With these preliminary results, we considered that route B was the most promising one, so we decided to drive the research on this direction.
C. Michael addition via iminium activation
Before studying the Michael addition, the synthesis of the pro- bis(nucleophile) 8b had to be scaled up. Only a longer reaction time (3 hours) was needed to obtain 8b in a gram scale using the conditions optimized in the Chapter II of the present Thesis (NaOH as a base in MeNO2 as solvent, Scheme 2.6, page 34).
According to our previous experience, the optimal conditions for aliphatic and aromatic enals are different from each other.93 As we have observed, the aliphatic enals are less reactive than the aromatic ones, whereas the aromatic enals offer problems of reversibility, leading easily to a decrease of the ee with time.
The results obtained for the aliphatic enals are summarized in Table 4.1. To check the viability of the reaction in enantioselectivity terms, we selected up to five proline-type catalysts (51a-f) using CH2Cl2 as solvent. We chose trans-2-hexenal 48a (R = Pr) as a model electrophile. Due to the instability of the resulting aldehyde in the HPLC columns, the enantiomeric excesses were determined on the cyclic acetal obtained from the reaction between the Michael adduct and 1,2-ethanediol (See experimental part).
We obtained the best result with catalyst 51a, which afforded the Michael adduct 52a in 69% yield and a promising 44% ee (entry 1). Unfortunately, the catalysts that had proven to be more effective for this type of transformations in the bibliography (catalysts 51d and 51e, entries 4 and 5), did not seem to be reactive enough. In the case of catalyst 51e, not even traces of the desired product were observed in the reaction crude, leaving nucleophile 8b unaffected.
Table 4.1. Catalyst screening for the Michael addition via iminium activation.
a isolated yield
Based on our previous experience, we considered that the use of TBAB as additive93 could be a good option to improve the yield of the reaction with the aliphatic enals. In fact, catalysts 51b,d,e which turned out to be sluggish in the absence of TBAB (entries 1, 4 and 5), afforded full conversions (entries 6 to 9). We obtained the highest ee (86%) with catalyst 51e (entry 7). Catalyst 51f, bearing the bulkier OTBDMS group did not afford complete conversion after 48 h (entry 9).
In all the cases we obtained adduct 52a as a mixture of diastereomers, with a 62:38 ratio in the particular case of the catalyst 51e. The rest of the catalysts afforded similar values. Nevertheless, as it was mentioned
Entry Catalyst Additive Time (h) Conv. (%) ee(%)
1 51a ---- 24 100 (69)a 44 2 51b ---- 72 23 n.d. 3 51c ---- 24 90 -41 4 51d ---- 24 68 n.d 5 51e ---- 120 0 n.d 6 51b TBAB 48 100 (68)a 80 7 51d TBAB 24 100 (68)a 79 8 51e TBAB 48 100 (69)a 86 9 51f TBAB 48 66 n.d.
above (see Scheme 4.11), the diastereoselectivity at this point lacked of importance as we expected epimerization towards the most stable isomer after cyclization.
Once we had achieved a good yield and ee, and based on our previous experience with this type of reactions93,128 we made some modifications in order to further improve the results. Firstly, we considered EtOH as a suitable solvent for the process.129 Secondly, we reduced the catalyst loading and tested an acidic and a basic additive (Table 4.2).
Table 4.2. Screening of conditions for Michael addition via iminium activation.
a 2 equiv of aldehyde were used. b isolated yield
The reduction of the amount of catalyst until 20 mol% slowed down the reaction: we did not observe completion after 48 h (entry 1). The reaction also became slower when using EtOH due to the very low solubility of nucleophile 8b in this medium. Trying to take advantage of the proven beneficial effect of EtOH,129 we tried different mixtures of CH2Cl2 and EtOH (entries 2 and 3). A CH2Cl2 : EtOH = 1:1 mixture raised both ee and
128 M. B. Cid, S. Duce, S. Morales, E. Rodrigo, J. L. García Ruano, Org. Lett. 2010, 12, 3586. 129 Sara Duce Igeño (2014): ‘Arylacetic acid derivatives as nucleophiles in iminium ion activation catalysis, Synthetic applications and mechanistic insights’ (PhD dissertation)
Universidad Autónoma de Madrid, Spain
Entry Solvent Additive Time
(h) Conv. (%) ee (%) 1 CH2Cl2 TBAB (1 equiv) 48 86 ---
2 CH2Cl2 / EtOH (1:4) TBAB (1 equiv) 96 82 ---
3 CH2Cl2 / EtOH (1:1) TBAB (1 equiv) 48 100 (95)b 90
4 CH2Cl2 / EtOH (1:1) TBAB (1 equiv) 24 83 ---
5a CH2Cl2 / EtOH (1:1) TBAB (1 equiv) 48 71 ---
6 CH2Cl2 / EtOH (1:1) Ph-CO2H (20 mol%) 48 46 ---
yield (entry 3). However, even under these conditions, the reaction needed long times to be completed, since after 24 h the reaction had not finished (entry 4). We also tried to reduce the amount of aldehyde, but the reaction was very slow (entry 5). Finally, neither the acidic nor the basic additive offered any improvement (entries 6 and 7), although in the case of LiOAc the results were very similar than in the case of TBAB (entries 3 and 7).
In order to work in the later intramolecular Julia-Kocienski olefination, we scaled the reaction up to 1 gram of pro-bis(nucleophile) 8b using conditions of Table 4.2, entry 3. We did not observe variation either of yield or enantiomeric excess in this scaled-up process.
In the case of the aromatic enals, the previously optimized conditions for the aliphatic enals were not valid. Firstly, when the Michael reaction was carried out using the best conditions obtained so far (1 equivalent of TBAB), we observed that the ee decreased with time. This is a consequence of the reversibility of the reaction, a behaviour observed when aromatic enals take part in Michael additions via iminium activation.93a, 128, 130
In this case, the use of LiOAc as additive could solve partially the problem (Scheme 4.18). With enal 48h, we left the reaction for 4 days until it was completed, achieving a good yield (89%), but a moderate ee (56%). Stopping the reaction at a shorter time (4 h) the ee raised up to 89%, although the yield was lower (55%) due to incomplete conversion.