Michael
Waibel
and
Nicolai
Cramer
*
Laboratory of Organic Chemistry, ETH Zurich, HCI H 304, Wolfgang-Pauli-Strasse 10, 8093 Zurich, Switzerland. E-mail: nicolai.cramer@org.chem.ethz.ch; Fax: +41 44 632 1328; Tel: +41 44 632 6412
First published on 31st August 2010
Rhodium-catalyzed desymmetrizations of meso-tert-norbornenols by retro-allylations generate allyl-rhodium species that allow for a rich and diverse downstream reactivity.
Herein, we report the complementary reaction profile of allyl rhodium species 3s/3p generated by desymmetrization of bicyclic meso-tert-norbornenols 1 and illustrate their synthetic potential and variable reactivity (Scheme 1).
Scheme 1 Desymmetrization of meso-tert-norbornenols 1 by metal promoted retro-allylation. |
A simultaneous coordination of the alkoxide and the olefin moiety7 of 4 to a chiral Rh(I)-complex induces an enantioselective retro-allylative C–C bond cleavage leading to allyl rhodium species either as their σ- (5s) or a π-bound (5p) complex (Scheme 2).8β-Hydride elimination forms regioisomeric dienes 69 and a rhodium hydride species, which in turn, is prone to re-addition to the formed diene.10 Re-additions in a 1,2- or 1,4-fashion could lead to enolate 7 and subsequent metal or base promoted isomerization yields more stable enone 8 as the terminal product. In view of the facile reversibility and the small differences in activation barriers of such additions/eliminations of rhodium hydrides to 6, we sought to explore conditions to address the selective formation of specific product branches.
Scheme 2 Proposed pathway for the formation of enone 8. |
Refluxing a solution of 4a in the presence of 2.5 mol% [Rh(cod)(OH)]2, 6 mol% (R)-Binap and caesium carbonate11 in toluene gave ring opened enone 8a in 79% yield, albeit in a moderate enantiomeric ratio of 75∶25 (Table 1, entry 1). Screening of a range of chiral diphosphines (Scheme 3) led to the identification of Josiphos L6 as a promising ligand giving 8a with an er of 91∶9 (entry 9). Performing the reaction at 115 °C and replacing toluene with chlorobenzene increased the yield from initially 36% to 74% (entry 10).
Entry | Ligand L* | % Yieldb | er c |
---|---|---|---|
a Conditions: 0.05 mmol 4a, 2.5 mol% [Rh(OH)(cod)]2, 6.0 mol% L*, 1.0 equiv. Cs2CO3, toluene (0.15 M), 110 °C, 12 h. b Isolated product 8a. c By HPLC with a CSP (sign of the optical rotation). d At 120 °C. e PhCl instead of toluene. f At 115 °C. | |||
1 | (R)-Binap | 79 | 75∶25 (−) |
2 | (R)-H8-Binap (L1) | 83 | 78∶22 (−) |
3 | (R)-Segphos (L2) | 89 | 76∶24 (−) |
6d | (R)-DTBM-Segphos (L3) | 37 | 93∶7 (−) |
7 | (R)-DM-Segphos (L4) | 83 | 82∶18 (−) |
8 | (R)-Difluorphos (L5) | 83 | 79∶21 (−) |
9 | L6 | 36 | 91∶9 (+) |
10e,f | L6 | 74 | 91∶9 (+) |
11e,f | ent-L7 | 62 | 61∶39 (−) |
12e,f | ent-L8 | 71 | 90∶10 (−) |
13e,f | ent-L9 | 65 | 74∶26 (−) |
Scheme 3 Utilized ligands L* (DTBM = 3,5-tBu-4-MeO-C6H2). |
With the optimized conditions, we then explored the influence of substituent R on the tertiary alcohol (Table 2). Aromatic groups with different steric and electronic properties are well tolerated and have little impact on the reactivity and selectivity of the process (entries 1–5). Although substituent R is oriented away from the reaction site, small alkyl groups cause diminished enantioselectivities (entries 6 and 7).
Entry | 4 | R | 8 | % Yieldb | er c |
---|---|---|---|---|---|
a Conditions: 0.05 mmol 4, 2.5 mol% [Rh(OH)(cod)]2, 6.0 mol% L6, 1.0 equiv. Cs2CO3, PhCl (0.15 M), 115 °C, 12 h. b Isolated product 8. c By HPLC with a CSP (sign of the optical rotation). d ent-L6 was used. | |||||
1 | 4a | Ph | 8a | 74 | 91∶9 (+) |
2d | 4b | 4-MeO–Ph | ent-8b | 65 | 91∶9 (−) |
3 | 4c | 4-F–Ph | 8c | 84 | 91∶9 (+) |
4 | 4d | 2-Me–Ph | 8d | 80 | 88∶12 (+) |
5 | 4e | 1-Naphthyl | 8e | 88 | 90∶10 (+) |
6 | 4f | Me | 8f | 85 | 80∶20 (+) |
7 | 4g | Bu | 8g | 86 | 85∶15 (+) |
Noteworthy, when the base was omitted in the retro-allylation reaction of 4a, regio-isomeric β-hydride elimination of 5 became the dominant pathway and led to presumed diene 9 (Scheme 4). Subsequent 1,4-addition might form allyl-rhodium species 10, which again undergoes H elimination accounting for the observed diene 11, formed in 77% yield and an er of 78∶22. On the other hand, when vinyl substituted tert-alcohol 4h was submitted to the reaction conditions, formation of diene 13 as well as aromatic product 14 was observed (Scheme 5). This suggests a preferential re-addition of the rhodium hydride species to the terminal, activated olefin instead of the cyclohexadiene moiety of 6h.12 Enolate 12 might subsequently isomerize to the more stable enone 13 or aromatize to arene 14. Fine-tuning of the reaction condition by the addition of ten equivalents of cyclohexene mitigated the undesired oxidation to arene 14.
Scheme 4 Proposed reaction pathways for the formation of diene 11. |
Scheme 5 Pathways for the formation of diene 13. |
Modifications of the bicyclic framework 1, for example 15 lacking the tetrahydrofuran reveal that subtle differences of the substrate have a significant impact on the reactivity profile of the allyl rhodium species (Scheme 6). In addition, slight modifications of the reaction conditions impact the fate of the putative organometallic species 16 as well.
Scheme 6 Proposed reaction pathways for the formation of 18 or 20. |
For example, performing the retro-allylation of 15 with [Rh(OH)(cod)]2, L6 and caesium carbonate (conditions A), presumably forms regio-isomeric diene 17. The concomitantly formed rhodium hydride species then selectively reduces the carbonyl group of 17 yielding 18 over the previously dominant 1,4-reduction of the diene moiety. In contrast, [Rh(OAc)(C2H4)2]2, L6 and 4 Å molecular sieves instead of caesium carbonate (conditions B) selectively provide diene 20 with an exo-methylene group by the following mechanistic scenario: a 1,4-hydrorhodation of 17 leads to alkyl rhodium species 19, which in turn undergoes β-alkoxide elimination to yield 20.13 A range of secondary alcohols 18 and methylene cyclohexenes 20 can be accessed from norbornenols 15 using these complementary conditions (Table 3). The observed enantioselectivities for the formation of 20 (entries 2, 4, 6, 8) are excellent and generally significantly higher than those for the formation of 18 (entries 1, 3, 5, 7), demonstrating a pronounced effect of the added base on the selectivity.11 The relative configuration of the secondary alcohol function of 18 was assigned by X-ray crystal structure analysis of the Diels–Alder adduct 21 obtained from 18a and 4-phenyl-3H-1,2,4-triazoline-3,5-dione (PTAD) (Scheme 7).§14 This selectivity suggests a facial reduction of the carbonyl group as depicted for 17 (Scheme 5).
Entry | 15 | Condition | R | 18/20 | % Yieldb | er c |
---|---|---|---|---|---|---|
a Conditions A: 0.05 mmol 15, 5 mol% [Rh(OH)(cod)]2, 12.0 mol% ent-L6, 1.0 equiv. Cs2CO3, PhCl (0.15 M), 120 °C, 12 h; Conditions B: 0.05 mmol 15, 5 mol% [Rh(OAc)(C2H4)2]2, 12.0 mol% ent-L6, 20 mg 4 Å MS, PhCl (0.15 M), 120 °C, 12 h. b Isolated product 18 or 20. c Determined by HPLC with a CSP (sign of the optical rotation). d With L6. | ||||||
1 | 15a | A | Ph | 18a | 76 | 78∶22 (+) |
2 | 15a | B | Ph | 20a | 68 | 95∶5 (−) |
3d | 15b | A | 4-MeO–C6H4 | ent-18b | 75 | 88∶12 (−) |
4 | 15b | B | 4-MeO–C6H4 | 20b | 82 | 98∶2 (−) |
5 | 15c | A | 4-Cl–Ph | 18c | 72 | 89∶11 (+) |
6 | 15c | B | 4-Cl–Ph | 20c | 71 | 99∶1 (−) |
7d | 15d | A | 3,5-Me–C6H3 | ent-18d | 75 | 88∶12 (−) |
8 | 15d | B | 3,5-Me–C6H3 | 20d | 81 | 98∶2 (−) |
Scheme 7 Determination of the relative configuration of 18 as its Diels–Alder adduct 21. |
In summary we showed that bicyclic meso-tert-norbornenols can be desymmetrized by retro-allylation mechanisms with chiral rhodium(I) catalysts. Subtle differences of the substrate structure and more importantly of the reaction conditions lead to diverging reaction pathways and hold the promise of a further rich downstream chemistry. Ongoing research is directed towards a deeper understanding and control of the individual steps as well as the development of synthetic applications.
N. C. thanks the Fonds der Chemischen Industrie for a Liebig-Fellowship. M. W. thanks the Roche Research Foundation and the Alexander von Humboldt Foundation for fellowships. We are grateful to Solvias for MeOBiphep, Takasago for Segphos and H8-Binap and Umicore for rhodium salts.
Footnotes |
† This article is part of the ‘Emerging Investigators’ themed issue for ChemComm. |
‡ Electronic supplementary information (ESI) available: Experimental procedures and characterization data. CCDC 779921. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c0cc01950j |
§ Crystallographic data for 21: C25H27N3O5, M = 449.51, triclinic, space groupP, a = 7.9939 (5) Å, b = 11.9064 (9) Å, c = 13.8485 (12) Å, α = 68.816 (3)°, β = 77.480 (4)°, γ = 72.174 (3)°, V = 1161.6 (2) Å3, Z = 2, Dcalc = 1.285 Mg m−3, T = 100 K, reflections collected: 6601, independent reflections: 4009 (Rint = 0.062), R(all) = 0.0818, wR(gt) = 0.1508. CCDC 779921. |
This journal is © The Royal Society of Chemistry 2011 |