Fabrication and optical sensing properties of mesoporous silica nanorod arrays

Abstract

In this study, a mesoporous silica nanorod (MSNR) array on a gold (Au) film was fabricated via the Stöber-solution growth method using a porous anodic alumina (PAA) membrane as a template. The typical length and diameter of the MSNRs are approximately 675 nm and 65 nm, respectively, which are in accordance with the characteristics of the PAA template. In the Kretschmann configuration, the MSNR array with the Au film exhibits typical optical waveguide (OWG) characteristics, and an obvious waveguide coupling dip corresponding to the TM₁ mode under p-polarized light was observed in the reflectivity spectrum. A resolution of the refractive index (RI) as high as 3.6 × 10⁻⁸ RIU was achieved in the fabricated device, which is unprecedented for OWG systems due to the unique structure of the waveguiding layer. As an example to demonstrate the sensitivity of the OWG sensor, (1-hexade)cyltrimethylammonium bromide (CTAB), a small surfactant molecule, is detected in water with a limit of detection (LOD) of 623 nM. Two heavy metal ions (Ag⁺ and Pb²⁺) were also detected using the OWG sensor after modification using 3-MPTMS, a LOD of 200 nM was determined, which is the lowest LOD achieved so far for a label free sensing system.

---

Introduction

Nanoporous anodic alumina (PAA) membranes containing well ordered, densely packed, and perpendicular pore structures have been widely used for fabricating one-dimensional nanostructures.^1–5 PAA templates and some other nanostructures have been applied to the field of optical waveguide (OWG) sensors in the biomedical and environmental field.^6–12 However, the processes of fabricating these nanostructures are time consuming and expensive, and some of the nanostructures suffer poor stability in acid or alkali solutions, which result in a deterioration of their sensing performances. To enhance their pH stability, a protective coating layer is usually applied on the surfaces of the nanostructures, such as SiO₂, TiO₂ etc.^13,14 In addition, intrinsically stable nanostructures, such as mesoporous silica (MS), a nanoporous TiO₂ film, or a hydrogel layer, can also be applied to OWG sensor.^15–20 In our previous work, a silica nanotube (SNT) array was fabricated on an Au film using the surface sol–gel (SSG) method; the array was then used as a label-free OWG biosensor, which exhibited high sensitivity, with the potential to be used as a versatile sensing platform.²¹ However, the procedures to fabricate the SNT array are complicated and tedious due to the slow increase of the coating layer for one SSG cycle; this slow and complex fabrication process restricts the range of applications of the SNT array.

The principle of an OWG sensor is based on the detection of the changes of the refractive index (RI) between the interface of a solid and a liquid. Besides the label-free character shared by surface plasma resonance (SPR) approach, OWG sensor features higher sensitivity and lower limit of detection (LOD) because of the sharper attenuation dip in the reflectivity spectrum.²² In addition, the effective refractive index of the waveguiding layer is of significance for the performance of an OWG sensor; and to achieve a sharper attenuation dip in the reflectivity spectrum, the refractive index should be as close as possible to that of water,²³ which is commonly used as the dispersion media of analytes. Although many efforts have been made to address this issue, limited progress has been achieved.

Herein, we present a simple and effective method to fabricate a highly stable, well-aligned, relatively loosely arranged MSNR array on an Au film as a waveguiding layer for an OWG sensor produced using the Stöber-solution growth method; this growth method is a facile and effective method for the preparation of MS through a self-assembly process using silica precursors and surfactant in an aqueous ethanol solution,^24–26 together with the use of PAA membrane as a template. This approach decreases the occupation ratio of silica in the layer, which leads to the reduction of the effective refractive index of the waveguiding layer and the increase of the surface area for analyte adsorbance, both of which will improve the performance of the sensor. The typical length and the diameter of the MSNR formed are 675 ± 25 nm and 65 ± 5 nm, respectively. The as-fabricated sensor shows a RI resolution of 3.6 × 10⁻⁸ RIU and a low LOD for CTAB of 623 nM. Upon modification of the waveguiding layer using 3-MPTMS, the sensor responds to a minimum concentration of 200 nM of aqueous Ag⁺ and Pb²⁺ solutions at visible wavelength, which is unprecedented performance for the label-free OWG sensing system.

Experimental

Materials

The K9 glass substrates (22 × 22 × 0.5 mm³) were purchased from Dongguan City Hong Cheng Optical Co., Ltd (China). (1-hexade)cyltrimethylammonium bromide (CTAB), tetraethoxysilane (TEOS), (3-mercaptopropyl)trimethoxysilane (3-MPTMS) were purchased from Alfa Aesar (Tianjin, China). All the other organic solvents and chemical reagents were purchased from Shenzhen Tianxiang Huabo Co., Ltd (China) and Xiya reagents (Sichuan, China). Milli-Q water with a resistivity of 18.2 MΩ cm was used in all the experiments.

Modification of the Au substrate

The modification of the Au substrate with hydroxyl was performed for enhanced attachment of the PAA template membrane. Briefly, 2 nm of Cr and 40 nm of Au were successfully sputtered onto a cleaned K9 glass substrate. Next, the Au film was immediately immersed into 20 mM 3-MPTMS solution (dissolved in ethanol) for over 3 h.²⁷ Subsequently, the Au film was removed and then rinsed with ethanol and water alternately several times. Finally, the Au film was hydrolyzed by 0.1 M HCl solution for 1–10 h to obtain a hydrophilic surface for next step experiments.

Fabrication of the through-hole PAA film on the Au substrate

A classic two-step anodization method was used to fabricate the PAA film.¹ Before anodization, the Al foil (1.2 × 3 × 0.5 mm³) was electropolished at 20 V in a mixed solution of HClO₄ : CH₃CH₂OH = 1 : 4 at 15 °C for 7 min to make its surface smooth.²⁸ The Al foil was then anodized at 40 V in 0.3 M oxalic acid solution at 15 °C. After removing the alumina formed during the first anodization, the time of the second anodization is set to be 8 minutes. After anodization, the pores of the PAA template were widened in a H₃PO₃ (5 wt%) solution at 33.6 °C for 10 min. Next, the remaining Al was detached from the template by floating on the saturated HgCl₂ solution and the subsequent barrier layer was removed by floating on the H₃PO₃ (5 wt%) solution at 30 °C for 35 min, after which the through-hole PAA film was formed. Next, the PAA film was placed on the hydroxyl modified Au substrate in a mixed acetone and water solution, and kept in a desiccator. To ensure the through-hole PAA film firmly adhered to the Au substrate, the template was dried under vacuum at 120 °C for 20 min to dehydrate the interface between the PAA film and the Au substrate.

Fabrication of the MSNR array film on the Au substrate

The Stöber-solution was prepared according to the literature with some modification.²⁴ First, a mixed solution consisting of 0.08 g of CTAB, 15 mL of ethanol, 35 mL of water and 5 μL of concentrated ammonia (25 wt%) was stirred at 63 °C for 10 min. Next, 40 μL of TEOS was added to the mixed solution quickly, and then the solution was stirred for another 5 min to form the Stöber solution. The Au substrate with the through-hole PAA film was then immersed in the precursor solution and kept quiescent at 63 °C for 72 h. Subsequently, the template was removed, followed by rinsing with water, and then drying in an oven at 100 °C for at least 8 h. To remove the surfactant in the as-formed MSNR array, the solvent extraction method was used.²⁹ Briefly, the substrate with the PAA template and imbedded MSNR array was immersed into a 0.1 M HCl ethanol solution under moderate stirring for 10 min, followed by respective washing with ethanol and water, and then the substrate was dried by N₂. To ensure that the surfactant of CTAB was removed thoroughly, the extraction step was repeated three times. Finally, the entire substrate was immersed into H₃PO₃ (12 wt%) solution at room temperature for 24 h to dissolve the PAA template, resulting in the MSNR array with a MS cap on the Au substrate. The scheme of the fabrication procedure is shown in Fig. 1.

Fig. 1 Illustration (not to scale) of the formation of the MSNR array on the Au substrate.

Functionalization of the surface of the MSNR array with 3-MPTMS

For the detection of heavy metal ions, immobilization of 3-MPTMS to the surface of the MSNR was performed. Before modification, the MSNR array on the Au substrate was dehydrated at 50 °C in vacuum to remove the water molecules. Next, the substrate with the MSNR array was placed into 26 mL of ethanol, to which 1 mL of 3-MPTMS was added. The mixture containing the substrate was kept at room temperature for 24 h under argon atmosphere. Subsequently, the substrate was removed and rinsed with copious amount of ethanol and then dried by blowing with N₂ gas gently. Finally, the substrate was kept at 70 °C for 2 h in vacuum to be dealcoholized.

Optical waveguide spectroscopy (OWS)

The detection configuration is based on the Kretschmann prism coupling technique³⁰ shown in Fig. 2(a). Briefly, the multilayer film was attached to a K9 equilateral glass prism via index matching oil (n = 1.515). The light from a He–Ne laser (λ = 632.8 nm) first passed through a beam splitter to separate the light into two beams with one beam for the substrate and the other as the reference beam to increase the signal-to-noise-ratio. The polarization of the incident light can be adjusted by the polarizer and an analyzer was set in front of the detector to make the polarization of the reflected light match that of the incident light. The reflectivity of the OWG sensor was normalized using the intensity of reflected light for a bare K9 glass substrate. The sample solutions flowed through the MSNR array by a flow cell with a volume of 2 μL (16 × 1.3 × 0.1 mm³) at a rate of 50 μL per minute using a syringe pump. In the experiment, the angular reflectivity spectrum and the reflectivity kinetic curves were recorded.

Fig. 2 (a) Scheme of OWG sensor configuration based on the Kretschmann prism coupling technique. (b) and (c) are the top view and cross-sectional view of typical SEM images of the MSNR array film respectively. The inset in (b) is the SEM image of the top surface under high magnification.

Results and discussion

Characterization of MSNR array

Fig. 2(b) shows the top view SEM images of the MSNR array on the Au substrate. As can be seen from Fig. 2(b), the surface of the MS cap was relatively smooth and showed good continuity. Based on the fabrication methods, the as-formed silica is not solid but is mesoporous, as shown in the inset of Fig. 2(b), and the porosity of the silica was estimated to be 49.66% according to the MS formed by the Stöber-solution growth with CTAB as the surfactant. This mesoporous character is extremely useful for OWG sensing as it reduces the overall RI of the waveguiding layer and allows for more analyte adsorption. Fig. 2(c) shows the cross-sectional view of the MSNR array adhered to the Au substrate. As can be seen, the MSNR array was highly ordered with a MS layer on the top. Due to the capillary force during the drying procedure, some MSNRs may be in contact each other in the middle part. In contrast, the top and bottom parts of the MSNRs remain intact via the support of the cap and substrate respectively, which illustrates that the MS cap is essential to keep the silica rods from collapsing. By estimating from the cross-sectional SEM image, the typical length and diameter of the MSNRs are approximately 675 ± 25 nm and 65 ± 5 nm, respectively, and the MS cap is 115 ± 15 nm in thickness. The distance between two MSNRs is approximately 102 nm, which is in accordance with the parameters of the PAA template used. Thus, the remaining ca. 40 nm distances in adjacent MSNRs is sufficient for ions or most large proteins to adsorb onto the surface of MSNRs. Note that the parameters of the MSNR array film with the MS cap can be precisely adjusted by controlling the experiment condition.

Liquid exchange study

To evaluate the sensitivity of the OWG sensor with the MSNR array, we measured the angular reflectivity spectrum (reflectivity vs. incident angle) with water (n = 1.333) and ethanol (n = 1.36) flowing through the surface of the sensor (Fig. 3). Using Fresnel calculations, various transverse electric (TE) and transverse magnetic (TM) modes can be simulated based on the RI and the thickness of each layer of the OWG sensor.^6,8 In the study, by carefully tuning the diameter and length of the MSNRs and the thickness of the MS cap of the MSNR array, only one resonance dip (the range of incident angle is 60–70°) was observed in water condition under both s- and p-polarized light, as shown in Fig. 3(a) and (b), respectively. According to the Fresnel calculations, the resonance dips correspond to the TE₀ and TM₁ mode excited by s- and p-polarized light. When ethanol was substituted for water, we can found that both TE₀ and TM₁ modes shifted to the larger incident angles due to the larger RI of ethanol. In addition, we can also see that the Fresnel calculations fit well with the experiment results of water and ethanol based on the five phase structure derived from the four phase structure.²³ For the TE₀ mode, the reflectivity value calculated was lower than the experimental data because of the roughness of the surface of the MSNR array.

Fig. 3 Angular reflectivity spectra measured for water and ethanol using OWG sensor under (a) s-polarized and (b) p-polarized light.

In the Fresnel calculations, the five phase structure is adopted. The structure consists of a glass substrate, Au layer, the MSNR array, the MS cap and the sample solution, as can be seen in Fig. 2(a). The complex RI and thickness of Au layer used in the calculation were set to be 0.183 + i3.09 and 40 nm, respectively. The MSNR array with the surrounding medium and the MS cap serve as the waveguiding layer of the OWG sensor. In addition, the diameter and length of the MSNRs were set to be 66 nm and 700 nm, respectively, and the thickness of the MS cap was set to be 115 nm, which was in agreement with the results obtained from SEM images. Because the silica formed by surfactant-templated synthesis method contains mesopores,^24,31 the effective RI of the MS (n_ms) can be described according to Bruggeman approximation:³²

where f_silica is the volume fraction of silica in the MS (f_silica = 50.34%), n_silica and n_m are the RI of solid silica (n_silica = 1.457) and the surrounding medium (n_m = 1.333 for water or n_m = 1.36 for ethanol), respectively.

In addition, as part of the waveguiding layer, the effective RI of the MSNR array with the surrounding medium (n_m) can be described according to the Maxwell-Garnett theory:³³

For p-polarized light:

n² = n_ms²f_rod + n_m²(1 − f_rod) (2)

For s-polarized light:

where n_ms is the effective RI of the MS in eqn (1), and f_rod is the volume fraction of the MSNRs in this layer.

As an important comparative parameter to judge the sensitivity of sensing devices, the figure of merit (FOM) which takes the sharpness of the resonance dip into account can be defined as follows:

FOM = Δθ/(Δn × τ)

where Δθ is the shift of resonance dips, Δn is the RI variation of the different sample solutions, and τ is the full-width of the resonance dips at half-minimum (FWHM) in the reflectivity spectrum. The parameters of both modes calculated for FOM are summarized in Table 1.

Table 1 FOM and parameters used in the calculation of FOM

OWG mode	Δθ (deg)	Δn (RIU)	τ (deg)	FOM (RIU^−1)
TE0	2.17	0.027	0.1	804
TM1	2.06	0.027	0.2	381

---

From the results, we can see that the calculated FOM of TE₀ and TM₁ modes reached 804 reciprocal refractive index units (RIU⁻¹) and 381 RIU⁻¹ respectively, which are much higher than that of other OWG sensors using polycyanurate nanorod array or SPR based on regular angular modulation.^11,34 According to the FOM values, the sensitivity obtained using TE₀ mode with sharper coupling dip is higher than that of TM₁ mode. However, the minimum value of the reflectivity for TM₁ mode is much smaller than that of TE₀ mode, as can be seen from Fig. 3(a) and (b). This result indicates that the coupling efficiency and the dynamic detection range of TM₁ mode are higher than those of the TE₀ mode in the sensor.³⁴ Thus, the TM₁ mode was chosen in the following experiments. In addition, to make sufficient use of the sharp coupling dip of OWG sensor, reflection kinetic curves were adopted to detect ultrasmall RI change and low concentration of sample solution, which both lead to an inconspicuous shift of the waveguide coupling dips.

Ethanol water detection based on the RI sensitivity of the sensor

To test the accuracy of the OWG sensor, a refractometric study was performed, in which the RI changes can be measured. In the experiment, a series of ethanol solution in water with increased concentrations were prepared: 0.1 wt% (Δn = 2.7 × 10⁻⁵ RIU), 0.2 wt% (Δn = 5.4 × 10⁻⁵ RIU), 0.3 wt% (Δn = 8.1 × 10⁻⁵ RIU), 0.4 wt% (Δn = 10.8 × 10⁻⁵ RIU). Next, the ethanol solutions were introduced into the flow cell in sequence. To establish a stable baseline, water was first passed through the surface of sensor via the flow cell for a few minutes; subsequently a clear waveguide coupling dip corresponding to TM₁ mode can be observed in the angular reflection spectrum at θ_OWG = 61.86°, as shown in Fig. 4(a). To perform the refractometric study, the incident angle of the light was fixed at θ_OWG = 61.86°, and the reflectivity was recorded in real time for the different concentrations of ethanol.

Fig. 4 (a) Angular reflectivity spectrum measured with water using the OWG sensor under p-polarized light. (b) Reflectivity kinetic curve measured with a series of concentrations of ethanol in water. The blue part indicates that the solution flowing through MSNR array was water. The parts of (1), (2), (3), and (4) represent 0.1 wt%, 0.2 wt%, 0.3 wt%, and 0.4 wt% of aqueous ethanol solution, respectively. The inset is the magnification of the reflectivity signal during the first twelve minutes. (c) Reflectivity change (ΔR) as a function of RI change calculated from (b). The black line is the best linear fit with the experimental data.

After the baseline from water was obtained (Fig. 4(b)), 0.1 wt% ethanol was substituted for water and an abrupt increase of ca. 0.08 in the reflectivity was observed and then remained stable, indicating the complete exchange of the two solutions. The increase of reflectivity is due to the larger refractive index of 0.1 wt% ethanol than pure water. As the concentration of ethanol increases, the reflectivity increases. In contrast, when the lower concentration of ethanol was introduced into the flow cell, the reflectivity decreased abruptly and then remained stable. In Fig. 4(c), the reflectivity change as the function of ethanol concentration was plotted. From the results, the slope of the reflectivity change (ΔR) with the concentration of ethanol is k = d(ΔR)/d(Δn) = 3507.4 RIU⁻¹. In the experiment, the variations in the reflectivity signal were acquired using the data obtained from the baseline and the typical standard deviation is σ = 1.2608 × 10⁻⁴ (see the inset in Fig. 4(b)), which gives the RI resolution of 3.6 × 10⁻⁸ RIU (defined as σ/k).¹⁹ This value was much higher than the RI resolution of the hydrogel waveguide sensor and SPR sensor based on regular angular modulation.¹⁹ The high RI resolution of the present OWG sensor can be ascribed to the enhanced field strength of the TM₁ mode and the small FWHM of the reflectivity spectrum, which are related to the unique structure of the waveguiding layer.

Detection of CTAB molecules adsorption

CTAB is a small cationic surfactant molecule that can be easily adsorbed to a negatively-charged surface via electrostatic attraction.³⁵ Because the isoelectric point (I.E.P) of silica is 4.2,³⁶ which is lower than the pH of the CTAB solution (dissolved in water), the surface of silica is negatively-charged and can absorb CTAB molecules. To investigate the ability of the OWG sensor for detecting CTAB molecules, a series concentration of CTAB solutions (0.1, 1, 2, 5 and 10 μM) were introduced into the flow cell. In this experiment, the reflectivity kinetic curve was also recorded as before, as shown in Fig. 5(a). After the baseline was established for 10 min using water, each of the concentrations of CTAB was passed through the surface of OWG sensor for 10 min followed by rinsing with water for 10 min.

Fig. 5 (a) Reflectivity kinetic curve measured using the OWG sensor with different concentrations of CTAB aqueous solutions (0.1, 1, 2, 5 and 10 μM). In between the injection of CTAB solution, the sensor surface was rinsed with water (blue parts). (b) Reflectivity change (ΔR) as a function against the logarithm of the concentration of CTAB solutions (c_CTAB). The black solid line is the linear fit.

As can be seen from Fig. 5(a), when 0.1 μM aqueous solution was introduced into the flow cell after the baseline was established followed by rising with water, no significant increase in the reflectivity signal was found in 10 min. However, when the concentration of aqueous CTAB solution was changed to 1 μM, a rapid increase in the reflectivity was observed. Subsequently, the continuous increase of the concentration of CTAB caused an obvious increase in the reflectivity. This response is due to the adsorption of the CTAB molecules onto the surface of the MSNRs, which increases the effective RI of the waveguiding layer and results in the shift of TM₁ mode to larger incident angles. During the water rinsing that occurred between two concentrations of aqueous CTAB solutions, however, the reflectivity signal decreased due to the process of desorption, and the decreasing rate gradually increase with increasing CTAB concentration. Fig. 5(b) shows the plot of the ΔR against the logarithm of the concentration of CTAB solutions (log(c_CTAB)). The black solid line is the linear fit, from which we can see that ΔR and log(c_CTAB) exhibit a good linear relationship. According to the 3σ edit rule (σ = 1.2608 × 10⁻⁴ RIU), the linear fit reaches 3σ gives the LOD of the aqueous CTAB solution; the LOD was estimated to be 10^−0.2056 = 0.623 μM. This value is lower than the result of 0.02 mM with the SPR sensor based on hierarchical MSF,³⁵ indicating the much higher sensitivity of our OWG sensor.

Detection of heavy metal ion adsorption

Due to the large surface area and the possible chemical modifications of silica, the MSNR array can also be used to adsorb bio-molecules and heavy metal ions. In the following experiment, a new MSNR array waveguide modified with 3-MPTMS was used to investigate the response of the OWG sensor for the adsorption of heavy metal ions (Ag⁺ and Pb²⁺); the illustration of optical setup is shown in Fig. 6(a). A typical reflectivity spectrum of the 3-MPTMS modified OWG sensor is shown in the inset in Fig. 6(b).

Fig. 6 (a) The illustration of OWG sensor with the MSNR array modified with 3-MPTMS as the waveguiding layer. (b) Reflectivity kinetic curve monitored for different concentrations of Ag⁺ ions using OWG sensor with (black line) and without (red line) 3-MPTMS modification. The inset is the typical angular reflectivity spectrum measured with water using the OWG sensor with 3-MPTMS modification under p-polarized light. (c) Reflectivity kinetic curve monitored for different concentrations of Pb²⁺ ions with (black line) and without (red line) 3-MPTMS modification. (d) Reflectivity change (ΔR) as a function of the logarithm of the concentration of the heavy metal ions (Ag⁺ and Pb²⁺) calculated from (b) and (c). The black line is the leaner fit.

The procedures to detect heavy metal ions are similar to that of CTAB. A series of aqueous AgNO₃ solutions (0.1, 0.2 and 0.5 μM) and Pb(NO₃)₂ solutions (0.1, 0.2, 0.5 and 1 μM) were prepared as the sample solutions. The incident angle of light was fixed at the angle of resonant dip (inset in Fig. 6(b)) and the reflectivity of the OWG sensor was recorded in real-time (see Fig. 6(b) and (c)). After the baseline was first established with water for approximately 10 min, different concentrations of AgNO₃ or Pb(NO₃)₂ solution were introduced according to the order from lower to higher concentration. After each injection, the sensor surface was rinsed with water. From the raw reflectivity, the kinetic curve monitored for Ag⁺ ions (black line in Fig. 6(b)), a slight increase of reflectivity was observed when 0.1 μM AgNO₃ solutions were introduced into the flow cell. However, when using 0.2 μM AgNO₃ solutions instead, an obvious increase of reflectivity can be seen. With increasing concentration of AgNO₃ solutions, a similar trend in reflectivity was observed, but with a higher rate of increase of reflectivity (Fig. 6(b)). This observation confirmed the gradual attachment of Ag⁺ ions to the thiol terminals on the surface of OWG sensor. After rinsing with water, the reflectivity slightly decreased due to the removal of nonspecific bound Ag⁺ ions. The obvious increase in the present case indicates the LOD for Ag⁺ ions of the present OWG sensor is at least 0.2 μM. To demonstrate that the Ag⁺ ions adsorbed to the surface of MSNR array were selectively attached to the thiol terminals, a reference experiment was performed using a MSNR array waveguide without the modification of 3-MPTMS. As can be seen from the red line in Fig. 6(b), no significant increase of reflectivity was observed, even when the concentration of the injected solution of AgNO₃ is as high as 10 μM, 100 μM and 1 mM, indicating the adsorption of Ag⁺ ions through electrostatic attraction can be neglected.

Detection of Pb²⁺ ions was also performed using the same procedures as that of Ag⁺ ions, as shown in Fig. 6(c) (black line). From the reflectivity kinetic curve, an obvious increase of reflectivity was also observed for 0.2 μM Pb(NO₃)₂, indicating the minimum concentration to produce a positive response for the present OWG sensor. This value is much lower than that of the sensor for adsorbing Pb²⁺ only via electrostatic attraction at 1 μM.^18,35 Similar to that of Ag⁺, the reference experiment results also reveal that Pb²⁺ ions are indeed selectively attached to the thiol terminals of the surface of MSNR array (see red line in Fig. 6(c)). In addition, compared to the two ions in the reflectivity kinetic curves and the quasilinear dependence of ΔR on the logarithm of the heavy metal ions (Ag⁺ and Pb²⁺) concentration (Fig. 6(d)), the 3-MPTMS modified MSNR array waveguide shows more sensitive response of absorbing Ag⁺ ions than Pb²⁺ ions, which is in agreement with the results reported before.³⁷

Conclusions

In summary, we report a facile method to prepare a MSNR array on a Au film for the fabrication of a highly sensitive OWG sensor; the sensor exhibited high values of 804 and 381 RIU⁻¹ as the sensing FOM through liquid exchange experiments under TE₀ and TM₁ modes, respectively. Benefitting from the mesoporous structure of the arrays and the relatively low effective refractive index of the waveguiding layer, a high RI resolution of 3.6 × 10⁻⁸ RIU was obtained due to enhanced field strength of the TM₁ mode and the small FWHM of the reflectivity spectrum. The experiments also demonstrate that LOD for both surfactant small molecules and heavy metal ions reached 623 nM and 200 nM, respectively. To the best of our knowledge, these results are better than the results of any OWG and SPR systems that are based on identical configuration for label-free detection. The presented MSNR array in the study thus provides a novel and practical strategy for high-sensitivity label-free sensing and may find broad applications, including real-time biosensing.

Acknowledgements

This work was supported by the National Nature Science Foundation of China (grants no. 21273126, 21573124) and Fundamental Research Program of Shenzhen (JCYJ20140509172959966). Part of this work was supported by the Open Research Fund Program of the State Key Laboratory of Low-Dimensional Quantum Physics (KF201311).

References

1. H. Masuda and K. Fukuda, Science, 1995, 268, 1466.
2. Q. Y. Lu, F. Gao, S. Komarneni and T. E. Mallouk, J. Am. Chem. Soc., 2004, 126, 8650.
3. J. M. Moon and A. Wei, J. Phys. Chem. B, 2005, 109, 23336.
4. B. Luo, D. Yang, M. Liang and L. Zhi, Nanoscale, 2010, 2, 1661.
5. C. Bae, H. Kim, Y. Yang, H. Yoo, J. M. M. Moreno, J. Bachmann, K. Nielsch and H. Shin, Nanoscale, 2013, 5, 2825.
6. K. H. A. Lau, L. S. Tan, K. Tamada, M. S. Sander and W. Knoll, J. Phys. Chem. B, 2004, 108, 10812.
7. K. Hotta, A. Yamaguchi and N. Teramae, ACS Nano, 2012, 6, 1541.
8. Y. Fan, K. Hotta, A. Yamaguchi and N. Teramae, Opt. Express, 2012, 20, 12850.
9. A. Yamaguchi, K. Hotta and N. Teramae, Anal. Chem., 2008, 81, 105.
10. Y. Fan, K. Hotta, A. Yamaguchi, Y. Ding, Y. H. He, N. Teramae, S. Q. Sun and H. Ma, Appl. Phys. Lett., 2014, 105, 031103.
11. A. Gitsas, B. Yameen, T. D. Lazzara, M. Steinhart, H. Duran and W. Knoll, Nano Lett., 2010, 10, 2173.
12. A. V. Kabashin, P. Evans, S. Pastkovsky, W. Hendren, G. A. Wurtz, R. Atkinson, R. Pollard, V. A. Podolskiy and A. V. Zayats, Nat. Mater., 2009, 8, 867.
13. Y. Fan, Y. Ding, Y. F. Zhang, H. Ma, Y. H. He and S. Q. Sun, RSC Adv., 2014, 4, 62987.
14. L. K. Tan, H. Gao, Y. Zong and W. Knoll, J. Phys. Chem. C, 2008, 112, 17576.
15. Q. S. Huo, D. Y. Zhao, J. L. Feng, K. Weston, S. K. Buratto, G. D. Stucky, S. Schacht and F. Schuth, Adv. Mater., 1997, 9, 974.
16. A. Yamaguchi, H. Arafune, K. Hotta, T. Itoh and N. Teramae, Anal. Sci., 2011, 27, 597.
17. P. J. Cameron, A. T. A. Jenkins, W. Knoll, F. Marken, E. V. Milsom and T. L. Williams, J. Mater. Chem., 2008, 18, 4304.
18. Z. Zhang, D. F. Lu and Z. M. Qi, J. Phys. Chem. C, 2012, 116, 3342.
19. Y. Wang, C. J. Huang, U. Jonas, T. Wei, J. Dostalek and W. Knoll, Biosens. Bioelectron., 2010, 25, 1663.
20. Q. Zhang, Y. Wang, A. Mateescu, K. Sergelen, A. Kibrom, U. Jonas and J. Dostalek, Talanta, 2013, 104, 149.
21. Y. Fan, Y. Ding, H. Ma, N. Teramae, S. Q. Sun and Y. H. He, Biosens. Bioelectron., 2015, 67, 230.
22. F. C. Chien and S. J. Chen, Biosens. Bioelectron., 2004, 20, 633.
23. N. Skivesen, R. Horvath and H. C. Pedersen, Sens. Actuators, B, 2005, 106, 668.
24. Z. G. Teng, G. F. Zheng, Y. Q. Dou, W. Li, C. Y. Mou, X. H. Zhang, A. M. Asiri and D. Y. Zhao, Angew. Chem., Int. Ed., 2012, 51, 2173.
25. M. Grun, I. Lauer and K. K. Unger, Adv. Mater., 1997, 9, 254.
26. B. Guan, Y. Cui, Z. Ren, Z. A. Qiao, L. Wang, Y. Liu and Q. Huo, Nanoscale, 2012, 4, 6588.
27. W. R. Thompson, M. Cai, M. Ho and J. E. Pemberton, Langmuir, 1997, 13, 2291.
28. Y. Peng, H. L. Zhang, S. L. Pan and H. L. Li, J. Appl. Phys., 2000, 87, 7405.
29. A. Walcarius, E. Sibottier, M. Etienne and J. Ghanbaja, Nat. Mater., 2007, 6, 602.
30. H. P. Ho, S. Y. Wu, M. Yang and A. C. Cheung, Sens. Actuators, B, 2001, 80, 89.
31. A. Yamaguchi, F. Uejo, T. Yoda, T. Uchida, Y. Tanamura, T. Yamashita and N. Teramae, Nat. Mater., 2004, 3, 337.
32. Z. M. Qi, I. Honma and H. Zhou, Anal. Chem., 2006, 78, 1034.
33. D. E. Aspnes, Thin Solid Films, 1982, 89, 249.
34. T. D. Lazzara, K. H. A. Lau and W. Knoll, J. Nanosci. Nanotechnol., 2010, 10, 4293.
35. J. Li, D. F. Lu, Z. Zhang, Q. Liu and Z. M. Qi, Sens. Actuators, B, 2014, 203, 690.
36. S. Fukuzaki, H. Urano and K. Nagata, J. Ferment. Bioeng., 1996, 81, 163.
37. T. Kumeria, M. M. Rahman, A. Santos, J. Ferré-Borrull, L. F. Marsal and D. Losic, ACS Appl. Mater. Interfaces, 2014, 6, 12971.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra18629c

---