Redox Cycloisomerization Approach to 1,2-Dihydropyridines

Abstract

The phosphine-catalyzed synthesis of 1,2-dihydropyridines via an alkyne isomerization/electrocyclization sequence is described. Propargylidenecarbamate substrates were prepared following a one-pot procedure between a terminal alkyne, a benzonitrile and a chloroformate in the presence of trimethylaluminum. This methodology gives access to a diverse set of 2,6-disubstitued 1,2-dihydropyridines in high yield. The products can be easily converted into substituted piperidines or pyridines and this methodology was applied to the synthesis of indolizidines.

Graphical Abstract

Heterocycles containing nitrogens are found abundantly in a broad array of natural products and pharmaceutically active substances.¹ Thus, an important amount of therapeutic agents contains polysubstituted piperidine or pyridine scaffolds.² They are also important precursors to complex biologically active molecules and are frequently utilized as building blocks.³ As a result, numerous methods have been reported for the synthesis of substituted piperidines ^4,5 and pyridines.⁵

1,2-Dihydropyridines are extremely useful and reactive synthetic intermediates as they can easily give access to a large variety of polysubstituted 6-membered N-heterocycles.^5,6,7 In comparison to 1,4-dihydropyridines, relatively few general methods for their selective synthesis have been reported in the literature rendering 1,2-dihydropyridines as underutilized chemical intermediates by the synthetic community.⁶ Major reasons are the poor stability upon storage of 1,2-dihydropyridines having no withdrawing group at the nitrogen atom and most reported syntheses give a limited product scope along with the formation of 1,4-dihydropyridines as side products.⁶ Traditional ways to prepare 1,2-dihydropyridines include pyridinium salt dearomatization via addition of Grignard⁸ or organocuprate reagents ⁹ as well as pericyclic reactions.⁶ Ellman and co-workers reported the Rhcatalyzed addition of alkynes to α,β-unsaturated N-benzyl aldimines via C-H activation followed by 6π-electrocyclization.¹⁰ Similarly, Okamura and co-workers described the 6π-electrocyclization of 1-azatrienes prepared from primary amines and 2,4-dienals.^11,12,13 These approaches are highly efficient and occurred under mild conditions but remain limited to alkyl imines. Thus, while the importance of accessing 6-membered nitrogen containing rings via 1,4-dihydropyridines (Hantzsch reaction) is well demonstrated,^7,14 having available a corresponding reaction via selective 1,2-dihydropyridines would open complementary regioselectivity in further derivatization as well as constitute a new approach to pyridines.

We previously reported the isomerization of electron poor alkynes 1 to dienes 3 catalyzed by nucleophilic phines.^15,16 Alkynones,¹⁶ alkynoates¹⁶ and alkynamides ¹⁷ were efficiently isomerized to the corresponding (E,E)-1,3-diene products which proceed through the formation of an allene intermediate 2 (Scheme 1). As part of our ongoing program aimed at creating structural complexity by catalytic isomerization of readily available unsaturated molecules,18 we now report the isomerization of propargylidenecarbamates 4 to 1,3-dienimines 5 followed by 6π-electrocyclization to obtain 1,2-dihydropyridines 6.

Scheme 1.

Phosphine-Catalyzed Isomerization of Alkynes to 1,3-Diene

In order to increase the synthetic utility of this methodology, we first investigated an efficient route to access unprecedented protected ynimines 4. Following a modified procedure recently reported by Lee and coworkers, we found that ynimines 4 could be easily accessed utilizing a one-pot procedure illustrated scheme 2.¹⁹ Non-enaminizable substrates 4 were prepared in good to high yields by addition of alkynyldimethyl-aluminum reagents (generated in-situ by reaction of AlMe₃ with a terminal alkyne) to nitrile 8 in hexanes/toluene followed by addition of the desired chloroformate 10 to dimethylaluminum iminate intermediate 9.²⁰

Scheme 2.

One-Pot Substrate Synthesis

To test our hypothesis concerning the conversion of 4 to 6, different phosphine donors were screened under various reaction conditions. Initial experiments with triphenylphosphine as catalysts in toluene gave full conversion but low isolated yields and significant decomposition was observed (Table 1, entries 1 and 2). The use of more nucleophilic phosphines increased the amount of decomposition products and only low amounts of 13 were obtained (entries 3 and 4).

Table 1.

Selected Phosphine Optimization

entry	cat. (mol %)a	time	conv %	yield %b
1	PPh3 (30)	48h	100	21
2	PPh3 (50)	48h	100	33
3	Ph2PMe (30)	24h	100	-c
4	Ph2PAllyl (30)	24h	85	20
5	dppp (10)	48h	80	10c
6	dppp (30)	24h	100	76
7d	dppp (30)	16h	100	<10e
8	dppp/AcOH(30)	1h	100	-f
9	dppp/(PhS)2 (30/10)	16h	30	<5
10	dppe (30)	24h	80	34
11	dppf (30)	16h	100	<10c

^adppp: bis(diphenylphosphino)propane; dppe: bis(diphenylphosphino)ethane; dppf: 1,1’-bis(diphenylphosphino)ferrocene.

^bIsolated yield.

^cComplex mixture of products.

^dReaction performed at 80 °C.

^eFull conversion to diene 12 was determined by ¹H NMR spectrum analysis of the crude material.

^fExtensive decomposition was observed.

Next we decided to explore the use of bidentate phosphines in the reaction. Their advantage can be explained by the ability of the second phosphine to act as a base^16b and therefore promote proton transfer within the diene chain. Gratifyingly, bidentate phosphine dppp (30 mol %) gave the desired 1,2-dihydropyridine 13 in 76% yield after 24 h at 110 °C (entry 6). The difference in reactivity between dppp and dppe (entries 6 and 10) remains unclear but suggests that the second phosphine of dppe does not promote proton transfer as similar results were obtained when monodentate allyldiphenylphosphine was used (entry 4). It is also noteworthy that 11 can easily isomerize to 12²¹ at lower temperature (80 °C) over 16 h but cyclization to 13 proved to be much slower.²² We also investigated the addition of a proton source in catalytic amount but 11 rapidly decomposed above 60 °C under acidic conditions (entry 8).

With these optimized conditions in hand, we investigated the product scope of this new cascade phosphine-catalyzed isomerization/6π-electrocyclization transformation (Scheme 3). To our delight, products 6 can be isolated in high yield starting from 4 with a large range of substituent patterns. The reaction tolerates the presence of functional groups present in various molecules of biological interest such as piperonyl (17), thiophene (18), indole (19) and benzopyran (20). We also screened different carbamates that could be easily removed, thus adding synthetic versatility to the products. The reaction proceeds well to give benzyl (14), allyl (22 and 25), ethyl and methyl carbamates of dihydropyridines; however, tert-butyl carbamate (15) was not obtained as slow decomposition of the starting material was observed, presumably due to steric hindrance proximal to the alkyne moiety and slow thermal removal of the Boc group.

Scheme 3.

Product Scope

The alkyl chain was also successfully functionalized with different types of substituents. Interestingly, this reaction tolerated a large variety of synthetically useful functional groups such as terminal alkyne 27 and silyl protected alcohol 26.

To further explore the synthetic utility of 1,2-dihydropyridine scaffolds, we investigated the functionalization of the 3,5-diene moiety. First, oxidation of various protected 1,2-dihydropyridines to pyridines was attempted under various reaction conditions (2,3-dichloro-5,6-dicyanobenzoquinone, o-chloranil and Mn(OAc)₂).²³ Unexpectedly, 6 reacted poorly and decomposed when reactions were run at higher temperature. Therefore, we envisioned a non-oxidative approach which would best suit our needs to access substituted pyridines. To the best of our knowledge, Pd-catalyzed deallylation/β-hydride elimination of allyl vinylcarbamate to vinyl imine has not been reported. We hypothesized that the β-hydride elimination should be highly favored in the case of a 1,2-dihydropyridine. Satisfyingly, when 22 and 25 were treated with a catalytic amount of Pd(OAc)₂ as a convenient pre-catalyst for Pd(0) in MeCN at 80 °C, pyridines 28 and 29 were obtained in excellent yield. This reaction performed without any phosphine ligand in contrast to Tsuji's conditions for his decarboxylative dehydrogenation of allyl β-ketoesters.²⁴

Depending on the nature of the carbamate, different levels of hydrogenation could be achieved in presence of molecular hydrogen and Pearlman's catalyst. Rapid removal of the Cbz group favored formation of piperidine cis-31 with excellent yield and dr, whereas the presence of ethylcarbamate facilitated hydrogenolysis of 13 to give 30. We were also able to selectively and quantitatively hydrogenate the olefin at the 3,4-positions of the 1,2-dihydropyridine system using Lindlar's catalyst to give 32. It should also be noted that the diene offers an opportunity to further functionalize the ring carbons. For example, chemoselective and diastereoselective dihydroxylation of 23 provides 33 in 9:1 dr.

To demonstrate the synthetic utility of this new methodology, we investigated a synthesis of indolizidine 39.²⁵ Starting from readily available nitrile 34, terminal alkyne 35 and CbzCl, 1,2-dihydropyidine 37 was easily obtained in two steps using the transformation reported herein. Dihydropyridine 37 was hydrogenated in presence of 5 wt % of Pd(OH)₂/C to give cis-piperidine 38. Finally a one-pot bromination²⁶/cyclization gave 39 in high yield (89%). In just four steps, we were able to access compound 37, which is a useful synthetic intermediate for the preparation of the histamine H₃ receptor agonists 40, used for treatment of pain and sleeping disorders.^25,27

In summary, we have developed an efficient synthesis of 1,2-dihydropyridines starting from propargylidenecarbamates catalyzed by bisphosphine dppp. Protected propargyl imines 4 were prepared in a one-pot procedure from commercially available reagents. The transformation proved to be high yielding and provided easy access to motifs found in molecules of biological interests. We also reported useful functionalizations of the reactive 1,2-dihydropyridine core structure to synthetize, piperidine, pyridine or non-cyclic carbamates. The applicability of the method has also been demonstrated by a concise synthesis of indolizidine 39.

Scheme 4.

Synthetic Applications

Scheme 5.

Concise Synthesis of Indolizidine 39