sensors Article
A New Hydrogen Sensor Based on SNS Fiber Interferometer with Pd/WO3 Coating Jin-xin Shao 1 1
2
*
ID
, Wen-ge Xie 1
ID
, Xi Song 1
ID
and Ya-nan Zhang 1,2, *
ID
College of Information Science and Engineering, Northeastern University, Shenyang 110819, China; [email protected] (J.S.); [email protected] (W.X.); [email protected] (X.S.) State Key Laboratory of Synthetical Automation for Process Industries, Shenyang 110819, China Correspondence: [email protected]; Tel.: +86-024-8368-7266
Received: 27 July 2017; Accepted: 15 September 2017; Published: 18 September 2017
Abstract: This paper presents a new hydrogen sensor based on a single mode–no core–single mode (SNS) fiber interferometer structure. The surface of the no core fiber (NCF) was coated by Pd/WO3 film to detect the variation of hydrogen concentration. If the hydrogen concentration changes, the refractive index of the Pd/WO3 film as well as the boundary condition for light propagating in the NCF will all be changed, which will then cause a shift into the resonant wavelength of interferometer. Therefore, the hydrogen concentration can be deduced by measuring the shift of the resonant wavelength. Experimental results demonstrated that this proposed sensor had a high detection sensitivity of 1.26857 nm/%, with good linearity and high accuracy (maximum 0.0055% hydrogen volume error). Besides, it also possessed the advantages of simple structure, low cost, good stability, and repeatability. Keywords: fiber interferometer; no core fiber (NCF); hydrogen sensor; Pd/WO3 film
1. Introduction Nowadays, energy has become an increasingly important issue in human development. Scientists believe that hydrogen will become another new source of energy to replace petroleum [1,2]. Although hydrogen has a lot of advantages, such as reproducibility, no pollution, high burning ratio, and so on [3], it is explosive and flammable due to its high diffusion coefficient (0.16 cm2 /s in air), low ignition energy (0.018 mJ), and high combustion heat (285.8 kJ/mol). For example, when hydrogen leaks into the air and the concentration of it reaches 4~74.5%, the mixture will be in an explosive state at room temperature and atmospheric pressure, and it will also be detonated when exposing to open flame [4]. Thus the wide application of hydrogen energy is limited. Therefore, it is of great value to develop a safe and reliable hydrogen sensor with high sensitivity to monitor the hydrogen concentration in real time. Hydrogen sensors can be divided into catalyst (Pellistor) type, thermal conductivity type, electrochemical type, optical fiber type, and so on [5]. Pellistor-typed catalytic hydrogen sensors are based on the theory that combustion of hydrogen and oxygen on the sensor surface can release heat for measuring. However, they have a high error rate, high power consumption, and a high risk of explosion [6]. Thermal conductivity typed sensors for hydrogen analysis were initially used for airships since 1913. The principle of this sensor is to measure heat loss from the detector to the surrounding gas [7]. The detection range can often cover 1–100%, but cannot be used when the concentration of hydrogen is low. Electrochemical hydrogen sensors use the solid electrolyte to react with hydrogen and then to change the current in the circuit [8]. Finally, the concentration of hydrogen can be determined by observing the change of the output voltage. Using an electrical signal, however, is easy to cause an
Sensors 2017, 17, 2144; doi:10.3390/s17092144
www.mdpi.com/journal/sensors
Sensors 2017, 17, 2144
2 of 10
explosion and compromise the measurement. By contrast, the measurement principle of an optical fiber hydrogen sensor is that the reaction between hydrogen and hydrogen sensitive material deposited on the fiber will change some properties of the hydrogen sensitive material, and then the wavelength or intensity of optical signal in the optical fiber will also be changed [9]. By observing the variations of the optical signal, the hydrogen concentration can be deduced. It should be mentioned that the most likely reason for high hydrogen leak explosions is due to some form of electrostatic charging [10]. However, due to the intrinsic safety of light signals, this problem can be avoided in optical fiber hydrogen sensors even when the optical fiber is broken. So it is obvious that optical fiber hydrogen sensors can realize the measurement of hydrogen concentration with high safety and accuracy. Broadly speaking, the measurement technologies of optical fiber hydrogen sensors contain a micro-mirror, evanescent field, surface plasmon resonance (SPR), interference, fiber Bragg grating (FBG), and so on [11]. A micro-mirror sensor was first put forward by M. A. Butler in 1994 [12]. The palladium film coated on the end of optical fiber would react with hydrogen and then the reflectivity of the film will be changed. So the relationship between hydrogen concentration and the intensity of reflected light can be established. The evanescent field sensor was first proposed by Tabib Azar who coated the sensitive material on the fiber to cause an intensity change in the optical signal [13]. However, the light intensity will easily be interrupted by surroundings, which will then induce a measurement error. The surface plasmon resonance (SPR) hydrogen sensor is realized by monitoring the shift of the resonance spectrum under the condition that the hydrogen sensitive film is properly coated onto the surface of a metal film [14]. This sensor has a high sensitivity, but the deposition of the essential metal film is expensive and difficult. The interference sensor was first proposed by Butler who used an optical fiber Mach–Zehnder (MZ) interferometer sensor [15]. When the Pd layer, which is deposited on the sensor, reacts with hydrogen, its volume will increase and the refractive index will decrease, which can cause a phase change in the optical signal [16]. For the FBG coated with the sensitive materials (for example Pd), the reflecting signal of FBG will be changed when the Pd film reacts with hydrogen. The hydrogen sensor based on FBG has a fast response speed, but the measurement sensitivity is relatively low. By utilizing the polishing, etching, or tapering technology, the measurement sensitivity can be enhanced, but the life of sensor will be shortened [17]. In comparison to other technologies, the sensor based on interference theory possesses particular advantages of simple structure, low cost, high sensitivity, and so on. In recent years, a lot of interferometer structures have been proposed [18]. Upon these methods, this paper provides a new thought of hydrogen sensor by simply splicing a no core fiber between two single mode fibers. The surface of the NCF is coated with Pd/WO3 film to detect the variation of hydrogen concentration. Simulation and experimental results demonstrated that this simple structure can realize a high sensitivity of 1.26857 n m/%, with good linearity, high stability, and good responsiveness. 2. Structure and Principle of Sensor Figure 1 shows the schematic structure of our sensing system. The structure consists of an amplified spontaneous emission source (ASE) with an operating wavelength ranging from 1520 nm to 1570 nm, an optical spectrum analyzer (OSA) with spectral resolution of 0.02 nm, and a hydrogen sensing probe. The hydrogen sensing head is constituted by two sections of single mode fiber (SMF) and one section of no core fiber (NCF) with Pd/WO3 coating, which is shown in Figure 2. The core and cladding diameters of SMF and NCF that used are 9/125 µm and 125/125 µm, respectively. These three fibers were spliced together by using a commercial fusion machine. In this structure, the light emitted from ASE feeds into the SMF, and propagates with the fundamental mode. When it travels from SMF to the NCF, multiple high-order modes are excited. Then the fundamental and high-order modes meet in the second welding zone and interfere with each other [19]. Finally, the interference spectrum is received by the OSA. To measure the hydrogen sensing property of the proposed sensor in an experiment, the sensing head was inserted into a closed air chamber. The hydrogen concentration of the air chamber can be controlled by adjusting the flow velocities of hydrogen and nitrogen with
Sensors 2017, 17, 2144
3 of 10
Sensors2017, 2017,17, 17,2144 2144 Sensors
of10 10 33of
two mass flow controllers. In the measurement, the air chamber is evacuated to a vacuum state by a vacuum stateby byaavacuum vacuumpump pumpeach eachtime time beforethe thehydrogen hydrogenconcentration concentrationisischanged. changed.By Bythis this vacuum vacuum state pump each time before the hydrogenbefore concentration is changed. By this way, we can control way, we can control the hydrogen concentration in the air chamber accurately. way, we can control the hydrogen in the air chamber accurately. the hydrogen concentration in the concentration air chamber accurately.
Schematicstructure structureof system. Figure1.1.Schematic ofhydrogen hydrogensensing sensingsystem. system. Figure
Figure2.2.Specific structureof ofSNS SNSsensing sensingstructure. structure. Specificstructure structure. Figure
Accordingto tomultiple multiplehigh-order high-orderLP LP(linear (linearpolarization) polarization)modes modespropagation propagationtheory, theory,βββmmmand andβββnnn According theory, According to multiple high-order areassumed assumedto tobe bethe thelongitudinal longitudinal propagation constants for and order modes inNCF. theNCF. NCF. So are be the longitudinal propagation constants mmand nnorder modes in the So are assumed propagation constants forfor m and n order modes in the So their their relationship can be expressed as [20] their relationship can be expressed as [20] relationship can be expressed as [20] 2
2
u2 −uu2n 2 nu n β − β == uum2mm−− ββmmm−− ββnnn = 2k a22 n
core k000aa2nncore 22k core
(1) (1) (1)
whereaaaisis the core radius ofofthe the NCF, coreis the core refractive index ofNCF, NCF, 2π/λ theisisisthe freewhere NCF, nncore core refractive index of kk00==k2π/λ isisthe freewhere isthe thecore coreradius radiusof the NCF, ncoreisthe is the core refractive index of NCF, is 0 = 2π/λ space wave number, u m and u n are normalized transverse propagation constants and they can be space wavewave number, um and un are normalized transverse propagation constants free-space number, um and un are normalized transverse propagation constantsand andthey theycan canbe be calculatedby bythe thenext nextformula formula[21] [21] calculated calculated by the next formula [21] u x = (2x − 0.5) · π/2 (2) (2xx−−0.5) 0.5)⋅⋅ππ//22 (2) uuxx ==(2 (2) If the phase difference between the two modes is an integral multiple of π, then the constructive interference of these modes occurs. Particularly, the resonant dips of the interference spectrum are thephase phase differencebetween between thetwo twomodes modesisisan anintegral integralmultiple multiple of π, then the constructive IfIfthe generated when difference the phase differencethe is an odd multiple of π [22], namely of π, then the constructive interference of these modes occurs. Particularly, the resonant dips ofthe theinterference interferencespectrum spectrumare are interference of these modes occurs. Particularly, the resonant dips of generatedwhen whenthe thephase phasedifference difference(isis an odd of namely generated an βm −odd β n )multiple ·multiple L = (2Nof +ππ1[22], π namely (3) )[22],
−ββn))⋅⋅LL=（ =（ 2N+1 ） (3) By solving the combined equations (1) and (3), can 2N+1 ） ππ the wavelengths of resonant dips(3) ((ββofmm−Formulas n be expressed as 8Formulas (2N + 1)n(1) a2 (3), the wavelengths of resonant dips By solving solving the the combined combined equations equations of Formulas and By (1)core and (3), the wavelengths of resonant dips λ = of (4) (m − n)[2(m + n) − 1] L canbe beexpressed expressedas as can where L is the length of NCF, and N is a natural number. 22 8(2N++1) 1)nncoreaa 8(2N core (4) λλ == (4) )[2(mm++nn))−−1]1]LL ((mm−−nn)[2(
whereLLisisthe thelength lengthof ofNCF, NCF,and andN Nisisaanatural naturalnumber. number. where So for different constructive interferences, easy to to get get the the wavelength wavelength shift shift of of resonant resonant So for different constructive interferences, itit isis easy spectrum spectrum
Sensors 2017, 17, 2144
4 of 10
So for different constructive interferences, it is easy to get the wavelength shift of resonant spectrum 16ncore a2 ∆λ = λ N − λ N −1 = (5) ( m − n ) · [2( m + n ) − 1] · L Therefore, by coating Pd/WO3 film on the surface of the NCF, we can establish a relationship between hydrogen concentration and the interference spectrum. A common method called the sol-gel method is widely used to manufacture Pd/WO3 film, and the film can be coated on the surface of NCF by dip-coating method. Both of these methods established in our group. In our previous work, we illustrated these methods in detail [23]. These methods have the advantages of low cost, simple fabrication, and high success ratio. Both palladium and WO3 have good hydrogen sensitivity: palladium can absorb hydrogen with high selectivity, while WO3 film has good adhesion and mechanical properties [24]. So if the palladium and WO3 film exposed to the hydrogen, with palladium as the catalyst, hydrogen will be absorbed and decomposed. Hydrogen atoms which transferred to the surface of WO3 by palladium will spread along the inner holes of WO3 , and then cause the variation of both volume and refractive index of Pd/WO3 film. By coating Pd/WO3 on the surface of NCF, the refractive index and stain of NCF surface will change with the variation of hydrogen concentration. However, the variations of ncore , L, and a of NCF will change little and can be ignored in the theory. However, the propagation constants and mode field will be reformed when the refractive index of the Pd/WO3 film is changed. From Formula (3), when the propagation constants are changed, the phase condition of the constructive (or destructive) interference will alter and the natural number, N, will also change. Therefore, the main reason caused the wavelength difference is the variation of the natural number N. In summary, if the hydrogen concentration changes, the refractive index of the sensitive film will also be changed. Then the boundary condition for light propagating in the NCF will also change and cause the shift in resonant wavelength. It has also been proven that the resonant wavelength of this SNS interferometer structure has an approximately linear change with the variation of refractive indices of the sensitive film [25]. So if the refractive index of the sensitive film is linearly proportional to the hydrogen concentration, this sensor will gain a good linearity, and it will be proved by the experiment in the following. 3. Simulation Analysis From the above analyses, it is obvious that this interferometer structure must create an interference phenomenon to obtain the hydrogen concentration. Besides, the larger the wavelength shift in resonance spectrum is, the higher the hydrogen sensor sensitivity will be. So we should first ensure the length of NCF to obtain a better interferometer effect and a higher sensitivity. In scientific research, modeling and simulation are also important parts to prove this hypothesis. So we used professional simulation software called Rsoft to imitate how the resonant spectrum will be influenced when the length of NCF changes. As shown in Figure 3, when the length of NCF changes from 3 cm to 6 cm with an interval of 1 cm, the simulated interference spectra have different forms, which is mainly due to the final interference spectrum being a combination of the interference spectra of different high-order modes and fundamental mode. Therefore, most interference spectra contain many resonant dips of different shapes. For different lengths of NCF, the high-order modes that generate interference may be different, so the interference spectra are also different. Particularly, the interference spectrum with NCF length of 4 cm has only one resonant dip in the wavelength range of 1540–1570 nm, which is mainly due to there being only one high-order mode to interfere with the fundamental mode in this situation. This smooth and pure spectrum is beneficial for sensors to observe the wavelength shift along with hydrogen concentration variation. Therefore, we will further investigate the refractive index property of this interferometer with NCF length around 4 cm in the following sections.
Sensors Sensors2017, 2017,17, 17,2144 2144 Sensors 2017, 17, 2144
55ofof10 10 5 of 10 3cm 3cm 4cm 4cm 5cm 5cm 6cm 6cm
0.25 0.25
Normalized Power Normalized Power
0.20 0.20 0.15 0.15 0.10 0.10 0.05 0.05 0.00 0.00 1540 1540
1550 1550
Wavelength/nm Wavelength/nm
1560 1560
1570 1570
Figure 3. Simulation spectra of SNS interferometer with different lengths of NCF. Figure Simulation spectra SNS interferometer with different lengths NCF. Figure 3. 3. Simulation spectra of of SNS interferometer with different lengths ofof NCF.
As shown in Figure 4, the resonant wavelength of interferometer will shift to the long AsAsshown in Figure4,4, the resonant wavelength of interferometer will shift to wavelength the long shown in Figure wavelength of interferometer will shift to long wavelength direction whenthe theresonant refractive index of the layer increases, which is the consistent with the wavelength direction when the refractive index of the layer increases, which is consistent with the direction when theThe refractive indexbetween of the layer increases, which is consistent with thecan theoretical theoretical analysis. relationship refractive indices and resonant wavelength be seen theoretical analysis. The relationship between refractive indices andwavelength resonant wavelength can be seen 5, analysis. The relationship between refractive indices and resonant can be seen in Figure in Figure 5, which shows high sensitivity of 1524 nm/RIU and good linearity of 0.98495. inwhich Figureshows 5, which shows high sensitivity of 1524and nm/RIU and goodoflinearity of 0.98495. high sensitivity of 1524 nm/RIU good linearity 0.98495. Then, considering that the length of NCF is difficult to control precisely, we compared the Then, considering that the is difficult totocontrol precisely, we the Then, considering that thelength lengthofofNCF NCF difficult control precisely, wecompared compared the refractive index sensitivity of interferometer whenisthe length of NCF is slightly changed from 3.8 cm refractive index sensitivity ofofinterferometer when the length ofofNCF isisslightly changed from 3.8 cm refractive index sensitivity interferometer when the length NCF slightly changed from 3.8 cm to 4.0 cm and then to 4.2 cm. As shown in Figure 6, with the increase of the length of NCF, the change toto4.0 cm and then to 4.2 cm. As shown in Figure 6, with the increase of the length of NCF, the change 4.0 cm and then sensitivity to 4.2 cm. As shown in Figure 6, with the of the of length of NCF, the change of refractive index is very little. So considering theincrease convenience sensor fabrication and ofofrefractive index sensitivity isisvery little. SoSoconsidering the convenience ofofsensor fabrication and refractive index sensitivity very little. considering the convenience sensor fabrication and the error of the welding zone, it is reasonable to choose 4.0 cm as the length of NCF in the following the error ofofthe welding zone, it itisisreasonable totochoose 4.0 cm asasthe length ofofNCF ininthe following the error the welding zone, reasonable choose 4.0 cm the length NCF the following experimental study. experimental experimentalstudy. study.
Normalized Power Normalized Power
0.20 0.20
0.15 0.15
when the NCF is 4.0cm when the NCF is 4.0cm Refractive index of the NCF's surface Refractive index of the NCF's surface 1.9935 1.9935 1.994 1.994 1.9945 1.9945 1.995 1.995 1.9955 1.9955 1.996 1.996
0.10 0.10
0.05 0.05 1550 1550
1555 1555
1560 1560
Spectral wavelength/nm Spectral wavelength/nm
1565 1565
Figure4.4.Resonant Resonantspectra spectraofofinterferometer interferometerunder underdifferent differentrefractive refractiveindices indiceswhen whenthe thelength lengthofofNCF NCF Figure Figure 4. Resonant spectra of interferometer under different refractive indices when the length of NCF cm. isis44cm. is 4 cm.
Sensors 2017, 17, 2144
6 of 10
Sensors 2017, 17, 2144 Sensors 2017, 17, 2144 1564.0
6 of 10 6 of 10
1564.0 1563.5
1563.0 1562.5 Resonant Wavelength/nm
Resonant Wavelength/nm
1563.5 1563.0
1562.5 1562.0
simulation point fitting curve simulation point Linearity(R-Square): 0.98495 fitting sensitivity: 1524curve nm/RIU Linearity(R-Square): 0.98495 sensitivity: 1524 nm/RIU
1562.0 1561.5 1561.5 1561.0 1561.0 1560.5 1560.5 1560.0 1560.0 1559.5 1559.5
1.9935 1.9935
1.9940
1.9945
1.9950
1.9955
Refractive Index/RIU 1.9940 1.9945 1.9950
1.9960
1.9955
1.9960
Refractive Index/RIU
Figure 5. Relationship between resonant wavelength and refractive indices when the length of NCF is 4Figure cm. 5. Relationship between resonant wavelength and refractive indices when the length of NCF is Figure 5. Relationship between resonant wavelength and refractive indices when the length of NCF 4 cm. is 4 cm. 1564 15621564
15581560 15561558
NCF's length 3.8cm NCF's length 4.0cm NCF's length 3.8cm NCF's length 4.2cm NCF's length nm/RIU 4.0cm Sensitivity:1524 NCF's length 4.2cm Sensitivity:1661.1 nm/RIU Sensitivity:1524 nm/RIU Sensitivity:1697.1 nm/RIU Sensitivity:1661.1 nm/RIU Sensitivity:1697.1 nm/RIU
Resonant Wavelength/nm
Resonant Wavelength/nm
15601562
15541556 15521554 15501552 15481550 15461548 15441546 1544
1.994
1.995
1.994 Refractive Index/RIU 1.995
1.996 1.996
Refractive Index/RIU Figure 6. Relationship between resonant wavelength and refractive indices with different lengths Figure 6. Relationship between resonant wavelength and refractive indices with different lengths of NCF. of NCF. Figure 6. Relationship between resonant wavelength and refractive indices with different lengths of NCF.
4. Experimental Test 4. Experimental Test 4. After Experimental Test the procedures we designed, we can get the interference spectra of the implementing After implementing the procedures we designed, we can get the interference spectra of the sensor for different hydrogenthe concentrations in experiment, as shown Figure 7. With the increase implementing procedures we we can getinthe interference of the sensorAfter for different hydrogen concentrations indesigned, experiment, as shown in Figure 7. Withspectra the increase in hydrogen concentration, the resonant wavelength shifts in short wavelength direction. This is sensor for different hydrogen in experiment, as short shownwavelength in Figure 7.direction. With the increase in hydrogen concentration, theconcentrations resonant wavelength shifts in This is because that the refractive index of Pd/WO 3 film will decrease when the hydrogen concentration is in hydrogen concentration, the of resonant shifts inwhen shortthe wavelength This isis because that the refractive index Pd/WOwavelength film will decrease hydrogendirection. concentration increasing. Therefore, the experimental result33isfilm consistent with the simulation result concentration as shown in is because that the refractive index of Pd/WO will decrease when the hydrogen increasing. Therefore, the experimental result is consistent with the simulation result as shown in Figure 4. Similarly, by quantifying the relationship thewith resonant wavelength and hydrogen increasing. Therefore, the experimental result isbetween consistent the simulation result ashydrogen shown in Figure 4. Similarly, by quantifying the relationship between the resonant wavelength and concentration, as shown in Figure 8, we can get the measurement sensitivity ofand hydrogen Figure 4. Similarly, by quantifying the relationship between the resonant wavelength hydrogen concentration, as shown in Figure 8, we can get the measurement sensitivity of hydrogen concentration concentration (1.26857shown nm/%) with good linearity of 0.99641. response timesensitivity of this kindofof hydrogen sensor concentration, in Figure 8, we can theThe measurement (1.26857 nm/%) as with good linearity ofthe 0.99641. Theget response time ofsensing. this kind of Pd/WO sensor primarily primarily depends on the sensitivity of material that is used for The 3 based concentration nm/%) withmaterial good linearity of 0.99641. The response time of this kind of sensor depends on the(1.26857 sensitivity of the the that is for sensing. The Pd/WO hydrogen 3 based hydrogen sensor will enter into steadyofstate in noused more than 30 min, this has been proven in3 our primarily depends on the sensitivity the material that is used for sensing. The Pd/WO based sensor will enter into theBesides, steady state in no more than 30 min, this has been proven in our previous previous manuscript [26]. the response time of hydrogen sensor based on Pd/WO 3 may be hydrogen sensor will enter the steady state in no more than 30 min, this has been proven in our manuscript [26]. Besides, theinto response of hydrogen sensor based on [28], Pd/WO bebeimproved 3 may improved bymanuscript UV-light irradiation [27]the ortime doping graphene quantum dots which tested previous [26]. Besides, response time of hydrogen sensor based on will Pd/WO 3 may be in our future work. improved by UV-light irradiation [27] or doping graphene quantum dots [28], which will be tested in our future work.
Sensors 2017, 17, 2144
7 of 10
Sensors 2017, 17, 2144
7 of 10
Sensors 2017, 17, 2144 by UV-light irradiation
[27] or doping graphene quantum dots [28], which will be
tested7 of in10our
future work. Volume Fraction of Hydrogen
-28 -28 -29 Transmission/dB Transmission/dB
-30 -31 -32 -33 -34
Volume Fraction0‰H of Hydrogen 2
-29
2‰H 0‰H 2 2
-30
4‰H 2‰H 2 2
-31
8‰H 6‰H 2 2
6‰H 4‰H 2 2 10‰H 8‰H 2 2
-32
10‰H2
-33 -34 -35
-35 1558
1559
1558
1559
1560
1561
1560 1561 Wavelength/nm
1562 1562
Wavelength/nm
Figure 7. Experimental spectra of sensor for different hydrogen concentrations. Figure 7. Experimental spectra of sensor for different hydrogen concentrations. Figure 7. Experimental spectra of sensor for different hydrogen concentrations. 1561.2 1561.2 1561.0
Experiment Point Fitting Curve Experiment Point Linearity(R-Square): 0.99641 Fitting Curve sensitivity: 1.26857 nm/1% Linearity(R-Square): 0.99641 sensitivity: 1.26857 nm/1%
Resonant Wavelength/nm
Resonant Wavelength/nm
1561.0 1560.8 1560.8 1560.6 1560.6 1560.4 1560.4 1560.2 1560.2 1560.0 1560.0 1559.8 1559.8
0.0 0.0
0.2
0.4
0.6
0.8
1.0
0.2
0.4Fraction0.6 0.8 1.0 Volume of Hydrogen/% Volume Fraction of Hydrogen/% Figure Relationship between between resonant resonant wavelength Figure 8. 8. Relationship wavelength and and hydrogen hydrogen concentration. concentration.
Figure 8. Relationship between resonant wavelength and hydrogen concentration.
To investigate investigatethethe stability ofproposed the proposed hydrogen sensor, we the observed resonant To stability of the hydrogen sensor, we observed resonantthe wavelength To investigate the every stability ofhydrogen the proposed hydrogen sensor, observed wavelength of sensor min when the concentration hydrogen concentration ofwe air chamber isthe stable 1%min. for of sensor every 5 min when5 the of air chamber is stable at 1% resonant forat30 wavelength of sensor every 5 min when the hydrogen concentration of air chamber is stable at 1% for 30 min. From9,Figure 9, see we that can see the maximum of the resonant wavelength 0.007 nm, From Figure we can thethat maximum change change of the resonant wavelength is 0.007isnm, which 30 will min. From Figure 9, we 0.0055% can see that the maximum change thecan resonant wavelength 0.007 nm,has which willonly cause only hydrogen volume error. of Therefore, we can say thatisthe proposed cause 0.0055% hydrogen volume error. Therefore, we say that the proposed sensor which will cause only 0.0055% volume error. Therefore, we little can has say that influence the sensor has goodBesides, stability. has been demonstrated that humidity little on this good stability. it Besides, has hydrogen beenitdemonstrated that humidity has influence on proposed this kind of sensor has good stability. Besides, it has been demonstrated that humidity has little influence on this of kind of[29], sensor [29],means whichthat means that the humidity compensation may not befor required most sensor which the humidity compensation may not be required most offor hydrogen kind of sensor [29], which means that the humidity compensation may beof required for of on hydrogen sensing applications. Besides, a Teflon can be used tonot decrease effect of H2O sensing applications. Besides, a Teflon layer can be layer used to decrease the effect Hthe themost hydrogen 2 O on hydrogen sensing applications. Besides, a Teflon layer can be used to decrease the effect of H 2 O on the hydrogen sensitive coating layer if the humidity variation is large [30]. sensitive coating layer if the humidity variation is large [30]. the hydrogen sensitive coating layer if the humidity variation is large [30].
Sensors2017, 2017, 17,2144 2144 Sensors Sensors 2017,17, 17, 2144
of10 10 888of of 10
1560.1 1560.1 Experiment Point Experiment Point the Average Line the Average Line Data Extreme Difference: 0.007nm Data Extreme Difference: 0.007nm
ResonantWavelength/nm Wavelength/nm Resonant
1560.0 1560.0
1559.9 1559.9
1559.8 1559.8
1559.7 1559.7
0 0
5 5
10 15 20 10 15 20 Time after Steady-State/min Time after Steady-State/min
25 25
30 30
Figure 9. Stability curve of the proposed hydrogen sensor. Figure Figure9.9.Stability Stabilitycurve curveof ofthe theproposed proposedhydrogen hydrogensensor. sensor.
Then we further investigate repeatability of For one we the Then investigate thethe repeatability of the sensor. For one we conducted the same Thenwe wefurther further investigate the repeatability of the the sensor. sensor. Forsensor, one sensor, sensor, we conducted conducted the same experimental tests three times. Then we got the sensitivity curves of sensor for each test, as experimental tests three times. Then we got the sensitivity curves of sensor for each test, as shown same experimental tests three times. Then we got the sensitivity curves of sensor for each test,in as shown in Figure 10. From this figure, we can confirm that the proposed sensor also has a good Figure 10. From this figure, we can confirm that the proposed sensor also has a good repeatability. We shown in Figure 10. From this figure, we can confirm that the proposed sensor also has a good repeatability. We can also see the lines parallel but the of lines can also see that the three are almost parallel but the intercepts these lines increase regularly. repeatability. We can alsolines seethat that thethree three linesare arealmost almost parallelof but theintercepts intercepts ofthese these lines increase regularly. Namely, the origins of the resonant wavelengths are not the same in each Namely, the origins of the resonant wavelengths are not the same in each measurement, which is mainly increase regularly. Namely, the origins of the resonant wavelengths are not the same in each measurement, due the isisdifferent This problem can due to that the which temperature is different in each test. This problem can in be resolved analyzing measurement, whichisismainly mainly dueto tothat that thetemperature temperature different ineach eachtest. test.by This problemthe can be resolved by analyzing the relationship between the change of the resonant wavelength and the relationship between the change of the resonant wavelength and the change of hydrogen concentration. be resolved by analyzing the relationship between the change of the resonant wavelength and the change concentration. Besides, temperature influence by Besides, this temperature influence can also bethis compensated by measuring the also hydrogen concentration changeof ofhydrogen hydrogen concentration. Besides, this temperature influencecan can alsobe becompensated compensated by measuring the hydrogen concentration and temperature, simultaneously [31]. and temperature, simultaneously [31]. measuring the hydrogen concentration and temperature, simultaneously [31]. First experiment First experiment Second experiment Second experiment Third experiment Third experiment Sensitivity: 1.23714 nm/1% Sensitivity: 1.23714 nm/1% Sensitivity: 1.19714 nm/1% Sensitivity: 1.19714 nm/1% Sensitivity: 1.26857 nm/1% Sensitivity: 1.26857 nm/1%
ResonantWavelength/nm Wavelength/nm Resonant
1561.0 1561.0
1560.5 1560.5
1560.0 1560.0
1559.5 1559.5
0.0 0.0
0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8 Volume Fraction of Hydrogen/% Volume Fraction of Hydrogen/%
1.0 1.0
Figure 10. Repeatability curve of the proposed hydrogen sensor. Figure Figure10. 10.Repeatability Repeatabilitycurve curveof ofthe theproposed proposedhydrogen hydrogensensor. sensor.
5. 5.5.Conclusions Conclusions Conclusions In summary, this paper proposed an in-line SNF interferometer structure to measure hydrogen In Insummary, summary,this thispaper paperproposed proposedan anin-line in-lineSNF SNFinterferometer interferometerstructure structureto tomeasure measurehydrogen hydrogen concentration. Then the hydrogen sensing properties of this structure had beenbyproven by concentration. Then the hydrogen sensing properties of this structure had been proven theoretical concentration. Then the hydrogen sensing properties of this structure had been proven by theoretical theoretical derivation, simulation calculation, and experimental test. Results demonstrated that derivation, derivation, simulation simulation calculation, calculation, and and experimental experimental test. test. Results Results demonstrated demonstrated that that the the proposed proposed sensor could achieve high-sensitivity hydrogen concentration detection (1.26857 nm/%). sensor could achieve high-sensitivity hydrogen concentration detection (1.26857 nm/%). Also, Also, the the
Sensors 2017, 17, 2144
9 of 10
the proposed sensor could achieve high-sensitivity hydrogen concentration detection (1.26857 nm/%). Also, the proposed sensor shows high precision and accuracy (maximum 0.0055% hydrogen volume error). Meanwhile, it has the advantages of simple structure, low cost, good linearity, stability, and repeatability. Acknowledgments: This work was supported in part by the National Science Foundation for Distinguished Young Scholars of China under grant 61425003, the National Natural Science Foundation of China under grants 61703080 and 61273059; the Fundamental Research Funds for the Central Universities under grant N160404012; the Liaoning Province Natural Science Foundation under grant 20170540314; the major scientific project of undergraduate under grant ZD1712; and State Key Laboratory of Synthetical Automation for Process Industries under grant 2013ZCX09. Author Contributions: The presented work is a result of the intellectual contribution of the whole team. Jin-xin Shao designed the sensor structure and wrote the manuscript; Wen-ge Xie built the simulation model and analyzed the simulation results; Xi Song performed the experiments and analyzed the experimental results; Ya-nan Zhang guided the experiments and gave some advice. All authors contributed and approved the final manuscript. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Jacobson, M.Z.; Colella, W.G.; Golden, D.M. Clearing the air and improving health with hydrogen fuel-cell vehicles. Science 2005, 308, 1901–1905. [CrossRef] [PubMed] Dedes, E.K.; Hudson, D.A.; Turnock, S.R. Assessing the potential of hybrid energy technology to reduce exhaust emissions from global shipping. Energy Policy 2012, 40, 204–218. [CrossRef] Dincer, I. Renewable energy and sustainable development: A crucial review. Renew. Sustain. Energy Rev. 2000, 4, 157–175. [CrossRef] Silva, S.F.; Coelho, L.; Frazão, O. A review of palladium-based fiber-optic sensors for molecular hydrogen detection. IEEE Sens. J. 2012, 12, 93–102. [CrossRef] Hübert, T.; Boon-Brett, L.; Black, G.; Banach, U. Hydrogen sensors—A review. Sens. Actuators B Chem. 2011, 157, 329–352. [CrossRef] Firth, J.G.; Jones, A.; Jones, T.A. The principles of the detection of flammable atmospheres by catalytic devices. Combust. Flame 1973, 20, 303–311. [CrossRef] Pollak-Diener, G.; Obermeier, E. Heat-conduction microsensor based on silicon technology for the analysis of two-and three-component gas mixtures. Sens. Actuators B Chem. 1993, 13, 345–347. [CrossRef] Korotcenkov, G.; Han, S.D.; Stetter, J.R. Review of electrochemical hydrogen sensors. Chem. Rev. 2009, 109, 1402–1433. [CrossRef] [PubMed] Dai, J.; Zhu, L.; Wang, G. Optical fiber grating hydrogen sensors: A review. Sensors 2017, 17, 577. [CrossRef] [PubMed] Astbury, G.R.; Hawksworth, S.J. Spontaneous ignition of hydrogen leaks: A review of postulated mechanisms. Int. J. Hydrogen Energy 2007, 32, 2178–2185. [CrossRef] Yang, M.; Dai, J. Fiber optic hydrogen sensors: A review. Photonic Sens. 2014, 4, 300–324. [CrossRef] Butler, M.A. Micromirror optical-fiber hydrogen sensor. Sens. Actuators B Chem. 1994, 22, 155–163. [CrossRef] Tabib, A.M.; Sutapun, B.; Petrick, R. Highly sensitive hydrogen sensors using palladium coated fiber optics with exposed cores and evanescent field interactions. Sens. Actuators B Chem. 1999, 56, 158–163. [CrossRef] Bevenot, X.; Trouillet, A.; Veillas, C. Surface plasmon resonance hydrogen sensor using an optical fibre. Meas. Sci. Technol. 2001, 13, 118. [CrossRef] Butler, M.A.; Ginley, D.S. Hydrogen sensing with palladium-coated optical fibers. J. Appl. Phys. 1988, 64, 3706–3712. [CrossRef] Yang, M.; Sun, Y.; Zhang, D. Using Pd/WO3 composite thin films as sensing materials for optical fiber hydrogen sensors. Sens. Actuators B Chem. 2010, 143, 750–753. [CrossRef] Dai, J.; Yang, M.; Yu, X. Greatly etched fiber Bragg grating hydrogen sensor with Pd/Ni composite film as sensing material. Sens. Actuators B Chem. 2012, 174, 253–257. [CrossRef] Cai, L.; Zhao, Y.; Li, X.G. Applications of modal interferences in optical fiber sensors based on mismatch methods. Instrum. Sci. Technol. 2015, 43, 1–20. [CrossRef]
Sensors 2017, 17, 2144
19. 20. 21. 22. 23. 24. 25. 26.
27. 28. 29.
30. 31.
10 of 10
Raghunandhan, R.; Chen, L.H.; Long, H.Y. Chitosan/PAA based fiber-optic interferometric sensor for heavy metal ions detection. Sens. Actuators B Chem. 2016, 233, 31–38. [CrossRef] Mohammed, W.S.; Mehta, A.; Johnson, E.G. Wavelength tunable fiber lens based on multimode interference. J. Lightwave Technol. 2004, 22, 469. [CrossRef] Zhao, Y.; Cai, L.; Li, X.G.; Meng, F.C.; Zhao, Z. Investigation of the high sensitivity RI sensor based on SMS fiber structure. Sens. Actuators A Phys. 2014, 205, 186–190. [CrossRef] Zhao, Y.; Li, X.G.; Cai, L. A reflective intensity modulated fiber tilt angle sensor based on an all-photonic crystal fiber interferometer. Sens. Actuators A Phys. 2016, 244, 106–111. [CrossRef] Zhang, Y.N.; Peng, H.; Zhou, T.; Zhang, L.; Zhang, Y.; Zhao, Y. Hydrogen sensor based on high-birefringence fiber loop mirror with sol-gel Pd/WO3 coating. Sens. Actuators B Chem. 2017, 248, 71–76. [CrossRef] Fardindoost, S.; Rahimi, F.; Ghasempour, R. Pd doped WO3 films prepared by sol-gel process for hydrogen sensing. Int. J. Hydrogen Energy 2010, 35, 854–860. [CrossRef] Liu, X.; Zhang, X.; Liu, Y. Multi-point fiber-optic refractive index sensor by using coreless fibers. Opt. Commun. 2016, 365, 168–172. [CrossRef] Zhang, Y.; Wu, Q.; Peng, H. Photonic crystal fiber modal interferometer with Pd/WO3 coating for real-time monitoring of dissolved hydrogen concentration in transformer oil. Rev. Sci. Instrum. 2016, 87, 125002. [CrossRef] [PubMed] Zhong, X.X.; Yang, M.H.; Huang, C.J.; Wang, G.P.; Dai, J.X.; Bai, W. Water photolysis effect on the long-term stability of fiber optic hydrogen sensor with Pt/WO3 . Sci. Rep. 2016, 6, 39160. [CrossRef] [PubMed] Fardindoost, S.; Zad, A.I.; Hosseini, Z.S.; Hatamie, S. Detecting hydrogen using graphene quantum dots/WO3 thin films. Mater. Res. Express 2016, 3, 116407. [CrossRef] Dai, J.; Yang, M.; Yang, Z.; Li, Z.; Wang, Y.; Wang, G.; Zhang, Y.; Zhuang, Z. Performance of fiber Bragg grating hydrogen sensor coated with Pt-loaded WO3 coating. Sens. Actuators B Chem. 2014, 190, 657–663. [CrossRef] Coelho, L.; de Almeida, J.; Santos, J.L. Fiber optic hydrogen sensor based on an etched Bragg grating coated with palladium. Appl. Opt. 2015, 54, 10342–10348. [CrossRef] [PubMed] Wang, M.; Wang, D.N.; Yang, M.; Cheng, J.; Li, J. In-line Mach-Zehnder Interferometer and FBG with Pd film for simultaneous hydrogen and temperature detection. Sens. Actuators B Chem. 2014, 202, 893–896. [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/).
A New Hydrogen Sensor Based on SNS Fiber Interferometer with Pd/WO3 Coating Jin-xin Shao 1 1
2
*
ID
, Wen-ge Xie 1
ID
, Xi Song 1
ID
and Ya-nan Zhang 1,2, *
ID
College of Information Science and Engineering, Northeastern University, Shenyang 110819, China; [email protected] (J.S.); [email protected] (W.X.); [email protected] (X.S.) State Key Laboratory of Synthetical Automation for Process Industries, Shenyang 110819, China Correspondence: [email protected]; Tel.: +86-024-8368-7266
Received: 27 July 2017; Accepted: 15 September 2017; Published: 18 September 2017
Abstract: This paper presents a new hydrogen sensor based on a single mode–no core–single mode (SNS) fiber interferometer structure. The surface of the no core fiber (NCF) was coated by Pd/WO3 film to detect the variation of hydrogen concentration. If the hydrogen concentration changes, the refractive index of the Pd/WO3 film as well as the boundary condition for light propagating in the NCF will all be changed, which will then cause a shift into the resonant wavelength of interferometer. Therefore, the hydrogen concentration can be deduced by measuring the shift of the resonant wavelength. Experimental results demonstrated that this proposed sensor had a high detection sensitivity of 1.26857 nm/%, with good linearity and high accuracy (maximum 0.0055% hydrogen volume error). Besides, it also possessed the advantages of simple structure, low cost, good stability, and repeatability. Keywords: fiber interferometer; no core fiber (NCF); hydrogen sensor; Pd/WO3 film
1. Introduction Nowadays, energy has become an increasingly important issue in human development. Scientists believe that hydrogen will become another new source of energy to replace petroleum [1,2]. Although hydrogen has a lot of advantages, such as reproducibility, no pollution, high burning ratio, and so on [3], it is explosive and flammable due to its high diffusion coefficient (0.16 cm2 /s in air), low ignition energy (0.018 mJ), and high combustion heat (285.8 kJ/mol). For example, when hydrogen leaks into the air and the concentration of it reaches 4~74.5%, the mixture will be in an explosive state at room temperature and atmospheric pressure, and it will also be detonated when exposing to open flame [4]. Thus the wide application of hydrogen energy is limited. Therefore, it is of great value to develop a safe and reliable hydrogen sensor with high sensitivity to monitor the hydrogen concentration in real time. Hydrogen sensors can be divided into catalyst (Pellistor) type, thermal conductivity type, electrochemical type, optical fiber type, and so on [5]. Pellistor-typed catalytic hydrogen sensors are based on the theory that combustion of hydrogen and oxygen on the sensor surface can release heat for measuring. However, they have a high error rate, high power consumption, and a high risk of explosion [6]. Thermal conductivity typed sensors for hydrogen analysis were initially used for airships since 1913. The principle of this sensor is to measure heat loss from the detector to the surrounding gas [7]. The detection range can often cover 1–100%, but cannot be used when the concentration of hydrogen is low. Electrochemical hydrogen sensors use the solid electrolyte to react with hydrogen and then to change the current in the circuit [8]. Finally, the concentration of hydrogen can be determined by observing the change of the output voltage. Using an electrical signal, however, is easy to cause an
Sensors 2017, 17, 2144; doi:10.3390/s17092144
www.mdpi.com/journal/sensors
Sensors 2017, 17, 2144
2 of 10
explosion and compromise the measurement. By contrast, the measurement principle of an optical fiber hydrogen sensor is that the reaction between hydrogen and hydrogen sensitive material deposited on the fiber will change some properties of the hydrogen sensitive material, and then the wavelength or intensity of optical signal in the optical fiber will also be changed [9]. By observing the variations of the optical signal, the hydrogen concentration can be deduced. It should be mentioned that the most likely reason for high hydrogen leak explosions is due to some form of electrostatic charging [10]. However, due to the intrinsic safety of light signals, this problem can be avoided in optical fiber hydrogen sensors even when the optical fiber is broken. So it is obvious that optical fiber hydrogen sensors can realize the measurement of hydrogen concentration with high safety and accuracy. Broadly speaking, the measurement technologies of optical fiber hydrogen sensors contain a micro-mirror, evanescent field, surface plasmon resonance (SPR), interference, fiber Bragg grating (FBG), and so on [11]. A micro-mirror sensor was first put forward by M. A. Butler in 1994 [12]. The palladium film coated on the end of optical fiber would react with hydrogen and then the reflectivity of the film will be changed. So the relationship between hydrogen concentration and the intensity of reflected light can be established. The evanescent field sensor was first proposed by Tabib Azar who coated the sensitive material on the fiber to cause an intensity change in the optical signal [13]. However, the light intensity will easily be interrupted by surroundings, which will then induce a measurement error. The surface plasmon resonance (SPR) hydrogen sensor is realized by monitoring the shift of the resonance spectrum under the condition that the hydrogen sensitive film is properly coated onto the surface of a metal film [14]. This sensor has a high sensitivity, but the deposition of the essential metal film is expensive and difficult. The interference sensor was first proposed by Butler who used an optical fiber Mach–Zehnder (MZ) interferometer sensor [15]. When the Pd layer, which is deposited on the sensor, reacts with hydrogen, its volume will increase and the refractive index will decrease, which can cause a phase change in the optical signal [16]. For the FBG coated with the sensitive materials (for example Pd), the reflecting signal of FBG will be changed when the Pd film reacts with hydrogen. The hydrogen sensor based on FBG has a fast response speed, but the measurement sensitivity is relatively low. By utilizing the polishing, etching, or tapering technology, the measurement sensitivity can be enhanced, but the life of sensor will be shortened [17]. In comparison to other technologies, the sensor based on interference theory possesses particular advantages of simple structure, low cost, high sensitivity, and so on. In recent years, a lot of interferometer structures have been proposed [18]. Upon these methods, this paper provides a new thought of hydrogen sensor by simply splicing a no core fiber between two single mode fibers. The surface of the NCF is coated with Pd/WO3 film to detect the variation of hydrogen concentration. Simulation and experimental results demonstrated that this simple structure can realize a high sensitivity of 1.26857 n m/%, with good linearity, high stability, and good responsiveness. 2. Structure and Principle of Sensor Figure 1 shows the schematic structure of our sensing system. The structure consists of an amplified spontaneous emission source (ASE) with an operating wavelength ranging from 1520 nm to 1570 nm, an optical spectrum analyzer (OSA) with spectral resolution of 0.02 nm, and a hydrogen sensing probe. The hydrogen sensing head is constituted by two sections of single mode fiber (SMF) and one section of no core fiber (NCF) with Pd/WO3 coating, which is shown in Figure 2. The core and cladding diameters of SMF and NCF that used are 9/125 µm and 125/125 µm, respectively. These three fibers were spliced together by using a commercial fusion machine. In this structure, the light emitted from ASE feeds into the SMF, and propagates with the fundamental mode. When it travels from SMF to the NCF, multiple high-order modes are excited. Then the fundamental and high-order modes meet in the second welding zone and interfere with each other [19]. Finally, the interference spectrum is received by the OSA. To measure the hydrogen sensing property of the proposed sensor in an experiment, the sensing head was inserted into a closed air chamber. The hydrogen concentration of the air chamber can be controlled by adjusting the flow velocities of hydrogen and nitrogen with
Sensors 2017, 17, 2144
3 of 10
Sensors2017, 2017,17, 17,2144 2144 Sensors
of10 10 33of
two mass flow controllers. In the measurement, the air chamber is evacuated to a vacuum state by a vacuum stateby byaavacuum vacuumpump pumpeach eachtime time beforethe thehydrogen hydrogenconcentration concentrationisischanged. changed.By Bythis this vacuum vacuum state pump each time before the hydrogenbefore concentration is changed. By this way, we can control way, we can control the hydrogen concentration in the air chamber accurately. way, we can control the hydrogen in the air chamber accurately. the hydrogen concentration in the concentration air chamber accurately.
Schematicstructure structureof system. Figure1.1.Schematic ofhydrogen hydrogensensing sensingsystem. system. Figure
Figure2.2.Specific structureof ofSNS SNSsensing sensingstructure. structure. Specificstructure structure. Figure
Accordingto tomultiple multiplehigh-order high-orderLP LP(linear (linearpolarization) polarization)modes modespropagation propagationtheory, theory,βββmmmand andβββnnn According theory, According to multiple high-order areassumed assumedto tobe bethe thelongitudinal longitudinal propagation constants for and order modes inNCF. theNCF. NCF. So are be the longitudinal propagation constants mmand nnorder modes in the So are assumed propagation constants forfor m and n order modes in the So their their relationship can be expressed as [20] their relationship can be expressed as [20] relationship can be expressed as [20] 2
2
u2 −uu2n 2 nu n β − β == uum2mm−− ββmmm−− ββnnn = 2k a22 n
core k000aa2nncore 22k core
(1) (1) (1)
whereaaaisis the core radius ofofthe the NCF, coreis the core refractive index ofNCF, NCF, 2π/λ theisisisthe freewhere NCF, nncore core refractive index of kk00==k2π/λ isisthe freewhere isthe thecore coreradius radiusof the NCF, ncoreisthe is the core refractive index of NCF, is 0 = 2π/λ space wave number, u m and u n are normalized transverse propagation constants and they can be space wavewave number, um and un are normalized transverse propagation constants free-space number, um and un are normalized transverse propagation constantsand andthey theycan canbe be calculatedby bythe thenext nextformula formula[21] [21] calculated calculated by the next formula [21] u x = (2x − 0.5) · π/2 (2) (2xx−−0.5) 0.5)⋅⋅ππ//22 (2) uuxx ==(2 (2) If the phase difference between the two modes is an integral multiple of π, then the constructive interference of these modes occurs. Particularly, the resonant dips of the interference spectrum are thephase phase differencebetween between thetwo twomodes modesisisan anintegral integralmultiple multiple of π, then the constructive IfIfthe generated when difference the phase differencethe is an odd multiple of π [22], namely of π, then the constructive interference of these modes occurs. Particularly, the resonant dips ofthe theinterference interferencespectrum spectrumare are interference of these modes occurs. Particularly, the resonant dips of generatedwhen whenthe thephase phasedifference difference(isis an odd of namely generated an βm −odd β n )multiple ·multiple L = (2Nof +ππ1[22], π namely (3) )[22],
−ββn))⋅⋅LL=（ =（ 2N+1 ） (3) By solving the combined equations (1) and (3), can 2N+1 ） ππ the wavelengths of resonant dips(3) ((ββofmm−Formulas n be expressed as 8Formulas (2N + 1)n(1) a2 (3), the wavelengths of resonant dips By solving solving the the combined combined equations equations of Formulas and By (1)core and (3), the wavelengths of resonant dips λ = of (4) (m − n)[2(m + n) − 1] L canbe beexpressed expressedas as can where L is the length of NCF, and N is a natural number. 22 8(2N++1) 1)nncoreaa 8(2N core (4) λλ == (4) )[2(mm++nn))−−1]1]LL ((mm−−nn)[2(
whereLLisisthe thelength lengthof ofNCF, NCF,and andN Nisisaanatural naturalnumber. number. where So for different constructive interferences, easy to to get get the the wavelength wavelength shift shift of of resonant resonant So for different constructive interferences, itit isis easy spectrum spectrum
Sensors 2017, 17, 2144
4 of 10
So for different constructive interferences, it is easy to get the wavelength shift of resonant spectrum 16ncore a2 ∆λ = λ N − λ N −1 = (5) ( m − n ) · [2( m + n ) − 1] · L Therefore, by coating Pd/WO3 film on the surface of the NCF, we can establish a relationship between hydrogen concentration and the interference spectrum. A common method called the sol-gel method is widely used to manufacture Pd/WO3 film, and the film can be coated on the surface of NCF by dip-coating method. Both of these methods established in our group. In our previous work, we illustrated these methods in detail [23]. These methods have the advantages of low cost, simple fabrication, and high success ratio. Both palladium and WO3 have good hydrogen sensitivity: palladium can absorb hydrogen with high selectivity, while WO3 film has good adhesion and mechanical properties [24]. So if the palladium and WO3 film exposed to the hydrogen, with palladium as the catalyst, hydrogen will be absorbed and decomposed. Hydrogen atoms which transferred to the surface of WO3 by palladium will spread along the inner holes of WO3 , and then cause the variation of both volume and refractive index of Pd/WO3 film. By coating Pd/WO3 on the surface of NCF, the refractive index and stain of NCF surface will change with the variation of hydrogen concentration. However, the variations of ncore , L, and a of NCF will change little and can be ignored in the theory. However, the propagation constants and mode field will be reformed when the refractive index of the Pd/WO3 film is changed. From Formula (3), when the propagation constants are changed, the phase condition of the constructive (or destructive) interference will alter and the natural number, N, will also change. Therefore, the main reason caused the wavelength difference is the variation of the natural number N. In summary, if the hydrogen concentration changes, the refractive index of the sensitive film will also be changed. Then the boundary condition for light propagating in the NCF will also change and cause the shift in resonant wavelength. It has also been proven that the resonant wavelength of this SNS interferometer structure has an approximately linear change with the variation of refractive indices of the sensitive film [25]. So if the refractive index of the sensitive film is linearly proportional to the hydrogen concentration, this sensor will gain a good linearity, and it will be proved by the experiment in the following. 3. Simulation Analysis From the above analyses, it is obvious that this interferometer structure must create an interference phenomenon to obtain the hydrogen concentration. Besides, the larger the wavelength shift in resonance spectrum is, the higher the hydrogen sensor sensitivity will be. So we should first ensure the length of NCF to obtain a better interferometer effect and a higher sensitivity. In scientific research, modeling and simulation are also important parts to prove this hypothesis. So we used professional simulation software called Rsoft to imitate how the resonant spectrum will be influenced when the length of NCF changes. As shown in Figure 3, when the length of NCF changes from 3 cm to 6 cm with an interval of 1 cm, the simulated interference spectra have different forms, which is mainly due to the final interference spectrum being a combination of the interference spectra of different high-order modes and fundamental mode. Therefore, most interference spectra contain many resonant dips of different shapes. For different lengths of NCF, the high-order modes that generate interference may be different, so the interference spectra are also different. Particularly, the interference spectrum with NCF length of 4 cm has only one resonant dip in the wavelength range of 1540–1570 nm, which is mainly due to there being only one high-order mode to interfere with the fundamental mode in this situation. This smooth and pure spectrum is beneficial for sensors to observe the wavelength shift along with hydrogen concentration variation. Therefore, we will further investigate the refractive index property of this interferometer with NCF length around 4 cm in the following sections.
Sensors Sensors2017, 2017,17, 17,2144 2144 Sensors 2017, 17, 2144
55ofof10 10 5 of 10 3cm 3cm 4cm 4cm 5cm 5cm 6cm 6cm
0.25 0.25
Normalized Power Normalized Power
0.20 0.20 0.15 0.15 0.10 0.10 0.05 0.05 0.00 0.00 1540 1540
1550 1550
Wavelength/nm Wavelength/nm
1560 1560
1570 1570
Figure 3. Simulation spectra of SNS interferometer with different lengths of NCF. Figure Simulation spectra SNS interferometer with different lengths NCF. Figure 3. 3. Simulation spectra of of SNS interferometer with different lengths ofof NCF.
As shown in Figure 4, the resonant wavelength of interferometer will shift to the long AsAsshown in Figure4,4, the resonant wavelength of interferometer will shift to wavelength the long shown in Figure wavelength of interferometer will shift to long wavelength direction whenthe theresonant refractive index of the layer increases, which is the consistent with the wavelength direction when the refractive index of the layer increases, which is consistent with the direction when theThe refractive indexbetween of the layer increases, which is consistent with thecan theoretical theoretical analysis. relationship refractive indices and resonant wavelength be seen theoretical analysis. The relationship between refractive indices andwavelength resonant wavelength can be seen 5, analysis. The relationship between refractive indices and resonant can be seen in Figure in Figure 5, which shows high sensitivity of 1524 nm/RIU and good linearity of 0.98495. inwhich Figureshows 5, which shows high sensitivity of 1524and nm/RIU and goodoflinearity of 0.98495. high sensitivity of 1524 nm/RIU good linearity 0.98495. Then, considering that the length of NCF is difficult to control precisely, we compared the Then, considering that the is difficult totocontrol precisely, we the Then, considering that thelength lengthofofNCF NCF difficult control precisely, wecompared compared the refractive index sensitivity of interferometer whenisthe length of NCF is slightly changed from 3.8 cm refractive index sensitivity ofofinterferometer when the length ofofNCF isisslightly changed from 3.8 cm refractive index sensitivity interferometer when the length NCF slightly changed from 3.8 cm to 4.0 cm and then to 4.2 cm. As shown in Figure 6, with the increase of the length of NCF, the change toto4.0 cm and then to 4.2 cm. As shown in Figure 6, with the increase of the length of NCF, the change 4.0 cm and then sensitivity to 4.2 cm. As shown in Figure 6, with the of the of length of NCF, the change of refractive index is very little. So considering theincrease convenience sensor fabrication and ofofrefractive index sensitivity isisvery little. SoSoconsidering the convenience ofofsensor fabrication and refractive index sensitivity very little. considering the convenience sensor fabrication and the error of the welding zone, it is reasonable to choose 4.0 cm as the length of NCF in the following the error ofofthe welding zone, it itisisreasonable totochoose 4.0 cm asasthe length ofofNCF ininthe following the error the welding zone, reasonable choose 4.0 cm the length NCF the following experimental study. experimental experimentalstudy. study.
Normalized Power Normalized Power
0.20 0.20
0.15 0.15
when the NCF is 4.0cm when the NCF is 4.0cm Refractive index of the NCF's surface Refractive index of the NCF's surface 1.9935 1.9935 1.994 1.994 1.9945 1.9945 1.995 1.995 1.9955 1.9955 1.996 1.996
0.10 0.10
0.05 0.05 1550 1550
1555 1555
1560 1560
Spectral wavelength/nm Spectral wavelength/nm
1565 1565
Figure4.4.Resonant Resonantspectra spectraofofinterferometer interferometerunder underdifferent differentrefractive refractiveindices indiceswhen whenthe thelength lengthofofNCF NCF Figure Figure 4. Resonant spectra of interferometer under different refractive indices when the length of NCF cm. isis44cm. is 4 cm.
Sensors 2017, 17, 2144
6 of 10
Sensors 2017, 17, 2144 Sensors 2017, 17, 2144 1564.0
6 of 10 6 of 10
1564.0 1563.5
1563.0 1562.5 Resonant Wavelength/nm
Resonant Wavelength/nm
1563.5 1563.0
1562.5 1562.0
simulation point fitting curve simulation point Linearity(R-Square): 0.98495 fitting sensitivity: 1524curve nm/RIU Linearity(R-Square): 0.98495 sensitivity: 1524 nm/RIU
1562.0 1561.5 1561.5 1561.0 1561.0 1560.5 1560.5 1560.0 1560.0 1559.5 1559.5
1.9935 1.9935
1.9940
1.9945
1.9950
1.9955
Refractive Index/RIU 1.9940 1.9945 1.9950
1.9960
1.9955
1.9960
Refractive Index/RIU
Figure 5. Relationship between resonant wavelength and refractive indices when the length of NCF is 4Figure cm. 5. Relationship between resonant wavelength and refractive indices when the length of NCF is Figure 5. Relationship between resonant wavelength and refractive indices when the length of NCF 4 cm. is 4 cm. 1564 15621564
15581560 15561558
NCF's length 3.8cm NCF's length 4.0cm NCF's length 3.8cm NCF's length 4.2cm NCF's length nm/RIU 4.0cm Sensitivity:1524 NCF's length 4.2cm Sensitivity:1661.1 nm/RIU Sensitivity:1524 nm/RIU Sensitivity:1697.1 nm/RIU Sensitivity:1661.1 nm/RIU Sensitivity:1697.1 nm/RIU
Resonant Wavelength/nm
Resonant Wavelength/nm
15601562
15541556 15521554 15501552 15481550 15461548 15441546 1544
1.994
1.995
1.994 Refractive Index/RIU 1.995
1.996 1.996
Refractive Index/RIU Figure 6. Relationship between resonant wavelength and refractive indices with different lengths Figure 6. Relationship between resonant wavelength and refractive indices with different lengths of NCF. of NCF. Figure 6. Relationship between resonant wavelength and refractive indices with different lengths of NCF.
4. Experimental Test 4. Experimental Test 4. After Experimental Test the procedures we designed, we can get the interference spectra of the implementing After implementing the procedures we designed, we can get the interference spectra of the sensor for different hydrogenthe concentrations in experiment, as shown Figure 7. With the increase implementing procedures we we can getinthe interference of the sensorAfter for different hydrogen concentrations indesigned, experiment, as shown in Figure 7. Withspectra the increase in hydrogen concentration, the resonant wavelength shifts in short wavelength direction. This is sensor for different hydrogen in experiment, as short shownwavelength in Figure 7.direction. With the increase in hydrogen concentration, theconcentrations resonant wavelength shifts in This is because that the refractive index of Pd/WO 3 film will decrease when the hydrogen concentration is in hydrogen concentration, the of resonant shifts inwhen shortthe wavelength This isis because that the refractive index Pd/WOwavelength film will decrease hydrogendirection. concentration increasing. Therefore, the experimental result33isfilm consistent with the simulation result concentration as shown in is because that the refractive index of Pd/WO will decrease when the hydrogen increasing. Therefore, the experimental result is consistent with the simulation result as shown in Figure 4. Similarly, by quantifying the relationship thewith resonant wavelength and hydrogen increasing. Therefore, the experimental result isbetween consistent the simulation result ashydrogen shown in Figure 4. Similarly, by quantifying the relationship between the resonant wavelength and concentration, as shown in Figure 8, we can get the measurement sensitivity ofand hydrogen Figure 4. Similarly, by quantifying the relationship between the resonant wavelength hydrogen concentration, as shown in Figure 8, we can get the measurement sensitivity of hydrogen concentration concentration (1.26857shown nm/%) with good linearity of 0.99641. response timesensitivity of this kindofof hydrogen sensor concentration, in Figure 8, we can theThe measurement (1.26857 nm/%) as with good linearity ofthe 0.99641. Theget response time ofsensing. this kind of Pd/WO sensor primarily primarily depends on the sensitivity of material that is used for The 3 based concentration nm/%) withmaterial good linearity of 0.99641. The response time of this kind of sensor depends on the(1.26857 sensitivity of the the that is for sensing. The Pd/WO hydrogen 3 based hydrogen sensor will enter into steadyofstate in noused more than 30 min, this has been proven in3 our primarily depends on the sensitivity the material that is used for sensing. The Pd/WO based sensor will enter into theBesides, steady state in no more than 30 min, this has been proven in our previous previous manuscript [26]. the response time of hydrogen sensor based on Pd/WO 3 may be hydrogen sensor will enter the steady state in no more than 30 min, this has been proven in our manuscript [26]. Besides, theinto response of hydrogen sensor based on [28], Pd/WO bebeimproved 3 may improved bymanuscript UV-light irradiation [27]the ortime doping graphene quantum dots which tested previous [26]. Besides, response time of hydrogen sensor based on will Pd/WO 3 may be in our future work. improved by UV-light irradiation [27] or doping graphene quantum dots [28], which will be tested in our future work.
Sensors 2017, 17, 2144
7 of 10
Sensors 2017, 17, 2144
7 of 10
Sensors 2017, 17, 2144 by UV-light irradiation
[27] or doping graphene quantum dots [28], which will be
tested7 of in10our
future work. Volume Fraction of Hydrogen
-28 -28 -29 Transmission/dB Transmission/dB
-30 -31 -32 -33 -34
Volume Fraction0‰H of Hydrogen 2
-29
2‰H 0‰H 2 2
-30
4‰H 2‰H 2 2
-31
8‰H 6‰H 2 2
6‰H 4‰H 2 2 10‰H 8‰H 2 2
-32
10‰H2
-33 -34 -35
-35 1558
1559
1558
1559
1560
1561
1560 1561 Wavelength/nm
1562 1562
Wavelength/nm
Figure 7. Experimental spectra of sensor for different hydrogen concentrations. Figure 7. Experimental spectra of sensor for different hydrogen concentrations. Figure 7. Experimental spectra of sensor for different hydrogen concentrations. 1561.2 1561.2 1561.0
Experiment Point Fitting Curve Experiment Point Linearity(R-Square): 0.99641 Fitting Curve sensitivity: 1.26857 nm/1% Linearity(R-Square): 0.99641 sensitivity: 1.26857 nm/1%
Resonant Wavelength/nm
Resonant Wavelength/nm
1561.0 1560.8 1560.8 1560.6 1560.6 1560.4 1560.4 1560.2 1560.2 1560.0 1560.0 1559.8 1559.8
0.0 0.0
0.2
0.4
0.6
0.8
1.0
0.2
0.4Fraction0.6 0.8 1.0 Volume of Hydrogen/% Volume Fraction of Hydrogen/% Figure Relationship between between resonant resonant wavelength Figure 8. 8. Relationship wavelength and and hydrogen hydrogen concentration. concentration.
Figure 8. Relationship between resonant wavelength and hydrogen concentration.
To investigate investigatethethe stability ofproposed the proposed hydrogen sensor, we the observed resonant To stability of the hydrogen sensor, we observed resonantthe wavelength To investigate the every stability ofhydrogen the proposed hydrogen sensor, observed wavelength of sensor min when the concentration hydrogen concentration ofwe air chamber isthe stable 1%min. for of sensor every 5 min when5 the of air chamber is stable at 1% resonant forat30 wavelength of sensor every 5 min when the hydrogen concentration of air chamber is stable at 1% for 30 min. From9,Figure 9, see we that can see the maximum of the resonant wavelength 0.007 nm, From Figure we can thethat maximum change change of the resonant wavelength is 0.007isnm, which 30 will min. From Figure 9, we 0.0055% can see that the maximum change thecan resonant wavelength 0.007 nm,has which willonly cause only hydrogen volume error. of Therefore, we can say thatisthe proposed cause 0.0055% hydrogen volume error. Therefore, we say that the proposed sensor which will cause only 0.0055% volume error. Therefore, we little can has say that influence the sensor has goodBesides, stability. has been demonstrated that humidity little on this good stability. it Besides, has hydrogen beenitdemonstrated that humidity has influence on proposed this kind of sensor has good stability. Besides, it has been demonstrated that humidity has little influence on this of kind of[29], sensor [29],means whichthat means that the humidity compensation may not befor required most sensor which the humidity compensation may not be required most offor hydrogen kind of sensor [29], which means that the humidity compensation may beof required for of on hydrogen sensing applications. Besides, a Teflon can be used tonot decrease effect of H2O sensing applications. Besides, a Teflon layer can be layer used to decrease the effect Hthe themost hydrogen 2 O on hydrogen sensing applications. Besides, a Teflon layer can be used to decrease the effect of H 2 O on the hydrogen sensitive coating layer if the humidity variation is large [30]. sensitive coating layer if the humidity variation is large [30]. the hydrogen sensitive coating layer if the humidity variation is large [30].
Sensors2017, 2017, 17,2144 2144 Sensors Sensors 2017,17, 17, 2144
of10 10 888of of 10
1560.1 1560.1 Experiment Point Experiment Point the Average Line the Average Line Data Extreme Difference: 0.007nm Data Extreme Difference: 0.007nm
ResonantWavelength/nm Wavelength/nm Resonant
1560.0 1560.0
1559.9 1559.9
1559.8 1559.8
1559.7 1559.7
0 0
5 5
10 15 20 10 15 20 Time after Steady-State/min Time after Steady-State/min
25 25
30 30
Figure 9. Stability curve of the proposed hydrogen sensor. Figure Figure9.9.Stability Stabilitycurve curveof ofthe theproposed proposedhydrogen hydrogensensor. sensor.
Then we further investigate repeatability of For one we the Then investigate thethe repeatability of the sensor. For one we conducted the same Thenwe wefurther further investigate the repeatability of the the sensor. sensor. Forsensor, one sensor, sensor, we conducted conducted the same experimental tests three times. Then we got the sensitivity curves of sensor for each test, as experimental tests three times. Then we got the sensitivity curves of sensor for each test, as shown same experimental tests three times. Then we got the sensitivity curves of sensor for each test,in as shown in Figure 10. From this figure, we can confirm that the proposed sensor also has a good Figure 10. From this figure, we can confirm that the proposed sensor also has a good repeatability. We shown in Figure 10. From this figure, we can confirm that the proposed sensor also has a good repeatability. We can also see the lines parallel but the of lines can also see that the three are almost parallel but the intercepts these lines increase regularly. repeatability. We can alsolines seethat that thethree three linesare arealmost almost parallelof but theintercepts intercepts ofthese these lines increase regularly. Namely, the origins of the resonant wavelengths are not the same in each Namely, the origins of the resonant wavelengths are not the same in each measurement, which is mainly increase regularly. Namely, the origins of the resonant wavelengths are not the same in each measurement, due the isisdifferent This problem can due to that the which temperature is different in each test. This problem can in be resolved analyzing measurement, whichisismainly mainly dueto tothat that thetemperature temperature different ineach eachtest. test.by This problemthe can be resolved by analyzing the relationship between the change of the resonant wavelength and the relationship between the change of the resonant wavelength and the change of hydrogen concentration. be resolved by analyzing the relationship between the change of the resonant wavelength and the change concentration. Besides, temperature influence by Besides, this temperature influence can also bethis compensated by measuring the also hydrogen concentration changeof ofhydrogen hydrogen concentration. Besides, this temperature influencecan can alsobe becompensated compensated by measuring the hydrogen concentration and temperature, simultaneously [31]. and temperature, simultaneously [31]. measuring the hydrogen concentration and temperature, simultaneously [31]. First experiment First experiment Second experiment Second experiment Third experiment Third experiment Sensitivity: 1.23714 nm/1% Sensitivity: 1.23714 nm/1% Sensitivity: 1.19714 nm/1% Sensitivity: 1.19714 nm/1% Sensitivity: 1.26857 nm/1% Sensitivity: 1.26857 nm/1%
ResonantWavelength/nm Wavelength/nm Resonant
1561.0 1561.0
1560.5 1560.5
1560.0 1560.0
1559.5 1559.5
0.0 0.0
0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8 Volume Fraction of Hydrogen/% Volume Fraction of Hydrogen/%
1.0 1.0
Figure 10. Repeatability curve of the proposed hydrogen sensor. Figure Figure10. 10.Repeatability Repeatabilitycurve curveof ofthe theproposed proposedhydrogen hydrogensensor. sensor.
5. 5.5.Conclusions Conclusions Conclusions In summary, this paper proposed an in-line SNF interferometer structure to measure hydrogen In Insummary, summary,this thispaper paperproposed proposedan anin-line in-lineSNF SNFinterferometer interferometerstructure structureto tomeasure measurehydrogen hydrogen concentration. Then the hydrogen sensing properties of this structure had beenbyproven by concentration. Then the hydrogen sensing properties of this structure had been proven theoretical concentration. Then the hydrogen sensing properties of this structure had been proven by theoretical theoretical derivation, simulation calculation, and experimental test. Results demonstrated that derivation, derivation, simulation simulation calculation, calculation, and and experimental experimental test. test. Results Results demonstrated demonstrated that that the the proposed proposed sensor could achieve high-sensitivity hydrogen concentration detection (1.26857 nm/%). sensor could achieve high-sensitivity hydrogen concentration detection (1.26857 nm/%). Also, Also, the the
Sensors 2017, 17, 2144
9 of 10
the proposed sensor could achieve high-sensitivity hydrogen concentration detection (1.26857 nm/%). Also, the proposed sensor shows high precision and accuracy (maximum 0.0055% hydrogen volume error). Meanwhile, it has the advantages of simple structure, low cost, good linearity, stability, and repeatability. Acknowledgments: This work was supported in part by the National Science Foundation for Distinguished Young Scholars of China under grant 61425003, the National Natural Science Foundation of China under grants 61703080 and 61273059; the Fundamental Research Funds for the Central Universities under grant N160404012; the Liaoning Province Natural Science Foundation under grant 20170540314; the major scientific project of undergraduate under grant ZD1712; and State Key Laboratory of Synthetical Automation for Process Industries under grant 2013ZCX09. Author Contributions: The presented work is a result of the intellectual contribution of the whole team. Jin-xin Shao designed the sensor structure and wrote the manuscript; Wen-ge Xie built the simulation model and analyzed the simulation results; Xi Song performed the experiments and analyzed the experimental results; Ya-nan Zhang guided the experiments and gave some advice. All authors contributed and approved the final manuscript. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Jacobson, M.Z.; Colella, W.G.; Golden, D.M. Clearing the air and improving health with hydrogen fuel-cell vehicles. Science 2005, 308, 1901–1905. [CrossRef] [PubMed] Dedes, E.K.; Hudson, D.A.; Turnock, S.R. Assessing the potential of hybrid energy technology to reduce exhaust emissions from global shipping. Energy Policy 2012, 40, 204–218. [CrossRef] Dincer, I. Renewable energy and sustainable development: A crucial review. Renew. Sustain. Energy Rev. 2000, 4, 157–175. [CrossRef] Silva, S.F.; Coelho, L.; Frazão, O. A review of palladium-based fiber-optic sensors for molecular hydrogen detection. IEEE Sens. J. 2012, 12, 93–102. [CrossRef] Hübert, T.; Boon-Brett, L.; Black, G.; Banach, U. Hydrogen sensors—A review. Sens. Actuators B Chem. 2011, 157, 329–352. [CrossRef] Firth, J.G.; Jones, A.; Jones, T.A. The principles of the detection of flammable atmospheres by catalytic devices. Combust. Flame 1973, 20, 303–311. [CrossRef] Pollak-Diener, G.; Obermeier, E. Heat-conduction microsensor based on silicon technology for the analysis of two-and three-component gas mixtures. Sens. Actuators B Chem. 1993, 13, 345–347. [CrossRef] Korotcenkov, G.; Han, S.D.; Stetter, J.R. Review of electrochemical hydrogen sensors. Chem. Rev. 2009, 109, 1402–1433. [CrossRef] [PubMed] Dai, J.; Zhu, L.; Wang, G. Optical fiber grating hydrogen sensors: A review. Sensors 2017, 17, 577. [CrossRef] [PubMed] Astbury, G.R.; Hawksworth, S.J. Spontaneous ignition of hydrogen leaks: A review of postulated mechanisms. Int. J. Hydrogen Energy 2007, 32, 2178–2185. [CrossRef] Yang, M.; Dai, J. Fiber optic hydrogen sensors: A review. Photonic Sens. 2014, 4, 300–324. [CrossRef] Butler, M.A. Micromirror optical-fiber hydrogen sensor. Sens. Actuators B Chem. 1994, 22, 155–163. [CrossRef] Tabib, A.M.; Sutapun, B.; Petrick, R. Highly sensitive hydrogen sensors using palladium coated fiber optics with exposed cores and evanescent field interactions. Sens. Actuators B Chem. 1999, 56, 158–163. [CrossRef] Bevenot, X.; Trouillet, A.; Veillas, C. Surface plasmon resonance hydrogen sensor using an optical fibre. Meas. Sci. Technol. 2001, 13, 118. [CrossRef] Butler, M.A.; Ginley, D.S. Hydrogen sensing with palladium-coated optical fibers. J. Appl. Phys. 1988, 64, 3706–3712. [CrossRef] Yang, M.; Sun, Y.; Zhang, D. Using Pd/WO3 composite thin films as sensing materials for optical fiber hydrogen sensors. Sens. Actuators B Chem. 2010, 143, 750–753. [CrossRef] Dai, J.; Yang, M.; Yu, X. Greatly etched fiber Bragg grating hydrogen sensor with Pd/Ni composite film as sensing material. Sens. Actuators B Chem. 2012, 174, 253–257. [CrossRef] Cai, L.; Zhao, Y.; Li, X.G. Applications of modal interferences in optical fiber sensors based on mismatch methods. Instrum. Sci. Technol. 2015, 43, 1–20. [CrossRef]
Sensors 2017, 17, 2144
19. 20. 21. 22. 23. 24. 25. 26.
27. 28. 29.
30. 31.
10 of 10
Raghunandhan, R.; Chen, L.H.; Long, H.Y. Chitosan/PAA based fiber-optic interferometric sensor for heavy metal ions detection. Sens. Actuators B Chem. 2016, 233, 31–38. [CrossRef] Mohammed, W.S.; Mehta, A.; Johnson, E.G. Wavelength tunable fiber lens based on multimode interference. J. Lightwave Technol. 2004, 22, 469. [CrossRef] Zhao, Y.; Cai, L.; Li, X.G.; Meng, F.C.; Zhao, Z. Investigation of the high sensitivity RI sensor based on SMS fiber structure. Sens. Actuators A Phys. 2014, 205, 186–190. [CrossRef] Zhao, Y.; Li, X.G.; Cai, L. A reflective intensity modulated fiber tilt angle sensor based on an all-photonic crystal fiber interferometer. Sens. Actuators A Phys. 2016, 244, 106–111. [CrossRef] Zhang, Y.N.; Peng, H.; Zhou, T.; Zhang, L.; Zhang, Y.; Zhao, Y. Hydrogen sensor based on high-birefringence fiber loop mirror with sol-gel Pd/WO3 coating. Sens. Actuators B Chem. 2017, 248, 71–76. [CrossRef] Fardindoost, S.; Rahimi, F.; Ghasempour, R. Pd doped WO3 films prepared by sol-gel process for hydrogen sensing. Int. J. Hydrogen Energy 2010, 35, 854–860. [CrossRef] Liu, X.; Zhang, X.; Liu, Y. Multi-point fiber-optic refractive index sensor by using coreless fibers. Opt. Commun. 2016, 365, 168–172. [CrossRef] Zhang, Y.; Wu, Q.; Peng, H. Photonic crystal fiber modal interferometer with Pd/WO3 coating for real-time monitoring of dissolved hydrogen concentration in transformer oil. Rev. Sci. Instrum. 2016, 87, 125002. [CrossRef] [PubMed] Zhong, X.X.; Yang, M.H.; Huang, C.J.; Wang, G.P.; Dai, J.X.; Bai, W. Water photolysis effect on the long-term stability of fiber optic hydrogen sensor with Pt/WO3 . Sci. Rep. 2016, 6, 39160. [CrossRef] [PubMed] Fardindoost, S.; Zad, A.I.; Hosseini, Z.S.; Hatamie, S. Detecting hydrogen using graphene quantum dots/WO3 thin films. Mater. Res. Express 2016, 3, 116407. [CrossRef] Dai, J.; Yang, M.; Yang, Z.; Li, Z.; Wang, Y.; Wang, G.; Zhang, Y.; Zhuang, Z. Performance of fiber Bragg grating hydrogen sensor coated with Pt-loaded WO3 coating. Sens. Actuators B Chem. 2014, 190, 657–663. [CrossRef] Coelho, L.; de Almeida, J.; Santos, J.L. Fiber optic hydrogen sensor based on an etched Bragg grating coated with palladium. Appl. Opt. 2015, 54, 10342–10348. [CrossRef] [PubMed] Wang, M.; Wang, D.N.; Yang, M.; Cheng, J.; Li, J. In-line Mach-Zehnder Interferometer and FBG with Pd film for simultaneous hydrogen and temperature detection. Sens. Actuators B Chem. 2014, 202, 893–896. [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/).
