Catalysis of linear alkene metathesis by Grubbs-type ruthenium alkylidene complexes containing hemilabile α,α-diphenyl-(monosubstituted-pyridin-2-yl)methanolato ligands

Four new Grubbs-type precatalysts [RuCl(H2IMes)(O^N)(=CHPh)], where [O^N = α,α-diphenyl-(3-methylpyridin-2-yl)methanolato, α,α-diphenyl-(4-methylpyridin-2-yl)methanolato, α,α-diphenyl-(5-methylpyridin-2-yl)methanolato and α,α-diphenyl-(3-methoxypyridin-2-yl)methanolato] were synthesized and tested for their activity, stability and selectivity in the 1-octene metathesis reaction. Overall the precatalysts showed good activity and high stability for the metathesis of 1-octene at temperatures above 80 °C and up to 110 °C. Selectivities towards the primary metathesis products, i.e., 7-tetradecene and ethene, above 85% were obtained with all the precatalysts at 80 and 90 °C. High selectivities were also observed at 100 °C for the 4-Me- and 3-OMe-substituted precatalysts. With an increase in temperature an increase in isomerisation products and secondary metathesis products were observed with the latter reaching values >20% for the 3-OMe- and 3-Me-substituted precatalysts at 110 and 100 °C, respectively. All the precatalysts exhibits first-order kinetics at 80 °C with the 3-substituted precatalysts the slowest. The behaviour of the 3-substituted precatalysts can be attributed to electronic and steric effects associated with the adjacent bulky phenyl groups.


Introduction
The alkene metathesis reaction is now well established as a powerful synthetic tool in organic and polymer chemistry [1,2]. The development of metal alkylidene precatalysts based on ruthenium, starting with the so-called Grubbs 1 (1) and 2 (2) metal carbenes, played a major role to extend the versatility of the reaction including the application of these in industrial processes ( Figure 1). Of course, the role of the so-called Schrock metal carbenes based on tungsten and molybdenum should not be ignored in the success story of the alkene metathesis reaction but it is not the focus of this article.  The large number of ruthenium alkylidene precatalysts that has been developed is based on the design concepts illustrated in Scheme 1 [3]. The design concept C is of interest because of the potential hemilabile nature and latent metathesis activity of these complexes [4]. Of particular interest to us are the ruthenium alkylidene complexes containing the pyridinyl alcoholato bidentate ligands investigated by a number of research groups [5]. Scheme 1: Design concepts for ruthenium alkylidene precatalysts [3].
We investigated a number of Grubbs 1-and  metal carbenes with pyridinyl alcoholato ligands for the 1-octene metathesis reaction ( Figure 3) [10][11][12][13][14]. The incorporation of pyridinyl-alcoholato ligands in the Grubbs-type precatalysts has shown an increase in the thermal stability, activity and lifetime of the precatalysts when compared to 1 and 2 [10]. The pyridinyl-alcoholato Grubbs 2-types exhibited higher activities and selectivities than the Grubbs 1-types and were investigated in more detail. It is clear from the results that the chelating ability of the pyridinyl alcoholato ligands combined with the NHC ligand is responsible for the activity and improved stability of the precatalyst at high temperatures. In general 5d performed the best in the 1-octene metathesis reactions when compared to complexes 5a-c and 5e-h. The catalytic perfor-mance could be further tuned by the incorporation of an electron-donating (e.g., OMe, 5k) or electron-withdrawing (e.g. Cl, 5i and 5j) group at the 2-or 4-position of one of the α-phenyl groups of 5d [14]. At 80-110 °C these complexes showed improved catalytic performance in the metathesis of oct-1-ene. At 110 °C complex 5k, with 96% conversion and 95% selectivity towards the primary metathesis products tetradec-7-ene and ethene, outperformed the other complexes. In a computational study the improved catalytic performance was attributed to strengthening of the Ru-N bond due to steric repulsion between the substituted phenyl group and the NHC ligand [14]. An 8-quinolinolate Grubbs 2-type derivative, patented by Slugovc and Wappel [15] for use in ROMP reactions, was found to be inactive (<1% conversion) for 1-octene metathesis at 60 °C [12].
Schachner et al. [16] evaluated the catalytic activity of 5b, 5d and related complexes for the ROMP of cyclooctene, CM of hex-5-enyl acetate with dec-5-ene and the RCM of hex-5-en-1yl undec-10-enoate. Superior (CM, RCM) to moderate (ROMP) activities were observed for most of these precatalysts. An interesting result was the very high affinity ("stickiness") to untreated, unmodified and commercially available chromatography-grade silica. This was exploited further by Cabrera et al. [17,18] when 5b and related complexes were investigated as heterogeneous precatalysts in biphasic RO-RCM and CM reactions. The substrate and catalyst were adsorbed on a thin layer silica plate and developed in EtOAc/hexane (1:7 v/v) for the CM of methyl 9-dodecene and in hexane for the RO-RCM of cis-cyclooctene.
The above-mentioned studies clearly illustrate the versatility and use of ruthenium alkylidene complexes with pyridinyl-alcoholato ligands. In principle these studies had one approach in common concerning the pyridinyl-alcoholato ligand, and that was to focus on substituents on the α-carbon of the ligand. To our knowledge, there are no reports on investigations of electronic and/or steric effect(s) of pyridinyl substituents on the chelation efficiency of pyridinyl alcoholato ligands, and subsequently its metathesis activity. Therefore, in this paper, we investigated the influence of a monosubstituent on the pyridinyl moiety on the 1-octene metathesis activity of a Grubbs 2-type precatalyst with an α,α-diphenyl methanolato ligand. For the synthesis of the pyridinyl methanol compounds, commercially available substituted bromopyridines were reacted with benzophenone followed by a reaction of the lithiated alcohol with 2.
Four new ruthenium alkylidene complexes, i.e., 6-9 (Figure 4), were successfully obtained and investigated as precatalyst in 1-octene metathesis in the temperature range 40-110 °C. The stability, selectivity and turnover frequency (TOF) of 2 increased upon substituting Me and OMe groups on the various positions of the pyridine ring of the pyridinyl-alcoholato ligands at high temperatures (80-110 °C). The increase in stability is attributed to the electronic and steric influence of the Me and OMe groups on Ru-N chelation. The activity of the precatalysts also showed a significant improvement upon increasing the reaction temperature from 40 to 110 °C. The increase in the activity of the precatalysts is relatively low in the 40-60 °C range, but a high activity difference is observed upon increasing the temperature in 10 °C intervals between 70 and 110 °C.

Results and Discussion
A mixture of products, summarised in Table 1, is obtained during the metathesis of 1-octene, i.e., primary metathesis products (PMPs), isomerisation products (IPs) and secondary me-  tathesis products (SMPs). The PMPs, 7-tetradecene (cis and trans) and ethene, forms as a result of the self-metathesis (SM) of 1-octene. Simultaneously 1-octene is isomerised to 2-, 3-and 4-octene (IPs). The subsequent SM and CM reactions of the internal alkenes yield alkenes (cis and trans) in the C 3 -C 13 range (SMPs).
All the reactions were followed by GC at regular sampling intervals until 540 min. Because the observed formation of IPs is mostly below 2% and never above 4% it is also not shown in the figures.

Effect of the reaction temperature
The results of the metathesis of 1-octene at temperatures 40-100 °C are presented in Figure 5 and Table 2   The selectivity towards PMPs showed a dramatic increase upon increasing the temperature from 40 to 80 °C (23-97%); however, it showed a decrease going from 80 (97%) to 90°C (89%),  and then to 100 °C (87%). An overall assessment of the results show that at 80 °C the catalyst showed a high selectivity for PMPs with a negligible amount of SMPs and IPs. Although the activity of the precatalyst increased a great deal at 90 and 100 °C, the selectivity for PMPs decreased as a result of the high amount of SMPs and IPs formation. The TOF increased significantly as a result of increasing the temperature. The highest TOF increase was observed upon increasing the temperature from 80 to 90 °C. Generally, precatalyst 7 showed very good activity, selectivity and stability at high temperatures.
The results of the metathesis of 1-octene at temperatures 40-100 °C are presented in Figure 6 and Table 3 for precatalyst 8. A similar overall trend for 8 is observed, i.e., very low reaction rates at temperatures below 60 °C with a rapid increase in reaction rates above 70 °C resulting in 1-octene conversions above ca. 70% after 540 min. Although the formation of PMPs equilibrated quickly at ca. 70% after ca. 150 min at 100 °C and at ca. 65% only after ca. 400 min for 90 °C it did not equilibrate at 80 °C even after 540 min.
The formation of SMPs is very low (below 4%) for 8 in the temperature range 40-80 °C after 540 min, while it is relatively high at 14 and 21% at 90 and 100 °C, respectively. In the same period the formation of IPs remained below 3% even at the high temperatures. At 100 °C and 540 min, a larger amount of SMPs is formed for precatalyst 8 than that of 7. Table 3 summarises the overall catalytic performance of precatalyst 8 at 420 min. The PMPs and SMPs formation, TON and TOF all show a direct relationship with temperature. Precatalysts 7 and 8 share similarities in having the same temperature range for the highest PMPs formation, i.e., 70 to 80 °C at 420 min. The biggest difference, however, is observed for 8 (37%). Relatively higher SMPs are formed for 8 (11%, 19%) than that of 7 (8%, 11%) at 90 and 100 °C, respectively. The relatively low PMPs formation of 8 compared to that of 7 is due to the relatively high SMPs and IPs formations with precatalyst 8. The IPs formation in 8 follows a similar pattern to that of 7, i.e., it showed an increase upon increasing the temperature from 40 to 50 °C, followed by a decrease from 50 to 60 °C and then an increase from 60 to 100 °C. The selectivity in 8 increased upon increasing the temperature from 40 to 70 °C, and then showed a decrease from 70 to 100 °C.
Summarising the comparisons of precatalysts 7 and 8, it is noted that precatalyst 7 showed better activity, selectivity and stability in the 60-100 °C temperature range, except for 80 °C, at 420 min. It also showed higher TOF at 60, 90 and 100 °C at 420 min. According to a DFT study by Getty et al. [19] the more positively charged the Ru, the slower the initiation rate of the catalyst. The calculated Mulliken atomic charge of Ru in 7 (0.934) is less positive than in 8 (0.976).
The results of the metathesis of 1-octene at temperatures of 60 to 110 °C are presented in Figure 7 and Figure 8 for precatalysts 6 and 9, respectively. Because of their low activity and high stability, the metathesis reactions were done between 60 and 110 °C. Metathesis of 1-octene by the 3-Me-substituted precatalyst 6 showed an increase in the activity of the precatalyst upon increasing the temperature from 60 to 110 °C. A large increase in the rate of the metathesis reaction is observed upon increasing the temperature from 80 to 90 °C. Although the activity of the precatalyst has shown an increase upon increasing the temperature from 60 to 110 °C, a very high (ca. 45%) increase in the PMPs formation is observed upon increasing the temperature from 80 to 90 °C at 540 min. The PMPs formation did not equilibrate at 90 °C and this shows the stability of precatalyst 6 at high temperatures.   The selectivity toward PMPs (67.2%) and the SMPs (28.8%) are relatively high at 110 °C. The TOF is also directly related to the reaction temperature. The TOF of precatalyst 6 are generally lower than for precatalysts 7 and 8. This, therefore, shows the relatively high stability of precatalyst 6 compared to those of 7 and 8. As a result of having a more positive Ru charge, precatalyst 6 showed a low initiation rate. This is also in agreement with the DFT study of Getty et al. [19], i.e., precatalyst 6 (0.988) has more positive Mulliken's atomic charge on Ru than both 7 (0.934) and 8 (0.976). Its high activity at 110 °C with 69% selectivity is, however, remarkable for linear alkene metathesis catalysed by ruthenium alkylidene precatalysts.
The 3-OMe-substituted precatalyst 9 showed a negligible activity for the metathesis of 1-octene at 60 °C (Figure 8), similar to the 3-Me-substituted precatalyst 6. The overall activity of the precatalyst, however, showed a significant increase upon increasing the temperature from 60 to 110 °C. In a similar way to that of 6, the largest increase in the activity of the precatalyst is observed upon increasing the temperature from 80 to 90 °C (ca. 38%) at 420 min. The activity of the catalyst showed a small difference between 100 and 110 °C, on the overall metathesis reaction.
During the course of PMPs formation, high catalytic activity for 9 is observed within 200 min at temperatures above 90 °C (ca. 60%), while the activity of the precatalyst showed a dramatic increase from 70 to 90 °C after ca. 500 min, similar to that of 6 (see Figure 7b). For both 6 and 9 the highest PMPs (>60%) is observed from 90 to 110 °C after 420 min. The rate of formation of SMPs is very high for 9 at 110 °C within 420 min. This is the reason for the decrease in the formation of PMPs from 71% at 100 °C to 64.2% at 110 °C. Table 5 presents the overall catalytic performance of precatalyst 9 at 420 min. Firstly, PMPs formation increased from 0.5 to 71% when increasing the temperature from 60 to 100 °C; however, from 100 to 110 °C the PMPs yield decreased from 71 to 64.2%. The reason for this is the formation of a very large amount of SMPs (23.2%) and IPs (2.6%), which is, more than twice the amount at 100 °C. Similarly, the selectivity and TOF showed a decrease when going from 100 to 110 °C.
The selectivity showed a dramatic (50%) increase upon increasing the temperature from 60 (43%) to 80 °C (93%) and then begins to decrease to 91% (at 90 °C), 85% (at 100 °C) and finally to 71% (at 110 °C). Although the catalyst showed signif-  icant stability and very good (71%) selectivity at 110 °C, it is advisable not to go beyond 100 °C, as the formation of SMPs and IPs doubled that will affect the overall PMPs yield. It is also worthwhile to note the decrease in the turnover frequency at 110 °C.
A general comparison of the overall performance of the precatalysts, in terms of PMPs, SMPs, IPs, selectivity, TON and TOF, exhibits the decreasing order of 7 > 9 > 8 > 6 at 60, 90 and 100 °C. The order, however, changes at 80 °C to 8 > 7 > 9 > 6 and at 70 °C to 7 8 > 9 > 6. In all cases, the small amounts of SMPs and IPs are positive for the application of these systems at higher temperatures. Overall precatalyst 7 performed the best at all temperatures (except at 80 °C). In an attempt to understand the significance of these results, DFT calculations were performed on the precatalysts.
Precatalyst 6 showed the lowest activity of all precatalysts in the specified temperature ranges. It is also worthwhile to note that increasing the reaction temperature showed a significant increase in the activity of 6. Precatalyst 9, on the other hand, showed better performance at high temperatures (≥70 °C) compared to 6. This may be explained by the longer Ru-N bond (2.181 Å) in the geometry-optimised structure (Figure 9) of precatalyst 9 compared to that of the Ru-N bond (2.166 Å) of 6.
A longer bond suggests a weaker Ru-N chelation thus a more active hemilabile complex. The difference in the Ru-N bond length may be attributed to the electron-withdrawing inductive effect of the OMe group making the Ru-N chelation weaker. Furthermore, a type of orbital interaction between the oxygen of the 3-OMe group and the two α-phenyl rings, i.e., an oxygen lone pair-aromatic π interaction illustrated in Figure 10, may add to the inductive effect. The longer Ru-O bond (2.031 Å), shorter C α -O bond (1.420 Å) and C α -C 2 bond (1.541 Å) observed in precatalyst 9 when compared to the corresponding bonds in 6, i.e., 2.028, 1.425 and 1.544 Å, respectively, supports such a premise. It may also be a plausible explanation for the envelope geometry of the five-membered ruthenacycle. In addition, the relatively low ruthenium metal positive charge on 9 would cause it to have a high initiation rate constant [19]. On the other hand, the 3-Me group in 6 will strengthen the Ru-N chelation via inductive electron-donation and steric repulsion between the methyl group and the two phenyl rings. As a result of the steric interaction 6 has a planar five-membered ruthenacycle geometry (Figure 9). In the absence of substituents on the pyridinyl moiety it is expected that the resulting precatalyst will be more active at lower temperatures. This is indeed the case when 5d is used as catalyst.
As we have discussed earlier, the 4-Me-substituted precatalyst 7 has shown better catalytic performance in all temperatures under investigation except at 80 °C. The reason for this is that the Me group is, relatively speaking, further removed from the pyridine nitrogen so that the inductive electron-donation by the methyl group cannot significantly influence the electron density on the pyridine nitrogen. There is also no steric effect that would interfere with the Ru-N bond strength. The strengthening effect on the Ru-N chelation would, therefore, possibly be low compared to the other precatalysts.
If this is a plausible explanation for the relatively better performance of the 4-Me-substituted precatalyst 7, one might ask what about the difference between the 3-Me-, 6, and 5-Mesubstituted, 8, precatalysts that are at the same distance from the pyridine nitrogen? In the optimised structure of 6, the Me group is in a crowded environment due to its proximity to the two α-phenyl groups, which upon opening the Ru-N chelation, would even become more sterically crowded. This results in a planar geometry of the five-membered ruthenacycle while 8 exhibits an envelope geometry. The Ru-N (2.179 Å) bond length in 8 is longer and the C α -O (1.417 Å) and C α -C 2 (1.532 Å) bonds are shorter than the corresponding bonds in 6.
In order to overcome the combined effect of the resistance that resulted from the steric crowdedness and the inductive electrondonation by the 3-Me group and open the strong Ru-N chelation, it needs relatively high energy. In 8 the methyl group is in exactly the opposite orientation to the two α-phenyl groups. Therefore, the steric crowdedness that is observed in 6 that will lead to steric resistance to open the Ru-N chelation does not exist. Thus 8 is more susceptible to hemilability than 6 and exhibits higher activity. Therefore, for 4-Me-and 5-Me-subtstituted precatalysts, only the inductive electron-donation effect of the methyl group is the reason for the increased stability. In the 3-Me-and 3-OMe-substituted precatalysts, however, the steric effect and orbital interactions work towards the stability of the precatalyst in addition to inductive effects.

Stability of precatalysts
In previous studies [11,12,14] we investigated the stability of pyridinyl-alcoholate Grubbs-type precatalysts as seen in the improved catalytic lifetimes of these complexes. Plots of ln([starting material]) versus time, proposed by Grubbs and co-workers [20], were used as a measure of the stability of the precatalyst, i.e., a linear plot indicates a reaction with pseudofirst order rate kinetics, while a curved plot points towards catalyst decomposition. We used the conversion of 1-octene at a Ru/1-octene molar ratio of 1:9000 and a reaction temperature of 80 °C to compare the stability of 5d with that of 5i, 5j and 5k [14]. In Figure 11 the literature data (% 1-octene conversion and ln(% 1-octene)) of 5d is compared with that of precatalysts 6 -9 at a Ru/1-octene molar ratio of 1:9000 and a reaction temperature of 80 °C over 540 min.
The overall activity order of the catalysts follows the order 5d > 8 > 7 > 9 > 6 up to ca. 540 min. The order 8 > 5d 7 > 9 > 6 is observed for both the overall metathesis and the PMPs formation. All the precatalysts exhibits first-order kinetics over the first ca 540 min when the ln(% 1-octene) plots (Figure 11b) are considered. The substituted precatalysts show better stability than 5d, thus longer lifetimes, with 6 and 9 the slowest and 8 close to but slower than 5d.
It is interesting to note that the stability of 5j and 5k correlates very well with that of 7, while 5i is more stable than 7 but less than 9 (comparison of current results with results in [14]). This clearly indicates that a substituent on one of the α-phenyl groups or the pyridinyl moiety has a stabilising effect on the corresponding precatalyst with a substituent on the 3-position (6 and 9) of the pyridinyl rendering the precatalyst the most stable. The latter two is also active at higher temperatures. Table 6 presents the overall catalytic performance of precatalysts 5d, and 6-9 at a Ru/1-octene molar ratio of 1:9000, 80 °C   and 420 min. According to these results precatalyst 5d shows the highest PMPs, TON and TOF. Although it has relatively high SMPs compared to most of the precatalysts, its overall performace prevails over the other precatalysts. The second best performance was observed for 8, as it resulted in relatively high PMPs, TON and TOF compared to the rest of the precatalysts, although its SMPs ranks as first. The rest of the precatalysts can be ranked in a decreasing order of activity of 7 > 9 > 6. It is clear from the data in Table 6 that the unsubstituted precatalyst 5d is more active compared to the substituted precatalysts at 420 min. This will only be due to the substituent effect on the activity of the precatalyst.

Effect of catalyst concentration
Earlier studies indicated that 80 °C is the optimum temperature for 5d [10,11]. It was therefor decided to investigate the effect of the concentration of the precatalyst on the metathesis of 1-octene at 80 °C. Precatalyst 8 was chosen for this investigation at Ru/1-octene molar ratios of 1:6000, 1:9000, 1:10000 and 1:15000. Table 7 presents the overall catalytic performance of precatalyst 8 at different Ru/1-octene molar ratios, 80 °C and 420 min. With a decrease in precatalyst concentration a direct relationship was observed with the conversion of 1-octene and PMPs, they all decreased, while the TON and TOF increased. The SMPs and IPs did not follow a specific trend while the selectivity remained the same, i.e., 92%, at all the concentrations.

H NMR investigation of precatalyst 7 and 5d
Proton nuclear magnetic resonance spectrometry ( 1 H NMR) is a powerful tool to study ruthenium alkylidene complexes and was used to study 1-octene metathesis in the presence of 1 and 2 [10,21,22]. Cl 2 ) could be clearly distinguished using the carbene-H α signals; they appeared as a singlet, triplet and singlet in the δ 18.5-20.2 ppm region, respectively. We also investigated 5a and observed five carbene-H α signals attributing three to the alkylidene species when the pyridinyl-alcoholato ligand was in the "closed" (coordinated) position; δ CHPh 18.05 ppm, δ CHHx 16.71 ppm and δCHH 16.08 ppm [10]. The other two was attributed to the benzylidene (δ CHPh 19.48 ppm) and methylidene (δ CHH 19.76 ppm) species in the "open" (uncoordinated) position with the uncoordinated heptylidene signal not appearing probably due to the fast reaction of this species. Four signals at δ 9.48 ppm (for the coordinated ligand), 9.05 ppm, 9.22 ppm, and 9.71 ppm attributed to the H α signals of the pyridine ring were also observed. The latter three signals overlapped too much to be useful.
We performed a 1 H NMR investigation of the metathesis of 1-octene by precatalyst 7 in the temperature region 60-90 °C in order to gain some insight into the reaction mechanism. The carbene-H α 1 H NMR signals at 90 °C over a period of 345 min are presented in Figure 12. Three signals attributed to the benzylidene (δ 17.33 ppm, singlet), heptylidene (δ 16.85 ppm, triplet) and methylidene (δ 15.68 ppm) were observed. A small signal at δ 16.66 ppm appeared at 270 min and was not assigned (inter alia multiplicity not discernable). A different development of carbene signals over time is observed than what was reported before for 5a [10], i.e., the methylidene signal starts to appear at 12 min while the heptylidene signal only starts to appear at 165 min. The benzylidene signal rapidly declines after 194 min and is not observed at 345 min. No clear indication of an "open" or "closed" complex was observed, so it assumed that the signals represent the "closed" species. It can be concluded that the alkylidene species of the pyridinyl-alcoholato Grubbs 2-type precatalysts are quite stable at high temperatures explaining the activity of these precatalyst at high temperatures and the slow rate of disappearance/formation of these signals confirms the longer lifetimes observed in the catalytic reactions.  The H α pyridine ring 1 H NMR signals at 90 °C at 345 min are presented in Figure 13. Five H α signals of the pyridine ring that are not observed at the beginning of the reaction were observed at δ 9.57 ppm (doublet), δ 9.22 ppm (doublet), δ 9.08 ppm (unknown multiplicity), δ 8.91 ppm (doublet) and δ 8.85 ppm (doublet). The signal at δ 9.70 ppm (singlet) was the only signal observed at the beginning of the reaction. These signals is probably due to the "open" and "closed" pyridinyl-alcoholato ligands of alkylidene species and a possible assignment is shown in Figure 13. Further research is required to gain a comprehensive understanding of the operation of the active species of the pyridinyl-alcoholato ruthenium alkylidene precatalysts.
Effect of solvent on 1-octene metathesis using precatalyst 7 Because toluene-d 8 was used in the 1 H NMR study it was decided to investigate if toluene as solvent has any effect on the 1-octene metathesis reaction using precatalyst 7. Results of this investigation are presented in Table 8.
An increase in SMPs formation is the only difference that was observed when toluene was used as solvent with an 8.9% in-crease at 420 min. This affected the other performance indicators, i.e., PMPs, S, TON and TOF; lower values than the neat reactions were obtained. The results suggest that no significant solvent effect appears to exist .However, the increase in SMPs (associated with an increase in IPs) indicates decomposition of the precatalyst to active isomerisation species, probably metal hydride species. In our NMR study no indication of the existence of metal hydride species was found.

Conclusion
The aim of our research is to control the Ru-N bond strength of the bidentate hemilabile pyridinyl-alcoholato ligands in precatalyst 5d in an attempt to synthesise a precatalyst with high performance for linear alkene metathesis at high temperatures. To reach this aim, we synthesised ruthenium alkylidene precatalysts by substituting one of the hydrogens of the pyridine ring of the bidentate pyridinyl-alcoholato ligand by Me and OMe groups. We synthesised the 3-, 4-, and 5-methyl and 3-methoxysubstituted 5d precatalysts.

Experimental procedures
Precatalyst synthesis: The well-established methods of Herrmann et al. [23] and Van Der Schaaf et al. [7] were used to synthesize precatalysts 6-9. This is illustrated in Scheme 2.

Metathesis reactions:
The metathesis reactions were carried out in 5 mL small-scale glass reactors. The reactor containing a small magnetic bar was flushed with nitrogen and an appropriate amount of precatalyst added by weighing. Once again, the contents of the reactor were carefully flushed with nitrogen and the reactor was sealed. The sealed reactor was placed in an aluminium block on a magnetic stirrer. The temperature was set to the desired temperature and allowed to stabilise prior to the reactor being placed in the block. The temperature was regulated throughout the reaction using a temperature controller fitted with a thermocouple. After one minute of heating nonane (0.25 mL) was added via gastight syringe (1 mL) as an internal standard, followed by the addition of 1-octene (5 mL) via gastight syringe (5 mL). Samples (0.1 mL) were withdrawn at time intervals for ca. 520 min with a gastight syringe (1 mL), transferred to a GC vial (1 mL), quenched with toluene (0.3 mL) and tert-butyl hydrogen peroxide (2 drops), and then injected into a GC/FID by auto sampler. The metathesis reaction was terminated after 1440 minutes and analysed by GC/FID. Some samples were also analysed by GC/MSD. Each experiment was repeated at least three times.

H NMR investigation of metathesis reaction:
An NMR tube was placed in a Schlenk tube, evacuated with a vacuum pump and then flushed with a stream of argon. The same procedure was repeated and then 12 mg (0.015 mmol) of precatalyst 7 was added to the NMR tube. Once again, the contents of the NMR tube were flushed with argon and toluene-d 8 (0.65 mL) added. The catalyst was dissolved by shaking the contents and 1-octene (0.1 mL, 0.64 mmol) added immediately before putting the tube in the spectrometer for temperature ranges 30-50 °C. 1 H NMR spectra were recorded at 5-6 minute intervals for 5-8.5 h. For temperature ranges 60-90 °C, 1 H NMR of the precatalyst was done alone before adding the 1-octene. The precatalyst (11.5 mg, 0.014 mmol) and anthracene (5.2 mg, 0.03 mmol) were mixed in the metathesis reaction where anthracene was used as an internal standard.