Facile Synthesis of 3-Amido-Dienynes via a Tandem α-Propargylation–Isomerization of Chiral Allenamides and their Applications in Diels-Alder Cycloadditions

Abstract

A series of de novo 3-amido-dienynes was synthesized via tandem α-propargylation–isomerization of chiral allenamides with moderate E/Z ratio. Reactivities of E-and Z-isomers were examined.

Graphical abstract

Recently, we developed a highly E-selective allenamide 1,3-H shift reaction under thermal or acid-promoted condition, which afforded an efficient route to access valuable 2-amidodiene building blocks.^1,2 Furthermore, chiral 3-amidotrienes such as 2 could be prepared by the reactions starting from α-allyl-allenamides 1 (path a, scheme 1), which could be facilely converted to cyclic 2-amidodienes 4 and further applied in various different transformations.^{1, 3}

Scheme 1.

3-amidotrienes formation via 1,3-H-shift and the design to 3-amidodenynes

While much efforts have been dedicated to 3-amidotrienes, we were also interested in the reactivity and potential application of its structural sibling 3-amidodienynes, which could potentially be obtained following a similar α-propargylation–1,3-H shift sequence of allenamides (path b, scheme 1). Comparing to 3-amidotrienes, precedents of 3-amino-or 3-amido-dienynes are relative rare.^4–7 An elegant example of 3-amidodienynes was reported recently by Huang and coworker in which they developed an interesting route to 1-silyloxy-3-aminodienynes, and subsequently, applied these electronically enriched 3-aminodienyens in different ring-formation reactions.⁷ Herein, we wish to disclose our recent studies on the preparation of 3-amidodienyens and exploration of the reactivity of both E-and Z-isomers.

Given that 3-amidodienynes could be obtained via a 1,3-H-shift of the corresponding α-propargyl allenamides, we commenced our study through examining the propargylation of allenamides using 6a bearing an i-Pr substituent on the oxazolidinone ring. To our surprise, under the standard alkylation conditions (n-BuLi, HMPA, propargyl bromide), in addition to the expected α-propargyl allenamide product 4a, two new products were isolated from the reaction mixture. These two additional products turned to be the E/Z isomers of the desired 3-amidodienynes 5a (entry 1, table 1). The direct formation of 3-amidodienynes could be promoted by base-promoted in situ 1,3-H-shift of α-propargyl allenamide in the presence of excess n-BuLi used for allenamide deprotonation, or even the α-deprotonated allenamide, although the exact mechanism remains unclear.⁸ Efforts to improve the conversion toward 3-amidodienyne formation, such as increasing the amount of n-BuLi and prolonged reaction time, proved to be fruitless.

Table 1.

Propargylation of allenamides^a

Entry	allenamide	Propargyl bromide	4/ E-5/ Z-5 (Yields %)b
1	6a (R = i-Pr)	7a (R1 = Me)	4a/ E-5a/ Z-5a (34/15/26)
2c	6a (R = i-Pr)	7b (R1 = TMS)	4b/ E-5b/ Z-5b (0/15/26)
3	6b (R = Bn)	7a (R1 = Me)	4c/ E-5c/ Z-5c (38/14/23)
4	6c (R = CHPh2)	7a (R1 = Me)	4d/ E-5d/ Z-5d (41/12/19)
5	6c (R = CHPh2)	7c (R1 = Ph)	4e/ E-5e/ Z-5e (0/13/22)

^aReaction Conditions: n-BuLi (1.5 equiv.) was added to allenamide 6 (1.0 equiv) in HMPA (1.5 equiv) and THF (0.1 M) at −78°C. Propargyl bromide 7 (1.5 equiv) was added 45 min later and the reaction was stirred at −78 °C for 1 h and then warmed to rt overnight.

^bIsolated yields in the bracket.

^cAll yields were overall yields of E/Z-5b (R¹ = H) after removal of TMS of the directly formed 3-amidodienynes products with TBAF since TMS protected 3-amidodienyne isomers were inseparable. TMS = trimethylsilyl.

Although this tandem propargylation–1,3-H-shift sequence afforded 3-amidodienynes in moderate E/Z ratios, we were satisfied with the efficiency that both isomers could be obtained in one-pot to set the stage for evaluation of their respective reactivities and potential applications. By employing the standard conditions, we examined tandem propargylation–1,3-H-shift reactions using combinations of different chiral allenamides and substituted propargyl bromides (entries 2–5, table 1). Generally, all these reactions proceeded smoothly in the standard conditions with different oxazolidinone auxiliaries. Interestingly, when either TMS (entry 2, table 1) or Ph (entry 5, table 1) substituted propargyl bromides was used, only 3-amidodienyne products were found in moderate yield with no α-propargyl allenamides isolated. This could a result of enhanced activity of propargylic protons in the initial α-propargylated allenamide toward base-promoted 1,3-H-shift.

With these 3-amidodienynes in hand, their potential synthetic applications were investigated. We envisioned that these highly electron-rich Z-3-amidodienynes could serve as suitable diene components in a Diels-Alder cycloaddition reaction with the alkyne motif preserved as a useful handle for further transformations. Towards this goal, Diels-Alder reaction of Z-5a with methyl vinyl ketone (MVK) was explored. Considering the potential labile property of the substrate in the presence of Lewis acids, conventional thermal condition was selected for the Diels-Alder reaction and Z-5a was found to react smoothly with MVK at 130 °C in the presence of BHT (2,6-Di-tert-butyl-4-methylphenol) to afford the desired product 9a in 56% yield with a diastereomeric ration of 8:1 (entry 1, table 2). Inspired by these results, the Diels-Alder reactions were further examined with other Z-3-amidodienynes (entries 2–6, table 2) under the standard thermal conditions. It was readily noticeable that the auxiliary employed played a key role in the reaction selectivity (entries 1,3, and 4). While Z-5c bearing a Bn substituted Evans auxiliary gave inferior selectivity comparing to Z-5a with an i-Pr substituent, the reaction with Z-5d incorporated with a Sibi auxiliary provided the cycloaddition product as a single isomer.⁹ Similar selectivity results were also observed in our previous studies.^3c

Table 2.

Diels-Alder reaction of 3-amidodienynes^a

^aReaction Conditions: A solution of Z-3-amidodienyne (1 qeuiv), MVK (10 equiv) and BHT (2.0 equiv) in toluene (3 mL) in a sealed tube was heated at 130 °C overnight.

^bIsolated yields and ratios of endo-II/endo-I in the bracket.

^c3 equiv. of dienophile was used.

^dThe reaction was sluggish.

While it would appear that these cycloadditions were compatible with variations on the alkyne motif (entries 2 and 5), the reaction with a terminal alkyne substrate gave relatively lower yield (entry 2). Other dienophiles such as 8b and 8c were also examined (entries 6 and 7). Dienophile 8b with p-bromophenyl substitution was tolerant and it offered the desired product 9f in good yield and selectivity under standard conditions. On the other hand, when 8c with terminal phenyl substituent on the double bond (entry 7), the reaction became very sluggish even at higher temperature. The prolonged reaction time resulted in a messy reaction mixture with decomposition of the starting material, although the desired product was detected from the reaction mixture by mass spectrometry. This outcome suggests that the cycloaddition was quite sensitive to the β-substitutions of these enones. We also attempted the reaction with 2-cyclohexen-1-one (results not shown in the table), but no desired product was found with only slow decomposition of the starting material.

NOE experiment using cycloadduct 9f confirmed its 1,3-cis relative stereochemistry, thereby suggesting an endo-selectivity for these cycloadditions. Furthermore, based on the mechanistic model from our previous studies using related amino-dienes, these Diels-Alder products would be assigned as endo-II cycloaddition products as shown in scheme 2.^3c,d.

Scheme 2.

Structure assignment of the Diels-Alder reaction products

Having established a reliable Diels-Alder reaction of Z-3-amidodienynes established, we then focused on the reaction of E-3-amidodienynes. Given the similarity to 3-amidotrienes, these E-3-amidodienynes were postulated to be intrinsic candidates for electrocyclizations or similar transformations.¹⁰ Thus, we attempted to directly cyclize E-5a various conditions (Ti(OPr)4, (JohnPhos)AuCl/AgSbF₆, PtCl₄ conventional heating, photo-activation conditions and so on), but none of these conditions led to desired products with decompositions being the case in most attempts (reaction a, scheme 3). When the direct cyclization failed to offer the desired product, a sequence of Pauson-Khand reaction-6π-electrocyclization was envisioned and explored (reaction b, scheme 3).¹¹ Albeit in low yield (in 28% yield un-optimized), E-5a did react with norbornene to afford the desired Pauson-Khand product 11. However, analysis selectivity was poor (roughly 1:1), which could be due to the long distance between the chiral auxiliary and the actual rsite of bond of formation. Unfortunately, initial attempts at 6π-electrocyclizations of 11 to give the octahydro-fluroenone derivative 12 proved to be challenging. Efforts are currently underway to achieve such synthetic design.

Scheme 3.

Reactions of E-3-amidodienynes

We have described here a tandem propargylation–1,3-H-shift sequence to access both E-and Z-isomer of chiral 3-amidodienynes. A series of 3-amidodienynes were synthesized with this method. An application of Z-3-amidodienynes in Diels-Alder cycloadditions gave endo-II products in good yields and excellent selectivity. Reactivity of the corresponding E-3-amidodienynes toward electrocyclization was also examined but was unsuccessful. Further efforts in developing synthetic applications of these de novo chiral 3-amindo-dienynes will be reported in due courses.