864
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 11, NO. 4, JULY/AUGUST 2005
Detecting Hybridization of DNA by Highly Sensitive
Evanescent Field Etched Core Fiber Bragg
Grating Sensors
Athanasios N. Chryssis, Student Member, IEEE, Simarjeet S. Saini, Member, IEEE, Sang M. Lee, Hyunmin Yi,
William E. Bentley, and Mario Dagenais, Senior Member, IEEE
Abstract—Highly sensitive fiber Bragg grating sensors were developed by etching away the cladding and part of the core of
the fiber and detecting the change of Bragg wavelength due to
the change of index of the surrounding medium. A sensitivity of
1394 nm/riu was achieved when the diameter of the grating core
was 3.4 µm and the index of the surrounding medium was close to
the index of the core of the fiber. Assuming a detectable spectral resolution of 0.01 nm realized in the experiment, the sensor achieves a
minimum detectable index resolution of 7.2 × 10−6 . Higher sensitivity at lower surrounding index was achieved by using higher
order modes excited in the Bragg grating region. The use of the
fiber Bragg grating sensor was further investigated to detect hybridization of DNA. Single stranded DNA oligonucleotide probes
of 20 bases were immobilized on the surface of the fiber grating
using relatively common glutarahyldehyde chemistry. Hybridization of complimentary target single strand DNA oligonucleotide
was monitored in situ and successfully detected. The demonstrated
fiber Bragg grating sensors provide an elegant method to monitor biological changes in an in situ manner, and provide temporal
information in a single experiment.
Index Terms—Biological sensor, chemical sensor, evanescent
field, fiber Bragg grating (FBG).
I. INTRODUCTION
T is important to monitor the physical, chemical, and biological parameters in natural environments for ecosystems
protection. Fiberoptic biosensors based on evanescent wave interactions have often been used in medical and environmental
diagnostics. Fiber optic sensors offer numerous advantages over
electric transducers due to their small size, high sensitivity, and
possibility of distributed measurements [1]. Different optical
sensing technologies have been developed capable of detecting
a broad range of biological reagents including antibodies [2], [3],
nucleic acids [4], toxins [5], pesticides [6], explosives [7], and
other small molecules.
A common approach used with optical fiber biological sensors
is the chemical modification of the silica surface allowing for the
immobilization of a high density of biological molecules; e.g.,
I
Manuscript received December 17, 2004.
A. N. Chryssis, S. S. Saini, S. M. Lee, and M. Dagenais are with the Department of Electrical and Computer Engineering, University of Maryland, College
Park, MD 20742 USA (e-mail: chryssis@glue.urnd.edu; ssaini@altanetcomm.
com; slee43@umd.edu; dage@eng.umd.edu).
H. Yi and W. E. Bentley are with the Department of Chemical Engineering
and Center for Agricultural Biotechnology, University of Maryland, College
Park, MD 20742 USA (e-mail: hyunmin@umd.edu; bentley@umd.edu).
Digital Object Identifier 10.1109/JSTQE.2005.857724
antibodies over the surface of the fiber. Different immobilization methods have been reported including avidin bridging [8],
covalent immobilization using heterobifunctional cross-linking
agents [9], [10], and noncovalent attachment via adsorption and
gel entrapment [11]. A target antigen specific to an immobilized
antibody is incubated with an auxiliary label such as fluorophore
or radioisotope, prior to detection. For detection, a sandwich
immunoassay is carried out on the fiber surface generating a
real time fluorescence or radioactive signal as the labeled antigen specifically interacts with the immobilized antibody. The
main disadvantage of this method is the requirement of the
antigen labeling, which requires additional reagents. Also, real
time monitoring of the biological process is not possible in this
method.
Another method to monitor biological changes is to monitor
the change of the refractive index that occurs in the evanescent
field upon binding of the target antigen on the derivatized surface. In this way, direct detection of molecular interactions can
be carried out. Specifically, fiber Bragg grating (FBG) sensors
have attracted considerable attention [12]. Gratings are simple
intrinsic sensing elements which can be photoinduced into a silica fiber [13]. The change of wavelength due to the changes in
measurand is used for sensing applications. Of particular interest is the use of FBGs as chemical and biological sensors where
the change of wavelength is induced by changing the composition of the surrounding medium around the sensor and detecting
the change of the Bragg wavelength. In order to measure small
changes in composition, it is important for the optical mode to
penetrate evanescently into the surrounding solution. Meltz et
al. proposed a scheme for chemical sensing based upon Bragg
grating in- and out-coupling for increased fluorescence excitation [14]. A sensor based upon long-period grating-assisted
coupling to the cladding modes of a fiber was presented in [15].
A method of increased sensitivity to the surrounding index by
etching the fiber close to the core diameter was presented by
Asseh et al. [16]. A sensitivity of 2.66 nm/riu was achieved
with the diameter of the fiber etched to 11 µm. The sensitivity
was increased to 7.3 nm/riu by etching the fiber to a diameter
of 8.3 µm in [17]. Schroeder et al. [18] proposed a method of
side-polishing the fiber and achieved a sensitivity of 340 nm/riu.
In this paper, a high sensitivity of 1394 nm/riu is demonstrated by etching the core of the fiber Bragg grating. With
the wavelength resolution of the setup to be 0.01 nm, an index
change of 7 × 10−6 was detectable. However, the high sensitivity is only achieved when the index of the surrounding medium
1077-260X/$20.00 © 2005 IEEE
CHRYSSIS et al.: DETECTING HYBRIDIZATION OF DNA BY HIGHLY SENSITIVE EVANESCENT FIELD ETCHED CORE FBG
is close to that of the fiber core. It is also shown that when the
index of the surrounding medium is less than that of the fiber
core, the sensitivity can be increased by monitoring higher order modes rather than the fundamental one. Finally, the use of
sensor to detect hybridization of DNA is demonstrated. Single
stranded DNA oligonucleotide probes of 20 bases were immobilized on the surface of the fiber grating using relatively common
glutarahyldehyde chemistry. Hybridization of a complimentary
target single strand DNA oligonucleotide was monitored in situ
and successfully detected. Measurements of change in Braggwavelength were recorded with time at every step of the process.
These temporal measurements should provide a better understanding of various chemical and biological processes and allow
for further optimization of the processes.
The paper is organized as follows. In Section II, a generalized theory for etched FBG sensors is developed and described.
The theory can be used to calculate the Bragg wavelength for
a given diameter and index of surrounding medium, and is also
used for understanding the sensitivity of the sensor. The experimental results obtained for characterizing the sensors with
different surrounding indices are described in Section III. Increased sensitivity from detection of higher order modes is also
presented. Controlled etching of the fiber to the desired core diameter is also described in this section. Section IV describes the
experiments for the hybridization of DNA and its detection by
using the sensor. The various chemical and biological processes
are detailed. Results are also presented for the shift of Bragg
wavelength at each step of the experiment. Section V discusses
some of the future improvements and the experiments that need
to be executed. Finally, the paper is concluded in Section VI
with a brief summary of results.
II. THEORY OF ETCHED FIBER BRAGG GRATING SENSORS
The basic principal of operation of FBG sensors is to monitor
the change in the peak of the Bragg wavelength as the index of
the measurand is changed. We defined a figure of merit for the
sensor called “sensitivity” that is defined as the change in Bragg
wavelength in nanometers due to change of refractive index of
1 and its unit is nm/riu (where riu is the refractive index unit).
In order to detect small changes in index of the surrounding
medium, it is important to increase the sensitivity of the sensor.
Intuitively, the sensitivity should increase as more of the optical mode penetrates evanescently into the surrounding solution.
This can be achieved by etching the core of the fiber and letting
the mode expand out of the core. The sensitivity should also increase as the index of the surrounding medium approaches that
of the core of the fiber as the index step is reduced and the mode
expands out of the core. In order to quantify the performance
of the sensor for different etch depths and surrounding medium
index, a graphical solution was developed to calculate the Bragg
wavelength.
Fig. 1 shows the standard relationship between the normalized effective index “b” and the normalized frequency “V ”. The
values of b and V can be calculated by solving the mode equations for the fiber [19], and for a fiber with a core diameter of
865
Fig. 1. Normalized propagation versus V of a fiber. Constant b curves for the
grating are also plotted. The intersection gives the Bragg wavelength of the
grating.
“a” are given by
(β/k)2 − n22
n2eff − n22
=
(1)
2
n2
n21 − n22
1/2
2πa 2
n1 − n22
v=
(2)
λ
where n1 is the index of the core, n2 is the index of the clad
(surrounding medium), neff is the effective index of the propagating mode, and β is the propagation constant of the mode and
k is the wave number. The value of n1 for standard single mode
fiber is 1.4504. The Bragg wavelength reflected by a grating is
given by
b=
λ = 2neff Λ
(3)
where Λ is the grating pitch. By substituting neff as λ/2Λ and
substituting the value of λ in terms of V from (2) in (1), the
effective index b can be written in terms of the grating parameters
as
πa 2
λ 2
n1 − n22 − n22
− n22
2Λ
vΛ
=
.
(4)
b=
n21 − n22
n21 − n22
The normalized effective index versus the normalized frequency V is plotted for different modes by solving the modal
equations for a fiber and also in terms of the grating parameters
as given by (4). These curves are plotted in Fig. 1 for different fiber diameters and a value of n2 of 1.33 (the value for the
etchant at 1550 nm). The wavelength window between 1.5 and
1.6 µm for different fiber diameters is also added for reference.
The intersection of the two curves gives the Bragg wavelength
reflected under given conditions. As the core of the fiber is
etched, the Bragg wavelength shifts toward shorter wavelengths
for the same surrounding index. This property can be used to in
situ monitor the diameter of the core during the etching process
and very good control can be achieved. Also, the wavelength
changes faster for an equivalent change of the surrounding index for fiber gratings with reduced core. This is because the
866
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 11, NO. 4, JULY/AUGUST 2005
Fig. 3.
Fig. 2.
Shift in Bragg wavelength as the core of the fiber is etched.
Experimental setup for in situ monitoring of fiber etch.
evanescent field penetrates further into the surrounding medium
for small diameter fibers.
If the value of V is less than 2.405, then only a single mode
exists and a single reflected peak is observed. This condition is
achieved when the index of the surrounding medium is close
to that of the core of the fiber; i.e., 1.4504 for standard single
mode fiber. However, in chemical and biological sensing, this
may not be always achievable; e.g., if, the index of acetone
is close to 1.34 at the wavelength of 1550 nm. In that case,
the value of normalized frequency V is greater than 2.405 and
multiple modes can exist in the fiber. For a given diameter, the
higher order modes shift more in wavelength with changing
index of the surrounding mode. This is because the evanescent
field penetrates more in the surrounding medium for higher
order modes compared to the fundamental mode. Also, higher
order modes have reflection peaks at a wavelength shorter than
that of the fundamental mode.
III. SENSOR CHARACTERIZATION
Commercially available FBGs were used for measuring the
sensitivity of the sensor for different diameters. The bandwidth
of the grating was 0.4 nm. There were two gratings in a single fiber with peak Bragg wavelengths at 1536 and 1564 nm
for the two gratings. Only the 1564-nm FBG was etched, and
the reflection from the 1533-nm grating was used as a control
signal to compensate for any change in temperature during the
experiment. The FBGs were further mounted in temperature
compensated housing for fiber protection and for reducing the
effect of change in temperature. The experiments were carried
out in a fume hood.
A. Etching of the Fiber Bragg Grating Sensor
The clad of the fiber grating was initially etched to approximately 20 µm in buffered oxide etch (BOE) at a rate of
800 Å/min. This etch was controlled by visual inspection by
measuring the diameter of the fiber under a microscope. The
FBG was then etched controllably by measuring the Bragg
wavelength in situ. The setup for the measurement is shown
in Fig. 2. Broadband light from an erbium doped amplifier is
coupled into the sensor using a 3-dB coupler. The reflected light
Fig. 4. Microscope photograph of an etched fiber. A tapered etch is observed
at the interface of the etch.
is measured using an optical spectrum analyzer and the peak reflected wavelength is monitored. An index of 1.33 is estimated
for the BOE at a wavelength of 1550 nm. The shift of the Bragg
wavelength as the fiber core is etched is shown in Fig. 3. A
theoretical fit calculated using the graphical solution described
previously is also shown in the figure. A good agreement is
achieved between the theory and the experimental values. Using the in situ monitoring, fibers with core diameters as small
as 3 µm were reproducibly etched with precision of ±0.2 µm.
The microscope image for an etched fiber is shown in Fig. 4.
A plastic jacket is initially removed from the part of the fiber
that needs to be etched. A taper is observed between the etched
fiber and the unetched fiber. This is because of the creeping of
the etchant under the plastic jacket due to capillary effects. The
length of the taper is approximately 100 µm and is nonadiabatic.
Hence, if higher order modes exist in the etched fiber (as determined by the value of V), they will be excited. If the taper shape
and length can be controlled, an almost adiabatic transformation
may be achieved between the etched and the unetched fiber and
a single mode can be excited [20]. This will be investigated in
detail in the future.
B. Sensitivity of the Etched Core Fiber Bragg Grating Sensor
The peak Bragg wavelength for the reflected spectrum for
the fundamental mode as the index of the surrounding medium
CHRYSSIS et al.: DETECTING HYBRIDIZATION OF DNA BY HIGHLY SENSITIVE EVANESCENT FIELD ETCHED CORE FBG
867
Fig. 7. Measurable change of index for sensors with different diameters.
0.01-nm resolution of peak wavelength detection is assumed.
Fig. 5. Change in Bragg wavelength with changing surrounding index for
different diameters of fiber.
Fig. 8. Sensitivity plots for the first order and second order modes as the index
of the surrounding medium is changed. Theoretical values are also plotted along
with the experimental values.
Fig. 6.
Sensitivity for fibers with different core diameters.
is changed is plotted in Fig. 5. Two sensors with diameters
3.4 and 4.0 µm were used for the measurements. Very stable
known index fluids from Cargille were used as the surrounding medium. The indices of the fluids were calculated at the
wavelength region of 1.55 µm using Cauchy relationships. As
the surrounding index is increased and approaches the core index, the Bragg wavelength increases. Also, the rate of change in
wavelength with surrounding index increases as the surrounding
index approaches the clad index. This rate of change in Bragg
wavelength with surrounding index defines the sensitivity of the
sensor and is plotted for two different sensors with diameters 4.0
and 3.4 µm in Fig. 6. Theoretical calculations are also plotted
in the figure. A maximum sensitivity of 1394 is achieved for a
3.4-µm sensor, which is at least four times larger than previously
reported values [18]. The sensitivity of the sensor also increases
as the diameter of the core is decreased. A good agreement is
achieved between the theoretical and experimental values. To
measure the value of the peak wavelength, an optical spectral
analyzer was used with a resolution of 0.01 nm. A minimum
change of the surrounding index of 7.2 × 10−6 can be resolved
by the sensor with a diameter of 3.4 µm, as shown in Fig. 7. It is
also to be noted that the index sensitivity increases with reduced
diameter, and also as the surrounding index approaches that of
the core index.
Sensitivity of a higher order mode was also measured for fiber
grating with a diameter of 4 µm. The sensitivity of the sensor for
the fundamental and the second-order mode as the index of the
surrounding medium changed is plotted in Fig. 8. Theoretical
curves are also plotted in the same figure with the bold curves
representing the second order mode. The sensitivity for both the
fundamental and second order mode increases as the index of
the surrounding medium is increased. Also, the sensitivity for
the second order mode is larger by at least a factor of two compared to the first order mode. However, after a particular index
of the surrounding medium, the second order mode achieved
cutoff and no Bragg reflection is observed. This cutoff results
from a smaller index step between the core and the surrounding
and the V parameter becoming less than 2.405. An experimental
peak sensitivity value of maximum sensitivity of 404 nm/riu is
achieved for the second order mode with the surrounding index
of 1.4098. The sensitivity of the first order mode at the same
surrounding index is only 171 nm/riu. The sensitivity of the
first order mode keeps on increasing and achieves a value of
546.4 nm/riu at surrounding index of 1.429, and should increase
further as the index approaches 1.4504.
From Fig. 8, it is realized that the first order mode achieves the
absolute maximum sensitivity. However, this high sensitivity is
achieved only when the surrounding index is close to that of the
core of the fiber. This may not always be possible in chemical
and biological sensing where the index of the chemical and
biological reagents can be less. When the surrounding index is
small enough compared to the core such that higher order modes
exist, then the sensitivity of the higher order modes is larger than
868
Fig. 9.
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 11, NO. 4, JULY/AUGUST 2005
Reflected spectrum of the grating in DI water.
Fig. 10. Shift of wavelength with time as the sensor is immersed in APTES.
Most of the change takes place in the first 10 min.
that of the fundamental order mode. Hence, in order to achieve
maximum sensitivity to changes of index of the surrounding
medium, the change of wavelength of the highest order which
exists under given conditions should be measured.
IV. DETECTION OF DNA HYBRIDIZATION
The large value of the sensitivity obtained in the sensor can
be used to measure small changes in index of the surrounding
medium. This can be used to measure the presence of chemical
or biological agents. Detection of hybridization of a target DNA
with a probe DNA was detected using the etched core FBG
sensor. Probe DNA was initially attached to the surface of the
fiber after chemical treatment of the fiber using common glutarahyldehyde chemistry. The shift of the Bragg wavelength is
measured during and after the hybridization of the DNA. In this
section, the various steps of the process and the measurements
in each of the steps are described.
Fig. 11. Measurement of the spectrum in DI water before silanization, after
silanization and after glutarahyldehyde chemistry on the sensor surface. An
increase in wavelength is observed after every step.
A. Preparation of Etched Core Fiber Bragg Grating
For this experiment, a grating similar to that used in Section
III was used. One of the two fiber Bragg gratings was etched to a
core diameter of 5 µm. The peak wavelength of the fundamental
mode was 1553.07 nm in deionized (DI) water at room temperature. The reflection of the fundamental mode of the grating
in DI water is shown in Fig. 9. It is seen from the figure that
the response of the grating is flat at the peak wavelength for
approximately 0.2 nm and, hence, there could be error in the
measurement of the peak wavelength, especially if the shift is
small. Hence, the shift of the wavelength of the first dip was
monitored as shown in Fig. 9. The reflection from the unetched
grating was used as a control signal to compensate for any temperature change. An experimental setup similar to that shown in
Fig. 2 was used to measure the reflected spectrum.
For this experiment, the fiber grating was freshly etched in
BOE and then cleaned in DI water. For previously etched fiber,
the surface may need to be cleaned with an acid wash such as
dilute nitric acid or sulphuric acid as shown in [21]. Silanization
of the fiber grating surface was performed by immersion in fresh
1% 3-Aminopropyl-triethoxysilane (APTS) in water for 30 min
at room temperature. The shift in wavelength of the first dip as
a function of time is shown in Fig. 10. A wavelength increase of
22 pm is observed after 30 min from the start of the silanization.
As seen in Fig. 10, the biggest shift was seen in the first ten
minutes of the immersion.
The sensor was then rinsed in DI water before the silanization
was activated for single strand DNA assembly (ssDNA) by glutarahyldehyde chemistry [22], [23]. The reflected spectrum was
measured for the modified sensor in DI water. The sensor was
then immersed in 5 ml of 0.1 % glutarahyldehyde solution for
30 min at room temperature. An increase of 7 pm was observed
in wavelength from the start of the glutarahyldehyde process
until the end.
The sensor was again rinsed in DI water, and the reflected
signal was measured. The spectrum is shown in Fig. 11. The
original spectrum in DI water and also the spectrum in DI water
CHRYSSIS et al.: DETECTING HYBRIDIZATION OF DNA BY HIGHLY SENSITIVE EVANESCENT FIELD ETCHED CORE FBG
869
Fig. 12. Binding test: the fluorescence on the surface of the etched fiber demonstrates successful chemical binding.
Fig. 14. Reflected spectrum when the sensor is immersed in the probe DNA
solution and after overnight immersion.
Fig. 13.
Chemically processed sensor in Dl water and SSC buffer.
after silanization, are also shown in the same figure. An increase
in the wavelength was observed which is consistent to the change
in the wavelength observed during the chemical processes. The
change in wavelength shows that the surface of the fiber has
been modified. For one sensor, a fluorescence tag was attached
to the glutarahyldehyde. The fluorescence of the fiber surface
was measured and is shown in Fig. 12. Uniform fluorescence on
the surface of the etched fiber demonstrates successful chemical
binding.
B. Immobilization of Probe DNA
The sensor was then equilibrated in a saline sodium citrate
(SSC) buffer (20x concentrated, molecular biology grade) with
0.1 M MgCl2 for 10 min. To prepare this buffer, the SSC buffer
was first diluted with DI water. After autoclaving, autoclaved 4
M MgCl2 solution was added and the pH was adjusted to 7.1. For
the rest of the experiment, the SSC buffer was used to rinse the
fiber instead of DI water. Fig. 13 shows the reflected spectrum
of the grating in the SSC buffer. The Bragg wavelength for the
sensor in SSC buffer is larger by 65 pm as compared to that in
Dl water due to the different refractive index of the SCC buffer.
Fig. 15. Sensor immersed in SSC buffer before and after the immobilization
of the probe DNA.
After equilibrating the sensor in SSC buffer, it was immersed
in a 3 ml solution containing 20 µg/ml of the amine-terminated
ssDNA surface probe. Specifically; the ssDNA was a 20-base
sequence complementary to a region in the dnaK gene of E.
coli, as shown in Table I. Details of the DNA sequences can be
seen in [22]. The sensor was kept immersed overnight at 4 ◦ C.
The Bragg wavelength at the start of the immersion and after
the overnight immobilization is shown in Fig. 14. A wavelength
increase of 146 pm was observed after the overnight deposition.
Also, the peak intensity decreased after the immobilization of
the probe DNA. This could be due to the surface roughness on
the fiber surface caused by immobilized probe DNA leading to
increased radiation loss.
The fiber was then rinsed in SSC buffer to remove any
unattached probe DNA from the fiber surface. The Bragg wavelength of the sensor in SSC buffer before and after the immobilization of the DNA is shown in Fig. 15. An increase in
wavelength of 51 pm was observed in the SSC buffer after the
870
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 11, NO. 4, JULY/AUGUST 2005
TABLE I
OLIGONUCLEOTIDE SEQUENCES AND 5’ MODIFICATIONS. UNDERLINED SEQUENCES REPRESENT FIVE-BASE COMPLEMENTARY PORTIONS
BETWEEN dnaK OLIGONUCLEOTIDES
Fig. 16.
Shift of wavelength during the hybridization of the target.
immobilization of the probe DNA, showing that the surface of
the fiber has been modified.
C. Hybridization
Hybridization reactions were performed with purified 20-base
ssDNA targets. Specifically, a complementary 20-base E-coli
dnaK target was used, as shown in Table I. Hybridization was
carried out by immersing the sensor in 1 ml of Sigma PerfectHyb
hybridization buffer containing the target ssDNA. Before the
immersion, the sensor was rinsed with the hybridization buffer
to remove any traces of the SSC buffer, and also to equilibrate
the sensor. The reflected spectrum was also measured in the
hybridization buffer as a control signal to be compared to after
the hybridization has been achieved.
After rinsing the sensor in the hybridization buffer, it was
immersed in a 1 ml solution containing 0.7 µg/mL of the target ssDNA for 60 min. The change of Bragg wavelength was
measured with time and the shift is shown in Fig. 16. A total
decrease of 29 pm is observed in the 60 min of hybridization.
Most of the change takes place in the first few minutes of the
hybridization process showing that the hybridization takes place
very quickly. Data in the first minute was not taken due to the
capture time of the OSA, and we believe that there was appreciable change of wavelength during that interval. In future
experiments, the results can be taken with better temporal resolution with a modified signal processing method, allowing for a
better understanding of the hybridization process. The ability to
detect changes in situ while conducting the experiment is a very
powerful aspect of the etched-core fiber Bragg grating sensor.
After the hybridization process, the sensor was again rinsed
with the hybridization buffer and the reflected spectrum was
Fig. 17. Sensor immersed in Hybridization buffer before and after the hybridization of the target DNA with the probe DNA on the surface of the fiber.
measured with the sensor immersed in the hybridization buffer.
The reflected spectrum with and without hybridization of the
target DNA is shown in Fig. 17. A decrease in wavelength
of 73 pm was observed in the spectrum before and after the
hybridization of the DNA. The decrease in wavelength is larger
change than the one seen while the hybridization was taking
place, and it could be due to the fact that we did not measure
the data in the first minute of the hybridization process.
V. FUTURE WORK
It has been shown that hybridization of DNA can be detected
by using an etched core fiber Bragg grating sensor. The reflected
spectrum in the system was detected by an optical spectrum analyzer with a resolution of 0.01 nm. The signal processing system
can be improved by using a scanning fiber Fabry Perot filter demodulation system as described in [21]. It may be possible to
achieve a higher resolution than with the OSA, and this is being
investigated.
Another important experiment will be to investigate a lower
concentration of target DNA that can be detected by the sensor.
The current sensor has a diameter of 5 µm and has a sensitivity of
29 nm/riu for the index of the hybridization buffer. It has already
been shown that the sensitivity can be increased by etching the
diameter further. Also, the sensitivity can be increased further
by using the higher order modes instead of the fundamental
mode. In order to detect lower concentrations of target DNA
and investigate the limit of detection of the concentration of the
target DNA, it will be important to incorporate the previously
discussed changes.
It will also be interesting to look into the stability and reliability of the sensor for repeated use by regeneration of surface
CHRYSSIS et al.: DETECTING HYBRIDIZATION OF DNA BY HIGHLY SENSITIVE EVANESCENT FIELD ETCHED CORE FBG
activity. It has previously been shown that the target nucleic
acid from the probe can be removed by subjecting the sensor
to denaturing conditions of 4 M urea at 65 ◦ C for 30 min [23].
The sensor can be exposed to a sequence of hybridization and
denaturation steps, and the repeatability of the change of signal
can be measured to investigate whether the sensor can be used
multiple times.
VI. CONCLUSION
In conclusion, highly sensitive etched core fiber Bragg grating
sensors have been demonstrated which detect the change of
index of the surrounding medium by measuring the change of
the Bragg wavelength. It was demonstrated that the sensitivity
of the sensor can be increased by etching the core of the fiber.
A high sensitivity of 1394 nm/riu was achieved with a fiber
core of 3.4 µm diameter. This high sensitivity is achieved when
the index of the surrounding medium is close to that of the
core. It was shown that when the surrounding index is less than
that of the core, the sensitivity of higher order modes is larger
than the fundamental mode. A sensitivity of 404 nm/riu was
achieved with the second order mode compared to a sensitivity
of 171 nm/riu at a surrounding index of 1.4098. Hence, in order
to achieve maximum detection sensitivity to changes of index
of the surrounding medium, the change of wavelength of the
highest order which exists under the experimental conditions
should be measured.
The Bragg sensor was used to detect hybridization of single
strand DNA. Probe DNA was immobilized on the surface of the
fiber using the common and simple glutarahyldehyde chemistry.
Target DNA was then hybridized and the change of wavelength
was measured. A change of wavelength of 73 pm was observed
after 60 min of the hybridization corresponding to change of
index of 2.5 × 10−3 . Most of the change occurred in the first
minute. The ability to achieve temporal measurements as various
steps are undertaken allows for a powerful tool to analyze the
chemical efficiency of the different steps. This should allow
for further optimization of the different steps in the process of
immobilization of DNA.
Since the change of wavelength is being detected, the DNA
does not need to be labeled with a fluorescent tag, which can
be time consuming and expensive. Also, multiple sensors with
different probe DNA immobilized on the surface can be prepared
on a single fiber with the signal being wavelength multiplexed
by having gratings at different wavelengths. This will allow for
detection of multiple reagents by the same sensor and a single
signal processing setup.
We believe that the etched core fiber Bragg grating sensor
can be a powerful and versatile sensor to detect chemical and
biological reagents, and it permits the in situ monitoring of
bioprocesses.
REFERENCES
[1] D. Yeung, A. Gill, C. H. Maule, and R. J. Davies, “Detection and quantification of biomolecular interactions with optical biosensors,” Trends Anal.
Chem., vol. 14, pp. 49–56, 1995.
[2] B. Lu, C. Lu, and Y. Wei, “A planar quartz wave-guide immunosensor
based on Tirf principle,” Anal. Lett., vol. 25, pp. 1–10, 1992.
871
[3] R. Cush, J. M. Cronin, W. J. Stewart, C. H. Maule, J. Molley, and N.
J. Goddard, “The resonant mirror: A novel optical biosensor for direct
sensing of biomolecular interactions Part I: Principle of operation and
associated instrumentation,” Biosens. Bioelectron., vol. 8, pp. 347–364,
1993.
[4] C. R. Graham, D. Leslie, and D. J. Squirrell, “Geneprobe assays on a fibreoptic evanescent wave biosensor,” Biosens. Bioelectron., vol. 7, pp. 487–
493, 1992.
[5] R. A. Ogert, J. E. Brown, B. R. Singh, L. C. Shriver-Lake, and F. S. Ligler,
“Detection of Clostridium botulinum toxin a using a fiber optic-based
biosensor,” Anal. Biochem., vol. 205, pp. 306–312, 1992.
[6] F. F. Bier and R. D. Schmid, “Real time analysis of competitive binding
using grating coupler immunosensors for pesticide detection,” Biosens.
Bioelectron., vol. 9, pp. 125–130, 1994.
[7] F. S. Ligler, J. P. Golden, L. C. Shriver-Lake, R. A. Ogert, D. Wijesuria,
and G. P. Anderson, “Fiber-optic biosensor for the detection of hazardous
materials,” Immunomethods, vol. 3, pp. 122–127, 1993.
[8] E. R. Goldman, E. D. Balighian, H. Mattousi, M. K. Kuno, J. M. Mauro,
P. T. Tran, and G. P. Anderson, “Avidin: A natural bridge for quantum dot
antibody conjugates,” J. Amer. Chem. Soc., vol. 124, pp. 68–78, 1987.
[9] S. K. Bhatia, L. C. Shriver-Lake, K. J. Prior, J. H. Georger, J. M. Calvert,
R. Bredehorst, and F. S. Ligler, “Use of thiol-terminal silanes and heterobifunctional crosslinkers for immobilization of antibodies on silica
surfaces,” Anal. Biochem., vol. 178, pp. 408–413, 1989.
[10] A. Ahluwalia, D. DeRossi, C. Ristori, A. Schirone, and G. Serra, “A
comparative study of protein immobilization techniques for optical immunosensors,” Biosens. Bioelectron., vol. 7, pp. 207–214, 1991.
[11] A. Subramanian, S. J. Kennel, P. I. Oden, K. B. Jacobson, J. Woodward,
and M. Doktycz, “Comparison of techniques for enzyme immobilization
on silicon supports,” Enzyme Microb. Technol., vol. 24, pp. 26–34, 1999.
[12] A. D. Kersey, M. A. Davis, H. J. Patrick, M. LeBlanc, K. P. Koo, C. G
Askins, M. A. Putnam, and E. J. Friebele, “Fiber grating sensors,” J.
Lightw. Technol., vol. 15, no. 8, pp. 1442–1463, Aug. 1997.
[13] K. O. Hill, Y. Fujii, D. C. Johnson, and B. S. Kawasaki, “Photo-sensitivity
in optical fiber waveguides: Application to reflection filter fabrication,”
Appl. Phys. Lett., vol. 32, p. 647, 1978.
[14] G. Meltz, W. W. Morey, and J. R. Dunphy, “Fiber Bragg grating chemical
sensor,” in Chemical, Biochemical, and Environmental Fiber Sensos III,
Proc. SPIE (Int. Soc. Opt. Eng.) 1587, 1991, pp. 350–361.
[15] V. Bhatia and A. M. Vengsarkar, “Optical fiber long-period grating sensors,” Opt. Lett., vol. 9, no. 21, pp. 692–694, 1996.
[16] A. Asseh, S. Sandgren, H. Ahlfeldt, B. Sahlgren, R. Stubbe, and G. Edwall,
“Fiber optical bragg grating refractometer,” Fiber Integr. Opt., vol. 17,
pp. 51–62, 1998.
[17] D. A. Pereira, O. Frazao, and J. L. Santos, “Fiber Bragg grating sensing
system for simultaneous measurement of salinity and temperature,” Opt.
Eng., vol. 43, no. 2, pp. 299–304, 2004.
[18] K. Schroeder, W. Ecke, R. Mueller, R. Willsch, and A. Andreev, “A fiber
Bragg grating refractometer,” Meas. Sci. Technol., vol. 12, pp. 757–764,
2001.
[19] G. Keiser, “Optical fiber communication,” ch. 2, 3rd ed. New York:
McGraw-Hill, 2000, p. 55.
[20] T. A. Birks and Y. W. Li, “The shape of fiber tapers,” J. Lightwave Technol.,
vol. 10, no. 4, pp. 432–438, Apr. 1992.
[21] M. P. Delisa, Z. Zhang, M. Shiloach, S. Pilevar, C. C. Davis, J. S. Sirkis,
and W. E. Bentley, “Evanescent wave long-period fiber Bragg grating as an
immobilized antibody biosensor,” Anal. Chem., vol. 72, pp. 2895–2900,
2000.
[22] H. Yi, L. Wu, R. Ghodssi, G. W. Rubloff, G. F. Payne, and W. E. Bentley,
“A robust technique for assembly of nucleic acid hybridization chips
based on electrochemically templated chitosan,” Anal. Chem., vol. 76,
pp. 365–372, 2004.
[23] H. Yi, L. Wu, J. J. Summer, J. B. Gillespie, G. F. Payne, and W. E. Bentley, “Chitosan scaffolds for biomolecular assembly: Coupling nucleic
acid probes for detecting hybridization,” New York, Wiley Interscience,
pp. 646–652, Jun. 2003.
Athanasios N. Chryssis (S’03) was born in Athens, Greece. He received the
B.S. degree in physics from National & Kapodistrian University of Athens in
2002. He is working toward the Ph.D. degree in electrophysics at the Department of Electrical and Computer Engineering, University of Maryland, College
Park.
He is currently working at the Photonic Switching and Integrated Optoelectronics Lab. as a Research Assistant at the University of Maryland at College
872
IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 11, NO. 4, JULY/AUGUST 2005
Park. His research interests include monolithic and hybrid integration of optoelectronics, biophotonics, and chemical and biological sensors.
Simarjeet S. Saini (SM’94–M’00) was born in Punjab, India, on April, 1975.
He earned the B.Tech. (Hons.) degree from the Department of Electronics and
Computer Engineering, Indian Institute of Technology, Kharagpur, India, in
1996 and the Ph.D. degree from Department of Electrical Engineering, the University of Maryland, College Park, in 2001.
He is currently working as a Research Associate in the Department of Electrical Engineering at the University of Maryland, College Park. He worked as Lead
Optoelectronic Engineer at Covega (formerly Quantum Photonics) from 2000 to
2004, where he was responsible for the design and development of Covega’s InP
optical chips. He has over 50 publications in various peer reviewed journals and
conferences and has written articles for international trade journals. He has two
U.S. patents granted and six patent applications. His research interests include
monolithic and hybrid integration of optoelectronics, biophotonics, chemical
and biological sensors, design and fabrication of high-power semiconductor
devices, bandgap engineering, and high-speed WDM devices and networks.
Dr. Saini is a member of the Optical Society of America and was a recipient of the International Society for Optical Engineering (SPIE) Educational
Scholarship in 1999. He also received a U.S. Army Research Fellowship from
1996–1998 for research and academic excellence.
Sang M. Lee received the B.S. degree in 1979, the M.S. degree in 1984, and the
Ph.D. degree in 1988 from the Department of Mechanical Engineering, Pusan
National University, Pusan, South Korea.
He is currently working as a Research Associate in the Department of Electrical and Computer Engineering at the University of Maryland, College Park, MD.
His research interests include fiber Bragg grating based chemical and biochemical sensors, biophotonics, sensor demodulation, and multiplexed sensor system
network. He has published publications journals and conferences related optical
sensors. He was on the faculty of Yonam College of Engineering, Chinju, South
Korea, from 1988 to 1995. He was a Visiting Scholar at Ohio Sate University
in 1990 and 1992 on leave from Yonam College of Engineering. From 1995
to 1999, was at the University of Maryland, College Park, and Korea Institute
of Science and Technology, Seoul, South Korea, where he developed optical
fiber Bragg grating and silica-based planar waveguide grating sensors. He also
developed the microfluidic bio-MEMS sensor at the University of Cincinnati,
Cincinnati, OH, from 1999 to 2001. Prior to joining the University of Maryland,
he worked as Senior Optical Process Development Engineer at Ciena Corp.
from 2001 to 2002 where he was responsible for optical process development
for optical test stations and WDM-based optical system integration and test.
Hyunmin Yi, photograph and biography not available at the time of publication.
William E. Bentley received the B.S. and M.Eng. degrees in chemical engineering from Cornell University, Ithaca, NY, and the Ph.D. degree from the
University of Colorado, in 1982, 1983, and 1989, respectively, all in chemical
engineering.
He is currently Director of the Bioengineering Graduate Program. Since
1989, he has been with University of Maryland’s Chemical Engineering
(now Chemical and Biomolecular Engineering) Department. He is the Herbert
Rabin Distinguished Professor of Engineering at the A. James Clark School of
Engineering, University of Maryland, College Park. He is author of over 125
publications in the general area of metabolic engineering and the production of
recombinant proteins in bacteria, insect cells, and insect larvae. He is cofounder
of an emerging protein production company, Chesapeake PERL, Inc., Savage,
MD.
Dr. Bentley is a Fellow of the American Institute for Medical and Biological
Engineering, and the American Association for the Advancement of Science.
Mario Dagenais (A’84–SM’88) received the Ph.D. degree in physics from the
University of Rochester, Rochester, NY, in 1978. He then spent two years at
Harvard University, Cambridge, MA.
He joined GTE Laboratories where he worked for seven years. Since 1987, he
has been with the University of Maryland, College Park, where he holds the position of Professor of Electrical and Computer Engineering. His research interests
are in the area of photonic switching, photonic integrated circuits, bio-sensing,
and optoelectronic packaging. In particular, he has been actively involved with
the development of Bragg grating bio-sensors, high power semiconductor laser
sources, tunable lasers, semiconductor optical amplifiers (SOAs), superluminescent LEDs, detectors, modulators, optical switches, and in the integration of
these components. He has published more than 200 papers.