Resonant Light Scattering Toward Optical Fiber Humidity Sensors

The deposition of tetrakis (4-sulonatophenyl) porphyrin (TPPS) thin film on optical fibers presents many possibilities for sensing applications. The J-form aggregation with a narrow and sharp spectral feature at about 490 nm and its sensitivity to humidity have been discussed; a fast change of wavelength occurs according with variation in the humidity level. The reproducibility and high sensitivity of TPPS-coated fibers, along with the capabilities of optical fibers, suggest the device as a good candidate for humidity sensing in harsh environments.


Introduction
The measurement and control of humidity levels in different environments are important in many fields, including agriculture, environmental science, and the oil and gas industries. To ensure safety, increase the lifetime of materials, and improve irrigation in agricultural fields, different application-specific techniques for humidity control with different sensitivities are used.
Optical fiber sensors are small in size, insensitive to electromagnetic interference, and capable of multi-parameter sensing, and are thus good candidates for environmental monitoring [1]. Chemical optical fiber sensors, among the most intriguing applications of optical fibers, have attracted significant research attention. Chemical sensors cannot be efficient and valuable without the right selectivity. Fiber optics are potentially good candidates for selective and sensitive sensors by combining the optical transducer with active materials selected and arranged to detect specific chemicals [2,3].
Recent investigations into the deposition of TPPS molecules on optical fibers using UV lasers have shown new possibilities for optical sensing [30][31][32][33]. The TPPS molecule can switch between different aggregate forms in response to the presence of acidic or basic vapors, namely the J-and H-type aggregates, which present distinctive spectral features [34].
In the present work, the spectral features of the J-aggregate are investigated as a function of the local RH. This extends the functionality of thin TPPS layers integrated with optical fibers. The integration of such a simple and inexpensive humidity sensor with an optical fiber transmission line permits employment in internet-of-things (IoT) applications, in harsh environments and in irrigation as a component of intelligent agricultural engineering. In intelligent farms, many humidity sensors are required to transmit humidity data through the internet and thereby control farm irrigation.

Theoretical background
In the mid-1930s, Scheibe [35,36] and Jelley [37] independently discovered the appearance of a narrow absorption band with an increase in the concentration of pseudoisocyanine (PIC) dye in water. The narrow-band absorption is redshifted relative to the monomer band. This effect is ascribed to the optical excitation of formed aggregates, which are clusters of molecules of intermediate size between crystals and isolated molecules. Aggregates are formed by the dipole-dipole interaction of monomers [38,39]. Based on the spectral shift of the aggregate absorption peak relative to the maximum absorption of the monomer, the aggregates are classified as J-and H-types, corresponding to red and blue spectral shifts, respectively. J-and H-aggregates are further distinguished by the angle θ between the molecular transition dipole moment and the long aggregate axis. In some substances, such as PIC and TPPS, both J-and H-aggregates are found. The Jand H-aggregates of TPPS form in basic and acidic solvents, respectively [30][31][32][33], and show the angles 54.7 θ >  and 54.7 θ <  . Dipole-dipole interactions between two or more monomers in an aggregate unit cell [40,41] cause the coupling of excitons and vibrational levels of the monomers. The coupling coefficients depend on the intermolecular distance, the number of interacting molecules, and the angle between the transition dipole moment and the aggregate axis. In the absence of exciton coupling coefficients, the exciton states are degenerate, and only one maximum is observed in the absorption spectrum. The coupling of excitons causes energy level splitting, which introduces several peaks to the absorption spectrum. Both theoretical and experimental investigations show that the J-band appears with exciton coupling and without vibrational-mode coupling of monomers. In the J-aggregates, the energy transfer by excitons along the monomers in the cooperative excitonic state is sufficiently rapid to suppress vibrational relaxation before energy is transferred to the next monomer; therefore, the effects of vibrational level broadening are negligible. However, the H-aggregate absorption spectrum cannot show the absence of vibrational mode coupling. The J-aggregate shift is determined by the coupling strength, which is simply given as [41] , nm n m where nm V is the matrix element of the total dipole-dipole interaction operator V , and n denotes the position of the nth monomer in a periodic lattice with one monomer per unit cell. C is affected by environmental disturbances. Suitable methods can be employed to deposit Jor H-aggregate in the solid form. Experimental investigations have shown that the coupling strength C in J-aggregated TPPS varies with the environmental RH, with the absorption peak variation demonstrating this sensitivity. The Kramer-Korning relation also predicts that increased RH affects the refractive index of J-aggregated TPPS; thus, the reflectance coefficient also changes with RH. Variations of both the absorption frequency peak and reflectance versus RH can be employed in the design of new humidity sensors. In the following sections, such a humidity sensor design is proposed, and its capability is verified experimentally.

UV-induced TPPS deposition on the fiber tip
Various controllable coating techniques for porphyrin molecules have been reported for use on optical fiber tips [42,43]. In this section, we describe the development of the UV-induced deposition of TPPS molecules on optical fiber (200 μm core, Ocean Optics P200-2-UV-VIS) end [32,33]. Briefly, the fiber is cleaved by a standard cleaver (Xinyufei Fiber Optic Cleaver D6) and has been kept clean in dichloromethane (CH 2 Cl 2 ) TPPS molecular film deposition obtained by dipping the cleaved tip of the multimode fiber in a tetrabutylammonium solution in CH 2 Cl 2 with a concentration of 5 mol/L -10 mol/L, followed by irradiation with a He-Cd 325 nm 9.5 mW UV laser (Melles Griot) coupled to the fiber optics.

Sample preparation and characterization
We employ a nonstandard multi-mode ultraviolet/visible optical fiber with a 200 μm core prepared by a fiber cleaver, as described in Section 3, for TPPS deposition. The core and cladding diameters of the UV/VIS fiber are 200 μm and 220 μm, respectively. The optical fiber guides UV radiation to the fiber tip, thus stimulating the local precipitation of molecules for deposition on the fiber core. The multi-mode fiber is situated in a diluted solution (5 μmol/L − 10 μmol/L) of the tetrabutylammonium salt of TPPS in CH 2 Cl 2 during UV laser irradiation. A 9.5 mW He-Cd laser is used as the UV radiation source.
Investigating several irradiation durations shows that 60 min UV irradiation provides the best morphological results. The microscopic observation of the deposited distal end is shown in Fig. 1(c). The self-aligned thin film on the fiber core is detectable in Figs. 1(b) and 1(c), with the separation line matching the core-cladding border because of UV photochemical effects during deposition, as previously reported [32,33].
To characterize the TPPS thin film deposited on the fiber tip, a reflectance scattering spectroscopy is employed, comparing the scattering spectra obtained before and after TPPS deposition. Figure 2 shows a schematic of the experimental setup for the reflectance scattering spectroscopy. A visible light source in the range of 400 nm -800 nm with a nearly Gaussian distribution is used. The spectrum of the light source is saved for data normalization.
As shown in Fig. 2  The reflectance spectra of the uncoated and TPPS-coated fibers are shown in Fig. 3. Spectral analysis of the film shows differences in optical amplitude, but the presence of Soret bands at 420 nm and 490 nm [30], and characteristic of the J-aggregate, are not observable; no special feature is observed in the blue band.
As shown in Fig. 3, J-aggregate bands [28] are not observed in the reflectance spectrum. The absorption spectrum of TPPS J-aggregates is depicted in Fig. 4 [40]. Post-deposition treatment is performed to obtain the J-and H-aggregate bands in the reflectance spectrum [32]. The reflectance spectrum is observed at each step of treatment. In this preparation, all physical and chemical parameters are fully controlled. As detailed in Section 5, the temperature and humidity during the treatment are fixed at 22 ℃ and 80%, respectively. The post-deposition treatment is given by the following steps: (1) The thin film is exposed directly to the ammonia vapor for 2 min to form an H-aggregate film; (2) The deposited fiber is placed in air for 5 minutes; (3) The film is exposed to the hydrochloric acid for protonation. The remaining basic and acidic species react to produce salt and water on the thin film. Steps (1) to (3) are repeated until the resonance light scattering peak at about 490 nm, indicating J-aggregate formation, reaches its maximum intensity in agreement with the results reported elsewhere. The strength of the Soret band is low relative to that of the 490 nm band [30]. The J-and H-aggregates and corresponding energy levels [31] are presented in Fig. 5.
This treatment is previously reported for an acidand base-detecting sensor [32]. We examine the ability of the prepared J-aggregate TPPS-coated optical fiber sample to measure humidity [32,33,44].

Sensing characterization of humidity variations
The sensing capability of the TPPS-coated fiber tip to RH is explored by monitoring the changes in its reflected spectra. The experiment is performed in a controlled humidity chamber, in which both temperature and humidity are measured and controlled by an ALTEC TH135 sensor. The RH and temperature resolutions are 1% and 0.1 ℃, with the corresponding dynamic ranges of 0% to 99% and −100 ℃ to 200 ℃, respectively. A double control and backup sensor (Novus Logbox RHT) is used with the RH resolution and dynamic range of 0.1% and 0% to 100% and the temperature resolution and dynamic range of 0.1 ℃ and −40 ℃ to 80 ℃. The temperature is maintained at 22 ℃, and the desired RH value is obtained by controlling the humidity flow rate. The prepared fiber tip is located within the chamber as shown in Fig. 6. The reflectance spectrum measuring setup is the same as that shown in Fig. 2  Humidity promotes the formation of the J-aggregate [30,33], and the fiber is in the acidic (protonated) form. Therefore, the initial spectrum shows a peak at about 490 nm, attributed to the growth J-aggregate inducing resonant light scattering (RLS) at 490 nm [32] of approximately + 30% relative to the baseline. The RLS peak amplitude increases significantly because it is very narrow, with a bandwidth of about 12 nm; it reaches a maximum reflectance peak power almost 20% higher than the reference value presented in Fig. 3.
This peak indicates well-ordered J-aggregates; the functional film responds to RH by changing aggregate qualities; as noted previously, J-aggregate formation is favored by aqueous media [30,32,46].
The experimental setup allows sample exposure to a controlled environmental RH. Different RH levels are achieved within 1% precision. We study the reflectance spectrum around 490 nm under RH variations from 0% to 85%, at which the deposited J-aggregate film is stable on the fiber tip. The electrostatic binding of the TPPS J-aggregate to the fiber tip is low enough that the TPPS is washed away from the optical fiber tip under high RH. Therefore, the dynamic range of the proposed humidity sensor has a maximum limit of 85%, above which the coated thin film is destroyed. The reflectance spectra of the TPPS-coated film on the fiber end under different RH values are shown in Fig. 7. Both the refractive index and absorption coefficient depend on the environmental humidity. The coated thin film can be considered a low-quality factor random Fabry-Pérot (F-P) cavity structure [47]. The quasi-periodic variation of the reflected spectra in all curves of Fig. 7 arises from the low-quality-factor resonance frequency of the F-P structure.
No strict rule appears in the variation of spectral peak intensity with RH. As an example, the spectral peak intensities corresponding to 30%, 38%, and 50% RHs are equivalent, although the spectra are otherwise different. The peaks of the measured spectra depend on the number of experiments in each sequence, as well as the ages of given samples.
All experiments are performed at the room temperature and ambient pressure, and all experimental parameters except humidity are held constant. An RLS band from 488 nm to 493 nm is observed in Fig. 7. Because the vibrational and excitonic levels of the J-aggregate structure are decoupled, the measured results are expected to be independent of temperature and pressure; experimental verification of this expectation is under investigation and will be reported in the near future.
The variation of the RLS peak amplitude with RH is more sensitive to noise than the peak frequency shift, which is a characteristic of J-aggregates. Because of the weak electrostatic force between the TPPS J-aggregate and the fiber tip, for high RH, the TPPS molecules gradually separate from the fiber tip. In other words, aging occurs at high RH. To evaluate the effects of aging on the reflectance spectrum, a coated fiber is measured for several sequential weeks. The amplitude of the resonance peak shows a strong dependence on aging, but the redshift wavelength is independent of aging.
This effect is in agreement with the hysteresis result observed in our experiments. The resonance frequency does not change, but the amplitude decreases slowly after each experiment. The response time of wavelength change is about two minutes.
The resonance peak amplitude decreases over subsequent weeks. At high RH, aging appears sooner than that at low RH. The lifetime of the sensor is defined by the time average at which the RLS peak intensity reaches half of its initial value. In our experiments, the lifetime at high RH (> 50%) is seven to ten days; at low RH (< 25%), no variation in RLS peak intensity is observed. This indicates that the humidity sensor is very durable, such that aging is not a problem in sensor applications. However, the aging effect renders the amplitude of the resonance frequency peak unsuitable for humidity measurement; the redshift wavelength is more appropriate. Although aging does not affect the redshift frequency, it decreases the lifetime of the humidity sensor. This problem can be solved by using a suitable interface and protective layers in high-RH conditions. Nanoscale scratches on the optical fiber tip before deposition can improve the aging effect, for scratch dimensions of < 490 nm. This improvement arises from an increase in the electrostatic force between the TPPS film and fiber tip. These effects are under both theoretical and experimental investigations and will be reported in near future. Meanwhile, we expect to improve the film quality and stability by mixing TPPS with different polymers. In the next section, we will discuss wavelength shifts of the RLS peak.

Chromatic response of coated fiber tip to humidity
The wavelengths of RLS for different RH levels are measured and are presented in Fig. 8

Discussion and suggestion
This preliminary study outlines the responsiveness of TPPS thin films on optical fiber tips to humidity. The thin film can be removed over several cycles of exposure to high humidity. Aging decreases the intensity of the resonance frequency peak over several weeks, which requires further investigation. Other aspects for future study include improving the thickness control of the TPPS film, studying the response time, and improving the mobility of the fiber head by using different polymer coatings in the fiber cladding.

Conclusions
Our experimental analysis has shown that the amplitude is unsuitable for sensing the humidity, while sensors based on wavelength changes are viable. This is attributed to the water solubility of porphyrin and the promotion of J-aggregate formation by humidity. Different RH levels act as different microscale solvents; therefore, we expect different J-aggregate arrangements with different coupling coefficients. This is in agreement with the results reported previously [45,46]. The measured wavelength of the RLS peak of the J-aggregate as a function of RH can be fitted by several polynomials. The best-fitting polynomial is of the fourth order. For low (< 25%) and high (> 50%) RHs, a linear approximation works well. The sensitivity / r RH λ ∂ ∂ is a function of RH, corresponding approximately to 1/12 nm / RH% for 1% RH.
Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.