[A0078]
A divergent, versatile synthesis of d-aminoalkenes and related compounds from carbohydrates; selective formation of polyalkoxypiperidines and polyalkoxypyrans.
Robert T. C. Brownlee2, Jason Dang2, Helmut M. Hügel1, Andrew B. Hughes2,* and Nuriman2
1Department of Applied Chemistry, RMIT, Plenty Rd, Bundoora,
Victoria 3083, Australia
2Department of Chemistry, La Trobe University, Bundoora,
Victoria 3083, Australia
E-mail: [email protected]
Received: 9 August 2000 / Uploaded: 14 August
A versatile synthesis of xylose derived d-aminoalkenes is presented. We have previously reported on the stereoselective synthesis of polyhydroxypiperidines.1 An extension to the methodology to incorporate more exotic N-substituted compounds was sought. This communication reports, in particular, the result of the completion of a synthetic sequence for the preparation of carbohydrate derived d-aminoalkenes.
Scheme 1 summarises the general d-aminoalkene preparation employed previously.1,2 In Scheme 1 the final reductive cleavage/reductive amination employs 16 equivalents of the amine.3 This excess limits the amines that might be installed in the resultant d-aminoalkenes to those that are cheap and readily available. A second feature of this reductive amination is the use of zinc dust to effect the reductive cleavage. This tends to limit the double bond formed to a terminal alkene, which may prevent some lines of synthetic elaboration.
The Scheme 1 approach starts with hexose sugars and so all the carbons of the final aminoalkene skeleton are present. In our approach, reported herein, the alkene was formed separately using Wittig chemistry and so the starting material required was a pentose sugar. In the initial scheme (Scheme 2) the aim was to prepare the d-aminoalkene 11 prepared previously for the purpose of comparison with Scheme 1.
D-Xylose has the requisite configurations of the secondary alcohols for construction of the aminoalkene 1a. Thus commercial methyl b-D-xylopyranoside 2 was perbenzylated (74%) under standard conditions and then the glycoside 34 was hydrolysed with a mixture of sulfuric acid and aqueous acetic acid to give the hemiacetal 4 (84%). The aldehyde was hen trapped as the dithiopropyl acetal 5 (86%). Elaboration to the double bond could then proceed. The 1˚ alcohol 5 was oxidised with pyridinium chlorochromate (PCC) in dichloromethane to give the aldehyde 6 (79%). This reaction was also attempted using standard Swern conditions but the yield was very low. The methylene ylid derived from treatment of methyl triphenylphosphonium iodide with n-butyl lithium in THF underwent a Wittig reaction with the aldehyde 6 to give the expected alkene 7 (80%). The dithioacetal was removed using mercuric chloride/mercuric oxide in water/acetonitrile to give the aldehyde 8 (92%).
At this point it was attractive to complete the sequence by reductive amination to afford the d-aminoalkene 1a. There are exemplary reductive aminations in the literature.5 However, despite our best efforts, reductive amination of the aldehyde 8, particularly with simple 1˚-alkyl amines, gave poor yields of aminated material. There is evidence that reductive amination reactions do not proceed well in the case of the simple alkyl amines6 though our continuing researches in this area will be reported in due course.
The aldehyde 8 was, however, reduced efficiently with sodium borohydride to give the alcohol 9 (95%) (Scheme 3). This alcohol was activated toward displacement by formation of the corresponding methanesulfonate 10 (75%). Heating the sulfonate ester in toluene with 5 equivalents of various amines caused displacement reactions to give the required d-aminoalkenes 1a and 11-16.
Spectroscopic data for the aminoalkene 1a were in complete agreement with the literature.1 The data for aminoalkenes 11-16 were in accord with expectations. The reason for making these compounds is to study their mercuriocyclisation reactions. Another possibility for their cyclisation is an intramolecular palladium mediated amination. However, amino groups bind tightly to palladium and can act as catalyst poisons requiring the reactions to be conducted stoichiometrically. This propensity needs to be reduced for the development of an efficient reaction catalytic in palladium. N-Acylated or sulfonylated amines or amides are one solution to this problem.7
Thus the aldehyde 8 was oxidised with silver nitrate to give the acid 17 (60%) (Scheme 4). A dicyclohexyl carbodiimide (DCC) mediated coupling of butyl amine then gave the amide 18 (80%) which was the first intermediate desired for proposed palladiocyclisations.
Finally, Banwell et al. recently reported stereoselective intramolecular Michael cyclisations of d-hydroxy and d-amino E- and Z-a, b-unsaturated esters.8 This report inspired us to consider similar transformations, which applied to sugar derived intermediates would give access to polyhydroxypyrans and polyhydroxypiperidines. Accordingly, the aldehyde 6 undergoes a Horner-Wadsworth-Emmons (HWE) reaction with triethylamine in the presence of lithium bromide to give a 78:22 mixture (75% combined yield) of the enoates 19 and 20 (19 d 7.06, 1H, dd, J5,6 16.0 and J5,4 5.5 Hz, H5, 6.09, 1H, d, J6,5 16.0 Hz, H6; 20 d 6.36, 1H, dd, J5,6 11.8 and J5,4 8.6 Hz, H5, 5.86, 1H, d, J6,5 11.8 Hz, H6) (Scheme 5). The enoates 19 and 20 were separated by flash column chromatography. In complementary fashion the HWE reaction employing fresh n-butyl lithium as base gave reversed selectivity for the mixture of enoates 19 and 20 (33:67) (80% combined yield).
The dithioacetal 19 was treated with mercuric chloride and mercuric oxide in water/acetonitrile to give the aldehyde 21 (90%) (Scheme 5). Reductive amination of the aldehyde 21 with butyl amine produces a secondary amine, which undergoes a Michael reaction in situ with the a,b-unsaturated ester to give a mixture of the piperidines 23 and 24 (75% combined yield). Similarly, the dithioacetal 20 was converted to the aldehyde 22 (92%) and then subjected to reductive amination with butyl amine to give the same piperidines 23 and 24 (58% combined yield).
Reductive amination/cyclisation with ammonium acetate was also trialed. The aldehyde 21 gave a mixture of the piperidines 25 and 26 (58% combined yield). While the aldehyde 22 gave a mixture of the piperidines 25 and 26 (52% combined yield). The data (NMR in particular) on the compounds 23-26 are complex. The results of analysis of the NMR spectra to determine the configuration of the new asymmetric centre in each compound will be reported in due course. The production of the piperidines 25 and 26 and their N-butyl derivatives 23 and 24 represents a successful approach to the synthesis of homo-azasugars 9 akin to such biologically active polyhydroxypiperidines as deoxynojirimycin.10 Thus they are of interest for their potential SAR as well as their elaboration to other natural products.
Figure 1: COSY spectra of compounds 27 and 28 acquired at 400 MHz.
Omission of the amine from the reduction of compounds 21 and 22 facilitates the Michael cyclisation to form the pyrans 27 and 28. In the case of compound 21 a 5:1 mixture of pyrans 27 and 28 was formed in 82% yield. Compound 22 gave exclusively pyran 28 (88%) under the same reaction conditions.
The assignments of the 1H NMR spectra and determination of the configuration of the C-5 of compounds 2711 and 28 12 were achieved through the analysis of COSY spectra (Figure 1). It is clear that the chemical shifts of the ring protons are considerably different in the two isomers. The assignment of structures 27 and 28 was based on the expected coupling patterns of the H5 protons. For compound 27, H5 is expected to have a di-axial coupling to H4 (9.0 Hz), whereas in pyran 28 axial-equatorial coupling (3.6 Hz) is present.
Our results in the formation of the pyrans 27 and 28 from the enoates 21 and 22 appear to be consistent with an extended chair-like conformation of the substrate during cyclisation possibly under chelation control. Thus we have successfully met the aim of designing a divergent synthesis of a range of N-alkylated d-aminoalkenes. And demonstrated the versatility of the sequence with the preparation of a d-amidoalkene for proposed palladiocyclisations. Finally, the formation of the a,b-unsaturated esters 21 and 22 allowed the generation of piperidines 23-26 and the novel pyrans 27 and 28. Further variations of this chemistry that will give access to stereoisomers such as the corresponding ribose derived species and compounds such as isomeric d-amidoalkenes and more substituted alkenes are being elaborated and will be reported at a later date.
Acknowledgement
We thank the University of Jember, Indonesia for the award of a DUE project scholarship (Nuriman).
References
1. H. M. Hügel, A. B. Hughes and K. Khalil, Aust. J. Chem., 1998, 51, 1149.
2. H. M. Hügel, A. B. Hughes and K. Khalil, Unpublished Data.
3. R. C. Bernotas and B. Ganem, Tetrahedron Lett., 1985, 26, 1123.
4. All new compounds exhibited satisfactory spectroscopic and exact mass or microanalytical data. Yields refer to spectroscopically and chromatographically homogeneous materials.
5. See reference 6, p 898-899; F. Manescalchi, A. R. Nardi and D. Savoia, Tetrahedron Lett., 1994, 35, 2775; S. A. Godleski and E. B. Villhauer, J. Org. Chem., 1986, 51, 486; T. Baasov, and M. Sheves, J. Am. Chem. Soc., 1985, 107, 7524; J. H. Billman and A. C. Diesing, J. Org. Chem., 1957, 22, 1068.
6. J. March, In �Advanced Organic Chemistry�; Reactions, Mechanisms and Structure�, 4th Ed., 1992, Wiley Interscience, New York.
7. L. S. Hegedus, In �Comprehensive Organic Synthesis�, vol. 4 p. 559-563, B. M. Trost and I. Fleming (eds), 1991, Pergamon Press Ltd, Oxford.
8. M. G. Banwell, B. D. Bissett, C. T. Bui, H. T. T. Pham and G. W. Simpson, Aust. J. Chem., 1998, 51, 9.
9. A. Kilonda, F. Compernolle, S. Toppet and G. J. Hoornaert, Tetrahedron Lett., 1994, 35, 9047; C. Herdeis and T. Schiffer, Tetrahedron, 1996, 52, 14745; F. Compernolle, G. Joly, K. Peeters, S. Toppet and G. Hoonaert, Tetrahedron, 1997, 53, 12739.
10. A. B. Hughes and A. J. Rudge, Nat. Prod. Rep., 1994, 11, 135.
11. Compound 27 1H NMR (400 MHz, CDCl3) d 7.34-7.11, 15H, m, ArH; 4.98, 1H, d, JAB 10.9 Hz, OCHHPh; 4.92, 1H, d, JAB 11.1 Hz, OCHHPh; 4.83, 1H, d, JAB 10.9 Hz, OCHHPh; 4.70, 1H, d, JAB 11.6 Hz, OCHHPh; 4.61, 1H, d, JAB 11.6 Hz, OCHHPh; 4.60, 1H, d, JAB 11.1 Hz, OCHHPh; 4.10, 2H, q, J 7.1 Hz, OCH2CH3; 3.95, 1H, dd, J1,2 5.1, J1,1� 11.5 Hz, H1; 3.68-3.56, 3H, m, H2, H4, H5; 3.27, 1H, apparent t, J 9.0 Hz, H3; 3.20, 1H, apparent t, J 10.6 Hz, H1�; 2.73, 1H, dd, J6,5 3.4, J6,6� 15.2 Hz, H6; 2.35, 1H, dd, J6�,5 8.7, J6�,6 15.2 Hz, H6�; 1.21, 3H, t, J 7.1 Hz, CH2CH3.
Compound 27 1H NMR (400 MHz, C6D6) d 7.40-7.04, 15H, m, ArH; 5.02, 1H, d, JAB 11.3 Hz, OCHHPh; 4.97, 1H, d, JAB 11.4 Hz, OCHHPh; 4.83, 1H, d, JAB 11.3 Hz, OCHHPh; 4.57, 1H, d, JAB 11.4 Hz, OCHHPh; 4.44, 1H, JAB 11.9 Hz, OCHHPh; 4.36, 1H, d, JAB 11.9 Hz, OCHHPh; 4.01, 2H, q, J 7.0 Hz, CH2CH3; 3.96-3.88, 2H, m, H1, H5; 3.65, apparent t, J 8.7 Hz, H3; 3.61-3.54, 1H, m, H2; 3.34, 1H, apparent t, J 9.0 Hz, H4; 3.17, 1H, apparent t, J 10.7 Hz, H1�; 2.86, 1H, dd, J6,5 2.9, J6,6� 15.2 Hz, H6; 2.53, 1H, dd, J6�,5 8.4, J6�,6 15.1 Hz, H6�; 0.99, 3H, t, J 6.9 Hz, CH2CH3.
12. Compound 28 1H NMR (400 MHz, CDCl3) d 7.24, 15H, m, ArH; 4.60-4.43, 6H, m, OCH2Ph � 3; 4.25, 1H, ddd, J5,6 8.5, J5,6� 5.5, J5,4 3.6 Hz, H5; 4.08-4.00, 2H, m, OCH2CH3; 3.73-3.70, 2H, m, H1, 1�; 3.63, 1H, apparent t, J 5.3 Hz, H3; 3.44, 1H, dd, J4,5 3.6, J4,3 5.4 Hz, H4; 3.37, 1H, dd, J2,3 4.4, J2,1 9.1 Hz, H2; 2.70, 1H, dd, J6,5 8.4, J6,6� 15.9 Hz, H6; 2.56, 1H, dd, J6�,5 5.5, J6�,6 15.9 Hz, H6�; 1.66, 3H, t, J 7.2 Hz, CH2CH3.
All comments on this poster should be sent by e-mail to (mailto:[email protected] ona.edu)
[email protected] with A0078 as the message subject of your e-mail.