Introduction

The back reaction takes 200 s, or 2x105 times longer (3). The molecular causes of this vectorial proceeding of the reaction are unknown. It is, however, of great interest to evaluate these causes. The conversion of short-lived excited states into long-lived redox pairs is identical to the conversion of light energy into storable chemical energy.

Several other applications of molecular structures containing porphyrins are realized in nature: "Special pairs" of two partly overlapping chlorophyll chromophores are photoxidized in photosystem I (6), porphyrin-steroid complexes are probably involved in the stereoselective hydroxylation of steroids by cytochrome P 450 (7), and heme-oxygen complexes play a prominent role in biological oxygen transport (8) and oxidation (9).

Reported attempts to produce complex porphyrins-structures fall into one of four categories:

1) A porphyrin molecule, usually dissolved in an organic solvent, aggregates spontaneously with a second porphyrin or a large-surface hydrocarbon, e.g. a steroid, or an electron-acceptor chromophore, e.g. a nitroaromat;

2) A porphyrin is covalently bound to a second porphyrin, or to a redox chromophore, e.g. a quinone, or to a redox-active metal complex;

3) A porphyrin is dissolved in micelles or bilayer lipid membranes, and the other redox components are dissolved in the aqueous phases or localized on the surface of the hydrophobia aggregate as head groups;

4) to arrange porphyrins and other reactants, e.g. electron acceptors, in an artificial and well-defined order in natural proteins.

The given categories (1)-(4) of organized porphyrins-systems containing porphyrins also constitute the approximate chronological order in the development of model systems. In the sixties and seventies the molecular complexes and aggregates of porphyrins and chlorophylls were investigated in great detail, mostly by spectroscopic means. The present times sees an explosion of the syntheses of covalent pairs and kinetic measurements of charge separation reactions. Experiments which try to arrange porphyrins and other components regioselectively in synthetic membranes are just beginning.

There are shortly reviewed and evaluated the results obtained with molecular complexes and covalent pairs, and then concentrate on new developments of synthetic membranes.

The porphyrin structures.

The knowledge on the geometrical and electronic structures of these porphyrin structures has been summarized in the years 1976-1978 in three reviews (7, 10, 11).

(1) Porphyrins of the etio-type (eight 6-pyrrolic substituents, no methine bridge substituents) tend to form cofacial dimers and molecular complexes with steroids. The binding force is somewhat less than 40 kj mol and is mainly caused by van der Waals type interactions (12). Steroids may block the catalytic action of metalloporphyrins in autoxidations of olefins (13). Other effects of complex formation on the chemical reactivity of the components are not known.

More pronounced and therefore better known are physical changes after the formation of molecular complexes of metalloporphyrins. NMR ring current effects and ESR-dectable metal interactions have often been applied in structural analyses. Long wavelength-shifts in visible spectra are observed in dimers of non-symmetric dipolar chromophores, such as chlorophylls and protochlorophylls. An even stronger electronic interaction is observed, if both porphyrin chromophores are oxidized to the cation radical state. The dimer is then hold together by a binding force of 70 kj mol , the unpaired electrons of the radicals become paired and a strong, new absorption band close to 900 nm is observed (7, 10, 11).

Light-induced charge-separation has been found, when the excited triplet and the ground states of zinc octaethylporphyrin form an encounter complex. The initial ion-pair, ZnOEP-ZnOEP, recombines with rate constants in excess to 2 x 10 -1 M. It is therefore not possible to couple chemical redox reactions to the photo-initiated electron transfer (14).

(2) Small molecules are only tightly bound to porphyrin chromophores, if they act as electron acceptors in organic charge transfer complexes. The organic charge transfer complexes have again a face-to-face structure. Examples are complexes with flavins (7), quinones (7), and nitroaromatics (15). The distance between the chromophore planes is again in the range 4-5 A. There has, however, been one report in which a very efficient charge transfer occurs over a distance of approximately 10 A. In a pyrazine-bridged heme dimer light-induced electron transfers were made responsible for a strong absorption band near 800 nm (16).

A dilute chlorophyll-quinone solution in ethanol showed the photoinduced electron transfer from excited chlorophyll triplet states to quinone (17).

(3) Electropositive metal ions, e.g. Mg, bind to polarized carbonyl substituents on the periphery of a second porphyrin molecule. Dimers with an almost orthogonal orientation of the porphyrin planes are thus formed (11).

Such dimers can be detected and analyzed in organic solvents by concentration dependent NMR-measurements. Addition of water leads to extensive re-orientations of both chlorophyll chromo-phores in respect to each other.

Covalent Pairs of porhyrins

Katz et al. have synthesized a photochlorophyllide dimer connected by an ethylene glycol unit (6). This dimer is in an open configuration A in methylene chloride (preferably with L = pyridine) and folds to conformation В if a hydrogen bonding ligand such as ethanol is present.

Photoexcitation with laser pulses gave excited S,. states in high yields and significant triplet populations in the open conformer A. In contrast the folded pair В exhibited an unusually short fluorescence life time and a correspondingly low quantum yield. One may therefore conclude that dimers with strong electronic interactions and short porphyrin-porphyrin distance are not useful in charge-separation experiments. The same conclusion has obviously been reached by several other workers, who have prepared different cofacial bis-porphyrins with short porphyrin-porphyrin distances. No extraordinary photochemical properties have been reported for the dimers 2 (18) and 3(19).

It is, however, remarkable that dimer J exhibits a red-shift of the Soret and a-bands (420 and 670 nm; explained as due to weak incipient charge transfer interactions), whereas the Soret band in dimer 3 is shifted to the blue ( A 376 nm max M = 2H); probably due to excitonic interaction of both chromophores).

This shows that the differences in the electronic structures of etio-type and meso-tetraphenyl porphyrins (20) also play a role in the electronic interactions in dimers. Collman has also used bis-metallic dimers for the 4-electron reduction of oxygen. Accordingly dimers with short distances between the porphyrins can be very useful in the promotion of redox-reactions (= electron transfer), but photo reactions are rather quenched than promoted.

Experiments with covalently connected porphyrins and quinones, e.g. 4, 5 and 6 have been more successful. In the cofacial adduct 4 both components are held rigidly at a center-to-center distance of 10 A. Irradiation at 470 nm produced a porphyrin triplet state which was quantitatively converted to a charge separated. The decay was about a hundred fold slower(21).

The quinone adducts 1 and 6 did not show long-lived charge separation in irradiation at room temperature. If the diamide 5 and diester 6 however, are irradiated in frozen solvents an ESR-spectrum arises, which corresponds to a 1:1 mixture of porphyrin and quinone radicals. This signal is formed essentially irreversible in the diamide linked compound whereas in the diester linked analogon 6 rapid decay of the ESR signal after cessation of the irradiation is observed. From the dipolar splitting of the signa an average distance of 10 - 12 A between the two unpaired electrons was estimated (22) .

The carotene-porphyrin-quinone (C-P-Q) adduct 7 is a concise structural sign for biological photosynthesis: electron donor (carotene), photo catalyst (porphyrin) and electron acceptor (quinone) are closely attached to each other. Irradiation with visible light leads to a very fast charge separation (30 ps), whereas recombination of charges is 2 x 10 times slower (60 s) (23). This compares well with the natural system. The key to obtaining long lifetimes of the charge-separated state appears to be the interposition of a neutral porphyrin and perhaps amide bonds between widely separated ions.

Porphyrins in membranes.

The water phases 1 and 9 are not organized and can only be used as reservoirs for water-soluble reagents, such as redox-active metal ions. The head group regions 3 and 7 have been functionalized with several redox active chromophores, e.g. phenylene diamine, viologen, diazonium benzene and quinone groupings. Since the outside region 3 can be regioselectively oxidized or reduced with water-soluble, membrane-impermeable reagents, one can easily produce unsymmetrical membranes, e.g. with an electron acceptor on the outside surface and a donator on the inside.

The distance of both head group regions can be varied from 18 A for the thinnest known monolayer membrane, to about 70 A for the thickest monolayer membrane. Bilayer lipid membranes have been made up to 100 A thick (25).

The charge of the head groups may be positive, negative or neutral. Charged solutes in the aqueous phase may therefore be adsorbed and localized in regions 2 and 8. Such adsorbed molecules are characterized by a tight adherence to the vesicle membrane, e.g. in Sephadex chromatography, and by the accessibility to water-soluble reagents. An example is the adsorption of meso-tetra (methylpyridinium)porphyrin 8 to dihexadecyl-phosphate vesicles (26). The particle load on the vesicle surface should, however, not exceed a few percent, since vesicle precipitation is observed at high porphyrin concentrations.

Regions 4 and 6 have also been be selectively occupied. This has been demonstrated with octaacetic acid porphyrin 1, an analogon of uro- and coproporphyrins (24,27). If this porphyrin is dissolved in vesicular solutions at pH 7 it is located in the bulk water phase as an oligoacetate. Acidification to pH = 5 leads to a half-neutralization of the acetate side-chains and raonoprotonation of the porphyrin centre. Further acidification, which would lead to precipitation in pure water, drives the neutral porphyrin base into the vesicle membrane. From a bathochromic shift of the a-band (616-625 nm) and from the fact that the porphyrin base is not protonated even at pH 1 , it is concluded that porphyrin 1 has migrated into the hydrophobic membrane. Rising the pH in the bulk aqueous phase, leads to a quantitative migration of the porphyrin back into the aqueous phase. The side-chains are therefore accessible to hydroxyl ions, the porphyrin's center is not accessible to protons. It should be noted, that acidification of a porphyrin in the aqueous phase leads to deprotonation of the porphyrin cation, which then migrates into the membrane. The described phenomena taken together clearly indicate, that the porphyrin is localized in region 4.

Region 6 can be occupied, if one sonicates porphyrin 1 and vesicles together at pH = 7, separates the outside porphyrin by gel chromatography and acidifies the solution.

The selective occupation of the hydrophobic region with porphyrins can be achieved in a trivial, non-specific way and in a more sophisticated manner. The trivial kind has often be realized with water-insoluble porphyrins, such as magnesium octaethylporphyrin (MgOEP). This porphyrin can be co-sonicated with any vesicle forming amphiphile and is then dissolved in any part of the hydrophobic membrane (region 4-6). Since MgOEP is well soluble in petrol ether, one may assume that it is evenly distributed over the whole width of the membrane. Experiments on black lipid membranes have indeed shown, that MgOEP and its cation radical freely diffuse through the membrane (28-30). If such a porphyrin is dissolved in a vesicle membrane with reactive head groups, one can assume that (1) the average porphyrin-head group distance is about half of the thickness of the membrane (2) the porphyrin and the head groups are mostly aligned perpendicular to the membrane plane. Both generalizations will only describe an average situation.

Large fluctuations and overall deviations may occur. The locality of a hydrophobia porphyrin chromophore gets more restricted, if it bears one or two charged substituents in one hemisphere, e.g. positively charged viologen or cholyl hydrazone derivatives (32,33). If the porphyrin 11 is dissolved in positively charged or neutral vesicles, sodium dithionite reduces the substituent to the radical. In negatively charged vesicles this becomes impossible. The porphyrin ligand of 11 is totally inaccessible to protons or other water-soluble-reagents. One may conclude that the porphyrin is buried within the hydrophobic membrane (regions 4 and 5), whereas the bipyridinium substituent is in the head group region 3 (31 ). The non-flexible bond between porphyrin and substituent should lead to a defined orientation of the porphyrin perpendicular to the membrane plane and a distance of about 4 A between the porphyrin centre and vesicle surface. Porphyrin 1 can also be dissolved in vesicles which contain reactive head groups. If all the outer head groups of the vesicle are active, as electron acceptors, e.g. viologens or quinones, then the distance to the porphyrin donor would be as short as 2 A. If the porphyrin is located at the"outside region 5" and the reactive head groups are on the inside surface (region 7), then the porphyrin acceptor distance has a length equal to the membrane thickness minus 7 A. It has been shown, that totally unsymmetric monolayer vesicle membranes can be prepared either from bolaamphiphiles with one large and one small head group or by regioselective precipitation. The small head groups are all at the vesicle at the small, inner surface, the large head groups are all outside (34). A water-soluble viologen bolaamphiphile vesiculates spontaneously, if titrated with perchlorate (35) .

A porphyrin-loaded vesicle has been realistically visualized by many scientists and therefore it have been a starting point for complex systems modelled after the photosystems of nature.

If one wants to work with a vesicle, in which only a few percent of the head groups are reactive, then the problem of domain formation becomes important. This problem is solved in nature by the application of protein complexes, in which all redox- and photo-reactive components are concentrated. In artificial membranes without proteins one may enforce domain formation by coulombic or charge transfer interactions at the surface. An example is the polymeric, blue charge transfer complex between viologen head groups of bolaamphiphiles and benzidine. The formation and dissociation of this "Scheibe-type complex is highly dependent of temperature, counterions and pH and can thus be regulated (36). The formation of porphyrin domains, e.g. with negatively charged porphyrin 8 parallel and positively charged porphyrins 1.1 or 12 perpendicular to the membrane should also be possible. If the latter porphyrins were connected via a long chain to a polar quinone molecule on the other side of the membrane, the organization of a synthetic analogue would be perfect. Another possibility is the introduction of bolaamphiphiles with one polar and one apolar chain. The apolar chain will dissolve within the membrane, the polar chain will aggregate and form pores (37).

All presuppositions to synthesize complex assemblies on the basis of well organized vesicle membranes are nowadays fulfilled (36, 37). It is only a question of time, until all the ready-made parts can be assembled in rational (and reproducible) ways. Is there any chance to make these reactive vesicles work as light-driven vectorial redox-chainsand to run interesting irreversible chemical reactions with solar energy?

Multicomponent systems of a relatively low degree of organization have been obtained and tested. Mauzerall has reviewed in 1975 the work with black lipid membranes, where dissolved chlorophyll produces photopotentials of a few millivolts. More recent work shows that chlorophyll derivatives, with positively charged substituents give higher photovoltages (18 compared to 11 mV), because the charged chlorophyll molecules are more concentrated in the surface area than the neutral chlorophyll (3O).

Calvin has extensively discussed vesicles which have EDTA as reducing agent entrapped in the vesicle aqueous volume (region 9), hydrophobic ruthenium complexes as electron carriers in regions 4-6 and methylviologen as electron acceptors together with water-soluble zinc porphyrins as sensitizers in the bulk aqueous phase. Although this arrangement with the final acceptor in the bulk phase should not be very efficient one obtained some photoreduction of the viologen, indicating light-induced electron transfer through the membrane (38). Fendler has recently given more direct evidence for this phenomenon (39).

Tollin et al. have studied the photoreaction of chlorophyll and quinone in vesicle membranes. They found that the lifetime of the charge separated state Chl-Q can be extended if negatively charged amphiphiles are mixed into electroneutral vesicles. The negative surface charge presumably leads to an expulsion of the negatively charged quinone radicals into the water-phase. The back-reaction is significantly retarded. In another system the chlorophyll sensitized vectorial electron transport across a bilayer membrane was investigated. A water-soluble quinone was entrapped in region 9, chlorophyll was dissolved in the vesicle membrane and glutathione was added as reducing agent to the bulk water phase (region 1). It was stated that 20 % of all the photons absorbed by the vesicle system resulted in electron transfer across the membrane from glutathione to the entrapped quinone (40, 41).

Hurst et al. have adsorbed the zinc complex of the tetra-cationic porphyrin 2 on negatively charged vesicles and claimed the formation of a ZnP-ZnP pair with a decay rate of T= 20 s (26). This compares very favorably to the 20 ps observed in homogeneous solution and is even ten times longer then with the carotene porphyrin-quinone triad.

Fendler has arranged all parts of the Shilov system for the photoreduction of water (reducing agent, sensitizer, redox catalyst and colloidal platinum in different compartments of vesicles. He found again that separated units are cooperating well through the membrane, if membrane soluble electron transporting molecules are present (42) .

None of the examined systems has a high quantum yield or is very long-lived. Better organized systems should improve the efficiency. More stable dyes in connection with inorganic systems preferably bound to regions 2 and 8, should provide longevity.

Because of space and time - limitations we have only discussed organized porphyrin systems in respect to photoinduced electron transfer reactions. Since vesicles have also extensively been applied in work on oxygen reactions with metalloporphyrins, we shall give two hints on very recent work on this closely related topic. Tsuchida has synthesized vesicles with artificial hemes which reversibly bind molecular oxygen in aqueous media (8). Whitten has shown that photooxidations of porphyrins in vesicles are predominantly operating with the intermediacy of superoxide, whereas in organic solvents most of the products arise from singlet oxygen attack (43). These examples indicate that slow, secondary reactions of porphyrins can be controlled as well by vesicles as the fast primary electron transfer reactions.

References.

1. Mathis, P., Breton, J., Vermeglio, A., and Yates, M.: FEBS Lett. 63, 171-173 (1976)

2. Ganago, I.B., Klimov, V.V., Ganago, A.O., Shuvalov, V.A.,and Erokhin, Y.E.: FEBS Lett. 140, 127-130 (1982)

3. Witt, H.T.: Biochim. Biophys. Acta 505, 355-427 (1979)

4. Clayton, R.K., in "The Photosynthetic Bacteria".

Clayton, R.K., Sistrom, W.R., Eds.; Plenum Press, New York, 387-396 (1978)

5. Trissl, H.W., Graber, P.: Biochim. Biophys. Acta 595, 96-108 (1980)

6. Hunt, J.E., Katz, J.J., Svirmickas, A., Hindman, J.C.: J.Am.Chem.Soc. 106, 2242-2250 (1984)

7. Fuhrhop, J.-H.: Angew. Chem. Intern.Ed.Engl. 15, 648-659, 1976)

8. Tsuchida, E., Nishide, H., Juasa, M.: J. Chem. Soc., Chem. Commun., 96-98 (1984)

9. White, R.E., Coon, M.J.: Annu.Rev.Biochem. 49, 315-356, 1980)

10. White, W.I.: D. Dolphin (Ed), The Porphyrins, Vol. V,Academic Press, New York, 303-341 (1978)

11. Katz, J.J., Shipman, L.L., Cotton, T.M., Janson,J.Am.Chem.Soc., 106, 402-458 (1984)

12. Sudhindra, B.S., Fuhrhop, J.-H.: Internat. J. Quantum Chem. 20, 747-753 (1981)

13. Fuhrhop, J.-H., Baccouche, M., Grabow, H.: J.Mol. Catal.7, 245-256 (1980)

14. Ballard, S.G., Mauzerall, D.C.: J.Chem.Phys. 72, 933-947, 1980)

15. Chandrashekar, Т.К., Krishnan, V.:Inorg. Chim.Acta 62, 259-264 (1982)

16. Fuhrhop, J.H., Baccouche, M., Btinzel, M.: Angew. Chem. Internat. Ed. 19, 322-323 (1980)

17. Tollin, G.F., Castelli, G., Cheddar, G., Rizzuto, P.: Photochem.Photobiol. 29, 147-152 (1979)

18. Kagan, N.E., Mauzerall, D., Merrifield, R.B.: J.Am.Chem. Soc. 99, 5484-5486 (1977)

19. Collman, J.P., Bencosme, C.S., Branes, C.E., Miller, B.D.: J. Am. Chem. Soc. 1 05, 2704 - 2710 (1983)

20. Felton, R.H., Dolphin, D., Borg, D.C., Fajer, J.: J. Am. Chem. Soc. 91, 196 (1969)

21. Lindsey, J.S., Mauzerall, D., Linschitz, H.: J. Am. Chem. Soc. 105, 6528 - 6529 (1983)

22. Me Intosh, A.R., Siemiarczuk, A., Bolton, J.R., Stillman,M.J., Ho, T.-F., Weedon, A.C.: J. Am. Chem. Soc. 105,7215-7223 (1983)

23. Moore, T.A., Gust, D., Mathis, P., Mialocq, J.-C.,Chachaty, C., Bensasson, R. V. , Land, E.J., Doizi, D.,Liddell, P.A., Lehman, W.R., Nemeth, G.A., Moore, A.L.: Nature 307, 630-632 (1984)Routes to Functional Porphyrin Assemblies

24. Fuhrhop, J.-H., Mathieu, J.: Angew. Chem. Internal. Ed. Engl. 2jb 100-113 (1984)

25. Hurst, J.K., Lee, L.Y.C., Gratzel, M.: J. Am. Chem. Soc. 105, 7048-7056 (1983)

26. Mauzerall, D., Hong, F.T.: in Smith, K. (ed), Porphyrins and Metalloporphyrins, Elsevier, Amsterdam 701-728 (1975)

27. Hani, A., Mauzerall, D.: Biophys. J. 35, 79-92 (1981)

28. Krakover, Т., Hani, A., Mauzerall, D.: Biophys. J. 35,93-98 (1981)

29. Fuhrhop, J.-H., Wanja, U., Biinzel, M.: Liebigs Ann.Chem. 426-432 (1984)

30. Loser, A., Mauzerall, D.: Photochem. Photobiol. 38, 355-361 (1983)

31. Fuhrhop, J.-H., Mathieu, J.: J. Chem. Soc. Commun. 144-145, (1983)

32. Fuhrhop, J.-H., Fritsch, D., Schmiady, H., Tesche, В.: J. Am. Chem. Soc. 166, 1998-2001 (1984)

33. Calvin, M.: Photochem. Photobiol. 37, 349-360 (1983)

34. Tricot, Y.M., Fendler, J.H.: J.Am.Chem.Soc. 106, 2477(1984)

35. Hurley, J.K., Castelli, F., Tollin, G.: Photochem. Photobiol. 32 79-86 (1980)

36. Ford, W.E., Tollin, G.: Photochem. Photobiol. 38, 441-449 1983)

37. Tunuli, M.S., Fendler, J.H.: J.Am.Chem.Soc. 103, 2507, 1981,

38. Krieg, M., Whitten, D.G.: J.Am.Chem.Soc. 106, 2477-2479