Phosphorus dendrimers as supports of transition metal catalysts

Inorganica Chimica Acta xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Inorganica Chimica Acta journal homepage: www.elsevier.com/locate/ica

Review

Phosphorus dendrimers as supports of transition metal catalysts Anne-Marie Caminade ⇑, Armelle Ouali, Régis Laurent, Jean-Pierre Majoral CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne, BP 44099, F-31077 Toulouse Cedex 4, France Université de Toulouse, UPS, INPT, F-31077 Toulouse Cedex 4, France

a r t i c l e

i n f o

Article history: Received 28 August 2014 Received in revised form 23 October 2014 Accepted 28 October 2014 Available online xxxx Keywords: Dendrimers 3d elements 4d elements Ligands Homogeneous catalysis

a b s t r a c t Dendrimer are hyperbranched macromolecules, which can be used as support of catalytic entities. This review focusses on polyphosphorhydrazone (PPH) dendrimers, bearing as terminal groups, transition metal complexes of 3d elements (Cu, Sc) and 4d elements (Pd, Ru, Rh). These complexes have been used as catalysts for various organic reactions, in particular C–C couplings. Most of these dendritic complexes can be recovered and re-used several times (up to 12 successive runs with the same efficiency). In several cases, a dendritic (or dendrimer) effect is observed, i.e., an increased efficiency and/or enantioselectivity when comparing a monomeric catalyst with increasing generations of the dendrimers, using in all cases the same number of catalytic entities. Ó 2014 Elsevier B.V. All rights reserved.

Anne-Marie Caminade is Director of Researches at the CNRS in Toulouse since 1997 (first class since 2004) and presently head of the ‘‘Dendrimers and Heterochemistry’’ group at the LCC Toulouse since 2006. After two PhDs from the University of Toulouse (1984 and 1988) and two Post-docs (IFPParis and Von Humboldt fellow in Saarbrücken), she was recruited at the CNRS in 1985. She developed several aspects of phosphorus chemistry, including low coordinated compounds, transition metals coordination, and macrocycles syntheses. Her current research interest is on the synthesis, reactivity, and applications of dendrimers in particular as catalysts, for nanomaterials and for biology. She is the co-author of more than 390 publications in journals, 40 book chapters and 30 patents (h index 56).

Armelle Ouali received her PhD from the University of Montpellier in 2005 in the field of copper-catalyzed arylation of nucleophiles (Dr Marc Taillefer). After a post-doc in the synthesis of phosphorous and silicon-based dendrimers for biological applications (Drs Jean-Pierre Majoral and Anne-Marie Caminade, Toulouse), she moved to the University of California for a post-doc in carbene chemistry (Prof. Guy Bertrand, Riverside). She joined the group of Dr Anne-Marie Caminade in 2008 where she is involved in the development of new catalytic systems for various applications in organic synthesis. Her current interest is the design of dendrimeric metal-based or organic catalysts.

⇑ Corresponding author at: CNRS, LCC (Laboratoire de Chimie de Coordination), 205 route de Narbonne, BP 44099, F-31077 Toulouse Cedex 4, France. E-mail addresses: [email protected] (A.-M. Caminade), [email protected] (A. Ouali), [email protected] (R. Laurent), jean-pierre. [email protected] (J.-P. Majoral). http://dx.doi.org/10.1016/j.ica.2014.10.035 0020-1693/Ó 2014 Elsevier B.V. All rights reserved.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

2

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Régis Laurent was born in Porvoo (Finland) and studied chemistry at the University of Toulouse (France) where he received his PhD in 1994 under the supervision of Pr. J. Dubac and Pr A. Laporterie, working on microwave activation in organic chemistry. After post-doctoral studies for the BASF Company at the University of Saarbrücken (Germany) in the group of M. Veith, and an assistant researcher-teacher position (ATER) at the University of Toulouse in the group of L. Gorrichon, he got in 1996 a position at the CNRS in the group of J.-P. Majoral in Toulouse. He is currently ‘‘Chargé de Recherches’’. His research interests are in the synthesis and characterization of phosphorus-containing dendrimers and dendrons, their applications in catalysis, asymmetric catalysis, catalysis in water, and in their incorporation in inorganic materials. He is in charge of the technological platform Technopolym (polymer characterization).

Jean-Pierre Majoral is Emeritus Director of Research at the CNRS in Toulouse. His research interest is focused on the design and the properties of macromolecules such as phosphorus dendrimers and hyperbranched polymers. Main efforts are directed to the use of dendrimers in medicinal chemistry and material sciences. Emphasis is also laid on immobilization of molecular and macromolecular organo- and metal catalysts and their use for fine chemical synthesis. He is a member of several Academies of Sciences worldwide and an author of 545 publications and 45 patents.

Contents 1. 2.

3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dendritic complexes of elements from the second transition series (4d elements) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Palladium complexes of dendrimers as catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Ruthenium complexes of dendrimers as catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.3. Rhodium complexes of dendrimers as catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dendritic complexes of elements from the first transition series (3d elements) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Copper complexes of dendrimers as catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Scandium complexes of dendrimers as catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction Dendrimers are macromolecules constituted of branches emanating radially from a central core. Dendrimers induce a large and increasing interest in the scientific community, due to the numerous properties they have in different fields, in particular for catalysis, for the elaboration of (nano)materials, and in biology/nanomedicine [1]. They are synthesized step-by-step, most generally by the repetition of a sequence of two reactions that induces a multiplication of the number of terminal functions, creating what is called a new ‘‘generation’’. Scheme 1 displays the principles for the divergent synthesis of dendrimers. The starting point is a multi-functional core, whose functions are activated, deprotected, or modified in the first step. A branched monomer is used in the second step to react with the modified core, affording the first generation dendrimer. The first generation has the same type of functions than the initial core, but the number of functions is multiplied, most generally by two, eventually by three [2], exceptionally by five [3]. If this sequence of reactions is absolutely quantitative, then the process can be used again, to afford the second generation, then the third, and so on. High generations have been attained for some dendrimers, generation 10 for PAMAM (polyamidoamine) dendrimers [4] and poly-L-lysine dendrimers

00 00 00 00 00 00 00 00 00 00 00

[5], generation 12 for PPH (polyphosphorhydrazone) dendrimers [6], and very recently generation 13 for polytriazine dendrimers [7] and polyphenylene dendrimers [8]. Obtaining these high generations is a long process, but most of the properties of dendrimers can be attained with lower generations, most often not higher than generations 4 or 5. A large number of publications and patents about dendrimers is related to catalysis. Indeed, dendrimers are considered in many cases as soluble supports of catalytic entities, most generally of transition metal complexes. The very first example in this field concerned a generation 1 carbosilane dendrimer ended by nickel derivatives, and used for the catalysis of the Kharasch addition [9]. The dendritic catalyst was less efficient than the monomeric catalyst, but it was proposed that only the dendrimer could be recovered and reused, using a membrane reactor. However, in most cases of re-uses of the catalysts, a bad solvent for the dendrimer is added to the reaction media after the catalytic experiments, to precipitate the dendritic catalysts. In most cases, this bad solvent is miscible with the solvent used for the catalysis, but it sufficiently modifies the media to induce the precipitation of the dendrimers. This is one of the major advantage when using a dendritic catalyst, the possibility to recover it easily, and to reuse it [10]. This aspect is particularly interesting when considering the

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

3

Scheme 1. Principle of synthesis of dendrimers. A new generation is created each time the number of terminal groups is multiplied.

cost of the catalyst (many transition metals are expensive, and dendrimers are also expensive, due to their lengthy method of synthesis), and the decrease of wastes (if the catalyst is easily taken off the products, their purification is facilitated). Of course the dendritic structure should remain (almost) intact for an efficient recycling. Different types of dendrimers have been used as catalysts, and the topic has been frequently reviewed [11–23]. In this paper we will present what has been done with phosphorus dendrimers of type PPH (polyphosphorhydrazone) as soluble supports of transition metal catalysts. The polyphosphorhydrazone dendrimers are synthesized by the repetition of two quantitative steps, starting most generally from a trifunctional core (SPCl3) [24] or a hexafunctional core (N3P3Cl6) [25]. The first step is the reaction of 4hydroxybenzaldehyde in basic conditions with the P-Cl functions. The second step is the condensation of the phosphorhydrazide H2NNMeP(S)Cl2 with the aldehyde functions, affording again P-Cl2 functions, as for the core, but with a number of functions multiplied by two. Both steps are quantitative, and can be repeated several times, affording at each step either chloride or aldehyde terminal functions. Both steps have been repeated up to the obtaining of the twelfth generation starting from the trifunctional core (the highest obtainable, Scheme 2A) [6], and the eighth generation starting from the hexafunctional core (presumably not the highest generation obtainable, Scheme 2B) [25]. In all cases, the structure is ascertained by NMR, and particularly by 31P NMR, which is an extraordinary tool for such purpose [26]. To differentiate both families in this review, the dendrimers built from the trifunctional core will be noted G, those built from the hexafunctional core will be noted Gc, when the terminal groups are P(S)Cl2. They will be noted G0 and Gc0 , respectively, when the terminal groups are aldehydes. In all cases, the number refers to the number of the generation n. For a given generation, the series Gcn (or Gc0 n) has twice the number of terminal groups compared to the series Gn (or G0 n). Both the expanded and linear structures are shown for several dendrimers (Scheme 2). In all the other schemes, the linear form will be used, but the reader should keep in mind that it represents the 3-dimensional structure of dendrimers. Both P(S)Cl2 and aldehyde functions are particularly reactive and versatile; both have been used for the grafting of ligands. The complexation of metals necessitates ligands that have to be especially engineered to be grafted to the dendrimers, in particular to avoid the possibility of interaction of a single metallic entity with two branches of a dendrimer or between two dendrimers. As many different types of ligands have been used, in each case the grafting of the ligand will be presented first, followed by the use of the corresponding complexes for catalysis. The ligands used are very often phosphine derivatives, as they constitute the most versatile ligands for most transition metals, in particular for 4d elements [27,28]. These ligands can be monodentate (one phosphine), or most frequently bidentate (two phosphines, or one phosphine

and one nitrogen), eventually tridentate (3 double bonds or 3 nitrogen atoms). This review will be organized depending on the type of metals, and on the transition series to which they pertain (4d and 3d elements). It must be emphasized that in all cases the comparison of the efficiency between dendrimers and monomers or between several generations of dendrimers is done by considering the same number of catalytic centers. In the absence of any effect of the dendrimer, the efficiency should be identical. However, in many cases a difference is observed in the efficiency when the generation of the dendrimer increases. Such effect, which can be positive or negative, is called a dendritic (or dendrimer) effect [16,29]; it is illustrated in Fig. 1. 2. Dendritic complexes of elements from the second transition series (4d elements) 4d elements are certainly the most widely used transition metals for catalysis. In the case of PPH dendrimers, most of the complexes are of palladium (as for many other types of dendrimers [30]), and to a lesser extend of ruthenium and rhodium. 2.1. Palladium complexes of dendrimers as catalysts Most of the ligands used for the complexation of palladium are of bidentate type, principally P,N and P,P ligands. A few more specific and recent examples of tri and monodentate ligands will be also given. All ligands can be grafted to the surface of dendrimers starting either from the aldehyde or the P(S)Cl2 terminal functions, that are ‘‘naturally’’ occurring during the synthesis of PPH dendrimers (see Scheme 2). Condensation reactions of amines are generally carried out with the aldehydes, whereas phenols are used for the reactions with the P(S)Cl2 groups. Both methods have been applied for the grafting of P,X (X = N, S) ligands. Starting from the aldehyde terminal groups, the condensation with the ligand (2S)-2-amino-1-(diphenylphosphinyl)-3-methylbutane afforded the chiral dendrimer 1-G3, ended by 24 P,N ligands (Scheme 3A) [31]. A chiral ferrocenyl P,S ligand, functionalized with a phenol group, was used for the substitution reaction on P(S)Cl2 terminal groups, affording a series of chiral dendrimers from generation 1 to generation 4 (Scheme 3B) [32]. Both types of compounds 1-G3 and 2-Gcn (n = 1–4) were used as ligands for the asymmetric allylic alkylation of rac-(E)-diphenyl-2propenyl acetate. The complexation was carried out in situ by mixing the dendrimers ended by the ligands with [Pd(g3-C3H5)Cl]2, using a stoichiometry of one Pd per P,X (X = N, S) ligand. In both cases the efficiency of the dendrimers is compared with that of the corresponding monomer M (methoxy derivative). The efficiency (yield and enantiomeric excess) of the P,N dendritic ligand 1-G3 is comparable to that of the monomer. However, only the dendritic complex could be recovered and reused two times, with

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

4

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Scheme 2. Step-by-step synthesis of polyphosphorhydrazone dendrimers, starting from (A) a trifunctional core (SPCl3) or (B) a hexafunctional core (N3P3Cl6).

the same efficiency for runs 1, 2, and 3 (Fig. 2, left) [31]. The efficiency of the series of P,S dendritic ligand 2-Gcn is also comparable to that of the corresponding monomer, and identical for all generations (Fig. 2, right) [32]. No dendritic effect could be observed in this case. Only one terminal group of the dendrimers D is shown, representative of all the terminal groups; this representation will be used in all the Figures.

Another type of P,N ligand was grafted to the P(S)Cl2 groups of the first generation Gc1, affording dendrimer 3-Gc1. The monomer 3-M was also synthesized (Scheme 4) [33]. The palladium derivatives of both compounds 3-M and 3-Gc1 were obtained by reaction with PdCl2 (one Pd per P,N ligand) and were isolated. The complexes were then used as catalysts for Stille couplings with various substrates. The rate of conversion was

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

5

Fig. 1. Illustration of the dendritic effect.

monitored by 1H NMR. Comparison of the efficiency (rate of conversion and yield) of the monomeric and dendritic catalysts afforded contrasted results, as the dendrimer is the least efficient in one case (Fig. 3A), the most efficient in another case (Fig. 3C), and monomer and dendrimer have the same efficiency in a third case (Fig. 3B) [33]. Various types of diphosphine ligands were used to decorate the surface of PPH dendrimers, thanks in all cases to the reactivity of Ph2PCH2OH (obtained by reaction of Ph2P-H with formaldehyde) with NH2 groups. In a first example, a two-step process was applied, starting from aldehyde functions. In the first step, a large excess of hydrazine was used, affording primary hydrazone terminal groups. In the second step, Ph2PCH2OH reacted with all the NH2 functions, affording azadiphosphine terminal groups. This sequence of reactions has been carried out with generations 1 to 4, affording the series 4-Gn (n = 1–4) (Scheme 5) [34]. Various complexes of these dendrimers were synthesized and isolated (Pd, Pt, Rh [34] and Ru [35] derivatives). In particular the PdCl2 and Pd(OAc)2 complexes of the third generation were used for catalyzed Stille coupling, and their recovery and reuse was tested. The Pd(OAc)2 derivatives were more efficient than the PdCl2 derivatives for the recycling, as the yield remained quantitative, even after 2 re-uses (Fig. 4) [36]. Other types of diphosphine terminal groups were grafted to the dendrimers from the P(S)Cl2 terminal groups, using phenol derivatives. The diphosphino phenols were obtained by reaction of

Ph2PCH2OH with tyramine, or with L-tyrosine methyl ester. The grafting of these phenols was carried out from generation 0 to generation 3, affording respectively the series 5-Gcn (n = 0–3) and 6Gcn (n = 1–3) [37] (Scheme 6). The Pd(OAc)2 complexes of both families of dendrimers 5-Gcn and 6-Gcn were isolated (1 Pd per diphosphine pincer ligand), as well as of the corresponding monomers, and used for Heck and Sonogashira reactions. In the case of Heck couplings, a slightly positive dendritic effect was observed with the complexes of the 5-Gcn family, for the efficiency (percentage of conversion) from the monomer (81% conversion) to the first generation dendrimer (95% conversion), but a negative effect is observed with the third generation (77% conversion). The complexes of the series 6-Gcn, in which the catalytic sites should be more sterically hindered, are less efficient (Fig. 5A). The same types of dendritic diphosphino complexes were used for catalyzing Sonogashira couplings. A very slight positive dendritic effect was observed with the complexes of the 5-Gcn family, from the monomer (66% conversion) to the third generation dendrimer (72% conversion). On the contrary, with the 6-Gcn family, a detrimental influence of the third generation was observed (Fig. 5B) [37]. For another example of Heck-type couplings, we have synthesized dendrimers having as terminal groups triazatriolefinic macrocycles, which are known to be able to complex Pd(0) and Pt(0), through the 3 double bonds [38]. These macrocycles can be functionalized by a primary amine, which is suitable for a condensation

Scheme 3. Grafting of P,X (X = N or S) ligands as terminal groups of dendrimers, suitable for the complexation of palladium. (A) Condensation reaction with aldehyde terminal groups. (B) Reaction of phenol on P(S)Cl2 terminal groups.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

6

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Fig. 2. Asymmetric allylic alkylation of rac-(E)-diphenyl-2-propenyl acetate by palladium complexes of chiral dendritic ligands 1-G3 and 2-Gcn (n = 1–4) and by the corresponding monomers left: recycling experiments carried out with Pd complex of 1-G3.

reaction with the aldehyde terminal functions of the dendrimers. As the imine functions are generally easily cleavable with water, the imine functions were reduced with NaBH4, to afford stable amino groups. Such process has been applied to the G0 n and Gc0 n families, affording in particular the dendrimers 7-G0 0 and 7-Gc0 4 shown in Scheme 7 [39]. The reaction of the dendrimers ended by macrocycles (7-G0 and 7-Gc4) with a stoichiometric amount of Pd(dba)2 (dba = dibenzylidene acetone), afforded discrete complexes which could be isolated (1 Pd per macrocycle). However, when a larger amount of Pd(0) was used, palladium nanoparticles (NPs) were obtained, with a narrow size distribution centered around 3 or 4 nm, and stabilized by the dendritic structures. The discrete Pd complex of 7-G0, as well as the Pd nanoparticles stabilized by 7-G0 and by 7-Gc4 were used as catalysts in Mizoroki–Heck couplings. In the first case (discrete complexes), the reaction occurred in homogeneous conditions, in the second case (Pd NPs) the reaction occurred in heterogeneous conditions. In both cases, the catalysts could be recovered and reused (5 runs), even the small Pd complex of 7G0. The efficiency (rate of conversion and yield) of the latter remained constant, whereas an increase in the catalytic efficiency

was observed with the number of recycling experiments in the case of the Pd NPs stabilized with 7-G0. Observation of the catalysts after each catalytic experiments showed a decrease of the size of the NPs, which is presumably responsible for the increased efficiency. The same effect was observed with the Pd NPs stabilized with 7-Gc4, but with a lower intensity, presumably because the NPs entrapped inside the dendrimer were more protected in this case than in the case of the small 7-G0 (Fig. 6) [39]. It must be emphasized that many examples of dendrimers encapsulating nanoparticles have been successfully used for catalysis [40,41]. The decrease in the size of the Pd nanoparticles observed in the previous case could be related to a leaching phenomenon, which is also an important drawback of the use of transition metal for homogeneous catalysts. Indeed, such leaching increases the difficulty of purification of products, or can induce the dispersion of potentially toxic metals in the environment, and anyhow decreases the lifetime of the catalysts. It was questionable if the grafting of the catalytic entities to dendrimers could decrease the leaching phenomenon. To answer this question, the most widely used phosphine (PPh3) was grafted on the surface of the first generation dendrimer, using its phenol analog, to afford dendrimer 8-Gc1. A more

Scheme 4. Synthesis of a phenol phosphine ligand and its grafting as terminal group of a first generation dendrimer. Structure of the corresponding monomer.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

7

Fig. 3. Stille couplings with various substrates, catalyzed by the PdCl2 complexes of monomer 3-M and dendrimer 3-Gc1.

sophisticated equivalent, a thiazolyl phosphine, was also synthesized and grafted to the Gc1 dendrimer, to afford 9-Gc1, ended by 12 thiazolyl phosphines (Scheme 8). Both families were grown up to the third generation 8-Gc3 and 9-Gc3, respectively [42].

Dendrimers 8-Gc1 and 9-Gc1, as well as their corresponding monomers (PPh3 and methoxy derivative of the thiazolyl phosphine) were used for the in situ complexation of Pd(OAc)2 (1 Pd per phosphine) affording catalysts suitable for Suzuki couplings.

Scheme 5. Diphosphines grafted at the surface of dendrimers, using a two-step process from aldehyde terminal groups.

Fig. 4. Stille coupling with Pd complexes of dendrimers ended by aminodiphosphines.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

8

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Scheme 6. Synthesis of monomeric phenol diphosphines, and their grafting as terminal groups of third generation dendrimers.

When carried out in water at room temperature, no difference in the efficiency (percentage of reaction after a given time) was found for the triphenyl phosphine family, whereas the monomer was found slightly more efficient than the dendrimers for the thiazolyl phosphine family (Fig. 7A). The possibility to recycle the first generations was tested in mixture H2O/THF (2:5). The recycling efficiency was low with the Pd complex of 8-Gc1 as shown by the

decrease of the percentage of reaction, from 95% at the first run to 50% at the third run. On the contrary, the efficiency of the Pd complex of 9-Gc1 remained constant after several re-uses (percentage of reaction: 95% at run 1, 96% at run 5) (Fig. 7B). In order to account for the striking difference observed between both families, the Pd leaching in the H2O/THF mixture was measured by ICP-MS. The results given in Fig. 7 displays both a ligand effect (thiazolyl-

Fig. 5. (A) Heck reaction catalyzed by the Pd(OAc)2 complexes of dendrimers ended by diphosphines. (B) Sonogashira reaction catalyzed by the Pd(OAc)2 complexes of dendrimers ended by diphosphines.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

9

Scheme 7. Synthesis of dendrimers ended by triazatriolefinic macrocycles, usable as tridentate ligands.

Fig. 6. Efficiency of the discrete Pd(0) complex of dendrimer 7-G0, or of Pd nanoparticles obtained in the presence of dendrimers ended by triazatriolefinic macrocycles for Heck couplings.

Scheme 8. Synthesis of first generation dendrimers ended by diphenylphosphine derivatives.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

10

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

phosphine better than PPh3), and a dendritic effect (dendrimers better than monomers). In the case of 9-Gc1, which combines both positive effect, the leaching of Pd was undetectable [42]. In all the previous experiments, the recovery of the dendritic catalyst was carried out by adding a bad solvent (most generally diethylether) to the solution after the end of the reaction. This method is simple and generally powerful, as shown for instance in Fig. 7. However, another way for recycling the dendritic catalyst could be used, using magnetic nanoparticles covered by graphene. For this purpose, a very special type of dendrimer was synthesized, with one function of the core being different from all the other ones, thanks to the possibility to react a single Cl on XPCl2 (X = O, S) functions [43], even on N3P3Cl6 [44]. One equivalent of a phenol bearing a pyrene group was first grafted to the cyclotriphospha-

zene, then 5 equivalents of the phenol of triphenyl phosphine reacted with the 5 remaining Cl, affording the dendrimer 10-Gc0 (Scheme 9). Using hydroxybenzaldehyde instead of the phosphine, followed by the condensation with H2NNMe-P(S)Cl2, and then the reaction with the phenol phosphine, afforded the corresponding first generation dendrimer [45]. The phosphine groups of 10-Gc0 was used for complexing Pd(OAc)2, and the pyrene group was used to interact by p-stacking with the graphene layers surrounding the magnetic cobalt nanoparticles. The dendritic complexes interacting with the nanoparticles were then used for catalyzing the synthesis of the anti-inflammatory drug felbinac. When heating, the dendrimers dissociated from the magnetic NPs, thus the catalysis occurred in homogeneous conditions. On cooling after the completion of the

Fig. 7. Suzuki couplings using the Pd complexes of dendrimers shown in Scheme 8. (A) In water. (B) In water/THF (2:5). Pd leaching in the water/THF solution, depending on the ligand.

Scheme 9. Specific functionalization on the cyclotriphosphazene core, for the grafting of a single pyrene derivative.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

11

Fig. 8. Recycling and re-use of a specially engineered catalytic dendrimer interacting with magnetic nanoparticles, used for magnetic recovery.

catalysis, the dendrimers deposited again onto the magnetic NPs, which could be simply recovered using a magnet, and re-used. The recovery was very powerful, as shown by the quantitative yield in felbinac obtained even for the 12th run (Fig. 8). In this case, and contrarily to the case shown previously in Fig. 6, the NPs have no catalytic properties, they are used only as temporary supports of the catalytic dendrimers, for their magnetic recovery [45]. 2.2. Ruthenium complexes of dendrimers as catalysts Many ligands used for the complexation of palladium could be useful also for the complexation of ruthenium. The diphosphine pincers 4-Gn (n = 1, 3) shown in Scheme 5 were particularly studied. The corresponding Ru complexes were isolated by reaction with RuH2(PPh3)4. These complexes are efficient catalysts for the Knoevenagel condensation of malononitrile with cyclohexanone. Both the generations 1 and 3 induced a quantitative conversion, for the first and second runs. Better results were obtained with the Ru complex of the generation 3 compared to that of generation 1 for the third run, the larger dendrimer being more easily recovered (Fig. 9) [36]. In all the previous cases, the catalytic entities are linked to the surface of the dendrimer, to maximize the possibilities of interactions between catalytic sites, which might induce positive effects

on the efficiency of catalysis. However, grafting a single catalytic entity at the core of dendrimers could also modify the efficiency or selectivity of the catalysis. In order to check this possibility, a dendron (dendritic wedge) possessing an activated vinyl group at the core was first synthesized. Addition of methylhydrazine to this core, followed by the reaction with Ph2PCH2OH afforded a dendron possessing a single diphosphine at the core, surrounded by 3 layers (three generations) of branching units, namely compound 11-G3 (Scheme 10) [46]. The efficiency (conversion and diastereoisomeric ratio) of the single ruthenium complex at the core of dendron 11-G3 was compared to that of the ruthenium complexes of dendrimer 4-G3 for diastereoselective Michael additions. Of course, as in all the other catalytic experiments, the number of catalytic sites is the same: when one equivalent of dendrimer 4-G3 is used, the comparison is done with 24 equivalents of 11-G3. Both compounds have the same conversion efficiency, and no influence on diastereoisomeric ratio could be observed (Fig. 10). Both dendrimers could be recovered and re-used two times (three runs) with almost the same efficiency, and no influence of the recycling on the enantioselectivity [36]. Most of the catalytic experiments shown above were carried out in organic solvents in homogeneous conditions, as these dendrimers are not soluble in water. Water solubility can be attained

Fig. 9. Knoevenagel condensation catalyzed by Ru complexes of dendrimers ended by diphosphines.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

12

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Scheme 10. Diphosphine at the core of a dendron (dendritic wedge).

Fig. 10. Michael addition catalyzed with RuH2(PPh3)2 complexes of diphosphine at the surface of dendrimer or core of dendron.

when having positive (ammoniums) or negative (phosphonates or carboxylates) charges on the surface of the dendrimers [47]. Phosphatriazaadamantane (PTA) [48] is an interesting ligand in the perspective of obtaining water-soluble dendritic complexes, as one of the nitrogen can be easily alkylated, affording an ammonium. The terminal groups of the dendrimers have to be modified in order to become an alkylating agent for PTA. Starting from aldehyde terminal groups, their reduction with BH3SMe2 afforded alcohol terminal groups, which were transformed into chloride by reaction with SOCl2. The reaction of these chloride terminal groups with PTA gave the water soluble-dendritic phosphines 12-Gcn (n = 1–3) (Scheme 11) [49]. The reaction of dendrimers 12-Gcn and of the corresponding monomer 12-M with [Ru(p-cymene)Cl2]2 afforded water-soluble complexes, which were used in biphasic catalytic experiments.

Isomerization of allylic alcohol was carried out in immiscible mixtures of water/heptane, upon strong stirring, the catalyst being in water, and the reagents and product in heptane. A clear dendritic effect was observed on the efficiency (percentage of conversion) of catalysis when going from the monomer (38%) to the third generation complex (98%), which was mainly situated at the interface between both solvents. Recycling experiments were carried out with the first generation, with an excellent efficiency of recycling up to the fourth run (Fig. 11) [49]. In view of the increased efficiency (percentage of conversion) observed with the third generation in which the density of catalyst is certainly the highest, we tried to increase the density on the surface, with a lower number of step for the synthesis of the dendrimer. For this purpose, the number of terminal groups at a given generation was multiplied by two, using 5-hydroxy-dim-

Scheme 11. Synthesis of water-soluble dendrimers ended by PTA derivatives.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

13

Fig. 11. Isomerization of allylic alcohol catalyzed with RuCl2(p-cymene) complexes of dendrimers ended by PTA.

ethylisophthalate. Starting from 12 Cl, the reaction afforded 24 ester terminal groups. Their reduction with LiAlH4, followed by the reaction with SOCl2, and finally with PTA gave dendrimer 13Gc1 (Scheme 12). This compound has the same size than 12-Gc1, but the same number of PTA groups than 12-Gc2 [50]. The efficiency of dendrimer 13-Gc1 was compared with that of 12-Gc1 and 12-Gc2 for the hydration of alkynes, catalyzed by the corresponding ruthenium complexes. The efficiency (percentage of conversion) of the catalysts increased with an increase of the density of terminal groups, as shown in Fig. 12 [50]. 2.3. Rhodium complexes of dendrimers as catalysts All the rhodium complexes that were used were based on monophosphine derivatives [51]. In particular, the three PTA

derivatives of dendrimers used in Fig. 12 for the complexation of ruthenium were also used for the complexation of rhodium. The corresponding Rh complexes were used for catalyzing the isomerization of allylic alcohol to ketone; all dendritic catalysts were poorly efficient for this reaction as shown in Fig. 13 [50]. Despite the previous example, dendritic derivatives of rhodium complexes can be very efficient catalysts. Phosphoramidite ligands have shown excellent catalytic properties in combination with rhodium [52,53]. Grafting phosphoramidite ligands as terminal groups of dendrimers was carried out in three steps, starting from the aldehyde terminal groups of dendrimers Gc0 n. The first step was the condensation with butylamine, followed by the reduction of the imine functions with NaBH4. The last step was the reaction of the secondary amine with the chlorophosphite derived from (S)BINOL, affording dendrimers 14-Gcn (n = 1–3) (Scheme 13) [54].

Scheme 12. Synthesis of dendrimers functionalized by two PTA derivatives at each terminal group.

Fig. 12. Hydration of alkyne, catalyzed by RuCl2(p-cymene) complexes of dendrimers ended by PTA (see structures in Schemes 10 and 11). Influence of the density of catalysts on the percentage of conversion.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

14

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Fig. 13. Isomerization of allylic alcohol to ketone with rhodium complexes of PTA derivatives as terminal groups of dendrimers.

Scheme 13. Grafting of chiral phosphoramidite ligands as terminal groups of dendrimers.

The complexes obtained by reaction of the dendrimers 14-Gcn with [Rh(C2H4)2Cl]2 were used in particular for catalyzing [2+2+2] cycloaddition reactions between N-tosyl 1,6-diyne and phenylacetylene. Better yields were obtained with the dendrimer 14-Gc3[Rh(C2H4)2Cl]48 (96%) than with the corresponding monomer 14-M[Rh(C2H4)2Cl] (61%); furthermore, the dendritic complex was efficiently recovered and re-used (95% yield for run 3). The most striking results with these complexes were obtained for the catalyzed [2+2+2] cycloaddition reaction between N-tosyl 1,6-diyne and 2-methoxynaphthalene alkynyl derivatives. The yield was improved on going from the monomer (49%) to all

generations of the dendrimers (97–99%), but the most important results concerned the enantioselectivity. Indeed, the monomer had no effect on the enantioselectivity, whereas all the dendrimers induced enantiomeric excesses above 97% for all generations. These high enantiomeric excesses remained constant with recycling, at least up to run 3, for all generations (Fig. 14). This unprecedented enhancement of the stereoselectivity induced by the dendrimers might be due to the large number of chiral ligands in close proximity, as a branch model (two phosphoramidite complexes) was not more efficient than the monomer [54].

Fig. 14. Rhodium complexes of phosphoramidite ligands for catalyzed [2+2+2] cycloaddition reactions. Influence of the dendrimers on the enantiomeric excess.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

3. Dendritic complexes of elements from the first transition series (3d elements) Contrarily to ligands for elements of the second transition series, which are preferentially phosphines, ligands for elements of the first transition series are mainly based on nitrogen atoms, eventually on oxygen atoms, generally included in a double bond with carbon. 3.1. Copper complexes of dendrimers as catalysts b-Diketones are efficient and versatile ligands for main group and transition metals, as well as for lanthanides. A b-diketone functionalized with a phenol was used for reacting with the P(S)Cl2 terminal functions of dendrimers, affording the series of dendrimers 15-Gcn (n = 1–4), ended by a diketone, which is predominantly in the enol form obtained through intramolecular hydrogen transfer (Scheme 14) [55]. Dendrimers 15-Gcn were used for the in situ complexation of CuI, using either 2 or 4 diketone ligands per Cu. These complexes catalyzed O-arylations of 3,5-dimethylphenol by aryl bromides. The monomer 15-M and all the generations of the dendrimers gave the same results, with a yield in diaryl ether of about 35% with two diketones per Cu, and about 55% with four diketones per Cu. Surprisingly, attempts to recover and reuse the dendritic catalysts were unsuccessful. Analyses of all the components after the catalytic reaction demonstrated that the dendrimer was broken

15

during the catalysis experiments, in particular at the level of the P–O linkages of the surface, regenerating the monomeric ligand including a phenol group. Furthermore, this ligand can be engaged in the catalysis as a reagent (Fig. 15) [55]. The cleavage of dendrimers under catalysis is exceptional, but it should be envisaged in all cases in which there is no dendritic effect (positive or negative), especially if the reuse of the dendrimer complex is not possible. Bis(oxazolines) are privileged chiral ligands, suitable for forming complexes with a variety of metals, for many different types of catalyzes. Bis(oxazoline) ligands have been grafted in four steps onto the surface of PPH dendrimers, starting from aldehyde terminal groups. As already reported in Scheme 11, the first steps are the reduction of the aldehydes to benzyl alcohols, which then reacted with SOCl2. The substitution of the terminal benzyl chlorides with NaN3 afforded benzyl azides on the surface of the dendrimers, suitable for the famous ‘‘click’’ [56] reactions. A bis(oxazoline) derivative bearing an alkyne was the precursor for the click reaction with the azides, in the presence of CuI (5%) as catalyst. The reaction has been carried out up to the third generation, affording the series of dendrimers 16-Gcn (Scheme 15) [57]. The dendrimers 16-Gcn were used as ligands for copper(II)-catalyzed asymmetric benzoylation of diols. The maximum yield of such reaction is intrinsically 50%. One equivalent of CuCl2 per bis(oxazoline) ligand was used. Starting from a linear diol, the reaction proceeded with about 30% yield for all generations, but only 16-Gc1 and 16-Gc2 gave interesting enantiomeric excesses. All the generations could be recovered and re-used two times with

Scheme 14. Synthesis of dendrimers ended by b-diketones, mainly in the enol form.

Fig. 15. (A) Catalyzed arylations using ligands 15-Gcn complexing copper. (B) Cleavage of the dendrimer periphery and subsequent copper catalyzed arylation of the released ligands.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

16

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

Scheme 15. Dendrimers ended by bis(oxazoline) ligands, synthesized via a ‘‘click’’ reaction.

no difference in yield and enantioselectivity compared to the first run (Fig. 16A). Starting from a cyclic diol, both the yields and enantioselectivities (close to 100%) were improved (Fig. 16B) [57]. Another example of N,N ligand was obtained by reaction of a phenol functionalized by a pyridine-imine ligand, with the P(S)Cl2 terminal groups of dendrimers Gcn. The reaction was carried out with generation 1 to 3, affording the series of dendritic ligands 17-Gcn (Scheme 16) [58]. In situ complexation of CuI with the dendritic ligands 17-Gcn (n = 1–3) afforded catalysts suitable for coupling 3,5-dimethylphenol with PhI, but also for coupling pyrazole with PhI and PhBr. In the case of PhI, the CuI complex of the monomer 17-M was not efficient, whereas all the dendrimers were very efficient (95% yield)

(Fig. 17A). In the case of PhBr, an increase of the efficiency (yield) was observed on going from the monomer 17-M (inefficient) to the third generation 17-Gc3 (80% yield), illustrating a true positive dendritic effect (Fig. 17B) [58].

3.2. Scandium complexes of dendrimers as catalysts Scandium is a rare earth element which has found applications as catalyst for organic reactions and polymerizations. In view of its complexation, terpyridine ligands were grafted as terminal groups of dendrimers, from generation 1 to generation 4, affording the 18Gcn (n = 1–4) family (Scheme 17) [59].

Fig. 16. Copper(II)-catalyzed asymmetric benzoylation of diols, and recycling experiments with bis(oxazoline) dendritic ligands. (A) Linear diol as reactant. (B) Cyclic diol as reactant.

Scheme 16. Synthesis of dendrimers ended by pyridine–imine ligands.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

17

Fig. 17. Catalyzed coupling of pyrazole with PhI (A) and PhBr (B).

Scheme 17. Synthesis of dendrimers ended by terpyridine ligands.

Fig. 18. Scandium complexes of terpyridine dendritic ligands used for catalyzing Friedel–Crafts acylations under microwave irradiation. Recycling experiments carried out using different substrates at each run.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035

18

A.-M. Caminade et al. / Inorganica Chimica Acta xxx (2014) xxx–xxx

The scandium complexes of dendrimers 18-Gcn (1 Sc per terpyridine ligand) were used as catalysts for Friedel–Crafts acylations under microwave irradiation. Acylation of anisole was carried out with all the generations of the dendritic Scandium complexes. A positive dendritic effect was observed, the complex of 18-Gc4 being the most efficient. Thus only this complex was used for an unprecedented recycling experiment, carried out by using different substrates at each step, excepted at runs 4 and 12, which afford the desired product with the same efficiency on the yield, illustrating the efficiency of the recycling (Fig. 18) [59]. 4. Conclusion In this review we have illustrated the most important properties of dendritic catalysts with a special type of dendrimers, the polyphosphorhydrazone (PPH) dendrimers, ended by various complexes of the first and second transition series (3d and 4d elements). The most important property of dendritic catalysts is certainly the possibility to recover them by precipitation and to re-use them in several other catalytic run. In some cases, 12 successive runs have been carried out with no decrease in the efficiency of the dendritic catalysts. The efficiency of the recovery is due to two factors: the slightly lower solubility of the dendritic catalysts compared to the products, and a decreased leaching compared to a monomeric catalysts; the leaching can even become undetectable in some cases. Another property of dendritic catalysts is the ‘‘dendritic effect’’, which concerns an increased efficiency (rate of reaction, yield, enantioselectivity, etc.) on going from the monomer to generations 3 or 4 of the dendrimers, always considering the same number of catalytic sites. Such effect can be dramatic, as illustrated with a monomeric complex being non active whereas all the generations of the dendritic catalysts are very active. The same remark can be done concerning the enantiomeric excess. Undoubtedly, gathering several catalytic entities in close proximity in a controlled way thanks to the dendritic structure modifies the catalytic properties, in a way that cannot be attained with (disordered) polymeric catalysts [60]. A step further consists in using dendrimers as organocatalysts, without any metal [21,61]. Undoubtedly, dendrimer complexes used as catalysts afford a renewal in the field of transition metal catalysts. Thousands of papers and hundreds of patents have already been published, emphasizing the importance of this topic in the scientific community, and hopefully in specialized industries. Acknowledgement Thanks are due to the CNRS for financial support. References [1] A.M. Caminade, C.O. Turrin, R. Laurent, A. Ouali, B. Delavaux-Nicot (Eds.), Dendrimers. Towards catalytic, material and biomedical uses, John Wiley & Sons, Chichester (UK), 2011, pp. 1–528. [2] G.R. Newkome, C. Shreiner, Chem. Rev. 110 (2010) 6338. [3] V. Maraval, A.M. Caminade, J.P. Majoral, J.C. Blais, Angew. Chem., Int. Ed. 42 (2003) 1822. [4] D.A. Tomalia, A.M. Naylor, W.A. Goddard, Angew. Chem., Int. Ed. Engl. 29 (1990) 138. [5] R.G. Denkewalter, J. Kolc, W.J. Lukasavage, US. Patent 4 289 872, 1981. [6] M.L. Lartigue, B. Donnadieu, C. Galliot, A.M. Caminade, J.P. Majoral, J.P. Fayet, Macromolecules 30 (1997) 7335. [7] J. Lim, M. Kostiainen, J. Maly, V.C.P. da Costa, O. Annunziata, G.M. Pavan, E.E. Simanek, J. Am. Chem. Soc. 135 (2013) 4660. [8] T.T.T. Nguyen, M. Baumgarten, A. Rouhanipour, H.J. Rader, I. Lieberwirth, K. Mullen, J. Am. Chem. Soc. 135 (2013) 4183. [9] J.W.J. Knapen, A.W. van der Made, J.C. de Wilde, P.W.N.M. van Leeuwen, P. Wijkens, D.M. Grove, G. van Koten, Nature 372 (1994) 659. [10] E. de Jesus, J.C. Flores, Ind. Eng. Chem. Res. 47 (2008) 7968. [11] R. Kreiter, A.W. Kleij, R. Gebbink, G. van Koten, Dendrimers Iv 217 (2001) 163.

[12] D. Astruc, F. Chardac, Chem. Rev. 101 (2001) 2991. [13] R. van Heerbeek, P.C.J. Kamer, P. van Leeuwen, J.N.H. Reek, Chem. Rev. 102 (2002) 3717. [14] L.J. Twyman, A.S.H. King, I.K. Martin, Chem. Soc. Rev. 31 (2002) 69. [15] A.M. Caminade, V. Maraval, R. Laurent, J.P. Majoral, Curr. Org. Chem. 6 (2002) 739. [16] B. Helms, J.M.J. Frechet, Adv. Synth. Catal. 348 (2006) 1125. [17] D. Mery, D. Astruc, Coord. Chem. Rev. 250 (2006) 1965. [18] J.N.H. Reek, S. Arevalo, R. Van Heerbeek, P.C.J. Kamer, P. Van Leeuwen, Dendrimers in catalysis, in: Advances in Catalysis. vol. 49, 2006, Elsevier Academic Press, pp. 71–151. [19] A.M. Caminade, P. Servin, R. Laurent, J.P. Majoral, Chem. Soc. Rev. 37 (2008) 56. [20] A. John, F.M. Nachtigall, L.S. Santos, Curr. Org. Chem. 16 (2012) 1776. [21] A.M. Caminade, A. Ouali, M. Keller, J.P. Majoral, Chem. Soc. Rev. 41 (2012) 4113. [22] D. Wang, D. Astruc, Coord. Chem. Rev. 257 (2013) 2317. [23] R.S. Bagul, N. Jayaraman, Inorg. Chim. Acta 409 (2014) 34. [24] N. Launay, A.M. Caminade, R. Lahana, J.P. Majoral, Angew. Chem., Int. Ed. Engl. 33 (1994) 1589. [25] N. Launay, A.M. Caminade, J.P. Majoral, J. Organomet. Chem. 529 (1997) 51. [26] A.M. Caminade, R. Laurent, C.O. Turrin, C. Rebout, B. Delavaux-Nicot, A. Ouali, M. Zablocka, J.P. Majoral, C. R. Chim. 13 (2010) 1006. [27] A.M. Caminade, J.P. Majoral, R. Mathieu, Chem. Rev. 91 (1991) 575. [28] A.M. Caminade, R. Laurent, B. Chaudret, J.P. Majoral, Coord. Chem. Rev. 178– 180 (1998) 793. [29] A.M. Caminade, A. Ouali, R. Laurent, C.O. Turrin, J.P. Majoral, Chem. Soc. Rev. ‘‘in press’’ doi: 10.1039/C4CS00261J. [30] R. Andres, E. de Jesus, J.C. Flores, New J. Chem. 31 (2007) 1161. [31] R. Laurent, A.M. Caminade, J.P. Majoral, Tetrahedron Lett. 46 (2005) 6503. [32] L. Routaboul, S. Vincendeau, C.O. Turrin, A.M. Caminade, J.P. Majoral, J.C. Daran, E. Manoury, J. Organomet. Chem. 692 (2007) 1064. [33] M. Koprowski, R.M. Sebastian, V. Maraval, M. Zablocka, V. Cadierno, B. Donnadieu, A. Igau, A.M. Caminade, J.P. Majoral, Organometallics 21 (2002) 4680. [34] M. Bardaji, M. Kustos, A.M. Caminade, J.P. Majoral, B. Chaudret, Organometallics 16 (1997) 403. [35] M. Bardaji, A.M. Caminade, J.P. Majoral, B. Chaudret, Organometallics 16 (1997) 3489. [36] V. Maraval, R. Laurent, A.M. Caminade, J.P. Majoral, Organometallics 19 (2000) 4025. [37] P. Servin, R. Laurent, A. Romerosa, M. Peruzzini, J.P. Majoral, A.M. Caminade, Organometallics 27 (2008) 2066. [38] M. Moreno-Manas, R. Pleixats, A. Roglans, R.M. Sebastian, A. Vallribera, ARKIVOC (2004) 109. [39] E. Badetti, A.M. Caminade, J.P. Majoral, M. Moreno-Manas, R.M. Sebastian, Langmuir 24 (2008) 2090. [40] R.M. Crooks, M.Q. Zhao, L. Sun, V. Chechik, L.K. Yeung, Acc. Chem. Res. 34 (2001) 181. [41] R.W.J. Scott, O.M. Wilson, R.M. Crooks, J. Phys. Chem. B 109 (2005) 692. [42] M. Keller, A. Hameau, G. Spataro, S. Ladeira, A.M. Caminade, J.P. Majoral, A. Ouali, Green Chem. 14 (2012) 2807. [43] M. Slany, A.M. Caminade, J.P. Majoral, Tetrahedron Lett. 37 (1996) 9053. [44] O. Rolland, L. Griffe, M. Poupot, A. Maraval, A. Ouali, Y. Coppel, J.J. Fournie, G. Bacquet, C.O. Turrin, A.M. Caminade, J.P. Majoral, R. Poupot, Chem. Eur. J. 14 (2008) 4836. [45] M. Keller, V. Collière, O. Reiser, A.M. Caminade, J.P. Majoral, A. Ouali, Angew. Chem., Int. Ed. 52 (2013) 3626. [46] V. Maraval, R. Laurent, B. Donnadieu, M. Mauzac, A.M. Caminade, J.P. Majoral, J. Am. Chem. Soc. 122 (2000) 2499. [47] A.M. Caminade, J.P. Majoral, Prog. Polym. Sci. 30 (2005) 491. [48] M.Y. Darensbourg, D. Daigle, Inorg. Chem. 14 (1975) 1217. [49] P. Servin, R. Laurent, L. Gonsalvi, M. Tristany, M. Peruzzini, J.P. Majoral, A.M. Caminade, Dalton Trans. (2009) 4432. [50] P. Servin, R. Laurent, H. Dib, L. Gonsalvi, M. Peruzzini, J.P. Majoral, A.M. Caminade, Tetrahedron Lett. 53 (2012) 3876. [51] M. Slany, M. Bardaji, A.M. Caminade, B. Chaudret, J.P. Majoral, Inorg. Chem. 36 (1997) 1939. [52] S. Perreault, T. Rovis, Chem. Soc. Rev. 38 (2009) 3149. [53] J.F. Teichert, B.L. Feringa, Angew. Chem., Int. Ed. 49 (2010) 2486. [54] L. Garcia, A. Roglans, R. Laurent, J.P. Majoral, A. Pla-Quintana, A.M. Caminade, Chem. Commun. 48 (2012) 9248. [55] M. Keller, M. Ianchuk, S. Ladeira, M. Taillefer, A.M. Caminade, J.P. Majoral, A. Ouali, Eur. J. Org. Chem. (2012) 1056. [56] V.V. Rostovtsev, L.G. Green, V.V. Fokin, K.B. Sharpless, Angew. Chem., Int. Ed. 41 (2002) 2596. [57] A. Gissibl, C. Padie, M. Hager, F. Jaroschik, R. Rasappan, E. Cuevas-Yanez, C.O. Turrin, A.M. Caminade, J.P. Majoral, O. Reiser, Org. Lett. 9 (2007) 2895. [58] A. Ouali, R. Laurent, A.M. Caminade, J.P. Majoral, M. Taillefer, J. Am. Chem. Soc. 128 (2006) 15990. [59] A. Perrier, M. Keller, A.M. Caminade, J.P. Majoral, A. Ouali, Green Chem. 15 (2013) 2075. [60] C. Liang, J.M.J. Frechet, Prog. Polym. Sci. 30 (2005) 385. [61] M. Keller, A. Perrier, R. Linhardt, L. Travers, S. Wittmann, A.M. Caminade, J.P. Majoral, O. Reiser, A. Ouali, Adv. Synth. Catal. 355 (2013) 1748.

Please cite this article in press as: A.-M. Caminade et al., Inorg. Chim. Acta (2014), http://dx.doi.org/10.1016/j.ica.2014.10.035