Open Access
Add to BibSonomy
Abstract
Refractive-index (RI)-based sensing is a major optical sensing modality that can be implemented in various spectral ranges. While it has been widely used for sensing of biochemical liquids, RI-based gas sensing, particularly small-molecule gases, is challenging due to the extremely small RI change induced by gas concentration variations. We propose a RI-based ultracompact fiber-optic differential gas sensor that employs metal-organic-framework (MOF)-based dual Fabry-Perot (FP) nanocavities. A MOF is used as the FP cavity material to enhance the sensitivity as well as the selectivity to particular gas molecules. The differential sensing scheme leverages the opposite change in the cavity-length-dependent reflection of the two FP cavities, which further enhances the sensitivity compared with single FP cavity based sensing. For proof-of-concept, a fiber-optic CO2 sensor with ZIF-8-based dual FP nanocavities was fabricated. The effective footprint of the sensor was as small as 157 µm2 and the sensor showed an enhanced sensitivity of 48.5 mV/CO2Vol%, a dynamic range of 0-100 CO2Vol%, and a resolution of 0.019 CO2Vol% with 1 Hz low-pass filtering. Although the current sensor was only demonstrated for CO2 sensing, the proposed sensor concept can be used for sensing of a variety of gases when different kinds of MOFs are utilized.
© 2020 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Gas sensors play an important role in various applications, including combustion control [1], indoor air quality control [2], agriculture [3], aquaculture [4], food processing [5], and medical care [6]. Fiber optic sensors are an attractive solution to in-situ and real-time gas sensing in these applications due to their advantages of robustness to harsh (e.g., high temperature and corrosive) environments, remote sensing capability, multiplexing capability, and immunity to electromagnetic interference [7,8].
Among different fiber-optic sensing modalities, refractive-index (RI)-based sensing has been widely employed to measure different parameters [9–13]. The RI-based sensing systems can be designed to work in a variety of optical spectral ranges, including the near-infrared wavelengths (e.g., 1.3 µm and 1.55 µm) that are commonly used for optical communications. Therefore, by taking advantage of the low-cost and compact optoelectronic devices used for optical communications [14–17], it is possible to develop cost-effective and compact RI-based sensing systems.
However, RI-based gas sensing is challenging. Because the RI change induced by the variation of gas concentration is extremely small, RI-based gas sensing often suffers from poor sensitivity. Moreover, RI-based gas sensing itself cannot provide selective sensing of a specific gas molecule or compound from a gas mixture. For these reasons, infrared absorption spectroscopy, which is based on the measurement of molecular fingerprint from gas molecular vibrational and rotational bands, has been widely used for selective and sensitive gas sensing [18–23]. Some efforts have recently been made to the development of near-infrared absorption gas sensors [24–26]. However, most of the infrared absorption-based gas sensors operate in the mid-infrared range, where the fundamental vibrational and rotational bands are located [18], and thus, these sensors require mid-infrared spectrometers that are often bulky and expensive.
Recent advances in metal-organic-framework (MOF) technology have shed light on the RI-based gas sensing [27,28]. MOFs are nanoporous crystalline structures made of metal ions and organic linkers. By engineering the pore size, MOF can be tailored to selectively adsorb specific gas molecules. Furthermore, the large surface area of the porous MOF structure helps increase the gas concentration in the MOF, and thus amplify the RI change induced by gas concentration variations.
There have been some efforts devoted to the development of RI-based gas sensors leveraging nanoporous MOFs [29–32]. However, most of them have been focused on sensing of large-molecule gases such as volatile organic compounds, and sensing of small-molecule gases such as carbon dioxide (CO2), nitrogen (N2), and oxygen, still remains a challenge due to the even smaller changes in the RI. To address this challenge, there is a need to develop a novel RI-based sensing modality for selective and sensitive sensing of small-molecule gases. Recently, Kim et al. developed an evanescent wave fiber-optic CO2 gas sensor with a MOF film coated on a 5-cm-long fiber side-wall [33]. Here, we report an ultracompact fiber-optic differential gas sensor with MOF-based dual Fabry-Perot (FP) nanocavities. By using differential sensing from two FP cavities that incorporate the MOF as the cavity material, the sensor has the capability of selective measurement of small-molecule gases with enhanced sensitivity compared with single FP cavity based sensing. For proof-of-concept, a fiber-optic CO2 sensor is developed with ZIF-8-based dual FP nanocavities. ZIF-8 is selected as the MOF since its pore size is suitable for the selective adsorption of CO2 molecules [23,33]. Furthermore, ZIF-8 is thermally stable up to 550 °C [34,35], which enables CO2 sensing in relatively high-temperature environments sections.
2. Operation principle
The CO2 gas sensor consists of two fiber-optic FP nanocavities made of ZIF-8 thin films of different thicknesses, which allows differential sensing of CO2 gas, as shown in Fig. 1(a). Each FP cavity has two mirrors: the first mirror is the fiber end-face and the second mirror is the top surface of the ZIF-8 film. The reflection spectrum of the FP cavity is determined by the mirror reflectivities and the optical path difference of the cavity ( = 2×nZIF-8×cavity length) [36]. From Fresnel’s law for normal incidence, the reflectivity of each mirror can be expressed as [37]
Fig. 1. (a) Schematic of the fiber-optic differential CO2 sensor with ZIF-8-based dual FP nanocavities. (b) Working principle of the differential CO2 sensor. The reflection intensities of the FP cavities 1 (intensity increase) and 2 (intensity decrease) change to opposite directions when the ambient gas changes from N2 (solid red and blue curves) to CO2 (dashed red and blue curves) over the entire wavelength range of interest. By measuring the differential intensity from the two FP cavities with a balanced photoreceiver, the sensitivity can be enhanced. The solid black curve represents the spectral profile of a broadband light source.
3. Experimental section
3.1 Materials
Zinc nitrate hexahydrate (Zn(NO3)2·6H2O) and 2-Methylimidazole (2-mIm) were purchased from Sigma-Aldrich. A single-mode optical fiber (SMF-28, Corning Inc.) was used for the sensor.
3.2 Fabrication of ZIF-8-based fiber-optic Fabry-Perot nanocavities
Each optical fiber was cleaved, so that the fiber end-face was perpendicular to its longitudinal direction. The cleaved angle was measured by using the imaging system of a fiber optic fusion splicer (Quantum Type-Q102-CA core alignment fusion splicer, Sumitomo Electric Lightwave), and it was less than ±1.0°. The cleaved optical fiber was cleaned with a Piranha solution for 1 hour followed by deionized water rinse and N2 blow dry. A ZIF-8 thin film was coated on the fiber end-face as follows. First, methanolic solutions of Zn(NO3)2·6H2O (1.25 mM) and 2-mIm (2.5 mM) were prepared. Next, the optical fiber tip was immersed in a fresh mixture of the two methanolic solutions for 30 minutes at room temperature. The fiber tip was then washed with fresh methanol to remove the unreacted zinc ions and 2-mIm followed by N2 blow dry. The ZIF-8 coating process was repeated until the designed film thickness (i.e., cavity length) was achieved.
3.3 Characterization of ZIF-8 films
To characterize the dimension and properties of the coated ZIF-8 film, a dummy ZIF-8 film was coated on a silicon substrate, which was fabricated together with the optical fiber samples. The thickness and RI of the ZIF-8 film were measured by spectroscopic ellipsometry (M-2000 Ellipsometer, J.A. Woollam Co.). A Drude-Lorentz model was utilized to fit the measured data in the wavelength range of 800 nm to 1690 nm. The crystal structure of the ZIF-8 film was analyzed by using an X-ray diffractometer (D8 Advance with LynxEye, Bruker) with Cu Κα radiation (wavelength λ = 1.5418 Å) in the 2θ range of 5° to 60° (with a step size of 0.05°). The amorphous halo in the X-ray diffraction (XRD) pattern was removed by using XRD software (Advanced TOPAS, Bruker). The ellipsometry and XRD experiments were carried out in ambient air at room temperature and atmospheric pressure.
3.4 Optical simulations
By using Finite-Difference Time-Domain (FDTD) software (FDTD solutions, Lumerical Inc.), the reflection spectrum of the ZIF-8 FP nanocavity was obtained. FDTD simulation was used since it provides the reflection spectrum over a wide wavelength range with a single simulation run. The FP cavity was simulated by using a two-dimensional FDTD model, as shown in Fig. 2(a). Bloch and perfectly matched layer (PML) boundary conditions were used for the x and y boundaries, respectively. A plane wave normal to the ZIF-8 film was used for the light source. The RIs of the ZIF-8 film in N2 and CO2 were modeled based on the dispersion relation provided in Fig. 2(b). Note that the dispersion relation of the ZIF-8 film in N2 was obtained from the measurement of the 9-cycle-coated ZIF-8 film and the RI change of the ZIF-8 film due to the CO2 adsorption was modeled as +0.005 refractive index unit (RIU) [33,38]. The RI of silica was adopted from the material database of the FDTD software. The ambient RIs of N2 and CO2 were modeled to be 1.000294 RIU and 1.000439 RIU, respectively [39,40].
Fig. 2. (a) Schematic of the FDTD simulation model to obtain the reflection spectrum of the ZIF-8 FP nanocavity. (b) Refractive indices of the ZIF-8 film in N2 and CO2 as a function of wavelength used for the FDTD simulation models. The black curve is the measured data of the 9-cycle-coated ZIF-8 film in ambient air and the RI difference between the black and red curves is 0.005 RIU. The RI of the ZIF-8 film in N2 may be slightly different from that in ambient air. However, it should be noted that considering the gas composition of ambient air (N2 78.08%, O2 20.95%, Ar 0.93%, CO2 0.035%) [41] and the CO2 selectivity of ZIF-8 (over N2, O2, and Ar) [24,33], the error would be much smaller than the RI change of the ZIF-8 film (i.e., +0.005 RIU) when the ambient gas changes from N2 to CO2.
3.5 Sensor characterization
The fabricated CO2 gas sensor was characterized by using the experimental setup shown in Fig. 3. A broadband source (EBS300006, Exalos) with a center wavelength of 1570 nm and a 3 dB bandwidth of 110 nm was used. The input light was coupled to the differential gas sensor through three 1×2 fiber-optic couplers (TW1550R5F1, Thorlabs), and the reflected light from the sensor was coupled to a 10 MHz balanced fiber-optic photoreceiver (2117-FC, New Focus). The photoreceiver gain was set to be 1×104 and an electronic various optical attenuator (V1550F, Thorlabs) was coupled to one of the photoreceiver input ports for balancing purpose. The output signal from the photoreceiver was collected by using a data acquisition system (Picoscope 6000, Picotechnology) with a sampling rate of 750 samples per second. CO2 sensing was carried out with N2 as a carrier gas. The N2 and CO2 gas flows were controlled with mass flow controllers (6A0107BV-NC and 6A0107SV-CA, Dakota Instruments). The gas flowrate was set to be 7 l/min if not mentioned otherwise. A solenoid valve (2W025-08, ATO Inc.) was installed in the CO2 gas tube to characterize the dynamic response of the gas sensor. Additionally, the reflection spectra of the individual fiber-optic FP nanocavities were obtained by using a tunable laser source (TSL-510, Santec) and a bare optical fiber without a ZIF-8 film as a reference, which is not shown in Fig. 3. The sensor characterization was carried out in the atmospheric-pressure, room-temperature condition. Before the characterization, all the sensors were preheated at 100 °C for 1 hour to avoid possible moisture adsorption on the outer surface of the ZIF-8 film [42].
Fig. 3. Experimental setup for sensor characterization. BBS: broadband light source, BPR: balanced photoreceiver, DAQ: data acquisition instrument, MFC: mass flow controller, VOA: variable optical attenuator, SMF: single-mode optical fiber, SV: solenoid valve.
4. Results and discussion
FDTD simulations were carried out to investigate how the FP cavity length (i.e., thickness of the ZIF-8 layer) affects the refection characteristic of the sensor. The simulated fringe pattern of reflection spectra for a wide range of FP cavity lengths from zero to 5 µm is shown in Fig. 4(a), which is used for sensor design. It can be seen that the free-spectral-range decreases as the cavity length increases. Figure 4(b) shows the reflection spectra obtained for two representative FP cavities (thicknesses of 330 nm and 650 nm) before (pure N2) and after CO2 exposure. Both reflection spectra exhibit redshift after the CO2 exposure (i.e., increase of 0.005 RIU). However, over the wavelength range of 1510 nm to 1620 nm, which is the typical 3 dB bandwidth of a broadband light source for optical communications, the reflected intensity obtained from the two FP cavities changes to opposite directions: the reflection intensity of the 330-nm-long cavity increases while that of the 650-nm-long cavity decreases [Fig. 4(c)]. Moreover, the change in the net reflection intensity over the wavelength range of 1510 nm to 1620 nm with respect to the cavity length was obtained, which exhibits an oscillatory behavior [Fig. 4(d)]. Therefore, if the lengths of the two FP cavities are chosen to ensure that the changes in the net reflection intensity take one of the maxima (positive) and one of the minima (negative), respectively, the sensitivity of the CO2 sensor can be enhanced by using the differential output from the two FP cavities. For proof-of-concept, the lengths of the two FP cavities are designed to be 330 nm and 650 nm so that the changes in the net reflection intensity reach the first local maximum and minimum, respectively. Note that the maximum sensitivity can be achieved by using two FP cavities with cavity lengths of 3900 nm and 4200 nm, which provide the global maximum and minimum in the net reflection intensity change, respectively. The maximum sensitivity is 2.6 times higher than that of the proof-of-concept design. However, fabrication of such a long FP cavity requires a large number of ZIF-8 coating cycles.
Fig. 4. (a) Simulated reflection spectra of FP cavities with different cavity lengths in N2. The color bar indicates the normalized reflection intensity. (b) Simulated reflection spectra 330-nm and 650-nm-long FP cavities in N2 (solid red and blue curves) and CO2 (dashed red and blue curves). (c) Zoom-in spectra from 1510 nm to 1620 nm showing the opposite change in the reflection of the two FP cavities. (d) Simulated changes in the net reflection intensity (spectral range: 1510 nm – 1620 nm) of FP cavities with different cavity lengths. The normalized reflection intensity was obtained by normalizing the amount of power reflected from the FP cavity against the source power. The net reflection intensity was obtained by integrating the simulated normalized reflection intensity from 1510 nm to 1620 nm.
The fiber-optic FP nanocavity was fabricated by coating a ZIF-8 film on the end-face of a single-mode optical fiber. Figure 5(a) shows a representative scanning electron microscopy (SEM) image of the fiber end-face coated with a ZIF-8 thin film. The thickness of the ZIF-8 film was controlled by the number of coating cycle. The film thickness increased linearly with the increase of coating cycles [Fig. 5(b)]. The thickness increase rate was measured to be approximately 33 nm per coating cycle. As the film thickness increased, the RI of the fabricated ZIF-8 film also increased. The dependency of RI on coating cycles is believed to be related to the crystalline size of the coated ZIF-8. The ZIF-8 crystalline size was found to increase with increasing coating cycles [Fig. 5(c)], which could make the ZIF-8 film denser by reducing the spacing between the neighboring crystals. Furthermore, the XRD patterns of the ZIF-8 films exhibited multiple peaks with the peak locations maintained independent of the coating cycles [Fig. 5(d)], which confirmed a cubic crystal structure of the coated ZIF-8. The locations and intensities of the peaks agreed well with the XRD patterns in the literature [29,30,43].
Fig. 5. (a) Representative SEM image of the ZIF-8 thin film fabricated on the end-face of a single-mode optical fiber (9 cycle coating). (b) Measured thicknesses and RIs for the ZIF-8 thin films fabricated with different coating cycles on a Si substrate. (c) SEM images showing the ZIF-8 thin films with different coating cycles on a Si substrate. The scale bar is 1 µm. (d) Representative XRD pattern of the ZIF-8 thin film on a Si substrate (13 cycle coating).
The performance of the fabricated fiber-optic FP cavities with different ZIF-8 coating cycles were characterized in terms of their response to the change of ambient gas from N2 to CO2. A bare optical fiber without a ZIF-8 film was used as a reference, which was coupled to the minus port of the photoreceiver. The output of the photoreceiver with respect to the coating cycle is plotted in Fig. 6(a), which agrees well with the simulation results [Fig. 4(d)].
Fig. 6. Measured responses of the fabricated fiber-optic FP cavities when the ambient gas switched from N2 to CO2. The sample size for the mean and standard deviation was three. (a) Responses of the FP cavities with respect to different coating cycles. (b) Normalized reflection spectra of the FP cavities with 9-cycle and 13-cycle-coated ZIF-8 films in N2 (solid red and blue curves) and CO2 (dashed red and blue curves). The normalized reflection spectra were obtained by normalizing the signals reflected from the fiber-optic FP cavities against the signal reflected from a bare optical fiber without a ZIF-8 film. (c) 10% - 90% response and recovery time of the FP cavities with respect to different coating cycles.
The maximum and minimum outputs were obtained with the 9-cycle and 13-cycle-coated FP cavities, whose cavity lengths were approximately 260 nm and 425 nm, respectively. The normalized reflection spectra of the 9-cycle and 13-cycle-coated FP cavities were measured, which confirmed that the reflections of the two FP cavities changed to the opposite directions over the spectral range of 100 nm [Fig. 6(b)]. The 10% - 90% response and recovery times of the FP cavities with different ZIF-8 coating cycles were measured [Fig. 6(c)]. It was found that the influence of the cavity length (i.e., the number of coating) on the response and recovery times was not significant and the average response and recovery times of the FP cavities with the six different coating cycles were approximately 164 msec and 364 msec, respectively. The variation in errors of the response and recovery times for different cavity lengths is believed to result from the fabrication error in the ZIF-8 film thickness, the errors in the gas flow controller operation and the solenoid valve operation, and the noise in the measurement system.
The differential gas sensor was developed by using two FP cavities with 9-cycle and 13-cycle ZIF-8 coatings. The response of the developed differential sensor was characterized and compared with the responses of each individual FP cavity [see Fig. 7(a)]. The sensor had a linear response to CO2 gas concentration of 0 - 100 CO2Vol%. The CO2 sensitivity was measured to be 48.5 mV/CO2Vol%, which was higher than those of the individual FP cavities (40.7 mV/CO2Vol% and -7.7 mV/CO2Vol%, respectively). These results confirm that the differential sensing scheme based on the opposite cavity-length-dependent reflection changes from two FP cavities helps enhance the CO2 sensitivity, compared with single FP cavity based sensing. The 10% - 90% response and recovery times were measured to be 141 msec and 310 msec [Fig. 7(b)], which were close to those of the individual FP cavities. Note that the response and recovery times were significantly influenced by the experimental setup. As shown in Fig. 7(c), the response and recovery times decrease with increasing gas flowrate. Based on exponential fitting of the measured data, the response and recovery times can be extrapolated to be approximately 3 sec at a 1 ml/min of flowrate. At a flowrate higher than 10 l/min, the response time converges to be approximately 100 msec, which is the response time of the solenoid valve. The measured root-mean-square (rms) noise of the steady-state sensor response was measured to be 14.07 mV for a 1 sec period, which rendered a resolution of 0.290 CO2Vol%. Note that the resolution can be improved by applying a low-pass filter to the sensor output. Figure 7(d) shows the filtered rms noises and the corresponding resolutions with respect to different cutoff frequencies. For example, with a 1 Hz low-pass filter (i.e., quasi-static measurement), the sensor has a resolution of 0.019 CO2Vol% ( = 0.93 mV). The resolution can be further improved by enhancing the sensitivity through the optimization of the FP cavity lengths since it is expressed as the noise divided by the sensitivity. For example, the simulation results in Fig. 4(d) show that the sensitivity of the optimized design (i.e., t1 = 3900 nm and t2 = 4200 nm) is 2.6 times higher than that of the current proof-of-concept design. Given the filter noise of 0.93 mV with a 1 Hz low-pass filter, the enhanced sensitivity can render a resolution of 0.007 CO2Vol%.
Fig. 7. Measured responses of the fabricated fiber-optic differential CO2 sensor. (a) Sensor response under different CO2Vol% conditions compared with the responses obtained with the two single FP cavities. (b) Dynamic response of the sensor when the flowrate ratio of N2 and CO2 changed from 1:0 to 1:1 and then back to 1:0. (c) 10% - 90% response and recovery times of the sensor measured under different gas flowrates. (d) Low-pass-filtered rms noises of the steady-state sensor response and the corresponding resolutions obtained with different cutoff frequencies.
5. Conclusion
We have reported an ultracompact fiber-optic differential gas sensor with MOF-based dual FP nanocavities. The proof-of-concept sensor demonstrated the selective and sensitive measurement of CO2 gas by using ZIF-8 as a cavity material and the differential sensing scheme based on the opposite cavity-length-dependent reflection changes from two FP cavities. Since the proposed sensor concept is based on sensing of gas concentration induced RI change, it can utilize a wide variety of spectral range, including the optical communication bands. Moreover, the effective footprint of the sensor is as small as 157 µm2 (π × (5 µm)2 × 2 EA), and the MOF-based fiber-optic FP nanocavities can be batch-fabricated by using a simple dip-coating method. Although the current sensor demonstrates the measurement of CO2, the proposed differential sensor concept can be used for sensing a variety of gases when different kinds of MOFs are adopted.
Funding
Office of Naval Research High-Temperature Energy Systems (HiTES) program; U.S. Department of Agriculture; National Institute of Food and Agriculture Sustainable Agricultural Systems (SAS) program.
Disclosures
The authors declare no conflicts of interest.
References
1. J. Mulrooney, J. Clifford, C. Fitzpatrick, and E. Lewis, “Detection of carbon dioxide emissions from a diesel engine using a mid-infrared optical fibre based sensor,” Sens. Actuators, A 136(1), 104–110 (2007). [CrossRef]
2. J. Kwon, G. Ahn, G. Kim, J. C. Kim, and H. Kim, “A study on NDIR-based CO2 sensor to apply remote air quality monitoring system,” in 2009 Iccas-Sice (IEEE, 2009), pp. 1683–1687.
3. D. Chaudhary, S. Nayse, and L. Waghmare, “Application of wireless sensor networks for greenhouse parameter control in precision agriculture,” Intl. J. Wireless Mobile Netw. 3(1), 140–149 (2011). [CrossRef]
4. T. J. Pfeiffer, S. T. Summerfelt, and B. J. Watten, “Comparative performance of CO2 measuring methods: Marine aquaculture recirculation system application,” Aquacultural Eng. 44(1), 1–9 (2011). [CrossRef]
5. S. Neethirajan, D. Jayas, and S. Sadistap, “Carbon dioxide (CO 2) sensors for the agri-food industry—a review,” Food Bioprocess Technol. 2(2), 115–121 (2009). [CrossRef]
6. M. Folke, L. Cernerud, M. Ekström, and B. Hök, “Critical review of non-invasive respiratory monitoring in medical care,” Med. Biol. Eng. Comput. 41(4), 377–383 (2003). [CrossRef]
7. K. Grattan and T. Sun, “Fiber optic sensor technology: an overview,” Sens. Actuators, A 82(1-3), 40–61 (2000). [CrossRef]
8. B. Lee, “Review of the present status of optical fiber sensors,” Opt. Fiber Technol. 9(2), 57–79 (2003). [CrossRef]
9. H. Tazawa, T. Kanie, and M. Katayama, “Fiber-optic coupler based refractive index sensor and its application to biosensing,” Appl. Phys. Lett. 91(11), 113901 (2007). [CrossRef]
10. X. Fan, I. M. White, S. I. Shopova, H. Zhu, J. D. Suter, and Y. Sun, “Sensitive optical biosensors for unlabeled targets: A review,” Anal. Chim. Acta 620(1-2), 8–26 (2008). [CrossRef]
11. H. Bae, D. Yun, H. Liu, D. A. Olson, and M. Yu, “Hybrid miniature Fabry–Perot sensor with dual optical cavities for simultaneous pressure and temperature measurements,” J. Lightwave Technol. 32(8), 1585–1593 (2014). [CrossRef]
12. Z. Zhang, Y. Chen, H. Liu, H. Bae, D. A. Olson, A. K. Gupta, and M. Yu, “On-fiber plasmonic interferometer for multi-parameter sensing,” Opt. Express 23(8), 10732–10740 (2015). [CrossRef]
13. H.-T. Kim and M. Yu, “Lab-on-Fiber Nanoprobe with Dual High-Q Rayleigh Anomaly-Surface Plasmon Polariton Resonances for Multiparameter Sensing,” Sci. Rep. 9(1), 1–9 (2019). [CrossRef]
14. B. Jalali, S. Yegnanarayanan, T. Yoon, T. Yoshimoto, I. Rendina, and F. Coppinger, “Advances in silicon-on-insulator optoelectronics,” IEEE J. Sel. Top. Quantum Electron. 4(6), 938–947 (1998). [CrossRef]
15. G. Masini, L. Colace, and G. Assanto, “Si based optoelectronics for communications,” Mater. Sci. Eng., B 89(1-3), 2–9 (2002). [CrossRef]
16. S. Fama, L. Colace, G. Masini, G. Assanto, and H.-C. Luan, “High performance germanium-on-silicon detectors for optical communications,” Appl. Phys. Lett. 81(4), 586–588 (2002). [CrossRef]
17. L. Tsybeskov and D. J. Lockwood, “Silicon-germanium nanostructures for light emitters and on-chip optical interconnects,” Proc. IEEE 97(7), 1284–1303 (2009). [CrossRef]
18. J. Hodgkinson and R. P. Tatam, “Optical gas sensing: a review,” Meas. Sci. Technol. 24(1), 012004 (2013). [CrossRef]
19. N. Gayraud, J. Stone, W. MacPherson, A. George, J. Knight, D. Hand, and D. Reid, “Mid-infrared methane detection in a photonic bandgap fiber using a broadband optical parametric oscillator,” Opt. Express 15(18), 11219–11224 (2007). [CrossRef]
20. N. Gayraud, ŁW Kornaszewski, J. M. Stone, J. C. Knight, D. T. Reid, D. P. Hand, and W. N. MacPherson, “Mid-infrared gas sensing using a photonic bandgap fiber,” Appl. Opt. 47(9), 1269–1277 (2008). [CrossRef]
21. S.-S. Kim, N. Menegazzo, C. Young, J. Chan, C. Carter, and B. Mizaikoff, “Mid-infrared trace gas analysis with single-pass Fourier transform infrared hollow waveguide gas sensors,” Appl. Spectrosc. 63(3), 331–337 (2009). [CrossRef]
22. S. Herminjard, L. Sirigu, H. P. Herzig, E. Studemann, A. Crottini, J.-P. Pellaux, T. Gresch, M. Fischer, and J. Faist, “Surface plasmon resonance sensor showing enhanced sensitivity for CO 2 detection in the mid-infrared range,” Opt. Express 17(1), 293–303 (2009). [CrossRef]
23. X. Chong, Y. Zhang, E. Li, K.-J. Kim, P. R. Ohodnicki, C.-H. Chang, and A. X. Wang, “Surface-Enhanced Infrared Absorption: Pushing the Frontier for On-Chip Gas Sensing,” ACS Sens. 3(1), 230–238 (2018). [CrossRef]
24. X. Chong, K.-J. Kim, P. R. Ohodnicki, E. Li, C.-H. Chang, and A. X. Wang, “Ultrashort near-infrared fiber-optic sensors for carbon dioxide detection,” IEEE Sens. J. 15(9), 5327–5332 (2015). [CrossRef]
25. X. Chong, E. Li, K. Squire, and A. X. Wang, “On-chip near-infrared spectroscopy of CO2 using high resolution plasmonic filter array,” Appl. Phys. Lett. 108(22), 221106 (2016). [CrossRef]
26. X. Chong, K.-J. Kim, E. Li, Y. Zhang, P. R. Ohodnicki, C.-H. Chang, and A. X. Wang, “Near-infrared absorption gas sensing with metal-organic framework on optical fibers,” Sens. Actuators, B 232, 43–51 (2016). [CrossRef]
27. B. Chen, S. Xiang, and G. Qian, “Metal− organic frameworks with functional pores for recognition of small molecules,” Acc. Chem. Res. 43(8), 1115–1124 (2010). [CrossRef]
28. L. E. Kreno, K. Leong, O. K. Farha, M. Allendorf, R. P. Van Duyne, and J. T. Hupp, “Metal–organic framework materials as chemical sensors,” Chem. Rev. 112(2), 1105–1125 (2012). [CrossRef]
29. G. Lu and J. T. Hupp, “Metal− organic frameworks as sensors: a ZIF-8 based Fabry− Pérot device as a selective sensor for chemical vapors and gases,” J. Am. Chem. Soc. 132(23), 7832–7833 (2010). [CrossRef]
30. J. Hromadka, B. Tokay, S. James, R. P. Tatam, and S. Korposh, “Optical fibre long period grating gas sensor modified with metal organic framework thin films,” Sens. Actuators, B 221, 891–899 (2015). [CrossRef]
31. J. Tao, X. Wang, T. Sun, H. Cai, Y. Wang, T. Lin, D. Fu, L. L. Y. Ting, Y. Gu, and D. Zhao, “Hybrid photonic cavity with metal-organic framework coatings for the ultra-sensitive detection of volatile organic compounds with high immunity to humidity,” Sci. Rep. 7(1), 41640 (2017). [CrossRef]
32. R. Zhang, D. Zhang, Y. Yao, Q. Zhang, Y. Xu, Y. Wu, H. Yu, and G. Lu, “Metal–Organic Framework Crystal-Assembled Optical Sensors for Chemical Vapors: Effects of Crystal Sizes and Missing-Linker Defects on Sensing Performances,” ACS Appl. Mater. Interfaces 11(23), 21010–21017 (2019). [CrossRef]
33. K.-J. Kim, P. Lu, J. T. Culp, and P. R. Ohodnicki, “Metal–organic framework thin film coated optical fiber sensors: a novel waveguide-based chemical sensing platform,” ACS Sens. 3(2), 386–394 (2018). [CrossRef]
34. K. S. Park, Z. Ni, A. P. Côté, J. Y. Choi, R. Huang, F. J. Uribe-Romo, H. K. Chae, M. O’Keeffe, and O. M. Yaghi, “Exceptional chemical and thermal stability of zeolitic imidazolate frameworks,” Proc. Natl. Acad. Sci. 103(27), 10186–10191 (2006). [CrossRef]
35. S. Gadipelli, W. Travis, W. Zhou, and Z. Guo, “A thermally derived and optimized structure from ZIF-8 with giant enhancement in CO 2 uptake,” Energy Environ. Sci. 7(7), 2232–2238 (2014). [CrossRef]
36. M. Yu and B. Balachandran, “Acoustic measurements using a fiber optic sensor system,” J. Intell. Mater. Syst. Struct. 14(7), 409–414 (2003). [CrossRef]
37. F. J. Arregui, Y. Liu, I. R. Matias, and R. O. Claus, “Optical fiber humidity sensor using a nano Fabry–Perot cavity formed by the ionic self-assembly method,” Sens. Actuators, B 59(1), 54–59 (1999). [CrossRef]
38. J. Cookney, W. Ogieglo, P. Hrabanek, I. Vankelecom, V. Fila, and N. E. Benes, “Dynamic response of ultrathin highly dense ZIF-8 nanofilms,” Chem. Commun. 50(79), 11698–11700 (2014). [CrossRef]
39. E. R. Peck and B. N. Khanna, “Dispersion of nitrogen,” J. Opt. Soc. Am. 56(8), 1059–1063 (1966). [CrossRef]
40. J. Old, K. Gentili, and E. Peck, “Dispersion of carbon dioxide,” J. Opt. Soc. Am. 61(1), 89–90 (1971). [CrossRef]
41. F. T. Mackenzie and J. A. Mackenzie, Our changing planet: an introduction to earth system science and global environmental change (Prentice Hall, Upper Saddle River, NJ, 1998).
42. F. Tian, A. M. Mosier, A. Park, E. R. Webster, A. M. Cerro, R. S. Shine, and L. Benz, “In situ measurement of CO2 and H2O adsorption by ZIF-8 films,” J. Phys. Chem. C 119(27), 15248–15253 (2015). [CrossRef]
43. M. C. McCarthy, V. Varela-Guerrero, G. V. Barnett, and H.-K. Jeong, “Synthesis of zeolitic imidazolate framework films and membranes with controlled microstructures,” Langmuir 26(18), 14636–14641 (2010). [CrossRef]
