The Development of Molecular Wires PART 11: ROLE OF AND OSMIUM POLYPYRIDINE COMPLEXES FOR FAST VECTORIAL ELECTRON TRANSFER By V. Grosshenny, A. Harriman, M. Hissler and R. Ziessel Ecole Europkenne de Hautes Etudes des Industries Chimique de Strasbourg, Universitk Louis Pasteur, Laboratoire de Chimie d’Electronique et de Photonique Molkculaires, Strasbourg, France

The concludingpart of thispaper on the use of ruthenium(ZI) and osmium(II) polypyridyl complexes, as molecular sized terminal subunits that are linked together bypolyyne bridges functioning as molecular girders to retain the stereo- chemical rigidity, deals with the process of electron transfer between the subunits and considers the benefits conferred by the use of polyyne bridges. The ruthe- nium and osmium complexes have properties which aid the selective promo- tion of an electron from the metal to the bridging ligand, together with amenable absorption and emission spectral pro+, and facile oxidation-reduction processes. This makes them promising candidates for vectorial electron transfer. Future work to extend the lengths of the linkages, to ensure unidirectional and long- range electron tunnelling, and to anchor the wires to supports is discussed. These are the necessary requirements for the development of molecular wiring made from these materials forfuture use with molecular-scale electronic devices.

The first part of this paper introduced the sub- metal centres, must occur between more widely- ject of molecular wires and considered the struc- spaced reactants and it displays a more signifi- ture and of the complexes that can cant attenuation factor. be used for them, and other materials currently Interestingly, the insertion of a (I1) believed to be the best for this purpose (13). bis-trialkylphosphine complex into the polyyne Here, the topic is further illustrated by reference bridge, see Figure 6 and Panel 6, has the effect to light-induced electron transfer. of inhibiting through-bond electron-transfer The selective population by an electron of the reactions occurring from the terminal ruthe- ethynyl-substituted terpyridyl ligand, followed nium(I1) polypyridine complex (14). This by electron delocalisation over an extended n*- situation arises because the platinun(I1) centre orbital is of extreme importance for the design donates charge to the ethynyl-substituted of effective molecular wires. It gives direction bipyridyl ligand, making its reduction potential to the electron flow, since, upon excitation of more negative than that of the corresponding the ruthenium(I1) complex, an electron is platinum(I1)-free ligand, see Panel 6. pushed along the molecular axis towards the Consequently, excitation of the ruthenium(I1) acceptor. In principle, this electron can approach complex results in selective electron donation close to the cation that is co-ordinated to the from the metal centre to the unsubstituted lig- second terpyridyl ligand, so that the actual elec- and. The electron is localised at this ligand and tron-transfer step might occur over a relatively is unable to migrate to an ethynyl-substituted short distance. However, there is no suggestion ligand because of the unfavourable thermody- that the extended x*-orbital includes the sec- namic position; thus the photophysical proper- ond terpyridyl ligand bound to the bridge. The ties remain virtually unchanged from those of corresponding hole transfer, which involves both the unsubstituted complex.

Platinum Metals Rev., 1996,40, (2), 72-77 72 Fig. 6 Molecular structures of the trinudear tmns-Ru"J't"and trans-Ru"Pt"0s"complexes

The same situation is observed for the anal- polyyne wire (7), as illustrated in Figure 8. Here, ogous osmium(II) complexes, see Figure 6, but excitation of one of the chromophores results not for the rhenium0 bipyridine complex shown in the formation of the corresponding triplet in Figure 7. In this latter case, excitation of excited state of the metal complex, which is of the rheniumm chromophore results in the trans- metal-to-ligand charge-transfer character. These fer of an electron from the (1) centre to binuclear complexes luminesce at much longer the co-ordinated bipyridine ligand. wavelength than do the reference compounds, Although the presence of a central platinum@) due to the lower energy of the charge-transfer complex lowers the reduction potential of the state. Furthermore, the triplet lifetimes of the bipyridine ligand because of charge injection, binuclear complexes are significantly longer, rel- the metal-to-ligand charge-transfer transition ative to analogous mononuclear ruthenium(I1) still occurs in the platinum(I1)-contaig rhe- polypyridine complexes of similar energy, at nium(I) complexes, with the result that the pro- both room temperature and 77 K. moted electron resides in an extended n*-orbital. In these polyyne-bridged binuclear complexes This causes a substantial enhancement in the it appears that, under illumination, an triplet lifetimes of the platinum(I1)-containing rhenium(1) complexes (- 80 ns) relative to that of the unsubstituted complex (- 40 ns), as mea- sured in solution at room temper- ature. The stabilised triplet state arises because of decreased vibronic coupling between triplet and ground states, and is a common feature of ethynyl-substituted polypyridine ligands.

Symmetrical Binuclear Complexes A special case can be made for binuclear com- Fig. 7 Molecular structure of the trinuclear plexes possessing identical ruthenium(I1) trans-Re'J't" complex polypyridine complexes at opposite ends of a I

Platinum Metals Rm., 1996, 40, (2) 73 Panel 6

The o-alkynyl-bipyridine derivatives of was prepared by a catalytic homo-coupling platinum(II) were stereoselectivelyprepared reaction of the ethynyl-substitutec from the ethynyl-substitutedbuilding blocks bipyridine in the presence of CuCl, and nuns-[Ptn(P"Bu3),C1,], using a CuI catal- tetramethylethylenediamine complexes ysed reaction. The platinum(I1)-free ligand see Chart A, below.

Br

(II) KF, MeOH 3 84%

Chart A 98% NQ)

The parent &compound of the metallo- the heterotrinuclear complex RunPtnOsn,sel synthon was similarly prepared (85 per Chart B, below. cent yield) from the parent cis- W"(P "B~,>,Cl,l. \ Selective com- plexation of one of the two bipy- ridine domains with complex ~-rRU(bPY),CU. 2H,O (bpy is 2,2'-bipyridine), afforded either the dinuclear PtnRunor the tri- nuclear Pt1'Ru1lZ complexes. Further com- plexation of the free bipyridine subunit within the PtnRuncom- plex with the

cis- [OS(bPY)2C121 UUULI afforded, in a RuPtOs '86% clean reaction,

Platinum Metals Rev., 1996,40, (2) 74 Fig. 8 Molecular structures of the ethynyl-substituted, diethynyl-substituted, diphenyl -substituted and diphenyldiaeetylene-substihrted dinuclear ruthenium(I1) complexes

electron is promoted from one of the ruthe- erties of the mononuclear reference compounds. niumm centres to a n*-orbital associated with Overall, it is clear that polyynes provide sev- the bridging ligand. As suggested earlier, this eral important benefits when covalently attached n*-orbital is not restricted to the polypyridine to transition metal polypyridine complexes of ligand but, because of conjugation, it extends the type considered here. They can cause a over part of the bridging polyyne chain. marked increase in triplet lifetime, thereby Due to the similarity of the terminal groups, favouring reactions with adventitious quenchers, photon migration can take place in which simul- and can induce a sense of directionality into taneous electron- and hole-transfer steps serve intramolecular energy- or electron-transfer steps. to alternate the photon between the two termini, Polyynes operate as molecular girders to retain see Panel 7. In this manner, the mplet state shut- stereochemical rigidity and, at the same time, tles between the terminal metal complexes, each they provide highly effective electron and hole journey taking about 1 ns, and can react with channels. a quencher at either site. This is a remarkable process, made possible by the strong Future Work on Molecular Wires electronic communication provided by the It is now necessary to fabricate longer polyyne polyyne bridge, and closely resembles the bridges and to ensure that the excellent elec- natural light harvesting complexes found in pho- tronic conductivity is not restricted to short tosynthetic bacteria (1 5). chain lengths. In this respect, it is important Decreasing the extent of electronic coupling to realise that the energy of the bridge is likely between the ruthenium(I1) polypyridine com- to decrease with increasing length and, in terms plexes, by inserting a bridging phenyl ring, see of superexchange theory, the degree of through- Figure 8, severely inhibits triplet migration and bond electronic coupling will increase as the the photophysical properties of such binuclear energy of the bridge approaches that of the complexes resemble the photophysical prop- excited triplet state. Each added ethynyl group,

Platinum Metals Rev., 1996, 40, (2) 75 Panel 7

According to superexchange theory, the ate energy gap would be between HOMOS electronic matrix coupling element (V,.,,), a associated with bridge and triplet. The mag- term that quantifies the amount of electronic nitudes of C, and Cpare controlled by the communication between the terminal metal nature of the reactants, especially stereo- complexes, can be expressed in the follow- chemical aspects, and can be estimated by ing form: detailed molecular orbital calculations. For an homologous series of donor-accep- C,CB/6E V,, = tor compounds, V, is predicted to decrease where C, and C@are the atomic orbital coef- exponentially with increasing length (R) of ficients describing coupling between the the bridge according to: triplet excited state and the bridge, and V,, = VoAoexp [-pR/2] between the bridge and the final acceptor, respectively; 6E represents the energy gap Here, VDA0refers to the electronic coupling between appropriate orbitals localised on the matrix element when the reactants are in triplet state and on the bridge. For electron orbital contact. For the polyyne-bridged mol- transfer, 6E refers to the difference in energy ecular systems considered in this work, it between LUMOs associated with bridge and appears that V,, will decrease with increas- triplet while for hole transfer the appropri- ing R due to the effective decreased overlap of wavefunctions associated with the reacting species. However, the energy transfer gap between triplet state and bridge (6E) ) +Cz\electron will also decrease with increasing R (see the Chart) because of conjugation, such that V,, might actu- ally increase with increasing length of the bridge. These hole transfer offsetting properties give rise to particu- larly small p values, at least over modest separations.

therefore, should act co-operatively to decrease if these complexes possess unique properties the energy of the bridge and this effect might that make them particularly attractive subunits offset the decrease in the magnitude of elec- for molecular-scale electronic devices. The tronic coupling that occurs upon increasing the importance of these complexes lies in the fact length of the bridge, see Panel 7. Such prop- that their triplet excited states, often long lived erties might serve to minimise the attenuation and formed in quantitative yield under visible- factors for through-bond electron and hole trans- light illumination, are of metal-to-ligand charge- fer. As such, polyynes possess many of the req- transfer character. This property facilitates selec- uisites for good molecular wires. tive promotion of an electron from the metal It is also instructive to consider the role of the centre to the bridging ligand, thereby provid- metal complex in such systems and to enquire ing the impetus for vectorial electron transfer.

Platinum Metals Rev., 1996, 40, (2) 76 Other common types of chromophore do not attempts to provide both protection and inher- provide this directionality which is inherent in ent redundancy. ruthenium(II), osmium(I1) and rhenium(1) Other types of bridge, such as those formed polypyridine complexes. Of these complexes, from polycondensed aromatic residues (4-6) or those based on ruthenium(I1) or osmium(I1) DNA (1 6),provide a more stable environment look to be the more promising because of their and exhibit interesting properties. amenable absorption and emission spectral pro- Theoretical models, predicting very small files and their facile oxidation-reduction attenuation factors for electron tunnelling processes. through certain materials, are also beginning to We should also consider, at this stage, the fea- appear (17) and major ndvances are taking place sibility of constructing molecular-scale wires in the synthetic practices used to fabricate that promote unidirectional and long-range elec- supramolecular entities (9). These various fac- tron tunnelling. Several such systems have been tors combine to ensure that appropriate mole- described (9) and many more seem certain to cular wires will soon be available. The next step emerge in the near future. Perhaps the most will involve anchoring the wires to a macroscopic effective bridges are those based on polyenes support and, by using scanning tunnelling elec- (2), where charge transfer has been observed to tron microscopy and time-resolved reflectance occur over distances in excess of 30 A. These , the means to study such inter- spacer moieties are probably too unstable for actions are in place. One might argue, there- practical purposes, hence our utilisation of fore, that the age of molecular electronic devices polyynes, although there have been innovative is drawing ever closer.

References 13 V. Grosshenny, A. Harriman, M. Hissler and R. 15 G. McDermott, S. M. Prince, A. A. Freer, A. M. Ziessel, Platinum Metals Review, 1996, 40, (l), Hawthornthwaite-Lawless, M. Z. Papiz, R. J. 26-35 Cogdell and N. W. Isaacs, Nature, 1995,374,5 17 14 M. Hissler and R Ziessel, 3 Chem. SOC.,Ddkm 16 A- M- Brunand A. Hdan,3Am. Chem. Trans., 1995, 893; A. Harriman, M. Hissler, R. 1994,116,10383 Ziessel, A. De Cian and J. Fisher, 3. Chem. SOC., 17 A. K. Felts, W. Thomas Pollard and R A. Friesner, Dalton Trans., 1995, 4067 J. Phys. Chem., 1995,99,2929 Preparation of Platinudolvmer NanocomDosites 1 J I A simple method of producing metaVpolymer structure and has a high permeation rate in poly- composites, in which nanosized metal particles mers. is also a versatile process are homogeneously distributed throughout the solvent and it is clean and environmentally polymer, would promote the development of a friendly. new generation of useful materials and devices. The precursor, dimethyl(cyc1ooctadiene)plat- These devices would utilise the specific com- inum(II), was dissolved into the carbon diox- positions and size-dependent electrical, chem- ide, and then sequentiallyimpregnated into, and ical and magnetic properties of the composite. reduced in, thick films of the polymers poly(4- Some syntheses for specific applications have methyl- 1-pentene) and poly(tetrafluoroethy1- been reported, but no general synthesis for such ene). The reduction to metallic platinum was a nanocomposite material has emerged. by hydrogenolysis and thermolysis. This resulted Now, however, scientists at the University of in platindpolymer composites containing plat- Massachusetts, U.S.A., have developed a inum clusters, of sizes &om 15 to over 100 nm syrithesis for making platinum/polymer (depending upon preparation) being dispersed composites a. J. Watkins and T. J. McCarthy, throughout the films. The sizes and distribu- “Polymer/Metal Nanocomposite Synthesis in tions of the platinum clusters were measured by Supercritical COT, Chem. Mum.,1995,7, (1 l), SEM and TEM. 1991-1994). Their method involves the use of In general, adjusting the permeation and reduc- a supercritical fluid, in this case carbon diox- tion rates controls the size and distribution of ide. As a supercritical fluid, carbon dioxide offers the platinum clusters, and hther work is being great control over composite composition and done to produce gradient structures.

Platinum Metals Rev., 1996, 40, (2) 77