U.S. patent application number 10/676352 was filed with the patent office on 2005-03-31 for double resonance interrogation of grating-coupled waveguides.
Invention is credited to Gollier, Jacques, Mozdy, Eric J., Piech, Garrett A..
Application Number | 20050070027 10/676352 |
Document ID | / |
Family ID | 34377369 |
Filed Date | 2005-03-31 |
United States Patent
Application |
20050070027 |
Kind Code |
A1 |
Gollier, Jacques ; et
al. |
March 31, 2005 |
Double resonance interrogation of grating-coupled waveguides
Abstract
A method for using a double resonance effect within a
grating-coupled waveguide (GCW) sensor, as generated from a light
beam with a given span of wavelengths or angles, is provided. The
method can be used for label-independent detection of biological
and chemical agents, to interrogate biological-binding events or
chemical reactions within a sensing region at increased
sensitivity, and with decreased sensitivity to environmental
perturbations. Also described is an optical interrogation system
incorporating the method.
Inventors: |
Gollier, Jacques; (Painted
Post, NY) ; Mozdy, Eric J.; (Elmira, NY) ;
Piech, Garrett A.; (Horseheads, NY) |
Correspondence
Address: |
CORNING INCORPORATED
SP-TI-3-1
CORNING
NY
14831
|
Family ID: |
34377369 |
Appl. No.: |
10/676352 |
Filed: |
September 30, 2003 |
Current U.S.
Class: |
436/518 |
Current CPC
Class: |
Y10S 436/805 20130101;
G01N 21/7743 20130101 |
Class at
Publication: |
436/518 |
International
Class: |
G01N 033/543 |
Claims
We claim:
1. A label-independent detection system for detecting biological or
chemical agents, the detection system comprises: 1) a substrate
surface having a sensing region with a bio- or chemo-responsive
layer; 2) an optical interrogation apparatus for monitoring said
bio- or chemo-responsive layer, said optical interrogation
apparatus comprising a grating-coupled waveguide structure, a light
source, an optical delivery system, and a detection instrument,
wherein more than one direction of propagation is used in said
waveguide to generate a sensor response for either a given angle or
wavelength.
2. The detection system according to claim 1, wherein for a given
angle or wavelength, two resonances exists as a result of light
propagation in two different, symmetrical directions in the
waveguide.
3. The detection system according to claim 1, wherein said sensor
response is generated simultaneously using more than one direction
of propagation.
4. The detection system according to claim 1, wherein said sensor
response is generated in sequence using more than one direction of
propagation.
5. The detection system according to claim 1, wherein an angular
shift as measured using both propagation directions as a function
of refractive index change greater than a sensitivity obtainable
from using only one direction of propagation.
6. The detection system according to claim 5, wherein an angular
shift as measured using both propagation directions as a function
of refractive index change improves interrogation signal-to-noise
sensitivity of said apparatus by a factor of at least about {square
root}2.
7. The detection system according to claim 1, wherein a spectral
shift as measured using both propagation directions as a function
of refractive index change improves an observed signal to noise
ratio in said system by a factor greater than that achievable from
using only one propagation direction.
8. The detection system according to claim 7, wherein a spectral
shift as measured using both propagation directions as a function
of refractive index change improves an observed signal to noise
ratio in said system by a factor of at least about {square
root}2.
9. The detection system according to claim 1, wherein signal from
different propagation directions are used to mitigate system
sensitivity to environmental perturbations.
10. The detection system according to claim 9, wherein a difference
in resonant peak locations is insensitive to an angular position of
said sensor.
11. The detection system according to claim 9, wherein the average
of resonant peak locations is insensitive to an angular position of
said sensor.
12. The detection system according to claim 1, wherein signal from
different propagation directions, together with mathematical
corrections for waveguide dispersion, are used to mitigate system
sensitivity to environmental perturbations.
13. The detection system according to claim 12, wherein an average
of resonant peak locations, modified by an appropriate waveguide
dispersion correction, is insensitive to an angular position of
said sensor.
14. The detection system according to claim 1, wherein said system
further includes an air-fluid delivery system, comprising either
macro or micro-fluidic passages designed to deliver biological or
chemical analytes to said sensing region.
15. A method of detecting biological or chemical agents, the method
comprises: providing a sensor system having a evanescent-field
sensing region comprising a substrate surface having at least a
bio- or chemo-responsive layer; generating a double resonance
within a grating-coupled waveguide of said system for either a
given angle or wavelength; exposing an individual sensing region to
an environment with analytes; and monitoring a response from said
sensor system.
16. The method according to claim 15, wherein an angular shift as
measured using both propagation directions as a function of
refractive index change doubles (2.times.) interrogation
sensitivity of said apparatus.
17. The method according to claim 15, wherein a spectral shift as
measured using both propagation directions as a function of
refractive index change improves an observed signal to noise ratio
in said system by a factor of at least about {square root}2.
18. The method according to claim 15, wherein said method uses
either a mean or difference of the resonance modes in a detection
system.
19. The method according to claim 15, wherein said substrate is
modified with one or more materials, which enhance stable
immobilization of said bio- or chemo-responsive layer.
20. A biosensor comprising: 1) a substrate surface having a sensing
region with a bio- or chemo-responsive layer; 2) an optical
interrogation apparatus for monitoring said bio- or
chemo-responsive layer, said optical interrogation apparatus
comprising a grating-coupled waveguide structure, a light source,
and an optical delivery system, wherein more than one direction of
light propagation is used in said waveguide to generate a sensor
response for either a given angle or wavelength, and a signal from
different propagation directions are used to mitigate sensitivity
to environmental perturbations.
21. The biosensor according to claim 20, wherein a spectral shift
as measured using both propagation directions as a function of
refractive index change improves an observed signal to noise ratio
in said system by a factor greater than that achievable from using
only one propagation direction.
22. The biosensor according to claim 20, wherein an angular shift
as measured using both propagation directions as a function of
refractive index change greater than a sensitivity obtainable from
using only one direction of propagation.
23. The biosensor according to claim 22, wherein an angular shift
as measured using both propagation directions as a function of
refractive index change improves interrogation signal-to-noise
sensitivity of said apparatus by a factor of at least about {square
root}2.
24. The biosensor according to claim 22, wherein an angular shift
as measured using both propagation directions as a function of
refractive index change doubles (2.times.) interrogation
sensitivity of said biosensor.
Description
FIELD OF INVENTION
[0001] The present invention pertains in general to a sensor used
for label-independent detection of biological and chemical agents.
More particularly, the invention relates to 1) a method for using a
double resonance effect within a grating-coupled waveguide (GCW)
sensor, as generated from a light beam with a given span of
wavelengths or angles, to interrogate biological-binding events or
chemical reactions within a sensing region at increased
sensitivity, and with decreased sensitivity to environmental
perturbations, and 2) an optical interrogation system incorporating
the method.
BACKGROUND
[0002] Evanescent field-based sensors are fast becoming a
technology of choice for accurate label-free detection of a
biological, biochemical, or chemical substance (e.g., cells,
spores, biological or drug molecules, or chemical compounds). This
technology typically involves using a grating-coupled waveguide
(GCW) to sense a concentration change, surface adsorption,
reaction, or the mere presence of a biological or chemical
substance at the waveguide surface. These detectable events are
manifested as a change in the effective refractive index of a
waveguide mode that partially or completely overlaps the sensing
region (waveguide substrate). To generate the evanescent or optical
field, an optical interrogation system uses optical elements, such
as a grating or prism, to couple a light beam from a light source
in and out of an optical mode in the waveguide of the GCW sensor.
The optical interrogation system also includes a detector that
receives the light beam coupled out from the waveguide. The angle
or wavelength of the emitted light beam is analyzed to determine
the effective refractive index of the waveguide. Changes in the
angle or wavelength of the probe light, for example, indicate
changes of the waveguide effective index that result from activity
at the sensor surface.
[0003] In determining the effective refractive index of the GCW
sensor, the principles of optical-physics dictate that the light
beam received by the detector had interacted with the waveguide
under a resonant condition, where the wavevectors of a diffraction
grating, incoming light beam, and guided mode all sum to zero,
thereby allowing one to probe the effective index of the mode,
which changes together with the surface index. This resonant
condition occurs only for a specific wavelength and angle of the
incoming light, and changes in this angle or wavelength correspond
to changes in the effective refractive index of the waveguide
caused by the concentration changes, surface adsorption, or
reactions of biological or chemical substances in the sensing
region of the GCW sensor. Thus, the optical interrogation system is
used to sense a change in the effective index of the GCW sensor
which enables one to determine whether or not a substance of
interest is located within the sensing region of the GCW
sensor.
[0004] For this technology to be viable, one must have an optical
interrogation system and in particular a detector capable of
accurately monitoring the resonant angle, the wavelength, or both.
In particular, the optical interrogation system must emit a light
beam that interacts with the GCW sensor, and must in turn receive
the light beam coupled-out of the GCW sensor and process that light
beam to detect in real time any changes in the resonant angle
and/or wavelength of the light beam. While there are many
approaches for accomplishing these tasks, each has unique
challenges associated with implementation, since the light beam
output from the GCW sensor may be relatively weak and the presence
of multiple sources of noise can degrade the light beam, especially
in high-throughput screening applications.
[0005] Evanescent- or optical-field sensors have demonstrated both
high sensitivity and an ability to detect binding reactions of as
little as about 250 Da molecular weight (e.g., biotin binding to
streptavidin). In recent years, the biological, pharmaceutical, and
other research communities have begun to recognize that optical
field-based sensors can be useful, high-throughput research tools
to measure a variety of biological or biochemical functions. GCW
sensors are particularly attractive for use in high-throughput
screening applications, where the absence of fluorescent tags and
the possibility of reduced false-negatives would provide a large
cost advantage. For this reason, microtiter well plates, also known
as microplates, have caught the attention of researchers as a
promising platform for such sensors, where 96 or 384 individual
wells provide the high-throughput access demanded by the industry.
When applied in the context of a microplate, the waveguide and
diffraction grating of the GCW sensor are preferably located in the
bottom of each well (e.g., the diffraction grating may be stamped
or otherwise molded into the well bottom, and the waveguide is
subsequently applied on top of the diffraction grating). The wells
themselves are typically composed of an optically transparent,
low-birefringence, preferably low-cost plastic that is typically
about several hundreds of microns to about a few millimeters thick.
Plates fabricated on glass substrates also are suitable for these
applications.
[0006] In the context of a high-throughput screening application,
microtiter well plates will be handled by various types of robotic
instruments. During the course of robotic manipulations fluids will
be added and removed from individual wells, assay protocols may
require incubation periods; hence, the microplate will likely as a
consequence be inserted and reinserted into the sensor or detection
device more than once during a single assay measurement cycle.
Since the resonance condition of the GCW sensor is critically
dependent on the angle of the light striking the microplate,
repositioning of the plate in the detection device will manifest
itself (at least in part) in the form of angular noise in the
sensor instrument. This environmental perturbation can in fact be
much larger than and overwhelm the sought-after response of the GCW
sensor to true biological or chemical changes in the sensing
region. Thus, this problem can work at cross-purposes with a sensor
that is designed purposefully to enhance or maximize sensitivity to
its biochemical environment. In other words, the sensor's
extra-sensitivity can exacerbate environmental background noise.
The simultaneous desire for an extremely responsive sensor with
high biochemical sensitivity and need for low susceptibility to
environmental responses place unique constraints on the system
designer. The present invention can balance these two competing
requirements, and the optical interrogation system, GCW sensor, and
method of the present invention successfully address and satisfy
this difficult problem.
SUMMARY OF THE INVENTION
[0007] The present invention relates in part to a method for
increasing the biochemical sensitivity of an evanescent- or
optical-field sensor having a grating-coupled waveguide (GCW)
structure, as well as decreasing its sensitivity to environmental
perturbations. The method, in part, involves generating a double
resonance effect by using more than one propagation direction in
the waveguide of the sensor from which is derived an output signal
from the sensor. One can produce a sensitivity or improved
signal-to-noise ratio (SNR) of greater than that obtained from
using only a single propagation direction for biological or
chemical sensors of grating-coupled waveguide (GCW) systems.
Moreover, appropriate use of the resonances from the separate
propagation directions can also aid in mitigating angular
misalignments of the GCW sensor.
[0008] Due to symmetry in the propagation direction of the
waveguide, a typical sensor device can provide two reflected angles
or wavelengths for a given input optical beam near normal
incidence. That is, for a given angle or wavelength, two resonances
can exist simultaneously in the waveguide as a result of light
propagation in two different, symmetrical directions. These
different signal propagation directions in the waveguide may be
excited simultaneously or in sequence by the detection
instrument.
[0009] With a wider field of interrogation, the two reflected
resonances will display opposite directional responses to index
changes in angle-space, while in wavelength-space the resonances
will display similar directional responses. By detecting both
resonances from an optical beam at either a given incident angle or
wavelength, one can calculate or derive indirectly the signal
sensitivity. The present technique takes advantage of either the
arithmetic mean or the difference of the resonance modes in a
detection system. The method can produce greater sensitivity than
that which is obtainable using only one propagation direction. When
much of the system noise on each resonance is common-mode, one can
approximately double (2.times.) the sensitivity of the sensor to
refractive index changes of a superstrate by employing the two
resonances together. In the case of uncorrelated noise on each
resonance, the sensitivity can be increase by a factor of about
{square root}2. In addition, because the two propagating modes are
symmetric about an incidence angle of zero degrees (normal to the
waveguide), one can ascertain information about the absolute angle
of the GCW sensor by considering both resonances together. In a
situation where the instrument monitors the resonance wavelength,
the average of the two resonance signals is insensitive to angle
shifts. In a situation where the detection instrument monitors the
angle of the resonances, the average signal will indicate the zero
angle position. Hence, the difference between the peaks also would
be insensitive to angle changes. Using the correct parameter in
each situation, the absolute angle of the GCW substrate (i.e.,
microplate) can be factored out or ignored.
[0010] Further, according to another aspect, the invention pertains
to a label-independent detection system that can exploit the double
resonance phenomenon. Such a system may use an evanescent- or
optical-field for detecting biological or chemical agents. An
evanescent- or optical-field sensor may comprises a substrate, a
diffraction grating, and a waveguide film, wherein the grating
and/or waveguide film form a waveguide, and an optical signal
propagated in more than one direction is used to derive an output
signal from the sensor. The detection system may comprise: a
substrate surface having a sensing region with a bio- or
chemo-responsive layer, and an optical interrogation apparatus for
monitoring the bio- or chemo-responsive layer. The optical
interrogation apparatus further includes a grating-coupled
waveguide structure, a light source, an optical delivery system,
and a detection instrument, wherein a light beam having more than
one direction of propagation is used in the waveguide to generate a
sensor response for either a given angle or wavelength. The
detection system may further include an air-fluid delivery system,
comprising either macro or micro-fluidic passages designed to
deliver biological or chemical analytes to the sensing region.
[0011] Alternatively, a method for using an evanescent-field or
optical-field sensor like that described for detecting biological
or chemical agents comprises: providing a sensor system having a
optical-field sensing region comprising a substrate surface having
at least a bio- or chemo-responsive layer; generating a double
resonance within a grating-coupled waveguide of said system for
either a given angle or wavelength; exposing an individual sensing
region to an environment with analytes; and monitoring a response
from the sensor system. The substrate can be modified with one or
more coatings or layers of materials with desired surface
chemistry, which enhance stable immobilization of said bio- or
chemo-responsive layer.
[0012] Additional features and advantageous of the present
invention will be revealed in the following detailed description.
Both the foregoing summary and the following detailed description
and examples are merely representative of the invention, and are
intended to provide an overview for understanding the invention as
claimed. Reference to the accompanying figures and the following
detailed description may convey a better understanding of the
present invention.
BRIEF DESCRIPTION OF THE FIGURES
[0013] FIG. 1 is a diagram of the basic components of an optical
interrogation system and GCW sensor in accordance with the present
invention.
[0014] FIG. 2 is a graph that illustrates the relationship between
the resonant angle and resonant wavelength of the GCW sensor shown
in FIG. 1;
[0015] FIG. 3 is a graph used to help describe how a spectral
interrogation approach can be used by the optical interrogation
system to determine the resonant wavelength of the GCW sensor shown
in FIG. 1.
[0016] FIG. 4 is a graph used to help describe how an angular
interrogation approach can be used by the optical interrogation
system to determine the resonant angle of the GCW sensor shown in
FIG. 1.
[0017] FIG. 5 is a graph that illustrates the resonant wavelength
(reflection anomaly) of an exemplary GCW sensor having a substrate
made from cyclo-olefin and a waveguide film made from
Ta.sub.2O.sub.5.
[0018] FIG. 6 is a graph illustrating the relationship between the
resonant angle and wavelength of an exemplary GCW sensor that has
two different cover indices.
[0019] FIG. 7 illustrates the concept of forward and backward
propagation in the GCW waveguide. FIG. 7A shows the convention for
forward (+) and reverse (-) propagation given certain input and
output directions. FIG. 7B shows the result of a mirror reflection
of FIG. 7A about a vertical axis (normal incidence axis), showing
the symmetry possible when considering input from both sides of the
sensor. FIG. 7C shows the doubly degenerate case when the
input/output beams strike the GCW sensor at normal incidence,
simultaneously exciting oppositely directed propagating modes.
[0020] FIG. 8 shows the mirror symmetry produced by reflecting the
theoretical curves of FIG. 6 about the zero-degree (normal
incidence) axis, again explicitly showing the resonance conditions
possible when striking the GCW sensor from both positive and
negative angles. Superimposed on this figure is a vertical line
segment representing the single-wavelength, multi-angular content
of a typical angular interrogation system (such as in FIG. 9A). The
circles show intersections of this optical beam with the resonance
curves, indicating the resonance locations for the system. As the
GCW superstrate index changes (due to a biological reaction, for
example), the resonances move from the dashed curves to the solid
curves, or vice versa. This demonstrates how the apparent sensor
response doubles compared to a single resonance system when the
difference between resonance locations is considered under this
scheme.
[0021] FIG. 9 shows two implementations of the optical detection
system for GCW sensors. FIG. 9A is a schematic of an angular
interrogation optical system, where a single-wavelength laser beam
is focused to generate a collection of angles on the GCW sample,
and the reflected beam is analyzed by a CCD camera. FIG. 9B is a
schematic of a spectral interrogation system, where a
multi-wavelength beam impinges on the GCW sample from a single
angle, and the reflected beam is analyzed by a spectrograph.
[0022] FIG. 10 shows the mirror symmetry produced by reflecting the
theoretical curves of FIG. 6 about the zero-degree (normal
incidence) axis, again explicitly showing the resonance conditions
possible when striking the GCW sensor from both positive and
negative angles. Superimposed on this figure is a horizontal line
segment representing the multi-wavelength, single-angle content of
a typical spectral interrogation system (such as in FIG. 7B). The
circles show intersections of this optical beam with the resonance
curves, indicating the resonance locations for the system. As the
GCW superstrate index changes (due to a biological reaction, for
example), the resonances move from the solid curves to the dashed
curves, or vice versa. This demonstrates how averaging could be
used to reduce the noise typically encountered in a single
resonance system, since two resonances move together under this
scheme.
[0023] FIG. 11 is a schematic representation of the angular
misalignment of the GCW sensor under the angular interrogation
instrument scheme. FIG. 11A shows an aligned sensor illuminated
with a cone of light (gray region), where the double resonances
appear on each side of the normal to the surface. FIG. 11B shows
the effect of misalignment, where the grayed lines indicate the
unaligned state; the resonances continue to appear on each side of
the normal, while the normal has simply shifted relative to the
input beam. The average position of the resonances has therefore
shifted, but their difference is unaffected. This insensitivity
scheme only applies when the angular misalignment is small compared
to the total angular extent of the input beam. As this figure
shows, when the tilt causes one of the resonances to move beyond
the boundary of the input light cone, this resonance will no longer
be excited by the optical system.
[0024] FIG. 12 depicts sample modes of a GCW structure used in the
example calculation of the dispersion correction factor necessary
to correctly insulate the GCW sensor from angular
misalignments.
[0025] FIG. 13 shows the n.sub.eff vs. .lambda. curve for the
experimental GCW structure used in the example calculation of the
dispersion correction factor necessary to correctly insulate the
FIG. 12 GCW sensor from angular misalignments.
[0026] FIG. 14 shows effective index dispersion
(dn.sub.eff/d.lambda.) vs. wavelength used in the example
calculation of the dispersion correction factor necessary to
correctly insulate the FIG. 12 GCW sensor from angular
misalignments.
[0027] FIG. 15 is a graph of a second order effective index
dispersion (d.sup.2n.sub.eff/d.lambda..sup.2) vs. wavelength used
in the example calculation of the dispersion correction factor
necessary to correctly insulate the FIG. 12 GCW sensor from angular
misalignments.
[0028] FIG. 16 shows the calculated residual wavelength error in
the average resonance under the double resonance scheme using
wavelength interrogation and the example sensor from FIG. 12.
[0029] FIG. 17 represents a typical single-resonance image. This
resonance was found using a He--Ne laser at a wavelength of
.lambda.=633 nm, and required use of an incidence angle of
.about.5', and the incoming cone of angles was .+-.0.7.degree..
[0030] FIG. 18 represents a double-resonance image generated
experimentally. Using a diode laser of .lambda.=660 run, one is
able to view the double resonance at near normal incidence
excitation of the waveguide.
[0031] FIG. 19 represents the sucrose data collected using a double
resonance, angular detection system. The image shows the result of
exposing the GCW superstrate to a sucrose series at 0-10-20%
concentration, where the gaps between each segment represents the
time required to pipette the solutions onto the sensor. The
location for the two separate peaks is shown, along with the peak
difference.
[0032] FIG. 20 represents the calculated baseline sensitivity over
time, showing a detailed close-up of the sample with 0%
concentration of sucrose (pure water) from FIG. 19, where the peak
difference pixel units have been converted to index of refraction
using the index data versus sucrose concentration of FIG. 19.
[0033] FIG. 21 represents an observed spectral double resonance.
The two peaks at about 883 and 893 nm, separated by about 10 nm,
are riding upon a large specular reflection background from a
superluminescent diode source (.about.50 nm bandwidth).
[0034] FIG. 22 shows the output of a GCW sensor interrogated with a
broadband, white-light optical spectrum, used to investigate this
GCW sensor's environmental stability against angular misalignment.
The four peaks represent two sets of double resonances, one set
each for TE (right) and TM (left) modes of the waveguide.
[0035] FIG. 23 shows the change or evolution of the spectral
location of each resonance peak shown in FIG. 22, as the GCW sample
is tilted relative to the input light beam. The outermost two
curves (A, and a') represent the TE double resonances, while the
innermost (B, top, and b', bottom) show the TM cases.
[0036] FIG. 24 is a graph showing the experimental average
resonance under spectral interrogation scheme as a function of
angle. Note that the average resonance location is not constant as
would be expected from simple theory. The dispersion of the
waveguide must be taken into account to explain the deviation.
DETAILED DESCRIPTION OF THE INVENTION
Section I--Definitions
[0037] Before describing the present invention in detail, this
invention is not necessarily limited to specific compositions,
reagents, process steps, or equipment, as such may vary. As used in
this specification and the appended claims, the singular forms "a,"
"an," and "the" include plural referents unless the context clearly
dictates otherwise. It is also to be understood that the
terminology used herein is for the purpose of describing particular
embodiments only, and is not intended to be limiting. All technical
and scientific terms used herein have the usual meaning
conventionally understood by persons skilled in the art to which
this invention pertains, unless context defines otherwise.
[0038] The term "air-fluid delivery system" as used herein refers
to a fluidic (i.e., gaseous or liquid) system that can collect
samples of biological or chemical analytes from the atmosphere or
surrounding environs, and deliver the samples to a sensor.
[0039] The term "analyte" or "target" as used herein refers to a
biological molecule or chemical entity, molecule, or compound to be
detected.
[0040] The term "biological molecule" or "biomolecule" refers to
any kind of biological entity, including, such as, modified or
unmodified nucleotides, nucleosides, peptides, polypeptides,
proteins, protein domains, fusion proteins, antibodies, membrane
proteins, lipids, lipid membranes, cellular membranes, cell lysate,
oligosaccharides, or polysaccharides, or lectins.
[0041] The term "bio-responsive" or "chemo-responsive" as used
herein refers to the ability to adsorb, desorb, react with, or bind
a biological or chemical species.
[0042] The term "bio- or chemo-responsive layer" as used herein
refers to a biological or chemical species, usually coated on a top
surface of a GCW sensor (in the superstrate, or sensing region, of
the sensor), used to promote binding, adsorption/desorption, or
reaction with the biological or chemical species to be
detected.
[0043] The term "evanescent" as used herein refers to that portion
of an optical field where the effective index of the optical field
exceeds the local index of the medium, thereby necessitating an
exponentially decaying field in space, according to Maxwell's
Equations.
[0044] The term "evanescent-field sensor" or "optical-field sensor"
as used herein refers generally to any kind of sensor where an
evanescent or confined optical field interacts with a medium to be
sensed, and changes in the refractive index or optical field can be
detected to indicate properties or changing characteristics of the
medium. The optical field confinement is typically accomplished
with a waveguide structure, where one or both extremeties of the
waveguide mode are evanescent fields.
[0045] The term "fluid" or "film of fluid" as used herein refers to
a material or medium that can flow such as a gas, a liquid, or a
semisolid.
[0046] The term "functionalization" as used herein relates to
modification of a solid substrate to provide a plurality of
functional groups on the substrate surface. The phrase
"functionalized surface" as used herein refers to a substrate
surface that has been modified to have a plurality of functional
groups present thereon. The surface may have an amine-presenting
functionality (e.g., .gamma.-amino-propylsilane (GAPS)) coating, or
may be coated with amine presenting polymers, such as chitosan or
poly(ethyleneimine).
[0047] The term "microspot" refers to a discrete or defined area,
locus, or spot on the surface of a substrate, containing biological
or chemical probe material. The corresponding microspots are
referred to as "probe microspots," and these microspots are
arranged in a spatially addressable manner to form a microarray.
One or more micropots, as in an array, constitute a sensing
region.
[0048] The terms "nucleoside" and "nucleotide" refer to moieties
which contain not only the known purine and pyrimidine bases, but
also other heterocyclic bases that have been modified. Such
modifications include methylated purines or pyrimidines, acylated
purines or pyrimidines, or other heterocycles. In addition, the
terms "nucleoside" and "nucleotide" include those moieties that
contain not only conventional ribose and deoxyribose sugars, but
other sugars as well. Modified nucleosides or nucleotides also
include modifications on the sugar moiety, e.g., wherein one or
more of the hydroxyl groups are replaced with halogen atoms or
aliphatic groups, or are functionalized as ethers, amines, or the
like. As used herein, the term "amino acid" is intended to include
not only the L-, D- and nonchiral forms of naturally occurring
amino acids (alanine, arginine, asparagine, aspartic acid,
cysteine, glutamine, glutamic acid, glycine, histidine, isoleucine,
leucine, lysine, methionine, phenylalanine, proline, serine,
threonine, tryptophan, tyrosine, valine), but also modified amino
acids, amino acid analogs, and other chemical compounds which can
be incorporated in conventional oligopeptide synthesis, e.g.,
4-nitrophenylalanine, isoglutamic acid, isoglutamine,
.epsilon.-nicotinoyl-lysine, isonipecotic acid,
tetrahydroisoquinoleic acid, .alpha.-aminoisobutyric acid,
sarcosine, citrulline, cysteic acid, t-butylglycine,
t-butylalanine, phenylglycine, cyclohexylalanine, .beta.-alanine,
4-aminobutyric acid, and the like.
[0049] The term "probe" refers to either a natural or synthetic
molecule, which according to the nomenclature recommended by B.
Phimister (Nature Genetics 1999, 21 supplement, pp. 1-60.), is
immobilized to a substrate surface. A probe may be either a natural
or synthetic, bio- or chemo-reactive molecule, which has been
immobilized to a substrate surface constituting part of a sensing
medium. Preferably, probes are arranged in a spatially addressable
fashion to form an array of microspots. A set of probes can bind or
otherwise react with analytes. Examples of probes which may be
employed according to this invention may include, but are not
limited to, antibodies, (e.g., monoclonal antibodies and antisera
reactive with specific antigenic determinants), glycolipids
including gangliosides, pharamaceutical or toxin molecules,
polynucleotides, peptide nucleic acid (PNA), peptides, proteins,
cofactors, lectins, polysaccharides, viruses, cells, cellular or
lipid membranes, membrane immuno-receptors, and organelles. For
chemical detection, the probes may include a polymer matrix, or a
ligand-gated ion channel membrane. Preferably, probes are arranged
in a spatially addressable manner to form an array of microspots.
When the array is exposed to a sample of interest, molecules in the
sample selectively and specifically binds to their binding partners
(i.e., probes). The binding of a "target" to the probes occurs to
an extent determined by the concentration of that "target" molecule
and its affinity for a particular probe.
[0050] The term "receptor" as used herein refers to a molecule that
has an affinity for a ligand. Receptors may be naturally-occurring
or man-made molecules. They may be employed in their unaltered
state or as aggregates with other species. Examples of receptors
which may be employed according to this invention may include, but
are not limited to, antibodies, monoclonal antibodies and antisera
reactive with specific antigenic determinants, pharamaceutical or
toxin molecules, oligonucleotides, polynucleotides, DNA, RNA,
peptide nucleic acid (PNA), peptides, polypeptides, protein
domains, proteins, fusion proteins, cofactors, lectins,
oligosacharides, polysacharides, viruses, cells, cellular
membranes, cell membrane receptors, and organelles. Receptors are
sometimes referred to in the art as anti-ligands. A
"ligand-receptor pair" is formed when two molecules have combined
through molecular recognition to form a complex.
[0051] The term "sample" as used herein relates to a material or
mixture of materials, typically, although not necessarily, in fluid
form, containing one or more components of interest.
[0052] The term "sensing region" as used herein refers to an area
or window on a surface of an optical-field sensor where analytes
may attach and be detected. Over the total surface of an
optical-field sensor, there is at least one, preferably a plurality
of sensing regions that may be each optically accessed in sequence
or in parallel. In other words, a sensing region is analogous to a
slide or a frame of film.
[0053] The term "substrate" or "substrate surface" as used herein
refers to a solid or semi-solid material which can form a stable
support and can function as an optical component of an
optical-field sensor. The substrate surface can be selected from a
variety of materials, including, for instance, glass,
glass-ceramic, metals, polymers, plastics, or combinations of
these.
[0054] The term "target(s)," "target moieties," "target analyte,"
"biological target," or "chemical target" refers to a solvated
particle, molecule or compound of interest in a sample that is to
be detected and identified. Suitable targets include organic and
inorganic molecules, biomolecules. In a perferred embodiment, the
target may be an environmental pollutant (e.g., such as pesticides,
insecticides, toxins, etc); a chemical (e.g., solvents, polymers,
organic materials, etc); a therapeutic molecule (e.g., therapeutic
and abuse drugs, antibiotics, etc); a biomolecule (e.g., hormones,
cytokines, proteins, peptides, protein domains, fusion proteins,
nucleotides, oligonucleotides, DNA, RNA, peptide nucleotide acids
(PNA), genomic DNA, lipids, lipid membranes, carbohydrates,
cellular membrane antigens, receptors or their ligands, etc); whole
cells (e.g., pathogenic bacteria, eukaryotic cells, etc); a virus;
or spores, etc.
Section II--Description
[0055] The present invention pertains, in part, to a sensor, sensor
system, and method of detecting analytes by means of using
evanescent- or optical-field waveguides. Reference to the
accompanying figures and the following detailed description may
convey a better understanding of the present invention. FIG. 1 is a
schematic representation of a generic optical-field sensor device.
An optical confinement layer, such as a grating-coupled waveguide
(GCW), provides a zone through which an optical mode propagates. A
grating-coupled waveguide structure can be designed to act as a
filter either in wavelength or angle-space, reflecting only a
particular, resonant wavelength or angle from an input optical
beam. GCWs offer good sensitivity to the surrounding environment or
medium (index). Grating coupled waveguides can provide a very
narrow spectral response to incident light. As a result, GCWs have
been applied to a wide variety of applications, including optical
filters, laser cavity mirrors, and biosensors, among others.
[0056] A light source populates a mode of the waveguide and
provides the evanescent optical field that penetrates into a medium
(superstrate) to be sensed. A change in the mass or refractive
index of the sensing medium causes a corresponding change in the
properties of the field in the optical confinement layer. In other
words, as the sensor surface encounters different biological or
chemical molecules, the evanescent field registers changes in
response, which is manifested as a change in the effective index of
the waveguide mode that can be monitored. For biosensing uses, the
optical response of the GCW varies as different biological or
chemical target species are brought into contact with the device.
Because the evanescent field extends into the medium, some of these
targets may adsorb or bind with probe materials on the surface of
the sensor or otherwise interact, thereby altering the refractive
index and the evanescent-field. Due to the physics of the
diffraction grating (or other coupling methods), an incident or
input probe light beam can be coupled into the waveguide to create
this confined optical energy (i.e., waveguide mode); and likewise,
the grating allows a subsequent output coupling of a light beam
from the waveguide that contains the desired information. This
information is typically in the form of either a change in the
wavelength or angle of the output light, since interaction between
the confined optical mode and the probe beam requires precise
matching of the wavelength and/or angle parameters as dictated by
the refraction/diffraction physics employed in either the prism,
grating, dielectric stack, or etc. to extract information from the
sensor. It should be noted that, while the evanescent tail is
typically the portion of the of the waveguide mode that interacts
with a biological sample, GCW sensors can be designed such that any
portion of the mode (guided or evanescent) can overlap a sensing
region. This invention applies equally to sensors that use any or
all parts of the confined optical energy for sensing.
[0057] GCW devices are used directly to monitor biological or
chemical assays in a label-free format, where the expense and
experimental perturbations of fluorescent dyes are completely
avoided. The sensitivity of the device is measured as the amount of
angular (wavelength) shift the reflected beam incurs for a given
cover index change. (See e.g., U.S. Pat. Nos. 6,455,004, 5,738,825,
4,952,056, or 4,815,843, to Tiefenthaler et al., the contents of
which are incorporated herein by reference.)
[0058] GCWs, however, should be optimized in terms of sensitivity
to biological or chemical analytes so as to enhance their
functional performance. According to the present invention, we
explain and describe a double resonance phenomenon, and methods for
exploiting the double resonance in order to either double the
sensitivity or improve the signal-to-noise of biological sensors.
The method improves instrument/sensor sensitivity by exploiting the
optical symmetry in the waveguide design. Moreover, the double
resonance can be useful to mitigate over sensitivity to undesirable
perturbations, such as angular or waveguide jitter in an optical
system. In other words, for a given index change, the basic
symmetrical design of GCWs can promote greater sensitivity of the
sensor for angular interrogation, while improving signal-to-noise
ratios by filtering background signal or rejecting environmental
noise.
[0059] A. Sensor Sensitivity
[0060] According to one aspect, the present invention provides an
evanescent-field optical sensor for detecting chemical, biochemical
or biological substances in a sample. As FIG. 1 shows, the sensor
100 includes a waveguiding structure 110 formed by a layer of
material (e.g., waveguiding film) 106 covering a substrate, wherein
the waveguiding film has a refractive index greater or higher than
the refractive index of the substrate; a diffraction grating 108
contained in or near the waveguiding structure 110; and a bio- or
chemo-responsive layer 103 covering the waveguiding film 106 in a
region around the diffraction grating 108, wherein the bio- or
chemo-responsive layer is capable of binding with the substances to
be assayed. Preferably, the bio- or chemo-responsive layer has a
thickness of less than about one wavelength. Preferably the
waveguide film is made of a dielectric material such as
Ta.sub.2O.sub.5, TiO.sub.2, TiO.sub.2-SiO.sub.2, HfO.sub.2,
ZrO.sub.2, Si, SiO.sub.2, Al.sub.2O.sub.3, Si.sub.3N.sub.4, HfON,
SiON, scandium oxides or mixtures thereof. The diffraction grating
108 is formed within a substrate 112 or waveguide film 106 by
embossing, holography, or other methods. The diffraction grating
108 can thereby be located above, below, or even within the
waveguide film 106. Moreover, a diffraction grating 108 need not be
in direct physical contact with a waveguide film 106, simply near
enough to cause optical influence on the waveguide mode.
Furthermore, due to effective-index waveguiding, the diffraction
grating itself can be fabricated with appropriately high enough
index to serve as the waveguide itself without the need for an
additional waveguide film deposition. The substrate is preferably
made of fused silica, a glass, or plastic material, such as
cyclic-olefin copolymer (COC). For example, the GCW sensor 100 can
have a cyclo-olefin substrate 112, which has an index n.sub.s=1.53,
a grating pitch .LAMBDA.=538 nm, a grating thickness is t.sub.g=10
nm, a waveguide index n.sub.f=2.01, a waveguide thickness
t.sub.f=110 nm, and a superstrate index that is nominally the index
of water (the solvent in which most experiments are performed,
n.sub.c.gtoreq.1.33). Of course, these particular values for the
physical properties are only nominal values, since all material
properties change with temperature and wavelength and other sensor
designs may employ different materials.
[0061] The presence of biological or chemical analytes 102 on or
near the sensing region alters the index of refraction at the
surface 104 of the GCW sensor 100. Thus, to detect the biological
analytes 102, the GCW sensor 100 is investigated with a light beam
126 emitted from the light source 122 and then a reflected light
beam 128 received at the detection system 124 is analyzed to
determine if there are any changes (.about.1 part per million) in
the refractive index caused by the presence of the biological
substance 102 in the sensing region 103 of the GCW sensor 100. In
one embodiment, the top surface 104 may be coated with biochemical
compounds (not shown) that only allow surface attachment of
specific complementary biological substances 102 which enables an
GCW sensor 100 to be created that is both highly sensitive and
highly specific. In this way, the optical interrogation system 120
and GCW sensors 100 may be used to detect a wide variety of
analytes 102. If the GCW sensors 100 are arranged in arrays they
may be used to enable high throughput drug or chemical screening
studies.
[0062] Furthermore, it should be appreciated that the interrogation
instruments are not necessarily limited to working with reflected
signal. An interrogation instrument can also work with waveguide
coupled light (peak) or transmitted light (dip). Aspects of these
light sources merely changes the location of the detection optics
with a reader instrument without changing the principle attributes
of the present invention, while providing possibly a means of
avoiding spurious optical signals, such as unwanted substrate
reflections, etc.
[0063] The sensitivity of the GCW sensor 100 may be best understood
by analyzing the structure of the diffraction grating 108 and the
waveguide 110. The physical operation of GCW devices can be
understood as an interaction between a free-space light field and
the environmentally-sensitive GCW waveguide modes. This interaction
is made possible by the diffraction grating, designed to diffract
light of specific wavelengths at specific angles to their incoming
propagation vectors. As will be described more fully herein, the
angular interrogation slope (AIS) or wavelength interrogation slope
(WIS) are primary measures of sensor response, depending upon the
particular parameters of experimental interrogation scheme. AIS and
WIS, respectively, refer to the amount of sensor angle or
wavelength shift that occurs in response to a unit change of
refractive index at the sensor surface.
[0064] In the GCW device, a particular wavelength incident at a
particular angle will diffract directly into the fundamental mode
of the waveguide, and propagate for some (short) distance. The
coupling between grating and waveguide preserves momentum, and
detailed mathematics can be found in the literature; for brevity,
we will simply mention that the difference in the real part of the
x-propagation coefficient between the free-space and waveguide mode
will equal the wavevector of the grating. The light beam 126 shone
on the diffraction grating 108 can only be coupled into the
waveguide 110 if its wave vector satisfies the following resonant
condition as shown in Equation No. 1:
k.sub.x'k.sub.x-.kappa. [1]
[0065] where k.sub.x' is the x-component of the incident wave
vector, k.sub.x is the guided mode wave vector, and .kappa. is the
grating vector. The grating vector .kappa. is defined as a vector
having a direction perpendicular to the lines of the diffraction
grating 108 and a magnitude given by 2.pi./.LAMBDA. where .LAMBDA.
is the grating period (pitch). This expression may also be written
in terms of wavelength .lambda. and incident angle .theta. as shown
in Equation No. 2: 1 2 n inc sin = 2 n eff - 2 [ 2 ]
[0066] Where .theta. is the angle of incidence of the light beam
126, n.sub.inc is the index of refraction of the incident medium,
.lambda. is the wavelength of the light 126, and n.sub.eff is the
effective index of refraction of the waveguide 110. The effective
index of the waveguide 110 is a weighted average of the indices of
refraction that the optical waveguide mode field or fundamental
mode "sees" as it propagates through the waveguide 110. The
fundamental mode preferably has a spatial extent that is wider than
the waveguide 110 itself, the extent depending on the refractive
index of the substrate 112. In particular, the fundamental mode has
an evanescent wave/tail that extends into the superstrate 103
(sensing region) which "sees" any surface changes created when the
biological substance 102 approaches or comes in contact with the
top surface 104 of the GCW sensor 100.
[0067] The previous expression shown in Equation No. 2 may be
rewritten in the more convenient form shown in Equation No. 3: 2
sin = n eff - [ 3 ]
[0068] which is the equation of a line where sin .theta. being the
y axis, .lambda. being the x-axis, .LAMBDA.n.sub.eff the
x-intercept, and -1/.LAMBDA. the slope. To obtain Equation No. 3,
n.sub.inc has been set to 1 so that it could be removed from
Equation No. 2. This approximation is used since air
(n.apprxeq.1.0003) is the most common incident medium. This
relation is pictured in the graph shown in FIG. 2. When a
biological substance 102 binds to the surface 104, the effective
index of the waveguide 110 is altered which leads to the shifting
the resonant wavelength or resonant angle of the GCW sensor 100.
This shifting can be seen as a shift of the x-intercept in the line
shown in FIG. 2.
[0069] In other words, 3 g - x = 2 [ 4 ]
[0070] where .beta..sub.g is the waveguide propagation constant,
.beta..sub.x is the free-space propagation constant parallel to the
grating vector and waveguide mode constant .beta..sub.g, and
.LAMBDA. is the grating period. The same grating that couples this
particular wavelength into the grating will also serve to couple
this light back out of the waveguide, according to the same
diffraction angle laws that governed the input coupling. The net
result is the angular redirection of a narrow wavelength band of
light incident on the GCW device. This narrowband response is often
referred to as a Wood anomaly. The design of the device determines
the input angle and wavelength for waveguide coupling, as well as
the output angle. This type of functionality is analogous to
directional optical filtration, with obvious applications wherever
optical filters are needed.
[0071] GCW devices with an ability to tune the location, in both
spectral wavelength and angle, of the above resonance with the
index of refraction of the waveguide superstrate are useful in
biosensors. As mentioned before, typically the evanescent tail of
the propagating waveguide mode senses the superstrate index
changes, thereby altering the guided mode's effective index. This
changes the resonance condition of the GCW according to Equation
No. 4, above, and the resonance thus shifts to a new wavelength or
angle location.
[0072] Referring to FIG. 1, the resonant condition (e.g., resonant
wavelength or resonant angle) of such a GCW 100 may be interrogated
to determine refractive index changes by observing the reflected
light 128 from the GCW 100. There are two different modes of
operation for monitoring refractive index changes--spectral
interrogation or angular interrogation. In spectral interrogation,
a nominally collimated, broadband beam of light 126 is sent into
the GCW 100 and the reflected light 128 is collected and monitored
with a spectrometer 124 (for example). By observing the spectral
location of the resonant wavelength (peak), one can monitor binding
or refractive index changes on or near the surface 104 of the GCW
100. The spectral interrogation concept is graphically represented
in the graph shown in FIG. 3. Conversely, in angular interrogation,
a nominally single wavelength of light 126 is focused to create a
range of illumination angles and directed into the GCW 100. The
reflected light 128 is monitored with a CCD camera or other optical
detector 124. By monitoring the position of the resonant angle
reflected by the GCW 100, one can monitor binding or refractive
index changes on or near the surface 104 of the GCW 100. The
angular interrogation concept is graphically represented in the
graph shown in FIG. 4.
[0073] In order to maintain simplicity and efficiency of operation,
the devices employed for biosensing are usually designed such that
only the zeroth diffracted orders of the grating propagate in free
space, while what would be the .+-.1 orders couple to the
fundamental mode of the waveguide. The higher diffraction orders
are avoided by designing a sub-wavelength grating (i.e., grating
pitch is smaller than the desired operating wavelength). The
coupling efficiency is large since multiple orders do not remove
power from the system. Moreover, since only the zeroth reflected
and transmitted beams exist in free space, the GCW can thereby
produce nearly total reflection or transmission of the desired
(anomalous) wavelength. FIG. 5 shows a GSOLVER (rigorous
coupled-wave analysis, or RCWA code) analysis of the structure of
the device illustrated in FIG. 1, when the input light angle is
about 3.degree.. The reflected beam (at 3.degree. from the normal)
occurs in the vicinity of about 824 nm for incident TE light and at
a cover index of 1.33 (water).
[0074] As mentioned above, GCW sensors 100 are used in biosensing
applications because they enable one to determine the location of
the resonance angle/wavelength 502 which enables one to calculate
the refractive index of the superstrate 103 and then determine
whether or not a biological substance 102 is located on the GCW
sensor 100. This is all possible because the evanescent tail of the
propagating fundamental mode in the waveguide 110 senses index
changes in the superstrate 103 caused by the presence of the
biological substance 102. The index change in the supersrate 103
changes the resonance condition of the GCW 100 according equation
no. 1 and then the resonance 502 shifts to a new wavelength or
angle location. The location of the shifted resonance indicates the
current index of the superstrate 103 which indicates whether or not
the biological substance 102 is in the superstrate 103 of the GCW
100. It has been shown that the resonance 502 can shift hundreds of
nanometers for a unit change in the refractive index of the
superstrate 103 (see FIG. 2). The relationship between angle and
wavelength is displayed in the graph of accompanying FIG. 6, for a
BIOS-1-GCW sensor 100. The different curves show behavior for both
TE and TM polarizations for two different cover indices.
[0075] The curves displayed in FIG. 6 were calculated using
rigorous coupled-wave analysis (RCWA), and present the unique
solution in wavelength and angle space for a given input beam
orientation. At any given wavelength and input beam orientation,
the optical energy will couple to only one propagation direction in
the waveguide, depicted by the solid line waveguide mode in FIG.
7A. Positive angles in FIG. 6 represent this solid (+) mode of FIG.
7A: an incident beam whose propagation vector x-component is in the
same direction as the waveguide mode, equivalent to .beta..sub.g
and .beta..sub.x having the same sign in Equation No. 4. Likewise,
negative angles in FIG. 6 represent an incident beam propagation
vector x-component oriented in the opposite direction relative to
the waveguide mode; this is depicted in FIG. 7A by the dashed (--)
waveguide mode, and is equivalent to .beta..sub.g and .beta..sub.x
having opposite signs in Equation No 4. While both directions are
displayed in FIG. 7A, it should be emphasized that these
oppositely-directed waveguide modes do not necessarily exist
simultaneously for the same input wavelength; FIG. 6 should again
be consulted for the proper wavelengths given inverse angles.
[0076] The single solution angle/wavelength curves of FIG. 6 can be
reflected about the zero-angle axis to produce the mirror
solutions. One may rewrite Equation No. 4 in terms of the incident
(or reflected) angle .theta. and the effective index of the
waveguide n.sub.eff: 4 sin = n eff - [ 5 ]
[0077] The "double" resonance condition for the grating sensor
arises from the fact that there is a second resonance condition
obtained by setting .theta. equal to -.theta. in the above
equation. As will be described in detail later, this second
resonance arises from the symmetry of the sensor: 5 sin = - n eff +
[ 6 ]
[0078] Since the sensor is symmetrical in the horizontal direction,
the geometry of FIG. 7A can be reflected about a vertical line to
produce its mirror image, resulting in duplicate positive-angle,
waveguiding directions as shown in FIG. 7B. This situation is
easily realized in the laboratory by launching a converging or
diverging optical beam that contains enough angular spread to
satisfy both the positive and negative conditions simultaneously,
or by launching two separate input beams into the sensor device at
the appropriate (mirror) angles. The limiting case of this geometry
is the auto-symmetric normal-incidence launch depicted in FIG. 7C.
In this case, a single input beam is able to excite both
waveguiding directions simultaneously.
[0079] A double resonance concept can be applied for both angular
and wavelength-based interrogation methods. On one hand, if one
assumes that the resonant coupling angle is the observable, and
that wavelength is fixed, then one can subtract Equations Nos. 5
and 6 to obtain: 6 difference sin + - sin - = 2 ( n eff - ) [ 7
]
[0080] Equation No. 7 represents the angular "difference"
observable that may be used as an effective index measure in the
angular interrogation system. As was mentioned, the signal is the
difference between the two resonant angles. As the cover index
changes, the two peaks will either move together or apart (the
wavelength does not change in the angular interrogation scheme). By
observing both the positive and negative resonances, one doubles
the amount of signal available, and if the noise in each
measurement is uncorrelated, the improvement in signal to noise
ratio is about a factor of {square root}2.about.1.414213. For
example, FIG. 8 illustrates the phenomenon by an angular
interrogation scheme for two different superstrate indices of
refraction.
[0081] In a system for an angular interrogation device, such as
depicted in FIG. 9A, a collection of angles is incident upon the
sensor at a single wavelength. This corresponds to the
finite-length vertical lines .alpha. and .beta. in FIG. 8, where
the length of the line represents the angular extent of the
incoming beam. These lines graphically illustrate the basic core of
the present double resonance concept. That is, for a given input
wavelength, two resonances exist simultaneously due to the
possibility of two propagation directions in the waveguide. One
should ensure that the incoming optical beam contains enough
angular or wavelength spectrum to excite both propagation
directions, either simultaneously or in sequence. For instance, the
dashed line (n=1.45) of FIG. 8 represents data which suggest that a
662 nm input beam focused to contain at least .+-.2.degree. will
excite simultaneously both resonances for a given superstrate
index.
[0082] According to desired operations of the sensor, one should
focus attention on how the resonance locations change with index
changes. This is demonstrated by observing the movement of the
resonances when shifting from the green curves to the solid line
(n=1.33 superstrate index). The arrows in FIG. 8 indicate that the
resonances move in opposite directions relative to each other when
the superstrate index is varied. Recalling that the sensor was
designed to target a specific AIS (angular interrogation slope--the
angular shift of the resonance for given superstrate refractive
index change) for any single resonance, having two resonances move
equal distances in opposite directions implies a doubling of the
device sensitivity as per Equation [7], thus generating an
improvement in the signal-to-noise of about {square root}2 when
each peak suffers from uncorrelated (Gaussian) noise.
[0083] In wavelength-based interrogation, the input beam is
directed at a unique angle, but contains a band of different
wavelengths. The analysis of this scenario is depicted in FIG. 10,
where a horizontal line corresponds to single-angle,
multi-wavelength interrogation. As FIG. 10 illustrates, under
wavelength or spectral interrogation the double resonances shift in
parallel with each other. In other words, the foci of the
resonances move in the same direction as each other.
[0084] In order to analyze this spectral system, one can again
subtract Equations Nos. 5 and 6 (where the angle is now constant
and the wavelength varies) to obtain: 7 mean + + - 2 = n eff [ 8
]
[0085] The meaningful signal for wavelength interrogation is the
mean (average) wavelength of the two spectral resonances. As the
cover index changes, the mean value of the two spectral peaks will
either shift to higher or lower wavelength. For instance, with
regard to a general Gaussian-distributed noise source, the
improvement in signal-to-noise averaging occurs as the square root
of n ({square root}n), where n is the number of averages. This
calculation implies that, for wavelength interrogation, just as in
the case of the angular measurement, one can improve the
signal-to-noise-ratio by a factor of about {square root}2 for
uncorrelated noise.
[0086] B. Robustness Against Environmental Noise
[0087] We note that the use of the double resonance enables either
the angular or spectral-wavelength sensors to reject certain types
of noise. Most prominent among external noise sources is the
absolute angular position of the GCW sensor in the detection
instrument. In particular, given a fixed optical beam, how will the
detected resonance signal vary when the GCW sensor is tilted? The
angular interrogation system will be insensitive to this noise
provided the angular misalignment does not exceed the total angular
content of the input beam, since both resonances move the same
direction when the sensor is tipped, and the difference in angular
peak locations is therefore constant. This is demonstrated
schematically in FIG. 11 are affected equally, thus canceling out
the each other.
[0088] Analogously, for the spectral sensor, the observation of the
mean of the double resonance signal will allow one to reject any
angular jitter in the apparatus, since this type of jitter will
cause the peaks to move together or apart, but will not affect
their mean value (see FIG. 10). This angular jitter is an example
of common mode noise that the double resonance design will robustly
reject, possibly improving the observed signal to noise in the
system by much more than square root of (2), depending on the
magnitude of the various noise sources.
[0089] According to these characteristics of the invention, signals
from different propagation directions can be employed to mitigate
the system's sensitivity to extraneous environmental perturbations.
In other words, the difference present in resonant peak locations
and/or the average pf resonant peak locations are made less
dependent on the vagaries of substrate placement and less sensitive
to angular position of the sensor, while heightening the ability of
the detection instrument to sense the true signals. Similarly, the
differences present in resonant peak locations and/or the average
of resonant peak locations are insensitive to wavelength noise.
[0090] C. Deviations from Simple Theory
[0091] The analysis of the above section, particular the results in
Equations [7] and [8], are somewhat simplified for ease of
understanding. A more detailed analysis starting with equations [5]
and [6] can indicate what other factors must be considered for
robust, accurate use of the double resonance technique.
[0092] For the angular measurement system, a single wavelength is
used, and the angles used to couple into the waveguide are equal
and opposite (.theta..sub.1=-.theta..sub.2), resulting in equal
waveguide effective indices. Equations [5] and [6] therefore
become: 8 sin 1 = n eff - sin 2 = - n eff + [ 9 ]
[0093] The difference of these two equations is therefore 9 1 - 2 =
sin - 1 ( n eff - ) - sin - 1 ( - n eff + ) observable = 2 sin - 1
( n eff - ) [ 10 ]
[0094] meaning that one must first take the sin of 1/2 of the
observable parameter to deduce true changes in superstrate index
(the wavelength and grating pitch do not change during the course
of a measurement). This still displays the doubled response
characteristic of the simplified analysis and reduces to the
simplified result of Equation [7] in the case of small angles.
[0095] In the case of the wavelength interrogation system, the
input/output angles are fixed, while the wavelength of each
separate resonance is different. Likewise, due to the different
wavelengths, the effective indices of the separate waveguide modes
are different. Reflecting this in equations [5] and [6], 10 sin = n
eff , 1 - 1 sin = - n eff , 2 + 2 [ 11 ]
[0096] Next, we define the observable parameter of interest, the
average wavelength: 11 a = 1 + 2 2 [ 12 ]
[0097] It is the derivative of this quantity with respect to angle
that the double resonance technique hopes to minimize and thereby
make the sensor more tolerant to angular misalignments.
[0098] Next, since the two resonance wavelengths change when the
angle changes (according to Equation [11]), the dispersion in the
waveguide system will cause the effective indices to likewise
change. Since the average wavelength is expected to vary slowly, we
expand the effective indices around this average wavelength, 12 n
eff = n a + n eff | a + 2 2 2 n eff 2 | a + = n a + D a + 2 2 D a '
+ [ 13 ]
[0099] where D.sub.a is the effective index dispersion evaluated at
.lambda.=.lambda..sub.a, and the term .DELTA..lambda. has opposite
sign for the two resonances:
.DELTA..lambda..sub.1=.lambda..sub.1-.lambda..sub.avg
.DELTA..lambda..sub.2=.lambda..sub.avg-.lambda..sub.2=-.DELTA..lambda..sub-
.1 [14]
[0100] We can now substitute the effective index expansion [13] (to
second order) into each resonance condition [11], and use these in
the average wavelength expression [12] 13 a = 2 ( n a + 1 D a + 1 2
2 D a ' - sin + n a - 1 D a + 1 2 2 D a ' + sin ) = ( n a + 1 2 2 D
a ' ) [ 15 ]
[0101] Taking the derivative of the above expression with respect
to angle, we obtain the parameter of interest: 14 a = 1 D a ' ( 1 -
a ) [ 16 ]
[0102] In order to evaluate this expression, we must evaluate the
individual resonance .lambda..sub.1 derivative, 15 1 = ( n eff1 -
cos ) = ( [ n a + 1 D a + 1 2 2 D a ' ] - cos ) = ( D a ( 1 - a ) +
1 D a ' ( 1 - a ) - cos ) [ 17 ]
[0103] Collecting terms on the left-hand side, we can evaluate this
derivative: 16 1 [ 1 - D a - 1 D a ' ] = - a [ D a + 1 D a ' ] -
cos [ 18 ] 1 = cos + a ( D a + 1 D a ' ) D a + 1 D a ' - 1
[0104] The derivative of the resonance average is therefore 17 a =
1 D a ' ( cos + a ( D a + 1 D a ' ) D a + 1 D a ' - 1 - a ) = 1 D a
' ( cos D a + 1 D a ' - 1 - a ( D a + 1 D a ' D a + 1 D a ' - 1 - 1
) ) = 1 D a ' ( cos D a + 1 D a ' - 1 - a ( 1 D a + 1 D a ' - 1 ) )
[ 19 ]
[0105] Now we can again arrange terms on the left-hand side to
solve for the derivative: 18 a ( 1 1 D a ' - 1 D a + 1 D a ' - 1 )
= cos D a + 1 D a ' - 1 a = cos D a + 1 D a ' - 1 1 D a ' - 1 = cos
D a - 1 1 D a ' = 1 D a ' cos D a - 1 [ 20 ]
[0106] Hence, we have the average resonance drift with respect to
angle expressed in terms of the average effective index dispersion
and its derivatives, the grating pitch (.LAMBDA.), and
.DELTA..lambda..sub.1, the 1/2 difference in resonance wavelengths.
Since .DELTA..lambda..sub.1, involves both .lambda..sub.1 and
.lambda..sub.a, we need to re-express this quantity in terms of
only average values that drift much more slowly with changes in
angle, 19 1 = 1 - a = - ( 2 - a ) = ( n eff1 - sin ) - a ( n a + 1
D a - sin ) - a 1 [ 1 - D a ] = ( n a - sin ) - a 1 = n a - sin - a
1 - D a [ 21 ]
[0107] We now have an expression for .DELTA..lambda..sub.1 in terms
only of average parameters that vary slowly. Substituting this
value into the expression for the average wavelength drift [20], we
arrive at the final expression for the drift: 20 a = cos D a ' ( n
a - sin - a 1 - D a ) D a - 1 = cos D a ' ( sin + a - n a ) ( D a -
1 ) 2 [ 22 ]
[0108] Equation [22] shows that, in order for a system designer to
use the double resonance technique in the spectral interrogation
system embodiment, they may need to consider and compensate for the
dispersion of the waveguide. In fact, depending upon the system,
one may even need to consider higher-order terms in the Taylor
series used in Equation [13].
[0109] In order to gain some understanding of this correction in
the spectral interrogation approach, one can evaluate the value of
the average wavelength error [22] using a model sensor:
1 n.sub.sub n.sub.wg t.sub.wg [nm] .LAMBDA. [nm] t.sub.g [nm]
.theta. [.degree.] 1.527 2.07 181 500 50 1.9
[0110] where n.sub.sub is the substrate index, n.sub.wg is the
waveguide index, t.sub.wg is the waveguide thickness, A is the
grating pitch, t.sub.g is the grating thickness, and .theta. is the
nominal angle the input light beam makes with the GCW sensor.
[0111] This information is first used to calculate the modes of the
structure, seen in FIG. 12. These are the modes as calculated at an
operating wavelength of 830 nm. In the case of a double resonance,
we will have two resonances from opposite propagation directions
with different wavelengths. For the device of this example
calculation with the input angle above, the resonances are equally
spaced about an average wavelength value of .about.801 nm (TM) and
.about.857 nm (TE) (this value depends upon the GCW design as well
as the choice of input angle, etc.). We therefore need to know the
effective index values for the modes at these average wavelengths
in order to evaluate [10]. In addition, we need to know two orders
of dispersion (derivatives of the effective index vs. wavelength
curve) at these average values. In order to evaluate these
derivatives, we have to calculate the effective index as a function
of wavelength; this data is shown in FIG. 13, as calculated using
the device specifications above. With the curve in FIG. 13 (and two
orders of derivatives shown in FIGS. 14 and 15), the average
wavelength drift can then be evaluated as a function of angle,
shown ultimately in FIG. 16.
[0112] FIG. 16 shows the existence of some residual average
wavelength shift that must be accounted for in exact application of
the double resonance scheme for angular insensitivity. At zero
angle, the two wavelengths are degenerate, and the wavelength error
is therefore zero. As the angle grows however, the two resonance
wavelengths separate, and according to Equation [22] the effect of
dispersion is to cause the average resonance wavelength to shift
from the ideal center position. The TM correction is larger than
the TE by about a factor of 2.3, mainly because the TM mode is
closer to cutoff, and therefore has more wavelength dispersion. At
an incidence angle of 1.9.degree. for example, the TE wavelength
error is 18.6 pm/.degree., whereas the TM wavelength error is 43.6
pm/.degree..
[0113] The following empirical section will demonstrate
experimentally the observation of this average resonance error.
Section III--Empirical
[0114] A. Sensitivity Improvement
[0115] In the laboratory, we have demonstrated the double resonance
concept and other principles of the present invention using both a
commercially-available BIOS-1 sensor and sensors fabricated by
Corning Incorporated. FIG. 17 shows an image of a typical single
GCW resonance under an angular interrogation scheme as bright
vertical streak. This resonance was found using a HeNe laser where
the wavelength (.lambda.=633 nm) required the use of a relatively
high angle of .about.5.degree., and the incoming cone of angles was
only .+-.0.7.degree.. In order to view the double resonance, a
.lambda.=660 nm diode laser was employed, allowing near normal
incidence excitation of the waveguide. FIG. 18 shows the
experimental double resonance, in this case consisting of vertical
shadows due to the presence of a large background reflection.
[0116] Aside from a simple demonstration that the double resonance
is feasible, an experiment was performed with a series of sucrose
solutions of varying refractive index applied to the sensor
surface. This study was performed to address a concern that, while
the symmetric peaks double the response to superstrate index
changes, the noise of the system may likewise be doubled. If the
peak detection is noise limited, and if both peaks contain
uncorrelated Gaussian noise, then one would expect the peak
separation noise variance to be two-times that of either peak
separately. In such a situation, the reduction of the noise would
be greater than single resonance operation by a factor of about
{square root}2 (since the standard deviation is the square root of
the variance). If the noise in each peak is partially or completely
correlated, the resulting signal-to-noise improvement can exceed a
factor of {square root}2. The noise associated with difference in
signal can be less than the noise of either individual peak. A
worst possible case is when the noise of each peak is
anti-correlated, possibly negating the signal to noise benefit due
to the double resonance technique, since the noise can double along
with signal.
[0117] The angular-based experiment involved monitoring the
separation between the two resonances as a function of time. Since
each resonance responds according to the device's AIS, the peak
separation is the measure of the relative index of refraction of
the superstrate. If the change in the peak separation per unit
index change is called .delta.p, and the standard deviation of the
time-resolved peak location for a given constant cover index is
used to quantify the system noise, then the minimum resolution of
the systems can be experimentally calculated as: 21 n = Noise [
Pixels ] p [ Pixel / RIU ] [ 23 ]
[0118] where the peak location difference is measured in pixels on
a CCD camera. The resonant wavelength or angle manifests itself as
a narrow band of wavelength or angles on a screen through the CCD
detection. FIG. 19 shows the result of a 0-10-20% sucrose series,
where the gaps between each segment represents the time required to
pipette the solutions onto the sensor. The location for the two
separate peaks is shown, along with the peak difference. It is
immediately apparent that the difference peak moves twice as far as
either peak individually, confirming the index sensitivity
hypothesis.
[0119] FIG. 20 shows a close-up view of the 0% sucrose (pure water)
time history from FIG. 19, where the pixel units at the peak
difference have been converted into index of refraction units using
the index data versus sucrose concentration of FIG. 19. FIG. 20
shows that the baseline sensitivity is quite good, and the standard
deviation of the data yields a detection limit of about
7.times.10.sup.-6 refractive index unit (RIU). To quantify the
impact of the peak difference operation on the noise of the system,
the standard deviation (C) of the two peaks separately were also
calculated. This permits one to compare the variances of each data
set. If the noise on each peak reflects uncorrelated Gaussian
noise, then the variances should be additive as a result of the
difference operation. If the difference variance, however, is less
than the sum of the variances of each peak, then much of the noise
would be correlated (common-mode) noise, and the differencing
operation would significantly improve the noise of the system.
Table 1 presents the data about variance for both peaks and their
difference in the sucrose series.
2TABLE 1 Variance of Peaks Left Right Difference 6.28 .times.
10.sup.-3 1.43 .times. 10.sup.-2 4.95 .times. 10.sup.-3
[0120] This data clearly shows that much of the noise from each
peak was common-mode, since the difference variance is in fact less
than either peak individually. This phenomenon permits one to
achieve excellent minimum sensitivity and performance.
[0121] The observation of the angular difference, or the spectral
mean, allows one to create a reader system that more fully utilizes
the resonant energy, has enhanced sensitivity, and has an increased
robustness to specific forms of common mode noise. An example of
this spectral double resonance is shown in FIG. 21. Just as in the
angular case, one can analyze the system noise to show noise
improvement from observing the mean wavelength signal. Table 2
presents a summary of the range and mean variance values in terms
of pixels for both spectral peaks.
3TABLE 2 Variance of Peaks (pixels) Lower Upper Mean .0241 .021
.012
[0122] B. Environmental Insensitivity Improvement
[0123] Using a very broadband spectral interrogation system, the
reduction of environmental noise was tested, particularly the
misalignment of the GCW sensor angle. FIG. 22 shows the two sets of
double resonances (TM, two left; TE, two right) resulting from the
sensor. As the plate was tilted, the location of each resonance was
monitored, tracing out the curves of FIG. 23. This graph shows the
susceptibility of GCW sensor systems to such misalignment: one
degree of sensor tilt results in over 5 nm of resonance location
change, whereas typical systems are capable of 0.1 pm resolution! A
one-degree tilt therefore exceeds the required system noise limit
by a factor of 50,000.
[0124] Using the methodology of the double resonance, the two TE
(or TM) resonance positions were subtracted from each other to
correct this angular misalignment. FIG. 24 shows the result of this
calculation, resulting in a factor of 50 reduction of the spectral
deviation of the observable parameter for a one degree
misalignment. If the misalignment were entirely corrected, these
curves would be horizontal lines, unwavering with angular detuning.
The downward trend of the graphs however indicates the presence of
the dispersion term indicated by Equation [15] for the spectral
system. In order to flatten these curves to an acceptable level
(for this particular sensor), the dispersive term should be
accounted for in this measurement.
[0125] C. Dispersion Correction Comparison
[0126] As discussed in Section II, Part B above, the double
resonance scheme can be used to make the system insensitive to
environmental perturbations, but under the spectral interrogation
scheme one should take waveguide dispersion into account for
accurate compensation. The data from FIG. 24 clearly supports this
conclusion.
[0127] From FIG. 24, we can calculate the slope of the average
resonance curves and compare with the theory of Section II, Part B.
For the TE mode, the experimental slope is .about.64 pm/degree
while for the TM mode the slope is .about.184 pm/degree. While
these values are somewhat higher than the predictions above, it is
expected that the experimental system added some offset to the
values. For example, the GRIN lenses used to deliver the light from
the optical fiber source to the GCW sensor are known to have some
chromatic aberration that causes similar shifts of resonance
wavelength as angle is varied. On the whole, the experimental data
above supports the theory quite well, especially with respect to
the zero-shift at zero-angle feature, as well as the factor of
.about.2.8 difference in slopes between the TE and TM curves.
[0128] In summary, the double resonance technique appears to be
useful for both angular and spectral interrogation of resonant
waveguide gratings sensors. The present invention has been
described both in general and in detail by way of examples. Persons
skilled in the art will understand that the invention is not
limited necessarily to the specific embodiments disclosed.
Modifications and variations may be made without departing from the
scope of the invention as defined by the following claims or their
equivalents, including equivalent components presently known, or to
be developed, which may be used within the scope of the present
invention. Hence, unless changes otherwise depart from the scope of
the invention, the changes should be construed as being included
herein.
* * * * *