Open Access
Add to BibSonomy
Abstract
The reported chalcogenide (ChG) rectangular waveguide sensors with a small evanescent field need a large waveguide length to obtain an enhanced light-gas interaction effect. To make such sensors compact and improve the light-gas interaction effect, a microcavity-enhanced absorption spectroscopy technique for methane (CH4) detection was proposed using a mid-infrared chalcogenide/silica-on-fluoride horizontal slot-waveguide racetrack resonator. For the horizontal slot waveguide, an equivalent sensor model (ESM) and related formulations were proposed to simplify the analysis of the racetrack resonator sensor model (RRSM), and the ESM was verified through a comparison between the theoretical result of ESM and the simulation result of RRSM based on the finite element method (FEM). Due to the use of a chalcogenide/silica-on-fluoride horizontal slot-waveguide structure, the waveguide parameters were optimized to obtain a high power confinement factor of 44.63% at the wavelength of 3291 nm, which is at least 5 times higher than other ChG rectangular waveguides. The waveguide length is reduced at least 30 times due to the use of the optimized chalcogenide/silica-on-fluoride horizontal slot-waveguide and racetrack resonator. The limit of detection (LoD) is 3.87 ppm with an intrinsic waveguide loss of 3 dB/cm and an amplitude coupling ratio of 0.1 for the resonator. The response time is less than 5 µs due to the small light-gas interaction area. The influences of environmental pressure and waveguide intrinsic loss on the sensing characteristics were discussed. The compact racetrack resonator sensor structure and equivalent analytical model can also be adopted in the design of an on-chip waveguide sensor for the detection of other gas species.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Most gas species have unique absorption in the mid-infrared band (2.5−20 µm), and the absorption intensity is at least two orders larger than that in the near-infrared band. However, the reported mid-infrared gas sensor systems based on discrete modules or devices with bulk size are power-consuming and sensitive to vibration [1–9]. Some other gas sensing techniques (e.g., electrochemical, catalysis) require a high operation temperature (> 200°C) for activating functional materials [10]. As a comparison with the discrete sensor system, a chip-scale optical waveguide gas sensor is mini in size with small power consumption [11–13]. Moreover, optical waveguide sensors can be used for label-free, non-invasive and non-damage sensing [14–15]. Based on refractive index variation or absorption spectroscopy, solid, liquid and gaseous analyte can be qualitatively or quantitatively analyzed using optical waveguide sensors [16–18].
Waveguide material and structure are two major considerations in the design of an optical waveguide sensor. Waveguide material should be transparent in the mid-infrared to reduce absorption loss. Chalcogenide (ChG) glasses with a high refractive index (> 2) have a low absorption loss below 12 µm [19], which are suitable as core layer material. Calcium fluoride (CaF2) with a low refractive index (∼ 1.4) has low absorption loss below 9 µm [20], which is suitable as lower buffer layer material. The widely used silicon-on-insulator (SOI) platform is unsuitable for mid-infrared applications because the transparent waveband of the silica (SiO2) is limited to ∼ 3.6 µm [21]. On the other hand, waveguide structure should have a high power confinement factor (PCF) within the target gas area for strong light-gas interaction. The easily fabricated rectangular waveguide is the most commonly used structure, but its PCF is < 10% [19,22–23]. To solve this problem, horizontal slot waveguide was proposed to reduce scattering loss and obtain a high PCF [20,24]. Using wet etching technique, horizontal slot waveguide is easy to fabricate without considering the depth-to-width ratio of the slot. In addition, spiral and flexural waveguide structures were proposed to achieve large interaction length with a compact size [25–26]. Superior to spiral and flexural waveguide structures, optical microring resonator is capable of shortening the device length by at least 10 times by enabling light circling in the ring microcavity [27].
In order to obtain a large light-gas interaction length, a novel mid-infrared ChG/silica-on-fluoride horizontal slot-waveguide racetrack resonator gas sensor was proposed, which reveals a high PCF of > 40% and a reduction in the device length by > 30 times. The novel aspects of this racetrack resonator gas sensor include: (1) Instead of vertical slot-waveguide based on dry-etching technique, horizontal slot-waveguide was adopted to obtain a high PCF as well as a low scattering loss due to wet-etching technique. (2) SiO2 near the horizontal slot was removed by hydrofluoric acid (HF) to reduce absorption loss and enhance the light-gas interaction for dropping the limit of detection (LoD). (3) An equivalent sensor model (ESM) and related formulations were proposed to simplify the analysis of the racetrack resonator sensor model (RRSM), and the ESM was verified through a comparison between the theoretical result of the ESM and the simulation result on the RRSM using COMSOL Multiphysics. The structural parameters of the horizontal slot waveguide and racetrack resonator were optimized for improving the PCF and reducing the sensor size. Methane (CH4) was selected as the target gas to estimate the sensing characteristics of the racetrack resonator gas sensor. Effects of operation temperature, environmental pressure and waveguide intrinsic loss on the sensor behavior were discussed to prove the robustness of the ChG/silica-on-fluoride horizontal slot-waveguide racetrack resonator.
2. ChG/silica-on-fluoride horizontal slot-waveguide
2.1 Waveguide structure
CH4 is selected as the target gas and the absorption line located at λ = 3291 nm is used as the target line for detection [28]. The structure of the proposed ChG/silica-on-fluoride horizontal slot-waveguide is shown in Fig. 1(a). The waveguide consists of an upper confined layer, an upper loaded strip core layer, a supporting layer, a horizontal slot, a lower strip core layer, and a lower buffer layer/substrate. The ChG material, As2Se3 is used as the core material with a refractive index n1 = 2.80 (@ 25 °C) at the wavelength of 3291 nm. SiO2 is adopted as the supporting layer, and the SiO2 near the slot is removed for PCF enhancement and decreasing the absorption loss by keeping it away from the slot. Then the effects of the supporting layer on the mode can be neglected. The fluoride material, CaF2 is used as the lower buffer layer as well as the substrate with a refractive index n2 = 1.41 (@ 25 °C). The target gas, with a refractive index of n3 = 1.0, is filled in the horizontal slot and used as the upper confined layer above the upper loaded strip. The waveguide parameters include the upper loaded strip width w1, upper strip height h1, suspended waveguide width w2, suspended waveguide height h2, slot height h3, and lower strip height h4. In the next, the slot waveguide structure will be optimized for guiding single-mode propagation and enhancing PCF in the target gas area.
Fig. 1. (a) The cross-section structure of the horizontal slot-waveguide. (b) Optical mode field distribution of the quasi-TM0 mode of horizontal slot-waveguide, where w1 = 2 µm, w2 = 6 µm, h1 = 0.42 µm, h2 = 0.1 µm, h3 = 0.15 µm, and h4 = 0.4 µm.
2.2 Waveguide cross-section optimization
There are two goals for optimizing the ChG/silica-on-fluoride horizontal slot-waveguide racetrack resonator. One goal is to meet the single-mode condition with increasing the PCF. The other goal is to increase the effective path length Leff and reduce the waveguide length. Here, COMSOL Multiphysics was used for analysis based on the finite element method (FEM). The PCF within the target gas area is defined as
For the quasi-TM0 mode of the straight horizontal slot-waveguide, we optimized the upper strip, the slot and the lower strip in turn to increase PCF with the same method in our previous work [28]. The optimized results of the slot waveguide parameters are shown in Table 1, where w1 = 2 µm, w2 = 6 µm, h1 = 0.42 µm, h2 = 0.1 µm, h3 = 0.15 µm and h4 = 0.4 µm. The effective refractive index Neff and PCF of the ChG/silica-on-fluoride horizontal slot-waveguide are 1.5925 and 44.63%, respectively. With the optimized structure, the mode field distribution of the quasi-TM0 mode of the horizontal slot waveguide is shown in Fig. 1(b). It can be found that a large amount of light is distributed in the target gas area to improve the sensing performance.
Table 1. Optimized parameters of the ChG/silica-on-fluoride horizontal slot-waveguide.
3. Horizontal slot-waveguide racetrack resonator gas sensor
3.1 Sensor structure
The top view of the horizontal slot-waveguide racetrack resonator sensor is shown in Fig. 2(a). The output power from the drop port is used for gas detection. The following parameters should be optimized for small bending loss αb and large effective path length Leff, including the straight waveguide length of the racetrack waveguide L1, the radius of the racetrack waveguide r, the coupling distance d, the straight waveguide length L2, the amplitude coupling ratio κ and the amplitude transmission ratio t. The system-level diagram of the horizontal slot-waveguide racetrack resonator sensor is shown in Fig. 2(b). A commercially available interband cascade laser (ICL, TO66, Nanoplus, Germany) can be used as the mid-infrared laser with an output power P0. The waveguide can easily be butt coupled using fluoride fibers. A single mode fluoride fiber (Thorlabs, ZrF4 fiber patch cable) can be used to couple the incident light into the waveguide. The output light of the waveguide can be coupled by multimode fluoride fiber (Thorlabs, InF3 fiber patch cable). The coupling loss of each facet is considered to be 5 dB, which refers to the simulation results of a horizontal rib slot waveguide [20]. The output laser power P from the drop port is derived by a HgCdTe detector (PDA10JT, Thorlabs, USA) with a specific noise equivalent power (NEP) and bandwidth B. The signal output from the detector is processed by an electrical circuit board (ECB). A data acquisition (DAQ) card (USB6211, National Instrument, USA) is used to sample the output signal from the detector and deliver a signal to drive the ICL. Two temperature controllers (TC1, TC2) are used to control the temperature of the laser and waveguide sensor, respectively. A current driver (CD) is used to control the current and to tune the laser wavelength. A laptop as well as a LabVIEW platform is used for data processing and control the DAQ card. For sensing application, a polydimethylsiloxane (PDMS) gas cell can be bonded on the integrated sensor. An inlet and outlet are added to the PDMS gas cell, and the injection and ejection of the target gas into the waveguide can be realized through a pump.
Fig. 2. (a) Top view of the horizontal slot-waveguide racetrack resonator gas sensor. (b) The system-level diagram of the horizontal slot-waveguide racetrack resonator gas sensor. ICL: interband cascade laser; TC: temperature controller; DAQ: data acquisition card; CD: current driver; ECB: electrical circuit board; PC: personal computer.
3.2 Sensing theory
3.2.1 Effective path length
Define b4 and a1 as the mode field amplitude of the input straight waveguide and output straight waveguide, respectively. The amplitude transfer function of the racetrack waveguide is determined by the amplitude coupling ratio κ and the amplitude transmission ratio t, shown as
Because of light-gas interaction, the output power from the drop port of the racetrack resonator changes with gas concentration and obeys the Lambert-Beer law [22]
To make the device compact, the minimum value of L2 is L2 = r + L1/2. Then Lt can be written as
3.2.2 ESM and verification
With the obtained Leff from Section 3.2.1, the RRSM in Fig. 3(a) can be equivalent to a straight waveguide sensor model (ESM) as shown in Fig. 3(b). The straight waveguide sensor is made up of an input straight waveguide, an output straight waveguide and the equivalent straight waveguide for gas sensing. For the two sensor models, the output power can be calculated using Eq. (7) and COMSOL Multiphysics, respectively, to check the accuracy of the equivalent model and theory. The input power of the sensor is set to 10 mW and the waveguide loss only origins from gas absorption. Based on the optimization in Section 2, the racetrack waveguide structural parameters are r = 108 µm, L1 = 16 µm and d = 1.5 µm. Curves of the output power P versus CH4 concentration C for the two models are shown in Fig. 3(c). As can be seen, the output power of the ESM shows good agreement in variation trend with that of the RRSM, and the slight difference probably results from the errors in reducing the three-dimensional waveguide model to two-dimensional waveguide model. This confirms the accuracy of the ESM.
Fig. 3. (a) Racetrack resonator sensor model (RRSM). (b) Equivalent straight waveguide sensor model (ESM). (c) Curves of output power P versus CH4 concentration C using the two models shown in Fig. 3(a) and 3(b), where P0 = 10 mW, r = 108 µm, L1 = 16 µm and d = 1.5 µm.
3.2.3 Key characteristics: dynamic range and limit of detection
With the verified ESM (Fig. 3(b)), we can analyze the key characteristics of the sensor using a similar technique proposed in [28]. The LoD satisfies
3.3 Sensor optimization
Now we optimize the racetrack waveguide sensor with the optimized waveguide structure in Section 2. Curves of the refractive index difference between the straight waveguide and the bending waveguide ΔNeff and waveguide bending loss αb versus ring radius r are shown in Fig. 4(a). With the increase of r, ΔNeff and αb decrease first and then become stable. ΔNeff equals zero when r > 20 µm, and the αb of a typical SOI bending waveguide with r > 5 µm can be ignored [29]. This is because the PCF of the proposed waveguide is large, so a large radius is required for the racetrack resonator waveguide to confine the mode propagation along the waveguide.
Fig. 4. (a) Curves of ΔNeff and αb versus r. (b) Curves of Leff and E versus κ where αint = 3 dB/cm and C = 50 ppm. (c) Curves of Leff and E versus Lr, where αint = 3 dB/cm, C = 50 ppm and κ = 0.1. (d) Curves of κ versus L1, where d = 1.5 µm. (e) Curves of κ versus d at different L1 from 15 µm to 17 µm.
The micro-cavity enhancement factor is defined as E, which is expressed as E = Leff / Lr. Curves of Leff and E versus mode amplitude coupling ratio κ are shown in Fig. 4(b), where αint = 3 dB/cm and C = 50 ppm. κ determines the coupling effect between the straight waveguide and racetrack waveguide. When κ is small, light tends to circulate in the racetrack waveguide. So Leff and E decrease with the increase of κ. However, if κ is < 0.1, it will be difficult for the coupling between the straight waveguide and racetrack waveguide. So we select κ = 0.1 to obtain a large Leff and a large E. Curves of Leff and E versus Lr are shown in Fig. 4(c), where αint = 3 dB/cm, C = 50 ppm and κ = 0.1. With the increase of Lr, Leff increases due to the increased light-gas interaction length. Because of the increased mode loss mainly resulting from αint other than gas absorption, Leff increases more slowly than Lr which leads to the decrease of E. In order to ignore the influence of αb and achieve a large Leff and E, Lr is selected to be 710 µm (a ring waveguide with the radius of ∼ 110 µm).
Finally, we optimize L1 to ensure κ = 0.1. Increasing PCF and L1 will improve the coupling effect, and the microring resonator can be fabricated without electron beam lithography when the coupling distance d > 1 µm. Curve of κ versus L1 is shown in Fig. 4(d), where d = 1.5 µm. As can be seen, κ increases with the increase of L1. Curves of κ versus d at different L1 from 15 µm to 17 µm are shown in Fig. 4(e). It can be seen that κ decreases with the increase of d because the increase of d weakens the interaction between waveguides. A larger L1 can enhance the interaction between waveguides which leads to the increase of κ. With increasing L1, d can be decreased to keep κ unchanged; however, the increase of L1 makes the footprint of the device large. A small d increases the difficulty of device fabrication, and so d is set to 1.5 µm. When L1 = 16 µm, r = 108 µm and d = 1.5 µm, κ reaches 0.1. The optimized parameters of the racetrack resonator sensor are shown in Table 1.
3.4 Wavelength tuning characteristics of the optimized resonator
The transmittance spectrum of the drop port is shown in Fig. 5 for the optimized waveguide sensor with an operation temperature of 25 °C. The resonance wavelength is near the absorption peak at 3291 nm selected for CH4 detection.
Fig. 5. Curves of resonant peaks of the racetrack resonator at different temperatures. Inset: The thermal-optical tuning coefficient (dλ/dT) fitting curve.
The refractive indices of waveguide materials are influenced by temperature, so the resonant peak can be tuned to the absorption peak by controlling the temperature of the sensor. The thermo-optic coefficient (dn/dT) of As2Se3 and CaF2 are ∼ 4×10−5 °C−1 and ∼ −8.2×10−6 °C−1, respectively [28,30]. Assuming that the temperature of the waveguide and substrate is uniform, within the temperature range from −30 °C to 70 °C, curves of the resonant peaks at different temperature are shown in Fig. 5. The resonant peak has a red shift with the increase of T. Fitted by least square method, the thermo-optic tuning coefficient (dλ/dT) is only ∼ 0.032 nm/°C due to the low thermo-optic coefficient of the core layer. The obtained fitting curve is shown in the inset of Fig. 5. The full width at half maximum (FWHM) of the resonant peak is ∼ 0.03 nm. The simulated absorption spectra of CH4 and H2O from HITRAN database are shown in Fig. 6(a), where T = 298 K, P = 1 atm, L = 10 cm and the concentration of CH4 and H2O are 1000 ppm and 2% (atmospheric concentration level), respectively. The influence of H2O can be ignored by drying the injected gas [1]. The FWHM of the CH4 absorption peak is ∼ 0.14 nm which is 4 times larger than the FWHM of the resonant peak. Setting the temperature to 27.5 °C, the resonant peak can be located at the CH4 absorption peak wavelength. The ICL emission spectrum measured by Fourier transform infrared (FTIR) spectrometer (Thermo Fisher Scientific, model Nicolet iS50) is shown in Fig. 6(b), where T = 15 °C. The output wavelength of the ICL can be located at the CH4 absorption peak wavelength by adjusting current. The line width of the ICL is narrower than the FWHM of resonant peak and the FWHM of CH4 absorption peak. So the αgas(C) of the CH4 absorption peak can be approximated as a fixed value for theoretical calculation and simulation study.
Fig. 6. (a) The simulated absorption spectra of CH4 and H2O, where T = 298 K, P = 1 atm, L = 10 cm and the concentration of CH4 and H2O are 1000 ppm and 2%, respectively. (b) The ICL emission spectrum measured by Fourier transform infrared (FTIR) spectrometer, where T = 15 °C.
3.5 Fabrication process
A diagram of the fabrication process of the integrated sensor is shown in Fig. 7. First, a 0.4 µm thick SiO2 film is deposited on CaF2 substrate by chemical vapor deposition (CVD). The mask on partial SiO2 is formed by photolithography and the SiO2 supporting layer is obtained by inductively coupled plasma (ICP) etching. Then, a ChG strip with a width of 2 µm and a height of 0.4 µm is fabricated by lift-off technique. Next, a 0.15 µm thick SiO2 film is fabricated on ChG by CVD. In the following, a 0.52 µm thick ChG film is fabricated on SiO2 by thermal evaporation and the air groove of the coupling region with a 1.5 µm width is formed by ICP etching. The ChG racetrack disk and strips with a height of 0.42 µm are removed by ICP etching. Finally, the SiO2 slot is removed by HF etching and the etching rate can be 100 nm/min by controlling the HF concentration and temperature. The suspended width of the ChG microcavity can be < 10 µm [31] and a pedestal waveguide similar to the suspended waveguide is proposed for liquid sensing [32], so the ChG layers can be held up well after the HF etching and the suspended slot waveguide structure can be as stable as the conventional rectangular waveguide and the vertical slot waveguide. Because the coupling distance with a width of 1.5 µm can be reached without electron beam lithography, this kind of racetrack resonator can be fabricated in large scale.
Fig. 7. Fabrication process of the waveguide sensor. (a) Depositing SiO2 film. (b) Etching SiO2. (c) Fabricating photoresist mask. (d) Depositing ChG film and removing photoresist. (e) Depositing SiO2 film. (f) Depositing and etching ChG film. (g) Etching ChG film. (h) Etching SiO2.
The key parameters of the waveguide include h3 and w1. h3 and w1 influence the limiting effect of slot on light. Therefore, the SiO2 fabricated by CVD and the ChG strip fabricated by lithography and etching have great influence on waveguide sensor performance. Lithography and etching processes need to be optimized to obtain almost vertical waveguide sidewalls. To reduce the fabrication error, a more precise mask alignment system and masks can be used according to the optimized fabrication conditions. Every fabrication step should be monitored to characterize the device sample.
4. Waveguide gas sensor characteristics
The output power of ICL (Nanoplus, Germany) is set to 10 mW. A HgCdTe detector (Thorlabs, PDA10JT, USA) is used to obtain the signal from the drop port with an operation wavelength from 2 µm to 5.4 µm, a NEP of 2.08×10−11 WHz−1/2 and a B of 160 kHz.
4.1 Leff and DR
Curves of Leff and E versus C are shown in Fig. 8(a). Leff and E become stable and larger than 2.4 cm and 33, respectively, when C < 1000 ppm. Curves of D1(C), PCFαgas(C) and –ln[D1(C)/D0] versus C under the optimal waveguide parameters are shown in Fig. 8(b). D1(C) decreases nonlinearly with the increase of C leading to the increase of –ln[D1(C)/D0]. But the increase rate of –ln[D1(C)/D0] decreases with the increase of C, and –ln[D1(C)/D0] < PCFαgas(C) when C > 66000 ppm. So with the increase of C, Leff and E gradually decrease, and Leff and E decrease to ∼ 0.8 cm and 11.27, respectively, at a pure CH4 environment.
Fig. 8. (a) Curves of Leff and E versus C, where κ = 0.1, r = 108 µm and L1 = 16 µm. (b) Curves of D1(C), PCFαgas(C) and –ln[D1(C)/ D0] versus C under the optimal waveguide parameters.
With mature fabrication technique, the intrinsic loss of the ChG rectangular waveguide can satisfy αint < 3 dB/cm [33–35]. In this work, we assume that αint = 3 dB/cm. Based on Eq. (12), the LoD is calculated to be 3.87 ppm, and the corresponding Leff and E are 2.4 cm and 33.83, respectively. Cmax is calculated to be 100% according to Eq. (13). Therefore, the DR of the sensor system is [3.87 ppm, 100%].
Curves of the full width at half maximum (FWHM) and quality factor (Q) of the microcavity versus C are shown in Fig. 9. The FWHM and Q decrease with the increase of C because of gas absorption. A lower Q leads to the decrease of Leff as shown in Fig. 8(a). Even though Leff decreases, the increase of C will cause the output power to decrease continuously. However, a larger Leff makes the output power change more obviously with C, and the change rate decreases with the decrease of Leff, causing the drop of the sensitivity.
4.2 Response time
Gas diffusion time determines the response time of the sensor. A detailed procedure for the determination of the response time can be seen in our previous work [28]. Under the condition of static free diffusion, only the gas diffusion process in the horizontal slot region is considered, and the diffusion time needed in other area is ignored. Through COMSOL Multiphysics simulation, curve of waveguide output power P1 versus t is shown in Fig. 10(a), where the initial gas concentration in the slot region C0 is 5000 ppm and the initial concentration in other area C1 is 0 ppm. The 10%−90% rise time is determined to be 4.5 µs. Curve of P1 versus t is shown in Fig. 10(b), where C0 is 0 ppm and C1 is 5000 ppm. The 90%−10% fall time is 4.4 µs.
Fig. 10. (a) Curve of P1 versus t when the initial CH4 concentration in the slot C0 = 5000 ppm and the initial concentration in other region is 0 ppm. (b) Curve of P1 versus t when the initial CH4 concentration in the slot C0 = 0 ppm and the initial concentration in other region is 5000 ppm.
4.3 Influence of environmental pressure variation
The environmental pressure variation will influence the normalized gas absorption coefficient αgas and therefore influence LoD. Curves of αgas and LoD versus gas pressure are shown in Fig. 11(a), where T = 25 °C. The increase of pressure leads to the decrease of LoD. However, the variation of LoD is < 1 ppm with a variation of pressure from 0.6 atm to 1.4 atm. So the influence of environmental pressure variation can be ignored.
Fig. 11. (a) Curves of αgas and LoD versus environmental pressure, where T = 25 °C. (b) Curves of Leff, E and LoD versus αint, where κ = 0.1, r = 108 µm and L1 = 16 µm.
4.4 Influence of waveguide intrinsic loss
The structure of the ChG/silica-on-fluoride horizontal slot-waveguide is different from conventional ChG rectangular waveguide. Within the range of 1 − 10 dB/cm, curves of E, Leff and LoD versus αgas are shown in Fig. 11(b). With the increase of αint, E and Leff drop. The decrease of Ps leads to the increase of LoD. Though the sensing characteristics become worse, LoD is still < 15 ppm. So the sensor can be used for monitoring CH4 leakage for safety operation. If αint is reduced to 1 dB/cm, Leff and E can increase to 5.38 cm and 75.7, respectively, and LoD can drop to 1.12 ppm. Therefore, reducing αint is an effective method to improve the sensor performance. The effect of coupling loss on the sensing characteristics is consistent with that of the αint, because they directly affect the output power. The coupling loss decreases the output power, which leads to the increase of the LoD and the decrease of Leff.
4.5 Comparison of PCF among this horizontal slot waveguide and other waveguide structures
Other simulation methods such as finite difference method (FDM) and finite-difference time-domain (FDTD) were reported for calculating PCF. In comparison with these methods, FEM is more precise and consistent with the experimental results [25]. So we use FEM to calculate the PCF of other ChG waveguide structures reported in [19,22–23], and the results are shown in the rightmost column in Table 2. As can be seen, the PCF of the proposed ChG/silica-on-fluoride horizontal slot-waveguide racetrack resonator is at least 5 times larger than other ChG rectangular waveguide. Though the suspended ChG slot waveguide has the largest PCF, its scattering loss is larger than the horizontal slot waveguide due to dry etching. Moreover, the suspended ChG slot waveguide with large PCF is unsuitable to be used in the microring resonator because of large bending loss. The racetrack waveguide reduces the length of the device by more than 30 times when the CH4 concentration is less than 40000 ppm.
Table 2. Comparison among the proposed sensor and other reported ChG waveguide sensors.
Microring resonators are widely used in refractive index sensing. A refractive index sensor obtains the wavelength drift according to the change of the refractive index of analyte. The sensitivity of a reported refractive index waveguide sensor is ∼ 103 nm/RIU, and the refractive index difference between CH4 and air is ∼ 4×10−4, which leads to a wavelength drift of only ∼ 0.4 nm for a refractive sensor and the signal variation (Δλ/λ) is only ∼ 10−4. The light intensity decreases exponentially with the increase of gas concentration and the signal variation (ΔP/P) for CH4 at LoD is ∼ 4×10−2. So this sensor can make the change of signal more sensitive and the light intensity change is easier to detect through a detector.
Our group proposed a CH4 sensor with a 16 m multi-pass gas cell (MPGC, physical size: 45 × 11 × 11 cm3, Egold Technology, Wuhan, Hubei Province, China) at the same wavelength and the LoD of this sensor is ∼ 43.9 parts-per-billion (ppb) [3]. The waveguide sensor proposed in this paper with a Leff of 2.4 cm can achieve a LoD of 3.87 ppm. The PCF in free-space is 1 and that of our waveguide sensor is 44.63%. The comparison results are shown in Table 3. The low PCF, low Leff and waveguide loss lead to a higher LoD than the free-space CH4 sensor. But the LoD×Leff of the waveguide sensor is less than that of the free-space CH4 sensor, which indicates a good performance of the proposed sensor structure.
Table 3. Comparison between free-space CH4 sensor and this waveguide sensor.
5. Conclusions
A novel ChG/silica-on-fluoride horizontal slot-waveguide racetrack resonator was proposed for microcavity-enhanced absorption spectroscopic CH4 sensing. The waveguide structural parameters were optimized for single-mode propagation, large PCF and Leff with a small footprint size. The PCF of the proposed waveguide is 44.63%, which is at least 5 times larger than other reported ChG rectangular waveguides. CH4 is used for evaluating the sensing characteristics of the sensor. The waveguide length is reduced at least 30 times for compactness of the sensor. The LoD is determined to be 3.87 ppm with an intrinsic waveguide loss of 3 dB/cm. The response time of the sensor is as short as 5 µs owning to the small light-gas interaction area. The variation of environmental pressure has less effect on the sensing characteristics, and a small intrinsic loss is preferred for improving the sensor performances. With a change of resonant wavelength, the proposed ESM, related formulations and analytical methods can also be adopted to the design of a chip-scale waveguide sensor for the detection of other gas species.
Funding
National Natural Science Foundation of China (61604151, 61627823, 61775079, 61960206004); Key Science and Technology R&D program of Jilin Province, China (20180201046GX, 20190101016JH, 20200401059GX).
Disclosures
The authors declare no conflicts of interest.
References
1. C. Zheng, W. Ye, N. P. Sanchez, C. Li, L. Dong, Y. Wang, R. J. Griffin, and F. K. Tittel, “Development and field deployment of a mid-infrared methane sensor without pressure control using interband cascade laser absorption spectroscopy,” Sens. Actuators, B 244, 365–372 (2017). [CrossRef]
2. B. Li, C. Zheng, H. Liu, Q. He, W. Ye, Y. Zhang, J. Pan, and Y. Wang, “Development and measurement of a near-infrared CH4 detection system using 1.654 µm wavelength-modulated diode laser and open reflective gas sensing probe,” Sens. Actuators, B 225, 188–198 (2016). [CrossRef]
3. F. Song, C. Zheng, W. Yan, W. Ye, Y. Wang, and F. K. Tittel, “Interband cascade laser based mid-infrared methane sensor system using a novel electrical-domain self-adaptive direct laser absorption spectroscopy (SA-DLAS),” Opt. Express 25(25), 31876–31888 (2017). [CrossRef]
4. M. Dong, C. Zheng, D. Yao, G. Zhong, S. Miao, W. Ye, Y. Wang, and F. K. Tittel, “Double-range near-infrared acetylene detection using a dual spot-ring Herriott cell (DSR-HC),” Opt. Express 26(9), 12081–12091 (2018). [CrossRef]
5. Q. He, C. Zheng, M. Lou, W. Ye, Y. Wang, and F. K. Tittel, “Dual-feedback mid-infrared cavity-enhanced absorption spectroscopy for H2CO detection using a radio-frequency electrically-modulated interband cascade laser,” Laser Phys. 26(12), 15436–15444 (2018). [CrossRef]
6. X. Yin, H. Wu, L. Dong, B. Li, W. Ma, L. Zhang, W. Yin, L. Xiao, S. Jia, and F. K. Tittel, “ppb-Level SO2 Photoacoustic Sensors with a Suppressed Absorption–Desorption Effect by Using a 7.41 µm External-Cavity Quantum Cascade Laser,” ACS Sens. 5(2), 549–556 (2020). [CrossRef]
7. H. Wu, L. Dong, X. Yin, A. Sampaolo, P. Patimisco, W. Ma, L. Zhang, W. Yin, L. Xiao, V. Spagnolo, and S. Jia, “Atmospheric CH4 measurement near a landfill using an ICL-based QEPAS sensor with V-T relaxation self-calibration,” Sens. Actuators, B 297, 126753 (2019). [CrossRef]
8. L. Dong, F. K. Tittel, C. Li, N. P. Sanchez, H. Wu, C. Zheng, Y. Yu, A. Sampaolo, and R. J. Griffin, “Compact TDLAS based sensor design using interband cascade lasers for mid-IR trace gas sensing,” Opt. Express 24(6), A528–A535 (2016). [CrossRef]
9. L. Dong, C. Li, N. P. Sanchez, A. K. Gluszek, R. J. Griffin, and F. K. Tittel, “Compact CH4 sensor system based on a continuous-wave, low power consumption, room temperature interband cascade laser,” Appl. Phys. Lett. 108(1), 011106 (2016). [CrossRef]
10. S. Akbar, P. Dutta, and C. Lee, “High-temperature ceramic gas sensors: a review,” Int. J. Appl. Ceram. Technol. 3(4), 302–311 (2006). [CrossRef]
11. W. Lai, S. Chakravarty, X. Long, C. Lin, and R. Chen, “On-chip methane sensing by near-IR absorption signatures in a photonic crystal slot waveguide,” Opt. Lett. 36(6), 984–986 (2011). [CrossRef]
12. R. Dwivedi and A. Kumar, “A compact and ultra-high sensitive RI sensor using modal interference in an integrated optic waveguide with metal under-cladding,” Sens. Actuators, B 240, 1302–1307 (2017). [CrossRef]
13. J. Alaman, M. Lopez-Valdeolivas, R. Alicante, and C. Sanchez-Somolinos, “Optical Planar Waveguide Sensor with Integrated Digitally-Printed Light Coupling-in and Readout Elements,” Sensors 19(13), 2856 (2019). [CrossRef]
14. T. Jin, L. Li, B. Zhang, H. G. Lin, H. Wang, and P. T. Lin, “Optical-feedback cavity-enhanced absorption: a compact spectrometer for real-time measurement of atmospheric methane,” Sci. Rep. 7(1), 5836 (2017). [CrossRef]
15. J. Su, “Label-Free Biological and Chemical Sensing Using Whispering Gallery Mode Optical Resonators: Past, Present, and Future,” Sensors 17(3), 540 (2017). [CrossRef]
16. C. Chen, D. A. Mohr, H. Choi, D. Yoo, M. Li, and S. Oh, “Waveguide-Integrated Compact Plasmonic Resonators for On-Chip Mid-Infrared Laser Spectroscopy,” Nano Lett. 18(12), 7601–7608 (2018). [CrossRef]
17. D. A. Mohr, D. Yoo, C. Chen, M. Li, and S. Oh, “Waveguide integrated mid-infrared plasmonics with high-efficiency coupling for ultracompact surface-enhanced infrared absorption spectroscopy,” Opt. Express 26(18), 23540–23549 (2018). [CrossRef]
18. C. Ranacher, C. Consani, A. Tortschanoff, R. Jannesari, M. Bergmeister, T. Grille, and B. Jakoby, “Mid-infrared absorption gas sensing using a silicon strip waveguide,” Sens. Actuators, A 277, 117–123 (2018). [CrossRef]
19. A. Gutierrez-Arroyo, E. Baudet, L. Bodiou, J. Lemaitre, I. Hardy, F. Faijan, B. Bureau, V. Nazabal, and J. Charrier, “Optical characterization at 7.7 µm of an integrated platform based on chalcogenide waveguides for sensing applications in the mid-infrared,” Opt. Express 24(20), 23109–23117 (2016). [CrossRef]
20. B. Kumari, R. K. Varshney, and B. P. Pal, “Design of chip scale silicon rib slot waveguide for sub-ppm detection of N2O gas at mid-IR band,” Sens. Actuators, B 255, 3409–3416 (2018). [CrossRef]
21. V. M. Lavchiev and B. Jakoby, “Photonics in the Mid-Infrared: Challenges in Single-Chip Integration and Absorption Sensing,” IEEE J. Sel. Top. Quantum Electron. 23(2), 452–463 (2017). [CrossRef]
22. J. Charrier, M. L. Brandily, H. Lhermite, K. Michel, B. Bureau, F. Verger, and V. Nazabal, “Evanescent wave optical micro-sensor based on chalcogenide glass,” Sens. Actuators, B 173, 468–476 (2012). [CrossRef]
23. A. Gutierrez-Arroyo, E. Baudet, L. Bodiou, V. Nazabal, E. Rinnert, K. Michel, B. Bureau, F. Colas, and J. Charrier, “Theoretical study of an evanescent optical integrated sensor for multipurpose detection of gases and liquids in the Mid-Infrared,” Sens. Actuators, B 242, 842–848 (2017). [CrossRef]
24. D. M. Kita, J. Michon, S. G. Johnson, and J. Hu, “Are slot and sub-wavelength grating waveguides better than strip waveguides for sensing?” Optica 5(9), 1046–1054 (2018). [CrossRef]
25. L. Tombez, E. J. Zhang, J. S. Orcutt, S. Kamlapurkar, and W. M. J. Green, “Methane absorption spectroscopy on a silicon photonic chip,” Optica 4(11), 1322–1325 (2017). [CrossRef]
26. Z. Han, P. Lin, V. Singh, L. Kimerling, J. Hu, K. Richardson, A. Agarwal, and D. T. H. Tan, “On-chip mid-infrared gas detection using chalcogenide glass waveguide,” Appl. Phys. Lett. 108(14), 141106 (2016). [CrossRef]
27. A. Nitkowski, L. Chen, and M. Lipson, “Cavity-enhanced on-chip absorption spectroscopy using microring resonators,” Opt. Express 16(16), 11930–11936 (2008). [CrossRef]
28. M. Pi, C. Zheng, R. Bi, H. Zhao, L. Liang, Y. Zhang, Y. Wang, and F. K. Tittel, “Design of a mid-infrared suspended chalcogenide/silica-on-silicon slot-waveguide spectroscopic gas sensor with enhanced light-gas interaction effect,” Sens. Actuators, B 297, 126732 (2019). [CrossRef]
29. S. Ghosh and B. M. A. Rahman, “Design of on-chip hybrid plasmonic Mach-Zehnder interferometer for temperature and concentration detection of chemical solution,” Sens. Actuators, B 279, 490–502 (2019). [CrossRef]
30. A. Savchenkov and A. Matsko, “Calcium fluoride whispering gallery mode optical resonator with reduced thermal sensitivity,” J. Opt. 20(3), 035801 (2018). [CrossRef]
31. D. Kita, H. Lin, J. Li, Z. Han, P. Su, T. Gu, A. Agarwal, A. Yadav, K. Richardson, and J. Hu, “Suspended chalcogenide microcavities for ultra-sensitive chemical detection,” IEEE Sens. (2016).
32. P. T. Lin, H. G. Lin, Z. Han, T. Jin, R. Millender, L. C. Kimerling, and A. Agarwal, “Label-Free Glucose Sensing Using Chip-Scale Mid-Infrared Integrated Photonics,” Adv. Opt. Mater. 4(11), 1755–1759 (2016). [CrossRef]
33. X. Gai, S. Madden, D. Y. Choi, D. Bulla, and B. Luther-Davies, “Dispersion engineered Ge11.5As24Se64.5 nanowires with a nonlinear parameter of 136W−1m−1 at 1550 nm,” Opt. Express 18(18), 18866–18874 (2010). [CrossRef]
34. J. F. Viens, C. Meneghini, A. Villeneuve, T. V. Galstian, E. J. Knystautas, M. A. Duguay, K. A. Richardson, and T. Cardinal, “Fabrication and characterization of integrated optical waveguides in sulfide chalcogenide glasses,” J. Lightwave Technol. 17(7), 1184–1191 (1999). [CrossRef]
35. X. Gai, D. Y. Choi, S. Madden, and B. Luther-Davies, “Polarization-independent chalcogenide glass nanowires with anomalous dispersion for all-optical processing,” Opt. Express 20(12), 13513–13521 (2012). [CrossRef]
