Open Access
Add to BibSonomy
Abstract
We develop a front-to-end solution where the shift of chromaticity from scattering of plasmonic nanoparticles is used as the reporter for nano-environmental refractive index variation. By co-projecting possible power combinations of RGB LEDs and digitized color grid density of CCD with various luminance onto the CIE 1931 chromaticity diagram, optimum condition for nanoenvironment sensing can be achieved. The highest resolution for local refractive index change is 0.0021 per distinguishable color, which is higher than that of a typical handheld spectrometer by 4.8 times. This result shows great potential in simplifying nano-environment sensing instruments and is particularly useful for multi-point dynamical process.
© 2019 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Nanoenvironment sensing is a substantial topic since local refractive index, PH value [1,2], charge density, and so on may dictate chemical reactions [3–5], interfacial physics [6–9], and spectroscopic properties. Of particular interest, noble nanoparticles (NPs) due to their relatively large volume density of free electrons, can be easily excited by visible light, forming the so-called surface plasmons [10]. The manifestation of collective oscillation of electrons is extremely sensitive to local changes of environment and is widely exploited as a nano-scale probe, contrast agent [9,11–15], and heat transducer for molecular identification [4,16,17], labeling/imaging [5,18,19], and photothermal therapeutics [20–23] to name a few. From optics point of view, most of the environmental variables can all be turned into effective refractive index (RI) variations. In general, depending on the substances to be probed, there are three popular sensing regions, namely, RI ~1 for gas sensing [7,24]; RI~1.33 for bio-chemicals in water [25,26]; and RI >1.5 for sensing with high index substrate such as glass [27,28].
So far, measuring spectral shift from scattering of plasmonic structures such as monomer, dimer, and trimer turns out to be the mainstream method to sense the local RI change [24,25,29]. Despite the resolution of available spectrometers can constantly achieve 0.01 nm, the inherent broad spectral response from plasmonic structures poses difficulties to determine the peak position and thereby limits the accuracy [24,25,30]. Besides, the built-in diffraction element, focusing mirrors, and slits altogether results in a bulky package which hinders compact integration. Moreover, to simultaneously gain information from different positions, the spectrometer based system needs to add on extra optics, or multiple scanning with different integration times is unavoidable which goes against real time probing. By contrary, charge coupled device (CCD) and complimentary metal-oxide-semiconductor (CMOS) based imaging systems hold advantages such as light-weight, compact, and are superior for obtaining position dependent information in one go. However, the spectral resolution is considered far lower than that of spectrometers since only 3 color channels (red, green, and blue) are available to turn the associated refractive index change ΔRI into color difference. As a result, one can rarely find relevant papers addressing the possibility to determine ΔRI as small as 0.01 with CCDs, except for cases with eye-sensorable color change associated with relatively large ΔRI [16,17]. Up to now, only very few people consider CCD based system a superior alternative for spectrometers [13,31–34].
Not to take it for granted, in this study, we challenge this issue by assessing the chromatic resolution comprehensively from various aspects. Unlike spectrometers which disperse different colors to independent pixels by gratings, CCDs dump all the received signal to pixels with R, G, and B color filters. As a result, the detailed spectrum information is merged into the three color channels and the light intensity affects the value of R, G, and B channels as well. Consequently, it is not difficult to imagine that upon conversion, the chromaticity coordinate is intensity dependent. This is the biggest difference between spectrometer and CCD based measurements since for spectrometers, the spectral features and shape will not change with light intensity. Moreover, for the same ΔRI, the color difference subject to different illuminations may be distinct [34]. Nevertheless, this property may potentially be useful for tracking dynamic process with much relaxed requirements for the illuminant itself, as long as it is reasonably stable during the measuring period [9,14,15,33]. In addition, the shift of chromaticity coordinate in response to ΔRI may be maximized by combinations of different illuminants [34–37]. There are massive research focusing on color mixing [34,38], color rendering [19,39–42], and visual color differences [35–37]. However, no systematic studies have ever been conducted for interrogating the limitation of CCD’s chromatic resolution. Here we use plasmonic monomer and dimer as the model system since their scattering spectra are extremely sensitive to the surrounding medium. We simultaneously project the spectral shift and digitized color grid density onto the chromaticity coordinate subject to various illuminating conditions. The rule of thumb to optimize the chromatic resolution is derived. It is found that under certain circumstances, CCD based system can be superior, reaching a resolution as high as 480 distinguishable color grids/RIU, where the RIU is defined as the refractive index change by 1. The result renders this method as a poor man’s alternative for nanoenvironment sensing and dynamic process monitoring applications.
2. Results and discussion
2.1 Correspondence between scattering spectrum and chromaticity coordinate
The normalized scattering spectra from plasmonic monomer and dimer are calculated using generalized multi-particle Mie theory (GMM) [43]. For monomer and dimer consist of Au nanoparticle (NP) with diameter d = 100 nm and inter-particle distance d = 2 nm, the calculated scattering spectra for surrounding media with RI change from n = 1.00 to n = 1.60 are shown in Fig. 1(a). The refractive index of Au in the calculation was extracted from the result of Johnson and Christy, 1972 [44]. It is found that for the case of plasmonic monomer in air (n = 1), the scattering peaked at the wavelength λ = 530 nm. With the increasing of the RI of surrounding medium, this peak red-shifted to longer wavelength with a rate of 85 nm/RIU. For the case of plasmonic dimer, apart from the scattering peak associated with the monomer’s response, an additional peak appears at longer wavelength corresponding to the gap plasmonic mode [45,46]. This mode as compared with the monomer mode, exhibits larger spectral shift (250 nm/RIU) subjecting to the same ΔRI. This has been attributed to the highly concentrated electromagnetic field which is largely amplified due to the excitation of surface plasmons. We now move onto the practical case where the illuminant is a standard D65 light with spectrum shown in Fig. 10(a) in the Appendix [47]. We can as well calculate the realistic scattering spectra for the two cases, as shown in Fig. 1(b). It is found that the resultant scattering spectra are modulated with many amplified features, reflecting the characteristics of the light source distributed in λ = 400-630 nm.
Fig. 1 (a) GMM calculated scattering spectra of gold nanoparticles in different surrounding media with RI vary from n = 1 to n = 1.6. The upper row is for the case of monomer, and the lower row is for dimers with 2 nm gap distance. (b) The scattering spectra of nanoparticles illuminated by standard D65 light source. (c) Trajectories of the chromaticity shift (black/red curve) when monomer/dimer was illuminated by D65 light with surrounding’s RI changes from n = 1 to n = 1.6.
Next we turn to project the obtained spectra onto CIE 1931 chromaticity coordinate. According to the definition [47], we can convert any spectral curve into the so-called tristimulus values by Eq. (1):
Where L(λ) is any power spectral density to be projected, , , are the CIE 1931 matching functions [47,48], as shown in Fig. 10(b) in the Appendix. By Eq. (2), we can further project the coordinates onto a 2D plane, namely, the CIE 1931 chromaticity coordinate. For plasmonic monomer and dimer, the trajectory of chromaticity shift associated with the ΔRI of surroundings subject to D65 light illumination are depicted in Fig. 1(c).
Compared with the scattering spectrum, there is a distinct difference. In scattering spectrum, the spectral shift is larger for the gap plasmonic mode of a dimer. However in chromaticity coordinate, the color shift spans a larger range for plasmonic monomer. The reason is that the construction of CIE 1931 chromaticity diagram is primarily from the standpoint of human vision. Upon conversion, any spectral feature longer than λ = 740 nm contributes nothing to the color shift. This suppresses the contribution from the gap plasmonic mode. Consequently, for plasmonic dimers, the color shift is firstly biased by the spectral shift of the monomer mode, moving from pinkish to greenish color and eventually back into the green-yellow range. As for the case of plasmonic monomer, since the scattering spectrum is single peaked, the evolution of color shifts monotonically from green at λ = 530 for n = 1 to orange-yellowish color at λ = 580 nm for n = 1.6. From this standpoint, plasmonic monomer is more appropriate for ΔRI sensing as compared to dimers. Hereafter, we focused on the case of monomer unless otherwise stated.
2.2 Light source dependent chromaticity
To further understand the color shift associated with different illuminants, we use the standard halogen lamp of Olympus microscope IX73 as the white light source to simulate the scattered chromaticity of plasmonic monomer and dimer (diameter d = 100 nm) under different illuminating intensities. Figure 2(a) gives four different illuminating power spectra used in the simulation, Figs. 2(b) and 2(c) are the corresponding chromaticity shift for plasmonic monomer and dimer, respectively. It is found that with the increasing of the illuminating powers (from P1 to P4), these curves shift from orange to yellow-green color. This can be attributed to the chromaticity shift of the illuminants in accordance with their spectral subject to different illuminating powers. It should be noted that apart from chromaticity shift, some curves are stretched or contracted indicating that the sensitivity of ΔRI by manifestation of color difference may be optimized with specially tailored spectrum of illuminant.
Fig. 2 (a) The spectra of halogen lamp at four different powers. (b)–(c) Trajectories of the chromaticity shift of monomer and dimer illuminated by the four light sources, respectively. The bottom row is the enlarged view with chromaticity coordinates of the corresponding light sources designated by points.
2.3 Standardization and enhancement of color difference
According to previous discussion, the detection of ΔRI by virtue of the color difference from plasmonic NPs is limited to visible range and is highly dependent on the spectrum of illumination. In order to accurately determine the ΔRI and maximize the resolution of the associated color difference, it is on high demand to establish a guideline to operate this sensing system. Here we use a tunable illuminant consisting of 3 commercially available LEDs emitted at 3 primary colors, namely, Red (R), Green (G), and Blue (B). The corresponding spectra are shown in Fig. 10(c) in the Appendix. We assume that the power of individual color can be independently adjusted. The goal is to find the relative power spectral density ratio (a:b:c) between the RGB LEDs under which the color difference of the plasmonic NPs surrounded with different RI can be maximized. In principle, to find the corresponding (a:b:c) for each chromaticity coordinate (x,y), we can use Eq. (2) and Eq. (3) and scan all possible power combinations (a:b:c) to achieve this goal. Unfortunately, it is by no means straightforward since the conversion involves integration which is neither just linear superposition nor simple algebra. In order to completely and seamlessly cover the whole chromaticity diagram without gaps, we must treat this as an inverse problem. We first start form a point (x,y) with a given Y, a quantity proportional to the luminance of color. In general, one can treat the value of Y as a parameter where perfect transmission/reflection out of the scatterer gives Y = 100 [46]. Next, the corresponding values for X, Y, and Z can be found by Eq. (2). We then plug the X, Y, and Z values in Eq. (3) to find the required power ratio (a:b:c) for producing this pre-selected chromaticity (x,y). This process repeats until chromaticity coordinates are fully filled.
The result for Y = 1 is depicted in Fig. 3. The triangle is the gamut within which colors can be produced and the vertices are the coordinates for pure R, G, and B colors. It is worth noting that the weighting of power spectral density for the blue color c(x,y) is much larger than that of red a(x,y) and green b(x,y). In addition, c(x,y) is highly localized around the vertex, while a(x,y) and b(x,y) are dispersed and monotonically decrease toward their complementary colors with relatively higher rate. This is why we need to carry out inverse conversion to avoid incomplete and inhomogeneous filling grids on CIE 1931 chromaticity diagram. Note here the pre-set value Y will not affect the relative ratio a(x,y):b(x,y):c(x,y) distributed on the x-y color space. Once we have built up the RGB power weighting basis for any point on the chromaticity diagram, the chromaticity coordinates for plasmonic NPs can be straightforwardly obtained via convolving their scattering spectra with that of the illuminant. As the RI of environment changes, the color difference (CD) can be calculated. The result was quantified by normalizing to that obtained with standard D65 light source ΔE0, and was defined as the color difference enhancement (CDE). Figures 4(a)–4(c) show the CDE for ΔRI = 1.0-1.1, ΔRI = 1.3-1.4, and ΔRI = 1.5-1.6, respectively.
Fig. 3 RGB power distribution for producing the CIE 1931 chromaticity diagram, where (a), (b), and (c) for R, G, and B LED, respectively.
Fig. 4 Color difference enhancement (CDE) of Au monomer illuminated by different power combinations of RGB LEDs. The values at each coordinate (x,y) are referenced to that illuminated by standard D65 light, and the Euclidean distance ΔE = [(Δx)2 + (Δy)2]1/2 was used to define the color difference.
2.4 Color gamut and chromatic resolution
Knowing the rule of thumb to optimize CD, we now turn to the discussion of measurable color gamut and the achievable chromatic resolution based on color CCDs. Consider a 24-bit color CCD, the recorded standard RGB (sRGB) values are transferred to linear RGB values, and further into tristimulus X, Y, Z according to Eq. (4) and Eq. (5) [49], respectively.
Using Eq. (2), we can project the CCD measured colors in form of RGB values onto the chromaticity diagram with any specified Y value. It is found that with the increasing of Y, the resolvable color gamut decreases, as shown in Fig. 5 for the three primary colors. This is attributed to quick saturation of the RGB values which results in very limited color difference around the white point.
Fig. 5 The resolvable color gamut of a 24-bit color CCD for individual R, G, and B channel at Y = 3, 30, and 70.
To derive the chromatic resolution of CCD based system, it should be noted that chromaticity is characterized by the RGB values. Therefore, realistic color difference cannot be continuous. Instead, the grey-scale has been digitized into 0-255 grids per color channel. We assume the value of RGB changes by 1 per color channel as distinguishable colors. In this way, we may re-assign all the even/odd values in RGB channels as 0/1 which can effectively reduce the original 24-bit data into 3-bit format without affecting the chromatic resolution. According to Table 1, new values are assigned to original values in RGB channels. In this situation, Fig. 5 was discretized where individual R, G, and B channels with the same Y value were merged into one in advance, as shown in Figs. 6(a)–6(c). Here, the grid size at each coordinate (x,y) represents the chromatic resolution of local chromaticity. Namely, the neighboring grid stands for a distinguishable color. It is once again found that with the increase of the luminance, the resolvable color gamut decreases. Besides, the size and shape of each grid is different, indicating the chromatic resolution is coordinate dependent. To further quantify the chromatic resolution (CR) within the color gamut, we counted the total number of grids per area (ΔxΔy = 0.03 × 0.03), i.e., the color grid density (CGD) is used as an indicator. The reason to choose such an area size is because the Euclidean distance, namely, the color difference produced by D65 light for ΔRI = 0.1 is about ΔE0 = 0.03. As shown in Figs. 6(d)–6(f), with the increase of Y value, the averaged CR becomes higher, more delocalized, and more homogeneously distributed on the chromaticity diagram.
Table 1. Re-assigned RGB values for converting 24-bit into 3-bit data.
Fig. 6 The chromatic mosaic (grid) which shows the minimally distinguishable color based on a 24-bit color CCD for luminance (a) Y = 3, (b) Y = 30, and (c) Y = 70. The grid density is calculated over an area of ΔxΔy = 0.03 × 0.03 for (d) Y = 3, (e) Y = 30, and (f) Y = 70, respectively.
2.5 Resolving power of environment ΔRI
We now illustrate how to translate CD into the corresponding ΔRI based on plasmonic NPs. Consider a plasmonic monomer immersed in a medium with RI varies from 1.3 to 1.4 and illuminated by standard D65 light source. Figure 7(a) shows the calculated scattering spectrum when the environmental RI changes from 1.3 to 1.4 in steps of 0.01. It is found that the scattering peak shifts from λ = 550 nm to λ = 560 nm. Following the process described previously, the chromaticity coordinates associated with the scattering spectra for different surrounding media can be calculated. As shown in Figs. 7(b)–7(d) where the minimally resolvable color grids are superimposed on CIE 1931 chromaticity diagram along with the calculated chromaticity coordinates for Y = 3, Y = 30, and Y = 70, respectively. It should be noted that the color difference characterized by the Euclidean distance is nearly linearly proportional to the spectral shift for the present case. It is found that for ΔRI = 0.1, the total number of grids which cover the span of the chromaticity shift are 13/33/48, corresponding to an averaged resolution of Δn = 0.0077/0.0030/0.0021 per grid (or 130/330/480 grids/RIU) for Y = 3/30/70, respectively. Compare to a handheld spectrometer with a typical spectral resolution about 1 nm, only 10 scattering peaks are distinguishable. Our result gives a higher resolution by 4.8 times which is surprisingly high considering CCD is a grating-free device. The only drawback is probably the inherently nonlinear distribution of color grids. Note that the resolution demonstrated here is not even optimized yet. Since the CDE in Fig. 4 and the CGD in Figs. 6(d) and 6(f) are both projected onto the chromaticity diagram, one may select appropriate working point where the product of the two quantities is maximized to further raise the resolution. This rule of thumb enables one to determine the optimal chromaticity of the illuminant and thereby the relative power ratio (a:b:c) between the RGB LEDs from Fig. 3. Practically, the convolution between Figs. 4(a)–4(c) and Figs. 6(d)–6(f) yields the map of color grid density which can be viewed as a shortcut to use, as shown in Figs. 8(a)–8(c). Note here the light source should be replaced by the scattered light from the plasmonic NPs under study. In order to make comparison with Figs. 7(b)–7(d) where the chromatic resolution was characterized by the number of grids covering the total span of the shift of chromaticity, the 1-dimensional line density of grid was obtained by taking square root of the 2-dimensional area density of grid. The obtained line density associated with each chromaticity coordinate can be treated as the enhanced resolution achieved by different combinations of RGB illuminants, as shown in Fig. 8. The result shows that with intermediate power of illumination (Y = 30), the resolution for ΔRI is higher and covers a larger area on the chromaticity diagram.
Fig. 7 (a) Scattering spectrum of Au monomer surrounded by a medium with RI changes from n = 1.3 to n = 1.4, and illuminated by D65 light. (b)–(d) CCD based chromatic resolution characterized by the number of grids covering the span of the color shift for Y = 3, Y = 30, and Y = 70, respectively.
Fig. 8 The effective chromatic resolution characterized by the line grid density which was produced by the convolution between the RGB enhanced color difference and CCD’s chromatic resolution for (a) Y = 3, (b) Y = 30, and (c) Y = 70. The environmental refractive index changes from n = 1.3 to n = 1.4.
Figure 9 outlines the entire process of nanoenvironment sensing introduced in this study. Essentially, the plasmon enhanced CD behaves as the reporter in response to the environmental ΔRI. The CD can be characterized by the convolution of the illuminant, scattering signal, and CCD’s response. By projecting the abovementioned quantities altogether onto the chromaticity diagram, one may find optimal conditions to achieve the highest possible chromatic resolution. It should be noted here that we have illustrated an example for minimal ΔRI measurement based on Au nanoparticles in water based environment (n = 1.3 to n = 1.4). The purpose is to show that color CCDs can achieve higher resolution in some cases which has never been considered possible. For researchers who are interested in other wavelength range based on different sensing structures, Fig. 9 is a guideline which may facilitate the development of deep-learning capable algorithms to optimize the chromatic resolution for specific applications.
Fig. 9 The process to optimize experimental parameters for achieving the highest possible chromatic resolution.
3. Conclusions
We develop a comprehensive colorimetry where the color difference as a consequence of nano-environmental refractive index variation is used as the reporter. By co-projecting power combinations of RGB LEDs, scattering spectrum from plasmonic monomer, and digitized color grid of CCD altogether onto CIE 1931 chromaticity diagram, front-to-end conditions for the illumination and detector can be iteratively optimized. The highest achieved resolution for local refractive index change is 0.0021, corresponding to 480 color grids/RIU, which is higher than any handheld grating-based spectrometer by 4.8 times. The present result provides an alternative way to minimize nano-environmental sensing instruments and is particularly useful for multi-point dynamic process monitoring applications.
Appendix
Funding
Ministry of Science and Technology (MOST), Taiwan, Republic of China. (MOST 107-2112-M-008-010-).
References
1. J. K. Lim and S. W. Joo, “Gold nanoparticle-based pH sensor in highly alkaline region at pH > 11: surface-enhanced Raman scattering study,” Appl. Spectrosc. 60(8), 847–852 (2006). [CrossRef] [PubMed]
2. W. Luo, Q. Cui, K. Fang, K. Chen, H. Ma, and J. Guan, “Responsive hydrogel-based photonic nanochains for microenvironment sensing and imaging in real time and high resolution,” Nano Lett., (posted 11 January 2018, in press).
3. R. Semenyshyn, M. Hentschel, C. Stanglmair, T. Teutsch, C. Tarin, C. Pacholski, H. Giessen, and F. Neubrech, “In-vitro monitoring conformational changes of polypeptide monolayers using infrared plasmonic nanoantennas,” Nano Lett., (posted 2 August 2018, in press).
4. D. A. Stuart, A. J. Haes, A. D. McFarland, S. Nie, and R. P. Van Duyne, “Refractive index sensitive, plasmon resonant scattering, and surface enhanced Raman scattering nanoparticles and arrays as biological sensing platforms,” Proc. SPIE 5327, 60–73 (2004). [CrossRef]
5. J. Ma, L. Zhan, R. S. Li, P. F. Gao, and C. Z. Huang, “Color-encoded assays for the simultaneous quantification of dual cancer biomarkers,” Anal. Chem. 89(16), 8484–8489 (2017). [CrossRef] [PubMed]
6. H. Li, Y. Xu, J. Xiang, X. F. Li, C. Y. Zhang, S. L. Tie, and S. Lan, “Exploiting the interaction between a semiconductor nanosphere and a thin metal film for nanoscale plasmonic devices,” Nanoscale 8(45), 18963–18971 (2016). [CrossRef] [PubMed]
7. J. Qin, Y. H. Chen, B. Ding, R. J. Blaikie, and M. Qiu, “Plasmonic gas sensing based on cavity-coupled metallic nanoparticles,” J. Phys. Chem. C 121(44), 24740–24744 (2017). [CrossRef]
8. K. Lodewijks, W. Van Roy, G. Borghs, L. Lagae, and P. Van Dorpe, “Boosting the figure-of-merit of LSPR-based refractive index sensing by phase-sensitive measurements,” Nano Lett. 12(3), 1655–1659 (2012). [CrossRef] [PubMed]
9. M. Liu, Q. Li, L. Liang, J. Li, K. Wang, J. Li, M. Lv, N. Chen, H. Song, J. Lee, J. Shi, L. Wang, R. Lal, and C. Fan, “Real-time visualization of clustering and intracellular transport of gold nanoparticles by correlative imaging,” Nat. Commun. 8, 15646 (2017). [CrossRef] [PubMed]
10. K. A. Willets and R. P. Van Duyne, “Localized surface plasmon resonance spectroscopy and sensing,” Annu. Rev. Phys. Chem. 58(1), 267–297 (2007). [CrossRef] [PubMed]
11. S. Vial, Y. Berrahal, M. Prado, and J. Wenger, “Single-step DNA detection assay monitoring dual-color light scattering from individual metal nanoparticle aggregates,” ACS Sens. 2(2), 251–256 (2017). [CrossRef] [PubMed]
12. D. Jana, C. Matti, J. He, and L. Sagle, “Capping agent-free gold nanostars show greatly increased versatility and sensitivity for biosensing,” Anal. Chem. 87(7), 3964–3972 (2015). [CrossRef] [PubMed]
13. Y. Liu and C. Z. Huang, “Single scattering particles based analytical techniques,” Chin. Sci. Bull. 58(17), 1969–1979 (2013). [CrossRef]
14. X. Y. Wan, L. L. Zheng, P. F. Gao, X. X. Yang, C. M. Li, Y. F. Li, and C. Z. Huang, “Real-time light scattering tracking of gold nanoparticles- bioconjugated respiratory syncytial virus infecting HEp-2 cells,” Sci. Rep. 4(4), 4529 (2014). [PubMed]
15. W. Ye, M. Götz, S. Celiksoy, L. Tüting, C. Ratzke, J. Prasad, J. Ricken, S. V. Wegner, R. Ahijado-Guzmán, T. Hugel, and C. Sönnichsen, “Conformational dynamics of a single protein monitored for 24 h at video rate,” Nano Lett. 18(10), 6633–6637 (2018). [CrossRef] [PubMed]
16. M. O. Noor and U. J. Krull, “Camera-based ratiometric fluorescence transduction of nucleic acid hybridization with reagentless signal amplification on a paper-based platform using immobilized quantum dots as donors,” Anal. Chem. 86(20), 10331–10339 (2014). [CrossRef] [PubMed]
17. Z. Yuan, C. C. Hu, H. T. Chang, and C. Lu, “Gold nanoparticles as sensitive optical probes,” Analyst (Lond.) 141(5), 1611–1626 (2016). [CrossRef] [PubMed]
18. S. H. Wang, C. W. Lee, A. Chiou, and P. K. Wei, “Size-dependent endocytosis of gold nanoparticles studied by three-dimensional mapping of plasmonic scattering images,” J. Nanobiotechnology 8(1), 33 (2010). [CrossRef] [PubMed]
19. T. Li, X. Wu, F. Liu, and N. Li, “Analytical methods based on the light-scattering of plasmonic nanoparticles at the single particle level with dark-field microscopy imaging,” Analyst (Lond.) 142(2), 248–256 (2017). [CrossRef] [PubMed]
20. X. Huang, P. K. Jain, I. H. El-Sayed, and M. A. El-Sayed, “Plasmonic photothermal therapy (PPTT) using gold nanoparticles,” Lasers Med. Sci. 23(3), 217–228 (2008). [CrossRef] [PubMed]
21. X. Huang and M. A. El-Sayed, “Plasmonic photo-thermal therapy (PPTT),” Alexandria Med. J. 47(1), 1–9 (2011). [CrossRef]
22. A. F. Bagley, S. Hill, G. S. Rogers, and S. N. Bhatia, “Plasmonic photothermal heating of intraperitoneal tumors through the use of an implanted near-infrared source,” ACS Nano 7(9), 8089–8097 (2013). [CrossRef] [PubMed]
23. S. Wang, H. Xu, and J. Ye, “Plasmonic rod-in-shell nanoparticles for photothermal therapy,” Phys. Chem. Chem. Phys. 16(24), 12275–12281 (2014). [CrossRef] [PubMed]
24. J. M. Bingham, J. N. Anker, L. E. Kreno, and R. P. Van Duyne, “Gas sensing with high-resolution localized surface plasmon resonance spectroscopy,” J. Am. Chem. Soc. 132(49), 17358–17359 (2010). [CrossRef] [PubMed]
25. C. P. Byers, B. S. Hoener, W. S. Chang, M. Yorulmaz, S. Link, and C. F. Landes, “Single-particle spectroscopy reveals heterogeneity in electrochemical tuning of the localized surface plasmon,” J. Phys. Chem. B 118(49), 14047–14055 (2014). [CrossRef] [PubMed]
26. B. Doherty, M. Thiele, S. Warren-Smith, E. Schartner, H. Ebendorff-Heidepriem, W. Fritzsche, and M. A. Schmidt, “Plasmonic nanoparticle-functionalized exposed-core fiber-an optofluidic refractive index sensing platform,” Opt. Lett. 42(21), 4395–4398 (2017). [CrossRef] [PubMed]
27. A. R. Rashed, B. Gudulluoglu, H. W. Yun, M. Habib, I. H. Boyaci, S. H. Hong, E. Ozbay, and H. Caglayan, “Highly-sensitive refractive index sensing by near-infrared metatronic nanocircuits,” Sci. Rep. 8(1), 11457 (2018). [CrossRef] [PubMed]
28. M. Piliarik, P. Kvasnička, N. Galler, J. R. Krenn, and J. Homola, “Local refractive index sensitivity of plasmonic nanoparticles,” Opt. Express 19(10), 9213–9220 (2011). [CrossRef] [PubMed]
29. T. Y. Tseng, P. J. Lai, and K. B. Sung, “High-throughput detection of immobilized plasmonic nanoparticles by a hyperspectral imaging system based on Fourier transform spectrometry,” Opt. Express 19(2), 1291–1300 (2011). [CrossRef] [PubMed]
30. P. Chen and B. Liedberg, “Curvature of the localized surface plasmon resonance peak,” Anal. Chem. 86(15), 7399–7405 (2014). [CrossRef] [PubMed]
31. J. Zhou, G. Lei, L. L. Zheng, P. F. Gao, and C. Z. Huang, “HSI colour-coded analysis of scattered light of single plasmonic nanoparticles,” Nanoscale 8(22), 11467–11471 (2016). [CrossRef] [PubMed]
32. J. Zhou, P. F. Gao, H. Z. Zhang, G. Lei, L. L. Zheng, H. Liu, and C. Z. Huang, “Color resolution improvement of the dark-field microscopy imaging of single light scattering plasmonic nanoprobes for microRNA visual detection,” Nanoscale 9(13), 4593–4600 (2017). [CrossRef] [PubMed]
33. Z. Gu, C. Jing, Y. L. Ying, P. He, and Y. T. Long, “In situ high throughput scattering light analysis of single plasmonic nanoparticles in living cells,” Theranostics 5(2), 188–195 (2015). [CrossRef] [PubMed]
34. X. Cheng, D. Dai, Z. Yuan, L. Peng, Y. He, and E. S. Yeung, “Color difference amplification between gold nanoparticles in colorimetric analysis with actively controlled multiband illumination,” Anal. Chem. 86(15), 7584–7592 (2014). [CrossRef] [PubMed]
35. R. W. Pridmore and M. Melgosa, “Effect of luminance of samples on color discrimination ellipses: analysis and prediction of data,” Color Res. Appl. 30(3), 186–197 (2005). [CrossRef]
36. W. R. J. Brown, “The influence of luminance level on visual sensitivity to color differences,” J. Opt. Soc. Am. 41(10), 684–688 (1951). [CrossRef] [PubMed]
37. F. Carreño and J. M. Zoido, “The influence of luminance on color-difference thresholds,” Color Res. Appl. 26(5), 362–368 (2001). [CrossRef]
38. S. A. Burns, “Subtractive color mixture computation,” http://scottburns.us/subtractive-color-mixture/, (accessed January 2019).
39. F. Carreño and J. M. Zoido, “Intra-observer and inter-observer variability of color-matching experimental data Variabilidad intra e inter-observadores en las igualaciones de color,” Opt. Pur. y Apl. 37(1), 67–75 (2004).
40. Z. Dong, J. Ho, Y. F. Yu, Y. H. Fu, R. Paniagua-Dominguez, S. Wang, A. I. Kuznetsov, and J. K. W. Yang, “Printing beyond sRGB color gamut by mimicking silicon nanostructures in free-space,” Nano Lett. 17(12), 7620–7628 (2017). [CrossRef] [PubMed]
41. R. Lupton, M. R. Blanton, G. Fekete, D. W. Hogg, W. O’Mullane, A. Szalay, and N. Wherry, “Preparing red-green-blue images from CCD data,” Publ. Astron. Soc. Pac. 116(816), 133–137 (2004). [CrossRef]
42. S. Alrasheed and E. Di Fabrizio, “Plasmonic nanospherical dimers for color pixels,” Nanomater. Nanotechno. 8(5), 1–7 (2018).
43. Y. L. Xu and B. Å. S. Gustafson, “A generalized multiparticle Mie-solution: further experimental verification,” J. Quant. Spectrosc. Radiat. Transf. 70(4–6), 395–419 (2001). [CrossRef]
44. P. B. Johnson and R. W. Christy, “Optical constants of the noble metals,” Phys. Rev. B 6(12), 4370–4379 (1972). [CrossRef]
45. H. Wang, “Plasmonic refractive index sensing using strongly coupled metal nanoantennas: nonlocal limitations,” Sci. Rep. 8(1), 9589 (2018). [CrossRef] [PubMed]
46. H. Kollmann, X. Piao, M. Esmann, S. F. Becker, D. Hou, C. Huynh, L. O. Kautschor, G. Bösker, H. Vieker, A. Beyer, A. Gölzhäuser, N. Park, R. Vogelgesang, M. Silies, and C. Lienau, “Toward plasmonics with nanometer precision: nonlinear optics of Helium-Ion milled gold nanoantennas,” Nano Lett. 14(8), 4778–4784 (2014). [CrossRef] [PubMed]
47. R. W. G. Hunt and M. R. Pointer, Measuring Colour, 4th ed. (Wiley, 2011).
48. T. Smith and J. Guild, “The C.I.E. colorimetric standards and their use,” Trans. Opt. Soc. 33(3), 73–134 (1931). [CrossRef]
49. M. Stokes, M. Anderson, S. Chandrasekar, and R. Motta, “A standard default color space for the internet – sRGB, Ver. 1.10, 1996,” https://www.w3.org/Graphics/Color/sRGB.html, (accessed January 2019).
