Shobhon
Aich
a,
Rajesh
Nandi
a,
Nirbhik
Chatterjee
b,
Krishnanka S.
Gayen
c,
Subhasis
Pal
a and
Dilip K.
Maiti
*a
aDepartment of Chemistry, University of Calcutta, 92 A. P. C. Road, Kolkata-700009, India. E-mail: dkmchem@caluniv.ac.in; Fax: +91-33-2351 9755; Tel: +91-33-2350 1014
bDepartment of Chemistry, Kanchrapara College, North 24 parganas-743145, India
cDepartment of Chemistry, Raja Peary Mohan College, West Bengal, India
First published on 14th February 2023
We developed an efficient and straightforward I2-catalyzed strategy for the synthesis of functionalized α-amidohydroxyketones and symmetrical and unsymmetrical bisamides using incipient benzimidate scaffolds as starting materials and moist-DMSO as a reagent and solvent. The developed method proceeds through chemoselective intermolecular N–C-bond formation of benzimidates and the α-C(sp3)–H bond of acetophenone moieties. The key advantages of these design approaches include broad substrate scope and moderate yields. High-resolution mass spectrometry of the reaction progress and labeling experiments provided suitable evidence regarding the possible mechanism. 1H nuclear magnetic resonance titration revealed notable interaction between the synthesized α-amidohydroxyketones and some anions as well as biologically important molecules, which revealed a promising recognition property of these valuable motifs.
“Recognition” is one of the “hotspots” of supramolecular chemistry15–18 and ensures wide applicability in various fields. Numerous modes of noncovalent interactions between a host and guest are predicted as major reasons for their origination. In most cases, the receptor parts are cage-like motifs or acyclic compounds capable of providing suitable orientation that are responsible for appropriate interaction with the guest. Hydroxyl groups bearing α-ketoamide have a unique architecture so we explored their recognition properties using 1H nuclear magnetic resonance (1HNMR) titration.
Our initial attempts for the construction of α-amidohydroxyketone (3aa) employing ethyl benzimidate (1a) and acetophenone (2a) covered several prospective catalysts (10 mol%), such as Cu(OTf)2, CuCl2, CuI, AgOTf, AgOAc and Zn(OTf)2 (Table 1, entry 1), in DMSO/H2O at 120 °C, but none of them provided a suitable outcome. Lewis acids such as ZnCl2, Ni(OTf)2, NiCl2·4H2O and AlCl3 exhibited promising results, but the yields were low (Table 1, entries 2–5). Only FeCl3·6H2O led to marginal improvement of the yield (30%, Table 1, entry 6). Gratifyingly, when we shifted our focus to molecular iodine (I2, 10 mol%) in DMSO/H2O, the yield was enhanced markedly (80%) with concomitant reduction of the reaction time (12 h, Table 1, entry 7). In the absence of iodine (I2), the reaction did not progress (Table 1, entry 8). This clearly indicated that molecular iodine was essential for a strategy for the construction of a N–C bond. Detailed solvent studies (Table 1, entries 9–16) showed DMSO/H2O to be the best choice (Table 1, entry 7) in comparison with other common polar and nonpolar organic solvents. Variation in catalyst load had a detrimental effect on the yield of the product (Table 1, entries 17 and 18). Hence, the optimum reaction conditions were 10 mol% iodine (I2) in DMSO at 120 °C to deliver product 3 in good yield (Table 1, entry 7). A gram-scale synthesis under standard reaction conditions afforded the anticipated product 3aa in good yield (75%) within 14 h (Table 1, entry 19), which demonstrated the applicability of this approach in academia and industry. We showed that using an exact amount of moisture (Table 1, entries 20–22) was suitable for the new reaction, and a DMSO:water ratio of 10:1 (v/v, Table 1, entry 21) was employed.
Entry | Catalyst | mol% | Solvent | Conditions (°C) | Time (h) | Yieldb (%) |
---|---|---|---|---|---|---|
a Reactions were carried out using ethyl benzimidate (1a, 1 mmol) and acetophenone (2a, 1 mmol) and I2 catalyst (0.1 mmol). b Yield of the isolated product after column purification. c No reaction. d Unreacted 1a and 2a were recovered. e Gram-scale synthesis. f DMSO (2 ml) and water (0.1 ml). g DMSO (2 ml) and water (0.2 ml). h DMSO (2 ml) and water (0.3 ml). | ||||||
1 | Zn(OTf)2 | 10 | DMSO/H2O | 120 | 24 | NRc |
2 | ZnCl2 | 10 | DMSO/H2O | 120 | 24 | 20d |
3 | Ni(OTf)2 | 10 | DMSO/H2O | 120 | 24 | 25d |
4 | NiCl2·4H2O | 10 | DMSO/H2O | 120 | 20 | 24d |
5 | AlCl3 | 10 | DMSO/H2O | 120 | 24 | 24 |
6 | FeCl3·6H2O | 10 | DMSO/H2O | 120 | 24 | 30d |
7 | I 2 | 10 | DMSO/H 2 O | 120 | 12 | 80 |
8 | — | — | DMSO/H2O | 120 | 24 | NRc |
9 | I2 | 10 | THF/H2O | Reflux | 24 | NRc |
10 | I2 | 10 | Dioxane/H2O | Reflux | 24 | 23 d |
11 | I2 | 10 | DCE/H2O | Reflux | 20 | NRc |
12 | I2 | 10 | DCM/H2O | Reflux | 22 | NRc |
13 | I2 | 10 | DMF/H2O | 120 | 20 | 22d |
14 | I2 | 10 | EtOH/H2O | Reflux | 12 | NRc |
15 | I2 | 10 | Toluene/H2O | Reflux | 24 | NRc |
16 | I2 | 10 | Benzene/H2O | Reflux | 24 | NRc |
17 | I2 | 7 | DMSO/H2O | 120 | 20 | 55 |
18 | I2 | 15 | DMSO/H2O | 120 | 12 | 65 |
19 | I2 | 10 | DMSO/H2O | 120 | 14 | 75e |
20 | I2 | 10 | DMSO/H2O | 120 | 12 | 45f |
21 | I2 | 10 | DMSO/H2O | 120 | 12 | 80g |
22 | I2 | 10 | DMSO/H2O | 120 | 12 | 79h |
After this successful survey, we proceeded to scrutinize the substrate scope utilizing benzimidate (1a) and a wide range of acetophenones (2a–j) under the developed reaction conditions. The desired α-amidohydroxyketone derivatives (3aa–aj) were obtained with good-to-high yields (72–85%, Scheme 2). To our delight, the reactions with sterically hindered ortho-substituted acetophenones (2b and 2c) were also well tolerated to furnish the crowded 3ab and 3ac, respectively. Replacement of acetophenone by cyclohexanone or cyclopentanone did not provide the expected product but instead led to complex reaction mixture.
To explore the substrate scope further, reactions between electronically enriched and poor benzimidates and electronically poor, sterically hindered acetophenone analogues were examined to obtain the desired functionalized α-amidohydroxyketones. Gratifyingly, electron-rich p-methylbenzimidate (1b) and p-methoxybenzimidate (1c) were compatible with electronically poor, electronically rich and sterically hindered ketones to provide the desired products 3ba–cj with significant yields. Again, electronically poor benzimidates (1d-p-NO2 and 1g-m-B(OH)2) also provided delightful outcomes (3de, 3gd). Heterocyclic imidates, such as ethyl thiophene-2-carbimidate 1e (3ea–ei, 76–78%) and ethyl furan-2-carbimidate 1f (3fa, 80%) were well tolerated. However, we did not obtain the desired product for pyridine or quinoline. All synthesized compounds were fully characterized by spectroscopic data (ESI) and the structural skeleton of 3ea was confirmed with the help of single-crystal X-ray diffraction analysis (Scheme 3).
To acquire a clear-cut vision of the mechanistic pathway, we pursued another important study whereby, under the developed reaction conditions, H2O was replaced by D2O. 1H NMR spectroscopy of the corresponding product (3aa(d), Scheme 4) clarified the absence of an O–H proton (ESI†), which indicated the involvement of nucleophilic attack by H2O (or D2O) during the course of the reaction.
During our survey, we noticed that the benzimidate moiety itself underwent unique coupling under the developed reaction conditions to provide gem-bisamide (Schemes 5 and 6), and the source of the methylene group was from DMSO. This prediction was justified by the fact that when DMSO-d6 was used instead of DMSO, there was no relevant peak of the –CH2 group in 1H NMR spectroscopy (ESI). It had been replaced by –CD2. Unlike previous reports6–9 our strategy was also capable of providing unsymmetrical gem-bisamides (Scheme 6).
In addition to the experiment described in Scheme 4, high-resolution MS was undertaken to find suitable evidence of the intermediates formed during the course of the reaction. This allowed justification of the reaction mechanism (calculated and observed values are given in parenthesis, Scheme 7 and ESI†). At the very beginning, acetophenone is converted to intermediate A in the presence of the catalyst (molecular iodine). Then, nucleophilic attack by the benzimidate and concomitant release of HI generates intermediate B. Hydrolysis of B results in formation of the iodo amidoketone C. Finally, release of the iodide ion from C provides the new iminium intermediate D and simultaneously leads to aerial oxidation to regenerate molecular iodine, and the cycle proceeds as usual. Nucleophilic attack upon D by a water molecule furnishes the desired architecture 3aa. A similar mode of investigation was carried out for the synthesis of bisamide 4cc.
In this regard, the proposal (Scheme 8) was that first DMSO is converted to A′ in the presence of molecular iodine. Now, as usual, the nucleophilic imidate reacts with A′ to provide substituted sulfoxide B′. Oxidation of B′ generates the sulfone C′. Elimination of MeSO2− from C′ results in the active iminium intermediate D′. Finally, a second molecule of the imidate moiety provides condensation with D′ to furnish the desired bisamide 4cc.
Entry | Aniona | Notable change in O–H and C–H regions in 1H NMR after addition of 40 μl of anion solution | Disappearance of O–H peak in 1H NMR after addition of 40 μl of anion solution | Notable change in aromatic and N–H regions in 1H NMR after addition of 40 μl of anion solution |
---|---|---|---|---|
a Anion solution = 1 mg in 1 ml of distilled water (same for uracil and thymine, for adenine solvent is 1:1 distilled water and DMSO-d6). Substrate solution = 5 mg in 0.5 ml of DMSO-d6. NMR titration was done by gradual addition of the anion solution (from 10 μl to 40 μl) to the substrate solution. | ||||
1 | OAc− | Yes | Yes | Yes |
2 | Cl− | Yes | Yes | Yes |
3 | CN− | Yes | Yes | No |
4 | H2PO4− | Yes | Yes | No |
5 | F− | Yes | No | No |
6 | NO3− | No | — | — |
7 | NO2− | No | — | — |
8 | HPF6− | No | — | — |
9 | HCO3− | No | — | — |
10 | Adenine | Yes | No | Yes |
11 | Thymine | Yes | Yes | Yes |
12 | Uracil | Yes | Yes | No |
13 | ATP | Yes | Yes | No |
Quite interestingly, thymine (Table 2, entry 11) was similar to chloride and acetate, and had a notable appearance in the entire 1H NMR spectrum (ESI†).
The 1H NMR spectrum of 3aa revealed the –OH and –CH peaks to have merged (Fig. 2b) but their appearances were quite distinctive for 3de (Fig. 2a). This scenario is relevant for other products in which the aromatic rings are substituted. Hence, we extended our NMR-titration study to observe the change in the 1H NMR spectrum of 3aa. After addition of Cl(−) solution (30 μl) to the sample solution, the –OH and –CH peaks began to separate (Fig. 2c) and the gap became slightly wider upon addition of Cl(−) (40 μl) (ESI†) but, unlike 3de, the –OH peak did not vanish.
Fig. 2 (a) 1H NMR of 3de, (b) 1H NMR of 3aa and (c) 1H NMR of 3aa after addition of Cl(−) solution (30 μl). |
Next, to establish the competence of the recognition property of synthesized products, we conducted a comparative 1H NMR-titration study between 3aa and a similar compound (compound 6), in which a methylene group is adjacent to –NH instead of a hydroxyl-bearing methine. We synthesized 6 according to the literature.19 Even after addition of Cl(−) solution (40 μl), a change in the 1H NMR spectrum of 6 was not observed (Fig. 3). Hence, the presence of the –CH(OH) part in our synthesized α-amidohydroxyketones had a profound effect upon the recognition properties.
Footnote |
† Electronic supplementary information (ESI) available: Details experimental procedures and full spectroscopic data of all the new synthesised compounds are provided. CCDC 2152756. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3ob00165b |
This journal is © The Royal Society of Chemistry 2023 |