Open Access
Add to BibSonomy
Abstract
Semiconductor-based surface enhanced Raman scattering (SERS) substrate design has attracted much interest due to the excellent photoelectronic and biochemical properties. The structural change caused by twin in semiconductor will have an influence on improving the Raman signals enhancement based on the chemical mechanism (CM). Here, we demonstrated the twin in semiconductor ZnSe nanowires as an ultrasensitive CM-based SERS platform. The SERS signals of the rhodamine 6G (R6G) and crystal violet (CV) molecules adsorbed on twin-ZnSe nanowires could be easily detected even with an ultralow concentration of 10−11 M and 10−8 M, respectively, and the corresponding enhancement factor (EF) were up to 6.12 × 107 and 3.02 × 105, respectively. In addition, the charge transfer (CT) between the twin-ZnSe nanowires and R6G molecule has been demonstrated theoretically with first-principles calculations based on density-functional theory (DFT). These results demonstrated the proposed ZnSe nanowires with twin as SERS substrate has a broader application in the field of biochemical sensing.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Surface-enhanced Raman spectroscopy (SERS) can achieve an ultrasensitive and nonspecific detection of biochemical molecules which has an extensive application in biosensing, food safety, environmental monitoring, and diagnosis of diseases fields [1–5]. The enhanced mechanisms of SERS have always been the focus of attention, which will contribute to obtaining excellent behaviors in SERS measurements. Generally, researchers believe that two mechanisms, electromagnetic enhancement mechanism (EM) and chemical enhancement mechanism (CM), are responsible for SERS [6–8]. EM originates from the several orders-of-magnitude enhancements of the local electromagnetic fields contributed from the localized surface plasmon resonance (LSPR) excited by the laser, while CM is caused by charge transfer (CT) between the absorbed molecules and SERS substrate [9–12]. Metallic nanostructures show the most powerful SERS performances with the enhancement factors (EF) even up to 106-1011 owing to the existence of EM [13,14]. However, SERS substrates fabricated by using noble metal nanostructures are expensive, and they lack of biocompatibility and high stability caused by oxidization. Besides, the side-reactions of catalytic performance from metal have an influence on the true fingerprint information of probe molecules [15]. These disadvantages hamper the real application of the metal-based SERS substrates. To overcome these obstacles, various semiconductors have been introduced into SERS substrates, where CM plays an important role in Raman scattering enhancement [16–20]. Compared to noble metals, semiconductors as SERS substrates have spectral reproducibility, high uniformity, good anti-interference, resistance to degradation, and excellent biocompatibility.
However, the limit of detection (LOD) and EF of semiconductor SERS substrate are about 10−3 M and 10-100, respectively, and such a SERS substrate with poor enhancement performance cannot be effectively applied in chemical and biological sensing. Thus, it is an extremely urgent task to substantially improve the performance of semiconductor SERS substrate. Recently, many results indicated that changing the crystal structure and modulating band gap are the promising strategies to enhance the charge-transfer resonance and exciton resonance efficiency between the molecules and the substrate [21–27]. With these methods, the practical application of semiconductor nanostructures with excellent SERS performances in biochemical sensors may be comparable to or even exceed that of metallic materials.
Besides, one-dimensional (1D) semiconductor nanomaterials have many interesting properties that are not observed in bulk or 3D materials. For example, electrons in nanowires are quantum confined laterally and thus occupy energy levels that are different from the traditional continuum of energy bands found in bulk materials. The specific surface area of 1D nanomaterial is larger than that of 2D and 3D nanomaterials, leading to the serious incoordination of surface atoms and providing the richer active sites for the adsorbed molecules [23]. Additionally, it has been proved in previous work that twin crystal has a special-low energy interface, and twin boundary has little influence on the scattering of conducting electrons. Therefore, twin can efficiently increase the CT between the absorbed molecules and substrate due to the defect of energy level [22,24]. Thus, combining the merits of 1D nanomaterial with twin will improve the charge-transfer efficiency, and further promote the performance of semiconductor SERS substrate.
Zinc selenide (ZnSe) is an important direct band gap semiconductor material, with a room temperature band gap energy at about 2.67 eV. This suggests that ZnSe is a broad potential material for short-wavelength optoelectronic devices. Recently, visible light emission and detection from ZnSe nanowires have been realized [28–30]. In addition, a study has shown that ZnSe film can be used for SERS due to CM [31]. However, there is little report on the SERS of ZnSe nanowires, especially by changing the crystal structure and modulating band gap, which is highly desired to further help the EF break through 106.
In this work, the effect of twin on SERS in ZnSe semiconductor nanowires was studied. We fabricated twin-ZnSe nanowires by physical vapor deposition (PVD) for semiconductor SERS-active substrate that possesses unparalleled SERS activity with an EF up to 6.12 × 107. The enhancement mechanism based on the CT arising from the defect band and large specific surface area in twin-ZnSe nanowires substrate was demonstrated by the experimental results and the first-principles methods based on density functional theory (DFT) simulation. Furthermore, the twin has a special low-energy interface and little scattering effect on conducting electrons. The electrons of twin can transfer to the probe molecules more easily, while enhancing the SERS signal. The ZnSe nanowires with twin as a SERS substrate have a broader application potential and can be further used as a biochemical sensor.
2. Experimental setup
2.1 Fabrication of the twin-, wurtzite- and zinc blende-ZnSe nanostructures
The growth of the ZnSe nanowires and nanoribbons was carried out in a horizontal tube furnace with a quartz tube through a PVD method. In detail, high-purity ZnSe powder (0.03 g) as the source in quartz boat was placed at the center of the furnace, the SiO2 (300 nm)/Si substrates coated with gold nanoparticles (the diameter of 20 nm) catalyst was placed in the downstream of the source. The distance (D) between the source and substrate was in a range from 6 cm to 11 cm. The fabricated temperature can be controlled by adjusting the D between the SiO2/Si substrate and the source as shown in Fig. 1. The size of the silicon wafer is 1 cm × 1 cm. Prior to the experiment, the quartz boat and Si wafers were ultrasonically cleaned with acetone, ethanol and deionized water. A carrier gas of high-purity argon was continuously introduced into the tube at a flow rate of 10 standard cubic centimeters per minute (sccm) and a pressure of 10 Pa throughout the whole experiment. The temperature of the central zone was heated to reach a temperature of 1000 °C for 20 °C/min and kept at this temperature for 30 min. Therefore, the temperature will change with increasing distance from 6 cm to 11 cm. In addition, to demonstrate the CM from the twin-ZnSe nanowires, the Ag film (The thickness is about 30 nm) deposited on SiO2/Si substrate by thermal evaporation, and the SiO2 (300 nm)/Si substrates coated with gold nanoparticles (the diameter of 20 nm) catalyst as the contrast substrates were also fabricated.
2.2 Characterization
The morphologies of the synthesized ZnSe nanostructures were characterized by utilizing scanning electron microscopy (SEM, ZEISS, Sigma 500) with an accelerating voltage of 5 kV. Transmission electron microcopy (TEM, JEOL, JEM-2100) with an accelerating voltage of 200 kV was used to obtain TEM and high resolution TEM (HRTEM) images. The X-ray diffractometer (XRD, Rigaku, Dmax-2200) was used to characterize the structural properties of the ZnSe nanostructures. The Xray and ultraviolet photoelectron spectroscopy (XPS and UPS) were employed to obtain qualitative analysis and investigate the valence band. Photoluminescence spectra (PL) of ZnSe nanostructures were obtained, in which the laser wavelength of 325 nm was chosen. Aqueous solution of R6G molecules with different concentrations (10−12 M to 10−6 M) was obtained via sequential diluting processes. For each SERS detection, 10 µL R6G molecules solution was dropped on the fabricated samples followed by the air-dry. SERS performances were characterized by Raman spectrometer (Horiba, LabRAM, HR Evolution), in which the laser wavelength of 532 nm, the laser excitation energy of 0.48 mW, the spot size of 1 µm and the diffraction grid of 1800 gr/mm were chosen. The integration time was 4 s, and a 50 × objective was used throughout the experiment. In Raman measurements, we try to choose an individual nanowire that are not tangled or disjoint with other wires.
2.3 Simulation calculations
In our works, first-principles calculations were performed based on DFT by using Vienna Ab initio Simulation Package (VASP) with the projector augmented wave (PAW) method [32,33]. The cutoff kinetic energy of 400 eV was chosen for the plane wave basis. For the exchange-correlation energy, we employed the Perdew-Burke-Ernzerhof (PBE) functional with generalized gradient approximation (GGA) [34]. For the one-dimensional nanoribbons, a vacuum thickness of 15 Å in two aperiodic directions was used to eliminate the interactions between neighboring layers. Based on the test of convergence accuracy, we use (9 × 1 × 1) Monkhorst-Pack k-point mesh to represent the reciprocal space of the unit cell. By using the conjugated gradient algorithm, the crystal structure is fully optimized, the convergence criteria for energy and force are set to be 10−4 eV and 0.01 eV Å−1, respectively.
3. Results and discussion
3.1 Characterization of ZnSe nanostructures
The surface morphology characterizations of samples obtained from different positions were presented in Fig. 2(a)–(c). In the case of D = 6 cm, the nanowire had a uniform diameter of 120 nm with a smooth edge. The straightforward nanoribbon with the width of 1.6 µm was obtained at D = 8 cm (Fig. 2(c)). Figure 2(b) showed the “saw blades-like” hybrid nanowire. The width of the sawtooth was about 163 nm, and that of the backbone was about 359 nm. The specific surface area of the “saw blades-like” hybrid structure is larger than that of nanowires and nanoribbons, which is more advantageous for the adsorption of the probe molecules, thereby improving the charge transfer efficiency. The nanoribbons with a similar morphology as samples presented in Fig. 2(c) were fabricated at a low temperature position, but the diameters became wider with the decrease of temperature (Fig. 2(d)–(f)). Representative ZnSe nanostructures were further characterized by XRD. A strong signal at 2θ = 33.0° is assigned to (402) lattice planes of SiO2 (JCPDS 44-1394) substrate. The XRD spectrum of the nanowire sample obtained at D = 6 cm exhibited two main peaks at 2θ = 27.4° and 45.4° assigned to (002) and (110) lattice planes, which can be indexed as the hexagonal wurtzite structure of ZnSe (JCPDS 15-0105) (Fig. 3(a)), with lattice parameters a = 3.996 Å, b = 3.996 Å and c = 6.550 Å. For the sample grown at D = 7 cm (Fig. 3(b)), except for the (002) and (110) peaks shown in Fig. 3(a), two new characteristic peaks at 2θ = 27.2° and 45.1° are assigned to (111) and (220) lattice planes of the zinc blende structure of ZnSe (JCPDS 37-1463). Figure 3(c) and Fig. 3(d)–(f) show the pure zinc blende structure samples grown in the range of 8-11 cm. The samples grown at 8 and 9 cm had a zinc blende structure with a (111) lattice plane. The (220) lattice plane was the primary growth orientation for the samples grown at D = 10 and 11 cm.
Fig. 2. SEM images of the obtained ZnSe nanostructures from D = (a) 6 cm, (b) 7 cm, (c) 8 cm, (d) 9 cm, (e) 10 cm and (f) 11 cm, respectively.
Fig. 3. XRD patterns of the obtained ZnSe nanostructures from D = (a) 6 cm, (b) 7 cm, (c) 8 cm, (d) 9 cm, (e) 10 cm and (f) 11 cm, respectively.
In order to further give creditable evidence for the successful fabrication of ZnSe nanostructures, XPS was employed to further obtain qualitative analysis. Figures 4(a) and (b) present the high resolution XPS spectra of Se and Zn, respectively. The detailed Se 3d peak can be decomposed into two peaks at 54.4 eV and 58.8 eV, which illustrates the presence of Se 3d5/2 and Se 3d3/2 orbital quantum number of negative divalent selenium ion (Se2-) from ZnSe. The Zn 2p peak could be fitted into only two doublets arising from Zn 2p3/2 and Zn 2p1/2 orbitals located at 1021.7 eV and 1044.8 eV, which can be indexed to Zn-Se bonding in ZnSe. Thus, the wurtzite, zinc blende and hybrid structured ZnSe nanomaterials could be successfully fabricated by changing the distance, i.e., growth temperature.
The low magnification TEM images of different ZnSe nanostructures were exhibited in Fig. 5(a)–(c). The corresponding typical local high resolution TEM (HRTEM) images were shown in Fig. 5(d)–(f), respectively. The interplanar spacings shown in Figs. 5(d) and (f), attributing to the (002) plane of wurtzite ZnSe structure and the (111) plane for zinc blende ZnSe structure. Compared to the HRTEM images in Figs. 5(d) and (f), a hybrid structure was obviously observed as shown in Fig. 5(e). It revealed the high dislocation density at the grain boundaries and the high density stacking faults in the twin domains. Two adjacent face-centered cubic domains from a (111) twin boundary (TB) shown in Fig. 5(e). Twin-ZnSe nanowires have a larger specific surface area than smooth nanowires and nanoribbons, probably impacting on the CT efficiency and magnifying the molecular polarization, ultimately resulting in the enhanced SERS signals for the adsorbed probe molecules. The selected-area electron diffraction (SAED) patterns (Figs. 5(g) and (i)) verified that wurtzite and zinc blende are correspond with (002) and (111), respectively. In Fig. 5(h), the SAED pattern arises unambiguously from the (111) twinning structure commonly appearing in the face-centered cubic structure systems. All diffraction spots in the present pattern can be easily identified by twinned nanocrystals. Raman spectra of the prepared wurtzite-, twin-, and zinc blende-ZnSe were further collected as shown the inset in Figs. 5(d), (e) and (f). The spectrum of wurtzite-ZnSe nanowire exhibited two characteristic peaks at ∼203 and ∼251 cm-1 attributed to the scattering of the transverse optic (TO) and longitudinal optic (LO) phonon modes of ZnSe, respectively [35]. In the inset of Fig. 5(e), an obvious blue shift (the characteristic peaks at 213 cm-1 and 258 cm-1) relative to the samples of D = 6 cm could be observed for the ZnSe nanostructure prepared at the lower growth temperature, which may be attributed to the different surface morphologies and caused by the quantum confinement effects and large specific surface area, and further indicates the ZnSe is polycrystalline nanostructure [23].
Fig. 5. TEM and HRTEM characterizations of the obtained ZnSe nanostructures. TEM images, HRTEM images and the corresponding SAED patterns of the obtained ZnSe nanostructures grown at D = 6 cm (a, d, and g), 7 cm (b, e, and h), and 8 cm (c, f, and i), respectively. Insets in (d), (e) and (f) are the Raman spectra of ZnSe nanostructures grown at D = 6 cm, 7 cm and 8 cm, respectively.
As shown in Fig. 6, the characteristic peaks of the Raman spectra of the samples (D = 8, 9, 10, 11 cm) gradually red-shifted compared to the sample grown at D = 7 cm. The different Raman peaks of the various ZnSe nanostructures are attributed to the changes in the energy levels of the material. For the twin-ZnSe nanowires, there is may be defect band (DB) that causes a larger shifting of Raman peak. There is the same distance between the two Raman peaks of the fabricated ZnSe nanostructures grown at the D = 8, 9, 10, 11 cm, which could be attributed to the same zinc blende structure in these samples.
Fig. 6. Raman spectra of the ZnSe nanostructures fabricated at different locations (D = 9, 10 and 11 cm).
To analyze defect-related emission in ZnSe, PL measurements of the different ZnSe nanostructures were carried out. As shown in Fig. 7(a), the PL emission spectra of different ZnSe nanostructures comprise two emission bands. One is a narrow near band edge (NBE) peak centered at 460.7 nm corresponding to 2.69 eV, and the other one is a broad deep level emission (DLE) at 633.5 nm, 624.4 nm and 632.2 nm for wurtzite-, twin- and zinc blende ZnSe nanostructures, respectively. There is a blue-shift compared to that of bulk ZnSe (the band gap of 2.67 eV). This bandgap extension in ZnSe nanowires was caused by strong quantum confinement [36,37]. The DLE depends on the preparation conditions, i.e., the growth temperature in our work. For samples with wurtzite- and zinc blende-ZnSe nanostructures, the PL spectra were dominated by NBE emission with a lower emission from the DLE states (black and blue curves in Fig. 7(a)). However, there was almost equivalent intensity of both peaks in the PL spectrum of twin-ZnSe nanowires. The relative intensities of DLE and NBE peaks and their ratios were collected to plot the histogram and broken line as illustrated in Fig. 7(b). Obviously, the maximum ratio can be obtained from twin-ZnSe nanowires, which could be attributed to the obvious twin defects introduced by the fabrication process [38,39]. Therefore, such a defect will lead to the defect band as a transition level that will make the charge transfer easier between twin-ZnSe nanowires and probe molecules, and one can obtain the optimal SERS performance from twin-ZnSe nanowires.
Fig. 7. (a) PL spectra of different ZnSe nanostructures, (b) the collected relative intensities of NBE and DLE peaks and their ratios versus different ZnSe nanostructures.
3.2 SERS performances of ZnSe nanostructures
To verify the hypothesis mentioned above, the R6G was adopted as a probe molecule to appraise the SERS performance of different ZnSe nanostructures, and the schematic diagram was shown in Fig. 8(a). As shown in Fig. 8(b), the SERS spectra of R6G with the concentration of 10−6 M deposited on wurtzite-, zinc blende-, and twin-ZnSe nanostructures exhibited all of the characteristic Raman peaks of R6G [40]. Three prominent peaks at 612, 773, and 1362 cm-1 of R6G molecules are assigned to C-C-C ring in-plane bend, C-H on-plane bend, and aromatic C-C stretching vibration modes, respectively. The intensities of the prominent peaks at 612, 773, and 1362 cm-1 (Fig. 8(c)) were chosen as interior labels to quantitatively compare SERS enhancement of the ZnSe substrates with different crystal nanostructures. The Raman peak intensities of R6G on the twin-ZnSe nanowires substrate were much higher than those on the pure wurtzite and zinc blende samples, which verified our previous hypothesis. In addition, obvious positive shifts of the binding energy for all the Se and Zn elements in twin-ZnSe adsorbed by R6G molecules were observed compared with bare twin-ZnSe (Fig. 9). This observation is consistent with the decline of the valence band in twin-ZnSe adsorbed by R6G relative to the Fermi level, which is due to the transfer of electron from R6G to twin-ZnSe. It was obvious that the optimal SERS activities could be achieved, which could be attributed to the more CT due to the existence of the defect band induced by the structural defect in the twin-ZnSe nanowires (Fig. 2(b) and Fig. 5(e)).
Fig. 8. SERS performance on different substrates. (a) Schematic diagram of the twin-ZnSe nanowires for Raman enhancing of R6G molecules. (b) The SERS spectra of R6G molecules (10−6 M) detected on wurtzite-, twin-, and zinc blende-ZnSe nanostructure substrate, respectively. (c) The corresponding SERS intensities histograms of Raman peaks (612, 773, and 1362 cm-1) for R6G. (d) SERS spectra of R6G (10−6 M) molecules adsorbed on twin-ZnSe nanowires, Ag film and Au nanoparticles, and the normal Raman spectrum of R6G (10−3 M) molecules on the SiO2/Si substrate.
Fig. 9. High resolution XPS spectra of the (a) Se 3d and (b) Zn 2p in twin-ZnSe nanowires adsorbed by R6G molecules.
Figure 8(d) presented the spectra of R6G molecules with the concentration of 10−6 M adsorbed on the twin-ZnSe nanowires, Ag film surface and SiO2/Si substrate with sporadic Au nanoparticles, and the spectrum of R6G molecules (10−3 M) adsorbed on the SiO2/Si substrate. When the R6G molecules were adsorbed SiO2/Si substrate with sporadic Au nanoparticles, there was no obvious Raman spectral characteristic peak, which proved that the pre-deposited gold nanoparticles on the SiO2/Si substrate had no effect on the SERS. The SERS activity was mainly derived from twin-ZnSe nanowires, which could eliminate the influence of the gold catalyst. The SERS spectrum of R6G adsorbed on Ag surface exhibited several characteristic peaks located at 610, 769, 1506, and 1647 cm-1, which were same as that of traditional Raman spectrum of R6G molecules on SiO2/Si substrate. However, the characteristic peaks for the R6G molecules adsorbed twin-ZnSe nanowires are at 612 cm-1 (C-C-C ring in-plane bend), 773 cm-1 (C-H on-plane bend), 1509 cm-1 and 1649 cm-1 (aromatic C-C stretching). The difference of the characteristic bands of these two spectra could be attributed to the different modes of the R6G molecules on the surface of different materials. For the traditional Raman spectrum of R6G molecules on SiO2/Si substrate and Ag surface, the relative intensities of several characteristic peaks at 1180 cm-1, 1506 cm-1 (aromatic C-C stretching), and 1647 cm-1 (aromatic C-C stretching) were much lower than those at 610, 769 and 1359 cm-1. However, for the R6G adsorbed on the twin-ZnSe nanowires, the relative intensities of the characteristic peaks (at 1182, 1509, and 1649 cm-1) were almost of the same as that at 773 and 1362 cm-1. The obvious peaks at 1182 cm-1, 1509 cm-1 (aromatic C-C stretching) and, 1649 cm-1 (aromatic C-C stretching) were correlated with the C-C mode resulting in these more pronounced Raman peaks associated with C-C modes. It could be attributed to the selective enhancement by the CT mechanism through Herzberg-Teller contributions [22].
Figure 10(a) showed the SERS spectra of R6G molecules with the concentration from 10−12 M to 10−6 M. The characteristic peaks of the R6G could be straightforwardly distinguished even at the extremely low concentration of 10−11 M, which demonstrated the LOD for R6G molecules is 10−11 M and the desired sensitivity. EF is an effective method to evaluate the contribution of ZnSe nanostructures to the enhanced Raman spectra of R6G molecules. The EF could be calculated according to the equation [26]:
Where ISERS and INR are the intensities of the same band of SERS spectra and normal Raman (pure silicon substrate), and NSERS and NNR are respectively R6G dye molecules number excited by SERS laser spot, and normal Raman condition. NSERS is equal to cVNAA1/Asubstrate, where c is the concentration (10−11 M) of R6G molecules with the droplet volume (V = 2 µL) drop-casted on SERS substrate whose surface area is Asubstrate ≈ 4.58 × 10−8 cm2. NNR is written as NNR = ρHNAA1/M, where ρ and M are the density (1.15 g/cm3) and molar mass (479.02 g/mol) of R6G molecules, respectively, and H is the penetration depth (21 µm) of laser with 532 nm wavelength [30]. And, NA and A1 are the Avogadro constant and the laser spot area, respectively. The EF can be easily calculated as 6.12 × 107. The EF was compared to the previous works of semiconductor SERS substrates as shown in Table 1.Fig. 10. SERS performance of twin-ZnSe nanowires. (a) Sensitivity: SERS spectra of R6G molecules (10−12 M to 10−6 M) on twin-ZnSe nanowires substrate. (b) Quantifiability: Raman intensities of R6G molecule at peak 612, 1182, 1362 and 1649 cm-1 as the function of concentration. (c) SERS spectra of CV molecules (10−7 M to 10−3 M) on twin-ZnSe nanowires substrate. (d) Raman intensities of CV molecule at peak 904 and 1614 cm-1 as the function of concentration. (e)Spot-to-spot uniformity: ten SERS spectra of the R6G molecules (10−8 M) detected from ten random spots on one substrate. (f)Substrate-to-substrate reproducibility: The SERS intensity histograms of the R6G (10−8 M) peaks at 613 cm-1 (red), 774 cm-1 (blue), and 1363 cm-1 (green) collected from 8 batches.
Table 1. Comparison of our work with other semiconductor-SERS substrates reported
The excellent sensitivity and EF could be attributed to the coupling of molecular resonance and photo-induced charge transfer resonance. Figure 10(b) showed the relationship between the relative intensity of the characteristic peak at 612 cm-1 and the molecular concentration from 10−11 M to 10−6 M fits well with linear curve in a log-log scale (The linear equation is log(I) = 4.67 + 0.25 × log(C), I and C is the relative SERS intensity of the peak and the R6G concentration, respectively; linear coefficient R2 is ∼0.991), which demonstrated a great potential of twin-ZnSe nanowires substrate for the application in quantitative detection. In Fig. 10(b), linearity in the fitting data should be observed at each concentration of analytes in experiment, however, compared to that of the 613 cm-1 peak, there is a larger deviation between fitting and experimental data under some concentrations for 1182, 1362 and 1649 cm-1 peaks. This phenomenon can be attributed to the following reasons: (1) Compared to EM, CM is lack of universality for enhancing SERS signals of every peak of R6G molecules; (2) The CT between ZnSe and R6G lead to the selective enhancement for different vibrational modes of R6G, which further demonstrate the CM from twin-ZnSe nanowires SERS substrate. The error bars in Fig. 10(b) indicated that the standard deviation of the peak intensities over different positions is very small. Similarly, crystal violet (CV) is chosen as a probe molecule to verify the excellent sensitivity and ability of quantitative detection of the proposed twin-ZnSe nanowires SERS substrate as shown in Figs. 10(c) and (d). Obviously, the fingerprint peaks at 904 cm−1 and 1614 cm−1 were detected even up to the minimum detected concentration is 10−7 M. In the detection range, SERS sensitivity at 1614 cm−1 increases linearly with the concentration of CV (Fig. 10(d)). The relation can be depicted with log(I) = 4.30 + 0.29 log(C), and the LOD of the proposed SERS substrate was calculated by the 3σ criterion and was found to be 0.79 × 10−8 M [47]. The The detailed calculation method of EF for CV molecules is the same as that of R6G. The relative intensity (≈113) of the peak at 1614 cm−1 for CV bulk [48]. ρ and M are the density (1.19 g/cm3) and molar mass (407.99 g/mol) of CV molecules, respectively. The EF can be easily calculated as 3.02 × 105. The EF was also compared to the previous works of semiconductor SERS substrates as shown in Table 1. To further investigate the uniformity of twin crystal structure as SERS substrate from spot to spot, the SERS spectra of the R6G molecules (10−6 M) from ten random points on one twin crystal sample were detected and shown in Fig. 10(e). The same intensity of each vibrational mode demonstrated a good uniformity of the proposed twin-ZnSe nanowires SERS substrate, which could be attributed to the homogeneous lattice structure. Ten batches of twin-ZnSe nanowires SERS substrates were fabricated to investigate the substrate-to-substrate reproducibility measurement of the R6G molecule (10−8 M). As shown in Fig. 10(f), the intensities of the three different peaks at 612, 773, and 1362 cm-1 were collected to plot the histogram, and the relative standard deviations (RSD) of 5.95%, 6.58%, and 11.48% corresponding to the above three peaks could be obtained, which were calculated with the standard formula [8]:
3.3 Mechanism of the SERS enhancement of ZnSe nanostructures
Figure 11 showed the charge difference distributions for R6G molecule chemically adsorbed on (002) lattice plane of wurtzite, (111) lattice plane of zinc blende, and (220) lattice plane of zinc blende, respectively. The charge accumulation and the depletion were labeled as yellow and blue distributions, respectively. Thus, compared to the CT process only in R6G molecule (Fig. 11(a)), there were the CTs between (111) or (220) lattice plane of zinc blende and R6G molecule as shown in Figs. 11(b) and (c), which may be attributed to facet-dependent interfacial CT process [21]. For the (111) lattice plane, the amounts of electrons transferred to R6G molecule were larger than that from the (220) and the (002) lattice plane. Significantly, twin-ZnSe nanowires were formed by stacking (111) planes, which are more likely to improve the CT efficiency. The theoretical calculation results agreed with the SERS behaviors obtained from the experiment above (Fig. 10).
Fig. 11. DFT simulation results verification of CT mechanism. Charge difference distributions for R6G molecule chemically adsorbed on (a) (002) lattice plane of wurtzite, (b) (111) lattice plane of zinc blende, and (c) (220) lattice plane of zinc blende, respectively. (d) Schematic illustration of energy-level for illustrating the CT between ZnSe and R6G under the excitation of 532 nm wavelength laser.
According to the research of Chen et al., the conduction band (CB) of semiconductor ZnSe locate at -4.09 eV [49]. In our work, by studying the PL spectrum of ZnSe nanowires, the band gap of ZnSe nanowires is 2.69 eV. Therefore, according to the research of Chen et al., the conduction band (CB) of semiconductor ZnSe locates at -4.09 eV [49]. In our work, by studying the PL spectrum of ZnSe nanowires, the band gap of ZnSe nanowires is 2.69 eV. Therefore, according to equation:
Where EVB, ECB and Eg are the value of valence band (VB), CB, and band gap, respectively. The VB of ZnSe nanowires locates at -6.78 eV. The DB related to twin are well separated from the bottom of CB, and below the CB minimum. R6G as a typical SERS probe molecule has the highest-occupied molecular orbital (HOMO) and lowest-unoccupied molecular orbital (LUMO) levels at -5.70 and -3.40 eV, respectively [50]. We obtain the energy level diagram as shown in Fig. 11(d). There are several processes that can cause SERS: (1) molecule resonance-Raman processes (µmol), (2) the photon induced CT (µPICT) between the substrate and molecule, and (3) exciton resonance (µex) [41]. In the case of all used laser of 532 nm, the band-band transition energy of the R6G molecule can be calculated as 2.3 eV (Fig. 11(d)), which excellently fits with the excitation laser energy at 532 nm wavelength. Under these conditions, when the excitation light (532 nm) irradiates the R6G molecule, a resonance-Raman (µmol) will take place in the band gap of the R6G molecule. Due to the size effect of the material, exciton resonance (µex) may occur in the twin-ZnSe nanowires under laser irradiation. Additionally, when radiated by the incident light, the electron occupying the ground state of the molecule was directly excited from the filled highest occupied molecular orbital of the molecule (HOMO) to the energy level of the CB of twin-ZnSe nanowires. The excited electrons are immediately transferred back to the ground vibrational energy level of the molecule, and some photons are released, resulting in CT (µPICT) between ZnSe nanowires and R6G molecules. Since the excitation light source is 532 nm (about 2.3 eV), which is insufficient to excite electrons from ZnSe CB to its VB, the surface defect energy level may act as a bridge during charge transfer. The light source excites electrons from the VB to the surface defect energy level between the ZnSe band gap, and then transfers them to the LUMO of the adsorbed molecule for charge transfer (µPICT) (Fig. 11(d)). Furthermore, the twin has a special low-energy interface and little scattering effect on conducting electrons. The electrons of twin can transfer to the probe molecules more easily, enhancing the SERS signal. Thus, experimental and theoretical simulations have demonstrated the molecular polarization originated from the CT can result in a strong SERS signal of the vibrational modes of probe molecule, which is consistent with the optimal SERS activity obtained from the twin structure. As a result, the SERS activity can be improved with resonance-Raman process and CT between the substrate and molecule.4. Conclusion
In summary, we investigated the SERS characteristics of the different ZnSe nanostructures. The optimal SERS performances from the ZnSe nanowires of twin structures were obtained. Due to the defect band arising from twin, the enhanced charge-transfer resonance in twin-ZnSe nanowires substrate has been demonstrated. The EF is estimated to reach up to 6.12 × 107, preceding many other reported semiconductor SERS substrates. The impressive EF can be attributed to the coupling of molecular resonance, photo-induced charge transfer resonance, and exciton resonance. This study proposed a method to tailor the SERS performance of twin-ZnSe nanowires, which promotes the developing of semiconductor SERS.
Funding
National Natural Science Foundation of China (11774208, 11974222); Natural Science Foundation of Shandong Province (ZR2018MA040); Postgraduate Education Quality Enhancement Program of Shandong Province (SDYY18064).
Disclosures
The authors declare no conflicts of interest.
References
1. S. Y. Ding, J. Yi, J. F. Li, B. Ren, D. Y. Wu, R. Panneerselvam, and Z. Q. Tian, “Nanostructure-based plasmon-enhanced Raman spectroscopy for surface analysis of materials,” Nat. Rev. Mater. 1(6), 16021 (2016). [CrossRef]
2. J. F. Li, Y. F. Huang, Y. Ding, Z. L. Yang, S. B. Li, X. S. Zhou, F. R. Fan, W. Zhang, Z. Y. Zhou, D. Y. Wu, B. Ren, Z. L. Wang, and Z. Q. Tian, “Shell-isolated nanoparticle-enhanced Raman spectroscopy,” Nature 464(7287), 392–395 (2010). [CrossRef]
3. C. H. Li, A. H. Liu, C. Zhang, M. H. Wang, Z. Li, S. C. Xu, S. Z. Jiang, J. Yu, C. Yang, and B. Y. Man, “Ag gyrus-nanostructure supported on graphene/Au film with nanometer gap for ideal surface enhanced Raman scattering,” Opt. Express 25(17), 20631–20641 (2017). [CrossRef]
4. Q. Hao, H. Huang, X. Fan, Y. Yin, J. Wang, W. Li, T. Qiu, L. B. Ma, P. K. Chu, and O. G. Schmidt, “Controlled Patterning of Plasmonic Dimers by Using an Ultrathin Nanoporous Alumina Membrane as a Shadow Mask,” ACS Appl. Mater. Interfaces 9(41), 36199–36205 (2017). [CrossRef]
5. H. Ma, Y. F. Chen, H. Wang, X. L. Wang, X. L. Zhang, X. X. Han, C. Y. He, and B. Zhao, “Charge-transfer effect on surface-enhanced Raman spectroscopy in Ag/PTCA: Herzberg-Teller selection rules,” J. Phys. Chem. C 121(46), 25788–25794 (2017). [CrossRef]
6. Z. G. Dai, X. H. Xiao, W. Wu, Y. P. Zhang, L. Liao, S. S. Guo, J. J. Ying, C. X. Shan, M. T. Sun, and C. Z. Jiang, “Plasmon-driven reaction controlled by the number of graphene layers and localized surface plasmon distribution during optical excitation,” Light: Sci. Appl. 4(10), e342 (2015). [CrossRef]
7. C. Zhang, C. H. Li, J. Yu, S. Z. Jiang, S. C. Xu, C. Yang, Y. J. Liu, X. G. Gao, A. H. Liu, and B. Y. Man, “SERS activated platform with three-dimensional hot spots and tunable nanometer gap,” Sens. Actuators, B 258, 163–171 (2018). [CrossRef]
8. C. H. Li, J. Yu, S. C. Xu, S. Z. Jiang, X. W. Xiu, C. S. Chen, A. H. Liu, T. F. Wu, B. Y. Man, and C. Zhang, “Constructing 3D and Flexible Plasmonic Structure for High-Performance SERS Application,” Adv. Mater. Technol. 3(11), 1800174 (2018). [CrossRef]
9. Z. L. Zhang, Y. R. Fang, W. H. Wang, L. Chen, and M. T. Sun, “Propagating Surface Plasmon Polaritons: Towards Applications for Remote-Excitation Surface Catalytic Reactions,” Adv. Sci. 3(1), 1500215 (2016). [CrossRef]
10. C. H. Li, S. C. Xu, J. Yu, S. Z. Jiang, A. H. Liu, Z. Li, S. Z. Zhang, X. F. Zhao, C. Zhang, and B. Y. Man, “3D hybrid MoS2/AgNPs/inverted pyramid PMMA resonant cavity system for the excellent flexible surface enhanced Raman scattering sensor,” Sens. Actuators, B 274, 152–162 (2018). [CrossRef]
11. M. Liu, Y. Shi, G. P. Zhang, Y. H. Zhang, M. M. Wu, J. F. Ren, and B. Y. Man, “Surface-Enhanced Raman Spectroscopy of Two-Dimensional Tin Diselenide Nanoplates,” Appl. Spectrosc. 72(11), 1613–1620 (2018). [CrossRef]
12. J. Yu, Y. Guo, H. J. Wang, S. Su, C. Zhang, B. Man, and F. C. Lei, “Quasi Optical Cavity of Hierarchical ZnO Nanosheets@Ag Nanoravines with Synergy of Near-and Far-Field Effects for In Situ Raman Detection,” J. Phys. Chem. Lett. 10(13), 3676–3680 (2019). [CrossRef]
13. Y. Tian, H. F. Wang, L. Q. Yan, X. F. Zhang, A. Falak, Y. J. Guo, P. P. Chen, F. L. Dong, L. F. Sun, and W. G. Chu, “A Generalized Methodology of Designing 3D SERS Probes with Superior Detection Limit and Uniformity by Maximizing Multiple Coupling Effects,” Adv. Sci. 6, 1900177 (2019). [CrossRef]
14. X. F. Zhao, C. H. Li, Z. Li, J. Yu, J. Pan, H. P. Si, C. Yang, S. Z. Jiang, C. Zhang, and B. Y. Man, “In-situ electrospun aligned and maize-like AgNPs/PVA@Ag nanofibers for surface-enhanced Raman scattering on arbitrary surface,” Nanophotonics 8(10), 1719–1729 (2019). [CrossRef]
15. H. Li, Q. Zhang, C. C. R. Yap, B. K. Tay, T. H. T. Edwin, A. Olivier, and D. Baillargeat, “From Bulk to Monolayer MoS2: Evolution of Raman Scattering,” Adv. Funct. Mater. 22(7), 1385–1390 (2012). [CrossRef]
16. W. Li, R. Zamani, P. Rivera Gil, B. Pelaz, M. Ibáñez, D. Cadavid, A. Shavel, R. A. Alvarez-Puebla, W. J. Parak, J. Arbiol, and A. Cabot, “CuTe nanocrystals: shape and size control, plasmonic properties, and use as SERS probes and photothermal agents,” J. Am. Chem. Soc. 135(19), 7098–7101 (2013). [CrossRef]
17. D. Qi, L. Lu, L. Wang, and J. Zhang, “Improved SERS sensitivity on plasmon-free TiO2 photonic microarray by enhancing light-matter coupling,” J. Am. Chem. Soc. 136(28), 9886–9889 (2014). [CrossRef]
18. Z. Q. Tian, B. Ren, and D. Y. Wu, “Surface-enhanced Raman scattering: from noble to transition metals and from rough surfaces to ordered nanostructures,” J. Phys. Chem. B 106(37), 9463–9483 (2002). [CrossRef]
19. B. Ren, G. K. Liu, X. B. Lian, Z. L. Yang, and Z. Q. Tian, “Raman spectroscopy on transition metals. Analytical and bioanalytical chemistry,” Anal. Bioanal. Chem. 388(1), 29–45 (2007). [CrossRef]
20. Y. H. Zhang, Y. Shi, M. M. Wu, K. Zhang, B. Y. Man, and M. Liu, “Synthesis and surface-enhanced Raman scattering of ultrathin SnSe2 nanoflakes by chemical vapor deposition,” Nanomaterials 8(7), 515 (2018). [CrossRef]
21. X. T. Wang, W. X. Shi, Z. Jin, W. F. Huang, J. Lin, G. S. Ma, S. Z. Li, and L. Guo, “Remarkable SERS activity observed from amorphous ZnO nanocages,” Angew. Chem., Int. Ed. 56(33), 9851–9855 (2017). [CrossRef]
22. L. B. Yang, D. Yin, Y. Shen, M. Yang, X. L. Li, X. X. Han, X. Jiang, and B. Zhao, “Mesoporous semiconducting TiO2 with rich active sites as a remarkable substrate for surface-enhanced Raman scattering,” Phys. Chem. Chem. Phys. 19(28), 18731–18738 (2017). [CrossRef]
23. L. B. Yang, M. D. Gong, X. Jiang, D. Yin, X. Y. Qin, B. Zhao, and W. D. Ruan, “Investigation on SERS of different phase structure TiO2 nanoparticles,” J. Raman Spectrosc. 46(3), 287–292 (2015). [CrossRef]
24. J. Lin, Y. Shang, X. X. Li, J. Yu, X. T. Wang, and L. Guo, “Ultrasensitive SERS Detection by Defect Engineering on Single Cu2O Superstructure Particle,” Adv. Mater. 29(5), 1604797 (2017). [CrossRef]
25. Z. H. Zheng, S. Cong, W. B. Gong, J. N. Xuan, G. H. Li, W. B. Lu, F. X. Geng, and Z. Zhao, “Semiconductor SERS enhancement enabled by oxygen incorporation,” Nat. Commun. 8(1), 1993 (2017). [CrossRef]
26. L. Tao, K. Chen, Z. F. Chen, C. X. Cong, C. Y. Qiu, J. J. Chen, X. M. Wang, H. J. Chen, T. Yu, W. G. Xie, S. Z. Deng, and J. B. Xu, “1T’ transition metal telluride atomic layers for plasmon-free SERS at femtomolar levels,” J. Am. Chem. Soc. 140(28), 8696–8704 (2018). [CrossRef]
27. J. Lin, W. Hao, Y. Shang, X. T. Wang, D. L. Qiu, G. S. Ma, C. Chen, S. Z. Li, and L. Guo, “Direct experimental observation of facet-dependent SERS of Cu2O polyhedral,” Small 14(8), 1703274 (2018). [CrossRef]
28. D. Wisniewski, K. Byrne, C. Fernandes, C. Stewart, C. F. de Souza, and H. E. Ruda, “Fingerprinting Electronic Structure in Nanomaterials: A Methodology Illustrated by ZnSe Nanowires,” Nano Lett. 19(4), 2259–2266 (2019). [CrossRef]
29. B. Xiang, H. Z. Zhang, G. H. Li, F. H. Yang, F. H. Su, R. M. Wang, J. Xu, J. W. Lu, X. C. Sun, Q. Zhao, and D. P. Yu, “Green-Light-Emitting ZnSe Nanowires Fabricated via Vapor Phase Growth,” Appl. Phys. Lett. 82(19), 3330–3332 (2003). [CrossRef]
30. Z. M. Mu, Q. Zheng, R. P. Liu, M. W. I. Malik, D. S. Tang, W. C. Zhou, and Q. Wan, “1D ZnSSe-ZnSe Axial Heterostructure and its Application for Photodetectors,” Adv. Electron. Mater. 5(4), 1800770 (2019). [CrossRef]
31. S. K. Islam, M. Tamargo, R. Moug, and J. R. Lombardi, “Surface-enhanced Raman scattering on a chemically etched ZnSe surface,” J. Phys. Chem. C 117(44), 23372–23377 (2013). [CrossRef]
32. G. Kresse and J. Furthmüller, “Efficient Iterative Schemes for Ab Initio Total-Energy Calculations Using a Plane-Wave Basis,” Phys. Rev. B 54(16), 11169–11186 (1996). [CrossRef]
33. G. Kresse and D. Joubert, “From ultrasoft pseudopotentials to the projector augmented-wave method,” Phys. Rev. B 59(3), 1758–1775 (1999). [CrossRef]
34. J. P. Perdew, K. Burke, and M. Ernzerhof, “Generalized Gradient Approximation Made Simple,” Phys. Rev. Lett. 77(18), 3865–3868 (1996). [CrossRef]
35. C. X. Shan, Z. Liu, X. T. Zhang, C. C. Wong, and S. K. Hark, “Wurtzite ZnSe nanowires: growth, photoluminescence, and single-wire Raman properties,” Nanotechnology 17(22), 5561–5564 (2006). [CrossRef]
36. D. Li, G. Xing, S. L. Tang, X. H. Li, L. Z. Fan, and Y. C. Li, “Ultrathin ZnSe nanowires: one-pot synthesis via a heat-triggered precursor slow releasing route, controllable Mn doping and application in UV and near-visible light detection,” Nanoscale 9(39), 15044–15055 (2017). [CrossRef]
37. P. Poulopoulos, B. Lewitz, A. Straub, S. D. Pappas, S. A. Droulias, S. Baskoutas, and P. Fumagalli, “Band-gap tuning at the strong quantum confinement regime in magnetic semiconductor EuS thin films,” Appl. Phys. Lett. 100(21), 211910 (2012). [CrossRef]
38. Y. Q. Wang, U. Philipose, T. Xu, H. E. Ruda, and K. L. Kavanagh, “Twinning modulation in ZnSe nanowires,” Semicond. Sci. Technol. 22(3), 175–178 (2007). [CrossRef]
39. U. Philipose, P. Sun, T. Xu, H. E. Ruda, and L. Yang, “Structure and photoluminescence of ZnSe nanostructures fabricated by vapor phase growth,” J. Appl. Phys. 101(1), 014326 (2007). [CrossRef]
40. P. Hildebrandt and M. Stockburger, “Surface-enhanced resonance Raman spectroscopy of Rhodamine 6G adsorbed on colloidal silver,” J. Phys. Chem. 88(24), 5935–5944 (1984). [CrossRef]
41. S. Cong, Y. Yuan, Z. Chen, J. Hou, M. Yang, Y. Su, Y. Zhang, L. Li, Q. Li, F. Geng, and Z. Zhao, “Noble metal-comparable SERS enhancement from semiconducting metal oxides by making oxygen vacancies,” Nat. Commun. 6(1), 7800 (2015). [CrossRef]
42. R. Haldavnekar, K. Venkatakrishnan, and B. Tan, “Non plasmonic semiconductor quantum SERS probe as a pathway for in vitro cancer detection,” Nat. Commun. 9(1), 3065 (2018). [CrossRef]
43. X. X. Li, Y. Shang, J. Lin, A. R. Li, X. T. Wang, B. Li, and L. Guo, “Temperature-Induced Stacking to Create Cu2O Concave Sphere for Light Trapping Capable of Ultrasensitive Single-Particle,” Adv. Funct. Mater. 28(33), 1801868 (2018). [CrossRef]
44. Y. Wang, W. Ruan, J. Zhang, B. Yang, W. Xu, B. Zhao, and J. R. Lombardi, “Direct observation of surface-enhanced Raman scattering in ZnO nanocrystals,” J. Raman Spectrosc. 40(8), 1072–1077 (2009). [CrossRef]
45. X. Fu, F. Bei, X. Wang, X. Yang, and L. Lu, “Surface enhanced Raman scattering of 4-mercaptopyridine on sub-monolayers of α-Fe2O3 nanocrystals (sphere, spindle, cube),” J. Raman Spectrosc. 40(9), 1290–1295 (2009). [CrossRef]
46. L. Jiang, T. You, P. Yin, Y. Shang, D. Zhang, L. Guo, and S. Yang, “Surface-enhanced Raman scattering spectra of adsorbates on Cu2O nanospheres: charge-transfer and electromagnetic enhancement,” Nanoscale 5(7), 2784–2789 (2013). [CrossRef]
47. E. Desimoni and B. Brunetti, “About estimating the limit of detection by the signal to noise approach,” Pharm. Anal. Acta 6(4), 1–4 (2015). [CrossRef]
48. D. Maznichenko, K. Venkatakrishnan, and B. Tan, “Stimulating multiple SERS mechanisms by a nanofibrous three-dimensional network structure of titanium dioxide (TiO2),” J. Phys. Chem. C 117(1), 578–583 (2013). [CrossRef]
49. W. Chen, N. Zhang, M. Y. Zhang, X. T. Zhang, H. Gao, and J. Wen, “Controllable growth of ZnO–ZnSe heterostructures for visible-light photocatalysis,” CrystEngComm 16(6), 1201–1206 (2014). [CrossRef]
50. X. Ling, L. M. Xie, Y. Fang, H. Xu, H. L. Zhang, J. Kong, M. S. Dresselhaus, J. Zhang, and Z. F. Liu, “Can graphene be used as a substrate for Raman enhancement?” Nano Lett. 10(2), 553–561 (2010). [CrossRef]
