Design of Chiral Auxiliaries for the Allene Ether Nazarov Cyclization

Link: JACS ASAP

April R. Banaag and Marcus A. Tius*

Department of Chemistry, University of Hawaii, 2545 The Mall, Honolulu, Hawaii 96822, and Cancer Research Center of Hawaii, 1236 Lauhala Street, Honolulu, Hawaii 96813

This is an interesting article describing a cool experimental result and identifying a key factor influencing enantioselectivity of a Nazarov cyclization of lithium allene species. This led to a basis for designing a chiral auxiliary on the allene moiety based on stereochemistry of the sugar molecules.

The premise of the studies was based on the following Nazarov cyclization to form alpha-methylene cyclopentenone (eq 1).

Following deprotonation with alkyllithium, allenyl lithium 1 could add to enamide 2. After a mild acid-induced Nazarov cyclization, cyclopentenone 4 could be obtained. If a chiral sugar molecule (a pyranose) was attached as an R1 group in 1, this could lead intermediate 3 to a selective conrotation to give 4 enantioselectively, and pyrilium cation as a by-product.

Keith A. Woerpel has demonstrated that the C-3 and C-4 substituents exert a large influence on the conformation of the tetrahydropyran oxocarbenium ions relative to their uncharged precursors (a) Ayala, L.; Lucero, C. G.; Romero, J. A. C.; Tabacco, S. A.; Woerpel, K. A. J. Am. Chem. Soc. 2003, 125, 15521-15528. b) Shenoy, S. R.; Woerpel, K. A. Org. Lett. 2005, 7, 1157-1160.) Specifically, C-3 and C-4 alkoxy groups have pseudoaxial preference in the oxocarbenium ion.

In testing this principle, lithium species 5 was treated with enamide 6 to give intermediate 7, which upon acid-workup gave R-(9) in high yield (84%) and er (93/7).

As seen from Scheme 1, the presence of C-4 axial group in transition state 8 blocked the back face of developing cyclopentenone and induced counterclockwise conratation to give 9 in good er. The criteria for the model to be valid are the 3,4,5-triaxial conformation of the pyran ring in 7 must be energetically accessible and one must assume a late transition state for the cyclization.

The model was further tested based on the ground of three hypotheses/predictions: 1) C-3 OTBS group is not important in the outcome, 2) C-4 OTBS is needed to effect the selectivity, and 3) locking C-4 OTBS group in equatorial position in the pyran ring will erode the selectivity.

As seen below, when 11 (missing C-3 OTBS) was reacted with enamide 6, R-(9) was obtained in good yield and unaffected er. In this case, 11 was also used to prove that oxygen atom in the pyran ring was needed to transfer the stereochemical information from the auxiliary to the product as its cyclohexyl derivative (without oxygen in the ring) delivered R-(9) in low yield and low er (55.5/44.5).

When 12 (missing C-4 OTBS) was treated with 6, lower yield and er of 9 were obtained.

When C-4 alkoxy was locked in equatorial position as in 13, 9 was obtained in lower er.

Because 1,3-diaxial interaction was required, therefore even though the conformation of C-4 was locked in the equatorial position in 14, by placing C-3 axial OTBS group, 9 could still be delivered in high er.

The stereochemistry of product could be inverted by inverting the sterechemistry of the anomeric C-1 allene from alpha- to beta-position. This allowed both enantiomers of cyclopentenone to be made from the D-sugar. The C-3 and C-5 substituents in this series were cis in relationship and it was expected that when the pyran ring was inverted in the transition state, these two groups would become cis diaxial to influence stereochemical outcome of the product.

But when both 15 and 16 were treated with enamide 6, S-(9) was obtained with the results as shown below. It was concluded that in this series, the pyran ring was not inverted in the transition state or the stereochemical outcomes would have been more similar. The reason for this difference in conformational preference between the alpha- and the beta-series was not known but it was suspected that the 1,3-diaxial interaction in the alpha-series may have affected the ring inversion in the transition state. This destabilizing 1,3-diaxial interaction was not present in the beta-series.

The stereochemical outcome of this series could be predicted using transition state 17 in Scheme 2. Therefore, the electron pair donation from oxygen in the pyran ring restricted the conformation of the transition state, bringing the cyclization intermediate closer to the C-4 OTBS group. This led the conrotation of 17 to occur in the counterclockwise fashion leading to S-(9).

Enamide 6 was by no means the best substrate for 5, or 14 and 16. Compound 5 was screened with 6 and other eight 2,3-disub morpholine enamides with good to excellent ers. More screening results will be reported in the future. In this work, one key factor, among many others, had been identified as contributing to the stereoselectivity of Nazarov cyclization. This would lay groud work for future studies in designing effective chiral auxiliary.