Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2021 Jul 12;6(29):18928–18938. doi: 10.1021/acsomega.1c02179

Quantitative Surface-Enhanced Raman Spectroscopy for Field Detections Based on Structurally Homogeneous Silver-Coated Silicon Nanocone Arrays

Hao Fu †,‡, Haoming Bao †, Hongwen Zhang †,*, Qian Zhao †, Le Zhou †,‡, Shuyi Zhu †,‡, Yi Wei †,‡, Yue Li †, Weiping Cai †,‡,*
PMCID: PMC8320141  PMID: 34337232

Abstract

graphic file with name ao1c02179_0009.jpg

Practical application of surface-enhanced Raman spectroscopy (SERS) is greatly limited by the inaccurate quantitative analyses due to the measuring parameter’s fluctuations induced by different operators, different Raman spectrometers, and different test sites and moments, especially during the field tests. Herein, we develop a strategy of quantitative SERS for field detection via designing structurally homogeneous and ordered Ag-coated Si nanocone arrays. Such an array is fabricated as SERS chips by depositing Ag on the template etching-induced Si nanocone array. Taking 4-aminothiophenol as the typical analyte, the influences of fluctuations in measuring parameters (such as defocusing depth and laser powers) on Raman signals are systematically studied, which significantly change SERS measurements. It has been shown that the silicon underneath the Ag coating in the chip can respond to the measuring parameters’ fluctuations synchronously with and similar to the analyte adsorbed on the chip surface, and the normalization with Si Raman signals can well eliminate the big fluctuations (up to 1 or 2 orders of magnitude) in measurements, achieving highly reproducible measurements (mostly, <5% in signal fluctuations) and accurate quantitative SERS analyses. Finally, the simulated field tests demonstrate that the developed strategy enables quantitatively analyzing the highly scattered SERS measurements well with 1 order of magnitude in signal fluctuation, exhibiting good practicability. This study provides a new practical chip and reliable quantitative SERS for the field detection of real samples.

1. Introduction

Surface-enhanced Raman spectroscopy (SERS) can greatly amplify the weak inelastic light scattering of analytes and offer their molecular fingerprints,1−6 and hence, it is promising to become an ultrasensitive and accurate detection technique applied in the fields of chemistry,7,8 biological sciences,4,9 materials,10,11 and environment.12−14 However, in many cases, quantitative SERS is restricted by the difficulty in obtaining uniform and highly reproducible SERS signals.15,16 The signal fluctuations are mainly attributed to two aspects: the nonuniform SERS chips (substrates), including their structural and composition heterogeneities, and the fluctuations of the measuring parameters (such as focusing depth and laser power).17−19 The structurally nonuniform SERS substrates would induce inhomogeneous distribution of hot spots, which results in poor signal uniformity and eventually difficulty in quantitative detection. The fluctuations of measuring parameters, which easily occur among different operators, different Raman spectrometers, and different test sites and moments, would induce great uncertainty of the Raman measurements, especially during the on-site detection or field tests.20,21 The fluctuations in measured Raman signals could be up to 1 to 2 orders of magnitude21 (also see the following Sections 3.2.3 and 3.4.2), which make it impossible to carry out quantitative analysis. Various attempts have been made to try to overcome the above obstacles for quantitative SERS.

For structural uniformity, many works utilized ordered plasmonic metal arrays as SERS substrates via self-assembling and patterning fabrications.22−29 Such arrays ensure that the hotspots and the adsorbed target molecules (or analyte) are homogeneously distributed on the chips at the scale of the incident light spot. For instance, Liu et al. fabricated Ag nanoplate-built hollow microsphere arrays based on organic colloidal template-assisted electrodeposition, which showed good signal reproducibility in detecting trace cyanide.30 Zhu et al. reported that the ordered Ag nanorod bundle arrays exhibited high signal homogeneity and reproducibility and could quantitatively analyze the phenolic pollutants.22 To a certain extent, these substrates could basically meet the requirements of signal uniformity with less than 15% in relative standard deviation (RSD).31 However, it is still difficult for these structurally uniform SERS chips to avoid the signal uncertainty induced by the measuring parameter’s fluctuations from different operators and Raman spectrometers as well as different test sites and moments during the nonlab or field tests.

Some researchers tried to reduce the interferences induced by the fluctuations in measuring parameters via modifying the SERS substrates with internal standard (IS) substances, which is named as the IS method.20,32−38 During SERS measurements, the IS substances and the analytes are distributed in the similar microenvironment and show synchronous evolutions in Raman signal intensity when the measuring parameters fluctuate.15,20,21,32,36−40 Thus, using the Raman signals of IS substances to normalize the analyte’s signals could effectively improve the reproducibility in measurements.15,32,37,40 According to the position of the IS substance on the SERS substrates, the IS method can be further classified into external mode and embedded mode. For the external mode, the plasmonic metal SERS substrates are generally modified or mixed with IS substances, including organic molecules (e.g., rhodamine 6G,32 acetone,35 thiolate ligands,41 and 4-mercaptopyridine42) and inorganic substances (e.g., graphene,37,40 silicon,43,44 and Mxene33). For instance, Yang et al.42 utilized 4-mercaptopyridine as the IS substance to normalize the SERS signals of analytes and effectively reduce the signal’s fluctuation, realizing quantitative analysis of various drugs. As for the embedded mode, usually, some special organic IS molecules are immobilized at the interfaces between the core parts and shell layers in the composite plasmonic metal nanoparticles (NPs).21,36,39,45 For instance, Shen et al.20 embedded some organic molecules (such as 4-mercaptopyridine, 4-mercaptobenzoic acid, or thiophenol) into the interfaces between Au cores and Ag shells to form core-molecule-shell NPs, obviously decreasing the fluctuations in measurements and realizing quantitative SERS analysis in a large range of analyte concentrations during lab tests.

In principle, the IS method, whether it is the external mode or the embedded mode, needs one-to-one distribution of the IS substance and the analyte on the SERS substrate to ensure that the normalized signal of the analyte maintains a stable value. Otherwise, even if there is no fluctuation of the measuring parameters, the inconsistent distribution between the IS substance and the analyte can still lead to poor reproducibility of the normalized signals, which would make the IS method lose its calibration function. What is more, the one-to-one distribution of the IS substance and the analyte on the substrate is usually very difficult to be achieved. In addition, the organic IS molecules in the external mode may compete with the target molecules to occupy the sites on the hot spots and generate interference to Raman bands, resulting in reduced detection sensitivity and difficulty in signal recognition. For the embedded mode, it is highly dependent on the complicated core-molecule-shell nanostructures and the homogeneous modification of IS molecules in the plasmonic gaps, which increase the difficulty in the substrate fabrication. All in all, the accurate quantitative SERS detection, especially during the field tests which often involve different operators, different Raman spectrometers, and different test sites and moments and hence produce highly scattered measurements, is still expected and remains to be a bottleneck for SERS practical applications.

Herein, we present a facile strategy for accurate and quantitative SERS for field tests via designing a structurally homogeneous Ag-coated silicon nanocone array (Ag/SiNCA). Such an array is fabricated as the SERS chips by depositing Ag on the template etching-induced Si NCA. Typically, 4-aminothiophenol (4-ATP) is taken as the analyte. It is demonstrated that the Si right below the Ag coating can respond to the fluctuations of the measuring conditions (such as focusing depth, laser power, operators, and instruments) synchronously with and similar to the adsorbed analyte, and the big fluctuations in measurements can well be eliminated by Si signal normalization or calibration, achieving excellent signal reproducibility (mostly, <5% in signal fluctuations) and accurate quantitative SERS detection. Finally, the simulated field tests show that the presented strategy is suitable for the quantitative detection on-site with the highly scattered SERS measurements (1 or 2 orders of magnitude in signal fluctuation), exhibiting good practicability. This study provides not only a new practical chip but also effective quantitative SERS for the field tests with large fluctuations in measuring conditions.

2. Quantitative Strategy and SERS Chip Design

As mentioned above, although the current SERS chips can mostly ensure the homogeneous distribution of the hotspots on the substrates at the scale of the incident spot, it is difficult to effectively eliminate the signal fluctuation induced by the variations in measuring parameters. The IS methods could effectively reduce the disturbance from the variations in measuring conditions but need using some organic molecules, which would cause complex Raman spectral patterns, in addition to the difficulty in obtaining the one-to-one distribution of the IS substance and the analyte.

Since the characteristic Raman band of silicon is at 520 cm–146,47 and located in the silent region (usually, below 1000 cm–1) of many organic analytes, if the silicon is used as the IS substance, its Raman band would not overlap with the vibrations of the analyte. We can thus strategize quantitative SERS by designing a Si-embedded plasmonic metal nanostructured array as the SERS chip, as illustrated in Figure 1. In this chip, a thin layer of plasmonic metal film is uniformly coated on an ordered Si nanostructured array (Figure 1a). Obviously, such a SERS chip can not only ensure the uniform distribution of the hotspots and the adsorbed analyte on the surface due to its ordered structure but also achieve the one-to-one distribution of the analytes and the Si substance due to the consistent structure between the Si array and the thin plasmonic metal film. In this case, the Raman signal intensities of the Si (ISi) and the adsorbed analyte (IA) could synchronously and similarly change with fluctuant measuring parameters (the inset of Figure 1a,b). If using ISi to normalize IA or the ratio IA/ISi as the Raman measurements, it is expected that we could effectively eliminate the interferences from the fluctuations of the measuring parameters (Figure 1c) and achieve the highly reproducible measurements and hence the quantitative SERS detection.

Figure 1.

Figure 1

Open in a new tab

Schematic illustration for the strategy of quantitative SERS based on Si-embedded plasmonic metal nanostructured array. (a) Thin plasmonic metal film-coated ordered Si-nanostructured array as a SERS chip. Inset: the anatomy of an analyte-adsorbed building block in the array. (b) Schematic spectral measurements at different sites or by different operators or on different Raman spectrometers. (c) Schematic histogram of the ratio IA/ISi corresponding to the Raman spectra in (b).

According to the strategy in Figure 1, the thin plasmonic metal film-coated Si nanostructured arrays are crucial. The building blocks in the Si array should be arranged in an ordered pattern, with much smaller size than the laser spot so that the good structural uniformity is ensured at the scale of the laser spot. Also, the plasmonic metal coating should be compact and uniform and appropriate in thickness. An overly thick coating would mask the Si signal, while an excessively thin metal coating would lead to low SERS activity. On these bases, here we design a Ag/SiNCA, which is fabricated by depositing Ag on the template etching-induced Si array (see the Experimental Section), and take it as the SERS chip and 4-ATP molecules as the typical analyte to systematically study the influences of fluctuations in measuring parameters on Raman signals and demonstrate the effectiveness and practicability of this strategy.

3. Results and Discussion

3.1. Morphology and Evaluation of SERS Chips

The self-assembled polystyrene (PS) colloidal monolayer was first fabricated on a Si wafer (3 cm × 3 cm) (Figure S1). Si-ordered NCA was then prepared by reactive ion etching (RIE) of the PS-covered Si wafer. The Ag/SiNCA was thus fabricated via depositing a thin layer of Ag on the Si NCA, showing a green color which arises from the diffraction of a periodic structure,48,49 as demonstrated in the inset of Figure 2a. The Ag coating was ca. 20 nm in deposition thickness. Field emission scanning electron microscopic (FESEM) observations show that the Ag/SiNCA consists of conic building blocks, which are 450 nm in height, 500 nm in period, and homogeneously arranged in an ordered hexagonal pattern, as illustrated in Figure 2a,b. Correspondingly, the energy-dispersive spectroscopic (EDS) measurements indicate that the array contains only Ag and Si, and the elemental mapping analysis shows that Si and Ag are uniformly distributed on the array (Figure S2). The optical absorption spectra, based on the diffuse reflection spectral measurements, show that there are two peaks around 460 and 530 nm for the Ag/SiNCA, while the bare Si NCA assumes high absorptivity in the whole measured optical region, as illustrated in Figure S3a. The optical absorption peak at 460 nm and 530 nm should be attributed to the Fabry–Perot (FP) mode of the Ag coating, and the hybrid mode of the FP and surface plasmonic resonance, respectively.46,50,51

Figure 2.

Figure 2

Open in a new tab

Morphological observation and SERS signal uniformity of the as-prepared Ag/SiNCA. (a,b) FESEM images (tilted view and cross-sectional view, respectively). The inset in (a) is a photo of the as-prepared Ag/SiNCA with 3 cm × 3 cm. (c) SERS spectrum of the Ag/SiNCA-based chip after soaking in 4-ATP solution (10–7 M) and exciting at 532 nm. (d) SERS spectra from 30 randomly selected sites on the chip after soaking in 10–7 M 4-ATP solution. (e,f) Histograms for the peak intensities of Si at 520 cm–1 and 4-ATP at 1435 cm–1 before and after normalized by Si peak intensity [data from the SERS spectra in (d)]. (g) RSD values of 4-ATP peak intensities at 1075, 1390, and 1435 cm–1 before and after Si-signal normalization.

Such an Ag/SiNCA was then cut into pieces (3 mm × 3 mm) as SERS chips. Figure 2c presents the Raman spectrum of the chip after soaking in the 4-ATP solution with 10–7 M and exciting at 532 nm. The peaks at 1075 and 1575 cm–1 belong to the vibrations of C–S and C–C, respectively,52−55 which are from a1 modes of 4-ATP molecules. The other three prominent characteristic peaks at 1140, 1390, and 1435 cm–1 are assigned to the b2 modes of 4-ATP.52−55 The sharp Raman peak at 520 cm–1, which is far away from the fingerprint area of the 4-ATP molecules, belongs to the Raman vibration of the Si crystal.46,47,50 In addition, it has been shown that the excitation at 532 nm is optimal compared with that at 633 nm or 785 nm (Figure S3b). Hence, an excitation wavelength at 532 nm was chosen to demonstrate the SERS performance of the chip. Then, the characteristic peaks of 4-ATP at 1075 and 1575 cm–1 were chosen to evaluate the enhancement factor (EF) of the chip according to previous reports.30,56 The EF values of the chip are 1.5 × 106 and 2.3 × 106 for the peaks at 1075 and 1575 cm–1, respectively, as shown in Figure S3c,d (the details of the EF determination are seen in the Supporting Information).

In order to demonstrate the Raman signal uniformity of the Ag/SiNCA-based SERS chip, the Raman spectra were measured from 30 randomly selected sites on the 4-ATP solution-soaked chip, as shown in Figure 2d. All these Raman spectra are similar in pattern, but there are small differences in peak intensities. The RSD in intensity was calculated to be 4.8 and 5.4% for the peaks at 520 and 1435 cm–1, respectively, as illustrated in Figure 2e. Such low RSD values indicate that the structure and hot spots on the SERS chip are homogeneously distributed at the scale of the incident laser spot. In addition, from the histograms in Figure 2e, it is noticed that the Si Raman peak at 520 cm–1 exhibits signal fluctuations similar to the characteristic vibration of 4-ATP at 1435 cm–1 and the nearly synchronous change with 4-ATP peaks. Such signal fluctuation could mainly originate from the fluctuations of measuring parameters (such as focusing depth and laser power. see the next section). Further, the difference in the 4-ATP Raman signals measured at different sites on the chip can be reduced if the intensity (ISi) of the Si peak at 520 cm–1 is used to normalize the intensities (IA) of the 4-ATP peaks in the same Raman spectrum or the ratio IA/ISi is used as the peak intensity of 4-ATP. Typically, for the peak at 1435 cm–1, the RSD value of the intensity was decreased from 5.4 to 4.1% after such Si signal normalization, as shown in Figure 2f. Similarly, the RSD values of the normalized intensities for the peaks at 1075 and 1390 cm–1 were also obviously decreased (Figure 2g). We also tested the analytes without sulfhydryl or isothiohydrogen, such as rhodamine 6G (R6G) and crystal violet (CV), as illustrated in Figures S4 and S5, respectively. The results are similar to those for 4-ATP, or the reproducibility is quite good and the Si-signal normalization can significantly improve the reproducibility of Raman measurements.

Finally, the influence of the Ag coating thickness on the SERS performances was examined for the Ag/SiNCA-based chips. When the thickness of the silver coating was increased from 10 to 40 nm, the EF value to the analyte (4-ATP) increased from 1.1 × 106 to 2.7 × 106 for the peak at 1075 cm–1, while the Raman signal intensity of Si was ever-decreasing, as shown in Figure S6. Such thickness dependence is attributed to the contribution of the Ag coating to the field enhancement and its masking effect on Si. It should be noted that the over-thin Ag coating (say, 10 nm or less) can cause the strong Si signal that may mask the analyte’s signals, especially when the concentration is low, while the over-thick coating (say, 40 nm or larger) would lead to the weak Si signal that is not conducive to the quantitative analysis of the analyte with a high concentration. Therefore, the thickness of the plasmonic metal coating on the SiNCA should be balanced according to need. In the next sections, only the 20 nm-thick Ag/SiNCA was used.

In addition, the 250 nm PS-covered Si wafer without etching and the bare Si wafer were also deposited with a 20 nm thick silver film on them (denoted as Ag/PS array and Ag NP film, respectively). The corresponding morphology and uniformity in the Raman signal are illustrated in Figures S7 and S8, showing less homogeneous structures and much lower EF values (∼105 in order of magnitude) than those of the Ag/SiNCA-based chip (Figure S9 and Table S1).

3.2. Antimeasuring Condition’s Fluctuations

In practical applications or field tests, the measurement often involves different operators and test sites, different environments, and Raman spectrometers as well as different test moments, which unavoidably lead to fluctuations in the measurement parameters, such as focusing depth and laser power. Occasionally, the Raman intensity was normalized with the laser power and integration time.57 However, Raman is a scattering process and the collected signals depend highly on the measuring conditions. If the measuring condition is changed during the test, such as changing the objective, the normalized intensity will still fluctuate. Such fluctuations would induce highly scattered measurements and hence quantitative analyses would be difficult. Here, we demonstrate the validity of the Ag/SiNCA chip-based quantitative SERS strategy even in the case with significant measuring parameters’ fluctuations or variations.

3.2.1. Focusing Depth Variation

Focusing depth is an important parameter in Raman spectral measurement. Here, we adjusted the position of the SERS chip in the Z-axis direction and kept the laser beam fixed to simulate the fluctuation of the focusing depth, as schematically shown in Figure 3a. The optimal position, at which the chip is located and the strongest Raman signal is obtained, is defined as the focal plane or zero defocusing distance. Figure 3b presents the Raman spectra, under different defocusing distances (from −50 to 50 μm), for the Ag/SiNCA-based chip after soaking in 10–7 M 4-ATP solution and drying. A slight variation of the focusing depth induces a significant change in the intensity of Raman signals. The Raman signals from 4-ATP and Si synchronically decay with the rising defocusing distance, and the maximum fluctuation in the Raman signal is more than 1 order of magnitude in the used defocusing range, as shown in Figure 3c (the upper frame). However, if the Raman peak intensities of 4-ATP are normalized by the Si signal in the corresponding spectrum, the influence of the variation in the focusing depth on the measured Raman signals can be nearly completely eliminated, as demonstrated in Figure 3c (the lower frame). The normalized intensities (or IA/ISi) of the characteristic peaks of 4-ATP are nearly independent of the defocusing distance varied in the large range from −50 to 50 μm, while the RSD values for the absolute peak intensities are as high as ∼85%, as shown in Figure 3d. These indicate that the defocusing-induced fluctuation of Raman signals can be well eliminated via using the Ag/SiNCA-based chip and Si signal normalization.

Figure 3.

Figure 3

Open in a new tab

Defocusing-induced SERS-signal fluctuation and its normalization. (a) Schematic illustration of adjusting the defocusing distance. (b) SERS spectra of the 10–7 M 4-ATP solution-soaked Ag/SNCA-based chip under the defocusing distances varied from −50 to 50 μm. (c) Upper frame: the absolute peak intensities at 520 and 1435 cm–1 vs the defocusing distance; lower frame: the Si signal-normalized peak intensities (IA/ISi) vs the defocusing distance [the data from (b)]. (d) RSD values of the 4-ATP peak intensities before and after Si-signal normalization [the data from (b) and the lower frame in (c)]. (e,f) RSD values of the 4-ATP peak intensities before and after Si-signal normalization for the 10–7 M 4-ATP solution-soaked Ag/PS array-covered Si wafer and the Ag NP film-covered Si wafer, respectively (the data from Figures S10 and S11, and defocusing range: −10 to 10 μm).

In contrast, for the Ag/PS array-covered Si wafer, the Raman signals from 4-ATP show quite different evolutions with the defocusing distance from those of Si, as shown in Figure S10. The RSD values of the 4-ATP peak intensities are higher than 30% and similar before and after Si-signal normalization, as shown in Figure 3e corresponding to the defocusing range from −10 to 10 μm. As for the Ag NP film-covered Si wafer, the RSD values of the normalized 4-ATP peak’s intensities are significantly higher than those for the Ag/SiNCA-based chip (Figures 3f and S11).

3.2.2. Laser Power Fluctuation

The laser power is also an important parameter in Raman measurement, and the intensity of the Raman signal increases with increasing power.18 Generally, the incident laser power would inevitably fluctuate or vary among the different operators and Raman spectrometers as well as the different test sites and environments, resulting in fluctuations of Raman spectral measurements. The influence of the power variation on the Raman signals was thus examined. We fixed the laser spot on the 4-ATP solution-soaked Ag/SiNCA chip and measured the Raman spectra under different laser powers varying in the range from 2.5 to 500 W/cm2, as shown in Figure 4a. The Raman peak intensities of 4-ATP and Si increase synchronously with the increasing laser power, as shown in Figure 4a or more intuitively in Figure 4b. The characteristic peaks of 4-ATP and Si significantly fluctuate in intensity up to about 160 times when the laser power was varying from 2.5 to 500 W/cm2. However, after the normalization by the Si signal, the intensities of the 4-ATP characteristic peaks are nearly independent of the power variations in the large range from 2.5 to 500 W/cm2, and the RSD values are less than 8.5% (Figure 4c). This suggests that the Si-signal normalization can effectively overcome the disturbance from laser power fluctuations for the Ag/SiNCA-based chip. For the Ag/PS array-covered Si wafer, however, the normalized intensities of the 4-ATP peaks are still highly scattered with RSD values more than 40%, probably due to the light absorption of the PS monolayer between the Ag coating and the Si wafer, as illustrated in Figure S12.

Figure 4.

Figure 4

Open in a new tab

Laser power variation-induced SERS signal fluctuation and its normalization. (a) Schematic illustration of the laser spot, with powers varied from 2.5 to 500 W/cm2, on the 10–7 M 4-ATP solution-soaked Ag/SiNCA-based chip and the corresponding SERS spectra. (b) Absolute peak intensities at 520 and 1435 cm–1 vs the laser power [data from (a)]. (c) Si signal-normalized peak intensities (IA/ISi) vs the laser power.

3.2.3. Interferences from Operators and Instruments

Comprehensively, the Raman measurements could significantly fluctuate among different operators and instruments as well as at different test moments and sites, which would induce fluctuations in both laser power and focusing depth. Figure 5a shows the intensity histogram of the 4-ATP characteristic peak at 1435 cm–1, measured for the 10–7 M 4-ATP solution-soaked Ag/SNCA-based chip by different operators or at different moments or on different Raman spectrometers but under the same measuring requirements (power, focusing depth, integral time, etc.). The results measured by operator O1 for the 10–7 M 4-ATP-spiked samples using lake (Dongpu Lake in Hefei, China) and ground water are also given in Figure 5a. The corresponding spectral measurements are shown in Figure S13. Obviously, the measured results are highly scattered among different operators or different spectrometers. The maximal difference among these measured results was up to 6 times, and the RSD value was up to 40.3% for the peak intensity at 1435 cm–1 in this group of experiments. In other words, the measured results are of big uncertainty. However, after Si-signal normalization, the measurements are highly reproducible and almost the same among these measuring conditions, with only a small fluctuation (RSD = 2.0% for the peak at 1435 cm–1), as clearly shown in Figure 5b. For the analyte R6G, the measured intensities of the main peak are also highly scattered (RSD > 30%) among different conditions, but the Si signal-normalized results show highly consistent measurements (RSD = 1.6%), as shown in Figure S14. These indicate that the difference in measurements among the different operators and Raman spectrometers as well as the different test moments and aqueous solutions and so forth can be eliminated almost completely via using the Ag/SiNCA-based chip and normalizing with the Si signal.

Figure 5.

Figure 5

Open in a new tab

Histograms of the 4-ATP characteristic peak intensity at 1435 cm–1, measured for the 10–7 M 4-ATP solution-soaked chip by different operators or at different moments or on different Raman spectrometers but under the same measurement requirements before (a) and after (b) Si-signal normalization. O1, O2, O3, and O4 represent the results obtained by different operators; O1-2 and O1-3 represent the measurements obtained by operator O1 at different test moments; I2 represents the measurements obtained by operator O1 on another Raman spectrometer; LW and GW are the results measured by operator O1 for the 4-ATP spiked samples using lake (Dongpu Lake in Hefei, China) and ground water, respectively (All data are from Figure S13).

3.3. Synchronous Response-Induced Good Reproducibility

As mentioned above, via using Ag/SiNCA-based chip and Si-signal normalization, we can effectively eliminate the interferences from the fluctuations of the measuring conditions and ensure excellent reproducibility in measurements and hence the quantitative SERS analyses. This is attributed to the special structure of the Ag/SiNCA-based chips. Here, the highly ordered structure ensures that the hotspots are homogeneously distributed and the analyte molecules are adsorbed uniformly on the chip’s surface at the scale of laser spot. The Si NCA, which is right underneath the Ag coating film and has a focusing depth similar to the Ag coating film, can respond to the fluctuations of the measurement parameters (or the focusing depth and laser power) synchronously with and similar to the analyte adsorbed on the Ag film due to their one-to-one distribution (Figure S15a). The disturbances from the fluctuations of the measuring parameters can thus be eliminated effectively via Si-signal normalization, leading to good signal reproducibility (Figures 3c, 4c, and 5b). For the Ag/PS array-covered Si wafer, however, the Ag coating layer has a significantly different focusing depth from the Si wafer due to the PS monolayer between them (Figure S15b). In this case, there is no one-to-one distribution between the adsorbed analyte and the silicon, or they would not synchronously response to the fluctuations of the measurement parameters, leading to the highly scattered measurements even after Si-signal normalization (Figures 3e, S10c and S12c).

3.4. Quantitative SERS Analysis and Practical Application

The variations in the measuring conditions could induce significant fluctuation or uncertainty in the Raman measurements, which does not allow the quantitative analysis. However, if using the Ag/SiNCA-based chip and normalizing with the Si signal, we can well eliminate the disturbance from fluctuations in measuring parameters and achieve the highly reproducible measurements and hence quantitative SERS analyses.

3.4.1. Lab Test

The Raman spectra were first measured by the same operator on the same Raman spectrometer at the lab for the Ag/SiNCA-based chip after soaking in 4-ATP solutions with different concentrations, as shown in Figure 6a. The characteristic peaks of 4-ATP remain distinguishable even when the concentration was as low as 10–10 M (or 12.5 ppt), indicating the high sensitivity of the Ag/SiNCA-based chip. Figure 6b shows the concentration-dependent intensities of the 4-ATP peaks at 1075, 1390, and 1435 cm–1. These characteristic peaks increased in intensity with the rising 4-ATP concentration (C) up to 5 × 10–7 M, and the higher concentration (10–6 M or higher) led to the insignificant change in the peak intensity (or the signal saturation). The peak intensities are obviously scattered. However, the Si signal-normalized characteristic peak intensities show good linear relation with the logarithmic concentration in the range from 5 × 10–9 M (∼0.6 ppb) to 10–6 M (∼0.1 ppm), as illustrated in Figure 6c. In addition, we performed such a test with a handheld Raman spectrometer (type: ATR6500) and also achieved the good linear relation between the normalized peak intensities and the logarithmic concentration, as shown in Figure S16. According to such a linear relation, we can thus quantitatively determine the 4-ATP concentration. These demonstrate that the accurate quantitative SERS detections can be achieved based on the Ag/SiNCA-based chip and Si-signal normalization.

Figure 6.

Figure 6

Open in a new tab

Quantitative SERS analyses of 4-ATP by lab tests. (a) SERS spectra of the Ag/SiNCA-based chip after soaking in the 4-ATP solutions with different concentrations. (b) Intensities of the 4-ATP peaks at 1075, 1390, and 1435 cm–1 as a function of the logarithmic concentration (C). (c) Plots of the Si signal-normalized peak intensities vs the logarithmic concentration. The lines are the linear fitting results.

Similarly, for some other target molecules which can be adsorbed on the plasmonic metal surface, we can also achieve accurate quantitative detection in the lab. Typically, R6G and CV were selected as target molecules because of their affinity to the metallic surface through electrostatic interaction. Figure S17a,c shows the concentration-dependent SERS spectra of R6G and CV, respectively. Here, the prominent peaks at 612 cm–1 and 913 cm–1 were chosen as the identification positions for the quantitative analyses of R6G and CV, respectively.37,39,45,58 The intensities of these peaks increase with the increasing concentration up to 10–4 M. The Si signal-normalized intensities show much better linear relations with the logarithmic concentration than those without Si-signal normalization for both analytes, as demonstrated in Figure S17b,d.

3.4.2. Practical Applications— Simulated Field Tests

Finally, the practicability of the proposed strategy was examined by simulating the field test which would lead to the badly reproducible measurements. The spectral measurements were carried out for the Ag/SiNCA-based chip after soaking in the 4-ATP spiked solutions with different concentrations, by different operators, at different moments, or on different Raman spectrometers but under the same measurement requirements. Here, the 4-ATP-spiked solutions were prepared by using deionized water or lake (Dongpu Lake in Hefei, China) water or groundwater. The corresponding Raman spectra are shown in Figure S18. We can thus make plots of the characteristic peak intensity versus logarithmic concentration. Representatively, Figure 7a shows the results for the peak at 1435 cm–1 under different measuring conditions. The data points are very scattered, and the measurements seem to be highly uncertain and unreliable. After the normalization using the corresponding Si signals, however, all these scattered data points, from the different operators and Raman spectrometers as well as different water, are close to or fall on a straight line, showing independence of the varied measuring conditions, as clearly shown in Figure 7b. Similarly, for the characteristic peaks at 1075 and 1390 cm–1, we can also obtain the good linear relations after the corresponding Si-signal normalization (Figure S19). In addition, the parameter values obtained by the linear fitting for the plots of IA/ISi versus Log(C) are close to those of the corresponding lines shown in Figure 6c, as listed in Table S2. All these results demonstrate the practicability of the strategy, as shown in Figure 1, and the possibility of the accurate and quantitative SERS-based detection on-site.

Figure 7.

Figure 7

Open in a new tab

Quantitative SERS detection of 4-ATP in the simulated field tests (see the text in detail). (a) Raman intensities of the 4-ATP peak at 1435 cm–1 vs logarithmic concentration. (b) Plot of the Si signal-normalized peak intensities at 1435 cm–1 vs the logarithmic concentration. The line is the linear fitting result. The meanings of O1, O2, O3, O4, O1-2 and O1-3, I2, LW, and GW are seen in the caption of Figure 5. All data are from Figure S18.

4. Conclusions and Remarks

In summary, we have developed a strategy of reliable quantitative SERS for the field tests based on the Ag/SiNCA-based chips, which were fabricated by depositing a thin layer of uniform Ag film on the template etching-induced Si NCAs. The prepared Ag/SiNCA-based chips are structurally homogeneous and ordered, which ensures that the hot spots and the adsorbed analytes are uniformly distributed on the chip surface, exhibiting good uniformity of SERS signals with an RSD value less than 4.1%. Since the Si right below the Ag-coating film can correspond to the adsorbed analyte one-to-one, the SERS measurements can be normalized by the corresponding Si signals, achieving quantitative analyses. The validity of the strategy has been demonstrated by taking 4-ATP, CV, and R6G as the typical analytes. Due to the homogenous and ordered structure of the Ag/SiNCA-based chip, the Si NCA underneath the Ag coating can respond to the fluctuations of the measuring parameters (such as laser power and focusing depth), possibly from different operators and instruments as well as different measuring moments and sites, and so forth, synchronously with and similar to the adsorbed analyte, which can well eliminate the signal fluctuations in measurements just by the Si-signal normalization and achieve good signal reproducibility (mostly, <5% in signal fluctuations). Finally, the practicability of the proposed strategy was demonstrated by simulating the field tests. Although the original measurements from the different operators and Raman spectrometers and so forth were highly scattered and uncertain, the Si signal-normalized results show independence of the varied measuring conditions and can give highly certain measurements. We have thus realized the accurate and reliable quantitative SERS detection on-site and demonstrated the good practicability of the presented strategy.

Furthermore, for the strategy of quantitative SERS, as shown in Figure 1, the other plasmonic metal (Au or Au–Ag alloys)-coated Si nanostructured arrays, in addition to Ag, could also be used as the SERS chips. Typically, Au-coated Si NCAs have been fabricated and used as the chip. We have obtained similar results, as demonstrated in Figure S20. Au-based SERS chips have a lower EF than Ag-based ones but have a higher stability because Ag suffers easily from oxidation.59

Finally, it should be mentioned that the preconditions of IS-based quantification are consistency (or uniformity) in the signals from both IS and analytes on a whole SERS chip and synchronization in the signal change of both IS and analytes with the variations in measuring conditions. Both the preconditions are indispensable. If the signals from either the IS or the analytes are inconsistent on the SERS chip, we cannot obtain good reproducibility. For instance, for the analytes which interact weakly with plasmonic metals, they would hardly be adsorbed on the chip after the soaking process or could be only in-homogeneously distributed on the chip after dropping the analyte-contained solution on it and drying. In this case, we can not achieve good reproducibility and quantification analyses if using the above chip without any modification. Nevertheless, for these weakly interacted analytes, the quantitative SERS detection could also be achieved if we appropriately surface-modify the chip in such a way that the analytes can be uniformly adsorbed on the modified chip, which is in progress. All in all, this work provides practical chips and an effective route for reliable quantitative SERS detection. Especially, this strategy is suitable for the quantitative SERS detection in the field tests, which often involve different operators and Raman spectrometers or environmental conditions as well as different test sites and measuring moments and could produce highly uncertain measurements.

5. Experimental Section

5.1. Chemicals and Materials

The PS sphere (PS, 500 and 250 nm in diameter) suspension (5% wt) and the (100)-plane polished silicon wafers were purchased from Shanghai Huge Biotechnology Co., Ltd and Zhejiang Lijing Silicon Material Co., Ltd, respectively. Ethanol, acetone, 4-ATP, R6G, and CV were bought from Alfa Aesar Corporation with analytical purity. The sulfur hexafluoride (SF6) etching gas was offered by Nanjing Special Gas Factory Co., Ltd. The deionized water used in our experiments had a resistivity of 18.2 MΩ·cm at 25 °C and was produced in a Millipore Milli-Q system.

5.2. Fabrication of the Ag/SiNCA

The Ag/SiNCA was fabricated via PS colloidal template-assisted RIE process and sputtering deposition technique, as schematically illustrated in Figure S21.

First, a uniform PS colloidal monolayer template was prepared on a Si wafer through an air–water interface self-assembly method, according to our previous works.60−62 In brief, the PS suspension was diluted into the ethanol to form a mixture with a volume ratio of 1:1 and slowly dropped on a deionized water film-covered hydrophilic silicon wafer (3 cm × 3 cm) from its edge. Due to the energy minimization, the PS self-assembled on the air–water interface and formed a close-packed monolayer. After removing the water with a filter paper and drying naturally, the ordered PS colloidal monolayer template was fabricated on the Si wafer (Figure S21a).

The as-prepared PS-covered Si wafer was then heated at 70 °C for 20 min to fix the PS spheres on the wafer firmly and etched with SF6 plasma in a RIE machine (200 W in power and 60 sccm in gas flow rate). After etching for 150 s, a highly ordered Si NCA was obtained with the much smaller etched (or residual) PS on its top (Figure S21b). Subsequently, the etched sample was rinsed with ethanol and annealed at 600 °C for 2 h to remove the residual PS (Figure S21c). Finally, a thin silver film was coated on the Si NCA via sputtering deposition for a certain time at the rate of 0.5 nm/s in a VTC-16-SM magnetron sputtering instrument, and the Ag/SiNCA was thus prepared (Figure S21d).

Also, the as-prepared PS-covered Si wafer without etching and the bare Si wafer were deposited with a thin silver film on them for reference.

5.3. Characterization

The morphological observations and composition analyses were performed on a field emission scanning electron microscope (FEI Sirion 200) equipped with an energy dispersion spectroscope (Oxford IE250X-Max50). The optical absorption spectra based on the diffuse reflection spectral measurements were recorded on a Shimadzu UV-2600 spectrometer.

5.4. Raman Spectral Measurements

The as-obtained Ag/SiNCA was cut into 3 mm × 3 mm pieces as SERS chips. Such Ag/SiNCA-based chips were immersed into analyte (4-ATP, R6G, or CV) aqueous solutions (50 mL in volume) with certain concentrations. After soaking for 24 h, the chips were taken out, rinsed with water, and dried in the flow of N2. Then, Raman spectral measurements were performed on a confocal microprobe Raman spectrometer (Renishaw inVia). The laser powers were 1 mW for 532 nm, 1.7 mW for 633 nm, and 2.5 mW for 785 nm. The spot of incident laser beam on chips was ca. ∼5 μm in diameter, and the acquisition time of Raman signals was 1 s. All obtained Raman spectra were given after baseline corrections.

Acknowledgments

This work is financially supported by the National Key Research and Development Program of China (grant no. 2017YFA0207101), the Natural Science Foundation of China (grant nos. 11974352, 51771182, and 52001305), and the CAS/SAF International Partnership Program for Creative Research Teams.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.1c02179.

  • Determination details of the EF; values of the parameters for the results from the lab tests and the simulated field tests; preparation strategy, PS colloidal monolayer, EDS elemental mappings, and optical absorption spectra of the Ag/SiNCAs; Raman spectra of the 4-ATP solution-soaked Ag/SiNCA under different excitation wavelengths and its EF values at the peaks 1075 and 1575 cm–1; SERS spectra of the 4-ATP solution-soaked Ag/SiNCA-based chips with different Ag coating thicknesses and its EF values; morphology observation, SERS signal uniformity, EF values, and the defocusing and laser power-dependent SERS spectra for the reference substrates; SERS spectra of the 10–7 M 4-ATP-soaked Ag/SiNCA-based chip and reference substrates measured by different operators or on different Raman spectrometers and the plots of the SERS intensities of 4-ATP versus the logarithmic concentration; concentration-dependent SERS spectra of the R6G and CV solution-soaked Ag/SiNCA-based chips and the plots of the SERS intensities of R6G and CV versus the logarithmic concentration; and quantitative SERS analysis of 4-ATP concentration using Au-coated Si NCA-based chip (PDF)

Author Contributions

§ H.F. and H.B. contributed equally.

The authors declare no competing financial interest.

Supplementary Material

ao1c02179_si_001.pdf (2.3MB, pdf)

References

  1. Wang X.; Huang S.-C.; Hu S.; Yan S.; Ren B. Fundamental understanding and applications of plasmon-enhanced Raman spectroscopy. Nat. Rev. Phys. 2020, 2, 253–271. 10.1038/s42254-020-0171-y. [DOI] [Google Scholar]
  2. Ding S.-Y.; Yi J.; Li J.-F.; Ren B.; Wu D.-Y.; Panneerselvam R.; Tian Z.-Q. Nanostructure-based plasmon-enhanced Raman spectroscopy for surface analysis of materials. Nat. Rev. Mater. 2016, 1, 16021. 10.1038/natrevmats.2016.21. [DOI] [Google Scholar]
  3. Lee H. K.; Lee Y. H.; Koh C. S. L.; Phan-Quang G. C.; Xing Y. L. Designing surface-enhanced Raman scattering (SERS) platforms beyond hotspot engineering: emerging opportunities in analyte manipulations and hybrid materials. Chem. Soc. Rev. 2019, 48, 731–756. 10.1039/C7CS00786H. [DOI] [PubMed] [Google Scholar]
  4. Wang Z.; Zong S.; Wu L.; Zhu D.; Cui Y. SERS-activated platforms for immunoassay: probes, encoding methods, and applications. Chem. Rev. 2017, 117, 7910–7963. 10.1021/acs.chemrev.7b00027. [DOI] [PubMed] [Google Scholar]
  5. Li J.-F.; Anema J. R.; Wandlowski T.; Tian Z.-Q. Dielectric shell isolated and graphene shell isolated nanoparticle enhanced Raman spectroscopies and their applications. Chem. Soc. Rev. 2015, 44, 8399–8409. 10.1039/c5cs00501a. [DOI] [PubMed] [Google Scholar]
  6. Liu Z.; Yang Z.; Peng B.; Cao C.; Zhang C.; You H.; Xiong Q.; Li Z.; Fang J. Highly sensitive, uniform, and reproducible surface-enhanced Raman spectroscopy from hollow Au-Ag alloy nanourchins. Adv. Mater. 2014, 26, 2431–2439. 10.1002/adma.201305106. [DOI] [PubMed] [Google Scholar]
  7. Zhao Y.; Du L.; Li H.; Xie W.; Chen J. Is the Suzuki-Miyaura cross-coupling reaction in the presence of Pd nanoparticles heterogeneously or homogeneously catalyzed? An interfacial surface-enhanced Raman spectroscopy study. J. Phys. Chem. Lett. 2019, 10, 1286–1291. 10.1021/acs.jpclett.9b00351. [DOI] [PubMed] [Google Scholar]
  8. Li Y.; Hu Y.; Shi F.; Li H.; Xie W.; Chen J. C-H arylation on nickel nanoparticles monitored by in situ surface-enhanced Raman spectroscopy. Angew. Chem., Int. Ed. 2019, 58, 9049–9053. 10.1002/anie.201902825. [DOI] [PubMed] [Google Scholar]
  9. Kim N.; Thomas M. R.; Bergholt M. S.; Pence I. J.; Seong H.; Charchar P.; Todorova N.; Nagelkerke A.; Belessiotis-Richards A.; Payne D. J.; Gelmi A.; Yarovsky I.; Stevens M. M. Surface enhanced Raman scattering artificial nose for high dimensionality fingerprinting. Nat. Commun. 2020, 11, 207. 10.1038/s41467-019-13615-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Sheng S.; Wu J.-b.; Cong X.; Li W.; Gou J.; Zhong Q.; Cheng P.; Tan P.-h.; Chen L.; Wu K. Vibrational properties of a monolayer silicene sheet studied by tip-enhanced Raman spectroscopy. Phys. Rev. Lett. 2017, 119, 196803. 10.1103/physrevlett.119.196803. [DOI] [PubMed] [Google Scholar]
  11. Sheng S.; Wu J.-B.; Cong X.; Zhong Q.; Li W.; Hu W.; Gou J.; Cheng P.; Tan P.-H.; Chen L.; Wu K. Raman spectroscopy of two-dimensional borophene sheets. ACS Nano 2019, 13, 4133–4139. 10.1021/acsnano.8b08909. [DOI] [PubMed] [Google Scholar]
  12. Li J.; Yan H.; Tan X.; Lu Z.; Han H. Cauliflower-inspired 3D SERS substrate for multiple mycotoxins detection. Anal. Chem. 2019, 91, 3885–3892. 10.1021/acs.analchem.8b04622. [DOI] [PubMed] [Google Scholar]
  13. Tian L.; Su M.; Yu F.; Xu Y.; Li X.; Li L.; Liu H.; Tan W. Liquid-state quantitative SERS analyzer on self-ordered metal liquid-like plasmonic arrays. Nat. Commun. 2018, 9, 3642. 10.1038/s41467-018-05920-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Tian Y.; Wang H.; Yan L.; Zhang X.; Falak A.; Guo Y.; Chen P.; Dong F.; Sun L.; Chu W. A generalized methodology of designing 3D SERS probes with superior detection limit and uniformity by maximizing multiple coupling effects. Adv. Sci. 2019, 6, 1900177. 10.1002/advs.201900177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Bell S. E. J.; Sirimuthu N. M. S. Quantitative surface-enhanced Raman spectroscopy. Chem. Soc. Rev. 2008, 37, 1012–1024. 10.1039/b705965p. [DOI] [PubMed] [Google Scholar]
  16. Ding Q.; Wang J.; Chen X.; Liu H.; Li Q.; Wang Y.; Yang S. Quantitative and sensitive SERS platform with analyte enrichment and filtration function. Nano Lett. 2020, 20, 7304–7312. 10.1021/acs.nanolett.0c02683. [DOI] [PubMed] [Google Scholar]
  17. Bell S. E. J.; Charron G.; Cortés E.; Kneipp J.; Chapelle M. L.; Langer J.; Procházka M.; Tran V.; Schlücker S. Towards reliable and quantitative surface-enhanced Raman scattering (SERS): From key parameters to good analytical practice. Angew. Chem., Int. Ed. 2020, 59, 5454–5462. 10.1002/anie.201908154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Wei H.; Leng W.; Song J.; Willner M. R.; Marr L. C.; Zhou W.; Vikesland P. J. Improved quantitative SERS enabled by surface plasmon enhanced elastic light scattering. Anal. Chem. 2018, 90, 3227–3237. 10.1021/acs.analchem.7b04667. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Dos Santos D. P.; Temperini M. L. A.; Brolo A. G. Intensity fluctuations in single-molecule surface-enhanced Raman scattering. Acc. Chem. Res. 2019, 52, 456–464. 10.1021/acs.accounts.8b00563. [DOI] [PubMed] [Google Scholar]
  20. Shen W.; Lin X.; Jiang C.; Li C.; Lin H.; Huang J.; Wang S.; Liu G.; Yan X.; Zhong Q.; Ren B. Reliable quantitative SERS analysis facilitated by core-shell nanoparticles with embedded internal standards. Angew. Chem., Int. Ed. 2015, 54, 7308–7312. 10.1002/anie.201502171. [DOI] [PubMed] [Google Scholar]
  21. Shi Y.; Chen N.; Su Y.; Wang H.; He Y. Silicon nanohybrid-based SERS chips armed with an internal standard for broad-range, sensitive and reproducible simultaneous quantification of lead(ii) and mercury(ii) in real systems. Nanoscale 2018, 10, 4010–4018. 10.1039/c7nr07935d. [DOI] [PubMed] [Google Scholar]
  22. Zhu C.; Meng G.; Zheng P.; Huang Q.; Li Z.; Hu X.; Wang X.; Huang Z.; Li F.; Wu N. A hierarchically ordered array of silver-nanorod bundles for surface-enhanced Raman scattering detection of phenolic pollutants. Adv. Mater. 2016, 28, 4871–4876. 10.1002/adma.201506251. [DOI] [PubMed] [Google Scholar]
  23. Im H.; Bantz K. C.; Lee S. H.; Johnson T. W.; Haynes C. L.; Oh S.-H. Self-assembled plasmonic nanoring cavity arrays for SERS and LSPR biosensing. Adv. Mater. 2013, 25, 2678–2685. 10.1002/adma.201204283. [DOI] [PubMed] [Google Scholar]
  24. Peng B.; Li G.; Li D.; Dodson S.; Zhang Q.; Zhang J.; Lee Y. H.; Demir H. V.; Yi Ling X.; Xiong Q. Vertically aligned gold nanorod monolayer on arbitrary substrates: Self-assembly and femtomolar detection of food contaminants. ACS Nano 2013, 7, 5993–6000. 10.1021/nn401685p. [DOI] [PubMed] [Google Scholar]
  25. Ding Q.; Kang Y.; Li W.; Sun G.; Liu H.; Li M.; Ye Z.; Zhou M.; Zhou J.; Yang S. Bioinspired brochosomes as broadband and omnidirectional surface-enhanced Raman scattering substrates. J. Phys. Chem. Lett. 2019, 10, 6484–6491. 10.1021/acs.jpclett.9b02380. [DOI] [PubMed] [Google Scholar]
  26. Akinoglu G. E.; Hutchison J. A. Perspective—quasi-babinet complementary plasmonic templates: A platform to perform spectroelectrochemistry. ECS J. Solid State Sci. Technol. 2021, 10, 035005. 10.1149/2162-8777/abeb52. [DOI] [Google Scholar]
  27. Schmidt M. S.; Hübner J.; Boisen A. Large area fabrication of leaning silicon nanopillars for surface enhanced Raman spectroscopy. Adv. Mater. 2012, 24, OP11–OP18. 10.1002/adma.201103496. [DOI] [PubMed] [Google Scholar]
  28. Akinoglu E. M.; Luo L.; Dodge T.; Guo L.; Akinoglu G. E.; Wang X.; Shui L.; Zhou G.; Naughton M. J.; Kempa K.; Giersig M. Extraordinary optical transmission in nano-bridged plasmonic arrays mimicking a stable weakly-connected percolation threshold. Opt. Express 2020, 28, 31425–31435. 10.1364/oe.403034. [DOI] [PubMed] [Google Scholar]
  29. Akinoglu G. E.; Mir S. H.; Gatensby R.; Rydzek G.; Mokarian-Tabari P. Block copolymer derived vertically coupled plasmonic arrays for surface-enhanced Raman spectroscopy. ACS Appl. Mater. Interfaces 2020, 12, 23410–23416. 10.1021/acsami.0c03300. [DOI] [PubMed] [Google Scholar]
  30. Liu G.; Cai W.; Kong L.; Duan G.; Li Y.; Wang J.; Zuo G.; Cheng Z. Standing Ag nanoplate-built hollow microsphere arrays: Controllable structural parameters and strong SERS performances. J. Mater. Chem. 2012, 22, 3177–3184. 10.1039/c1jm14296h. [DOI] [Google Scholar]
  31. Huang J.-A.; Zhang Y.-L.; Ding H.; Sun H.-B. SERS-enabled lab-on-a-chip systems. Adv. Opt. Mater. 2015, 3, 618–633. 10.1002/adom.201400534. [DOI] [Google Scholar]
  32. Zhang D.; Xie Y.; Deb S. K.; Davison V. J.; Ben-Amotz D. Isotope edited internal standard method for quantitative surface-enhanced Raman spectroscopy. Anal. Chem. 2005, 77, 3563–3569. 10.1021/ac050338h. [DOI] [PubMed] [Google Scholar]
  33. Zheng F.; Ke W.; Shi L.; Liu H.; Zhao Y. Plasmonic Au-Ag janus nanoparticle engineered ratiometric surface-enhanced Raman scattering aptasensor for Ochratoxin A detection. Anal. Chem. 2019, 91, 11812–11820. 10.1021/acs.analchem.9b02469. [DOI] [PubMed] [Google Scholar]
  34. Hahm E.; Cha M. G.; Kang E. J.; Pham X.-H.; Lee S. H.; Kim H.-M.; Kim D.-E.; Lee Y.-S.; Jeong D.-H.; Jun B.-H. Multilayer Ag-Embedded Silica Nanostructure as a Surface-Enhanced Raman Scattering-Based Chemical Sensor with Dual-Function Internal Standards. ACS Appl. Mater. Interfaces 2018, 10, 40748–40755. 10.1021/acsami.8b12640. [DOI] [PubMed] [Google Scholar]
  35. Yu F.; Su M.; Tian L.; Wang H.; Liu H. Organic Solvent as Internal Standards for Quantitative and High-Throughput Liquid Interfacial SERS Analysis in Complex Media. Anal. Chem. 2018, 90, 5232–5238. 10.1021/acs.analchem.8b00008. [DOI] [PubMed] [Google Scholar]
  36. Jiang X.; Tan Z.; Lin L.; He J.; He C.; Thackray B. D.; Zhang Y.; Ye J. Surface-enhanced Raman nanoprobes with embedded standards for quantitative cholesterol detection. Small Methods 2018, 2, 1800182. 10.1002/smtd.201800182. [DOI] [Google Scholar]
  37. Tian H.; Zhang N.; Tong L.; Zhang J. In situ quantitative graphene-based surface-enhanced Raman spectroscopy. Small Methods 2017, 1, 1700126. 10.1002/smtd.201700126. [DOI] [Google Scholar]
  38. Tian H.; Li H.; Fang Y. Binary thiol-capped gold nanoparticle monolayer films for quantitative surface-enhanced Raman scattering analysis. ACS Appl. Mater. Interfaces 2019, 11, 16207–16213. 10.1021/acsami.9b02069. [DOI] [PubMed] [Google Scholar]
  39. Wang Y.; Zhang M.; Feng L.; Dong B.; Xu T.; Li D.; Jiang L.; Chi L. Tape-imprinted hierarchical lotus seedpod-like arrays for extraordinary surface-enhanced Raman spectroscopy. Small 2019, 15, 1804527. 10.1002/smll.201804527. [DOI] [PubMed] [Google Scholar]
  40. Zou Y.; Chen L.; Song Z.; Ding D.; Chen Y.; Xu Y.; Wang S.; Lai X.; Zhang Y.; Sun Y.; Chen Z.; Tan W. Stable and unique graphitic Raman internal standard nanocapsules for surface-enhanced Raman spectroscopy quantitative analysis. Nano Res. 2016, 9, 1418–1425. 10.1007/s12274-016-1037-6. [DOI] [Google Scholar]
  41. Chen H.-Y.; Lin M.-H.; Wang C.-Y.; Chang Y.-M.; Gwo S. Large-scale hot spot engineering for quantitative SERS at the single-molecule scale. J. Am. Chem. Soc. 2015, 137, 13698–13705. 10.1021/jacs.5b09111. [DOI] [PubMed] [Google Scholar]
  42. Yan X.; Li P.; Zhou B.; Tang X.; Li X.; Weng S.; Yang L.; Liu J. Optimal hotspots of dynamic surfaced-enhanced Raman spectroscopy for drugs quantitative detection. Anal. Chem. 2017, 89, 4875–4881. 10.1021/acs.analchem.6b04688. [DOI] [PubMed] [Google Scholar]
  43. Peksa V.; Jahn M.; Štolcová L.; Schulz V.; Proška J.; Procházka M.; Weber K.; Cialla-May D.; Popp J. Quantitative SERS analysis of azorubine (E 122) in sweet drinks. Anal. Chem. 2015, 87, 2840–2844. 10.1021/ac504254k. [DOI] [PubMed] [Google Scholar]
  44. Patze S.; Huebner U.; Weber K.; Cialla-May D.; Popp J. TopUp SERS substrates with integrated internal standard. Materials 2018, 11, 325. 10.3390/ma11020325. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Lin S.; Hasi W.; Lin X.; Han S.; Xiang T.; Liang S.; Wang L. Lab-on-capillary platform for on-site quantitative SERS analysis of surface contaminants based on Au@4-MBA@Ag core-shell nanorods. ACS Sens. 2020, 5, 1465–1473. 10.1021/acssensors.0c00398. [DOI] [PubMed] [Google Scholar]
  46. Mehrvar L.; Sadeghipari M.; Tavassoli S. H.; Mohajerzadeh S.; Fathipour M.; Hajihosseini H. r. Fabrication of Ag-modified nanocone frustum arrays with controlled shape as active substrates for surface-enhanced Raman scattering. J. Raman Spectrosc. 2019, 50, 1416–1428. 10.1002/jrs.5662. [DOI] [Google Scholar]
  47. Piscanec S.; Cantoro M.; Ferrari A. C.; Zapien J. A.; Lifshitz Y.; Lee S. T.; Hofmann S.; Robertson J. Raman spectroscopy of silicon nanowires. Phys. Rev. B: Condens. Matter Mater. Phys. 2003, 68, 241312. 10.1103/physrevb.68.241312. [DOI] [Google Scholar]
  48. Kumar K.; Duan H.; Hegde R. S.; Koh S. C. W.; Wei J. N.; Yang J. K. W. Printing colour at the optical diffraction limit. Nat. Nanotechnol. 2012, 7, 557–561. 10.1038/nnano.2012.128. [DOI] [PubMed] [Google Scholar]
  49. James T. D.; Mulvaney P.; Roberts A. The plasmonic pixel: large area, wide gamut color reproduction using aluminum nanostructures. Nano Lett. 2016, 16, 3817–3823. 10.1021/acs.nanolett.6b01250. [DOI] [PubMed] [Google Scholar]
  50. Mehrvar L.; Sadeghipari M.; Tavassoli S. H.; Mohajerzadeh S.; Fathipour M. Optical and Surface Enhanced Raman Scattering properties of Ag modified silicon double nanocone array. Sci. Rep. 2017, 7, 12106. 10.1038/s41598-017-12423-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Zhou L.; Yu X.; Zhu J. Metal-core/semiconductor-shell nanocones for broadband solar absorption enhancement. Nano Lett. 2014, 14, 1093–1098. 10.1021/nl500008y. [DOI] [PubMed] [Google Scholar]
  52. Wang T.; Hu X.; Dong S. A renewable SERS substrate prepared by cyclic depositing and stripping of silver shells on gold nanoparticle microtubes. Small 2008, 4, 781–786. 10.1002/smll.200700888. [DOI] [PubMed] [Google Scholar]
  53. Ye J.; Hutchison J. A.; Uji-i H.; Hofkens J.; Lagae L.; Maes G.; Borghs G.; Van Dorpe P. Excitation wavelength dependent surface enhanced Raman scattering of 4-aminothiophenol on gold nanorings. Nanoscale 2012, 4, 1606–1611. 10.1039/c2nr11805j. [DOI] [PubMed] [Google Scholar]
  54. Fonsaca J. E. S.; Hostert L.; Orth E. S.; Zarbin A. J. G. Tailoring multifunctional graphene-based thin films: from nanocatalysts to SERS substrates. J. Mater. Chem. A 2017, 5, 9591–9603. 10.1039/c7ta01967j. [DOI] [Google Scholar]
  55. Note: 4-ATP can self-assemble to a monolayer on Ag surface and be chemically transformed into 4,4′-dimercaptoazobenzene during SERS tests due to the plasmon-mediated chemical reactions [seeHuang Y.-F.; Zhu H.-P.; Liu G.-K.; Wu D.-Y.; Ren B.; Tian Z.-Q. When the signal is not from the original molecule to be detected: chemical transformation of para-aminothiophenol on Ag during the SERS measurement. J. Am. Chem. Soc. 2010, 132, 9244–9246. The Raman spectra of 4-ATP on Ag surface are actually 4,4′-dimercaptoazobenzene signals. However, the obtained Raman signals is reproducible at a given 4-ATP concentration and can reflect indirectly the information of 4-ATP 10.1021/ja101107z. [DOI] [PubMed] [Google Scholar]
  56. Kim Y.-K.; Ok G.; Choi S.-W.; Jang H.; Min D.-H. The interfacing structural effect of Ag/graphene oxide nanohybrid films on surface enhanced Raman scattering. Nanoscale 2017, 9, 5872–5878. 10.1039/c7nr00308k. [DOI] [PubMed] [Google Scholar]
  57. Wu K.; Li T.; Schmidt M. S.; Rindzevicius T.; Boisen A.; Ndoni S. Gold nanoparticles sliding on recyclable nanohoodoos—engineered for surface-enhanced Raman spectroscopy. Adv. Funct. Mater. 2017, 28, 1704818. 10.1002/adfm.201704818. [DOI] [Google Scholar]
  58. Jin B.; He J.; Li J.; Zhang Y. Lotus seedpod inspired SERS substrates: A novel platform consisting of 3D sub-10 nm annular hot spots for ultrasensitive SERS detection. Adv. Opt. Mater. 2018, 6, 1800056. 10.1002/adom.201800056. [DOI] [Google Scholar]
  59. Fan M.; Lai F.-J.; Chou H.-L.; Lu W.-T.; Hwang B.-J.; Brolo A. G. Surface-enhanced Raman scattering (SERS) from Au: Ag bimetallic nanoparticles: the effect of the molecular probe. Chem. Sci. 2013, 4, 509–515. 10.1039/c2sc21191b. [DOI] [Google Scholar]
  60. Li Y.; Lee E. J.; Cai W.; Kim K. Y.; Cho S. O. Unconventional method for morphology-controlled carbonaceous nanoarrays based on electron irradiation of a polystyrene colloidal monolayer. ACS Nano 2008, 2, 1108–1112. 10.1021/nn8001483. [DOI] [PubMed] [Google Scholar]
  61. Dai Z.; Li Y.; Duan G.; Jia L.; Cai W. Phase diagram,design of monolayer binary colloidal crystals,and their fabrication based on ethanol-assisted self-assembly at the air/water interface. ACS Nano 2012, 6, 6706–6716. 10.1021/nn3013178. [DOI] [PubMed] [Google Scholar]
  62. Fu R.; Liu G.; Jia C.; Li X.; Tang X.; Duan G.; Li Y.; Cai W. Fabrication of silver nanoplate hierarchical turreted ordered array and its application in trace analyses. Chem. Commun. 2015, 51, 6609–6612. 10.1039/c4cc10009c. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao1c02179_si_001.pdf (2.3MB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES