Remote stereoselective deconjugation of α,β-unsaturated esters by simple amidation reactions

The amidation of macrocyclic conjugated esters affords in one-pot single (chiral) β,γ-unsaturated diastereomers via effective remote stereocontrol.


General Remarks
NMR spectra were recorded on 300, 400 or 500 MHz spectrometer at 20 °C unless otherwise stated. 1 H-NMR chemical shifts are given in ppm relative to Me4Si with the solvent resonance used as the internal standard (CDCl3 δ = 7.26 ppm). 13 C-NMR (125 or 101 MHz) chemical shifts were given in ppm relative to Me4Si with the solvent resonance used as the internal standard (CDCl3 = 77.16 ppm). IR spectra were recorded using an ATR sampler and are reported in wave numbers (cm -1 ). Melting points (Mp) were measured in open capillary tubes and were uncorrected. Optical rotations were measured in a thermostated (20 °C) 10.0 cm long microcell at 589 nm (Na). Electrospray mass spectra (ESI +) were obtained by the department of Mass Spectrometry of the University of Geneva. All reactions involving air sensitive compounds were carried out under dry N2 or argon by means of an inert gas/vacuum double manifold line and standard Schlenk techniques. Flash column chromatography was performed with silica gel 40-63 or alumina (neutral Brockmann I, 50-200 μm).

2.a. General synthesis of macrocycles
To the suspension of starting macrocycle 1 (0.1 mmol) and aniline (0.4 mmol) in 1.0 mL of dry THF was added tBuOK (44.8 mg, 0.4 mmol) at -100 °C. After 2 minutes, the cooling bath (EtOH, N2 liquid) was removed and the reaction was allowed to reach room temperature on its own. It was stirred for additional 2 hours. Without further treatment or work-up, the reaction mixture was purified by column chromatography (SiO2) or preparative TLC.

2.b. General synthesis of perimidine derivatives
To the suspension of starting macrocycle (2o, 5o or 7o) (0.1 mmol), 1,8-diaminonaphthalene (0.22 mmol) and t-BuOH (0.1 mmol) in 1.0 mL of dry THF was added tBuOK (44.8 mg, 0.44 mmol) at -100 °C. After 2 minutes, the cooling bath (EtOH, N2 liquid) was removed and the reaction was allowed to reach room temperature on its own. It was stirred for additional 2 hours. Without further treatment or work-up, the reaction mixture was purified by column chromatography (SiO2) or preparative TLC.

Computational Details.
Geometry optimizations have been performed with the Gaussian 09 (1) package at the B3PW91 (2) level of hybrid density functional theory. The potassium atom was represented by the relativistic effective core potential (RECP) from the Stuttgart group and the associated basis sets (3) augmented by a d polarization function (4). The remaining atoms (H, C, O, N) were represented by a 6-31G** basis set (5). All energies reported in the present work are Zero-point energies corrected. Demodulation was performed by a lock-in amplifier (SR830 DSP). An optical low-pass filter (< 1800 cm-1) in front of the photoelastic modulator was used to enhance the signal/noise ratio. Spectra were recorded with a transmission cell equipped with CaF2 windows and a 0.2 mm Teflon spacer. Solutions of (-)-4a and (+)-4a in CD2Cl2 at concentrations of 7 mg in 700 l CD2Cl2 were measured under identical conditions and subtracted to each other in order to eliminate artifacts. Samples were measured at a resolution of 4 cm -1 by averaging about 24'000 scans for both enantiomers. Spectra are presented without further data processing.

IR and VCD calculations
Calculations were performed for (R,R)-4a. The geometry optimizations, vibrational frequencies, IR absorption and VCD intensities were calculated with Density Functional Theory (DFT) using the B3PW91 functional and a 6-31+G(d,p) basis set. Frequencies were scaled by a factor of 0.98. IR absorption and VCD spectra were constructed from calculated dipole and rotational strengths assuming Lorentzian band shape with a half-width at half maximum of 4 cm -1 . The crystal structure served as the starting point for the geometry optimization, which was done for the molecule and its complex with water. All calculations were performed using Gaussian09. 1 Figure S5: Optimized structures of (R,R)-4a·H2O (left) and (R,R)-4a (right).
The IR spectrum clearly shows the coexistence of 4a and its complex with water (4a·H2O) in solution.
This is evidenced by the double band between 1550 cm -1 and 1580 cm -1 due to the N-H deformation S-68 vibration of the amide bond, which shifts to higher wavenumbers upon complex formation (see IR Figure). Figure S6: IR spectra of 4a. Bottom: experimental spectrum (7 mg in 700 ml CD2Cl2), top: calculated spectra for 4a and its complex with water. Dashed lines indicate the N-H vibrations that are sensitive to complex formation.

Crystallographic data
All data were collected on an Agilent Supernova diffractometer equipped with an ATLAS CCD detector using Cu radiation. Integration and data reduction were carried out in the crysalis Software 1 [ref1]. Refinement was made using SHELXL 2 . Material for publication was prepared using the Olex2 3 software.

Figure S5
Comments on the X-ray structure: One of the CF3 group is disordered and was refined using two components, the one with the smaller occupancy was refined isotropically. Geometrical restrains (1-2 and 1-3 bond lengths) were applied as well as restraints on the anisotropic displacement parameters. Comments on the X-ray structure: Part of the macrocycle is disordered and was refined using two components. Geometrical restraints were applied on 1-2 and 1-3 distances as well as restraints on anisotropic displacement parameters.
A disordered toluene molecule is also present. It was modelled using three rigid bodies. Restraints were applied on anisotropic displacement parameters.
The water molecule was refined as a rigid-body.

Figure S7
Comments on the X-ray structure: Three disordered CF3 groups were refined as two components each. Geometrical restraints on C-F and F-F distances were applied, as well as restraints on anisotropic displacement parameters. There is still a small peak present close to the dichloromethane molecule that may be due to a small disorder in the solvent model. Inclusion of this disorder in the model did not improve it significantly, so that this disorder was not included in the final model.