Organic-Inorganic Semiconductor Heterojunction P3HT@Ag₂NCN Composite Film as a Recyclable SERS Substrate for Molecule Detection Application

Abstract

Semiconductor composite materials have attracted interest from surface-enhanced Raman scattering (SERS) substrate research. Here, we investigate an organic-inorganic semiconductor heterojunction P3HT@Ag₂NCN composite film as a recyclable SERS substrate for molecule detection application. Our study shows that the SERS substrate of the composite P3HT@Ag2NCN composite film has high sensitivity, excellent signal reproducibility, and is reusable. Significant π-stacking of the probe molecules with the thiophene π-cores molecules from P3HT plays an important role in the large SERS enhancement by the charge transfer mechanism. Due to physical interaction between P3HT and Ag₂NCN, the organic-inorganic semiconductor heterojunction structure further improves charge transfer efficiency and the SERS property. Our results show that the enhancement factor (EF) of P3HT@Ag₂NCN composite films (EF = 6147 ± 300) for the probe molecule methylene blue is more than 7 times that of P3HT substrate (EF = 848 ± 85) and is about 75 times that of Ag₂NCN nanorods (EF = 82 ± 8). In addition, the SERS substrates of the P3HT@Ag₂NCN composite film also display excellent reusability and signal reproducibility (RSD < 4.8%). Our study opens up a new opportunity for designing an ideal SERS substrate with high sensitivity, selectivity, long-term stability, low cost, and reusability.

1. Introduction

SERS technology has a wide range of applications as a fast [1], lossless [2], and high sensitivity [3] trace detection tool. Due to their localized surface plasmon resonance (LSPR), noble metal nanoparticles display excellent SERS properties [4,5,6,7,8]. Previous theoretical and experimental research has shown the importance of the shape, distance, and size of noble metal nanoparticles for their LSPR effects [9,10,11,12,13]. Previous studies [14,15,16] have shown various methods for the synthesis of metal particles with different shapes of nanostructures used as SERS substrates, including nanorods, nanotriangles, nanocubes, and nanocages. Nanostructures with higher curvature, such as nanoflowers and nanostars, can help to increase the sharpness and roughness of the substrate material. These dense, sharp parts and suitable nano-gaps can increase the number of “hotspots”, enhance the surrounding electric field, and even achieve single-molecule detection. Although they have sensitive and efficient response, their instability in air and cost limit their practical application.

The advantages, e.g., high selectivity, low cost, and high stability, of inorganic semiconductor substrates have attracted interest in the field of SERS substrate research [17,18,19,20,21,22,23,24]. Yang et al. [23] prepared TiO₂ nanoparticles with different crystallinities by a sol-hydrothermal method and performed SERS experiments using ciprofloxacin as a probe molecule. However, there is still a big gap in Raman enhanced performance between inorganic semiconductors and noble metal nanoparticles. In recent years, organic semiconductor SERS substrates, due to their easy preparation, structural versatility, high selectivity, and long-term stability, have opened up new fields for the development of SERS substrates [25,26,27,28,29,30]. Yilmaz et al. [25] confirmed that nanoscale organic semiconductors, i.e., diperfluorohexyl tetrathiophene molecules with hydrophobic properties, can be used as SERS substrates, and the enhancement factor for the probe molecule methylene blue can reach 3.4 × 10³. Our previous study showed that poly(3-hexylthiophene) (P3HT) films with an optimized crystal structure can improve the charge transfer of the substrate molecule and the probe molecule via more effective chemisorption of the latter with the former [26]. The SERS substrate of P3HT film has excellent signal reproducibility (RSD < 10%) and can retain good Raman enhancement after one year. However, the Raman enhanced performance of organic semiconductor materials still cannot achieve that of noble metal nanoparticles. The function of a single material cannot fit all the criteria of an ideal SERS substrate, including low cost, high sensitivity, reproducibility, and long-term stability. Some works have stated that composite materials are a good choice [31,32,33,34,35,36,37,38,39,40,41]. The development of noble metal nanoparticles and semiconductor composites can expand the theory and application of SERS technology. Zhao et al. [38] prepared TiO₂ nanofiber membranes with a diameter of about 200 nm using electrospinning technology and then surface-modified the fibers with silver nanoparticles by a chemical plating method. The composite SERS substrate showed good SERS activity for the three probe molecules, i.e., 4-mercaptobenzoic acid, rhodamine 6G, and 4-aminothiophenol. Due to the high photocatalytic performance of TiO₂, the SERS substrate can be recycled by soaking it in water and irradiating it with ultraviolet light. At present, noble metal nanoparticles have been composited with semiconductor materials and have become a hot topic in SERS research because of their simple recovery process and excellent performance. In noble metal nanoparticles/semiconductor composite materials, noble metals exhibit a strong LSPR effect in the visible light region, which can increase light absorption. At the same time, the Fermi energy level of noble metals is lower than that of semiconductors, which can promote the separation of photogenerated electrons from holes and improve the charge transfer (CT) efficiency. Samir Kumar et al. [35] used grazing angle deposition (GLAD) technology to fabricate a novel Ag nanoparticle-modified TiO₂ nanorod array substrate that can be used for photocatalysis as well as surface-enhanced Raman scattering. Stavytska-Barba et al. [36] deposited poly(3-hexylthiophene)/[6,6]-phenyl-C61-butyrate methyl ester (P3HT/PCBM) on triangular silver nanoprisms, achieving excellent SERS signals. Wei et al. [39] showed that a MoS₂@TiO₂@Au platform can rapidly degrade crystal violet under solar irradiation and enable reproducible and sensitive SERS analyses. Wang et al. [40] reported that a AgNPs/TiO₂ NSAs/CP hybrid can renew and be reused five times by UV irradiation due to a photocatalysis induced self-cleaning feature. Ag-NP-decorated ZnO nanoflower arrays were demonstrated to possess a self-cleaning function, enabled by UV irradiation via the photocatalytic degradation of the analyte molecules [41]. However, our understanding of the SERS mechanism of composite materials remains limited. A deeper understanding is required to address many open questions regarding the SERS properties of composite materials.

Here, we investigate the SERS properties of organic-inorganic semiconductor heterojunction P3HT@Ag₂NCN composite films. [NCN]²⁻ in Ag₂NCN is similar to O²⁻ in metal oxides, but compared with Ag₂O, the 2p orbital energy level of O is lower than those of C and N, and the 2p orbitals of [NCN]²⁻ constitute a higher valence band energy level. Additionally, the band gap will be narrower, so Ag₂NCN should theoretically have excellent charge transfer (CT) efficiency and photocatalytic properties. Our study shows that a P3HT@Ag₂NCN composite film is an ideal SERS substrate, successfully overcoming the shortcomings of the poor stability, low sensitivity, high cost, and non-reusability of traditional SERS substrates. Our results show that there is an obvious physical interaction between Ag₂NCN and P3HT. Our previous paper [26] showed that significant π-stacking of the probe molecules with the thiophene π-cores molecules from P3HT plays an important role in SERS enhancement via the charge transfer mechanism. An organic-inorganic semiconductor heterojunction structure further improves charge transfer (CT) efficiency and the SERS property. The enhancement factor of P3HT@Ag₂NCN composite films (6147 ± 300) for probe molecule methylene blue was more than seven times that of the P3HT substrate (848 ± 85). In addition, the SERS substrates of the P3HT@Ag₂NCN composite film also displayed excellent reusability. The study contributes to the rational design of an ideal SERS substrate with high sensitivity, selectivity, long-term stability, low cost, and reusability.

2. Experimental Section

2.1. Materials

Regioregular poly(3-hexylthiophene-2,5-diyl) (P3HT, M_w = 54~75 kg/mol, Mw/Mn = 2.2) was supplied by Sigma-Aldrich (St. Louis, MO, USA). Reagents silver nitrate, chloroform, cyanamide, n-hexane, toluene, acetonitrile, sulfuric acid (98%), hydrogen peroxide (30%), and ammonia were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Methylene blue (MB, 98.5%) was purchased from Amresco (Radnor, PA, USA).

2.2. Preparation of P3HT@Ag₂NCN Composite Film

The preparation process of the P3HT@Ag₂NCN composite film is shown in Figure 1. Firstly, 5.1 g of silver nitrate was dissolved in 300 mL of ultrapure water, and then 600 mL ammonia water (3 mol/L) was slowly added. Next, 300 mL of an aqueous cyanamide solution (0.9 wt%) was slowly added. After stirring for 0.5 h, filtration was carried out. The yellow powder-silver cyanamide was washed several times with ultrapure water and dried at 60 °C in a vacuum oven. To further optimize the crystal structure of the silver cyanamide, the compound was placed in a hydrothermal kettle for 4 h at 200 °C and then filtered, washed, and dried.

A chloroform solution of P3HT/Ag₂NCN with a concentration of 8 mg/mL was prepared. This solution was spin-coated on a glass wafer to fabricate the P3HT@Ag₂NCN composite film. The spin speed was 4000 rpm, and the spin time was 60 s. The residual solvent was removed by putting the films in a vacuum oven for 24 h at room temperature. The P3HT@Ag₂NCN composite films were immersed in different mixed solvents to regulate their crystal structures, according to a method described in a previous paper [42]. After the P3HT@Ag₂NCN composite film had been prepared, 5 μL of dye solution (different concentrations) was placed on the P3HT@Ag₂NCN composite film and left to evaporate and dry.

2.3. Raman Spectroscopy

Raman measurements were performed at room temperature with a commercial confocal Raman microscope system (LABRAM HR 800 UV, Horiba Jobin Yvon, Bensheim, Germany) equipped with a Leica microscope (50× objectives). A laser (532 nm) was used as the light source for excitation. A D2 filter was used; the exposure time was 5 s and the accumulation number was 1. The laser power used in the measurement was 50 micro-watts.

2.4. Characterization Techniques

The morphologies of the P3HT@Ag₂NCN composite films were observed by field emission scanning electron microscope (SEM, JSM-7500F, Japanese, JEOL, Tokyo, Japan). An Escalab 250Xi instrument was used for X-ray photoelectron spectroscopy (XPS, Thermo Scientific, Waltham, MA, USA). The binding energy scale referenced to the C1s level of the carbon overlayer at 284.8 eV. The UV-Visible absorption spectra were recorded using an ultraviolet-visible spectrophotometer (TU-1901, Beijing PERSEE, Beijing, China) in the wavelength range of 230–850 nm. The infrared spectrum was recorded using an infrared spectrometer (Nicolet iS50,Thermo Scientific, Waltham, MA, USA) in the wavenumber range of 600–4000 nm.

2.5. SERS Enhancement Factor (EF)

The SERS EF was calculated by the formula [43]:

$$\mathrm{EF}=\frac{{\mathrm{N}}_{Reference}\times {\mathrm{I}}_{compositefilm}}{{\mathrm{N}}_{compositefilm}\times {\mathrm{I}}_{Reference}}$$

(1)

where I_{compositefilm} and I_Reference are the SERS intensity and the unenhanced normal signal intensity, respectively, N_{compositefilm} is the number of probed molecules in SERS, and N_Reference is the number of probed molecules in normal Raman measurements.

3. Result and Discussion

Ag₂NCN nanorods were prepared by the chemical precipitation method, and their crystal structure was further optimized by the hydrothermal method. Figure 2 shows SEM images of Ag₂NCN nanorods at different magnifications before and after optimization. As shown in Figure 2a–c, the untreated Ag₂NCN nanorods presented a round rod shape. In contrast, as shown in Figure 2d–f, the nanorods after hydrothermal treatment showed a square column shape; we can see clear edges and corners in Figure 2d–f. This result shows that Ag₂NCN nanorods recrystallized in the hydrothermal process at high temperature, which improved their crystallization ability and optimized their crystal structure. To explore the change in the crystal structure of Ag₂NCN nanorods before and after treatment, we characterized them using XRD. Figure 3 shows the XRD spectra of silver cyanamide before and after treatment. The crystal plane diffraction peak position of the Ag₂NCN nanorods prepared in the experiment was consistent with the standard card (JCPDS No.70–5232 [44]). Among them, the characteristic diffraction peaks of the crystal plane were 2θ = 12.20°, 19.14°, 24.72°, 28.94°, 32.21°, 32.74°, 37.55°, 38.94°, 40.66°, respectively, corresponding to (100), (110), (200), (210), (102), (021), (300), (220), (022) crystal planes of monoclinic silver cyanamide. At (100) crystal plane (12.20°), the full widths at half maximum of the peak before and after treatment were 0.25° and 0.16°, respectively. At (110) crystal plane (19.14°), the full widths at half maximum of the peak before and after treatment were 0.24° and 0.18°, respectively. As shown in Figure 3, all of the XRD diffraction peaks of the Ag₂NCN nanorods were sharper, and the full widths at half maximum of the peak became smaller, after hydrothermal treatment, indicating that the crystallinity of Ag₂NCN nanorods had been improved by the hydrothermal treatment.

UV-vis DRS was measured to investigate the optical properties of the Ag₂NCN nanorods. Figure 4 shows the UV diffuse reflection curve and K-M equation curve of the nanorods before and after treatment. Figure 4a shows that the nanorods had an absorption range from 250–530 nm. This relatively wide light absorption range indicated that more charge carriers can be generated under the excitation of light, reducing the probability of electron-hole recombination and thus improving the efficiency of photocatalytic conversion.

In order to further obtain the band gap width of the Ag₂NCN nanorods, we converted their UV diffuse reflectance spectrum into a K-M equation curve, according to the Kubelka Munk function:

(Ahν)^1/n = C(hν − E_g)

(2)

In the Kubelka Munk function, A is the absorbance, h is the Planck constant, ν is the optical frequency, 1/n is related to the type of semiconductor (generally 2 for a direct transition semiconductor), C is a constant (generally 1), and E_g is the band gap width. As shown in Figure 4b, the band gap width of the untreated Ag₂NCN nanorods was about 2.15 eV. After optimization, the band gap width was reduced to 2.02 eV. This showed that the optimization of the crystal structure decreased the band gap width, which enhanced the photocatalytic activity.

We mixed P3HT with Ag₂NCN nanorods in chloroform to obtain P3HT@Ag₂NCN composite films by a spin coating process. We then characterized the morphology of the nanorods via SEM. We used the amount of Ag₂NCN nanorods with a mass fraction of 10% in the P3HT@Ag₂NCN composite film. Figure 5a shows an SEM image of the P3HT@Ag₂NCN composite film. As shown, the Ag₂NCN nanorods were uniformly distributed on the substrate, and the surface and edges of the nanorods became blurred, which may have been due to the P3HT film covering the Ag₂NCN particles. Next, we characterized the substrate surface using EDS elemental analysis; see Figure 5b–e. As shown, the surface of the P3HT@Ag₂NCN composite film was composed of Ag, C, N, and S. The uniform distribution of S indicated that the P3HT film was evenly distributed on the Ag₂NCN.

Table 1. Mass distribution of elements in EDS.

Element	Wt%	Wt% Sigma
C	9.45	0.27
N	11.15	0.41
Ag	76.56	0.61
S	2.84	0.26
Total	100.00

shows the mass distribution of each element in the EDS. Among them, the contents of Ag, C, N, and S were 76.56%, 9.45%, 11.15%, and 2.84%, respectively.

We also performed infrared spectroscopic characterization of the P3HT@Ag₂NCN composite film; see Figure 6. As shown, the absorption peaks of the composite film were the superposition of the absorption peaks of Ag₂NCN and P3HT, and there were no new peaks. In addition, it may clearly be seen that the vibration peaks of the symmetrical and asymmetrical structures belonging to the [NCN]²⁻ at 1963.48 and 632.98 cm⁻¹ in Ag₂NCN had been shifted toward a high wave number at 1966.75 and 634.04 cm⁻¹ in the P3HT@Ag₂NCN composite film. Additionally, aliphatic C-H telescopic vibrations at 2923 cm⁻¹ and extra-hydrocarbon bending vibrations on the thiophene ring at 821 cm⁻¹ in P3HT were slightly shifted toward a high wave number at 2926 and 823 cm⁻¹. These results indicate the presence of physical interaction between Ag₂NCN and P3HT.

To further explore the interaction between P3HT and Ag₂NCN in P3HT@Ag₂NCN composite films, we characterized the films by XPS; see Figure 7. From Figure 7b–d, we can observe that the binding energy of cyanide silver Ag 3d_3/2 and Ag 3d_5/2 was shifted from 368.5 eV and 374.5 eV in Ag₂NCN to 367.75 eV and 373.70 eV in P3HT@Ag₂NCN composite films. Additionally, the binding energy of graphite carbon was reduced from 284.81 eV to 284.67 eV, and that of C-N was reduced from 397.93 eV to 397.54 eV. The slight shift in binding energy indicated physical interaction between P3HT and Ag₂NCN.

Next, we measured the SERS performance of the P3HT@Ag₂NCN composite substrate. The MB was chosen as probe molecule. Firstly, we assessed the SERS performance of the Ag₂NCN nanorods substrate. Figure 8 shows the SERS spectra of the Ag₂NCN nanorods before and after treatment. As shown, the original Ag₂NCN nanorods achieved effective detection of MB, and the characteristic peak strength at 1621 cm⁻¹ in the case of 10⁻⁵ M remained clearly visible. After hydrothermal treatment, the SERS ability of Ag₂NCN nanorods was improved; this manifested in a significant increase in strength. Therefore, the optimization of the crystal structure of Ag₂NCN nanorods greatly improved their SERS performance, and the detection limit of MB reached 10⁻⁵ M.

In order to explore the SERS performance of the P3HT@Ag₂NCN composite substrate, we used composite films with different mass fractions of Ag₂NCN nanorods as the SERS substrate and MB as the probe molecule. Figure 9 shows the SERS spectra of MB (10⁻⁴ M) on the composite films with different mass fractions of Ag₂NCN nanorods. As shown, the characteristic peak signal of the composite substrate with 5% addition was relatively weak. In the case of the composite film with 5% Ag₂NCN nanorods, the probe molecules were not adsorbed on the position of Ag₂NCN nanorods. With a 10% addition, the intensity of the characteristic peak at 1621 cm⁻¹ was relatively high. When the addition amount was more than 10%, the signal intensity at the characteristic peak at 1621 cm⁻¹ was almost the same. Therefore, in the subsequent experiments, we used a 10% mass fraction of Ag₂NCN nanorods.

Figure 10a shows the SERS spectra of MB with different concentrations on the composite films. No Raman enhancement was seen on the glass wafer. The obvious Raman peak at 1622 cm⁻¹, which resulted from the C-C ring stretching of MB, was observed on the composite films at an excitation wavelength of 532 nm (see Figure 5a). In the case of the P3HT@Ag₂NCN composite substrate, the Raman peak at 1622 cm⁻¹ was seen at 10⁻⁴ M and remained until 10⁻⁸ M. The detection limit of the P3HT@Ag₂NCN composite substrate was 10⁻⁸ M, and the enhancement factor was about 6147. Our previous paper [26] reported that the enhancement factor of the pure P3HT film is only 848 and the detection limit is 10⁻⁶ M. The SERS performance of Ag₂NCN nanorods was also found to be weak and its enhancement factor was only 82. Clearly, the result indicates that the SERS performance of the P3HT@Ag₂NCN composite substrate was far higher than that of thin pure P3HT films and Ag₂NCN nanorods. Additionally, Figure 10b,c showed an MB (10⁻⁵ M) SERS spectra of 10 random points on the composite substrate. As shown in Figure 10b,c, the relative standard deviation (RSD) of these SERS intensities was only 4.8%. Therefore, the composite substrate had a high degree of uniformity. The result indicates that the composite substrate not only had an excellent SERS property, but also a high degree of uniformity. Our previous paper [26] showed that significant π-stacking of the probe molecules with the thiophene π-cores molecules from P3HT plays an important role in SERS enhancement by the charge transfer mechanism. More effective chemisorption of the MB molecules with proper molecular orientations could facilitate the charge transfer of substrate and MB molecules. Our results show that there was an obvious physical interaction between Ag₂NCN and P3HT. The proposed organic-inorganic semiconductor heterojunction structure further improves charge transfer (CT) efficiency and the SERS property.

In addition, we mixed Ag₂NCN nanorods as a photocatalytic material with P3HT to make a composite film in order to achieve the reusability of the composite substrate based on its photocatalytic properties. We also evaluated the potential photocatalytic properties of the Ag₂NCN nanorods via the degradation of MB under visible light irradiation. We employed light illumination (λ > 420 nm) to drive the degradation of MB at room temperature. The reaction was monitored using UV-vis spectroscopy. Figure 11a presents the UV-vis spectra over time. As shown, the characteristic absorption peak intensities of MB significantly decreased with increasing reaction time. Figure 11b shows the kinetics of the peak photocatalytic performance of the Ag₂NCN nanorods. The photocatalytic efficiency of the Ag₂NCN nanorods was 92.8%. We used light illumination (λ > 420 nm) to illuminate the composite substrate of the SERS experiment for 8 h in order to perform SERS testing again and observed a change in the characteristic peak. Figure 12a shows the SERS spectra of the composite film during the illumination cycle. We observed that the composite substrate after the SERS experiment disappeared after illumination, indicating that the Ag₂NCN nanorods had achieved catalytic degradation of MB during illumination. Then, after the same position had been added to the MB solution, its characteristic peak appeared to be stable, indicating that the composite substrate can be reused. After four SERS cycles using the complex substrate shown in Figure 12b, stable SERS activity remained. The results show that the SERS substrates of the P3HT@Ag₂NCN composite film had high sensitivity, good uniformity, and excellent reusability.

4. Conclusions

In summary, an organic-inorganic semiconductor P3HT@Ag₂NCN composite film fabricated by the spin-coating process displayed good SERS properties and was shown to be reusable. The SERS substrate not only had a high sensitivity (the enhancer factor was 6147 ± 300), but also had excellent signal reproducibility (RSD = 4.8%). The enhancement factor of P3HT@Ag₂NCN composite films (6147 ± 300) for MB was more than 7 times that of the P3HT substrate (848 ± 85) and about 75 times that of Ag₂NCN nanorods (82 ± 8). Significant π-stacking of the probe molecules with the thiophene π-cores molecules from P3HT played an important role in the large SERS enhancement by the charge transfer mechanism. There was physical interaction between P3HT and Ag₂NCN nanorods. Thus, our organic-inorganic semiconductor P3HT@Ag₂NCN heterojunction structure further improved the charge transfer efficiency and the SERS property. In addition, stable SERS activity remained after four illumination cycles, demonstrating excellent reusability. The approach described in this study provides great potential to develop an ideal SERS substrate with high sensitivity, selectivity, long-term stability, low cost, and reusability.