Optical surface wave-based sensors dominate optical sensing technology oriented on structures and devices enabling sensing of various physical quantities. This research field has been represented preferably by surface plasmon resonance (SPR) [1] but in recent years is accompanied by sensors utilizing Bloch surface waves (BSWs). The BSWs are related to one-dimensional photonic crystals (1DPhCs) whose symmetry in the periodic modulation of the refractive index (RI) is broken [2]. Light confinement in BSWs, which is caused by total internal reflection from the homogeneous layer and the 1DPhC characterized by a photonic bandgap, occurs near the interface between the 1DPhC and a sensed medium. The BSW propagates along the interface between a 1DPhC and an external dielectric medium, and this is in contrast with propagation of the surface plasmon wave along the interface between a thin metal film, such as gold or silver, and a dielectric medium. Moreover, BSWs can be excited by both TE and TM waves [3] at any wavelength by suitably changing the geometry and materials of the 1DPhC. In addition, if wavelength interrogation is used, BSW-based sensors enable sharper resonances than conventional SPR sensors and are an ideal candidate for various sensor applications [4–17]. The main limitation of some BSW-based sensors is that the resonances exhibit shallow dips and are characterized by lower sensitivity compared to SPR and other sensors [5–8,18,19].

In this Letter, a new sensing concept based on resonances supported by a microcavity resonator in the Kretschmann configuration is presented. Using a suitable angle of incidence and wavelength interrogation, we show for a 1DPhC employed to form a microcavity that TE BSW excitation shows up as a sharp dip within the 1DPhC bandgap. We express the sensitivity and the figure of merit for gaseous analyte and demonstrate that when the angle of incidence is decreased, both TE and TM cavity-mode resonances are resolved and they exhibit ultrahigh sensitivity and figure of merit, confirmed experimentally for the TM wave.

To measure responses of both TE and TM waves, we consider an experimental setup shown schematically in Fig. 1. It consists of a white-light source with launching optics, an input optical fiber, and a collimating lens (CL). The collimated beam passes through a polarizer oriented $45^\circ$ with respect to the plane of incidence so that TE and TM waves are generated with equal amplitudes. These waves are coupled to the first 1DPhC using a right-angle prism. The second 1DPhC is parallel placed to the first one to form a microcavity resonator whose thickness can be varied. The reflected light passes through an analyzer oriented either $0^\circ$ or $90^\circ$ with respect to the plane of incidence so that the spectral reflectances for either TM or TE waves are measured. The light is launched via a microscope objective (MO) into a read optical fiber of a spectrometer, connected via a USB to a notebook.

 

Fig. 1. Experimental setup to measure the spectral reflectance responses of a 1DPhC microcavity resonator in the Kretschmann configuration.

Download Full Size | PDF

We consider that the 1DPhC under study comprises six bilayers of TiO$_2$/SiO$_{2}$ with thicknesses of 100 nm/85 nm and a termination layer of TiO$_2$ with a thickness of 60 nm. The layers are deposited on a BK7 glass substrate, and the 1DPhC is employed in the Kretschmann configuration with a coupling prism made of BK7 glass. If the dispersion characteristics of the layers and prisms are the same as presented in a previous paper [20], the 1DPhC is characterized at the angle of incidence $\theta =50^\circ$ (above the critical angle) by a bandgap for the TE wave over a wavelength range of approximately 470–780 nm. To model the reflectance response for both TE and TM waves, a transfer matrix method [2] was used. A sharp dip within the bandgap of the reflectance spectrum is revealed for the TE wave when the cavity thickness is 5.5 µm and a cavity medium (analyte) is air ($n=1$), as demonstrated in Fig. 2. In the same figure are depicted the spectral reflectances for the RI of the analyte increased up to 1.005. This sharp dip with a full with at half maximum (FWHM) of 3.7 nm shifts toward longer wavelengths (from 700.69 nm to 701.05 nm) as the RI of analyte increases.

 

Fig. 2. Theoretical reflectance of the TE wave as a function of the wavelength for different refractive indices of the gaseous analyte in the microcavity when $\theta =50^\circ$.

Download Full Size | PDF

To ensure the response of the 1DPhC microcavity, the sensitivity $S_n$ to the RI $n$, defined as the change in the position of the dip $\delta \lambda _r$ with respect to the change in the RI $\delta n$ of gaseous analyte, $S_n=\delta \lambda _r/\delta n$, is specified. Because the dependence of the resonance wavelength on the RI of the gaseous analyte can be well approximated by a linear function, the sensitivity to the RI $S_n$ reaches approximately 72 nm/RIU. Another important parameter used for the sensing setup evaluation is the figure of merit (FOM), defined as the sensitivity divided by the FWHM of the dip, FOM$=S_n/$FWHM, and it reaches 19.5 RIU$^{-1}$.

To confirm that the sharp dip in the reflectance spectrum of the microcavity is related to the BSW excitation, the intensity distribution shown in Fig.S1, Supplement 1, clearly demonstrates that the enhanced optical field, characterized by nearly a seventy-fold enhancement of the optical field with respect to the incident field, is confined to the surface of the first 1DPhC, and its exponential tail is effectively far from the second 1DPhC. The BSW-based response is the same as with a single 1DPhC in the Kretschmann configuration, similarly as in [15].

To demonstrate the new concept, we decrease the angle of incidence $\theta$ below the critical angle and vary the thickness of the microcavity to obtain the satisfactory response. One of the acceptable responses is obtained for $\theta =40.61^\circ$ and the cavity thickness of 5.5 µm, and Fig. 3 shows the theoretical spectral reflectances for the TE wave when the RI of gaseous analyte changes in a range of 1–1.0005. First, for the same angle of incidence and a single 1DPhC, no dip is present in the bandgap in a wavelength range of approximately 510–755 nm, as illustrated in the black curve in Fig. 3. Second, due to the 1DPhC microcavity, a very sharp dip with an FWHM of 0.13 nm shifts significantly toward longer wavelengths as the RI increases. The corresponding resonance wavelength as a function of the RI is shown in Fig. 4 giving sensitivity and FOM of 52,300 nm/RIU and 402,300 RIU$^{-1}$, respectively.

 

Fig. 3. Theoretical reflectance of the TE wave as a function of the wavelength for different refractive indices of gaseous analyte in the microcavity when $\theta =40.61^\circ$. The black curve corresponds to a single 1DPhC.

Download Full Size | PDF

 

Fig. 4. Resonance wavelength as a function of the RI of gaseous analyte for TE and TM waves with solid lines as fits.

Download Full Size | PDF

These ultrahigh values are due to the cavity mode supported by the microcavity, as illustrated in Fig.S2, Supplement 1, which shows more than a 20,000-fold enhancement of the optical field with respect to the incident field. In addition, the detection limit (DL) reaches 1.9$\times 10^{-7}$ RIU when we consider that the resonance wavelength is resolved with a precision of 0.01 nm.

For the same angle of incidence $\theta =40.61^\circ$, the spectral reflectance of the TM wave is computed when RI of gaseous analyte changes in the range from 1 to 1.0005, and the corresponding dips are shown in Fig. 5. Once again, the bandgap over a wavelength range of approximately 560–720 nm is illustrated by the black curve in the figure, and the reflectance is lowered due to the decreased angle of incidence. A well-pronounced dip with an FWHM of 6.5 nm is due to the 1DPhC microcavity, and it shifts toward longer wavelengths as the RI of analyte increases. The dependence of the resonance wavelength of the cavity mode on the RI is shown in Fig. 4. The sensitivity and FOM are as high as 14,000 nm/RIU and 2154 RIU$^{-1}$, respectively.

 

Fig. 5. Theoretical reflectance of the TM wave as a function of the wavelength for different refractive indices of gaseous analyte in the microcavity when $\theta =40.61^\circ$. The black curve corresponds to a single 1DPhC.

Download Full Size | PDF

These values are lower when compared to those for cavity-mode resonances of TE waves what can be simply justified by only more than a 250-fold enhancement of the optical field with respect to the incident field, as shown in Fig.S3, Supplement 1.

The experimental confirmation of the theoretical results has been done using the setup shown in Fig. 1, which included a halogen lamp (LS-1, Ocean Optics), a linear polarizer and analyzer (LPVIS050, Thorlabs), a right-angle BK7 glass prism (PS914, Thorlabs) with index-matching fluid (Cargille, $n_D$=1.516), and a read optical fiber (M14L02, Thorlabs) of a spectrometer (USB4000, Ocean Optics). The 1DPhCs employed in a microcavity were prepared by means of vacuum electron beam evaporation using a coating system (BAK550, Balzers). The fabrication procedure is described in detail in previous papers [16,20]. The prepared 1DPhC was inspected by scanning electron microscopy with an SEM image shown in [20] (detail is shown in Fig.S4, Supplement 1). The 1DPhCs were glued to a kinematic mount (KC1-T/M, Thorlabs) and translator (SM1ZA, Thorlabs) enabling to adjust the mirror tilt and to form a microcavity with a controllable thickness. A humid air, whose RH was varied from the starting RH value (room RH of air) by a humidifier comprising a two-line peristaltic pump [15], was injected into the microcavity via the inlet in the ring surrounding it (see Fig.S5, Supplement 1).

To confirm the sensing ability of a microcavity resonator in the Kretschmann configuration, a smallest cavity thickness of approximately 7 µm was adjusted as it results from the measured normal incidence spectrum [19,20] shown in Fig.S6, Supplement 1. In the first step, the external angle of incidence [15] $\alpha _1=3.8^\circ$ ($\theta \approx 42.49^\circ$) was adjusted to excite the BSW. The BSW excitation is manifested by a narrow dip for TE ($s$-polarized) wave, whose FWHM is 6.3 nm, as shown in Fig. 6 for the reflectance ratio $R_s(\lambda )/R_p(\lambda )$. This sharp dip shifts from 748.26 nm to 749.13 nm (see Fig.S7, Supplement 1) as the RH of air increases, but the sensitivity to the RH, defined analogously to the RI sensitivity as $S_{\rm RH}=\delta \lambda _r/\delta {\rm RH}$, decreases. This decrease can be attributed to a porous termination layer of the 1DPhC [20].

 

Fig. 6. Measured reflectance ratio $R_s(\lambda )/R_p(\lambda )$ as a function of the wavelength for different RHs of air when $\alpha _1=3.8^\circ$.

Download Full Size | PDF

When the external angle of incidence $\alpha$ is increased (the internal angle $\theta$ decreases) from $10.3^\circ$ to $12.3^\circ$, no dip is revealed for the TE wave due to a too low resolving power of the spectrometer, but the microcavity resonances for the TM ($p$-polarized) wave show up as several dips. Three dips are for $\alpha _3=11.3^\circ$ (see Fig. 7) and the most sensitive central resonance dip is redshifted as the RH increases. In Fig. 8, showing the resonance wavelength as a function of the RH of air also for $\alpha _4=12.3^\circ$, together with fits according to $a+b*\exp (c/{\rm RH})$, where $a$, $b$, and $c$ are constants, it results that the sensitivity to the RH (see Fig. 9) at an RH of 50%RH is enhanced from 0.073 nm/%RH for the BSW resonance to 1.367 nm/%RH for the TM cavity-mode resonance. Once again, the sensitivity decreases with the RH of air. In addition, FOM of the RH measurements, defined as FOM$=S_{\rm RH}/$FWHM, can reach 0.012 %RH$^{-1}$ for the BSW resonance and 0.149 %RH$^{-1}$ for the cavity-mode resonance. The measurements for angles $\alpha _2=10.3^\circ$ and $\alpha _4=12.3^\circ$ confirm lower sensitivities (see Fig. 9).

 

Fig. 7. Measured reflectance ratio $R_p(\lambda )/R_s(\lambda )$ as a function of the wavelength for different RHs of air when $\alpha _3=11.3^\circ$.

Download Full Size | PDF

 

Fig. 8. Resonance wavelength as a function of the RH of air for angles of incidence $\alpha _3$ and $\alpha _4$. The solid lines are fits.

Download Full Size | PDF

 

Fig. 9. Sensitivity as a function of the RH of air for four different angles of incidence $\alpha _i$.

Download Full Size | PDF

In conclusion, a new sensing concept based on resonances supported by a 1DPhC microcavity resonator in the Kretschmann configuration has been demonstrated. The sensing concept, which utilizes wavelength interrogation, was demonstrated for a 1DPhC comprising six bilayers of TiO$_2$/SiO$_{2}$ and a termination layer of TiO$_2$. We revealed that Bloch surface TE wave can be excited for a suitable angle of incidence, but sensing of gaseous analytes gives the sensitivity and FOM not too high. When the angle of incidence is decreased, cavity-mode resonances are resolved for both TE and TM waves, and the sensitivity and FOM for the TE wave are enhanced to 52,300 nm/RIU and 402,300 RIU$^{-1}$, respectively. Due to a limited resolving power of a spectrometer, the cavity-mode resonance was revealed experimentally for the TM wave only with the sensitivity and FOM for a humid air as high as 1.367 nm/%RH and 0.149 %RH$^{-1}$, respectively. The proposed concept can be extended to other 1DPhCs enabling ultrahigh-sensitive analysis of different gaseous analytes.