WO1996041177A1 - Simultaneous assay method using lanthanide chelates as the luminophore for multiple labels - Google Patents

Abstract

An electrochemiluminescent (ECL) detection method for simultaneous assays. The method employs an excitation process whereby one or more luminophores are excited through an antenna mechanism using an oxidative or reductive coreactant. The luminophores can be lanthanide chelates and a persulfate system is used to generate ECL. The ECL emissions may be separated either by measuring at different emission wavelengths or by electrode potentials.

Description

SIMULTANEOUS ASSAY METHOD USING LANTHANIDE

CHELATES AS THE LUMINOPHORE FOR MULTIPLE LABELS

BACKGROUND OF THE INVENTION

Field of the Invention

The present invention relates to electrochemiluminescence (ECL) detection methods.

More particularly the invention has to do with a system using multiple ECL labels for simultaneous assays.

Description of Related Art

In recent years there has been a great deal of interest in the use of lanthanides in fluoroimmunoassays. This is due to the intense, delayed emission of the lanthanides which allows the development of time-resolved fluorometric assays which are competitive with radioimunoassays. Soini, E.; Hemmila, Clin, Chem. 1979, 25, 353. Since efficient population of the lanthanide excited state cannot be achieved via direct excitation, excited state generation typically is accomplished via sensitization from a ligan excited state.

Emission intensity is also dependent upon the extent of non-radiative deactivation of the lanthanide excited state by solvent interactions. Therefore, the more luminescent lanthanide complexes are composed of heteroaromatic ligands which encapsulate the metal. The photophysical properties of encapsulated lanthanides have recently been reviewed.

Sabbatini, N.; Guardigli, M.; Lehn. J.-M., Coord. Chem. Rev. 1993, 123, 201.

There is very little mention in the scientific literature of lanthanide chelates used as luminophores in ECL processes. The lanthanides are not readily excited by ECL processes. This accounts for the relative silence in the literature, not because people are unaware of the materials. Lanthanides are very common materials used in television displays and time resolved fluorescence techniques.

Volger and Kunkely (ACS Symposium Series 333, pp 155-168, 1987) describe an ECL process using Tb(III)(TTFA)₃(o-phen), Tb(IlI)(TTFA)₄-, and Eu(III)(TTFA)₃(o-phen) where TTFA = thenoyltrifluoroacetonate and o-phen = 1,10 phenanthroline. Their ECL process used oscillating electrode voltages as the means for excitation. Weak emissions were observed. Our process utilizes a different means of excitation which produces much better emission intensities.

Hemingway, Park, and Bard (J. Am. Chem. Soc. 97, 200, 1975) describe an ECL process where the lanthanide Eu(DBM)₃PIP, DBM = dibenzoylmethide, PIP = piperidine, is activated by an energy transfer process from the ECL luminophore. In this case, the lanthanide is excited indirectly through formation of an exciplex.

A cathodic electroluminescence technique using lanthanides has been described by Kankare (for example see Anal. Chim. Acta 266, 205, 1992, Anal. Chim. Acta 256, 17, 1992). Cathodic electroluminescence is a different way of exciting than ECL techniques.

A common method of exciting lanthanide chelates is time resolved fluorescence; for example see I. Hemmilä, S. Dakubu, V.M. Mukkala, H. Siitari, T. Lovgren, Anal. Biochem. 137, 335-343 (1984).

An object of the invention was to develop an ECL detection method for simultaneous assays and we have developed a method which employs an excitation process for certain luminophores; particularly the lanthanides. A new excitation method for the lanthanides was needed because previous methods do not provide satisfactory intensities for simultaneous assays. Lanthanides, as luminophores, possess emission characteristics which enable the development of simultaneous assay. For example, four labels present in one reaction volume (such as in a simultaneous assay), each emission must be separately quantifiable. Using our new excitation process and lanthanides, the ECL emissions may be separated either by measuring at different emission wavelengths or by electrode potentials.

We have now discovered that ECL can be obtained from various lanthanide complexes and with different ligands. The level of the ECL signal for the Tb(2)³⁺ complex is high enough to yield the lowest detection limit observed to date. We employ an "Antenna ECL" mechanism which is based upon lanthanide sensitization by another excited state, an analogous mechanism by which lanthanide emission is obtained in fluoroescence immunoassays. Additionally, this mechanism suggests that it is possible to tune the ECL peak potentials without substantially changing the emission wavelengths by modifying the ligand reduction potentials. This demonstrates that ECL could compete with the lanthanide based fluoro-immunoassays.

In summary, we have developed a new ECL process for: (1) excitation of one or more luminophores through an "antenna" mechanism, using an oxidative or reductive coreactant; and

(2) excitation of one or more lanthanide chelates through an "antenna" mechanism, using an oxidative or reductive coreactant.

SUMMARY OF THE INVENTION

According to the invention, the ECL of lanthanides is achieved by "Antenna ECL". This process consists of the generation of a ligand excited state via the reaction of a reduced ligand with an oxidizing radical, the "electrochemical excitation" (ECX) steps. Energy transfer from the ligand excited state to the emissive state (i.e., lanthanide excited state) yields the characteristic lanthanide emission. Shown below are the reactions involved in the Antenna ECL process. In these reactions L represents the generic ligand, La represents a generic lanthanide metal, and an example of a coreactant which produces an oxidizing radical is presented, peroxydisulfate.

La(L)³⁺ +e- La(L)²⁺

S₂O₈ ²- +e- s₂o₈ ^3"

S₂O₈ ³- so₄ ²- + C-₄-

La(L)²⁺ +O₄- La(L*)³⁺ + S0₄

La(L*)³⁺ La*(L)³⁺

La*(L)³⁺ La(L)³⁺ + hv

wherein La represents lanthanide metal, L represents ligand, L represents a reduced ligand, L* represents a ligand excited state and La* represents a lanthanide excited state. This process has been demonstrated with five lanthanide complexes consisting of four lanthanides (Sm, Eu, Tb, and Dy) and two ligands (1 and 2), illustrated below. The complexes were Sm(2)³⁺, Eu(1)³⁺, Tb(2)³⁺, Tb(1)³⁺, and Dy(2)³⁺. LIGANDS (1 AND 2)

Each of these complexes gave ECL upon reduction of the ligand in the presence of an oxidizing radical.

We have developed the ability to access lanthanide f-f excited states through excitation of the ligand. (For discussion of lanthanide f-f excited states see N. Sabbatini, M. Guardigli, J.M. Lehn; Coord. Chem. Rev.123, 201-228 (1993), A.P.B. Sinha; Spectroscopy in Inorganic Chemistry, 2, 255-288 ( 1971 ) and J.C.G. Bunzli, D. Wessner; Coord. Chem. Rev.60 , 191-253 (1984.) The Antenna ECL process allows access to excited states which cannot be accessed directly via the normal ECL processes. This feature allows the coupling of a species (such as the antenna or ligand) which is efficient at ECX and/or with the desired reduction potential (or peak potential) to another species (such as the lanthanide metal) with the desired excited state properties (i.e., highly emissive, appropriate wavelength of emission, or non-emissive).

We also have developed the ability to tune the potential at which ECL is observed through changes in the antenna (ligand) without substantially changing the emission wavelengths. The first step of the Antenna ECL process involves a ligand reduction.

Altering the ligand structure changes the reduction potential and correspondingly, the electrode potential for onset of ECL (excitation potential). The emission spectra is unchanged because it is solely related to the type a metal lanthanide present in the chelate. This feature is illustrated by Tb(1)³⁺ and Tb(2)³⁺. Using an ORIGEN® Analyzer (available from IGEN, Inc., 16020 Industrial Drive, Gaithersburg, MD 20877 U.S.A.) the peak potential observed for Tb(1)³⁺ was ca. -3 V while that for Tb(2)³⁺ was approximately (sometimes abbreviated herein as "ca.") -5 V; however, both complexes have the

characteristic Tb³⁺ emission bands.

During the development of this invention, ECL from a Dy³⁺ and Sm³⁺ complex was first observed. Additionally, we observed for the first time a lanthanide ECL with a non- oscillating potential. This is one of three examples of aqueous reductive ECL, the other reported examples are Cr(phen)₃ ³⁺ (phen = 1 , 10-phenanthroline) by F. Bollette, M. Ciano, V. Balzani, and N. Serpone, Inorganica Chimica Acta, 1982, 62, 207 and Ru(bpz)₃ ²⁺ (bpz = 2,2'-bipyrazine) by S. Yamazaki-Nishida, Y. Harma, and K. Yamashita, Journal of

Electroanalytical Chemistry, 1990, 283, 455.

The lanthanide luminophores of the invention are used in a similar manner as with Ru(bpy)₃ ²⁺. Linker arms for attachment to various biological molecules such as antibodies are used and sandwich assays for the analytes can be carried out in the usual manner.

BRIEF DESCRIPTION OF THE DRAWINGS

Figure 1 illustrates the ligand structures 1 and 2.

Figure 2 is a cyclic voltammogram of Dy(2)³⁺ in an acetonitrile solution 0.1 molar ("M") in tetrabutylammonium perchlorate ("TBAP") and millimolar ("mM") in La³⁺ recorded on platinum electrodes in volts ("V") vs. 3M Ag/AgCl.

Figure 3 is a cyclic voltammogram of Eu(2)³⁺ using the same type of solution and electrodes as Figure 2.

Figure 4 is a graph of corrected ECL vs. potential (millivolts, abbreviated "mV") for aqueous La(1)³⁺ using Ramp ECL (wherein the potential was continuously increased at a rate of 4800 mV/s) with 1000 nanomolar ("nM") Tb(1)³⁺ and 10,000 nM Eu(1)³⁺.

Figure 5 is a graph of corrected ECL counts vs. time (centiseconds, abbreviated cs") for aqueous La(1)³⁺ using ECL Step wherein a potential pulse was applied at -5000 mV for 1000 milliseconds ("ms")

Figure 6 is a calibration curve for aqueous La(1)³⁺. Figure 7 is a graph of corrected ECL counts vs. potential (mV) for aqueous La(2)³⁺ using Ramp ECL.

Figure 8 is a graph of corrected ECL counts vs. time (cs) for aqueous La(2)³⁺ using ECL Step (-5V).

Figure 9 is a calibration curve for aqueous La(2)³⁺.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The four lanthanide complexes; Tb(2)³⁺, Dy(2)³⁺, Sm(2)³⁺, and Eu(1)³⁺; can be quantified by reductive ECL in 0.1 M phosphate buffer (pH 6) which is 0.1% in surfactant and 50 micromolar (" M") in potassium persulfate. A modified ORIGEN Analyzer was used for these analyses. The Analyzer was equipped with gold electrodes and a filter wheel running at ca. 10 hertz ("Hz") which had four narrow-band interference filters (613 nanometers ("nm") for Eu³⁺ , 545 nm for Tb³⁺, 644 nm for Sm³⁺, and 573 nm for Dy³⁺). The filter wheel was placed between the electrode and the photomultiplier tube ("PMT") (PMT model R1104 available from Hamamatsu, 360 Foothill Road, P.O. Box 6910, Bridgewater, NJ 08807 U.S. A.). A potential step of -5 V for 10 seconds ("s") was applied to generate the ECL from these complexes. Since the ECL from these complexes does not totally decay over 10 s, this long of a pulse can be utilized for integration of the signals at each wavelength. Fluorescence discrimination of these lanthanides has been described by Y.-Y. Xu and I.A. Hemmilä, Analytica Chimica Acta, 1992, 256, 9.

Based upon the highly luminescent properties of encapsulated lanthanides (La) and the narrowness of the emission bands (ca. 50 nm), six lanthanide complexes have been prepared with Eu³⁺, Tb³⁺, Dy³⁺, Sm³⁺, and two bipryidine based ligands, 1 and 2 (see Figure 1). The complexes prepared were, Eu(1)³⁺, Eu(2)³⁺, Tb(1)³⁺, Tb(2)³⁺, Sm(2)³⁺, and Dy(2)³⁺. The ECL from these complexes was evaluated using a persulfate system.

The mechanism by which ECL is generated with the persulfate system is slightly different from that in the previously studied tripropylamine ("TPA") system. In the persulfate system negative potentials are applied to the electrochemical cell; thereby, reducing the label and persulfate. The reduction of persulfate involves the injection of an electron into the peroxide bond which then cleaves yielding SO₄ ^2- and SO₄-, a strong oxidant. The radical anion then oxidizes the reduced label yielding sulfate and excited state label. These reactions are summarized in Equations (1), (2) and (3) below with Ru(bpy)₃ ²⁺ (bpy = 2,2'-bipyridne) as the label.

(1) S₂O₈ ^2- + e- SO₄ ²- + SO₄-

(2) Ru(bpy)₃ ²⁺ + e- Ru(bpy)₂(bpy-)⁺

(3) Ru(bpy)₂(bpy-) + SO₄- Ru(bpy)₃ ²⁺* + SO₄ ²-

EXAMPLES

All reference to water indicate deionized water. Aqueous stock solutions of Eu(1)³⁺ (0.662 mM) and Tb(1)³⁺ (0.986 mM) were used as received. Eight complexes with ligand 2 were prepared by a previously published procedure (Prodi, L.; Maestri, M.; Ziessel, R.; Balzani, V., Inorg. Chem, 1991, 30, 3798) and were used as received. These complexes were [Eu(2)]Cl₃, [Eu(2)](PF₆)₃, [Tb(2)]CI₃, [Eu(2)](PF₆)₃, [Dy(2)]CI₃, [Dy(2)](PF₆)₃,

[Sm(2)]Cl₃, and [Sm(2)](PF₆)₃. Three aqueous stock solutions were prepared for each of the chloride salts. In 5.50 milliliters ("mL") of water, 10.7 mg of [Tb(2)]Cl₃ was dissolved to make a 2.16 millimolar ("mM") solution. A 100 microliter ("μL") aliquot was diluted to 100 mL with water to make a 2.16 micromolar ("μM") solution, a 10 mL aliquot of this was diluted to 100 mL with water to make a 0.216 μM solution. In 6.64 mL of water, 12.0 mg of [Dy(2)]Cl₃ was dissolved to make a 2.00 mM solution. A 100 μL aliquot of the 2.00 mM solution was diluted to 100 mL with water to make a 2.00 μM solution; a 10 mL aliquot of the 2.00 μM solution was diluted to 100 mL with water to make a 0.200 μM solution. In 7.02 mL of water, 12.5 mg of [Sm(2)]Cl₃ was dissolved to make a 2.00 mM solution. A 100 μL aliquot of the 2.00 mM solution was diluted to 100 mL with water to make a 2.00 μM solution; lOmL of the 2.00 μM solution was diluted to 100 mL with water to make a 0.200 μM solution. In 6.50 mL of water, 1 1.6 mg of [Eu(2)]Cl₃ was dissolved to make a 2.00 mM solution. A 100 μL aliquot of the 2.00 mM solution was diluted to 100 mL with water to make a 2.00 μM solution and a 10 μL aliquot was diluted to 100 mL with water to make a 0.200 μM solution.

Photophysics

Ultraviolet-visible ("UV-visible") spectra were recorded with a HP 8452A Diode

Array spectrophotometer (available from Hewlett Packard Company, 2101 Gaither Road, Rockville, MD 20850 U.S.A.). A 100 μL aliquot of each mM aqueous solutions was diluted with 5.61 mL of water to make the following aqueous solutions: 37.8 μM Tb(2)³⁺ and 35.0 μM of Eu(2)³⁺, Sm(2)³⁺, and Dy(2)³⁺. UV-visible spectra (180-820 nm) were measured for each of the ca. 35 μM aqueous La(2)³⁺ solutions.

Steady state fluorescence and quantum yield measurements were made on a Perkin- Elmer LS-5 Fluorescence spectrophotometer equipped with a Kipp and Zonen X-Y recorder model BD-91 (available from VWR Scientific, P.O. Box 626, Bridgeport, NJ 08014 U.S.A.). A scan speed of 60 nm/minute ("nm/ min") was employed with the emission and excitation slits at 5 millimeters ("mm"). For quantum yield measurements, the solution absorbencies were measured with a Perkin-Elmer Lambda 4B UV-VIS spectrophotometer (available from Perkin-Elmer Corp., 761 Main Ave., Norwalk, CT 06859 U.S.A.). All solutions for the excited state measurements were air equilibrated. Aqueous solutions of the La(2)³⁺ complexes were prepared such that each had an absorbance of ca. 0.10 at 320 nm. The absorbencies were 0.105 (Tb³⁺), 0.101 (Dy³⁺), 0.106 (Sm³⁺), and 0.105 (Eu³⁺). With an excitation wavelength of 320 nm the emission spectra for each complex was measured from 400 nm to 635 nm. The excitation spectra were recorded for each complex with an emission wavelength of 544 nm (Tb³⁺), 575 nm (Dy³⁺), 568nm (Sm³⁺), and 620 nm (Eu³⁺). The instrument scale used in recording the excitation and emission spectra for each complex was 0.05 (Tb³⁺), 0.05 (Eu³⁺), 1.00 (Dy³⁺), and 2.00 (Sm³⁺). Electrochemistry

Cyclic voltammograms were recorded with a EG&G PAR 173

Potentiostat/Galvanostat (available from EG&G Princeton Applied Research, P.O. Box 2565, Princeton, NJ 08543 U.S.A.) controlled by a EG&G PARC 175 Universal

Programmer (available from EG&G Princeton Applied Research, P.O. Box 2565, Princeton, NJ 08543 U.S.A.) and plotted with a Kipp and Zonen X-Y recorder model

BD-91. Platinum working and counter electrodes were used and the voltages were referenced to a 3M Ag/AgCl electrode. The solutions were prepared from JT Baker reagent grade acetonitrile (available from VWR Scientific, P.O. Box 626, Bridgeport, NJ 08014 U.S.A.) stored over molecular sieves (4 ), which was 0.1 M in tetrabutyl ammonium perchlorate and purged with nitrogen. The cyclic voltammograms for each of the

[La(2)](PF₆)₃ complexes were measured from 2.5 V to -2.3 V. The complex concentrations were 2.3 mM (Tb³⁺), 2.1 mM (Dy³⁺), 1.9 mM (Sm³⁺), and 2.6 mM (Eu³⁺). ECL Measurements

ECL measurements were made with a modified ORIGEN Analyzer. The

modifications included a flow cell with gold electrodes, a positioning plate (which centered the PMT over the working electrode), and a Hamamatsu R268 (blue sensitive PMT)

(available from Hamamatsu, 360 Foothill Road, P.O. Box 6910, Bridgewater, NJ 08807 U.S.A.). A 50 μM persulfate buffer solution (see below) was used. ORIGEN cell cleaner (available from IGEN, Inc.) was used without modification. The heaters were set at 35 C and a preoperative potential ("POP") of O V was used. (POP is the initial potential at the beginning of a step or ramp.)

A 50 μM persulfate buffer solution was prepared from 10.2 mg of potassium persulfate (available from Aldrich, 1001 West Saint Paul Ave., Milwaukee, WI 53233 U.S.A.) and 750 mL of a 0.1 M phosphate buffer (pH 6.1-6.0) solution that was 0.1% in Triton X100 (available from Sigma Chemical Co., P.O. Box 14508, St. Louis, MO 63178 U.S.A.). Aliquots of the Eu(1)³⁺ aqueous stock, 151 μL and 15.1 μL, were diluted with 10 mL of the persulfate buffer to make 10 μM and 1 μM Eu(1)³⁺ persulfate buffer solutions. A 100 μL aliquot of the 10 μM Eu(1)³⁺ persulfate buffer solution was diluted with 10 mL of persulfate buffer to make a 100 nM Eu(1)³⁺ persulfate buffer solution. Aliquots of the Tb(1)³⁺ aqueous stock, 101 μL and 10.1 μL, were diluted with 10 mL of the persulfate buffer to make 10 μM and 1 μM Tb(1)³⁺, persulfate buffer solutions. Aliquots, 100 μL and 10 μL, of the 10 μM Tb(1)³⁺ persulfate buffer solution were diluted with 10 mL of persulfate buffer to make 100 nM and 10 nM Tb(1)³⁺ persulfate buffer solutions.

The instrument program to measure ECL was used with a ramp to -5 V and a rate of 4800 mV/s. 800 V was applied to the PMT. The following solutions were analyzed:

persulfate buffer, 100 nM Eu(1)³⁺, 1 μM Eu(1)³⁺, 10 μM Eu(1)³⁺, 10 nM Tb(1)³⁺, 100 nM Tb(1)³⁺, 1 μM Tb(1)³⁺, and 10 μM Tb(1)³⁺.

Potential step analyses were performed with a step to -5 V and a pulse width of 10 s on persulfate buffer, 100 nM Eu(1)³⁺, 1 μM Eu(1)³⁺, 10 μM Eu(1)³⁺, 10 nM Tb(1)³⁺, 100 nM Tb(1)³⁺, 1 μM Tb(1)³⁺, and 10 μM Tb(1)³⁺. As above, 800 V was applied to the PMT.

The ECL from the La(2)³⁺ complexes was studied with a ramp to -5 V at a rate of 4800m V/s as well as with a step potential to -5 V for 1 s. For both types of analyses 800 V were applied to the PMT. A 50 μM persulfate buffer solution was prepared as above by dissolving solid K₂S₂O₈ in a 0.1 M phosphate buffer (pH 6) which was 0.1% in Triton X100. Aliquots of the three aqueous Tb(2)³⁺ stock solutions; 0.216 μM (50 μL and 250 μL), 2.16 μM (50 μL, 125 μL, 250 μL, and 500 μL), and 2.16 mM (5 μL and 250 μL); were diluted with 10 mL of persulfate buffer to make the following Tb(2)³⁺ - persulfate solutions: 1.08 nM, 5.4 nM, 10.8 nM, 27 nM, 108 nM, 1080 nM and 10800 nM. Three tubes of each of these solutions and three tubes of persulfate buffer were analyzed by ECL with the potential ramp and potential step. Aliquots of two aqueous Dy(2)³⁺ stock solutions; 2.0 μM (50 μL, 125 μL, 250 μL, and 500 μL) and 2.0 mM (5 μL and 50 μL); were diluted with 10 mL of persulfate buffer to make the following Dy(2)³⁺ - persulfate solutions: 10 nM, 25 nM, 50 nM, 100 nM, 1000 nM, and 10000 nM. Three tubes of each of these solutions and three tubes of persulfate buffer were analyzed by ECL with the potential ramp and potential step. Aliquots of the three aqueous Eu(2)³⁺ stock solutions; 0.2 μM (50 μL and 250 μL), 2.0 μM (50 μL, 250 μL, and 500 μL), and 2.0 mM (5 μL and 50 μL); were diluted with 10 mL of persulfate buffer to make the following Eu(2)³⁺ - persulfate solutions: 1.0 nM, 5.0 nM, 10 nM, 25 nM, 50 nM, 100 nM, 1000 nM, and 10000 nM. Three tubes of each of these solutions and three tubes of persulfate buffer were analyzed by ECL with the potential ramp and potential step. Aliquots of two aqueous Sm(2)³⁺ stock solutions; 2.0 μM (50 μL, 125 μL, 250 μL, and 500 μL) and 2.0 mM (5 μL and 50 μL); were diluted with 10 mL of persulfate buffer to make the following Sm(2)³⁺ - persulfate solutions: 10 nM, 25 nM, 50 nM, 100 nM, 1000 nM, and 10000 nM. Three tubes of each of these solutions and three tubes of persulfate buffer were analyzed by ECL with the potential ramp and potential step. A second set of Sm(2)³⁺ - persulfate buffer solutions were prepared by diluting with 10 mL of persulfate buffer a 500 μL aliquot of the 2.0 μM aqueous solution, 5 μL and 50 μL aliquots of the 2.0 mM aqueous solution, and 250 μL and 500 μL aliquots of the 10000 nM Sm(2)³⁺ - persulfate buffer solution to make 100 nM, 250 nM, 500 nM, 1000 nM, and 10000 nM Sm(2)³⁺ - persulfate buffer solutions. Both sets of Sm(2)³⁺ - persulfate buffer solutions were analyzed in triplicate with three tubes of persulfate buffer by ECL with a potential ramp and potential step.

Photophysics

The absorption maxima and extinction coefficients for the La(2)³⁺ complexes are found in Table I. The extinction coefficients were calculated assuming a formula for each complex of [La(2)]Cl₃. These data are consistent with complexation with the free ligand. There is the expected red shift in the ligand - * bands as previously reported for the Tb(2)³⁺ and Eu(2)³⁺ complexes. Prodi, L.; Maestri, M; Ziessel, R.; Balzani, V., Inorg. Chem, 1991, 30, 3798. However, the extinction coefficients for the Tb(2)³⁺ and Eu(2)³⁺ complexes listed in Table I are 20% - 40% higher than previously reported. Id. This may be accounted for by working with a more pure sample. This hypothesis is substantiated by the questionably pure elemental analyses reported by Prodi, et al. for the [Eu(2)]Cl₃ and [Tb(2)]Cl₃ complexes. Id.

The emission maxima are listed in Table I and are consistent with the reported maxima for these lanthanides. Xu, Y.-Y.; Hemmila, LA. Anal. Chim. Acta 1992, 256, 9. Two expected emission bands were not reported, this was due to interference by the first harmonic of the excitation wavelength. The relative quantum efficiencies, (La), were determined by the equation:

{phi (Tb)} over {Area(Tb)}~=~Ieft

[{{Scale(Tb)} over {Scale(La)}} right]~x~left

[{{phi (La)} over {Area(La)} } right]

where (Tb) is the reported quantum efficiency for Tb(2)³⁺, 0.37 ± 0.1³, area represents the area under the emission bands and is estimated by the weight of the paper within the emission bands, and Scale is the instrument scaling factor. Each weight used in the calculation was an average weight from four separate emission spectra. The relative quantum efficiencies are reported in Table I. Also included are the deviations which are based upon the reported deviation in the Tb(2)³⁺ quantum efficiency and the standard deviations in each average weight. The largest portion of the listed quantum efficiencies was due to the reported 30% deviation in the quantum efficiency for Tb(2)³⁺. Additionally, the quantum efficiency for the Sm(2)³⁺ complex is expected to be higher than reported in Table I since the largest emission band for this complex was obscured by the second harmonic of the excitation wavelength. The quantum efficiency measurements will be repeated with a bandpass filter between the source and the sample to eliminate the second harmonic of the excitation wavelength.

Also included in Table I are the photophysical data for Tb(1)³⁺ and Eu(1)³⁺ as reported by Alpha et al. Alpha, B.; Ballardini. R.; Balzani, V.; Lehn, J.-M; Perathoner, S.; Sabbatini, N., Photochem. Photobiol. 1990, 52, 299.

Electrochemistry

No oxidative redox processes were observed for the La(2)³⁺ complexes within the 0.0 V to +2.5 V potential window. There were, however, several reductive redox processes within the 0.0 V to -2.3 V potential window. Three of the complexes have similar processes and cyclic voltammograms; Dy(2)³⁺, Sm(2)³⁺ , and Tb(2)³⁺. Figure 2 contains the cyclic voltammogram of Dy(2)³⁺ which is representative of this group of complexes. These complexes have a group of three reductive waves between -1.0 V and -1.5 V. These waves correspond to one electron reductions for each of the bpy pendent arms of ligand 2. The final one electron reductive wave is being assigned as the second reduction of one bpy arm, the other second reductions are presumed to be just beyond the potential window. The cyclic voltammogram of Eu(2)³⁺ is found in Figure 3 and shows a one electron reduction at -0.28 V and a set of three-one electron reductions between -1.7 V and -2.1 V. The first reduction is assigned as the Eu^III/II couple based upon previously reported Eu^III/II couple. Alpha, B.; Lehn, J.-M.; Methis, G., Angew. Chem. Int. Ed. Engl. 1987, 26, 266. The set of three reductions are assigned as the first reductions of each bpy arm. The set of bpy reductions for Eu(2)³⁺ are shifted to more negative potentials as compared to those of Tb(2)³⁺, Sm(2)³⁺, and Dy(2)³⁺; since it is more difficult to reduce a +2 complex than a +3 complex. Table II lists the electrochemical data for the La(2)³⁺ complexes as well as the redox couple assignments.

ECL Measurements

Plots of the point-by-point background corrected ECL counts vs. potential curves for the La(l) complexes are shown in Figure 4. These results show that both complexes have similar potential responses. The onset of ECL for both complexes occurs at ca.

-1.4 V. Peak potentials are observed at ca. -2.8 V. However, the ECL vs. potential curve for the Tb(1)³⁺ complex shows a shoulder at ca. - 1.7 V where this is not observed in the Eu(1)³⁺ case. The shoulder, if real, is presumably due to ECL generated by the mediated

mechanism. The potential step-ECL decay curves for the La(1)³⁺ complexes (Figure 5) show a biphasic decay. A very rapid decay is observed after the potential was applied, followed by a pseudo-plateau which starts to decay to baseline at ca. 8 s. This final decay is presumed to be due to consumption of persulfate. ECL vs. La(1)³⁺ concentration curves are found in Figure 6. These data are based upon the potential step results. These data show a linear response over the concentration range

( 10000 nM) studied. A notable exception from the linear response is the 10000 nM Tb(1)³⁺ data point. This lower than expected signal (47%) may be due to changes in the ECL mechanism from the direct mechanism (reduced label reacting with reduced persulfate) to the mediated (reduced persulfate oxidizing the label which is then reduced at the electrode) mechanism which is expected at μM label concentrations. Detection limits for these complexes have been estimated. An estimate of the system noise, 10% the background ECL, was used to determine the detection limits. Listed in Table III are the peak potentials and detection limits for these complexes.

Point-by-point background corrected ECL vs. potential curves for the Tb(2)³⁺, Sm(2)³⁺, and Dy(2)³⁺ complexes are shown in Figure 7. The curves are similar for these complexes. The ECL commences at ca. -2.2 V and increases until the edge of the potential window (-5 V) is reached. The peak potentials for these complexes are -5 V and are presumably associated with the ligand reduction. A potential step-ECL decay curve for each of these complexes is shown in Figure 8. The curves for the Tb(2)³⁺ and Sm(2)³⁺ complexes show that the ECL is noisy but constant over the 1 s pulse width; however, the ECL of Dy(2)³⁺ decays over the pulse width. ECL vs. label concentration curves for Tb(2)³⁺, Sm(2)³⁺, and Dy(2)³⁺ are shown in Figure 9. The Dy(2)³⁺ and Sm(2)³⁺ complexes have a linear response over the concentration range studied (100 nM to 10000 nM for Dy³⁺ and 250 nM to 10000 nM for Sm³⁺). The response for the Tb(2)³⁺ complex is linear between 5 nM and 1000 nM; however, the 10000 nM data show a 90% drop-off from expected. The source of this large deviation from the expected is unknown. No ECL was observed for the Eu(2)³⁺ complex over the concentration range studied (1 nM to 10000 nM). This has been attributed to the negative shift in the ligand reductions being large enough to put them outside the potential window. Table III lists the detection limits for the La(2)³⁺ complexes as well as the peak potentials.

Discussion

The proposed mechanism by which ECL is generated by these lanthanide compounds is complex. This must be the case for Tb(1)³⁺, Tb(2)³⁺, Sm(2)³⁺, and Dy(2)³⁺ since the ECL stems from an f-f excited state. This excited state cannot be populated in a redox process analogous to Equation (3) since these complexes do not have a metal based reduction within the working potential window. However, there is a ligand based reduction within the potential window. Therefore, we are proposing that the light is obtained via Antenna ECL. The following steps summarize in Equations (4), (5) and (6) below the Antenna ECL mechanism.

(4) La(L)³⁺ +e- La(L)²⁺

(5) La(L)²⁺ +O₄- La(L*)³⁺ +SO₄ ²

(6) La(L*)³⁺ La*(L)³⁺ In Equation (4) the ligand (L) is reduced and following the homogeneous electron transfer, an intraligand excited state is formed, Equation (5), via electrochemical excitation (ECX). This excited state sensitizes the f-f excited state in an intramolecular energy transfer step, Equation (6), analogous to the mechanism by which f-f emission is observed in a

fluorescence experiment. This mechanism is supported by the change in ECL peak potential with a change in ligand between ligands 1 and 2. This new ECL mechanism is analogous to the mechanism by which the efficient lanthanide fluorescence is obtained with these complexes. That is the highly absorbing ligand bound to the lanthanide sensitizes the f-f excited state. The notable difference being the mechanism of ligand excitation. In the fluorescence a technique, ligand excitation occurs via absorption of a photon. Since a large majority of biologically relevant species and the media absorb the light needed for ligand excitation fewer photons will produce the ligand excited state, and since many of these species will also emit there will be a high background signal. Therefore, this excitation mechanism can be a disadvantage for the use of lanthanides in immunoassays. However, fewer biologically relevant species and the media will emit via ECX. If ECX and Antenna ECL are used to produce the lanthanide emission one would expect to have an assay with fewer interferences.

A comparison of the quantum efficiencies and ECL efficiencies is useful. The ECL efficiency is defined as the moles of photons per 2 moles of electrons (2 electrons are needed to produce one reduced labels and one reduced persulfate). The quantum efficiency is defined as the moles of photons per mole of excited states. These two equations are identical when all reduced labels react with all reduced persulfate to produce a label excited state. Therefore, a comparison of these two values can gauge the efficiency of the electron transfer reactions. However, in the aqueous ECL reactions the media is also being reduced. This prevents an exact determination of the ECL efficiencies. If one uses ECL detection limits in place of ECL efficiencies, a trend analysis can be made which will eliminate interferences due to media reduction. Dy(2)³⁺, and Sm(2)³⁺. However, the remaining complexes do not allow the trend for various reasons. The complex Tb(1)³⁺ does not follow the trend since complexes of ligand 1 and complexes of ligand 2 are expected to have differences in electron transfer rates since the ligands have different reduction potentials (see Table II). The Eu³⁺ comp lexes cannot be compared to the others due to some major differences. All of these differences are due to the presence of the Eu^III/II couple prior to the ligand reductions. Different electron transfer rates are expected since the first ligand reduction is for Eu²⁺ complex instead of Eu³⁺ complex; therefore, the ligand reductions occur at more negative potentials than the other complexes with the same ligand. Also differences in excited state or in the mechanism of excited state formation are expected. Since the ligand excited state is produced following Eu reduction, either a second oxidation must occur prior to Eu^III excited formation or a Eu^II excited state will be formed which is expected to have different photophysical properties such as emission efficiencies and maxima.

Claims

What is claimed is:

1. An electrochemiluminescent process for excitation of one or more luminophores through an antenna mechanism using an oxidative or reductive coreactant, wherein each luminophore is a complex with one or more ligands, comprising

reacting one or more reduced luminophores with an oxidizing radical, or reacting one or more oxidized luminophores with a reducing radical and generating a luminophore excited state and

measuring the electrochemiluminescent emissions.

2. The process of claim 1 wherein the luminophores are lanthanide chelates.

3. The process of claim 2 wherein the coreactant is a persulfate.

4. The process of claim 3 wherein the reactions in the process are as follows:

La(L)³⁺ +e- La(L)²⁺

S₂O₈ ²- +e- S₂O₈ ^3-

S₂O₈ ³- SO₄ ²- + O₄-

La(L)²⁺ +O₄- La(L*)³⁺ + SO₄ ²

La(L*)³⁺ La*(L)³⁺

La*(L)³⁺ La(L)³⁺ +hv

wherein La represents lanthanide metal, L represents ligand, L represents a reduced ligand, L* represents a ligand excited state and La* represents a lanthanide excited state.

5. The process of claim 2 wherein the lanthanides are selected firom the group consisting of Sm, Eu, Tb, and Dy.

6. The process of claim 2 wherein the ligands are selected from the group consisting of

7. The method of claim 5 wherein the lanthanide chelates are complexes selected from the group consisting of Sm(2)³⁺, Eu(1)³⁺, Tb(2)³⁺, Tb(1)³⁺, and Dy(2)³⁺.

8. An electeochemiluminescent process for excitation of one or more lanthanide chelates, each lanthanide chelate comprising a lanthanide complex with one or more ligands, through an antenna mechanism using an oxidative or reductive coreactant comprising

reacting one or more reduced ligands with an oxidizing radical or reacting one or more oxidized ligands with a reducing radical and generating a ligand excited state; and

measuring the energy transfer from the ligand excited state to the emissive state.

9. The process of claim 8 wherein two or more lanthanide chelates are employed and the electrochemiluminescent emissions are measured at different emission wavelengtias or different electrode potentials.

10. The process of claim 8 wherein the coreactant is a persulfate.

11. In a system for electrochemiluminescent excitation and emissions measurement employing a luminophore and a coreactant, the improvement comprising employing a lanthanide chelate as the luminophore and a persulfate as the coreactant.

12. The system of claim 11 wherein two or more luminophores are employed and the emissions are measured at different emission wavelengths or different electrode potentials.

13. A method of simultaneously determining the presence of two or more different analytes which comprises

contacting a sample suspected of containing one or more analytes with two or more different lanthanide chelates in the presence of a coreactant under conditions for measuring electrochemilumin escence,

measuring electrochemilumin escence at different emission wavelengths or different electrode potentials,

comparing the measurements to a standard.

14. The method of claim 13 wherein the coreactant is a persulfate.

15. A kit for simultaneously determining the presence of two or more different analytes which comprises

two or more different lanthanide chelates in premeasured amounts;

a coreactant in a premeasured amount; and

a reference standard,

wherein the premeasured amounts are sufficient to perform a single sample measurement.

16. The kit of claim 15 wherein the coreactant is a persulfate.

17. The kit of claim 15 further comprising a means of generating

electrochemilurninescence and a means of measuring electrochemiluminescence at different emission wavelengths or different electrode potentials.