Design of an integrated optics for transglutaminase conformational change

2.1 Architecture of SPR-based waveguide

SPR-based waveguide is the heart of the integrated optics. We consider an SPR-based optical waveguide consisting of two identical Ag strips and an SPR multilayer section located on the SiO₂ substrate. The Si core of the SPR section is coated with two metal layers (Cr and Ag layers), above which lies an intermediate layer Ta₂O₅. As a result, the fundamental TM polarized SP mode can be supported effectively on the surface of the multilayer. Two identical Ag strips are placed on opposite sides of the SPR section with a distance of nanometers; thus, the TE mode is confined to the lateral nanoslots. In the simulation, thicknesses of silica substrate, Si region, metallic layers, and the intermediate layer are taken as h_SiO2=3 μm, h_Si=550 nm, h_Cr=h_Ag=h_Ta2O5=5 nm. The width of the Si core and the thickness of the Ag strip are set as w_Si=750 nm and h_Ag=565 nm. The lateral nanoslot formed between the Si rib and Ag strip is taken as w_slot=50 nm. The wavelength-dependent refractive indices of Si [13] and SiO₂ [14] regions are achieved by a previous literature. The Cr layer is modeled using experimental data from Johnson and Christy [15] and that of the Ta₂O₅ intermediate is gained using literature data [16]. The Drude-Sommerfeld mode is adopted to characterize the dielectric function of Ag [17]. The operation wavelength spans 1.280 μm≤λ≤1.315 μm. (central wavelength 1.3 μm). The wavelength selection [18] is based on a tradeoff between the absorption loss caused by aqueous solution and the broadband light sources. Water solution optical absorption is smaller at wavelength of 1.3 μm compared to the 1.55 μm wavelength commonly used for operation. The cross-section view of the proposed structure is shown in Figure 1A. For such a configuration, the SPR-based waveguide is found to strongly support both TE and TM polarized modes, which are presented in the field profile (Figure 1B) obtained via simulation based on FEM.

Figure 1:

The SPR-based waveguide consisting of two identical silver strips and between which an SPR multilayer section is located. Highly confined hybrid modes with dual polarizations are supported in this structure. Thus, the refractive index of biolayer and its thickness can be measured individually and without ambiguity. The greatest strength of such technique lies in monitoring protein conformational change. (A) Cross-section view of SPR-based waveguide and (B) field profile of SPR-based waveguide for dual polarization.

2.2 Theory of SPR-based waveguide

For the SPR multilayer section, incident light is fed to the Si core and excited a TM polarized SPW on the surface of the multilayer when the phase velocities of the optical mode matches that of the SPW. Because the refractive index of Ag (the real part) is significantly different from the index of dielectric waveguide (Si core), layer Cr is introduced to greatly promote the phase match between the guided mode and SP mode. As a result, the electromagnetic field is enhanced at the surface of the multilayer and the sensitivity is improved as well. Meanwhile, a TE polarized SPW is excited in the lateral nanoslots between the Ag strip and Si rib. Two measurements simultaneously provided by dual polarization optogeometrical properties (density and thickness) of the protein layers can be determined without ambiguity. As a result, a dip in the output spectrum is observed at specific wavelength for dual polarization and the coupling of electromagnetic energy between a guided mode and an SP mode occurs. The amplitude of resonance dip strongly relies on the refractive index of the sensed medium. Any changes of the sensed medium will alter the resonance condition, which results in a shift of the resonance dip. Evidently, nanoslots are able to store electromagnetic energy resulting in subwavelength optical guiding with lower propagation loss. The dielectric discontinuity at the Si core-solution interface produces a polarization charge [19] that interacts with the plasma oscillations of the metal-solution interface. The gap region can be equated with an effective optical capacitance [19] resulting in strong field confinement and longer propagation distance. We take intensity to evaluate refractive index changes. Power change detection is more sensitive to a small change in the refractive index of the ambient medium [12]. Such detection can be achieved simply using an optic-electric detector, thus significantly cutting the experiment cost. With the introduction of MZI, not only the common mode noise and drift in the detection can be reduced greatly but also the control trial can be done.

2.3 Modal investigation

In this section, we will discuss the influence of the intermediate layer on the sensitivity of SPR-based waveguide. Cr film is capable of promoting the conversion of the guided mode in the Si rib to the SP mode on the outer surface of the multilayer and supporting the fundamental SP mode. The field profile is shown in Figure 2. The profile of the electromagnetic field, especially for the TM mode, is smooth and continuous, thereby supporting the fundamental mode on the Cr layer surface. However, the one without Cr film supporting multimode might lead to spurious measurement. The sensitivities of the biosensor are improved by introducing a Ta₂O₅ intermediate layer for dual polarizations and the results are shown in Figure 3. The amplitude of intensity is increased with regard to the structure with Ta₂O₅ film; thus, the sensitivity is expected to be enhanced. The increased values of intensity (dP) are 0.02 and 0.05 for dual polarization. Moreover, we consider the influence of multilayer thickness on the effective mode index n_eff and propagation loss of SPR-based waveguide for both TE and TM modes. The thickness of the metallic layer plays a key role in the excitement of SPW [20]. If the metal is too thin, the SPW will be greatly damped due to radiation damping into the Si rib. The thicker metallic film might prevent the excitation of SPW because of the adsorption in the metal. To simplify the research, thickness per layer is set as an equal value. The data are plotted in Figure 4. Effective mode indices increase with thinner multilayer. The propagation losses achieve the minimum decay value when the multilayer thickness decreases. Thus, the optimum structure of the proposed SPR waveguide with thickness of 5 nm per film (Ag, Cr, and Ta₂O₅) is adopted and subsequently implemented on the arms of MZI.

Figure 2:

Field profile of SPR-based waveguide with respect to TM mode: (A) with Cr layer and (B) without Cr layer.

Figure 3:

Output spectrum of Tris buffer calibration for SPR-based waveguide with or without Ta₂O₅ layer: (A) TE mode and (B) TM mode.

Figure 4:

The effective mode index n_eff dependent on the thickness of multilayer. (A) for TE mode and (B) for TM mode. The propagation loss dependent on the thickness of multilayer. (C) for TE mode and (D) for TM mode.

3.1 Architecture of integrated optics

The optimized SPR-based waveguide is incorporated onto the arms of MZI afterward and the model is sketched in Figure 5. The lengh of the device is 80 μm (L=80 μm). The sensing and reference arms with identical length L_m=20 μm are separated by the distance D=15 μm. The Y-splitter is composed of S-shaped bent wires with a bending radius of 23 μm (r₀=23 μm). The length (l) and width (w) of SPR-based waveguide are 1 and 7 μm, respectively. Cytop (http://www.bellexinternational.com/products/cytop/) cladding, which aims to avoid external interference, is employed to cover the designed structure with window opened for injecting solution.

Figure 5:

3D schematic structure of integrated MZI with SPR-based waveguide. Inset figure amplifies the interaction area.

3.2 Protein analysis based on integrated optics

Generally, the tTG-Ca²⁺ binding event can be simplified to four distinct steps: (1) Tris buffer flowing over the waveguide surface for initial calibration; (2) BS₃, an amine-amine linker immobilized on the waveguide surface; (3) capture of protein tTG onto the surface via the BS₃ layer; and (4) introduction of Ca²⁺ to bind with protein tTG. To prevent unspecific interaction and external effects, the control surface is designed without adding protein solution. During the biological event, the thickness a and refractive index n_l of the protein layer with known numerical value were obtained from typical literature data [21] and collected in Table 1. Comparing the covering solution exposed to the two arms, one can find that Tris buffer instead of tTG protein is flowed over the reference arm in step 3. In this way, the parallel control trial is realized to prevent unspecific binding from biological material and error from geometrical difference. In the biological event, bulk solution remains constant with a refractive index of 1.3337. Resonance phenomena appear but without marked intensity difference in the output spectrum with respect to the pure SPR-based waveguide. The reason probably comes from that the bulk refractive index remains unchanged and the bulk solution has dominant influence on the effective mode index of the waveguide. Moreover, the difference of optogeometrical properties (refractive index and thickness) between the tTG layer and layer tTG bound with Ca²⁺ is too tiny (δ_a=0.5 nm) to be identified by the pure waveguide. To maximize the sensitivity, we incorporated the proposed waveguide onto MZI. As expected, resonance dips with different intensity at specific wavelength are presented in the output spectrum. The magnitude of the resonance dip in the received power is dependent on the conversion of SPW. The MZI improves the efficiency of the mode conversion and thus functions as an amplifier with regard to the sensitivity. The results are shown in Figure 6. It is obvious that resonance dips with different amplitude representing each step in the molecular interaction.

Figure 6:

Output spectrum of protein tTG interaction with Ca²⁺ for the integrated MZI biosensor: (A) TE mode and (B) TM mode.

3.3 Characterization of protein conformation

To extract biolayer thickness and refractive index, we investigated the relation between power variations and properties of the protein layer (refractive index and thickness) at resonance dips for dual polarizations and defined the expression as follows:

where ΔP_i (i=TE, TM) is the normalized power variation at resonance dip (herein, λ=1.305 μm. for TE mode and 1.303 μm for TM mode), ∂P_i/∂n_l (i=TE, TM) are the power changes caused by refractive index difference, and ∂P_i/∂a (i=TE, TM) are the power variations due to adlayer thickness changes. To investigate the effect of the biolayer refractive index on the intensity, we detect different values of intensity caused by the changes of n_l at various resonance wavelengths. Similarly, the influence of biolayer thickness on the intensity can be obtained. The terms ∂P_i/∂n_l and ∂P_i/∂a with fitting formulas are sketched in Figure 7. It is noteworthy that the terms Δn_l and Δa are the shifts in adlayer properties for each step compared to Tris buffer calibration. The calculated values by the mean of Eq. (1) share the similar results compared to experimental results [21]. The calculated adlayer index of protein tTG is 1.418 (1.4184 experimental data [21]) and the tTG layer thickness is 5.25 nm (5.261 nm experimental data [21]). After Ca²⁺ binding, the index and thickness of protein tTG are 1.426 and 4.76 nm, respectively, compared to experimental results (1.428 for biolayer index and 4.761 nm for biolayer thickness). Given the values of thickness and refractive index of biolayer, it is able to calculate the surface mass density through De Feijter’s formula [22]: G=a(n_l−n_s)/(dn/dc), where dn/dc is the refractive index increment, which in the case of protein is 0.186 cm³ g⁻¹. The densities of adlayer contributed by tTG protein and tTG-Ca²⁺ compound are 2.38 and 2.36 cm³ g⁻¹, respectively. Integrated with adlayer thickness and density, we can conclude that a thin dense layer formed as Ca²⁺ binds to tTG protein. This phenomenon can be explained as that tTG protein undergoes conformational change on binding to Ca²⁺.

Figure 7:

Partial derivative of the power versus adlayer index at λ=1.305 μm for TE mode (A) and 1.303 μm for TM mode (B) and partial derivative of the power versus adlayer thickness at λ=1.305 μm for TE mode (C) and 1.303 μm for TM mode (D).