On the Dehydrocoupling of Alkenylacetylenes Mediated by Various Samarocene Complexes: A Charming Story of Metal Cooperativity Revealing a Novel Dual Metal σ-Bond Metathesis Type of Mechanism (DM|σ-BM)

Abstract

The prevailing reductive chemistry of Sm(II) has been accessed and explored mostly by the use of samarocene precursors. The highly reducing character of these congeners, along with their Lewis acidity and predominantly ionic bonding, allows for the relatively facile activation of C–H bonds, as well as peculiar transformations of unsaturated substrates (e.g., C–C couplings). Among other important C–C coupling reactions, the reaction of phenylacetylene with different mono- or bimetallic samarocene complexes affords trienediyl complexes of the type {[(C₅Me₅)₂Sm]₂(µ-η²:η²-PhC₄Ph)}. In contrast, when t-butylacetylene is used, uncoupled monomers of the type (C₅Me₅)₂Sm(C≡C–^tBu) were obtained. Although this type of reactivity may appear to be simple, the mechanism underlying these transformations is complex. This conclusion is drawn from the density functional theory (DFT) mechanistic studies presented herein. The operating mechanistic paths consist of: (i) the oxidation of each samarium center and the concomitant double reduction of the alkyne to afford a binuclear intermediate; (ii) the C–H scission of the acetylinic bond that lies in between the two metals; (iii) a dual metal σ-bond metathesis (DM|σ-SBM) process that releases H₂; and eventually (iv) the C–C coupling of the two bridged μ-alkynides to give the final bimetallic trienediyl complexes. For the latter mechanistic route, the experimentally used phenylacetylene was considered first as well as the aliphatic hex-1-yne. More interestingly, we shed light into the formation of the mono(alkynide) complex, being the final experimental product of the reaction with t-butylacetylene.

1. Introduction

The versatility of coupling reactions mediated by single-electron transfer (SET) from Sm(II) complexes is clearly illustrated by the chemistry of SmI₂ [1,2]. This coupling chemistry has been expanded by the use of (Cp*)₂Sm(THF)_n (where Cp* = C₅Me₅; n = 0–2) [3,4,5,6,7]. Interestingly, both the Lewis acidity and the ionic bonding of these Sm(II) complexes offer ligation of substrates that can undergo facile activation of C–H bonds, or partial reduction of unsaturated compounds [8]. These two elementary steps can combine to achieve dehydrocoupling of alkynes, as was initially reported by Evans in the case of phenylacetylene. This led to the characterization of a new class of dinuclear Sm(III) trienediyl complexes, {[(Cp*)₂Sm]₂(µ-η²:η²-PhC₄Ph)} [9]. These trienediyl species were obtained by four different reaction procedures: (a) the reaction of (Cp*)₂Sm[CH(SiMe₃)]₂ with HC≡CPh; (b) thermolysis of (Cp*)₂Sm(C≡CPh)(THF) at 120–145 °C; (c) the reaction of [(Cp*)₂Sm(μ-H)]₂ with HC≡CPh; and (d) the reaction of (Cp*)₂Sm with HC≡CPh. This reactivity was also extended to include aliphatic terminal alkynes, HC≡CR (R = CH₂CH₂Ph, ⁱPent, and ⁱPr), leading to similar dinuclear trienediyl complexes [7]. However, when a more bulky acetylide is used (R = ^tBu), the reaction halts at the formation of the alkynide monomers, e.g., the uncoupled dimer of the type [(Cp*)₂Sm(C≡C–^tBu)]₂, which was prepared from the use of two different precursors; [(Cp*)₂Sm^III(μ-H)]₂, and (Cp*)₂Sm^II(THF) [7]. Alkyne dehydrocoupling has also been observed with copper salts [10] and titanium complexes [11].

Around the same period of time as the Evans report, two related contributions appeared in the literature. In particular, Teuben et al. [12] reported a study on early lanthanide carbyls, e.g., (Cp*)₂LnCH(SiMe₃)₂, that promote C–C coupling of terminal alkynes (HC≡CR) to give the corresponding trienediyl bimetallic species, {[(Cp*)₂Ln]₂(µ-η²:η²-RC₄R)} (Ln = Ce, R = Me or ^tBu; Ln = La, R = Me). At the same time, Marks et al., also reported the reaction of (Cp*)₂LaCH(SiMe₃) with HC≡CR (R = ^tBu and Ph) to form the corresponding trienediyl dinuclear complexes [13]. The latter group concluded that the uncoupled dimer [(Cp*)₂La(C≡C–R)]₂ is the direct kinetic precursor of this reaction, and that neither a redox-active lanthanide ion nor a phenyl substituent on the alkyne is required for its completion. In a recent report from our group, we were able to shed light onto this type of reactivity by means of computational techniques [14]. Nevertheless, up until now, there has not been a clear picture of the alkyne coupling mechanism leading to dinuclear Sm(III) trienediyl complexes starting from low-valent samarocene (Figure 1).

Figure 1. Reaction of samarocene with phenylacetylene to afford {[(C₅Me₅)₂Sm]₂(µ-η²:η²-PhC₄–Ph)} complex and H₂.

Figure 1. Reaction of samarocene with phenylacetylene to afford {[(C₅Me₅)₂Sm]₂(µ-η²:η²-PhC₄–Ph)} complex and H₂.

The bimetallic mechanism which was previously proposed in the case of N-anchored tris(aryloxide) uranium(III) can serve as a base [15], but how acetylenic C–H bonds break to release H₂ still needs to be answered. Hence, in this paper we attempt to understand the performance of samarocene in the formation of the trienediyl dinuclear complex, {[(Cp*)₂Sm]₂(µ-η²:η²-PhC₄–Ph)}, with plausible energetic reaction profiles being computed and suggested at the DFT (B3PW91) level of theory. The formation of the uncoupled complexes, instead of the trienediyl one where ^tBuC≡CH was used, is also discussed. Based on these studies, novel transition states which lead to unique reaction pathways in the domain of f-block chemistry are revealed.

2. Results and Discussion

2.1. Dehydrocoupling of PhC≡CH Using Cp*₂Sm

Based on chemical intuition and previous knowledge on related mechanisms [15], three different types of reaction sequence are envisioned for this particular reactivity (Figure 2).

Figure 2. Possible reaction paths A, B, and C leading to the formation of the corresponding trienediyl complex using PhC≡CH as the substrate and (Cp*)₂Sm as a precursor.

Figure 2. Possible reaction paths A, B, and C leading to the formation of the corresponding trienediyl complex using PhC≡CH as the substrate and (Cp*)₂Sm as a precursor.

All these sequences share a common initiation step, the double SET one, in which a doubly reduced phenylacetylene unit contained in a dinuclear samarium(III) complex is formed. From this “key intermediate” the reaction can proceed either through: (A) a direct insertion of PhC≡CH that leads to an alkynyl bis-Sm(III) complex; (B) a peculiar direct H–H coupling type of Transition State (TS); (C) a C–H activation leading to a μ-H:μ-C motif. Energetically the most favorable pathway relies on the C–H activation that corresponds to mechanism C and will be discussed in detail afterwards, together with the other routes. The whole energy profile for mechanism C is depicted in Figure 3.

Figure 3. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the corresponding trienediyl complex using PhC≡CH as substrate and (Cp*)₂Sm as a precursor (mechanism C). Cp* ligands are omitted for clarity. Numbers in blue correspond to natural charges.

Figure 3. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the corresponding trienediyl complex using PhC≡CH as substrate and (Cp*)₂Sm as a precursor (mechanism C). Cp* ligands are omitted for clarity. Numbers in blue correspond to natural charges.

As already reported by us in a recent theoretical work, the one electron reduction of phenylacetylene, induced by the coordination to the samarium center, is a highly exothermic process [16]. This is the outcome of high level ab-initio calculations (CAS-SCF), but it is also based on the newly introduced theoretical methodology that takes into account the “HOMO-LUMO gap” energy difference. As the overall redox process is known to be exothermic, it contributes to the total energy stabilization of the newly formed “key intermediate” [17]. Hence, the latter species will be the reference point for the whole mechanism. In the 1-Ph intermediate, the α- and β-carbons of the alkylenic substrate bind to each distinct Sm center in an E-configuration, with the Z-isomer being less stable by 3.6 kcal·mol⁻¹ in accordance with earlier structural interpretation [18]. Interestingly, in 1-Ph, the C–C bond distance of the sandwiched alkyne is 1.36 Å which is 0.15 Å longer than in free phenylacetylene. This is a strong indication of the double reduction of the alkyne. In addition, the acetylenic C–H bond points towards one Sm center and is elongated by 0.06 Å with respect to free phenylacetylene, being readily available for a potential bond scission. At that point, the subsequent C–H bond activation (TS₁₂-Ph) yielding a μ-hydride-µ-alkynyl dinuclear complex (2-Ph) can take place with a moderate energy barrier of 17.5 kcal·mol⁻¹ with respect to the “key intermediate”. Hoffmann et al., showed in a seminal work that extra stabilization is achieved by the mixing of the 1s orbital of the hydride with the available empty d_z² and d_yz orbitals, of 1a₁ and 1b₂ symmetry respectively, of each Sm(Cp*)₂ fragment [19]. This can nicely explain the relative low energy barrier computed for this step. It is noteworthy that this bridged dimer shares features with geometrically close related yttrium complexes that were previously reported [20,21]. The reaction proceeds via TS₂₃-Ph that corresponds to the activation of the acetylenic proton of a second PhC≡CH molecule by the bridging hydride of complex 2-Ph. This step is better described as a six-member σ-bond metathesis showing a negligible activation energy barrier. The alternation of the natural charge signs between the six atoms in TS₂₃-Ph (See Figure 3) and the respective shortening and lengthening of the bonds involved, confirms its σ-bond metathesis character. This is in line with previous studies [22,23,24]. To the best of our knowledge, this Dual Metal σ-Bond Metathesis transition state (DM|σ-BM) is the first report of such a type that implies two lanthanide centers working in concert. This type of transition state can be also found in other related lanthanide types of reactivity [25]. The latter step affords a μ-alkynyl-η¹-alkynyl dihydrogen complex (3-Ph) that is almost isoenergetic with the “key intermediate”. In the subsequent step an isomerization takes place which induces the release of H₂, and leads to the formation of bis-µ-phenylacetylenyl complex, 4-Ph. The formation of the latter requires a very low activation barrier (Δ_rH^# = 3.2 kcal·mol⁻¹). From 4-Ph, the final trienediyl dinuclear complex 5-Ph is obtained by the C–C homocoupling of the two terminal µ-alkynyl moieties via TS₄₅-Ph. Interestingly, similar transition states were postulated to proceed in analogous systems, as for instance the bis-μ-alkynyl titanocene complexes [26]. In addition, analogous structures were structurally isolated and characterized for other lanthanide centers [12,27,28,29]. As already reported from our group, the energy barrier for the C–C coupling step is surprisingly low, being only 10.7 kcal·mol⁻¹ [14]. This is odd since two negatively charged moieties have to be homocoupled. One way to accomplish this is to reduce the electron density of the negatively charged α-carbons of the alkynyls. This is achieved by the nucleophilic assistance from the β-carbon of each triple bond which accumulates enough negative charge, resulting in a stronger interaction with the second samarium center. Finally, the rate determining step of the entire mechanism is the scission of the C–H acetylenic bond (17.5 kcal·mol⁻¹), with the overall exothermicity of the mechanism being 24.7 kcal·mol⁻¹ with respect to 1-Ph intermediate.

The other two pathways depicted in Figure 1 were considered starting from 1-Ph. In the first case (mechanism A), and inspired by our recent mechanistic work [15], the possibility of a direct C–C coupling in 1,1-fashion was investigated. This led to the corresponding Sm(III) bis-vinyl complex 6-Ph (Figure 4). Although this step is strongly exothermic (−31.0 kcal·mol⁻¹), the activation barrier (Δ_rH^‡ = 24.8 kcal·mol⁻¹) is much higher than the highest barrier of mechanism C, by more than 7 kcal·mol⁻¹. The reaction involves an ionic transition state, TS₁₆-Ph, in which the incoming triple bond is polarized by one samarium center, allowing the formation of a lone-pair at the β-carbon atom and an empty sp²-orbital at its α-carbon that overlaps with the filled sp²-orbital of the α-carbon of the bridged reduced triple bond. This transition state closely resembles the one reported for the C–C coupling of the hex-1-yne using N-anchored tris-aryloxide complexes of U(III) [15]. Then the reaction proceeds through the activation of the C–H bond in close vicinity to the Sm center, surmounting an even higher activation barrier than the previous step (Δ_rH^‡ = 25.7 kcal·mol⁻¹). The geometry of the intermediate that results, 7-Ph, corresponds to a weakly coordinated complex, in which two fragments develop an H–H interaction (Figure 4). The latter undergoes a proton transfer to give the trienediyl compound upon release of H₂, via a transition state of relatively low activation barrier (Δ_rH^‡ = 6.9 kcal·mol⁻¹). However, this mechanism can be ruled out from further consideration since it suffers kinetically from two relatively high activation barriers of ca. 24 kcal·mol⁻¹, and thermodynamically by the formation of the highly stable bis-vinyl complex 6-Ph. The latter corresponds to the lowest energy point of the mechanism, being 6.3 kcal·mol⁻¹ lower than the final trienediyl complex, as depicted in Figure 4.

Figure 4. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the corresponding trienediyl complex using PhC≡CH as substrate and (Cp*)₂Sm as a precursor (mechanism A). Cp* ligands are omitted for clarity.

Figure 4. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the corresponding trienediyl complex using PhC≡CH as substrate and (Cp*)₂Sm as a precursor (mechanism A). Cp* ligands are omitted for clarity.

Finally, from “key intermediate” 1-Ph, a mechanism that eliminates H₂ and homocouples the two alkynyls in a concerted homolytic manner (mechanism B) was attempted. Despite our efforts, we were unable to locate any transition state of this type. Instead, we found the direct formation of H₂, but without any C–C coupling, leading to 5-Ph after reorganization (Figure 5, mechanism B). By examining the nature of each hydrogen at the transition state, this can be regarded as an H–H coupling type of TS. Interestingly, the corresponding natural charges of the two hydrogens are −0.01 and 0.10 |e| (Figure 5). The first charge corresponds to the acetylenic bridged hydrogen, while the second belongs to the newly inserted alkyne. Nevertheless, the activation energy barrier for this step (Δ_rH^‡ = 27.2 kcal·mol⁻¹) allows us to discard this pathway from further consideration.

Figure 5. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the corresponding trienediyl complex using PhC≡CH as substrate and (Cp*)₂Sm as a precursor (Mechanism B). Cp* ligands are omitted for clarity. Numbers in blue correspond to natural charges.

Figure 5. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the corresponding trienediyl complex using PhC≡CH as substrate and (Cp*)₂Sm as a precursor (Mechanism B). Cp* ligands are omitted for clarity. Numbers in blue correspond to natural charges.

In contrast, when a monometallic mechanism is operating, the C–C coupling can occur in 1,1 or 1,2 fashion (Figure S1 in supplementary information). In particular, addition of a second alkyne molecule to the (Cp*)₂Sm^III(HC≡CPh) intermediate can take place in two different ways; in 1,1 and 1,2 fashion, affording the corresponding bis-vinyl species. The activation barriers were found to be 9.8 and 5.9 kcal·mol⁻¹ respectively, with the exothermicities of each one being 31.1 and 27.7 kcal·mol⁻¹ with respect to the monoreduced intermediate, (C₅Me₅)₂Sm(HC≡CPh). The computed small energy difference between the two transition states, as well as the preference for 1,2 C–C coupling, cannot account for the observed experimental selectivity which is found to be the reverse. On top of that, another weak point of the monometallic mechanism is the fact that after the formation of bis-vinyl monomeric complex, an oxidation of a second Sm^II(Cp*)₂ has to take place. This will result in the formation of the same energetically buried bis-vinyl intermediate, 6-Ph, as described previously in mechanism A. Since the formation of the latter is undesired, for the reasons described above, the existence of a monometallic pathway can also be ruled out.

2.2. Dehydrocoupling of ^tBuC≡CH Using Cp*₂Sm

We then turned our attention to include the more bulky H–C≡C–^tBu alkyne that is found to give monomeric acetylide complexes rather than bimetallic trienediyl complexes [7], with the mechanism C applied to the new alkyne (Figure 6).

Figure 6. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the uncoupled alkynide complex using ^tBuC≡CH as substrate and (Cp*)₂Sm as a precursor. Cp* ligands are omitted for clarity.

Figure 6. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the uncoupled alkynide complex using ^tBuC≡CH as substrate and (Cp*)₂Sm as a precursor. Cp* ligands are omitted for clarity.

Unlike the previous case, the first step corresponds to SET but with a concomitant dissociation of a THF molecule; the latter molecule being coordinated to the precursor complex Sm(Cp*)₂. It is expected that the high energy gain observed in this step will be sufficient for the easy dissociation of the THF molecule. The latter is computed to cost around 10 kcal·mol⁻¹ as it is depicted in Figure 7a. In addition, the isodesmic reaction of the net exchange of THF with the t-butylacetylene costs only 7.3 kcal·mol⁻¹, indicative of the low amount of energy needed for such a reaction (Figure 7b). It should be noted that all the attempts to locate a local minimum that corresponds to the “key intermediate” with at least one THF coordinated to a samarium center were not successful, due to steric reasons. Hence, most likely the THF dissociates to yield the more energetically stable bimetallic intermediate 1-^tBu.

Figure 7. (a) Dissociation reaction of THF from the mono-solvated samarocene complex; (b) Exchange reaction of mono-solvated (THF) samarocene complex with ^tBuC≡CH.

Figure 7. (a) Dissociation reaction of THF from the mono-solvated samarocene complex; (b) Exchange reaction of mono-solvated (THF) samarocene complex with ^tBuC≡CH.

Therefore, as in the phenylacetylene case, the “key intermediate” 1-^tBu is considered as the starting point of the whole enthalpy profile. It is noteworthy that the first two steps are found to possess similar energy variations with that of phenylacetylene. This was not expected initially since the bulkier ^tBuC≡CH will introduce more steric repulsions when it is sandwiched between the two samarocene moieties. This was especially anticipated in the C–H activation step, being in very close proximity to the two pentamethylcyclopentadienyl groups. Nevertheless, the inherent flexibility of the system probably allows the minimization of such repulsive forces by directing the methyl groups of the ^tBu to lie in an optimal position among the four Cp* ligands. The only, but major differentiation with respect to the phenylacetylene case is the geometry of the product of the DM|σ-BM step. The outcome of this process, computationally, is the disruption of the bimetallic complex into two monomeric alkynyl complexes, with one bearing a weakly coordinated dihydrogen molecule. Hence, instead of the generation of a μ-alkynyl-η¹-alkynyl dihydrogen-like complex, as is found in the PhC≡CH case, e.g., 3-Ph, the formation of the two monomers leads to a substantial energy stabilization (being 10.3 kcal·mol⁻¹ lower than the reference point). Then the system releases H₂ very easily to gain an extra energy stabilization of 1.4 kcal·mol⁻¹ and to yield consequently the well separated alkynyl complexes 3′-^tBu. The latter were experimentally observed as the only product of such reactivity [7]. Following this, the reaction can potentially proceed into the subsequent C–C coupling requiring a considerable amount of energy (Δ_rH^‡ = 27.7 kcal·mol⁻¹), kinetically hardly accessible [14]. Overall, the formation of the separated alkynyl species corresponds to the lowest point of the whole reaction route, being in perfect agreement with the experimental observations.

2.3. Dehydrocoupling of Hex-1-yne Using Cp*₂Sm

To understand the effect induced by an alkyl substituent on the alkyne, we considered computationally the aliphatic hex-1-yne in order to check how this differentiates from the bulky ^tBuC≡CH, and the PhC≡CH in which there is an extended conjugation of the triple bond with the phenyl ring. Even though the gain in energy is large for the double reduction of the hex-1-yne, as it was computed in a previous work by our group [16], it is half of that obtained for the phenylacetylene, possibly due to the delocalization of the extra electron(s) into the phenyl ring. In Figure 8, the most favorable reaction pathway is presented, reverting again to mechanism C.

Figure 8. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the trienediyl complex using ⁿBuC≡CH as substrate and (Cp*)₂Sm as a precursor. Cp* ligands are omitted for clarity. Numbers in blue correspond to natural charges.

Figure 8. Enthalpy energy ΔH (kcal·mol⁻¹) profile for the formation of the trienediyl complex using ⁿBuC≡CH as substrate and (Cp*)₂Sm as a precursor. Cp* ligands are omitted for clarity. Numbers in blue correspond to natural charges.

The geometry of the “key intermediate”, 1-ⁿBu, is similar to that found in the two previously described cases with the α- and β-carbons of the doubly reduced triple bond adopting a zig-zag geometry. In the same way, the acetylenic C–H bond is elongated by 0.08 Å (with respect to the free hex-1-yne) being already prepared for the forthcoming intramolecular C–H bond scission. Next, the aforementioned C–H bond activation takes place yielding a µ-hydride-µ-alkynyl dinuclear intermediate with an enthalpy activation barrier of 14.4 kcal·mol⁻¹. This activation barrier is lower than the two previous by almost 4 kcal·mol⁻¹. This is probably attributed to the reduced steric hindrance that substituents as ⁿBu-, and Ph- to a smaller extent, are inducing compared to the ^tBu- one. The reaction can further proceed with the addition of a second hex-1-yne equivalent to form the bis-µ-alkynyl complex, 4-ⁿBu, after release of H₂. This is achieved by passing through the TS₂₃-ⁿBu, which corresponds to a barrierless process. The last step corresponds to the homo C–C coupling of the two µ-alkynyl moieties to afford the final trienediyl dinuclear complex. The activation barrier of this step (Δ_rH^‡ = 16.9 kcal·mol⁻¹), is the highest of the entire mechanism, being consequently the rate determining step, contrary to the previous cases. Finally, this process is exothermic by 16.4 kcal·mol⁻¹, while the overall reaction exothermicity is found to be 23.6 kcal·mol⁻¹. Therefore, we conclude that the tested alkyne can be possibly used experimentally to afford, upon reaction with the Cp*₂Sm complex, the corresponding trienediyl species. Concluding, apart from some small stabilization observed at the C–H bond activation transition state and the formation of the uncoupled dimeric species, 4-ⁿBu, with respect to the other two cases, the ⁿBu- substituent does not play any decisive role in the feasibility of such reactivity.

2.4. Dehydrocoupling of Various Terminal Alkynes Using [Cp*₂Sm(μ-H)₂]₂ Complex

As described in the introduction, two experiments were conducted by Evans et al., for the formation of the potential trienediyl complexes using ^tBuC≡CH. The first was already investigated in detail in Section 2.2. In the second experiment, the bis-hydride dimer of Sm(III) was used, but as in the first case, it did not lead to the expected homocoupled product but instead to the formation of the uncoupled alkynide complexes as major product. It should also be noted that when PhC≡CH is considered using the same precursor, [Cp*₂Sm(μ-H)₂]₂, it yields the corresponding trienediyl complex. Therefore, a mechanistic investigation at the B3PW91 level of theory was carried out in order to discover the reason for this experimentally observed discrepancy and to propose a rational mechanism. As a reasonable mechanistic scenario we envisioned two consecutive (DM|σ-BM–H₂ release) steps (Figure 9a). Interestingly, for the phenylacetelyne case, the first DM|σ-BM type of transition state surmounts a relative low activation barrier (Δ_rH^‡ = 10.5 kcal·mol⁻¹). The intermediate 2-Ph formed after the H₂ release (the isomerization step) is energetically stable enough, being 7.9 kcal·mol⁻¹ lower than the starting material. The following part of the mechanism is essentially the same as proposed in mechanism C and consequently will not be discussed further. Finally, it is worth noting that a potential equilibrium between the dimer 9 and its mono-hydride structure was found computationally in favor of the dimer by 16.5 kcal·mol⁻¹ in terms of enthalpy energy, in direct agreement with the experimentally observed thermal stability of 9 [18]. Hence, any mechanistic scenario starting from the monomeric complex was not considered in our studies.

Figure 9. Enthalpy energy ΔH (kcal·mol⁻¹) profiles for the formation of the corresponding trienediyl complexes using (a) PhC≡CH, and (b) ^tBuC≡CH as substrates, and [(Cp*)₂Sm(μ-H)]₂ as a precursor. Cp* ligands are omitted for clarity.

Figure 9. Enthalpy energy ΔH (kcal·mol⁻¹) profiles for the formation of the corresponding trienediyl complexes using (a) PhC≡CH, and (b) ^tBuC≡CH as substrates, and [(Cp*)₂Sm(μ-H)]₂ as a precursor. Cp* ligands are omitted for clarity.

The same energy reaction pattern was considered in the case of ^tBuC≡CH. Again, the first step corresponds to a low energy process, even lower than for the phenylacetylene, with sufficient energy stabilization upon H₂ release, being almost isoenergetic to that obtained for PhC≡CH. However, unlike PhC≡CH, the second DM|σ-BM step yields the uncoupled alkynyl complex, as was the case in the divalent samarocene, from which the C–C coupling is not kinetically easily accessible (Figure 9b). Hence, such type of two consecutive (DM|σ-BM–H₂ release) steps can serve as an explanation of the experimentally observed reactivity, shedding ample light into this peculiar reactivity in lanthanide chemistry.

3. Computational Section

All the quantum calculations, required for the delineation of the intermediate and transition state molecular structures, were performed at the density functional theory level using the B3PW91 [30,31] hybrid functional as implemented in the Gaussian program code [32]. The basis set used for the samarium atoms was the Stuttgart–Dresden–Koln (large-core) [33], augmented by an f polarization function in conjunction with the appropriate scalar relativistic pseudopotential [34]. For all the remaining atoms the all-electron double zeta basis set 6–31 G (d,p) was used [35,36]. Enthalpy energies were obtained at T = 298.15 K based on the harmonic approximation. Intrinsic Reaction Paths (IRPs) were traced from the various transition structures to verify the reactant to product linkage [37,38]. Natural population analysis (NPA) was performed using Weinhold’s methodology [39,40].

4. Conclusions

Although the reaction of a low-valent samarocene with terminal alkynes affording a trienediyl binuclear complex and H₂ seems simple, the mechanism underlying this transformation is mechanistically complicated. In the present DFT mechanistic study, the cooperativity of the two lanthanides is highlighted, with a unique mechanistic scenario being proposed, involving novel type of transition states, such as the DM|σ-SBM, for example. In this particular saddle point of the potential energy surface, the two samarium centers are working synergistically to facilitate H₂ elimination, by forming a six-member transition state. It is worth noting that this type of transition state consists of a novel paradigm of σ-bond metathesis that can be added to the already well-established related types [41]. Among other mechanistic paths, the operating mechanism consists of (i) the oxidation of each samarium center and the concomitant double reduction of the alkyne to afford a binuclear intermediate; (ii) the C–H scission of the acetylinic bond that lies in between the two metals; (iii) a DM|σ-SBM process that releases the H₂ molecule; (iv) the C–C coupling of the two bridged μ-alkynides to give the final experimentally observed bimetallic trienediyl complex. For the latter mechanistic route the experimentally used substrate, phenylacetylene, was considered first, and afterwards the hex-1-yne in order to check the effect of the alkyl substituent. For the aliphatic alkyne, the applied mechanism leads to the same conclusions in terms of enthalpy of reaction, making this process feasible experimentally. On the other hand, the lack of obtaining the trienediyl analogs when the ^tBuC≡CH is used, as reported experimentally, is due to the thermodynamic preference for disruption of the bimetallic product of the DM|σ-SBM step over the homocoupling of the two bulky alkynyl moieties. We strongly believe that this computational contribution sheds light into the mechanism at work in this peculiar reactivity, and will help in the direction of additional understanding in the future of organolanthanide chemistry.