Silver-Free Gold-Catalyzed Heterocyclizations through Intermolecular H-Bonding Activation

Modulable monosulfonyl squaramides have been shown to exert activation of gold(I) chloride complexes through H-bonding in an intermolecular way. Combinations of (PPh3)AuCl or IPrAuCl complexes and an optimal sulfonyl squaramide cocatalyst bearing two 3,5-bis(trifluoromethyl)phenyl groups efficiently catalyzed diverse heterocyclizations and a cyclopropanation reaction, avoiding in all cases undesired side reactions. Computational studies indicate that the Au–Cl bond breaks by transligation to the triple bond in a ternary complex formed by the actual AuCl···HBD catalyst and the substrate.


Synthesis of organocatalyst IV
1-Adamanthyl amine (135 mg, 0.9 mmol) was added to a suspension of precursor P2 (121 mg, 0.3 mmol) in CH2Cl2 (4.5 mL, 0.07 M) at room temperature. The resulting mixture was stirred at the same temperature overnight. Then, H2O was added (4.5 mL), the mixture was acidified with a saturated solution of NaHSO4 until pH 1-2, and extracted with CH2Cl2 (2 x 10 mL). The combined organic layers were washed with H2O (1 x 15 mL), dried over MgSO4 and concentrated under reduced pressure. The resulting residue was purified by column chromatography on silica gel (EtOAc/acetone 4/1) to afford IV as an off-white solid (127 mg, 0.24 mmol, 81%). 1 H NMR (500 MHz, DMSO-d6):  [3,phenyl]methanamine (205 mg, 0.8 mmol) was added to a suspension of 13 [13] (271 mg, 0.8 mmol) in MeOH (4 mL, 0.2 M) at room temperature. The resulting mixture was stirred at the same temperature overnight. Then, the precipitated formed was filtered and washed with a cold mixture of pentane/Et2O (1/1, 2 x 5 mL) to afford VII as a white solid (351 mg, 0.64 mmol, In a flame-dried schlenk tube, the corresponding freshly distilled anhydrous solvent (2.0 mL, 0.1 M) was added to a mixture of 1 (32 mg, 0.2 mmol), the corresponding [LAuCl] complex (0.01 mmol, 5 mol%), and the corresponding activator (0.01 mmol, 5 mol%) at room temperature. An additive was added if it is specified. The resulting solution was stirred for 16 h at the specified temperature (sand bath). Then, the reaction was quenched with Et3N (1.4 μL, 0.01 mmol) and the solvent was eliminated under reduced pressure. NMR yield was determined by 1 H NMR (CDCl3) using 1,3,5trimethoxybenzene as internal standard (see Table S1).
In a flame-dried schlenk tube, the corresponding freshly distilled anhydrous solvent (M) was added to a mixture of 3 (43 mg, 0.2 mmol), the corresponding [LAuCl] complex (0.01 mmol, 5 mol%), and the corresponding activator (0.01 mmol, 5 mol%) at room temperature. The resulting solution was stirred for 24 h at the specified temperature using sand bath. Then, the reaction was quenched with Et3N (1.4 μL, 0.01 mmol) and the solvent was eliminated under reduced pressure. NMR yield was determined by 1 H NMR (CDCl3) using 1,3,5-trimethoxybenzene as internal standard (see Table S2).

S10
Spectroscopic data of product 4 are in accordance with those reported in literature. [15] The following diagnostic signals were integrated to quantify the amount of starting material and product: for substrate 3, peak at 3.00 ppm (dt, J = 13.8, 6.2 Hz, 1H); for product 4, peak at 6.00 ppm (s, 1H); for 1,3,5-trimethoxybenzene, peak at 6.09 ppm (s, 3H). are in accordance with those reported in literature. [15] 8. General procedure for tandem cycloisomerization and nucleophilic addition H), 1.00 (q, J = 6.5 Hz, 1 H). The experimental data are in accordance with those reported in literature. [17] S13

Computational methods
All of the calculations were performed using the Gaussian16 program. [18] Computations were done using wb97xd functional [19] in conjunction with standard basis sets def2SVP and def2TZVP. [20] Geometry full optimizations were made at wb97xd/def2SVP level. Single point calculations using def2TZVP basis set were carried out over optimized geometries to obtain the energy values. Solvent effects (toluene) were considered using the SMD model. [21] The nature of stationary points was defined on the basis of calculations of normal vibrational frequencies (force constant Hessian matrix).
The optimizations were carried out using the Berny analytical gradient optimization method. [22] Minimum energy pathways for the reactions studied were found by gradient descent of transition states in the forward and backward direction of the transition vector (IRC analysis). [23] Analytical second derivatives of the energy were calculated to classify the nature of every stationary point, to determine the harmonic vibrational frequencies, and to provide zero-point vibrational energy corrections. The thermal and entropic contributions to the free energies were also obtained from the vibrational frequency calculations, using the unscaled frequencies. Correction to free energy was made by substracting Strans contribution and considering a 1 M concentration. [24] Structural representations were generated using CYLView. [25]

ELF analysis
The electronic structures of stationary points were analyzed by the topological analysis of the gradient field of electron localization function (ELF) [26] developed by Silvi and Savin. [27] The ELF study was performed with the TopMod program [28] using the corresponding wavefunctions of the all structures of the IRC. The topological analysis of the gradient field of ELF has showed to be a powerful tool for the study of the bonding changes along an organic reaction. [29] NCI Calculations NCI (non-covalent interactions) were computed using the methodology previously described. [30] Quantitative data were obtained with the NCIPLOT4 program. [31] A density cutoff of ρ=0.5 a.u. was applied and isosurfaces of s(r) =0.5 were colored by sign(λ2)ρ in the [-0.03,0.03] a.u. range using VMD software. [32] s(r) against sign(λ2)ρ(r) plots were generated with gnuplot software. [33] S14

Molecular Dynamics
MD simulations were carried out with AMBER20 suite of programs. [34] Parameters for trisaccharides were generated with the antechamber module using the general Amber force field (GAFF2), [35] with partial charges calculated using AM1-BCC method. The system to be studied was neutralized if necessary, and immersed in a chloroform box of 12 Å. A two-stage geometry optimization approach was carried out: (i) minimization of only the positions of solvent molecules executed by 500 cycles of steepest descent minimization followed by 500 cycles of conjugate gradient minimization and (ii) unrestrained minimization of all the atoms in the simulation cell executed by 2500 cycles of steepest descent minimization followed by 2500 cycles of conjugate gradient minimization. After system optimization, running of MD simulations was started on the systems by gradually heating each system in the NVT ensemble from 0 to 300 K for 100 ps using a Langevin thermostat with a coupling coefficient of 1.0/ps. Harmonic restraints of 10 kcal·mol-1 were applied to the solute, and the Langevin temperature coupling scheme [36] was used to control and equalize the temperature. The time step was kept at 2 fs during the heating stages, allowing potential inhomogeneities to self-adjust.
Water molecules are treated with the SHAKE algorithm such that the angle between the hydrogen atoms is kept fixed. Long-range electrostatic effects are modelled using the particle-mesh-Ewald method. [37] Then 5 ns of density equilibration with a force constant of 2.0 kcal/mol·Å 2 was performed by releasing all the restraints. Finally, production trajectories were then run for 100 ns under the same simulation conditions with an integration time step of 0.5 fs, recording geometry every 0.05 ps and with snapshots written each 2 ps, producing 50,000 frames per simulation. All MD simulations were replicated three times to ensure feasibility. In this reaction, a simultaneous migration of a bond seems to be necessary to take place when the cyclization is produced. However, any attempt of locating a single transition structure in which both cyclization and migration takes place at the same time, failed. On the other hand, the reaction proceeds in a similar way to the formation of oxazolines, that is forming a reactive intermediate (IN1b) that cyclizes (through TS2b). A subsequent 1,2-migration through a third transition structure TS3b would yield the product. The energy barrier for the last migration step (18.7 kcal/mol) is the rate limiting stage of the process. Consequently, we have, in this case, a two-step reaction, being the second step the migration to form the spiro derivative. Finally, the catalytic cycle would continue as usual with a deprotonation and protodeauration, yielding the spirocycle 4 (Scheme S4). Similarly, to the model reaction showed in Scheme S3, the Au-Cl bond is not completely broken in the first step and only S18 after the second transition structure can be considered that the Au atom is forming a bond with the triple bond (now becoming a double bond).

Topological analyses
The electron localization function (ELF) was introduced by Becke and Edgecombe