1 Introduction

The detection of humidity plays an important role in various fields, such as biomedicine, food processing, air conditioning control, industrial water flow, agricultural production and many other fields [1]. In particular, special environmental conditions including high electromagnetic interference, narrow installation space and load limitation put forward more stringent requirements for the sensing performance of humidity sensors in their sensitivity, resolution and response time [2,3,4].

The optical fiber humidity sensors have received extensive attention in recent years, due to their characteristics of small size, high sensitivity, fast response, and less susceptibility to electromagnetic interference [5]. Furthermore, it is easier to series-connect a large number of different sensors (temperature, humidity, PH, magnetic field, pressure) in a same optical fiber system, reducing the need for multiple wiring of traditional electronic humidity sensors [6, 7]. Therefore, a variety of optical fiber humidity sensors have been proposed for solving biomedical problems and controlling industrial production. Being one of the most typical structure to explore the fiber humidity sensors, the fiber Fabry–Perot interferometer (FPI) can be classified into several types, such as intrinsic FPI (IFPI), extrinsic FPI (EFPI) and in-line FPI (ILFPI) [8,9,10]. These structures usually contain an open cavity or thin film, which will be either greatly affected by fluctuations or fragile [11]. Therefore, it is necessary to develop an optical fiber humidity sensor with compact structure, high reliability, high sensitivity and low cost.

Electrolytes, high molecular polymers, porous metal oxides and semiconductor materials are common moisture-sensitive materials, which have been widely concerned by researchers in different fields and used in many applications [12,13,14]. Among them, hygroscopic polymer materials have a wide range of relative humidity, good film-forming properties, and good adhesion properties [15], which are more suitable for exploring the optical fiber RH sensors. polymethyl methacrylate (PMMA) solution was chosen because PMMA has a wide range of applications, mainly related to the physical characteristics of high transparency, light weight, high hardness, rapid condensation in the air, and will not cause interference to the outside world [16]. As the water content increases, the deformation behavior of PMMA changes significantly, from brittle deformation mode to ductile deformation mode [17, 18]. By monitoring the movement of the interference wave, the relative humidity information of the measured environment can be extracted. Coreless fiber (CLF) has no fiber core and only silica cladding. This kind of waveguide-less structure helps to reduce back reflection and prevent damage to the end face of the fiber in high-power applications [19].

This paper presents an EFPI optical fiber sensor based on PMMA coating. The interference-based optical fiber sensor has the highest accuracy and is easy to measure various parameters. The relative humidity response has been studied in detail from the aspects of sensitivity, stability, time response and static characteristics.

2 Sensor production and experimental system

2.1 Working principle and sensor design

2.1.1 Fiber structure and working principle

The sensing probe has a simple structure and very small volume, composed by single-mode fiber (SMF, Coning SMF-28, core diameter: ~ 8.3 μm; outer diameter: ~ 125 μm), CLF (CL 1010-A, outer diameter: 125 ± 1 μm) and PMMA, as illustrated in Fig. 1.

Fig. 1
figure 1

Schematic of EFPI humidity sensor

Full size image

In this fiber structure, the humidity change can be determined by real-time monitoring the FP interference spectrum change as a function of the width D of FP cavity between CLF and SMF end face. The probe is formed by fusion splicing one end of the CLF with SMF, then aligning its other end with another SMF, between them the PMMA solution was filled. This EFPI optical fiber sensor is composed of a smooth-end optical fiber and a PMMA diaphragm close to it. The gap between their reflective surfaces will form a Fabry–Perot (FP) cavity. Since the core diameter for a SMF is only 5–8 μm, the dislocation of the fiber cores for the two free standing SMFs will seriously affect the quality of the transmission light signal. The introduction of CLF can effectively expand the diameter of the light beam and result in a more stable extinction ratio of the interference spectrum. Furthermore, the sensing performance can be improved when the wide beam signal passes through the cavity. Because the sensitive material PMMA will expand after absorbing the water molecules, the thickness will be different in the film interior. Compared to the too thin diameter of the two fiber cores of SMFs, the influence of the local inhomogeneity of PMMA film on the interference spectrum can be effectively reduced.

2.1.2 Fiber structure optimization

For the proposed humidity sensor, the length of CLF and the width of the FP cavity will be the two important parameters, which will exert a significant impact on the FP interference spectra. To optimize the CLF length, the interference spectra for the length of 1 mm, 2 mm, 3 mm and 4 mm have been compared in Fig. 2.

Fig. 2
figure 2

Transmission spectra of with a the different length (1 mm, 2 mm, 3 mm and 4 mm) of CLF and b the air gap width (20 μm, 50 μm, 80 μm and 110 μm) of FPI cavity

Full size image

It is found that the spectra contrast drops quickly when the CLF length is greater than 2 mm. Furthermore, the relative error will be lower for a longer fiber, the CLF length was finally determined to be 2 mm. Then the CLF and a single-mode fiber were aligned with a fusion splicer, and the air gap cavity was used as the FPI and adjusted from 20 to 110 μm with a step of 30 μm. By experimental recording and comparing the interference spectra, it is found that the shorter cavity has the higher accuracy. The best value of 50 μm was obtained due to its higher contrast of the interference spectra and less distortion of the waveform.

2.1.3 Fabrication of sensor probe

To fabricate this PMMA-based EFPI humidity sensor as shown in Fig. 1, one cut of CLF with the length of ~ 2 mm was spliced onto the SMF firstly, where the cut length can be precisely controlled by the 3-dimension fiber adjustment with the precision of 1 μm. Then, the CLF end and another SMF was aligned in the splicer with the air gap of ~ 50 μm, when the gap size can be adjusted and observed in the microscope screen, meanwhile, the transmission spectra can be real-time recorded by a spectrometer. The fabrication humidity sensor is shown in Fig. 3.

Fig. 3
figure 3

Microscope picture of EFPI humidity sensor with PMMA thickness of ~ 50 μm

Full size image

Finally, the PMMA solution (prepared following the method in [20]) was dipped to fill in the air gap and cover the outer surface of the connection region of CLF and SMF. After the PMMA solution becomes solid, it will be used as the moisture-sensitive material. Furthermore, it bonds together the CLF and single-mode fiber to form the PMMA-based EFPI humidity sensor.

2.2 Experimental setup

The humidity measurement system is shown in Fig. 4. It consists of a spectrum analyzer (AQ6370D), an ASE broadband light source (KG-ASE-CL-D-13-FC/APC), and an airtight chamber, used as the humidity control box, in which the PMMA-based EFPI humidity sensor and an electronic hygrometer are placed. The ASE broadband light source outputs the light single with the flat pattern waveform from 1528 to 1603 nm, which is launched into the sensor to produce the interference spectrum being recorded by the spectrum analyzer.

Fig. 4
figure 4

Experimental setup for measuring humidity

Full size image

The different humidity environment was supplied by the cotton balls, which was soaked the saturated K2SO4 solution and put into the airtight chamber quickly to increase the RH from 25 to 80%RH. An electronic hygrometer real-time recorded the RH value for the humidity increasing every 5%RH, which was used as the reference of humidity sensor.

3 Humidity sensing performance

When the ambient temperature was maintained as a constant and the spectra waveform was recorded after the environmental humidity becoming stable, the airflow and the random vibration of fiber sensor can be ignored. As the humidity increased, the wavelength red-shift phenomenon can be observed in the spectra. During the humidity increasing process, more water molecules were absorbed by PMMA, increasing its refractive index. The mechanism lies in that the adsorbed water in the hydrophilic polymer containing pyridine or hydroxyl group binds to the polymer to a higher degree. PMMA material contains a large number of micropores and is easier to absorb the water molecules in air, resulting in its higher density and larger refractive index, which further caused the wavelength red-shift phenomenon.

In Fig. 5, the experimental sensing curves for either humidity rising or decreasing process has been linear fitted and compared. The humidity sensitivity of 0.1747 nm/%RH was obtained with a linear relationship, indicating the good stability and accuracy of the humidity sensor for measuring the environmental humidity. During the experiment, the ambient humidity was 67%RH on the experimental day. To study the humidity response time, the humidity was changed around 67%RH with the difference of 10%RH, when the spectra were recorded and observed continually for 14 min. As illustrated in Fig. 6, the humidity in the airtight charmer was firstly increased by 10%RH, the spectra waveform was recorded for 14 min once the airtight chamber was opened to the air environment; then the humidity was decreased by 10%RH to observe the spectra change.

Fig. 5
figure 5

Humidity sensing performance of PMMA-based EFPI humidity sensor during rising and decreasing process in range of 25%RH to 80%RH

Full size image
Fig. 6
figure 6

Response time of humidity fiber sensor

Full size image

It is found in Fig. 6 that during either the humidity rising or falling process, the response time was determined to be ~ 4.5 min when the sensing curve became stable.

4 Discussions

The influence of the temperature on the sensing performance of FP humidity sensor was studied in the range of 30 °C-80°C step by 5 °C. The spectra were recorded when the temperature in the incubator became stable, so as to reduce the possible error caused by the fluctuation of environmental airflow during the temperature changing process. When the humidity was 67%RH, ignoring the impact from airflow and vibration, the wavelength blue-shifted firstly, and then turn round as red-shift at a certain temperature. It is because that either humidity or temperature change can result in the wavelength shift, but contributed in the opposite. The equilibrium point appeared at ~ 45℃, as shown in Fig. 7.

Fig. 7
figure 7

Transmission power and wavelength as a function of environmental temperature during 30–80 °C at humidity of 67%RH.

Full size image

When the temperature became higher than 65 °C, the hydrogen bonds existing between the main chains of the PMMA copolymer are strongly attracted, resulting in a fast red shift of the wavelength, as well as the enhancement of the transmission optical intensity. Therefore, during the temperature interval of 30 °C to 65 °C, the influence of humidity is greater than that of temperature. Especially, the temperature effect can be ignored for the environmental temperature lower than 40 °C.

5 Conclusion

In summary, a RH sensor based on the FPI structure is proposed. PMMA was coated as a moisture-sensitive material to form an EFPI, and its moisture-sensitive characteristics and temperature effects have been analyzed. Wavelength monitoring results show that under the conditions of the FP cavity length of 50 μm and the coreless fiber length of 2 mm, the proposed sensor has a good stability and sensitivity in the range of 25–80%RH. The corresponding humidity sensitivity is 0.1747 nm/%RH, and response time is 4.5 min. Due to the low cost of the PMMA material and the good sensing properties, this RH fiber sensor has a promising aspect for determining the environmental humidity in some special or bad conditions.