Journal of Sensors

Fiber and Integrated Waveguide-Based Optical Sensors

Guest Editors: Valerio Pruneri, Christos Riziotis, Peter G. R. Smith, and Athanasios Vasilakos

Fiber and Integrated Waveguide-BasedOptical Sensors

Journal of Sensors

Fiber and Integrated Waveguide-BasedOptical Sensors

Guest Editors: Valerio Pruneri, Christos Riziotis,Peter G. R. Smith, and Athanasios Vasilakos

Copyright 2009 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in volume 2009 of Journal of Sensors. All articles are open access articles distributed under the CreativeCommons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the originalwork is properly cited.

Editor-in-ChiefFrancisco J. Arregui, Public University of Navarre, Spain

Associate Editors

Francesco Baldini, ItalyAmine Bermak, Hong KongMehmet Can Vuran, USAAndrea Cusano, ItalyCristina E. Davis, USAUtkan Demirci, USAJiri Homola, Czech RepublicBernhard Jakoby, AustriaK. Kalantar-Zadeh, AustraliaChalla S S R Kumar, USAHiroki Kuwano, Japan

Yongxiang Li, ChinaFranco Maloberti, ItalyIgnacio R. Matias, SpainMike McShane, USAIgor L. Medintz, USATommaso Melodia, USAJoan Ramon Morante, SpainS. C. Mukhopadhyay, New ZealandK. S. Narayan, IndiaMichele Penza, ItalyPavel Ripka, Czech Republic

Giorgio Sberveglieri, ItalyPietro Siciliano, ItalyWeilian Su, USAIsao Takayanagi, JapanMaria Tenje, DenmarkAthanasios T. Vasilakos, GreeceWojtek Wlodarski, AustraliaStanley E. Woodard, USAHai Xiao, USA

Contents

Fiber and Integrated Waveguide-Based Optical Sensors, Valerio Pruneri, Christos Riziotis,Peter G. R. Smith, and Athanasios VasilakosVolume 2009, Article ID 171748, 3 pages

Surface Plasmon Resonance-Based Fiber Optic Sensors: Principle, Probe Designs, and SomeApplications, B. D. Gupta and R. K. VermaVolume 2009, Article ID 979761, 12 pages

Microstructured and Photonic Bandgap Fibers for Applications in the Resonant Bio- and ChemicalSensors, Maksim SkorobogatiyVolume 2009, Article ID 524237, 20 pages

Properties of Specialist Fibres and Bragg Gratings for Optical Fiber Sensors, John CanningVolume 2009, Article ID 871580, 17 pages

Highly Sensitive Sensors Based on Photonic Crystal Fiber Modal Interferometers, Joel Villatoro,Vittoria Finazzi, Gonccal Badenes, and Valerio PruneriVolume 2009, Article ID 747803, 11 pages

Gemini Fiber for Interferometry and Sensing Applications, E. Zetterlund, A. Loriette, C. Sterner,M. Eriksson, H. Eriksson-Quist, and W. MargulisVolume 2009, Article ID 196380, 7 pages

Photonic Crystal Fiber Sensors for Strain and Temperature Measurement, Jian Ju and Wei JinVolume 2009, Article ID 476267, 10 pages

Photonic Crystal Fiber Temperature Sensor Based on Quantum Dot Nanocoatings, Beatriz Larrion,Miguel Hernaez, Francisco J. Arregui, Javier Goicoechea, Javier Bravo, and Ignacio R. MatasVolume 2009, Article ID 932471, 6 pages

Temperature Sensor Based on Ge-Doped Microstructured Fibers, Salvador Torres-Peiro, Antonio Dez,Jose Luis Cruz, and Miguel Vicente AndresVolume 2009, Article ID 417540, 5 pages

Wide and Fast Wavelength-Swept Fiber Laser Based on Dispersion Tuning for Dynamic Sensing,Shinji Yamashita, Yuichi Nakazaki, Ryosei Konishi, and Osamu KusakariVolume 2009, Article ID 572835, 12 pages

Plastic Optical Fibre Sensors for Structural Health Monitoring: A Review of Recent Progress,K. S. C. Kuang, S. T. Quek, C. G. Koh, W. J. Cantwell, and P. J. ScullyVolume 2009, Article ID 312053, 13 pages

Fabry-Perot Fiber-Optic Sensors for Physical Parameters Measurement in Challenging Conditions,Eric PinetVolume 2009, Article ID 720980, 9 pages

Covalent Attachment of Carbohydrate Derivatives to an Evanescent Wave Fiber Bragg GratingBiosensor, Christopher J. Stanford, Geunmin Ryu, Mario Dagenais, Matthew T. Hurley, Karen J. Gaskell,and Philip DeShongVolume 2009, Article ID 982658, 7 pages

Remote System for Detection of Low-Levels of Methane Based on Photonic Crystal Fibres andWavelength Modulation Spectroscopy, J. P. Carvalho, H. Lehmann, H. Bartelt, F. Magalhaes, R.Amezcua-Correa, J. L. Santos, J. Van Roosbroeck, F. M. Araujo, L. A. Ferreira, and J. C. KnightVolume 2009, Article ID 398403, 10 pages

Fiber-Optic Aqueous Dipping Sensor Based on Coaxial-Michelson Modal Interferometers, Paola Barrios,David Saez-Rodrguez, Amparo Rodrguez, Jose Luis Cruz, Antonio Dez, and Miguel Vicente AndresVolume 2009, Article ID 815409, 4 pages

Surface Plasmon Resonance-Based Fiber and Planar Waveguide Sensors, Raman Kashyap andGalina NemovaVolume 2009, Article ID 645162, 9 pages

Planar Bragg Grating SensorsFabrication and Applications: A Review, I. J. G. Sparrow, P. G. R. Smith,G. D. Emmerson, S. P. Watts, and C. RiziotisVolume 2009, Article ID 607647, 12 pages

Design and Fabrication of Slotted Multimode Interference Devices for Chemical and Biological Sensing,M. Mayeh, J. Viegas, P. Srinivasan, P. Marques, J. L. Santos, E. G. Johnson, and F. FarahiVolume 2009, Article ID 470175, 11 pages

Hindawi Publishing CorporationJournal of SensorsVolume 2009, Article ID 171748, 3 pagesdoi:10.1155/2009/171748

Editorial

Fiber and Integrated Waveguide-Based Optical Sensors

Valerio Pruneri,1 Christos Riziotis,2 Peter G. R. Smith,3 and Athanasios Vasilakos4

1The Institute of Photonic Sciences (ICFO) and ICREA, Mediterranean Technology Park, Avenue del Canal Olmpic s/n,08860 Castelldefels (Barcelona), Spain

2Photonics for Nanoapplications Laboratory, Theoretical and Physical Chemistry Institute, National Hellenic Research Foundation(NHRF), 48 Vassileos Constantinou Avenue, 11635 Athens, Greece

3Optoelectronics Research Centre, University of Southampton, Southampton SO17 1BJ, UK4Department of Electrical and Computer Engineering, National Technical University of Athens, 15780 Athens, Greece

Correspondence should be addressed to Christos Riziotis, riziotis@eie.gr

Received 15 December 2009; Accepted 15 December 2009

Copyright 2009 Valerio Pruneri et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Over the last years, a large part of the activity in appliedphotonics and especially in ber or integrated waveguide-based devices has been transferred partially from the photon-ics telecommunications industry towards the optical sensorsresearch. Further to the necessity due to telecommunicationssector turn down, it has been proven that this shift has beenwelcomed by the development in relevant industrial sectors(pharmaceutical, medical) where new requirements for veryaccurate control of the manufacturing process are required.This increasing research eort on all-optical sensors tech-nology, combined with emerging and demanding applica-tions, has demonstrated a promising technological platformcharacterized by unique sensitivity, compactness, reliability,electromagnetic immunity, and low cost, promoting themto a preferable solution for real-world applications, frommechanical sensing to chemical/biochemical and pharma-ceutical industry. The inherent also capability of photonicstechnology for the ecient sensing-signal transmissionthrough optical bers suggests an enhanced functionalityfrom a systems perspective, by enabling the high-speedinterconnection of multiple remote sensing points, eitherthrough a single readout and administration unit, or througha distributed network. Furthermore the need for develop-ment of large-scale ad hoc sensor networks requires reliableautonomous and controllable sensing nodes and opticalsensors exhibit very attractive and unique characteristics toplay key role in this area. Emerging technologies combiningnew design concepts and operational approaches such asmicrostructured bers (PCFs), tapered nanobers, Bragggratings, and long-period gratings, interferometric devices,

as well as Surface Plasmon Resonance (SPR) devices haveshown a strong impetus for novel applications. A criticalissue which could dramatically enhance the performance ofsuch functional devices is the use of novel polymers andnanostructured materials able to improve the sensitivity andexpand also sensors selectivity range.

This special issue is completely devoted in this dynamicarea of optical sensors, aiming to broadly cover aspects suchas material properties, fabrication techniques, modeling,optimization, and novel applications. Hosting 17 represen-tative papers demonstrates successful engineering of novelber and planar optical sensors for a variety of applicationssuch as physical, chemical, and biosensing. Selected papershave been invited in order to provide the state of the art andcurrent trends in distinct hot areas.

The issue begins with a group of review papers on keyber-based sensor categories, with the rst invited paperfrom Gupta and Kumar which gives an up-to-date reviewon ber optic-based surface plasmon resonance sensorsdemonstrating applications in measuring various physical,chemical, and biochemical parameters. Various designs ofthe ber optic SPR probe were reported there for theenhancement of sensors sensitivity. The second invitedpaper by Skorobogatiy presents a comprehensive reviewon microstructured optical ber and photonic bandgap(PBG) ber-based resonant optical sensors. Two sensorarchitectures are discussed where in the rst one wereemployed hollow core photonic bandgap bers where core-guided mode is strongly conned in the analyte-lledcore. The second sensor case employed metalized photonic

2 Journal of Sensors

bandgap waveguides and bers, where core-guided modewas phase matched with a plasmon propagating at themetalized ber/analyte interface. Operational regions of theresonant sensors are reviewed covering a wide range ofwavelengths from the visible to terahertz. Canning presentsin his invited paper the state-of-the-art work on sensorsbased on Bragg gratings and microstructured bers. Thepaper presents the current situation and the measurementschallenges with temperature and strain and discusses as wellnovel engineering concepts such as an integrated lab in aber. Another architecture based on modal interferometerson PCFs is reviewed by Villatoro et al. as a solution for highlysensitive sensors of low thermal sensitivity and applicationsranging from strain, temperature, to refractive index andvolatile organics. Sagnac, Mach-Zehnder, or Michelson-likeinterferometers can be fabricated by dierent postprocessingtechniques such as grating inscription, tapering or cleaving,and splicing. Margulis et al. propose in their invited researchpaper a novel and alternative interferometric architecturebased on the plurality of individual ber preforms drawntogether but nearly maintaining their original shape. Thepotential temperature independence and the ease of splicingof Gemini bers make them an attractive solution for sensorsdevelopment in a monolithic-robust multicore ber design.

The following three papers present architectures andapplications of PCFs for physical parameters sensing. Therst paper by Ju and Jin is an invited review on photoniccrystal bers utilizing either long-period gratings or in-bermodal interferometers for strain and temperature sensing.The paper suggests that air-silica PCF sensors are comparableor better than those implemented in conventional single-mode bers but the temperature sensitivities of the PCFsensors are much lower. Next paper by Larrion et al.presents a novel conguration of a temperature sensor basedon a PCF with quantum dot nanocoatings in its innerholes, deposited by means of the Layer-by-Layer technique.The paper studies for a temperature range from 40 to70C the optical properties of these sensors and introducesalso the consideration of the full width at half maximum(FWHM) as a new insightful characterization parameter.Another conguration of a temperature sensor with a low-cost amplitude interrogation technique is presented in thepaper by Torres-Peiro et al. based on the cuto properties ofthe fundamental mode in a liquid lled Y-shaped Ge-dopedmicrostructured ber. The sensitivity is mainly determinedby the thermo-optic coecient of the lling liquids andvalues of 25 nm/C with detection limit of about 0.001C arereported.

The next three papers present dierent approaches formeasuring also physical parameters but using conventionalbers. The invited review by Yamashita presents a novelwide and fast wavelength-swept ber laser architecture fordynamic and accurate ber sensing, based on the dispersiontuning technique, by modulating the loss/gain in the disper-sive laser cavity. Kuang et al. present in their invited reviewa low-cost platform based on plastic optical bers-POF forthe quite important area of Structural Health Monitoring(SHM). Between dierent possible schemes, the intensity-based interrogation capability of POF sensors provides

additional a very favorable solution of low-cost SHM systemsimplementation. The next paper by Pinet gives from anindustrial and applications perspective the state-of-the-artsituation on miniaturized sensors based on Fabry-Perot beroptic implementation. This alternative architecture providesa mature and reliable technology for strain, temperature,pressure, displacement, or refractive index measurements,and with a lot of applications ranging from industrial tomedical.

The following three papers are devoted to ber opticbased chemical and biosensing, where the rst invited paperby Dagenais presents a promising sensor based on an etchedBragg grating functionalized with glucopyranosyl-siloxaneconjugate in order to interact selectively with glucose bindingproteins, altering thus the surfaces refractive index. Nextpaper by Carvalho et al. presents a hollow-core PCF fordetecting low levels of methane. A prototype is demonstratedin a portable system incorporating an interrogation schemebased on the Wavelength Modulation Spectroscopy tech-nique. In the next paper by Andres et al., another simpletechnique based on single mode ber at 850 nm is presented,where a long-period grating as an equivalent 3 dB beamis used to form a coaxial-Michelson modal interferometer.Direct dipping of the sensor head in water solutions permitsthe measure of small refractive index changes.

The nal group of papers is devoted to the integratedoptics approach which oers a promising technologicalplatform towards robust and multifunctional sensors phys-ically compatible with microuidics. The invited paperby Kashyap reviews the state of the art on Surface Plas-mon Resonance-based devices focusing especially on theirintegrated platform approach as a route for low cost,mass production, disposable sensors. The next paper bySparrow et al. reviews the emerging class of integratedBragg grating-based sensors, and examines in detail thearchitecture, fabrication, and applications of special sensorschips fabricated by the exible direct UV writing techniqueon silica-on-silicon wafers. A variety of applications rangingfrom temperature, refractive index, chemical, and biosensingare described and also indicative specic problems in foodindustry for monitoring long-term industrial process suchas fermentation are demonstrated. The nal invited paperis by Mayeh et al. describing another class of integrateddevices, based on the operational principle of slottedmultimode interference (MMI). The paper presents designand modeling results towards the tuning of the device inorder to detect protein-based molecules or water-solublechemical or biological materials. Fabrication of the devicehas been demonstrated in silicon oxynitride (SiON) as highlystable to the reactivity with biological agents and processingchemicals.

Introducing this special issue to the Journal of Sensors,we would like to thank all the authors for their prompt andvaluable contributions and also the reviewers for their criticalhelp, necessary to achieve a high level of papers qualityand make thus possible the completion of this special issue.Also we would like to thank the Editor in Chief ProfessorFrancisco J. Arregui and the Editorial Board for approvingthis special issue as well as the Journals sta for their

Journal of Sensors 3

professionalism and eective consideration of all the detailsduring the preparation of the issue.

Valerio PruneriChristos RiziotisPeter G.R. Smith

Athanasios Vasilakos

Hindawi Publishing CorporationJournal of SensorsVolume 2009, Article ID 979761, 12 pagesdoi:10.1155/2009/979761

Review Article

Surface Plasmon Resonance-Based Fiber Optic Sensors:Principle, Probe Designs, and Some Applications

B. D. Gupta and R. K. Verma

Department of Physics, Indian Institute of Technology Delhi, New Delhi 110016, India

Correspondence should be addressed to B. D. Gupta, bdgupta@physics.iitd.ernet.in

Received 26 January 2009; Accepted 2 June 2009

Recommended by Christos Riziotis

Surface plasmon resonance technique in collaboration with optical ber technology has brought tremendous advancements insensing of various physical, chemical, and biochemical parameters. In this review article, we present the principle of SPR techniquefor sensing and various designs of the ber optic SPR probe reported for the enhancement of the sensitivity of the sensor. Inaddition, we present few examples of the surface plasmon resonance- (SPR-) based ber optic sensors. The present review mayprovide researchers valuable information regarding ber optic SPR sensors and encourage them to take this area for furtherresearch and development.

Copyright 2009 B. D. Gupta and R. K. Verma. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

1. Introduction

Surface plasmon resonance (SPR) is one of the most promis-ing optical techniques that nd applications in dierentelds. The rst sensing application of SPR technique wasreported in 1983 [1]. Since then, numerous SPR sensingstructures for chemical and biochemical sensing have beenreported. In SPR technique, a TM (transverse magnetic) orp-polarized light causes the excitation of electron densityoscillations (known as surface plasmon wave, SPW) at themetal-dielectric interface. When the energy as well as themomentum of both, the incident light and SPW, match,a resonance occurs which results in a sharp dip in thereected light intensity. The resonance condition dependson the angle of incidence, wavelength of the light beam,and the dielectric functions of both the metal as well as thedielectric. If the wavelength is kept constant and the angle ofincidence is varied, then the sharp dip appears at a particularangle and the method is called angular interrogation. Inanother method, called spectral or wavelength interrogation,the angle of the incident beam is kept constant and thewavelength is varied. In this method, resonance occurs at aparticular wavelength. The resonance parameter (angle orwavelength) depends on the refractive index of the dielectricmedium. Change in refractive index changes the value of the

resonance parameter. To excite surface plasmons, generally,a prism is used [29]. The prism-based SPR sensing devicehas a number of shortcomings such as its bulky size andthe presence of various optical and mechanical (moving)parts. Further, the prism-based SPR sensing device cannotbe used for remote sensing applications. These shortcomingscan be overcome if an optical ber is used in place ofprism. The additional advantage of optical ber is that theSPR probe can be miniaturized which can be advantageousfor samples which are available in minute quantity orare costly. Due to these advantages the surface plasmonresonance-based optical ber sensors have drawn a lotof attention [1023]. Both experimental and theoreticalinvestigations have been reported in literature on the SPR-based ber optic sensors. The performance of these sensorsis, generally, evaluated in terms of sensitivity and signal-to-noise ratio (or detection accuracy). As is known, thehigher the values of these parameters, the better is thesensor.

The present review begins with a section on principle ofthe sensing technique. The section includes the descriptionof the performance parameters of the sensor: sensitivityand signal-to-noise ratio or detection accuracy. In the nextsection, we present the simplest SPR-based ber optic sensorand discuss its modulation scheme. Section 4 of the present

2 Journal of Sensors

review deals with various designs of the SPR probe studiedto enhance the performance of the ber optic sensor. Thedesigns include tapered, U-shaped and side-polished. Theadvantages of bimetallic coatings, addition of dopants inber core and the choice of the metals for coating have beendiscussed in the same section. In the last section, we reviewsome of the SPR-based ber optic sensors in details that havebeen reported in literature.

2. Principle

A metal-dielectric interface supports charge density oscil-lations along the interface which are called surface plasmaoscillations. The quantum of these oscillations is given thename as surface plasmon. The surface plasmons are accom-panied by a longitudinal (TM- or p-polarized) electric eldwhich decays exponentially in metal as well as in dielectricmedium. The electric eld has its maximum at metal-dielectric interface. The TM-polarization and exponentialdecay of electric eld are found by solving the Maxwellequation for semi-innite media of metal and dielectric withan interface of metal-dielectric. The propagation constant(KSP) of the surface plasmon wave propagating along themetal-dielectric interface is given by

KSP = c

(msm + s

)1/2, (1)

where m and s are the dielectric constants of metal andthe dielectric medium, respectively, is the frequency ofincident light, and c is the velocity of light. From (1) it maybe noted that the propagation constant of surface plasmonwave depends on the dielectric constants of both the metaland the dielectric medium.

The surface plasmons can be excited by light withsame polarization state as that of surface plasmons. Thepropagation constant (Ks) of the light wave with frequency propagating through the dielectric medium is given by

Ks = c

s. (2)

Since m < 0 (for metal) and s > 0 (for dielectric),for a given frequency, the propagation constant of surfaceplasmon (KSP) is greater than that of the light wave indielectric medium (Ks). To excite surface plasmons, twopropagation wave-vectors should be equal. Hence, the directlight cannot excite surface plasmons at a metal-dielectricinterface. To excite surface plasmons the momentum andhence the wave vector of the exciting light in dielectricmedium should be increased. This can be done if instead ofa direct light, evanescent wave is used to excite the surfaceplasmons. To obtain the evanescent wave for the excitationof surface plasmons, a prism with high dielectric constant isused.

When a light beam is incident through one of the twosides of the prism at an angle greater than the criticalangle at prism-air interface the total internal reection oflight beam takes place. In the condition of total internalreection light beam does not return exactly from the

p-polarized light Reflected light

SPW

EW

p

m

s

Figure 1: Kretschmann conguration for the excitation of surfaceplasmon at metal-dielectric interface [23]. c IEEE.

interface. Instead it returns after penetrating in the lowerrefractive index medium (air in this case). The eld in thelower refractive index medium is called evanescent eld andthe wave corresponding to this is called evanescent wave.The evanescent wave propagates along the prism-air interfaceand decays exponentially in the rarer medium (air). Thepropagation constant of the evanescent wave at prism-airinterface is given by

Kev = cp sin , (3)

where p represents the dielectric constant of the materialof the prism and is the angle of incidence of the beam.Increase in the dielectric constant of the prism increases thepropagation constant of the evanescent wave and hence thiscan be made equal to propagation constant of the surfaceplasmon wave to satisfy the surface plasmon resonancecondition. Kretschmann and Reather [24] devised a prism-based conguration, shown in Figure 1, to excite the surfaceplasmons using evanescent wave. In this conguration thebase of the glass prism is coated with a thin layer ofmetal (typically around 50 nm). The metal layer is kept indirect contact with the dielectric medium of lower refractiveindex (such as air or some other dielectric sample). Whena p-polarized light beam is incident through the prismon the prism-metal layer interface at an angle equal toor greater than the critical angle, the evanescent wave isproduced at the prism-metal interface. The excitation ofsurface plasmons occurs when the wave vector of evanescentwave exactly matches with that of the surface plasmonsof similar frequency. This occurs at a particular angle ofincidence res. Thus the resonance condition for surfaceplasmon resonance is

cp sin res =

c

(msm + s

)1/2. (4)

The excitation of surface plasmons at metal/dielectric inter-face results in the transfer of energy from incident light tosurface plasmons, which reduces the intensity of the reectedlight. If the intensity of the reected light is measured as afunction of angle of incidence for xed values of frequency,metal layer thickness and dielectric layer thickness then a

Journal of Sensors 3

Re

ecta

nce

(R)

0

1

Angle ()

res

Figure 2: SPR spectrum [23]. c IEEE.

sharp dip is observed at resonance angle, res, due to anecient transfer of energy to surface plasmons as shownin Figure 2. The minimum of the reected intensity can bequantitatively described with the help of Fresnels equationsfor the three-layer system.

For a given frequency of the light source and the dielectricconstant of metal lm one can determine the dielectricconstant (s) of the sensing layer adjacent to metal layerusing (4) if the value of the resonance angle (res) is known.The resonance angle is experimentally determined by usingangular interrogation method. It is very sensitive to variationin the refractive index of the sensing layer. Increase inrefractive index of the sensing layer increases the resonanceangle.

Sensitivity and detection accuracy or signal-to-noiseratio (SNR) are the two parameters that are used to analyzethe performance of an SPR sensor. For the best performanceboth the parameters should be as high as possible. Sensitivityof an SPR sensor utilizing angular interrogation methoddepends on the amount of shift of the resonance angle witha change in the refractive index of the sensing layer. For agiven refractive index change if the shift in resonance angleincreases this means an increase in the sensitivity of thesensor.

Figure 3 shows a plot of reectance as a function of angleof incidence of the light beam for sensing layer with refractiveindices ns and ns + ns. Increase in refractive index by nsshifts the resonance angle by res. Thus the sensitivity of anSPR sensor utilizing angular interrogation method is denedas

Sn = resns

. (5)

The detection accuracy or the SNR of an SPR sensordepends on how accurately and precisely the sensor candetect the resonance angle and hence, the refractive index ofthe sensing layer. The narrower the width of the SPR curve,the higher is the detection accuracy. Therefore, if 0.5 is theangular width of the SPR curve corresponding to reectance0.5, the detection accuracy of the sensor is assumed to be

Re

ecta

nce

(R)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

res

0.5

ns ns + ns

Figure 3: SPR spectra for two dierent refractive indices of thesensing layer [23]. c IEEE.

inversely proportional to 0.5 (Figure 3). The SNR of theSPR sensor with angular interrogation is, thus, dened as[25]

SNR = res0.5

. (6)

3. Fiber Optic SPR-Based Sensors

In the case of a prism-based SPR sensor evanescent waverequired to excite surface plasmons is resulted due to the totalinternal reection taking place at the prism-metal interfacewhen the angle of incidence of the beam is greater thanthe critical angle. The evanescent wave is also present inan optical ber because the light guidance in an opticalber occurs due to the total internal reection of theguided ray at the core-cladding interface. In the case ofoptical ber the evanescent wave propagates along the core-cladding interface. Therefore, to design an SPR-based beroptic sensor, the prism can be replaced by the core of anoptical ber. To fabricate an SPR-based ber optic sensor,the silicon cladding from a small portion of the ber,preferably from the middle, is removed and the unclad coreis coated with a metal layer. The metal layer is further,surrounded by a dielectric sensing layer as shown in Figure 4.In an SPR-based ber optic sensor, all the guided raysare launched and hence, instead of angular interrogation,spectral interrogation method is used. The light from apolychromatic source is launched into one of the ends of theoptical ber. The evanescent eld produced by the guidedrays excites the surface plasmons at the metal-dielectricsensing layer interface. The coupling of evanescent eldwith surface plasmons strongly depends on wavelength, berparameters, probe geometry, and the metal layer properties.Unlike prism-based SPR sensor, the number of reections formost of the guided rays is greater than one in SPR-based beroptic sensor geometry. The smaller the angle of incidenceat the interface, the larger is the number of reections perunit length in the ber. In addition, the number of reectionsfor any ray also depends on the length of the sensing region

4 Journal of Sensors

SPW

Cladding

Core

Metal layer

Sensing layer

Input light

Transmitted light

Figure 4: A typical probe of an SPR-based ber optic sensor [23].c IEEE.

and the ber core diameter. The number of reections isone of the important parameters that aect the width ofthe SPR curve. The intensity of the light transmitted afterpassing through the SPR sensing region is detected at theother end of the ber as a function of wavelength. The SPRspectrum thus obtained is similar in shape to that shownin Figure 2. The sensing is accomplished by observing thewavelength corresponding to the dip in the spectrum (calledresonance wavelength). A plot of resonance wavelength withthe refractive index of the sensing layer is the calibrationcurve of the SPR sensor. The sensitivity and the detectionaccuracy are determined in the same way as determined inthe case of angular interrogation. The angles are replacedby wavelengths in the denitions of sensitivity and detectionaccuracy.

4. SPR Probe Designs

Sensitivity, detection accuracy, reproducibility, and operatingrange of a sensor are the important parameters to compareit with other sensors. The best sensor is the one that hashigh sensitivity, detection accuracy and operating range,in addition to giving reproducible results. To achieve thisvarious modications have been carried out in the design ofber optic SPR probe. We review some of these modicationsin what follows.

4.1. Bimetallic Coating. For metallic coating on prism baseor ber core either silver or gold is used. Gold demonstratesa higher shift of resonance parameter to change in refractiveindex of sensing layer and is chemically stable. Silver, onthe other hand, displays a narrower width of the SPR curvecausing a higher SNR or detection accuracy. The sharpnessof the resonance curve depends upon the imaginary partof the dielectric constant of the metal. Silver having thelarger value of the imaginary part of the dielectric constantshows narrower width of the SPR curve causing a higherSNR or detection accuracy. On the other hand, the shiftof the resonance curve depends on the real part of thedielectric constant of the metal. The real part of the dielectricconstant is large in the case of gold than the silver and hencegold demonstrates a higher shift of resonance parameter tochange in refractive index of sensing layer. The chemicalstability of silver is poor due to its oxidation. The oxidationof silver occurs as soon as it is exposed to air and especiallyto water, which makes it dicult to give a reproducibleresult and hence the sensor remains unreliable for practical

applications. Therefore, the treatment of silver surface bya thin and dense cover is required. In this regard, a newstructure of resonant metal lm based on bimetallic layers(gold as outer) on the prism base with angular interrogationmethod was reported [25]. The new structure displayeda large shift of resonance angle as gold lm, and alsoshowed narrower resonance curve as silver lm along withthe protection of silver lm against oxidation. The samestructure with spectral interrogation was extended to SPR-based optical ber sensor for the selected and all guidedrays congurations [19]. The sensitivity and SNR wereevaluated numerically for dierent ratios of the thicknessesof silver and gold layers. Figures 5(a) and 5(b) show thevariations of SNR and sensitivity with percentage of silverin bimetallic combination, respectively [19]. As expected theSNR increases with the increase in the silver thickness. Thevariation is almost the same for both kinds of launching butthe values of SNR for all guided rays launching are about1.5 times higher than those corresponding to selected rayslaunching. As far as sensitivity is concerned, it decreases asthe silver layer thickness increases or gold layer thicknessdecreases. Its variation with silver thickness is almost samein the two cases but in terms of values, the selected raylaunching has higher value than the all guided rays launchingcase.

4.2. Choice of Metals. The capability of other metals such ascopper (Cu) and aluminium (Al) for SPR sensor applicationshas also been analyzed. Both metals have the ability to beused for an SPR sensor. The copper has some limitationslike silver. It is chemically vulnerable against oxidation andcorrosion, therefore, its protection is required for a stablesensing application. The SPR sensing capabilities of dierentbimetallic combinations made out of Ag, Au, Al, and Cuwere theoretically investigated for the design of SPR-basedber optic sensors [26]. Figures 6(a) and 6(b) show thevariation of sensitivity and SNR with the ratio of inner layerthickness to total bimetallic thickness for dierent bimetalliccombinations, respectively [26]. The gure predicts thatthe sensor with single gold layer is the most sensitivewhereas the sensor with single aluminium layer is theleast. Further, Cu-Al combination provides the minimumsensitivity for any ratio of their corresponding thicknessvalues. In all the combinations with gold, Ag-Au and Cu-Au combinations provide good sensitivity for the smallthickness of the inner layer while the Au-Al combinationprovides larger sensitivity than all other combinations forthe larger thickness of inner gold layer. It implies that athick Au layer with very thin cover of Al layer (around 24 nm) provides quite a large sensitivity. As far as variationof SNR with inner layer fraction is concerned, Cu-Al isbetter among all the bimetallic combinations while the Ag-Au combination provides the minimum values of SNR.To achieve highest SNR, one should choose a thin Culayer and a much thicker covering of Al layer. This studyimplies that there is no single combination of metals thatprovides high values of both SNR and sensitivity simultane-ously.

Journal of Sensors 5

Sign

al/n

oise

rati

o

0.70.80.9

11.11.21.31.41.51.61.71.81.9

2

Percentage of silver thickness

0 20 40 60 80 100

Selected rayAll guided rays

(a)

Sen

siti

vity

(mm

/RIU

)

5

5.2

5.4

5.6

5.8

6

6.2

6.4

Percentage of silver thickness

0 20 40 60 80 100

Selected rayAll guided rays

(b)

Figure 5: Variation of (a) signal-to-noise ratio and (b) sensitivity with percentage of silver in bimetallic layer for two dierent kinds of lightlaunching [19]. c Elsevier.

Ag-Au

Ag-Al

Cu-Au

Cu-Al

Au-Al

Sen

siti

vity

(m

/RIU

)

1.9

2

2.1

2.2

2.3

2.4

2.5

2.6

2.7

2.8

2.9

3

Inner layer thickness/total bimetallic thickness

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

(a)

Ag-Au

Ag-Al

Cu-Au

Cu-Al

Au-AlSN

R

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

1.1

Inner layer thickness/total bimetallic thickness

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

(b)

Figure 6: Variation of (a) sensitivity and (b) SNR with inner layer fraction for dierent bimetallic combinations [26]. c AIP.

4.3. Eect of Dopants. It may be noted from (4) that the SPRcondition depends upon the refractive index of the materialof the ber core. Therefore, if an optical ber is fabricated byadding dopants in the ber core the sensitivity of the sensorcan be enhanced or tuned. Generally, an optical ber withpure silica core is used for SPR-based sensors. Sharma et al.[27] carried out theoretical modeling and analysis of SPR-based ber optic sensor to evaluate the eect of dopants onthe sensitivity and SNR. Germanium oxide (GeO2), boronoxide (B2O3), and phosphorus pent-oxide (P2O5) were usedas dopants for pure silica. Figure 7 shows the variation ofsensitivity with the refractive index of the sensing layer fordierent dopants with concentrations [27]. The simulationpredicts an increase of about 50% in sensitivity betweenB2O3 (5.2) and GeO2 (19.3) dopants. Moreover, as thedoping concentration of GeO2 is increased from a low of6.3 mole % to a high of 19.3 mole %, a noticeable decrease in

sensors sensitivity is obtained. Further, the eect of dopantson the sensitivity of the ber optic SPR sensor was reportedto be same irrespective of whether a single metal layer or abimetallic conguration is used.

4.4. Tapered Probe. Several groups have worked on theimprovement of the sensitivity of a ber optic SPR sensorby changing the shape or the geometry of the ber opticSPR probe. Tapering the ber optic SPR probe was one ofthe modications reported in literature [18, 28]. A typicaltapered ber optic SPR probe is shown in Figure 8 [29]. Theuses of dual-tapered and tetra-tapered ber optic SPR probesfor gas and liquid sensing have also been reported [18].Changing the prole of the tapered SPR probe also aects thesensitivity of the sensor. Surface plasmon resonance-basedtapered ber optic sensor with three dierent taper proles,

6 Journal of SensorsSe

nsi

tivi

ty(

m/R

IU)

1.8

2.2

2.6

3

3.4

3.8

4.2

4.6

5

Sensing layer refractive index (ns)

1.333 1.338 1.343 1.348 1.353 1.358 1.363 1.368

B2O3 (5.2)

Pure SiO2

P2O5 (10.5)

GeO2 (6.3)

GeO2 (19.3)

Figure 7: Variation of sensitivity with sensing layer refractive index.A: Silica doped with GeO2 (19.3); B: Silica doped with GeO2 (6.3);C: Silica doped with P2O5 (10.5); D: Pure silica with no doping; andE: Silica doped with B2O3 (5.2). Numbers in brackets are the molarconcentrations of dopant in mole % [27]. c Elsevier.

Gold layer

Sensing region

Cladding

Core2i

20

L

Figure 8: A typical SPR-based ber optic sensor with tapered probe[29]. c Elsevier.

namely, linear, parabolic, and exponential-linear, shown inFigure 9, was analyzed theoretically [29]. Figure 10 shows thevariation of the sensitivity of the tapered ber optic SPRprobe with taper ratio for these three taper proles [29].Theoretical analysis predicts an increase in the sensitivitywith the increase in the taper ratio. The study further showsthat, for a given taper ratio, the exponential-linear taperprole provides the maximum sensitivity. The increase insensitivity occurs because of the decrease in the angle ofincidence of the guided rays with the normal to the core-cladding interface in the tapered region.

To further enhance the sensitivity, an SPR probe ofuniform core (with metallic coating) sandwiched betweentwo unclad tapered ber regions, shown in Figure 11, wasproposed [30]. The taper region 1 brings down the anglesof the bound rays in the ber close to the critical angle ofthe unclad tapered region while the taper region 2 reconvertsthe angles of these rays to their initial values so that all theguided rays can propagate up to the output end of the ber.This was achieved by choosing the minimum allowed value

Exponential-linear

Linear

Parabolic

(z)

0

i

Z

L

Figure 9: Three dierent taper proles showing the linear,parabolic, and exponential-linear [29]. c Elsevier.

Sen

siti

vity

(m

/RIU

)

2.8

3

3.2

3.4

3.6

3.8

4

4.2

4.4

4.6

4.8

Taper ratio (i/0)

1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3

Exponential-linearParabolicLinear

Figure 10: Variation of sensitivity with taper ratio for threedierent taper proles [29]. c Elsevier.

of the radius of the uniform core in the sensing region. In thesensing region rays propagate close to the critical angle of theregion. Figure 12 shows the variation of sensitivity with taperratio for this kind of probe [30]. The sensitivity increaseswith an increase in the taper ratio as reported in otherstudies [29]. However, a signicant sensitivity enhancementof more than 5 times was obtained for a taper ratio of 2.0 incomparison to conventional (TR = 1) ber optic SPR sensor.

4.5. U-Shaped. The angle of incidence of the ray with thenormal to the core-cladding interface can be brought closeto the critical angle by using a U-shaped probe. An SPR-based ber optic sensor with uniform semimetal coated U-shaped probe, shown in Figure 13, was analyzed using a bi-dimensional model [31]. To make the analysis simpler, all

Journal of Sensors 7

Sensing region

Taperedregion 2

Taperedregion 1

Au layer

SPW

Core

Cladding

EW2i

i

a

20

z = 0 z = L z = 2L z = 3LFigure 11: A novel SPR-based ber optic sensor. Sensing probe of uniform core radius is sandwiched between two tapered ber regions[30]. c IEEE.

Sen

siti

vity

(m

/RIU

)

2.75

4

5.25

6.5

7.75

9

10.25

11.5

12.75

14

15.25

Taper ratio (i/0)

1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2

Figure 12: Variation of sensitivity of the SPR-based ber opticsensor with taper ratio [30]. c IEEE.

the guided rays of the p-polarized light launched in theber and their electric vectors were assumed to lie in theplane of bending of the U-shaped probe. Figure 14 showsthe variation of the sensitivity of the probe with bendingradius for dierent values of the sensing length of the probe.Increase in sensitivity with the decrease in the bending radiuswas obtained. The increase in sensitivity was up to a certainvalue of the bending radius below that it starts decreasingsharply. The decrease in sensitivity occurs because the angleof incidence of the last ray becomes less than the critical anglerequired for a ray to be guided in the bent region. Thus,there exists an optimum value of the bending radius at whichthe sensitivity of the SPR sensor based on U-shaped probeacquires a maximum value. The trend of variation was samefor all the three values of the sensing length used. This is dueto the independence of the angle of incidence on the sensinglength. For a given bending radius the sensitivity increasesas the sensing length increases. The increase in sensitivity,as mentioned earlier, is due to the decrease in the angle ofincidence. The decrease in angle of incidence increases the

Sensingmedium

Gold-layer

Cladding

Core2

h

R0

Figure 13: A typical U-shaped ber optic SPR probe [31]. c IOP.

number of reections which causes the broadening of SPRcurve and hence the decrease in the detection accuracy orSNR of the sensor. The enhancement in sensitivity obtainedwas much more compared to the decrease in the detectionaccuracy and hence the decrease in detection accuracy canbe tolerated. In fact the maximum sensitivity achieved wasseveral times more than that reported for an SPR-based beroptic tapered probe.

4.6. Side-Polished Fiber. The SPR probes reported above usemultimode optical bers. A single mode ber has also beenused to fabricate ber optic SPR probe. Surface plasmonresonance sensor using side-polished single mode opticalber and a thin metal over layer is shown in Figure 15[32]. The design of the probe is slightly dierent from thatshown in Figure 4. In this conguration, the guided modepropagating in the ber excites the surface plasmon waveat the interface between the metal and a sensing medium.The resonance occurs if the two modes are closely phasematched. Such single-mode optical ber based SPR sensoris more sensitive and more accurate in comparison to thosewith multi-mode optical bers. However, their fabrication ismuch more complex and sophisticated compared with thosethat use multi-mode bers. The advantage of side-polishedhalf block SPR sensor is that it requires a very small amountof sample for measuring the refractive index.

8 Journal of Sensors

Sen

siti

vity

(m

/RIU

)

0

510

1520

2530

3540

4550

5560

6570

75

Bending radius (cm)

0.9 1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2

L = 1 cmL = 2 cmL = 3 cm

Figure 14: Variation of sensitivity with bending radius for threedierent values of the sensing length [31]. c IOP.

Gold thin film

Core

Cladding

Silica block

of a fiber

Figure 15: Side-polished single mode ber optic SPR probe [23].c IEEE.

Recently SPR-based side-polished multimode opticalber sensors have been reported [33, 34]. In these sensorsthe ber was side-polished until half the core was closed. Thisincreased the sensing area which is an advantage. Apart fromside-polished single mode ber, D-type single mode opticalbers have also been used for sensing applications utilizingSPR technique [35, 36]. These bers also improve sensitivity.The other designs for SPR-based ber optic sensor includesSPR probe at one of the ends of the ber with the reectingend face [37, 38] and a ber tip [39, 40]. The photonic band-gap ber based SPR sensors have also been reported veryrecently [41, 42].

4.7. Eect of Skew Rays. Most of the theoretical studieson ber optic SPR sensors using ray optics and reportedin literature do not consider skew rays in the analysis.These studies consider the propagation of only meridionalrays in the ber that makes the analysis simpler. However,apart from meridional rays, skew rays also exist in theber depending on the light launching conditions. Theserays follow a helical path inside the ber. To specify thetrajectory of a skew ray, a second angle (known as skewnessangle), in addition to the inclination of the ray with the

axial direction of the ber, is required. Recently, the eectof skew rays on the sensitivity and the SNR of a beroptic SPR sensor were studied using spectral interrogationmethod [43]. Figures 16(a) and 16(b) show the variation ofsensitivity and SNR with skewness parameter, respectively[43]. Both the sensitivity and the SNR decrease as the valueof skewness parameter increases irrespective of the metalused for coating. The sensitivity is better in the case ofgold whereas silver demonstrates better SNR as expected.The decrease in sensitivity for highest value of skewnessparameter is about 6% in the case of both the metals. Theeect of skew rays is more on SNR than on the sensitivity. Inthe case of gold lm it is about 40% while in the case of silverit is around 30%.

5. Some Applications

The surface plasmon resonance-based ber optic sensorshave large number of applications for quantitative detectionof chemical and biological species. These include foodquality, medical diagnostics and environmental monitoring.These are detected by changing the refractive index ofthe medium around the metallic coating. The measurandchanges the refractive index of the medium directly orindirectly. Here we present few SPR-based ber optic sensors.

5.1. Temperature. Since the refractive index of a mediumdepends on its temperature, the SPR technique can beapplied to sense the temperature of a medium. The tem-perature sensor based on surface plasmon resonance wasproposed using a coupling prism and angular interrogationmode of operation [44]. It was suggested that for SPRtechnique to be used in temperature sensing, the sensinglayer of large thermo-optic coecient (such as titaniumdioxide or silicon acrylate) should be used. Further, thepenetration of surface plasmon wave should be restrictedonly to metal and sensing layers by taking an appropriatethickness of the two layers. The analysis was later extendedto SPR-based ber optic remote sensor for temperaturedetection [45]. The sensing layer was assumed to be ofTiO2 (titanium dioxide). The outermost ambient mediumwas chosen as air, which adds to the exibility of thesensors design. Figure 17 shows the variation in resonancewavelength with temperature for silver and gold [45]. Forboth the metals, resonance wavelength shifts to shorter sidewith the increase in the temperature. However, there is noappreciable dierence in resonance wavelength (and hence,in temperature sensitivity) for two metals used. This wassuggested to be due to the similarity in the variation ofphysical properties of both the metals with temperature.Figure 18 depicts the corresponding variation in SPR curvewidth with temperature [45]. The SPR curve width increaseswith the increase in the temperature. This was reportedto happen because the imaginary (absorption) part of themetal dielectric function increases with temperature. Theeect of other parameters such as numerical aperture andthe ratio of sensing region length to ber core diameter onthe sensitivity and detection accuracy were also reported.

Journal of Sensors 9

Silver

Gold

Sen

siti

vity

2.5

2.6

2.7

2.8

2.9

3

3.1

3.2

Skewness parameter

0 0.2 0.4 0.6 0.8 1

(a)

Silver

Gold

SNR

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Skewness parameter

0 0.2 0.4 0.6 0.8 1

(b)

Figure 16: Variation of (a) Sensitivity and (b) SNR of an SPR-based ber optic sensor with skewness parameter for two dierent metals[43]. c OSA.

Res

onan

cew

avel

engt

h(n

m)

557

558

559

560

561

562

563

564

565

566

Temperature (K)300 330 360 390 420 450 480 510 540 570 600

SilverGold

Figure 17: Variation of resonance wavelength with temperature fortwo metals: silver and gold [45]. c Elsevier.

It was also reported that the ber with smaller numericalaperture provides larger detection accuracy without aectingthe temperature sensitivity. Similarly, the small values of theratio of sensing region length to ber core diameter give highdetection accuracy. For SPR-based ber optic temperaturesensor smaller sensing region and highly multimode opticalber was recommended.

5.2. Naringin. The processing of citrus fruit juice has facedformidable problems in terms of bitterness, thereby aectingits consumer acceptability. The bitterness is caused by exces-sive naringin contents in fruit juice. The presence of naringinin fruit juice can be detected using an SPR-based ber opticsensor. An SPR-based ber optic sensor relying on spectralinterrogation method was reported for the detection of

SPR

curv

ew

idth

(nm

)

56

57

58

59

60

61

62

63

Temperature (K)300 330 360 390 420 450 480 510 540 570 600

SilverGold

Figure 18: Variation of SPR curve width with temperature for twometals: silver and gold [45]. c Elsevier.

naringin [21]. The SPR probe was prepared by immobilizingthe enzyme naringinase on the silver-coated core of theoptical ber. To immobilize, gel entrapment technique wasused. The experimental set up for the characterization ofthe SPR-based ber optic sensor is shown in Figure 19 [21].Light from a broadband source (tungsten halogen lamp)was focused on the input face of the ber using a circularslit and a microscope objective. The probe was mountedin a ow cell to keep the sample around the probe. Thespectral distribution of the transmitted power for the samplearound the probe was determined using a monochromator,a silicon detector and a power meter. The spectral distri-bution of transmitted light so obtained for a sample wasdivided by the corresponding spectral distribution obtainedwithout any sample around the probe. The resultant SPRspectrum so obtained does not depend on the source spectral

10 Journal of Sensors

Enzymes immobilized fiber

Si detector

Monochromator

Flow cell

LampMicroscope

objective

Figure 19: Experimental set up of the SPR-based ber optic sensor [21]. c Elsevier.

SPR

wav

elen

gth

(nm

)

500

520

540

560

580

600

620

640

Concentration of naringin (gm/100 mL)

0 0.05 0.1 0.15 0.2

Figure 20: Variation of resonance wavelength with the concentra-tion of naringin in buer solution [21]. c Elsevier.

output, spectral sensitivity of photodetector and the spectralabsorbance of the ber. From the SPR spectrum resonancewavelength was determined. The calibration curve of theSPR-based ber optic sensor obtained for the detection ofnaringin is shown in Figure 20 [21]. As the concentration ofnaringin increases the resonance wavelength increases. Theincrease in resonance wavelength was reported to be dueto an increase in the refractive index of the immobilizedlayer which may occur due to the formation of a complexof naringin with the enzyme naringinase in the lm.

5.3. Pesticides. Surface plasmon resonance-based ber opticsensor for the detection of organophosphate pesticide, chlor-phyrifos, was also reported recently using similar experimen-tal method as used for the detection of naringin [46]. Theprobe was prepared by immobilizing acetylcholinesteraseenzyme on the silver-coated core of the ber. The principleof detection of pesticide was slightly dierent from thatof naringin. It was based on the principle of competitivebinding of the pesticide (acting as an inhibitor) for thesubstrate (acetlythiocholine iodide) to the enzyme. Figure 21shows SPR spectra for pesticide concentration of 1.0 Mand a substrate concentration of 2.5 mM for the two cases

A BNor

mal

ised

outp

ut

pow

er

0.6

0.65

0.7

0.75

0.8

0.85

0.9

0.95

1

SPR wavelength (nm)

450 550 650 750 850

Figure 21: Variation of normalized output power as a functionof wavelength for a substrate concentration (2.5 mM) and 1.0 Mconcentration of pesticide (A) without enzyme and (B) withenzyme (B) [46]. c Elsevier.

(A) without enzyme (control experiment) and (B) withenzyme in the lm [46]. The dip in control experimentoccurs due to the refractive index of the gel. The dierencein two spectra implies that the enzyme is playing a role inchanging the refractive index of the gel in the presence ofpesticide. The resonance wavelength was obtained around590 nm for 1.0 M pesticide concentration. Figure 22 showsthe calibration curve of the sensor [46]. The trend is oppositeto that reported for the detection of naringin. In this case,the SPR wavelength decreases with the increase in theconcentration of the pesticide for the xed concentration ofthe substrate. The results appear to be due to the decreasein the refractive index of the lm as the concentration ofpesticide increases.

Surface plasmon resonance-based ber optic sensorshave also been reported for other chemical species. Theseinclude measurements of salinity in water [22, 37], refractiveindices of alcohols [17], BSA [34], vapor and liquid analyses[18]. In addition a single-mode waveguide surface plasmonresonance sensor has been developed for biomolecular inter-action analysis [47]. Recently, ber optic SPR sensors havebeen reported that uses either a series of long period grating[48] or Bragg grating [49]. The advantage of these designs isthat these oer multiple sensing channels capability.

Journal of Sensors 11SP

Rw

avel

engt

h(n

m)

550

560

570

580

590

600

610

620

630

Log (concentartion of pesticide) (moles/liter)

8 7 6 5 4 3 2

Figure 22: Calibration curve of SPR-based ber optic pesticidesensor [46]. c Elsevier.

6. Summary

In this review article we have rst described the principleof surface plasmon resonance technique. The technique iswell established and will remain unchanged for years. Inthe beginning, the technique was used for prism-based SPRsensors but later when the advantages of optical ber wererealized, it was applied to optical ber based sensors. Recentlythere have been signicant advancements in both the designof ber optic SPR sensors and strategies for the enhancementof the performance of the sensors particularly sensitivity. Theprobe designs include tapered, U-shaped, and side-polishedbers. These developments are likely to drive future trends inthe research and development of optical ber sensors. Theunique optical properties of metal nanoparticles have alsoattracted the sensor community to develop localized surfaceplasmon resonance- (LSPR-) based sensors. The LSPR-basedsensors have few advantages over bulk SPR-based sensors.However, the pace of collaboration of LSPR with optical berfor sensing is at present slow but may gain momentum incoming years.

Acknowledgment

The present work is partially supported by the Departmentof Science and Technology (India).

References

[1] B. Liedberg, C. Nylander, and I. Lunstrom, Surface plasmonresonance for gas detection and biosensing, Sensors andActuators, vol. 4, pp. 299304, 1983.

[2] J. van Gent, P. V. Lambeck, H. J. M. Kreuwel, et al.,Optimization of a chemooptical surface plasmon resonancebased sensor, Applied Optics, vol. 29, pp. 28432849, 1990.

[3] E. Stenberg, B. Persson, H. Roos, and C. Urbaniczky, Quan-titative determination of surface concentration of proteinwith surface plasmon resonance using radiolabeled proteins,

Journal of Colloid and Interface Science, vol. 143, no. 2, pp. 513526, 1991.

[4] G. Dougherty, A compact optoelectronic instrument witha disposable sensor based on surface plasmon resonance,Measurement Science and Technology, vol. 4, no. 6, pp. 697699, 1993.

[5] S. Ekgasit, A. Tangcharoenbumrungsuk, F. Yu, A. Baba, andW. Knoll, Resonance shifts in SPR curves of nonabsorbing,weakly absorbing, and strongly absorbing dielectrics, Sensorsand Actuators B, vol. 105, no. 2, pp. 532541, 2005.

[6] J.-J. Chyou, C.-S. Chu, F.-C. Chien, et al., Precise determi-nation of the dielectric constant and thickness of a nanolayerby use of surface plasmon resonance sensing and multiexper-iment linear data analysis, Applied Optics, vol. 45, no. 23, pp.60386044, 2006.

[7] H. P. Chiang, C.-W. Chen, J. J. Wu, et al., Eects oftemperature on the surface plasmon resonance at a metal-semiconductor interface, Thin Solid Films, vol. 515, no. 17,pp. 69536961, 2007.

[8] J. Le Person, F. Colas, C. Compere, et al., Surface plasmonresonance in chalcogenide glass-based optical system, Sensorsand Actuators B, vol. 130, no. 2, pp. 771776, 2008.

[9] W. Feng, L. Shenye, P. Xiaoshi, C. Zhuangqi, and D. Yongkun,Reective-type conguration for monitoring the photo-bleaching procedure based on surface plasmon resonance,Journal of Optics A, vol. 10, no. 9, Article ID 095102, 2008.

[10] R. C. Jorgenson and S. S. Yee, A ber-optic chemical sensorbased on surface plasmon resonance, Sensors and Actuators B,vol. 12, no. 3, pp. 213220, 1993.

[11] R. D. Harris and J. S. Wilkinson, Waveguide surface plasmonresonance sensors, Sensors and Actuators B, vol. 29, no. 13,pp. 261267, 1995.

[12] W. B. Lin, N. Jarezic-Renault, A. Gagnaire, and H. Gagnaire,The eects of polarization of the incident light-modeling andanalysis of a SPR multimode optical ber sensor, Sensors andActuators A, vol. 84, no. 3, pp. 198204, 2000.

[13] R. Slavk, J. Homola, J. Ctyroky, and E. Brynda, Novel spec-tral ber optic sensor based on surface plasmon resonance,Sensors and Actuators B, vol. 74, no. 13, pp. 106111, 2001.

[14] M. Piliarik, J. Homola, Z. Mankova, and J. Ctyroky, Sur-face plasmon resonance sensor based on a single-modepolarization-maintaining optical ber, Sensors and ActuatorsB, vol. 90, no. 13, pp. 236242, 2003.

[15] D. J. Gentleman, L. A. Obando, J.-F. Masson, J. R. Holloway,and K. S. Booksh, Calibration of ber optic based surfaceplasmon resonance sensors in aqueous systems, AnalyticaChimica Acta, vol. 515, no. 2, pp. 291302, 2004.

[16] A. K. Sharma and B. D. Gupta, Absorption-based ber opticsurface plasmon resonance sensor: a theoretical evaluation,Sensors and Actuators B, vol. 100, no. 3, pp. 423431, 2004.

[17] M. Mitsushio, S. Higashi, and M. Higo, Construction andevaluation of a gold-deposited optical ber sensor system formeasurements of refractive indices of alcohols, Sensors andActuators A, vol. 111, no. 2-3, pp. 252259, 2004.

[18] Y.-C. Kim, W. Peng, S. Banerji, and K. S. Booksh, Taperedber optic surface plasmon resonance sensor for analyses ofvapor and liquid phases, Optics Letters, vol. 30, no. 17, pp.22182220, 2005.

[19] A. K. Sharma and B. D. Gupta, On the sensitivity andsignal to noise ratio of a step-index ber optic surfaceplasmon resonance sensor with bimetallic layers, OpticsCommunications, vol. 245, no. 16, pp. 159169, 2005.

12 Journal of Sensors

[20] B. D. Gupta and A. K. Sharma, Sensitivity evaluation ofa multi-layered surface plasmon resonance-based ber opticsensor: a theoretical study, Sensors and Actuators B, vol. 107,no. 1, pp. 4046, 2005.

[21] Rajan, S. Chand, and B. D. Gupta, Fabrication and characteri-zation of a surface plasmon resonance based ber-optic sensorfor bittering componentNaringin, Sensors and Actuators B,vol. 115, no. 1, pp. 344348, 2006.

[22] N. Diaz-Herrera, O. Esteban, M. C. Navarrete, M. Le Haitre,and A. Gonzalez-Cano, In situ salinity measurements inseawater with a bre-optic probe, Measurement Science andTechnology, vol. 17, no. 8, pp. 22272232, 2006.

[23] A. K. Sharma, R. Jha, and B. D. Gupta, Fiber-optic sensorsbased on surface plasmon resonance: a comprehensive review,IEEE Sensors Journal, vol. 7, pp. 11181129, 2007.

[24] E. Kretchmann and H. Reather, Radiative decay of non-radiative surface plasmons excited by light, Naturforsch, vol.23, pp. 21352136, 1968.

[25] S. A. Zynio, A. V. Samoylov, E. R. Surovtseva, V. M. Mirsky,and Y. M. Shirshov, Bimetallic layers increase sensitivity ofanity sensors based on surface plasmon resonance, Sensors,vol. 2, no. 2, pp. 6270, 2002.

[26] A. K. Sharma and B. D. Gupta, On the performance of dif-ferent bimetallic combinations in surface plasmon resonancebased ber optic sensors, Journal of Applied Physics, vol. 101,no. 9, Article ID 093111, 6 pages, 2007.

[27] A. K. Sharma, Rajan, and B. D. Gupta, Inuence of dopantson the performance of a ber optic surface plasmon resonancesensor, Optics Communications, vol. 274, no. 2, pp. 320326,2007.

[28] B. Grunwald and G. Holst, Fibre optic refractive indexmicrosensor based on white-light SPR excitation, Sensors andActuators A, vol. 113, no. 2, pp. 174180, 2004.

[29] R. K. Verma, A. K. Sharma, and B. D. Gupta, Surface plasmonresonance based tapered ber optic sensor with dierent taperproles, Optics Communications, vol. 281, no. 6, pp. 14861491, 2008.

[30] R. K. Verma, A. K. Sharma, and B. D. Gupta, Modeling oftapered ber-optic surface plasmon resonance sensor withenhanced sensitivity, IEEE Photonics Technology Letters, vol.19, no. 22, pp. 17861788, 2007.

[31] R. K. Verma and B. D. Gupta, Theoretical modeling of a bi-dimensional U-shaped surface plasmon resonance based breoptic sensor for sensitivity enhancement, Journal of Physics D,vol. 41, Article ID 095106, 2008.

[32] J. Homola and R. Slavik, Fibre-optic sensor based on surfaceplasmon resonance, Electronics Letters, vol. 32, no. 5, pp. 480482, 1996.

[33] H.-Y. Lin, W.-H. Tsai, Y.-C. Tsao, and B.-C. Sheu, Side-polished multimode ber biosensor based on surface plasmonresonance with halogen light, Applied Optics, vol. 46, no. 5,pp. 800806, 2007.

[34] Y.-C. Lin, Y.-C. Tsao, W.-H. Tsai, T.-S. Hung, K.-S. Chen,and S.-C. Liao, The enhancement method of optical berbiosensor based on surface plasmon resonance with coldplasma modication, Sensors and Actuators B, vol. 133, no.2, pp. 370373, 2008.

[35] M.-H. Chiu, C.-H. Shih, and M.-H. Chi, Optimum sensitivityof single-mode D-type optical ber sensor in the intensitymeasurement, Sensors and Actuators B, vol. 123, no. 2, pp.11201124, 2007.

[36] M.-H. Chiu and C.-H. Shih, Searching for optimal sensitivityof single-mode D-type optical ber sensor in the phasemeasurement, Sensors and Actuators B, vol. 131, no. 2, pp.596601, 2008.

[37] D. J. Gentleman and K. S. Booksh, Determining salinityusing a multimode ber optic surface plasmon resonance dip-probe, Talanta, vol. 68, no. 3, pp. 504515, 2006.

[38] H. Suzuki, M. Sugimoto, Y. Matsui, and J. Kondoh, Eectsof gold lm thickness on spectrum prole and sensitivity of amultimode-optical-ber SPR sensor, Sensors and Actuators B,vol. 132, no. 1, pp. 2633, 2008.

[39] K. Kurihara, H. Ohkawa, Y. Iwasaki, O. Niwa, T. Tobita,and K. Suzuki, Fiber-optic conical microsensors for surfaceplasmon resonance using chemically etched single-modeber, Analytica Chimica Acta, vol. 523, no. 2, pp. 165170,2004.

[40] T. Abrahamyan and Kh. Nerkararyan, Surface plasmonresonance on vicinity of gold-coated ber tip, Physics LettersA, vol. 364, no. 6, pp. 494496, 2007.

[41] A. Hassani and M. Skorobogatiy, Design criteria formicrostructured-optical-ber-based surface-plasmon-resonance sensors, Journal of the Optical Society of America B,vol. 24, no. 6, pp. 14231429, 2007.

[42] B. Gauvreau, A. Hassani, M. F. Fehri, A. Kabashin, and M.Skorobogatiy, Photonic bandgap ber-based surface plasmonresonance sensors, Optics Express, vol. 15, no. 18, pp. 1141311426, 2007.

[43] Y. S. Dwivedi, A. K. Sharma, and B. D. Gupta, Inuence ofskew rays on the sensitivity and signal-to-noise ratio of a ber-optic surface-plasmon-resonance sensor: a theoretical study,Applied Optics, vol. 46, no. 21, pp. 45634569, 2007.

[44] S. K. Ozdemir and G. Turhan-Sayan, Temperature eectson surface plasmon resonance: design considerations for anoptical temperature sensor, Journal of Lightwave Technology,vol. 21, no. 3, pp. 805814, 2003.

[45] A. K. Sharma and B. D. Gupta, Theoretical model of a beroptic remote sensor based on surface plasmon resonance fortemperature detection, Optical Fiber Technology, vol. 12, no.1, pp. 87100, 2006.

[46] Rajan, S. Chand, and B. D. Gupta, Surface plasmon resonancebased ber-optic sensor for the detection of pesticide, Sensorsand Actuators B, vol. 123, no. 2, pp. 661666, 2007.

[47] T. Matsushita, T. Nishikawa, H. Yamashita, J. Kishimoto,and Y. Okuno, Development of new single-mode waveguidesurface plasmon resonance sensor using a polymer imprintprocess for high-throughput fabrication and improved designexibility, Sensors and Actuators B, vol. 129, no. 2, pp. 881887, 2008.

[48] Y.-J. He, Y.-L. Lo, and J.-F. Huang, Optical-ber surface-plasmon-resonance sensor employing long-period ber grat-ings in multiplexing, Journal of the Optical Society of AmericaB, vol. 23, no. 5, pp. 801811, 2006.

[49] B. Spackova, M. Piliarik, P. Kvasnicka, C. Themistos, M.Rajarajan, and J. Homola, Novel concept of multi-channelber optic surface plasmon resonance sensor, Sensors andActuators B, vol. 139, no. 1, pp. 199203, 2009.

Hindawi Publishing CorporationJournal of SensorsVolume 2009, Article ID 524237, 20 pagesdoi:10.1155/2009/524237

Review Article

Microstructured and Photonic Bandgap Fibers for Applicationsin the Resonant Bio- and Chemical Sensors

Maksim Skorobogatiy

Departement de Genie Physique, Ecole Polytechnique de Montreal, C. P. 6079, succ. Centre-ville, Montreal, QC, Canada H3C 3A7

Correspondence should be addressed to Maksim Skorobogatiy, maksim.skorobogatiy@polymtl.ca

Received 3 March 2009; Accepted 2 June 2009

Recommended by Christos Riziotis

We review application of microstructured and photonic bandgap bers for designing resonant optical sensors of changes in thevalue of analyte refractive index. This research subject has recently invoked much attention due to development of novel bertypes, as well as due to development of techniques for the activation of ber microstructure with functional materials. Particularly,we consider two sensors types. The rst sensor type employs hollow core photonic bandgap bers where core guided mode isconned in the analyte lled core through resonant eect in the surrounding periodic reector. The second sensor type employsmetalized microstructured or photonic bandgap waveguides and bers, where core guided mode is phase matched with a plasmonpropagating at the ber/analyte interface. In resonant sensors one typically employs bers with strongly nonuniform spectraltransmission characteristics that are sensitive to changes in the real part of the analyte refractive index. Moreover, if narrowabsorption lines are present in the analyte transmission spectrum, due to Kramers-Kronig relation this will also result in strongvariation in the real part of the refractive index in the vicinity of an absorption line. Therefore, resonant sensors allow detection ofminute changes both in the real part of the analyte refractive index (106104 RIU), as well as in the imaginary part of the analyterefractive index in the vicinity of absorption lines. In the following we detail various resonant sensor implementations, modesof operation, as well as analysis of sensitivities for some of the common transduction mechanisms for bio- and chemical sensingapplications. Sensor designs considered in this review span spectral operation regions from the visible to terahertz.

Copyright 2009 Maksim Skorobogatiy. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

1. Introduction

R&D into ber-optic bio- and chemosensors (FOSs) hasmade a lot of progress during the last ten years. This isdue to the appealing properties of FOS such as immunityto electromagnetic interference, safety in explosive environ-ments, and potential to provide continuous quantitative andqualitative real-time analysis. Chemically sensitive thin lmsdeposited on selected areas of optical bers can inuencethe propagation of light in such bers depending on thepresence or absence of chemical molecules in the surround-ing environment [1]. A wide range of optical sensors hasbeen developed for selective biomolecule detection. Mostof them have reliability issues as they employ very fragileantibodies as sensing elements. These sensors include highrefractive index waveguides [2], surface plasmon resonancesensors [3], resonant mirrors [4], and classical ber-optical

sensors [5, 6]. Most optical sensors are based on evanescentwave sensing, where the perturbations in the refractive indexclose to the sensor surface are probed by the exponentiallydecaying optical wave. Such sensors have proven to be highlysensitive in detection of small targets such as proteins andviruses, but they experience diculties in detecting largertargets such as bacteria (0.55 m) since in that case muchlarger penetration of the evanescent eld into analyte isrequired [7].

Microstructured Optical Fibers (MOFs), and PhotonicBandgap (PBG) bers which are a subset of MOFs, promisea viable technology for the mass production of highlyintegrated and intelligent sensors in a single manufacturingstep. In standard total internal reection (TIR) ber-basedevanescent-wave sensors the ber polymer jacket is strippedand the ber cladding is polished to the core in order toobtain an overlap between the optical eld and analyte,

2 Journal of Sensors

with sensor sensitivity proportional to such an overlap.Compared to the conventional solid core optical bers,MOFs oer a number of unique advantages in sensingapplications. A dening feature of a microstructured beris the presence of air holes running along its entire length.Fiber optical properties are then determined by the size,shape, and relative position of the holes. Particularly, aunique ability of MOFs is to accommodate biological andchemical samples in gaseous or liquid forms inside of the airholes in the immediate vicinity of the ber core [810]. Inthis context a MOF is used simultaneously as a light guideand as a uidic channel. The MOFs unique architecturemakes it a very promising sensing platform for chemical andbiological detection. First, MOFs naturally integrate opticaldetection with the mirouidics, allowing for continuous on-line monitoring of dangerous samples in real-time withoutexposing the personnel to danger. In addition, the samplescan be transferred in the enclosed MOF optouidic systemfor further conrmation analysis, for example, polymerasechain reactions (PCRs) if needed. Such channels can befurther functionalized with biorecognition layers that canbind and progressively accumulate target biomolecules, thusenhancing sensor sensitivity and specicity. Second, theMOF hole size is in sub-100 m rage, leading to verysmall uid samples required for sensing. Third, MOF-based sensors can be coiled into long sensing cells (10 m),thus dramatically increasing their sensitivity. The same isimpossible to achieve with traditional TIR ber sensors asside polishing step limits sensor length to several cm. Forth,the desired MOFs can be mass-produced using commercialber draw tower in a very cost-eective manner. Fifth, theMOFs can potentially be scaled up into a two-dimensionalarray with small dimensions, which is suitable for makinginto portable point-of-care devices for simultaneous on-site detection of dierent kinds of analytes. Sixth, PhotonicBandgap bers can be designed to guide light directly in theiranalyte-eld hollow cores [11]. In such bers light-analytecoupling is considerably stronger than that in evanescentsensors.

In this paper we review several MOF and PBG ber-based resonant optical sensors, which have recently invokedstrong interest due to development of novel ber types,as well as due to development of techniques for activationof the ber microstructure with functional materials. Inresonant sensors one typically employs bers with stronglynonuniform spectral transmission characteristics that aresensitive to changes in the real part of the analyte refractiveindex. Moreover, if narrow absorption lines are presentin the analyte transmission spectrum, due to Kramers-Kronig relation this will also result in strong variation inthe real part of the refractive index in the vicinity of anabsorption line. Therefore, resonant sensors allow detectionof minute changes both in the real part of the analyterefractive index, as well as in the imaginary part of theanalyte refractive index in the vicinity of absorption lines.Although the operational principle of almost all ber-basedresonant sensors relies on strong sensitivity of the bertransmission losses to the value of the analyte refractiveindex, particular transduction mechanism for biodetection

can vary. Consider, for example, the case of a hollow corePBG ber featuring an analyte lled core. In one sensorimplementation one can label the target biomolecules withhighly absorbing particles of known absorption spectra, suchas metal nanoparticles or quantum dots. The presence ofsuch particles in the aqueous ber core can then be quan-tied by detecting appearance of the absorption lines in theber transmission spectrum, or through resonant changes inthe ber transmission losses induced by variations in the realpart of the core refractive index. In another implementation,a functional layer that binds specic biomolecules can bedeposited on the inside of the hollow ber core. Biomoleculebinding events to such a layer can then be detectedthrough resonant changes in the ber transmission lossesinduced by variations in the real part of the layer refractiveindex.

In what follows we discuss two types of resonant sensors.One such sensor type relies on changes in the radiation lossesof a leaky core mode due to changes in the real part ofan analyte refractive index. Such a leaky mode is typicallyconned inside an analyte lled ber core by a resonantreector cladding. The term leaky mode generally refersto the guidance mechanism where the eective refractiveindex of a propagating mode is smaller than that of theber cladding. Such unusual modes are called leaky modesas, outside of a waveguide core, they do not exhibit atraditional evanescent decay into the cladding, but ratherthey radiate slowly (leak) into the cladding. Unlike in thecase of common TIR bers, leaky modes in PBG bers areconned by the bandgap of a microstructured reector. Fora particular value of an analyte refractive index geometry ofsuch a ber is chosen to provide strong optical connementof the leaky core mode. An example of a resonant sensordescribed above is a photonic bandgap ber featuring ahollow core lled with analyte. When changing the real partof an analyte refractive index, resonant condition for modeconnement will change, resulting in strong changes in themodal radiation loss (see Figure 1(a)). Detection of changesin the transmitted intensities can be then reinterpretedin terms of the changes in the real part of an analyterefractive index. Interestingly, the same sensor can also beused in a standard nonresonant interrogation mode for thedetection of changes in the imaginary part of the analyterefractive index (analyte absorption), see Figure 1(b). Evenwhen operated in a nonresonant regime, sensitivity of thehollow core PBG ber-based sensors is, generally, superiorto that of traditional TIR ber-based sensors due to greatlyimproved modal overlap with analyte.

Second sensor type considered in this review is operatedin the vicinity of a phase matching wavelength betweena Gaussian-like core-guided-mode and some other (high-order) mode that shows high sensitivity of its propagationproperties to changes in the real part of the analyte refractiveindex Figure 1(b). For example, by activating ber surfacewith a thin metal layer, at a specic resonant wavelengthone can induce strong optical loss of a core-guided-modedue to coupling to an absorbing plasmon mode propagatingat the metal/analyte interface. As plasmon mode is largelydelocalized in the analyte region, wavelength of phase

Journal of Sensors 3

Core mode

Waveguidecore filled

with analyte

CladClad

Mod

allo

ss na

na

Re(n)

na + Re(n)

na

Mod

allo

ss

na

na + iIm(n)

(a)

Cor

e m

ode

Plas

mon

Phasematching

point

Plasmon Core mode

Core

Phasematching

point

Clad

naMod

allo

ss

An

alyt

ena

n a+

Re(n)

Metal

(b)

Figure 1: Operational principles and schematics of the two types of the resonant optical sensors. (a) Analyte-lled hollow photonic bandgapber-based sensor. Transmission loss through such a sensor is very sensitive to the values of both the real and imaginary parts of the analyterefractive index. (b) Sensor operating near the phase matching point of a core-guided-mode and a second mode featuring large overlap withthe analyte region. In the case of a phase matching with a plasmon mode, propagation loss of a core-guided-mode is strongly dependent onthe real part of the analyte refractive index.

matching between the two modes will be very sensitive to thevalue of the real part of analyte refractive index.

2. Detection Strategies forAbsorption-Based Sensors

We now remind the reader some general facts aboutamplitude-based and spectral-based detection methodolo-gies. Particularly, we focus on ber-based sensors that relyon detection of changes in the transmitted light intensity inthe presence of a target analyte.

In the amplitude-based detection methodology oneoperates at a xed wavelength and records changes in theamplitude of a signal, which are then reinterpreted in termsof changes in the analyte refractive index. To characterizesensitivity of a ber-based sensor of length L, one employsan amplitude sensitivity function Sa(,L), which is denedas a relative change in the intensity P(, ,L) of a transmittedlight for small changes in the measurand . Note that can be any parameter that inuences transmission propertiesof a ber sensor. This includes concentration of absorbingparticles in the analyte, thickness of a biolayer that canchange due to capture of specic biomolecules, as well as realor imaginary parts of the analyte refractive index. Amplitudesensitivity is then dened as

Sa(,L) = lim 0

P(, ,L) P(0, ,L) P(0, ,L) =

P(, ,L) /|=0P(0, ,L)

.

(1)

Denoting, (, ) to be the ber propagation loss at a xedvalue of a measurand, light intensity at the ber output canbe written as:

P(, ,L) = Pin() exp((, )L), (2)where Pin() is the light intensity launched into a ber.Substituting (2) into (1), amplitude sensitivity function canbe then expressed as:

Sa(,L) = (, )|=0 L . (3)

As follows from (3), sensor sensitivity is proportional to thesensor length L. In turn, as follows from (2), the maximalsensor length is limited by the absorption loss of a ber.Dening Pdet() to be the power detection limit at whichchanges in the light intensity can still be detected reliably, themaximal sensor length allowed by the power detection limitcan be calculated from (2) as

L = log(Pin()/Pdet())(0, )

. (4)

Dening det() = log(Pin()/Pdet()), maximal sensitivityallowed by the power detection limit can be written using (3)as

Sa() = det()(, ) /|=0(0, )

. (5)

In all the simulations that follow we assume that det() = 1,which allows us to characterize an inherent sensitivity of asensor system, while separating it from the issue of a power

4 Journal of Sensors

budget that might bring additional sensitivity enhancement.Finally, given sensor amplitude sensitivity, to estimate sensorresolution of a measurand , one can use expression (1).Assuming that the minimal detectable relative change in thesignal amplitude is (P/P)min (which is typically on the orderof 1% if no advanced electronics is used), then the minimumvalue of a measurand that can be detected by such a sensor is

min = (P/P)minSa()

. (6)

Another popular sensing methodology is spectral-based. Ituses detection of displacements of spectral singularities inthe presence of a measurand with respect to their positionsfor a zero level of a measurand. This sensing approach isparticularly eective in the resonant sensor congurationsthat feature sharp transmission or absorption peaks in theirspectra. Dening p() to be the position of a peak in asensor transmission spectrum as a function of a measurandvalue , spectral sensitivity function can be dened as

S = ()|=0 . (7)

Given sensor spectral sensitivity, to estimate sensor resolu-tion of a measurand , one can use expression (7). Thus,assuming that the minimal detectable spectral shift in thepeak position is (p)min (which is typically on the orderof 0.1 nm in the visible spectral range if no advanced opticaldetection is used), then the minimum value of a measurandthat can be detected by such a sensor is

min =(p

)min

S. (8)

3. Sensing Using Analyte-Filled Hollow CorePhotonic Bandgap Fibers

We now describe the rst resonant sensor type basedon hollow core photonic bandgap bers lled with ana-lyte. In their crossection PBG bers can contain periodicsequence of micron-sized layers of dierent materials [11,12] (Figure 2(a)), periodically arranged micron-sized airvoids [1315] (Figure 2(b)), or rings of holes separatedby nanosupports [16, 17] (Figure 2(c)). PBG bers arecurrently available in silica glass, polymer and specialtysoft glass implementations. The key functionality of suchbers is their ability to guide light directly in the hollowor liquid-lled cores with refractive index smaller than therefractive index of a surrounding cladding material. Unlikemicrostructured bers, PBG bers conne light in theirhollow cores by photonic bandgap eect, rather than bytotal internal reection. Practically, bandgaps are denedas frequency regions of enhanced ber transmission, andthey are the result of destructive interference of the core-guided light inside of the ber-microstructured cladding.When launching spectrally broad light into a PBG ber, onlythe spectral components guided by the ber bandgaps willreach the ber end, while all the spectral components not

located within the bandgaps will be irradiated out near theber coupling end. Moreover, even in the absence of bermaterial losses, core-guided-modes always exhibit radiationloss. This is a direct consequence of guidance in a core withrefractive index smaller than that of a cladding. As we willsoon see, core mode radiation loss can be very sensitiveto the value of the real part of the refractive index of thematerial lling the ber core, which can be utilized for sensorapplications. Finally, PBG bers have a tendency to improvethe beam quality of guided light, while being eectively singlemode in the limit of long propagation distances. This is aconsequence of the fact that radiation losses (and, generally,absorption losses too) of the core-guided-modes of a PBGber are strongly dierentiated with only a few low-ordermodes having small propagation losses. Thus, when excitingseveral modes at the ber input end, only the modes havingthe lowest losses will survive till the ber end. For historicalreference we mention that before the invention of the all-dielectric PBG bers, guidance in the hollow core bers hasbeen demonstrated in the context of metal coated capillaries[18, 19].

We now detail some of the advantages oered by thehollow core PBG bers for sensing applications. One hasto distinguish two modes of operation of such sensors.First, is sensing of changes in the imaginary part of theanalyte refractive index (analyte absorption) by detectingthe presence and strength of the narrow absorption bandsin the ber transmission spectrum. This is the simplest,nonresonant application of the hollow PBG bers for opticalsensing in which one only takes advantage of large overlapof the core-guided leaky mode with analyte. In such sensors,signal strength due to analyte absorption, as well as sensorsensitivity are directly proportional to the sensor length.Recently, several experimental implementations of suchabsorption-based sensors have been demonstrated [10, 2023]. Second mode of operation of a PBG ber-based sensoris sensing of changes in the real part of the analyte refractiveindex by detection of shifts in the ber bandgap position.As it will be explained in the following, such a sensoroperates in the resonant regime with sensitivity that is largelyindependent of the sensor length.

3.1. Nonresonant Sensing. Classic perturbation theory con-siderations [24] predicts that changes in the eective refrac-tive index of a guided mode ne are related to the changesin the refractive index na of analyte inltrating the ber,through the overlap factor f dened as

ne = na f = Re(na) f + i Im(na) f

f =

analytedA|E|2Re(z crossectiondAEt Ht) ,

(9)

where Et, Ht are the transverse electromagnetic elds ofa ber mode, while E is a complete electric eld of amo