Sheng-Han Huang,
Yu-Wei Shih,
Wen-Tse Huang,
Deng-Hong Li and
Te-Fang Yang*
Department of Applied Chemistry, National Chi Nan University, 1, University Road, Puli, 545, Nantou, Taiwan. E-mail: tfyang@ncnu.edu.tw
First published on 21st September 2016
Addition reaction of 3,4-dihydroisoquinolines with γ-hydroxy-α,β-unsaturated ketones furnished a variety of diastereomerically pure oxazolo[2,3-a]tetrahydro-isoquinolines, which could be purified by either recrystallization or solvent evaporation. The reaction proceeded smoothly under “green” conditions without an additive and catalyst, giving the target molecules in good to excellent yields.
Especially, oxazolo[2,3-a]tetrahydroisoquinolines have been important intermediates for the synthesis of isoquinoline alkaloids. By starting from isochroman, as shown in Scheme 1, Yamato et al. accomplished the diastereoselective synthesis of oxazolo[2,3-a]tetrahydroisoquinoline,12 which was obtained in three steps then converted to chiral 1-substituted tetrahydroisoquinolines. In addition, 3-phenyl oxazolo[2,3-a]tetrahydroisoquinoline was afforded by the condensation of ortho-vinylbenzaldehyde with (R)-phenylglycinol in hot 1,4-dioxane or DMSO.13 Furthermore, a similar synthetic methodology was applied to the preparation of oxazolo[2,3-a]tetrahydroisoquinolone, a tricyclic lactam.14 On the other hand, Grigg and co-workers studied the reaction of two equivalents of pyridine-2-carbaldehyde with 1,2,3,4-tetrahydroisoquinoline in warm acetonitrile, and found that dipyridyl oxazolo[2,3-a]tetrahydroisoquinoline was furnished.15 The mechanism of this reaction involved a formation of an intermediate called azomethine ylide. Recently, Houk, Seidel and co-workers described that the reaction of salicylaldehyde with tetrahydrohydroisoquiline could provide diphenolic oxazolo[2,3-a]tetrahydroisoquinoline, which was a [3 + 2] adduct.16 In addition, it is particularly noteworthy that Matsubara and co-workers carried out the asymmetric cycloaddition reaction of N-tosyl imines with γ-hydroxy-α,β-unsaturated ketones in the presence of a bifunctional catalyst derived from cinchonidine.8,17 All the interesting contents of reports mentioned above prompted us to search for a sustainable and efficient methodology for the synthesis of oxazolo[2,3-a]tetrahydroisoquinolines. Thus, we report herein the results of the addition of 3,4-dihydroisoquinolines with γ-hydroxy-α,β-unsaturated ketones (Scheme 1). To the best of our knowledge, this is the first example of the synthesis of oxazolo[2,3-a]tetrahydroisoquinoline derivatives via a reaction carried out under environmentally friendly conditions.18
For the condition screening, 3,4-dihydroisoquinoline (1a) and (E)-4-hydroxy-1-phenylbut-2-en-1-one (2a) were utilized as the reactants of the model reaction (Table 1). It is noteworthy that Jacobsen and co-workers reported that 3,4-dihydroisoquinoline could be activated with acetic acid to an iminium ion, which then reacted with a dienanine in toluene to afford benzoquinolizidine.19 At beginning of our investigation, hence, toluene was selected as the solvent for the addition of 3,4-dihydroisoquinoline (1a) with (E)-4-hydroxy-1-phenylbut-2-en-1-one (2a) in the presence of acetic acid (15 mol%), which might function as the additive or catalyst (entry 1). Then, it was found that product 3a could be also obtained in an acceptable yield in the absence of acetic acid (entries 2, 3). The yield of 3a remained almost unchanged either with or without acetic acid even if the reaction time was decreased to 15 min (entries 4, 5). Compared with toluene, both ether and EtOAc are better solvents for the reaction (entries 6–8).
Entry | Solventa | Additiveb | Time | Yieldc (%) |
---|---|---|---|---|
a Conditions: 3,4-dihydroisoquionline (0.76 mmol) and (E)-4-hydroxy-1-phenylbut-2-en-1-one (0.76 mmol) were mixed in solvent (1 mL).b AcOH (15 mol%) was added.c Isolated yield. | ||||
1 | Toluene | AcOHb | 1 h | 76 |
2 | Toluene | None | 1 h | 73 |
3 | Toluene | None | 3 h | 76 |
4 | Toluene | AcOHb | 15 min | 75 |
5 | Toluene | None | 15 min | 73 |
6 | EtOAc | AcOHb | 15 min | 93 |
7 | EtOAc | None | 15 min | 93 |
8 | Ether | None | 15 min | 83 |
9 | H2O | None | 15 min | 75 |
10 | MeOH | AcOHb | 15 min | 71 |
11 | MeOH | None | 15 min | 73 |
12 | EtOH | None | 15 min | 63 |
13 | n-BuOH | None | 15 min | 73 |
14 | DMF | None | 15 min | 55 |
Especially, the reaction furnished 3a in an excellent yield in EtOAc at room temperature in the absence of additive and catalyst (entry 7). Polar solvents, such as methanol (entries 10, 11) and ethanol (entry 12) could not help improve the efficiency of the addition. Butanol (entry 13) and DMF (entry 14) turned out to be relatively inappropriate solvents for the reaction.
The NMR technique and HRMS experiments were utilized for the characterization of the isolated product. Furthermore, the relative stereo-structure of a single crystal of compound 3a was confirmed with X-ray crystallography (Fig. 1). It revealed that the orientations of the two protons on chiral centres C-9 and C-11 were anti to each other.15a
For the investigation of the scope of substrates, at first, the reactions of various 3,4-dihydroisoquinoline20 derivatives with (E)-4-hydroxy-1-phenylbut-2-en-1-one (2a) were performed under the optimized conditions shown in Table 1 (entry 7). As presented in Table 2, all the reactions were finished at room temperature in 15 min, providing the corresponding products in good to excellent yields. It was found that the reaction tolerated both activating groups (3a–3h) and deactivating groups (3i–3l), which were on the benzo-moiety of the parent dihydroisoquinoline, regardless of their substitution positions. As expected, 6,7-dimethoxy-3,4-dihydroisoquinoline also underwent the reaction to smoothly afford product 3d (95% yield). Notably, it seemed that the phenyl group at C-7 atom of the molecule (3h) did not cause any steric hindrance to the addition. The diastereomeric ratio (dr) for each product was determined based on its proton NMR spectrum. It is noteworthy that the addition reaction was highly diastereoselective (3f:
dr > 25
:
1).
a dr = diastereomeric ratio determined with 1H-NMR spectrum. |
---|
![]() |
With the good results obtained from the above experiments in hand, we subsequently made a move to the exploration of the scope of γ-hydroxy-α,β-unsaturated ketones.21 As presented in Table 3, the addition of various 3,4-dihydroisoquinoline derivatives to the p-methoxyphenyl ketone gave the corresponding oxazolo[2,3-a]tetrahydroisoquinolines (4a–4d) in high to excellent yields. Moreover, products 4e–4g and 4h–4k were obtained easily from the reaction of the p-chlorophenyl ketone and that of the p-nitrophenyl ketone, respectively. Thus, it should be noted that the addition reaction tolerated both activating and deactivating groups on the para-position of phenyl moiety of the applicable γ-hydroxy-α,β-unsaturated ketones 2. Gratifyingly, the reaction of 2-naphthyl ketone 2 with dihydroisoquinolines 1 also proceeded very well and smoothly afforded target molecules 4l–4p (90–93% yield). On the other hand, the reactivity of ketones 2 which possess an alkyl group (R2) instead of an aryl was examined. It was observed that methyl ketone 2 reacted with the dihydroisoquinolines efficiently, providing expected products 4q–4t in high yields. Furthermore, compounds 4u–4x were also easily furnished from the consecutive additions, indicating that the reaction tolerated steric-demanding t-butyl unit on the ketone molecule. The relative stereochemistry of some products was confirmed with NOESY experiments, which provided the information that the NOE between H-9 and H-11 was not observed.
a dr = diastereomeric ratio, which was determined with 1H-NMR spectrum. |
---|
![]() |
As shown in Scheme 2, the addition reaction could be efficiently scaled up under the standard conditions. Reaction of 6-chloro-3,4-dihydroisoquinoline (12.7 mmol) with 2a (12.7 mmol) took 15 min at room temperature, and provided oxazolo[2,3-a]tetrahydroisoquinoline 3j in 85% yield (3.55 g). One might expertize on the potential application of this methodology to the preparation of key intermediates for some natural products or biological active compounds.
Two plausible mechanisms for the title reaction are illustrated in Fig. 2. At first, it was speculated that the title reaction consisted of two consecutive additions. For route a, regular nucleophilic addition of 1 with 2 furnished structure A,8 which immediately underwent intramolecular aza-Michael addition to give species B. Then, an intramolecular proton-transfer from the ammonium moiety to enolate proceeded in B to afford enol C. Product 4 was then obtained after a tautomerism occurred to enol C. At beginning of the formation of structure A, the oxygen atom of the hydroxyl group in 2 attacked the carbon atom of imino group in 1 from one side of the conjugate system. Consequently, the nitrogen atom on the secondary amino group in A attacked the β-carbon of the unsaturated ketone moiety predominantly from the opposite side during the formation of B. Thus, the reaction turned out to be highly diastereoselective. For route b, intermolecular aza-Michael addition of 1 to 2 provided zwitterion D, which immediately underwent tautomerism to furnish zwitterion E. Then, cyclization of E smoothly afforded product 4. The rational of the stereochemistry involved in route b could be similar to that involved in route a.
In order to further figure out the exact mechanism for the title reaction, two control experiments were performed under the standard conditions, as shown in Scheme 3. However, both mixture of 1 with 4-hydroxy-1-phenylbut-an-1-one (5)22 and mixture of 1 with (E)-1-phenylpent-2-en-1-one (6)23 revealed no reaction at all. Thus, it is presumed that the attack by the oxygen atom in 2 to the carbon atom of imino group in 1 and the attack by the nitrogen atom of imino group in 1 to the β-carbon of the unsaturated ketone moiety in 2 occurred simultaneously and followed two opposite directions. This situation could be the rational of the stereochemistry involved in the concerted mechanism (route c, Fig. 3). Therefore, route c cannot be excluded while route a or route b (Fig. 2) exhibits the origin of diastereoselectivity for the reaction.
In conclusion, we developed an efficient method for the diastereoselective synthesis of oxazolo[2,3-a]tetrahydro-isoquinolines. The addition reaction of 3,4-dihydro-isoquinolines with γ-hydroxy-α,β-unsaturated ketones could complete under “green” conditions without additive and catalyst, producing the target molecules in good to excellent yields. The applications of this [3 + 2] cycloaddition reaction to the syntheses of analogues of oxazolo[2,3-a]tetrahydro-isoquinolines and other heterocyclic compounds are under investigation.
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedure, characterization, NMR spectra, single-crystal XRD data. CCDC 1483927–1483929. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra20708a |
This journal is © The Royal Society of Chemistry 2016 |