light driven organometallic catalysis by

www.rsc.org/dalton | Dalton Transactions

FRONTIERS

Inspired by nature: light driven organometallic catalysis by heterooligonuclear Ru(II) complexes Sven Rau,*a Dirk Walthera and Johannes G. Vosb Received 2nd November 2006, Accepted 3rd January 2007 First published as an Advance Article on the web 19th January 2007 DOI: 10.1039/b615987g In view of diminishing resources and an ever increasing demand for energy attention has been directed towards the development of new photocatalytic systems modelling natural photosynthesis. Recent investigations have shown that supramolecular devices consisting of photosynthetic reaction centres and catalyst metals—connected via bridging ligands—are capable of performing light driven catalytic reactions such as hydrogen production, reduction of CO2 , and conversion reactions of olefins.

Introduction In the light of the rapidly increasing need for sustainable energy interest in the conversion of solar energy is growing. Solar energy is utilised as a source for thermal or electrical energy (photovoltaic cells),1 however, its application in light driven catalytic systems a

Institut f¨ur Anorganische und Analytische Chemie, Friedrich-Schiller-Universit¨at, Lessingstr. 8, 07743, Jena, Germany. E-mail: [email protected]; Fax: +49 3641 948102; Tel: +49 3641 948113 b National Centre for Sensor Research, School of Chemical Sciences, Dublin City University, Dublin 9, Ireland. E-mail: [email protected]; Fax: +353 1 7005503; Tel: +353 1 7005307

may open a route towards highly energy efficient chemical reactions. In this area only a relatively small number of publications have appeared.2 Most of these deal with homogeneous photocatalytic systems based on intermolecular energy or electron transfer processes. The general observation with this approach, which relies on the collision induced electron transfer between the different species involved, has been that catalytic efficiencies are low and that the presence of various reaction intermediates limits its applicability. There is, however, considerable potential for artificial photocatalytic systems, inspired by the natural photosynthetic process. Here the supramolecular assembly of light harvesting units and catalytic centres allows light driven reactions

Sven Rau studied chemistry at the Friedrich-Schiller-University of Jena and Dublin City University. He obtained his Diploma in chemistry in 1997 and his PhD in 2000. His research interests are focused on the synthesis and characterisation of photoredoxactive metal complexes their applications in light driven catalytic conversions, interaction with surfaces and with DNA. Prof. D. Walther received his PhD in 1968. In 1992 he became the chair of Inorganic chemistry at the University of Jena (Germany) and from 1994 to 1997 he was the director of the Institute of Inorganic and Analytical Chemistry in Jena. From 2002–2003 he was Vice-president of the Gesellschaft Deutscher Chemiker (Society of German Chemists). His research interests are focused on the activation of carbon dioxide at transition metal centres, supramolecular organometallic chemistry and oligometal complexes as homogeneous catalysts. Han Vos is Professor of Inorganic Chemistry at Dublin City University. His research interests are in the design of supramolecular systems containing transition metal complexes. Of particular relevance are the synthesis, photophysical and electrochemical properties of dinuclear and polymeric ruthenium and osmium polypyridyl complexes both in solution and when immobilised on solid substrates. The applications of these compounds as optical and electrochemical sensors and in molecular electronics are investigated.

Sven Rau This journal is © The Royal Society of Chemistry 2007

Dirk Walther

Johannes G. Vos Dalton Trans., 2007, 915–919 | 915

such as the oxidation of water. It is anticipated that artificial assemblies copying the fundamental biological construction principles would result in an intramolecular system consisting of a light harvesting unit and a catalytic centre. (Fig. 1.).

the successful design of such a light driven catalytic assembly is the tuning of electronic communication between the three components allowing the desired vectorial electron or energy transfer processes from antenna to catalytic centre to occur. There are two discernable pathways for light driven catalysis: First, systems which catalyse reactions but where no net electron transfer between the catalyst and the substrate occurs, in other words no photoredox reaction takes place. In this case the photoeffect may be explained by energy transfer to, or triplet sensitisation of M or the substrate, see eqn (1). Ru(II)BM + hm → Ru(II)*BM → Ru(II)BM*

Fig. 1 Essential construction requirements and potential light driven catalysis.

Artificial photosynthetic systems based on covalently linked ruthenium–manganese centres aimed at reproducing water splitting have shown intramolecular electron transfer from the manganese to the photooxidised ruthenium, but no catalytic activity was observed.3 An intriguing option is the photocatalytic oxidation of secondary alcohols using a dinuclear ruthenium complex, however, only a small catalytic activity was observed.4 In addition the potential of catalytic devices based on the ligand dppz, where dppz is dipyrido[3,2-a:2
,3
-c]phenazine and its analogues capable of storing up to four photogenerated electrons has been proposed but until very recently no catalytic systems were reported.5 In this overview, the work so far carried out on the potential of heterooligonuclear metal complexes containing both a light harvesting system and a catalytic centre will be reviewed. The limited number of working catalysts known today permits a glimpse at an exciting field of research situated between inorganic photochemistry and organometallic catalysis which is still in its infancy. Principles and design The basic design of a covalently bound multinuclear assembly capable of acting as a photoinduced catalyst for intramolecular processes is shown in Fig. 1. Such an assembly consists of three components, a light harvesting centre Ru, a bridging ligand B and a catalytic centre M. Ruthenium polypyridyl complexes possess attractive photophysical and electrochemical properties and an extensive synthetic chemistry exists which allows their systematic tuning, such as controlling the nature of the excited 3 MLCT state.6 The nature of the catalytic centre, M, will be specific to the reaction to be catalysed. Properties of the metal centre, such as available redox states and nature of the coordinative bond at the catalytic centre determine the basic catalytic functions. Although these two components are important factors for efficient light absorption and catalytic activity, the assembly will only work when an appropriate bridging ligand or linker B, is chosen. Critical for 916 | Dalton Trans., 2007, 915–919

(1)

The second group comprises of catalytic reactions where the redox state of the substrate is altered by intramolecular photoinduced electron transfer between the antenna and the catalytic centre M. At present the catalytic cycle can only be closed if a sacrificial agent is utilised to reform the photo oxidised or reduced Ru centre. The use of such sacrificial agents is detrimental for the overall energy balance of the reaction, however, intelligent utilisation of the full redox power of the photoexcited ruthenium might allow for the coupling of catalytic oxidation with catalytic reduction. Natural photosynthesis applies this concept yielding O2 and reduced organic matter. A general reaction pathway involving electron transfer from a Ru(II) antenna centre to the catalytic centre, eqn (2), and reduction by the sacrificial donor D is shown below, eqn (3); Ru(II)BM + hm → Ru(II)*BM → Ru(III)B(−)M

(2)

Ru(III)B(−)M + D → Ru(II)B(−)M → Ru(II)BM−

(3)

Synthetic aspects The synthesis of heteroleptic ruthenium complexes with potential bridging ligands starts most often with [(Rbpy)2 RuX2 ] which can be obtained in high yields and purity using microwave assisted reactions.7 In the following step the bridging ligand can be coordinated at this centre to form Ru–B and subsequent purification using HPLC methods results in the pure complex.8 The introduction of the catalytic metal centre (by the “complex as ligand approach”) may require more effort. In a recent example introduction of allyl–palladium units was possible using LiBu in dry deoxygenated solvents resulting in the formation of a heterotetranuclear Ru2 Pd2 -complex.9 This complex can be utilised as a thermal catalyst in the carbon–carbon bond forming Heckreaction (TON 4300). (a)

Photocatalytic reactions based on energy transfer

Dimerisation of 1-olefins. The heterobinuclear complex 1 (Scheme 1), recently reported by Akita et al.10 serves as a precatalyst for the photocatalytic dimerisation of 1-methylstyrene. Mechanistic investigations (scheme 2) suggest that the catalytic cycle starts with the palladium-hydrido species, B. Insertion of 1-methyl-styrene into the Pd–H bond then yields the complex C, which has been isolated and characterised. Only the subsequent insertion of the second substrate molecule in the Pd–C bond is light activated and catalysed by the excited state C*. The catalytic cycle is closed by thermal b-hydride elimination leading to the organic This journal is © The Royal Society of Chemistry 2007

Scheme 1 Photocatalyst 1 (left) for the dimerisation of 1-olefins and photocatalyst 2 (right) for the trans–cis isomerisation of 4-cyano-stilbene.

Fig. 2 Structures of the complexes 3–6.

Scheme 2 Hypothetical mechanism of the photocatalytic dimerisation of a-methylstyrene.10

product and regeneration of B. Until now no detailed properties of the excited state C* are known. The attainable turn over numbers of more than 90 within 4 h and the high selectivity of the reaction rival the most advanced thermal catalytic systems.11–16 Trans–cis-isomerisation of trans-4-cyanostilbene. Osawa et al. have reported that the binuclear photocatalyst 2 (Scheme 1) consisting of a phosphin-substituted ruthenium light harvester and a Cp(CO)Ru fragment acts as a photocatalyst for the catalytic conversion of trans-4-cyanostilbene to cis-4-cyanostilbene.17 Preliminary mechanistic investigations suggest that after coordination of trans-4-cyanostilbene at the catalyst centre an intramolecular triplet–triplet sensitisation is the light driven step which results in the formation of the product cis-4-cyano-stilbene. In comparison with this intramolecular reaction intermolecular systems show lower efficiencies proceeding via a triplet–triplet sensitisation.18,19 (b)

Photocatalytic reactions based on electron transfer

Photocatalytic CO2 -reduction with supramolecular Ru(II)–Re(I) catalysts. The catalytic potential of bi- and tetranuclear Ru(II)– Re(I) complexes for the conversion of CO2 into CO has been demonstrated by Ishitani et al.20 They constructed the heterooligonuclear complexes 3–6 and could for the first time show how this activity is influenced by the nature of the bridging ligand and the ligands at the light harvesting ruthenium complex. 3 to 6 comprise all of the light harvesting ruthenium complex which is connected via different bridging ligands to a [(bpy)Re(I)Cl(CO)3 ] unit. The latter which is well-known as a catalytic centre for CO2 conversion, (Fig. 2).21 Investigations of the photocatalytic properties of these compounds using light of k > 480 nm in DMF/TEOA (here This journal is © The Royal Society of Chemistry 2007

acting as base) with 1-benzyl-1,4-dihydronicotinamide (BNAH) as sacrificial donor suggests that upon light absorption a ruthenium 3 MLCT state is generated which is localised on the ruthenium coordinated bipyridine part of the bridging ligand. Subsequently, the photooxidised Ru(III) is reduced by BNAH. The following intraligand electron transfer is relatively slow and results in the migration of the electron to the rhenium coordinated bipyridine part of the bridging ligand. The so formed [(R-bpy• -)Re(I)Cl(CO)3 ] moiety is known to react with CO2 upon Cl ligand loss21 a process which proved to be the rate limiting step in this catalysis. As the reduction of CO2 requires two electrons it is assumed that the second electron transfer follows a similar route. The strong influence of the bridging ligand on the photocatalytic activity is illustrated by investigation of 6 which contains a fully conjugated bridging ligand and displays only a very small reactivity. Introduction of CF3 substituents in the ruthenium coordinated bipyridine ligands (complex 5) results in an exclusive light induced electron transfer to the (CF3 )2 bpy instead of the bridging ligand. Consequently, the photocatalytic activity of complex 5 is significantly lower if compared with complex 3. Photocatalytic hydrogen production and related reactions. The possibility to produce molecular hydrogen with multicomponent photocatalytic systems based on intermolecular electron transfer events has been known for nearly 30 years.22 A detailed description of an intramolecular concept combining all the essential components of the redox chain in one molecule was first put forward in 2003.23 However, no photocatalytic hydrogen has yet been reported using this photocatalyst, containing an iron hydrogenase moiety (Fig. 3, complex 7). In 2006 Sakai et al. investigated the heterobinuclear ruthenium– platinum photocatalyst 8 (Fig. 3) which showed in water in the presence of EDTA as sacrificial donor and light (k ≥ 390 nm) only a very small catalytic formation of hydrogen with an activity of 2.4 mol H2 per mol cat (TON = 2,4).24 No mechanistic details are currently available for this system. The catalytically more active supramolecular ruthenium– palladium catalyst 9 (Fig. 3) has achieved a TON(H2 ) of 56 if irradiated with visible light (k ≥ 450 nm) in acetonitrile using triethylamine (TEA) as sacrificial donor.25 In addition to the photocatalytic hydrogen production, the heterobinuclear complex also catalyses the selective hydrogenation of diphenylacetylene to cis-stilbene under irradiation with visible light (k ≥ 450 nm) Dalton Trans., 2007, 915–919 | 917

Fig. 3 Heterooligonuclear complexes 7–10 containing a light harvesting unit and a potential catalytic centre, L is 2,2
-bipyridine, L
is 4,4
-di-t-butyl-2,2
-bipyridine.

and use of TEA (TON = 63). The fact that only cis-stilbene is produced and the absence of any catalytic activity of the palladium free mononuclear ruthenium complex proves that the catalytic palladium site is involved in this reaction. Preliminary investigations of this process suggest an essential involvement of the tetrapyridophenazine bridging ligand, TPPHZ, which serves as electron reservoir. Replacement of this ligand by bipyrimidine results in complete loss of catalytic activity. DFT calculations show that the photo reduction of the phenazine induces a dissociation of the chloride ligand from the palladium centre which results in subsequent charge transfer to the palladium centre. This restores the photoactive Ru–TPPHZ unit which is now in position to accept a second electron which subsequently enables the formation of molecular hydrogen from the doubly reduced catalyst. The relevance of this theoretical finding for the catalytic process could be experimentally proven: Addition of chloride ions results in the inhibition of hydrogen production. Brewer et al. have reported the heterotrinuclear complex 10 (Fig. 3) which contains two photoactive Ru-centres and one catalytic Rh(III) centre26 capable of performing visible light induced photocatalytic hydrogen production in acetonitrile water mixtures with N,N-dimethylaniline as sacrificial donor. Unfortunately no detailed information on the amount of hydrogen produced or the underlying mechanisms are available.27 Comparison of the active photocatalysts 3–6 and 8–10 shows that the bridging ligand obviously influences the catalytic activity to a great extent and the combination of bridging ligand and catalytic metal centre has to be carefully tuned in order to achieve an active catalyst.

Outlook The heterooligonuclear complexes presented here are the first supramolecular catalysts consisting of a light harvesting unit, a linker and a reaction centre which can use light as a tool to selectively control organometallic catalysis. Recent synthetic advances have led to the synthesis of heterotrinuclear complexes with one light harvesting unit and two catalytic palladium centres.28 Within this class two groups are clearly discernable by the redox processes involved: in the first transformation of a substrate by energy transfer, and in the second utilisation of electrons transferred via the catalyst to the substrate take place. 918 | Dalton Trans., 2007, 915–919

For the first class of catalysis it is at this stage unclear which design principles have to be implemented to synthesise active catalysts where excited state energy of the light harvester can be utilised for a catalytic synthesis with highly valuable reaction products. A highly attractive, not realised reaction, would for example be the polymerisation of unsaturated substrates where under the influence of light the monomer m can be introduced in a stereochemically defined manner which is different from the “dark” reaction. Another similar approach could be the copolymerisation of different monomers m, m
where the ratio of m/m
in the copolymer can be tuned using light. For the construction of supramolecular catalysts of the second group which involves photo redox reactions it is evident that any active catalyst must contain a bridging ligand which permits a photo electron transfer and forms a very stable bond to the catalytically active metal centre in different oxidation states. As shown for the complex 9 bridging ligands might in addition serve as intermediary electron storage site. A more detailed understanding of the potentially very complex photoinitiated processes especially on a very short timescale will help in the development of potential photocatalysts as has been shown in a preliminary investigation for a tetranuclear Ru2 Pd2 complex.29 A more fundamental problem is the utilisation of sacrificial donors in the catalytic preparation of hydrogen or conversion of carbon dioxide. These will have to be replaced by water as sacrificial donor or by electron donors which are converted into useful products under simultaneous formation of H2 or CO. In this direction, the research towards light driven oxidation of a coordinated metal centre may pave the foundation for generating photooxidising catalysts.3 In conclusion, the photocatalytic reactions using heterooligonuclear complexes is an emerging field at the interface of two well established areas, the organometallic catalysis and the photo redox chemistry of transition metal complexes.

Acknowledgements S. R. and D. W. acknowledge financial support by the DFG, SFB 436.

References 1 (a) B. O’Regan and M. Gr¨atzel, Nature, 1991, 353, 737; (b) M. K. Nazeeruddin, Coord. Chem. Rev., 2004, 248, 1161. 2 E. Amouyal, Sol. Energy Mater. Sol. Cells, 1995, 38, 249. ˚ kermark and S. Styring, Chem. Soc. ¨ 3 L. Sun, L. Hammarstrom, B. A Rev., 2001, 30, 36. 4 J. H. Alstrum-Acevedo, M. K. Brennaman and T. J. Meyer, Inorg. Chem., 2005, 44, 6802 and references cited therein. 5 R. Konduri, H. Ye, F. M. MacDonnell, S. Serroni, S. Campagna and K. Rajeshwar, Angew. Chem., Int. Ed., 2002, 17, 3185. 6 J. G. Vos and J. M. Kelly, Dalton Trans., 2006, 4869. ¨ 7 S. Rau, B. Sch¨afer, A. Grussing, S. Schebesta, K. Lamm, J. Vieth, H. ¨ Gorls, D. Walther, M. Rudolph, U. W. Grummt and E. Birkner;, Inorg. Chim. Acta, 2004, 357, 4496. 8 L. O’Brien, M. Duati, S. Rau, A. L. Guckian, T. E. Keyes, N. M. ¨ and J. G. Vos, Dalton Trans., 2004, 514. O’Boyle, A. Serr, H. Gorls ¨ ¨ 9 D. Walther, L. Bottcher, J. Blumhoff, S. Schebesta, H. Gorls, K. Schmuck, S. Rau and M. Rudolph, Eur. J. Inorg. Chem., 2006, 12, 2385. 10 A. Inagaki, S. Edure, S. Yatsuda and M. Akita, Chem. Commun., 2005, 5468. 11 (a) S. R. Stoyanov, J. M. Villegas and D. P. Rillema, Inorg. Chem., 2002, 41, 2941; (b) W. Matheis and W. Kaim, Z. Anorg. Allg. Chem., 1991,

This journal is © The Royal Society of Chemistry 2007

12 13 14 15 16 17 18 19 20 21

593, 147; (c) R. Sahai, L. Morgan and D. P. Rillema, Inorg. Chem., 1988, 27, 3495. (a) P.-R. Zhang, J.-Y. Du Yang, Fan G. Zou and J. Tang, Chin. J. Chem., 2005, 23, 581; (b) Q. Cai, J. Li, F. Bao and Y. Shan, Appl. Catal., A, 2005, 279, 139. (a) B. Rac, G. Mulas, A. Csongradi, K. Loki and A. Molnar, Appl. Catal., A, 2005, 282, 255; (b) C. Peppe, E. Schulz Lang, F. M. de Andrade and L. B. de Castro, Synlett, 2004, 1723. M. Higashimura, K. Imamura, Y. Yokogawa and T. Sakakibara, Chem. Lett., 2004, 33, 728. T. Tsuchimoto, S. Kamiyama, R. Negoro, E. Shirakawa and Y. Kawakami, Chem. Commun., 2003, 852. M. Saito, Y. Nishibayashi and S. Uemura, Organometallics, 2004, 23, 4012. M. Osawa, M. Hoshino and Y. Wakatsuki, Angew. Chem., Int. Ed., 2001, 40, 3472. M. Wrighton and J. Merkham, J. Phys. Chem., 1973, 77, 3042. G. Gennari, G. Gallazzo and G. Cauzzo, Gazz. Chim. Ital., 1980, 110, 259. B. Gholamkhass, H. Mametsuka, K. Koike, T. Tanabe, M. Furue and O. Ishatani, Inorg. Chem., 2005, 44, 2326. J. Hawecker, J.-M. Lehn and R. Ziessel, J. Chem. Soc., Chem. Commun., 1983, 69, 536; J. Hawecker, J.-M. Lehn and R. Ziessel, Helv. Chim. Acta, 1986, 1990; K. Koike, H. Hori, M. Ishizuka, J. R. Westwell, K. Takeuchi, T. Ibusuki, K. Enjouji, H. Konno, K. Sakamoto and

This journal is © The Royal Society of Chemistry 2007

22 23

24 25

26 27 28 29

O. Ishitani, Organometallics, 1997, 16, 5724; D. Walther, M. Ruben and S. Rau, Coord. Chem. Rev., 1999, 182, 67–100. J.-M. Lehn and J.-P. Sauvage, Nouv. J. Chim., 1977, 1, 449; J. I. Goldsmith, W. R. Hudson, M. S. Lowry, T. H. Anderson and S. Bernhard, J. Am. Chem. Soc., 2005, 127, 7502. ˚ kermark and L. Sun, Angew. Chem., Int. S. Ott, M. Kritikos, B. A ¨ L. Hammarstrom, ¨ J. Ed., 2004, 43, 1006; H. Wolpher, M. Borgstrom, ˚ kermark, Inorg. ¨ Bergquist, V. Sundstrom, S. Styring, L. Sun and B. A ˚ kermark and S. Ott, Chem. Commun., 2003, 6, 989; and; L. Sun, B. A Coord. Chem. Rev., 2005, 249, 1653. H. Ozawa, M.-A. Haga and K. Sakai, J. Am. Chem. Soc., 2006, 128, 4926. S. Rau, B. Sch¨afer, D. Gleich, E. Anders, M. Rudolph, M. Friedrich, H. ¨ Gorls, W. Henry and J. G. Vos, Angew. Chem., Int. Ed., 2006, 45, 6215; presented in part at Chemiedozententagung Munich, book of abstracts p. 92, March, 2005. S. Swavey and K. Brewer, Inorg. Chem., 2002, 41, 4044; M. Elvington and K. J. Brewer, Inorg. Chem., 2006, 45, 5242. K. J. Brewer and M. Elvington, US 20060120954A1, 2006. M. Gagliardo, G. Rodriguez, H. H. Dam, M. Lutz, A. L. Spek, R. W. A. Havenith, P. Coppo, L. De Cola, F. Hartl, G. P. M. van Klink and G. van Koten, Inorg. Chem., 2006, 45, 2143. ¨ B. Dietzek, W. Kiefer, J. Blumhoff, L. Bottcher, S. Rau, D. Walther, U. Uhlemann, M. Schmitt and J. Popp, Chem.–Eur. J., 2006, 12, 5105.

Dalton Trans., 2007, 915–919 | 919