Molecular hydrogen and water activation by transition metal frustrated Lewis pairs containing ruthenium or osmium components: catalytic hydrogenation assays

Abstract

The transition metal frustrated Lewis pair compounds [(Cym)M(κ³S,P,N-HL1)][SbF₆] (Cym = η⁶-p-MeC₆H₄iPr; H₂L1 = N-(p-tolyl)-N′-(2-diphenylphosphanoethyl)thiourea; M = Ru (5), Os (6)) have been prepared from the corresponding dimer [{(Cym)MCl}₂(μ-Cl)₂] and H₂L1 by successive chloride abstraction with NaSbF₆ and AgSbF₆ and NH deprotonation with NaHCO₃. Complexes 5 and 6 and the previously reported phosphano–guanidino compounds [(Cym)M(κ³P,N,N′-HL2)][SbF₆] [H₂L2 = N,N′-bis(p-tolyl)-N′′-(2-diphenylphosphanoethyl) guanidine; M = Ru (7), Os (8)] and pyridinyl–guanidino compounds [(Cym)M(κ³N,N′,N′′-HL3)][SbF₆] [H₂L3 = N,N′-bis(p-tolyl)-N′′-(2-pyridinylmethyl) guanidine; M = Ru (9), Os (10)] heterolytically activate H₂ in a reversible manner affording the hydrido complexes [(Cym)MH(H₂L)][SbF₆] (H₂L = H₂L1; M = Ru (11), Os (12); H₂L = H₂L2; M = Ru (13), Os (14); H₂L = H₂L3; M = Ru (15), Os (16)). DFT calculations carried out on the hydrogenation of complex 7 support an FLP mechanism for the process. Heating 9 and 10 in methanol yields the orthometalated complexes [(Cym)M(κ³N,N′,C-H₂L3_-H)][SbF₆] (M = Ru (17), Os (18)). The phosphano–guanidino complex 7 activates deuterated water in a reversible fashion, resulting in the gradual deuteration of the three cymene methyl protons through sequential C(sp³)–H bond activation. From DFT calculations, a metal–ligand cooperative reversible mechanism that involves the O–H bond activation and the formation of an intermediate methylene cyclohexenyl complex has been proposed. Complexes 5–10 catalyse the hydrogenation of the C=C double bond of styrene and a range of acrylates, the C=O bond of acetophenone and the C=N bond of N-benzylideneaniline and quinoline. The C=C double bond of methyl acrylate adds to catalyst 9, affording complex 19 in which a new ligand exhibiting a fac κ³N,N′,C coordination mode has been incorporated.

---

Introduction

There are known examples of intra- and intermolecular combinations of Lewis acids and bases that in solution do not form the corresponding adducts usually due to steric hindrance. These acid–base couples are referred to as frustrated Lewis pairs (FLP). The first examples of FLP systems were based on representative elements, boron for the acidic centre and phosphorus or nitrogen for the basic centre. A milestone in the development of such systems was the discovery in 2006 by Stephan and co-workers that such species were capable of activating reversibly and heterolytically the hydrogen molecule under mild conditions.¹ Shortly after, it was found that FLP species could activate many other small (CO₂, CO, SO₂, N₂O, NO) and organic (olefins, alkynes) molecules in a concerted and cooperative manner, following new reaction pathways.²

Much less developed, but becoming increasingly important, are FLPs in which one of the components is a transition metal fragment. They are called transition metal frustrated Lewis pairs (TMFLPs). Introducing a transition metal into the system gives the FLP, on the one hand, greater structural diversity and, on the other hand, access to fundamental reactions of transition metal compounds which are characteristic of homogeneous catalysis.³

An important feature of some FLP systems is that they can show their activity even if the classical Lewis acid–base adduct is stable, provided that the dissociated form is thermally accessible. Such a type of FLP is denominated as a masked FLP.⁴ In this regard, we have recently reported the preparation of masked TMFLPs based on the phosphano–guanidine H₂L2 and pyridinyl–guanidine H₂L3 ligands depicted in Scheme 1.⁵ These are half-sandwich (ring) M-complexes (A, B in Scheme 1) with deprotonated monoanionic HL2 or HL3 species acting as the κ³ ligand. In these monoanionic ligands, the central nitrogen atom may adopt sp³ hybridization allowing both the phosphorus (HL2) or the pyridine nitrogen (HL3) and the iminic nitrogen atom to bond with the metal in a fac κ³-coordination mode. The resulting compounds are stable saturated 18 electron species containing a strained M–N–C–N four-membered metalacycle whose inherent tension favours the thermal access to active FLP species through the breaking of one of its metal–nitrogen bonds.

Scheme 1 Phosphano–guanidine H₂L2 and pyridinyl–guanidine H₂L3 ligands and derived Rh, Ir, Ru and Os complexes A and B.

In particular, we have shown that rhodium^5a,d and iridium^5d complexes adopting the half-sandwich geometry depicted in Scheme 1 behave as masked TMFLPs activating dihydrogen and H₂O (or D₂O) in a reversible manner, causing the gradual deuteration of the Cp* groups when D₂O was activated. However, the study of the reactivity of these phosphano– and pyridinyl–guanidino complexes has sometimes been hindered by the metalation of one of the ortho carbons of the p-tolyl group, masking the processes under study and even preventing them, in some cases.

To overcome these drawbacks, we envisaged the possibility of employing the thiourea derived H₂L1 ligand shown in Scheme 2 in which one of the N(p-tolyl) groups of the guanidine moiety in H₂L2 and H₂L3 has been replaced by a sulphur atom. Formally, H₂L1 retains all the desirable characteristics to generate TMFLPs and avoids the problem of orthometalation.

Scheme 2 The phosphano–thiourea ligand H₂L1.

Herein, we report on (i) the preparation and characterization of the new complexes [(Cym)M(κ³S,P,N-HL1)][SbF₆] (M = Ru (5), Os (6)), (ii) the reaction of complexes 5 and 6, as well as that of the related ruthenium^5c and osmium^5b guanidinato compounds [(Cym)M(κ³P,N,N′-HL2)][SbF₆] (M = Ru (7), Os (8)) and [(Cym)M(κ³N,N′,N′′-HL3)][SbF₆] (M = Ru (9), Os (10)) with hydrogen, (iii) the activation of water by the phosphano–guanidinato ruthenium complex 7, (iv) DFT calculations for the activation mechanisms of dihydrogen and water and (v) the catalytic activity of complexes 5–10 in the hydrogenation of C=C, C=O or C=N bonds.

Results and discussion

Synthesis of the complexes [(Cym)M(κ³S,P,N-HL1)][SbF₆] (M = Ru (5), Os (6))

Reaction of the dimers [{(Cym)MCl}₂(μ-Cl)₂] (M = Ru and Os; Cym = η⁶-p-MeC₆H₄iPr)⁶ with N-(p-tolyl)-N′-(2-diphenylphosphanoethyl)thiourea (H₂L1)⁷ in the presence of NaSbF₆ affords [(Cym)MCl(κ²S,P-H₂L1)][SbF₆] (M = Ru (1) and Os (2)). Compounds 1 and 2 react with AgSbF₆ rendering the dicationic complexes [(Cym)M(κ³S,P,N-H₂L1)][SbF₆]₂ (M = Ru (3) and Os (4); Scheme 3). For details about the synthesis and characterization of the ligand H₂L1 and complexes 1–4, see the ESI.†

Scheme 3 Synthetic route to complexes 5 and 6.

Addition of solid NaHCO₃ to methanol solutions of complexes 3 and 4 results in partial deprotonation of the coordinated ligand H₂L1 and formation of the monocationic complexes [(Cym)M(κ³S,P,N-HL1)][SbF₆] (M = Ru, 5; Os, 6; Scheme 3).

A broad and weak band around 3340 cm⁻¹ in the IR spectrum and a singlet at ca. 7.35 ppm in the ¹H NMR spectrum indicate the presence of the remaining NH group. An IR band at around 655 cm⁻¹ is attributed to the uncoordinated SbF₆ counter-anion. At room temperature, the ³¹P{¹H} NMR spectrum consists of a singlet at 55.90 (complex 5) and that at 21.16 ppm (complex 6) is assigned to the phosphano substituent of the HL1 ligand.⁸

The crystal structure of complexes 5 and 6 has been determined by single-crystal X-ray diffraction analysis. As both compounds show similar structural parameters, we will focus our discussion on the structure of the ruthenium complex 5 (for relevant structural parameters of compound 6 see the ESI†). A hexahapto coordinated Cym ring and monodeprotonated H₂L1 acting as a fac κ³S,P,N ligand make up the coordination sphere of the ruthenium cation. The metal is a stereogenic centre and the compound crystallises as a racemate. The S_Ru enantiomer⁹ is shown in Fig. 1. As expected, bond angles within the four-membered metalacycle Ru–S(1)–C(25)–N(1) reveal the strain of this ring and, remarkably, the pyramidal arrangement around the N(1) atom (Σ°N(1) = 347.1(16)°) allows the fac coordination of the HL1 ligand.

Fig. 1 Molecular structure of the cation of complex 5. For clarity minor part of disordered fragment and all the hydrogen atoms (except that of the NH group) are omitted. Selected bond lengths (Å) and angles (°): Ru–P(1) 2.3103(13), Ru–N(1A) 2.194(14), Ru–S(1) 2.3905(15), Ru–Ct 1.6669(5), N(1A)–C(25A) 1.318(14), N(2A)–C(25A) 1.358(11); Ru–S(1)–C(25A) 84.6(3), S(1)–C(25)–N(1A) 107.0(9), S(1)–C(25A)–N(2A) 128.4(7), N(1A)–C(25A)–N(2A) 124.6(9), Ru–N(1A)–C(24A) 123.1(9), Ru–N(1A)–C(25A), 103.4(9), and C(24A)–N(1A)–C(25A) 120.6(10). Ct represents the centroid of the Cym ring.

Reaction of complexes 5, 6 and [(Cym)M(κ³-HL)][SbF₆] (HL = HL2; M = Ru (7), Os (8). HL = HL3; M = Ru (9), Os (10)) with molecular hydrogen

A common structural feature of complexes 5 and 6, as well as of those previously reported^5b,c7–10 (Scheme 4), is the presence of a strained four-membered metalacycle in the cation. We assume that the cleavage of one metal–nitrogen bond within the four-membered metalacycle, favoured by the associated strain-release, could generate a TMFLP in which the metal and the nitrogen would play the role of the acidic and basic centre, respectively. Hence, complexes 5–10 could be considered as masked TMFLP species. On these grounds, we tested the reaction of these compounds with molecular hydrogen to see if they were capable of promoting the heterolytic cleavage of the hydrogen molecule showing an FLP behaviour.

Scheme 4 Reaction with the hydrogen of complexes 5–10.

Indeed, treatment of THF-d₈ solutions of complexes 5–10 with hydrogen gas (initial pressure, 5 bar) resulted in the formation of the corresponding metal hydrido-complexes [(Cym)MH(κ²-H₂L)][SbF₆] (11–16; Scheme 4). Formally, the heterolytic cleavage of the molecule of hydrogen gives rise to hydridic M–H and protic N–H bonds. It is notable that while in the phosphano–guanidino compounds 7 and 8 and in the pyridinyl–guanidino complexes 9 and 10 the metal–imino nitrogen bond is broken, with subsequent protonation of the resulting N(p-tolyl) group, in the reaction with hydrogen of the phosphano–thiourea compounds 5 and 6, the cleavage of the metal–imidic nitrogen bond takes place and the metal–sulfur bond remains intact (Scheme 4).¹⁰

Slow reaction rates were observed. Several hours to some days of reaction are necessary to achieve complete conversion working in the 298–373 K range of temperature (see the Experimental section). In the hydrogenation of the pyridinyl–guanidino compounds 9 and 10, the concomitant formation of an orthometalated compound in 10% and 20% molar ratio, respectively, was observed. For the alternative preparation and characterisation of these orthometalated side-products (compounds 17 and 18) see the following.

Hydrogenation is reversible. Fig. 2 shows the evolution of the ³¹P{¹H} spectrum of a sample of complex 5 after treatment with molecular hydrogen and, then, after the hydrogen pressure was removed. In the absence of hydrogen, the hydrogenation product 11 spontaneously converts completely into starting compound 5, after 6 days at 363 K. Due to the reversibility of the reaction, removing of volatiles from the reaction medium mostly afforded the starting materials sometimes containing variable amounts of the corresponding hydrido compounds. For this reason, to obtain satisfactory NMR data for the hydrido compounds, spectra have to be recorded under hydrogen pressure.

Fig. 2 Reversible hydrogenation of complex 5. The ³¹P{¹H} spectrum of (a) starting complex 5; (b) after 1 day of treatment with H₂ (5 bar) at 363 K, (c) after one more day in the absence of H₂, (d) after 5 more days without H₂.

Two singlets in the 6.64–9.53 ppm region of the ¹H NMR spectrum of complexes 11–16 are indicative of the presence of two NH functionalities in the molecule. A signal in the high field region of the ¹H NMR spectrum is attributed to the hydrido ligand. For the phosphano–thiourea and phosphano–guanidino compounds 11–14 this signal appears as a doublet, due to the coupling to the phosphorus nucleus in the −8.83 to −10.30 ppm region. For the pyridinyl–guanidino complexes 15 and 16 a singlet at −5.34 and −5.29 ppm, respectively, is assigned to the metal–hydrido resonance. At room temperature, the ³¹P{¹H} spectrum of complexes 11–14 consists of one singlet confirming the presence of the PPh₂ group in the cation.

The crystal structure of complexes 13 and 14 was elucidated by X-ray diffraction analysis. Both compounds have been found to be isostructural, crystallising in the centrosymmetric P2₁/c space group with two independent molecules in their asymmetric unit. One of the independent cations of compound 13 is depicted in Fig. 3. It shows a three-legged piano stool geometry formed by the η⁶-coordinated Cym fragment, the hydrido and the H₂L2 ligand κ² P,N coordinated (an analogous representation of complex 14 is included in the ESI†). Geometrical parameters of the CN₃ guanidine fragment indicate a slight partial double character for the two C–NH(p-tolyl) bonds [N(2)–C(25) 1.364(3), N(3)–C(25) 1.367(3)] together with a comparatively shorter distance of the C(25)–N(1) bond [N(1)–C(25) 1.304(3)]. A planar geometry around the coordinated nitrogen of the phosphano–guanidine ligand [Σ°N(1) = 360(3)°] was determined.

Fig. 3 Molecular structure of one of the two crystallographically independent cations of complex 13. For clarity all the hydrogen atoms, except for the hydrido and hydrogen atoms of NH groups, together with minor component of disordered fragments are omitted. Selected bond lengths (Å) and angles (°): Ru(1)–N(1) 2.1245(17), Ru(1)–P(1) 2.2645(6), Ru(1)–H(1A) 1.608, Ru(1)–Ct(1) 1.7397(1), N(1)–C(25) 1.304(3), N(2)–C(25) 1.364(3), N(3)–C(25) 1.367(3), Ct(1)–Ru(1)–N(1) 130.36(1), Ct(1)–Ru(1)–P(1) 133.54(1), Ct(1)–Ru(1)–H(1) 120, P(1)–Ru(1)–N(1) 81.63(5), P(1)–Ru(1)–H(1A) 88.2, and N(1)–Ru(1)–H(1A) 88.4. Ct represents the centroid of the Cym ring.

DFT calculations for the hydrogenation of complex 7

The Gibbs free energy profiles for the hydrogenation of 7 were calculated by DFT methods, at the B3LYP-D3/311G(d,p)/LanL2TZ(f) level (see the ESI†). Fig. 4 shows the intermediates and transition states along with the relative Gibbs free energies for the process. Dissociation of the terminal Ru–N bond of 7 affords the true TMFLP species A through the transition state TS_7-A. Intermediate A interacts with H₂ rendering A·H₂ in which the hydrogen molecule approximates one of its hydrogen atoms to the free nitrogen (N⋯H: 2.593 Å). Subsequently, the formation of an N–H bond and coordination of the remaining hydrogen atom with the metal with the concomitant H–H bond rupture yield hydride complex 13. The activation Gibbs free energies of the direct and reverse processes (21.2 and 28.6 kcal mol⁻¹, respectively) are coherent with the slow experimental reaction rates observed and parallel to those proposed for main group FLPs^11a,b and TMFLPs.^11c The electronic rearrangement that takes place throughout the hydrogenation process is remarkable. Indeed, the CN bond distances suggest that while the C–N3 bond remains single throughout the process (Fig. 4), the C–N1 and C–N2 bonds change from single to double and from double to single, respectively going from 7 to 13.

Fig. 4 Gibbs free energy profile (kcal mol⁻¹) for the reaction with hydrogen of complex 7 [B3LYP-D3/311G(d,p)/LanL2TZ(f) level, in THF (SMD), 298 K].

Orthometalation reactions

Heating methanol solutions of [(Cym)M(κ³N,N′,N′′-HL3)][SbF₆] (M = Ru (9), Os (10)) at 338 K affords the orthometalated complexes [(Cym)M(κ³N,N′,C-H₂L3_-H)][SbF₆] (M = Ru (17), Os (18)), respectively (Scheme 5).

Scheme 5 Preparation of the orthometalated complexes 17 and 18.

As the most relevant NMR data, the ¹H NMR spectrum in CD₂Cl₂, shows two singlets at 7.09 and 6.54 ppm for complex 17 and at 7.09 and 6.49 ppm for 18 that denote the presence of two inequivalent NH groups. In addition, a singlet at 151.47 (17) and at 140.26 (18) ppm in the ¹³C{¹H} NMR spectrum is assigned to the corresponding M–C carbon resonance.

Single crystals of 17 were grown from methanol/diethyl ether/n-pentane solutions. The complex crystallises with two crystallographically independent, but chemically identical molecules in the unit cell. As only slight structural differences are found between them, description will be focused on one of them. The cation exhibits a three-legged-piano stool geometry with an η⁶ coordinated Cym ligand. The metalated pyridinyl–guanidine ligand occupies three mutually cis positions at the metal centre rendering two fused metalacycles, namely the five-membered ring Ru(1)–N(1)–C(15)–C(16)–N(2) and the six-membered ring Ru(1)–N(2)–C(17)–N(3)–C(18)–C(19). The metal centre is stereogenic and, as the molecule crystallises in the centrosymmetric space group P1̄, both enantiomers, namely S_Ru (shown in Fig. 5) and R_Ru, are present in the unit cell. In the formation of complexes 17 and 18 from 9 and 10, new metal–carbon and N–H bonds form and, hence, formally speaking, the metalation reaction involves the activation of an aromatic C(sp²)–H bond by metal–ligand cooperation. Seemingly, these structural features are closely related to the formation of a less strained six-membered ring in 17 instead of the four-membered ring present in the starting materials.

Fig. 5 Molecular structure of one of the two crystallographically independent cations of complex 17. For clarity all the hydrogen atoms are omitted, except those of NH groups. Selected bond lengths (Å) and angles (°): Ru–N(1) 2.0914(19), Ru–N(2) 2.0781(19), Ru–C(19) 2.069(2), Ru–Ct 1.7118(1), N(2)–C(17) 1.305(3), N(3)–C(17) 1.358(3), N(4)–C(17) 1.372(3), Ct–Ru–N(1) 127.57(1), Ct–Ru–N(2) 132.45(1), Ct–Ru–C(19) 129.28(1), N(1)–Ru–N(2) 76.18(7), N(1)–Ru–C(19) 90.05(8), and N(2)–Ru–C(19) 83.67(8). Ct represents the centroid of the Cym ring.

Water and C(sp³)–H bond activation by the phosphano–guanidino complex 7

The phosphano–guanidino compound [(Cym)Ru(κ³P,N,N′-HL2)][SbF₆] (7) activates the polar O–H bond of water. Although addition of excess of water to THF solutions of complex 7 does not produce significant changes in its ¹H and ³¹P NMR spectra, NMR and mass spectrometry measurements clearly indicate that complex 7 reacts with deuterated water in a reversible fashion resulting in the complete deuteration of the Me group of the Cym ligand. The ³¹P{¹H} NMR spectrum of THF-d₈/D₂O (75%/25%, v/v) solutions does not change over time but the ¹H NMR spectrum shows a gradual decrease of the singlet attributed to the Me protons of the Cym ligand together with the concurrent appearance of a broad signal at almost the same chemical shift (Fig. 6a). The concomitant variation of the mass spectra of the cation of 7 over time is shown in Fig. 6b. This whole set of data evidences the progressive deuteration of the methyl protons of the Cym ring. The deuteration process is clean, with only isotopologues of compound 7 with different degrees of deuteration being detected by NMR spectroscopy.

Fig. 6 Evolution of the ¹H NMR spectra (a) and of the mass spectra (b) of solutions of complex 7 in THF-d₈/D₂O (75%/25%, v/v).

A few examples of related H/D exchanges involving the methyl groups of the Cp* ligand in half-sandwich rhodium(III) or iridium(III) complexes have been reported.^5a,d,12 Usually, the intervention with a strong external base is necessary^12a–d but mediation of the basic component of an FLP species can also promote this exchange.^5a,d,12e However, under similar conditions, deuteration of the methyl protons of the Cp* ligand was not observed in Cp*Ru(II) homologues^12b and, as far as we know, no cases of methyl proton deuteration of Cym ligands have been described so far. Notably, the H/D exchange observed in 7 does not need the addition of an external base.

Kinetic measurements indicate that the deuteration process obeys a pseudo-first-order rate law with k_obs values from 3.88 × 10⁻⁵ to 2.32 × 10⁻⁴ s⁻¹, in the 358–373 K temperature range. The linear fitting of ln(k_obs/T) versus 1/T gives a ΔG^‡ value of 28(1) kcal mol⁻¹ at 273 K. The formation of 7-d₃ from 7 is reversible. At 358 K, a THF-d₈/H₂O (75%/25%, v/v) solution of 7-d₃ evolves to 7 with an observed pseudo-first-order rate constant of 0.65 × 10⁻⁵ s⁻¹ (see the ESI†). The high measured ratio, k_H/k_D ≈ 6, indicates that the rate-determining step for the exchange process is the C–H(D) bond cleavage.

The mechanism of this H/D exchange process was explored by DFT calculations in order to shed light on the deuteration of 7 in the presence of D₂O. Fig. 7 shows the Gibbs free energy profile together with the involved intermediates and transition states. Formation of the unsaturated TMFLP species A has been discussed above. A reacts with H₂O giving the species A·H₂O in which the incoming water molecule forms an N2⋯H1–O hydrogen bond [N2⋯H1: 1.888 Å, O–H1: 0.985 Å, N2⋯O: 2.872 Å, N2–H1–O: 176.7°]. The Ru–O distance, 3.470 Å, excludes any significant metal–hydroxo bond interaction. Intermediate A·H₂O evolves to B that contains an O–H ligand (Ru–O: 2.091 Å) and a N2–H1⋯O hydrogen bond [N2–H1: 1.040 Å, O⋯H1: 1.726 Å, N2⋯O: 2.697 Å, N2–H1–O: 153.6°]. The transition state from A·H₂O to B, TS_A-B, shows a geometry rather similar to A·H₂O, with the N2⋯H1 distance shortened to 1.795 Å. Finally, B evolves to the methylene cyclohexenyl ruthenium(II) derivative through the transition state TS_B–C. The newly formed water molecule in C is bound to the ruthenium with a rather long Ru–O bond (2.326 Å); a search in the Cambridge Structural Database (CSD)¹³ for (Cym)RuOH₂ fragments showed Ru–OH₂ bond distances between 2.118 and 2.204 Å, with a mean value of 2.146 Å. An N2–H1⋯O hydrogen bond is still present [N2–H1: 1.016 Å, O⋯H1: 1.949 Å, N2⋯O: 2.925 Å, N2–H1–O: 160.4°]. The calculated Gibbs free energy profile (Fig. 7) indicates that the formation of species C should be reversible. Therefore, the H/D exchange process under discussion should result in the progressive H/D exchange at the methyl of the Cym ligand.

Fig. 7 Gibbs free energy profile [B3LYP-D3/311G(d,p)/LanL2TZ(f) level, in THF (SMD), 298 K] for the hydrogen exchange at the methyl Cym ligand of complex 7.

In summary, complex 7 belongs to the class of thermally induced FLPs (masked FLP) and presents the advantage of being air and moisture stable. The existence of a strong ring strain within the Ru–N–C–N four-membered cycle most probably facilitates the reversible cleavage of the Ru–N(sp²) bond and makes accessible frustrated Lewis pair sites available for the reversible activation of both polar, H–O–H, and nonpolar, H₂, bonds. When D₂O was employed, the generated Ru–OD intermediate is responsible for an unprecedented H/D exchange of the methyl protons of the Cym ligand.

Catalytic hydrogenation assays

The masked FLP complexes 5–10 have been tested as catalysts in the hydrogenation of the C=C double bond of styrene and a range of acrylates, the C=O bond of acetophenone and the C=N bond of N-benzylideneaniline and quinoline. Under the standard conditions employed (5 bar hydrogen, 5 mol% catalyst loading, at 363 K in THF as the solvent) all the complexes are active catalysts in all the tested hydrogenations. However, slow hydrogenation rates were observed, several hours (even a few days) being needed to achieve complete conversion (see the ESI†). Both, the starting complex 5–10 and the corresponding hydrido compounds 11–16 were detected under catalytic conditions indicating that both the hydrogenation of the catalyst and the hydrogen transfer to the substrate are slow steps in the catalytic cycle. Side reactions also affect negatively the catalytic outcome. Thus, when complexes 9 and 10 were used as catalysts, variable amounts of the corresponding metalated complexes 17 and 18 (see the Experimental section and the ESI†) were detected. Additionally, when the hydrogenation of methyl acrylate was mediated by the pyridinyl–guanidino ruthenium compound 9, single-crystals of the complex 19 were isolated from the catalytic medium.

Fig. 8 shows a view of the cation of complex 19 including a selection of the bond lengths and angles of the complex. Along with the η⁶-Cym arene ligand, a fac κ³N,N′,C coordinated ligand completes the coordination sphere of the metal. Scheme 6 shows a plausible explanation for the formation of this trihapto ligand. Cleavage of the Ru–N(p-tolyl) bond of the masked FLP 9 affords active FLP 9A in equilibrium with its tautomer 9B whose formation involves the dearomatisation of the pyridine ring of the ligand. Reaction with methyl acrylate, accompanied by the rearomatisation of the pyridine ring, gives complex 19. The proposed path for the formation of complex 19 from 9 indicates that the C=C double bond of olefins bearing electron-withdrawing groups can be activated by complex 9 following a metal–ligand cooperative dearomatisation/aromatisation mechanism.¹⁴

Fig. 8 Molecular structure of the cation of the complex 19. For clarity all the hydrogen atoms are omitted, except for those of NH bonds. Selected bond lengths (Å) and angles (°): Ru–N(1) 2.110(5), Ru–N(2) 2.088(5), Ru–C(32) 2.203(7), Ru–Ct 1.704(3), N(2)–C(17) 1.320(8), N(3)–C(17) 1.355(8), N(4)–C(17) 1.358(8), Ct–Ru–N(1) 131.83(16), Ct–Ru–N(2) 135.43(17), Ct–Ru–C(32) 131.5(2), N(1)–Ru–N(2) 76.6(2), N(1)–Ru–C(32) 82.1(2), and N(2)–Ru–C(32) 77.5(5). Ct represents the centroid of the Cym ring.

Scheme 6 Proposed formation pathway of complex 19 from 9.

In complex 19, the ruthenium atom and the carbon atoms C(16) and C(32) (see Fig. 8) are stereogenic centres. The configuration of the C(16) carbon is predetermined by the configuration at the metal: the R configuration at the ruthenium exclusively induces R configuration at the C(16) carbon and vice versa. However, in principle, the C(32) carbon may freely adopt both R and S configurations. Hence, diastereomers of configuration R_Ru,R_C(16),R_C(32) and R_Ru,R_C(16),S_C(32) and their corresponding enantiomers S_Ru,S_C(16),S_C(32) and S_Ru,S_C(16),R_C(32) can be obtained. Indeed, the ¹H NMR spectrum of the detected compound 19 shows the presence of two diastereomers in about 55/45 molar ratio. The complex crystallises in the chiral P2₁2₁2₁ space group of the orthorhombic crystal system and only the R_Ru,R_C(16),S_C(32) isomer,⁹ depicted in Fig. 8, is present in the unit cell. Therefore, the compound crystallises as a conglomerate.¹⁵ The other diastereomer detected in solution by ¹H NMR spectroscopy has to be (R_Ru,R_C(16),R_C(32))-19.

Notably, the CO₂Me protons of the minor diastereomer are strongly shielded (δ(CO₂Me)_m, 3.21 ppm) with respect to those of the major diastereomer (δ(CO₂Me)_M, 3.88 ppm). We assume that this shielding is produced by the diamagnetic current of the pyridinyl ring of the trihapto ligand and, therefore, the configuration of the C(32) carbon in the minor diastereomer has to be S, because only in this configuration the CO₂Me group can be affected by the pyridinyl ring current. Consequently, the configuration of the most abundant isomer in solution is (R_Ru,R_C(16),R_C(32))-19.

Conclusions

Phosphano–thiourea H₂L1, phosphano–guanidino H₂L2 and pyridinyl–guanidino H₂L3 derivatives are well suited for the preparation of masked FLPs containing ruthenium and osmium fragments as the acidic component and nitrogen as the basic counterpart. The derived species 5–10 activate molecular dihydrogen following FLP pathways. The phosphano–guanidino complex 7 activates deuterated water and the resulting nucleophilic Ru–OD containing fragment is able to abstract a proton of the methyl group of the p-cymene ligand. The reversibility of the process results in the sequential H/D exchange up to complete deuteration of this methyl group. One of the p-tolyl groups of complexes 9 and 10 undergoes the metalation reaction of one of its ortho C(sp²)–H bonds rendering the saturated compounds 17 and 18. The C=C double bond of methyl acrylate diastereoselectively adds to the ruthenium complex 9, giving rise to new C–C and Ru–C bonds. Remarkably, the study of the reactivity of TMFPL species with H₂ and H₂O has allowed us to detect the activation of a variety of chemical bonds (H–H, O–H, C(sp³)–H, C(sp²)–H and C(sp²)=C(sp²)) by metal–ligand cooperation mechanisms. DFT calculations support the FLP behaviour of the complexes in both hydrogen and water activation.

Experimental

All preparations have been carried out under argon. All solvents were treated in a PS-400-6 Innovative Technologies Solvent Purification System (SPS) and degassed prior to use. Infrared spectra were recorded on a PerkinElmer Spectrum-100 (ATR mode) FT-IR spectrometer. Carbon, hydrogen and nitrogen analyses were performed using a PerkinElmer 240 B microanalyser. ¹H, ¹³C and ³¹P NMR spectra were recorded on a Bruker AV-300 spectrometer (300.13 MHz), Bruker AV-400 (400.16 MHz) or Bruker AV-500 (500.13 MHz). In both ¹H NMR and ¹³C NMR measurements the chemical shifts are expressed in ppm downfield from SiMe₄. The ³¹P NMR chemical shifts are relative to 85% H₃PO₄. J values are given in Hz. COSY, NOESY, HSQC and HMBC ¹H–X (X = ¹H, ¹³C, ³¹P) correlation spectra were obtained using standard procedures. Mass spectra were obtained with a Micro Tof-Q Bruker Daltonics spectrometer.

Preparation of the complexes [(Cym)M(κ³S,P,N-HL1)][SbF₆] (M = Ru (5), Os (6))

To a solution of the corresponding complex [(Cym)M(κ³S,P,N-H₂L1)][SbF₆]₂ (0.50 mmol) in methanol (20 mL), 42.6 mg (0.50 mmol) of solid NaHCO₃ were added. The resulting suspension was stirred for 24 h and then vacuum-evaporated to dryness. The residue was extracted with dichloromethane and the resulting solution was concentrated under reduced pressure to ca. 2 mL. The slow addition of n-pentane led to the precipitation of an orange (complex 5) or yellow (complex 6) solid, which was washed with n-pentane (3 × 5 mL) and vacuum-dried. Crystals of 5 and 6 suitable for X-ray diffraction analysis were obtained by crystallisation from CH₂Cl₂/n-pentane (5) or CH₂Cl₂/n-hexane (6) solutions.

Complex 5. Yield: 330.9 mg, 78%. Anal. calcd for C₃₂H₃₆N₂F₆PRuSSb: C, 45.3; H, 4.3; N, 3.3; S, 3.8. Found: C, 45.0; H, 4.5; N, 3.0; S, 3.9. HRMS (μ-TOF), C₃₂H₃₆N₂PRuS, [M – SbF₆]⁺, calcd: 613.1375, found: 613.1361. IR (cm⁻¹): ν(NH) 3347 (w), ν(SbF₆) 655 (s). ¹H NMR (500.10 MHz, CD₂Cl₂, RT): δ = 7.65–7.35 (m, 10H, PPh₂), 7.38 (s, 1H, NH), 7.02, 6.68 (AB system, J(A,B) = 8.4 Hz, 4H, p-tolyl), 5.66, 5.47 (AB system, J(A,B) = 6.2 Hz, 2H, H_A, H_B, Cym), 5.25 5.06 (AB system, J(A,B) = 5.9 Hz, 2H, H_A′, H_B′, Cym), 4.06 (dm, J = 40.0 Hz, 1H, NCH₂), 3.28 (m, 1H, NCH₂), 2.73 (m, 2H, PCH₂), 2.64 (sp, 1H, CH iPr), 2.27 (s, 3H, Me p-tolyl), 2.13 (s, 3H, Me Cym), 1.29 (d, J = 6.9 Hz, 3H, Me iPr), 1.26 (d, J = 6.9 Hz, 3H, Me iPr). ¹³C{¹H} NMR (125.77 MHz, CD₂Cl₂, RT): δ = 185.51 (C=S), 138.07 (d, J = 47.2 Hz), 135.14–129.51 (PPh₂), 137.08 (CMe), 133.69 (CN), 130.43 (CH₃, CH₅), 129.91 (d, J = 46.3 Hz), 123.81 (CH₂, CH₆) (p-tolyl), 111.52 (C–iPr, Cym), 102.24 (C–Me, Cym), 91.74 (d, J = 6.4 CH_A), 90.48 (d, J = 4.1 CH_B), 89.70 (d, J = 3.1 CH_B′), 88.05 (CH_A′), 53.50 (d, J = 4.8, CH₂N), 32.01 (d, J = 35.1, CH₂P), 31.12 (CH iPr), 22.55, 22.15 (Me iPr), 19.98 (Me p-tolyl), 18.18 (Me Cym). ³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, RT): δ = 55.90 (s).

³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, 173 K): δ = 66.61(bs), 61.12 (bs), 56.64 (bs).

Complex 6. Yield: 365.7 mg, 78%. Anal. calcd for C₃₂H₃₆N₂F₆OsPSSb: C, 41.0; H, 3.9; N, 3.0; S, 3.4. Found: C, 40.9; H, 3.9; N, 2.9; S, 3.6. HRMS (μ-TOF), C₃₂H₃₆N₂OsPS, [M − SbF₆]⁺, calcd: 703.1946, found: 703.1943. IR (cm⁻¹): ν(NH) 3336 (w), ν(SbF₆) 654 (s). ¹H NMR (500.10 MHz, CD₂Cl₂, RT): δ = 7.64–7.28 (m, 10H, PPh₂), 7.34 (s, 1H, NH), 7.00 (A part of an AB system, 2 H, H₃, H₅), 6.73 (B part of an AB system, J(A,B) = 8.2 Hz, 2 H, H₂, H₆, p-tolyl), 5.75 (d, J = 5.9 Hz, 1H, H_B), 5.49 (d, 1H, H_A), 5.25 (bs, 1H, H_B′), 5.17 (d, J = 5.8 Hz, 1H, H_A′), 4.08 (dm, J = 27.4 Hz, 1H), 3.17 (m, 1H (NCH₂)), 2.75, 2.60 (2 × m, 2H, PCH₂), 2.60 (m, 1H, CH iPr), 2.25 (s, 3H, Me p-tolyl), 2.23 (s, 3H, Me Cym), 1.29 (d, J = 7.0 Hz, 3H, Me iPr), 1.24 (d, J = 6.9 Hz, 3H, Me iPr). ¹³C{¹H} NMR (125.77 MHz, CD₂Cl₂, RT): δ = 191.75 (C=S), 137.58 (d, J = 55.9 Hz), 135.46–129.56, 128.77 (d, J = 53.0 Hz) (PPh₂), 137.04, 133.23, 130.39 (CH₃, CH₅), 123.43 (CH₂, CH₆) (p-tolyl), 102.72 (C–Me, Cym), 93.78 (C–iPr, Cym), 83.84 (d, J = 6.5 Hz, CH_A), 83.09 (d, J = 4.2 Hz, CH_B), 82.04 (CH_B′), 79.51 (CH_A′), 56.63 (d, J = 3.6 Hz, CH₂N), 34.35 (d, J = 40.5 Hz, CH₂P), 32.82 (CH iPr), 24.15, 23.75 (Me iPr), 21.39 (Me p-tolyl), 19.30 (Me Cym). ³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, RT): δ = 21.16 (s).

³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, 173 K): δ = 32.83 (bs), 26.84 (bs), 22.81 (bs).

Formation of the complexes [(Cym)MH(κ²-H₂L)][SbF₆] (H₂L = H₂L1, M = Ru (11), Os (12); H₂L = H₂L2, M = Ru (13), Os (14); H₂L = H₂L3, M = Ru (15), Os (16))

A high-pressure NMR tube containing a solution of the corresponding [(Cym)M(κ³-HL)][SbF₆] (5–10; 0.045 mmol) in THF-d₈ (0.45 mL), was pressurised with H₂ (5 bar) and the resulting solution was monitored by NMR spectroscopy. After several hours of reaction at temperatures ranging from 298 to 373 K (24 h at 263 K (5), 5 days at 273 K (6), 17 h (7) and 22 h (8) at 263 K, and 18 h (9) and 22 h (10) at 298 K) complete conversion was achieved in all cases. The resulting hydrido complexes [(Cym)MH(κ²-H₂L)][SbF₆] (11–16) were characterised by NMR spectroscopy, under H₂ pressure. Crystals suitable for X-ray diffraction analysis were obtained by crystallisation from THF/diethyl ether/n-pentane (complex 13) and dichloromethane (complex 14) solutions.

Complex 11. HRMS (μ-TOF), for C₃₂H₃₈N₂PRuS, [M – SbF₆]⁺, calcd: 615.1531, found: 615.1522. ¹H NMR (500.10 MHz, THF-d₈, RT): δ = 9.49 (s, 1H, p-tolylNH), 7.84–7.38 (m, 10H, PPh₂), 7.18 (t, J = 6.5 Hz, 1H, CH₂NH), 7.09, 6.49 (AB system, J = 8.1 Hz, 4H, p-tolyl), 6.01, 5.24 (AB system, J = 6.4 Hz, 2H, H_A, H_B, Cym), 5.55, 4.39 (AB system, J = 5.6 Hz, 2H, H_A′, H_B′, Cym), 4.22, 4.07 (2 × m, 2H, NCH₂), 2.94, 2.67 (2 × m, 2H, PCH₂), 2.28 (s, 3H, Me p-tolyl), 2.13 (s, 3H, Me Cym), 1.91 (sp, 1H, CH iPr), 1.28 (d, J = 6.8 Hz, 3H, Me iPr), 1.10 (d, J = 6.9 Hz, 3H, Me iPr), −9.32 (d, J = 47.2 Hz, 1H, Ru–H). ¹³C{¹H} NMR (125.77 MHz, THF-d₈, RT): δ = 179.45 (C=S), 138.12 (d, J = 50.9 Hz), 136.39 (d, J = 45.2 Hz), 133.62–128.94 (PPh₂), 138.07 (CMe), 133.90 (CN), 130.80 (CH₃, CH₅), 125.67 (CH₂, CH₆) (p-tolyl), 110.22 (C–iPr, Cym), 108.26 (C–Me, Cym), 93.71 (d, J = 5.2 CH_A), 92.53 (CH_A′), 89.42 (d, J = 6.1 CH_B), 85.45 (CH_B′), 43.69 (CH₂N), 32.14 (d, J = 31.7, CH₂P), 25.09 (CH iPr), 24.53, 22.75 (Me iPr), 20.69 (Me p-tolyl), 18.26 (Me Cym). ³¹P{¹H} NMR (202.46 MHz, THF-d₈, RT): δ = 51.53 (s).

³¹P{¹H} NMR (202.46 MHz, THF-d₈, 173 K): δ = 55.78 (bs), 52.04 (bs).

Complex 12. HRMS (μ-TOF), C₃₂H₃₈N₂OsPS, [M − SbF₆]⁺, calcd: 705.2137, found: 705.2103. ¹H NMR (500.10 MHz, THF-d₈, RT): δ = 9.53 (s, 1H, p-tolylNH), 7.81–7.33 (m, 10H, PPh₂), 7.28 (t, 1H, CH₂NH), 7.10, 6.53 (AB system J = 7.7 Hz, 4H, p-tolyl), 5.80, 5.07 (AB system, J = 6.2 Hz, 2H, H_A, H_B, Cym), 5.44, 4.27 (AB system, J = 5.2 Hz, 2H, H_A′, H_B′, Cym), 4.37, 3.94 (2 × m, 2H, NCH₂), 3.20, 2.77 (2 × m, 2H, PCH₂), 2.28 (s, 3H, Me p-tolyl), 2.26 (s, 3H, Me Cym), 1.90 (sp, 1H, CH iPr), 1.26 (d, J = 6.9 Hz, 3H, Me iPr), 1.14 (d, J = 6.9 Hz, 3H, Me iPr), −10.30 (d, J = 39.2 Hz, 1H, Os–H). ¹³C{¹H} NMR (125.77 MHz, THF-d₈, RT): δ = 177.49 (C=S), 138.65 (d, J = 55.78 Hz), 135.04 (d, J = 49.9 Hz), 134.05–128.81 (PPh₂), 138.10 (CMe), 130.83 (CH₃, CH₅), 125.60 (CH₂, CH₆) (p-tolyl), 101.52 (C–iPr, Cym), 101.15 (C–Me, Cym), 85.96 (d, J = 5.3 CH_A), 84.37 (CH_A′), 80.38 (d, J = 6.3 CH_B), 76.81 (CH_B′), 44.10 (CH₂N), 32.31 (d, J = 36.6, CH₂P), 31.52 (CH iPr), 24.58, 24.27 (Me iPr), 20.72 (Me p-tolyl), 17.67 (Me Cym). ³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, RT): δ = 5.64 (s).

³¹P{¹H} NMR (202.46 MHz, THF-d₈, 193 K): δ = 9.41 (bs), 7.00 (bs).

Complex 13. HRMS (μ-TOF), C₃₉H₄₅N₃PRu, [M − SbF₆]⁺, calcd: 688.2439, found: 688.2422. ¹H NMR (500.10 MHz, THF-d₈, RT): δ = 8.31 (s, 1H, NH trans Ru), 7.80–7.40 (m, 10H, PPh₂), 7.35 (s, 1H, NH trans CH₂), 7.12, 7.10, 6.99, 6.87 (4 × d, J = 8.4 Hz, 8H, p-tolyl), 5.40, 5.37 (AB system, J = 6.2 Hz, 2H, H_A, H_B, Cym), 5.24, 5.18 (AB system, J = 5.9 Hz, 2H, H_A′, H_B′, Cym), 3.95 (dm, J = 43.6 Hz, 1H), 3.21 (m, 1H, NCH₂), 2.97, 1.82 (2 × m, 2H, PCH₂), 2.28 (s, 6H, Me p-tolyl), 2.22 (sp, 1H, CH iPr), 2.15 (s, 3H, Me Cym), 1.26 (d, J = 6.9 Hz, 3H, Me iPr), 1.20 (d, J = 6.9 Hz, 3H, Me iPr), −8.83 (d, J = 42.9 Hz, 1H, Ru–H). ¹³C{¹H} NMR (125.77 MHz, THF-d₈): δ = 156.52 (C=N), 140.52 (d, J = 41.8), 137.77 (d, J = 58.5 Hz), 135.31–129.75 (PPh₂), 139.30, 138.60, 134.65, 133.66, 131.61, 131.29, 120.70, 120.29 (p-tolyl), 116.48 (C–iPr, Cym), 107.57 (C–Me Cym), 92.67 (CH_A′), 89.36, 88.51 (CH_A, CH_B), 84.76 (CH_B′), 57.53 (CH₂N), 34.84 (d, J = 30.2 Hz, CH₂P), 33.02 (CH iPr), 24.55, 24.51 (Me iPr), 21.35 (Me p-tolyl), 20.27 (Me Cym). ³¹P{¹H} NMR (202.46 MHz, THF-d₈. RT): δ = 68.80 (s).

Complex 14. HRMS (μ-TOF), C₃₉H₄₅N₃OsP, [M − SbF₆]⁺, calcd: 778.2960, found: 778.2964. ¹H NMR (500.10 MHz, CD₂Cl₂, RT): δ = 7.70–7.24 (m, 10H, PPh₂), 7.19 (s, 1H, NH trans CH₂), 7.09, 7.05, 6.87, 6.79 (4 × d, J = 8.0 Hz, 8H, p-tolyl), 6.65 (s, 1H, NH trans Os), 5.24 (d, J = 6.2 Hz, 1H, H_A), 5.20 (d, 1H, H_B), 4.92 (d, J = 5.9 Hz, 1H, H_A′), 4.87 (d, 1H, H_B′), 4.16 (m, J = 43.6 Hz, 1H), 3.15 (m, 1H, NCH₂), 2.99, 1.89 (2 × m, 2H, PCH₂), 2.27, 2.26 (2 × s, 6H, Me p-tolyl), 2.25 (sp, 1H, CH iPr), 2.20 (s, 3H, Me Cym), 1.25 (d, J = 6.9 Hz, 3H, Me iPr), 1.22 (d, J = 6.9 Hz, 3H, Me iPr), −9.90 (d, J = 34.6 Hz, 1H, Os–H). ¹³C{¹H} NMR (125.77 MHz, CD₂Cl₂): δ = 155.32 (C=N), 138.62 (d, J = 51.1), 135.22 (d, J = 64.3 Hz), 134.95–129.24 (PPh₂), 137.24, 136.15, 135.39, 134.19, 131.27, 130.72, 120.57, 119.80 (p-tolyl), 106.57 (C–Me Cym), 98.79 (C–iPr Cym), 83.81 (CH_A′), 81.30 (CH_A), 79.70 (CH_B), 74.17 (CH_B′), 59.52 (CH₂N), 35.38 (d, J = 33.6 Hz, CH₂P), 32.47 (CH iPr), 24.42, 24.26 (Me iPr), 21.26, 21.18 (Me p-tolyl), 19.65 (Me Cym). ³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, RT): δ = 29.89 (s).

³¹P{¹H} NMR (202.46 MHz, CD₂Cl₂, 173 K): δ = 30.63 (bs), 30.23 (bs).

Complex 15. HRMS (μ-TOF), C₃₁H₃₇N₄Ru, [M − SbF₆]⁺, calcd: 567.2056, found: 567.2059. ¹H NMR (500.10 MHz, THF-d₈, RT): δ = 8.59 (d, J = 5.7 Hz, 1H, H₆ Py), 8.42 (s, 1H, NH trans Ru), 7.79 (t, J = 7.8 Hz, 1H, H₄ Py), 7.34 (d, J = 7.9 Hz, 1H, H₃ Py), 7.28 (t, J = 6.5 Hz, 1H, H₅ Py), 7.14, 7.08 (A parts of an AB system, J(A,B) = 8.0 Hz, 4H, p-tolyl), 7.01(s, 1H, NH trans CH₂), 6.97 (B parts of an AB system, 4H, p-tolyl), 5.63 (d, J = 5.7 Hz, 1H, H_B), 5.50 (d, J = 6.1 Hz, 1H, H_A′), 5.30 (d, 2H, H_A, H_B′), 5.25 (d, J = 5.7 Hz, 1H, H_A), 4.85 (AB system, J(A,B) = 18.2 Hz, 2H, CH₂), 2.52 (sp, 1H, CH iPr), 2.29, 2.26 (2 × s, 6H, Me p-tolyl), 2.21 (s, 3H, Me Cym), 1.29 (d, J = 6.8 Hz, 3H, Me iPr), 1.26 (d, J = 6.8 Hz, 3H, Me iPr), −5.34 (s, 1H, Ru–H). ¹³C{¹H} NMR (125.77 MHz, THF-d₈, RT): δ = 162.73 (CCH₂), 155.25 (CH₆), 138.01 (CH₄), 123.94 (CH₅), 121.13 (CH₃) (Py), 137.95, 137.71, 133.12, 132.10, 130.31, 129.79, 120.06, 119.93 (p-tolyl), 110.89 (C–iPr Cym), 102.64 (C–Me Cym), 84.84 (CH_A′), 84.56, 80.56 (CH_A, CH_B′), 83.83 (CH_B), 61.39 (CH₂), 32.42 (CH, iPr), 24.05, 23.23 (Me iPr), 20.51, 20.49 (Me p-tolyl), 19.49 (Me Cym).

Complex 16. HRMS (μ-TOF), C₃₁H₃₇N₄Os, [M − SbF₆]⁺, calcd: 657.2628, found: 657.2631. ¹H NMR (500.10 MHz, THF-d₈, RT): δ = 8.81 (d, J = 5.8 Hz, 1H, H₆ Py), 8.46 (s, 1H, NH trans Os), 7.79 (t, J = 7.7 Hz, 1H, H₄ Py), 7.42 (d, J = 7.9 Hz, 1H, H₃ Py), 7.25 (t, J = 6.8 Hz, 1H, H₅ Py), 7.11, 6.95 (AB system, J(A,B) = 8.5 Hz, 4 H, p-tolyl), 7.07, 6.96 (AB system, J(A,B) = 8.5 Hz, 4 H, p-tolyl), 7.08 (s, 1H, NH trans CH₂), 5.71 (d, J = 5.5 Hz, 1H, H_B), 5.59 (d, J = 5.5 Hz, 1H, H_A′), 5.42 (d, 1H, H_B′), 5.34 (d, 1 H, H_A), 5.15, 4.78 (AB system, J(A,B) = 18.0 Hz, 2H, CH₂), 2.47 (sp, 1H, CH iPr), 2.31 (s, 3H, Me Cym), 2.28, 2.25 (2 × s, 6H, Me p-tolyl), 1.26 (d, J = 7.0 Hz, 3H, Me iPr), 1.24 (d, J = 7.0 Hz, 3H, Me iPr), −5.32 (s, 1H, Os–H). ¹³C{¹H} NMR (125.77 MHz, THF-d₈, RT): δ = 163.00 (C=N), 155.5 (CH₆), 153.75 (CCH₂), 138.11 (CH₄), 124.70 (CH₅), 120.79 (CH₃) (Py), 137.88 (CN), 137.68 (CN), 133.37 (CMe), 132.36 (CMe), 130.38, 129.88, 120.21, 120.04 (p-tolyl), 101.11 (C–iPr Cym), 94.3 (C–Me Cym), 75.99 (CH_A′), 75.50 (CH_A), 74.77 (CH_B), 71.94 (CH_B′), 62.76 (CH₂), 32.57 (CH, iPr), 24.23, 23.50 (Me iPr), 20.51, 20.50 (Me p-tolyl), 19.58 (Me Cym).

Preparation of orthometalated complexes [(Cym)M(κ³N,N′,C-H₂L3_-H)][SbF₆] (M = Ru (17), Os (18))

A solution of complex [(Cym)M(κ³N,N′,N′′-HL3)][SbF₆] (M = Ru (10), Os (11)) (0.12 mmol) in methanol (5 mL) was heated at 338 K for 8 (Ru) or 30 h (Os). The resulting solution was concentrated under reduced pressure to ca. 2 mL. The slow addition of n-pentane led to the precipitation of an orange (Ru) or yellow (Os) solid, which was washed with n-pentane (3 × 5 mL) and vacuum-dried. Crystals of complex 17 suitable for X-ray diffraction analysis were obtained by crystallization from methanol/diethyl ether/n-pentane solutions.

Complex 17. Yield: 88.4 mg, 92%. Anal. calcd for C₃₁H₃₅N₄F₆RuSb: C, 46.5; H, 4.4; N, 7.0. Found: C, 46.1; H, 4.5; N, 6.7. HRMS (μ-TOF), C₃₁H₃₅N₄Ru, [M − SbF₆]⁺, calcd: 565.1908, found: 565.1930. IR (cm⁻¹): ν(NH) 3386 (br), ν(N=C) 1641 (br), ν(SbF₆) 655 (s). ¹H NMR (500.10 MHz, CD₃OD, RT): δ = 8.87 (d, J = 5.5 Hz, 1H, H₆ Py), 7.76 (t, J = 7.8 Hz, 1H, H₄ Py), 7.71 (s, 1H, orthometalated p-tolyl), 7.34 (t, J = 6.3 Hz, 1H, H₅ Py), 7.27 (1H, H₃ Py), 7.27 (2H, p-tolyl), 7.00 (d, J = 8.4 Hz, 2H, p-tolyl), 6.69 (bs, 1H), 6.56 (d, 1H, J = 7.8 Hz, orthometalated p-tolyl), 5.78 (d, J = 5.8 Hz, 1H, H_B), 5.60 (d, J = 6.0 Hz, 1H, H_A′), 5.45 (d, 1H, H_B′), 5.18 (d, 1H, H_A), 4.79 (s, 2H, CH₂), 2.37 (s, 3H, Me p-tolyl), 2.31 (s, 3H, Me orthometalated p-tolyl), 2.31 (m, 1H, CH iPr), 1.68 (s, 3H, Me Cym), 1.07 (d, J = 6.9 Hz, 3H, Me iPr), 0.95 (d, J = 6.9 Hz, 3H, Me iPr). ¹³C{¹H} NMR (125.77 MHz, CD₃OD, RT): δ = 163.09 (CCH₂), 155.75 (CH₆), 139.87 (CH₄), 126.16 (CH₅), 121.75 (CH₃) (Py), 153.97 (C=N), 151.47 (C–Ru), 143.08, 140.44, 134.52, 126.01, 116.68 (orthometalated p-tolyl), 139.51, 139.44 (CN), 135.00, 132.11, 121.61 (p-tolyl), 111.28 (C–iPr Cym), 101.65 (C–Me Cym), 91.86 (CH_A′), 91.06 (CH_B), 84.71 (CH_B′), 83.65 (CH_A), 64.87 (CH₂), 33.08 (CH iPr), 23.82, 23.15 (Me iPr), 21.80, 21.65 (Me p-tolyl), 19.00 (Me Cym).

Complex 18. Yield: 90.3 mg, 86%. Anal. calcd for C₃₁H₃₅N₄F₆OsSb: C, 41.85; H, 4.0; N, 6.3. Found: C, 42.1; H, 4.0; N, 6.0. HRMS (μ-TOF), C₃₁H₃₅N₄Os, [M − SbF₆]⁺, calcd: 655.2472, found: 655.2500. IR (cm⁻¹): ν(NH) 3383 (br), ν(N=C) 1635 (m), ν(SbF₆) 655 (s). ¹H NMR (500.10 MHz, CD₃OD, RT): δ = 8.96 (d, J = 5.6 Hz, 1H, H₆ Py), 7.76 (t, J = 7.8 Hz, 1H, H₄ Py), 7.59 (s, 1H, orthometalated p-tolyl), 7.34 (t, J = 8.4 Hz, 1H, H₅ Py), 7.31 (d, 1H, H₃ Py), 7.27, 7.00 (2 × d, J = 8.4 Hz, 4H, p-tolyl), 6.63, 6.57 (2 × d, J = 7.9 Hz, 2H, orthometalated p-tolyl), 5.73 (d, J = 5.3 Hz, 1H, H_B), 5.67 (d, J = 5.5 Hz, 1H, H_B′), 5.59 (d, 1H, H_A′), 5.44 (d, 1H, H_A), 5.12 (d, J = 16.2 1H), 4.66 (d, 1H) (CH₂), 2.37, 2.26 (2 × s, 6H, Me p-tolyl), 2.26 (m, 1H, CH iPr), 1.79 (s, 3H, Me Cym), 1.08 (d, J = 6.9 Hz, 3H, Me iPr), 0.95 (d, J = 6.9 Hz, 3H, Me iPr). ¹³C{¹H} NMR (125.77 MHz, CD₃OD, RT): δ = 162.59 (CCH₂), 155.23 (CH₆), 140.09 (CH₄), 126.65 (CH₃), 121.52 (CH₅) (Py), 153.39 (C=N), 143.06, 140.26 (C–Os), 139.79, 135.05, 126.18, 116.23 (orthometalated p-tolyl), 139.41, 139.33 (CN), 135.19, 132.12, 121.76 (p-tolyl), 101.04 (C–iPr Cym), 93.49 (C–Me Cym), 81.42, 81.38 (CH_B, CH_A′), 76.18 (CH_B′), 74.90 (CH_A), 66.39 (CH₂), 33.14 (CH iPr), 23.96, 23.64 (Me iPr), 21.67, 21.64 (Me p-tolyl), 18.92 (Me Cym).

Preparation of [(Cym)Ru(κ³N,N′,C-H₂L3_-H)(CH₂CHCOOMe)][SbF₆] (19)

A high-pressure NMR tube containing complex 9 (13.3 mg, 0.015 mmol) and methyl acrylate (27.2 μL, 0.30 mmol) in THF-d₈ (0.45 mL) was pressurised with hydrogen gas (5 bar). The tube was heated for 10 h at 90 °C and the resulting solution was vacuum-evaporated to dryness. The slow addition of diethyl ether led to the precipitation of a brown solid which was washed with diethyl ether (3 × 1 mL) and vacuum-dried. Crystals of 19 suitable for X-ray diffraction analysis were obtained by crystallisation from THF-d₈ solutions. This compound was also prepared working under similar conditions but in the absence of hydrogen.

Complex 19. Yield: 11.9 mg, 90%. Anal. calcd for C₃₅H₄₁N₄O₂F₆RuSb·THF: C, 48.9; H, 5.1; N, 5.8. Found: C, 48.9; H, 5.0; N, 5.9. HRMS (μ-TOF), C₃₅H₄₁N₄O₂Ru, [M]⁺, calcd: 651.2268, found: 651.2259. IR (cm⁻¹): ν(NH) 3401 (br), ν(C=O) 1721, ν(C=N) 1655, ν(SbF₆) 655 (s).

R _Ru R _C(16) R _C(32)-19. ¹H NMR (500.10 MHz, CD₂Cl₂, RT): δ = 8.83 (d, J = 5.8 Hz, 1H, H₆ Py), 7.72 (t, J = 7.6 Hz, 1H, H₄ Py), 7.32 (t, J = 5.9 Hz, 1H, H₅ Py), 6.82 (1H, H₃ Py), 7.14–6.79 (m, 8 H, p-tolyl), 5.30 (d, J = 6.1 Hz, 1H, H_B), 5.26 (d, J = 6.1 Hz, 1H, H_A′), 5.13 (d, 1H, H_B′), 4.97 (d, 1H, H_A), 4.70 (d, J = 5.5 Hz, 1H, C*HN), 3.88 (s, 3H, OMe), 2.92 (sp, 1H, CH iPr), 2.42 (d, J = 9.4 Hz, 1H, C*HRu), 2.31 (s, 6H, Me p-tolyl), 2.30 (s, 3H, Me Cym), 1.96, 1.22 (2 × m, 2H, CH₂), 1.26, 1.19 (2 × d, J = 6.8 Hz, 6H, Me iPr). ¹³C{¹H} NMR (125.77 MHz, CD₂Cl₂, RT): δ = 188.31 (C=O), 164.22 (NCC*) (Py), 155.22 (C=N), 153.36(CH₆), 139.37 (CH₄), 124.20 (CH₅), 121.11 (CH₃) (Py), 135.76, 131.38, 120.86, 120.61 (p-tolyl), 106.94 (C–iPr Cym), 99.46 (C–Me Cym), 88.51 (CH_B), 88.22 (CH_A′), 83.73 (CH_A), 83.18 (CH_B′), 70.54 (C*N), 50.95 (OMe), 36.87 (CH₂), 31.03 (CH, iPr), 26.16 (C*Ru), 23.84, 21.31 (Me iPr), 21.31 (Me p-tolyl), 18.96 (Me Cym).

R _Ru R _C(16) S _C(32)-19. ¹H NMR (500.10 MHz, CD₂Cl₂, RT): δ = 8.47 (d, J = 5.6 Hz, 1H, H₆ Py), 8.33, 7.38 (2 × s, 2H, NH), 7.66 (t, J = 7.6 Hz, 1H, H₄ Py), 7.26 (t, J = 5.9 Hz, 1H, H₅ Py), 6.82 (1H, H₃ Py), 7.18–6.65 (m, 8 H, p-tolyl), 5.69 (d, J = 5.7 Hz, 1H, H_B), 5.13 (d, J = 6.0 Hz, 1H, H_A′), 5.09 (d, 1H, H_B′), 4.70 (d, 1H, H_A), 4.86 (bd, J = 3.9 Hz, 1H, C*HN), 3.98 (d, J = 6.0 Hz, 1H, C*HRu), 3.21 (s, 3H, OMe), 2.84 (sp, 1H, CH iPr), 2.40 (s, 3H, Me Cym), 2.25 (s, 6H, Me p-tolyl), 1.69 (m, 1H), 1.45 (bd, J = 13.1 Hz, 1H) (CH₂), 1.25, 1.21 (2 × d, J = 6.9 Hz, 6H, Me iPr). ¹³C{¹H} NMR (125.77 MHz, CD₂Cl₂, RT): δ = 182.11 (C=O), 163.97 (NCC*) (Py), 154.47 (C=N), 152.43 (CH₆), 138.56 (CH₄), 124.37 (CH₅), 120.98 (CH₃) (Py), 135.93, 130.92, 122.12, 120.98 (p-tolyl), 107.61 (C–iPr Cym), 103.43 (C–Me Cym), 87.55 (CH_B), 84.94 (CH_A′), 84.34 (CH_B′), 77.79 (CH_A), 71.02 (C*N), 51.30 (OMe), 35.01 (CH₂), 31.50 (CH, iPr), 27.69 (C*Ru), 23.84, 23.53 (Me iPr), 21.31 (Me p-tolyl), 18.89 (Me Cym).

General procedure for the catalytic hydrogenation reactions

A high-pressure NMR tube containing the catalyst (0.015 mmol) and the substrate to be hydrogenated (0.30 mmol) in THF-d₈ (0.45 mL) was pressurised with hydrogen gas (5 bar). The tube was heated at the appropriate temperature and the solution was monitored by NMR. Conversion values were determined by ¹H NMR.

Author contributions

S.B. and A.G.: synthetic and catalytic work. F.V. and R.R.: characterization and supervision of the work. P.G. and F.J.L.: crystallographic analysis. J.A.L.: DFT calculations. P.L. and D.C.: design and supervision of the work. P.G., F.J.L., J.A.L., P.L. and D.C.: writing of the manuscript. All authors have given their approval to the final version of the manuscript.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We thank the Ministerio de Economía y Competitividad of Spain (PID2021-122406NB-100, PID2020-119512GB-I00 and CTQ2018-095561-BI00) and Gobierno de Aragón (Grupo de Referencia: Catálisis Homogénea Enantioselectiva, E05_20R and E05_23R) for financial support. Amie Parker is acknowledged for experimental assistance.

References

1. G. C. Wel, R. R. San Juan, J. D. Masuda and D. W. Stephan, Science, 2006, 314, 1124–1126.
2. (a) Molecular Catalysis, Frustrated Lewis Pairs Vol. 2, ed. J. C. Slootweg and A. R. Jupp, Springer, Switzerland, 2021; (b) A. R. Jupp and D. W. Stephan, Trends Chem., 2019, 1, 35–48; (c) J. Paradies, Coord. Chem. Rev., 2019, 380, 170–183; (d) D. J. Scott, M. J. Fuchter and A. E. Ashley, Chem. Soc. Rev., 2017, 46, 5689–5700; (e) D. W. Stephan, Science, 2016, 354, aaf7229; (f) D. W. Stephan, J. Am. Chem. Soc., 2015, 137, 10018–10032; (g) D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2015, 54, 6400–6441; (h) D. W. Stephan, Acc. Chem. Res., 2015, 48, 306–316; (i) Topics in Current Chemistry, Frustrated Lewis Pairs II: Expanding the Scope, ed. G. Erker and D. W. Stephan, Springer, Heidelberg, 2013, vol. 334; (j) Topics in Current Chemistry, Frustrated Lewis Pairs I: Uncovering and Understanding, ed. G. Erker and D. W. Stephan, Springer, Heidelberg, 2013, vol. 332; (k) D. W. Stephan and G. Erker, Angew. Chem., Int. Ed., 2010, 49, 46–76.
3. (a) S. R. Flynn and D. F. Wass, ACS Catal., 2013, 3, 2574–2581; (b) H. B. Hamilton, A. M. King, H. A. Sparkes, N. E. Pridmore and D. F. Wass, Inorg. Chem., 2019, 58, 6399–6409; (c) A. M. Chapman, M. F. Haddow and D. F. Wass, J. Am. Chem. Soc., 2011, 133, 8826–8829; (d) Z. Jian, C. G. Daniliuc, G. Kehra and G. Erker, Organometallics, 2017, 36, 424–434; (e) A. T. Normand, C. G. Daniliuc, B. Wibbeling, G. Keh, P. Le Gendre and G. Erker, J. Am. Chem. Soc., 2015, 137, 10796–10808; (f) X. Xu, G. Kehr, C. G. Daniliuc and G. Erker, J. Am. Chem. Soc., 2013, 135, 6465–6476; (g) S. Zhang, A. M. Appel and R. M. Bullock, J. Am. Chem. Soc., 2017, 139, 7376–7387; (h) B. R. Barnett, M. L. Neville, C. E. Moore, A. L. Rheingold and J. S. Figueroa, Angew. Chem., Int. Ed., 2017, 56, 7195–7199; (i) S. Zhang, A. M. Appel and M. Bullock, J. Am. Chem. Soc., 2017, 139, 7376–7387.
4. (a) T. C. Johnstone, G. N. J. H. Wee and D. W. Stephan, Angew. Chem., Int. Ed., 2018, 57, 5881–5884; (b) S. Roters, C. Appelt, H. Westenberg, A. Hepp, J. C. Slootweg, K. Lammertsma and W. Uhl, Dalton Trans., 2012, 41, 9033–9045; (c) M. Boudjelel, E. D. S. Carrizo, S. Mallet-Ladeira, S. Massou, K. Miqueu, G. Bouhadir and D. Bourissou, ACS Catal., 2018, 8, 4459–4464; (d) M. J. Sgro and D. W. Stephan, Angew. Chem Int. Ed., 2012, 51, 11343–11345.
5. (a) M. Carmona, J. Ferrer, R. Rodríguez, V. Passarelli, F. J. Lahoz, P. García-Orduña, L. Cañadillas-Delgado and D. Carmona, Chem. – Eur. J., 2019, 25, 13665–13670; (b) A. Parker, P. Lamata, F. Viguri, R. Rodríguez, J. A. López, F. J. Lahoz, P. García-Orduña and D. Carmona, Dalton Trans., 2020, 49, 13601–13617; (c) E. T. Wilkinson, F. Viguri, R. Rodríguez, J. A. López, F. J. Lahoz, P. García-Orduña, P. Lamata and D. Carmona, Helv. Chim. Acta, 2021, 104, e2100044; (d) M. Carmona, R. Pérez, J. Ferrer, R. Rodríguez, V. Passarelli, F. J. Lahoz, P. García-Orduña and D. Carmona, Inorg. Chem., 2022, 61, 13149–13164.
6. (a) M. A. Bennet, T. N. Huang, T. W. Matheson and A. K. Smith, Inorg. Synth., 1982, 21, 75; (b) J. Cabeza and P. M. Maitlis, J. Chem. Soc., Dalton Trans., 1985, 573–578.
7. The ligand H₂L1 was prepared in situ from 2-(1-diphenylphosphano)ethylamine and p-tolylisothiocyanate following literature procedures: (a) H.-P. Deng and M. Shi, Eur. J. Org. Chem., 2012, 183–187; (b) Y.-L. Yang, C.-K. Pei and M. Shi, Org. Biomol. Chem., 2011, 9, 3349–3358; (c) H.-P. Deng, Y. Wei and M. Shi, Eur. J. Org. Chem., 2011, 1956–1960; (d) Y.-L. Shi and M. Shi, Adv. Synth. Catal., 2007, 349, 2129–2135.
8. Compounds 5 and 6 are fluxional in solution. From ³¹P{¹H} NMR measurements, an activation energy at the coalescence temperature of 9.50 ± 0.12 and 8.89 ± 0.12 Kcal mol⁻¹ for 5 and 6, respectively, has been calculated. DFT calculations reveal that these experimental values can be accounted for by assuming that the dynamic behaviour consist of rotation around the arene centroid-metal axis and about the SC–N(p-Tolyl) bond. For details see ESI.†.
9. (a) R. S. Cahn, C. Ingold and V. Prelog, Angew. Chem., Int. Ed. Engl., 1966, 5, 385–415; (b) V. Prelog and G. Helmchen, Angew. Chem., Int. Ed. Engl., 1982, 21, 567–583; (c) C. Lecomte, Y. Dusausoy, J. Protas, J. Tirouflet and A. Dormond, J. Organomet. Chem., 1974, 73, 67–76.
10. Compounds 11 and 12 are fluxional in solution. From ³¹P{¹H} NMR measurements, an activation energy at the coalescence temperature of 10.18 ± 0.12 and 10.02 ± 0.12 Kcal mol⁻¹ for 11 and 12, respectively, has been calculated. DFT calculations do not conclusively discriminate between a dynamic behaviour consisting of rotation around the arene centroid-metal axis and about the SC–N(p-tolyl) bond (as for complexes 5 and 6) or an exchange between two conformations within the seven-membered metalacycle involving the phosphano-thiourea ligand. For details see ESI.†.
11. (a) Topics in Current Chemistry, Frustrated Lewis Pairs I: Uncovering and Understanding, ed. G. Erker and D. W. Stephan, Springer, Heidelberg, 2013, vol. 332, p 157–212; (b) J. Paradies, Eur. J. Org. Chem., 2019, 283–294; (c) N. Hidalgo, J. J. Moreno, M. Pérez-Jiménez, C. Maya, J. López-Serrano and J. Campos, Chem. – Eur. J., 2020, 26, 5982–5993.
12. (a) J. W. Kang and P. M. Maitlis, J. Organomet. Chem., 1971, 30, 127–133; (b) G. Ciancaleoni, S. Bolaño, J. Bravo, M. Peruzzini, L. Gonsalvi and A. Macchioni, Dalton Trans., 2010, 39, 3366–3368; (c) S. Banerjee, J. J. Soldevila-Barreda, J. A. Wolny, C. A. Wootton, A. Habtemariam, I. Romero-Canelón, F. Chen, G. J. Clarkson, I. Prokes, L. Song, B. Peter, P. B. O'Connor, V. Schünemann and P. J. Sadler, Chem. Sci., 2018, 9, 3177–3185; (d) A. Sink, S. Banerjee, J. A. Wolny, C. Imberti, E. C. Lant, M. Walker, V. Schünemann and P. J. Sadler, Dalton Trans., 2022, 51, 16070–16081; (e) C. Ferrer, J. Ferrer, V. Passarelli, F. J. Lahoz, P. García-Orduña and D. Carmona, Organometallics, 2022, 41, 1445–1453.
13. Date of search: 20-07-2023. Version: 2022.3.0. C. R. Groom, I. J. Bruno, M. P. Lightfoot and S. C. Ward, Acta Crystallogr., Sect. B: Struct. Sci., Cryst. Eng. Mater., 2016, 72, 171–179.
14. C. Gunanathan and D. Milstein, Acc. Chem. Res., 2011, 44, 588–602.
15. E. L. Eliel, S. H. Wilen and L. H. Mander, Stereochemistry of Organic Compounds, Wiley-Interscience, New York, 1994, p. 159.

---

Footnote

† Electronic supplementary information (ESI) available: Preparation and characterization of H₂L1 and complexes 1–4. Kinetic studies for the H/D exchange of complex 7. NMR and DFT studies on the fluxionality of the complexes 5, 6, 11 and 12. NMR spectra of compounds H₂L1, 1–6 and 11–19. Crystallographic data of complexes 2, 4–6, 13, 14, 17 and 19. DFT calculations for the hydrogenation reaction and the H/D exchange process. CCDC 2240240, 2240241, 2240243–2240248. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3dt02339g

---