NON-LINEAR SIMULTANEOUS TWO-PHOTON EXCITATION ENERGY TRANSFER IN THE WRONG DIRECTION

The simultaneous two-photon excitation energy transfer (SEET) was demonstrated for 
the first time using trichromophoric model compounds. Two identical donors (A–antenna) 
were covalently linked to an energy acceptor unit (T–target) with different 
energy levels preventing energy transfer of a single photon. At high intensity illumination 
(laser exposure) of a trichromophoric system A∼T∼A (A–fluorescein, erythrosin; T-Estilbene), 
sufficient to excite both of the appended donor subunits, population of the 
target excited state may occur via simultaneous energy transfer of two photons, one from 
each donor. In order to restrict reverse energy transfer from the higher energy target to 
the lower energy donor(s) it is necessary that the excited target unit undergoes an 
efficient photoreaction. In the investigated case this was achieved by photoisomerization 
of the stilbene unit used for monitoring of the SEET.


INTRODUCTION
The phenomenon of radiationless electronic excitation energy transfer (EET) A -D A* D* + + has been intensively studied both theoretically and experimentally [1,2].The precondition for EET from D* (electronically excited donor molecule) to A (acceptor molecule) is a partial overlap of the donor fluorescence band FD() and the acceptor absorption band eA().The   EET process can be the result of long range Coulombic (dipol-dipol, F6rster mechanism) and short range electron-exchange interaction.According to F6rster [3], the rate coefficient for Coulomb excitation energy transfer kEET(C) is described by 9000 (InlO)2 o kEET(C)-128 7r 5 n 4 NA v 7-,o r 6 J, (1) where n2 is the orientation factor [2], ,0 is the donor fluorescence quantum yield in the absence of the acceptor, n is the refractive index, NAy is Avogadro's number, r is the distance between donor and acceptor and J is the overlap integral defined by J F D () eA () W4 d.
(2) According to Dexter [4], the rate coefficient for electron-exchange excitation energy transfer kEET(E) depends on the overlap of the emission spectrum and the absorption spectrum of A, too, kEET(E) 2 2 J, (3) where the parameter Z is obviously related to the electronic matrix element for electron-exchange energy transfer (Z 2 exp(-2r/l), is the van der Waals radius of the donor-acceptor pair).
The simultaneous two-photon excitation energy transfer (SEET) from two excited energy donors A* (A-antenna unit) to one energy acceptor (T-target unit) (4) is still unknown.In this paper we discuss the concept of SEET and investigate the deactivation behavior of the covalently linked AN TA system 1 and 3 containing fluorescein or erythrosin (antenna unit) and E-stilbene (target unit) (Tab.I).The SEET enables a deep UV photochemistry at high intensity long wavelength excitation.

CONCEPT OF SEET
The principle of SEET (eq.4) is shown in Figure 1.The energy of the excited state of the target unit is marked higher than the excited state energy of the antenna system.The requirements of a successful proof of the SEET are" 1. Provided that the SEET has taken place, the rate coefficient of the "common" energy transfer (eqs. 1, 3, 5) will be very high.Therefore, the proof of the existence of T* requires an adjustment of the spectroscopic properties of A and T resulting in a small overlap integral (eq.2) and a small deactivation rate coefficient of T* by EET (eq.5). 2. Resulting from the strong competition of the EET (eq.5) and the photophysical and/or photochemical deactivation of T* it is necessary to use target units characterized by a very high rate coefficient of the fluorescence of T* or by a rapid chemical deactivation of T*.In that case the chemical reaction is the indication of a previously occured SEET.The model compound E-4,4'-di(fluorescein-5"-yl-thioureanyl)-stilbene 1 (nomenclature according to Haugland and Larison [5]) was used to varify this concept.The fluorescein chromophore is the antenna unit, the E-stilbene represents the target unit.The successful SEET will be shown by the E,Z-isomerization of the stilbene at (high intensity) fluorescein excitation (eq.6). ( To prove the concept of SEET the corresponding dichromophoric systems ANT 2 and 4 (Tab.I) were also investigated.

EXPERIMENTAL Materials and Preparations
All preparations were carried out under yellow light.

Instrumentation and Measurements
The registration of the UV/VIS absorption spectra was realized with a U-3410 spectrophotometer (Hitachi) combined with a computer.The emission spectra were recorded using a fluorescence spectrophotometer MPF-2A (Hitachi-Perkin-Elmer).The solutions were thermo- stated.
All investigations were done using aqueous/ethanolic solutions (10:90).For the reason of the dependence of the excited state properties of the dyes upon the pH value (band position, life time and quantum yield of the fluorescence, Figs. 2, 3), all solutions were alkalized with sodium hydroxide to pH 8.
The low intensity photolysis was realized using a high pressure mercury lamp HBO 500 (Oriel) with controlled light intensity and a metal interference filter.The laser exposure experiments were carried out with an Ar +-Laser (Coherent, Inc., Aexc 488 nm for 1 and 2 and Aexc=514 nm for 3 and 4) and/or a dye laser (FL 80, ZWG, Aexc 500 nm for 1 and 2, Aexc 514 nm for 3 and 4 using Coumarin 307, Lambdachrome) pumped by an XeC1 excimer laser (EMG 103, Lambda Physik).

Spectroscopic Properties
The absorption spectra of the investigated compounds 1-4 (Tab.I) are the superposition of the corresponding chromophors (Fig. 4b).The partial spectra of the stilbene and the fluorescein (Fig. 4a) or erythrosin are characterized by separated bands.For the selection of   the model compound I and the other investigated AN TA systems the following aspects were considered: 1.The energy of the local excited states of A* and T* complies with the precondition shown in Figure 1.
2. The overlap of the absorption of the antenna chromophor and the fluorescence of the target chromophor (Fig. 4b) is very weak and, therefore, the resonance integral J will be very small (eq.2).Re- gardless whether the F6rster (eq. 1) or the Dexter (eq. 3) mechanism are considered the rate constant of an excitation energy transfer T*A is very low.
3. Conjugative interaction between the 7r-systems of A and T is not transmitted through the thiourea group.4. The rate constant of the E, Z-isomerization is very high, the spectroscopic properties of E-and Z-isomer differ [10].The isomers can be separated using chromatographic methods [11].

Deactivation Behavior at Low Intensity Excitation
At excitation of 1 using Aec 488 nm (local excitation of the antenna unit) only the fluorescence of a fluorescein derivative is observed (Fig. 4b).The fluorescence quantum yield is }88(1)= 0.11 :k: 0.01.The low quantum yield of 1 compared with fluorescein sodium (Fig. 3) is caused by the heavy atom effect of the covalently bonded sulfur.The intramolecular heavy atom quenching by the sulfur is observed also in the case of other fluorescein thiourea derivatives, in the case of fluorescein-5-isothiocyanate (Fig. 3) and at addition of thiourea to a solution of fluorescein sodium.The Stern-Volmer rate constant of this intermolecular heavy atom quenching is kq=2.9.10-91 mol-s-.At excitation of 1 using Aexc 333 nm (local excitation of the target unit) a dual fluorescence and a photoreaction are observed simulta- neously, in detail: the typical fluorescein fluorescence 333 wf,1 (1) 0.10 -+-0.02, a short wavelength fluorescence following to the absorption band of the stilbene chromophore (Fig. 4b), 333 7,2 (1) 0.010 4-0.002, the photoisomerization E-I-+Z-I(Fig.5).E-z-333 rl.) 0.020 +0.004.
The main deactivation process of the local excited target unit is the undesirable excitation energy transfer T*-A (EET).The quantum yields of the corresponding deactivation processes of the isolated excited target chromophor (4,4'-diaminostilbene) are /=0.12 and E_z=0.26.chromophores undergo the excitation energy transfer and only 10 % deactivate via the target specific fluorescence and chemically by photoisomerization.The results of the analysis of the photolyzed solutions using HPLC are shown in Figure 7a.The composition of the photostationary state described by the Zimmerman equation [7] is about 66 % Z-isomer and 34 % E-isomer.The compounds 2, 3 and 4 show a very similar deactivation behavior on excitation at Aexc 333 nm.In all cases the corresponding dual fluorescence and the isomerization are observed.The fluorescence of the antenna chromophor (fluorescein, erythrosin) is the dominating process.

Deactivation Behavior at High Intensity Excitation
The photolysis of I using the Ar + laser (488 nm, 1.6 W/spot 14.9 W cm-2) is shown in Figure 6.Beside the very strong fluorescence of the antenna unit the photoisomerization E-IZ-1 goes on.The quantum yield of this photoreaction depends on the intensity of the laser light, the relative quantum yields are _488 E_z(700 mW/spot)" _488 E--zl 600 mW/spot) 488 vez 170 mW/spot) 0.2" 0.01.The results of the HPLC analysis of the irradiated solutions are presented in Figure 7b.The composition of the photostationary state at laser photolysis (45 % E-isomer, 55 % Z-isomer) differs from that of lamp photolysis (target excitation, Fig. 7a).
At photolysis of 3 (/exc--500 nm, dye laser, or )kexc---514 nm, Ar + laser) the same non-linear behavior of the photoisomerization is observed, but in this case the ATA system is not stable to reach the photostationary state (photobleaching).
The laser irradiation of the reference compounds 2 and 4 using the same condition demonstrated in Figure 6 does not effect in a photoisomerization.The laserchemical photoisomerization of 1 and 3 by antenna excitation is absolutely dependent on the existence of two antenna units.

DISCUSSION
The scheme illustrates the deactivation processes of the AT*A system populated by monophotonic local excitation of the target (stilbene) unit.The unimolecular processes fluorescence, isomerization and EET compete.The photoisomerization of 1 and 3 at high intensity antenna excitation is explained by the model of the simultaneous two-photon excitation energy transfer (SEET) shown in scheme 2. The population of the twofold mono-excited state occurs consecutively.For a simultaneous two-photon absorption the used laser intensity is too low.The twofold mono-excited system A*TA* deactivates partly by SEET and populate AT*A.The formation of A T*A on this way requires two antenna units, the comparable experiments using 2 and 4 failed, which excludes any mechanism considering a possible two-photon excitation of one chromophor (A or T).The main de- activation process of AT*A is the common EET forming ATA* ("back transfer") followed by the antenna fluorescence.
At the laser irradiation experiment shown in Figure 6. the following experimental conditions were used" number of molecules per active volume, determined by concentra- tion, laser spot diameter and thickness of the cuvet: 2.4.1015,intensity per spot: 1600 mW/spot, diameter of the spot: 3.70 mm photons per second: 9.3.1017.
Assuming a life time of the mono-excited system of about ns (singlet state), during the laser exposure a steady-state concentration of A." TE".A hv A hv A A -/" TE'.A SCHEME 2 Deactivation at high intensity antenna excitation.about 108 molecules per active volume results.This concentration is too low for an effective absorption of a second photon within the life time.From this we conclude, that the absorbing intermediately formed mono-excited states are triplet states.The life time of the fluorescein triplet state (---5-10-4 S) guarantees a steady-state concentration for ATA* of about 2-1013 molecules per active volume and a second excitation with a probability of 0.0075.The SEET (from a twofold mono-excited triplet-triplet state) shows similarities to the known triplet-triplet annihilation process.

FIGURE 6
FIGURE 6 Irradiation spectra of 1, laser excitation of the antenna unit at/exc 488 nm (Ar laser).