Note Cite This: J. Org. Chem. 2019, 84, 1079−1084
pubs.acs.org/joc
Stereoselective Synthesis of 2,3,5-Trisubstituted Tetrahydrofurans Initiated by a Titanium−BINOLate-Catalyzed Vinylogous Aldol Reaction Patrick Hoffmeyer and Christoph Schneider* Institut für Organische Chemie, University of Leipzig, Johannisallee 29, D-04103 Leipzig, Germany
Downloaded via IOWA STATE UNIV on January 18, 2019 at 11:30:36 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
S Supporting Information *
ABSTRACT: The enantioselective synthesis of 2,3,5-trisubstituted tetrahydrofurans 3 has been achieved using a chiral titanium−BINOL complex as catalyst for the vinylogous Mukaiyama aldol reaction of bis(silyl) diendiolate 1 and an aldehyde. The ensuing BF3·OEt2-mediated Prins-type cyclization with a second aldehyde gave rise to 2,3,5-substituted tetrahydrofurans 3 with generally good yields and excellent stereocontrol. In this process, three new σ-bonds and three new stereogenic centers were generated in a one-pot process.
S
late 1 established in our group.13−17 On the basis of detailed mechanistic insight which we successfully obtained using labon-chip technology, we were able to develop an enantioselective version of this new process by taking advantage of methodology established by Keck et al. for similar vinylogous Mukaiyama aldol reactions (VMAR) of thioester-based dienolates. 18 Thus, a chiral titanium−BINOL catalyst generated in situ from Ti(OiPr)4, R-BINOL, and B(OMe)3 effected the VMAR of 1 and benzaldehyde in diethyl ether to furnish the intermediate vinylogous aldol product 2a. That in turn was treated with BF3·OEt2 and a second equivalent of the aldehyde in ethyl acetate to give rise to tetrahydrofuran 3a with good overall yield and enantioselectivity (Scheme 1). On that basis, we now report a quite generally applicable, straightforward, and enantioselective synthesis of 2,3,5-substituted tetrahydrofurans. In our first attempts to broaden the scope of this process, the enantiomeric ratios often happened to be significantly lower when we repeated the experiment. After extensive optimization, we found that, in our VMAR of bis(silyl) dienediolate 1 and aldehydes, the B(OMe)3 additive which was used in order to increase the reactivity of the catalyst had a detrimental effect on enantioselectivity. When we omitted it completely and ran the reaction under otherwise identical conditions just without B(OMe)3, the reaction became very reproducible and the yield
ubstituted tetrahydrofurans are common structural motifs found in several natural products and biologically active compounds. Therefore, in recent years, many efforts have been devoted toward the development of stereoselective methods to generate multisubstituted tetrahydrofurans. With regard to natural product synthesis and medicinal chemistry, enantioselective processes are particularly valuable but at the same time most challenging.1−3 The Prins cyclization using homoallyl alcohols and aldehydes has been shown to be a powerful tool for this purpose.4−8 In this context, Coates and co-workers developed a diastereoselective tetrahydrofuran synthesis employing cyclic allylsiloxanes with Lewis acids, but up to this point, no enantioselective modification has been reported.9 Suga et al. employed rhodium carbene complexes, leading to 1,3-dipoles, to generate tetrahydrofurans with good diastereoselectivity in a cycloaddition reaction. Nevertheless, an enantioselective modification required the use of a chiral auxiliary.10 Kočovský et al. developed a two-step allylation sequence of a bifunctional allyldisilane with aldehydes to generate tetrahydrofurans diastereo- and enantioselectively. However, the substrate scope was rather limited with only a very few enantioselective examples.11 Finally, the List group was able to generate tetrahydrofurans with high enantioselectivity by employing a novel Brønsted acid as chiral catalyst. The dienyl homoallylic alcohols used, however, had to be prepared separately prior to the actual synthesis.12 We recently reported a novel Lewis-acid-mediated one-pot process for the highly diastereoselective synthesis of 2,3,5trisubstituted tetrahydrofurans employing bis(silyl) dienedio© 2018 American Chemical Society
Received: November 5, 2018 Published: December 21, 2018 1079
DOI: 10.1021/acs.joc.8b02828 J. Org. Chem. 2019, 84, 1079−1084
Note
The Journal of Organic Chemistry Scheme 1. Synthesis of Enantioenriched Tetrahydrofuran 3a under Our Previous Conditions
Table 1. Substrate Scope with Different Aldehydesa−c
a Reaction conditions: 1 (1.50 mmol), aldehyde (0.75 mmol), BF3·OEt2 (0.75 mmol) in Et2O. bYields refer to chromatographically pure material over two steps. cDiastereomeric ratios determined by 1H NMR spectroscopy on crude products, and enantiomeric ratios determined by HPLC analysis on chiral stationary phases.
also increased substantially from 65 to 86%. In addition, extending the reaction time at −78 °C from 0.5 to 5 h ensured that the vinylogous aldol reaction proceeded to completion at low temperature and gave rise to an improved enantiomeric ratio of 94:6 (Table 1). Other chiral ligands with substituents
in the BINOL backbone and other solvents displayed diminished reactivity and enantioselectivity in this reaction. The reason for this improvement is not entirely clear at the moment. Yamamoto’s observation of a Brønsted-assisted Lewis acid catalyst prepared from BINOL and B(OMe)3 in situ for 1080
DOI: 10.1021/acs.joc.8b02828 J. Org. Chem. 2019, 84, 1079−1084
Note
The Journal of Organic Chemistry Diels−Alder and aldol reactions,19 however, suggests that this catalyst might have been formed in varying amounts, as well, and eventually foiled our VMAR. As a control experiment, this very catalyst was employed in the THF synthesis on purpose, yielding the desired product in only 22% yield and with 63:37 er; thus, the undesired formation of this catalyst during the reaction might be a reasonable explanation for the decreased er when using B(OMe)3 as an additive. With the optimized conditions in hand, the scope of the reaction was investigated for the synthesis of 2,5 identically substituted tetrahydrofurans. Therefore, several aromatic and heteroaromatic aldehydes were subjected to the reaction and delivered the products 3a−3f with good chemical yields and generally good diastereo- and enantiocontrol (Table 1). We then turned to the synthesis of tetrahydrofurans with different 2- and 5-substituents, which required the use of different aldehydes for the VMAR and the Prins reaction. Here, aromatic aldehydes were found to be most suitable as substrates for the VMAR and provided tetrahydrofurans with generally good yields and enantioselectivities irrespective of the nature of the second aldehyde used for the Prins reaction (e.g., 3g, 3i, 3j, 3l−3o). Although yields for heteroaromatic aldehydes were also good, the enantioselectivity was somewhat lower. Aliphatic aldehydes tend to give slightly poorer results as substrates for the VMAR as documented by tetrahydrofurans 3h and 3k. The α-branched aliphatic aldehydes showed a significant loss in yield and enantioselectivity. In cases of less satisfactory diastereoselectivity, the diastereomers could be easily separated by column chromatography with silyl-endcapped silica gel.20 Tetrahydrofuran 3n could be recrystallized from hexane/ MTBE/CH2Cl2 (12:1:1, v/v/v) to give crystals of >99% ee suitable for X-ray crystallographic analysis for the determination of the absolute configuration, which was used for all other products by analogy (see the Supporting Information for details). In summary, we have developed a highly diastereo- and enantioselective synthesis of trisubstituted tetrahydrofurans based upon a sequence comprising a vinylogous Mukaiyama aldol and Prins reaction. A chiral titanium−BINOL catalyst prepared easily in situ was employed to control the enantioselectivity of the reaction effectively. Different commercially available aromatic, aliphatic, and heteroaromatic aldehydes along with the readily available bis(silyl) dienediolate nucleophile 1 were employed in this process successfully and delivered the products in generally good yields. Key to the success of this process was a modification of the original procedure for the VMAR by leaving out the B(OMe)3 additive.
■
or with a solution of phosphomolybdic acid hydrate (5 g in 250 mL of ethanol). The nucleophile 1 was prepared using an improved procedure established in our group.14 1H and 13C NMR spectra were recorded in CDCl3 at 26 °C using a Mercury Plus 300 MHz and a Bruker Avance DRX 400 MHz spectrometer. Spectra were referenced to residual chloroform (7.26 ppm, 1H; 77.16 ppm, 13C). Chemical shifts are reported in parts per million; multiplicities are indicated by s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), and related permutations. All ESI mass spectra were recorded on a Bruker Daltonics Apex II FT-ICR. IR spectra were obtained using a Jasco 4100 FT-IR and UV spectra using a Jasco V630 UV−vis spectrometer. General Procedure for the Enantioselective Synthesis of 2,3,5-Trisubstituted Tetrahydrofurans. In an oven-dried 10 mL two-neck round-bottom flask under a light argon pressure, 150 mg of 4 Å molecular sieves and R-BINOL (0.43 mg, 0.15 mmol, 20 mol %) were suspended in 2.0 mL of anhydrous Et2O. Ti(OiPr)4 (22 μL, 0.075 mmol, 10 mol %) was added, and the suspension turned brownred. After being heated to 40 °C for 1 h, R1CHO (0.75 mmol, 1.0 equiv) was added at 25 °C and stirred for 20 min. The mixture was cooled to −78 °C; bis(silyl) dienediolate 1 (1.5 mmol, 2.0 equiv) was added dropwise and stirred for 5 h at −78 °C and then warmed to 25 °C. After 4 days, 10 mL of saturated NaHCO3 and 5 mL of dichloromethane were added. After extraction with dichloromethane (3 × 5 mL), the combined organic layers were dried over Na2SO4 and filtered and the solvent was removed under reduced pressure to afford an orange oil. Without further purification, the crude product was dissolved in 1.5 mL of anhydrous ethyl acetate. R2CHO (0.75 mmol, 1.0 equiv) and BF3·OEt2 (0.75 mmol, 1.0 equiv) were added dropwise and stirred for 16 h at 25 °C. Five milliliters of saturated NaHCO3 solution and 5 mL of dichloromethane were added. After extraction with dichloromethane (3 × 5 mL), the combined organic layers were dried over Na2SO4 and filtered and the solvent was removed under reduced pressure. The crude product was preabsorbed on deactivated silica gel and purified by flash column chromatography using EtSiCl3functionalized silica gel. Ethyl 2-((2S,3R,5R)-2,5-Diphenyltetrahydrofuran-3-yl)-2-oxoacetate (3a). According to the general procedure, the reaction was run with benzaldehyde (77 μL, 0.75 mmol, 1.0 equiv) and benzaldehyde (77 μL, 0.75 mmol, 1.0 equiv). Purification by flash column chromatography using EtSiCl3-functionalized silica gel starting from hexane to hexane/MTBE (20:1, v/v) yielded the title compound (211 mg, 0.65 mmol, 86%, 97:3 dr, 94:6 er) as a white solid: Rf (hexane/MTBE = 3:1, v/v) = 0.42; 1H NMR (400 MHz, CDCl3) δ = 7.61−7.51 (m, 2H), 7.48−7.24 (m, 8H), 5.41 (d, J = 9.0 Hz, 1H), 5.02 (dd, J = 10.0, 6.0 Hz, 1H), 4.58 (ddd, J = 9.0, 8.5, 8.0 Hz, 1H), 4.00−3.78 (m, 2H), 2.70−2.55 (m, 1H), 2.52−2.39 (m, 1H), 1.09 (t, J = 7.0 Hz, 3H) ppm; 13C{1H} NMR (101 MHz, CDCl3) δ = 193.3, 161.0, 140.6, 137.2, 128.6 (2 × PhCH), 128.5 (2 × PhCH), 128.4, 128.1, 127.7 (2 × PhCH) 126.6 (2 × PhCH), 82.5, 81.1, 62.4, 52.3, 35.6, 13.8 ppm. The NMR spectroscopic data are in agreement with those previously reported for the racemic compound.13 HPLC conditions: Chiracel ASH column, hexane/2propanol = 90:10; flow rate = 1.0 mL min−1, enantiomer 1 tR = 9.26 min, enantiomer 2 tR = 12.78 min; [α]D24 = +25.3 (c 0.85, CH2Cl2). Ethyl 2-((2S,3R,5R)-2,5-Di-p-tolyltetrahydrofuran-3-yl)-2-oxoacetate (3b). According to the general procedure the reaction was run with 4-methylbenzaldehyde (89 μL, 0.75 mmol, 1.0 equiv) and 4methylbenzaldehyde (89 μL, 0.75 mmol, 1.0 equiv). Purification by flash column chromatography using EtSiCl3-functionalized silica gel starting from hexane to hexane/MTBE (50:1, v/v) yielded the title compound (212 mg, 0.60 mmol, 80%, dr = 95:5, er = 85:15) as a colorless solid: Rf (hexane/MTBE = 3:1, v/v) = 0.47; 1H NMR (400 MHz, CDCl3) δ = 7.44 (d, J = 8.0 Hz, 2H), 7.20 (dd, J = 8.0, 7.0 Hz, 4H), 7.09 (d, J = 8.0 Hz, 2H), 5.36 (d, J = 9.0 Hz, 1H), 4.97 (dd, J = 10.0, 6.0 Hz, 1H), 4.54 (q, J = 8.5 Hz, 1H), 3.96−3.84 (m, 2H), 2.59 (ddd, J = 13.0, 10.0, 9.0 Hz, 1H), 2.45−2.39 (m, 1H), 2.37 (s, 3H), 2.30 (s, 3H), 1.09 (t, J = 7.0 Hz, 3H) ppm; 13C{1H} NMR (101 MHz, CDCl3) δ = 193.4, 161.1, 138.1, 137.7, 137.6, 134.3, 129.24 (2 × ArCH), 129.16 (2 × ArCH), 127.7 (2 × ArCH), 126.6 (2 ×
EXPERIMENTAL SECTION
General Methods. All reactions were carried out in oven-dried glassware under an Ar atmosphere unless otherwise noted. Solvents were distilled from the indicated drying reagents: dichloromethane (CaH2), diethyl ether (Na, benzophenone), tetrahydrofuran (Na, benzophenone). Other solvents were of technical grade and distilled from the indicated drying reagents: dichloromethane (CaH2), methyltert-butylether (KOH), n-hexane (KOH). Ethyl acetate (Acros Organics, extra dry, with molecular sieves, water
