Alkene Synthesis Using Phosphonium Ylides as

DOI: 10.1002/ajoc.201800586 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Communication

Olefin Synthesis

Alkene Synthesis Using Phosphonium Ylides as Umpolung Reagents Minh Duy Vu, Wei-Lin Leng, Hao-Cheng Hsu, and Xue-Wei Liu*[a] original Wittig reaction conditions. Schlosser and co-workers successfully employed lithium salt effect to obtain (E)-alkene from non-stabilized ylides.[12] More recently, Tian’s modification allows tunable selectivity for semi-stabilized ylides.[13] However, up to date, the synthesis of (Z)-alkene, an isomer with higher in energy yet more difficult to obtain, from stabilized phosphonium ylides is still elusive. Other than Wittig reaction, autoxidation of non-stabilized ylides to form symmetrical alkene were also reported since 1977,[14] however, the reaction gave complex mixture with low yielding of olefin. Noiret then developed symmetrical (Z)-olefin synthesis from non-stabilized phosphonium ylides through oxidation in 1996,[15] in which, classical Wittig reaction was believed to be the key step determining the selectivity in this coupling protocol. To the best of our knowledge, there has been no further extension to the less reactive stabilized phosphorane up to date. Since 2008, the use of photosensitizers as redox-catalyst in organic transformation has been growing tremendously.[16] Upon activated by visible light, the excited state photoredox catalyst might undergo single electron transfer events resulting in either oxidizing or reducing the organic substrate, lead to novel reactivity. Various examples of oxidative activation to form radical cation from unsaturated system were nicely demonstrated by Fukuzumi,[17] Nicewicz[18] and Yoon[19] et al. The first application of photoredox chemistry on sulfonium ylides has been reported by Xiao and coworkers for the cyclization of transient radical cation to form 3-acyl oxindoles.[20] Quick growth of the newly developed photoredox catalysis triggered our attention to discover new reactivity of the well-established phosphonium ylide chemistry. Our primary aim was to seek for novel umpolung reactivity of phosphonium ylides after undergoing the hypothesized single electron oxidation. In this article, we would like to present our unprecedented discovery on the formation of selective (Z)-alkene starting from stabilized and semi-stabilized ylides under visible-light mediated photoredox conditions. Our study was initiated by probing the reactivity of commercially available phosphorane 1 a under photoredox condition. The selection of catalysts was firstly limited to highly oxidative complexes such as Ir[dF(CF3)ppy]2(dtbbpy) PF6 (E1/2ox = 1.21 V vs SCE)[21] and the acridinium organic dye (E1/2ox = 2.06 V vs SCE)[22] (entry 1–2). However, complex product mixture was obtained probably due to over oxidation. To our surprise, while screening with more reductive photo-catalysts (entry 3–5), alkene product was obtained in

Abstract: We report a transformation between a wellknown nucleophile, phosphonium ylide, and an electrophile. Subsequent ylide-ylide coupling achieves unusual (Z)selectivity as compared to the conventional Wittig olefination. The synthesis of medium-sized ring olefins has also been demonstrated, together with a one-pot olefination from activated halides.

Phosphonium ylide, an organic structure containing C P polarized double bond, was first synthesized over a century ago.[1] Since then, the reactivity of these compounds has been well investigated in both mechanistic and synthetic aspects.[2] Being easily accessible from other functionalities such as organohalides and alcohols, ylides have shown their wide applicability in a variety of organic transformations as the role of nucleophile. Its nucleophilicity has been demonstrated through the reactions with typical electrophiles,[3] for example, nitriles,[4] diazo compounds,[5] azides[6] and acyl chlorides[7]. Moreover, it has also been utilized in Michaeltype and Mannich-type reactions, with or without stereocontrol, to construct new C C bond.[8] The most outstanding application of phosphonium ylides is undeniably Wittig olefination, which was discovered by Georg Wittig in 1954.[9] The reaction proceeds chemoselectively between an aldehyde or ketone and a phosphonium ylide (phosphorane) under mild condition, and so far is still one of the best choices for synthetic chemists to approach alkenes. Nevertheless, the stereoselectivity of final olefin products is highly dependent on starting phosphoranes. Extensive scope studies have shown that, under the classic Wittig reaction condition, stabilized ylides usually give dominant (E)-alkene, while non-stabilized ylides give the (Z)-isomer.[10] The phenomenon has been explained by many groups through experimental data as well as DFT calculation.[11] In order to achieve selectively the remaining isomer from the same type of ylides, chemists have developed modification to the

[a] M. Duy Vu, W.-L. Leng, H.-C. Hsu, Prof. X.-W. Liu Division of Chemistry and Biological Chemistry School of Physical and Mathematical Sciences Nanyang Technological University 21 Nanyang Link, Singapore 637371 E-mail: [email protected] Supporting information for this article is available on the WWW under https://doi.org/10.1002/ajoc.201800586 Asian J. Org. Chem. 2018, 7, 1 – 5

1

These are not the final page numbers! ��

© 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

good to excellent yield. Interestingly, strong reducing photoredox catalyst such as Ir(ppy)3 facilitated the formation of (E)alkene, while Ru complexes, on the other hand, gave predominantly (Z)-isomer. 1 Changing the ligand on the catalyst could affect the selectivity outcome. Despite possessing similar redox profiles, Ru(phen)3Cl2 gave exclusively (Z)isomer, while Ru(bpy)3(PF6)2 resulted in a mixture of isomers. This interesting observation has not been rationalized yet. Solvent evaluation revealed that the transformation proceeded most efficiently in non-polar halogenated solvents (entry 6–10). Control experiments have also been conducted to confirm the necessity of all the components and parameters. The presence of photo-catalyst is indeed crucial for this novel reactivity, since the reported photo-reactivity of phosphorane and related compounds undergo different pathways.[23] Apart from phosphoranes, phosphonium salts could also be utilized to give comparable results (deprotonation in situ using strong base ie. NaH or LiO tBu). In the case of semi-stabilized ylides, phosphonium salts were applied directly in the presence of excess sodium hydride (in mineral oil) and catalytic amount of crown ether (15 C5).[24]

Scheme 1. Homo-coupling scope of various phosphonium ylides, phosphonium salts and phosphonate. All reactions were carried out under open air at 0.1 mmol scale, using 34 W blue LED with cooling fan. Isolated yields were reported, and olefin isomeric ratios were determined from analyses of 500 MHz 1H-NMR spectra of crude mixtures. [a] Starting from phosphonium salt. [b] Starting from HWE phosphonate.

Table 1. Selected results for optimization study.

Entry

Catalyst (1 mol%)

Solvent

Results[a]

1 2 3 4 5 6 7 8 9 10 11[b] 12[c] 13

Ir[dF(CF3)ppy]2(dtbbpy)PF6 MesAcr + Me (5 mol%) Ir(dtbbpy)2(ppy)PF6 Ir(ppy)3 Ru(bpy)3(PF6)2 Ru(phen)3Cl2 Ru(phen)3Cl2 Ru(phen)3Cl2 Ru(phen)3Cl2 Ru(phen)3Cl2 Ru(phen)3Cl2 Ru(phen)3Cl2 –

DCM DCM DCM DCM DCM DCM THF CHCl3 DMF MeCN DCM DCM DCM

ND ND 28% (> 20E : 1Z) 50% (> 20E : 1Z) 98% (1E : 4Z) 95% (> 20Z : 1E) ND 89% (10Z : 1E) NR 77% (> 20Z : 1E) Trace Trace Trace

coupling reaction smoothly to give (Z)-alkene as major product in excellent yields, despite the electronic property difference of substituents on benzene ring. Aliphatic keto-stabilized phosphorane (1 n) also yielded comparable result. On the other hand, esterstabilized phosphoranes, as well as Horner-Wadsworth-Emmons (HWE) phosphonate reagent gave high (E)-selectivity under the same conditions. Benzylic semi-stabilized phosphorane (1 f, 1 g and 1 h) could undergo reaction smoothly as well, albeit diminished selectivity. Interestingly, for the case of allylic phosphorane (1 i) tri-ene product, subsequent electrocyclic reaction occurred to yield substituted 1,3-cyclohexadiene in moderate yield. Our reaction scope was also successfully extended to ringclosing reaction for the syntheses of medium-sized rings. Starting from simple aliphatic diols, 9- and 10-membered-ring dilactones were synthesized efficiently over 2 steps, thereby demonstrating the utility of our method (Scheme 2a). These lactones resemble reported bioactive natural products such as the families of Pyrenolides[25] and Stagonolides.[26] Although moderate yield was achieved, the selectivity of (Z)-isomer was almost exclusive and other side products observed were likely dimers and trimers (LC–MS). Further applications on crosscoupling of the phosphorane, as well as phosphonium salts have also been demonstrated (Scheme 2b). It was found that the phosphoranes formed in situ by deprotonation gave similar reaction efficiency and the excessive amount of strong base used did not have an adverse effect on the reaction. A few semi-stabilized ylides were tested under our reaction condition and could successfully yield non-symmetrical stilbene analogues predominantly in (Z) form. Notably, a series of challenging cis-1,4-ketoester could also be synthesized in very high selectivity. Our results also showed that the cross-coupling between two different keto-stabilized ylides proceeded comparably well, yielding 1,4-dionene. With the

All reactions were carried out under open air at 0.1 mmol scale, using 34 W blue LED with cooling fan. [a] Isolated yield of 2 aa was reported, olefin isomeric ratio was determined from analysis of 500 MHz 1H-NMR spectra of crude mixture. [b] Reaction in the dark. [c] Reaction in Ar atmosphere. ND: not determined. NR: no reaction.

With the optimized condition in hand, we proceeded to evaluate the substrate scope of the reaction. To test for the reactivity of various types of ylides under photoredox condition, we started with homo-coupling of phosphonium ylides (or phosphonium salt precursors) and the results are illustrated in Scheme 1. Generally, symmetrical alkenes could be obtained in good to excellent yields, predominantly in (Z)-selectivity. Ketostabilized phosphoranes (1 b, 1 c, 1 d and 1 e) underwent the 1

Strong reducing photoredox catalyst may undergo single electron transfer to the highly electron-deficient alkene products. Hence, rotation around the double bond enables isomerization to the more stable (E)-isomer.

Asian J. Org. Chem. 2018, 7, 1 – 5

www.AsianJOC.org

2



Communication 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

success on utilizing phosphonium salts, we anticipated that the activated halides could be applied in the syntheses of olefins via one-pot reaction. Indeed, (Z)-stilbenes could be obtained starting from benzylic halides in synthetically useful yield. (E)-4 a and (E)4 b were undetectable in crude NMR as well as in the isolated products. Intriguingly, (Z) to (E) isomerization was observed while monitoring the formation of stilbene 4 c during the reaction. This information suggests the (Z) selectivity originates from reaction mechanism, not through photo-isomerization. To the best of our knowledge, the direct synthesis of stilbene from benzyl halides was only reported in recent work done by Walsh et al. so far.[27] To investigate the reaction mechanism, cyclic voltammetry experiment has been conducted to measure the oxidation potential of the model substrate 1 a (E1/2ox = 1.12 V vs SCE). The result showed that single electron oxidation of the ylide by [Ru (phen)33 + ] is thermodynamically favorable. This also indicates an oxidative quenching pathway of the excited photoredox catalyst [Ru(phen)32 + ]* (E1/2ox = 1.26 V vs SCE). Hence, we hypothesized that, upon single electron oxidation, the radical cation intermediate I will be intercepted by another ylide molecule presence in the solution to form either radical cation IIA (path A) or IIB (path B) through nucleophilic attack. The quenching of radical cation IIA with radical anion oxygen followed by triphenylphosphine oxide elimination facilitates the alkene formation. On the other hand, radical cation IIB resulting from SN2-type reaction can undergo single electron reduction (IIIB) and subsequent phosphine elimination to form the alkene product.

Figure 1. Overview of Phosphonium ylide chemistry.

However, monitoring the reaction by 31P-NMR did not show any trace of PPh3 during reaction but Ph3PO was observed instead. Therefore, we presume that path A is the operative pathway, even though the formation of dioxygen-bridgedphosphine IIIA has not been reported in preceding literature. To explain for the intriguing (Z)-selectivity in most of the cases, we have looked up seminal works on the possible isomerization of olefin under visible-light photoredox condition. The first visiblelight photoredox up-hill catalysis done by Weaver et al. featured a facile isomerization of several (E)-aminoalkenes.[28] The work done by Gilmour and co-workers on Riboflavin catalyzed alkene isomerization presented another approach to (Z) olefin from (E)-counterpart.[29] Wang and co-workers reported the isomerization of (E)-1,4enediones.[30] Additionally, Liang and co-workers demonstrated that the use of organic dye also help to facilitate 1,4-enedione isomerization to favorable (Z)-configuration.[31] Therefore, the probability of our alkene products undergoing photo-isomerization could not been excluded. However, while the reaction progress of 1 a in NMR tube was monitored, the formation of cisalkene 2 aa was continuously observed over time. Photo-isomerization test was also conducted, starting from purified (E)-2 aa, only partial isomerization occurred under our optimized photoredox condition (Figure 2). Prolonged reaction time merely led to the decomposition of the alkene substrate. Hence, we propose that, apart from photo-isomerization, the approach of the ylides in the transition state may play significant role in the observed selectivity. The phosphorane (nucleophile) and its cationic radical form (electrophile) should approach each other in such a way that large substituents stay opposite in space to minimize steric repulsion. As a result, dioxygen-bridge IIIA intermediate will have a

Scheme 2. Intramolecular reaction and cross-coupling reaction of phosphonium ylides/ phosphonium salts. All reactions were carried out under open air at 0.1 mmol scale, using 34 W blue LED with cooling fan. Isolated yields were reported, and olefin isomeric ratios were determined from analyses of 500 MHz 1H-NMR spectra of crude mixtures. [a] Examples of ring-closing reaction starting from simple diols. [b] Reactions between different phosphonium ylides (2 : 1 equivalent). [c] One-pot reaction of activated halides (bromides) (2 : 1 equivalent). Asian J. Org. Chem. 2018, 7, 1 – 5

www.AsianJOC.org

Figure 2. Experimental evidence for the photoisomerization of alkene products

3



Communication 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Conflict of Interest

cis-conformation, following which the concerted removal of phosphine oxide molecules gives rise to (Z)-alkene product (Scheme 3). In summary, a practically simple method to synthesize alkene from phosphonium ylide or phosphonium salts was reported. Our reactions undergo a mild condition providing a unique (Z)-selectivity for stabilized phosphoranes. Detailed mechanistic study and further expansion to other type of ylides have been carrying out in our laboratory.

The authors declare no conflict of interest. Keywords: ylides · photoredox diastereoselectivity · alkenes

catalysis

·

umpolung

·

[1] A. Michaelis, H. V. Gimborn, Ber. Dtsch. Chem. Ges. 1894, 27, 272–277. [2] a) O. I. Kolodiazhnyi, in Phosphorus Ylides, Wiley-VCH Verlag GmbH, 2007, pp. 1–8; b) A. W. Johnson, Ylides and imines of phosphorus, J. Wiley, 1993. [3] R. Appel, R. Loos, H. Mayr, J. Am. Chem. Soc. 2009, 131, 704–714. [4] R. G. Barnhardt, W. E. McEwen, J. Am. Chem. Soc. 1967, 89, 7009–7014. [5] Y. V. Tomilov, D. N. Platonov, D. V. Dorokhov, O. M. Nefedov, Tetrahedron Lett. 2007, 48, 883–886. [6] P. Ykman, G. L’Abbé, G. Smets, Tetrahedron 1971, 27, 845–849. [7] A. M. A. Rocha Gonsalves, A. M. T. D. P. V. Cabral, T. M. V. D. D. Pinho e Melo, T. L. Gilchrist, Synthesis 1997, 1997, 673–676. [8] O. I. Kolodiazhnyi, in Phosphorus Ylides, Wiley-VCH Verlag GmbH, 2007, pp. 9–156. [9] G. Wittig, U. Schöllkopf, Chem. Ber. 1954, 87, 1318–1330. [10] B. E. Maryanoff, A. B. Reitz, Chem. Rev. 1989, 89, 863–927. [11] a) R. Robiette, J. Richardson, V. K. Aggarwal, J. N. Harvey, J. Am. Chem. Soc. 2006, 128, 2394–2409; b) E. Vedejs, M. J. Peterson, in Top. Stereochem., John Wiley & Sons, Inc., 2007, pp. 1–157; c) P. A. Byrne, D. G. Gilheany, Chem. Soc. Rev. 2013, 42, 6670–6696. [12] M. Schlosser, K. F. Christmann, Angew. Chem. Int. Ed. 1966, 5, 126–126; Angew. Chem. 1966, 78, 115–115. [13] D.-J. Dong, H.-H. Li, S.-K. Tian, J. Am. Chem. Soc. 2010, 132, 5018–5020. [14] S. H. Pine, E. Fujita, J. Org. Chem. 1977, 42, 1460–1461. [15] S. Poulain, N. Noiret, H. Patin, Tetrahedron Lett. 1996, 37, 7703–7706. [16] a) C. K. Prier, D. A. Rankic, D. W. C. MacMillan, Chem. Rev. 2013, 113, 5322–5363; b) M. H. Shaw, J. Twilton, D. W. C. MacMillan, J. Org. Chem. 2016, 81, 6898–6926. [17] S. Fukuzumi, K. Ohkubo, Chem. Sci. 2013, 4, 561–574. [18] K. A. Margrey, D. A. Nicewicz, Acc. Chem. Res. 2016, 49, 1997–2006. [19] M. A. Ischay, T. P. Yoon, Eur. J. Org. Chem. 2012, 2012, 3359–3372. [20] X.-D. Xia, L.-Q. Lu, W.-Q. Liu, D.-Z. Chen, Y.-H. Zheng, L.-Z. Wu, W.-J. Xiao, Chem. Eur. J. 2016, 22, 8432–8437. [21] M. S. Lowry, J. I. Goldsmith, J. D. Slinker, R. Rohl, R. A. Pascal, G. G. Malliaras, S. Bernhard, Chem. Mater. 2005, 17, 5712–5719. [22] D. A. Nicewicz, T. M. Nguyen, ACS Catal. 2014, 4, 355–360. [23] M. Dankowski, in Organophosphorus Compounds (1993), John Wiley & Sons, Ltd, 2006, pp. 325–343. [24] G. Bellucci, C. Chiappe, G. L. Moro, Tetrahedron Lett. 1996, 37, 4225– 4228. [25] M. Nukina, T. Sassa, M. Ikeda, Tetrahedron Lett. 1980, 21, 301–302. [26] a) A. Evidente, A. Cimmino, A. Berestetskiy, A. Andolfi, A. Motta, J. Nat. Prod. 2008, 71, 1897–1901; b) A. Evidente, A. Cimmino, A. Berestetskiy, G. Mitina, A. Andolfi, A. Motta, J. Nat. Prod. 2008, 71, 31–34. [27] M. Zhang, T. Jia, H. Yin, P. J. Carroll, E. J. Schelter, P. J. Walsh, Angew. Chem. Int. Ed. 2014, 53, 10755–10758; Angew. Chem. 2014, 126, 10931– 10934. [28] K. Singh, S. J. Staig, J. D. Weaver, J. Am. Chem. Soc. 2014, 136, 5275– 5278. [29] a) J. B. Metternich, R. Gilmour, J. Am. Chem. Soc. 2015, 137, 11254– 11257; b) J. B. Metternich, R. Gilmour, J. Am. Chem. Soc. 2016, 138, 1040–1045. [30] K. Xu, Y. Fang, Z. Yan, Z. Zha, Z. Wang, Org. Lett. 2013, 15, 2148–2151. [31] D. Wei, F. Liang, Org. Lett. 2016, 18, 5860–5863.

Experimental Section To an oven-dry 8 mL vial equipped with a rubber septum and magnetic stir bar was charged Ru(phen)3.Cl2.xH2O (1 mol %) and Phosphorane 1 a (1 eqv.). Freshly distilled dichloromethane was then added to form a clear orange solution. A small needle poked through rubber septum ensures air contact while minimizing solvent evaporation. The reaction mixture was stirred under blue LED irradiation for 12 hours. Surrounding temperature was maintained by a cooling fan placed on top. The reaction mixture was then concentrated in vacuo and purified by column chromatography (hexane/EtOAc) affording alkene 2 aa as white solid (95%, > 20Z:1E).

Acknowledgements We gratefully acknowledge National Research Foundation (NRF2016NRF-NSFC002-005) and the Ministry of Education (MOE 2013-T3-1-002), Singapore for financial support of this research.

Manuscript received: October 10, 2018 Accepted manuscript online: October 15, 2018 Version of record online: ■■■, ■■■■

Scheme 3. Plausible mechanism and explanations for the unique (Z)selectivity. Asian J. Org. Chem. 2018, 7, 1 – 5

www.AsianJOC.org

4



Communication 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

COMMUNICATION Olefin Synthesis M. Duy Vu, W.-L. Leng, H.-C. Hsu, Prof. X.-W. Liu* 1–5

Intriguing! A mild and easy method to access olefins through phosphonium ylides via photoredox catalysis, transforming this wellknown nucleophile to an electrophile (umpolung) and subsequent ylideylide coupling, has been developed.

Unusual (Z)-olefin selectivity is achieved in comparison to the conventional Wittig olefination. The construction of medium-sized ring olefins has also been demonstrated through one-pot in situ ylide generation/olefination from activated halides.

Alkene Synthesis Using Phosphonium Ylides as Umpolung Reagents