WO1994017388A1

WO1994017388A1 - Ion selective fluorosensor based on the inner filter effect

Info

Publication number: WO1994017388A1
Application number: PCT/US1994/000459
Authority: WO
Inventors: Otto S. Wolfbeis
Original assignee: Fci-Fiberchem, Inc.
Priority date: 1993-01-26
Filing date: 1994-01-12
Publication date: 1994-08-04

Abstract

An ion selective optrode or optical chemical sensor makes use of a primary absorber dye (14) whose color changes with the analyte concentration, and a fluorophore (18) whose excitation and/or emission is modulated by the changing absorption of the dye (14). In a membrane (10), an ion carrier (12) captures target ions X+ from a solution (16). Through charge exchange or co-extraction, the absorber changes its absorption characteristics and acts as an inner filter. The sensing scheme provides high flexibility in the selection of dyes and results in detection limits on the order of 1 νmole. The preferred absorber dye is 4-(2'-4'-dinitrobenzeneazo)-2-octadecylaminocarbonyl-1-naphthol.

Description

ION SELECTIVE FLUOROSENSOR BASED ON THE INNER FILTER EFFECT

BACKGROUND OF THE INVENTION

The invention relates generally to optical chemical sensors, and more specifically to ion selective optical chemica sensors.

Reliable and sensitive ion sensors are highly needed fo various applications including clinical and environmental io sensing. A number of schemes for optically sensing alkali an alkaline earth ions have been developed in the past few years. While some are based on the use of chro o-ionophores or crow ethers, the most promising sensing schemes at present appea to rely on the use of neutral ion carriers. Three differen methods based on the use of ion carriers (such as valinomycin) are known. One uses potential-sensitive indicators, anothe uses proton carrier dyes, and a third uses lipophilize acridinium dyes whose fluorescence is highly polarity sensitive.

The recently introduced ion-exchange or co-extraction sensin schemes have provided a most promising way for sensing number of ions. In an ion exchange system, the carrie mediated transport of a cation into a pvc membrane is couple to the release of a proton from a lipophilic proton carrie contained in the sensing membrane. This deprotonation cause a change in the optical properties of the proton carrier t occur. In a co-extraction system, the carrier mediate transport of an ion (e.g., anion) into a membrane is coupled to the transport of a counter ion (e.g., proton) into the membrane to a counterion carrier which thereby changes optical properties. Most sensors of that kind are based on absorbance measurements. However, fluorosensors are preferred because of the intrinsic sensitivity of fluorescence, its inertness to turbidity effects, and its flexibility with respect to geometric arrangements.

Fluorescent dyes fulfilling the following requirements are desirable in order to be useful in carrier-based ion sensors:

(a) longwave excitation and emission wavelengths so as to make the sensor LED-co patible, i.e., excitable by inexpensive light-emitting diodes which, in addition, can be modulated at extremely high frequencies; (b) solubility in plasticized pvc; (c) high lipophilicity in order to keep the dye in the lipophilic pvc membrane; (d) appropriate pK_a values so as to cover a wide dynamic range and to achieve low detection limits; (e) high photostability; (f) high fluorescence quantum yield. However, no fluorescent dye has been found that fulfills all these requirements.

Charge exchange and coextraction mechanisms in membranes, and related chromoionophore based optical sensors have been studied by Simon et al, e.g, as described in "Design and Characterization of a Novel Ammonium Ion Selective Optical Sensor Based on Neutral Ionophores," Seiler et al, Anal. Sci. (Japan) , Vol. 5 (1989) 557-561; and "Carriers For Chemical Sensors: Design Features of Optical Sensors (Optodes) Based on Selective Chro oionophores, " Morf et al, Pure & Applie Chem., Vol. 61 (1989) 1613-1618. Fluorescence sensors ar represented by Kawabata et al, "Fiber-optic Potassium Io Sensor Using Alkyl-Acridine Orange in Plasticized Poly(viny chloride) Membrane," Anal. Chem., Vol. 62 (1990) 1528-1531 i which fluorescence of a hydrophobic probe varies as a resul of ion exchange; Suzuki et al, "Fiber-Optic Magnesium an Calcium Ion Sensor Based on a Natural Carboxylic Polyethe Antibiotic," Anal. Chem., Vol. 61 (1989) 382-384 in which th fluorescer complexes with the target ion; and He et a (Wolfbeis) , "Fluorescence Based Optrodes for Alkali Ions Base on the Use of Ion Carriers and Lipophilic Acid/Bas Indicators," SPIE Vol. 1368 Chemical, Biochemical, an Environmental Sensors II (1990) 165-174 which uses a selectiv ion carrier with a fluorescent proton carrier.

Another type of ion sensor is based on fluorimetri measurement of membrane potential, as shown by Wolfbeis et al "Optical Sensors: An Ion-Selective Optrode for Potassium, Analytica Chimica Acta, 198 (1987) 1-12; Shaffar et al, " Calcium-Selective Optrode Based on Fluorimetric Measuremen of Membrane Potential," Analytica Chimica Acta, 217 (1989) 1 9; and Wolfbeis et al, U.S. Patent 4,892,640 (1990) .

Energy transfer and inner filter effects have been used t produce absorbance modulated fluorescence detection method and sensors, such as Jordan et al, "Physiological pH Fiber optic Chemical Sensor Based on Energy Transfer," Anal. Chem. Vol. 59 (1987) 437-439; Walt, U.S. Patent 4,822,746 (1989) Gabor et al, "Sensitivity Enhancement of Fluorescent p Indicators by Inner Filter Effects," Anal. Chem. Vol. 6 (1991) 793-796. These techniques require an overlap betwee the fluorescer excitation or emission band and absorbe absorption band; the two dyes must also have appropriate p values.

The absorption type charge exchange system of Simon et al, an the fluorescence type of Wolfbeis et al each have limitation in that all the desirable characteristics including io specificity, high quantum yield, proper pK_, insolubility stability and long wavelength operation may not be found i a single absorber or fluorescer. The inner filter effec methods of Walt et al have not been adapted to these charg exchange types of solid state optical sensor.

Accordingly, the invention may provide an improved io specific optical chemical sensor, an ion sensor based on th inner filter effect, and an ion sensor based on charg exchange and/or co-extraction the inner filter effect.

SUMMARY OF THE INVENTION According to one aspect of the invention, there is provide an optical chemical sensor, comprising: a membrane; an io selective carrier in the membrane for extracting a target io from a solution into the membrane when the membrane is place in contact with the solution; a charge exchanging o counterion coextracting absorber in the membrane whic maintains electrical neutrality by charge exchange o counterion coextraction when target ions are extracted by th carrier and thereby charges its absorbance; a fluorescer i the membrane having an excitation band or emission band whic overlaps an absorption band of the absorber; whereb fluorescent emission from the fluorescer is modulated b changes in the absorbance of the absorber caused by the targe ion.

According to another aspect of the invention, there i provided an absorber dye comprising 4-(2'-4' dinitrobenzeneazo) -2-octadecylamindocarbonyl-l-naphthol.

The invention is an ion selective optrode or optical chemica sensor that preferably exploits the inner filter effect (IFE with fluorescence. The sensing scheme may make use of neutra ion carriers which extract target ions into a membrane Transduction may be based on the use of two different dyes an absorber and a fluorescer. The absorber is ion specific the fluorescer is not. The absorber may also be a charg exchange carrier or coextraction counterion carrier. Th absorber may act as the proton carrier or receiver. Th capture of an ion by the ion carrier causes the absorber dy to give up or capture a proton to maintain electrica neutrality in the system. This changes the absorptio characteristics of the dye. The other may be a stable and pH independent fluorophore bound to minute particles containe in the sensor membrane. Fluorescence varies through the IF as a result of the varying absorption of the dye which in tur is modulated by the ion concentration. The sensor materia is preferably extremely sensitive to the ion, e.g., it fully reversibly responds over the 1 μ mole to 10 m mole concentration range with a useful dynamic range of 1 μM to 1 m mole. The sensing approach may be generic in that it can be applied to almost any species for which respective carriers are known. Thus, by replacing the potassium ion carrier (valinomycin) by carriers for ammonium and calcium ion, respective ion sensors may be obtained without requiring any changes in the optical system.

Many of the problems resulting from the use of one single dye (where a number of compromises has to be made) can be overcome by making use of two dyes, the first serving as the charge donor or receiver that can be optimized with respect to absorption wavelengths, pK₃ value, lipophilicity, and photostability, the other being an extremely stable and strongly fluorescent dye contained in minute beads and whose excitation and/or emission bands overlap the absorption band of the deprotonatable or protonatable absorber dye. The sensor material has an ion-dependent absorption which results in a variable excitation or emission efficiency for the fluorophore because of the so-called inner filter effect (IFE) . The absorber-modulated intensity of the fluorescence of the particles provides analytical information.

This approach for sensing alkali ions is preferably different in that (a) synthetic charge carriers/coextractors may be used and the recognition process and ion transport are coupled to a protonation/deproton-ation step; (b) solid particles may be used which are incorporated into plasticized pvc membranes and (c) the sensing scheme clearly may be based on an IFE an not on energy transfer.

The invention is a solid state optical sensor. The membran material can be obtained in a simple manner, e.g., al components are dissolved/suspended in a solvent such a tetrahydrofurane, and then spread as a membrane or placed o a waveguide structure. The particles cannot be washed ou because they are mechanically incorporated into the pv membrane. This method has almost all the advantages o fluorescence-based sensors and is broadly applicable; it i a generic approach which can be adapted to almost any ion fo which a carrier is known.

BRIEF DESCRIPTION OF THE DRAWINGS Fig. 1 is a schematic view of an ion selective membrane.

Figs. 2A,B illustrate the ion extraction and charge exchang mechanism.

Fig. 3 shows the chemical structure of the lipophilize proton-carrying dye (KFU 111) .

Fig. 4 shows the absorption spectra of KFU ill in plasticized pvc membrane containing PTCB and valinomycin, i contact with a 100 m mole aqueous solution of KCf of pH 7.41 and with plain potassium-free buffer of pH 7.41. Fig. 5 shows the excitation and emission spectra o FluoSphere™ particles contained in a plasticized pvc membran containing PTCB, valino ycin and KFU 111 contacted with a 200 m mole KCf solution of pH 5.22 and with plain buffer of p 5.22 demonstrating the dramatic inner filter effect caused b the blue form of the absorber dye.

Fig. 6 shows the response time, relative signal change, an reversibility of the potassium sensor (a) in the presence an (b) absence of dye KFU 111 in the membrane at pH 5.82, excitation/-emission wavelengths set to 560/605 nm. The small signal change that can be seen in the dotted line was obtained after exposure of the sensor to 200 m mole potassium and is comparable in intensity with the signal obtained with 1 μ potassium when the absorber dye KFU 111 is added.

Fig. 7 shows the pH dependence of the calibration curve of the potassium-sensitive membrane, excitation/emission wavelengths set to 560/605 nm, absorption measured at 640 nm: (a) fluorescence measurements at pH 5.22; (b) fluorescence measurements at pH 5.82; (c) absorption measurements at pH 7.41; (d) absorption measurements at pH 5.22.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS As shown in Fig. 1, an ion selective membrane 10 contains ion selective carrier molecules 12 and deprotonatable absorber molecules 14 homogeneously distributed therein. A sample solution 16 contacts membrane 10. The carrier molecules 12 attract ions X⁺ from solution 16 into membrane 10. To maintain charge neutrality, the absorber molecules 1 deprotonate, losing protons H⁺ to solution 16. A pluralit of discrete fluorescent particles 18 are also distribute

(non-homogeneously) through membrane 10. Membrane 10 i mounted on a glass substrate 20. Excitation light from sourc

20 incident on membrane 10 excites particles 18 to fluoresce

Detector 24 detects fluorescence from particles 18

Deprotonated absorber molecules 14 in membrane 10 modulate th fluorescence from particles 18 as a function of th concentration of target ions X^'1" which have been drawn int membrane 10 by ion carriers 12. Deprotonated molecules 1 modulate fluorescence by either absorbing excitation light o fluorescent emission or both.

The detection system is shown with membrane 10 mounted o glass substrate 20, e.g., in a flow-through cell. However other measurement configurations can be used, e.g. placin membrane 10 on an optical fiber. Any suitable source 22 an detector 24 can be used; however, the system can be designe for long wavelength sources (e.g. LEDs) and detectors.

The details of the charge exchange mechanism are further show in Figs. 2A,B. As shown in Fig. 2A, membrane 10 contain carrier molecules 12 and proton carrying absorber molecule 14 (A"-H⁺) while sample solution 16 contains target ion (cations) X^"1" (and corresponding anions Y^") . When solution 1 contacts membrane 10, carrier molecules 12 extract ions X from solution to form ion-carrier complexes 12', as shown i Fig. 2B. Simultaneously absorber molecules 14 lose proton H⁺ to solution 16, leaving deprotonated absorber molecules 14 (A^") .

Although illustrated specifically with a deprotonatable charg exchange molecule, and cationic species X⁺, more generalize ion carriers and charge exchange or coextraction molecules ca be used. For an ion carrier which captures the anioni species Y^", e.g., nitrate (NO-,^") , the coextraction absorbe molecules (A) would be protonated instead of deprotonated i.e. protons would also be extracted from the solution int the membrane, producing protonated molecules A-H⁺. The charg exchange/coextraction molecules could also exchange or captur ions other than protons.

Counterion co-extraction is the preferred mechanism for anio sensing. The ion carrier will carry the anion into th membrane. The membrane cannot release another anion (in orde to maintain charge neutrality) because the anions in th membrane are too lipophilic. Instead, a counterion, e.g. proton, will be co-extracted from the sample solution into th membrane where it protonates the absorber dye, giving rise t a color change.

Electronically, the charge exchange or coextraction molecule function to maintain charge neutrality when target ions ar extracted into the membrane. The charge exchange/coextractio molecules also have an optical function, to change absorbance as a result of the charge exchange or coextraction, so tha the fluorescent emission of the fluorescent particles in th membrane may be modulated by the inner filter effect. Thi requires that the absorption band of the charge exchang absorber molecules overlaps the excitation and/or emissio band of the fluorescer.

However, other requirements can be optimally allocated betwee the absorber and fluorescer. The fluorescer can be chose with high quantum yield, long wavelength excitation an emission, stability and insolubility, while the absorber ca be selected for proper pK_α and insolubility.

An illustrative potassium sensor was prepared and tested a follows.

Reagents. Poly(vinyl chloride) (pvc, high molecular grade), valinomycin, potassiumtetrakis(4-chlorophenyl) borate (PTCB), 2-nitrophenyloctyl ether (NPOE) , and tetrahydrofurane (THF) were obtained from Fluka AG (Buchs, Switzerland) . Th fluorophore particles (FluoSphere™) were purchased fro Molecular Probes, Inc. (Eugene, OR) . The excitation/emissio maxima of the particles are at 580/605 nm, respectively. Th high-purity potassium chloride was obtained from Merc (Darmstadt, FRG) . All buffer components (citric acid, tris, HCf) were from Merck. Aqueous buffers were 0.05 M in eac citric acid and Tris, and the pH was adjusted with 1 M NaO or HCf to the appropriate value.

Syntheses of lipophilic pH indicators. The pH-sensitiv absorber dye, 4 - ( 2 ' , 4 ' -dinitrobenzeneazo ) -2 octadecylamidocarbonyl-1-naphthol (KFU 111) was prepared b the following procedure:

(a) Synthesis of precursor dye 4-(2 ', 4 '-dinitrobenzeneazo) -1 hydroxynaphthalene-2-carboxylic acid (KFU 110) . Solution A. 16 g of concentrated sulfuric acid wer stirred and cooled to below 5° C in an ice bath. Then, 1. g (20 m mole) of finely ground sodium nitrite were adde slowly. The temperature of the mixture was kept at below 5 C until the addition of nitrite was complete. Thereafter, th temperature was slowly raised to 70° C and the mixture becam clear. After cooling to 20° C, 3.7 g (20 m mole) finel ground 2,4-dinitroaniline were added slowly, the mixtur stirred for 10 min. , and finally poured onto 48 g ice.

Solution B. 3.7 g (20 m mole) 1-hydroxynaphthalene 2-carboxylic acid were dissolved in 200 ml water containin 30 g sodium carbonate. The residue was filtered and th filtrate cooled to 0° C.

Solution A was dropped into Solution B within about 30 min under strong stirring. The resulting precipitate wa filtered, washed three times with 500 mf of 1 M sulfuric acid and dried at 50° C for 24 hours to yield 6.5 g (84%) of crud yellow precursor dye. Elemental analysis: calc. C 53.41, 2.64, N 14.66; found C 53.04, H 2.74, N 13.38. The absorptio maxima are at 485 nm in water of pH 4.0, and at 615 nm a aqueous solution of pH 10.00. The pK., as calculated from th titration curve, was found to be 8.98 ± 0.06 at 21° C and a ionic strength of 100 mole sodium chloride.

(b) Synthesis of 4-(2',4 '-Dinitrobenzeneazo) -2-octadecyl amidocarbonyl-l-naphthol (KFU 111) 0.38 g crude KFU 110 and 0.28 g octadecylamine were dissolve in 200 ml THF containing 0.22 g N,N'-dicyclohexyl-carbodi imide and stirred at room temperature for 4 hours. Th suspension was filtered in order to remove the whit precipitate (N,N'-dicyclohexylurea) . The solvent wa evaporated and the yellow residue redissolved in 150 m dichloromethane. This solution was washed with 5% aqueou sodium carbonate solution until the organic phase becam colorless. The solvent was evaporated and the residu purified by flash chromatography (ethyl acetate/petrol ether 1:4, v/v) to yield 100 mg (15%) of the dye. The chemica structure of KFU 111 is shown in Fig. 3. Elemental analysis calc. C 66.33 H 7.47 N 11.05; found C 65.91 H 7.56 11.21.

The absorption maximum of the base form of KFU 111 i chloroform (to which a drop of triethylamine was added i order to create the blue anion) is at 640 nm, the mola absorption coefficient is 97 000 M^"1 cm'.

Instruments. Optical measurements were performed using a Aminco SPF 500 spectrofluorometer equipped with a tungste halogen lamp as a light source. Data were transferred ont a HP 9825A desk calculator and plotted. Sample solutions wer pumped through the flow-through cell containing the sensin membrane, using an auto-sampler. Absorption measurements wer performed with a Perkin-Elmer Lambda 5 spectrophotometer. p measurements for both the buffer preparations and p determinations were performed at 21° C with a Metrohm p meter (Metrohm, Buch, Switzerland) calibrated with Aldrich pH standards of pH 4.00, 7.00 and 10.00, respectively.

Membrane Preparation. The membranes were prepared from a batch of 2.4 mg pvc, 0.4 mg PTCB, 1.8 mg valino ycin, 30 mg of the aqueous suspension of the FluoSphere™ particles, 0.4 mg KFU 111 and 6.0 mg NPOE, all of which were dissolved in 1.5 ml freshly distilled tetrahydrofurane. 100 μl of this solution were pipetted onto a 12 X 50 mm dust-free glass plate which then was placed in a THF-saturated atmosphere in an exsiccator. After about 15 min. the glass plate with the sensing membrane on it was removed and placed in ambient air for 15 min. for complete drying. Before measurements, the membrane was placed in a 0.1 M potassium chloride solution for activation. From the amount of materials employed, the thickness of the membranes is estimated as between 2 and 3 μm. Although plasticized pvc is preferred, other hydrophobic membranes could be used, with added plasticizer to increase permeability.

Experimental Procedure. The glass plate with the pvc membrane of about 2 μm thickness (as calculated from the amount of materials used) was mounted in a flow-through cell. Standard solutions, prepared from double distilled water, were pumped over the sensing membranes, and fluorescence intensity was observed at excitation/emission wavelengths of 550/605 nm. The photostability of fluorophore and absorber were tested by illuminating the dry sensing membrane with a 250-Watt tungsten halogen lamp at 550 nm and a bandpass of 8 nm. Leaching effects were tested by passing a continuous flow of pH 5. buffer over the membrane at a rate of 1.5 mf/min.

Selection of Absorber and Fluorophore. The approach present here requires the presence of two dyes, one acting as t analyte-sensitive primary absorber, the other as the analyt independent fluorophore whose excitation or emission intensi is modulated by the varying absorption of the prima absorber. This requires the absorption band of the absorb to overlap with the excitation and/or fluorescence emissi band of the fluorophore. The primary absorber homogeneously distributed in the sensor material, while t fluorophore is present in the form of small particles coated beads which cannot be washed out of the pvc membra (although it also could be homogeneous) . Both KFU 110 and K ill match the absorption (excitation) band of the FluoSpher particles, thus giving rise to a significant IFE in t presence of potassium ion. KFU 111 is one of the preferr dyes because the octadecyl side chain provides both hi solubility in pvc and insolubility in water. An excess absorber to fluorescer is preferred in order to strong modulate the fluorescent signal.

The fluorophores are commercially available. The FluoSpher particles are Latex microspheres with surface modification in typical sizes from 0.01 to 10.0 μ , available with bl (ex. 360/em. 415 nm) , yellow-green (490/515) , oran (530/560) , red (580/605) , crimson (625/645) and dark r (650/690) emissions. Other manufactures of fluoresce microspheres include Polysciences, Seradyne, Duke Science an American Colors.

The Sensing Scheme. Neutral ion carriers such as valinomyci are known to specifically recognize and bind alkali ions an to transport them into the plasticized pvc membrane. In th case of ion-selective electrodes, a potential is created whic is measured by potentiometry. In the case of optrodes several options exist: One is to measure the potential b optical means, the other is to couple the ion transport int the membrane to a proton transport out of the membrane. Thi is the mechanism of this sensing scheme: Upon potassiu transport into the plasticized pvc membrane, a proton i released by the dye (the primary absorber in the membrane into the sample solution. The absorber dye thereby undergoe a measurable change in its optical properties. Fig. 4 show the absorption spectra of a membrane containing KFU 111, bu no fluorescent particles. The maxima of the alkaline for (which is present in the membrane when it is contacted wit a 100 m mole solution of potassium ion) and the acidic for (present when contacted with plain, potassium-free buffer) ar at 640 nm and 480 nm, respectively. The alkaline form has strong shoulder at 605 nm.

In Fig. 5, the fluorescence spectra of the same membrane afte addition of fluorophore particles are given. Due to spectra distortion caused by the IFE, the maxima of excitation an emission are at 580 nm and 605 nm now. The emission ban matches the absorption of KFU 111. When this membrane i contacted with 200 m mole potassium solution, the pH-sensitiv dye becomes fully deprotonated, resulting in an increase i the 640-nm band which absorbs the emission of the fluorophore As a result, the fluorescence intensity at 605 nm i decreased. In the presence of 200 m mole potassium, th excitation and emission bands are even more distorted.

Dynamic Response of the Sensor Membrane. Fig. 6 shows th dynamic response of a sensing membrane when exposed to variou levels of potassium ion. The signal is fully reversible, wit response time (of a 2-μm membrane) on the order of 6 min. fo the full signal change to occur in the forward direction, bu only 2 min. for the reverse response. The response tim increases proportionally with the square of the thickness o the sensing layer. The sensor responds to potassium over th 1 μ mole to about the 10 m mole concentration range, with detection limit (defined as 3 times the background) of aroun 1 μ mole potassium. Detection limits are virtually unaffecte by the thickness of the sensing layer. Even with membrane as thin as 1 μm, a signal-to-noise ratio of > 500 was obtaine with these sensor membranes. The dotted line in Figure shows the response of the membrane without a pH-sensitiv absorber dye being added. The membrane has little respons even to 200 m mole potassium. The weak signal observed ma be due to the fact that the fluorophore is a weak potentia sensitive dye.

Fig. 7 shows the pH dependence of the response curves in th presence and absence of fluorophore. The dynamic range i shifted to the lower potassium concentration range in th presence of the fluorophore. Furthermore, the linear rang of curve (b) (ranging from 10 μ mole to 3 m mole) occurs a a much lower concentration range than that of curve (c) (10 μ mole - 100 m mole) .

Sensor Stability. The shelf lifetime of the sensor membran exceeds 3 months when stored in the dark at 4° C. Th limiting factors determining the stability of the membran include (a) the photostability of the primary absorber (whic is moderate) , (b) the photostability of the fluorescen particles (which is excellent) , (c) the leaching of valinomycin, borate, and plasticizer from the membrane when under operation (or in contact with aqueous standards) . The sum of effects result in a sensor drift of -4% over a period of 2.5 hours when sensing aqueous samples. Photobleaching of KFU 111 seems to be responsible for most of the drift.

The Sensing Scheme. The results demonstrate the feasibility of sensing alkali ions via the inner filter effect (IFE) of fluorescence by making use of fluorescently dyed particles added to the sensor material. The scheme offers an attractive alternative to absorptiometric, reflectometric, or fluorimetric techniques using one single dye. By making use of two dyes (a primary absorber and a fluorescer) , a more versatile system is obtained which can be optimized in terms of the properties of both the absorber and fluorescer. An absorber dye needs to be adjusted to the proper pK_a value (in order to make it an ideal proton carrier) and made highly lipophilic so as to remain in the pvc phase. The fluorophor is expected to display a high quantum yield and to match th absorption band of one of the two forms of the primar absorber. Unlike the absorber dye, there is no need for th fluorophore to be homogeneously distributed in the senso material. A commercially available fluorescer material wit a quantum yield close to unity is used, in the form of smal particles or coated beads.

The only additional requirement for this approach over the on where a single dye acts as both a proton carrier and fluorophore is an overlap of the absorption of the primar absorber with the excitation or emission band of th fluorophore. This is rather easy to accomplish; the absorbe dye and the FluoSphere particles used here are a well matche pair of dyes. In addition, both are LED-compatible.

In cases where charge exchange cannot occur, e.g., whe capturing anions, the absorber dye must be a charg coextractor which captures the counterion from the solutio and thereby changes optical absorption properties.

Sensor Features. There are a number of promising aspects o this type of sensor:

(1) It allows every absorption-based sensor to b made a fluorescent sensor, even if no appropriate fluorescen primary absorber can be found. (2) The tremendous wealth and variety of neutra ion carriers which have been optimized for electrochemica sensing purposes in the past 20 years can be exploited for optical sensing purposes as well. Hence, by simply replacing the potassium carrier by another one, the sensor can be made responsive to any other analyte for which an ion carrier iε known. For example, nonactin is a carrier for the ammonium ion. A list of ion carriers is found in the Fluka catalogue "Ionophores for Ion-Selective Electrodes and Optrodes." In instrumentation design, this has the tremendous advantage of an identical opto-electronic system for all kinds of sensors. (3) The linear range of the response curve is distinctly shifted to the lower concentrations range; this is important in practice where blood samples will be diluted with buffer in order to adjust a constant pH and to reduce the lipophilicity of the blood sample (which tends to extract carriers and plasticizers) ; hence, this sensor material is ideally suited for determination of potassium in diluted serum. Serum dilution also is known to considerably prolong the operational lifetime of sensor membranes.

(4) It also overcomes a fundamental limitation of the previous approach of Simon et al, an insufficient sensitivity at near-neutral pH.

Comparison with Other Methods. Dynamic ranges (DR) , limits of detection (LD) , and response times (RT) of the membranes used in this method have been compared with those of other methods, described above, (Table 1) . The data show the method presented here to have a much smaller limit of detection, typically 1 μ mole. This probably can be improved even more by increasing the absorber dye-to-fluorophore ratio. The low detection limit overcomes a major problem of the Simon typ optrodes based on the ion exchange mechanism whose limits o detection can be on the order of 10 to 50 μ mole. Th response times of the present invention, however, are longer, probably due to the heterogeneity of the sensing membrane. The fluorophore beads are not homogeneously dispersed, s diffusion is slow and not free in all directions. .

Table 1

The improvement in the limit of detection caused by the inne filter effect was first reported by Walt et al. In that case, however, pH modulates both dyes employed; while the absorbanc of one species (the absorber) goes in one direction, that o the fluorophore goes in the other. The enhanced sensitivit of the present system, where only one dye is analyte sensitive, is attributed to the relative excess of primar absorber over the fluorophore. As a result, even smal changes in the fraction of the acid-to-base form of the dy cause a substantial change in the optical density of th membrane and, hence, in its permeability for fluorescen light. The fluorescent particles act as a light source insid the membrane, and the dye (KFU 111) as the modulator that act as the screen for fluorescence.

SUBSTITUTESHEET(RUtE26) Changes and modifications in the specifically describe embodiments may be carried out without departing from th scope of the invention which is intended to be limited onl by the scope of the appended claims.

Claims

1. An optical chemical sensor, comprising: a membrane; an ion selective carrier in the membrane fo extracting a target ion from a solution into th membrane when the membrane is placed in contac with the solution; a charge exchanging or counterion coextractin absorber in the membrane which maintains electrica neutrality by charge exchange or counterio coextraction when target ions are extracted by th carrier and thereby charges its absorbance; a fluorescer in the membrane having an excitatio band or emission band which overlaps an absorptio band of the absorber; whereby fluorescent emission from the fluorescer i modulated by changes in the absorbance of th absorber caused by the target ion.

2. The sensor of Claim 1 wherein the carrier an absorber are homogeneously distributed in th membrane.

3. The sensor of Claim 2 wherein the fluorescer i non-homogeneously distributed in the membrane.

4. The sensor of Claim 3 wherein the fluorescer is i the form of microspheres in sizes from 0.01 to 10. μm distributed in the membrane.

5. The sensor of Claim 1 wherein the absorber i insoluble and is selected for proper pK_a and th fluorescer is insoluble and is selected for hig quantum yield, stability and long wavelengt excitation and emission.

6. The sensor of Claim 1 wherein the membrane include a plasticizer.

7. The sensor of Claim 1 further comprising a sourc for exciting the fluorescer and a detector fo detecting fluorescent emission from the fluorescer.

8. The sensor of Claim 1 wherein the membrane is a plasticized PVC membrane.

9. The sensor of Claim 1 wherein the absorber is 4- ( 2 ' - 4 ' - d i n i t r o b e n z e n e a z o ) - 2 - octadecylamindocarbonyl-1-naphthol.

10. The sensor of Claim 1 wherein: a. The target ion is potassium and the ion carrier is valinomycin; b. The target ion is ammonium and the ion carrier is nonactin. 26 fluorescer is insoluble and is selected for high quantum yield, stability and long wavelength excitation and emission.

15. The sensor of Claim 12 wherein the optical means includes a LED light source.

16. The sensor of Claim 12 wherein the membrane is a plasticized PVC membrane.

17. The sensor of Claim 12 wherein the absorber is 4- ( 2 ' - 4 ' - d i n i t r o b e n z e n e a z o ) - 2 - octadecylamindocarbonyl-1-naphthol.

18. The sensor of Claim 12 wherein: a. The target ion is potassium and the ion carrier is valinomycin; b. The target ion is ammonium and the ion carrier is nonactin.

19. The sensor of Claim 12 wherein the target ion is potassium, the membrane is a plasticized PVC membrane, the absorber is 4-(2'-4'- dinitrobenzeneazo) -2-octadecylamindocarbonyl-l- naphthol, and the ion carrier is valinomycin.

20. An absorber dye comprising 4-(2'-4'- dinitrobenzeneazo) -2-octadecylamindocarbonyl-l- naphthol. 25

11. The sensor of Claim 1 wherein the target ion potassium, the membrane is a plasticized P membrane, the absorber is 4-(2'-4' dinitrobenzeneazo) -2-octadecylamindocarbony1-1 naphthol, and the ion carrier is valinomycin.

12. An optical chemical sensor, comprising: a plasticized membrane; an ion selective carrier homogeneously distribut in the membrane; a lipophilic charge exchanging or counteri coextracting absorber homogeneously distributed i the membrane; a lipophilic fluorophore in the membrane; optical means aligned with the membrane fo inputting an excitation signal to the fluoropho and detecting fluorescent emissions from t fluorophore; wherein the absorber is selected to chan absorption in response to target ion capture by th carrier and to filter the excitation or fluorescen emissions of the fluorophore.

13. The sensor of Claim 12 wherein the fluorophore i in the form of particles which are disperse through the membrane.

14. The sensor of Claim 1 wherein the absorber i insoluble and is selected for proper pK_a and th