The structural parameters of the sensor have an impact on the detection performance of methane gas concentration and temperature. To optimize the sensing characteristics of the sensor, this paper discusses the main structural parameters of the sensor, including the diameters of the air holes d1, d2, and d3, the lattice spacing P, the thicknesses of the metal coated by the two measurement channels t1 and t2, and the thicknesses of the sensitive film t3 and t4.
Figure 6 illustrates the impact of varying the diameter of the air holes (d1) on the loss peaks and resonance wavelengths of the two channels. When d1 is varied from 1.1 to 1.3 µm, the resonance wavelengths of both channels undergo a red shift. Additionally, the loss peaks initially increase and then decrease. However, the effect on channel Ⅰ is more significant than that on channel Ⅱ. This is because an increase in the diameter of the air hole reduces the refractive index of the fiber core, which alters the phase-matching conditions. This results in the core mode and the SPP mode needing to reach a phase match at longer wavelengths. Furthermore, the enhancement of the swift wave and the SPW interaction causes more energy transfer from the fiber core to the SPP mode. Nevertheless, as the diameter of the d1 continues to increase, a greater proportion of light is leaked from the fiber core and distributed around the air hole, rather than near the metal layer. This results in a weakening of the strength of the SPR and a reduction in the loss peak. The diameter of d1 was optimized to 1.2 µm in consideration of the requisite sensitivity and detection accuracy.
Figure 7 illustrates the impact of varying the air pore diameter (d2) on the system. For channel Ⅱ, as d2 is varied from 1.4 to 1.6 µm, the resonance wavelength undergoes a redshift, accompanied by an initial increase and subsequent decrease in the loss peak. This phenomenon can be attributed to the same underlying mechanism as observed in the variation of d1. For channel Ⅰ, the resonance wavelength is blue-shifted, and the loss peak decreases. This is since the air holes are primarily aligned in proximity to the channel Ⅰ. An increase in d2 impedes the penetration of the swift wave, resulting in a reduction in fiber core energy transferred to the SPP mode. Consequently, the resonance strength is diminished. Here, the optimal diameter for d2 is determined to be 1.5 µm.
The diameter of the outermost air hole of the PCF, designated as d3, exerts a profound influence on the sensor performance. As illustrated in Fig. 8, the resonance wavelengths of both channels exhibit a blueshift when the diameter is varied from 1.7 to 1.9 µm. Additionally, the resonance peak is observed to disappear at d3 = 1.9 µm for channel Ⅰ. This is because d3 is located in the outermost layer, close to the metal layer. Consequently, its alteration has a smaller impact on the core mode's effective refractive index and a more significant effect on the SPP mode. This results in a shift in the phase-matching condition between the core mode and the SPP mode, enabling resonance at a shorter wavelength. Conversely, when d3 becomes excessively large, it diminishes the interaction between the SPP mode and the swift wave, leading to a reduction in the resonance peak. In consideration of these factors, the optimal value for d3 was determined to be 1.7 µm.
The gap between two adjacent air holes is also optimized, and Fig. 9 shows the loss characteristics of the proposed sensor P in the range of 2.9 variation to 3.1 µm. It can be seen that the P has no effect on the resonance wavelength of channel Ⅰ, and there is only a change in the loss peak, which indicates that the variation of P does not change the phase-matching condition of its core mode with the SPP mode. For channel Ⅱ, its resonance wavelength is red-shifted, and the loss first increases and then decreases, which is mainly because the P becomes larger to make the effective refractive index of the fiber larger, and the propagation constant of the SPW decreases, which makes the resonance wavelength move to the direction of longer wavelengths. For the loss value, the increase of lattice spacing makes more swift wave generated to interact with the SPP mode, but as the P continues to increase, the energy of optical leakage increases, and most of the energy has been scattered out before reaching the SPP mode. The comprehensive consideration selects P = 3.0 µm as its optimum value.
The thickness of the metal, as a necessary material for SPP excitation, also deeply affects the characteristics of the sensor. As shown in Fig. 10(a) and (b), the loss characteristics of the two channels with the thickness of the metal film, respectively. t1 mainly affects channel Ⅰ, and either too large or too small a thickness leads to the disappearance of the SPR phenomenon. The SPW generated by the too-thin metal film is not strong enough to excite and support the SPR. Too thick metal film increases the damping effect, resulting in increased energy dissipation, and most of the incident light is reflected or absorbed, which cannot reach the interface between the metal and the medium, resulting in insufficient SPW excitation. t2 mainly affects channel Ⅱ, whose SPR is the strongest at a thickness of 40 nm, and the loss peak decreases when t2 = 30 or 50 nm, for the same reason as the concentration channel. Considering the above reasons, t1 = 30nm and t2 = 40nm are chosen as the optimal values for the two channels, respectively.
v
Figure 10 (a) Variation of channel Ⅰ with Ag thickness loss characteristics (b) Variation of channel Ⅱ with Ag thickness loss characteristics
In order to achieve specific measurements of methane concentration and temperature, a sensitive film was used, the thickness of which is also discussed in this paper, as shown in Fig. 11(a) and (b). The t3 variation does not affect channel Ⅱ, which also indicates its insensitivity to temperature. It has less effect on channel Ⅰ, and its loss value increases and then decreases, but the change between the thicknesses of the adjacent films is not large, which indicates that when producing the sensors, their thicknesses are more tolerant and more achievable to produce, and t3 = 100nm is chosen as its optimum value. The t4 change has no effect on channel 1, while the temperature-induced loss peak first increases and then decreases with its thickness, and to produce a more pronounced resonance peak, we chose t4 = 1.0 µm as its optimum value.
After analyzing and optimizing the structure of the sensor, the main structural parameters of the sensor are finally determined as follows: d1 = 1.2 µm, d2 = 1.5 µm, d3 = 1.8 µm, t1 = 30 nm, t2 = 40 nm, t3 = 100 nm, t4 = 1.0 µm, and P = 3.0 µm. Under this structure, the proposed sensor demonstrates better sensing performance.
This paper compares and analyzes the sensing performance of the methane concentration temperature sensor proposed with that of existing sensors of the same type, as shown in Table 1. The initial two sensors employ the SPR sensing principle. Despite their concentration detection range being marginally broader than that of the sensor proposed in this paper, their sensitivity is inferior to that of the proposed sensor. The sensing principle employed by Ref. 16 is that of interference, which is less effective than the SPR type of sensor. Fiber optic sensors that are capable of measuring both methane concentration and temperature are scarce. While Ref. 22 can fulfill the requisite criteria, its average sensitivity is inferior to that of the proposed sensor. Consequently, the proposed sensor exhibits a notable advantage.
Table 1
A comparative analysis of the proposed sensor with recently reported sensors of the same type.
Ref. | Sensing Method | Concentration Range (%) | WS(nm/%) | Temperature Range(℃) | WS(nm/℃) |
[1] [12] | SPR SPR | 0 ~ 3.5 0 ~ 3.5 | 5.54 Ave. 8 Max. | \ \ | \ \ |
[13] | Sagnac Interference | 0 ~ 3.0 | 4.6 Ave. | \ | \ |
[16] [17] [19] Proposed | SPR SPR SPR SPR | 0 ~ 3.5 0 ~ 3.0 0.5 ~ 3.0 0.5 ~ 3.0 | 9.88 Ave. 14.19 Ave. 4.84 Ave. 96.57 Ave. | \ \ 10 ~ 30 0 ~ 50 | \ \ 0.58 6.86 |