**3. IMI structures for sensing performance**

The evanescent depth of SPR plays an essential role in determining the sensing performance of plasmonic devices [26]. Above all, an IMI structure can control the evanescent field by changing the thickness of the metal layer [27]. This IMI structure is applied to slightly asymmetric dielectric media [28]. The slightly asymmetric IMI structure can provide benefits for aqueous sensing solutions. To date, no attention has been paid to the SPR property based on a ZnO-related IMI structure. The employment of IMI geometry to SPR excitations on ZnO: Ga films remains unclear. In this section, we introduce asymmetric IMI structures consisting of water, ZnO: Ga, and a cytop polymer. In particular, water-based IMI structures are one of the interesting sensing platforms in SPR applications. The benefits of SPR are confirmed from the bulk sensitivity based on index changes. We discuss the sensing performance of ZnO-SPR based on the proof-of-concept of an IMI structure.

The fabrication of IMI samples was performed as follows. The refractive index of a cytop polymer is kept close to water, which can excite SPRs using IMI structures with ZnO: Ga films. A Cytop polymer (perfluropolymer) film (1.8 mm-thickness) was deposited on BK-7 glass substrates using a polymer content of 9% in fluoride solvent by a spin coating method (2500 rpm for 50s). The coated polymer films were annealed at 220°C for 2 h in the air to evaporate the solvent. ZnO: Ga films were fabricated on polymer/glass substrates using PLD at room temperature. ArF excimer laser pulses (193 nm, 5 Hz, and 1 J/cm2 ) were focused on ZnO: Ga targets in an O2 flow of 10−4 Pa [29].

The IMI sample SPR reflectance spectra are shown in **Figure 4(a)**. The peak positions gradually moved within the 4000 to 6000 cm−1 range with an increasing incident angle from 60.5<sup>o</sup> to 64o in 2o increments [29]. The SPR spectra were observed even at a film thickness of 22 nm. The SPR peak dependence on the incident angle of light for the IMI sample was higher than that for the single ZnO film. In addition, the IMI sample showed narrower spectral features than those of the single ZnO films. These behaviors were also confirmed by theoretical SPR spectra (**Figure 4(b)**) [29].

The SP mode of an IMI structure is separated into two types of plasmon branches comprising the short-range and long-range SP modes. The dispersion curves of both SP modes can be described using the Maxwell relations in a planar structure [30]:

$$
\left(\frac{\varepsilon\_2(\nu)}{\varepsilon\_1(\nu)} + \frac{\gamma\_z(\nu)}{\gamma\_1(\nu)}\right) \cdot \left(\frac{\varepsilon\_0(\nu)}{\varepsilon\_1(\nu)} + \frac{\gamma\_o(\nu)}{\gamma\_1(\nu)}\right) e^{\varepsilon\_{\gamma\_1(\nu)\varepsilon}} = \left(\frac{\varepsilon\_2(\nu)}{\varepsilon\_1(\nu)} \cdot \frac{\gamma\_z(\nu)}{\gamma\_1(\nu)}\right) \cdot \left(\frac{\varepsilon\_o(\nu)}{\varepsilon\_1(\nu)} \cdot \frac{\gamma\_o(\nu)}{\gamma\_1(\nu)}\right) \tag{2}
$$

$$\gamma\_i^2 = k^2 \text{ -b}\_i \left(\frac{v}{c}\right)^2 \text{ ( $\mathbf{i} = \mathbf{o}$ ,  $\mathbf{z}$ ,  $\mathbf{z}$ )}\tag{3}$$

where εi(*v*) (*i* = 0, 1, 2) represents the dielectric function in each layer, i.e., water, ZnO: Ga, and cytop polymer layers, respectively. **Figure 4(c)** shows a dispersion curve of the IMI sample, revealing two plasmon branches of short-range and longrange SP modes. The experimental data were similar for the long-range SP mode, which was due to the phase-matching of wave vectors of SPRs at the water–ZnO: Ga– cytop polymer interfaces. An SP wave of the long-range mode is expected to show a longer propagation distance than that of the single-mode. The following equation can

#### **Figure 4.**

*(a) Experimental and (b) calculated SPR reflectance spectra of the ZnO IMI sample with a film thickness (t) of 22 nm. Water was selected as the dielectric medium. (c) Experimental and calculated dispersion curves of the ZnO IMI sample (t = 22 nm). A dot line indicates light line in water medium. Long-range and short-range SP modes are represented in the figure. (d) a depth-dependent mean square evanescent field at calculated at 4500 cm−1 for the IMI sample (***Figures 4(d)** *and* **5(c)** *of [29]). Copyright by the American Institute of Physics.*

express the propagation distance (*L*prop): *L*prop = 1/2 × Im[*k*x] [31], which represents the length from the launch point where the evanescent field power decays by a factor of 1/*e*. The *L*prop value of the IMI sample was approximately 10 μm, which was longer than that of the single film (*L*prop = 3 μm). The difference in SPR response between the IMI sample and single films was related to the propagation distance.

**Figure 4(d)** shows a depth profile of a mean-square evanescent field at 4500 cm−1 of the *p*-polarized component [29]. The value of <*E*zz2 > was estimated to be 18.0, which was higher than that of the single film. Besides, the δw value in the water medium was in the vicinity of 1 μm. The penetration depth clearly expanded when using the IMI sample, relating to the long propagation distance and the high angle dependence of SPR reflectance spectra.

**Figure 5(a)** shows the SPR reflectance spectra (θ = 58.25o ) of the IMI sample, revealing a large Δ*v* of 82 cm−1 at 1 g/dL, which was higher than that of the single film [29]. Additionally, the sensitivity was measured at 4500 cm−1, which resulted in a large *S*exp value of 30,530 cm−1/RIU. This value was close to the theoretical estimation (*S*cal = 33,000 cm−1/RIU). The enhanced sensitivity was attributed to the evanescent field depth and longer propagation distance of the SPR, which was compared with the single film (**Figure 5(b)**) [29].

*Biological Sensing Using Infrared SPR Devices Based on ZnO DOI: http://dx.doi.org/10.5772/intechopen.104562*

#### **Figure 5.**

*(a) SPR reflectance spectra of the IMI sample measured with varying glucose content in water from 0, 1, and 5 g/dL. (b) Correlation between peak shift (Δv) and glucose content in water for the IMI sample. The change in refractive index (n) of a mixed solution consisting of glucose and water is also described in the upper horizontal axis. Black line and dots indicate calculated and experiment data (***Figure 4(b)** *and* **(c)** *of [29]). Copyright by the American Institute of Physics.*

In this section, we investigated the SPR properties and sensing performance of the IMI sample. The SPR spectra of the IMI sample displayed narrower features than those of the single film. This result provided the extended evanescent field depth and the long propagation distance. Consequently, the sensitivity of the IMI sample was markedly enhanced compared to that of the single film.

### **4. Dielectric-assisted ZnO-SPR**

In Sections 2 and 3, we introduced the SPR properties of the single film and IMI sample. The single film showed broad SPR resonances and weak evanescent fields. The use of an IMI structure increased the evanescent field because of the long-range SP mode. However, the IMI sample produced a large sensing volume on the sample surface due to the long penetration depth, where it is not easy to detect small changes in the refractive index near the sample surfaces. Therefore, there is a need to find new strategies for circumventing the structural limitations of ZnO-SPRs in terms of penetration depth and evanescent field. Consequently, we propose a new structural concept based on dielectric-assisted SPRs to overcome some difficulties associated with ZnO-SPRs for real-time monitoring of biological interactions.

When sufficiently thick dielectric layers are placed on Ag and Au-metal layer surfaces, waveguide modes are supported in addition to the conventional SP mode [32, 33]. However, the sensing performance of a waveguide is comparable to that of a classical SPR. However, the Ag and Au-based SPR devices with thin dielectric layers (e.g., Al2O3 and SiO2) show SP waves along the thin dielectric layer surfaces with a thickness insufficient to support a waveguide mode. The SP waves of dielectricassisted SPRs sufficiently penetrate the analyte region, unlike the waveguide mode. Therefore, the introduction of thin dielectric layers on top of Ag and Au metallic films has been reported to improve the sensing performance in the visible range [34].

In this section, we report on the capping of thin dielectric layers to a ZnO-SPR sensing platform to enhance the detection sensitivity in the IR range. Here, we define dielectric-assisted ZnO-SPR devices as "hybrid samples". The insertion of dielectric layers changes the *E*-field distribution and penetration depth, providing enhanced detection sensitivity. The improvement in sensing performance is discussed from the viewpoint of penetration depth and *E-*field of the SP waves. Finally, we introduce the sensing capabilities of the hybrid samples by measuring the biological interactions between biotin and streptavidin.

The fabrication of hybrid samples was conducted as follows. Cytop polymer films (2.2 μm thickness) were deposited on BK-7 glass substrates. Metallic ZnO: Ga films with a thickness of 22 nm were fabricated on the polymer-coated substrates using a PLD method at room temperature (RT). Fabrication conditions of the polymer and ZnO: Ga films were the same as those of the IMI sample. In this study, we selected Ga2O3 as a dielectric layer. 190 m-thick Ga2O3 dielectric layers were deposited over ZnO: Ga layers on the polymer/ glass substrates at RT in an O2 flow of 10−4 Pa using the PLD method [35].

**Figure 6(a)** shows a hybrid sample's experimental SPR reflectance spectra with a Ga2O3 thickness of 190 nm [35]. The peak positions moved within the 4000–5500 cm−1

#### **Figure 6.**

*(a) Experimental and (b) calculated SPR reflectance spectra of the hybrid sample with a Ga2O3 thickness of 190 nm. Water was selected as the dielectric medium. (c) Experimental and calculated dispersion curves of the IMI sample and hybrid sample. Black open and closed circles indicate experimental data. Dotted and straight lines represent calculation data. (d) a depth-dependent mean square evanescent field at calculated at 4500 cm−1 for the hybrid sample (***Figures 4(c)***,* **6(d)** *and* **(h)** *of [35]). Copyright by the American Chemical Society.*

#### *Biological Sensing Using Infrared SPR Devices Based on ZnO DOI: http://dx.doi.org/10.5772/intechopen.104562*

range as the incident angle increased in 2.5o increments. The dependence of the SPR peak on the incident angle of light for the hybrid sample was higher than that for the IMI sample. The hybrid sample exhibited narrower spectral features than those of the IMI sample. These SPR behaviors were further reproduced by theoretical SPR spectra (**Figure 6(b)**) [35]. The difference in SPR response between the hybrid and IMI samples is attributed to the propagation length. An experimental *L*prop of approximately 14 mm was obtained for the hybrid sample, which was greater than that of the IMI sample (*L*prop = 10 μm). **Figure 6(c)** shows the dispersion curves of the hybrid and IMI samples [35]. The dispersion curve of the IMI sample was close to the light line in free space due to the long-range SP mode. However, the dispersion curve of the hybrid sample shifted to higher SP wave vectors by insertion of a Ga2O3 layer on top of a ZnO: Ga layer surface, leading to changes in the optical properties of the SPR reflectance spectra. Besides, the use of Ga2O3 dielectric layer was effective in reducing the penetration depth in the water medium. **Figure 6(d)** shows a depth profile of a mean-square evanescent field at 4500 cm−1 of the *p*-polarized component [35]. The value of <*E*zz2 > was increased up to 34.5, which was higher than that of the IMI sample. Additionally, the *δ*w value in the water medium was suppressed to 400 nm. The reduced penetration depth was related to an increase in SP wave vector by inserting the dielectric layer to the IMI sample.

**Figure 7(a)** shows the SPR reflectance spectra of the hybrid sample [35]. The dip peaks shifted to lower wavenumbers with increasing glucose concentration. A wavenumber shift of 19 cm−1 at 1 g/dL was observed from the SPR reflectance spectra taken at *θ* = 63<sup>o</sup> . The *S*exp and *S*cal of the hybrid sample were 18,700 and 19,200 cm−1/ RIU, respectively. These values were smaller than those of the IMI sample [35]. (**Figure 7(b)**). This was due to the reduced penetration depth of the hybrid sample. Furthermore, there is a need to evaluate the sensitivity (*S*) and spectral linewidth (FWHM: full-width half-maximum) when considering the performance of a biosensor. The figure-of-merit (FoM) is defined by the following relation:

#### **Figure 7.**

*(a) SPR reflectance spectra of the hybrid sample measured with varying glucose content in water (0, 1, 1, 2 and 4 g/dL). (b) Correlation between peak shift (Δv) and glucose content in water for the hybrid sample. The change in refractive index (n) of a mixed solution consisting of glucose and water is also described in the upper horizontal axis. Black line and dots indicate calculated and experiment data (***Figure 8** *of [35]). Copyright by the American Chemical Society.*

$$Fo\mathbf{M} = \mathbf{S} / \text{FWHM} \tag{4}$$

This relation is used to quantify the general performance of a biosensor [36]. This normalization allows for comparison with other sensing platforms. The hybrid sample provided a higher FoM value (51.0 RIU−1) than did the IMI sample (36.1 RIU−1). The enhanced FoM was attributed to the narrowing of the spectral linewidth resulting from the insertion of the Ga2O3 layer.

### **5. Real-time monitoring of biological interactions**

We evaluated the SPR sensing performance of the hybrid sample using the biological interactions between biotin-PEG-DPPE and streptavidin. PEG and DPPE indicate poly(ethylene glycol) and 1, 2 dipalmitoyl-*sn*-glycerol-3-phosphatidylethanolamine, respectively. The high binding affinity and irreversible binding of the molecular pair of biotin-streptavidin is a powerful tool to measure changes in the SPR reflectance, which has medical applications such as antigen–antibody reactions and allergic reactions. We conducted surface modifications before the biological experiments using self-assembled monolayer (SAM) formation.

The Ga2O3 layer surface of the hybrid sample was chemically modified using a SAM of *n*-octadecylphosphonic acid [C18H37PO(OH)2:ODPA] to form a CH3 terminated SAM (CH3-SAM). This CH3-SAM is commonly used to obtain hydrophobic surfaces because of the strong hydrogen bonding acid–base character of the –PO(OH)2 group [37]. The hybrid samples were immersed in ODPA (5 mM in ethanol) at RT for 48 h after O2 plasma irradiation. The surface states and chemical composition of the SAM-coated sample were investigated using X-ray photoemission spectroscopy (XPS). **Figure 8(a)** shows the typical Ga(3d), P(2p), C(1 s), and O(1 s)

#### **Figure 8.**

*(a) XPS core-level spectra of Ga(3d), P(2p), C(1 s) and O(1 s)-related peaks following different surface modifications. The XPS spectra were taken after surface treatments of ODPA in ethanol, ethanol only and O2 plasma. (b) Ga(3d)- and C(1 s)-related peak intensities as a function of OPDA treatment time. (c) P(2p) related peak and intensity ratio of C(1 s) and P(2p) as a function of OPDA treatment time (Figure 10 of [35]). Copyright by the American Chemical Society.*

#### *Biological Sensing Using Infrared SPR Devices Based on ZnO DOI: http://dx.doi.org/10.5772/intechopen.104562*

peaks for different surface treatments [35]. The P(2*p*) peak of the ODPA-coated sample was observed at 136 eV, which was not obtained for the ethanol or O2 plasmatreated samples. This result clarified the formation of CH3-SAM on the sample surface. Immersion of the hybrid samples in a toluene solution of ODPA reduced the peak intensities related to Ga(3*d*) and Ga(3*p*) assigned to Ga–O within 20 min owing to the adsorption of ODPA (**Figure 8(b)**) [35]. The ratio of the surface carbon concentration to the phosphorous concentration was 20 ± 1, which was similar to 18, the value expected from the molecular formula of ODPA (**Figure 8(c)**). These XPS results revealed the formation of ODPA on the hybrid samples [35].

Biological interactions were conducted as follows [35]. The hybrid and IMI samples were washed with ethanol several times and then dried using nitrogen gas. Both samples were first exposed to a phosphate-buffered saline (PBS) solution with pH = 7.4 [PBS (1) process], until a stable SPR signal was obtained. Next, both samples were exposed to a solution of biotin-PEG-DPPE (100 μg/mL in PBS) for 35 min, followed by washing with PBS for 5 min [PBS (2) process]. After the sample surfaces were treated with biotin-PEG-DPPE, bovine serum albumin (BSA) was introduced to confirm non-specific protein adsorption to the sample surfaces [BSA process], followed by another PBS wash [PBS (3) process]. PBS solutions with different streptavidin concentrations (0–10 μg/dl in PBS) were then introduced [Streptavidin process].

#### **Figure 9.**

*Experimental SPR reflectance spectra of the hybrid sample (a) and the IMI samples (c) in each process in the biological reaction between biotin-PEG-DPPE and streptavidin of a concentration of 5 μg/mL. Detection of streptavidin at different concentrations using biotin-PEG-DPPE of the hybrid (b) and IMI samples (d). Differential reflectance (DR) was monitored at 4250 and 5250 cm−1 for the hybrid and IMI samples, respectively, (*Figure 13 *of [35]). Copyright by the American Chemical Society.*

The streptavidin concentrations were used to monitor the specific biological bonding of biotin and avidin [38].

**Figure 9(a)** shows the SPR reflectance spectra measured after each process [35]. The SPR peak position shifted to small wavenumbers after the surfaces were treated with biotin-PEG-DPPE and streptavidin. The SPR peak shift (Δ*v*) was observed to be 14 cm−1. The SPR peak shift was evaluated by monitoring the reflectance difference (Δ*R*) at 4250 cm−1 induced by biotin-streptavidin binding (**Figure 9(b)**) [35]. The Δ*R* values were changed slightly through surface treatment with biotin-PEG-DPPE. The streptavidin binding to biotin-PEG-DPPE remarkably depended on the streptavidin concentration. However, the IMI sample did not exhibit a significant change in Δ*R* during the biotin-streptavidin binding (**Figure 9(c)** and **(d)**) [35]. The time-dependent Δ*R* of the hybrid and IMI samples was attributed to the biotin-streptavidin binding. The use of the Ga2O3 layer on top of the ZnO: Ga layer surface successfully allowed for monitoring of the biological interactions. This resulted from the higher *E*-field and shorter field depth of the hybrid sample compared with those of the IMI sample.

The present detection limit of the hybrid samples was approximately 1 μg/ml (15 nM) due to the background noises. Industrial applications are expected to detect bio-molecule concentration at the ng/dl (~ pM) level. The spectral features of the present ZnO-SPRs were markedly influenced by the interface roughness, relating to the sensing activity for biomolecular detection. Recently, high-sensitive SPR detection at the pM levels of small biomolecules (e.g. biotin) using nanophotonic devices such as cavity structures and metamaterials with layer structures have been reported [39–41]. Recently, biological sensing based on ZnO-related SPR and metamaterials are reported by some papers [42–44]. Therefore, monitoring biological interactions at the pM level could be realized by enhancing the structural and crystalline quality of ZnO-SPR sensors. The detection sensitivity of ZnO-SPR is expected to improve the employment of new nano-plasmonic structures such as cavities and metamaterials.
