The Nature of the True Catalyst in Transfer Hydrogenation with Alcohol Donors Using (arene)2Ru2Cl4(II)/TsDPEN Precursor

Judith Toubiana, Liraz Medina, Yoel Sasson
2014 Modern Research in Catalysis  
The widespread precatalyst (prepared in-situ or ex-situ) (arene) RuTsDPEN advocated for highly effectual asymmetric transfer hydrogenation (ATH) reactions with 2-propanol as hydrogen donor at ambient conditions, is proven to be unstable under the strong reducing conditions prevailing in the reaction mixtures (blend of alcohol and a base such as KOH). We assert that the true catalysts are the ruthenium metal nanoclusters formed swiftly under the reducing conditions of these systems. The TsDPEN
more » ... stems. The TsDPEN ligand plays a critical role in the generation and formatting of the active catalyst including wreaking chiral properties to the so formed catalytic nanoparticles. Kinetic measurements, NMR, UV-visible spectroscopy, circular dichroism (CD) and TEM analyses corroborate this argument. * Corresponding author. J. Toubiana et al. 69 2-propanol as a hydrogen donor, with astounding TOF of 720,000 h −1 at 820 and 55,800 h −1 at 280 achieving >99% yields within minutes. With chiral ligands, these catalysts also exhibit high enantioselectivity with ee of up to 99% [6] [7] . A key milestone in the development of the contemporary transfer hydrogenation catalysis has been the contribution by Noyori and Ikariya et al. who in 1995 [8] [9] introduced the diamino ligand TsDPEN (along with amino alcohols) that when combined with a Ru(II)-arene precursor, generated in situ a catalyst that was proven active in the highly efficient asymmetric transfer hydrogenation (ATH) of ketones at room temperature (Scheme 1) [10]- [12] . This process and its equivalent, where formic acid is used as hydrogen donor, [10] were implemented in numerous synthetic procedures and were widely successfully applied on a commercial scale [13]- [17] . The contribution of Noyori et al. was not confined merely to the discovery and exploration of this particular catalytic system. He also authorized the archetype of CTH mechanisms with two novel paradigms in homogeneous catalysis by metal complexes. These are: a) The concept of "Outer Sphere" mechanism where the substrates do not directly coordinate to the metal center but act in response to interactions with the ligands only [18] and b) the idea of "Metal-Ligand bifunctional catalysis" where one of the ligands (such as primary or secondary amine) functions as a basic site that interact with the donor molecule via hydrogen bonding thus facilitating a proton transfer between the donor and the acceptor [19]- [21] . This novel mechanism was also proposed for the direct catalytic hydrogenation of ketones under hydrogen pressure [22] . Several experimental and theoretical studies supported the concerted hydrogen transfer process via the above mechanism [23] [24] . The concept of metal-ligand bifunctional catalysis was further developed by Handgraaf [25] and by Baratta [26] also for the alternative "inner sphere" mechanism where a metal alkoxide complex is a key intermediate. Noyori's outer sphere bifunctional mechanism was corroborated experimentally through isolation and identification of the presumed intermediates [27] . Starting with dichlororuthenium (p-cymene) (II) (complex 4) and TsDPEN as an auxiliary ligand, the following three complexes (1 -3) were synthesized and fully characterized (Figure 1) . The 18e complex 1 was considered as the catalyst precursor while 2 and 3 were advocated as reactive intermediates playing a major role in the ATH catalytic cycle [17] . The structure and chirality of the three complexes 1 -3 were confirmed via single crystal X-ray analysis and by 1 H NMR. Another important tool used for the identification of these species is electrospray ionization combined with a mass spectrometer (ESI-MS) [28]- [30] . A state of the art development in this field has been the ambient ionization method desorption electrospray ionization (DESI) coupled with high resolution MS [31]. This method was shown to provide a straightforward approach for intercepting reactive species in real time without prior sample preparation. It was indeed demonstrated that DESI can intercept CTH intermediates in solution on the millisecond time scale [32] . Numerous theoretical studies were also carried out to substantiate [33] [34] or to challenge [35] the concerted metal-ligand bifunctional CTH and ATH mechanisms. While practicing the standard Noyori's protocol in transfer hydrogenation of simple ketones such as acetophenone, using 2-propanol as a hydrogen donor (Scheme 1), and in replicating the preparation and characterization of the intermediates 1 -3, we were intrigued by several puzzling observations as follows: 1) The Ru(II)TsDPEN catalyst is not stable under the reaction conditions and rapidly loses activity. Only one reaction batch is typically viable [36] [37] . 2) Through the ATH process the color of the reaction solution is changing with time indicating a continuous alteration in the state of the catalyst (the substrates and products are obviously colorless). In other words, this catalytic system is not operating in steady state as should be expected in an archetypal catalytic process. 3) The preparation of the putative catalytic intermediates 1 -3, was not carried out under authentic CTH conditions (namely in the simultaneous presence of 2-propanol and KOH). We realized that upon exposure of 1 -3 to KOH dissolved in 2-propanol at ambient temperature, and these intermediates swiftly react and transform to other species. Upon inspection we came to the conclusion that the catalytic mechanism, originally proposed by Noyori and later adopted by numerous authors, cannot be correct simply since the catalyst precursor 4, the TsDPEN ligand and the intermediates 1 -3 are all unstable under the strong reducing conditions of the CTH reaction where a mixture of 2-propanol and KOH is applied. We believe that the true catalyst in Scheme 1 is ruthenium nanoclusters swiftly formed and uniquely shaped in the presence of the TsDPEN ligand under the reaction conditions. This assertion is corroborated by kinetic
doi:10.4236/mrc.2014.33010 fatcat:iuouar3txvcerfl3fsr7yfbabi