Reinhard W.
Hoffmann
*,
Jochen
Krüger
and
David
Brückner
Fachbereich Chemie, Philipps-Uniersität Marburg, Hans-Meerwein-Strasse, D-35032, Marburg, Germany. E-mail: rwho@chemie.uni-marburg.de; Fax:
+ 49 6421 2828917
First published on 29th September 2000
Alkoxyalkynes 9 may be hydroborated with pinacol borane, preferentially under Cp2ZrHCl catalysis, to give the vinylboronates 10. The latter, when subjected to the Matteson–Brown homologation with LiCH2Cl, give rise to the (E)-γ-alkoxyallylboronates 3 in good yield. This reaction sequence has been used to generate the (E)-γ-alkoxyallylboronates 14, 21, 26 and 31, which were the starting point for intramolecular allylboration reactions leading to the trans-disubstituted tetrahydropyrans 8 and 22, as well as hydrooxepans 27 and 32.
![]() | ||
Scheme 1 |
The application of these reactions in synthesis is, however, compromised by the fact that not all γ-heterosubstituted allylboronates of interest are equally well accessible as both the (Z)- and the (E)-isomers. Problems are associated with the generation of the (E)-γ-alkoxyallylboronates, for which our initial procedure4 is moderately practicable at best.5 More recently, Miyaura and co-workers described a second route to these species by an iridium-catalyzed isomerization of γ-alkoxyvinylboronates 2 (Scheme 2).6
Yet, when structures such as 4 were to be attained by allylmetallation reactions, chemists frequently resorted to a multistep route (Scheme 2) involving the (E)-γ-silylallylboronates 5 as reagents, followed by conversion of the silyl residue in the adduct
6 to the desired oxygen functionality.7 Obviously, the route ia
the γ-silylallyl boronates 5 is not viable when one is interested in
an intramolecular allylboration (Scheme 3) that should lead to
heterocyclic compounds such as 8.
![]() | ||
Scheme 2 |
![]() | ||
Scheme 3 |
In the search for another route to (E)-γ-alkoxyallylboronates such
as 7
we envisaged a homologation of vinylboronates to allylboronates,9iz. the conversion of 10 into 3 (Scheme 3), a method that is increasingly gaining importance.10 This route would rely
on the hydroboration of an alkoxyalkyne 9 to give 10. We hoped
that the direct hydroboration of readily accessible alkoxyalkynes
911 with pinacol borane12 should afford the desired vinylboronate 10.
![]() | ||
Scheme 4 |
The clean conversion of the alcohol 11ia the alkoxyalkyne 12
and the allylboronate 14 into the trans-vinyltetrahydropyranol 8 demonstrates the versatility of this route to (E)-γ-alkoxyallylboronates and their use in intramolecular allylboration
reactions. In view of the general interest in trans-fused annelated oligo-tetrahydropyrans,16 we wanted to explore the subsequent conversion of 8 into the bicyclic bis-tetrahydropyran 22. This required the conversion of the vinyl group in 8 into a latent aldehyde function, such as a dimethyl acetal. This was accomplished in a series of standard transformations (Scheme
5): protection of the alcohol as an acetate 15
(86%), hydroboration of 15 to give the primary alcohol 16 using di-sec-isoamylborane (70%), Swern oxidation of the primary alcohol
16 to an aldehyde followed by in situ acetalization to give 17
(90%), and finally, cleavage of the acetate in 17 by K2CO3 in methanol to give the alcohol 18 (92%).
![]() | ||
Scheme 5 |
The alcohol 18 is the starting point for a second round of intramolecular allylboration (Scheme 6). To this end, the alcohol 18 was converted to the alkoxyalkyne 19 as above (78%). We found that the hydroboration of alkoxyalkynes 9 to give the vinylboronate 10 (Scheme 3) could be carried out at room temperature when catalyzed with zirconocene hydrido chloride.17 This allowed us to similarly convert the alkoxyalkyne 19 in 87% yield to the vinylboronate 20. Subsequent transformation of 20 into the allylboronate 21 (95%) proceeded as usual. Liberation of the aldehyde from 21 with LiBF4 in moist acetonitrile18 initiated the intramolecular allylboration, which resulted in the formation of 76% of the desired bicyclic bis-tetrahydropyran 22. The latter was obtained as a single diastereomer, attesting to the high asymmetric induction from the stereocenters resident in the cyclization precursor 21. For alternate syntheses of 22 see ref. 19.
![]() | ||
Scheme 6 |
We recently showed20 that hydroformylation is a good way to generate an aldehyde function in the presence of an allylboronate moiety. This opens the way to a domino hydroformylation–allylboration–hydroformylation sequence, allowing rapid access to annelated heterocyclic structures. It was therefore tempting to combine this technique with the generation of (E)-γ-alkoxyallylboronates described above. Model studies of this point were carried out starting from the homoallylic alcohol 23 (Scheme 7). The latter was converted to the alkoxyalkyne 24 in 78% yield. Hydroboration of 24 was effected with pinacol borane catalyzed by zirconocene hydrido chloride17 (69%). Subsequent homologation of 25 afforded the allylboronate 26 (65% over two steps).
![]() | ||
Scheme 7 |
Hydroformylation of the latter (Scheme 7) was carried out in the presence of the BIPHEPHOS ligand21 to attain a high preference for the formation of the linear aldehyde over the branched aldehyde. This, however, entailed that the hydroformylation reaction became rather slow, requiring two days at 65°C and 5 bar of CO–H2. In the end this treatment resulted in the expected domino hydroformylation–allylboration–hydroformylation reaction to give 48% of the lactols 27, which were obtained as a mixture of anomers. The products obtained could be separated into two diastereomeric sets of lactols 27a and 27b, which were obtained in a ca. 1:1 ratio. To facilitate structure assignment, each of these mixtures was oxidized22 with the Dess–Martin reagent23 to a single lactone 28a and 28b, respectively. Assignment of the relative configuration was made by NMR NOE spectroscopy.
The fact that the two cyclization products 27a and 27b were obtained in a nearly 1:1 ratio indicates that asymmetric induction on the formation of a seven-membered hydro-oxepane ring was low.
At this point we did not know whether this was a singular event or characteristic of a more general feature. For this reason we investigated a second domino hydroformylation–allylboration–hydroformylation reaction leading to an annelated hydro-oxepane ring. To this end we converted the vinyltetrahydropyranol 8 into the alkoxyalkyne 29 (72%, Scheme 8). Zirconocene hydrido chloride catalyzed hydroboration,17 followed by homologation, furnished the allylboronate 31 (50% over two steps). The allylboronate was subjected to the hydroformylation conditions. The reaction turned out to be rather slow, requiring more than three days at 65°C. The products obtained (52%) were not the lactols, but the corresponding enol ethers 32. Apparently dehydration, to give 32, had occurred during the longer reaction period. The tricyclic compounds were again obtained as a 1:1 mixture of stereoisomers, presumably the trans-syn-trans32a and the trans-anti-trans isomer 32b. The individual structures were not assigned to the materials obtained.
![]() | ||
Scheme 8 |
This made it clear that intramolecular allylboration to form hydro-oxepane rings is stereounselective, showing no asymmetric induction from the resident stereocenters present in the cyclization precursors 26 or 31. This contrasts to the high asymmetric induction found on intramolecular allylboration to give the tetrahydropyran ring in the cyclization of 21 to 22.
In
conclusion, we have found a reliable route to (E)-γ-alkoxyallylboronates starting from alcohols, ia alkoxyalkynes, hydroboration and homologation. The γ-alkoxyallylboronates 13, 21, 26
and
31 served as substrates for intramolecular allylboration reactions,
giving rise to tetrahydropyrans 8 and 22 and the hydro-oxepane ring systems 27 and 32.
A solution of n-butyllithium in hexane (1.49 M, 4.36 mL, 6.50 mmol) was added under an argon atmosphere at − 105°C into a solution of the vinylboronate 13 (1.644 g, 5.03 mmol) and chloroiodomethane (473 μL, 6.5 mmol) in THF (25 mL). After reaching room temperature overnight pH 7 buffer solution (20 mL) was added, the phases were separated and the aqueous phase was extracted with ether (3 × 10 mL). The combined organic phases were washed with brine (10 mL), dried (Na2SO4) and concentrated. NMR analysis of the residue showed the presence of a 4:1 mixture of 14 :13, corresponding to a yield of 14 of 78%. The following NMR data of 14 could be recorded: 1H-NMR (300 MHz, CDCl3): δ = 1.45 (br d, J = 7.21 Hz, 2H), 4.74 (dt, J = 12.6 and 7.5 Hz, 1H), 6.18 (d, J = 12.6 Hz, 1H). 13C-NMR (75 MHz, CDCl3): δ = 24.7, 29.0, 52.6, 68.6, 83.1, 98.9, 104.3, 146.1.
The mixture of 13 and 14 obtained was taken up in acetonitrile (30 mL). Water (0.6 mL) and ytterbium triflate (312 mg, 0.1 equiv.) were added. After stirring for 12 h ether (20 mL) and saturated aqueous NaHCO3 solution (10 mL) were added. The phases were separated and the aqueous phase was extracted with ether (3 × 10 mL). The combined organic phases were washed with brine (10 mL), dried (Na2SO4) and concentrated by distillation over a short column. The residue was purified by flash chromatography with pentane–ether (1:1) and the eluates were concentrated at 0°C to leave the vinyltetrahydropyranol 8 (419 mg, 66%) as a single diastereomer. 1H-NMR (300 MHz, CDCl3): δ = 1.46 (m, 1H), 1.66–1.80 (m, 3H), 2.14 (m, 1H), 3.27–3.50 (m, 3H), 3.94 (m, 1H), 5.31 (ddd, J = 10.5, 1.6, 0.8 Hz, 1H), 5.38 (ddd, J = 17.3, 1.7, 1.2 Hz, 1H), 5.87 (ddd, J = 17.5, 10.6 7.0 Hz, 1H). 13C-NMR (75 MHz, CDCl3): δ = 25.3, 31.6, 67.4, 69.5, 84.0, 118.8, 136.2. These data corresponded to a sample that was prepared according to the procedure given by Yamamoto et al.15
Chloroiodomethane (105 μL, 1.4 mmol) was added to a solution of the vinylboronate 20 (380 mg, 1.11 mmol) in THF (5.5 mL) and the solution was cooled to − 105°C. n-Butyllithium (1.22 M in hexane, 1.18 mL, 1.44 mmol) was added and the mixture was allowed to reach room temperature overnight. tert-Butyl methyl ether (20 mL) was added and the mixture was poured into pH 7 buffer solution (20 mL). The phases were separated and the aqueous phase was extracted with tert-butyl methyl ether (3 × 20 mL). The combined organic phases were dried (Na2SO4) and concentrated, leaving the allylboronate 21 (375 mg, 95%) as a slightly yellowish oil. Rf(PE–EtOAc 5:1) = 0.37. 1H-NMR (300 MHz, CDCl3): δ = 1.24 (s, 12H), 1.32–1.51 (m, 3H), 1.58 (ddd, J = 14.4, 9.0, 3.7 Hz, 1H), 1.63–1.73 (m, 2H), 2.13–2.26 (m, 2H), 3.20–3.38 (m, 3H), 3.31 (s, 3H), 3.33 (s, 3H), 3.82–3.91 (m, 1H), 4.63 (dd, J = 8.3, 3.7 Hz, 1H), 4.92 (dt, J = 12.5, 7.6 Hz, 1H), 6.06 (dt, J = 12.5, 1.0 Hz, 1H). 13C-NMR (75 MHz, CDCl3): δ = 24.7 (4C), 25.3, 29.6, 35.3, 52.0, 53.2, 67.5, 77.1, 77.9, 83.2 (2C), 101.5, 101.6, 144.7.
The allylboronate 21 (61 mg, 0.17 mmol) was taken up in acetonitrile (0.86 mL). Water (17 μL) and lithium tetrafluoroborate (48 mg, 0.51 mmol) were added. After 1 day tert-butyl methyl ether (20 mL) and saturated aqueous NaHCO3 solution (10 mL) were added. The phases were separated and the aqueous phase was extracted with tert-butyl methyl ether (5 × 10 mL). The combined organic phases were dried (Na2SO4) and concentrated. Flash chromatography of the residue with pentane–diethyl ether (1: 5) furnished 22 (24 mg, 76%) as a colorless oil. Rf(diethyl ether) = 0.35. 1H-NMR (500 MHz, CDCl3): δ = 1.41–1.48 (m, 1H), 1.52 (q, J = 11.3 Hz, 1H), 1.69–1.77 (m, 2H), 1.83 (br d, J = 3.4 Hz, 1H), 2.06–2.11 (m, 1H), 2.40 (dt, J = 11.6, 4.3 Hz, 1H), 3.00–3.10 (m, 2H), 3.35–3.48 (m, 2H), 3.57 (dd, J = 8.7, 7.6 Hz, 1H), 3.89–3.94 (m, 1H), 5.35 (ddd, J = 10.5, 1.6, 0.8 Hz, 1H), 5.43 (ddd, J = 17.3, 1.6, 1.0 Hz, 1H), 5.85 (ddd, J = 17.4, 10.4, 7.2 Hz, 1H). NOE contacts between H-1 and H-3; H-4 and H-6; H-5ax and H-1; H-5ax and H-3. 13C-NMR (75 MHz, CDCl3): δ = 25.5, 29.2, 37.8, 67.8, 69.0, 76.7, 77.6, 83.8, 119.6, 135.6. MS (EI) m/z (%) 28 (33), 43 (30), 55 (32), 71 (34), 84 (100), 127 (44). HRMS (EI): C10H16O3 requires 184.1099; found 184.1101.
The vinylboronate 25 was converted to the allylboronate 26 as described for 22 (200 mg, 95%) as a yellowish oil. Rf(PE–EtOAc 5:1) = 0.73. 1H-NMR (200 MHz, CDCl3): δ = 0.83 (d, J = 6.8 Hz, 3H), 0.84 (d, J = 6.8 Hz, 3H), 1.17 (s, 12H), 1.35–1.45 (m, 2H), 1.65–1.86 (m, 1H), 2.16–2.25 (m, 2H), 3.32 (q, J = 5.7 Hz, 1H), 4.83 (dt, J = 12.1, 7.4 Hz, 1H), 4.92–5.06 (m, 2H), 5.76 (ddt, J = 17.1, 10.1, 7.0 Hz, 1H), 5.99 (dt, J = 12.3, 1.5 Hz, 1H). 13C-NMR (75 MHz, CDCl3): δ = 17.8, 18.4, 24.8 (4C), 31.1, 35.6, 83.1 (2C), 84.9, 100.9, 116.6, 135.1, 146.3.
Rh(CO)2acac (1.8 mg, 0.007 mmol) and BIPHEPHOS21 (10.7 mg, 0.014 mmol) were dissolved in THF (1 mL). After 10 min a solution of the crude allylboronate 26 (190 mg, 0.68 mmol) in THF (2 mL) was added. The mixture was treated in an autoclave with 5 bar of CO–H2 (1:1) for 2 days at 65°C. The mixture was concentrated and the residue was purified by flash chromatography with pentane–tert-butyl methyl ether (3:1) to give first the 1,3-cis-compound 27a (37 mg) as an anomeric mixture and then the 1,3-trans-compound 27b (33 mg) as an anomeric mixture.
27a: Rf(PE–EtOAc 5:1) = 0.24. 1H-NMR (300 MHz, CDCl3): characteristic signals δ = 0.86 (d, J = 6.8 Hz), 0.87 (d, J = 6.6 Hz, together 3H), 0.92 (d, J = 6.8 Hz), 0.94 (d, J = 6.8 Hz, together 3H), 4.76–4.82 (m, 0.55H), 5.20 (t, J = 3.1 Hz, 0.45H). 13C-NMR (75 MHz, CDCl3): characteristic signals δ = 90.8, 96.1.
27b: Rf(PE–EtOAc 5:1) = 0.17. 1H-NMR (300 MHz, CDCl3): characteristic signals δ = 0.84 (d, J = 6.6 Hz), 0.85 (d, J = 6.8 Hz, together 3H), 0.99 (d, J = 6.6 Hz), 1.00 (d, J = 6.8 Hz, together 3H), 4.74 (m, 0.55H), 5.19 (t, J = 2.5 Hz, 0.45H). 13C-NMR (75 MHz, CDCl3): characteristic signals δ = 90.3, 95.6.
To facilitate characterization the lactols 27 were oxidized to the lactones 28: 27a (28 mg, 0.13 mmol) was taken up in dichloromethane (1.3 mL). Pyridine (53 μL, 0.65 mmol) and Dess–Martin periodinane (66 mg, 0.16 mmol) were added at 0°C. After stirring for 8 h at room temperature the mixture was concentrated and the residue was purified by flash chromatography with pentane–tert-butyl methyl ether (5:1) to give the 1,3-cis-lactone 28a (15 mg, 54%) as a colorless oil. Rf(PE–EtOAc 5:1) = 0.29. 1H-NMR (500 MHz, CDCl3): δ = 0.88 (d, J = 6.8 Hz, 3H), 0.92 (d, J = 6.8 Hz, 3H), 1.55–1.83 (m, 5H), 1.86 (ddd, J = 9.5, 9.0, 7.6 Hz, 1H), 1.88 (ddd, J = 9.6, 9.1, 7.6 Hz, 1H), 2.11–2.19 (m, 2H), 2.55 (dt, J = 17.6, 7.7 Hz, 1H), 2.74 (ddd, J = 17.6, 8.9, 5.4 Hz, 1H), 3.27 (ddd, J = 8.7, 6.3, 5.4 Hz, 1H), 3.41 (td, J = 9.6, 6.2 Hz, 1H), 4.04 (td, J = 9.3, 4.8 Hz, 1H). NOE contacts between H-1 and H-3. 13C-NMR (125 MHz, CDCl3): δ = 18.4, 18.9, 19.4, 26.0, 27.9, 31.5, 32.3, 34.0, 78.0, 82.3, 86.1, 171.4. MS (EI) m/z (%): 41 (35), 43 (42), 55 (55), 57 (52), 82 (76), 85 (100), 109 (40), 169 (34).
The 1,3-trans-lactol 27b furnished in the same manner the 1,3-trans-lactone 28b (20 mg, 63%) as a colorless oil. Rf(PE–EtOAc 5:1) = 0.21. 1H-NMR (500 MHz, CDCl3): δ = 0.87 (d, J = 6.8 Hz, 3H), 0.99 (d, J = 6.8 Hz, 3H), 1.38–1.55 (m, 3H), 1.69 (ddt, J = 13.4, 8.2, 6.7 Hz, 1H), 1.80–1.87 (m, 1H), 1.93 (dtd, J = 13.7, 8.6, 6.8 Hz, 1H), 1.97–2.03 (m, 1H), 2.06–2.13 (m, 1H), 2.25–2.32 (m, 1H), 2.49 (ddd, J = 17.3, 9.0, 6.4 Hz, 1H), 2.72 (dt, J = 17.3, 6.6 Hz, 1H), 3.31 (ddd, J = 11.3, 8.4, 5.3 Hz, 1H), 3.67 (td, J = 8.7, 6.4 Hz, 1H), 4.04 (ddd, J = 10.4, 9.5, 4.9 Hz, 1H). No NOE contacts between H-1 and H-3. 13C-NMR (125 MHz, CDCl3): δ = 18.9, 19.7, 20.5, 27.1, 28.2, 31.9, 32.9, 35.4, 68.1, 82.3, 83.2, 171.1. MS (EI) m/z (%): 41 (35), 43 (40), 55 (56), 81 (27), 85 (73), 95 (34), 101 (35), 109 (29), 123 (47), 169 (100). HRMS(EI): C12H20O3 requires 212.1412; found 212.1411.
The vinylboronate 30 was converted into the allylboronate 31 as described for 22. The crude product was purified by filtration over silica gel with pentane–tert-butyl methyl ether (5:1). The allylboronate 31 (263 mg, 78%) was obtained as a colorless oil. Rf(PE–EtOAc 5:1) = 0.49. 1H-NMR (300 MHz, CDCl3): δ = 1.24 (s, 12H), 1.40–1.53 (m, 3H), 1.60–1.76 (m, 2H), 2.11–2.28 (m, 1H), 3.28–3.49 (m, 2H), 3.58–3.70 (m, 1H), 3.88–3.99 (m, 1H), 4.92 (dt, J = 12.3, 7.5 Hz, 1H), 5.20 (dt, J = 10.7, 1.7 Hz, 1H), 5.34 (dt, J = 17.3, 1.7 Hz, 1H), 5.97 (ddd, J = 17.5, 10.8, 5.5 Hz, 1H), 6.04 (dt, J = 12.4, 1.6 Hz, 1H). 13C-NMR (75 MHz, CDCl3): δ = 24.8 (4C), 25.1, 29.7, 67.3, 78.2, 80.5, 83.2 (2C), 101.9, 116.6, 136.1, 144.9.
The hydroformylation of the allylboronate 31 was carried out as described for 28 for 80 h at 65°C under 5 bar of CO–H2 (1:1). Flash chromatography of the crude product with pentane–tert-butyl methyl ether (5:1) furnished 32 (97 mg, 52%) as a 1:1 mixture of the two diastereomers. This mixture was not separated. Rf(PE–EtOAc 5:1) = 0.41. 1H-NMR (500 MHz, CDCl3): δ = 1.36–1.54 (m, 1.5H), 1.57–1.76 (m, 2.5H), 1.85–2.01 (m, 2H), 2.01–2.34 (m, 4H), 3.06 (ddd, J = 9.1, 7.2, 4.8 Hz, 0.5H), 3.13 (ddd, J = 10.4, 9.1, 5.0 Hz, 0.5H), 3.20 (ddd, J = 10.6, 9.1, 4.9 Hz, 0.5H), 3.26 (ddd, J = 11.1, 9.1, 4.1 Hz, 0.5H), 3.29–3.35 (m, 1H), 3.58–3.63 (m, 1H), 3.66 (td, J = 9.5, 2.7 Hz, 0.5H), 3.71 (td, J = 9.8, 5.7 Hz, 0.5H), 3.84 (ddt, J = 11.1, 4.0, 2.0 Hz, 0.5H), 3.89 (ddt, J = 11.3, 3.9, 2.0 Hz, 0.5H), 4.64 (td, J = 5.6, 2.1 Hz, 0.5H), 4.69 (td, J = 5.5, 2.3 Hz, 0.5H), 6.26 (dt, J = 5.9, 2.0 Hz, 0.5H), 6.28 (dt, J = 5.9, 1.9 Hz, 0.5H). 13C-NMR (75 MHz, CDCl3): δ = 25.7, 26.0, 28.24, 28.27, 28.8, 29.0, 29.7, 31.4, 31.9, 32.0, 67.3, 68.0, 71.5, 74.9, 78.2, 79.3, 79.7, 81.4, 81.8, 83.4, 99.0, 99.6, 143.89, 143.91. MS (EI) m/z (%): 41 (26), 43 (26), 55 (30), 71 (79), 81 (41), 94 (29), 97 (100), 141 (32), 210 (33). HRMS(EI): C12H18O3 requires 210.1256; found 210.1254.
Footnote |
† For part LII, see ref. 1. |
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2001 |