Sensitivity Enhancement of Transition Metal Dichalcogenides/Silicon Nanostructure-based Surface Plasmon Resonance Biosensor

In this work, we designed a sensitivity-enhanced surface plasmon resonance biosensor structure based on silicon nanosheet and two-dimensional transition metal dichalcogenides. This configuration contains six components: SF10 triangular prism, gold thin film, silicon nanosheet, two-dimensional MoS2/MoSe2/WS2/WSe2 (defined as MX2) layers, biomolecular analyte layer and sensing medium. The minimum reflectivity, sensitivity as well as the Full Width at Half Maximum of SPR curve are systematically examined by using Fresnel equations and the transfer matrix method in the visible and near infrared wavelength range (600 nm to 1024 nm). The variation of the minimum reflectivity and the change in resonance angle as the function of the number of MX2 layers are presented respectively. The results show that silicon nanosheet and MX2 layers can be served as effective light absorption medium. Under resonance conditions, the electrons in these additional dielectric layers can be transferred to the surface of gold thin film. All silicon-MX2 enhanced sensing models show much better performance than that of the conventional sensing scheme where pure Au thin film is used, the highest sensitivity can be achieved by employing 600 nm excitation light wavelength with 35 nm gold thin film and 7 nm thickness silicon nanosheet coated with monolayer WS2.


Material
Three-layer graphene was grown on high-purity (99.99%) copper foil by use of low-pressure chemical vapor deposition as the one used in our previous work [Zeng et al., Adv. Mater. 27, 6163, (2015)]. Glycerin (99%) was purchased from Sigma-Aldrich. Ultrapure deionized (DI) water was obtained by Spectra-Teknik water purification system.

SPR Setup
One phase-sensitive SPR setup ( Fig. S1a) was employed for the experimental investigation of phase. A 4mW 632.8nm He-Ne laser with a beam spot size of 1mm was used to excite the surface electromagnetic waves. P-polarized light and s-polarized light were splitted by a polarized beam splitter. In the beam of p-polarized light, the configuration was based on the well-known Krestchmann configuration including a right-angle BK7 coupling prism with 40 mm length of legs which was immobilized on a rotation stage. The thin Au film graphene stacks with a flow chamber was attached to the face of prism by optical matching oil (Cargille Labs). Sample solutions could be injected into the flow chamber by a syringe pump to interact with the sensing film. The incident p-polarized light passed through the sensing film and then reflected out. In the beam of s-polarized light, a galvo-mirror (Thorlabs, GVS001) was driven by sine wave signals oscillating at 86 Hz to generate optical path difference and obtain complete sine waves for the interference light of p-polarized light and s-polarized light. The intensity of final interfered light was detected by a photo detector and collected by a data acquisition card (NI PCI-6115) using a Labview program. A point-wise arcsine algorithm was used to extract the phase difference between p-polarized light and s-polarized light.
Another angle-sensitive SPR setup (Fig. S1b) was used for the measurement of reflectivity of the thin Au film graphene stacks. The incident light was set as the p-polarized light by a polarizer.
The light beam was focused on the homogeneous part of the graphene stacks on the Au film attached to the hypotenuse face of the prism and then reflected out. The incident angle could be changed by rotating the rotation stage. The intensity of the reflected light was changing with the incident angle, which could be monitored by a high-precision optical power meter (Newport 2832C). Then we could get the reflectance for each incident angle. Figure S1. Schematic diagram showing the setup of (a) phase-sensitive SPR setup and (b) anglesensitive SPR setup.

Result
Reflectance and differential phase were measured for both bare Au sensing film and 3-layer graphene-coated Au sensing film. In the measurement of phase sensitivity of bare Au sensing film and 3-layer graphene-coated Au sensing film, glycerin solutions (Sigma-Aldrich) with different weight ratios were injected into the flow chambers. The corresponding refractive index from pure DI water to 10% glycerin solution was measured to be from 1.332 to 1.344 by an abbe refractometer (2WAJ).
For 50nm bare Au sensing film with 2.5 nm titanium adhesion layer, experimental data of reflectance in the air with respect to different incident angle were measured using angle-sensitive SPR setup, where the blue line was fitted for eye guide (Fig. S2a). The measured resonance angle was 36.2º, which could well match the simulated data by solving the Fresnel's equation, estimated resonance angle was 36.1º, (Fig. S2b). Change in differential phase of different weight ratios of glycerin solutions was obtained by phase-sensitive SPR setup, where the blue line fitted the experimental data (Fig. S2c). The differential phase was linearly changed from DI water (1.332) to 1.5% glycerin solutions (1.3338), corresponding to a phase sensitivity of 16,944 Deg/RIU, which was in good consistent with the simulation result (Fig. S2d). Figure S2. The thickness of the bare Au sensing film is 50nm with 2.5 nm titanium adhesion layer. Variation of reflectance in the air, (a) experimental result and (b) simulation result, with respect to angle of incidence; Change in differential phase, (c) experimental result and (d) simulation result, for various weight ratios of glycerin solutions. Figure S3. Three-layer graphene was coated on the 50 nm bare Au sensing film with 2.5 nm titanium adhesion layer. Variation of reflectance in the air, (a) experimental result and (b) simulation result, with respect to angle of incidence; Change in differential phase, (c) experimental result and (d) simulation result, for various weight ratios of glycerin solutions.
For 3-layer graphene-coated Au sensing film, the reflectance in the air was measured as well and the obtained resonance angle was 36.5º ( Fig. S3a) in good agreement with the simulation result (estimated resonance angle 36.3º) (Fig. S3b). A phase sensitivity of 22,489 Deg/RIU could be calculated corresponding to the linearly phase change from DI water to 1.5% glycerin solutions ( Fig. S3c), matching well with the result of simulation (Fig. S3d). Both the reflectance and differential phase change were in good consistent with the simulation result for bare Au sensing film and 3-layer graphene-coated Au sensing film. The phase sensitivity of 3-layer graphene-coated Au sensing film was 1.33 times higher than the one of bare Au sensing film, which could validate that 3-layer graphene could enhance the performance of whole sensing structure.

Angular interrogation
A silicon nanosheet and 2D MX 2 heterostructure based surface plasmon resonance biosensor configuration is proposed. The four enhanced models, namely, silicon-WS 2 , silicon-WSe 2 , silicon-MoS 2 and silicon-MoSe 2 were analyzed individually through the transfer matrix method.
In order to study the angular sensitivity, the reflectivity, resonance angle shift as well as the FWHM of all the four silicon-MX 2 models were studied under five excitation wavelengths ( 600 nm, 633 nm, 660 nm, 785 nm and 1024 nm), as shown in Table S1-20. Therefore, the optimized parameters of each of the silicon-MX 2 enhanced models were obtained. Here, we plot the SPR curves before (the blue line) and after (the red line) the adsorption of the ssDNA with a refractive index change at 0.005, as shown in Fig. S4-S7. Specifically, in silicon-WS 2 enhanced model, with the optimized parameters of 35 nm gold, 7 nm silicon nanosheet and monolayer WS 2 and 600 nm excitation wavelength, the resonance angle (SPR angle) shift can reach to a maximum of 0.7786º, as shown in Fig. S4. In silicon-MoS 2 and silicon-MoSe 2 enhanced models, the optimized parameters were the same: 40 nm thick gold thin film, 7 nm silicon and monolayer WS 2 /MoSe 2 at 633 nm excitation wavelength. The resonance angle shift were 0.6586º and 0.6854º respectively, see Fig. S5 and Fig. S6. At 633 nm excitation wavelength, the optimized thickness parameters of the silicon-WSe 2 enhanced model were 40 nm gold, 7 nm silicon and bilayer WSe 2 , the resonance angle change reached to 0.7070º, as shown in Fig. S7. In the theoretical analysis, we considered the refractive index of 2D MX 2 materials is independent on the thickness in nanoscale range. Our theoretical studies were based on the experimental conditions and parameters. The multi-layer 2D MX 2 were fabricated through stacking each single 2D layer on the sensing substrate. Thus, the refractive index of each layer of the N-layered MX 2 thin film was consistent in the theoretical analysis. Figure S4. The reflectivity as a function of the incident angle before (blue line) and after (red line) the adsorption of the target analyte at 600 nm excitation wavelength under the optimized parameters of 35 nm gold thin film, 7 nm silicon nanosheet and monolayer WS 2 . Figure S5. The reflectivity as a function of the incident angle before (blue line) and after (red line) the adsorption of the target analyte at 633 nm excitation wavelength under the optimized parameters of 40 nm gold thin film, 7 nm silicon nanosheet and monolayer MoS 2 . Figure S6. The reflectivity as a function of the incident angle before (blue line) and after (red line) the adsorption of the target analyte at 633 nm excitation wavelength under the optimized parameters of 40 nm gold thin film, 7 nm silicon nanosheet and monolayer MoSe 2 . Figure S7. The reflectivity as a function of the incident angle before (blue line) and after (red line) the adsorption of the target analyte at 633 nm excitation wavelength under the optimized parameters of 40 nm gold thin film, 7 nm silicon nanosheet and bilayer WSe 2 .

Table S1
The optimized values of gold thin film, silicon nanosheet thickness and the number of WS 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 600nm excitation wavelength.

Table S2
The optimized values of gold thin film, silicon nanosheet thickness and the number of WS 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 633nm excitation wavelength.

Table S3
The optimized values of gold thin film, silicon nanosheet thickness and the number of WS 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 660nm excitation wavelength.

Table S4
The optimized values of gold thin film, silicon nanosheet thickness and the number of WS 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 785nm excitation wavelength.

Table S8
The optimized values of gold thin film, silicon nanosheet thickness and the number of MoSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 660nm excitation wavelength.

Table S10
The optimized values of gold thin film, silicon nanosheet thickness and the number of MoSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 1024nm excitation wavelength.

Table S12
The optimized values of gold thin film, silicon nanosheet thickness and the number of MoS 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 633nm excitation wavelength.

Table S14
The optimized values of gold thin film, silicon nanosheet thickness and the number of MoS 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 785nm excitation wavelength.

Table S16
The optimized values of gold thin film, silicon nanosheet thickness and the number of WSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 600nm excitation wavelength.

Table S17
The optimized values of gold thin film, silicon nanosheet thickness and the number of WSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 633nm excitation wavelength.

Table S18
The optimized values of gold thin film, silicon nanosheet thickness and the number of WSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 660nm excitation wavelength.

Table S19
The optimized values of gold thin film, silicon nanosheet thickness and the number of WSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 785nm excitation wavelength.

Table S20
The optimized values of gold thin film, silicon nanosheet thickness and the number of WSe 2 layers with corresponding change in resonance angle and FWHM in SPR curve for 1024nm excitation wavelength.

Phase interrogation
In order to further demonstrate the validity of our N-layer 2D models, we also studied the SPR phase signal changes with different refractive-index changes at the sensing surfaces. The thickness of the metallic silicon-MX 2 thin film are optimized under three excitation wavelengths, namely 633 nm, 660 nm, 1540 nm for the phase interrogation scheme. The phase sensitivity can reach to one order of magnitude higher than that of the conventional configuration, as shown in    The phase sensitivity as a function of the refractive index change of the biomolecular sample that ranges from 1×10 -5 RIU to 1×10 -4 RIU.

Table S21
The change in differential phase of various refractive index change of the biomolecular sample with optimized thickness of 44 nm gold and 5 nm silicon at 633 nm excitation wavelength in silicon-WSe 2 model.

Table S22
The phase sensitivity of various refractive index change of the biomolecular sample with optimized thickness of 44 nm gold and 5 nm silicon at 633 nm excitation wavelength in silicon-WSe 2 model.

Table S23
The change in differential phase of various refractive index change of the biomolecular sample with optimized thickness of 28 nm gold, 7 nm silicon at 660 nm excitation wavelength in silicon-MoSe 2 model.

Table S24
The phase sensitivity of various refractive index change of the biomolecular sample with optimized thickness of 28 nm gold and 7 nm silicon at 660 nm excitation wavelength in silicon-MoSe 2 model.

Table S25
The change in differential phase of various refractive index change of the biomolecular sample with optimized thickness of 31 nm gold and 4 nm silicon at 1540 nm excitation wavelength in silicon-WS 2 model.

Table S26
The phase sensitivity of various refractive index change of the biomolecular sample with optimized thickness of 31 nm gold and 4 nm silicon at 1540 nm excitation wavelength in silicon-WS 2 model.