Development of an In Situ Protocol for the Intramolecular Olefination of Oximes

Abstract

We herein report the development of a protocol for the olefination of oximes, resulting in cyclic alkene products. Our approach relies on the in situ formation of Ru alkylidenes from [RuCl₂(p-cymene)]₂ and 3,3-diphenylcyclopropene as carbene precursors and gives rise to the desired functionalized alkenes in yields up to 58%. Compared to our previously reported strategy for oxime olefination, it foregoes the use of commercially available Ru alkylidenes, including Hoveyda–Grubbs 2. This is desirable concerning the overall costs of this transformation and also provides important mechanistic insights into the development of future catalytic variants of this reaction as the alkylidenes required for efficient olefination can be generated in situ.

Graphical Abstract

Development of new olefin-forming reactions can provide unique entry points into pharmaceuticals, (1–3) natural products, (4,5) or alternatively as functional handles for subsequent reactivity. (6) The discovery and development of the olefin metathesis reaction promoted by transition metal alkylidenes enabled new synthetic disconnections and led to wide applications in academic and industrial settings. (6–8) Attempts to exploit metal alkylidene cross-metathesis reactivity between olefins and carbonyls (9,10) or imines (11–14) remain less developed due to the inherent challenges of post-metathesis turnover of strong M═O/M═NR bonds. (13,15) We recently reported the use of the air-stable, commercially available Hoveyda–Grubbs complex (17) (HG2, 2) for ring-closing olefination of olefins with hydrazones and oximes (Figure 1A). (18) This approach is based on the hypothesis that tuning of the polarity of the oxime or the hydrazone substituent enables desirable reactivity of the C═NX bond while retaining the olefin reactivity with Ru alkylidenes. (18) However, broad adoption of this synthetic strategy may face challenges associated with the requirement of stoichiometric amounts of HG2 (2) and the steep-associated costs ($290/g), (19) which make a catalytic protocol highly desirable. Toward this goal, previous efforts showed that the reaction does not result in the formation of M═NR species (Figure 1B) but instead forms trans-RuCl₂(dppe)₂ upon addition of excess 1,2-bis(diphenylphosphino)ethane (dppe). (18) Therefore, we postulate that the development of a catalytic protocol for the olefination of oximes and hydrazones will require the reformation of a ruthenium alkylidene from a Ru(II) species under the optimal reaction conditions. We herein report the development of a new protocol for the olefination of oximes that relies on the in situ generation of ruthenium alkylidenes from Ru(II) sources and expect that these results will spur further advances toward the ultimate goal of a catalytic procedure for the olefination of oximes and hydrazones (Figure 1C). This approach is the first example of successful oxime olefination relying on in situ-generated Ru-alkylidenes and provides access to desired cyclic alkene products in up to 58% yield. Furthermore, in situ oxime olefination is estimated to be more than 2.45 times less expensive than the use of commercial HG2 (2). (22,23).

Figure 1.

(A) Oxime olefination mediated by discrete Ru alkylidenes (Mes = 1,4,6-trimethylphenyl). (18) (B) Isolation of a known Ru(II) species after hydrazone olefination and possibility for a catalytic process. (C) This work: oxime olefination via in situ formation of Ru alkylidenes.

To develop a protocol for oxime olefination relying on the in situ generation of ruthenium alkylidenes, we were inspired by the >20 years of research in the formation of metathesis-active species for olefin metathesis from alkylidene-free Ru species. (20) Specifically, the phosphine-containing congeners of complex 8 have been demonstrated to have metathesis activity in the presence of a carbene precursor (20a–c) and utilized for ring-closing metathesis, (20a,c) ring-opening metathesis, (20a,c–e) and self-metathesis. (20f) Key studies in the generation of NHC-containing Ru species 10 by Grubbs, (20g) Noels, (20h) and Dixneuf (20i) for olefin metathesis demonstrated the viability of this approach for access to the analogous active species (20j) of the second generation of Ru-alkylidene catalysts. Recently, a report by Müller (21) systematically investigated in situ formation of Ru alkylidene species from carbene precursors and Ru(II) species for olefin metathesis relying on cyclopropenes and a suitable base (Figure 2A). To obtain proof of principle that ruthenium alkylidenes can indeed be regenerated in the olefination of oximes and hydrazones, (18) we added cyclopropene 7 as a known carbene precursor and PPh₃ as a reductant to our previously identified optimal reaction conditions relying on HG2 (2). Importantly, the desired olefination product 6 is formed with a 25% increase in yield and a TON of 1.25 based on HG (2) (Figure 2B), suggesting that an active ruthenium alkylidene capable of promoting the desired transformation is indeed regenerated.

Figure 2.

Literature precedent for in situ formation of Ru alkylidenes for olefin metathesis. (20,21).

Our subsequent efforts focused on identifying efficient reaction conditions for the in situ generation of ruthenium alkylidenes in the olefination of oximes. We first investigated various carbene precursors and NHC ligands building on the work reported by Müller and co-workers (21) (entries 1–4, Table 1). Specifically, di-Ph-cyclopropene 7 and Ph,Me-cyclopropene 13 provided the desired olefin product 6 in 6 and 2% yield, respectively, relying on SIPr·HCl as the imidazolinium salt (NHC·HCl). In comparison, using he alkylidene precursors, alkynes 14 and 15 (4-OMe) provided less than 1% yield of the olefin product under otherwise identical reaction conditions. Our general reaction setup relies on the exclusion of light due to a previous report by Day and Fogg, (24) which suggests that the SIMes-containing analogue of Ru complex 11 formed in situ is light-sensitive. Additionally, as cyclopropene 7 polymerizes in the presence of oxygen, reaction development was pursued under an inert atmosphere.

Table 1.

Evaluation of Carbene Precursors and NHC salts.

Conditions:

^aReactions were performed with 0.01 mmol [RuC1₂(p-cymene)]₂ di-mer, 0.02 mmol NHC·HCI, 0.02 mmol base, 0.06 mmol substrate 12, and 0.10 mmol alkylidene precursor in THE (0.01 M) at 60 °C in the dark under a N₂ atmosphere; a GC yield based on dodecane as an internal standard and reported relative to Ru(II).

Next, the imidazolinium salt SIMes·HCl was investigated due to its beneficial effect in previously established oxime olefination (18) relying on HG2 under otherwise identical conditions (entry 5, Table 1). This provided the desired olefin product in 9% yield (entry 5, Table 1). Evaluation of various bases for deprotonation of SIMes·HCl revealed that metal hexamethyldisilazide salts were generally superior to nBuLi and KOtBu, with the optimal result being LiHMDS, which provided 6 in 22% yield (entries 6–9, Table 1).

With a promising initial combination of a carbene precursor, NHC salt, and a base identified, we subsequently varied the ratio of NHC salt, base, and Ru(II) (Figure 3). This allowed investigation of the impact of differing amounts of SIMes carbene to Ru(II) on the yield of olefin product 6 to ultimately establish optimal reaction conditions.

Figure 3.

Identification of the optimum ratio of SIMes·HCl/LiHMDS/Ru(II).

Specifically, under conditions relying on a 1:1 molar ratio of SIMes·HCl/Ru(II) and varying amounts of LiHMDS, the yield of cyclopentene 6 follows this trend: 22% (1.00 equiv), 27% (2.00 equiv), and 17% (3.00 equiv). This suggests a maximum yield with 2 equiv of base, although further increased loading of base has a deleterious effect on olefination yield.

Under conditions relying on a 2:1 molar ratio of SIMes·HCl/Ru(II) and varying amounts of LiHMDS, the yields of cyclopentene 6 followed this trend: 45% (1.00 equiv), 44% (2.00 equiv), and 9% (3.00 equiv). This suggests that increased loading of SIMes·HCl (2-fold) leads to an ∼2-fold increase in olefination yield relative to conditions with equimolar loading (1:1:1 Ru(II)/SIMes·HCl/LiHMDS, 22% of 6).

Notably, under conditions relying on a 3:1 molar ratio of SIMes·HCl/Ru(II) and varying amounts of LiHMDS, the yields of cyclopentene 6 followed this trend: 25% (1.00 equiv), 58% (2.00 equiv), and 43% (3.00 equiv). Increased olefination yield with ≥2.00 equiv of LiHMDS suggests that at least 2.00 equiv of free SIMes carbene favored in situ formation of the Ru-alkylidene. Finally, the optimal ratio of SIMes·HCl/LiHMDS/Ru(II) was identified as 3:2:1, which provided product 6 in 58% yield. This result is consistent with the formation of the Ru-alkylidene from up to 2 equiv of SIMes·HCl and LiHMDS, while avoiding decreased product formation observed at increased LiHMDS loading.

With optimized conditions for in situ oxime olefination in hand, we next explored the substrate scope of this reaction for a variety of structurally diverse alkene products commonly used as benchmark systems in olefin metathesis reactions (Figure 4). (18,24–28) Olefination yields of substrates under in situ conditions (shown in blue) were compared to olefination with HG2 (2) (18) (run in benzene-d₆ with select examples in THF-d₈, shown in black, Figure 4). Notably, while benzoyl hydrazone 4 provides 82% of cyclopentene 6 when reacted with HG2 (2), no olefination was observed under in situ conditions. This interesting result is presumably due to deprotonation of the hydrazone N–H under the strongly basic reaction conditions. Based on this result, we focused on subsequent investigations on oxime substrates previously found to be productive in olefinations with HG2 (2). (18) Specifically, methyl- and benzyl-substituted oximes 16 and 17 led to the desired olefination product 6 in 31 and 27% yields, respectively (Figure 4). In comparison, oximes with more steric bulk containing tert-butyl and adamantyl substituents (18 and 12) resulted in increased yields of 49 and 58%, respectively.

Figure 4.

Exploration of scope for in situ olefination of oximes and hydrazones.

The corresponding one-carbon homologue of O-adamantyl oxime 9, 19, provided cyclohexene 20 in 41% yield under our optimal reaction conditions. This lower yield for the formation of a six-membered ring was consistent with the trend in olefination yield with HG2 (2) of the O-adamantyl oximes (olefination of 9 and 19) and the benzoyl hydrazones of our previous report. (18) Incorporation of α-substitution on the carbon backbone led to olefination with yields that were not synthetically useful (∼4%). This was observed for α-Ph-substituted oxime 31, which provided product 32 in 4% yield under in situ conditions compared to 43% of 32 with HG2 (2). Additionally, substrate 21 with incorporation of an aromatic ring into the backbone provided indene 22 in 4% yield, while olefination with HG2 (2) only provided 22 in 19% yield. These results suggest that olefination with an in situ generated alkylidene is sensitive to sterically bulky groups at the α-position of the oxime. Notably, incorporation of substitution at the β position in addition to esters (substrates 12 and 16–19) was well tolerated for in situ oxime olefination. The β-methyl-/phenyl-substituted oxime 33 and β-phenyl/phenyl oxime 35 provided 51% of 34 and 57% of 36, respectively. The increased olefination yields are presumably due to the influence of the “Thorpe-Ingold effect” (29) provided by the substituents at the β-position of oximes 33 and 35. Importantly, our optimal in situ olefination conditions tolerated suitably protected Lewis-basic amines in the substrate backbone. Specifically, the N-Ts protected amine 25 formed 47% of 26 under in situ conditions. The l-phenylalanine derived N-Ts protected adamantyl oxime 27 also formed cyclopentene 28 in 53% yield.

However, Boc-protected oxime 23 failed to form any of the desired product 24 under in situ conditions. This result suggests that the Boc group may be labile under strongly basic reaction conditions as recently reported by Tom et al. (30) and Mopatra et al. (31) Additionally, ether-containing oxime 29 provided none of the desired olefin 30 under either in situ or HG2 (2) conditions. (16) These results suggest that the in situ oxime olefination protocol is more sensitive to Lewis bases than olefination mediated by HG2 (2). (18) This sensitivity likely arises from the decreased stability toward Lewis bases for coordinatively unsaturated 14-electron Ru-alkylidene species 8 compared to coordinatively-saturated HG2 (2).

Our subsequent efforts focused on comparing the isolated yields of products obtained in our stoichiometric protocol (18) relying on HG2 (2) to our in situ protocol described herein. Notably, the isolated yields after chromatography with HG2 (2) and the in situ-generated Ru-alkylidene were 46 and 44%, respectively (Figure 5). These results are comparable to NMR yields obtained for 25 under the above conditions (Figure 4) and demonstrate that the additional reagents of the optimal conditions can be removed from the olefin product. This method tolerates utilization of various oximes (−OMe, −OBn, −O^tBu, and −OAd) with optimal performance from sterically bulky oximes (−OAd) to form six- and six-membered rings. However, in contrast to olefination with HG (2), (18) hydrazones do not undergo olefination. Substitution at the β-position of substrates (diPh, Me, and Ph) and the presence of Lewis-basic protected NTs amines were well tolerated under in situ olefination conditions (47–57% yield). Substrates where olefination was challenging were α-substituted with bulky groups (indenyl, −Ph, ∼4%), substrates with Lewis-basic ether backbones (0%), and −NBoc-protected substrates (0%). Despite being more substrate-specific, the method reported herein allows for the in situ generation of Ru-alkylidene species that result in the desired oxime olefination in up to 58% yield. Additionally, the modularity of this approach allowed investigation of transiently stable Ru alkylidenes for cross-olefination with oximes, which we expect to hold significance for the development of future catalytic variants and enable further advances.

Figure 5.

Isolated yield comparison for oxime olefination with HG2 (2) (18) and in situ-generated Ru-alkylidene.