1. Introduction

Waveguide integrated sensors have received great attentions for their applications in many areas, such as biotechnological and pharmaceutical research for target analysis, drug discovery, drug development, and disease diagnostics. Owing to the large refractive index and the low material absorption at infrared wavelengths, silicon nanophotonic sensor platforms in various extensions have been developed with the advantage of high sensitivity and measurement in presence of the sample without any rinsing [1–3].

Optical waveguide sensors employing the evanescent field typically obtain the refractive index of the analyte from the wavelength shift of the functional signal by using a spectrometer [4–6]. However, conventional spectrometers are large and expensive and have a performance that often exceeds the requirements for optical sensing applications. Recently, we presented an optical sensor design by integrating a circular-hole defect with an etched diffraction grating (EDG) spectrometer based on the resonant scattering effect [7]. In the present paper, by optimizing the device, we demonstrate an optical sensor with a miniature spectrometer-on-chip and a sensitivity of up to ~10000 nm/RIU. As an application example, the sensing chip will be used for distinguishing similar biodiesel types and accurately determining their concentrations in a diesel oil mixture. Compared with other integrated refractive index sensors with rather high sensitivity [8, 9], our designs can contribute to not only an accurate index measurement, but also a principal component analysis from the near-infrared spectrum by using the same chip. In addition, the sensors should also be compatible with microfluidic delivery systems that can be integrated directly on the sensor platform for lab-on-a-chip developments.

2. Design and simulation

Figure 1 presents a schematic illustration of the current spectrometer-on-chip biosensor by integrating a circular-hole defect with an EDG spectrometer. As shown, the proposed chip consists of a fiber coupler, an EDG, a sensing area with a circular-hole defect and arrayed photodetectors heterogeneously integrated on top of the output silicon waveguides. Light is first coupled from a fiber to the chip via a planar grating coupler. An input single mode fiber is aligned and glued on the top of the fiber coupler using UV-curable glue. An amplified spontaneous emission (ASE) source gives a broadband unpolarized light with spectral range 1610 nm–1690 nm. This unpolarized light is butt-coupled to the single mode fiber through a focusing gradient index lens. The light is guided trough the input waveguide and crosses the sensing area with a circular-hole defect. Next, the light is guided to the EDG spectrometer which separates and focused the light into different wavelength channels. Finally, InGaAs photodetectors can be integrated on top of the output waveguides to measure the optical power in the corresponding channel. However, in the paper, we do not integrate the arrayed photodetectors, but only connect each output port of the EDG with a power meter with ~ ± 2 nW resolution to measure the transmission power at the corresponding wavelength. We protect the chip area with SU-8, a photo-definable polymer and rectangular contact windows are opened on top of the sensing area, whose photograph can be seen at the lower-right corner of Fig. 1.

 

Fig. 1 The Schematic configuration of proposed optical sensor by integrating a circular-hole defect with an EDG spectrometer.

Download Full Size | PDF

A typical EDG is based on a Rowland mounting. The field propagating from an input waveguide to the free propagation region (FPR) is diffracted by each grating facet. It is then refocused onto an imaging curve and guided into the corresponding output waveguides according to the wavelengths. The circular-hole defect close to the input waveguide can result in strong resonances with the incident light, and the resonant wavelength is very sensitive to the refractive index of the analyte. The position of the resonant peak can be detected using the power meter. The evanescent tail of the optical modes in the silicon waveguides feels the analyst absorption processes. Therefore, the open sensing area covers the whole FSR so that the clear near-infrared absorption information can be achieved using the same chip.

As an example, two sensing chips with the same parameters will be designed: the central wavelength is 1650 nm; the refractive indexes of silica buffer layer and α-Si:H core layer are 1.46 and 3.58, respectively; the incident angle is 45 degrees; the diffraction order is 6; the channel number is 33; and the internal of output waveguides is 5μm. The only difference between the two devices is the channel interval. One is 2.4 nm and the other is 0.5 nm.

As the beginning step of the whole device fabrication, a 5 μm silica buffer layer (i.e., SiO₂) and a 220 nm α-Si:H core layer are successively deposited on a silicon wafer. Then a process of pattern generation will be carried out using the electron beam lithography based on negative resists with 50 nm resolution. Figure 2 gives some pictures of the fabricated device with a 2.4 nm channel interval (i.e., the region with the dotted square in Fig. 1).To increase the coupling efficiency, the width of each input and output waveguide is tapered from 500 nm to 2 μm through a linear taper (see enlarged picture at the upper-right corner in Fig. 2). The circular defect close to the input waveguide is an air hole, and has a 1.65 μm diameter (see enlarged picture at the middle-upper direction in Fig. 2). The circular-hole defect on the surface of the amorphous silicon layers can be fabricated based on a thermal annealing process [7]. We use a horizontal quartz tube oven with nitrogen atmosphere for the annealing process. The holding time at maximum temperature (i.e., 600 °C) is chosen as ~10 minutes to achieve sufficient temperature stabilization. In the process, the outdiffusion of the hydrogen in the α-Si:H thin film will result in circular-hole defects on the surface.

 

Fig. 2 Top-view SEM pictures of the fabricated device corresponding to the region with the dotted square in Fig. 1.

Download Full Size | PDF

In the previous works, we have shown that a circular-hole defect with ~1 μm diameter in an EDG spectrometer can induce strong resonant scattering at some special wavelengths [7]. Now, we will show that with further optimization of the defect diameter a higher sensitivity (more than 10000 nm/RIU) is achievable for the present sensing application. Based on the numerical method we presented for the sensing chip [7], Fig. 3(a) shows the scattering loss with different wavelengths and diameters of the air circular-hole defect for a given refractive index (i.e., 1.455) of the analyte. From this figure, one can see that there are some large loss peaks at the calculated wavelength range for some special defect diameters. Physically, one can consider that the loss peaks result from resonance between the incident wavelength and the physical structure of the air circular-hole defect. One can also see that the resonance depends on the geometrical structure of the defect. Among these loss peaks, four are marked with “1”, “2”, “3” and “4” for the convenience of the analysis below. The corresponding defect diameters for the four peaks are 0.54 μm, 1.62 μm, 2.61 μm and 3.06 μm, respectively.

 

Fig. 3 Numerical analysis of the influence of the circular-hole defect size on the sensing application. (a) The calculated scattering loss with different wavelengths and diameters of the air circular-hole defect for a given refractive index (i.e., 1.455) of the analyte. (b) The calculated resonant scattering wavelength varies as the refractive index of the analyte increases for four resonant peaks marked in (a).

Download Full Size | PDF

Figure 3(b) shows the calculated resonant scattering wavelength varies as the refractive index of the analyte increases using the parameters of the diameter of the circular-hole defect given by the four loss peaks marked in Fig. 3(a). From this figure, one can see that the resonant scattering wavelength noticeably varies as the refractive index of the analyte increases. Using the parameter of Peak “1”, the sensitivity will be considerable high (~12285 nm/RIU). However, from Fig. 3(a), one can also see that the full width at half maximum (FWHM) of the defect diameter for a resonant effect is very small (~0.12 μm) using the parameter of Peak “1”. Although we can obtain only one perfect circular defect in the FPR of the EDG spectrometer using the annealing process, we cannot accurately control its diameter. Therefore, using the parameter of the defect diameter given by Peak “2”, we can achieve a relatively high sensitivity (~9932 nm/RIU) but keep enough fabrication tolerances (the FWHM of the defect diameter is ~0. 7 μm). In addition, since the FWHM of the resonant wavelength for peak “2” is ~6.5 nm, a channel interval of the EDG spectrometer should be less than 2.4 nm to accurately measure the shift of the resonant scattering peak. In addition, considering the fabrication errors, two fabricated chips should have slightly different diameters of the circular-hole defect around 1.62 μm.

3. Result and discussion

In the paper, to verify the validity of the sensing chip we will measure the biodiesel concentration in a diesel oil mixture using two different spectrometer-on-chip devices. The measurement of the refractive index has been used for distinguishing different biodiesel samples [10]. However, some similar types (e.g., sunflower oil methyl ester (SuME) and sunflower ethyl ester (SuEE).) have so similar refractive indices that they are virtually indistinguishable from one another, especially at low blends with conventional commercial refractometers [8]. Now, we will show that we can clearly determine the blend level of these samples and gain further information on the blend component using the present sensing chip.

For the first characterization experiments, we will measure the transmission loss of three different SuEE concentrations in a diesel oil mixture based on the spectrometer-on-a-chip with ~1.65 μm defect diameter and 2.4 nm channel interval. Figure 4(a) shows the detected transmission loss at 33 channels of the EDG spectrometer for the three different blend levels. Note that, for the characterization of the chip, we fabricate another chip without any defect in the FPR using the same parameters, and measure the transmission loss of each channel when no analyte is dropped into the sensing box. Then, the transmission measurements in Fig. 4(a) are normalized by subtracting both measurements. In this way, the normalized spectrum already removes the waveguide propagation and the fiber coupler losses, and the resonant scattering loss and the absorption loss from the analyte will be more clearly illustrated. From the figure, one can see that the resonant scattering wavelength will shift towards a longer wavelength as the blend level decreases. The wavelength difference between pure SuEE and pure diesel oil is very large (~50 nm). Therefore, we can accurately measure such mixture percentiles as long as we design the sensing chip with a smaller channel interval and more output channels. In addition, from this figure, one also sees that there is an obvious absorption peak for the SuEE at ~1665 nm wavelength, whereas the absorption is insignificant for the pure diesel. Then, based on the same chip, we can also carry out a simple component analysis using the near infrared spectrum. In addition, we carry out eight independent measurements for the pure SuEE sample with the same sensing chip, and error bars are given in Fig. 4(a). The maximal error of the transmission loss measurement is ~0.42 dB. However, there is not any influence of the loss error on the final measurement of the refractive index since the position of the resonant peak is determinate with a given analyte for the chip, and only dependent on the channel interval of the EDG spectrometer (e.g., 2.4 nm for the present chip) as shown in Fig. 4(a).

 

Fig. 4 Measured results using the sensing chip with 2.4 nm channel interval. (a) The normalized transmission loss at 33 channels for three different SuEE concentrations in a diesel oil mixture. (b) The resonant scattering wavelength varies as the blend level of the SuEE in a diesel oil mixture increases. (c) The loss at the 1665 nm absorption peak varies as the blend level increases.

Download Full Size | PDF

Figure 4(b) shows the resonant scattering wavelength as a function of the SuEE concentration in a diesel oil mixture using the device. From this figure, one can see that the spectral position of the resonant scattering peak shifts linearly with the blend level. Figure 4(c) shows the measured loss at the 1665 nm absorption peak varies as the blend level. A singular point at 20% blend level in the figure occurs since the resonant scattering peak is also at ~1665 nm wavelength. Obviously, the identity of blend level can also be determined from the intensity of the absorption peak but with limited sensitivity.

As discussed above, we designed and fabricated an EDG spectrometer chip with a smaller channel interval (0.5 nm) to measure different biodiesel analyte at low blends. The defect diameter of the fabricated chip is ~1.63 μm. Figure 5 shows the measured resonant scattering wavelength varies as the blend level of mixtures of SuEE or SuME with diesel oil increases using the fabricated chip. At low blend levels, SuEE and SuME have so similar refractive index that they are virtually indistinguishable from one another using commercial refractometer [10]. However, from Fig. 5, one can see that two samples have significantly different resonant scattering wavelength at the same blend level, and can easily be distinguished using the present method.

 

Fig. 5 The resonant scattering wavelength varies as the blend level of the SuEE and SeMe in a diesel oil mixture increases based on the sensing chip with 0.5 nm channel interval.

Download Full Size | PDF

4. Conclusions

An optical sensing chip integrating a circular-hole defect and an EDG spectrometer based on silicon nanophotonic platforms were presented in this paper. By optimizing the defect diameter, we can achieve a sensing chip with ~10000 nm/RIU sensitivity. As examples, we designed and fabricated two different chips with 0.5 and 2.4 nm spectral intervals, respectively, and applied them into the measurements of the biodiesel concentration in a diesel oil mixture. Measured results show that our chips cannot only easily distinguish similar analyte at low blend levels, but accurately determine the blend level. Although we use the sensing chip for measuring and distinguishing different biodiesel samples as an example in the paper, the present method are equally applicable to other refractive-index-based sensing applications (e.g., label free biosensors).