Substituent Effect on the Formation and Their Reactivity of Platinum Carbenoids

Abstract

A propargylic ester containing a propargylic alkoxy group has been observed to preferentially undergo [1,2]-acyl shift over [1,3]-shift. In addition, the complementary and contrasting reactivity of vinyl vs. alkynyl platinum carbenoids has been discovered. Vinyl Pt-carbenoids are more prone to undergo [1,2]-H shift over addition to π-bonds whereas alkynyl Pt-carbenoids preferentially add to π-bonds.

Carbenes and carbenoids are versatile intermediates in organic chemistry. Usually, diazo compounds serve as precursors to these reactive species under photolytic, thermal, or transition metal-catalyzed conditions.^[1] And yet, due to the inconvenience of their preparation and their hazardous nature, alternative source for these species is highly desirable. In this context, an alkyne such as propargylic ester 1^[2] would be an ideal precursor to generate a metal-carbenoid 2 if the alkyne can be properly activated by electrophilic transition metals^[3] under appropriate conditions (Scheme 1). However, the potential of this transition metal-based alkyne activation is significantly reduced by the propensity of 1 to undergo a [1,3]-acyl shift to generate allene 3^[4] over a [1,2]-shift to form vinyl metal carbenoid 2, if the alkyne moiety is internal. In general, only terminal alkynes (R₃ = H)^[5] and internal alkynes with electron-withdrawing substituent (R₃ = CO₂R)^[6] were known to provide the metal carbenoid 2.

Scheme 1.

[1,2]- vs. [1,3]-acyl shift.

Recently, in the study of the metallotropic [1,3]-shift^[7] of electrophilic metal complexes, we have shown that an 1,3-diyne-containing system (R₃ = alkynyl group) provided strong preference for [1,2]-acyl shift, thereby selectively generating vinyl carbenoid 2.^[8] Recognizing the subtlety of the electronic influence by the R₃ substituent for the [1,2]- and [1,3]-acyl shift, we envisioned that the introduction of an electron-withdrawing oxygen substituent at the propargylic carbon of 4 would provide enough driving force for the formation of carbenoid 6 over 5 via an [1,2]-acyl shift (Scheme 2).^[9] Furthermore, the oxygen substituent would also allow an alternative trapping mechanism of metal carbenoid 6 to form 1,3-diene 7 through a [1,2]-hydride shift^[5f],[10] due to an electronic effect of the oxygen substituent. This would nicely complement the more prevailing trapping manifolds of reactions of 6 through [1,2]-acyl migration forming 8 or intra/intermolecular cyclopropanation generating 9. We report herein a prominent role of the propargylic oxygen substituent to induce [1,2]-acyl migration in the initiation step and a favorable [1,2]-hydride shift at the termination step. This ultimately allowed us to observe the conspicuous difference in trapping preference between vinyl and alkynyl carbenoids, where the former have a higher tendency to undergo [1,2]-hydride shift whereas the latter favours an addition to alkenes and alkynes.

Scheme 2.

Controlled [1,2]-acyl shift to form vinyl carbenoid and its trapping.

First, we examined substrate 10 containing propargylic methoxy substituent for its [1,2]-acyl shift and the termination behaviour of the resultant carbenoid. When 10 was subjected to a typical reaction conditions (5 mol % PtCl₂ under CO,^[11] toluene, 80 °C) the expected 1,3-diene 12 was obtained in high yield as a mixture of Z/E isomers (Scheme 3). Based on the known examples of intermolecular trapping of vinyl carbenoid of type 11 with styrene,^[12] we expected that in the presence of excess amount of styrene, it should deliver cyclopropanated product 13. However, even in the presence of excess amount of styrene the cyclopropanated product 13 was not obtained, instead still 12 was produced as the sole product. In comparison, under identical conditions, 1,3-diyne 14 did not lead to [1,2]-hydride shifted product 16 and most starting material was recovered unchanged. On the other hand, efficient intermolecular cyclopropanation occurred in the presence of styrene to yield product 17 in 76% (3.8:1). This result clearly suggests that vinyl carbenoid 11 preferentially undergoes [1,2]-hydride shift^[13] whereas alkynyl carbenoid 15 prefers to react with double bond of styrene, thereby indicating significant difference in reactivity between these carbenoids.^[12,14]

Scheme 3.

The reactivity difference between vinyl and alkynyl Pt-carbenoids.

This reactivity difference manifested by putative vinyl carbenoid 11 and alkynyl carbenoid 15 was further examined by using substrates 18 and 22 that contain an allyl ether moiety by which the respective carbenoid intermediates would be readily trapped intramolecularly (Scheme 4). We surmised that the reactions with these substrates would eliminate any uncertainty related to the entropy-based kinetic barrier, thereby showing the inherent reactivity difference more clearly. Under standard conditions (5 mol % PtCl₂ under CO, toluene, 60 °C), the reaction of substrate 18 selectively produced [1,2]-hydride shifted product 20 (91%, Z:E = 2.1:1) without any indication of formation of cyclopropanated product 21. On the other hand, the related 1,3-diyne 22 gave only cyclopropanated product 25 in 96% yield devoid of [1,2]-hydride shifted product 24.^[15] Again, this result is believed to be caused by the reactivity difference between vinyl carbenoid 19 and alkynyl carbenoid 23, which further corroborates the previous conclusion that vinyl Pt-carbenoids prefer [1,2]-hydride shift whereas alkynyl Pt-carbenoids favor cyclopropanation. It is worth emphasizing that propargylic acetoxy alkynes 10/18 and 14/22, although internal alkynes are known to promote allene formation, did not provide any allenes via [1,3]-acetoxy shift. They delivered products only derived from the putative vinyl Pt-carbenoids, which can be rationalized by the presence of electron-withdrawing methoxy and allyloxy substituents for the monoynes and the alkynyl substituent for the diynes.

Scheme 4.

The reactivity difference between vinyl and alkynyl Pt-carbenoids.

Having shown the salient reactivity difference between vinyl and alkynyl Pt-carbenoids, diverse substrates were employed to confirm the observed reactivity in broader context. Especially, the trapping of initially formed alkynyl Pt-carbenoids by tethered alkynes instead of alkenes in substrates 26f will further bolster their preferred reactivity toward π- bonds over [1,2]-hydride shift (Table 1). Substrate 26a that contains a tethered terminal alkene provided only [1,2]-hydride shifted product 27a in 70% yield as a mixture of E/Z isomers (entry 1). Similarly, a trisubstituted alkene-containing substrate 26b also gave 27b in 92% yield without any sign of cyclopropanated product (entry 2). An aromatic substituent^[16] in 26c did not perturb the reactivity of the corresponding vinyl Pt-carbenoid, producing 1,3-diene product 27c in 77% yield (entry 3). The reaction of symmetrical monoyne 26d yielded 27d via a consecutive [1,2]-OAc shift followed by a [1,2]-hydride shift in 75% yield (Z:E = 1.7:1) at 40 °C (entry 4). On the other hand, substrate 26e (entry 5), containing mono and diyne moieties that can potentially compete, yielded product 27e (Z-isomer, 67% at 47% conversion), where only the monoyne was transformed to a 1,3-diene moiety presumably via a vinyl Pt-carbenoid. This result suggests that a monoyne is more reactive than the corresponding diyne. In comparison, the reaction of symmetrical diyne 26f gave symmetrical endiyne 27f in 92% yield, which is the result of the favourable addition over a [1,2]-H shift of the initially formed alkynyl carbenoid onto the other 1,3-diyne moiety across the tether (entry 6). The differential reactivity of vinyl and alkynyl Pt-carbenoids was further examined using a diyne initiator moiety of substrates 26g and 26h. The reaction of 26g containing a monoyne terminating group gave 27g in 47% yield (Z:E = 1:1.2) via the penultimate vinyl Pt-carbenoid, which ultimately delivered the observed product via a [1,2]-hydride shift (entry 7). On the other hand, 26h with diyne terminating group, which eventually generated an alkynyl carbenoid at the penultimate stage of the catalytic cycle, did not give any product in the absence of styrene but yielded cyclopropanated product 27h in 26% yield in the presence of styrene (entry 8). Decomposition of products as well as other side reactions such as aromatization seem to account for the low yields of the last two reactions.

Table 1.

Products from [1,2]-H shift of Vinyl Carbenoids and Addition to Alkenes and Alkynes of Alkynyl Carbenoids.^a

Entry		Substrate	Temp (°C)	Product	Yield (%)b	Z/E ratioc
1	26a (R = H)		60		70d	1: 2.8e
2	26b (R = Me)		60		92	5.1: 1e
3	26c		40		77	2.2: 1
4	26d		40		75	1.7: 1
5	26e		60		67	1.0: 1
6	26f		80		92	–
7	26g		80		47	1: 1.2
8	26h		80		26f	–

^a)Conditions: 5 mol% PtCl₂, under CO, toluene.

^b)Isolated yield.

^c)Z/E ratio of crude products.

^d)Additional 20% of the corresponding hydrolyzed aldehyde was isolated.

^e)Z/E ratio of the methyl enol ether (*).

^f)Mixture of diastereomers in 2.9:1 ratio.

In conclusion, we have observed that propargylic esters containing a propargylic alkoxy group preferentially undergo [1,2]-acyl shift over [1,3]-shift with PtCl₂. In this study, we also have demonstrated that vinyl and alkynyl Pt-carbenenoids have significantly different reactivity: vinyl Pt-carbenoids have the propensity to undergo [1,2]-H shift whereas alkynyl Pt-carbenoids preferentially add to π-bonds in both intra- and intermolecular reactions. Further study to elucidate the origin of the difference between these carbenoids will be reported in due course.

Experimental Section

Procedures for PtCl₂-catalyzed reaction

PtCl₂ (5 mol %) was added to a solution of the alkyne in toluene (0.05 M), and CO was bubbled through the solution. The mixture was warmed to the appropriate temperature (40 °C, 60 °C, or 80 °C) and stirred until the reaction was complete (30 min∼2 hrs). The mixture was cooled to room temperature and the solvent was evaporated under vacuum. Purification by flash chromatography on silica gel (hexane/ethyl ether) afforded the product.