Liquefied Petroleum Gas Monitoring System Based on ... - MDPI

sensors Article

Liquefied Petroleum Gas Monitoring System Based on Polystyrene Coated Long Period Grating Flavio Esposito 1 ID , Aldobenedetto Zotti 2 , Giovanna Palumbo 1 , Simona Zuppolini 2 , Marco Consales 3,4 , Antonello Cutolo 3 , Anna Borriello 2 , Stefania Campopiano 1 ID , Mauro Zarrelli 2, * and Agostino Iadicicco 1, * 1

2 3 4

*

Department of Engineering, University of Naples “Parthenope”, 80143 Napoli, Italy; [email protected] (F.E.); [email protected] (G.P.); [email protected] (S.C.) Institute for Polymers, Composites and Biomaterials (IPCB)-CNR, 80055 Portici, Italy; [email protected] (A.Z.); [email protected] (S.Z.); [email protected] (A.B.) Optoelectronics group, Department of Engineering, University of Sannio, 82100 Benevento, Italy; [email protected] (M.C.); [email protected] (A.C.) Centro Regionale Information Communication Technology-CeRICT scrl, 82100 Benevento, Italy Correspondence: [email protected] (M.Z.); [email protected] (A.I.)

Received: 11 April 2018; Accepted: 3 May 2018; Published: 5 May 2018







Abstract: In this work, we report the in-field demonstration of a liquefied petroleum gas monitoring system based on optical fiber technology. Long-period grating coated with a thin layer of atactic polystyrene (aPS) was employed as a gas sensor, and an array comprising two different fiber Bragg gratings was set for the monitoring of environmental conditions such as temperature and humidity. A custom package was developed for the sensors, ensuring their suitable installation and operation in harsh conditions. The developed system was installed in a real railway location scenario (i.e., a southern Italian operative railway tunnel), and tests were performed to validate the system performances in operational mode. Daytime normal working operations of the railway line and controlled gas expositions, at very low concentrations, were the searched realistic conditions for an out-of-lab validation of the developed system. Encouraging results were obtained with a precise indication of the gas concentration and external conditioning of the sensor. Keywords: chemical sensors; gas detectors; long period gratings; optical fiber sensors; polymers

1. Introduction Hydrocarbons are a very dangerous species, even when considered at low concentrations, due to high flammable and volatile behavior. As their main application is for fuels, their demand is always increasing, which in turn increases their transport frequency on both roads and rails. For the safekeeping of both the environment and public health, a real-time monitoring of leaks during their transportation would be an appealing solution. Typical gas sensors are built on planar architectures and their working principle is based on the changes in electrical properties of specific materials (the most common are metal oxides), when they are exposed to the analyte gas species. One of the most important parameters is the lower explosive limit (LEL) and, as a consequence, the sensors are designed to detect concentrations below this value, which is different for each target. Unfortunately, commercial sensors generally require high operating temperatures and may not be suitable for use in hostile, harsh and remote environments. In addition, they should be able to operate safely even in an explosive atmosphere [1,2]. Currently, fiber optic technology represents a successful alternative, employed for the detection of very small amounts of chemical and gaseous species. In addition, optical fiber sensors have the Sensors 2018, 18, 1435; doi:10.3390/s18051435

www.mdpi.com/journal/sensors

Sensors 2018, 18, 1435

2 of 14

advantages of small size, immunity to electromagnetic interferences (EMI), and the possibility of both easy multiplexing and long distance monitoring [3,4]. Nowadays, different configurations and sensitive materials have been proposed for their realization. Among these technologies, we recall refractometric and interferometric configurations [5–7], ring resonators [8], etched/tilted fiber Bragg gratings (FBGs) [9,10], long-period gratings (LPGs) [11], photonic crystal fibers [12], surface plasmon resonance (SPR) [13,14], and lossy mode resonance (LMR) sensors [14]. Based on the number of studies carried out over the last decades, we believe that LPG technology is one of the most mature in this field. Special designs such as mode transition (MT) and dispersion turning point (DTP) have been investigated to increase their sensitivity as compared to bare gratings [15–20]. As a result, these devices have been used for the detection of several solvents, as well as aromatic and aliphatic hydrocarbons. For example, in [21], a LPG operating at DTP with alternate thin films of PAA/PAH was used to detect ammonia. The syndiotactic polystyrene (sPS) was proposed as a high refractive index layer (HRI) for the LPG in order to detect chloroform in water [22]. Metal organic framework thin films were used in [23] to detect methanol, ethanol, isopropanol, and acetone. Different LPGs were used to develop a system for indoor air quality monitoring in [24]. The selective detection of toluene was demonstrated in [25] by using an LPG coated with a calixarene. Whereas, in [26], an LPG was interrogated by ring-down spectroscopy to detect m-xylene, cyclohexane, trichloroethylene and commercial gasoline. Methane detection was achieved in [27] by using a LPG coated with polycarbonate/cryptophane-A HRI overlay, and in [28] through a coated LPG in photonic crystal fiber. Finally, such devices were also used for hydrocarbon detection in fuel, water and atmospheric environment [29,30]. In our previous work [31], we have presented a butane gas sensor based on a single-ended LPG coated with atactic polystyrene (aPS). The aPS was chosen in light of the affinity between its olefin chains and hydrocarbon molecules. Moreover, its thickness was designed in order to work in mode transition when air is the surrounding medium, in order to attain a significant sensitivity of the final device. Finally, the sensor was tested towards butane in a very low concentrations range, 0.1–1.0 vol %, by means of an in-lab procedure involving a gas test chamber. In this work, expanding on the results of previous characterizations, we report on the demonstration in a real scenario of a monitoring system for liquefied petroleum gas (LP gas), within the Italian funded project OPTOFER “Innovative optoelectronic technologies for the monitoring and diagnostics of the railway infrastructure”. LP gas is one of the dangerous species being transported with higher frequency (first in Italy) by railways lines. Specifically, the system was installed in a railway tunnel ~850 m long, located between Porta Rufina and Arco Traiano Station in Benevento (Italy), whereas the control and interrogation units were hosted in Porta Rufina Station, 300 m away from the tunnel entrance. The location is part of the railway line Avellino–Benevento run by Italian railways, with passenger trains powered by diesel motors. The monitoring system is based on the aPS coated LPG butane (the main component of LP gas) sensor, which is combined with two fiber Bragg grating sensors for the measurement and compensation of environmental temperature and humidity changes. The sensors are housed in a custom package, which was developed in order to ensure the operation in harsh conditions. Finally, the system was tested in different operating states: (i) with the railway line under normal operation; and (ii) under controlled exposition to the target gas. In Section 2, we describe the development of the gas monitoring system, focusing on the different components. In Section 3, we report on the system installation and show the experimental results obtained during the in-field tests performed at Porta Rufina Station. 2. Development of the Gas Monitoring System The gas monitoring system herein proposed is based on the architecture depicted in Figure 1, where two main components can be distinguished: the sensing unit (SU) and the control unit (CU). The SU is located in the railway tunnel and is composed of an aluminum package, hosting the following fiber sensors: (i) a LPG-based gas sensor and (ii) a FBG-based temperature and humidity sensor. The

Sensors 2018, 18, 1435

3 of 14

2018, 18,to x FOR PEER located REVIEW in the technical room of the railway station, by two 500 m long 3 of 14single SU isSensors connected the CU, 18,CU x FOR PEER REVIEW 3 of 14 mode Sensors fibers.2018, The comprises two main hardware systems, the HBM FS22 optical interrogator and a 500 m long single mode fibers. The CU comprises two main hardware systems, the HBM FS22 optical personal computer, respectively. The interrogator is controlled through a custom LabView software, 500 m long single fibers. The CUrespectively. comprises two main hardwareissystems, thethrough HBM FS22 optical interrogator and a mode personal computer, The interrogator controlled a custom whichinterrogator manages all the controls computer, and operations, starting from the acquisition of sensors spectra up to and a personal respectively. The interrogator is controlled through a custom LabView software, which manages all the controls and operations, starting from the acquisition of the real-time estimation of gas concentration. LabView software, manages all the controls and operations, starting from the acquisition of sensors spectra up towhich the real-time estimation of gas concentration. In the following, a detailed description concerning development of component each component sensors spectra up to the real-time estimation of gas concentration. In the following, a detailed description concerning thethe development of each is is reported. In the following, a detailed description concerning the development of each component is reported.

reported. Gas Monitoring System Gas Monitoring System

Control Unit (Technical room) Control Unit (Technical room)

Connecting optical fibers Connecting optical fibers PC PC

Gas Sensing Unit (Tunnel) Gas Sensing Unit (Tunnel) Package with sensors Package with sensors

FS22 Optical Interrogation FS22 Optical System Interrogation System

Figure 1. Schematicfigure figureand and photos photos of system. Figure 1. Schematic ofthe thegas gasmonitoring monitoring system. Figure 1. Schematic figure and photos of the gas monitoring system.

2.1. Gas Sensor Based on aPS Coated LPG 2.1. Gas Sensor Based on aPS Coated LPG 2.1. Gas Sensor Based on aPS Coated LPG The core of the SU consists of the single-ended LPG, coated with a thin layer of aPS as a sensitive The core the of the single-ended LPG, withaathin thin layer of aPS a sensitive Theasof core ofSU the consists SU consists of in the single-ended LPG, coated coated with layer aPS aas sensitive overlay, schematically depicted Figure 2a. The design and characterization ofofsuch aassensor has overlay, aspresented schematically depicted Figure 2a. The andcharacterization characterization of such aonsensor overlay, as schematically depicted in Figure 2a.paper Thedesign design and such a sensor has has been and discussed inin our previous [31]. Briefly, we focused theofattention LPG been technology, presented and and discussed in in our previous paper [31]. Briefly,we wefocused focused attention on LPG been presented discussed our previous paper [31].chemical, Briefly, thethe attention on [32– LPG since it has been employed in several physical, and biological applications technology, has been employed in several physical, chemical,and and biological applications [32– technology, since since it has employed in several biological [32–38]. 38]. It consists of aitbeen periodic perturbation, withphysical, period ofchemical, hundreds of μm, in the applications refractive index 38]. It consists of a periodic perturbation, with period of hundreds of μm, in the refractive index and/orof the of the optical fiber [39–41]. It consists a geometry periodic perturbation, with period of hundreds of µm, in the refractive index and/or and/or the of the optical fiber [39–41]. the geometry ofgeometry the optical fiber [39–41].

core cladding core cladding

(a) (a)

LP07 LP07

(b) (b)

Figure 2. Single ended long-period grating (LPG)-based gas sensor: (a) schematic figure; (b) power Figure 2.during Single ended long-period grating(LPG)-based (LPG)-based gas gas sensor: figure; (b) (b) power Figure 2. Single ended long-period grating sensor:(a) (a)schematic schematic figure; power spectrum the fabrication steps. spectrum during the fabrication steps. spectrum during the fabrication steps. For the purpose of this work, a commercially available LPG written by UV technique in a For SMF28 the purpose ofcorethis work, written by UV technique in a standard fiber (d = 8.2 μm,adcommercially clad = 125.0 μm,available and NA LPG = 0.12) was selected. The LPG was For the purpose of this work, a commercially available LPG written by UV technique in a standard standard with SMF28 fiber (d 8.2 μm μm,which dclad =presents 125.0 μm, and NA =bands 0.12) related was selected. The LPGwith was fabricated a period ofcore Λ = 360 attenuation to the coupling SMF28 fiber (dhigh µm, = 125.0 µm, NA = 0.12) was selected. The LPG was fabricated core clad fabricated with= a8.2 period ofdΛ = 360 μm in which presents attenuation bands related to with relatively order cladding modes the and NIR range. In this spectral range, thethe LPcoupling 07 band was with chosen arelatively period ofhigh Λthe =order 360 whichmodes presents attenuation bands related to range, the refractive coupling with cladding in the NIRrefractive range. Inindex this spectral the LP 07index. bandrelatively was due to highµm sensitivity to surrounding and overlay The high experimental order modes in spectrum the NIR In thisrefractive spectral range, the LP07 nm band was index. chosen due to chosencladding due totransmission the high sensitivity torange. surrounding index and refractive of the grating around the band atoverlay 1606 related to LP 07The is experimental spectrum of the grating around the band at 1606index. nm related to LP 07 is the high sensitivitytransmission to surrounding refractive index and overlay refractive The experimental

transmission spectrum of the grating around the band at 1606 nm related to LP07 is reported in

Sensors 2018, 18, 1435

4 of 14

2018, 18, x FOR PEER REVIEW a single-ended (or reflection) grating, since this approach 4 ofallows 14 FigureSensors 2b, black line. We developed simpler and strain-free packaging and at same time makes the integration with sensitive layers easier. reported in Figure 2b, black line. We developed a single-ended (or reflection) grating, since this In particular, for this aim, the fiber was cut along the grating and a silver mirror was deposited on the approach allows simpler and strain-free packaging and at same time makes the integration with fiber tip by using the Tollens’ test, according to the procedure reported in [31]. The resulting reflected sensitive layers easier. In particular, for this aim, the fiber was cut along the grating and a silver spectrum of the is reported in Figure with line.test, according to the procedure reported mirror wasLPG deposited on the fiber tip by2b using theblue Tollens’ For the The sensitive material, used the atactic polystyrene (aPS) due to blue the chemical affinity in [31]. resulting reflectedwe spectrum of the LPG is reported in Figure 2b with line. between itsFor olefin chains and hydrocarbon molecules [31]. Moreover, the aPS has a refractive index the sensitive material, we used the atactic polystyrene (aPS) due to the chemical affinity of between its olefin chains and hydrocarbon molecules Moreover,inthe aPSto has a refractive 1.55, which is higher than silica, and thus its thickness can[31]. be designed order tune the LPGindex working of mode 1.55, which is higher than The silica, and thus its thickness can be designedthickness in order to tune the LPG by point in transition region. design of the aPS overlay optimum was performed point intool mode transitioninregion. The design of the[42]. aPS The overlay optimum thickness was by using working our numerical developed Matlab environment model was particularized performed by using our numerical tool developed in Matlab environment [42]. The was using the parameters reported in [31], finding that aPS thickness has to be selected in the model range 320–380 particularized by using the parameters reported in [31], finding that aPS thickness has to be selected nm, resulting in a wavelength blue shift of the LP07 band in range 40–75 nm after the deposition. To in the range 320–380 nm, resulting in a wavelength blue shift of the LP07 band in range 40–75 nm after this aim, the film was deposited by using the dip-coating technique, starting from a 9.5% w/w aPS the deposition. To this aim, the film was deposited by using the dip-coating technique, starting from solution in chloroform at 60 ◦inCchloroform (aPS MW at = 280.000 g/mol, Sigma-Aldrich, Milan, Italy).Milan, The aPS a 9.5% w/w aPS solution 60 °C (aPS MW = from 280.000 g/mol, from Sigma-Aldrich, coatedItaly). LPG The wasaPS subjected to a blue shift of 53 nm for LP band, in good agreement with design coated LPG was subjected to a blue shift of0753 nm for LP07 band, in good agreement criteria, as design showncriteria, in Figure 2b with the red2bline. it addition, should be remarked that ourthat design with as shown in Figure withIn theaddition, red line. In it should be remarked ourindesign resulted inwavelength the resonance of thebeing final device being located in the of system the resulted the resonance of wavelength the final device located in the middle of middle the FS22 FS22 systemnm). rangeFinally, (1500–1600 nm). Finally, it is worth that the coated LPG exhibits a range (1500–1600 it is worth highlighting thathighlighting the coated LPG exhibits a very low decrease very low decrease in the band depth, achieved through the design of the single-ended LPG in overin the band depth, achieved through the design of the single-ended LPG in over-coupling state. coupling The sensorstate. response towards butane was characterized by using a custom-designed test chamber. The sensor response towards butane was characterized by using a custom-designed test The set-up is constituted by two lines: the first one for the butane and the second one for an inert chamber. The set-up is constituted by two lines: the first one for the butane and the second one for an gas (nitrogen), and the gas flow is tuned using mass flow controllers. The complete schematic of the inert gas (nitrogen), and the gas flow is tuned using mass flow controllers. The complete schematic experimental setup for gas testing cantesting be found in found [31]. The sensor thus to different of the experimental setup for gas can be in [31]. The was sensor wasexposed thus exposed to concentrations of butane, in the range 0.1–1.0 vol %.0.1–1.0 As anvol example, we report in variation different concentrations of butane, in the range %. As an example, weFigure report3a in the Figure 3a in the the spectral position of spectral the LP07position band when theLPLPG waswhen exposed butane concentrations of 0.5, variation in the of the 07 band the to LPG was exposed to butane concentrations 0.5,sensor 0.75, and 1.0 vol %. The sensoran practically exhibits an immediate reaction toand gas the 0.75, and 1.0 vol %. ofThe practically exhibits immediate reaction to gas presence, presence, and the following times were T measured: rising T90falling ~2–5 min and falling 10~15curve min. in following response times were response measured: rising and T10 ~15 min.TThe 90 ~2–5 min The curve in Figure 3b reports the gas concentration as a function of the negative wavelength shift Figure 3b reports the gas concentration as a function of the negative wavelength shift within the range within the range 0–1.0 vol %. As one can observe, the data fitting can be obtained with a 3rd degree 0–1.0 vol %. As one can observe, the data fitting can be obtained with a 3rd degree polynomial. Around polynomial. Around the concentration of 0.1 vol %, the slope is 0.5 vol %/nm, while it increases to 2.1 the concentration of 0.1 vol %, the slope is 0.5 vol %/nm, while it increases to 2.1 vol %/nm around vol %/nm around 1.0 vol % of butane. Finally, it should be highlighted that butane LEL is equal to 1.8 1.0 volvol % of butane. Finally, it should be highlighted that butane LEL is equal to 1.8 vol %, and thus we %, and thus we focused the attention on concentrations lower than about half the LEL. focused the attention on concentrations lower than about half the LEL.

2.1 vol%/nm

0.5 vol%/nm

0.5 vol% 0.75 vol%

1 vol%

(a)

(b)

FigureFigure 3. (a) Wavelength shiftshift of the LPLP band exposuretotodifferent different butane concentrations; 3. (a) Wavelength of the bandduring during the the exposure butane concentrations; 07 07 gas concentration versus (negative) wavelengthshift. shift. (b) gas(b) concentration versus (negative) wavelength 2.2. Gas Sensor Sensitivities to EnvironmentalConditions Conditions 2.2. Gas Sensor CrossCross Sensitivities to Environmental In our previous experimentation [31], tests were performed on an aPS coated LPG sensor under

In our previous experimentation [31], tests were performed on an aPS coated LPG sensor under controlled temperature and relative humidity (RH) conditions. In a real scenario, it is necessary to controlled temperature and relative humidity (RH) conditions. In a real scenario, it is necessary to deal with the variations in environmental conditions, especially if operation in harsh environments is

Sensors 2018, 18, 1435

5 of 14

Sensors 2018, 18, x FOR PEER REVIEW

5 of 14

deal with in environmental conditions, especially operation in harsh environments envisaged. For the thisvariations reason, the sensor was characterized towardsiftemperature and humidity changes, is envisaged. For this reason, the sensor was characterized towards temperature and ◦ ◦ considering values being compatible with the selected scenario: 0 C < T < 15 C and 75%humidity < RH < 95%. changes, considering values being compatible with the selected scenario: 0 °C < T < 15 °C and 75% < Concerning the temperature characterization, we used the setup reported in [43,44], which was RH < 95%. modified for the housing of single-ended sensors and where a Peltier cell-based system was employed Concerning the temperature characterization, we used the setup reported in [43,44], which was ◦ C. The wavelength shifts of the LP band versus to change the temperature in of thesingle-ended range 0–20 sensors modified for the housing and where a Peltier cell-based 07 system was temperature are reported in Figure 4a, where a linear behavior can be appreciated with employed to change the temperature in the range 0–20 °C. The wavelength shifts of the LP07sensitivity band ◦ C. ST = 0.23 nm/ versus temperature are reported in Figure 4a, where a linear behavior can be appreciated with For the humidity we used the setup depicted in Figure 4b. The setup consists of sensitivity ST = 0.23characterization, nm/°C. characterization, the (~20% setup depicted in Figure 4b. The setup consists to two lines: For the the firsthumidity one is connected to a drywe airused source RH), while the second one is connected of two lines: thelines first are one connected is connectedto toflow a drycontrollers air source (~20% RH), while the secondto one is connected a humidifier. Both which can be regulated change humidity to a humidifier. Both lines are connected to flow controllers which can be regulated to change from 20% to 95% in the sensor chamber (as measured from reference sensor). The results, in term of humidity from 20% to 95% in the sensor chamber (as measured from reference sensor). The results, the wavelength shift of the LP07 band versus humidity level, are reported in Figure 4c. The response is in term of the wavelength shift of the LP07 band versus humidity level, are reported in Figure 4c. The negligible below 70% RH, while it exhibits a quadratic-like behavior for higher humidity conditions, as response is negligible below 70% RH, while it exhibits a quadratic-like behavior for higher humidity from the blue fitting curve. For example, within the range 80–90%, the sensitivity SRH increases conditions, as from the blue fitting curve. For example, within the range 80–90%, the sensitivity SRHfrom 0.18 toincreases 0.34 nm/%RH. from 0.18 to 0.34 nm/%RH. PC

Optical Interrogation System Flow controller

Ref. sensors

LPG Sensor

Dry Air Source

Humidifier

(b)

0.34 nm/%RH

(a) 0.25 nm/%RH

0.18 nm/%RH

(c)

Figure 4. (a) Wavelength shift of the LPG sensor versus temperature; (b) schematic setup for humidity

Figure 4. (a) Wavelength shift of the LPG sensor versus temperature; (b) schematic setup for humidity characterization; (c) wavelength shift of the LPG sensor versus humidity. characterization; (c) wavelength shift of the LPG sensor versus humidity. From these results, it is clear that during the in-field testing, a compensation of temperature and humidity changes is it necessary. To this end, the FBG technology was considered in order and From these results, is clear that during thewell-settled in-field testing, a compensation of temperature to monitor real time variations of temperature and humidity level at the installation location. In fact, humidity changes is necessary. To this end, the well-settled FBG technology was considered in order such devices are normally sensitive to temperature changes, whereas they can be used for humidity to monitor real time variations of temperature and humidity level at the installation location. In fact, sensing when coated with hygroscopic materials by exploiting the strain induced by the latter when such devices are normally sensitive to temperature changes, whereas they can be used for humidity subjected to humidity changes [45]. Moreover, their interrogation is also possible by using the same sensing when coated with hygroscopic materials by exploiting the strain induced by the latter when instrumentation used for LPG. subjected In to particular, humidity we changes [45]. Moreover, their is also possible by using the designed a single-ended arrayinterrogation composed of two 0.5 cm long FBGs (spaced 1.0same instrumentation used for LPG. cm), namely FBG1 and FBG2, written in SMF28 fiber, and schematically reported in Figure 5. The first In particular, we designed a single-ended array composed two cm long FBGs (spaced 1.0 one is coated with a polyimide (PI) layer with thickness aroundof20 μm0.5 and resonance wavelength λ B1 cm), = 1572.067 nm.FBG2, The second bare and characterized by λ B2 =reported 1579.95 nm. The resonance namely FBG1 and writtenone in is SMF28 fiber, and schematically in Figure 5. The first array were in order to avoid spectral overlapping even during the one iswavelengths coated withofa the polyimide (PI)designed layer with thickness around 20 µm and resonance wavelength λB1 = 1572.067 nm. The second one is bare and characterized by λB2 = 1579.95 nm. The resonance wavelengths of the array were designed in order to avoid spectral overlapping even during the sensing phase; in addition, they can be potentially used in multiplexing with the LPG gas sensor. The spectrum

Sensors2018, 2018,18, 18,1435 x FOR PEER REVIEW Sensors

6 6ofof1414

sensing phase; in addition, they can be potentially used in multiplexing with the LPG gas sensor. The ofspectrum the arrayofisthe reported Figure in 5 as well.5Finally, FBGs were characterized, resulting in the array isin reported Figure as well.both Finally, both FBGs were characterized, resulting ◦ C—and humidity—S following sensitivities to temperature—S = S = 10.1 pm/ = 2.8 pm/%RH T1 T2 RH1 in the following sensitivities to temperature—S T1 = ST2 = 10.1 pm/°C—and humidity—S RH1 = 2.8 (FBG2 shows no humidity pm/%RH (FBG2 shows nodependency). humidity dependency).

Fiber jacket 900 μm

FBG1

FBG2

Figure 5. Fiber Bragg gratings (FBG) array for temperature and humidity compensation: schematic Figure 5. Fiber Bragg gratings (FBG) array for temperature and humidity compensation: schematic and power spectrum. and power spectrum.

2.3. Sensor Packaging 2.3. Sensor Packaging In order to employ the developed fiber optic sensing system in a real railway location scenario, In order to employ the developed fiber optic sensing system in a real railway location scenario, a custom aluminum package was properly designed and realized to host the two fibers containing a custom aluminum package was properly designed and realized to host the two fibers containing both the LPG- and the FBG-based sensors. The package was designed to address the following both the LPG- and the FBG-based sensors. The package was designed to address the following requirements: requirements: To enable its fixing on the inner wall of a railway tunnel; -To fixing on (prepared the inner wall a railway tunnel;allowing air exchange with the external Toenable protectitsthe fibers as inofFigure 6a) while To protect the fibers (prepared as in Figure 6a) while allowing air exchange with the environment; external To keepenvironment; both fibers in a strain-free state and to avoid their bending, as a consequence of the air To keep induced both fibers strain-free state and to avoid their bending, as a consequence of the air motion by in thea train transit; motion induced by the train transit; To avoid water infiltration inside the package. To avoid water infiltration inside the package. Figure 6b reports the CAD image of the designed package. It is mainly composed of three independent “sensing area”, containing sensitiveItarea of the fiber devices; the Figure 6bregions: reports (i) thethe CAD image of the designedthe package. is mainly composed of (ii) three “housing unit”, where the two optical fibers are positioned and kept fixed by means of properly independent regions: (i) the “sensing area”, containing the sensitive area of the fiber devices; (ii) designed magnetic clamps; thefibers “connection unit”, needed connect the optical fibers the “housing unit”, where theand two (iii) optical are positioned and kepttofixed by means of properly containing the sensors with the patch-cords travelling tunnelthe and reaching the designed magnetic clamps; andouter (iii) the “connection unit”,throughout needed to the connect optical fibers technical room where with the interrogation system is travelling located. throughout the tunnel and reaching the containing the sensors the outer patch-cords technical room where the interrogation system is located. Outer patch-cords To protect the fiber optic sensors as much as possible during train transit inside the tunnel, (a) the “sensing area” was designed so as to be covered with two independent perforated covers, put one on the other, with holes having a diameter of 2.5 mm. The holes of the two covers were designed in Connector such a way to be offset one from the other (i.e., to create a “labyrinth” type path) in order to avoid the air movement during train transit being able to enter directly into contact with the sensors. In this way, indeed, the air is forced to follow the labyrinth path, thus slowing down after the impact with the Connection Unit Sensitive undrilled part of the cover before coming into contact with the sensitive element area of the sensors. Housing Unit Magnetic clamps Sensitive elements

Sensing Area (b)

Perforated covers

Figure 6b reports the CAD image of the designed package. It is mainly composed of three independent regions: (i) the “sensing area”, containing the sensitive area of the fiber devices; (ii) the “housing unit”, where the two optical fibers are positioned and kept fixed by means of properly designed magnetic clamps; and (iii) the “connection unit”, needed to connect the optical fibers Sensorscontaining 2018, 18, 1435 the sensors with the outer patch-cords travelling throughout the tunnel and reaching the7 of 14 technical room where the interrogation system is located. Outer patch-cords

(a)

Connector

Connection Unit

Sensitive element

Sensors 2018, 18, x FOR PEER REVIEW

7 of 14

Housing Unit Magnetic clamps

Figure 6. (a) Picture of the fiber optic sensor prepared to be installed inside the package; Sensitive elements (b) CAD image of the designed package. Sensing Area

Perforated covers

(b) To protect the fiber optic sensors as much as possible during train transit inside the tunnel, the “sensing area” was designed so as to be covered with two independent perforated covers, put one on the other, holes having diameter of 2.5 mm. The of the two covers were designed in Figure 6. (a) with Picture of the fibera optic sensor prepared to holes be installed inside the package; (b) CAD such a way to be offset one from the other (i.e., to create a “labyrinth” type path) in order to avoid image of the designed package. the air movement during train transit being able to enter directly into contact with the sensors. In this way, indeed, the air is forced to follow the labyrinth path, thus slowing down after the impact with Some photos of of the package areinto reported in Figure 7a,b. The package was installed the undrilled part therealized cover before coming contact with the sensitive area of the sensors. on the inner of the railway atare 30 reported cm frominthe floor in aThe maintenance located at Somewall photos of the realizedtunnel, package Figure 7a,b. package wascavity installed on 50 mthe farinner fromwall theofentrance (Figure 7c,d). a position chosen considering that,atin50case the railway tunnel, at 30 Such cm from the floorwas in a maintenance cavity located m of a far from the entrance (Figure 7c,d). Such a creeks positionand wasdeposits chosen considering that,due in case of higher a leakage, leakage, liquefied petroleum gas occupies on the floor, to its density 2 liquefied petroleum gas occupies creeks and deposits on the floor, due to its higher density compared compared to air. Figure 7e,f show a schematic representation of the “sensing area” (~4 cm ), with the 2), with the two singleto air. Figure 7e,f showhosting a schematic representation the “sensing area” (~4 cmarray two single-ended fibers the LPG-based gasofsensor and the 2-FBGs for temperature and ended fibers hosting the LPG-based gas sensor and the 2-FBGs array for temperature humidity, humidity, respectively. The sensor package is connected with two approx. 500 m longand single mode fiber respectively. The sensor package is connected with two approx. 500 m long single mode fiber patchpatch-cords reaching the technical room with the interrogation unit. Finally, to avoid water infiltration cords reaching the technical room with the interrogation unit. Finally, to avoid water infiltration inside the package, a Plexiglas canopy was also realized and mounted on the aluminum package. inside the package, a Plexiglas canopy was also realized and mounted on the aluminum package.

(a)

(b)

(c)

(d)

aPS coated LPG

(e)

PI coated FBG Bare FBG

(f)

Figure 7. Photos of the realized package (c,d)its itsinstallation installationonon inner wall of Figure 7. Photos of the realized packagebefore before(a,b) (a,b)and and after after (c,d) thethe inner wall the railway tunnel; (e,f) schematic (not(not to scale) of of sensing of the railway tunnel; (e,f) schematic to scale) sensingarea. area. 2.4. Measuring Hardware and Software The CU is located in the technical room of Porta Rufina Station. It involves the interrogation system for the acquisition of sensor spectra, controlled through a PC with custom-developed software, and aimed at acquisition as well as real-time data analysis.

Sensors 2018, 18, 1435

8 of 14

2.4. Measuring Hardware and Software The CU is located in the technical room of Porta Rufina Station. It involves the interrogation system for the acquisition of sensor spectra, controlled through a PC with custom-developed software, and aimed at acquisition as well as real-time data analysis. As anticipated, for the acquisition system, we selected the commercial HBM FS22 8-channel optical interrogator operating in the wavelength range 1500–1600 nm, having a maximum resolution of 1 pm and maximum acquisition rate of 1 S/s. To meet the requirements of the OPTOFER project, a custom LabView software for the real-time control of the interrogator unit as well as data analysis was developed by Migma srl. The software permits the acquisition of sensor spectra connected to the different optical channels and is able to track the resonance wavelength of both FBGs and LPGs. In addition, a programming interface permits sensor data analysis to compensate the effects of humidity and temperature changes and thus to estimate the gas concentration. Moreover, the software can notify an alarm to a remote server if the concentration exceeds a user-defined threshold. At any rate, these aspects lie beyond the aim of this manuscript. 3. In-Field Testing Results The system was validated/tested in an operative railway scenario in Benevento (Italy) through different testing sessions. Each session was slightly longer than 24 h, across two different days. The sessions typically started in the mornings from 10:00 to 13:00, when the timetable of trains does not report any transits and was carried out until 13:00 of the following day. In particular, during the first day’s morning (Day 1), when it was possible to access the tunnel, the sensors were placed in the package (previously installed into the tunnel) and, after acclimatization time, we could simulate the presence of butane by injecting the gas in the proximity of the package. At about 13:00 of Day 1, the line was again operative, and thus we could monitor the sensor response during normal operations. During the second day’s morning (Day 2), additional gas tests were performed up to 13:00. During the night, the system was shut down for few hours. In the following, we report the main results of the performed in-field tests. 3.1. Testing Sessions In Figure 8a,b, the data acquired during a testing session conducted in January 2017 are reported. Figure 8a shows the resonance wavelength of the gas sensor considering both raw data (blue line) and temperature and humidity compensated curve (red line). The environmental parameters are also reported in Figure 8b, as acquired from the array of FBGs. It can be clearly observed that average humidity was around 85% during this session, also probably due to rainy days. For convenience, during the tests with butane, the data were acquired every 3 s, whereas this interval was increased during long-term monitoring period to 30 s maximum. The overall figure can be divided into three time-windows: the red marked boxes identify the time intervals when the operator could access the tunnel to simulate the gas leakage (Day 1 and Day 2 mornings, respectively); the second interval (in the blue box) was restricted for any operation due to the train transits and was considered for long-term monitoring. Commercial butane gas was employed for sensor testing through injections of an air–butane mixture at a few centimeters from the sensors package. To this end, butane was first poured into a syringe where it was mixed with air in order to strongly decrease the concentration. Note that this procedure induces a very small gas leak and therefore does not represent a risk for safety. As one can observe, the exposition to butane produces negative spikes in resonance wavelength, i.e., a response similar to what was observed during the chamber characterization. A good reversibility of the response was measured as well when the gas concentration is dispersed after injection, as the aPS film permits the release of the gas molecules, with recovery times in 15–20 min. We focus now on the time interval from 13:00 h of Day 1 to the following morning of Day 2 in Figure 8, when no leakage test could be performed due to train transits. In this period, the behavior of

Sensors 2018, 18, 1435

9 of 14

the (raw) resonance wavelength exhibits a slow monotonic increase of about 2.0 nm, with the presence of faster changes, principally due to the transit of trains (indicated by blue arrows in figure). The former effect is directly related to environmental changes and in particular to the relative humidity increasing from 81% to 88%, while the temperature variations were lower than 2 ◦ C. This effect is quasi-completely compensated in the red curve reported in the same figure. Concerning the second effect, the train Sensors 2018, 18, x FOR PEER REVIEW 9 of 14 transits induce faster changes in the humidity sensor response and faster changes in temperature curve. from compensated it can be noted that some cases response of the sensor theHowever, response of the sensor to trains data, is almost trivial, whereas in in other cases it is the more pronounced and to trains is almost trivial,compensated. whereas in other casesthat it isthis more pronounced notexhaust always completely not always completely We believe could be attributedand to the gas from diesel-powered trains and in theto case electric trains. Additionally, anothertrains compensated. We believe thatthus thiscould coulddisappear be attributed theofexhaust gas from diesel-powered eventcould resulting in a fastinchange of of theelectric raw resonant as well as theevent humidity curveinisa fast and thus disappear the case trains.wavelength Additionally, another resulting measured after midnight, which is not related to a train transit and is well compensated in themidnight, red change of the raw resonant wavelength as well as the humidity curve is measured after curve. which is not related to a train transit and is well compensated in the red curve. (a)

train

train train

train

train gas

gas

(b)

12:30

17:30 Day 1

22:30 0:00

3:00 9:00

13:00

Day 2

Figure 8. Results completein-field in-field testing testing session 2017: (a) (a) LPGLPG sensor response; Figure 8. Results of of complete sessionininJanuary January 2017: sensor response; (b) temperature humidity. (b) temperature andand humidity.

In Figure 9a,b, we report the results of a testing session conducted in March 2017, following the In Figure we report the results a testing session conducted in March following same above9a,b, reported procedure. Duringof this second test session, the recorded relative2017, humidity was the samelower, aboveatreported procedure. During second session, the recorded humidity around 70%, due to the milder this climate of thetest approaching spring season.relative The overall figure was lower, at be around 70%, duetime-windows, to the milder climate of resonance the approaching spring season. overall can divided in two where the wavelength moved in a The range 1.7 nmfigure wide. The one marked by the red where box identifies the morning of Daymoved 2, when leakage was wide. can be divided in two time-windows, the resonance wavelength in gas a range 1.7 nm simulated and usual behavior was observed. Note that gas tests were not performed in the first day and The one marked by the red box identifies the morning of Day 2, when gas leakage was simulated the was line was under maintenance and tests the access tunnel was forbidden. in this case, the usualbecause behavior observed. Note that gas weretonot performed in the Also, first day because the blue box window refers to long-term monitoring and presents train transits. It is important noting line was under maintenance and the access to tunnel was forbidden. Also, in this case, the blue box that, during the evening/night, several quick changes of humidity level were recorded, as from Figure window refers to long-term monitoring and presents train transits. It is important noting that, during 9b green line, probably associated to climate events (e.g., strong gusts of wind in the tunnel). the evening/night, several quick changes of humidity level were recorded, as from Figure 9b green Consequently, quick changes in sensor response (blue line in Figure 9a) are measured. It is worth line, highlighting probably associated climate (e.g., strong removed gusts of after windcompensation, in the tunnel). Consequently, that even iftosuch peaksevents were not completely probably due quicktochanges in sensor response (blue line in Figure 9a) are measured. It is worth highlighting the different response time to humidity of the gas sensor and the reference sensors, their that evenamplitudes if such peaks completely removed after compensation, probably due to the different (redwere line innot Figure 9a) were significantly attenuated. antooverall assessment, we sensor can state during thesensors, differenttheir sessions, the resonance response As time humidity of the gas andthat, the reference amplitudes (red line in wavelength of LPG-based attenuated. gas sensor moved in a range lower than 2 nm as a consequence of the Figure 9a) were significantly changes occurring in all the mentioned perturbations. Concerning the redsessions, curves, where the As an overall assessment, we can state that, during the different the resonance environmental condition changes have been compensated, we found that the resonance wavelength wavelength of LPG-based gas sensor moved in a range lower than 2 nm as a consequence of the changes moves in a range of ±0.3 nm (without considering gas leakages). As a result, we can assess that the occurring in all the mentioned perturbations. Concerning the red curves, where the environmental uncertainty in the resonant wavelength detection induces an error in concentration of ±0.14 vol %, condition changes have been compensated, we found that the resonance wavelength moves in a range i.e., ±8% of the butane LEL. Moreover, it should be remarked that testing was conducted in the worst case, i.e., the trains are powered by diesel motors and thus their emissions can actually be a source of disturbance.

Sensors 2018, 18, 1435

10 of 14

of ±0.3 nm (without considering gas leakages). As a result, we can assess that the uncertainty in the resonant wavelength detection induces an error in concentration of ±0.14 vol %, i.e., ±8% of the butane LEL. Moreover, it should be remarked that testing was conducted in the worst case, i.e., the Sensors 2018, 18, x FOR REVIEW 10 of 14 trains are powered byPEER diesel motors and thus their emissions can actually be a source of disturbance. Sensors 2018, 18, x FOR PEER REVIEW

10 of 14

(a) (a)

train train train train train

train

train

train

train

train

gas gas

(b) (b)

12:30

17:30

22:30 0:00

Day 1 17:30

12:30

3:00 9:00 Day9:00 2 3:00

22:30 0:00

Day 1

13:00 13:00

Day 2

Figure 9. Resultsofofcomplete complete in-field in-field testing testing session 2017: (a) LPG sensor response; Figure 9. Results sessionininMarch March 2017: (a) LPG sensor response; (b) temperature and humidity. Figure 9. Results of complete in-field testing session in March 2017: (a) LPG sensor response; (b) temperature and humidity. (b) temperature and humidity.

3.2. Gas Leakages Detection 3.2. Gas Detection 3.2.Leakages Gas Leakages Detection To better highlight the capability to detect gas leakages, Figure 10 compares the compensated To better highlight the the capability totodetect leakages, Figure compares compensated To highlight capability detect gas T2, leakages, Figure 1010compares thethe compensated curves of better three representative tests (identified asgas T1, and T3), conducted across time intervals of curves of three representative tests (identified T2, andthe T3), conducted across time intervals of three representative tests (identified asT1, T1, during T2, and T3), conducted across time intervals of of 45curves min. These tests correspond to the acquired as signal morning of three different sessions min.railway These correspond to the acquired signal duringthe theduring morning three at45the location. Compensation was signal well performed theof periods, assessions the 45 min. These teststests correspond to the acquired during morning ofthree threedifferent differentsessions at at the location. railwayconditions location. Compensation was well performed during the stable threeas periods, asthe the environmental for the shorter time interval were obviously more than the railway Compensation was well performed during the three periods, theduring environmental environmental conditions the shorter intervalmore were obviously more stable than during the whole monitoring session. Infor particular, thetime temperature and humidity levels were around 10 °C and conditions for the shorter time interval were obviously stable than during the whole monitoring whole monitoring session. particular, were around 10 °C and 80–85%, respectively, duringInthese tests. the temperature and humidity levels ◦ session. In particular, the temperature and humidity levels were around 10 C and 80–85%, respectively, 80–85%, respectively, during these tests.

during these tests.

T1 E1

E2

(a)T1

(b) E1

E2

(a)

(b) T1

E1

E2 T1

E1

E2

E3 T2

E3

E4 E4

T2

E3 E3

T2 T2

E4 E4

T3 T3

E5 E5

E6 E6 T3

E5 E5

E6

T3

E6

Figure 10. Results of in-field tests: (a) sensor response (after compensation); (b) corresponding butane concentration. Figure 10. Results of in-field tests: (a) sensor response (after compensation); (b) corresponding butane Figure 10. Results of in-field tests: (a) sensor response (after compensation); (b) corresponding concentration.

butane Forconcentration. simplicity, to highlight the expositions to butane, red boxes have been used in Figure 10a,b For simplicity, to E6. highlight expositions to butane, redby boxes have been used in 10a,b and labelled from E1 to The gasthe expositions were performed successive injections ofFigure air–butane and labelled from E1 to E6. The gas expositions were performed by successive injections of air–butane

Sensors 2018, 18, 1435

11 of 14

For simplicity, to highlight the expositions to butane, red boxes have been used in Figure 10a,b and labelled from E1 to E6. The gas expositions were performed by successive injections of air–butane mixtures. Consequently, the resonant wavelength moved from the baseline value by a step-wise increasing blue shift. It is worth noting that here the recoveries were in free evolution, in contrast to the chamber characterization setup used in laboratory. Nevertheless, a similar rising time of 2–5 min and recovery time of 15–20 min were measured, as reported in Section 2.1. In Figure 10b, we have reported the corresponding butane concentrations, as calculated from the responses of Figure 10a and the characterization presented in Figure 3b. It appears clear that within the range 0–10 min (i.e., considering E1, E3 and E5), the concentration associated to the single exposition step is always reduced from T1 to T3, whereas the total concentration of the sequence is kept constant at ~1 vol %. Above this range (>10 min), the sensor signal returns to zero, allowing a second sequence of testing (i.e., E2, E4, and E6) characterized by a slightly lower accumulated concentration. As an overall consideration, it can be reported that during the different performed tests, a negative wavelength shift from 0.1 to 1.1 nm was revealed by the developed sensor, validating its reliability in detecting gas up to the maximum used concentration of 1 vol %. 4. Conclusions In this work, the in-field demonstration of a fiber optic gas monitoring system, installed in an operative railway tunnel located in the south of Italy, was performed. As a target, the attention was focused on the butane, which is the main component of liquefied petroleum gas, since it is one of the dangerous species being transported with higher frequency (first in Italy) by railways lines. The developed system comprises a sensing unit, installed directly in the tunnel, and a control unit, located in the station technical room. Customized software was developed in order to obtain the real-time measurements of gas concentrations, temperature and humidity levels in the tunnel in normal operative conditions of train transits. Specifically, during the validation tests, we have considered two main conditions: with the railway line under normal operation, and in the presence of controlled gas expositions. The results demonstrated that the system is always capable of detecting the traces of the gas during controlled expositions, whereas, during normal operation (i.e., in presence of trains transits), we found some false positives that can be quantified in ±8% of the LEL (slightly lower than the same value in commercial sensors). It was demonstrated that, thanks to the advantages of all in-fiber technology, the system has the following important features: immunity to electromagnetic interferences, which are indeed a real problem in railways; operating in harsh conditions (even an explosive atmosphere); and operating at low temperature. The developed system could be easily scaled by introducing several sensing units due to easy multiplexing (in our case, up to eight measuring points are possible with the same setup) with the potential advantage of covering longer distances for monitoring purposes. Author Contributions: F.E. and A.Z. contributed equally to this work (both have to be considered as first author); F.E., G.P., S.C. and A.I. designed and developed the optical transducers; A.Z., S.Z., A.B. and M.Z. developed the sensitive material and the gas test chamber; F.E. and A.Z. fabricated the gas sensor and performed the characterizations; M.C. designed and developed the sensors package; F.E., M.C., A.C., S.C., M.Z. and A.I. performed the in-field tests; F.E., S.C., M.Z. and A.I. analyzed the data; F.E. took the lead in writing the manuscript; All the authors discussed the results and contributed to the final manuscript; A.C. supervised and managed the project OPTOFER. Funding: This research was funded by the Italian Ministry of Education, Universities and Research (MIUR) through the national project PON-OPTOFER (Tecnologie OPTOelettroniche innovative per il monitoraggio e la diagnostica dell’infrastruttura FERroviaria) PON03PE_00155_1. Acknowledgments: The authors would like to thank Migma Srl for developing the LabView software for the Control Unit, and would like to thank Ansaldo STS spa and Rete Ferroviaria Italiana spa for their support during the system installation. Conflicts of Interest: The authors declare no conflict of interest.

Sensors 2018, 18, 1435

12 of 14

References 1. 2. 3. 4. 5. 6. 7.

8. 9.

10. 11.

12. 13. 14.

15. 16. 17. 18.

19.

20.

21.

Kohl, D. Function and applications of gas sensors. J. Phys. D Appl. Phys. 2001, 34, R125–R149. [CrossRef] Liu, X.; Cheng, S.; Liu, H.; Hu, S.; Zhang, D.; Ning, H. A Survey on Gas Sensing Technology. Sensors 2012, 12, 9635–9665. [CrossRef] [PubMed] Hill, K.O.; Meltz, G. Fiber Bragg grating technology fundamentals and overview. J. Lightwave Technol. 1997, 15, 1263–1276. [CrossRef] Kersey, A.D.; Davis, M.A.; Patrick, H.J.; LeBlanc, M.; Koo, K.P.; Askins, C.G.; Putnam, M.A.; Friebele, E.J. Fiber grating sensors. J. Lightwave Technol. 1997, 15, 1442–1463. [CrossRef] Some, S.; Xu, Y.; Kim, Y.; Yoon, Y.; Qin, H.; Kulkarni, A.; Kim, T.; Lee, H. Highly Sensitive and Selective Gas Sensor Using Hydrophilic and Hydrophobic Graphenes. Sci. Rep. 2013, 3, 1868. [CrossRef] [PubMed] Liu, J.; Sun, Y.; Fan, X. Highly versatile fiber-based optical Fabry-Pérot gas sensor. Opt. Express 2009, 17, 2731. [CrossRef] [PubMed] Zotti, A.; Zuppolini, S.; Giordano, M.; Zarrelli, M.; Borriello, A.; De Luca, G. Optical aliphatic hydrocarbon gas sensor based on Titanium Dioxide thin film. In Proceedings of the 2015 18th AISEM Annual Conference, Trento, Italy, 3–5 February 2015. Sun, Y.; Fan, X. Optical ring resonators for biochemical and chemical sensing. Anal. Bioanal. Chem. 2011, 399, 205–211. [CrossRef] [PubMed] Wu, Y.; Yao, B.; Zhang, A.; Rao, Y.; Wang, Z.; Cheng, Y.; Gong, Y.; Zhang, W.; Chen, Y.; Chiang, K.S. Graphene-coated microfiber Bragg grating for high-sensitivity gas sensing. Opt. Lett. 2014, 39, 1235. [CrossRef] [PubMed] Yang, M.; Yang, Z.; Dai, J.; Zhang, D. Fiber optic hydrogen sensors with sol–gel WO3 coatings. Sens. Actuators B Chem. 2012, 166–167, 632–636. [CrossRef] Pilla, P.; Iadicicco, A.; Contessa, L.; Campopiano, S.; Cutolo, A.; Giordano, M.; Guerra, G.; Cusano, A. Optical Chemo-Sensor Based on Long Period Gratings Coated With delta Form Syndiotactic Polystyrene. IEEE Photonics Technol. Lett. 2005, 17, 1713–1715. [CrossRef] Pinto, A.M.R.; Lopez-Amo, M. Photonic Crystal Fibers for Sensing Applications. J. Sens. 2012, 2012, 598178. [CrossRef] Mishra, S.K.; Kumari, D.; Gupta, B.D. Surface plasmon resonance based fiber optic ammonia gas sensor using ITO and polyaniline. Sens. Actuators B Chem. 2012, 171–172, 976–983. [CrossRef] Usha, S.P.; Mishra, S.K.; Gupta, B.D. Fiber optic hydrogen sulfide gas sensors utilizing ZnO thin film/ZnO nanoparticles: A comparison of surface plasmon resonance and lossy mode resonance. Sens. Actuators B Chem. 2015, 218, 196–204. [CrossRef] Iadicicco, A.; Campopiano, S.; Giordano, M.; Cusano, A. Spectral behavior in thinned long period gratings: Effects of fiber diameter on refractive index sensitivity. Appl. Opt. 2007, 46, 6945–6952. [CrossRef] [PubMed] Cusano, A.; Iadicicco, A.; Pilla, P.; Contessa, L.; Campopiano, S.; Cutolo, A.; Giordano, M. Mode transition in high refractive index coated long period gratings. Opt. Express 2006, 14, 19–34. [CrossRef] [PubMed] Cusano, A.; Iadicicco, A.; Pilla, P.; Cutolo, A.; Giordano, M.; Campopiano, S. Sensitivity characteristics in nanosized coated long period gratings. Appl. Phys. Lett. 2006, 89, 201116. [CrossRef] Del Villar, I. Ultrahigh-sensitivity sensors based on thin-film coated long period gratings with reduced diameter, in transition mode and near the dispersion turning point. Opt. Express 2015, 23, 8389–8398. [CrossRef] [PubMed] Pilla, P.; Trono, C.; Baldini, F.; Chiavaioli, F.; Giordano, M.; Cusano, A. Giant sensitivity of long period gratings in transition mode near the dispersion turning point: An integrated design approach. Opt. Lett. 2011, 37, 4152–4154. [CrossRef] [PubMed] ´ Smietana, M.; Koba, M.; Mikulic, P.; Bock, W.J. Towards refractive index sensitivity of long-period gratings at level of tens of µm per refractive index unit: Fiber cladding etching and nano-coating deposition. Opt. Express 2016, 24, 11897–11904. [CrossRef] [PubMed] Wang, T.; Korposh, S.; James, S.; Tatam, R.; Lee, S.-W. Optical fiber long period grating sensor with a polyelectrolyte alternate thin film for gas sensing of amine odors. Sens. Actuators B Chem. 2013, 185, 117–124. [CrossRef]

Sensors 2018, 18, 1435

22.

23. 24.

25. 26.

27.

28.

29. 30.

31.

32.

33.

34. 35.

36.

37.

38. 39. 40. 41.

13 of 14

Cusano, A.; Iadicicco, A.; Pilla, P.; Contessa, L.; Campopiano, S.; Cutolo, A.; Giordano, M.; Guerra, G. Coated long-period fiber gratings as high-sensitivity optochemical sensors. J. Lightwave Technol. 2006, 24, 1776–1786. [CrossRef] Hromadka, J.; Tokay, B.; James, S.; Tatam, R.P.; Korposh, S. Optical fibre long period grating gas sensor modified with metal organic framework thin films. Sens. Actuators B Chem. 2015, 221, 891–899. [CrossRef] Hromadka, J.; Korposh, S.; Partridge, M.C.; James, S.W.; Davis, F.; Crump, D.; Tatam, R.P. Multi-parameter measurements using optical fibre long period gratings for indoor air quality monitoring. Sens. Actuators B Chem. 2017, 244, 217–225. [CrossRef] Topliss, S.M.; James, S.W.; Davis, F.; Higson, S.P.J.; Tatam, R.P. Optical fibre long period grating based selective vapour sensing of volatile organic compounds. Sens. Actuators B Chem. 2010, 143, 629–634. [CrossRef] Barnes, J.A.; Brown, R.S.; Cheung, A.H.; Dreher, M.A.; Mackey, G.; Loock, H.-P. Chemical sensing using a polymer coated long-period fiber grating interrogated by ring-down spectroscopy. Sens. Actuators B Chem. 2010, 148, 221–226. [CrossRef] Yang, J.; Zhou, L.; Huang, J.; Tao, C.; Li, X.; Chen, W. Sensitivity enhancing of transition mode long-period fiber grating as methane sensor using high refractive index polycarbonate/cryptophane A overlay deposition. Sens. Actuators B Chem. 2015, 207, 477–480. [CrossRef] Yang, J.; Che, X.; Shen, R.; Wang, C.; Li, X.; Chen, W. High-sensitivity photonic crystal fiber long-period grating methane sensor with cryptophane-A-6Me absorbed on a PAA-CNTs/PAH nanofilm. Opt. Express 2017, 25, 20258. [CrossRef] [PubMed] Falate, R.; Kamikawachi, R.C.; Müller, M.; Kalinowski, H.J.; Fabris, J.L. Fiber optic sensors for hydrocarbon detection. Sens. Actuators B Chem. 2005, 105, 430–436. [CrossRef] Monteiro-Silva, F.; Santos, J.L.; de Almeida, J.M.M.M.; Coelho, L. Quantification of Ethanol Concentration in Gasoline Using Cuprous Oxide Coated Long Period Fiber Gratings. IEEE Sens. J. 2018, 18, 1493–1500. [CrossRef] Esposito, F.; Zotti, A.; Ranjan, R.; Zuppolini, S.; Borriello, A.; Campopiano, S.; Zarrelli, M.; Iadicicco, A. Single-Ended Long Period Fiber Grating Coated with Polystyrene Thin Film for Butane Gas Sensing. J. Lightwave Technol. 2018, 36, 825–832. [CrossRef] Quero, G.; Consales, M.; Severino, R.; Vaiano, P.; Boniello, A.; Sandomenico, A.; Ruvo, M.; Borriello, A.; Diodato, L.; Zuppolini, S.; et al. Long period fiber grating nano-optrode for cancer biomarker detection. Biosens. Bioelectron. 2016, 80, 590–600. [CrossRef] [PubMed] ´ Janczuk-Richter, M.; Dominik, M.; Ro´zniecka, E.; Koba, M.; Mikulic, P.; Bock, W.J.; Ło´s, M.; Smietana, M.; Niedziółka-Jönsson, J. Long-period fiber grating sensor for detection of viruses. Sens. Actuators B Chem. 2017, 250, 32–38. [CrossRef] Chiavaioli, F.; Baldini, F.; Tombelli, S.; Trono, C.; Giannetti, A. Biosensing with optical fiber gratings. Nanophotonics 2017, 6, 663–679. [CrossRef] Esposito, F.; Ranjan, R.; Stăncălie, A.; Sporea, D.; Negu¸t, D.; Becherescu, N.; Campopiano, S.; Iadicicco, A. Real-time analysis of arc-induced Long Period Gratings under gamma irradiation. Sci. Rep. 2017, 7, 43389. [CrossRef] [PubMed] Stancălie, A.; Esposito, F.; Ranjan, R.; Bleotu, P.; Campopiano, S.; Iadicicco, A.; Sporea, D. Arc-induced Long Period Gratings in standard and speciality optical fibers under mixed neutron-gamma irradiation. Sci. Rep. 2017, 7, 15845. [CrossRef] [PubMed] Stăncălie, A.; Sporea, D.; Negu¸t, D.; Esposito, F.; Ranjan, R.; Campopiano, S.; Iadicicco, A. Long Period Gratings in unconventional fibers for possible use as radiation dosimeter in high-dose applications. Sens. Actuators A Phys. 2018, 271, 223–229. [CrossRef] Esposito, F.; Ranjan, R.; Campopiano, S.; Iadicicco, A. Arc-Induced Long Period Gratings from Standard to Polarization-Maintaining and Photonic Crystal Fibers. Sensors 2018, 18, 918. [CrossRef] [PubMed] Esposito, F.; Ranjan, R.; Campopiano, S.; Iadicicco, A. Experimental Study of the Refractive Index Sensitivity in Arc-induced Long Period Gratings. IEEE Photonics J. 2017, 9, 1–10. [CrossRef] Iadicicco, A.; Ranjan, R.; Esposito, F.; Campopiano, S. Arc-Induced Long Period Gratings in Polarization-Maintaining Panda Fiber. IEEE Photonics Technol. Lett. 2017, 29, 1533–1536. [CrossRef] Ranjan, R.; Esposito, F.; Iadicicco, A.; Campopiano, S. Arc-Induced Long Period Gratings in Phosphorus-Doped Fiber. IEEE Photonics Technol. Lett. 2017, 29, 611–614. [CrossRef]

Sensors 2018, 18, 1435

42. 43.

44. 45.

14 of 14

Del Villar, I.; Matías, I.; Arregui, F.; Lalanne, P. Optimization of sensitivity in Long Period Fiber Gratings with overlay deposition. Opt. Express 2005, 13, 56–69. [CrossRef] [PubMed] Ranjan, R.; Esposito, F.; Iadicicco, A.; Stancalie, A.; Sporea, D.; Campopiano, S. Comparative Study of Long-Period Gratings Written in Standard and Fluorine-Doped Fibers by Electric Arc Discharge. IEEE Sens. J. 2016, 16, 4265–4273. [CrossRef] Ranjan, R.; Esposito, F.; Campopiano, S.; Iadicicco, A. Sensing Characteristics of Arc-Induced Long Period Gratings in Polarization-Maintaining Panda Fiber. IEEE Sens. J. 2017, 17, 6953–6959. [CrossRef] Huang, X.F.; Sheng, D.R.; Cen, K.F.; Zhou, H. Low-cost relative humidity sensor based on thermoplastic polyimide-coated fiber Bragg grating. Sens. Actuators B Chem. 2007, 127, 518–524. [CrossRef] © 2018 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/).