separately added to the Ag nanoprism solutions. Afterwards, the etching reactions were conducted under the conditions identical to those described above.

K. Tan et al. / Biosensors and Bioelectronics 59 (2014) 227–232

2.4. Calculations The density functional theory (DFT) calculations were performed with the Vienna ab initio simulation package code that is particularly efficient with periodic metallic systems (Craig et al., 2005; Duncan et al., 2007; Kresse and Furthmuller, 1996a, 1996b; Kresse and Hafner, 1993), which were described in detail in Supplementary materials.

3. Results and discussion 3.1. Protecting effects of UA on H2O2 etching Fig. S1A is a representative low-magnification SEM image of the as-made sample, indicating that the products are well-defined triangle nanoprisms. Their mean side length is about 50 nm, and the thickness is only 5 nm. Herein, the ultrathin plate morphology causes very limited side area of the Ag nanoprisms, which is especially significant to the high sensitivity of UA assay, as described below. Fig. S1B displays the SPR absorption spectra of three different batches of Ag nanoprisms, which exhibit almost identical SPR profiles. Such highly controllable synthesis is beneficial to the applications of the Ag nanoprisms. From short to longer wavelength range (330, 472, 714 nm), the three SPR peaks are corresponding to out-of-plane quadrupole, in-plane quadrupole and in-plane dipole plasmon resonance of triangular nanoplates, respectively (Zhang et al., 2011). The peak position of the in-plane dipole resonance is extremely sensitive to the shape of Ag nanoplates, which can be employed to the design of wavelength-variations based chemosensors (Chen et al., 2013; Xia et al., 2013). As 40 μM H2O2, one of effectively etching agents for Ag NPs (He et al., 2012; Xia et al., 2013), was added, the Ag nanoprism solution turned from cyanine to mauve very rapidly (Inset of Fig. 1A, from right to left), and the color almost completely disappear after 15 min reaction. Such evolution behaviors could also be quantified by absorption spectra. As described in Fig. 1A, the SPR peak shifts from 706 to 546 nm within 60 min, and that its intensity decreases by 77%. Interestingly, as 1000 nM UA was pre-added, the case was rather different. As shown in Fig. 1B, the absorption spectra only exhibit very small changes (68 nm peak shifts, 22% intensity decrease) at same conditions. As a consequence, the solution color only turns from cyanine to blue within 60 min (Inset of Fig. 1B). The microcosmic difference of the two processes has been directly observed by SEM technique. In the absence of UA, the 50 nm triangle nanoprisms are well etched to 20 nm round nanodiscs (Fig. 1C). In contrast, for UA pre-added system, the morphology of the Ag nanoprisms kept rather well, even distinctly triangular

229

nanoplates can be easily found (Fig. 1D). These results clearly indicated that UA could effectively protect the Ag nanoprisms from H2O2 etching (Xia et al., 2013). We then used absorption technique to study the effects of UA concentrations on the etching dynamics of the Ag nanoprisms. As shown in Fig. 2A, in the absence of UA, the in-plane dipole resonance band linearly shifts to blue range in the first 5 min. Such shift rate is very fast and the slope is as high as 11.62. Then, the blue shift gradually decreases and levels off to a saturation value after 25 min. As UA was pre-added, the etching rate was decreased and the linear shift time was consequently extended. Furthermore, the more addition of UA, the better protection effect of the Ag nanoprisms. For example, the slope decreased to only 0.62 as the added UA reached 3000 nM. Intuitively, the slopes vs. UA concentrations can be used for the quantification of the analyte (Fig. S8). However, it exhibits rather large relative errors (15–20%). For obtaining an acute quantification, we further investigated the SPR shift (Δλ) of the Ag nanoprisms vs. UA concentrations (Fig. 2B). As described in Fig. 2C, the relationship between Δλ and UA concentrations is obviously nonlinear, the increase rate of Δλ dramatically decreases as UA concentration 4100 nM. However, c1/3 and Δλ exhibits a good linear relationship, which can be conveniently quantified UA concentrations ranging from 10 to 3000 nM (Inset of Fig. 2C), and relative errors are only 5–10%. In addition, as shown in the inset of Fig. 2B, the solutions exhibit distinctly different colors (cyanine, blue, purple and mauve) depending on the amounts of added analyte. Such wavelength shift based sensing strategy is not only more reliable for quantification, but promising for visualized UA assay by naked eyes. It is very worthy to note that UA can be well quantified even the concentration is as low as 10 nM (Fig. 2B and C), as far as we know, the detection limit is about 2 to 4 orders of magnitude lower than that of previous colorimetric sensing systems, and even 2 times lower than that of the most sensitive fluorescent sensor (Table. S1). 3.2. DFT calculations It is interesting to ask why only 10 nM UA causes a so distinct protecting effect for the etching (slope: from 11.62 to 8.29; Δλ: from 140 to 124 nm, 35 min reaction). In the assay experiments, the used Ag nanoprism concentration was about 56 pM (Zhang et al., 2011). So, for 10 nM UA sensing system, the molar ratio of the Ag nanoprism and the analyte was 1:180. Taking into account 35– 40 thousands of silver atoms on each plate surface, it is difficult to understand that less than one percent of UA can play so distinct role for the protection. To answer this question, DFT calculations have been performed to try to discern the interaction of UA and the Ag nanoprisms. Several matters should be clarified before

Fig. 1. Time-dependent SPR absorption spectra of the Ag nanoprisms upon incubating with H2O2 (40 μM) in the absence (A) and presence (B) of 1000 nM UA. SEM image of the Ag nanoplates etched by 1 h in the absence (C) and presence (D) of 1000 nM UA. The insets in (C) and (D) show the TEM images of the corresponding products. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

230

K. Tan et al. / Biosensors and Bioelectronics 59 (2014) 227–232

calculation. First of all, the as-prepared Ag nanoprisms are single crystalline with facet-centered cubic (fcc) lattice structures (Millstone et al., 2009). Furthermore, the two triangular bases are (1 1 1) facets, while the three side faces are dominated by (1 1 0) facets (Shahjamali et al., 2013). Then, UA can bind to the Ag nanoplate surface as they mix with each other, as indicated by a small SPR bathochromic shift of the Ag nanoprisms (The binding of UA leads to the increase of local dielectric constant around the Ag nanoplates and results in the SPR bathochromic shift, Fig. S9). Third, in the experimental conditions (pH ¼8.0), almost all UA molecules ionize one proton and form urate (pKa1 and pKa2 of UA are 5.40, and 9.80, respectively (Simic and Jovanovic, 1989)). Finally, based on SEM observation, H2O2 etching proceeds along to the Ag nanoprism side faces and the tips, because the products are round nanodiscs (If the two base faces are etched, the holes at triangular faces would appear (Shahjamali et al., 2012)). With these considerations in mind, the adsorption properties of urate on Ag(1 1 1) and Ag(1 1 0) surfaces are investigated with the Vienna ab initio simulation package code (Kresse and Furthmuller, 1996a, 1996b; Kresse and Hafner, 1993). It has been revealed that ionization from N3 position produces nearly the same stability as the N9 (Smith et al., 1988). Thus, two possible ionized forms of UA are considered upon adsorption on Ag surfaces. Interestingly, both (1 1 1) facets and (1 1 0) steps have their own adsorption preference, namely, on (1 1 1) surface, the binding energy of N9urate is 0.15 eV lower than that of N3-urate; while on (1 1 0) surface, N9-urate is 0.15 eV higher than N3-urate. Fig. 3 illustrates the most stable configurations of ionized UA adsorption on Ag surfaces, which indicates that both N3- and N9-urates adopt an “upright” structure with respect to the Ag surface through two adjacent atoms of N and O. Other possible adsorption configurations combined with the binding energies are also conducted, as shown in Fig. S10. The binding energies of N3- and N9-urates on Ag(1 1 1) and Ag(1 1 0) are calculated to be 1.38 and 1.70 eV, respectively. According to the Boltzmann–Gibbs statistics at a finite temperature T, the binding energy determines the density of the ionized UA on the given surface, and the ratio of the densities at a pair of surfaces equals ð1 Nð1 1 1Þ expðEb ¼ Nð1 1 0Þ expðEð1 b

Fig. 2. (A) Etching dynamics of the Ag nanoprisms in the different concentrations of pre-added UA. (B) Normalized SPR absorption spectra the Ag nanoprism incubating with 40 μM H2O2 in the presence of various concentrations of UA for 35 min. The inset shows a photograph of the corresponding reaction solutions. (C) Plots of SPR peak shift (Δλ) vs. concentrations of added UA. The inset of (C) exhibits linear relationship of Δλ and c1/3. Δλ ¼ λ0  λ, λ0 is the SPR peak of the original Ag nanoprisms, λ is the SPR peak of the Ag nanoprisms etched by H2O2 with the protection by different concentration of UA. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

1 1Þ

=kB TÞ

1 0Þ

=kB TÞ

¼ 3:86  10  6 ðT ¼ 298 KÞ;

where kB is the Boltzmann constant. Thus, UA prefers to ionize its proton at N9 position and to adsorb on more open surface of Ag (1 1 0). As a further support, coverage effects have been also performed to evaluate whether the preferential binding still holds true when the concentration of adsorbates increases. As seen in Fig. 3, the binding energy takes on a decrease on both two surfaces with the coverage increase, however, their decay tendency is rather different. The binding energy fades to 0 eV for N3-urate/ Ag(1 1 1) adsorption complex at 1/3 monolayer (ML); in contrast, there is still 0.83 eV for N9-urate/Ag(1 1 0) even at 1/2 ML coverage. This phenomenon further indicates that ionized UA preferential bind to Ag(1 1 0) surface. To understand the reason why the binding energy attenuation is different, the charge density differences of N3-urate/Ag(1 1 1) system at 1/3 ML coverage and N9urate/Ag(1 1 0) at 1/2 ML coverage were plotted, as shown in Fig. S11. It can be clearly seen that the charge redistribution occurs on the whole molecule of N3-urate when it adsorbs on Ag(1 1 1) at 1/ 3 ML coverage. Moreover, the charge accumulates and depletes on two opposite sides of the molecular plane, resulting in attraction interaction between two adjacent molecules (Fig. S11A). It was evidenced by the fact that the O-, N–Ag distances elongate to 2.44 and 2.41 Å, which are pronounced longer than the corresponding ones on surfaces at 1/6, 1/4 ML coverage. At variance, the charge

K. Tan et al. / Biosensors and Bioelectronics 59 (2014) 227–232

231

Fig. 3. Binding energies as a function of surface coverage. (A) N9-urate on Ag(1 1 0) surface and (B) N3-urate on Ag(1 1 1). The stepped Ag atoms are highlighted with orange. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

was symmetrically distributed along the two sides of molecular plane, indicating that it was less disturbed by the neighboring molecules of N9-urate when it adsorbs on Ag(1 1 0) at 1/2 ML coverage. It should be noted that the above calculation is based on clean Ag surface. In fact, citrate ions were used as capping agents in the Ag nanoprism preparation. Based on literature, citrate can bind more strongly to (1 1 1) facets than other ones of fcc Ag (Zeng et al., 2010). Obviously, the adsorption of citrate can effectively prevent UA from binding on the base (1 1 1) facets of the Ag nanoprisms. As a result, UA should have more obvious tendency to bind to the side (1 1 0) facets. On the other hand, as shown in Fig. S1A, the side length of the Ag nanoprisms is about 50 nm, and the thickness is as low as 5 nm. Geometrically, the side face area only takes 25.7% in the total surface area. Based on above analysis, the present system processes three features: (1) H2O2 etching proceeds along to the Ag nanoprism sides; (2) the very thin plate morphology of the Ag nanoprisms leads to their very small proportion of side area; (3) UA preferentially binds to side (1 1 0) facets of the Ag nanoprisms. So, even very small amounts of UA can play a significant role in the Ag nanoprism protection, which well prevents the Ag nanoprisms from etching and achieves ultrahigh sensitivity. 3.3. Selectivity In addition to sensitivity, selectivity is another key parameter in sensing. For the applications of blood UA assay, the potential interfering substances coexisting in serum, including various of amino acids, glutathione, urea, hypoxanthine, adenine, xanthine, ascorbic acid, dopamine, glucose had to be considered. Our previous study indicated several coexisting substances (proline, glutamic acid, tyrosine, ascorbic acid, and dopamine) at original concentrations in serum could prevent the Ag nanoprisms from H2O2 etching (Xia et al., 2013). However, as shown in Fig. 4, all the substances do not interfere UA determination as they are diluted by 200 times. These coexisting substances have N or S atoms, which would affect the etching/protection effects because they can also interact with Ag. However, they did not exhibit obvious interference in UA sensing after a sufficient dilution. The reason might be probably attributed to as follows: UA molecules have two binding sites, and they just well match with Ag(1 1 0) facets. Such multidentate binding effect is very strong (Behra et al., 2013) and can well protect the Ag nanoprisms from etching. While for the potential interfering substances, most of them have only one binding site (or their multiple binding sites do not match with Ag facets). On the one hand, the single binding effect has much lower affinity than that of multidentate ligand (UA). On the other hand, the interaction of single site binding and Ag might be

Fig. 4. Selectivity of Ag nanoprisms based system toward various potential interfering substances. The concentration of the used H2O2 is 40 μM. The concentrations of various of amino acids (from left to right) are 1.5, 0.5, 0.4, 0.6, 2.3, 1.4, 0.6, 0.8, 1.0, 0.1, 0.4, 1.3, 0.7, 0.8, 0.3, 0.4, 0.8, 1.25 μM; the concentrations of glutathione, urea, hypoxanthine, adenine, xanthine, ascorbic acid, dopamine and glucose are 5, 0.03, 0.02, 0.02, 0.01, 0.5 μM, 15 pM and 0.03 mM, respectively. The concentrations of potential interfering substances correspond to 200 times dilution of those in serum. The etching time is 60 min.

random, which did not have facet selectivity. As a consequence, the potential interfering substances could not effectively compete with UA and did not interfere the sensing at lower concentration levels. As described by Fig. 2B and C, the linear quantification range of the proposed method is 10–3000 nM. While the concentration levels of blood UA reach up to 100–400 μM (the levels of hyperuricemia patients are higher). So, dilution strategy is entirely feasible to eliminate these interferences. 3.4. Human serum sample assay Performances of the proposed sensing system for UA assay were tested in human serum. As we know, the concentration levels of blood UA are hundreds of μM, while the linear quantification range of the proposed method is 10–3000 nM. To avoid larger relative errors by extremely small volume of the used samples, all the serum samples were diluted 5 times by 0.01 M phosphate buffer solution (pH 7.4) before assay. Then, a 20 μL aliquot of pre-diluted serum samples was added in the Ag nanoprism solution, 15 min later, H2O2 was added and time dependent SPR shift was measured. As shown in Fig. 5A, five serum samples from normal people and hyperuricemia patients, were finally diluted 500 times, cause a gradual shift of the in-plane dipole band of the nanoplates. Their UA concentrations were then quantified by the SPR peak shift at 35 min reaction. The measured

232

K. Tan et al. / Biosensors and Bioelectronics 59 (2014) 227–232

potential application in point-of-care diagnostics, but well enriches anisotropic noble metal NP based sensing systems.

Acknowledgment This work is financially supported by the National Natural Science Foundation of China (21275001, Y.X.; 21203001, Y.H.), the Natural Science Foundation of Anhui Province (1208085QB20, Y.X.; 1208085QB37, Y.H.).

Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.03.048. References

Fig. 5. (A) SPR peak shift (Δλ) of the Ag nanoprisms incubating with five serum samples with different concentrations of UA. (B) The photographs of the Ag nanoprism based sensing systems for five samples. From left to right, the reaction times are 0, 2, 4, 8, 10, 20, 30, 45, 60 min. In the assay, all the samples are finally diluted to 500 times.

values of the five samples were 161, 240, 334, 416 and 698 μM, respectively, which were well in agreement with the results obtained by biochemistry analyzer (171, 229, 320, 443 and 732 μM) (Table. S2). In addition to accurate quantification, we also studied its visual detection performances. The etching processes were conducted and shown in Fig. 5B. It is found that the color change vs. etching time is highly dependent on UA contents. For normal people serum containing systems (the top three samples), the Ag nanoprisms exhibit an obvious color changes with reaction time (cyanine- blue- purple); in contrast, the colors of the Ag nanoprism solutions keep rather well for the systems containing the hyperuricemia patients’ serum (the bottom two samples). These preliminary results indicate that the proposed system cannot only accurately quantify blood UA but conveniently discriminate hyperuricemia patients from normal people by naked eyes. Because of simplicity and effectivity, the proposed UA sensor has great potential in the applications for point-of-care diagnostics, especially for the rural population in third world countries. 4. Conclusions In summary, a novel “facet dependent binding and etching” strategy has been proposed for blood UA assay, by using Ag nanoprisms as colorimetric reporters. It offered the advantages as follows: (1) simple; (2) sensitive; (3) practicable. So, the present work not only provides a UA colorimetric chemosensor for

Aslan, K., Zhang, J., Lakowicz, J.R., Geddes, C.D., 2004. J. Fluorescence 14, 391–400. Becker, M.A., Jolly, M., 2006. Rheum. Dis. Clin. North Am. 32, 275–293. Behra, R., Sigg, L., Clift, M.J.D., Herzog, F., Minghetti, M., Johnston, B., Petri-Fink, A., Rothen-Rutishauser, B., 2013. J. R. Soc. Interface 10, 20130396. Bera, R.K., Anoop, A., Raj, C.R., 2011. Chem. Commun. 47, 11498–11500. Bos, M.J., Koudstaal, P.J., Hofman, A., Witteman, J.C.M., Breteler, M.M.B., 2006. Stroke 37, 1503–1507. Chen, D., Wang, Q., Jin, J., Wu, P., Wang, H., Yu, S., Zhang, H., Cai, C., 2010. Anal. Chem. 82, 2448–2455. Chen, L., Fu, X., Lu, W., Chen, L., 2013. ACS Appl. Mater. Interfaces 5, 284–290. Craig, C.F., Duncan, W.R., Prezhdo, O.V., 2005. Phys. Rev. Lett. 95, 163001. Dehghan, A., Hoek, M., Sijbrands, E.J.G., Hofman, A., Witteman, J.C.M., 2008. Diabetes Care 31, 361–362. Duncan, W.R., Craig, C.F., Prezhdo, O.V., 2007. J. Am. Chem. Soc. 129, 8528–8543. He, H., Xu, X., Wu, H., Jin, Y., 2012. Adv. Mater. 24, 1736–1740. Huang, J., Liu, Y., Hou, H., You, T., 2008. Biosens. Bioelectron. 24, 632–637. Huang, M.H., Lin, P.-H., 2012. Adv. Funct. Mater. 22, 14–24. Kageyama, N., 1971. Clin. Chim. Acta 31, 421–426. Kannan, P., John, S.A., 2009. Anal. Biochem. 386, 65–72. Kim, H.N., Guo, Z., Zhu, W., Yoon, J., Tian, H., 2011. Chem. Soc. Rev. 40, 79–93. Kresse, G., Furthmuller, J., 1996a. Comput. Mater. Sci. 6, 15–50. Kresse, G., Furthmuller, J., 1996b. Phys. Rev. B: Condens. Matter 54, 11169–11186. Kresse, G., Hafner, J., 1993. Phys. Rev. B: Condens. Matter 47, 558–561. Liu, X., Zhang, S., Tan, P., Zhou, J., Huang, Y., Nie, Z., Yao, S., 2013. Chem. Commun. 49, 1856–1858. Malile, B., Chen, J.I.L., 2013. J. Am. Chem. Soc. 135, 16042–16045. Martinez-Pérez, D., Ferrer, M.L., Mateo, C.R., 2003. Anal. Biochem. 322, 238–242. Millstone, J.E., Hurst, S.J., Métraux, G.S., Cutler, J.I., Mirkin, C.A., 2009. Small 5, 646–664. Pollak, A., Blumenfeld, H., Wax, M., Baughn, R.L., Whitesides, G.M., 1980. J. Am. Chem. Soc. 102, 6324–6336. Rodríguez-Lorenzo, L., Rica, R., Álvarez-Puebla, R.A., Liz-Marzán, L.M., Stevens, M. M., 2012. Nat. Mater. 11, 604–607. Schumacher, H.R., 1988. Primer on the Rheumatic Disease, ninth ed. William Byrd press, USA. Shahjamali, M.M., Bosman, M., Cao, S., Huang, X., Cao, X., Zhang, H., Pramana, S.S., Xue, C., 2013. Small 9, 2880–2886. Shahjamali, M.M., Bosman, M., Cao, S., Huang, X., Saadat, S., Martinsson, E., Aili, D., Tay, Y.Y., Liedberg, B., Loo, S.C.J., Zhang, H., Boey, F., Xue, C., 2012. Adv. Funct. Mater. 22, 849–854. Simic, M.G., Jovanovic, S.V., 1989. J. Am. Chem. Soc. 111 (5778–5182). Smith, R.C., Gorea, J.Z., McKee, M., Hargis, H., 1988. Microchem. J. 38, 118–124. Stewart, M.E., Anderton, C.R., Thompson, L.B., Maria, J., Gray, S.K., Rogers, J.A., Nuzzo, R.G., 2008. Chem. Rev. 108, 494–521. Wang, J., Chatrathi, M.P., Tian, B., Polsky, R., 2000. Anal. Chem. 72, 2514–2518. Xia, Y., Ye, J., Tan, K., Wang, J., Yang, G., 2013. Anal. Chem. 85, 6241–6247. Xie, X., Xu, W., Liu, X., 2012. Acc. Chem. Res. 45, 1511–1520. Zeng, J., Zheng, Y., Rycenga, M., Tao, J., Li, Z.-Y., Zhang, Q., Zhu, Y., Xia, Y., 2010. J. Am. Chem. Soc. 132, 8552–8553. Zhang, Q., Li, N., Goebl, J., Lu, Z., Yin, Y., 2011. J. Am. Chem. Soc. 133, 18931–18939. Zhou, S., Zuo, R., Zhu, Z., Wu, D., Vasa, K., Deng, Y., Zuo, Y., 2013. Anal. Methods 5, 1307–1311. Zhou, Y., Wang, S., Zhang, K., Jiang, X., 2008. Angew. Chem. Int. Ed. 47, 7454–7456. Zhu, Y., Pandya, B.J., Choi, H.K., 2011. Arthritis Rheum. 63, 3136–3141.