SINGLE MODE FIBER OPTIC EVANESCENT WAVE REFRACTOMETER AND METHOD OF IMMMUNOASSAY MEASUREMENT OF A TARGET COMPONENT IN A FLUID
Field of the Invention
This invention is in the field of chemical and biological assay and is useful for measurement of variables which are addressable through sensitive measurements of the index of refraction. Brief Description of the Prior Art
Evanescent wave spectroscopy is not new. It is a variant of internal reflection spectroscopy and many configurations are thoroughly described in a text by N.J. Harrick. The concept works generally through the illumination of an optical interface which is in contact with a sample material; interaction with the sample occurs through evanescent wave sampling. Some configurations operate near the angle of total internal reflection and the sample index is tested either by measurement of the critical angle or by measurement of light loss as the solution index changes. Additional methods are provided through measurement of evanescent wave adsorption.
In bulk optical systems, the light intensity is low and suitable sensitivity is obtained by multiple exposures to the sample interface, usually by geometric constructions which provide multiple
reflections.
Some multimode fiber optic variants to the theme are described in Harrick's book. They operate on the same principal as the bulk optic devices with extended exposure to the fiber/
sample solution interface provided by the multibounce propagation method in the multimode fiber Sensing is via intensity vaπation at the output
Three construction vanants, based on multimode optical fibers, have been proposed in the literature as candidates for immunoassay In the first, the fiber cladding is stripped to expose the core, and antigens, labeled with fluorophores, are attached to the fiber surface The fluorophores are excited by the evanescent field and can be detected through reduction in the light level of by collection of the fluorescence A second type of fiber optic immunoassay sensor uses a coating deposited on the fiber tip that can be illuminated by an optical pulse, which in turn induces fluorescence which is reflected back up the fiber and detected A third type of fiber optic sensor for immunoassay involves a stπpped fiber core that has antibodies and antigens attached to the core/solution interface This sensor is used as one leg of a fiber optic Mach-Zender interferometer The binding of molecules to the surface duπng attachment of either antibody of antigen suffices to locally change the index of refraction at the core/solution interface This changes the phase velocity of the light on one leg of the device and interference fringes are observed at its output
Still other configurations of immuno ensors have been described such as surface plasmon
resonance iinmunosensors, and grating couplers used as integrated optical chemical sensors These and others are discussed in the book edited by Wolfbeis Velander and Murphy at Virginia Tech have proposed a fiber optic technique for an immunoassay that uses a grating superimposed on the fiber to scatter light into the cladding where it can sample the cladding/air interface An affinity aerogel coating is used to collect and concentrate target antigens which are measured through the absorption spectrum of the returned light
Still more methods based on evanescent field absorption in optical waveguides are described by G. Stewart.
Conventional evanescent wave spectrometry has been thoroughly researched and is well known in the literature. The techniques involved are also used in fiber optic sensing. The processes usually rely on absorption processes in regions of waveguides where the evanescent field penetrates the guide. The guides are arranged so that as many reflections as possible illuminate the sample interface. Even so, the places where ray optics allows interaction between the optical species and the sample are comparatively few and the illumination is weak compared to the single mode fiber optic coupler approach.
In large multimode optical fibers, a relatively large number, possibly hundreds, of spatial modes are supported. The modes can and do interfere with each other leading to extensive noise generation at the point of detection (i.e speckle). The detector can't distinguish between intensity vaπations due to the sample and intensity variations that occur due to random inference between the propagating optical modes
In contrast, single mode fibers only support one propagating mode Therefore random interference is impossible and no modal redistribution occurs due to envnonmental factors.
Another significant advantage of single mode fibers is that more than 90% of the optical energy can be forced into the evanescent field and that field surrounds the entire space immediately
surrounding the core.
Interferometric approaches are usually the most sensitive available, however they require exceptional mechanical stability. In the fiber optic case, the interferometπc technique suffers because the optical signal rotates due to birefπngence in the bent fiber and path stability becomes a phenomenal problem. The single mode fiber optic coupler sensor is also an interferometric
device, of sorts However, the two legs of the interferometer (the two propagating supermodes) are both contained within the device itself Because of this, the device is self-referencing and the noise associated with path instability is completely avoided
The technique used by Veander and Murphy requires a spec rophotoraeter to read Even m mi nia ture con figurat ion , this is hardl y a "point of care" device use ful in f ield envi ronment s
In a pπor patent, Gerdt and Herr correctly descπbed the significant benefits of the single mode fiber optic coupler sensor relative to current and pπor art The disclosure of the pπor patent, United States Patent 5,494,798, Gerdt and Herr, February 27, 1996, is incorporated by reference herein, as though recited in full To summanze, the benefits include 1 ) the high illumination levels exposed to the sample interface and the high sensitivity which result, 2) the single mode field exposed to the sample and the low noise which results with the elimination of other interfenng modal noise components in the measured signal, 3) a measurement which uses only the vaπation in propagation constant to sense the measurement and is inherently separable from the multitude of intensity noise sources which encumber the measurement, 4) a differential signal output which allows normalization of the measurement and isolation from other noise sources
In their patent, Gerdt and Herr proposed a method for immunoassay using the single mode fiber optic coupler sensor in conjunction with a surface coating of specific monoclonal or polyclonal antibodies to detect small concentrations of target antigens
There are three significant weaknesses in the method proposed which limit the ultimate sensitivity of the device and its use for quantitative measurement The first occurs because the optical source produces noise components in both intensity and frequency In their proposed
method, the intensity noise is removed by conventional difference/sum signal processing of the optical output Although the intensity noise components are removed by the method, the frequency noise components are enhanced along with the real signal No method was provided to address this noise component that becomes significant at higher values of solution index of refraction The second limitation occurs because the measured power splitting ratio is transcendental Single measurements of the splitting ratio are not unique and no method is provided to quantitatively assess the solution index of refraction The third limitation occurs because the gain and thus the noise figure in the initial detection stage is limited by the need to accept the full optical signal without clipping at large values ot the coupling ratio which occurs peπodically as the solution index is changed As a result, small signal changes must be measured on a very large background The signal, thus measured, is very small and system sensitivity suffers as the result The proposed system, as illustrated in Figure L, includes a reference channel that addresses these shortcomings of the proposed method of the pπor art
Summary of the Tnvention The invention relates broadly to a fiber optic sensor device, advantageously, a refractive index measurement device A light emitting member transmits light pieferably coherent light, to a light splitter The light splitter splits the light into a first portion that is transmitted to a first
fiber optic coupler having an input optical fiber member and an optical fiber member having a waist region The coupler preferably is formed from a plurality of single mode fiber optic fibers A first output optical fiber member emerges from the first fiber optic coupler waist region The light splitter is positioned to insert light into the first fiber optic coupler input optical fiber member
A reference fiber optic coupler includes an input optical fiber member and an optical fiber member having a waist region The reference fiber optic coupler is enclosed in a potting medium stable reference of constant refractive index A reference output optical fiber member emerges from the reference fiber optic coupler waist region The first fiber optic coupler and the reference fiber optic coupler are single mode couplers
The light splitter member inserts a second portion of the split light into the reference fiber optic coupler input optical fiber member
A light measuπng member is optically coupled to receive light emitted from the first output optical fiber member and the reference output optical fiber member The light measuπng member can be a light meter for measuπng the magnitude of light
A sum difference device processes the measured light emitted from the first output optical fiber member and from the reference output optical fiber member whereby the value deπved said measured light emitted from said first output optical fiber member is independent of vaπations in light from said light emitting member
The device of the present invention can be used for the immunoassay measurement of a target component Light output from two 01 more of the sample coupler optic fibers are
measured betore and after surrounding the coupler with the target component Light is measured from the reference coupler to establish a light reference value The presence or concentration of the target component is determined by measuπng changes in the output of light from the two or more optic fibers of the sample coupler due to the specific binding of the target component to the first immunoassay component as compared to the light reference value The step of measuπng changes in the output light preferably compπses a plurality of measurements made over a predetermine time interval and deteπnining the rate of change of the light output The step of
compaπng preferably includes companng the rate of change of the light output as an indication of the concentration of the target mateπal in the fluid
The present invention advantageously includes an immunoassay method for deteπmning a target analyte in a fluid sample The method does not use labeled reagents and is based upon measuπng changes in refractive index in an evanescent field surrounding a fiber optic coupler relative to light changes in a reference fiber optic coupler surrounded by a medium of stable refractive index The method includes immobilizing an antibody or antigen capable of specifically binding to the target analyte on the fiber optic coupler and illuminating at least one of the plurality of optical fibers to provide the surrounding evanescent field with a first portion of light from a light splitter
A first refractive index of the surrounding evanescent field is obtained by measuπng a first ratio of light output by the fiber optic coupler The fluid sample is brought into contact with the target fibei optic coupler such that any target analyte in the fluid sample specifically binds to the immobilized antibody or antigen A second refractive index of the surrounding field is obtained by measuπng a second ratio of light output by the target fiber optic coupler
The reference coupler is illuminated with a second portion ot light trom the light splitter to deteπnine a illumination reference level by measuring light output trom said rereience coupler
Brief Description of the Drawings
Figure 1 is a schematic illustration of the system of the present invention,
Figure 2 is a graphic illustration of the vaπation in the coupling coefficient with elongation duπng the fabπcation of the biconical fused tapered couplers,
Figure 3 is a graphic illustration of the variation in the coupling coefficient vs solution
index of refraction for biconical fused taper couplers drawn to different initial values in air,
Figure 4 is a graphic illustration of the vaπation in differential retardance and sensitivity vs solution index of refraction for couplers drawn to different initial values in air,
Figure 5 is a graphic compaπson of predicted values for coupling coefficient v. solution index of refraction and measured values for vaπous solutions,
Figure 6 is a graph of measured value for sensitivity for coupler pulled to initial value of 0 500 in air and potted vs an identical coupler pulled to an initial value of 0 500 in air and unspotted,
Figure 7 are graphs of FFT's of calculated values for sinusoidal vaπation in coupling coefficient as the optical source wavelength is modulated
Detailed Description of the Preferred Embodiments of the Invention
A single mode optical fiber is compπsed of three concentnc components The central component is called the core and is compπsed ot a mateπal, usually quartz, with an index of refraction near 1 46 The core contains almost all of the propagating light and is sized just large enough to contain only one propagating mode, the fundamental The core is surrounded by a concentnc region called the cladding The index of refraction of the cladding is slightly depressed relative to that of the core so that a guiding interface exists between the two regions The cladding is of sufficient size that the propagating mode in the coie is unaffected by
vaπations in the environment which might surround the fiber Quartz is susceptible to breakage if it is exposed to moisture, therefore the quartz components of the fiber are packaged in a third layer of a plastic mateπal which provides a hemietic seal and protects the fiber duπng handling. This layer is called the buffer
If the plastic buffer is removed and two fibers are placed adjacent to each other in a furnace of flame, they will fuse When the heated and fused fibers are gently pulled, they will
elongate resulting in a reduction of both the fiber diameter and separation between the two cores As this process continues, optical power in one of the fibers will begin to couple into the second fiber As the fibers are pulled still further, the power will transfer completely to the second fiber and then begin to transfer back to the first fiber again, Figure 2 This process will continue indefinitely, the light alternately transferπng from one fiber to the other, until the fibers become too small and break
Light, propagating in the cores of the two adjacent fibers, is descπbed by the wave equation, a physical interpretation of Maxwell's equations applying to electromagnetic phenomena The solution for this model is degenerate and results in two wave functions, one of which is symmetπc and the other anti-symmetric relative to the geometπc center of the pair ot fibers These wave functions are called the two "super modes" The propagating light at any point in the fibers is described by the vector sum of the two super modes and the optical power in each of the fibers is descπbed by the square of their sum The super modes extend beyond the fiber surface and into the surrounding mateπal, however one of the modes extends a little further than the other and as a result, they propagate at different speeds A phase difference, E, is accumulated between the super modes as they propagate over the fused length of the coupler That phase difference depends on the length ot the fused region, the wavelength ot the
illumination, and index of refraction of the core and cladding The division of power between the two fibers is descπbed by the coupling ratio, a , where
a = Sin2 (E)
In the fused region of the coupler, the fiber diameter is reduced from an initial value near 125 microns to about 15 microns The oπginal core diameter is reduced in proportion As the core is reduced, it can no longer contain the propagating mode and light spills into the
surrounding cladding. After fabrication, almost all of the light resides in the original cladding
material. The original cladding becomes the new core, and whatever surrounds the fused region, the sample solution, becomes the new cladding. As a result, the division of power between the
two fibers becomes a strong function of the index of refraction in the small volume surrounding the fiber surfaces.
A mathematical model, which approximately descπbes the coupling ratio in terms of the
fused length, the wavelength of the illumination, and the index of refraction of the core and
cladding has been derived and published by J. Bures, S. Lacroix, and J. LaPieπe, ( 1983). In their work, Bures, et. al. calculate the coupling ratio in terms of: the wave number,
2π k = λ the profile height parameter,
the waveguide parameter.
V = kr\n — 'I , ., I
the core parameter,
-{!+<!)
U = 2.405e
and the cladding parameter,
W = (v 2 - U 2 )ϊ
then the coupling coefficient can be calculated
and the differential phase, calculated over the interval where C
pp is non zero, is then
the resulting coupling ratio is a = Sιn2(θ)
and the sensitivity is then defined as
Sensitivity = — n = constant dλ
In the preceding deπvation, e is the wavelength of the illumination, r is the ratio of core
separation to the core radius, ncme and ncladώn are the indices of refraction of the core and
cladding respectively designates a partial denvative, and Ko and K are modified Bessel
functions of the second kind
Figure 3, illustrates the vaπation of the coupling ratio, a , with solution index for couplers pulled to a = 0 500 and 1 00 respectively, in air n = I 000 The figuie demonstrates that the correlation between coupling ratio and solution index is not unique A means is required to quantitatively define a unique value of index relative to a single measurement of the coupling
ratio
Figure 4a, illustrates the vaπation of the differential retardance, e, vs the solution index for the same two couplers pulled to initial values of the coupling ratio, 0 500 and 1 00 Here, the measured function is monotonic, however, its measured value must be deπved from the
measurement of the coupling ratio, which is not Since the retardance contains many cycles (N*2ό) it is still impossible to uniquely determine the index from the measured value
Figure 4b, illustrates the first partial deπvative of retardance relative to wavelength at constant index This curve relates the signal change (retardance) that would result from changes in wavelength at constant values of solution index If two couplers were simultaneously illuminated from the same source, the amount of change in signal (retardance) from the two couplers would depend on the index of refraction in each If one of the couplers was immersed in a standard mateπal of known index (the reference), the index of unknown sample could be uniquely and quantitatively deteπnined by the latio between the change in signal amplitude of the unknown and the change in signal amplitude of the reference This ratio is uniquely defined for all values of index below the index of retraction of the coupler itself At this value, dependent on the mateπal in the optical fiber, the light becomes ungmded and escapes Measurements are not possible above this value
Measurements, deπved from the evanescent wave refractometer must be made from observation ot the optical power in each of the fiber outputs and from calculations based on those obser ations As has already been discussed, the total optical power in each coupler, b. is
divided between its two output channels Ii and Remembeπng that the coupling ratio, a , is the fraction of the total power in the coupler coupled into channel 2
a = Sιn2(θ)
then the power in the channels is
/, = /0(l -Λιι2 (/9))
and I2 = I0Sιn2 {θ)
The coupling ratio, a , and the retardance, e , can be deπved from measurements of the channel power levels, and calculated from the difference between the two channels, A , divided by the sum of the two channels, O
Δ. = tn1{θ))- I0(l - Sιn1{θ)) Σ n2(θ))+ I0(l - Sιn 2(θ))
from which it can be shown that the coupling ratio
a = 0 5| 1 +
It is apparent that measurements of the coupling ratio of the retardance based on A/O are normalized to the instantaneous optical power and are, therefore immune to intensity noise in the source From the previous deπvation, however, it is also apparent that the measured values are still dependent on wavelength and that another means is required to remove that noise source Inclusion of the second optical channel, the reference, potted in a mateπal of constant index, provides that function
Based on this model, predictions of the \ aπation in coupling ratio vs solution index of refraction for a coupler pulled to an initial value of a = 0 5 in air, n = I 000 is plotted in Figure 5 and the ratio of signals obtained from a wavelength modulated source (relative sensitivity) for a
coupler in air and a coupler in solutions of higher index is plotted as Figure 6
A preferred embodiment of the evanescent wave refractometer system is schematically descπbed in Figure 1 It is compπsed of a pigtailed optical source (laser or light emitting diode) dπven by a controller, a thermoelectπc (Peltier) cooler and its theπnoelectπc controller, a splitter
(a 1 x N "tree" coupler with approximately equal outputs), a coupler used to measure the sample characteπstics, a potted coupler used as a stable reference, detectors and transimpedance amplifiers for each channel of optical output and sum/difference processing for the sample and reference channels. An oscillator that also provides a refeience signal for the dual channel Phase Sensitive Detector modulates the source controller A plotter, logger, scope, or other display apparatus is used to capture the measurement output.
A less complicated version of the system is currently in use In this system, output from the transimpedance amplifiers on each of the optical channels are independently collected by the data logger and data analysis is performed "off line" by computer based analysis packages Using the lesser embodiment, a test system was constructed using a sample coupler drawn to an initial value of a = 0 5 in air and a stable reference whose initial value of a = 0 5 in GE RTV12, n = I 420, at the optical wavelength of 1 300 im Data was collected at d c for the coupling coefficient obtained using several mateπals of known index These measurements are plotted on Figure 5 to illustrate the agreement obtained from the uncorrected measurement The value obtained using Hexane, n = 1 3745, shows significant error because Hexane rapidly attacks the tygon tubing used to construct the test apparatus
The optical source was modulated at I Hz and the channel output data on Figure 7 was collected for a sample coupler in air, n = I 000, and a reference coupler potted in GE RTVI 2, n = 1 402 Instantaneous values of retardance were calculated and the vanation in amplitude at the fundamental modulation frequency was deπved from it using the Fast Fourπer Transform method The ratio of signal amplitudes for that measurement is plotted on Figure 6 to illustrate the accuracy of the method The ratio of modulation signal amplitudes for the sample and
reference predict the index of refraction for the sample without the uncertainty produced by using the coupling coefficient by itself.
OVERVIEW
A single mode fiber optic system has been constructed for quantitative measurement of chemical and biological reaction vaπables which are available through direct measurement of the solution index of refraction Such vanables include measurement of reaction extent and reaction kinetics through measurement of the solution index of refraction which is dependent on concentrations of the products and reactants Further, through application of selected components to the sensing surface, concentrations of specific target components can be quantitatively measured with the extremely high sensitivity required for use in immunoassay
The system is based on the fundamental properties of a biconical, fused tapered coupler as descπbed by Gerdt As a pair of single mode optical fibers are fused and drawn, optical power in the first fiber of the pair will couple into the second fiber of the pair This division of power in the fibers is reflected in the coupling ratio (1 e the coupled power as a percentage of the total power in the pair of fibers ) The coupling ratio is dependent on the degree of elongation in the fusion zone, the separation of the fiber cores, the wavelength of the illumination, and the index of refraction in the environment surrounding the fused and elongated portion of the fiber (i e the w ist) If a coupler is diawn to and Initial Value (IV) of the coupling ratio in air and fixtured in a stable housing, any further change in power division is a function ot the wavelength and the index of lefraction in the sensing region
Sensitivity and repeatability of the system are limited by system noise The pπncipal noise source in the system is the optical source, typically a solid state laser or light emitting diode The noise is low frequency and demonstrates a 1/f characteπstic which is well documented in the literature A feature of the noise is that it contains components in both the amplitude domain (intensity noise) and n the frequency domain (frequency noise) The evanescent wave sensor as
described by Gerdt, et. al. provides means to remove the intensity noise component through application of a well known difference/sum technique to the differential outputs of the sensor. It, however, amplifies the frequency noise component which appears in the measurement of the coupling ratio which the fundamental output of the device. The proposed system provides additional means to remove the frequency noise component in the output by comparison of the measured coupling ratio of the sample sensor with that of a stable reference, simultaneously illuminated by the optical source. The stable reference is an additional coupler which has been potted in a material of known index. Its output signal, also sampled through the same difference/sum technique, varies only with the wavelength of the source and allows the frequency noise component to be subtracted from the measurement, thus eliminating optical system noise components to the limits of the detection apparatus.
The output from the evanescent wave sensor as described by Gerdt, et, al. is measured in teπus of the coupling ratio. This ratio is transcendental and repeats the same value at multiple values of the sample solution index of refraction. Thus the measurement of a single value of the coupling ratio cannot be uniquely correlated to a value of solution index if the measurement is not continuous (e.g. if there is a rapid change or if the measurement is lost duπng inteπnediate
steps in a chemical process). The inclusion of the stable reference in the measurement system allows quantitative evaluation of the sample index of refraction as a simple ratio of the variation in the signal from the sample sensor vs. the simultaneous signal variation from the stable reference as the source wavelength is changed.
If the frequency of the optical source is modulated (e.g. by intensity modulation of the laser or light emitting diode) then and a c technique of measurement is obtained by which the index of
refraction in the sample can be continuously and unambiguously measured as the ratio of signal levels in the sample channels relative to the signal levels in the stable reference. Additionally, the a/c technique allows the use of the homodyne detection scheme which moves the sampling measurement out of the high noise environment at low frequencies and provides means for substantial signal / noise improvement through application of phase sensitive demodulation to the much larger signal levels which become available.