Chemicals and materials

TiO₂ nanoparticles with diameters of 21 nm were obtained from AEROSIL. CYTOP (CTL-809M, with a glass transition temperature of 375 K) was purchased from ASAHI glass. P3HT, ethanol, 1,2-dichlorobenzene, CPF (analytical standard), and phosphate-buffered saline (PBS; pH 7.3) were purchased from Sigma-Aldrich.

Fabrication of sensing layers and experimental setup

Figure 1(a) shows a schematic of the fabricated plasmonic photoelectrode. The photoelectrode was constructed by spin-coating CYTOP onto a BK-7 glass slide at 3000 rpm for 30 s, and then drying the CYTOP at 100 °C for 15 min to obtain a 400-nm-thick film. A DVD-R disk was used as a template for nanoimprinting.^[ Reference Hara, Lertvachirapaiboon, Ishikawa, Ohdaira, Shinbo, Kato, Kaneko and Baba ^{34 ,} Reference Nootchanat, Pangdam, Ishikawa, Wongravee, Shinbo, Kato, Kaneko, Ekgasit and Baba ^{35 ]} The template was cut into a size of 2.5 cm × 2.5 cm and then washed with nitric acid, ethanol, and deionized water. The grating structure was subsequently etched in a 1 : 4 acetone:ethanol solution for 15 min before rinsing with deionized water. The DVD-R mold with the grating structure was stamped onto the CYTOP film with a force of 25 MPa at 110 °C for 90 min. The grating-structured CYTOP/glass substrate was coated with approximately 3 nm of Cr and 60 nm of gold using vacuum evaporation. Next, TiO₂ nanoparticles were dispersed in ethanol at a concentration of 3 mg/mL, spin-coated on the gold-coated grating structure at 3000 rpm for 30 s, and dried at 100 °C for 1 h. The sample was then annealed at 250 °C for 30 min, forming a 100-nm-thick TiO₂ film. P3HT was dissolved in 1,2-dichlorobenzene at a concentration of 10 mg/mL and deposited on the TiO₂ layer by spin-coating at 2000 rpm for 60 s prior to annealing at 120 °C for 30 min, thus producing a 50-nm-thick P3HT film. Finally, ~3-mm-diameter AuNPs were loaded onto the P3HT film by vacuum evaporation. A UV–vis spectrometer (V-650, JASCO) was used to record the absorption spectra of each layer during the preparation process. The surface morphology of the gold-coated grating substrate, gold-coated grating/TiO₂/P3HT, and the gold-coated grating/TiO₂/P3HT loaded with AuNPs were studied via atomic-force microscopy (SPM-9600 from Shimadzu Corporation). Results indicated that the grating pattern remained intact after coating with TiO₂/P3HT and AuNPs (see Supplementary Data, Fig. S1).

Figure 1.

Schematic diagrams of (a) the fabricated photoelectrode with a grating structure, (b) the PEC cell for CPF measurements, and (c) the CPF sensing mechanism.

A schematic diagram of the PEC measurements is shown in Fig. 1(b). A platinum wire was used as the counter electrode. A 0.1 M PBS solution was used as the electrolyte. In the PEC measurements, the fabricated cell with the active area of 10.17 mm² was mounted on a goniometer. A solar simulator (HAL-C100 from ASAHI Spectra USA Inc.) with a light intensity of 75 mW/cm² was used as the light source to illuminate the cell on its counter-electrode/glass-slide side.

The SPR reflectivity measurements were performed under p-polarized light illumination for excitation of the surface plasmons; s-polarized light was used for comparison. The white light was collimated by objective lenses and applied at a fixed incident light angle with varying wavelengths. A spectrometer was used to measure the reflected light.

SPR properties

Figure 2 shows the UV–vis spectra of TiO₂/P3HT/AuNPs, TiO₂, P3HT, and AuNPs on glass slides to determine which peaks in the spectrum originated from which layer in the composite layered structure. A Q-band peak at 550 nm with a shoulder at 600 nm originating from the P3HT film and a broad shoulder at 720 nm originating from the AuNPs were observed in the multilayered film.

Figure 2.

UV–vis spectra of TiO₂/P3HT/AuNPs, TiO₂, P3HT, and AuNPs.

The excitation of SPR was studied by measuring the reflectivity curves as a function of wavelength from the fabricated AuNP-loaded photoelectrode cell in PBS. The ratio of the light reflected by the fabricated cell to those of a flat gold reference specimen was taken as the reflectivity. Figure 3(a) shows the SPR reflectivity curves of the fabricated cell with a grating structure measured at incident angles ranging from 30° to 60° with p-polarized light irradiation. The results show a resonance dip for each incident angle due to propagating surface plasmon excitation, indicating that enhanced electric fields could be obtained in the cell. The observed resonance dips shifted toward shorter wavelengths as the incident angle increased. The reflectivity curve of the fabricated cell with a grating structure under s-polarized light, as well as that of the cell without the grating structure under p-polarized light, were also measured and are shown in Figs. 3(b) and 3(c). In these conditions, the reflectivity dips did not occur due to lack of excitation of propagating SPR. Comparing the surface plasmon dips [Fig. 3(a)] with the absorption spectrum of the multilayered film (Fig. 2), we found that the wavelengths at which dips occur (around 40°−60°) corresponded to the main absorption peak wavelength of the PEC cell, which originated from the absorption of P3HT.

Figure 3.

Reflectivity curves of the fabricated PEC cell obtained by illumination with (a) p-polarized light and (b) s-polarized light on the grating substrate, and (c) p-polarized light on a cell with a flat gold substrate, each measured at incident light angles ranging from 30° to 60°.

Short-circuit photocurrent measurements

To measure the electrical signal from the PEC sensor in the presence of CPF, the short-circuit current at 0 V was measured. The underlying mechanism that gives rise to the short-circuit current is that upon injection of CPF, P3HT generates exited electrons due to visible light irradiation, which are injected into the conduction band of TiO₂, and subsequently to the gold electrode. Simultaneously, holes in the valence band of TiO₂ migrate to the valence band of P3HT. The holes can react with PBS to generate hydroxyl radicals (•OH). This dissociation reacts with CPF, forming CPF•, which can amplify the sensor's response.

We first studied the effect of propagating surface plasmon excitation on the photocurrent at zero potential as a function of the incident light angle between 0° and 60°. It is noted that no current was observed under light irradiation in the absence of CPF. Figure 4 shows the enhancement factor of the short-circuit current in the fabricated cells with 5 µM CPF in PBS, where the enhancement factor is defined as the ratio of the current measured with propagating surface plasmon excitation (p-polarized light irradiation) to the current without surface plasmon excitation (s-polarized light irradiation). The enhancement factor gradually increased as the incident light angle increased from 0° to 40° and reached a peak at 50°, which coincides with the surface plasmon dip wavelength observed in the fabricated cell and with the absorption of the P3HT film. This result is reasonable because the light absorption of the P3HT film is enhanced by the strong electric field generated by the excitation of a propagating surface plasmon at the optimal incident angle.

Figure 4.

Enhancement factor of the short-circuit current, defined as the ratio of the current with surface plasmon excitation (p-polarized light irradiation) to the current without surface plasmon excitation (s-polarized light irradiation) in the presence of 5 µM CPF.

Short-circuit photocurrent enhancement for CPF detection by hybrid surface plasmon excitation

The short-circuit photocurrent signal enhancement by both propagating surface plasmon and localized surface plasmon excitations were studied. Photoelectrodes were prepared in four configurations: flat gold substrates and gold-coated grating substrates, each prepared with or without loaded AuNPs. The short-circuit photocurrent signal was measured as a function of the CPF concentration (0.01–16 µM), as shown in Fig. 5. The linearity of the signal as a function of the CPF concentration was observed in each photoelectrode sample. The flat gold-coated substrate provided a lower I _SC signal than the gold-coated grating substrate. This is because the strong electric field intensity generated by the propagating surface plasmon on the grating gold surface increases the light absorption^[ Reference Baba, Wakatsuki, Shinbo, Kato and Kaneko ^{29 ,} Reference Baba, Aoki, Shinbo, Kato and Kaneko ^{30 ]} of P3HT, which subsequently increases the number of holes generated in P3HT, leading to more reactions with PBS to generate more •OH and CPF• radicals. When AuNPs were loaded on P3HT, the resulting signal was enhanced compared with that without AuNPs. This may be because the localized surface plasmon excited around the AuNPs enhances the electric field in the P3HT film.^[ Reference Pangdam, Nootchanat, Ishikaw, Shinbo, Kato, Kaneko, Thammacharoen, Ekgasit and Baba ^{31 ]}

Figure 5.

Plots of the short-circuit photocurrent of the fabricated PEC cells with various compositions as a function of the CPF concentration. The short-circuit current was measured at an optimal incident angle of 50°.

Due to the synergetic effect of propagating surface plasmon and localized surface plasmon, an extremely low limit of detection of 7.5 nM was achieved. Although the sensitivity was still one or two orders lower than that obtained by enzyme/antibody-based immunoassay systems with the electrochemical technique,^[ Reference Sun, Wang, Wen, Gan, Lei and Liu ^{36 ]} surface enhanced Raman,^[ Reference Kim, Barcelo and Li ^{37 ]} and magnetic platform^[ Reference Pino, Ivandini, Nakata, Fujisjima, Merkoci and Einaga ^{38 ]} or by magnetic-based molecularly imprinted technique with Kretschmann-based SPR,^[ Reference Yao, Liang, Huang, Wang and Qiu ^{14 ]} our technique exhibited high sensitivity without using complicated systems. Furthermore, the signal intensity was amplified to μA, producing nA-level signals, which are significantly more than that obtained using a previously reported method.^[ Reference Li, Li, Xu and Hu ^{28 ]} Based on our study, the advantages afforded by the PEC sensor using the gold-coated grating structure and loaded AuNPs were effectively demonstrated, including enhancement of the detection limit and amplification of the short-circuit current.