Method for measuring oxygen concentration

The embodiments of the invention in which an exclusive property or privilege is claimed are defined as follows:

1. A method of measuring oxygen concentration in a fluid, comprising the steps of:

(a) contacting a test fluid with a sensor composition comprising a luminescent substance, whose luminescent emission is sensitive to quenching by oxygen, admixed in an oxygen-permeable matrix, the sensor composition exhibiting a nonexponential luminescent emission when irradiated with light containing wavelengths strongly absorbed by the luminescent substance,

(b) irradiating the sensor composition with light containing wavelengths strongly absorbed by the luminescent substance,

(c) terminating the irradiation of step (b),

(d) measuring the flux of luminescent light emitted by the sensor composition during at least two time intervals comprising at least one time interval subsequent to step (c),

(e) comparing the flux values measured in step (d) to obtain a ratioing R value,

(f) determining the oxygen concentration in the test fluid by comparing the ratioing R value obtained in step (e) with a similarly obtained ratioing R value for at least one reference fluid of known oxygen concentration.

2. The method of claim 1 wherein the at least two time intervals for flux measurement in step (d) substantially encompass the period of linear decay of luminescent emission by the sensor composition subsequent to step (c).

3. The method of claim 2 wherein the at least two time intervals subdivide the period of linear decay of luminescent emission by the sensor composition subsequent to step (c) into substantially equal time intervals.

4. The method of claim 2 wherein the at least two time intervals are two in number.

5. The method of claim 4 wherein the ratioing R value is given by:

R=(I.sub.1 -I.sub.2)/(I.sub.1 +I.sub.2)

where I.sub.1 is the flux of luminescent emission by the sensor composition measured during a first time interval and I.sub.2 is the flux of luminescent emission by the sensor composition measured during a second time interval.

6. The method of claim 1 wherein the at least two time intervals for flux measurement in step (d) substantially encompass the period of detectable luminescent emission by the sensor composition subsequent to step (c).

7. The method of claim 6 wherein at least one of the at least two time intervals substantially encompasses the period of linear decay of luminescent emission by the sensor composition subsequent to step (c).

8. The method of claim 6 wherein the ratioing R value is given by:

R=I.sub.1 /I.sub.2

wherein I.sub.1 is the flux of luminescent emission by the sensor composition measured during the period of linear decay of luminescent emission by the sensor composition subsequent to step (c) and I.sub.2 is the flux of luminescent emission by the sensor composition measured during any remaining period of detectable luminescent emission by the sensor composition.

9. The method of claim 1 wherein step (b) takes place during at least one of the at least two time intervals for flux measurement.

10. The method of claim 9 wherein at least one first time interval substantially encompasses the period leading up to an emission intensity plateau by the sensor composition during step (b) and at least one second time interval substantially encompasses the period of detectable luminescent emission by the sensor composition subsequent to step (c).

11. The method of claim 10 wherein the ratioing R value is given by:

R=I.sub.1 /I.sub.2

wherein I.sub.1 is the flux of luminescent emission by the sensor composition measured during the at least one first time interval and I.sub.2 is the flux of luminescent emission by the sensor composition measured during the at least one second time interval.

12. A method of measuring oxygen concentration in a fluid, comprising the steps of:

(a) contacting a test fluid with a sensor composition comprising a phosphorescent substance, whose phosphorescent emission is sensitive to quenching by oxygen, admixed in an oxygen-permeable matrix, the sensor composition exhibiting a nonexponential phosphorescent emission when irradiated with light containing wavelengths strongly absorbed by the luminescent substance,

(b) irradiating the sensor composition with light containing wavelengths strongly absorbed by the phosphorescent substance,

(c) terminating the irradiation of step (b),

(d) measuring the intensity of phosphorescent light emitted by the sensor composition at a plurality of times subsequent to step (c),

(e) fitting the measured intensity values as follows:

I(t)=A.sub.1 e.sup.-k.sbsp.1.sup.t +A.sub.2 e.sup.-k.sbsp.2.sup.t

wherein I(t) is the measured intensity at time t, e is the exponential function, and A.sub.1, k.sub.1, A.sub.2, and k.sub.2 are fitting parameters,

(f) determining the average decay rate of phosphorescent emission by the sensor composition, k, from the fitting parameters in step (e) as follows:

k=(A.sub.1 k.sub.1 +A.sub.2 k.sub.2)/(A.sub.1 +A.sub.2),

(g) determining the oxygen concentration in the test fluid by comparing the average decay rate determined in step (f) with a similarly obtained k value for at least one reference fluid of known oxygen concentration.

13. A method of measuring oxygen concentration in a fluid, comprising the steps of:

(a) contacting a test fluid with a sensor composition comprising a phosphorescent substance, whose phosphorescent emission is sensitive to quenching by oxygen, admixed in an oxygen-permeable matrix, the sensor composition exhibiting a nonexponential phosphorescent emission when irradiated with light containing wavelengths strongly absorbed by the phosphorescent substance,

(b) irradiating the sensor composition with light containing wavelengths strongly absorbed by the phosphorescent substance,

(c) terminating the irradiation of step (b),

(d) measuring the intensity of phosphorescent light emitted by the sensor composition at a plurality of times subsequent to step (c),

(e) fitting the measured intensity values as follows:

I(t)=A.sub.1 e.sup.-k.sbsp.1.sup.t +A.sub.2 e.sup.-k.sbsp.2.sup.t

wherein I(t) is the measured intensity at time t, e is the exponential function, and A.sub.1, k.sub.1, A.sub.2, and k.sub.2 are fitting parameters,

(f) determining the average decay time of phosphorescent emission by the sensor composition, .tau., from the fitting parameters in step (e) as follows:

.tau.=(A.sub.1 k.sub.1.sup.-1 +A.sub.2 k.sub.2.sup.-1)/(A.sub.1 +A.sub.2),

(g) determining the oxygen concentration in the test fluid by comparing the average decay time determined in step (f) with a similarly obtained .tau. value for at least one reference fluid of known oxygen concentration.

14. A method of measuring oxygen concentration in a fluid, comprising the steps of:

(a) contacting a test fluid with a sensor composition comprising a luminescent substance, whose luminescent emission is sensitive to quenching by oxygen, admixed in an oxygen-permeable matrix, the sensor composition exhibiting a nonexponential luminescent emission when irradiated with light containing wavelengths strongly absorbed by the luminescent substance,

(b) irradiating the sensor composition with light containing wavelengths strongly absorbed by the luminescent substance,

(c) terminating the irradiation of step (b),

(d) measuring the intensity of luminescent light emitted by the sensor composition at a plurality of times subsequent to step (c),

(e) from the measured intensities in step (d), determining the slope of emission decay during the period of linear decay of luminescent emission by the sensor composition,

(f) determining the oxygen concentration in the test fluid by comparing the slope value obtained in step (e) with a similarly obtained slope value for at least one reference fluid of known oxygen concentration.

15. A method of measuring oxygen concentration in a fluid, comprising the steps of:

(a) contacting a plurality of reference fluids of known oxygen concentration with a sensor composition comprising a luminescent substance, whose luminescent emission is sensitive to quenching by oxygen, admixed in an oxygen-permeable matrix, the sensor composition exhibiting a nonexponential luminescent emission when irradiated with light containing wavelengths strongly absorbed by the luminescent substance,

(b) irradiating the sensor composition with light containing wavelengths strongly absorbed by the luminescent substance,

(c) terminating the irradiation of step (b),

(d) measuring the intensity of luminescent light emitted by the sensor composition during a plurality of times subsequent to step c),

(e) from the measured intensities in step (d), determining the intensity versus time profile of emission decay for each reference fluid during the period of linear decay of luminescent emission by the sensor composition,

(f) contacting a test fluid of unknown oxygen concentration with the sensor composition and repeating steps (b) and (c) with the test fluid,

(g) measuring the time it takes for the intensity of luminescent light emitted by the sensor composition subsequent to step (f) to fall to a predetermined intensity value which can be referenced against the intensity versus time profiles determined in step (e), and

(h) determining the oxygen concentration in the test fluid by comparing the time value measured in step (g) with the intensity versus time profiles determined in step (e) for the reference fluids of known oxygen concentrations.

16. A method of measuring oxygen concentration in a fluid, comprising the steps of:

(a) contacting a plurality of reference fluids of known oxygen concentration with a sensor composition comprising a luminescent substance, whose luminescent emission is sensitive to quenching by oxygen, admixed in an oxygen-permeable matrix, the sensor composition exhibiting a nonexponential luminescent emission when irradiated with light containing wavelengths strongly absorbed by the luminescent substance,

(b) irradiating the sensor composition with light containing wavelengths strongly absorbed by the luminescent substance,

(c) terminating the irradiation of step (b),

(d) measuring the intensity of luminescent light emitted by the sensor composition during a plurality of times subsequent to step (c),

(e) from the measured intensities in step (d), determining the intensity versus time profile of emission decay of each reference fluid during the period of linear decay of luminescent emission by the sensor composition,

(f) contacting a test fluid of unknown oxygen concentration with the sensor composition and repeating steps (b) and (c) with the test fluid,

(g) measuring the intensity of luminescent light emitted by the sensor composition at a given time interval subsequent to step (f), which time interval falls within the period of linear decay for the sensor composition, and

(h) determining the oxygen concentration in the test fluid by comparing the intensity value measured in step (g) with the intensity versus time profiles determined in step (e) for the reference fluids of known oxygen concentrations.

17. The method of claims 1, 12, 13, 14, 15, or 16 wherein the test fluid is blood.

18. The method of claims 1, 12, or 13 wherein the test fluid is blood and the at least one reference fluid is air.

19. The method of claims 1, 12, 13, 14, 15, or 16 wherein the irradiating light and emitted light are transmitted by fiber optic means.

20. The method of claim 19 wherein the sensor composition is positioned in a light path of a single optical fiber.

21. The method of claims 1, 14, 15, or 16 wherein the luminescent substance is selected from the group consisting of phosphorescent and fluorescent substances.

22. The method of claim 21 wherein the luminescent substance is a metallo derivative of a porphyrin, chlorin, bacteriochlorin, or isobacteriochlorin.

23. The method of claim 22 wherein the luminescent substance is octaethylporphyrin, tetraphenylporphyrin, tetra(pentafluorophenyl)porphyrin, tetrabenzporphyrin, or the chlorins, bacteriochlorins, or isobacteriochlorins of said porphyrins.

24. The method of claim 22 wherein the luminescent substance is a platinum or palladium derivative of a porphyrin, chlorin, bacteriochlorin, or isobacteriochlorin.

25. The method of claim 21 wherein the luminescent substance is partially or wholly fluorine substituted.

26. The method of claims 1, 14, 15, or 16 wherein the luminescent substance comprises one or both of a platinum derivative and a palladium derivative of a tetra(pentafluorophenyl)porphyrin.

27. The method of claims 1, 12, 13, 14, 15 or 16 wherein the oxygen-permeable matrix is a plastic.

28. The method of claim 27 wherein the plastic comprises one or more of polyvinyl chloride, polymethyl metacrylate, cellulose acetate, and silicon-polybicarbonate copolymer, with or without a plasticizer.

29. The method of claim 12 or 13 wherein the phosphorescent substance is a metallo derivative of a porphyrin, chlorin, bacteriochlorin, or isobacteriochlorin.

30. The method of claim 29 wherein the phosphorescent substance is octaethylporphyrin, tetraphenylporphyrin, tetra(pentafluorophenyl)porphyrin, tetrabenzporphyrin, or the chlorins, bacteriochlorins, or isobacteriochlorins of said porphyrins.

31. The method of claim 29 wherein the phosphorescent substance is a platinum or palladium derivative of a porphyrin, chlorin, bacteriochlorin, or isobacteriochlorin.

32. The method of claim 29 wherein the phosphorescent substance is partially or wholly fluorine substituted.

33. The method of claim 12 or 13 wherein the phosphorescent substance comprises one or both of a platinum derivative and a palladium derivative of a tetra(pentafluorophenyl)porphyrin.

Description Submit all comments and votes

FIELD OF THE INVENTION

This invention relates generally to the measurement of oxygen concentration using the quenching of emission of a luminescent aromatic molecule embedded in a plastic medium.

BACKGROUND OF THE INVENTION

It is known that a luminescent aromatic molecule embedded in plastic is subject to quenching by oxygen present in the gas or liquid in contact with the plastic. This phenomenon was reported by Bergman (Nature 218:396, 1966), and a study of oxygen diffusion in plastic was reported by Shaw (Trans. Faraday Soc. 63:2181-2189, 1967). Stevens, in U.S. Pat. No. 3,612,866, ratios the luminescence intensities from luminescent materials dispersed in oxygenpermeable and oxygen-impermeable plastic films to determine oxygen concentration. Lubbers et al. in U.S. Pat. No. 4,003,707 proposed the possibility of positioning the emitting substance at the end of an optical fiber. Peterson et al. in U.S. Pat. No. 4,476,870 also employs the quenching of an emitting molecule in plastic at the end of an optical fiber. Both Lubbers and Peterson reference emission against scattered exciting light.

The quenching of the luminescence of an emitter at the end of an optical fiber has been used in temperature sensors. For temperature probes the emitters are generally solid phosphors rather than an aromatic molecule embedded in plastic, since access by molecules from the environment is not desirable. Various methods have been used to measure the amount of quenching: (i) Quick et al. in U.S. Pat. No. 4,223,226 ratios the intensity at one wavelength of the emission against another; (ii) Quick et al. also proposes determining the length of time it takes for the signal to fall from one level to another; (iii) Samulski in U.S. Pat. No. 4,245,507 (reissued as U.S. Pat. No. Re. 31,832) proposes to measure quenching by determining the phase of the emitted life. In a very recent patent for temperature sensing at the end of an optical fiber, Hirschfeld in U.S. Pat. No. 4,542,987 proposes, in addition to method (i), that (iv) emission lifetime be used to measure quenching and that (v) Raman scattered light can be used as a reference.

Eastwood and Gouterman (1970) noted generally with respect to Pd and Pt porphyrin complexes that their "relatively high [emission] yields and short triplet lifetimes . . . may make these systems useful as . . . biological probes for the presence of oxygen." More recently, Bacon and Demas in UK Patent Application No. 2,132,348A propose the use of, inter alia, porphyrin complexes of VO.sup.2+, CU.sup.2+, Pt .sup.2+, ZN.sup.2+ and Pd.sup.2+ or dimeric Rh, Pt, or Ir complexes of monitoring oxygen concentration by emission quenching of intensity or lifetime. Suitable ligands would reportedly be etioporphyrin, octaethylporphin, and porphin.

SUMMARY OF THE INVENTION

Methods and compositions are described for measuring oxygen concentration, particularly for monitoring oxygen in the blood with a fiber optic catheter. Oxygen concentration is determined by observing quenching of the emission from a luminescent (phosphorescent or fluorescent) molecule embedded in oxygen-permeable plastic. A test fluid of unknown oxygen concentration is contacted with a plastic film containing at least one luminescent substance. The film is subjected to irradiation over some period of time by light of a wavelength that is strongly absorbed by the luminescent substance, and a measure of the time dependence of luminescent emission intensity I(t) is obtained. Three modes of determining oxygen concentration from I(t) are described. (i) Subsequent to a brief (approximately 5 us) flash of light I(t.sub.i) is determined by use of a transient recorder and fit to Eq. (6). An average decay rate

k=(A.sub.1 K.sub.1 +A.sub.2 k.sub.2)/(A.sub.1 +A.sub.2) (7)

is determined, and k used for the Stern-Volmer plot of Eq. (1). (ii) The period of linear decay of luminescent emission is determined from the I(t) data, and that intensity versus time profile is referenced against similarly obtained profiles for reference fluids of known oxygen concentration, using slopes, intensity or time setpoints. (iii) The sample is irradiated for some time interval (generally under 50 us) using a flash lamp or a light emitted diode. Intensity segments of emission, I.sub.1 and I.sub.2, are determined during two time intervals defined with respect to the time of irradiation. These two intensity segments are compared to form a ratio R, and a calibration plot of R versus oxygen pressure is obtained using solutions of known oxygen concentration. Three different methods for determining I.sub.1, I.sub.2, and R are presented. These methods (i-iii) of measuring quenching are insensitive to variation in plastic thickness and emitter concentration of the probe and to decomposition during operation. Methods (i-iii) also take into account the non-exponential decay of the emission, and thus extend the pressure range over which the probe is sensitive. Method (iii) requires at most one point calibration against atmospheric oxygen.

Also disclosed are photostable luminescent molecules for use with the subject method. In a preferred embodiment platinum tetra(pentafluorophenyl)porphyrin, Pt(TFPP), serves as the luminescent oxygen quenchingsensitive molecule. Pt(TFPP) has a strong absorbance in the visible region, a strong phosphorescence with lifetime of roughly 100 .mu.s, and is photostable. Photostability is provided by the substitution of fluorine atoms in the periphery of the synthetic porphyrin ring. Other suitable fluorinated luminescent molecules include metallo derivatives, particularly platinum and palladium derivatives, of partially or fully fluorinated octaephylporphyrin, tetraphenylporphyrin, tetrabenzoporphyrin, or the chlorins, bacteriochlorins, or isobacteriochlorins thereof. The latter reduced porphyrins have the advantage that their absorption is red-shifted to a region for which light emitting diodes can be used for excitation.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram of a representative system 10 for measuring oxygen concentration by the emission lifetime method of this invention;

FIG. 2 is a graph showing phosphorescence emission of Pt(TFPP) decay curves as a function of molecular oxygen pressures, as described in Example 1;

FIG. 3 is a graph plotting k/k.sub.o and .tau..sub.o /.tau. versus pO.sub.2 for the decay curves shown in FIG. 2 and the data presented in Table 1;

FIG. 4 is a representative plot showing the two intensity segments I.sub.1 and I.sub.2 as used in Equation (9) in conjunction with system 10;

FIG. 5 is a plot of R versus pO.sub.2 for the embodiment shown in FIG. 4;

FIG. 6 is a block diagram of representative systems 10' and 10" suitable for monitoring oxygen concentration in the bloodstream;

FIG. 7 is a representative plot showing the two intensity segments I.sub.1 ' and I.sub.2 ' as used in Equation (10) in conjunction with system 10';

FIG. 8 is a plot of R' versus pO.sub.2 for the embodiment shown in FIG. 7;

FIG. 9 is a representative plot showing the two intensity segments I.sub.1 " and I.sub.2 " as used in Equation (11) in conjunction with system 10";

FIG. 10 is a plot of R" versus pO.sub.2 for the embodiment shown in FIG. 9;

FIG. 11 presents Stern-Volmer plots for Pt(TFPP) in various film matrices containing various amounts of plasticizer, as described in Example 3; and,

FIG. 12 presents Stern-Volmer plots of k/k.sub.o for films containing various mixtures of Pt(TFPP) and Pd(TFPP), as described in Example 4.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT

The present invention addresses several previously unrecognized problems inherent in the prior art. In particular, we have found that the molecules studied by Eastwood and Gouterman are subject to photodecomposition when exposed to light and oxygen for extended periods of time. For best photostability the previously unreported fluorinated derivative platinum tetra(pentafluorophenyl)porphyrin, Pt(TFPP), must be used. (As used herein, the term "tetra(pentafluorophenyl)porphyrin" refers to the identical compound as the term "tetraperfluorophenylporphyrin" as stated in our referenced prior application.)

Furthermore, we have found that the phosphorescent emission decay in plastic is not a simple exponential, and so conventional methods of analysis fail to provide a measure of quenching that is useful over a wide range of oxygen pressures. In the next paragraphs we review the luminescence and quenching of aromatic molecules and then present our methods for using the observed nonexponential decay to accurately and conveniently determine oxygen concentration.

Photoexcitation of many molecular systems leads to metastable excited states that relax to the ground state by emitting light. Metastable photoexcited states are of two types: (i) those with the same spin as the ground state and (ii) those of different spin from the ground state. Emission from metastable excited states of type (i) is called fluorescence and is generally accomplished in under 0.2 microseconds. Emission from metastable excited states of type (ii) is called phosphorescence and is generally accomplished in times ranging from 1 microsecond to 20 seconds.

Both types of emission are quenched by oxygen. The first order decay rate of emission (k) of an excited state in the presence of oxygen is given by:

k=k.sub.nat +k.sub.d +k.sub.q [O.sub.2 ]=k.sub.o +k.sub.q [O.sub.2 ]=1/.tau.(1)

wherein k.sub.nat is the natural radiative lifetime of the metastable excited state, k.sub.d is the rate of any intrinsic radiationless decay processes, k.sub.q is the quenching rate of oxygen, [O.sub.2 ] is the oxygen concentration, and k.sub.o is the decay rate in the absence of oxygen. The inverse of k is called the emission lifetime, .tau..

The amount of light emitted by a sample is called the quantum yield, .phi., is defined as:

.phi.=(photons emitted/photon absorbed)=k.sub.nat /k. (2)

It follows from Equations (1) and (2) that the quantum yield in the absence of oxygen, .phi..sub.o, divided by the quantum yield in the presence of oxygen, .phi., given by:

.phi..sub.o /.phi.=1+k.sub.q [O.sub.2 ]. (3)

Equations (1) and (3) are two forms of the Stern-Volmer equation for the effect of oxygen on quantum yield. This relationship is the basis for oxygen monitors that study the intensity of emission (I), which is directly proportional to quantum yield as follows:

I.sub.o /I=1+k.sub.q [O.sub.2 ] (4)

wherein I.sub.o and I are respectively the emission intensities in the absence and presence of oxygen.

It can be seen that oxygen quenching can be determined by monitoring either decay rate as in Equation (1) or decay intensity as in Equation (4). For both types of measurements it is necessary to calibrate the sensor against reference solutions of known oxygen concentrations in order to determine the quenching parameter k.sub.q. If decay rate is to be used, it is necessary to measure the emission intensity at a series of times, t.sub.1, t.sub.2, t.sub.3 . . . , after the exciting light has terminated. For a simple exponential decay two such timed measurements will suffice, and decay rate is determined from the equation:

k=(t.sub.2 -t.sub.1).sup.-1 1n[I(t.sub.1)/I(t.sub.2)]. (5)

If intensity is used as in Equation (4), then it is necessary to reference intensity I to the intensity I.sub.o in the absence of oxygen. As mentioned above, Stevens proposed obtaining I.sub.o from a second probe from which oxygen is excluded by an oxygen impermeable varnish. The Stevens configuration is impractical for fiber optic probes. Lubbers and Peterson determine I.sub.o from scattered light. Both methods of measuring I.sub.o have the intrinsic limitation that they do not take account of photodegradation of the sensing molecule that may occur during the operation and that would reduce light output. Also, depending on the ability to manufacture exactly identical sensors, each particular probe may need calibration. Accordingly, Peterson calibrates each probe at three points: no oxygen, oxygen at the pressure of air, and at an intermediate oxygen pressure.

In contrast to the intensity measurement of Equation (4), the decay rate measurement of Equation (5) avoids difficulties caused by variability in sensor manufacture and by photodegradation. However, the oxygen sensors of Stevens, Lubbers et al., and Peterson et al. are all based on the quenching of fluorescence, which decays generally in under 0.1 microsecond (.mu.s) and so requires fast electronics and fast light flashes to measure decay rate. This may account for their preferred use of the intensity ratio of Equation (4). The molecule used in the preferred embodiment of our oxygen sensor has a decay time in the 100 .mu.s time range, so that slower electronics and exciting light flashes can be used, making the study of quenching through decay rate very practical. Furthermore, the use of decay rate as in Equation (1) rather than intensity ratio as in Equation (4) in principle makes our sensor insensitive to both variation in the probe manufacture (e.g., small variations in plastic thickness and emitter concentration) and to the formation of nonluminescent photoproducts during operation. In principle no calibration is necessary, but in practice we have found that one point calibration in air increases the accuracy of measurement by different probes.

While Equations (1) and (4) are generally considered to be equivalent measures of oxygen quenching, a serious problem arises in determining decay rate of aromatic molecules in plastic, since we find that such decay is generally nonexponential. That is, the decay time of luminescent aromatic molecules in plastic cannot be fit by a simple exponential but must rather be considered as a sum of two exponentials:

I(t)=A.sub.1 e.sup.-k.sbsp.1.sup.t +A.sub.2 e.sup.-k.sbsp.2.sup.t (6)

wherein e is the exponential function. The four fitting parameters A.sub.1, k.sub.1, A.sub.2, k.sub.2 can be fit if I(t) is measured at many times t following an interval of photoexcitation. This type of determination requires considerable instrumentation and software analysis that may not be practical for routine operation. Furthermore, even given the knowledge of these four parameters, it is not clear how best to employ them to determine oxygen concentration because in the absence of an exponential decay the Stern-Volmer Equations (1) and (3) no longer apply.

To overcome these problems, we provide molecules having a decay time sufficiently long so that oxygen concentration can be conveniently determined through a measurement of the time dependence of emission intensity, i.e., I(t). We thereby avoid problems due to variations in probe construction and photodegradation that arise using methods that measure total intensity and its ratio to a reference, i.e., as in Equation (4). We also provide convenient methods of determining oxygen concentration from the function I(t), even though I(t) is non-exponential.

FIG. 1 shows a representative system 10 for determining the oxygen concentration of gaseous samples by measuring the quenching of various emitting sensor compositions in plastic. A flashing light source 12 (e.g., Strobotac Model No. 1538A) provides time dependent light excitation (indicated by dashed arrow 14) of an oxygen quenching-sensitive composition, which is sequestered in film 16 in this embodiment. Film 16 is mounted inside the fluid, depicted as vapor 18 here, that is to be sampled. Phosphorescent light (dashed arrow 20) emitted from film 16 impinges on a photodetector 22 (e.g., RCA 7265). A housing 24 containing windows 26, 28 can be used to isolate film 16 inside the fluid 18 being monitored, which is typically not identical with the environment of the rest of system 10. Windows 26, 28 can be made of quartz or glass.

Key components of system 10 are filters 30, 32. Filter 30 is a band-pass filter (e.g., Ealing 35-3649) that allows only shorter wavelength light 14, e.g., wavelengths in the range 480 to 600 nanometers (nm), to impinge on film 16. The range of band-pass filter 30 is chosen to match the region of strong absorption by the phosphorescing compound that is sequestered in film 16. Filter 32 is a cutoff filter (e.g., Corning 261) that allows only long wavelength light 20, e.g., wavelengths longer than 620 nm, to impinge on photodetector 22 and is chosen to allow the phosphorescent light 20 to reach the photodetector 22. In a preferred embodiment, the band-pass of filter 30 and the cutoff of filter 32 are complementary such that no light 14 from flashing light source 12 reaches photodetector 22.

The electric output (arrow 38) of photodetector 22 passes into preamplifier 40 of standard design. In the embodiment shown here, the output (arrow 42) of preamplifier 40 passes into a transient recorder 44 (e.g., Biomation Model No. 805) that is capable of sampling intensity, I(t.sub.1), at time intervals below a microsecond. The timing of system 10 is under control of microcomputer 36. In a preferred mode of operation that provides a zero baseline, microcomputer 36 puts out a trigger (arrow 46) to active transient recorder 44 and after a slight delay on the order of 20 microseconds puts out another trigger (arrow 48) to start flash light source 12. Transient recorder 44 thus collects data relating to I(t.sub.1), the intensity I of phosphorescence at various times t.sub.1 before and after the flash, which data are read (as indicated by arrow 50) into the computer 36.

Flash rates using system 10 are typically on the order of 100 per second. The time interval between successive flashes 14 is sufficiently long that a decay time and hence oxygen concentration can be calculated after each flash 14. It should be noted that the limiting time of response is set by the diffusion rate of oxygen into the monitor film 16, which is typically under one second. This diffusion rate becomes faster, permitting response times on the order of milliseconds, with thinner films 16 and by adding plasticizer to the carrier matrix, as described below. Thus, real-time measurements of oxygen concentrations can be made by this method.

Pursuant to this disclosure, measurement is made of the intensity, I(t.sub.i), of phosphorescence 20 emitted by the film 16 at a series of times, t.sub.1, t.sub.2, t.sub.3 . . . , after the flash 14. The light intensities I(t.sub.i) at any particular time t.sub.i following several flashes can e averaged. A decay rate, k, of the phosphorescence is calculated from these data by the computer 36 using various algorithms. In particular, with a full set of values I(t.sub.i) we can fit the decay to Eq. (6) above. Decay curves for Pt(TFPP) in polyvinyl chloride with plasticizer are shown in FIG. 2, wherein t' is the period of linear decay, i.e., where .DELTA.I/.DELTA.t has a constant slope at any particular oxygen pressure of interest, and wherein t" is the remaining period of the detectable emission, during which .DELTA.I/.DELTA.t is not constant. Representative values found for the fitting parameters A.sub.1, k.sub.1, A.sub.2, k.sub.2 are listed in Table 1.

TABLE 1 ______________________________________ Data analysis of PtTFPP in PLS/PVC.sup.a,b,c. p0.sub.2 A.sub.1 k.sub.1 A.sub.2 k.sub.2 - k - .tau. - k/k.sub.o .tau..sub.o /- .tau. ______________________________________ 1 1. 13.1 13.1 73.3 1. 1. 10 0.88 13.26 0.12 24.2 14.6 71.3 1.11 1.07 50 0.78 15.03 0.22 26.5 17.6 60.3 1.34 1.27 100 0.64 16.23 0.36 31.4 21.7 50.9 1.66 1.50 200 0.63 18.25 0.31 41.5 26.8 43.4 2.05 1.76 300 0.58 19.23 0.42 45.4 30.2 39.4 2.30 1.94 400 0.33 17.0 0.67 39.4 32.0 36.5 2.44 2.09 500 0.305 17.2 0.695 41.1 33.8 34.6 2.58 2.20 ______________________________________ .sup.a PLS/PVC = plasticized polyvinyl chloride; see Example 1 for details. .sup.b Data taken by system 10. .sup.c The k's are in (ms).sup.-1, and -.tau. is in .mu.s.

Two measures for the time decay can be defined: Average decay rate, k, is given by:

k=(A.sub.1 k.sub.1 +A.sub.2 k.sub.2)/(A.sub.1 +A.sub.2). (7)

Average decay time, .tau., corresponds to the normalized integral of the emission intensity, I(t), as follows:

.tau.=(A.sub.1 k.sub.1.sup.-1 +A.sub.2 k.sub.2.sup.-1)/(A.sub.1 +A.sub.2). (8)

.tau. represents the total amount of light emitted following termination of excitation, which has been the principal measure of oxygen quenching in the prior art; however, the prior art did not consider the disclosed nonexponential decay phenomenon. The value k corresponds to the normalized decay rate at t=0. in FIG. 3 we plot representative .tau..sub.o /.tau. and k/k.sub.o curves using the data listed in Table 1. These curves would be identical for an exponential decay. FIG. 3 shows that k/k.sub.o gives a more linear Stern-Volmer plot; hence k provides a more accurate determination of oxygen concentration at higher oxygen pressures.

The double exponential decay of Eq. (6) can be understood as resulting from two types of emitting molecules: Those with larger k.sub.1 are more subject and those with smaller k.sub.2 are less subject to quenching by oxygen. The average decay time, .tau., gives heavier weight to the unquenched molecules, whereas the average decay rate, k, gives heavier weight to the quenched population and provides a better measure of oxygen quenching.

In a related method, the slope of the emission profile during the period of linear decay is compared with similarly obtained slopes for fluids of known oxygen concentrations. Due to the double exponential nature of the luminescent decay curve, the referenced slope values must not encompass any of the tail region (t" in FIG. 2) of the emission profile. Once this relationship is established, the time it takes for the intensity of emission to fall to any particular level within time t' can provide a convenient readout of pO.sub.2. Alternatively, the intensity measured with the test fluid at any particular time t less than t' can be compared with standard curves of intensity versus time (again, less than t') for a series of fluids