Comparison of the sensing mechanisms and capabilities of three functional materials surface-modified TFBG sensors

Functional materials have attracted considerable attention recently, with the use of such materials in various types of sensors being especially notable. At present, there is considerable potential for the further development of their application in optical fiber sensors for strain monitoring, magnetic field and refractive index (RI) sensing,¹ humidity sensing,^2,3 biochemical sensing,⁴ gas sensing,⁵ and the sensing of glucose concentrations in vitro.⁶ The fundamental concept for such sensors primarily relies on using the light intensity, phase, wavelength, and polarization mode in the optical fibers for detection.⁷ At present, fiber sensors used for physical detection have been developed using fiber Bragg grating (FBG) structures fabricated on optic fibers. Furthermore, fiber optic sensors with tilted fiber Bragg grating (TFBG)⁸ structures have been further developed for application in various physical phenomena, and RI detection sensors were used to enhance signal sensitivity with controllable cross-sensitivities.⁹ 

Relative humidity (RH) is an important environmental parameter and the most basic abiotic factor. Various functional material coatings, such as those consisting of polymers, self-assembled compounds, and porous media, have been coated on the TFBG to increase the utility of TFBG sensors for RH sensing applications.¹⁰ Miao et al.¹¹ proposed an RH sensor based on the TFBG that was fabricated by utilizing polyvinyl alcohol (PVA) as the sensitive cladding film, as the RH increase in the PVA coating would result in the reduction of the refractive index of the fiber. The transmission power of the resulting TFBG has two different humidity ranges, and the sensitivity for the humidity ranges reached 2.52 dBm/% RH (at 20–74% RH) and 14.947 dBm/% RH (at 74–98% RH), respectively. In 2015, Lin et al.² demonstrated a fiber-optic humidity sensor that had been fabricated by coating the fiber Bragg grating (FBG) with a polyimide film. The characteristics of the sensor were improved by controlling the thickness and the degree of imidization of the polyimide. In the following year, Wang¹² introduced a sensor coated with a graphene oxide (GO) film to find out the amplitudes of the cladding mode resonances of the TFBG as they varied with the water sorption and desorption processes. They obtained a sensitivity of 0.129 dB/% RH in the cladding mode. In addition, researchers have widely used poly(3,4-ethylenedioxythiophene)/poly(styrenesulfonate) (PEDOT:PSS) to detect the physical properties.¹³ PEDOT:PSS will affect optical and electrical reactions due to the changes in humidity.¹⁴ This polymer is coated on optical and electronic components so that when the humidity changes, its optical or electrical changes are sensed.

To combine the advantages of the tilted optical fiber grating and a polymer coated as a smart material, a sulfated polyaromatic hydrocarbon derivative termed PAHP4 was designed through chemical engineering that configures the SO₃⁻ functional group and polyaromatic group into a simple polymer configuration that offers high sensitivity, high stability, and a repeatable sensing range for humidity sensing. Thus, a TFBG sensor coated with PAHP4 could be utilized in real-time RH monitoring and for evaluation in normal or extremely humid environments.

When the grating is perpendicular to the direction of the optical fiber core axis, mode coupling occurs between the core and the cladding of the fiber. The reflection resonance wavelength of the TFBG appears with transmission loss, and the Bragg resonance condition is generated by the coupling between the propagating core mode and the counter-propagating core mode. Therefore, the resonance wavelength of a tilt angle grating satisfying the Bragg condition is expressed by the following equation [Eq. (1)]:¹⁵ 

where Λ is the grating period and n_eff,core is the effective refractive index of the grating in the fiber core. For the TFBG, $Λcos θ$ can be used to modify the tilted angle of the grating in the propagating core mode and the counter-propagating core mode and the grating period along the fiber axis.¹⁶ Due to the tilt angle of the grating, a part of the light from the propagating core mode is coupled to the counter-propagating core mode and the resonance wavelength of the cladding mode can be determined by

where $Λg=Λcos θ$ and n_eff,clad,i is the effective refractive index of the ith cladding mode. The measurement results of a TFBG refractive index sensor can be verified by the resonance wavelength shift of the cladding mode using a fixed grating period, tilted angle, and Bragg period. In this study, according to Eqs. (1) and (2), a redshift occurs when the effective refractive index is increased. This phenomenon, in turn, can be used to measure humidity. More specifically, we can monitor humidity by measuring the wavelength shift and obtaining the correlation between the wavelength shift and the Bragg period.

A germanium-doped single-mode fiber (No. PS 1250/1500 purchased from FIBERCORE Corp.) was cleaned with alcohol after a wire stripper was employed to remove 5 cm of the protective layer of the optical fiber from its middle section, and then, the fiber was secured onto a triaxial micro-displacement platform. The TFBG was fabricated by phase-mask direct-writing 5 mm gratings into the germanium-doped single-mode fiber using an excimer KrF laser (COHERENT; Xantos XS 500 Lübeck, Germany). TFBG photo-writing was carried out using a phase-mask rotated to a tilt angle of 10°, with a 12 mJ excimer laser applied to the fiber. During the excimer laser treatment process, the resulting gratings were evaluated in real-time by tracking the spectra of the fiber using an optical spectrum analyzer (OSA). The experimental architecture is shown in Fig. 1.

Moderated lessening of the cladding diameter from 125 µm to 20 µm was accomplished by chemical etching of buffered oxide etch (BOE).

The polyaromatic hydrocarbon derivative PAHP4 (the chemical structure of which is shown in Fig. 2) with sulphofication was synthesized via a Suzuki palladium-catalyzed coupling chemical method using bis(di-tert-butyl(4-dimethylaminophenyl)phosphine) dichloropalladium(II) ((A-^taPhos)₂PdCl₂, Pd(amphos)Cl₂) that was purchased from Sigma-Aldrich.

The sensors were fabricated using functional thin films prepared by using a dip-coating technique. Prior to graphene oxide coating deposition, the employed TFBG was subjected to a surface modification technique accomplished by immersion in 5% (v/v) HNO₃ solution at room temperature for 1 h to remove the contaminant and then entirely rinsed with ethanol and deionized (DI) water several times for the grating region was soaked in the GO suspension for 3 min and was then dried in an oven at 100 °C for 10 min. The TFBG was coated with the PEDOT:PSS and PAHP4 functional material in the same way.

The sensors covered with GO, PEDOT:PSS, or PAHP4 were placed in a humidity chamber, into which a humidifier was placed to increase the humidity, after which pure N₂ gas at a concentration of 95% was injected to refresh the humidity chamber for the next cycle. More specifically, an increase in the humidity was accomplished by injecting H₂O vapors through the humidifier into the humidity chamber at the room temperature of 26 °C. The measurement range for the increasing humidity was from 20% RH to 80% RH, with measurements taken at every 10% RH interval. Figure 3 is a schematic diagram of the humidity sensor evaluation system. The system used consisted of an OSA (Anritsu, MS9710C), a superluminescent diode (SLD), a humidity controller, a humidifier, and two precision translation stages. The fiber was kept straight and stationary by fixing its two ends onto the two-stage platform.

The surfaces of TFBG sensors, each of which had a diameter of 20 µm, were, respectively, coated with three functional materials. The sensitivity levels of the three functional sensing materials for humidity detection and their effects on the TFBG fiber optic sensors were compared. Humidity measurements of RH from 20% to 80% RH were performed, and three cycles were performed to analyze the results for the conjugate spectrum of each of the fiber optic sensors. Figure 4 shows micro-scan images of the three functional layers coated on the surfaces of the TFBG sensors. Figure 4(a) is an image of GO on the surface of a sensor, showing a fish scale-like graphite and metallic luster profile, indicating a sheet-like stack with a thickness of ∼0.125 µm. Figure 4(b) shows the surfaces of TFBG sensors coated with PEDOT:PSS, and the surface of the given optical fiber had a dark blue coating with the thickness of about 0.013 µm. As shown in Fig. 4(c), the surface of the optical fiber coated with the layer of PAHP4 appeared glossy and the surface of the optical fiber had significant colorful mural-like traces. The thickness of the PAHP4 sensing layer was about 0.05 µm. The degree of thickness for the three material layers was GO > PEDOT:PSS > PAHP4.

A TFBG sensor was coated with GO for humidity measurements. As shown in the inset of Fig. 5, for the wavelength ranging from 1545.4 nm to 1546.3 nm, in the core mode, the sensitivity in the resonance wavelength was 0.00167 nm/% RH and the linearity was as high as 0.982 (R²), resulting in the resonance wavelength having a linear relationship with the change in humidity. The wavelength gradually shifted to longer when the humidity increased. The optical fiber was stretched as H₂O penetrated into the GO. As mentioned above, GO consisted of sheet-like stacks and there were abundant hydrophilic hydroxyl (–OH), carboxyl (–COOH) groups, all of which are good at forming hydrogen bonds with H₂O. Attracting hydroxyl groups attached to the GO sheet maintained a relatively interlayer distance and formed a two-dimensional GO nanochannel, allowing the low-friction flow of a single layer of H₂O.

In the cladding mode, as shown in the inset, the wavelength values range from 1511.7 nm to 1512.9 nm. The wavelength redshifted by 0.113 nm (as shown in the inset of Fig. 5) when the humidity was increased because the H₂O super-permeable channels of the fiber were stretched and they interacted with the thicker GO layer, which caused an external complex RI.¹⁷ Hydrophilic –OH and –COOH groups provide a super-permeable nanochannel for H₂O and accelerate the swift interaction between H₂O and the GO layer. The optical fiber will be affected by the environmental complex RI of the sensing layer, which will have an influence on the fiber core and cladding light propagation. The distance between GO layers causes changes in the RI when H₂O penetrates the GO layers, and the environmental RI will cause a wavelength shift in the cladding mode without affecting the transmission power.

By applying PEDOT:PSS as a surface-modified material for the humidity sensor in the process of H₂O absorption and dehumidification, the spectrum can be tracked and analyzed. In the cladding mode, it was indicated that the wavelength redshifted when the humidity was increased, the average drift was 0.229 nm accompanied with a gradual decrease in transmission loss, and the average transmission loss was about 0.908 dB [Fig. 6(a)]. The sensor has a wavelength sensitivity of 0.00367 nm/% RH (R² = 0.912); the transmission loss sensitivity was 0.015 dB/% RH (R² = 0.867). In the core mode, there is no shift in the wavelength when the humidity is changed, while the transmission loss decreases by degrees. The average loss over the three cycles decreases by 0.939 dB at 1544 nm, and the sensitivity of the loss was 0.0153 dB/% RH [Fig. 6(b)]. PEDOT:PSS is an aqueous dispersion material, that is, water molecules can easily enter the polymer forming hydrogen bonds with sulfate ions (–SO₃⁻) and sulfonic acid (−SO₃H), allowing it to cause a decrease in the RI when the humidity increases.

The humidity analysis results of the PAHP4 surface-modified fiber are shown in Fig. 7(a). When the humidity was increased, the wavelength redshifted and the transmission loss reduced in the cladding mode of the TFBG. For the Bragg mode, since a change in the environmental RI will not affect the core mode, there is no drift phenomenon for the Bragg wavelength peak. This result can also be used as a criterion for determining the cladding mode, and it can also indicate that the dew point temperature of the relative humidity has consistency.

We found that a more valuable average transmission loss sensitivity as high as 0.004 29 dB/% RH (R² = 0.848) was achieved when PAHP4 was used as the sensing layer. The sensor produced by coating PAHP4 on the TFBG had an excellent linear relationship of transmission loss to humidity variation. The cladding mode results of the humidity experiment are shown in Fig. 7(b). At a humidity of 20–60% RH, the average redshift over the three cycles was 0.003 17 nm. This differs from the redshift for a humidity of 60–80% RH for which the red-shifted distance rapidly increased, and the average shift was 0.0193 nm. The wavelength sensitivity of the sensor was 0.003 17 nm/% RH in the low humidity region (20–60% RH) and was 0.0193 nm/% RH in the high humidity region (60–80% RH). An average wavelength sensitivity of 0.0112 nm/% RH and loss sensitivity of about 0.002 72 dB/% RH were observed. The polymer material PAHP4 is a newly designed and synthesized material, and it has a similar sensing mechanism as PEDOT:PSS and GO. Upon interacting with all of them, H₂O molecules will enter the sensing interlayers and form hydrogen bonds with sulfate ions (SO₃⁻). The particular materials contained in the structures of the molecules create more water-absorbing functional groups (SO₃⁻), and hydrophobic polyaromatics are introduced. The optical fiber will be affected by the sensing mechanism of the sensing layer, which can be credited to the permeability between polymer layers. PAHP4 can also be used as a humidity sensor that has a better response and sensitivity. PAHP4 will be affected by the sensing mechanism, which can be credited to the permeability between polymer layers, as illustrated in Fig. 8. As the humidity increases, the RI of PAHP4 increases. The spectrometers used for tracking demonstrate that the resulting sensor has high sensitivity to humidity variation.

The effects of each type of sensing layer were compared when the humidity varied (Table I). The resonance wavelength of the core mode will shift due to the slight expansion of the sensor when GO is used as the sensing layer. There is no shift in the resonant wavelength when PEDOT:PSS or PAHP4 is used as the sensing layer. This phenomenon is attributed to the fact that the resonance wavelength of the core mode is not affected by the environmental RI.

The sensitivity comparison of functional materials was conducted by analyzing the resonance wavelength shifts of the cladding mode. Upon introducing SO₃H successfully into PAHP4, the sensor showed the best sensitivity and the sensitivity levels of the resonance wavelengths in the cladding mode were ranked as PAHP4 > PEDOT:PSS > GO. The PAHP4 develops the highest sensitivity for the surface-modified TFBG based humidity sensing, with a sensitivity of 0.0112 nm/% RH.

A highly sensitive, stable, and repeatable sensing material is used for the development of the high sensitivity surface-modified TFBG as a humidity sensor. PAHP4 was successfully designed and synthesized via chemical engineering. The water molecule can enter between the polymer chains and form hydrogen bonds with sulfate ions (SO₃⁻), which can be wetted by H₂O or used as a water-absorbing material. In the structure of PAHP4, abundant water-absorbing functional groups are created and this improves the permeability between the polymer layers, which could be credited to the introduction of hydrophobic polyaromatics. We successfully introduced the molecule containing the sulfonic acid (–SO₃H) functional and polyaromatic group into PAHP4, and the TFBG sensor coated with PAHP4 showed the highest sensitivity in the cladding mode of the TFBG. The average wavelength sensitivity was 0.0112 nm/% RH, and the linearity was 0.912.

The data that support the findings of this study are available within the article.

This study was financially supported by a grant from the Ministry of Science and Technology (Grant No. MOST 107-2221-E-992-043-MY3).

The authors declare no conflicts of interest.