Open Access
Add to BibSonomy
Abstract
A novel double spot-ring plane-concave multipass cell (DSPC-MPC) gas sensor was proposed for simultaneous detection of trace gases, which has lower cost and higher mirror utilization than the traditional multipass cell with 129 m, 107 m, 85 m, 63 m and 40 m effective optical path lengths adjustable. The performance of the DSPC-MPC gas sensor was evaluated by measuring CO and CH4 using two narrow linewidth distributed feedback lasers with center wavelengths of 1567 nm and 1653 nm, respectively. An adjustable digital PID laser frequency stabilization system based on LabVIEW platform was developed to continuously stabilize the laser frequency within ∼±30.3 MHz. The Allan deviation results showed that the minimum detection limits for CO and CH4 were 0.07 ppmv and 0.008 ppmv at integration times of 711 s and 245 s, respectively. The proposed concept of DSPC-MPC provides more ideas for the realization of gas detection under different absorption path lengths and the development of multi-component gas sensing systems.
© 2024 Optica Publishing Group under the terms of the Optica Open Access Publishing Agreement
1. Introduction
Carbon monoxide (CO) and methane (CH4) are two important trace gases in the atmosphere that contribute to air pollution and global warming [1–4]. As a highly flammable and explosive gas, CH4 poses a high risk of leakage in places such as natural gas pipelines, coal mines, and areas near swamps, and threatens the safety of personnel and equipment in coal mine production and natural gas pipeline transport at all times [5,6]. In addition, CH4 is also considered to be one of the main greenhouse gas in the atmosphere, with a warming capacity 84 times stronger than CO2, and its contribution to global warming due to the greenhouse effect reaches about 20% [7]. Since the Industrial Revolution, atmospheric CH4 concentrations have more than doubled from ∼0.7 parts-per-million in volume (ppmv) to ∼1.9 ppmv, with an average growth rate of 15-17 parts-per-billion per year (ppbv yr-1) [8,9]. The proportion of CH4 emissions from natural and anthropogenic sources is 40% and 60%, respectively [10], and any source of emissions that rapidly increase CH4 concentrations in atmospheric would be devastating for the Earth's climate and ecosystems [11]. At the 28th United Nations Climate Change Conference (COP28) in November 2023, over 100 countries signed a joint declaration, reaching a consensus on reducing CH4 and other greenhouse gas emissions to achieve short-term mitigation of global warming targets [12]. CO is a common atmospheric pollutant, which is a product of incomplete combustion of carbon under conditions of insufficient oxygen [13]. It is extremely toxic and can also combine with hemoglobin in the body, leading to a reduction in the oxygen-carrying capacity of the blood, which may lead to asphyxiation due to lack of oxygen in severe cases [14,15]. In addition, CO contributes to the greenhouse effect. Although it is not a direct greenhouse gas, it can react with hydroxyl radicals to produce carbon dioxide (CO2), which affects the lifetime of greenhouse gases [16]. Therefore, developing a multi-component trace gas sensing system for high-precision and high-sensitivity measurements of CO and CH4 not only enables the timely detection of their leaks but also allows for the determination of their sources and concentration levels, which is of great significance in preventing poisoning incidents, understanding global warming and climate change, and ensuring public and environmental safety.
Common trace gas detection technologies mainly include gas chromatography (GC) [17], chemiluminescence (CL) sensor technology [18], electrochemical (EC) sensor technology [19], and optical gas sensor technology [20–22]. However, GC has poor selectivity for different gases, the operation is troublesome and the results are easy to be interfered. CL sensing technology requires the addition of O3 to oxidize the target gas, making the results easily interfered. EC sensor technology is sensitive to humidity change and has short lifespan, which is not conducive to gas detection. In comparison, tunable diode laser absorption spectroscopy (TDLAS), as one of the most common optical gas sensing technologies, with its excellent molecular fingerprinting ability, as well as the advantages of high sensitivity, and fast response speed, has pushed forward the continuous progress in the field of gas analysis, which is widely used in environmental monitoring, industrial production, medical diagnosis, and security precautions [23–25].
TDLAS is usually combined with multi-pass gas cell (MPGC) to increase the effective interaction length of the sample gas with the laser beam, thus the signal-to-noise ratio (SNR) and the minimum detection limit (MDL) of the system are improved. Traditional MPGCs include Herriott cell [26], White cell [27], circular cell [28] and Chernin cell [29], however, reducing the cost of MPGC and improving mirror utilization has always been a difficult and unresolved key issue. In 2019, Haiyue Sun et al. demonstrated an acetylene (C2H2) sensor based on TDLAS technique using a Herriott cell with an effective optical path length (EOPL) of 10 m, which resulted in a MDL value of 1.3 ppbv when the averaging time was 270 s [30]. In 2024, Yahui Liu et al. developed a highly sensitive light-induced thermoelectric spectroscopy sensor based on a MPGC with dense spot pattern and a novel quartz tuning fork with low resonance frequency for the detection of C2H2, which achieved a MDL of 24.6 ppbv at an EOPL of 37.7 m [31]. The MPGC they designed can achieve high EOPL and high mirror utilization in a small volume, however, the EOPL is single and cannot be adjustable within the same MPGC. In 2010, J. Barry McManus et al. developed a high-sensitivity trace gas sensor using a astigmatic Herriott cell with an EOPL of 240 m, achieving high-precision measurements of gases such as formaldehyde (HCHO), hydrogen peroxide (H2O2), nitrous acid (HNO2), and nitrogen dioxide (NO2), with MDLs reaching the pptv level [32]. In 2024, Zhao Li et al. designed and constructed a new low-pressure TDLAS sensor for detecting dissolved CO and carbon dioxide (CO2) in transformer insulating oil using a White cell with an EOPL of 15 m, the MDLs of 0.147 ppmv and 0.9 ppmv were achieved [33]. These MPGCs can simultaneously detect multi-component gases. However, the double concave mirrors design of the traditional Herriott cell and the triple mirror design of the White cell often require relatively high costs. Therefore, a multi-component trace gas sensors based on a low-cost, OPLE adjustable MPGC urgently need to be developed to fill the gaps in previous research.
In this work, a multi-component trace gas sensing system based on a novel double spot-ring plane-concave multipass cell (DSPC-MPC) with 129 m, 107 m, 85 m, 63 m and 40 m EOPLs adjustable was developed, and CO and CH4 were simultaneously detected by use of two narrow linewidth NIR distributed feedback (DFB) lasers with centre wavelengths of 1567 nm and 1653 nm, respectively. Wavelength modulation spectroscopy with the second-harmonic detection technique (WMS-2f) employed to achieve high detection sensitivity. Software program written on the basis of LabVIEW replace hardware such as function generators and lock-in amplifiers, allowing for greater compactness and flexibility in the sensing system. A laser frequency stabilization system based on digital PID control was designed for laser frequency locking. The performance of the developed multi-component trace gas sensing system was evaluated by simultaneous detection of CO and CH4, verifying the stability and reliability of the sensing system.
2. Sensor system setup
2.1 Calculation and simulation of the DSPC-MPC
Unlike the traditional double concave spherical mirror Herriott cell and three-sided mirror White cell, the new DSPC-MPC consists of a concave spherical mirror and a plane mirror. The ray transmission model of a plane-concave optical resonant cavity is based on the spatial equations and the law of reflection [34–36]. As shown in Fig. 1, the concave mirror M1 with a curvature radius of R and the plane mirror M2 are placed symmetrically on the same axis, with a mirror spacing of D. Establish a x-y-z coordinate system with the center of the concave spherical surface as the origin, the spatial equations of M1 and M2 can be respectively represented as:
Fig. 1. (a) Theoretical model diagram of spatial ray transmission in a plane-concave mirror optical resonant cavity.
The initial incident position is defined as X0(x0, y0, z0), the direction vector of the incident light beam is (a, b, c), and the linear equation of the incident light beam can be represented as:
The position of the first reflected spot X1(x1, y1, z1) can be obtained from eq 2 and 3. Based on the principle of mirror reflection, the coordinates of the point XP, which is the symmetric point of X0 with respect to the normal line, are (2×1-x0, 2y1-y0, z0). Since XP is on the reflected light beam, the linear equation of the first reflected light beam can be determined by X1 and XP:
The position of the second reflected spot X2(x2, y2, z2) can be obtained from eq 1 and 4. The line between X2 and the center point of the spherical equation is the normal of the second reflection, as shown in eq 5. Assuming that XQ is a point of symmetry of X1 about the normal, the equation of the line determined by these two points is shown in equation 6. The line determined by XQ and X1 is perpendicular to the normal, as shown in eq 7, and the midpoint of XQ and X1 lies on the normal, as shown in eq 8. The coordinates of XQ can be determined by combining eq 7 and 8.
Then the linear equation of the second reflected light beam can be found from XQ and X2:
The position of the third reflected spot X3 can be obtained from eq 2 and 9. Finally, all spot position information on a plane-concave mirror can be obtained by repeating the above calculation.
Based on the optical ray transmission theory of the plane-concave optical resonant cavity, a program was written in Matlab for determining the parameters of mirror radius (r), two-mirror spacing (D), radius of curvature (R), and initial incidence position of the DSPC-MPC. The TracePro software was used to model the lenses to simulate ray-tracing based optical transmission characteristics. The model of the plane-concave mirror is shown in Fig 2, four coupling holes with a diameter of 3.6 mm are placed on the two mirrors, for the entry and exit of the two laser beams. Placing the two mirrors with coaxial symmetry, the midpoint of the line connecting the centers of the mirrors is selected as the origin to establish the x-y-z coordinate system, and the z-axis is the base axis of the symmetrical placement. The coordinates of the entrance hole on the concave spherical mirror are (0, 42, -250) and (0, -32, -250), which correspond to the exit hole coordinates (-5.5, 41, 250) and (6.75, -31.1, 250) on the plane mirror, respectively. The beams from light sources 1 and 2 will enter DSPC-MPC through the entrance holes 1 and 2, and be reflected back and forth between the mirrors, then pass through the exit hole. Different spot distributions can be obtained on the concave and plane mirrors by varying the direction vector of the incident beam, as shown in Fig. 3, and the corresponding parameters are given in Table 1. It can be seen that by varying the direction vector of the incident beam, the DSPC-MPC can realize 129 m, 107 m, 85 m, 63 m and 40 m EOPLs adjustable. In this work, the spot pattern shown in Fig. 3(a) and (b) was chosen for the experiment. It can be seen that EOPL of ∼129 m was achieved with back and forth of 256 passes. Each channel produces an EOPL of ∼64.5 m.
Fig. 2. (a) Model diagram of concave spherical mirror and plane mirror. (b) Model diagram of plane-concave optical resonant cavity.
Table 1. Parameters corresponding to different spot patterns of the DSPC-MPCa
2.2 Design of the DSPC-MPC
The structure of the DSPC-MPC consists of plane-concave mirror, mirror frame, hollow cylindrical tube, double-screw bolt, flange, and O-ring, as shown in Fig. 4(a). The two mirrors are ground and polished on a K9 glass substrate, with a central thickness of 6 mm each. The material of the hollow cylindrical tube is transparent polyvinyl chloride (PVC), with a thickness of 4 mm. The frame, double-ended bolts and flanges are made of stainless steel to ensure sufficient rigidity. The material of the O-ring is hydrogenated nitrile rubber, which is pressed by the flange at the connection between the gas cell and the outside to achieve excellent sealing. Fig. 4(b) shows the 3D model and simulated optical transmission characteristics of the DSPC-MPC, verifying the accuracy of the structural design. The dimensions of the whole mechanical structure are 60 cm × 14 cm × 11 cm, and the internal gas cell volume of the DSPC-MPC is ∼3.3 L considering the thickness of the hollow cylindrical tube is 4 mm.
Fig. 4. (a) Mechanical structure design of the DSPC-MPC. (b) The 3D model and simulated optical transmission characteristics of the DSPC-MPC.
The reflective surfaces of plane-concave mirrors need to be coated to realize the back-and-forth reflection of the beam in the DSPC-MPC. However, in actual processing, the reflectivity of the coated mirror surface cannot reach 100%. According to Beer-Lambert's Law, a laser beam with an intensity of I0 is injected into a sample gas, the absorbed light intensity can be expressed as:
where R is the mirror reflectance, n is the number of reflections, k is the gas spectral absorption coefficient, c is the gas concentration, and L is the EOPL. When the mirror spacing is D, L can be represented as D·n. Under weak absorption, Eq. 10 can be written as: where I/(I0kc) can be viewed as the SNR of the MPC. When the mirror spacing is 500 mm, the relationship between SNR and number of reflections under different reflectivity is shown in Fig. 5(a). It can be seen that the optimal SNR increases by ∼22.2 times when the reflectivity increases from 95% to 99.78%. In order to achieve high mirror reflectivity of the DSPC-MPC, special dielectric films are coated on the plane-concave mirrors, which can achieve high reflectivity while possessing corrosion resistance and anti-pollution properties. The reflectivity of the special dielectric films at different wavelengths is shown in Fig. 5(b), and it can be seen that the reflectivity reaches 99.78% at 1567 nm and 1653 nm.Fig. 5. (a) The relationship between SNR and number of reflections in a MPC under different reflectivity. (b) Reflectance of the special dielectric films at different wavelengths.
2.3 Spectral line pair selection
In general, gas molecules exhibit different absorption line strengths in different spectral ranges. According to the HITRAN database on the Spectrogram website [37], CO and CH4 have strong absorption near 1.5 µm and 1.6 µm, respectively. At the same time, in order to avoid the potential interference of water vapor (H2O) and CO2 during the measurement, the absorption spectra of CO, CH4, H2O and CO2 were simulated with the EOPL of 64.5 m at 1 atm and 298 K, as shown in Figs. 6(a) and 6(b). It can be seen that CO and CH4 have high absorption line intensities at wavenumbers of 6383.08 cm-1 and 6046.94 cm-1, respectively, and are almost unperturbed by other gases. Thus, two NIR DFB lasers with centre wavelengths of 1567 nm and 1653 nm were chosen to complete the detection in this work. After setting the temperature of the laser, it can be tuned around a specific wavelength by varying the injection current. The temperatures of the two DFB lasers were set to 22 °C and 24 °C, respectively, and the relationship between the laser wavelength and the injection current is given in Fig. 6(a) and 6(b). It can be observed that when the currents were 57 mA and 82 mA, respectively, the lasers can precisely target the absorption lines of CO and CH4.
Fig. 6. Simulated absorbance of 100 ppmv CO, 2 ppmv CH4, 1% H2O, and 380 ppmv CO2 at 298 K, 1 atm, and 64.5 m optical path length, and the current-wavenumber relationship for the DFB laser. (a) In the range of 6382.2-6384 cm-1. (b) In the range of 6046.35-6047.55 cm-1.
2.4 Experimental configuration
The experimental configuration of the dual-gas sensor is shown in Figure 7. Two NIR DFB lasers with centre wavelengths of 1567 nm (NEL, NLK1L5GAAA) and 1653 nm (NEL, NLK1U5FAAA) were selected for CO and CH4 detection and driven by laser controller 1 (SRS, LDC-501) and laser controller 2 (ILX Lightwave, LDC-3724C), respectively. The high-frequency sine-wave signals and low-frequency triangle-wave signals were generated by homemade digital function generator based on LabVIEW, and were superimposed by a digital adder program to modulate input of the laser controller through the analog output (AO) of a data acquisition (DAQ) card (NI, USB-6363) for tuning the output wavelength of the laser. The two laser beams were collimated by two custom-made fiber collimators, and then focused by two flat-concave lenses (THORLABS, LA5464) with a focal length of 500 mm into the DSPC-MPC, so that the focal point of the beams falls exactly on a plane mirror. The focused beam was detected by a beam quality analyzer (CINOGY, InGaAs-320) at a distance of 500 mm from the flat-concave lenses prior to the experiment, as shown in the inset of Fig. 7, with a beam diameter of 0.4 mm. After reflected back-and-forth in the DSPC-MPC, the two beams were focused by a flat-convex lens (THORLABS, LA5315) with a focal length of 20 mm and received by two photodetectors (THORLABS, PDA20CS2). The two signals were captured by the DAQ card and demodulate by a quadrature digital lock-in amplifier, then processed with PC.
2.5 Verification of the EOPL
The EOPL of the DSPC-MPC was experimentally validated. The temperature and current of the CO laser were set to 24 °C and 82 mA, respectively. A 50 Hz triangle-wave with an amplitude of 2.1 V was used to modulate the laser current, causing its output wavelength swept through the absorption peaks of 6000 ppmv CO (mixture with pure N2). The direct absorption signal detected by the detector is shown in Fig. 8(a), and a third-order polynomial was fitted to the unabsorbed region to obtain the background signal. It can be seen that the maximum absorbed signal (V1) and the corresponding unabsorbed signal (V2) are 1.792 V and 2.011 V, respectively. The bias voltage of the detector is 0.01 V (V0) when there is no laser irradiation. The absorbance α can be given by the following equation:
α was calculated to be 0.1156. Comparing with the HITRAN database on the Spectrogram website [37], EOPL can be determined as 64.55 m.The temperature and current of the CH4 laser were set to 22 °C and 57 mA, respectively. A 50 Hz triangle-wave with an amplitude of 1.8 V caused the output wavelength of the laser to sweep through the absorption peak of 80 ppmv CH4. The direct absorption signal and the background signal are shown in Fig. 8(b), with the maximum absorption signal of 0.818 V (V1) and the corresponding unabsorbed signal of 0.994 V (V2). According to Eq. 12, α is obtained as 0.197. Comparing the simulation results on the spectrogram website [37], the EOPL is determined to be 64.55 m.
3. Sensor optimization
3.1 Parameter optimization
The 2f signal amplitude is related to both the absorption gas pressure and the modulation amplitude of the sine-wave, it is necessary to measure the 2f signals of CO and CH4 at different pressures and modulation amplitudes to obtain the optimal modulation amplitude. It should be noted that pressure broadening may result in spectral overlap and enhance the absorption of other gases at the absorption peak of the target gas. In order to reduce interference and facilitate atmospheric monitoring, the pressure is set to be less than 1 atm in the measurement. The temperature and current settings of the laser are the same as the previous section. A 5200 Hz sine-wave is superimposed on a 50 Hz triangle-wave with an amplitude of 2.1 V was used to modulate the CO laser. 600 ppmv CO (mixture with pure N2) was measured at 296 K, and the relationship between the normalized 2f peak signals and the sine-wave modulation amplitudes at 0.6 atm, 0.7 atm, 0.8 atm, 0.9 atm, and 1 atm were shown in Fig. 9(a). It can be observed that the optimal modulation amplitude of 0.2 V is obtained at 1 atm.
Fig. 9. (a) Normalized 2f signal peaks of CO at different pressures and modulation amplitudes. (b) CH4 signals.
The temperature and current of the CH4 laser were set to 24°C and 82 mA, respectively. A 5200 Hz sine-wave is superimposed on a 50 Hz triangle-wave with an amplitude of 1.8 V was used to modulate the CH4 laser. 10 ppmv CH4 was obtained by dilution with pure N2, and the relationship between the normalized 2f peak signals and the amplitude of the sine-wave modulation was measured at 0.6 atm, 0.7 atm, 0.8 atm, 0.9 atm, and 1 atm, respectively. The results are shown in Fig. 9(b), which shows that the maximum signal of CH4 occurs at 1 atm and 0.2 V.
3.2 Laser frequency locking
65 ppmv CH4 gas (mixture with pure N2) was appraised stability of the laser output wavelength. Each 2f signal consists of 2100 sampling points, corresponding to a wavenumber range of 6045.22-6047.48 cm-1. With 930 s of continuous measurements, a total of 62 2f signals were collected, with a sampling interval of 15 s. Analyzing the 2f signal collected for the first time, its peak corresponds to the 990th sampling point, and 930 s later, the peak of the 2f signal collected corresponds to the 1397th sampling point. Using ΔX to denote the offset of the sampling points corresponding to the peak of the 2f signal, ΔX is 407, which corresponds to a wavenumber offset of 0.438 cm-1, as shown in Fig. 10. It can be observed that there is a significant drift of the laser emission frequency with time and the drift rate is 0.00047 cm-1/s.
The output wavenumbers of the CH4 laser at different temperatures and currents were measured by a wavelength meter (Bristol, 671series), as shown in Fig. 11, which shows that the wavenumber of the laser changes at rates of 0.4224 cm-1/°C and 0.0294 cm-1/mA for temperature and current, respectively. In practical measurements, the control accuracy of the laser controller for the laser temperature and current can be affected by factors such as working time and ambient temperature, making it unable to maintain the emission frequency of the laser at a constant value. The drift of laser frequency seriously threatens the stability and sensitivity of the sensor. In this work, we designed a digital PID laser frequency stabilization system based on LabVIEW platform, which realizes digital PID control through software, making it simple to design, easy to implement, and convenient for real-time adjustment. The flow of the laser frequency locking algorithm based on digital PID is shown in Fig. 12. First, laser-locked target frequency value f0 is set. The bias voltage Voff, as well as the thresholds Vmax and Vmin are set to ensure that [Voff +Vmin, Voff +Vmax] falls within the allowable input voltage range for laser modulation to avoid damaging the laser. The difference between the set frequency and the actual frequency of the laser is calculated as the error signal e(k). The PID parameters: Proportional Gain (Kp), Derivative Gain (Kd) and Integral Gain (Ki) are set in the control software of the upper computer respectively, and the PID control quantity u(k) is generated to satisfy:
Next, a threshold protection judgment is applied to the control quantity u(k). If u(k) > Vmax, the output signal V0 = Vmax; if u(k)< Vmin, the output signal V0 = Vmin; if Vmin ≤ u(k) ≤ Vmax, the output signal V0 = u(k). Finally, a DC bias Voff is added to the output signal V0 to obtain the output control voltage signal Vc = Voff + V0, which is then fed back to the laser controller to modulate the laser frequency for frequency locking.
The frequency fluctuations obtained from the change of sampling points at the peak of the 2f signal under free running and laser frequency locking techniques are shown in Figure 13. In the experiment, continuous measurements were made for ∼12 h with a sampling interval of 1 s. It can be seen that the laser frequency was effectively controlled within a range of 30.3 MHz after the laser frequency-locked technique was turned on, indicating a significant improvement in the stability of the sensor.
Fig. 13. Wavenumber variation at the peak of 2f signal with free running and laser frequency locking technique.
4. Sensor performance
4.1 Linear and time response
In the experiment, twelve different concentrations of CO gas (range from 450 ppmv to 2400 ppmv) were measured at 296 K and 1 atm, and 2f signals were shown in Fig. 14(a). One hundred points were measured continuously for each concentration with a sampling interval of 1 s, and the results of the peak 2f signals are shown in Fig. 14 (c). As can be seen from the inset in Fig. 14 (c), the standard deviation of the 450 ppmv is 2.57397 × 10−8. The linear responsivity (R2 = 0.9988 for the linear fit) was evaluated with each concentration was taken, and shown in Fig. 14 (e). Different concentrations of CH4 gas were obtained by diluting a 135 ppmv CH4/N2 mixture, and the 2f signals were measured separately as shown in Fig. 14 (b). Fig. 14 (d) shows the results of the continuous measurement of different CH4 concentrations, with 100 points sampled for each concentration and an average sampling time of 1 s. The inset shows the standard deviation of 135 ppmv CH4 is 7.28862 × 10−7. While diluting 50 ppmv of CH4 to 40 ppmv with pure N2 at a gas flow rate of 5 L/min, the peak 2f signal was monitored in real time, and it can be seen that the response time of the sensor is ∼3.5 s. The relationship between 2f signal of CH4 and the gas concentration is shown in Fig. 14 (f), and the R-square value of the linear fit was ∼0.999.
Fig. 14. (a), (b) 2f signals of CO and CH4 at different concentrations. (c), (d) Continuous measurement of the peak 2f signal of CO and CH4 at different concentrations. (Inset: Standard deviation of continuous measurements for a given concentration.) (e), (f) The linear response of the peak 2f signals of CO and CH4 to the corresponding concentrations.
4.2 Precision and stability
The SNR and the MDL are important parameters for evaluating system performance. Therefore, 2200 ppmv CO was measured and the 2f signal was shown in Fig. 15(a), the inset gives the standard deviation of the unabsorbed signal. The SNR was calculated as 23, which corresponds to a MDL of 95.6 ppmv. Fig. 15(b) shows the 2f signal of 2 ppmv CH4, the standard deviation of the unabsorbed portion as shown in the inset, and the SNR of the sensing system was calculated to be 30.25, which corresponds to a MDL of 0.066 ppmv. Based on these results, a noise equivalent absorption (NEA) of 8.74 × 10−5 Hz−1/2 was achieved.
600 ppmv CO was measured continuously for ∼2 h with a sampling interval of 1 s, and 7000 data points were obtained. The 2f peak signal were used for concentration inversion, and the concentration sequence of the 7000 points and the corresponding histogram are shown in Fig. 16(a). The histogram reflects the normal distribution of the measured CO concentration around the mean value, and a Gaussian fit was applied to the results with an R-square of 0.999. It can be seen that the data presents a good Gaussian distribution, with a half-width at half-maximum (HWHM) of 4.72 ppmv. Fig. 16(c) shows the concentration sequence and corresponding histograms of CH4 measured continuously for about 2-h at a concentration of 2 ppmv. The histogram was fitted with a Gaussian distribution, resulting in a HWHM of 0.46 ppmv and an R-squared of 0.999.
Fig. 16. (a), (c) Concentration sequence and histogram of CO and CH4 continuous measurements. (b), (d) Allan deviation corresponding to the concentration sequences of CO and CH4.
The Allan-Werle deviation was used to further analyze the stability and detection accuracy of the system. The Allan deviation results corresponding to the concentration sequences of successive measurements of CO and CH4 are shown in Fig. 16(b) and 16(d). As can be seen from the figures, the MDL for CO and CH4 were 7.3 ppmv and 0.24 ppmv, respectively, at an integration time of 10 s. In addition, the Allan deviations for CO and CH4 was proportional to (1/τ)1∕2 as the integration time increases before 711 s and 245 s. Therefore, 711 s and 245 s were used as the optimal integration time, corresponding to detection limits of 0.07 ppmv and 0.008 ppmv, respectively.
Table 2 gives a comparison between the sensor proposed in this paper and some advanced TDLAS-based gas sensors. In order to improve the comparability, the parameters for CH4 detection were selected to compare with other CH4 sensors. Although the sensor proposed in this paper does not stand out significantly enough in terms of MDL and NEA, in the future, targeting stronger gas absorption lines with mid-infrared lasers [44], increasing laser power through erbium-doped fiber amplifiers [45], or using a wide response bandwidth and small-sized quartz tuning fork instead of the used photodetector [46] can further improve the sensor's performance.
Table 2. Comparison of the proposed sensor with some advanced TDLAS-based gas sensorsa
4.3 Atmospheric CH4 concentration measurement
In order to verify the accuracy and stability of the sensor, 3-day real-time and continuous monitoring of atmospheric CH4 was carried out in front of Laser Research Institute of Qufu Normal University, and shown in Fig. 17. It is noteworthy that before each ventilation cycle, the air was filtered and dried using a Polytetrafluoroethylene (PTFE) tube containing silica particulate matter and filtering cotton to avoid contamination of the DSPC-MPC by atmospheric substances such as H2O and aerosols. From the measurement results, it can be seen that the CH4 concentration shows a good cyclic variation during the measurement period, and it reaches a high level in the morning from 9-12 a.m. and a low level in the early morning from 1-3 a.m, which may be caused by human activity. CH4 concentrations measured over the three days varied from 1.72-2.48 ppmv, 1.70-2.49 ppmv, and 1.68-2.49 ppmv, with mean values of 2.08 ppmv, 2.09 ppmv, and 2.07 ppmv, respectively.
5. Conclusion
The simulation and design of a novel DSPC-MPC was presented and applied to the development of a multi-component gas sensor, which has a simpler structure and lower cost while realizing 129 m, 107 m, 85 m, 63 m and 40 m EOPLs adjustable compared with the traditional double concave spherical mirror Herriott cell and three-sided mirror White cell. Simultaneous detection of CO and CH4 were accomplished using two DFB lasers with center wavelengths of 1567 nm and 1653 nm, respectively, target the double spot-ring with EOPLs of 64.5 m each. A simple and portable digital PID laser frequency stabilization system was developed based on LabVIEW platform, which can stabilize the laser frequency within $\sim$±30.3 MHz for a long time. The linear responses of the DSPC-MPC gas sensor were analyzed with different concentrations of CO and CH4, and the linear fits were all ∼0.999. The MDLs for CO and CH4 were calculated to be 95.6 ppmv and 0.066 ppmv, respectively, by comparing the 2f signal values of the absorbed and unabsorbed portions. The Allan deviation analysis corresponding to the concentration sequences of ∼2 h continuous measurements of CO and CH4 indicated that the MDLs of 0.07 ppmv and 0.008 ppmv at integration times of 711 s and 245 s, respectively. The real-time monitoring of CH4 concentration for three days further validated the effectiveness and practicality of the sensor. The sensor can be more widely used in various fields such as industrial safety, medical diagnosis, and aerospace.
Funding
Urgent Talents research program of Shandong Province (2023TT2308); Key Research and Development Plan of Ji Ning (20231103); Natural Science Foundation of Shandong Province (ZR2018MD015); Key Technology Research and Development Program of Shandong (2017GSF17125); National Natural Science Foundation of China (11104160).
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. M. Rigby, S. A. Montzka, R. G. Prinn, et al., “Role of atmospheric oxidation in recent methane growth,” Proc. Natl. Acad. Sci. U.S.A. 114(21), 5373–5377 (2017). [CrossRef]
2. I. Jonquières and A. Marenco, “Redistribution by deep convection and long-range transport of CO and CH4 emissions from the Amazon basin, as observed by the airborne campaign TROPOZ II during the wet season,” J. Geophys. Res. 103(D15), 19075–19091 (1998). [CrossRef]
3. H. Schaefer, S. E. M. Fletcher, C. Veidt, et al., “A 21st-century shift from fossil-fuel to biogenic methane emissions indicated by 13CH4,” Science 352(6281), 80–84 (2016). [CrossRef]
4. X. Lan, E. G. Nisbet, E. J. Dlugokencky, et al., “What do we know about the global methane budget? Results from four decades of atmospheric CH4 observations and the way forward,” Phil. Trans. R. Soc. A. 379(2210), 20200440 (2021). [CrossRef]
5. S. Iwaszenko, P. Kalisz, M. Słota, et al., “Detection of natural gas leakages using a laser-based methane sensor and UAV,” Remote Sensing 13(3), 510 (2021). [CrossRef]
6. L. Shi, J. Wang, G. Zhang, et al., “A risk assessment method to quantitatively investigate the methane explosion in underground coal mine,” Process Saf. Environ. Prot. 107, 317–333 (2017). [CrossRef]
7. O. Anisimov, “Potential feedback of thawing permafrost to the global climate system through methane emission,” Environ. Res. Lett. 2(4), 045016 (2007). [CrossRef]
8. T. Stocker, “Implications of climate science for negotiators,” Ferdi 31–48 (2015).
9. D. T. Shindell, G. Faluvegi, N. Bell, et al., “An emissions-based view of climate forcing by methane and tropospheric ozone,” Geophys. Res. Lett. 32 (2005).
10. M. Ak-Bhd, “WMO Greenhouse Gas Bulletin,” WMO: Geneva 15, 2078–0796 (2020).
11. X. Lan, S. Basu, S. Schwietzke, et al., “Improved Constraints on Global Methane Emissions and Sinks Using δ13C-CH4,” Global Biogeochemical Cycles 35(6), e2021GB007000 (2021). [CrossRef]
12. F. A. N. Xing, “The analysis of COP28 Global Stocktake outcome and global climate governance prospects,” Advances in Climate Change Research (2024).
13. R. Motterlini and L. E. Otterbein, “The therapeutic potential of carbon monoxide,” Nat. Rev. Drug Discovery 9(9), 728–743 (2010). [CrossRef]
14. M. R. Dent, J. J. Rose, J. Tejero, et al., “Carbon Monoxide Poisoning: From Microbes to Therapeutics,” Annu. Rev. Med. 75(1), 337–351 (2024). [CrossRef]
15. L. Eichhorn, M. Thudium, and B. Jüttner, “The Diagnosis and Treatment of Carbon Monoxide Poisoning,” Dtsch Arztebl Int. 115, 863 (2018). [CrossRef]
16. L. Liu, Q. Zhuang, Q. Zhu, et al., “Global soil consumption of atmospheric carbon monoxide: an analysis using a process-based biogeochemistry model,” Atmos. Chem. Phys. 18(11), 7913–7931 (2018). [CrossRef]
17. K. D. Bartle and P. Myers, “History of gas chromatography,” TrAC, Trends Anal. Chem. 21(9-10), 547–557 (2002). [CrossRef]
18. C. Dodeigne, L. Thunus, and R. Lejeune, “Chemiluminescence as diagnostic tool. A review,” Talanta 51(3), 415–439 (2000). [CrossRef]
19. M. Pumera, S. Sánchez, I. Ichinose, et al., “Electrochemical nanobiosensors,” Sens. Actuators, B 123(2), 1195–1205 (2007). [CrossRef]
20. Y. Zhao, Y. Qi, H. Lut Ho, et al., “Photoacoustic Brillouin spectroscopy of gas-filled anti-resonant hollow-core optical fibers,” Optica 8(4), 532 (2021). [CrossRef]
21. S. D. Russo, A. Zifarelli, P. Patimisco, et al., “Light-induced thermo-elastic effect in quartz tuning forks exploited as a photodetector in gas absorption spectroscopy,” Opt. Express 28(13), 19074 (2020). [CrossRef]
22. G. Li, Y. Wu, Z. Zhang, et al., “WMS based dual-range real-time trace sensor for ethane detection in exhaled breath,” Optics and Lasers in Engineering 159, 107222 (2022). [CrossRef]
23. S. Qiao, Y. Ma, Y. He, et al., “Ppt level carbon monoxide detection based on light-induced thermoelastic spectroscopy exploring custom quartz tuning forks and a mid-infrared QCL,” Opt. Express 29(16), 25100 (2021). [CrossRef]
24. G. Li, X. Zhang, Z. Zhang, et al., “A mid-infrared exhaled carbon dioxide isotope detection system based on 4.35 µm quantum cascade laser,” Opt. Laser Technol. 152, 108117 (2022). [CrossRef]
25. Y. Wang, Z. Zhao, Q. Luo, et al., “An improved residual oxygen detection method by using axis-sectional multi-reflection within optical-length-limited pharmaceutical vials,” Measurement 222, 113667 (2023). [CrossRef]
26. D. Herriott, H. Kogelnik, and R. Kompfner, “Off-Axis Paths in Spherical Mirror Interferometers,” Appl. Opt. 3(4), 523–526 (1964). [CrossRef]
27. J. U. White, “Long Optical Paths of Large Aperture,” J. Opt. Soc. Am. 32(5), 285–288 (1942). [CrossRef]
28. H. Chang, S. Feng, X. Qiu, et al., “Implementation of the toroidal absorption cell with multi-layer patterns by a single ring surface,” Opt. Lett. 45(21), 5897–5900 (2020). [CrossRef]
29. S. Chernin and E. Barskaya, “Optical multipass matrix systems,” Appl. Opt. 30(1), 51–58 (1991). [CrossRef]
30. H. Sun, Y. Ma, Y. He, et al., “Highly sensitive acetylene detection based on a compact multi-pass gas cell and optimized wavelength modulation technique,” Infrared Phys. Technol. 102, 103012 (2019). [CrossRef]
31. Y. H. Liu, S. D. Qiao, C. Fang, et al., “A highly sensitive LITES sensor based on a multi-pass cell with dense spot pattern and a novel quartz tuning fork with low frequency,” Opto-Electron. Adv. 7(3), 230230 (2024). [CrossRef]
32. J. Barry McManus, Mark S. Zahniser, and David D. Nelson, “Dual quantum cascade laser trace gas instrument with astigmatic Herriott cell at high pass number,” Appl. Opt. 50(4), A74–A85 (2011). [CrossRef]
33. Z. Li, Q. Zhang, Z. Wang, et al., “A highly sensitive low-pressure TDLAS sensor for detecting dissolved CO and CO2 in transformer insulating oil,” Opt. Laser Technol. 174, 110622 (2024). [CrossRef]
34. N. Yu, P. Genevet, M. A. Kats, et al., “Light propagation with phase discontinuities: generalized laws of reflection and refraction,” Science 334(6054), 333–337 (2011). [CrossRef]
35. Pramode Ranjan Bhattacharjee, “Exhaustive study of reflection and refraction at spherical surfaces on the basis of the newly discovered generalized vectorial laws of reflection and refraction,” Optik 123(5), 381–386 (2012). [CrossRef]
36. P. R. Bhattacharjee, “The generalized vectorial laws of reflection and refraction,” Eur. J. Phys. 26(5), 901–911 (2005). [CrossRef]
37. Available online: https://www.spectraplot.com/absorption (accessed on March 17, 2024).
38. Wolfgang Gurlit, R. Zimmermann, C. Giesemann, et al., “Lightweight diode laser spectrometer CHILD (Compact High-altitude In-situ Laser Diode) for balloonborne measurements of water vapor and methane,” Appl. Opt. 44(1), 91–102 (2005). [CrossRef]
39. D. M. Sonnenfroh, R. Wainner, M. G. Allen, et al., “Interband cascade laser–based sensor for ambient CH4,” Opt. Eng 49(11), 111118 (2010). [CrossRef]
40. J. B. McManus, J. Barry, D. D. Nelson, et al., “Pulsed quantum cascade laser instrument with compact design for rapid, high sensitivity measurements of trace gases in air,” Appl. Phys. B 92(3), 387–392 (2008). [CrossRef]
41. K. Liu, L. Wang, T. Tan, et al., “Highly sensitive detection of methane by near-infrared laser absorption spectroscopy using a compact dense-pattern multipass cell,” Sens. Actuators, B 220, 1000–1005 (2015). [CrossRef]
42. Lei Dong, L. Li, N. P. Sanchez, et al., “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]
43. Lance E. Christensen, Christopher R. Webster, and Rui Q. Yang, “Aircraft and balloon in situ measurements of methane and hydrochloric acid using interband cascade lasers,” Appl. Opt. 46(7), 1132–1138 (2007). [CrossRef]
44. Y. Ma, Y. Tong, Y. He, et al., “Compact and sensitive mid-infrared all-fiber quartz-enhanced photoacoustic spectroscopy sensor for carbon monoxide detection,” Opt. Express 27(6), 9302–9312 (2019). [CrossRef]
45. C. Zhang, S. Qiao, Y. He, et al., “Trace gas sensor based on a multi-pass-retro-reflection-enhanced differential Helmholtz photoacoustic cell and a power amplified diode laser,” Opt. Express 32(1), 848–856 (2024). [CrossRef]
46. C. Fang, T. Liang, S. Qiao, et al., “Quartz-enhanced photoacoustic spectroscopy sensing using trapezoidal-and round-head quartz tuning forks,” Opt. Lett. 49(3), 770–773 (2024). [CrossRef]
