A novel asymmetric route to 2-amino-1,2,3,4-tetrahydronaphthalenes

Article abstract of DOI:10.1016/S0040-4039(00)00327-0

A novel asymmetric route to 2-amino-1,2,3,4-tetrahydronaphthalenes has been demonstrated starting from phthalimidovinylglycinol (PVG). Functionalisation of PVG via Heck reaction, olefin hydrogenation and cyclisation provides the title products. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4039(00)00327-0

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 41 (2000) 3187–3191
A novel asymmetric route to
2-amino-1,2,3,4-tetrahydronaphthalenes
Michael C. J. Harris, Mark Jackson, Ian C. Lennon,∗ James A. Ramsden and Helen Samuel
Chirotech Technology Ltd, Cambridge Science Park, Milton Road, Cambridge CB4 0WG, UK
Received 2 February 2000; accepted 21 February 2000
Abstract
A novel asymmetric route to 2-amino-1,2,3,4-tetrahydronaphthalenes has been demonstrated starting from
phthalimidovinylglycinol (PVG). Functionalisation of PVG via Heck reaction, olefin hydrogenation and cyclisation
provides the title products. © 2000 Elsevier Science Ltd. All rights reserved.
The pharmacological activity of 2-amino-1,2,3,4-tetrahydronaphthalenes (2-aminotetralins) is well
known.1 The 2-aminotetralin structure is most commonly accessed from the corresponding 2-tetralone
either via reductive amination2 or oxime reduction.3 Two asymmetric routes have been reported, namely
the hydrogenation of an enamide derived from a 2-tetralone,4 and the cyclisation of homo-phenylalanine
derivatives.5 Here we report an alternative asymmetric route to 2-aminotetralins using readily available
and inexpensive starting materials.
The key starting material is phthalimidovinylglycinol (PVG) 1, readily available from racemic 3,4-
epoxy-1-butene by a palladium(0)-catalysed allylic substitution as described by Trost (Scheme 1).6 All of
the substrate is converted into optically enriched product, therefore, a mechanism for the equilibration of
the diastereoisomeric π-allyl intermediates is required. Two mechanisms for this equilibration have been
suggested, Pd–Pd substitution and an η3–η1–η3 process,7 however, we have no evidence as to which
is operational in this reaction. High enantiomeric excess product is obtained at high catalyst loading
(5 mol%),6 but both the enantiomeric excess and the rate of reaction drops as the catalyst loading is
decreased to 0.1 mol% (Table 1, entry 1).8 Recent literature reports have demonstrated that halides
can be added to increase regio-9 and enantioselectivity9,10 in palladium catalysed allylic substitution
reactions. We chose to study the effect of using a mild, soluble fluoride additive, tetrabutylammonium
(triphenylsilyl)difluorosilicate (TBAT)9,11 in the synthesis of PVG 1 (Table 1).
In order to achieve high conversion in dichloromethane with 0.1 mol% of catalyst prolonged heating at
reflux was required (Table 1, entry 1). Upon the addition of TBAT, both the rate and the enantioselectivity
of the reaction were increased (entry 2). Changing to the more environmentally responsible solvent,
∗
Corresponding author.
0040-4039/00/$ - see front matter © 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(00)00327-0
tetl 16607

3188
Scheme 1. Synthesis of PVG using Trost palladium catalyst
Table 1
Effect of additives in PVG synthesis
toluene, clearly demonstrated the positive effect of the additive. The standard conditions in toluene gave
poor selectivity (entry 3), whereas addition of TBAT increased the enantiomeric excess by more than
10% (entry 4).
After further process optimisation we found that catalytic lithium carbonate was superior to sodium
carbonate when toluene was used as the solvent. A dramatic increase in enantioselectivity was observed
(entry 5) and the amount of regioisomer 2 was significantly reduced. These conditions have been used
to prepare kilogram quantities of PVG 1 and are reproducible giving high yields and ee. Replacement of
TBAT with tetrabutylammonium chloride (TBAC) provides a cheaper alternative with a similar reaction
profile (entry 6).
Our synthetic route to aminotetralins from PVG 1 is shown in Scheme 2. The Heck reaction was found
to proceed best when catalysed by palladium acetate and tri(o-tolyl)phosphine.12 Significant amounts
of phenyl addition product13 were observed when triphenylphosphine was employed and palladium
precipitated prior to completion of the reaction. Triethylamine was found to be the optimal base.
Upon heating PVG 1 in the presence of a carbonate base significant ring opening of the phthalimide
was observed. A range of aromatic halides were used in the Heck coupling, giving access to several
unsaturated N-protected amino alcohols 4a–f (Table 2).14
Hydrogenation of the double bond proceeded smoothly over palladium on carbon provided
that the Heck product had been purified by crystallisation, otherwise the more robust 1,10-
bis(diisopropylphosphino)ferrocenyl rhodium catalyst15 was used. Cyclisation of the electron rich
substrates proceeded readily upon heating in the presence of trifluoromethanesulfonic acid.16
Unexpectedly, cyclisation of the p-methoxy compound 5a did not give the 7-methoxy-2-aminotetralin.

3189
Scheme 2. (i) 1 mol% Pd(OAc)2, 3 mol% P(o-tol)3, Et3N, DMF, 100–130°C; (ii) H2, Pd/C or [DiPFc]Rh[COD]BF4; (iii)
CF3SO3H, PhCl, 80°C; (iv) H2NNH2, EtOH
Table 2
Synthesis of aminotetralins from PVG

3190
Instead 6-methoxy-2-aminotetralin 8 was obtained, presumably via ipso attack and rearrangement
(Scheme 3).
Scheme 3. Cyclisation of para methoxy compound
The unexpected product was confirmed by preparation of the authentic 6- and 7-methoxy-2-
aminotetralins from the corresponding 2-tetralones.3 This rearrangement appears to be limited to the
cyclisation of the alcohols, as the corresponding acid17 and amino acid18 gave the expected 7-methoxy
product. Cyclisation of the m-methoxy compound 5d afforded a 3:1 mixture of the 6- and 8-methoxy-2-
aminotetralins resulting from cyclisation at both the ortho and para positions. No cyclised product was
observed with electron deficient aromatics 5b and 5c. The parent compound 7e (R1–R4=H) was readily
obtained from iodobenzene.
The aminotetralins were obtained in high enantiomeric excess (>97%)19 despite the PVG being
only 88–93% ee. Presumably recystallisation of the Heck product 4 and the phthalimide protected
aminotetralin 6, enhanced the enantiomeric excess of these intermediates, and there was no degradation
of optical purity in the cyclodehydration step.
To conclude we have demonstrated a short, novel and economic asymmetric synthesis of 2-
aminotetralins from PVG, which can be synthesised using the optimised conditions outlined in this
communication. In addition both enantiomers of vinylglycinol are readily available, >98% ee,20 from
PVG by hydrazinolysis, formation of the benzoic acid salt and crystallisation. Studies in the use of these
versatile building blocks are currently ongoing.
Acknowledgements
This work was carried out as part of a collaboration with Eastman Fine Chemicals. We would like to
thank Professor B. M. Trost for useful discussions. We are grateful to Catherine Rippé and Will Spearing
for the development of analytical methods.
References
1. van Vliet, L. A.; Tepper, P. G.; Dijkstra, D.; Damsma, G.; Wikstrom, H.; Pugsley, T. A.; Akunne, H. C.; Heffner, T. G.;
Glase, V.; Wise, L. D. J. Med. Chem. 1996, 39, 4233 and references cited therein.
2. Copinga, S.; Tepper, P. G.; Grol, C. J.; Horn, A. S.; Dubocovich, M. L. J. Med. Chem. 1993, 36, 2891.
3. Ye, Q.; Grunewald, G. L. J. Med. Chem. 1989, 32, 478.
4. Tschaen, D. M.; Abramson, L.; Cai, D.; Desmond, R.; Dolling, U.-H.; Frey, L.; Karady, S.; Shi, Y.-J.; Verhoeven, T. T. J.
Org. Chem. 1995, 60, 4324.
5. (a) Baxter, A. D.; Murray, P. J.; Taylor, R. J. K. Tetrahedron Lett. 1992, 33, 2331. (b) Nordlander, J. E.; Payne, M. J.;
Njoroge, F. G.; Vishwanath, V. M.; Han, G. R.; Laikos, G. D.; Balk, M. A. J. Org. Chem. 1985, 50, 3619.
6. Trost, B. M.; Bunt, R. C. Angew. Chem., Int. Ed. Engl. 1996, 35, 99.
7. Trost, B. M.; Lemoine, R. C. Tetrahedron Lett. 1996, 37, 9161.
8. Determined by HPLC: Chiracel OD, UV 254 nm, 1.0 ml/min, 10% IPA, 90% heptane. RT (R)=12.8 min, RT (S)=15.9 min.
9. Trost, B. M.; Toste, F. D. J. Am. Chem. Soc. 1999, 121, 4545.
10. Burckhardt, U.; Baumann, M.; Togni, A. Tetrahedron: Asymmetry 1997, 8, 155.

3191
11. Pilcher, A. S.; Ammon, H. L.; DeShong, P. J. Am. Chem. Soc. 1995, 117, 5166.
12. For a related Heck coupling with vinylglycinol derivatives, see: Crisp, G. T.; Glink, P. T. Tetrahedron 1992, 48, 3541.
13. Morita, D. K.; Stille, J. K.; Norton, J. R. J. Am. Chem. Soc. 1995, 117, 8576.
14. All new compounds gave satisfactory spectral and analytical data.
15. Burk, M. J.; Harper, T. G. P.; Lee, J. R.; Kalberg, C. Tetrahedron Lett. 1994, 35, 4963.
16. A cyclodehydration has been reported for 1-alkyl substituted tetralins: Kropp, P. J.; Breton, G. W.; Craig, S. L.; Crawford,
S. D.; Durland Jr., W. F.; Jones III, J. E.; Raleigh, J. S. J. Org. Chem. 1995, 60, 4146.
17. Silverman, I. R.; Daub, G. H.; Vander Jagt, D. L. J. Org. Chem. 1985, 50, 5550.
18. Moretti, G. P.; Foresta, P. WO 9833762.
19. GC method, Chiralsil Dex CB column: (R)-6-methoxy-2-aminotetralin, ee>98%, compound derivatised with trifluoroacetic
anhydride, 155°C for 25 mins, ramp to 200°C at 10°C/min, hold for 5 mins, RT (R)=24.0 min, RT (S)=23.5 min; (R)-2-
aminotetralin, ee 97%, compound derivatised with propionic anhydride, 165°C for 25 mins, ramp to 200°C at 10°C/min,
hold for 5 mins, RT (R)=24.6 min, RT (S)=23.8 min.
20. GC method, Chiraldex BTA, vinylglycinol was derivatised with acetic anhydride, 150°C for 20 mins, RT (S)=9.9 min, RT
(R)=10.2 min.