High-Throughput Sulfide Sensing with Colorimetric Analysis of Single

May 8, 2014 - ... detection of methylamphetamine with surface enhanced Raman scattering ..... Journal of Materials Chemistry C 2015 3 (23), 6035-6045 ...
0 downloads 0 Views 273KB Size
Download PDF
Technical Note pubs.acs.org/ac

High-Throughput Sulfide Sensing with Colorimetric Analysis of Single Au−Ag Core−Shell Nanoparticles Jinrui Hao, Bin Xiong, XiaoDong Cheng, Yan He,* and Edward S. Yeung State Key Laboratory of Chemo/Biosensing and Chemometrics, College of Chemistry and Chemical Engineering, College of Biology, Hunan University, Changsha, Hunan 410082, P. R. China S Supporting Information *

ABSTRACT: We present a high-throughput strategy for sensitive detection of H2S by using individual spherical Au−Ag core−shell plasmonic nanoparticles (PNPs) as molecular probes. This method is based on quantification of color variation of the single PNPs resulting from formation of Ag2S on the particle surface. The spectral response range of the 51 nm PNP was specifically designed to match the most sensitive region of color cameras. A high density of immobilized PNPs and rapid color RGB (red/green/blue) analysis allow a large number of individual PNPs to be monitored simultaneously, leading to reliable quantification of color change of the PNPs. A linear logarithmic dependence on sulfide concentrations from 50 nM to 100 μM was demonstrated by using this colorimetric assay. By designing PNPs with various surface chemistries, similar strategies could be developed to detect other chemically or biologically important molecules.

I

accompanied by color change. Moreover, with the availability of sensitive digital color cameras nowadays, the readout of the colors of single nanoparticles could be much more efficient than obtaining their individual spectra.11 For example, with color RGB (red/green/blue) analysis of the colors of plasmonic PNPs, Liu studied binding of thiols to gold nanoparticles (AuNPs).12 By mapping RGB values onto the chromaticity diagram, Jing successfully measured the diameters of single AuNPs.13 However, high-throughput single particle colorimetric analysis has not been reported so far. Although hydrogen sulfide (H2S) is a toxic gas with the disagreeable rotten egg smell, it is widely distributed in the human body and other biological systems, as well as in the ecosystem, and is an important index of the environment14 and one of the gaseous transmitters of signaling molecules in biology.15 During the past decade, various methods based on colorimetry, chromatography, etc. have been developed for H2S determination in the ensemble environment.16,17 Several organic fluorescent probes have been synthesized for monitoring intracellular sulfide with sensitivity down to μM level.18−20 To map local distribution of H2S in living cells, we have developed a highly sensitive H2S detection method by using single gold nanorod−silver (AuNR−Ag) core−shell PNPs as probes.10 With the transmission grating-based single particle spectral imaging method, we demonstrated real-time intracellular H2S detection with nM sensitivity. However, due to the restraint imposed by the grating, both the number of single PNP spectra that can be acquired in one image and the speed of spectral data analysis were very limited.

n recent years, single-particle spectral analysis based on the localized resonance scattering of metal plasmonic nanoparticles (PNPs) has attracted considerable attention due to promising applications in biological and chemical analysis.1,2 Compared to fluorescent molecules or quantum dots, PNPs are highly photostable and do not suffer photobleaching or intensity fluctuations. A plasmonic PNP, depending on its size, could be thousands to millions times brighter than a single fluorophore, allowing easy observation of single PNPs with dark-field microscopy. The spectral profiles of PNPs are highly dependent on their size, shape, composition, and surrounding environment. All these advantages allow individual plasmonic PNPs to be used as single optical spectral imaging probes.3 In particular, since every single PNP in the field-of-view can serve as an independent sensor with high signal-to-noise ratio (SNR), spatially resolved high-throughput detection could potentially be accomplished.4−6 However, the most commonly used single PNP spectral imaging setup is a dark-field microscope equipped with a spectrograph. Due to the necessity of an entrance slit, usually only the plasmonic scattering spectra of a thin slice of stationary nanoparticles can be obtained each time, which puts a strong limit on the throughput of single particle spectral measurement. To tackle this problem, we have developed a simple technique by placing a transmission grating (TG) before the CCD (charge coupled device) camera, which splits the optical image of each molecule into a zero-order focused spot and a first-order spectrally dispersed long streak, allowing the spectra of multiple single fluorescent molecules7,8 or plasmonic PNPs9,10 to be acquired simultaneously in real time. However, this method requires time-consuming, manual identification of first-order spectra and is not applicable in high particle density situations. On the other hand, it is well-known that PNPs with plasmonic absorption and scattering peaks in the visible region exhibit various colors, and spectral shifts of PNPs are often © 2014 American Chemical Society

Received: February 2, 2014 Accepted: April 28, 2014 Published: May 8, 2014 4663

dx.doi.org/10.1021/ac500376e | Anal. Chem. 2014, 86, 4663−4667

Analytical Chemistry

Technical Note

Figure 1. (A) Schematics of single particle sensing with dark-field microscopy and colorimetric analysis. (B) TEM images of the as-prepared spherical AuNP−Ag core−shell PNPs. (C) Normalized UV−vis spectra of core AuNPs (green), AuNP−Ag core−shell PNPs (blue), and AuNP−Ag2S core−shell PNPs (red). (D) Normalized UV−vis spectra of ∼12.5 pM AuNP−Ag core−shell PNP solution after adding 1.0, 10, 100, and 1000 μM of NaHS. (E) Selectivity and time-response tests. Legend: (1) HS− (10 μM), (2) cysteine, (3) glutathione, (4) S2O32−, (5) SO32−, (6) SCN−, (7) NO3−, (8) NO2−, (9) Cl−, (10) Br−, (11) I−, (12) CH3COO−, and (13) CO32−. The concentrations of these anions are all 100 μM unless indicated otherwise. The inset is the time-dependent increase of Δλmax after addition of 100 μM NaHS. (F) The color response regions of the spherical AuNP−Ag PNPs, 1, and the rod-shaped AuNR−Ag PNPs, 2, in the CIE 1931 colorimetric diagram.

ethanol solution for 3 h, followed by rinsing with ethanol several times and drying under a stream of N2. Finally, the silanized coverslip was functionalized with the AuNP−Ag core−shell PNPs by adding a drop of the diluted PNP solution of ∼34.5 nM onto the glass surface for ∼8 min, followed by rinsing with water to remove non-adsorbed PNPs and drying under a stream of N2. Dark-field imaging was performed using a Nikon 80i upright microscope equipped with a 100W halogen lamp, an oil immersion dark-field condenser, a 60× dark-field objective, and an Olympus DP72 color CCD camera. For sulfide detection, flow channels with a width of 8 mm and a total volume of 20 μL were made by using one PNP functionalized coverslip and one 22 × 50 mm coverslip with double-sided adhesive tapes. The CCD exposure time was 800 ms. All images were processed using ImageJ or Origin.

Herein, we improve the throughput of our PNP-based H2S sensing scheme by using RGB colorimetric analysis with spherical AuNP−Ag core−shell PNPs as the single particle probes. The reason that we choose spherical PNPs rather than the previous rod-shaped PNPs is for matching the spectral maximum of the probes with the most color-sensitive region of the color CCD camera. With simple and fast RGB analysis under dark-field microscopy, a large number of PNPs could be monitored simultaneously, allowing highly sensitive and highthroughput H2S sensing. By changing the surface chemistry of the PNPs, similar colorimetric assays could be developed to detect DNAs, proteins, and other small molecules.

■

EXPERIMENTAL SECTION The 50 nm AuNPs acting as the core were synthesized using a seed-mediated method from 18 nm AuNP seeds.21 In brief, 0.500 mL of 18 nm AuNP seeds, 0.165 mL of 2.428 × 10−2 M HAuCl4, and 0.24 mL of 0.01 M MSA (2-mercaptosuccunic acid) were sequentially added into 20 mL of DI water. The growth process lasted for 2 h. For the growth of Ag shell, 1.0 mL of the as-prepared MSA-modified AuNP solution was centrifuged twice at 6000 rpm and resuspended in 1.0 mL of DI H2O. Then, 5.0 mL of purified AuNP solution was mixed with 5.0 mL of 3% PVP (polyvinylpyrrolidone, MW 29 000) solution. After that, 10 μL of a freshly prepared 0.01 M [Ag(NH3)2]NO3 solution was added to the mixture while stirring, followed by addition of 50 μL of 0.01 M ascorbic acid. Finally, the solution was placed in the dark and kept undisturbed for 3−4 h. The freshly prepared AuNP−Ag core−shell PNPs were characterized using TEM (transmission electron microscopy) and UV−vis spectrometry. For immobilization of the PNPs onto the glass substrate, the coverslips (22 × 22 mm, Corning) were cleaned using chromic acid lotion to remove organic residues, followed by sonication in DI water for at least 3 times to remove excess dusts. After drying in an oven for 5−6 h at 80 °C, the coverslips were incubated in 0.1% v/v APTMS (3-aminopropyltriethoxysilane)

■

RESULTS AND DISCUSSION The basic principle of H2S detection using AuNP−Ag core− shell PNPs, in which the Au core acts as the signal reporter and the Ag shell serves as the sensing agent, has been described in detail previously10 (see the supplementary note and Figure S1 in the Supporting Information). In short, in the presence of O2, silver reacts with sulfide to form Ag2S. The large refractive index difference between Ag (∼0.17) and Ag2S (∼2.2) results in redshift of the PNP plasmonic spectrum, allowing highly sensitive detection of H2S. Figure 1A shows the schematic diagram of the optical setup. For colorimetric analysis, it is desirable that the target-induced spectral shifts lead to visually detectable color change, preferentially in the yellow-green region of 500−600 nm, the most sensitive region of human color perception, but AuNRs prepared with the conventional seed-mediated growth methods usually have a longitudinal spectral maximum larger than 600 nm. Thus, we decided to use AuNP spheres as the signal-reporter core, whose plasmonic peaks fall in the range of 510−580 nm depending on the size of spheres, though their spectral sensitivity toward local environ4664

dx.doi.org/10.1021/ac500376e | Anal. Chem. 2014, 86, 4663−4667

Analytical Chemistry

Technical Note

∼270 per 100 μm2. For the 60× objective and the DP72 color CCD camera with 1360 × 1024 pixels, that means the color change of ∼1000 single PNPs can be monitored simultaneously. In contrast, when a grating-based method is used, to avoid significant spectra overlap between different spots, no more than ∼50 single particles/molecules can be tracked each time. Figure 2A show typical dark-field images of the single

mental change is not as good as the AuNRs.22 The appropriate AuNP−Ag nanoprobes should appear green under dark-field microscopy and exhibit a green to yellow/red color change upon addition of sulfide. To determine the optimal size of the AuNP−Ag PNPs for colorimetric detection of sulfide, we performed a simulation on the color images of the PNPs according to the following formula23 (Figure S2 in the Supporting Information) 700

∫380

I(λ) × S(λ) × C(λ)dλ

where I(λ) is the spectral power distribution of the light source, S(λ) is the plasmonic scattering spectrum of AuNP−Ag PNPs obtained from DDA simulation, and C(λ) is the spectral response function of the respective red, green, and blue color channels of the CCD camera. We assume that the same amount of Ag is coated on the surface of both AuNPs. It can be seen that, for the PNP with a 50 nm AuNP core, after the Ag coating is completely reacted with sulfide, the spectral peak of the AuNP−Ag PNP changes from 515 to 557 nm, whose corresponding images exhibit a clear color change, but for the one with a 100 nm core, the shift is from 527 to 535 nm and the color change is not obvious. In that regard, we decided to synthesize PNPs with a 50 nm AuNP core. The PNPs with size smaller than 50 nm were not considered because we want the plasmonic scattering images of the single PNPs to be bright enough during the experiments to maintain a good signal-tonoise ratio. Figure 1B shows the TEM image of the as-prepared AuNP− Ag core−shell PNPs. The mild synthesis method results in monodispersed, mostly spherical PNPs with a thin layer of Ag covered on the surface. Statistical analysis shows that the diameter of the PNPs is 51.09 ± 3.05 nm. EDS (Energy dispersive spectroscopy) analysis confirmed that the PNPs are composed of Au and Ag with atomic ratio of Au/Ag = 3:1 (Figure S3 in the Supporting Information). From the EDS results, the average thickness of the Ag shell was calculated to be ∼2.6 nm. Figure 1C shows the UV−vis spectra of the AuNP, the AuNP−Ag PNP, and the AuNP−Ag2S PNP solution, respectively. Due to the difference in refractive indices, the deposition of Ag on AuNP surface leads to the plasmonic maximum (λmax) of the PNPs blueshift from 534 to 518 nm; subsequent formation of Ag2S on the surface results in λmax redshift to a longer wavelength. Figure 1D shows spectral change of the PNP solution (∼12.5 pM) upon addition of different concentrations of NaHS. A redshift of Δλmax= 63 to 581 nm was registered after adding 1000 μM NaHS. The detection limit in the bulk solution was found to be 1.0 μM with Δλmax= 1.0 nm. Satisfactory selectivity and fast response of the nanoprobe toward sulfide were also confirmed (Figure 1E). Notably, as shown in the CIE (International Commission on Illumination) 1931 chromaticity diagram24 in Figure 1F, the spectral response range of the AuNP−Ag PNPs falls in the yellow-green region, the perceptually most sensitive wavelength range of human vision. To study the performance of the AuNP−Ag PNPs for single particle colorimetric sensing of sulfide, NaHS solutions of various concentrations were applied and the changes of the PNP colors were monitored with dark-field microscopy in real time. Before experiments, the PNPs were immobilized on the glass substrate surface via a commonly utilized aminosilane APTMS. By optimizing the immobilization conditions, the surface density of the PNPs was reproducibly controlled at

Figure 2. (A) Dark-field images of the AuNP−Ag PNPs before and after adding 10 μM NaHS. The scale bar is 10 μm. (B) ΔR/G distribution of the single AuNP−Ag PNPs as a function of the added sulfide concentration after over 25 min of reaction. (C) Calibration curves of ΔR/G values at various time points. (D) Selectivity of the single AuNP−Ag core−shell PNPs toward other anions.

spherical AuNP−Ag PNPs before and after applying H2S. The PNP colors changed gradually from green to orange/red, indicative of accumulative formation of Ag2S on the PNP surface. The degree of color variation appeared uniform for most of the particles, suggesting that they had similar reactivity toward H2S and the design and synthesis of the nanoprobe were indeed successful. For quantitative evaluation of the color change, the total R (red), G (green), and B (blue) intensity values of each PNP spot are calculated. With increasing reaction time or sulfide concentration, the R value increases and the G and B values decreased gradually (Figure S4 in the Supporting Information). Without addition of sulfide, no fluctuation of R/G (Figure S5 in the Supporting Information) was observed. Therefore, the parameter R/G and its change Δ[R/G] could be used to depict the amount of PNP color variation. Figure 2B displays the results after applying 8 different concentrations of 0.05, 0.1, 0.5, 1.0, 5.0, 10, 50, and 4665

dx.doi.org/10.1021/ac500376e | Anal. Chem. 2014, 86, 4663−4667

Analytical Chemistry

■

100 μM NaHS to the immobilized PNPs for over 20 min, respectively. The R/G ratio distribution of the original PNPs was shown in Figure S6 in the Supporting Information. A clear trend of and Δ[R/G] enlargement was evident. The average Δ[R/G] value exhibits a roughly linear dependence on the logarithm of the applied H2S concentration (Figure S7 in the Supporting Information). Figure S8 in the Supporting Information shows the kinetic Δ[R/G] variation as a function of the reaction time at each of the 8 sulfide concentrations. Due to the “accumulative” nature of the nanoprobe, the spectral shift as well as the color change of the PNPs were all relatively fast initially and then slowed down gradually. Because the average Δ[R/G] variation rate or Ag2S formation rate on the PNP surface at each point along the reaction time is different for different concentrations of the applied sulfide, a set of calibration curves can be constructed at different time points (Figure 2C). Thus, for any PNP experiencing a color change, as long as we have its total reaction time and can follow the kinetic Δ[R/G] change from its initial point, the external sulfide concentration surrounding the particle can be deduced from the calibration curves.10 Figure 2D shows the selectivity test results of the single PNP in the presence of a 25 μM solution of S2O32−, SO32−, SCN−, NO3−, NO2−, Cl−, Br−, I−, CH3COO−, CO32−, and HS−, respectively, under the same experimental condition. No color change was observed for anions other than HS−, indicating that the PNPs are highly selective toward sulfide. Taken together, the experimental results from colorimetric analysis by using single spherical PNPs are consistent with those from spectral analysis with single rod-shaped PNPs in the previous report. Since each spherical PNP is shown as just one color spot and there is no interference between zero-order and first-order images of different nanoprobes, a much higher density of the surface immobilized PNPs could be used and high-throughput sensing could be accomplished with the ease of RGB calculation. Every Δ[R/G] distribution in Figure 2B was from 120 PNPs. The 8 curves in Figure S8 in the Supporting Information were each from consecutive measurements at 20 time points of 30 different PNPs and were the results of color analysis of 4800 spots in total. Such a large number of data ensures that the calibration curves in Figure 2C are statistically significant, but it would be extremely timeconsuming to analyze the spectra of individual spots manually. The only drawback of this single-particle colorimetric sensing approach so far is that the limit of detection cannot be lower than 50 nM. The reason is that color measurement provides only the intensity integration of the three R, G, and B channels and is intrinsically not as sensitive as grating-based spectral measurement in differentiating small spectral variations. Nevertheless, this single PNP assay is already 20 times more sensitive than UV−vis detection in the bulk solution and covers a large dynamic range of over 3 orders of magnitude, so it is adequate for practical H2S sensing in biological systems. It has been reported that the H2S concentration in the blood and tissue is in the range of 10−100 μM.25,26 Figure S9 in the Supporting Information shows that these PNPs can be used for H2S sensing in the cell culture medium and the blood serum. Another study aiming to improve the sensitivity of colorimetric analysis has been carried out and would be reported in a later publication.

Technical Note

CONCLUSION We have synthesized 51 nm spherical AuNP−Ag core−shell PNPs for H2S sensing based on Ag2S formation induced color change of the single PNPs. Their spectral response range was designed to match the most sensitive region of human perception of color. Individual PNPs were immobilized on glass substrate with high density. With colorimetric sensing under dark-field microscopy, each PNP becomes a spatially distinct individual H2S probe. As there is no spectrometer or grating, no restriction on the sensing area by the entrance slit, and no overlay between the zero and first order images, highthroughput detection of the target molecule can be achieved by imaging the color change of hundreds of nanoprobes simultaneously with no interference from each other. Importantly, because calculating the RGB intensity values and R/G ratios with a digital color camera are much simpler and time-efficient than manually identifying the spectral peak position with a grating-based setup, rapid color analysis of a large number of PNPs could be accomplished, allowing better statistics and reliable quantification of the response of the nanoprobes under each condition. With color imaging capabilities widely available in various consumer devices, developments of similar single particle colorimetric sensing strategies can be expected to detect other molecules for chemical and biological applications.

■

ASSOCIATED CONTENT

* Supporting Information S

Additional figures as mentioned in the main text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS This work was supported by NSFC 21127009, NSFC 91027037, NSFC 21221003, and Hunan University 985 fund. REFERENCES

(1) Stewart, M. E.; Anderton, C. R.; Thompson, L. B.; Maria, J.; Gray, S. K.; Rogers, J. A.; Nuzzo, R. G. Chem. Rev. 2008, 108, 494− 521. (2) Saha, K.; Agasti, S. S.; Kim, C.; Li, X.; Rotello, V. M. Chem. Rev. 2012, 112, 2739−2779. (3) Jain, P. K.; Huang, X. H.; El-Sayed, I. H.; El-Sayed, M. A. Acc. Chem. Res. 2008, 41, 1578−1586. (4) Novo, C.; Funston, A. M.; Mulvaney, P. Nat. Nanotechnol. 2008, 3, 598−602. (5) Anker, J. N.; Hall, W. P.; Lyandres, O.; Shah, N. C.; Zhao, J.; Duyne, R. P. V. Nat. Mater. 2008, 7, 442−453. (6) Zhang, L.; Li, Y.; Li, D. W.; Jing, C.; Chen, X.; Lv, M.; Huang, Q.; Long, Y. T.; Willner, I. Angew. Chem., Int. Ed. 2011, 50, 6789−6792. (7) Ma, Y. F.; Shortreed, M. R.; Yeung, E. S. Anal. Chem. 2000, 72, 4640−4645. (8) Li, J. W.; Lee, J. Y.; Yeung, E. S. Anal. Chem. 2006, 78, 6490− 6496. (9) Cheng, J.; Liu, Y.; Cheng, X. D.; He, Y.; Yeung, E. S. Anal. Chem. 2010, 82, 8744−8749. (10) Xiong, B.; Zhou, R.; Hao, J. R.; Jia, Y. H.; He, Y.; Yeung, E. S. Nat. Commun. 2013, 4, 1708−1716. 4666

dx.doi.org/10.1021/ac500376e | Anal. Chem. 2014, 86, 4663−4667

Analytical Chemistry

Technical Note

(11) Xiao, L. H.; Wei, L.; He, Y.; Yeung, E. S. Anal. Chem. 2010, 82, 6308−6314. (12) Liu, Y.; Ling, J.; Huang, C. Z. Chem. Commun. 2011, 47, 8121− 8123. (13) Jing, C.; Gu, Z.; Ying, Y. L.; Li, D. W.; Zhang, L.; Long, Y. T. Anal. Chem. 2012, 84, 4284−4291. (14) Morse, J. W.; Millero, F. J.; Cornwell, J. C.; Rickard, D. EarthSci. Rev. 1987, 24, 1−42. (15) Li, L.; Rose, P.; Moore, P. K. Annu. Rev. Pharmacol. Toxicol. 2011, 51, 169−187. (16) Davidson, J. M.; Pikramenou, Z.; Ponce, A.; Winpenny, R. E. P. Anal. Chem. 2009, 81, 3669−3675. (17) Hyšplera, R. r.; Ticháb, A.; Indrováb, M.; Zadáka, Z.; Hyšplerovác, L.; Gaspariča, J.; Churácě k, J. J. Chromatogr., B 2002, 770, 255−259. (18) Peng, H.; Cheng, Y.; Dai, C.; King, A. L.; Predmore, B. L.; Lefer, D. J.; Wang, B. Angew. Chem., Int. Ed. 2011, 50, 9672−9675. (19) Xuan, W.; Sheng, C.; Cao, Y.; He, W.; Wang, W. Angew. Chem., Int. Ed. 2012, 51, 2282−2284. (20) Sasakura, K.; Hanaoka, K.; Shibuya, N.; Mikami, Y.; Kimura, Y.; Komatsu, T.; Ueno, T.; Terai, T.; Kimura, H.; Nagano, T. J. Am. Chem. Soc. 2011, 133, 18003−18005. (21) Niu, J. L.; Zhu, T.; Liu, Z. F. Nanotechnology 2007, 18, 325607− 325613. (22) Chen, H. J.; Kou, X. S.; Yang, Z.; Ni, W. H.; Wang, J. F. Langmuir 2008, 24, 5233−5237. (23) Jenison, R.; Yang, S.; Haeberli, A.; Polisky, B. Nat. Biotechnol. 2001, 19, 62−65. (24) http://en.wikipedia.org/wiki/CIE_1931_color_space. (25) Li, L.; Bhatia, M.; Zhu, Y. Z.; Zhu, Y. C.; Ramnath, R. D.; Wang, Z. J.; Anuar, F. B. M.; Whiteman, M.; Salto-Tellez, M.; Moore, P. K. FASEB J. 2005, 19, 1196−1212. (26) Chen, Y. H.; Yao, W. Z.; Geng, B.; Ding, Y. L.; Lu, M.; Zhao, M. W.; Tang, C. S. Chest 2005, 128, 3205−3211.

4667

dx.doi.org/10.1021/ac500376e | Anal. Chem. 2014, 86, 4663−4667