Ru(II)Porphyrinate-based molecular nanoreactor for carbene insertion reactions and quantitative formation of rotaxanes by active-metal-template syntheses

Selectivity in N–H and S–H carbene insertion reactions promoted by Ru(II)porphyrinates currently requires slow addition of the diazo precursor and large excess of the primary amine and thiol substrates in the reaction medium. Such conditions are necessary to avoid the undesirable carbene coupling and/or multiple carbene insertions. Here, the authors demonstrate that the synergy between the steric shielding provided by a Ru(II)porphyrinate-based macrocycle with a relatively small central cavity and the kinetic stabilization of otherwise labile coordinative bonds, warranted by formation of the mechanical bond, enables single carbene insertions to occur with quantitative efficiency and perfect selectivity even in the presence of a large excess of the diazo precursor and stoichiometric amounts of the primary amine and thiol substrates. As the Ru(II)porphyrinate-based macrocycle bears a confining nanospace and alters the product distribution of the carbene insertion reactions when compared to that of its acyclic version, the former therefore functions as a nanoreactor.


SUPPORTING INFORMATION
-Carbonyl stretching frequency region of the FTIR-ATR spectrum of the porphyrinate product isolated after metalation reaction of macrocycle 7 with Ru3(CO)12.

Materials
All chemicals were purchased from Sigma-Aldrich and Labsynth and used without further purification unless otherwise noted. For moisture-sensitive reactions, solvents were freshly distilled. Dichloromethane (DCM), benzene and toluene (MePh) were dried over calcium hydride, whereas 1,4-dioxane and tetrahydrofuran (THF) were dried using the sodium/benzophenone system. Chloroform was dried over calcium chloride.
Anhydrous dimethylformamide (DMF) was used as received. Acetone, 1,2,4trichlorobenzene, dimethyl sulfoxide (DMSO), petroleum ether, methanol (MeOH) and ethanol (EtOH) were P.A. ACS grade reagents and were used as received. Technical grade ethyl acetate (EtOAc) and hexanes were used as received in extractions and as eluent in column chromatography purifications. Epoxidized soybean oil was prepared from commercially available refined soybean oil and performic acid [generated in situ from aqueous hydrogen peroxide (50% w/w) and formic acid (assay 85%)] as oxidizing agent. S1 The final product had an oxirane oxygen content of about 6% (w/w) as determined according to the American Oil Chemical Society Official Method Cd-9-57.
All syntheses were carried out using Schlenk line techniques. Moisture-sensitive liquids were transferred by cannula or syringe. The progress of the reactions was monitored by thin-layer chromatography (TLC) whenever possible. TLC was performed using precoated aluminum plates (SiliaPlate from Silicycle with 200 m thickness) containing a 254 nm fluorescent indicator. Preparative TLC was carried out on glass plates coated with silica gel 60 PF254 containing gypsum (CaSO4•0.5 H2O, 28-32%, w/w) with particle size 90%  55 m from Merck EMD Millipore Corporation. Column chromatography was carried out using Fluka Silica Gel 60 (230-400 mesh particle size). Melting points were determined using a PF1500 Farma Capillary Melting Point apparatus from Gehaka and were uncorrected. Diphenyldiazomethane was prepared according to literature methods. S2 S4 quartet and quintet; m, multiplet; br, broad), the values of the J-coupling constants in Hz (if applicable), the number of protons implied and finally the assignment. Residual solvent peaks and eventual aliphatic impurities were assigned according to literature. S3 Two-dimensional NOESY NMR spectra were acquired on a Bruker AVANCE 400 (400 MHz) using CDCl3 as deuterated solvent at 298 K and 400 ms mixing time.

Mass Spectrometry
Low resolution mass spectra (LRMS) were afforded from an Agilent GC-MS piece of equipment, model 5975C, equipped with a monoquadrupole detector in positive mode and electron impact (EI) source. High-resolution mass spectra (HRMS) were recorded on a Q-TOF (ESI-QTOF) equipment operating in positive mode. Low-resolution MALDI-TOF mass spectra were recorded in a Bruker Daltonics Microflex LT MALDI-TOF MS, while the high-resolution spectra were acquired in a Bruker Daltonics Autoflex III Smartbean equipped with LIFT TOF/TOF technology spectrometers. The mass spectra represent an average over 512 consecutive laser shots in linear mode using (1E,3E)-1,4diphenylbuta-1,3-diene (DPB) as matrix (unless otherwise noted), which was purchased from Aldrich. The mass scale was calibrated using the peptide calibration standard purchase from Bruker and -cyano-4-hydroxycinnamic acid (HCCA) as matrix. Data processing was carried out using the software package Compass for Flex series available from Bruker. Mentioned m/z values correspond to monoisotopic masses.

Fourier-Transform Infrared Spectroscopy -FTIR
FTIR spectra were obtained from an Agilent Cary 630 spectrometer equipped with Attenuated Total Reflection (ATR) accessory and 4 cm -1 resolution. All samples were analyzed in the solid state.

Steady-State Ultraviolet-Visible Absorption Spectroscopy -UV-Vis
UV-Vis spectra were obtained from an Agilent Cary 50 spectrometer with 1.0 nm resolution. All samples were analyzed in dichloromethane solutions (10 -5 mol/L) at room temperature in quartz cuvettes with 1 cm width. Figure S1 -Estimated cavity size afforded from the crystal structure of macrocycle 7. Using Mercury software, a centroid at the C26-C26 i bond on the phenanthrene moiety was calculated, which along with the centroid calculated from the 24 porphyrin atoms and the two centroids calculated from the 6 carbon atoms of the two phenyl spacers allowed the estimation of the cavity size in 7. Carbon atoms are shown in grey, nitrogen in blue and oxygen in red. Hydrogen atoms are omitted for clarity purposes. Ellipsoids are drawn at 50% probability levels. Symmetry code: i = 1/2-x, y, z. All resonances in the spectrum of macrocycle 7 are sharp and distinct and confirm the proposed hollow structure in solution. Furthermore, comparison between 1 H NMR spectra of 6 and 7 reveals that the phenanthrene protons facing the cavity (labelled as HL) are unusually shielded. That significant shielding (ΔL = 2.03 ppm) informs that the HL protons are at reach of the strong magnetic anisotropy of the porphyrin aromatic system in 7. In fact, all protons on the aromatic loop in 7 are affected by the porphyrin ring current effects (dotted lines). The shielding of the nuclei on the aromatic loop confirms the welldefined and relatively small central cavity in 7. Therefore, the aromatic loop provides the internal axial position of the resulting porphyrinate with effective steric shielding against upcoming bulky ligands in solution. Figure S3 -1 H NMR spectrum of the porphyrinate product isolated after metalation reaction of macrocycle 7 with Ru3(CO)12 (500 MHz, CDCl3, 298 K). To improve the solubility of the complex in the CDCl3 solvent, 0.5% (v/v) of regular methanol was added to the NMR sample, which appears as large singlets at  = 3.39 (CH3OH) and  = 1.24 (CH3OH). Residual solvent and aliphatic impurities: chloroform ( = 7.26 ppm), dichloromethane ( = 5.29 ppm), residual water in CDCl3/CH3OH solvent mixture ( = 1.77 ppm). After the metalation reaction, only one porphyrin product is isolated from the crude, despite all our efforts to identify the expected isomers 8 and 9 by TLC. The spectrum of the isolated porphyrin product is not consistent with a mixture of isomeric complexes as no duplication of signals are observed. As described in the manuscript, it turned out that the isolated product was complex 8. Figure S4 -13 C NMR spectrum of the porphyrinate product isolated after metalation reaction of macrocycle 7 with Ru3(CO)12 (125 MHz, CDCl3, 298 K). To improve the solubility of the complex in the CDCl3 solvent, 0.5% (v/v) of regular methanol was added to the NMR sample, which appears at  = 50.53 (CH3OH). Inset: resonance of the CO axial ligand. It turned out that this product was complex 8. to the porphyrin magnetic anisotropy, informs that those groups are held close to the Ru(II)porphyrinate core in 12, and confirms the axial coordination of the amino group to the Ru(II) ion. However, the Ru-NH2 resonance is not observed in the 1 H NMR spectrum of 12 in CDCl3. This lack of the Ru-NH2 signal is due to rapid hydrogendeuterium exchange reactions between the amino group and CDCl3 solvent molecules. Accordingly, the expected broad and strongly shielded Ru-NH2 signal is clearly observed in the spectrum of 12 in C6D6 at 298 K as shown below in Figure   S8.1.    i) benzene, rt, 4 h, quantitative relative to 11.

Synthesis of macrocycle 7
In a two-necked round bottom flask of 2 L capacity, compound 6 (0.115 g, 0.213 mmol, 1 equiv) and dipyrromethane S8 (0.063 g, 0.43 mmol, 2 equiv) were dissolved in 1 L of an oxygen-free solvent mixture composed of 750 mL of dichloromethane and 250 mL of chloroform at room temperature and inert atmosphere. TFA (0.745 g, 6.53 mmol, 30 equiv, 0.50 mL) was added and the red reaction mixture was stirred at room temperature for 20 h. DDQ (0.420 g, 1.85 mmol, 8.5 equiv) was added as a solid and the resulting black mixture was heated at reflux for 2 h. After cooling, the crude was neutralized with Et3N (5 mL). The solvent was removed under reduced S36 pressure and the resulting black paste was directly loaded into a column chromatography (SiO2) with dichloromethane. Flash chromatography using DCM as eluent removed black tar material. The crude product was concentrated under reduced pressure. Washing the crude with diethyl ether to remove impurities followed by filtration through paper yielded target macrocycle 7 as a purple solid in 20% (0.034 g).    Synthesis of 2-(3,5-di-tert-butylphenoxy)ethan-1-ol S2.

Crystal Data
The integration and corrections were performed by CCP4, S10 XIA2 0.5.653-g9f819c0c-dials-1.11, S11 DIALS 1.11.2-g01fb9e997-release. S12 Data merging and scale were performed using Aimless S13 and Pointless. S14 The number of reflections measured for the crystal of Olex2, S15 the structures were solved with the ShelXT S16 structure solution program using Intrinsic Phasing and refined with the XL S17 refinement package using Least Squares minimization. The position of all non-hydrogen atoms was refined anisotropically. The

S97
hydrogen atoms in the compounds were added to the structures in idealized positions and further refined according to the riding model. Uiso(H) = 1.2Ueq(C) for aromatic and N groups. Uiso(H) = 1.5Ueq(C) for methyl groups and O groups. Tables S1, S2 and S3 present crystal data, data collection and refinement data for macrocycle 7, respectively.
For macrocycle 11, tables S4, S5 and S6 gather the crystal data, data collection and refinement data, respectively.
Table S1 -Sample and crystal data for macrocycle 7.     Figure S100 -Estimated cavity size of macrocycle 7 afforded from the crystal structure.
Using Mercury software, a centroid at the C26-C26 i bond on the phenanthrene moiety was calculated, which along with the centroid calculated from the 24 porphyrin atoms and the two centroids of the phenyl spacers allowed the estimation of the cavity size in 7.
Carbon atoms are shown in grey, nitrogen in blue and oxygen in red. Hydrogen atoms are omitted for clarity purposes. Ellipsoids are drawn at 50% probability levels. Symmetry code: i = 1/2-x, y, z.  conformation with open-orange circles and closed-blue circles representing atoms lying above and below the mean plane, respectively. Although the cylindrical projection of the porphyrin group in 7 reveals a distortion pattern of a ruffled conformation, the very low deviation of the z-coordinate displacements for each atom (h, in Å) from the porphyrin ring mean plane (confirmed by the low root-mean-square out-of-plane value of 0.124 Å) informs that the porphyrin core is flat in 7. S18 Figure S103 -Estimated cavity size of macrocycle 11 afforded from the crystal structure.
Using Mercury software, a centroid at the C37-C42 bond on the phenanthrene moiety was calculated, which along with the centroid calculated from the 24 porphyrin atoms and the two centroids of the phenyl spacers allowed the estimation of the cavity size in 11. Carbon atoms are shown in grey, nitrogen in blue, oxygen in red, ruthenium in turquoise, chlorine in green. Hydrogen atoms are omitted for clarity purposes. Ellipsoids are drawn at 50% probability levels.