One-Pot Intermolecular Reductive Cross-Coupling of Deactivated

Letter Cite This: Org. Lett. 2018, 20, 5086−5089

pubs.acs.org/OrgLett

One-Pot Intermolecular Reductive Cross-Coupling of Deactivated Aldehydes to Unsymmetrically 1,2-Disubstituted Alkenes Anna I. Arkhypchuk,‡ Nicolas D’Imperio,‡ and Sascha Ott* Department of Chemistry − Ångström Laboratory, Uppsala University, Box 523, 75120 Uppsala, Sweden

Org. Lett. 2018.20:5086-5089. Downloaded from pubs.acs.org by KAOHSIUNG MEDICAL UNIV on 09/08/18. For personal use only.

S Supporting Information *

ABSTRACT: The phospha-Peterson reaction between a lithiated secondary phosphane, MesP(Li)TMS, and an aldehyde affords Mes-phosphaalkenes which, upon methanol addition and P-oxidation, react with a second carbonyl compound site specifically to produce unsymmetric alkenes. The E/Z selectivity of the one-pot cross coupling is largely determined by the electronic nature of the aryl substituent of the first aldehyde, with electron-donating groups giving rise to increased amounts of Z-alkenes.

T

Scheme 1. Synthetic Strategies for the Reductive Coupling of Two Carbonyl Compounds into Alkenesa

he olefination of carbonyl compounds is one of the most important transformations in organic chemistry, intimately linked to the names of Wittig, Horner, Wadsworth, and Emmons.1 Common in these name reactions is the presence of phosphorus groups in the olefination reagent, ylides (Wittig),2 phosphine oxides (Horner−Wittig, HW),3 or phosphonates (Horner−Wadsworth−Emmons, HWE),4 which act as reductants and oxygen acceptors to the carbonyl compound. The impact that these reactions has on fundamental and industrial chemistry cannot be overemphasized, but an interesting observation is that this type of chemistry has never been used to reductively couple two carbonyl compounds directly to alkenes. Such a reaction would be desirable because it would circumvent the synthesis of bromides or tosylate precursors that are typically necessary for the preparation of the olefination reagents. The only procedure for the reductive coupling of carbonyl compounds has until very recently been the McMurry coupling5 which uses low-valent titanium species as reductants to form Ti-coordinated carbonyl radicals that eventually dimerize to the desired alkene and TiO2.6 The McMurry coupling works well for many substrates7 but also suffers from general drawbacks that arise from the heterogeneous radical nature of the reaction, with long reaction times in high boiling solvents often leading to poor reproducibility. The most important disadvantage is the nonselectivity in the intermolecular coupling of two different carbonyl compounds. In such cases, the McMurry coupling will give rise to statistic mixtures of two symmetric alkenes in addition to the desired unsymmetric product (Scheme 1a).8 In a first proof-of-concept study, we have recently reported a conceptually different reductive aldehyde cross-coupling strategy that gives access exclusively to unsymmetric alkenes.9 Crucial to this one-pot methodology is the reaction of a phosphanylphosphonate10 reagent with a first aldehyde to produce a phosphaalkene intermediate. This phospha-HWE reaction11 proceeds with an Umpolung of the polarity of the carbonyl carbon, which promotes the ionic mechanism of the reaction. The phosphaalkene intermediate is subsequently © 2018 American Chemical Society

a

EWG = electron-withdrawing group.

activated by the addition of hydroxide to form a phosphine oxide (Scheme 1b) which reacts further with a second aldehyde to the unsymmetric alkene. The reaction works at ambient temperature within a few minutes and produces exclusively E-alkenes. Unfortunately, the substrate scope in the initial report was limited to aldehydes with electron-withdrawing groups (EWGs).9 Herein, we report a new procedure that offers significant improvements compared to the initially reported protocol in that it allows the reductive cross coupling of deactivated, electron-rich aldehydes and, to some extent, even ketones. This increased substrate scope is achieved by decreasing the steric bulk of the P-substituent at the phosphaalkene, as well as by an additional phosphorus oxidation that increases the acidity and nucleophilicity of the carbon center in the α-position of the resulting Received: June 5, 2018 Published: August 13, 2018 5086

DOI: 10.1021/acs.orglett.8b01754 Org. Lett. 2018, 20, 5086−5089

Letter

Organic Letters phosphinate (Scheme 1c). These modifications allow the direct, one-pot coupling of a vast variety of carbonyl compounds, giving access, for example, to biologically relevant stilbenoids such as resveratrol or pterostilbene.12 In the field of low-coordinate organo-phosphorus compounds, the use of the supermesityl group (2,4,6-tBu3Ph, Mes*) is a popular strategy to kinetically stabilize PC double bonds.13 While this strategy is valuable for the isolation and characterization of such compounds, it is disadvantageous for our purposes where the phosphaalkenes are intermediates and higher reactivity is desirable. Thus, a decreased kinetic stabilization as provided by the mesityl group (2,4,6(CH3)3Ph, Mes) was explored. A screening of literature reports of MesPCR 2 compounds shows that such phosphaalkenes are synthetically accessible but rather reactive. Characterized examples are almost exclusively those that carry two larger substituents at the phosphaalkene carbon,14 whereas those with only one such substituent are rare.15 The problem with these compounds that lack substantial steric stabilization is that they are prone to polymerization, in particular under the basic conditions that are frequently used for phosphaalkene preparation.16 Initial attempts to prepare the mesityl analogue of the Mes*phosphanylphosphonate reagent employed in our initial study were unsuccessful, and alternative access to Mes-phosphaalkenes had to be found. By adopting a procedure of Gates and co-workers,14 we were happy to find that a lithiated secondary phosphane MesP(Li)TMS (1, Scheme 2) was suitable for the conversion of aldehydes into Mes-phosphaalkenes 2 in a phospha-Peterson reaction.

Figure 1. 31P NMR spectroscopic investigation of the elementary steps of the reductive cross-coupling protocol. Shown is MesP(Li)TMS 1 (a) and its conversion to phosphaalkenes 2 (a → b) followed by the addition of MeOH to form 3 (b → c) which is oxidized to the phosphinate 4 (c → d). The last step of the procedure is the conversion of 4 to the alkene product and the phosphonate side product 5 (d → e). Conditions as in Scheme 2; R = 3,5-(OMe)C6H3.

corresponding phosphinites 3 (Scheme 2).17 The transformation of both E- and Z-2 is complete in less than 1 min at ambient temperature, as evident from the complete disappearance of any 31P NMR chemical shift beyond 200 ppm and the emergence of a diagnostic signal at ca. 130 ppm (Figure 1c). 17 Oxidation of 3 to the corresponding phosphinate 4 has been described before17 and could be applied to our one-pot procedure without difficulties. Conveniently, the addition of tBuOOH to etheral solutions of phosphinite 3 produces 4 at ambient temperature in a few minutes. The reaction is considerably faster than the oxidation of the second aldehyde that may already be present at this stage of the reaction. Intermediate 4 bears resemblance to HWE reagents, and a similar compound was indeed shown to react with aldehydes to form CC double bonds in the past.18 The addition of tBuOK and a second carbonyl to 4 leads to the desired unsymmetric alkene 6 and the phosphonate side product 5 (δ = +16 ppm, Figure 1e). Having established the general applicability and mechanism of the reductive cross-coupling procedure, it was investigated how the reduced steric demand of the P-substituent and the additional P-oxidation impact the substrate scope of the reaction. As discussed above, the previous procedure neither allowed the use of any deactivated, electron-rich benzaldehydes, nor any ketones as coupling partners. Important to note in this context is that olefinations of such compounds are generally difficult, and yields around 50% in classical HWE reactions of deactivated substrates are typical reference points.1 As shown in Table 1, 4-bromobenzaldehyde as an example of an electron-deficient benzaldehyde can be used as the first carbonyl compound, converted to the phosphaalkene, and then coupled with a second benzaldehyde also by the newly developed procedure. The coupling proceeds with a complete control of stereoselectivity, and the desired olefin is formed exclusively as the E-isomer (entries 1 and 2). Electron-deficient 3,5-dimethoxybenzaldehyde gives the same selectivity with different aldehydes, including deactivated ones that were previously not compatible as a second coupling partner

Scheme 2. Sequence for the Reductive Coupling of Two Carbonyl Compounds to Unsymmetric Di- and Trisubstituted Alkenes

Crucial to the success of this transformation is that it is conducted in diethyl ether, as other solvents such as THF give rise to a mixture of undesirable side products in addition to the phosphaalkene 2. Under the optimized conditions, 1 reacts with a variety of benzaldehydes bearing electron-withdrawing (EWGs) or electron-donating groups (EDGs) to the corresponding phosphaalkene at room temperature within minutes. The clean formation of 2 can conveniently be followed by 31P NMR spectroscopy which shows the emergence of two diagnostic downfield signals around 250 ppm that arise from the E- and Z-phosphaalkene isomers (shown for R = 3,5-(OMe)2C6H3 in Figure 1b). The stability of the produced phosphaalkenes against anionic polymerization16 depends strongly on the nature of the aryl substituent, with those carrying EWGs being less stable than the electronrich ones, as one may expect. Thus, the phosphaalkenes were quenched shortly after their formation by the addition of methanol to produce the 5087

DOI: 10.1021/acs.orglett.8b01754 Org. Lett. 2018, 20, 5086−5089

Letter

Organic Letters

using the initially reported sequence.9 The yields of the crosscoupled products are in the same range or sometimes higher than those typically reported for olefination reactions that involve deactivated aldehydes.1 Homocoupling products that would stem from aldehyde scrambling are not observed in any experiment described herein. When comparing entries 1−10, an interesting trend in E/Z selectivity of the formed stilbenes emerges, with the reaction becoming increasingly less selective for the E-isomer. This trend correlates with the electronic nature of the first aldehyde, with increasingly more deactivated aldehydes as the first coupling partner giving rise to more and more Z-products. In the most extreme case of the 4-methoxy-substituted benzaldehyde (entry 10), the Z-isomer is obtained in equal amounts as the E-isomer. No such correlation is found for the second aldehyde, the nature of which thus seems to be less important in determining the stereoselectivity of the stilbene product. Since the new cross-coupling protocol allows the use of aldehydes with EWG or EDG at either stage of the reaction, it should in principle be possible to direct the reaction toward the desired stilbene isomer by choosing the right order in which the aldehydes are added. Such possibilities, together with other means to direct E/Z ratios, are the subject of ongoing efforts, the results of which will be reported shortly. The yields reported in entries 1−10 in Table 1 are relative to the aldehydes, which were used in equimolar amounts. In case the second aldehyde is economically valuable, it can be used as the limiting substrate instead. For example, by adding the second aldehyde in a 1:2 ratio relative to the first one, the yields for the reductive cross coupling can be increased dramatically to up to 80% (entries 11−13). With the considerably higher substrate scope of the newly developed protocol, it was investigated whether the reactivity of the phosphinate intermediate would be sufficiently high to also promote the cross coupling of an aldehyde with a ketone. Thus, the coupling between the most reactive aldehyde with EWGs and acetone, cyclohexanone, and acetophenone was attempted. Simple adoption of the room-temperature protocol developed for the reactions in Table 1 did not afford any product. However, we were delighted to find that by raising the temperature during the reaction of the phosphinate and the ketones to refluxing THF the trisubstituted olefins indeed formed (Table 2). While the isolated yields are admittedly modest, it is important to realize that these results are the first of their kind, i.e., the first time that ketones have been reductively coupled to aldehydes under an ionic mechanism.

Table 1. Intermolecular Cross Coupling of Two Differently Substituted Benzaldehydes to Unsymmetric Stilbenes According to the Procedure Reported in Scheme 2

Table 2. Intermolecular Cross-Coupling of a Benzaldehyde with Ketones to Afford Trisubstituted Olefins According to the Procedure Reported in Scheme 2

a

Yields are calculated relative to the second aldehyde which was used in substoichiometric amounts (1:2 relative to the first aldehyde).

(entries 3−6). It is noteworthy that the products of these coupling reactions are precursors and derivatives of naturally occurring stilbenoids that have attracted considerable attention in recent years for various potential applications.19 One such stilbenoid, resveratrol,20 can be derived from the olefin synthesized in entry 3,21 while acidic cleavage of the THP group of the stilbene from entry 4 affords pterostilbene.22 The substrate scope is not limited to electron-deficient benzaldehydes in the first step, and the new procedure allows also the use of neutral (entries 7 and 8) and electron-rich aldehydes (entries 9 and 10) as the first carbonyl compound of the sequence. Such substrate combinations were not possible 5088

DOI: 10.1021/acs.orglett.8b01754 Org. Lett. 2018, 20, 5086−5089

Letter

Organic Letters

(2) (a) Wittig, G. Science 1980, 210, 600−604. (b) Kolodiazhnyi, O. I. In Phosphorus Ylides; Wiley-VCH: Weinheim, 2007; pp 359−538. (3) (a) Horner, L.; Hoffmann, H.; Wippel, H. G.; Klahre, G. Chem. Ber. 1959, 92, 2499−2505. (b) Clayden, J.; Warren, S. Angew. Chem., Int. Ed. Engl. 1996, 35, 241−270. (4) (a) Boutagy, J.; Thomas, R. Chem. Rev. (Washington, DC, U. S.) 1974, 74, 87−99. (b) Wadsworth, W. S., Jr. Org. React. 1977, 25, 7310. (5) (a) McMurry, J. E.; Fleming, M. P. J. Am. Chem. Soc. 1974, 96, 4708−4709. (b) McMurry, J. E. Chem. Rev. (Washington, DC, U. S.) 1989, 89, 1513−1524. (6) Ephritikhine, M. Chem. Commun. (Cambridge, U. K.) 1998, 0, 2549−2554. (7) Fürstner, A.; Bogdanović, B. Angew. Chem., Int. Ed. Engl. 1996, 35, 2442−2469. (8) McMurry, J. E.; Krepski, L. R. J. Org. Chem. 1976, 41, 3929− 3930. (9) Esfandiarfard, K.; Mai, J.; Ott, S. J. Am. Chem. Soc. 2017, 139, 2940−2943. (10) Esfandiarfard, K.; Arkhypchuk, A. I.; Orthaber, A.; Ott, S. Dalton Trans. 2016, 45, 2201−2207. (11) (a) Marinetti, A.; Mathey, F. Angew. Chem., Int. Ed. Engl. 1988, 27, 1382−1384. (b) Marinetti, A.; Bauer, S.; Ricard, L.; Mathey, F. Organometallics 1990, 9, 793−798. (c) Arkhypchuk, A. I.; Svyaschenko, Y. V.; Orthaber, A.; Ott, S. Angew. Chem., Int. Ed. 2013, 52, 6484−6487. (12) De Filippis, B.; Ammazzalorso, A.; Fantacuzzi, M.; Giampietro, L.; Maccallini, C.; Amoroso, R. ChemMedChem 2017, 12, 558−570. (13) (a) Klebach, T. C.; Lourens, R.; Bickelhaupt, F. J. Am. Chem. Soc. 1978, 100, 4886−4888. (b) Yoshifuji, M. J. Organomet. Chem. 2000, 611, 210−216. (14) Yam, M.; Chong, J. H.; Tsang, C. W.; Patrick, B. O.; Lam, A. E.; Gates, D. P. Inorg. Chem. 2006, 45, 5225−5234. (15) Bates, J. I.; Patrick, B. O.; Gates, D. P. New J. Chem. 2010, 34, 1660−1666. (16) (a) Tsang, C. W.; Yam, M.; Gates, D. P. J. Am. Chem. Soc. 2003, 125, 1480−1481. (b) Termaten, A.; van der Sluis, M.; Bickelhaupt, F. Eur. J. Org. Chem. 2003, 2003, 2049−2055. (c) Rawe, B. W.; Brown, C. M.; MacKinnon, M. R.; Patrick, B. O.; Bodwell, G. J.; Gates, D. P. Organometallics 2017, 36, 2520−2526. (17) Van der Knaap, T. A.; Klebach, T. C.; Lourens, R.; Vos, M.; Bickelhaupt, F. J. Am. Chem. Soc. 1983, 105, 4026−4032. (18) Horner, L.; Hoffmann, H.; Klink, W.; Ertel, H.; Toscano, G. Chem. Ber. 1962, 95, 581−601. (19) Csuk, R.; Albert, S.; Siewert, B.; Schwarz, S. Eur. J. Med. Chem. 2012, 54, 669−678. (20) Jang, M.; Cai, L.; Udeani, G. O.; Slowing, K. V.; Thomas, C. F.; Beecher, C. W. W.; Fong, H. H. S.; Farnsworth, N. R.; Kinghorn, A. D.; Mehta, R. G.; Moon, R. C.; Pezzuto, J. M. Science (Washington, DC, U. S.) 1997, 275, 218−220. (21) Li, Q. Q.; Shah, Z.; Qu, J. P.; Kang, Y. B. J. Org. Chem. 2018, 83, 296−302. (22) McCormack, D.; McFadden, D. J. Surg. Res. 2012, 173, 53−61.

The results should be considered as an outlook, and we are confident that higher yields for these kinds of substrate combinations are feasible using improved procedures in the future. In conclusion, we have reported a new methodology for the direct synthesis of unsymmetric alkenes from a variety of feedstock aldehydes. The attractiveness of the protocol is demonstrated by the one-step synthesis of stilbenoid natural products. Site selectivity is achieved by the sequential addition of the two carbonyl substrates at different stages of the reaction. Noteworthy is the fact that no homocoupled product was ever observed in any of the experiments described herein. The reaction is free of transition metals, proceeds at ambient temperature, and is not limited by the electronics of either of the aldehydes. Mechanistically, the one-pot protocol is a combination of a phospha-Peterson reaction that is followed by activation steps to afford phosphinates 4 that react with a second aldehyde in a HWE-like olefination. The high substrate scope is a result of the relatively moderate kinetic stabilization that is provided by the P-mesityl substituent, as well as the oxidation of an intermediate phosphinite 3 to phosphinate 4. The E/Z ratio of the produced alkenes is determined by the electronic nature of the first aldehyde. The substituent of the latter ends up at the phosphinates 4 in which EWGs give rise exclusively to E-alkenes, while destabilizing EDGs result in larger proportions of Z-alkenes. The high substrate scope of the developed protocol also allows the reductive cross coupling between an aldehyde and a ketone for the first time. Future work will include efforts to broaden the substrate scope as well as to improve the user friendliness of the intermolecular cross coupling even further.

■

ASSOCIATED CONTENT

* Supporting Information S

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.orglett.8b01754. Preparation of starting materials, general experimental procedures, and NMR spectroscopic data of final products (PDF)

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Sascha Ott: 0000-0002-1691-729X Author Contributions ‡

A.I.A. and N.D. contributed equally. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Notes

The authors declare no competing financial interest.

■ ■

ACKNOWLEDGMENTS Financial support from the Swedish Research Council is gratefully acknowledged. REFERENCES

(1) Edmonds, M.; Abell, A. In Modern Carbonyl Olefination; Takeda, T., Ed.; Wiley-VCH: Weinheim, 2004; pp 1−17. 5089

DOI: 10.1021/acs.orglett.8b01754 Org. Lett. 2018, 20, 5086−5089