sensors Article

Tilted Fiber Bragg Grating Sensor with Graphene Oxide Coating for Humidity Sensing Yung-Da Chiu 1 , Chao-Wei Wu 2 and Chia-Chin Chiang 1, * 1 2

*

ID

Department of Mechanical Engineering, National Kaohsiung University of Applied Sciences, No. 415, Jiangong Rd., Sanmin Dist., Kaohsiung 807, Taiwan; [email protected] Department of Aeronautical and Mechanical Engineering, Air Force Academy, Kaohsiung 807, Taiwan; [email protected] Correspondence: [email protected]; Tel.: +886-7-381-4526 (ext. 5340)

Received: 8 August 2017; Accepted: 14 September 2017; Published: 15 September 2017

Abstract: In this study, we propose a tilted fiber Bragg grating (TFBG) humidity sensor fabricated using the phase mask method to produce a TFBG that was then etched with five different diameters of 20, 35, 50, 55 and 60 µm, after which piezoelectric inkjet technology was used to coat the grating with graphene oxide. According to the experimental results, the diameter of 20 µm yielded the best sensitivity. In addition, the experimental results showed that the wavelength sensitivity was −0.01 nm/%RH and the linearity was 0.996. Furthermore, the measurement results showed that when the relative humidity was increased, the refractive index of the sensor was decreased, meaning that the TFBG cladding mode spectrum wavelength was shifted. Therefore, the proposed graphene oxide film TFBG humidity sensor has good potential to be an effective relative humidity monitor. Keywords: humidity sensor; tilted fiber Bragg grating; graphene oxide

1. Introduction After it was first developed, most optical fiber was used in telecommunications applications, but in recent years the development of fiber optic sensors has proceeded rapidly, especially with respect to temperature and stress sensors. Furthermore, fiber optic sensors for biomedical and home security purposes have also been developed in recent years. The mechanism by which such sensors provide measurements is very dependent on changes in the refractive index, such as in the case of LPG gas sensor measurements of CO2 [1] and U-shaped bending-induced interference sensor measurements of glucose solutions [2]. However, these previously developed sensing methods have not taken into account the impact of environmental humidity. Therefore, according to the following references, optical fiber sensors can be applied to provide a variety of different physical measurements, including measurements of humidity [3–8], strain, temperature [3,4], and RI, with some sensors even being versatile enough to measure several phenomena at once [9]. The present study proposes a humidity sensor that could potentially be used to aid other sensors in adjusting to environmental humidity. It should be noted, however, that various other types of fiber optic humidity sensors have previously been reported. For example, a side-polished fiber sensor was fabricated with a wheel side-polishing technique [10], a long period fiber grating sensor was fabricated using a multi-layer dip coating technique [11], and a hollow core fiber sensor was achieved at a resonant wavelength [12]. However, the methods used to fabricate those sensors are complicated. Therefore, the advantages of the sensor proposed in this study are its low manufacturing cost and short processing time. The proposed sensor utilizes a tilted fiber Bragg grating (TFBG) to achieve the humidity sensing function because TFBG can effectively measure the surrounding refractive index. This is because when the refractive index changes, the wavelength of the cladding mode of the TFBG will shift in a way that corresponds to the ambient humidity value. Sensors 2017, 17, 2129; doi:10.3390/s17092129

www.mdpi.com/journal/sensors

Sensors 2017, 17, 2129

2 of 11

Humidity sensors are often used in conjunction with cooled computers, heat treating furnaces, smelting furnaces, clean room controls, and dryers, as well as in applications relating to drug administration, food preservation, and climatic measurements. However, while some harsh environments, such as high temperature environments or environments with high levels of electromagnetic waves, will have an impact on the sensitivity of previously produced humidity sensors, optical fiber humidity sensors could potentially overcome those environmental challenges. In TFBG, there is a certain tilt angle between the grating plane and the fiber, so the transmission spectrum possesses many resonances [13], resulting in the occurrence of more complex mode coupling. Core and cladding mode coupling, including coupling involving the core mode and radiation mode, also occur in TFBGs. Since the tilt angle and refractive index modulation determine the coupling efficiency and the bandwidth of cladding mode resonances [14], the transmission characteristics of TFBG provide a great amount of information related to the fiber and grating structures. In 1996, Erdogan and Sipe [15], using a high-intensity ultraviolet radiation grating written in fiber, presented a detailed theory of how Bragg reflection and radiation mode coupling loss are combined in TFBG. In 2001, Li et al. [16] proposed a tilt-type fiber grating that uses the volumetric current (VCM) method to calculate radiation morphology (including wavelength dependence, azimuth distribution, and polarization dependence). In 2001, Feder et al. [17] used a spectrum analyzer and the radiation pattern of the core and the cladding mode of TFBG to provide multi-band Raman pump power measurements. In 2012, Wang et al. [18] introduced the sensing principles of a polyimide-coated FBG humidity sensor. The results obtained for this sensor, including a sensitivity of 2 pm/% RH and a measurement range of 30% to 80%, showed that polyimide resin is an ideal coating material for use in the manufacturing of an FBG humidity sensor. In 2013, Berruti et al. [19] proposed a polyimide-coated FBG humidity sensor with a wavelength sensitivity of 0.00213 nm/% RH in a humidity range of 0% to 75% at low temperatures. In 2014, Montero et al. [20] applied steel and carbon fiber polymers to FBG measuring a range of 30% to 90% RH in a temperature range of 10~70 ◦ C; in the range of 30% to 80% RH, the sensitivity wavelength was 0.001375 nm/% RH. In 2016, Wang et al. [21] proposed coating graphene oxide on TFBG as a means of sensing humidity via the adsorption and desorption of water by the graphene oxide film. With the use of cladding mode and graphene oxide film interaction between the detection wavelength at 1557 nm and the transmission loss change, a maximum sensitivity of 0.129 dB/% RH with a linear correlation coefficient in the humidity range of 10% to 80% and 99% was achieved, indicating that this humidity sensor has repetitive stability and other good characteristics. According to the above literature, the humidity measurement range of such sensors is about 30% to 80% RH, with the transmission loss changes yielding an average sensitivity of 0.00168 nm/% RH. As these sensors are made using the arc fusion splicing method, where the sensing layer has a hydrogel coating, the above method is highly time-consuming and involves a high degree of complexity. Therefore, this study proposes for the first time, to the best of our knowledge, a small diameter TFBG sensor (20 µm) with a graphene oxide coating for humidity sensing. The novelty of the proposed TFBG sensor is twofold: first, the sensitivity of the sensor is increased by reducing the fiber diameters via etching; second, we provide new coating technology used the piezoelectric inkjet can provide a uniform graphene oxide coating and fast processing, allowing it to be used as a humidity sensor. 2. Working Principle of the Graphene Oxide Humidity Sensor Mode fiber coupling between the wave vectors of all vector fields occurs along the fiber axis to form a uniform axial conduction mode structure that is defined by a flat phase plane perpendicular to the guide shaft, as shown in Figure 1. The resonant wavelength satisfying the Bragg condition for an ordinary fiber grating can be expressed by the following Equation [14]:   λ Bragg = ne f f ,core + ne f f ,core Λ g

(1)

Sensors 2017, 17, 2129

3 of 11

Sensors 2017, 17, 2129

3 of 11

𝜃 Cladding Core Λ𝑔 Figure 1. Schematic diagram of TFBG.

Figure 1. Schematic diagram of TFBG. The Bragg resonance conditions described above are produced by the coupling of mode-forward core modes and back-propagating core modes. For TFBG, of the Thepropagating Bragg resonance conditions described above are produced bythe theinclination couplingangle of mode-forward grating surface and the axis of the fiber, as well as the grating period along the fiber axis, can be of the propagating core modes and back-propagating core modes. For TFBG, the inclination angle modified according to the following Equation [14]: grating surface and the axis of the fiber, as well as the grating period along the fiber axis, can be 𝛬 modified according to the following Equation [14]: (2) 𝛬𝑔 = 𝑐𝑜𝑠𝜃

Substituting Equation (2) into Equation (1), yields Λ the following:

Λg =

cos θ

𝑐𝑜 𝑐𝑜 𝜆𝐵𝑟𝑎𝑔𝑔 = (𝑛𝑒𝑓𝑓 + 𝑛𝑒𝑓𝑓 )

Λ cos 𝜃

(2) (3)

Substituting Equation (2) into Equation (1), yields the following:

Due to the presence of the tilted angle, part of the light propagating in a forward direction  cladding mode  of through the core mode will be coupled to the Λ the backward propagation, and the λ Bragg =will nco + nco resonant wavelength of the cladding mode by the following equation [14]: ebe f f determined ef f

cos θ

(3)

Λ (4) 𝜆𝐶𝑙,𝑖 = (nco + ncl ) eff,i Due to the presence of the tilted angle, part ofeffthe light propagating in a forward direction through cos θ cl the corewhere modenwill be coupled to the cladding mode of the backward propagation, and the resonant eff,i is the effective refractive index of the ith cladding mode. As the effective refractive index wavelength of the mode will be by(2), theit following increases, thecladding wavelength redshifts, anddetermined from Equation can be seenequation that using [14]: a fixed grating

period to change the rotation angle can change the Bragg period. The purpose of this study was to investigate the cladding mode wavelength shift, and inserting the measurement data into cl by Λ λCl,i = (nco (4) eff + neff ,i ) cos Equation (4), we could verify whether the measurements provided by the TFBG humidity sensor θ were correct. where ncl thetoeffective refractive index of the ith cladding As the effective refractive Due the interaction between the cladding mode and the mode. graphene oxide, the resonance of the index eff ,i is cladding mode changes as the graphene oxide adsorption process proceeds, because the effective increases, the wavelength redshifts, and from Equation (2), it can be seen that using a fixed grating refractive the graphene changes the water molecules adsorbed. canstudy be period to changeindex the of rotation angleoxide can film change theasBragg period. Theare purpose of As this was seen from Equation (4), when the humidity increases, the refractive index changes, causing the to investigate the cladding mode wavelength shift, and by inserting the measurement data into wavelength to be shifted, and through this phenomenon, the changes in the TFBG sensor effectively Equation (4), we could verifyhumidity whetherchanges. the measurements provided by the TFBG humidity sensor correspond to the relative

were correct. 3. Experiment Due to the interaction between the cladding mode and the graphene oxide, the resonance of the cladding mode changes as the graphene oxide adsorption process proceeds, because the effective 3.1. Processing and Fabrication of TFBG refractive index of the graphene oxide film changes as the water molecules are adsorbed. As can In Equation this study, (4), the TFBG sensor was fabricated using single-mode fiber (SMF-28), the be seen from whenhumidity the humidity increases, the refractive index changes, causing the phase mask method, and piezoelectric inkjet technology. Using a wire stripper, 3 cm of the protective wavelength to be shifted, and through this phenomenon, the changes in the TFBG sensor effectively middle layer was stripped from optical fiber, after which the fiber was wiped clean with alcohol and correspond the on relative changes. then to placed a 3-axishumidity micro-platform. Next, the mask platform was rotated 10 degrees and excimer laser (Xantos XS, Excimer laser, COHERENT, Lübeck, Germany) irradiation was applied through

3. Experiment phase masks (1075.5 nm, Phase mask, Ibsen photonics, Farum, Denmark) to produce the grating

formation. An optical spectrum analyzer was then utilized to confirm the generation of TFBG in the

3.1. Processing andspectrum Fabrication of TFBG core of the structure, as shown in Figure 2. The length of the TFBG was 5 mm. The TFBG was processed using buffered oxide etch (BOE) to change the diameter of the fiber by etching.

In this study, the TFBG humidity sensor was fabricated using single-mode fiber (SMF-28), Approximately 3 cm of the protective layer at the center of a single-mode fiber section was removed, the phase mask method, and piezoelectric inkjet technology. Using a wire stripper, 3 cm of the protective middle layer was stripped from optical fiber, after which the fiber was wiped clean with alcohol and then placed on a 3-axis micro-platform. Next, the mask platform was rotated 10 degrees and excimer laser (Xantos XS, Excimer laser, COHERENT, Lübeck, Germany) irradiation was applied through phase masks (1075.5 nm, Phase mask, Ibsen photonics, Farum, Denmark) to produce the grating formation. An optical spectrum analyzer was then utilized to confirm the generation of TFBG

Sensors 2017, 17, 2129

4 of 11

in the core of the spectrum structure, as shown in Figure 2. The length of the TFBG was 5 mm. Sensors 2017, 17, 2129 4 of by 11 The TFBG was processed using buffered oxide etch (BOE) to change the diameter of the fiber Sensors 2017, 17, 2129 4 of 11 etching. Approximately 3 cm of the protective layer at the center of a single-mode fiber section was and the fiber was adhered to the etching board, which was then placed into a plastic box into which removed, and the fiber was adhered to the etching board, which was then placed into a plastic box and the wasto adhered to the etching board, which was then placed into a plastic box into which BOE wasfiber poured effect the etching. into which BOE was poured to effect the etching. BOEThe wasTFBG poured to effect was etchedthe to etching. 20 μm, and then piezoelectric inkjet technology (DMP-2850, Dimatix The TFBG was etched to 20 µm, and then piezoelectric inkjet technology (DMP-2850, Dimatix The TFBG was etched 20 μm, andInnovation, then piezoelectric inkjet CA, technology (DMP-2850, Dimatix Materials Printer, FUJIFILMtoValue From Santa Clara, USA) was used to spray an Materials Printer, FUJIFILM Value From Innovation, Santa Clara, CA, USA) was used to spray an Materialssolution Printer,of FUJIFILM Innovation, Santa Clara, CA, was used towas spray an aqueous grapheneValue oxideFrom onto the surface of the optical fiber.USA) The coated fiber then aqueous solution of graphene oxide onto the surface of the optical fiber. The coated fiber was then aqueous solution of graphene oxide onto the surface of the optical fiber. The coated fiber was then heated to 100 degrees on a baking plate to induce evaporation, allowing the graphene to become heated to 100 degrees on a baking plate to induce evaporation, allowing the graphene to become more heated to 100 attached degrees on a baking induceasevaporation, allowing the graphene to become more securely to the opticalplate fiber to surface, shown in Figure 3. securely attached to the optical fiber surface, as shown in Figure 3. more securely attached to the optical fiber surface, as shown in Figure 3. Periscopes Periscopes

Excimer Laser (246nm) Excimer Laser (246nm) Control Laser Control Laser

ASE Light Source ASE Light Source

Focal Lens Focal Lens

Optical Spectrum Analyzer Computer Optical Spectrum Analyzer 10°Computer

Cladding Core Cladding Core

TFBG TFBG

10°

Light Output Light Output Motorized Linear Stages Motorized Linear Stages

Mask Mask

Figure 2. 2. Processing and fabrication fabrication of of TFBG TFBG sensor. sensor. Figure 2. Processing and fabrication of TFBG sensor. CCD Camera Inkjet header CCD Camera Inkjet header

Pizeoelectric inkjet Pizeoelectric Printer inkjet Printer

200 nm 200 nm 200 nm 200nm nm 200 200nm nm 200 200 nm 200 nm 200 nm

inkjet thickness control inkjet thickness control

n ( ) (50n m) ( ) (50 m)

TFBG Coating Coating Graphene inkjet to optical TFBG Coating Coating Graphene inkjetfiber to optical Graphene Oxide Oxide Graphene Oxide Oxide fiber Figure 3. Flow chart of the process of using piezoelectric inkjet technology to coat the fiber with Figure Flow Figure3.3. oxide. Flow chart chart of of the the process process of of using using piezoelectric piezoelectric inkjet inkjet technology technology to to coat coat the the fiber fiber with with graphene graphene grapheneoxide. oxide.

3.2. The Setup of the Humidity Sensing Experiment 3.2. 3.2.The TheSetup Setupofofthe theHumidity HumiditySensing SensingExperiment Experiment The fiber was placed in a humidity control box. A light source (ASE-2200, ASE light source, The fiber aahumidity control source light TheTechnologies fiber was wasplaced placed in humidity control box. A A light light source (ASE-2200, ASE light source, NXTAR Inc.,in Tainan, Taiwan) was box. attached to one end(ASE-2200, of the box,ASE while an source, optical NXTAR Technologies Inc., Tainan, Taiwan) was attached to one end of the box, while an optical NXTAR Technologies Inc., Tainan, to Taiwan) was attached to one end of the box, of while an optical spectrum analyzer was attached the other. humidification control range 20–80% RH, spectrum analyzer was attached to the other. humidification control range of 20–80% RH, temperature spectrum analyzer was attached to the other. humidification control range of 20–80% temperature controller setting 27 °C, standard humidity sensor for compared to our TFBG RH, GO ◦ C, standard humidity sensor for compared to our TFBG GO humidity sensor, controller setting 27 temperature controller setting 27 °C, standard humidity sensorhumidity, for compared to our TFBG GO humidity sensor, humidifier to increase the environment relative dehumidification is the humidifier to increase the environment relative humidity, dehumidification is the dry air into the humidity humidifier toL/min) increase the environment humidity, dehumidification is the dry air intosensor, the (a flow rate of 10 humidity control Boxrelative to reduce humidity, as shown in Figure 4. (a flow rate of 10 L/min) humidity control Box to reduce humidity, as shown in Figure 4. dry air into the (a flow rate of 10 L/min) humidity control Box to reduce humidity, as shown in Figure 4. In this study, the TFBG-coated graphene oxide film coated on SMF-28 single-mode In this study, the TFBG-coated graphene oxide film coated on SMF-28 single-mode photosensitive fiber was etched with five different diameters of 20, 35, 50, 55 and 60 μm, and the photosensitive fiber etched at with five different diameters 20,and 35,a50, 55 and range 60 μm, relative humidity waswas measured a constant temperature of 27of°C humidity ofand 20%the to relative humidity was measured at a constant temperature of 27 °C and a humidity range of 20% to 80% RH. 80% RH.

Sensors 2017, 17, 2129

5 of 11

SensorsIn2017, 2129 this17, study,

5 of 11 the TFBG-coated graphene oxide film coated on SMF-28 single-mode photosensitive fiber was etched with five different diameters of 20, 35, 50, 55 and 60 µm, and the relative humidity Sensors 2017, 17, 2129 5 of 11 was measured at a constant temperature of 27 ◦ C and a humidity of 20% to 80% RH. Optical Spectrumrange Analyzer

Optical Spectrum Analyzer

Light Source Light Source

Humidity Controller Humidity Controller

ir

Humidity Sensor

ir

Humidity Sensor

Humidifier Humidifier

TFBG TFBG

Humidity Humidity System System

Experimental Figure 4. 4. Experimental forhumidity humiditysensing. sensing. Figure Experimental setup setup for

4. Results Discussion 4. and Discussion 4. Results Results andand Discussion 4.1. Graphene Oxide Coating 4.1. Graphene Oxide FilmFilm Coating In this study, piezoelectric inkjet technology was used to coat graphene oxide film on optical

In this study, piezoelectric inkjet technology was used to coat graphene oxide film on optical fiber. The thickness of the coated graphene oxide, as measured by scanning electron microscope The thickness thickness of of the the coated coated graphene graphene oxide, oxide, as as measured by scanning electron microscope fiber.(SEM) The imaging, was about 1 μm. Enlargement of the SEM image shown in Figure 5a revealed that imaging, was about 1 µm. Enlargement of the SEM image in Figure 5awas revealed that the (SEM)the imaging, was about 1 μm. Enlargement of the SEM image in Figure 5a coated revealed graphene film was in the form of a sheet, thereby proving thatshown theshown graphene oxide on that graphene film was in the form of a sheet, thereby proving that the graphene oxide was coated on the the graphene film was in the form of a sheet, thereby proving that the graphene oxide was coated the optical fiber sensor (Figure 5b). We sought to compare the effects of a uniform coating with those on optical sensor (Figure 5b). We We sought to compare the effects of a of uniform coating withdegree those of a the optical fiber sensor (Figure 5b). to compare the effects aresulted uniform with those of fiber a nonuniform coating. As shown in sought Figure 5c,d, the nonuniform coating incoating a slight of transmission variation relative to Figure the uniform but thecoating amount of resulted variation was minimal nonuniform coating. As shown in Figure 5c,d, thecoating, nonuniform resulted in ain slight degree of of a nonuniform coating. As shown in 5c,d, the nonuniform coating a slight degree and would not significantly affect the measurement results. transmission variation relative to the uniform coating, but the amount of variation was minimal and of transmission variation relative to the uniform coating, but the amount of variation was minimal the humidity themeasurement air more water molecules are adsorbed on the graphene would notWhen significantly affectin the results. and would not significantly affect theincreases, measurement results. oxide, the andhumidity the waterin molecules on the graphene layer will, in turn, change theongap between theoxide, increases, more water molecules are adsorbed theon graphene When the humidity inthe theairair increases, more water molecules are adsorbed the graphene graphene and the fiber. In a previous study of graphene oxide thin films [21], these changes were and theand water on the graphene layer will, layer in turn, change the gap between the graphene oxide, themolecules water molecules on the graphene will, in turn, change the gap between and the found to be due to the bonding between the dissociated H2O molecules (H2O = H+ + OH−) and the the fiber. In a previous study of graphene oxide thin films [22], these changes werethese found to be due to graphene and the fiber. In a previous study of graphene oxide thin films [21], changes were C−OH and C=O groups of the GO. Such an increase in interlayer spacing is also consistent with + + OH− ) and the C + − −and the bonding between the dissociated H O molecules (H O = H OH C=O found to be due to the bonding between the dissociated H 2 O molecules (H 2 O = H + ) and the 2 of graphene oxide 2 as a function of relative humidity. previous studies on the interlayer spacing groups and of theC=O GO.groups Such anofincrease in Such interlayer spacing in is also consistent with is previous studies on the C−OH the GO. an increase interlayer spacing also consistent with interlayerstudies spacingonofthe graphene oxide as a function of relative previous interlayer spacing of graphene oxide humidity. as a function of relative humidity.

(a)

(b)

(a)

(b)

50 μm Figure 5. Cont.

50 μm

Figure 5. Cont. Figure 5. Cont.

10 μm

10 μm

Sensors 2017, 17, 2129 Sensors 2017, 17, 2129 Sensors 2017, 17, 2129

6 of 11 6 of 11 6 of 11

(c) (c)

(d) (d)

Figure 5. 5. (a) (a) Scanning Scanning Electron Electron Microscope Microscope (SEM) (SEM) image of TFBG coated with graphene oxide film, Figure Electron Microscope (SEM) image image of of TFBG TFBG coated coated with with graphene graphene oxide oxide film, film; Figure 5. (a) Scanning (b) SEM image of graphene oxide film, (c) TFBG GO coating comparison, (d) TFBG GO uniform (b) SEM image of graphene oxide film; (c) TFBG GO coating comparison; (d) TFBG GO uniform (b) SEM image of graphene oxide film, (c) TFBG GO coating comparison, (d) TFBG GO uniform coating coating comparison. comparison.

4.2. Temperature Temperature Response of of the the TFBG TFBG Sensor Sensor 4.2. Temperature Response Response of the TFBG Sensor We conducted conducted an initial temperature experiment to verify verify the effect effect of temperature temperature on the theon TFBG We conducted an aninitial initialtemperature temperature experiment to verify the effect of temperature the experiment to the of on TFBG ◦C sensor, with the results being shown in Figure 6. As the temperature was raised from 40 °C to 100 °C, TFBG sensor, with the results being shown in Figure 6. As the temperature was raised from 40 sensor, with the results being shown in Figure 6. As the temperature was raised from 40 °C to 100 °C, ◦ C, the wavelength the100 wavelength was shifted shiftedwas fromshifted short wavelength to aa long long wavelength, wavelength, with the the wavelength wavelength to from a short wavelength to a long wavelength, with the the wavelength was from aa short wavelength to with ◦ sensitivity being 0.014 nm/°C. This can be compensated for by Bragg mode temperature wavelength sensitivity being 0.014 This nm/ C. be compensated Braggmode mode temperature sensitivity being 0.014 nm/°C. canThis be can compensated for for by by Bragg compensations. We will use use temperature temperature control control by by ± ±11degrees degrees C to reduce the effect of temperature. compensations. We will degrees C Cto toreduce reduce the the effect effect of of temperature. temperature. temperature control by ±1

Figure 6. 6. Effects of of temperature on on TFBG TFBG GO GO wavelengths. wavelengths. Figure Figure 6. Effects Effects of temperature temperature on TFBG GO wavelengths.

4.3. Comparison Comparison of of Different Diameters Diameters 4.3. 4.3. Comparison of Different Different Diameters The strength strength of of the the transmission, transmission, which which is is less less strong strong without without the the etching, etching, becomes becomes stronger stronger The The strength ofresulting the transmission, which is less strong without the aetching, becomes stronger after after the etching, in the wavelength being shifted from short wavelength to long afteretching, the etching, resulting in the wavelength beingfrom shifted from a short wavelength to aa long the resulting in the wavelength being shifted a short wavelength to a long wavelength. wavelength. After etching, the cladding mode is significantly enhanced, which improves the ability wavelength. After etching, the cladding mode is significantly enhanced, which improves After etching, the cladding mode is significantly enhanced, which improves the ability ofthe theability fiber of the fiber to measure the refraction index, as shown in Figure 7. Therefore, this study explored how of the fiber to measure theindex, refraction index,in asFigure shown7.inTherefore, Figure 7. Therefore, this study how explored how to measure the refraction as shown this study explored different different diameters of the TFBG affect its sensitivity. different diameters of the TFBG affect its sensitivity. diameters of thecoating TFBG affect its sensitivity. The TFBG graphene oxide humidity humidity sensor sensor has has the the best best sensitivity sensitivity in in the the number number of of The TFBG TFBG coating coating graphene graphene oxide oxide The humidity sensor has the best sensitivity in the number of diameters. In In this this paper, paper, the the sensitivity sensitivity of of the the sensor sensor to to humidity humidity changes changes was was measured measured with with diameters. diameters. In this paper, the sensitivity of the sensor tobe humidity changes was measured with4.1 different different diameters of 20, 35, 50 55 and 60 μm. As can seen from the discussion in section above, different diameters of 20, 35, 50 and 60can μm. As can be seen from the discussion in above, diameters of 20,of 35,water 50 55molecules and 60 55 µm. As be seen from the discussion in Section 4.1section above,4.1 when the when the level in the air is increased, the molecules will, in turn, increase the gap when the level of water molecules in the air is increased, the molecules will, in turn, increase the gap level of water molecules in the the fiber, air is increased, the molecules will, in turn, increase thethe gap between between the graphene graphene and thereby changing changing the external external refractive index of of TFBG. between the and the fiber,changing thereby the index the graphene and the fiber, thereby the external refractive refractive index of the TFBG.the TFBG.

Sensors 2017, 17, 2129 Sensors 2017, 17, 2129

7 of 11 7 of 11

Figure 7. TFBG spectrum, before etching (Black line), after etching (Red line). Figure 7. TFBG spectrum, before etching (Black line), after etching (Red line).

As shown in Figure 8a, when the diameter was 20 μm and the RH was at 20%, the wavelength

As innm, Figure 8a, when the diameter was 20 µm and wasmeaning at 20%,that the the wavelength wasshown 1535.962 whereas the wavelength was 1535.367 nm at an the RH RH of 80%, shift was 1535.962 nm, whereas thepm. wavelength 1535.367 nm atthe andiameter RH of 80%, of the wavelength was 595 As shownwas in Figure 8b, when was meaning 35 μm andthat the the RH shift 20%, the wavelength was As 1550.060 nm, the when wavelength was 1549.815 at anand RH the of RH of thewas wavelength was 595 pm. shown in whereas Figure 8b, the diameter was nm 35 µm 80%, meaning that the shift of the wavelength was 245 pm. As shown in Figure 8c, when the diameter was 20%, the wavelength was 1550.060 nm, whereas the wavelength was 1549.815 nm at an RH of 50 μmthat andthe theshift RH of was 20%, the wavelength was 1538.706 nm, whereas the was 80%, was meaning the wavelength was 245 pm. As shown in Figure 8c,wavelength when the diameter 1538.531 nm at an RH of 80%, meaning that the shift of the wavelength was 175 pm. As shown in was 50 µm and the RH was 20%, the wavelength was 1538.706 nm, whereas the wavelength was Figure 8d, when the diameter was 55 μm and the RH was 20%, the wavelength was 1538.231 nm, 1538.531 nm at an RH of 80%, meaning that the shift of the wavelength was 175 pm. As shown in whereas the wavelength was 1538.106 nm at an RH of 80%, meaning that the shift of the wavelength Figure 8d, when the diameter was 55 µm and the RH was 20%, the wavelength was 1538.231 nm, was 125 pm. Finally, as shown in Figure 8e, when the diameter was 60 μm and the RH was 20%, the whereas the wavelength was nm, 1538.106 nmthe at an RH of 80%, the shift the wavelength wavelength was 1539.980 whereas wavelength wasmeaning 1539.855 that nm when the of RH was 80%, was 125 pm. that Finally, as shown in Figure 8e, when the pm. diameter was 60 µm and the RH was 20%, meaning the shift of the whole wavelength was 125 the wavelength was 1539.980 nm, thewas wavelength was nm when the0.996. RH was At a diameter of 20 μm, thewhereas sensitivity −0.01 nm/% RH,1539.855 and the linearity was At a 80%, diameter of 35 μm, the sensitivity was −0.00425 nm/% RH, and the linearity was 0.990. At a diameter meaning that the shift of the whole wavelength was 125 pm. of 50 μm, the sensitivity was −0.0031 nm/% was RH, and the nm/% linearityRH, wasand 0.985. Atlinearity a diameter of 55 μm, At a At a diameter of 20 µm, the sensitivity −0.01 the was 0.996. the sensitivity −0.0022 nm/% and thenm/% linearity At a diameter of 60At μm, the diameter of 35 µm, was the sensitivity wasRH, −0.00425 RH,was and0.995. the linearity was 0.990. a diameter sensitivity was −0.0021 nm/% RH, and the linearity was 0.996. Through the experimental results of 50 µm, the sensitivity was −0.0031 nm/% RH, and the linearity was 0.985. At a diameter of detailed above, we found that the diameter of 20 μm had the best wavelength sensitivity, as shown 55 µm, the sensitivity was −0.0022 nm/% RH, and the linearity was 0.995. At a diameter of 60 µm, in Figure 9. From the results for the above five diameters, it can be seen that the refractive index the sensitivity was −0.0021 nm/% RH, and the linearity was 0.996. Through the experimental results decreases when the humidity is increased from 20% RH to 80% RH. The fact that the wavelength is detailed above, found that the to diameter of 20 µm had the best wavelength sensitivity, as so shown shifted fromwe a long wavelength a short wavelength can also be ascertained from Equation (4), in Figure 9. From the results for the above five diameters, it can be seen that the refractive this paper selected a 20 μm diameter TFBG humidity sensor to undergo the whole humidificationindex decreases when the humidity is increased from 20% RH to 80% RH. The fact that the wavelength is and dehumidification process. shifted from a long wavelength to a short wavelength can also be ascertained from Equation (4), so this Analysisaand of Humidification and Dehumidification paper4.4. selected 20 Discussion µm diameter TFBG humidity sensor to undergo the whole humidification and dehumidification process.sensitivity in the first cycle was −0.01 nm/% RH, and the linearity was 0.996. The wavelength The wavelength sensitivity in the second cycle was −0.01 nm/% RH, and the linearity was 0.996. The

4.4. Analysis andsensitivity Discussioninofthe Humidification Dehumidification wavelength third cycle wasand −0.01 nm/% RH, and the linearity was 0.996. The optimal wavelength sensitivity was −0.01 nm/% RH, and the linearity was 0.996. From the results for the three

The wavelength sensitivity in the first cycle was −0.01 nm/% RH, and the linearity was 0.996. cycles, it can be seen that the sensitivity and linearity were the best in the third cycle. From the error The wavelength sensitivity in the second cycle was −0.01 nm/% RH, and the linearity was 0.996. bars shown in Figure 10 the average error of about 0.000001 can be obtained. According to the Figure The wavelength sensitivity the third cycle was −to 0.01 nm/% RH, and themin, linearity was 0.996. 11, the response time for a in humidity change from 20% 80% RH was about 12.25 the recovery The optimal sensitivity wasin−Figure 0.01 nm/% RH, that andthe theresults linearity washumidification 0.996. From the time waswavelength about 21.75 min. The results 11 indicate for the results for the three cycles, it can be seen thatThis the in sensitivity and wereoxide the best in the third process exhibited very good reproducibility. turn shows thatlinearity the graphene film coating goodthe sensitivity and shown linearityinwhen used TFBG sensor. cycle.has From error bars Figure 10with the the average error of about 0.000001 can be obtained. According to the Figure 11, the response time for a humidity change from 20% to 80% RH was about 12.25 min, the recovery time was about 21.75 min. The results in Figure 11 indicate that the results for the humidification process exhibited very good reproducibility. This in turn shows that the graphene oxide film coating has good sensitivity and linearity when used with the TFBG sensor.

Sensors 2017, 17, 2129 Sensors 2017, 17, 2129

8 of 11 8 of 11

(a-1)

(b-1)

(c-1)

(d-1)

(e-1)

Figure8.8.TFBG TFBGhumidity humidity sensor sensor spectra spectra with Figure with different different diameters diametersof of(a) (a)20 20μm, µm;(b) (b)35 35μm, µm;(c) (c)50 50μm, µm; (d) 55 μm, and (e) 60 μm. Insets: (a-1) 1535.367 nm to 1535.962 nm, (b-1) 1549.815 nm to 1550.060 (d) 55 µm; and (e) 60 µm. Insets: (a-1) 1535.367 nm to 1535.962 nm; (b-1) 1549.815 nm to 1550.060nm, nm; (c-1)1538.2 1538.2nm nmtoto1539 1539nm; nm,(d-1) (d-1)1537.8 1537.8nm nmto to 1538.6 1538.6 nm; nm, and and (e-1) (e-1) 1539.4 1539.4 nm nm to (c-1) to 1540.4 1540.4 nm. nm.

Sensors 2017, 17, 2129

9 of 11

Sensors 2017, 17, 2129 Sensors 2017, 17, 2129

9 of 11 9 of 11

Sensors 2017, 17, 2129

9 of 11

Figure 9. Comparison of sensitivity of TFBG humidity sensors with different diameters. Figure 9. Comparison ofof sensitivity sensorswith withdifferent different diameters. Figure 9. Comparison sensitivityof ofTFBG TFBG humidity humidity sensors diameters. Figure 9. Comparison of sensitivity of TFBG humidity sensors with different diameters.

Figure 10. 3-cycle repeat test TFBG humidity wavelength results. Figure 10. 3-cycle repeat test TFBG humidity wavelength results. Figure 10. 3-cycle repeat test TFBG humidity wavelength results. Figure 10. 3-cycle repeat test TFBG humidity wavelength results.

Figure 11. Response and recovery time of the TFBG sensor when subjected to a step humidity change Figure 11.to Response and recovery time of the TFBG sensor when subjected to a step humidity change from 20% 80%. Figure 11. Response and recovery time of the TFBG sensor when subjected to a step humidity change Figure 11.20% Response from to 80%. and recovery time of the TFBG sensor when subjected to a step humidity change from 20% to 80%.

from 20% to 80%. After three cycles of testing, we can see that the relative humidity measured by the TFBG After three cycles of testing, we relative can see seehumidity that the the relative relative humidity humidity measured bythe thesensor, TFBG graphene to the measured by the standard humidity Afteroxide threesensor cycleswas of close testing, we can that measured by TFBG graphene oxide sensor was close to the relative humidity measured by the standard humidity sensor, After three cycles of TFBG testing, wetocan that humidity the oxide relative humidity measured byhumidity thefor TFBG graphene graphene oxide was close thesee relative measured bythe the standard sensor, which shows thatsensor the sensor with graphene coating has same accuracy humidity which shows the TFBG sensor with graphene oxide coating coating hasstandard thesame sameaccuracy accuracy for humidity which shows that theto TFBG sensor with graphene oxide has the for humidity sensing despite being cheaper. oxide sensor wasthat close the relative humidity measured by the humidity sensor, which sensing despite being cheaper. sensing despite being cheaper. shows that the TFBG sensor with graphene oxide coating has the same accuracy for humidity sensing

despite being cheaper.

Sensors 2017, 17, 2129

10 of 11

Sensors 2017, 17, 2129

10 of 11

From the analysis results detailed in in Section 4.3,4.3, it can be be seen thatthat a diameter of 20 yielded the From the analysis results detailed Section it can seen a diameter of µm 20 μm yielded best humidity sensitivity, which is why this study subjected a 20 µm diameter humidity sensor to three the best humidity sensitivity, which is why this study subjected a 20 μm diameter humidity sensor cycles of testing intesting order to anobtain analysis of sensitivity, linearity, and standard deviations for the to three cycles of in obtain order to an analysis of sensitivity, linearity, and standard deviations overall humidification and dehumidification process. As shown in Figure 11, as the humidity increased, for the overall humidification and dehumidification process. As shown in Figure 11, as the humidity the refractivethe index decreased, resulting in the wavelength being shifted from a long wavelength to a increased, refractive index decreased, resulting in the wavelength being shifted from a long short wavelength. Specifically, the wavelength was nm when the humidity was 20%, whereas wavelength to a short wavelength. Specifically, the1535.962 wavelength was 1535.962 nm when the humidity the wavelength was 1535.367 nm when the humidity was 80%. That is, as the humidity went was 20%, whereas the wavelength was 1535.367 nm when the humidity was 80%. That is, asfrom the 20% RH to 80% the total amount change in the wavelength wasin595 humidity wentRH, from 20% RH to 80% of RH, the total amount of change thepm. wavelength was 595 pm. We have and durability durability test. test. We haveprovided providedthe thepreliminary preliminaryexperimental experimental data data for for aa sensor sensor stability stability and According to the experimental results, the humidity sensitivity measured at 14 days was still quite close According to the experimental results, the humidity sensitivity measured at 14 days was still quite toclose the initial as shown Figurein12b. Furthermore, as shownasinshown Figure in 12a, the humidity to thesensitivity, initial sensitivity, asinshown Figure 12b. Furthermore, Figure 12a, the measurement results for 20% RH, 45% RH, and 80% RH showed good stability. It is thus proven that humidity measurement results for 20% RH, 45% RH, and 80% RH showed good stability. It is thus the proposed sensor exhibits good stability and durability. proven that the proposed sensor exhibits good stability and durability.

(a)

(b)

Figure12.12.TFBG TFBGsensor sensorstability stabilityand anddurability durability test test results results for for (a) (a) 20% 20% RH, RH, 45% 45% RH, RH, and and 80% 80% RH Figure RH measurements, (b) trial 1 and trial 2 (conducted 14 days after trial 1). measurements; (b) trial 1 and trial 2 (conducted 14 days after trial 1).

5. Conclusions 5. Conclusions In this study, a TFBG humidity sensor was fabricated by coating TFBG with a graphene oxide In this study, a TFBG humidity sensor was fabricated by coating TFBG with a graphene oxide film. film. The measurement results showed that the measurement sensitivity of the humidity was 4.695 The measurement results showed that the measurement sensitivity of the humidity was 4.695 times times higher than 0.00213 nm/% RH [7], 5 times higher than 0.002 nm/% RH [8], 7.273 times higher higher than 0.00213 nm/% RH [7], 5 times higher than 0.002 nm/% RH [8], 7.273 times higher than than 0.001375 nm/% RH [13], and 2.625 times higher than 0.00381 nm/% RH [21], while the 0.001375 nm/% RH [13], and 2.625 times higher than 0.00381 nm/% RH [21], while the wavelength wavelength sensitivity was −0.01 nm/% RH and the linearity was 0.996. This optical fiber sensing sensitivity was −0.01 nm/% RH and the linearity was 0.996. This optical fiber sensing system has system has high research and development value, as the TFBG can be used not only for relative high research and development value, as the TFBG can be used not only for relative humidity sensing, humidity sensing, but also in temperature, strain, and magnetic field sensing applications, a but also in temperature, strain, and magnetic field sensing applications, a versatility that suggests that versatility that suggests that its future is really limitless. its future is really limitless. Acknowledgments: This work was supported by the Ministry of Science and Technology, Taiwan (grant Acknowledgments: This work was supported by the Ministry of Science and Technology, Taiwan (grant number number MOST-106-2623-E-151-001-D and MOST- 106-2221-E-151-023). MOST-106-2623-E-151-001-D and MOST106-2221-E-151-023). Author Contributions:Chia-Chin Chia-Chin Chiang Chiang designed and experiments, analyzed the data, and Author Contributions: designedthe thestudy studymethods methods and experiments, analyzed the data, wrote thethe paper. Chao-Wei and wrote paper. Chao-WeiWu Wuconducted conductedthe theexperiments experimentsand andanalyzed analyzedthe theexperimental experimentaldata. data.Yung-Da Yung-Da Chiu Chiu conducted the experiments conducted the experimentsand andanalyzed analyzedthe theexperimental experimentaldata. data.

Conflicts ofof Interest: The authors Conflicts Interest: The authorsdeclare declarenonoconflict conflictofofinterest. interest.

References References 1. 1. Wu, C.;C.; Wu, C.;C.; Chiang, C. C. A ZnO Nanoparticle-Coated Long Period Fiber Grating as aas Carbon Dioxide Gas Wu, Wu, Chiang, A ZnO Nanoparticle-Coated Long Period Fiber Grating a Carbon Dioxide Sensor. Invention 2016, 1, 21. 1, [CrossRef] Gas Sensor. Invention 2016, 21. 2. 2. Fang, Y.;Y.; Wang, C.;C.; Chiang, C. C. A Small U-Shaped Bending-Induced Interference Optical Fiber Sensor for for the Fang, Wang, Chiang, A Small U-Shaped Bending-Induced Interference Optical Fiber Sensor Measurement of Glucose Solutions. Sensors 2016,2016, 16, 1460. [CrossRef] [PubMed] the Measurement of Glucose Solutions. Sensors 16, 1460. 3.

Woyessa, G.; Fasano, A.; Markos, C.; Rasmussen, H.K.; Bang, O. Low Loss Polycarbonate Polymer Optical Fiber for High Temperature FBG Humidity Sensing. IEEE Photon. Technol. Lett. 2017, 29, 575–578.

Sensors 2017, 17, 2129

3. 4.

5.

6.

7.

8. 9. 10.

11.

12.

13. 14. 15. 16. 17. 18. 19.

20. 21.

22.

11 of 11

Woyessa, G.; Fasano, A.; Markos, C.; Rasmussen, H.K.; Bang, O. Low Loss Polycarbonate Polymer Optical Fiber for High Temperature FBG Humidity Sensing. IEEE Photon. Technol. Lett. 2017, 29, 575–578. [CrossRef] Woyessa, G.; Nielsen, K.; Stefani, A.; Markos, C.; Bang, O. Temperature insensitive hysteresis free highly sensitive polymer optical fiber Bragg grating humidity sensor. Opt. Express 2016, 24, 1206–1213. [CrossRef] [PubMed] Alberto, N.; Tavares, C.; Domingues, M.F.; Correia, S.F.H.; Marques, C.; Antunes, P.; Pinto, J.L.; Ferreira, R.A.S.; Andre, P.S. Relative humidity sensing using micro-cavities produced by the catastrophic fuse effect. Opt. Quant. Electron. 2016, 48, 1–8. [CrossRef] Torres, D.; Elosua, C.; Villatoro, J.; Zubia, J.; Rothhardt, M.; Schuster, K.; Arregui, F. Photonic crystal fiber interferometer coated with a PAH/PAA nanolayer as humidity sensor. Sens. Actuators B Chem. 2017, 242, 1065–1072. [CrossRef] Urrutia, A.; Goicoechea, J.; Ricchiuti, A.L.; Barrera, D.; Sales, S.; Arregui, F.J. Simultaneous measurement of humidity and temperature based on a partially coated optical fiber long period grating. Sens. Actuators B Chem. 2016, 227, 135–141. [CrossRef] Ascorbe, J.; Corres, J.M.; Matias, I.R.; Arregui, F.J. High sensitivity humidity sensor based on cladding-etched optical fiber and lossy mode resonances. Sens. Actuators B Chem. 2016, 242, 7–16. [CrossRef] Alberto, N.J.; Marques, C.A.; Pinto, J.L.; Nogueira, R.N. Three-parameter optical fiber sensor based on a tilted fiber Bragg grating. Appl. Opt. 2010, 49, 6085–6091. [CrossRef] Huang, Y.; Zhu, W.; Li, Z.; Chen, G.; Chen, L.; Zhou, J.; Lin, H.; Guan, J.; Fang, W.; Liu, X.; et al. High-performance fibre-optic humidity sensor based on a side-polished fibre wavelength selectively coupled with graphene oxide film. Sens. Actuators B Chem. 2018, 255, 57–69. [CrossRef] Dissanayake, K.P.W.; Wu, W.; Nguyen, H.; Sun, T.; Grattan, K.T.V. Graphene oxide coated long period grating based fibre optic humidity sensor. In Proceedings of the 2017 25th Optical Fiber Sensors Conference (OFS), Jeju, Korea, 24–28 April 2017. Gao, R.; Lu, D.F.; Cheng, J.; Jiang, Y.; Jiang, L.; Qi, Z.-M. Humidity sensor based on power leakage at resonance wavelengths of a hollow core fiber coated with reduced graphene oxide. Sens. Actuators B Chem. 2016, 222, 618–624. [CrossRef] Dong, X.; Zhang, H.; Liu, B.; Miao, Y. Tilted fiber Bragg gratings: Principle and sensing applications. Photon. Sens. 2011, 1, 6–30. [CrossRef] Albert, J.; Shao, L.; Caucheteur, C. Tilted fiber Bragg grating sensors. Laser Photon. Rev. 2013, 7, 83–108. [CrossRef] Erdogan, T.; Sipe, J.E. Tilted fiber phase gratings. Opt. Soc. Am. A 1996, 13, 296–313. [CrossRef] Li, Y.; Froggatt, M.; Erdogan, T. Volume Current Method for Analysis of Tilted Fiber Gratings. J. Lightwave Technol. 2001, 19, 1580. [CrossRef] Feder, K.S.; Westbrook, P.S.; Ging, J.; Reyes, P.I.; Carver, G.E. In-Fiber Spectrometer Using Tilted Fiber Gratings. Photon. Technol. Lett. 2003, 15, 933–935. [CrossRef] Wang, L.; Fang, N.; Huang, Z. Polyimide-Coated Fiber Bragg Grating Sensors for Humidity Measurements; InTech Open Access: Rijeka, Croatia, 2012. Berrutia, G.; Consalesa, M.; Giordanob, M.; Sansoneb, L.; Petagnac, P.; Buontempod, S.; Breglioe, G.; Cusanoa, A. Radiation hard humidity sensors for high energy physics applications using polyimide-coated fiber Bragg gratings sensors. Sens. Actuators B Chem. 2013, 177, 94–102. [CrossRef] Montero, A.; Aldabaldetreku, G.; Durana, G.; Jorge, I.; de Ocáriz, I.S.; Zubia, J. Influence of Humidity on Fiber Bragg Grating Sensors. Adv. Mater. Sci. Eng. 2014. [CrossRef] Wang, Y.; Shen, C.; Lou, W.; Shentu, F.; Zhong, C.; Dong, X.; Tong, L. Fiber optic relative humidity sensor based on the tilted fiber Bragg grating coated with graphene oxide. Appl. Phys. Lett. 2016, 109, 031107. [CrossRef] Naik, G.; Krishnaswamy, S. Room-Temperature Humidity Sensing Using Graphene Oxide Thin Films. Sci. Res. 2016. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).