Comunicación de Resultados

Table 3.1 shows the results of the cyclisations of the 3-Me substituted precursors. A combined yield for both diastereoisomers is reported because they could not be fully separated. The ratio of cyclised to reduced products was determined from the ‘H NMR spectra of the crude product (acetonitrile extracts) after partitioning to remove the excess tin residues. 'H NMR spectra of the corresponding cyclohexane and hexane partitions were checked to ensure that no cyclised or reduced products had been taken up in these solvents. In all cases the hydrocarbon solvents were found to have removed only tin products from the mixture.

Chapter 3: Stereochemistry o f 5-exo cyclisations of amidyl radicals

Compound number R2 Yield of 73/% Ratio of 73:88 d.e./%

73a h-Bu 55 a 1 0

73b CH(CH3) 3 42 7.4:1 1 1

73c Me 47 17:1 17

73d CH2Ph 53 1 2 : 1 23

a negligible amount o f reduced

Table 3.1 - Cyclisations of N-alkyl-3-methylpent-4-enamides

The results indicate that the nature of the R group has little effect on the yield or diastereoselectivity of the cyclised products, while the ratio of cyclised/reduced products is significantly altered. The yield is seen to decrease slightly with the bulky secondary R group 73b, but this is a consequence of the poorer cyclised/reduced ratio. This could be due to the increased steric demand of the /V-isopropyl substituent, resulting in a slower rate of cyclisation. The /V-w-butyl group 73a, gives the best cyclised to reduced ratio but the worst diastereoselectivity, while the /V-benzyl group 73d, gives a high cyclised to reduced ratio and shows the greatest diastereoselectivity. It should be noted that the diastereoselectivity is much poorer than for simple hex-S-enyl radical cyclisations and is not in fact great enough to be synthetically useful. This is probably due to the fact that the transition states for

Chapter 3: Stereochemistry of 5-exo cyclisations of amidyl radicals

cyclisation of the amidyl radicals is much flatter than for simple alkyl radicals. This will lead to poorer stereoselectivity due to a lower energy difference between the alternative transition states, (See Figure 3.6, later).

Comparison of the 'H NMR spectra o f each cyclised product showed that the major isomer was the same in each of the four cases studied. These isomers were assigned to be trans on the basis of nOe data and by comparison with data from authentic samples previously reported.

The nuclear Overhauser effect (nOe) is used as an aid to determine which protons (or groups of protons) in a molecule are in close proximity to each other. If two protons (Ha and Hb) are within 3.5 Angstroms o f each other then irradiation of one can result in an enhancement o f the signal for the other proton. This is because each proton contributes to the others spin-lattice relaxation process. Double irradiation of Ha stimulates absorption and emission processes for Ha and this stimulation is transferred through space to the relaxation mechanism of Hb. The increase in intensity of the Hb signal can be from 1-50% but the observable enhancement is usually less than 20%. The effect is normally viewed by obtaining nOe difference spectra. A conventional spectrum (no irradiation) is first recorded followed by one with irradiation of a specific proton or group of protons. Subtraction of the former from the latter gives the nOe difference spectrum with only the enhanced peaks

94.95 remaining. ’

Chapter 3: Stereochemistry of 5-exo cyclisations of amidyl radicals

An nOe experiment was performed on the major isomer of jV-n-butyl-4,5- dimethylpyrrolidin-2-one, 73a, the results of which are shown in Figure 3.5. Irradiating at the 5-Me resonance for the major isomer resulted in a 2% enhancement of the 4-H resonance. Likewise, irradiating the 4-Me protons enhanced the signal for the 5-H proton by 2%. This suggests a tram stereochemistry. Irradiating at the 5-H resonance for the major isomer gave no enhancement of the 4-H signal which again is in accord with a tram stereochemistry.

Figure 3.5 - nOe Data obtained for JV-fi-butyl-4,5-dimethylpyrrolidin-2-one

7>«m-A'-benzyl-4,5-dimethylpyrrolidin-2-one (tram-73d) has been reported earlier by Takahota et al?b and a comparison of their published spectral details and the 5- values obtained for both the major and minor isomers produced in our experiment is given in Table 3.2. The data published on the tram isomer fits most closely with the major isomer in our work (particularly the position o f the methyl resonances) and so the major isomer was assigned as being tram.

Chapter 3: Stereochemistry o f 5-exo cyclisations o f amidyl radicals

Published 8-values for 8-values for major isomer 8-values for minor isomer

trans isomer (270 MHz) (400 MHz) (400 MHz) 1.03 (3H, d, J 7.0) 1.01 (3H, d, J 6.6, 4-Me) 0.97 (3H, d, J 7.0, 4-Me) 1.14 (3H, d, J 6.4) 1.12 (3H, d, J 6.3, 5-Me) 1.00 (3H, d, J 6.7, 5-Me) 1.65-2.21 (2H, m) 1.90 (1H, m, 4-H) 2.13 (1H, ddd, J 16.0, 8.6, 2.04 (1H, dd, J 16.8, 7.7, 0.9, 3-H) 3-H) 2.39 (1H, m, 4-H) 2.41-2.71 (1H, m) 2.63 (1H, ddd, J 16.8,8.4, 2.49 (1H, dd, J 16.0, 7.9, 1.1, 3-H) 3-H) 2.81-3.21 (1H, m) 3.00 (1H, q, J 6.2, 5-H) 3.47 (1H, q, J 6.7, 5-H) 3.93-4.94 (2H, AB q, J 3.95 (1H, d ,J 14.7, 3.91 (1H, d, J 14.5, 15.2) CH2Ph) CH2Ph) 4.94 (1H, d, J 14.7, 4.97 (1H, d, J 14.5, CH2Ph) CH2Ph)

Table 3.2 - A comparison of *H NMR spectral details obtained with published results (J values given are in Hertz).

The Irons stereoselectivity matches that predicted by the Beckwith model. Out of the four possible ‘chair-like’ transition states [(a-d), Figure 3.6], those with the

Chapter 3: Stereochemistry of 5-exo cyclisations of amidyl radicals

Beckwith] lead to the trans cyclised products. The minor cis products may arise from boat-like transition states as well as chair transition states with axial substituents, (Figure 3.6). The boat-like transition states (e-h), are usually assumed to be much higher in energy than the chair transition states (a-d), but the energy difference may be much less here for two reasons. Firstly, the carbonyl group flattens the transition state (c.f. with hex-5-enyl radical cyclisations) leading to a lower energy difference between alternative transition states. Secondly, an E amidyl radical geometry (the preferred conformation for secondary amides) places the N-R group pointing towards the double bond in the chair transition states (a and b), but away from it in the corresponding boat transition states (e and f)- Placing the carbon substituent in an equatorial position in the boat-like transition state (e) leads to a cis stereochemistry.

R2 R2

(a) (b) (c)

(d)

R2 R2

(e)

(0

(h)

Chapter 3: Stereochemistry of 5-exo cyclisations of amidyl radicals