Jan Otevrel‡
*,
David Svestka‡ and
Pavel Bobal*
Department of Chemical Drugs, Faculty of Pharmacy, Masaryk University, Palackeho 1946/1, 612 42 Brno, Czech Republic. E-mail: pbobal@icloud.com; otevrel@icloud.com
First published on 30th June 2020
We have identified a novel one-pot method for the synthesis of β-amino alcohols, which is based on C–H bond hydroxylation at the benzylic α-carbon atom with a subsequent nitrile or amide functional group reduction. This cascade process uses molecular oxygen as an oxidant and sodium bis(2-methoxyethoxy)aluminum hydride as a reductant. The substrate scope was examined on 30 entries and, although the respective products were provided in moderate yields only, the above simple protocol may serve as a direct and powerful entry to the sterically congested 1,2-amino alcohols that are difficult to prepare by other routes. The plausible mechanistic rationale for the observed results is given and the reaction was applied to a synthesis of a potentially bioactive target.
As the research in this field is vibrant, a number of possible routes for the β-amino alcohol synthesis including their enantioselective versions have been described and well-reviewed in the literature.2b,3 Accordingly, five general protocols have been established: (1) functional group interconversions of compounds with both heteroatoms already present in the molecule, such as a reduction of nitroaldols, cyanohydrins, azidohydrins, amino acids, amino ketones, hydroxy imines, etc.;4 (2) ring-opening of epoxides and aziridines;5 (3) addition of a single heteroatom to a molecule already containing oxygen or nitrogen functionality;2b (4) introduction of both heteroatoms in a single process, i.e. by aminohydroxylation or oxidative nitration;6 (5) C–C bond-forming reactions including 1,2-additions to α-hydroxy imines and α-amino aldehydes.7
Compared to other methods, the introduction of oxygen to a molecule already containing a nitrogen functionality is not a common approach for the synthesis of β-amino alcohols and hence remained considerably underexplored. Several reactions providing entry into this interesting pathway mainly comprise of the addition of O-nucleophiles to α,β- or β,γ-unsaturated amines3a,8 and nitroolefins,9 hydroboration–oxidation of enamines8 or the Schlenck ene reaction.10 Herein we disclose another, conceptually different method, which is based on a protocol including α-carbon atom hydroxylation and amide or nitrile functional group reduction.
In connection with our interest in the development of organocatalysts for the asymmetric aldol-type reactions, we have been continuously compelled to explore novel chiral auxiliaries for the catalyst design and screening. As part of the above synthetic efforts, we have prepared many chiral non-racemic amine scaffolds by reductions of the corresponding enantiopure amides.11
For such transformations, we normally utilized sodium bis(2-methoxyethoxy)aluminum hydride (SMEAH) as a reductant.12 In comparison to lithium aluminum hydride (LAH), SMEAH is much better soluble, easier to handle, insensitive to dry air, thermally stable (<205 °C), and non-violently reacting with water, which makes it an invaluable reducing agent for both laboratory and industrial use.13 Although it has been documented that SMEAH exhibited a unique reactivity and selectivity in some cases,13,14 we have been using it as an equivalent reagent to LAH for numerous amide reductions.
However, (S)-naproxamide subjected to the above reaction conditions gave a curious product, which was then, to our surprise, identified as 1,2-amino alcohol derivative (Fig. 1). From the initial insights, the formation of this unexpected substance seemed to be favored by a longer reaction time together with a large excess of SMEAH. As this represented, to the best of our knowledge, the only example of the organic transformation that enabled a direct conversion of the primary amide to the appropriate vic-amino alcohol derivative, it stimulated us into a research of this interesting phenomenon in more detail.
Fig. 2 Partial 1H NMR (400 MHz, DMSO-d6, 35 °C) spectra showing the 3a/4a ratio affected by the type of the complex metal hydride. |
Further, we examined other reaction solvents besides toluene, such as diethyl ether, glyme, dioxane, and THF. The highest selectivity towards the 1,2-amino alcohol formation was reached by performing the reduction in anhydrous THF (3a/4a = 92:8). Although a pretty good selectivity was obtained at this stage, we were still discouraged by a poor yield (21%) and a very long reaction time, which was necessary for completion of the reaction (>48 h). Finally, the optimization of the reaction atmosphere proved to be the major turning point. The switching of the previously used argon atmosphere to dry air and then oxygen dramatically enhanced both the reaction rate and the yield. Consequently, the desired product was furnished in 43% yield in less than 3 h even with the slightly improved selectivity (3a/4a = 94:6). Further attempts to increase the product yield were unsuccessful. Nevertheless, according to our experience, the analogous LAH reductions of 2-arylalkanoic acid amides or nitriles to amines also barely exceeded isolated yields of 50–60%.
In the following task, we investigated the scope of possible functional groups that undergo the β-amino alcohol-forming process. The obtained data is summarized in Scheme 1. Accordingly, the appropriate N-methylamide (1aa), hydrazide (1ab), hydroxamic acid (1ac), thioamide (1ad), and nitrile (2a) derivatives of the model compound were tested under the optimized reaction conditions. The above N-methylamide (1aa) provided a regular N-methylamine (4aa) instead of the desired N-methylamino alcohol, while the respective reaction of hydrazide (1ab) or hydroxamic acid (1ac) resulted only in a complete recovery of starting material. These experiments have confirmed that the corresponding N-substituted derivatives failed to provide β-amino alcohol products under the screening conditions. Amongst other tested functional groups, the best results in terms of the selectivity towards the formation of 3a were obtained with the amide (1a) and nitrile (2a) functions. Therefore these types of substrates were selected for the tests henceforth.
Following the optimization of the functional groups, the feasible substituents of the α-position of substrates were evaluated (Scheme 2). The α-oxidation of the compound 1b lacking the α-substituent turned out to be even more sensitive to the reaction atmosphere than 1a as the selectivity towards the 1,2-amino alcohol formation (3b/4b) was improved from 81:19 to 95:5 by replacing Ar by O2 (Fig. S10–11†). The control α-OH bearing substrate 1ba was reduced under the identical reaction conditions to the sole 3b in 36% yield. The SMEAH-mediated reduction of the α-methoxy and α-phenylsulfanyl functionalized derivatives 1bb and 1bc led at least partially to a displacement of the original α-substituent by OH group. On the contrary, the α-chlorine was preserved and 1bd yielded the regular 1,2-chloroamine 4bb in 57%. The α-alkyl and α-aryl derivatives 1c, 1cc, and 1d afforded the respective β-amino alcohols in excellent selectivities and modest yields. Surprisingly, reduction of the α-CF3 substituted 1ca provided the completely hydrodefluorinated derivative 3c identical to the product of the reaction of 1c. On the other hand, the corresponding α-OH analog thereof (1cb) did not exhibit any detectable defluorination under the above reaction conditions. According to these observations and the available literature,14d,15 we have postulated a plausible hypothesis that defluorination of the CF3 group of 1ca could be driven by a series of fluoride β-elimination – hydride addition events of the intermediary α-carbanion leading to 1c that is then converted to the β-amino alcohol 3c. This is also supported by the fact that the respective α,α-disubstituted derivative 1cb did not lose any of the fluorine atoms during the reduction.
Next, we tested also several nitrile derivatives 2b, 2ba, 2c, and 2d, which gave the results comparable with the amide partners regarding both the yield and the selectivity. The remaining amides 1e–h and also their nitrile counterparts 2e–h together with 1a–d and 2a–d are presented in Scheme 3.
In all the described experiments, the corresponding 1,2-amino alcohols were isolated by simple filtration of their oxalic acid salts. After isolation of the respective ammonium salts, the filtrate afforded the one-carbon-shorter alcohol as a predominant by-product in all the displayed entries. However, the corresponding dehomologated alcohols were isolated and purified only in the instance of the α-substituted derivatives (entries 1 and 3–6), where this method, in contrast to the α-unsubstituted 1b and 1g–h, can constitute at least an interesting route for the synthesis thereof.
As depicted in Schemes 1–3, the reaction tolerated well hydrogen, alkyl or aryl substituents at the benzylic position. Aromatic substitution may comprise alkyl and alkoxy groups. Aryl halides such as fluorine also remained practically untouched, however, in the case of aryl chloride, we have encountered dehalogenation on a significant level,16 which was also documented in the literature.17 In summary, it should be emphasized that although the reduction power of SMEAH is attenuated in comparison with the parent LAH due to the presence of two bulky and electron-withdrawing alkoxy groups, it is still a very powerful reducing agent and, as such, it is incompatible with almost any reducible functional group in a molecule. Thus, these properties considerably limit the chemoselectivity of the present process.13b,13c
To establish a plausible mechanistic proposal, a series of experiments with 1d, 2d, and other diphenyl substituted derivatives were performed. Initially, the respective products 3d and 4d were subjected to the standard reaction conditions (entries 1–2, Scheme 4). The full recovery of the unchanged material in both the cases supported that, according to our expectations, there is no detectable interconversion between the products and that the oxidation of the α-carbon precedes the functional group reduction step.
Scheme 4 Control experiments with 3d, 4d, 1d, 1da, 2d, and 2da. Notes: the yields were determined by RP-HPLC analysis using the external standard calibration. |
As mentioned above, the SMEAH-mediated reduction of 1d provided the oxalic acid salt of 3d as a sole product in the precipitate and 5c as a predominant side product in the filtrate.18
The subsequent analysis of the filtrate and particularly the fore fractions obtained during the purification of 5c by column chromatography (SiO2) n-heptane–EtOAc (9:1), revealed the minor and trace components 1da, 2d, 6, and 7, which were further confirmed by comparison with the authentic samples. The presence of 2d and 7 in the reaction mixture of 1d led us to assume that reduction of the primary amide likely involved intermediacy of the nitrile (Scheme 5).4 Moreover, if we consider that the respective nitrile 2d gave a similar product profile with the comparable selectivity in even better yield and shorter reaction time than the corresponding amide 1d (Scheme 4), it seems plausible to suppose that both 1d and 2d could share a common reduction pathway. The above hypothesis is also in connection with the aforementioned incapability of the secondary amides to undergo the present process.
It has been reported earlier that a striking difference between the reduction properties of SMEAH and LAH towards nitriles was found with reductions of compounds carrying hydrogen atom at α-carbon, which proceeded unsatisfactorily with SMEAH. Hence the majority of the starting nitrile was always recovered on workup.19 These findings were also validated by us when the above reaction was conducted under the strictly oxygen-free conditions.
Therefore we have hypothesized that the nitrile intermediate 2d, which can be only α-deprotonated but not efficiently further reduced by SMEAH, is present in the reaction mixture as a colored resonance-stabilized nitrile anion. The small portion of the reduced substrate 2d is likely responsible for a formation of the minor by-product 4d, which can become more significant under elevated temperature.20 Under basic conditions, the electron-rich nitrile anions are susceptible to autooxidation via single-electron transfer (SET).21 Hence, the anion of 2d is probably trapped by O2 to the corresponding α-hydroperoxy nitrile, which is then rapidly reduced to the cyanohydrin 2da and the successive species (Scheme 5). This presumption was supported by the identification of the corresponding radical homocoupling product 7 from the above reaction mixture.
Since cyanohydrins are formed reversibly, they are considered to be unstable in basic media (entry 6, Scheme 4). Especially for cyanohydrins that are sterically hindered, the position of the equilibrium is unsatisfactory for the effective synthesis thereof. Therefore, cyanohydrins encumbered with bulky substituents are rather prone to the elimination of cyanide ion and the formation of the corresponding dehomologated carbonyl compounds.22
Our findings indicate that the moderate yields of β-amino alcohols in the present transformation are probably caused by the previously mentioned retro-cyanohydrination (Scheme 5). These speculations were also supported by isolation of traces of the ketone 6 from the SMEAH-driven reductions of 1d, 1da, 2d, and 2da.23
Under the above conditions, the presumptive α-hydroxy nitrile intermediate 2da can undergo either reduction to the desired β-amino alcohol 3d or expulsion of cyanide anion to 6. The resulting degraded carbonyl compound 6 is then readily reduced to the alcohol 5c by the hydride. To further prove this hypothesis, the separately prepared cyanohydrin 2da was subjected to identical reaction conditions. Accordingly, 4d and 5c were isolated as the major products. However, in comparison with the reaction starting from 1d or 2d, the desired β-amino alcohol 4d was obtained in a somewhat decreased yield (21%). Expectably, the yield of the corresponding alcohol 5c was slightly increased in this case (63%). Thus we suppose that the continuous reduction of cyanohydrin generated in situ during the present process is likely beneficial to the batch reduction of 2da, as it eliminates the excessive decomposition of this relatively labile compound by retro-cyanohydrination. Besides that, we have found that the formation of 5c was favored over 3d mainly under a diminished reaction temperature (<0 °C). This likely corresponds to the instability of 2da in a basic environment for the extended period necessary for the slower reduction thereof under a decreased temperature.
We further speculated whether the hypothetically employed decyanation of 2da might be suppressed by raising the concentration of free cyanide in the reaction medium according to Le Chatelier's principle. Indeed, the addition of powdered NaCN (10 equiv.) to the reaction mixture of 2d furnished a slightly increased yield of 3d (50% → 61%) after the prolonged reaction time (3 h) with the improved selectivity expressed by 3d/4d (92:8 → 99:1, Fig. S35–37†).
Although the residues of 1da were also detected in the SMEAH-mediated reduction involving 1d, the autooxidation of 1d to 1da presumably constitutes only a less important path in the overall process. Based on the control experiment with sodium hydride, the reaction time longer than 12 h was required for completion in this case, which is in contrast to the rapid autooxidation of 2d under the same conditions (entries 3 and 5, Scheme 4). Moreover, the analogous reaction of 1a with NaH under O2 atmosphere did not provide any product even after 24 h, which suggests that α-oxidation of nitriles is considerably faster in comparison with the parent primary amides. The subjection of α-hydroxy amide 1da to the control experiment with NaH resulted in the complete recovery of starting material (entry 4, Scheme 4). On the other hand, the SMEAH-driven reaction thereof afforded both 3d and 5c as the major products in accordance with 1d, 2d, and 2da. Hence we suppose that the possible reaction pathways of all the aforementioned derivatives with SMEAH can plausibly proceed as depicted in Scheme 5.
The analysis of the stereochemical course of the reaction involving (S)-naproxamide (S)-1a gave us evidence for complete racemization of the optically pure substrates lacking the α-hydroxy group during the reaction. It is in agreement with our assumption that the initial α-proton abstraction and mainly the subsequent radical isomerization led to the loss of the stereochemical integrity (Scheme 6).24 The respective alcohol 5a was isolated from the reaction mixture of (S)-1a as a slightly enriched (R)-isomer (6% ee).25 This phenomenon is not yet clearly understood by us but might be attributed to the asymmetric reduction of the corresponding ketone 6 by a non-racemic complex of the hydride with the substrate or some chiral intermediate. To the contrary, (S)-mandelamide (S)-1ba subjected to identical reaction conditions furnished (S)-3b with retention of its configuration, which suggests that the α-hydroxy-substituted stereocenter remained unaffected.
To showcase the synthetic utility of the developed process, we have prepared the racemic fluoro-analog of the biarylpropylsulfonamide potentiator of AMPA receptors, LY-503430, by our reaction (Scheme 7).26 Accordingly, the respective β-amino alcohol product 3e was subjected to the fluorodehydroxylation reaction with the Olah's reagent to afford the respective 1,2-fluoroamine 9 in 69% yield. This intermediate was further converted to the sulfonamide 10 using isopropylsulfonyl chloride (y. 37%).
Scheme 7 Synthesis of the LY-503430 analog. Conditions: (a) 65% HF-pyridine, CH2Cl2, rt, 24 h (y. 69%); (b) i-PrSO2Cl, Et3N, DMAP, CH2Cl2, 0 °C, 3 h (y. 37%). |
Analytical data for 3a·H2C2O4: Mp 226–228 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.94 (s, 1H), 7.82 (dd, J = 8.7, 2.8 Hz, 2H), 7.59 (d, J = 8.6 Hz, 1H), 7.31 (d, J = 1.9 Hz, 1H), 7.17 (dd, J = 8.9, 2.3 Hz, 1H), 3.87 (s, 3H), 3.13, 3.10 (q, AB, JAB = 13.0 Hz, 2H), 1.57 (s, 3H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.2, 157.3, 140.3, 133.3, 129.5, 128.0, 126.7, 124.0, 123.5, 118.6, 105.6, 71.0, 55.1, 49.1, 27.4; IR (neat) /cm−1: 3459w, 2961w, 1606m, 1515m, 1386w, 1265m, 1179s, 1025m, 898m, 854m, 812m, 702s; HRMS (ESI-Q-TOF) m/z: calcd for C14H18NO2 [M + H]+ 232.1332, found 232.1328.
Analytical data for 5a:49 Mp 93–95 °C; 1H NMR (400 MHz, CDCl3) δ/ppm: 7.75–7.71 (m, 3H), 7.48 (dd, J = 8.6, 1.6 Hz, 1H), 7.18–7.13 (m, 2H), 5.04 (q, J = 6.5 Hz, 1H), 3.93 (s, 3H), 2.26 (br s, 1H), 1.58 (d, J = 6.4 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ/ppm: 157.6, 140.9, 140.0, 129.4, 128.7, 127.1, 124.4, 123.7, 118.9, 105.6, 70.4, 55.3, 25.0; IR (neat) /cm−1: 3326br, 2961w, 1633w, 1605s, 1485m, 1462m, 1391w, 1260m, 1216m, 1162s, 1072m, 1028s, 961w, 930w, 890m, 853s, 815s, 750w, 673w; HRMS (ESI-Orbitrap) m/z: calcd for C13H15O2 [M + H]+ 203.1067, found 203.10664.
Analytical data for rac-8a (0% ee): Yield: 100 mg (45%). Physical state: white powder. Mp 193–194 °C; [α]25D 0 (c 1.0, CHCl3). HPLC conditions: a Hypersil silica column (3 μm, 100 × 4.6 mm) used as a precolumn, which was connected via the standard blue PEEK capillary tubing (L 300 mm, ID 0.01′′, OD 1/16′′) to a Daicel Chiralpak IA column (5 μm, 250 × 4.6 mm), i-PrOH–n-heptane, 10:90, 0.5 mL min−1, 25 °C (230 nm), t1 = 36.36 min, t2 = 49.38 min. 1H NMR (400 MHz, CDCl3) δ/ppm: 7.71–7.69 (m, 2H), 7.60 (d, J = 8.6 Hz, 1H), 7.37 (dd, J = 8.6, 1.7 Hz, 1H), 7.22 (d, J = 2.4 Hz, 1H), 7.12 (dd, J = 9.0, 2.4 Hz, 1H), 6.97 (t, J = 6.2 Hz, 1H), 6.81 (s, 2H), 5.24 (s, 1H), 3.86 (s, 3H), 3.05 (dd, J = 6.2, 1.3 Hz, 2H), 2.40 (s, 6H), 2.18 (s, 3H), 1.44 (s, 3H); 13C NMR (100 MHz, CDCl3) δ/ppm: 157.0, 141.5, 141.1, 138.0, 134.2, 133.1, 131.4, 129.5, 127.9, 126.1, 124.2, 123.4, 118.3, 105.5, 72.7, 55.1, 53.5, 27.3, 22.5, 20.4; IR (neat) /cm−1: 3501w, 3296w, 2930w, 1605m, 1455w, 1378w, 1314s, 1261m, 1199m, 1171m, 1142s, 1084m, 1034m, 960w, 854s, 819s, 739w; HRMS (ESI-Orbitrap) m/z: calcd for C23H26NO4S [M − H]− 412.1588, found 412.1590.
Analytical data for (R)-5a were in accordance with a compound 5a reported above. [α]25D +5 (c 0.4, CH2Cl2), 6% ee; lit.49 [α]25D +26.7 (c 0.4, CH2Cl2), 70% ee. HPLC conditions: a Hypersil silica column (3 μm, 100 × 4.6 mm) used as a precolumn, which was connected via the standard blue PEEK capillary tubing (L 300 mm, ID 0.01′′, OD 1/16′′) to a Daicel Chiralpak IB column (5 μm, 250 × 4.6 mm), i-PrOH–n-heptane, 5:95, 0.5 mL min−1, 25 °C (230 nm), t1 = 28.50 min (S-isomer), t2 = 34.07 min (R-isomer).
Analytical data for 4aa·H2C2O4: Mp 209–211 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.82–7.77 (m, 2H), 7.72 (s, 1H), 7.42 (d, J = 9.0 Hz, 1H), 7.30 (d, J = 2.3 Hz, 1H), 7.16 (dd, J = 8.9, 2.5 Hz, 1H), 3.87 (s, 3H), 3.29–3.17 (m, 3H), 2.54 (s, 3H), 1.34 (d, J = 6.5 Hz, 3H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.3, 157.1, 137.6, 133.4, 129.0, 128.5, 127.2, 125.9, 125.4, 118.6, 105.8, 55.1, 54.4, 36.3, 33.1, 19.5; IR (neat) /cm−1: 2970br, 1715m, 1608s, 1485m, 1385w, 1263s, 1218s, 1162m, 1028s, 950w, 849s, 809m; HRMS (ESI-Q-TOF) m/z: calcd for C15H20NO [M + H]+ 230.1539, found 230.1545.
Analytical data for 4a·H2C2O4 (the authentic sample was prepared by a DIBAH-mediated reduction of 1a): Mp 225–226 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.81–7.78 (m, 2H), 7.70 (s, 1H), 7.42 (dd, J = 8.5, 1.7 Hz, 1H), 7.30 (d, J = 2.5 Hz, 1H), 7.16 (dd, J = 8.9, 2.5 Hz, 1H), 3.87 (s, 3H), 3.19–3.03 (m, 3H), 1.33 (d, J = 6.7 Hz, 3H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.2, 157.1, 137.7, 133.4, 129.0, 128.5, 127.1, 125.9, 125.4, 118.5, 105.8, 55.1, 44.9, 37.4, 19.2; IR (neat) /cm−1: 2959br, 1717w, 1606s, 1503m, 1458w, 1392w, 1263m, 1218s, 1197m, 1162m, 1030m, 928w, 849s, 815m, 705s; HRMS (ESI-Q-TOF) m/z: calcd for C14H18NO [M + H]+ 216.1383, found 216.1379.
Analytical data for 3b·H2C2O4 (in accordance with the authentic sample):50 Mp 232–235 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.41–7.35 (m, 4H), 7.33–7.28 (m, 1H), 4.81 (dd, J = 9.9, 2.8 Hz, 1H), 3.04 (dd, J = 12.6, 2.8 Hz, 1H), 2.84 (dd, J = 12.6, 9.9 Hz, 1H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 163.7, 141.9, 128.3, 127.6, 125.8, 69.1, 45.7; IR (neat) /cm−1: 2884br, 1582s, 1446w, 1296m, 1061m, 1020m, 780m, 700m; HRMS (ESI-Orbitrap) m/z: calcd for C8H12NO [M + H]+ 138.0913, found 138.0915.
Analytical data for (S)-8b (>99% ee):51 Yield: 55 mg (48%). Physical state: colorless oil. [α]25D +40 (c 1.0, CHCl3). HPLC conditions: a Hypersil silica column (3 μm, 100 × 4.6 mm) used as a precolumn, which was connected via the standard blue PEEK capillary tubing (L 300 mm, ID 0.01′′, OD 1/16′′) to a Daicel Chiralpak IA column (5 μm, 250 × 4.6 mm), i-PrOH–n-heptane, 10:90, 0.5 mL min−1, 25 °C (230 nm), t = 57.51 min (S-isomer). 1H NMR (400 MHz, CDCl3) δ/ppm: 8.08 (s, 1H), 7.35–7.26 (m, 5H), 6.95 (s, 2H), 4.94 (dd, 7.1, 4.2 Hz, 1H), 4.79–4.76 (m, 1H), 3.20 (ddd, J = 13.2, 8.5, 4.5 Hz, 1H), 3.00 (ddd, J = 13.2, 8.5, 4.5 Hz, 1H), 2.62 (s, 6H), 2.29 (s, 3H); 13C NMR (100 MHz, CDCl3) δ/ppm: 142.4, 140.9, 139.1, 133.5, 132.1, 128.8, 128.4, 125.9, 72.8, 49.8, 23.0, 21.0; IR (neat) /cm−1: 3313w, 1603w, 1453w, 1318m, 1188w, 1149s, 1055m, 917w, 852w, 756w, 700m; HRMS (ESI-Orbitrap) m/z: calcd for C17H20O3NS [M − H]− 318.1169, found 318.1172.
Analytical data for 4ba·H2C2O4: 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.44–7.33 (m, 5H), 4.46 (dd, J = 8.7, 3.4 Hz, 1H), 3.19 (s, 3H), 3.03–2.94 (m, 2H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.0, 133.8, 128.6, 128.4, 126.7, 56.2, 44.3; HRMS (ESI-Orbitrap) m/z: calcd for C9H14NO [M + H]+ 152.1070, found 152.1069.
Analytical data for 4bb·H2C2O4: Mp 173–175 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.48–7.37 (m, 5H), 4.28–4.25 (m, 1H), 3.75–3.66 (m, 2H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.2, 135.7, 128.5, 128.4, 127.4, 63.0, 56.0; IR (neat) /cm−1: 2968w, 1593m, 1529m, 1202s, 1057s, 765w; HRMS (ESI-Orbitrap) m/z: calcd for C8H11NCl [M + H]+ 156.0575, found 156.0570.
The reduction of 1c (194 mg, 1.3 mmol) was performed according to GP. The precipitated product was obtained as an inseparable mixture of 3c·H2C2O4/4c·H2C2O4 in a 95:5 ratio based on 1H NMR and 94:6 according to RP-HPLC. HPLC conditions: an YMC-Triart C18 column (3 μm, 150 × 2.0 mm), MeCN–50 mM HCOONH4 (pH 9.30), 30:70, 0.2 mL min−1, 40 °C, 210 nm, t1 = 2.77 min (3c), t2 = 4.76 min (4c). Yield: 101 mg (32%). Physical state: white hygroscopic powder. The crude alcohol 5b was isolated from the filtrate according to GP. Yield: 75 mg (47%). Physical state: yellowish oil.
Analytical data for 3c·H2C2O4: Mp 126–128 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.51–7.49 (m, 2H), 7.38–7.35 (m, 2H), 7.29–7.26 (m, 1H), 3.09, 2.98 (q, AB, JAB = 12.9 Hz, 2H), 1.49 (s, 3H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.4, 145.6, 128.1, 126.9, 124.9, 70.9, 49.1, 27.2; IR (neat) /cm−1: 3438br, 1713w, 1620w, 1514m, 1447w, 1394w, 1211m, 1062m, 760s; HRMS (ESI-Q-TOF) m/z: calcd for C9H14NO [M + H]+ 152.1070, found 152.1066.
Analytical data for 5b:52 1H NMR (400 MHz, CDCl3) δ/ppm: 7.37–7.32 (m, 4H), 7.29–7.25 (m, 1H), 4.88 (q, J = 6.5 Hz, 1H), 2.02 (s, 1H), 1.48 (d, J = 6.5 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ/ppm: 145.8, 128.5, 127.4, 125.3, 70.4, 25.1; IR (neat) /cm−1: 3329br, 2970w, 1493w, 1452w, 1368w, 1077m, 899m, 761m, 700s; HRMS (ESI) not detected.
Analytical data for 3ca·H2C2O4: Mp 235–237 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.61–7.60 (m, 2H), 7.46–7.39 (m, 3H), 3.39, 3.37 (q, AB, JAB = 14.4 Hz, 2H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.4, 135.1, 128.8, 128.3, 126.8, 125.1 (q, 1JCF = 287.1 Hz), 74.7 (q, 2JCF = 27.0 Hz), 43.1; 19F NMR (376 MHz, DMSO-d6) δ/ppm: −77.15 (s); IR (neat) /cm−1: 2900w, 1594m, 1506m, 1453m, 1299m, 1197m, 1152s, 1046m, 989m, 763s; HRMS (ESI-Q-TOF) m/z: calcd for C9H11F3NO [M + H]+ 206.0787, found 206.0795.
Analytical data for 3cb·H2C2O4: Mp 138–139 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.46–7.44 (m, 2H), 7.38–7.34 (m, 2H), 7.28–7.25 (m, 1H), 3.12, 3.10 (q, AB, JAB = 13.0 Hz, 2H), 1.81–1.72 (m, 2H), 1.25–1.13 (m, 3H), 0.85–0.79 (m, 1H), 0.76 (t, J = 7.2 Hz, 3H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.3, 143.1, 128.0, 126.8, 125.5, 73.5, 48.5, 24.7, 22.2, 13.8; IR (neat) /cm−1: 2930w, 1587m, 1504m, 1448m, 1218s, 1049w, 764w; HRMS (ESI-Q-TOF) m/z: calcd for C12H20NO [M + H]+ 194.1539, found 194.1546.
Analytical data for 3d·H2C2O4: Mp 182–184 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.50–7.48 (m, 4H), 7.36–7.33 (m, 4H), 7.27–7.24 (m, 2H), 3.68 (s, 2H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.1, 144.3, 128.2, 127.2, 125.8, 74.9, 47.3; IR (neat) /cm−1: 3443w, 1738m, 1614m, 1417m, 1224s, 715s; HRMS (ESI-Q-TOF) m/z: calcd for C14H15NO [M + H]+ 214.1226, found 214.1225.
Analytical data for 5c:52 Mp 60–61 °C; 1H NMR (400 MHz, CDCl3) δ/ppm: 7.39–7.31 (m, 8H), 7.28–7.24 (m, 2H), 5.84 (s, 1H), 2.23 (s, 1H); 13C NMR (100 MHz, CDCl3) δ/ppm: 143.8, 128.5, 127.5, 126.5, 76.2; IR (neat) /cm−1: 3378w, 1494m, 1446m, 1269w, 1181w, 1017m, 752m, 734s; HRMS (ESI-Orbitrap) m/z: calcd for C13H13O [M + H]+ 185.0961, found 185.0959.
Analytical data for 3e·H2C2O4: Mp 195–196 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.56–7.39 (m, 8H), 3.08 (q, AB, JAB = 13.5 Hz, 2H), 1.50 (s, 3H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 163.6, 158.9 (d, 1JCF = 245.6 Hz), 147.4 (d, 3JCF = 7.7 Hz), 134.8, 130.4, 128.60, 128.56, 127.8, 126.7 (d, 2JCF = 13.5 Hz), 121.6, 113.2 (d, 2JCF = 24.1 Hz), 70.7, 48.8, 27.4; 19F NMR (376 MHz, DMSO-d6) δ/ppm: −118.12 to −118.17 (m); IR (neat) /cm−1: 3465w, 2930w, 1615m, 1501m, 1410m, 1223m, 1107w, 1067m, 1009m, 882w, 831w, 766m; HRMS (ESI-Q-TOF) m/z: calcd for C15H17FNO [M + H]+ 246.1289, found 246.1290.
Analytical data for 5d:53 Mp 65–66 °C; 1H NMR (400 MHz, CDCl3) δ/ppm: 7.58–7.55 (m, 2H), 7.48–7.37 (m, 4H), 7.23–7.20 (m, 2H), 4.96 (q, J = 6.5 Hz, 1H), 2.08 (br s, 1H), 1.54 (d, J = 6.5 Hz, 3H); 13C NMR (100 MHz, CDCl3) δ/ppm: 157.7 (d, 1JCF = 248.5 Hz), 147.4 (3JCF = 6.7 Hz), 135.6, 130.7 (3JCF = 3.9 Hz), 128.9 (3JCF = 2.9 Hz), 128.4, 127.9 (d, 2JCF = 13.5 Hz), 127.6, 121.2, 113.1 (d, 2JCF = 24.1 Hz), 69.6, 25.2; 19F NMR (376 MHz, CDCl3) δ/ppm: −117.56 to −117.61 (m); IR (neat) /cm−1: 3324br, 1581w, 1484m, 1416m, 1268m, 1153w, 1127w, 1070m, 1010m, 939w, 868m, 839m, 766s; HRMS (ESI-Orbitrap) m/z: calcd for C14H12F [M − H2O + H]+ 199.0918, found 199.0917.
Analytical data for 3f·H2C2O4: Mp 193–195 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.40–7.38 (m, 2H), 7.16–7.14 (m, 2H), 3.05, 2.97 (q, AB, JAB = 12.8 Hz, 2H), 2.44 (d, J = 7.1 Hz, 2H), 1.88–1.78 (m, 1H), 1.48 (s, 3H), 0.87 (d, J = 6.6 Hz, 6H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 163.3, 142.7, 139.8, 128.6, 124.7, 70.7, 49.1, 44.1, 29.5, 27.2, 22.1; IR (neat) /cm−1: 3467m, 2954m, 1732w, 1615m, 1504s, 1365w, 1184s, 1063s, 1019s, 810w, 794m; HRMS (ESI-Q-TOF) m/z: calcd for C13H22NO [M + H]+ 208.1696, found 208.1689.
Analytical data for 4f·H2C2O4 (the authentic sample was prepared by a DIBAH-mediated reduction of 1f): Mp 190–191 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.19–7.11 (m, 4H), 3.00–2.95 (m, 3H), 2.43 (d, J = 6.9 Hz, 2H), 1.86–1.79 (m, 1H), 1.25–1.19 (m, 3H), 0.87 (d, J = 6.4 Hz, 6H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.0, 140.0, 139.6, 129.1, 126.7, 45.0, 44.1, 37.0, 29.4, 22.1, 19.1; IR (neat) /cm−1: 2866m, 1622m, 1574s, 1532s, 1446m, 1300s, 1186w, 1019w, 840w, 799w, 777s; HRMS (ESI-Q-TOF) m/z: calcd for C13H22N [M + H]+ 192.1747, found 192.1737.
Analytical data for 5e:54 1H NMR (400 MHz, CDCl3) δ/ppm: 7.28–7.25 (m, 2H), 7.13–7.11 (m, 2H), 4.86 (q, J = 6.5 Hz, 1H), 2.46 (d, J = 7.1 Hz, 2H), 2.15 (br s, 1H), 1.90–1.80 (m, 1H), 1.48 (d, J = 6.5 Hz, 3H), 0.90 (d, J = 6.5 Hz, 6H); 13C NMR (100 MHz, CDCl3) δ/ppm: 143.0, 141.0, 129.2, 125.2, 70.2, 45.0, 30.2, 25.0, 22.3; IR (neat) /cm−1: 3344br, 2952s, 2923m, 2867m, 1650w, 1513w, 1465m, 1366m. 1077s, 1008m, 898m, 846m; HRMS (ESI-Orbitrap) m/z: calcd for C12H19O [M + H]+ 179.1430, found 179.1424.
Analytical data for 3g·H2C2O4: Mp 139–141 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 8.17 (d, J = 8.2 Hz, 1H), 7.99–7.96 (m, 1H), 7.89 (d, J = 8.2 Hz, 1H), 7.73 (d, J = 6.8 Hz, 1H), 7.61–7.53 (m, 3H), 5.60 (dd, J = 9.6, 2.3 Hz, 1H), 3.16 (dd, J = 12.9, 2.3 Hz, 1H), 2.92 (dd, J = 12.6, 9.6 Hz, 1H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 163.6, 137.5, 133.2, 129.7, 128.7, 128.0, 126.3, 125.6, 125.4, 123.4, 122.7, 66.1, 45.4; IR (neat) /cm−1: 2899m, 1715w, 1652s, 1580m, 1524m, 1218m, 1150m, 1059m, 772s, 722s; HRMS (ESI-Q-TOF) m/z: calcd for C12H14NO [M + H]+ 188.1070 m/z, found 188.1072.
Analytical data for 3h·H2C2O4: Mp 168–170 °C; 1H NMR (400 MHz, DMSO-d6) δ/ppm: 7.47 (d, J = 4.8 Hz, 1H), 7.06–7.01 (m, 2H), 5.05 (d, J = 8.1 Hz, 1H), 3.13–3.10 (m, 1H), 2.95–2.90 (m, 1H); 13C NMR (100 MHz, DMSO-d6) δ/ppm: 164.5, 145.9, 126.9, 125.0, 123.9, 65.6, 45.8; IR (neat) /cm−1: 2892br, 1582s, 1446w, 1296m, 1232m, 1021w, 924w, 867w, 784w, 717m; HRMS (ESI-Orbitrap) m/z: calcd for C6H10NOS [M + H]+ 144.0478, found 144.0478.
Footnotes |
† Electronic supplementary information (ESI) available: ESI contains optimization, mechanistic experiments, HPLC, NMR, GC/MS, and HRMS data for compounds 1–10. See DOI: 10.1039/d0ra04359a |
‡ These authors contributed equally. |
This journal is © The Royal Society of Chemistry 2020 |