Ruthenium-based catalysts bearing unsymmetrical NHCs for specific olefin metathesis applications

Abstract

Novel ruthenium complexes for olefin metathesis incorporating unsymmetrical, backbone-substituted N-heterocyclic carbenes (uNHCs) were synthesised and fully characterised. The effect of NHC backbone configuration (syn or anti), in combination with a pendant N-benzyl substituent and increasingly sterically demanding N′-aryl groups, was systematically evaluated in representative metathesis reactions and in transformations targeting the valorisation of renewable substrates into industrially relevant products. The NHC substitution pattern significantly affected both the stability and catalytic behaviour of the resulting complexes. In comparison to the commercial benchmark Hoveyda–Grubbs second-generation catalyst (HGII), which remains generally superior in standard metathesis reactions, the newly developed complexes proved to be more efficient in selected sustainable processes such as the self-metathesis of eugenol acetate and the ethenolysis of ethyl oleate. Notably, the N-benzyl, N′-2,6-diisopropylphenyl catalyst featuring an anti NHC backbone achieved more than 90% selectivity towards the desired terminal olefins and a TON up to 12 900 in the ethenolysis reaction. Moreover, in the synthesis of degradable polymers via alternating ring-opening metathesis polymerisation of bio-based monomers (an exo-norbornene derivative and a cyclic enol ether), the analogous catalyst bearing a syn NHC backbone displayed higher Z-selectivity compared to HGII (Z double bond content 78% vs. 21%).

Unsymmetrical NHC–ruthenium catalysts enable sustainable olefin metathesis. Catalyst activity and selectivity are tuned by backbone configuration and the size of the N-aryl substituent (mesityl vs. diisopropylphenyl).

Introduction

Olefin metathesis is a well-established catalytic methodology for the efficient construction of carbon–carbon double bonds.^1,2 This synthetic approach aligns with the principles of green chemistry by promoting atom economy, operational simplicity and reduced by-product generation.^3,4 Accordingly, it is currently regarded as an indispensable tool in a variety of industrial sectors, including oleochemistry, materials, pharmaceuticals, agrochemicals, and perfumery.^1,2,5–10 The enormous popularity of olefin metathesis in several fields of chemical synthesis is closely related to the introduction of easy-to-handle, stable ruthenium-carbene catalysts, especially those supported by N-heterocyclic carbene (NHC) ligands^11,12 which are particularly interesting owing to their robustness, high tolerance towards functional groups, and broad applicability across different substrates, including bio-sourced olefins.^13–17 With the growing demand for sustainable chemical transformations, ruthenium-catalysed olefin metathesis has indeed gained an increasingly important role in the conversion of renewable raw materials into value-added products.^8,16,18 Further advancement in adopting this methodology, however, relies on the development of even more performant catalysts that combine high robustness and selectivity with cost-efficiency, thereby facilitating scalable and sustainable processes. Considerable efforts have been devoted, over the years, to structural modifications of ruthenium complexes to enhance their catalytic performance. In particular, the fine-tuning of the steric and electronic properties of the NHC ligands has been extensively explored. Variation of substituents on the NHC backbone and/or the nitrogen atoms has been shown to exert a substantial influence on the activity, stability, and selectivity of the resulting catalysts in a wide range of metathesis transformations.^11,12,19

Notably, the introduction of unsymmetrical NHCs (uNHCs), featuring structurally and sterically differentiated substituents at the nitrogen atoms, in close proximity to the ruthenium centre, has led to ruthenium complexes exhibiting enhanced stability and/or selectivity.^19–22 These complexes have found successful applications in challenging metathesis reactions where symmetrical analogues often fall short or fail.^{19,20,22–24} Such reactions include the synthesis of alternating copolymers,^25–27 self-metathesis of terminal olefins,^28–31 macrocyclization,^32,33 and ethenolysis.^34–36 Moreover, some uNHC–ruthenium complexes have been shown to perform well with polyfunctional substrates of pharmaceutical interest, even in green solvents, thus representing promising candidates for large-scale applications in drug production.^37–39 The combination of the dissymmetry arising from different N-substituents with NHC backbone modification may introduce further control over the steric and electronic properties of the carbene ligand, and consequently over the catalytic performance of the resulting complexes.^20,40 In this context, our group has contributed to the development of novel olefin metathesis ruthenium catalysts featuring uNHC ligands bearing alkyl and aryl substituents of varying steric bulk on the nitrogen atoms, as well as phenyl groups on the NHC backbone in either syn or anti configurations.^41–44 Complexes with an anti-NHC backbone generally exhibit enhanced thermal stability and higher activity relative to their syn analogues; however, these differences are mitigated by the introduction of highly branched, bulky N-alkyl substituents. Furthermore, variation of the N′-aryl substituent, specifically 2-isopropylphenyl versus mesityl, significantly influences not only on the stability of the corresponding complexes but also their activity and selectivity.⁴⁴ These uNHC ligand modifications result in enhanced performance in demanding transformations, including the ring-closing metathesis of sterically hindered diolefins,^41,42 the ethenolysis of ethyl oleate,⁴⁴ and the alternating polymerisation of norbornene with cyclopentene or cyclooctene.^45,46 Overall, these studies highlight the importance of considering both the nature of the NHC N-substituents and the backbone configuration in modulating catalyst performance.

To gain deeper insight into the relationship between steric modulation of uNHC ligands and the catalytic behaviour of the resulting ruthenium complexes, four new systems were designed and synthesised (Fig. 1). These complexes belong to the family of second-generation Hoveyda–Grubbs catalysts and feature syn and anti-oriented phenyl substituents on the NHC backbone that represent a substitution pattern readily accessible from commercially available and cost-effective diphenylethylenediamines. In addition, the ligands incorporate a dangling benzyl group as a constant feature on one nitrogen atom to provide conformational flexibility, and aryl substituents of increasing steric demand (mesityl or 2,6-diisopropylphenyl) on the other nitrogen atom to modulate the steric environment around the ruthenium centre. These structural elements were expected to influence catalytic behaviour and enable identification of key uNHC structural parameters that enhance catalyst performance, particularly in specific metathesis applications.

Fig. 1. Structures of ruthenium complexes used in this study.

The catalytic behaviour of the resulting complexes was evaluated in standard ring-closing metathesis reactions and in more challenging transformations involving renewable substrates, including the ethenolysis of ethyl oleate and the synthesis of degradable polymers via alternating ring-opening metathesis polymerisation of an oxanorbornene derivative with 2,3-dihydrofuran. The performance of these systems was compared to that of the commercially available second-generation Hoveyda–Grubbs catalyst (HGII) (Fig. 1).

Results and discussion

Synthesis of ruthenium complexes

The synthesis of ruthenium complexes 1 and 2 was accomplished through a sequence of four straightforward steps starting from commercial (R,R)- or meso-1,2-diphenylethylenediamine, according to modified literature procedures (Scheme 1).^41,44 The initial step for complexes 1a and 2a having an anti NHC backbone involved the monoalkylation of (R,R)-1,2-diphenylethylenediamine by a condensation reaction with benzaldehyde in water, followed by a reduction with sodium borohydride in methanol (A, 70% yield).⁴⁷ In the second step, the monoalkylated diamine A was arylated via a Pd-catalysed Buchwald–Hartwig cross coupling reaction with mesityl (for 1a) or 2,6-diisopropylphenyl (for 2a) bromide in toluene, affording the desired unsymmetrical diamines B and B′ in good yields (70–78%). Regarding complexes presenting a syn backbone, owing to the higher tendency to give dialkylated diamines,⁴¹ the initial step was the Buchwald–Hartwig monoamination of meso-1,2-diphenylethylenediamine with mesityl (for 1b) or 2,6-diisopropylphenyl (for 2b) bromide in toluene (65–68% yields). The resulting monoarylated diamines C and C′ were then alkylated through a condensation reaction with benzaldehyde, followed by reduction with sodium borohydride in methanol (E and E′, 64–76% yields). The third step, common to both the unsymmetrical substituted diphenylethylenediamines, involved the formation of the NHC ligand salts by cyclization in the presence of triethyl orthoformate and ammonium tetrafluoroborate, at 135 °C (D, D′, F and F′, 70–88% yields). Finally, in the fourth step, free NHCs generated in situ by treating the salts with potassium tert-amylate in toluene, were reacted with Hoveyda–Grubbs first generation catalyst (HGI), resulting in the formation of the desired complexes 1 and 2 which were isolated as air and moisture-stable green solids in moderate yields (44–54%). Complexes 1a and 2a, obtained from (R,R)-1,2-diphenylethylenediamine, are enantiopure, whereas complexes 1b and 2b, derived from meso-1,2-diphenylethylenediamine, are racemic mixtures.

Scheme 1. Synthetic pathways to the synthesis of ruthenium complexes 1 and 2 bearing uNHCs.

All complexes were characterized by 1D and 2D NMR spectroscopy and ESI-FT-ICR analysis. The solution structures of 1a, 1b, 2a and 2b, determined by NMR experiments recorded in C₆D₆ at room temperature, revealed the presence of only one rotational isomer for each complex. This isomer was identified by 2D NOESY experiments as the one in which the benzylidene moiety is located underneath the N-aryl group (Fig. S35, S40, S45 and S50). Moreover, analysis of ¹H NMR spectra indicated that both benzyl and aryl N-substituents undergo restricted rotation around the carbon–nitrogen bond of the NHC ligand, likely attributable to the steric influence of the phenyl substituents on the ligand backbone. In fact, the methylene protons of the benzyl group, as well as the methyl groups of the mesityl substituent or methine and methyl groups of 2,6-diisopropylphenyl substituent, appear as non-equivalent signals in the corresponding ¹H NMR spectra (Fig. S31, S36, S41 and S46).

As part of the solid-state structural analysis of the new complexes, crystallization experiments were carried out for all compounds. Single crystals of complexes 1a and 1b, suitable for X-ray measurements, were obtained at room temperature by vapor diffusion of pentane into the corresponding oversaturated benzene solutions (10 mg of 1a and 15 mg of 1b in 1.5 mL and 2.0 mL of benzene, respectively). Since the enantiopure ruthenium complex 1a did not yield crystals of sufficient quality for X-ray diffraction analysis, the crystallographic study was performed using crystals obtained from the corresponding racemic mixture. The X-ray molecular structures are shown in Fig. 2; crystallographic data and refinement details are reported in the SI (Table S1).

Fig. 2. ORTEP drawing of (a) compound 1a and (b) compound 1b. Ellipsoids are drawn at 20% probability level.

Crystals of 1a belong to centrosymmetric space group P1̄ and feature (R,R) and (S,S) configuration for C2 and C3 carbon atoms.

Crystals of 1b belong to centrosymmetric space group P2/c and feature (S,R) and (R,S) configuration for C2 and C3 carbon atoms.

In both X-ray molecular structures, we observe:

(i) One methylene hydrogen atom (H16) of the benzyl moiety features anagostic interactions with similar intramolecular C–H‧‧‧Ru distances (2.57 Å and 2.59 Å, respectively in 1a and 1b) and C–H‧‧‧Ru angles (118.6° and 118.3°, respectively in 1a and 1b), see also Fig. 2a, b and Tables S2, S3.⁴⁸

(ii) The mesityl aromatic C atoms interact with the alkylidene hydrogen atom (H32) through CH–π interaction, although in 1a three C atoms are at distance less than the sum of the van der Waals radii, while in 1b two C atoms are at distance less than the sum of the van der Waals radii (Fig. 2a, b and Tables S2, S3). This is also evident from the distance between H32 atom and the centroid of the mesityl aromatic plane in 1a and 1b (2.498 Å and 2.706 Å, respectively).

The thermal stability of the new complexes was evaluated by heating a 0.01 M C₆D₆ solution of each complex under nitrogen at 60 °C and monitoring decomposition by ¹H NMR spectroscopy over a period of one month, using tetrakis(trimethylsilyl)silane as an internal standard. Degradation profiles over 20 days are shown in Fig. 3. All complexes exhibited high stability; notably, the anti isomers were more stable than their syn counterparts, with this difference being more pronounced for complexes bearing less bulky N-mesityl substituents. Furthermore, 1a, 2a, and 2b were found to be more stable than the benchmark symmetrical catalyst HGII, retaining more than 90% of the intact complex after one month, whereas the least stable complex, 1b, showed only 65% of the initial complex remaining. As previously discussed in the literature, the stability of NHC–ruthenium complexes is strongly influenced by the steric properties of the NHC ligand, with more sterically demanding NHCs generally conferring enhanced stability.^44,49 In line with this trend, Grela et al. showed that, for related N-benzyl, N′-aryl complexes lacking backbone substitution, mesityl-substituted derivatives were less stable than their 2,6-diisopropylphenyl analogues.³⁰ These findings indicate that backbone stereochemistry plays a key role in modulating the steric environment of the NHC ligand and, consequently, the stability of the corresponding complexes.

Fig. 3. Stability tests of 1 and 2 in C₆D₆ (0.01 M) at 60 °C under nitrogen. Decomposition was monitored by ¹H NMR spectroscopy using tetrakis(trimethylsilyl)silane as an internal standard. The lines are intended as a visual aid only.

As for electronic properties, it is well known that the σ-donor properties of NHC ligands are crucial in controlling their interaction with transition metals, and consequently, in influencing the efficiency of the resulting complexes. Therefore, the σ-donor character of the NHCs of 1 and 2 was assessed using a ¹³C NMR-based method, which allows correlation of the ¹J_CH coupling constant between the carbon and the hydrogen atoms at the precarbenic position in azolium salts with the σ-electron donating ability of the corresponding NHCs.^50,51 Higher ¹J_CH values correspond to weaker σ-donating properties.⁵² The ¹J_CH values measured for the NHC precursors of 1, 2 and HGII (ref. 50) were nearly identical (206–207 Hz), suggesting that the resulting NHCs possess comparable σ-donor properties despite differences in substitution on the backbone and nitrogen atoms of the NHC ligand.

Catalysis

The catalytic behaviour of the newly synthesized complexes was investigated in ring-closing metathesis (RCM), cross metathesis (CM), self metathesis (SM) and ring-opening metathesis polymerisation (ROMP) of both model and bio-based substrates. For comparison, the results were benchmarked against those obtained with the commercially available HGII catalyst and, whenever possible, against those reported in the literature for other uNHC ruthenium catalysts.

Ring-closing metathesis (RCM) reactions

As a first step, complexes 1 and 2 were evaluated in the model RCM reaction of diethyl diallylmalonate (3) to give the cyclic disubstituted olefin 4.⁵³ All reactions were performed at 60 °C in C₆D₆ using 1 mol% catalyst loading and were monitored over time by ¹H NMR spectroscopy. The time-conversion profiles are shown in Fig. 4. Parallel reaction with the commercially available HGII catalyst, also included in the plot, was carried out at a lower loading (0.1 mol%). This reduced catalyst loading was chosen because of the exceptionally high activity of HGII in the RCM of diethyl diallylmalonate (3), which achieves complete conversion in less than 5 minutes at 1 mol%.

Fig. 4. Kinetic profiles for the RCM of diethyl diallylmalonate (3).

Both the NHC backbone configuration and the steric bulk of the aryl N-substituent were found to influence the reaction outcome. In particular, the less sterically encumbered N-mesityl catalysts 1a and 1b reached full conversion within 20 and 30 minutes, respectively, while the bulkier N-2,6-diisopropylphenyl complexes 2a and 2b required more than 2 h to complete cyclization. Moreover, anti catalysts revealed as more active than their syn analogues. All newly developed catalysts were less efficient than the commercially available symmetrical HGII, even when the latter was employed at a tenfold lower catalyst loading.

The catalytic performance of complexes 1 and 2 was further investigated in the RCM of (±)-linalool (5), a naturally occurring terpene alcohol.⁵⁴ This transformation is particularly significant, as it produces 1-methylcyclopent-2-en-1-ol (6) and isobutylene, key intermediates in the synthesis of renewable fuels and polymeric materials.^55,56 The RCM of 5 is characterized by significant steric hindrance, due to one double bond being trisubstituted and the other bond being attached to a tetrasubstituted carbon. Nonetheless, the reaction is facilitated by the presence of an allylic hydroxyl group, which can interact with the catalytic centre.⁵⁷ The kinetic profiles for the RCM of linalool (5) carried out at 60 °C in C₆D₆ in the presence of catalysts 1, 2 and HGII are sketched in Fig. 5. Notably, HGII exhibited the highest overall efficacy, achieving 100% conversion within merely 4 minutes. Catalysts 1a and 1b demonstrated comparable activity in the RCM of 5 achieving 100% conversion after 28 and 29 minutes respectively; however, both were slower than HGII. Concerning 2a and 2b, they exhibited significantly lower activity compared to 1a and 1b, indeed after 30 min no conversion was observed in both cases (Fig. 5A). The activity of these catalysts was therefore monitored every hour for the first four hours and subsequently after 24 hours. As shown in Fig. 5B, catalyst 2a with an anti backbone configuration achieved 96% conversion after 24 hours, whereas catalyst 2b, which possesses a syn backbone configuration, resulted in a 94% conversion after the same time. From these results, it is evident that the introduction of sterically encumbered aryl N-groups, such as 2,6-diisopropylphenyl, dramatically slows down the reaction rate. In contrast, the uNHC backbone configuration appears to play a negligible role overall.

Fig. 5. Time-conversion plot for the RCM of (±)-linalool (5): (A) promoted by complexes 1, 2 and HGII (over 30 min); (B) promoted by complexes 2a and 2b (over 24 h).

Cross metathesis (CM) and self metathesis (SM) reactions

Then, all catalysts were examined in the model CM of allyl benzene (7) and cis-1,4-diacetoxy-2-butene (8).⁵³ The reaction was carried out employing two equivalents of 8 with respect to 7 and 2.5 mol% of catalyst, in refluxing dichloromethane for 12 hours (Scheme 2). The results are summarized in Table 1.

Scheme 2. CM of allyl benzene (7) and cis-1,4-diacetoxy-2-butene (8).

Table 1. CM of 7 and 8 promoted by catalysts 1, 2, and HGII.

Entry	Catalyst	Isolated yield (%)	E/Z
1	1a	24	3.7
2	1b	21	2.9
3	2a	19	2.0
4	2b	10	1.8
5a	HGII	69	8.6

^aRef. 44.

All catalysts displayed lower activity than the benchmark HGII catalyst. Catalysts bearing the less sterically demanding N-mesityl substituent (1a and 1b) outperformed their N-2,6-diisopropylphenyl counterparts, and catalysts featuring an anti backbone were more active than their syn analogues (1b and 2b). Among the newly synthesized catalysts, catalyst 1a, characterized by an anti-backbone configuration alongside N-benzyl, N′-mesityl groups, achieved the highest conversion.

The backbone configuration in these complexes slightly affects the E/Z ratio, with the syn catalysts exhibiting lower E/Z ratios compared to their anti analogues. It should be noted that the structural variation introduced into the NHC ligand of 1 and 2 resulted in much lower activities than those reported for similar backbone-substituted N-alkyl, N′-aryl NHC–Ru catalysts tested in the same standard reaction conditions,⁴⁴ as well as than those of analogous N-benzyl catalysts previously reported by Grela et al., which, however, were evaluated under different conditions (1 mol% of [Ru], 3 equiv. Cis-1,4-diacetoxy-2-butene, toluene, c = 100 mM, 80 °C, 20 h).³¹ Nonetheless, we decided to investigate the catalytic potential of the newly synthesized complexes in more demanding SM and CM reactions, such as those involving compounds from renewable resources. In this context, phenylpropenoids such as eugenol and its derivatives, which are functionalized allylbenzenes derived from essential oils, represent suitable substrates for metathesis transformations, affording valuable polyfunctional olefins that can be further modified for a wide range of applications.^58–60

A major challenge associated with SM and CM metathesis reactions of these substrates is their tendency to undergo double bond isomerization during metathesis, therefore is important to identify catalysts which promote these transformations with high selectivity.^58,61,62 To this end, the efficiency of uNHC catalysts 1 and 2 was first tested in the self metathesis of eugenol acetate (10). The reaction was carried out using 2.5 mol% of catalyst, in toluene at 40 °C for 15 hours (Scheme 3), according to the procedure reported by Grela et al.³¹ The results are reported in Table 2.

Scheme 3. SM of eugenol acetate (10).

Table 2. SM of 10 promoted by catalysts 1, 2 and HGII.

Entry	Catalyst	Isolated yield (%)	E/Z
1	1a	49	4.0
2	1b	70	3.3
3	2a	79	3.0
4	2b	67	3.8
5	HGII	—a	—a

^aInseparable mixture of products obtained.

All the catalysts proved to be fully selective, affording the desired self metathesis product (11) in good isolated yields (49–79%) and without the formation of any undesired isomerized or shortened byproducts. Notably, the commercial HGII catalyst was found to be ineffective in this reaction, as previously reported.³¹ Indeed, it affords an inseparable mixture containing the desired product, its C–C double bond isomerized analogue, a shorter compound resulting from cross-metathesis between the product and its isomer, as well as traces of the corresponding stilbene.

Catalysts with anti backbone showed an appreciable difference in activity, with 2a, bearing the more sterically encumbered N-2,6-diisopropylphenyl NHC, exhibiting a higher conversion to the desired product. It should be noted that, under the same reaction conditions, the analogous N-benzyl, N′-2,6-diisopropylphenyl NHC catalyst, lacking backbone substitution, reported by Grela et al., furnished 85% yield.³¹

As previously observed, reactions conducted in the presence of catalysts 1b and 2b, featuring a syn NHC backbone, resulted in slightly lower E/Z ratios.

Eugenol acetate (10) was also employed in a CM reaction with cis-1,4-dichloro-2-butene (12), as this transformation affords an allylic chloride derivative (13) that can be further converted into a variety of useful chemicals.^63,64 The reaction was conducted using three equivalents of cis-1,4-dichloro-2-butene relative to eugenol acetate, in the presence of 2.5 mol% of catalyst, in refluxing dichloromethane for 12 hours (Scheme 4). All results, including those obtained with the benchmark HGII catalyst, are summarized in Table 3.

Scheme 4. CM of eugenol acetate (10) and cis-1,4-dichloro-2-butene (12).

Table 3. CM of 10 and 12 performed in the presence of catalysts 1, 2 and HGII.

Entry	Catalyst	Isolated yield (%)	E/Z
1	1a	55	9.2
2	1b	30	7.7
3	2a	38	2.4
4	2b	29	2.7
5a	HGII	52	7.4

^aRef. 49.

The newly synthesized catalysts were all active in this reaction, with anti catalysts performing better than their syn counterparts, as previously observed for similar systems.⁶⁵ Regarding N-aryl substitution, catalysts with bulkier N-2,6-diisopropylphenyl group showed higher Z-selectivity than the less encumbered N-mesityl catalysts. Among all the catalysts, catalyst 1a, featuring an anti backbone and an N-mesityl substituent, exhibited the highest activity, surpassing even the benchmark catalyst HGII.

A special form of CM reaction is represented by ethenolysis, which involves ethylene as one of the partners and is used to cleave internal carbon–carbon double bonds forming terminal olefins.^66,67

In the oleochemical industry, increasing interest in natural fats and fatty acids as renewable feedstocks has driven the development of catalytic processes to replace petrochemical-derived products.^9,18,59,68 Among these, ethenolysis of unsaturated fatty acid esters is one of the most extensively studied metathesis reactions, providing an efficient and atom-economical route to ω-unsaturated fatty substrates and short-chain 1-alkenes, high-value chemicals with diverse industrial applications.^66,67 In recent years, the ethenolysis of ethyl oleate, a widely used model for natural fatty acid glycerol esters, has attracted particular attention as a benchmark transformation. Numerous research groups have sought to optimize the reaction conditions and identify the most effective catalysts for this process.^{31,35,36,44,69} The challenges associated with ethenolysis primarily stem from the inherently low stability of the propagating methylidene species ([Ru] CH₂), which causes most conventional first- and second-generation catalysts to decompose rapidly in the presence of ethylene.^34,70 To date, ruthenium catalysts incorporating cyclic(alkyl)(amino) carbene (CAAC) ligands have delivered the most significant advancements in ethenolysis.^71–73

Nevertheless, the industrial application of these catalysts may be subject to operational limitations, including the necessity for ethylene of exceptionally high purity, typically above 99.99%.⁷¹ Alternative catalytic systems include unsymmetrical NHC ruthenium complexes, first reported by Grubbs as effective catalysts for ethenolysis reactions owing to their extended lifetimes in metathesis, likely attributable to the enhanced stability of their methylidene species.³⁴

Subsequently, additional ruthenium catalysts bearing uNHC ligands have been identified as suitable for carrying out ethenolysis reactions.^{26,31,35,36,44,68}

To determine which aspects of the uNHC ligand architecture contribute to enhanced performance in ethenolysis, all newly synthesized catalysts were evaluated in the ethenolysis of ethyl oleate (90–95%) using ethylene grade 3.0 (99.5%) at a pressure of 10 bar, under varying catalyst loadings and temperatures (Scheme 5). The results are reported in Table 4. First, the ethenolysis of ethyl oleate (14) was performed at a catalyst loading of 500 ppm and a temperature of 50 °C. All four catalysts were found to be more effective than the benchmark catalyst HGII. Among them, complexes containing N-2,6-diisopropylphenyl groups exhibited superior performance compared to those containing N-mesityl groups, with the syn complex 2b displaying the highest selectivity.

Scheme 5. Ethenolysis of ethyl oleate (14).

Table 4. Ethenolysis of 14 promoted by catalysts 1, 2 and HGII.

Entry	Catalysta	Cat./14 (ppm)	Temperature (°C)	Conversionb (%)	Selectivityc (%)	Yieldd (%)	TONe
1	1a	500	50	79	84	67	1330
2	1b	500	50	64	79	51	1020
3	2a	500	50	86	87	75	1500
4	2b	500	50	80	94	74	1490
5	HGII	500	50	71	43	30	600
6	1a	50	50	15	76	12	2320
7	1b	50	50	18	81	15	2920
8	2a	50	50	30	76	23	4560
9	2b	50	50	14	99	14	2780
10	1a	50	70	12	73	9	1800
11	1b	50	70	13	73	9	1830
12	2a	50	70	70	92	65	12 920
13	2b	50	70	20	99	20	3920

^aThe reactions were conducted under neat conditions utilizing 5.4 mmol of 14 (90–95%) at a pressure of 10 bar of ethylene grade 3 for 3 h. Dodecane (150 µL) was used as an internal standard.

^bDetermined by GC analysis. Conversion = 100 − (final moles of 14) × 100/[initial moles of 14].

^cDetermined by GC analysis. Selectivity = 100 × (moles of ethenolysis products 15 + 16)/[(moles of 15 + 16) + (2 × (moles of 17 + 18))].

^dYield = conversion × selectivity/100.

^eTON = yield × (initial moles of 14/moles of catalyst)/100.

This enhanced efficiency is likely attributable to the steric bulk of the N-2,6-diisopropylphenyl groups, which more effectively stabilizes the active methylidene species against decomposition than the N-mesityl substituents. The results obtained with HGII catalyst at a loading of 500 ppm were significantly inferior, making further evaluation at a reduced loading of 50 ppm unnecessary. Consequently, complexes 1 and 2 were evaluated in the ethenolysis of 14 at a catalyst loading of 50 ppm and a temperature of 50 °C. Notably, catalysts bearing N-mesityl substituents exhibited very similar catalytic behaviour, whereas those containing N-2,6-diisopropylphenyl groups displayed distinct trends depending on the NHC backbone configuration. In particular, the syn complex 2b showed higher selectivity but lower activity than its anti analogue 2a.

Following the evaluation at 50 °C, the reaction temperature was increased to 70 °C to investigate the effect of temperature on catalyst activity and selectivity while maintaining a catalyst loading of 50 ppm. It was observed that catalysts bearing N-2,6-diisopropylphenyl substituents outperformed those containing N-mesityl groups.

Notably, catalyst 2a, featuring an anti NHC backbone, displayed particularly significant results, achieving a TON of 12 920, which, to the best of our knowledge, represents the highest value recorded to date for this class of compounds under comparable reaction conditions.^35,36,44,74 These results suggest that the steric hindrance of the N-aryl substituents, combined with a definite stereochemistry of the NHC backbone, plays a significant role in stabilizing the ruthenium methylidene active species, thereby enhancing the catalytic performance in terms of both activity and selectivity.

Alternating ring-opening metathesis polymerisation (ROMP) reactions

Finally, our attention was focused on investigating catalytic behaviour of complexes 1 and 2 in specific ROMP applications. Indeed, although ROMP is a highly versatile polymerisation technique that enables the synthesis of polymers with tuneable molecular weights, sophisticated architectures, and functionalized macromolecules for a wide range of applications,^1,2,75–77 the synthesis of precision polymers via alternating ROMP copolymerisation of two distinct monomers, particularly using ruthenium-based catalysts, remains relatively limited.⁷⁸ In copolymerisations of cycloolefins with markedly different reactivities, such as norbornene with cyclooctene or cyclopentene, large excesses of the less reactive monomer are often required to achieve high degrees of alternation in the presence of ruthenium complexes bearing unsymmetrical NHC ligands.^25,79 Further design of this class of uNHC catalysts has enabled highly chemoselective alternating copolymerisations of norbornene with cyclooctene or cyclopentene, even at lower comonomer ratios.^45,46,80 As a consequence, the newly synthesized complexes were evaluated in the alternating ROMP of norbornene (NBE) and cyclooctene (COE) to obtain preliminary insights into the catalysts' ability to promote chemoselective copolymerisations. The ROMP reactions were conducted in CH₂Cl₂ at 30 °C at 0.1 mol% of catalyst loading, using a 1 : 1 comonomer ratio. The results are reported in Table 5. The ¹³C{¹H} NMR resonances relative to NBE-COE diads are in the range 128.5–129.0 ppm and 134.8–135.5 ppm. The signals for NBE-NBE are found at 132.9–133.4 ppm and 133.8–134.3 ppm, while the peaks attributable to COE-COE diads at 130.0 and 130.5 ppm. In the copolymers obtained with catalysts 2a and 2b, only NBE-NBE diads were detected. The ¹³C{¹H} NMR olefinic region spectra of obtained copolymers are shown in Fig. 6.

Table 5. ROMP copolymerization of NBE and COE performed in the presence of catalysts 1, 2 and HGII.

Entry	Catalyst	Poly(NBE)a,b [%]	Poly(COE)a,b [%]	Alternating diadsb [%]	M n c (g mol−1)	Ð c
1	1a	30	4	66 (54)	6.1 × 105	1.9
2	1b	34	2	64 (63)	6.1 × 105	2.0
3	2a	42	—	58 (58)	6.8 × 105	2.4
4	2b	56	—	44 (55)	9.5 × 105	2.0
5	HGII	37	47	16 (20)	2.7 × 105	1.8

^aFractions of homopolymer sequences in the copolymer.

^bDetermined using ¹³C NMR.

^cDetermined using CHCl₃ size-exclusion chromatography (SEC) calibrate utilizing polystyrene standards.

Fig. 6. Olefinic region of ¹³C{¹H} NMR (150 MHz) spectra of alternating NBE-COE copolymers.

As shown in Table 5 and Fig. 6, all four catalysts exhibited a strong tendency to form alternating NBE-COE sequences when an equimolar mixture of the two monomers was used. This behaviour stands in clear contrast to that of the benchmark symmetrical HGII catalyst, which once again confirms the inability of this class of complexes to promote alternating ROMP of simple cycloolefins.²⁵ Catalysts 1a,b bearing N-mesityl groups displayed higher chemoselectivity than their N-2,6-diisopropylphenyl counterparts 2a,b, suggesting that increased steric bulk on the N-aryl substituent is detrimental to chemoselectivity.

An even more pronounced influence of the NHC backbone configuration was observed for the N-2,6-diisopropylphenyl catalysts, with the anti isomer 2a outperforming its syn counterpart 2b.

This outcome is particularly intriguing, as previously reported N-cycloalkyl/N′-mesityl uNHC catalysts featuring syn-oriented phenyl groups on the backbone were identified as the most efficient at the same comonomer ratio.^45,46 Taken together, these results reinforce the notion that subtle modifications within the uNHC framework can lead to markedly different catalytic behaviours and provide valuable insights into the balance of ligand substituents required for effective catalyst design.

In pursuit of advanced polymeric materials, increasing attention has been directed towards the development of sustainable and environmentally friendly alternatives to petroleum-derived polymers.⁸¹ Biomass-derived polymers, particularly those that are readily degradable, offer a promising route to new materials.⁸² Alternating ring-opening metathesis polymerisation has emerged as an efficient strategy to access degradable polymers from renewable monomers with minimal synthetic complexity.^83,84 In particular, the ROMP of fully biomass-derived monomers, obtained from furan and maleic anhydride with complete atom economy and a remarkably low E-factor, together with a biomass-derived cyclic enol ether, affords alternating copolymers that undergo degradation into small molecules under acidic conditions, highlighting their potential as sustainable, stimuli-responsive materials.⁸⁵

Taking inspiration from this work, the catalytic behaviour of complexes 1a,b and 2a,b was evaluated in the synthesis of degradable polymers via the alternating ROMP of an oxanorbornene derivative (dimethyl 7-oxabicyclo[2.2.1]hept-5-ene-2,3-dicarboxylate, ONC) and a cyclic enol ether (2,3-dihydrofuran, DHF), two monomers with different reactivities.

In the previous work of Sun et al., high level of alternation (>94%) were reached in the presence of commercial GII catalyst using 2 equivalents of the less reactive DHF to suppress homopolymerisation of the oxanorbornene derivative. The alternating copolymers displayed a similar content of E and Z double bonds.⁸⁵ In the present study, the molar ratio of ONC to DHF was initially set to 1 : 1 and the ROMP copolymerisations were conducted at room temperature for 3 h in THF as the solvent.

The results are reported in Table 6 (entries 1–5).

Table 6. Alternating ROMP of ONC and DHF promoted by catalysts 1, 2 and HGII.

Entry	Catalyst	[ONC]/[DHF]/[cat]	Time (h)	Alternating diadsa [%]	M n b (g mol−1)	Ð b
1	1a	120 : 120 : 1	3	76 (53)	4.1 × 104	1.4
2	1b	120 : 120 : 1	3	68 (49)	5.9 × 104	2.4
3	2a	120 : 120 : 1	3	66 (75)	5.7 × 104	2.3
4	2b	120 : 120 : 1	3	82 (76)	3.4 × 104	2.0
5	HGII	120 : 120 : 1	3	84 (27)	1.0 × 104	1.7
6	2b	120 : 240 : 1	12	>98 (78)	3.5 × 104	1.9
7	HGII	120 : 240 : 1	12	>98 (21)	7.1 × 103	2.2

^aDetermined using ¹H-NMR. The value in parenthesis is the percentage of Z double bonds.

^bDetermined using THF size-exclusion chromatography (SEC) calibrate utilizing polystyrene standards.

The resulting copolymers exhibited controlled molecular weights and a notable degree of alternation, evaluated using ¹H NMR spectroscopy. The integration of proton signals for ONC and DHF repeating units revealed that all poly(ONC-alt-DHF) copolymers exhibited high degrees of alternation (66–82% of alternating diads), demonstrating excellent sequence control within the polymer structures. Interestingly, the symmetrical catalyst HGII and N-2,6-diisopropylphenyl catalyst 2b, bearing syn-related substituents on the backbone, were found to give the same degree of alternation but opposite E/Z stereochemistry of double bonds along the polymer chain (Fig. 7).

Fig. 7. ¹H NMR spectra (diagnostic region) of the alternating copolymers poly(ONC-alt-DHF) obtained with HGII and 2b, as well as the two homopolymers including poly(ONC) and poly(DHF), obtained with HGII, in CDCl₃ at 25 °C. The asterisk denotes the residual solvent signal.

The most effective catalysts HGII and 2b were also tested using an ONC/DHF comonomer ratio of 1 : 2, as described by Sun et al.⁸⁵ In the ¹H NMR spectra of copolymers obtained using a ONC/DHF comonomer ratio of 1 : 2, the characteristic olefinic proton resonances associated with poly(DHF) were absent, indicating that DHF homoaddition did not occur during polymerization (Fig. 5). Moreover, the olefinic proton signal characteristic of poly(ONC) was markedly reduced in the copolymer, suggesting substantial suppression of ONC homoaddition throughout the alternating ROMP process. Again, opposite E and Z double bonds content was observed. ¹³C NMR analysis corroborates the ¹H NMR results, as the copolymer spectra show no olefinic carbon signals attributable to poly(DHF) sequences (Fig. S66).

The thermal behaviour of the highly alternating copolymers obtained with HGII and 2b was evaluated by thermogravimetrical analysis (TGA) and differential scanning calorimetry (DSC). Both polymers displayed good thermal stabilities with decomposition temperatures (T_d, with 5% weight loss) of 240 °C and 290 °C, respectively (Fig. S67). In both cases, the thermograms revealed only a glass transition temperature (T_g), with no detectable melting or crystallization events (Fig. S68), indicating an amorphous nature. Notably, the T_g values were strongly dependent on the copolymer microstructure, amounting to 34 °C for the HGII-derived copolymer and 26 °C for the material obtained with 2b.

Overall, although alternating ROMP of ONC and DHF can be achieved using the commercially available symmetrical catalyst HGII, the use of unsymmetrical NHC-based complexes enables superior Z-selectivity. This enhanced stereocontrol translates into polymeric materials with distinct microstructures and, consequently, differentiated thermal properties and application potential. Further studies on the development of new degradable alternating copolymers with varying stereoregularity are currently underway.

Experimental

Diamines B, B′ and (E, E′) were synthesized according to literature procedures;^44,86,87 experimental details are reported in the SI.

General procedure for the synthesis of tetrafluoroborate salts⁴¹

The desired diamine (B, B′, E or E′, 1.0 equiv.) was dissolved in triethylorthoformate (8 equiv.) followed up by the addition of NH₄BF₄ (1.2 equiv.) and left to stir for 4 hours at 135 °C. After 3 hours, water was removed from the condenser, and the heating was maintained for an additional hour. Following this, the reaction's completion was assessed using TLC and triethylorthoformate was removed washing the raw product with a small amount of diethyl ether. Subsequently, the remaining solid was dried using a rotary evaporator and the crude product was purified via flash chromatography with a solvent system of petroleum ether/diethyl ether (1 : 1). The pure products were obtained as white solids.

D: yield 88%.

¹H NMR (600 MHz, CDCl₃, Fig. S15): δ = 8.55 (s, 1H, N–CH–N), 7.43–7.38 (overlapping signals, 5H, Ar–H), 7.30–7.29 (overlapping signals, 3H, Ar–H), 7.06–7.03 (overlapping signals, 3H, Ar–H), 6.96 (t, J = 7.7 Hz, 2H, Ar–H), 6.80 (br s, 1H, Ar–H), 6.75 (d, J = 7.3 Hz, 2H, Ar–H), 6.72 (br s, 1H, Ar–H), 6.33 (d, J = 11.8 Hz, 1H, N–CH–CH–N), 5.60 (d, J = 11.8 Hz, 1H, N–CH–CH–N), 5.17 (d, J = 14.2 Hz, 1H, CHH–Ph), 4.70 (d, J = 14.2, 1H, CHH–Ph), 2.38 (s, 3H, Ar–CH₃), 2.30 (s, 3H, Ar–CH₃), 2.15 (s, 3H, Ar–CH₃) ppm. ¹³C NMR (150 MHz, CDCl₃, Fig. S16): δ = 157.46 (N–CH–N), 140.28 (Cq), 136.18 (Cq), 134.64 (Cq), 134.45 (Cq), 132.93 (Cq), 132.20 (Cq), 130.46 (CH), 130.32 (CH), 130.22 (CH), 130.14 (CH), 129.53 (CH), 129.46 (CH), 129.33 (CH), 129.27 (CH), 128.60 (CH), 128.48 (CH), 127.11 (CH), 75.38 (N–CH–CH–N), 69.78 (N–CH–CH–N), 50.93 (CH₂–Ph), 20.89 (Ar–CH₃), 18.52 (Ar–CH₃), 17.74 (Ar–CH₃) ppm. HRMS: MALDI-FT-ICR calcd for [(M-BF₄⁻)]: 431.24818; found. 431.24738.

D′: yield 80%.

¹H NMR (400 MHz, CDCl₃, Fig. S19): δ = 8.67 (s, 1H, N–CH–N), 7.56–7.23 (overlapping signals, 14H, Ar–H), 7.20 (dd, J = 7.9, 1.3 Hz, Ar–H), 6.98 (dd, J = 7.9, 1.3 Hz, 1H, Ar–H), 6.92 (d, J = 7.4 Hz, 2H, Ar–H), 5.53 (d, J = 8.9 Hz, 1H, N–CH–CH–N), 5.30 (d, J = 14.4 Hz, 1H, CHH–Ph), 5.03 (d, J = 8.9 Hz, 1H, N–CH–CH–N), 4.54 (d, J = 14.4 Hz, 1H, CHH–Ph), 2.98 (hept, J = 6.8 Hz, 1H, CH(CH₃)₂), 2.44 (hept, J = 6.8 Hz, 1H, CH(CH₃)₂), 1.34 (d, J = 6.8 Hz, 3H, CH(CH₃)₂), 1.28 (d, J = 6.8 Hz, 3H, CH(CH₃)₂), 1.13 (d, J = 6.8 Hz, 3H, CH(CH₃)₂), 0.29 (d, J = 6.8 Hz, 3H, CH(CH₃)₂) ppm.¹³C NMR (150 MHz, CDCl₃, Fig. S20): δ = 157.52 (N–CH–N), 147.54 (Cq), 146.12 (Cq), 134.47 (Cq), 132.56 (Cq), 132.20 (Cq), 131.19 (Cq), 130.49 (CH), 130.37 (CH), 130.23 (CH), 129.58 (CH), 129.47 (CH), 129.33 (CH), 128.73 (CH), 127.90 (CH), 126.89 (CH), 124.86 (CH), 77.80 (N–CH–CH–N), 69.82 (N–CH–CH–N), 51.01 (CH₂–Ph), 28.96 (CH(CH₃)₂), 25.56 (CH(CH₃)₂), 24.90 (CH(CH₃)₂), 23.85 (CH(CH₃)₂), 22.29 (CH(CH₃)₂) ppm. HRMS: MALDI-FT-ICR calcd for [(M-BF₄⁻)]: 473.29513, found. 473.29432.

F: yield: 88%.

¹H NMR (600 MHz, CDCl₃, Fig. S23): δ = 8.56 (s, 1H, N–CH–N), 7.42–7.39 (overlapping signals, 5H, Ar–H), 7.29–7.26 (overlapping signals, 3H, Ar–H), 7.04 (br s, 3H, Ar–H) 6.96 (t, J = 7.5 Hz, 1H, Ar–H), 6.76 (d, J = 7.6 Hz, 4H, Ar–H), 6.32 (d, J = 12.2 Hz, 1H, NCHCHN), 5.61 (d, J = 12.1 Hz, 1H, NCHCHN), 5.17 (d, J = 14.4 Hz, 1H, CHH–Ph), 4.72 (d, J = 14.3 Hz, 1H, CHH–Ph), 2.34 (br s, 6H, Ar–CH₃), 2.15 (s, 3H, Ar–CH₃) ppm. ¹³C NMR (150 MHz, CDCl₃, Fig. S24): δ = 157.97 (N–CH–N), 138.58 (Cq), 133.49 (Cq), 133.30 (Cq), 131.10 (Cq), 129.74 (Cq), 129.39 (Cq), 129.24 (Cq), 128.68 (CH), 128.58 (CH), 128.50 (CH), 128.46 (CH), 128.37 (Cq), 128.21 (CH), 127.81 (CH), 127.27 (CH), 126.56 (CH), 71.69 (N–CH–CH–N), 67.34 (N–CH–CH–N), 50.61 (CH₂–Ph), 19.74 (Ar–CH₃), 18.39 (Ar–CH₃), 18.10 (Ar–CH₃) ppm.

MALDI-FT-ICR: calcd for [(M-BF₄⁻)]: 431.24818, found 431.24802.

F′: yield 70%.

¹H NMR (300 MHz, CDCl₃, Fig. S27): δ (ppm) 8.31 (s, 1H, NCHN), 7.26 (br s, 5H, Ar–H) 7.19–7.09 (overlapping signals, 3H, Ar–H), 6.98–6.96 (overlapping signals, 3H, Ar–H) 6.89 (t, J = 6.6, 2H. Ar–H), 6.81 (t, J = 7.7, 2H, Ar–H), 6.62 (d, J = 7.8 Hz, 2H, Ar–H), 5.85 (d, J = 12.3 Hz, 1H, NCHCHN), 5.60 (d, J = 12.3 Hz, 1H, NCHCHN), 5.04 (d, J = 14.3 Hz, 1H, CHH–Ph), 4.57 (d, J = 14.3 Hz, 1H, CH₂ CHH–Ph), 2.96–2.84 (m, 2H, CH–(CH₃)₂), 1.14–1.10 (overlapped d, 6H, CH₃), 1.01 (d, J = 6.7 Hz, 3H, CH₃), 0.75 (d, J = 6.7 Hz, 3H, CH₃). ¹³C NMR (150 MHz, CDCl₃, Fig. S28): δ (ppm) 159.42 (N–CH–N), 146.72 (Cq), 146.57 (Cq), 132.18 (Cq), 132.56 (Cq), 130.82 (CH), 130.44 (Cq), 129.93 (CH), 129.71 (CH), 129.63 (CH), 129.58 (CH), 129.47 (Cq), 129.26 (CH), 129.19 (Cq), 128.67 (Cq), 128.22 (CH), 127.92 (CH), 125.27 (CH), 124.93 (CH), 75.10 (NCHCHN), 68.33 (NCHCHN), 51.50 (CH₂–Ph), 29.16 (CH(CH₃)₂), 28.88 (CH(CH₃)₂), 25.81 (CH(CH₃)₂), 25.05 (CH(CH₃)₂), 24.09 (CH(CH₃)₂), 22.96 (CH(CH₃)₂).

MALDI-FT-ICR: calcd for [(M-BF₄⁻)]: 473.29513; found 473.29453.

General procedure for the synthesis of catalysts

Under a nitrogen atmosphere, the desired proligand salt (D, D′, F or F′, 1.2 equiv.) was dissolved in toluene (C = 0.02 M). Potassium tert-amylate (1.2 equiv.) was then added, followed by the addition of the Hoveyda–Grubbs first generation catalyst (HGI) (1.0 equiv.). The reaction mixture was heated for 50 minutes at 70 °C in an oil bath. After that, the mixture was purified using column chromatography with a solvent system of diethyl ether/n-hexane (1 : 5 to 1 : 2), employing TSI silica under a nitrogen atmosphere.

1a: yield = 54%.

¹H NMR (600 MHz, C₆D₆, Fig. S31) δ = 16.48 (s, 1H, Ru CH-oOiPrC₆H₄), 7.82 (d, J = 7.2 Hz, 2H, Ar–H), 7.51 (br s, 1H, Ar–H), 7.20 (dt, J = 11.0, 8.0 Hz, 4H, Ar–H), 7.13–7.03 (overlapping signals, 7H, Ar–H and CHH–Ph), 6.92–6.89 (m, 3H, Ar–H), 6.77 (s, 1H, Ar–H), 6.68 (td, J = 7.5, 0.9 Hz, 1H, Ar–H), 6.57 (s, 1H, Ar–H), 6.45 (d, J = 8.3 Hz, 1H, Ar–H), 5.21 (d, J = 13.6 Hz, 1H, CHH–Ph), 5.17 (d, J = 5.3 Hz, 1H, N–CH–CH–N), 4.81 (d, J = 5.3 Hz, 1H, N–CH–CH–N), 4.66 (hept, J = 6.2 Hz, 1H, O–CH(CH₃)₂), 2.49 (s, 3H, Ar–CH₃), 2.07 (s, 3H, Ar–CH₃), 1.81 (s, 3H, Ar–CH₃), 1.71 (d, J = 6.2 Hz, 3H, O–CH(CH₃)₂), 1.68 (d, J = 6.2 Hz, 3H, O–CH(CH₃)₂) ppm. ¹³C NMR (150 MHz, C₆D₆, Fig. S32): δ = 290.22 (Ru CH-oOiPrC₆H₄), 210.73, 153.64, 145.23, 141.29, 140.00, 139.58, 139.24, 138.96, 137.54, 136.26, 131.06, 130.76, 130.38, 130.12, 129.86, 129.73, 129.47, 129.33, 129.25, 127.97, 122.98, 113.75, 78.00, 75.93, 71.09, 54.99, 30.78, 22.74, 22.62, 21.53, 19.99, 19.68 ppm. ESI-FT-ICR calcd for [M + H]⁺:751.17957, found 751.18256 (Fig. S51).

2a: yield 47%.

¹H NMR (600 MHz, C₆D₆, Fig. S36): δ = 16.42 (s, 1H, Ru CH-oOiPrC₆H₄), 7.81 (d, J = 7.4 Hz, 2H, Ar–H), 7.50 (br s, 2H, Ar–H), 7.30 (t, J = 7.7 Hz, 1H, Ar–H), 7.20–7.04 (overlapping signals, 13H, Ar–H and CHHPh), 6.92–6.88 (m, 3H, Ar–H), 6.69 (t, J = 7.3 Hz, 1H, Ar–H), 6.44 (d, J = 8.4 Hz, 1H, Ar–H), 5.30 (d, J = 13.8 Hz, 1H, CHHPh), 5.03 (d, J = 5.4 Hz, 1H, N–CH–CH–N), 4.69 (d, J = 5.6 Hz, 1H, N–CH–CH–N), 4.64 (hept, J = 6.1 Hz, 1H, O–CH(CH₃)₂), 3.57 (hept, J = 6.6 Hz, 1H, Ar–CH(CH₃)₂), 3.34 (hept, J = 6.6 Hz, 1H, Ar–CH(CH₃)₂), 1.70 (d, J = 6.1 Hz, 3H, O–CH(CH₃)₂), 1.68 (d, J = 6.1 Hz, 3H, O–CH(CH₃)₂), 1.24 (d, J = 6.6 Hz, 3H, Ar–CH(CH₃)₂), 0.98 (d, J = 6.5 Hz, 3H, Ar–CH(CH₃)₂), 0.93 (d, J = 6.5 Hz, 3H, Ar–CH(CH₃)₂) ppm.¹³C NMR (150 MHz, C₆D₆, Fig. S37): δ = 286.98 (Ru CH-oOiPrC₆H₄), 211.71, 153.57, 150.31, 148.84, 143.67, 139.77, 139.60, 137.34, 135.63, 130.50, 129.88, 129.45, 129.27, 129.16, 129.01, 128.92, 128.76, 128.52, 126.73, 125.54, 125.24, 122.55, 122.40, 113.25, 80.27, 75.47, 71.56, 55.17, 28.84, 28.05, 25.56, 24.97, 24.84, 24.14, 24.05, 22.24 ppm. ESI-FT-ICR calcd for [M + H]⁺: 793.22660; found. 793.228 (Fig. S52).

1b: yield = 48%.

¹H NMR (600 MHz, C₆D₆, Fig. S41): δ = 16.52 (s, 1H, Ru CH-oOiPrC₆H₄), 7.82 (d, J = 7.7 Hz, 2H, Ar–H), 7.18–7.06 (overlapping signals, 10H, Ar–H, CHH–Ph), 6.74 (t, J = 7.5 Hz, 1H, Ar–H), 6.70 (t, J = 7.5 Hz, 1H, Ar–H), 6.63–6.59 (overlapping signals, 6H, Ar–H), 6.50 (d, J = 7.7 Hz, 2H, Ar–H), 6.47 (d, J = 8.2 Hz, 1H, Ar–H), 5.78 (d, J = 9.6 Hz, 1H, NCHCHN), 5.31 (d, J = 13.6 Hz, 1H, CHH–Ph), 4.92 (d, J = 9.5 Hz, 1H, NCHCHN), 4.69 (p, J = 6.0 Hz, 1H, O–CH(CH₃)₂), 2.52 (d, J = 10.4 Hz, 6H, Ar–CH₃), 1.96 (br s, 3H, Ar–CH₃), 1.73 (d, J = 5.9 Hz, 3H, O–CH(CH₃)₂), 1.71 (d, J = 6.1 Hz, 3H, O–CH(CH₃)₂) ppm.

¹³C NMR (151 MHz, C₆D₆, Fig. S42): δ = 290.62 (Ru CH-oOiPrC₆H₄), 215.01, 153.08, 144.72, 139.68, 138.24, 138.17, 137.44, 136.69, 133.74, 132.86, 130.51, 130.27, 130.04, 129.76, 129.48, 129.36, 129.07, 128.87, 127.52, 122.52, 122.45, 113.23, 75.44, 74.23, 67.85, 54.39, 22.17, 22.10, 20.82, 19.89, 19.92 ppm.

ESI-FT-ICR: calcd for [M + H]⁺: 751.17957, found: 751.18236 (Fig. S53).

2b: yield = 44%.

¹H NMR (600 MHz, C₆D₆, Fig. S46): δ = 16.46 (s, 1H, Ru CH-oOiPrC₆H₄), 7.75 (d, J = 7.1 Hz, 2H), 7.24 (s, 1H, Ar–H), 7.22 (s, 1H, CHH–Ph), 7.17 (s, 1H, Ar–H), 7.15–7.11 (overlapping signals, 6H, Ar–H), 7.07–7.02 (overlapping signals, 3H, Ar–H), 6.71 (t, J = 7.5 Hz, 1H), 6.63–6.61 (overlapping signals, 5H, Ar–H), 6.46 (d, J = 8.3 Hz, 1H, Ar–H), 5.72 (d, J = 9.6 Hz, 1H, NCHCHN), 5.29 (d, J = 13.5 Hz, 1H, CHH–Ph), 4.93 (d, J = 9.7 Hz, 1H, NCHCHN), 4.66 (p, J = 6.1 Hz, 1H, O–CH(CH₃)₂), 3.78 (p, J = 6.6 Hz, 1H, Ar–CH(CH₃)₂), 3.50 (p, J = 6.7 Hz, 1H, Ar–CH(CH₃)₂), 1.75–1.70 (m, 6H, Ar–CH(CH₃)₂), 1.38 (d, J = 6.8 Hz, 3H, O–CH(CH₃)₂), 1.21 (d, J = 6.3 Hz, 3H, O–CH(CH₃)₂), 1.11 (d, J = 6.3 Hz, 3H, O–CH(CH₃)₂), 1.03 (d, J = 6.7 Hz, 3H, O–CH(CH₃)₂) ppm. ¹³C NMR (151 MHz, C₆D₆, Fig. S47): δ 287.67 (Ru CH-oOiPrC₆H₄), 216.28, 153.67, 149.91, 149.42, 143.47, 137.38, 163.50, 133.88, 132.37, 130.89, 130.29, 129.75, 129.63, 129.45, 129.13, 129.00, 128.80, 128.72, 128.64, 127.60, 125.61, 125.18, 122.58, 122.52, 113.30, 77.33, 75.56, 68.18, 54.49, 30.56, 30.23, 29.12, 28.29, 28.10, 28.03, 26.64, 26.58, 25.80, 25.75, 24.75, 24.69, 22.32, 22.19 ppm.

ESI-FT-ICR: calcd for [M + H]⁺: 793.22660; found.793.22851 (Fig. S54).

General procedure for thermal stability tests

In a nitrogen-filled glovebox, a J-Young NMR tube was charged with 0.80 mL of C₆D₆, 8.0 µmol of catalyst, and 8.0 µmol of tetrakis(trimethylsilyl)silane (TMSS) as the internal standard. After acquisition of the ¹H NMR spectrum at t = 0, the NMR tube was kept in an oil bath at 60 °C for one month, and the decrease in the area of the diagnostic benzylidene proton signal of each catalyst was measured against the area of the methyl signals of TMSS at various time intervals.

General procedure for RCM reactions

In a nitrogen-filled glove box, an NMR tube equipped with a screw-cap septum was charged with 0.80 mL of a solution of catalyst (1 mol%) prepared in C₆D₆. After heating the NMR tube at 60 °C in NMR probe, 0.080 mmol of substrate (0.1 M) was added into tube. The conversion of each substrate was monitored by ¹H-NMR.

RCM of diethyl diallylmalonate (3)

A portion of 3 (19.3 µL) was added to a preheated NMR tube containing 0.80 mL of catalyst solution. The conversion to the cyclic product 4 was determined by integrating methylene protons of the reagent (δ 2.84 ppm as doublet of triplets) and the product (δ 3.14 ppm as singlet).

RCM of (±) linalool (5)

A 14.3 µL sample of 5 was introduced into a preheated NMR tube containing 0.80 mL of catalyst solution. The conversion to the cyclic product 6 was determined by integrating methylene protons of the linalool at δ 1.12 ppm (broad singlet) and of the product at δ 1.27 ppm (broad singlet).

Cross metathesis of allyl benzene (7) and cis-1,4-diacetoxy-2-butene (8)

Under a nitrogen atmosphere, 7 (29.0 mg, 33 µL, 0.249 mmol) and 8 (86.0 mg, 79 µL, 0.498 mmol) were introduced in a vial containing DCM (1.25 mL). Subsequently the catalyst (2.5 mol%) was added, the reaction mixture was heated at 40 °C for 12 h and then the desired product 9 purified by column chromatography with hexane/ethyl acetate 9/1 as eluent. E/Z ratio of 9 was determined integrating the signals at δ = 4.55 (d) and δ = 4.74 (d) ppm (chemical shifts in CDCl₃).

Self metathesis of eugenol acetate (10)

Under a nitrogen atmosphere, the catalyst (0.5 mol%) was added to a vial containing a solution of 10 (0.100 g, 0.500 mmol) in toluene (1.0 mL) solution. The reaction was heated at 40 °C for 15 h and the desired product 11 purified by column chromatography using n-hexane/diethyl ether (9 : 1) as eluent.

E/Z ratio of 11 was determined integrating the signals at δ = 3.36 (d) and δ = 3.50 (d) ppm (chemical shifts in CDCl₃).

Cross metathesis of eugenol acetate (10) and cis-1,4-dichloro-2-butene (12)

Under a nitrogen atmosphere, 10 (50.0 mg, 0.242 mmol) and 12 were introduced in a vial containing DCM (1.0 mL). Subsequently, the catalyst (2.5 mol%) was added, and the reaction was heated at 40 °C for 12 h. The desired product 13 was obtained after purification by flash chromatography (hexane : ethyl acetate, 9 : 1). E/Z ratio of 13 was determined integrating the signals at δ = 4.05 (d) and δ = 4.18 (d) ppm (chemical shifts in CDCl₃).

Ethenolysis of 14

In an autoclave, 14 (1.68 g, 5.4 mmol, 1 equiv.) and dodecane (150 µL) were introduced under a nitrogen atmosphere. At t₀, a sample was prepared using hexane (1.0 mL), followed by the addition of a toluene solution containing the catalyst (50 or 500 ppm). The autoclave was purged with ethylene three times and then pressurized to 150 psi. The reaction mixture was stirred at temperatures of 50 °C or 70 °C for 3 hours, afterwards it was cooled in an ice bath and quenched with ethyl vinyl ether. Subsequently, GC samples were prepared in hexane and stored at −20 °C until GC analysis could be performed.

Ring opening metathesis polymerisations

Alternating ROMP of NBE and COE

Under a nitrogen atmosphere, NBE (122 mg, 1.3 mmol) and COE (143 mg, 168 µL, 1.3 mmol) were dissolved in methylene chloride (2.5 mL). The reaction mixture was heated at 30 °C and subsequently methylene chloride solution of catalyst (1.3 µmol) was injected. The polymerisation was quenched by ethyl vinyl ether (∼1 mL), the polymer obtained was coagulated in methanol, recovered via filtration and dried under vacuum.

Alternating ROMP of ONC and DHF

ONC (100 mg, 0.47 mmol, 120 equiv.) and DHF (36 µL, 0.0.047 mmol, 120 equiv. or 72 µL, 0.0084 mmol, 240 equiv.) were dissolved in dry THF (2.0 mL). Subsequently, a solution of catalyst (0.8 mol%.) in dry THF (1.0 mL) was injected to the monomer solution. The reaction mixture was stirred for 3 or 12 h at room temperature. The polymerisation was quenched with excess ethyl vinyl ether (40 µL, 0.43 mmol, 110 equiv.) and precipitated into cold methanol to obtain the alternating copolymer.

ROMP of ONC

Under a nitrogen atmosphere, ONC (100 mg, 0.47 mmol, 120 equiv.) was dissolved in THF (2.0 mL). Separately, a solution containing catalyst HGII (2.5 mg, 0.3.9 µmol, 1.0 equiv.) and THF (1.0 mL) was prepared, and then quickly added to the monomer solution. The reaction mixture was stirred for 3 h at room temperature. The polymerisation was quenched with excess ethyl vinyl ether (840 µL, 0.43 mmol, 110 equiv.) and precipitated into cold methanol to obtain the homopolymer.

ROMP of DHF

For the synthesis of Poly(DHF), HGII (2 mg, 3.2 µmol, 1.0 equiv.) was quickly added to the degassed DHF (0.48 mL, 6.4 mmol, 2000 equiv.). The reaction mixture was stirred for 4 h at room temperature, then cold methanol was added. The resulting homopolymer was further analyzed by NMR analysis.

Conclusions

New Hoveyda–Grubbs-type complexes containing backbone-substituted N-benzyl, N′-aryl NHC ligands of increasing steric demand were obtained and fully characterised. Their catalytic potential was explored in standard olefin metathesis reactions, as well as in more challenging transformations involving renewable substrates, and compared with that of the commercially available symmetrical HGII catalyst. In RCM and CM reactions involving model substrates, the newly developed systems were found to be less efficient than HGII. The N-aryl substitution strongly affected catalyst activity, with complexes bearing less encumbered N-mesityl groups showing higher activity than those containing N-2,6-diisopropylphenyl groups. Regarding the NHC backbone configuration, anti catalysts performed better than their syn counterparts. In contrast, in the ROMP copolymerisation of simple cycloolefins with different reactivities, such as NBE and COE, the newly developed systems confirmed the superiority of unsymmetrical NHC catalysts in promoting alternating copolymerisation, with less hindered N-mesityl catalysts exhibiting higher chemoselectivity. Interestingly, in more demanding reactions involving renewable substrates, such as self-metathesis of eugenol acetate and ethenolysis of ethyl oleate, all complexes outperformed HGII. In particular, the more sterically demanding N-2,6-diisopropylphenyl catalysts proved to be the most effective in both transformations. Importantly, in ethenolysis performed under industrially relevant conditions, the influence of NHC backbone configuration became significant, as the anti N-2,6-diisopropylphenyl catalyst provided a TON of 12 920, representing, to the best of our knowledge, the highest value reported for unsymmetrical NHC systems.

Finally, in the synthesis of degradable polymers from bio-sourced monomers (ONC and DHF) via alternating ROMP, the syn N-2,6-diisopropylphenyl catalyst afforded copolymers with a high degree of alternation, even using equimolar amounts of comonomers, similarly to HGII. Notably, higher cis-selectivity with respect to HGII was observed (76–78% vs. 21–27% of Z olefin content, respectively).

Overall, these results further support the notion that Ru catalysts bearing unsymmetrical NHC ligands act as specialised systems for metathesis transformations, highlighting their potential in demanding and industrially relevant processes. At the same time, the present findings suggest that, as commonly observed, even within this specific class of unsymmetrical ruthenium catalysts, no single uNHC–ruthenium complex proved optimal for all investigated metathesis reactions. Importantly, the key structural features identified in this study, supported by theoretical investigations, may provide useful guidelines for catalyst selection and for the design of more efficient catalytic systems.

Author contributions

Rida Nadeem: investigation, methodology, data curation, formal analysis. Rubina Troiano: investigation, data curation, formal analysis, visualization, writing – review & editing. Assunta D'Amato, data curation, formal analysis, visualization, writing – review & editing. Consiglia Tedesco: investigation, data curation, formal analysis, writing – original draft. Fabia Grisi: conceptualization, investigation, data curation, formal analysis, validation, visualization, supervision, writing – original draft, writing – review & editing.

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting this article have been included as part of the supplementary information (SI). Supplementary information: experimental procedures for the synthesis of diamines, NMR characterization of diamines, NHC salts and catalysts, characterization data for polymers. See DOI: https://doi.org/10.1039/d6ra03171d.

CCDC 2543718 and 2543719 contain the supplementary crystallographic data for this paper.^88a,b