An HC-PCF Fluorescence Spectrocopy for Detection of Microsphere Samples Based on Refractive Index Scaling Law

doi:10.4236/opj.2011.12014

Paper Menu >>

Journal Menu >>

Optics and Photonics Journal, 2011, 1, 85-90

doi:10.4236/opj.2011.12014 Published Online June 2011 (http://www.SciRP.org/journal/opj/)

An HC-PCF Fluorescence Spectrocopy for Detection of

Microsphere Samples Based on Refractive Index Scaling Law

Vengalathunadakal K. Shinoj1, Vadakke. M. Murukeshan1

1School of Mechanical and Aerospace Engineering, Nanyang Technological University, Singapore.

E-mail: [email protected]

Received April 13, 2011; revised May 15, 2011; accepted May 23, 2011

Abstract

This paper illustrates an efficient fluorescence detection of micro particles using hollow-core photonic crys-

tal fibers (HC-PCFs) by applying the refractive index (RI) scaling law. The variations in the central wave-

length for different filling material indices are illustrated for most commonly available HC-PCFs that have

cladding made of pure fused silica with array of air holes running along the entire length of the fiber. The

proposed concept is verified by immobilizing fluorescent microsphere samples inside two HC-PCFs of dif-

ferent central wavelengths and the quantification of fluorescence inside the fibers is performed through spec-

troscopic analysis. The sensitivity has been compared for similar fiber with different dispersed media and

different fibers with same dispersed medium.

Keywords: Optical Fiber, Hollow-Core Photonic Crystal Fiber, Refractive Index Scaling, Fluorescence Spec-

troscopy

1. Introduction

It is reported that the detection sensitivity of small amount

of biological threats can be enhanced with simple inex-

pensive methods which can give brightest possible fluo-

rescence for detection using high throughput suspension

arrays [1,2]. Moreover, many biomolecules are not avail-

able in large quantities which limit the usage of reagents in

such a small environment. It is therefore important to de-

velop optical elements or systems which can be made effi-

cient and potent in small quantities and can be used re-

peatedly. The use of optical fibers for various sensing

purposes has been reported [3,4]. The emergence of mi-

crostructured optical fibers (MOFs) opens up new oppor-

tunities for novel fluorescent detection and relevant bio-

sensor design, which can solve the problems encountered

in conventional biosensors [5-7]. MOFs are characterized

as having a plurality of air holes running along the entire

length of the fiber [8]. The optical properties of this class

of fibers are determined by their geometry, size, and rela-

tive position of the air holes. Photonic crystal fibers (PCFs)

are one of the most prominent MOFs that have emerged in

recent years that could be engineered to have vastly dif-

ferent properties compared to conventional fibers [9,10].

Its guiding mechanism is based on the photonic bandgap

formed due to its high index contrast (commonly silica

and air in optical region) and from the wavelength-scale

microstructure. The mode propagation properties strongly

depend on wavelength, which in turn depends on the de-

sign, configuration and geometry of air holes [11]. Unlike

conventional fibers, photonic crystal fibers are made of

pure silica glass (SiO2) without any doping. Hence it is

biocompatible and chemically inert [12]. Further, the cap-

illary tubes present in the PCFs have a good surface-

to-volume ratio. The PCF-based sensor hence utilizes the

available sample volume much more efficiently.

The hollow-core PCFs (HC-PCFs) are comprised of an

air core with a cladding that consists of a two-dimensional

(2-D) periodic array of air inclusions in silica [13]. As

indicated by their name, HC-PCFs guide light in the air

core within certain bandgaps, which manifest as transmis-

sion windows in the transmission spectrum. The photonic

bandgap (PBG) property of the fiber is a function of both

its geometry and the refractive-index (RI) contrast [10,13].

The transmission bands, or transmission windows, of the

HC-PCF is decided by the spacing between the holes of

the capillaries (pitch), the hole diameters of the capillaries,

and/or the air-filling content within the inner cladding.

When the holey regions of HC-PCFs are filled with aque-

ous solution, the transmission window shows a blue shift

[14-18]. This approach is in analogous with the well-

known scaling laws that describe the shift in the PBG edge

86 V. K. SHINOJ ET AL.

which is derived from scalar waveguide approximation

[14,15]. By means of the scalar-wave approximation, sim-

ple index RI scaling laws have been derived to predict the

manner in which the photonic states of the fiber scale with

changes in the refractive-index contrast [14]. An experi-

mental demonstration of the shift in the PBG edge due to

refractive-index scaling using D2O-filled HC-PCFs has

reported based on the above approximation [16]. The ap-

plication of HC-PCF as a refractive index sensor based on

RI scaling laws has also been reported [17]. Recently, the

dependence of PBG edge shift on the physical measurands

such as strain, temperature, curvature, and twist are stud-

ied [18]. In this context, this paper investigates the influ-

ence of shift in central wavelength on the fluorescence

emission intensity in common fluorescence sensing stud-

ies employing HC-PCFs.

2. Theoretical Background

In Hollow-core Photonic crystal fibers (HC-PCFs) are

characterized as having a hollow-core surrounded by pat-

tern of air holes running along the entire length of the

fiber. Filling the holes of such a fiber with liquid will

change the refractive index of the holey region and

therefore will result in the shift of band gaps and their

operational bandwidths. The shift in bandgap can be es-

timated by refractive index scaling law which is derived

from scalar waveguide approximation [14,16].



022

















(1)

In Equation (1), ‘na’ represents the ambient index in-

side the holey region, which includes the core and the

holes inside the cladding. The refractive index of back-

ground material and infiltrated material is denoted as ‘nb’

and ‘nm’ respectively. Also, λ0 represents the central

wavelength of the fiber in air medium (n0). Hence for

hollow-core fibers with similar geometry profile, when

the refractive index of the filling material changes from

n0 to nm, the corresponding wavelength shift of the PBG

edge varies from λ0 to λ. Differentiating Equation (1),



22 22

ba bm

nnn nn



















(2)



mbm

nnn



 



(3)

where,









‘K’ is a positive constant determined by the refractive

index of the fiber material and the central wavelength.

(dλ/dnm) represents the refractive index sensitivity and its

negative value indicates that the PBG has a blue-shift in

wavelength with increase in index of the infiltrated ma-

terial. (dλ/dnm) varies with the ambient refractive index

(na) of the medium, background material index (nb) and

infiltrated material index (nm).

Most HC-PCFs have cladding made of pure fused sil-

ica (nb = 1.45) with array of air holes (na = 1) running

along the entire length of the fiber. In HC-PCF based

fluorescence sensing applications, the sample volume is

drawn into the fiber holes using capillary action. In gen-

eral, the fluorescence samples are dissolved/dispersed in

medium such as methanol, water or ethanol etc. Evalua-

tion of Equation (1) and Equation (3) for different filling

material indices ranging from 1.3 to 1.4 are performed

for two hollow-core fibers (from Crystal Fiber A/S) with

central wavelengths 830 nm (HC-800-01) and 1060 nm

(HC-1060-02) which are employed in the experimental

study (section 3). The obtained results given in Figure

1(a) and Figure 1(b) denote the variation in the central

wavelength and refractive index sensitivity, respectively,

for different filling liquid indices. Based on the RI scal-

ing law (Equation (1)), for a particular filling material,

the shifted wavelength (



) is proportional to central

wavelength (



0) of the HC-PCF. The variation of



with



0 is plotted for different filling material indices values

ranging from 1.3 to 1.4 in Figure 2. Here also we con-

sidered most commonly available HC-PCFs that have

cladding made of pure fused silica with array of air holes

running along the entire length of the fiber. It can be seen

that on increasing the filling material indices, the central

wavelength of a particular HC-PCF is shifted to the lower

wavelength region. This shift in central wavelength should

be a significant parameter to be considered in HC-PCF

based fluorescent sensors where fluorescent sample solu-

tions are infiltrated into the fiber holes. An experiment

has been performed to demonstrate the induced changes

in the sensitivity of the HC-PCF based fluorescence sen-

sors due to the shift in wavelength and is explained in the

following section.

3. Experimental Study

3.1. Materials and Methods

Two hollow-core fibers, HC-800-01 and HC-1060-02 are

selected for the experimental study. The scanning electron

micrograph (SEM) images of the HC-PCF facets are given

in Figure 3(a) and Figure 3(b). The HC-800-01 has an

approximate core diameter of 9.3 μm surrounded by a 40

μm-diameter microstructured cladding. It operates at a

V. K. SHINOJ ET AL.

Figure 1. (a) Central wavelength plot for HC-1060 (solid

circles) and HC-800 (solid rectangles) using Equation (1)

and (b) refractive index sensitivity plot for HC-1060 (solid

circles) and HC-800 (solid rectangles) using Equation (3),

for different filling indices between 1.3 and 1.4.

Figure 2. The shift in central wavelength (



0) of HC-PCFs

to the new wavelength (



) at various filling material indices.

center wavelength of 830 nm and exhibits full photonic

bandgap (high transmission range) extending from ap-

proximately 770 nm to 890 nm. The attenuation over this

range is less than 0.5 dB/m. While HC-1060-02 hollow-

core photonic bandgap presents a band larger than 100

nm centred at 1060 nm. The hollow core has a centre

core size of diameter 10 ± 1 µm surrounded by a micro-

structure comprised of eight periods of hexagonally

Figure 3. SEM images of HC-PCF with (a), central wave-

length 830 nm (HC-800) & (b) central wavelength 1060 nm

(HC-1060) and microscopic Side view of (c) cleaved HC-800

end (imaged with10X/0.3NA objective lens) and (d) green

fluorescent microspheres, of size 2μm, immobilized inside

the HC- PCF (imaged with 50X/0.75NA Objective lens).

packed cylinders with a period of 2.75 μm and a filling

fraction of around 90%. The cladding diameter is 123 ± 5

µm. Both the hollow-core fibers are cut into segments of

≈10cm length and one end of the fiber is cleaved carefully

using a fiber cleaver to produce a flat surface. Microscopic

side view of the cleaved fiber end is given in Figure 3(c).

The green fluorescent microspheres (Duke Scientific

Corp.), of diameter ≈2 µm, employed in this study are

internally-dyed polymer beads. The particles are in a solu-

tion of DI water and some surfactants. The green fluores-

cence labeled microsphere immobilized fiber that gives an

emission maximum wavelength at around 508 nm is ex-

cited with blue laser light (473 nm). In order to verify the

influence of photonic bandgap edge shift on the sensitivity

of fluorescence signal, two types of study has been per-

formed. In the first study, same fiber (with central wave-

length 830 nm) has been used for two different dispersion

media such as ethanol (n = 1.36) and distilled water (n =

1.33). In the second case, two fibers with different central

wavelengths (830 nm and 1060 nm) are considered with

sample particles are dispersed in same medium (ethanol).

The experiment is carried out on both fibers for same val-

ues of laser power and similar coupling efficiency in order

to compare the fluorescence collection efficiency.

The cleaved end of the HCPCFs segments were

dipped into the sample solution to allow the sample to

drawn into the fiber due to the capillary effect. The mi-

crosphere particles had nearly the same density as water

(1.05 g/cm3). Therefore, the particles would follow the

fluid flow arising from the capillary force. The presence

of the microsphere sample inside the fiber is detected

using a fluorescence microscope. The obtained fluores-

88 V. K. SHINOJ ET AL.

cent microscopic picture of fiber containing fluorescent

microspheres is shown Figure 3(d). The quantification

of fluorescence signal from both fibers is performed us-

ing spectroscopic analysis as described below.

3.2. Spectroscopic Analysis

Schematic diagram of the experimental setup is shown in

Figure 4. A continuous wave (CW) diode-pumped solid-

state (DPSS) 473 nm laser (output power 10 mW) is

coupled into the proximal end of PCF, immobilized with

fluorescence sample, using a high precision single mode

fiber coupling (FC) unit (Melles Griot Pte Ltd) with a

microscope objective (20X, 0.65NA (L1). The diverging

light beam emerging from the distal end of the sample

immobilized fiber is collimated using a microscope ob-

jective lens [Newport M-20X/0.4 (L2)]. The parallel

beam emerging from this objective lens is focused onto

the entrance slit of the high quantum efficiency spec-

trometer using another microscope objective lens [New-

port M-40X/0.65 (L3)]. The spectrometer is coupled to a

PC which displays the spectrum.

The fluorescent spectra obtained at identical conditions

from HC-800 fiber segments for ethanol and water dis-

persed microsphere samples are normalized as shown in

Figure 5. The water dispersed fiber gives better signal for

green fluorescence when compared to the ethanol dis-

persed fiber. This result is in agreement with the result

obtained in section 2 (Figure 1 and Figure 2). It can be

inferred from Figure 2 that for a fiber with central wave-

length (λ0) 830 nm (HC-800), the filling of water causes

the shift in central wavelength to an approximate value of

457 nm. But, the filling of ethanol shifts the central wave-

length from 830 nm to 397 nm, approximately. Hence the

water filled HC-800 has central wavelength nearer to the

green region which results in better sensitivity. The re-

sults are found to be reproducible for different fiber seg-

ments with same central wavelength (830 nm).

Figure 6 shows the normalized spectra obtained from

fibers with central wavelengths 830 nm and 1060 nm for

green fluorescent particles dispersed in ethanol medium.

It is vivid that the fiber with central wavelength 1060 nm

Figure 4. Schematic diagram—experimental set up used for

the spectral analysis.

Figure 5. The obtained fluorescence spectrum (normalized)

at 473 nm excitation from the HC-PCF with central wave-

length 830 nm filled with green fluorescent microparticles

dispersed in water (black solid rectangles) and ethanol (red

solid circles) [Inset: the wavelength region corresponding to

the fluorescence emission is expanded].

Figure 6. The obtained fluorescence spectrum (normalized)

at 473 nm excitation from the HC-PCFs with central wave-

lengths 830 nm (red solid circles) and 1060 nm (blue solid

triangles) filled with green fluorescent microparticles dis-

persed in ethanol [Inset: the wavelength region correspond-

ing to the fluorescence emission is expanded].

shows better fluorescent signal when compared with fi-

ber of central wavelength 830 nm. The higher intensity

obtained with 1060 nm fiber is also in agreement with

the results shown in section 2, corresponding to refrac-

tive index 1.36 which is refractive index of the dispersed

medium (ethanol). From Figure 2, it can be seen that for

a filling material index of 1.36, central wavelength of the

HC-1060 fiber will shift from 1060 nm to around 508 nm.

Whereas in the case of fiber with central wavelength at

830 nm (HC-800), the approximate value of shifted

wavelength is 400 nm for the infiltrated material index of

V. K. SHINOJ ET AL.

Figure 7. Comparison of green fluorescent signals obtained

from the HC-800 filled with green fluorescent microparticles

dispersed in water (black solid rectangles) and ethanol (red

solid circles), and HC-1060 filled with green fluorescent mi-

croparticles dispersed in ethanol (blue solid triangles).

1.36. The results are repeatable for different fiber segments

with the same central wavelengths. The fluorescent spec-

troscopic signals obtained (which are shown in Figure 5

and Figure 6 are plotted in Figure 7 for intensity com-

parison. The obtained results are found to be in accordance

with the simulation done based on RI scaling law.

4. Conclusions

In conclusion, an HC-PCF fluorescence spectroscopic

scheme has been illustrated on the basis of refractive

index scaling law. The variations in the central wave-

length for different filling material indices are analyzed

in the case of HC-PCFs with cladding made of pure

fused silica with array of air holes running along the en-

tire length of the fiber. A proof of concept study has been

performed by infiltrating fluorescence sample volume

inside HC-PCF and the quantification of fluorescence

intensity is analyzed using spectroscopic method. The

sensitivity has been compared for similar fiber with dif-

ferent dispersed media and different fibers with same

dispersed medium. The obtained experimental results are

in good agreement with the analytical simulation results.

These findings are expected to accelerate the R&D on

HC-PCF based ultrasensitive spectroscopic analysis and

relevant sensors for specific detection of biomolecules in

very low sample volumes.

5. Acknowledgements

The authors acknowledge the financial support received

through ASTAR-SERC and ARC (MOE). One of the

authors, V. K. Shinoj, would also like to acknowledge

Nanyang Technological University for the research stu-

dent support.

6. References

[1] E. Brooks Shera, N. Seitzinger, L. Davis, R. Keller and S.

Soper, “Detection of single fluorescent molecules,”

Chemical Physics Letters, Vol. 174, No. 6, 1990, pp.

553-557. doi:10.1016/0009-2614(90)85485-U

[2] A. Castro, and E. Shera, “Single-Molecule Detection:

Applications to Ultrasensitive Biochemical Analysis,” Ap-

plied Optics, Vol. 34, No. 18, 1995, pp. 3218-3222.

doi:10.1364/AO.34.003218

[3] Y. Zhao, and Y. Liao, “Discrimination Methods and De-

modulation Techniques for Fiber Bragg Grating Sensors,”

Optics and Lasers in Engineering, Vol. 41, No. 1, 2004,

pp. 1-18. doi:10.1016/S0143-8166(02)00117-3

[4] J. I. Peterson, and G. G. Vurek, “Fiber-Optic Sensors for

Biomedical Applications,” Science, Vol. 224, No. 4645,

1984, pp. 123-127. doi:10.1126/science.6422554

[5] J. Jensen, P. Hoiby, G. Emiliyanov, O. Bang, L. Pedersen,

and A. Bjarklev, “Selective Detection of Antibodies in

Microstructured Polymer Optical Fibers,” Optics Express,

Vol. 13, No. 15, 2005, pp. 5883-5889.

doi:10.1364/OPEX.13.005883

[6] C. Cordeiro, M. Franco, G. Chesini, E. Barretto, R. Lwin,

C. Brito Cruz and M. Large, “Microstructured-Core Opti-

cal Fibre for Evanescent Sensing Applications,” Optics

Express, Vol. 14, No. 26, 2006, pp. 13056-13066.

doi:10.1364/OE.14.013056

[7] A. S. Webb, F. Poletti, D. J. Richardson and J. K. Sahu,

“Suspended-Core Holey Fiber for Evanescent-Field Sens-

ing,” Optical Engineering, Vol. 46, No. 1, 2007, p.

010503. doi:10.1117/1.2430505

[8] M. van Eijkelenborg, M. Large, A. Argyros, J. Zagari, S.

Manos, N. Issa, I. Bassett, S. Fleming, R. McPhedran and

C. de Sterke, “Microstructured Polymer Optical Fibre,”

Optics Express, Vol. 9, No. 7, 2001, pp. 319-327.

doi:10.1364/OE.9.000319

[9] P. Russell, “Review: Photonic Crystal Fibers,” Science,

Vol. 299, 2003, pp. 358-362.

doi:10.1126/science.1079280

[10] J. C. Knight, J. Broeng, T. A. Birks and P. S. J. Russell,

“Photonic Band Gap Guidance in Optical Fibers,” Science,

Vol. 282, No. 5393, 1998, pp. 1476-1478.

doi:10.1126/science.282.5393.1476

[11] D. Mogilevtsev, T. A. Birks and P. S. J. Russell,

“Group-Velocity Dispersion in Photonic Crystal Fibers,”

Optics Letters, Vol. 23, No. 21, 1998, pp. 1662-1664.

doi:10.1364/OL.23.001662

[12] S. Padmanabhan, V. K. Shinoj, V. M. Murukeshan and P.

Padmanabhan, “Highly Sensitive Optical Detection of

Specific Protein in Breast Cancer Cells Using Microstruc-

tured Fiber in Extremely Low Sample Volume,” Journal

of Biomedical Optics, Vol. 15, No. 1, 2010, pp.

017005-017006. doi:10.1117/1.3302810

[13] R. F. Cregan, B. J. Mangan, J. C. Knight, T. A. Birks, P. S.

J. Russell, P. J. Roberts and D. C. Allan, “Single-Mode

V. K. SHINOJ ET AL.

Photonic Band Gap Guidance of Light in Air,” Science

Vol. 285, No. 5344, 1999, 1537.

doi:10.1126/science.285.5433.1537

[14] T. Birks, D. Bird, T. Hedley, J. Pottage and P. Russell,

“Scaling Laws and Vector Effects in Bandgap-Guiding

Fibres,” Optics Express, Vol. 12, No. 1, 2004, pp. 69-74.

doi:10.1364/OPEX.12.000069

[15] J. Joannopoulos, R. Meade and J. Winn, “Photonic Crys-

tals,” Princeton University Press, Princeton, 1995.

[16] G. Antonopoulos, F. Benabid, T. A. Birks, D. M. Bird, J.

C. Knight and P. S. J. Russell, “Experimental Demonstra-

tion of the Frequency Shift of Bandgaps in Photonic

Crystal Fibers Due to Refractive Index Scaling,” Optics

Express, Vol. 14, No. 7, 2006, pp. 3000-3006.

doi:10.1364/OE.14.003000

[17] J. Sun and C. C. Chan, “Photonic bandgap fiber for refrac-

tive index measurement,” Sensors and Actuators B:

Chemical, Vol. 128, No. 1, 2007, pp. 46-50.

doi:10.1016/j.snb.2007.05.037

[18] S. H. Aref, R. Amezcua-Correac, J. P. Carvalho, O.

Frazão, J. L. Santos, F. M. Araújo, H. Latifi, F. Farahi, L.

A. Ferreira and J. C. Knight, “Spectral Characterization of

a Photonic Bandgap Fiber for Sensing Applications,” Ap-

plied Optics, Vol. 49, No. 10, 2010, pp. 1870-1875

doi:10.1364/AO.49.001870