Site-Selective Double and Tetracyclization Routes to Fused Polyheterocyclic Structures by Pd-Catalyzed Carbonylation Reactions

In this contribution, we report novel palladium-catalyzed carbonylative cascade approaches to highly functionalized polyheterocyclic structures. The Pd-catalyzed carbonylative process involves the regioselective insertion of one to three CO molecules and the sequential ordered formation of up to eight new bonds (one C–O, two C–C, five C–N). The exclusive formation of six-membered heterocycles is elucidated by detailed modeling studies.


General Methods
All reagents were used as received from commercial sources without further purification. All solvents were dried over activated molecular sieves. All reactions were carried out in stainless steel autoclaves and analyzed by TLC (Thin Layer Chromatography) on silica gel 60 F254. Flash column chromatography was performed on silica gel 60 (70-230 mesh). Melting points were measured with an Electrothermal apparatus and are uncorrected.
GC analyses were performed with an Agilent Tenchnologies 7820A equipped with a FID detector and a 30 m capillary column. GC-MS analyses (m/z, relative intensity %) were performed with an Agilent Technologies 7820A gas chromatograph coupled to an 5977B mass selective detector (Agilent Technologies) working at 70 eV ionizing voltage. Exact masses were recorded on a LTQ ORBITRAP XL Thermo Mass Spectrometer (ESI source). IR spectra were run on a Nicolet FT-IR 5700 spectrophotometer paired with a Diamond Smart Orbit accessory. Unless otherwise indicated NMR spectra were recorded on Bruker AVANCE 300 and 400 spectrometers in deuterated chloroform, using the solvent residual signals as internal reference (7.26 and 77.00 ppm, respectively for 1 H and 13 C). Chemical shifts (δ) and coupling constants (J) are given in ppm and in Hz, respectively. The following abbreviations were used to explain the multiplicities: s=singlet, d= doublet, t=triplet, q=quartet, hept=heptet, m=multiplet, dd=double doublets, b=broad. (Table S1)   Table S1. [

General Procedures
General Procedure A for the alkylation of 2-haloanilines [2] To a round-bottom flask the aniline derivative (1 equiv), the benzaldehyde derivative (1 equiv) and MeOH (2 mL/mmol of aniline) were added. The resulting mixture was stirred at rt overnight. After removal of solvent under vacuum, the residue was dissolved in AcOH (2 mL/mmol of aniline), then NaBH4 (1.2 equiv) was added in portions at 0 °C. After stirring at rt for 1 h, the solvent was evaporated and the residue was dissolved in EtOAc.
A solution of NaOH (1N) was added to the mixture until pH 8−9. The two phases were stirred vigorously for 1 h then separated. The aqueous layer was extracted with EtOAc twice and the combined organic extracts were washed with brine, dried over anhydrous Na2SO4 and concentrated under vacuum to give the crude product, which was used for the next step without purification.

General Procedure B for the synthesis of substrates 1 (Sonogashira coupling)
To a solution of 2-iodoaniline (1 equiv) in triethylamine (4 mL/mmol), PdCl2(PPh3)2 (2 mol%), CuI (6 mol%) and propargylic amine/amide (1.2 equiv) were added. A proper amount of CH2Cl2 (1-2 mL/mmol) was added in order to obtain a homogeneous solution. The mixture was stirred at rt for 4-24 h. After filtration and evaporation of the solvent, the residue was diluted with EtOAc (40 mL) and washed with water (40 mL) and brine (30 mL). The organic layer was dried over anhydrous Na2SO4, filtered and the solvent was removed under reduced pressure. S12

Computational details
The calculations have been performed using the Gaussian09 program package. [3] We have applied the global hybrid, M06 functional containing exact exchange which has been shown to have good performance for Pdcatalyzed reactions. [4] A smaller basis set (B1) has been used for the geometry optimizations, frequency calculations and the calculations of the Gibbs free energy contributions at the experimental 353 K temperature and for the standard 1 M concentration. For the routes leading to formation of 2 the 6-31+G* basis set was used. For the exploration of the routes leading to 3 the 6-31G* basis set was employed. For CO we have taken into account the experimental 1.2 MPa. The larger 6-311++G(3df,3pd) basis set (B2) and the SMD implicit solvation model [5] has been applied to recalculate the optimized structures to obtain the solvent corrected The reactants and intermediates in several cases have large conformational freedom. To find the most stable conformers which are necessary to set the initial free energy levels of the profiles we have performed molecular mechanics simulations with the OPLS force field using the Macromodel software. [6] The conformational searches were performed with a mixed Monte Carlo torsional sampling and a low-mode sampling which together can efficiently explore the conformational space of the molecule in question. We obtained a few hundreds of conformers within a 10 kcal/mol window which then were the subject of clustering. From the located clusters the most stable structures were selected and DFT calculations were performed to find the lowest conformers.

Effect of increased pressure
It follows from the ideal gas model employed in our calculations that the effect of a pressure change can be calculated by the following equation: G = RT ln(p2/p1) where p2 and p1 are the new and the original pressures, respectively, T is the actual absolute temperature, R is the universal gas constant, whereas G is the change in the free energy. It can be seen that a 10-fold increase in the pressure yields a +1.6 kcal/mol destabilization in the reactant state (the TS is not affected because its concentration (pressure) is set by the reactant present in smaller amount), ie. the barrier becomes smaller implying a faster reaction.   As it is noted in the article we have explored different routes and selected the most favorable ones on the basis of the computed Gibbs free energy profiles. We found that reaction paths leading to indolization or quinoline framework require high activation free energies. In particular, we obtained 53.3 kcal/mol S87 barrier for the formation of indoline-2-one frame (from 1a'-2 toward 1a'-3') and 61.1 kcal/mol for the formation of quinoline-2-one scaffold (from 1a'-2 toward 1a'-3"). Therefore, we have excluded these paths. When exploring the initial amid-attack we found that an initial 5-exo-dig attack by the carbonyl group requires 29.8 kcal/mol activation free energy, much higher than that of the 6-endo-dig attack (16.7 kcal/mol). Finally, we have obtained that initial enolization is not necessary as the enol attacks on the activated triple bond leading to either 6-endo-dig or 5-exo-dig cyclization feature very high activation barriers (39.3 and 47.2 kcal/mol, respectively).

8.
Single Crystal X-ray Diffraction for 2g, 2o, 2s and 3d SC-XRD analyses were performed on single crystal samples of 2o, 2g, 2s and 3d on a Bruker D8 Venture diffractometer equipped with a kappa goniometer. Data collection were performed using microfocused MoKα radiation (λ = 0.71073 Å) under nitrogen flux with Oxford Cryosteam. Lorentz polarization and absorption correction were applied. Data were reprocessed using APEX v3 software. Structures were solved by direct methods using SHELXT [7] and refined by full-matrix least-squares on all F 2 using SHELXL [8] implemented in Olex2.21 [9] . For all samples anisotropic displacement parameters were refined except for hydrogen atoms. Samples 2o, 2g and 2s crystallizes in monoclinic system as anhydrous compound while 3d crystallizes as CHCl3 solvate in triclinic systems.
The X-ray crystallographic coordinates for structures reported in this paper have been deposited at the Cambridge Crystallographic Data Centre, under deposition numbers CCDC 1943405-1943408. These data can be obtained free of charge from the Cambridge Crystallographic Data Centre (http://www.ccdc.cam.ac.uk/data_request/cif).

ACKNOWLEDGMENT
Chiesi Farmaceutici SpA is acknowledged for the support for the D8 Venture X-ray equipment  Figure S7. Ellipsoid plot of 2g. All non-hydrogen atoms shown as ellipsoids at the 50% probability level. H atoms (isotropically refined) are reported in ball-and-stick style for the sake of clarity. Figure S8. Ellipsoid plot of 2o. All non-hydrogen atoms shown as ellipsoids at the 50% probability level. H atoms (isotropically refined) are reported in ball-and-stick style for the sake of clarity. Figure S9. Ellipsoid plot of 2s. All non-hydrogen atoms shown as ellipsoids at the 50% probability level. H atoms (isotropically refined) are reported in ball-and-stick style for the sake of clarity. Figure S10. Ellipsoid plot of 3d. All non-hydrogen atoms shown as ellipsoids at the 50% probability level. H atoms (isotropically refined) are reported in ball-and-stick style for the sake of clarity.