Mechanochemical Solvent‐Free Catalytic C−H Methylation

Abstract The mechanochemical, solvent‐free, highly regioselective, rhodium‐catalyzed C−H methylation of (hetero)arenes is reported. The reaction shows excellent functional‐group compatibility and is demonstrated to work for the late‐stage C−H methylation of biologically active compounds. The method requires no external heating and benefits from considerably shorter reaction times than previous solution‐based C−H methylation protocols. Additionally, the mechanochemical approach is shown to enable the efficient synthesis of organometallic complexes that are difficult to generate conventionally.


General information
Mechanochemical reactions were carried out using an InSolido Technologies IST636 mixer mill. Reaction vessels and milling medium used for mechanochemical reactions were purchased either from InSolido Technologies (Teflon TM vessels, 14 mL) or FormTech Scientific (stainless steel (SS) vessels, 14 mL). The average masses of SS balls used in the milling reactions were: • 10 mm diameter: mean weight of 16 balls: 3.94 g (range between 3.59-4.02 g).
1,3,5-Trimethoxybenzene or naphthalene internal standard was prepared as a stock solution (0.1 M in EtOAc). Starting materials were bought commercially and used without further purification or synthesized according reported literature procedures.
1,2-Dichloroethane (DCE), 1,4-dioxane and toluene were dried over 4 Å molecular sieves and stored under argon prior to use. EtOH and H2O were degassed with argon for half an hour before use. CDCl3 was stored over K2CO3 to neutralize residual acid. The other solvents were obtained commercially and used "as is" without further purification.
High-resolution electrospray ionisation mass spectrometry was performed on a micrOTOF II Focus instrument (Bruker Daltonics, Coventry, UK). High-resolution nanospray ionisation was performed on a Synapt G2S instrument (Waters, Manchester, UK) using a Triversa chip based nanospray source (Advion Biosciences, Norwich, UK). High-resolution electron ionisation mass spectrometry was performed at 70eV on a QExactive GC Orbitrap instrument (Thermo Scientific, Hemel Hempstead, UK). APCI was recorded on an Advion Expression L CMS with positive fragmentation. S3

Mechanochemical C-H methylation protocols:
General procedure 1: A stainless steel milling vessel (14 mL internal volume) was charged with substrate (0.3 mmol), MeB(OH)2 (1.2-4.0 equiv.), [Cp*RhCl2]2 (0.009 g, 0.015 mmol, 5 mol%), Ag2CO3 (0.124 g, 0.045 mmol, 1.5 equiv.) and one stainless steel ball (10 mm diameter). The vessel was mounted into the holding station of a mixer mill and milling was conducted at the indicated frequency for the indicated time. The vessel was then cooled to rt and the crude material was washed out with EtOAc or MeOH, filtered through a thin layer of Celite (3 cm), eluting with EtOAc (200 mL) or an EtOAc/MeOH mixture (1:1, 200 mL). 1,3,5-Trimethoxybenzene or naphthalene was added as a stock solution (0.1 M in EtOAc) for use as an NMR standard and the filtrate was concentrated under reduced pressure. The residue was further purified by silica-gel column chromatography to give the product.
General procedure 2: A Teflon TM milling vessel (14 mL internal volume) was charged with substrate (0.3 mmol), MeBF3K (1.5-6.0 equiv.), [Cp*RhCl2]2 (5 or 10 mol%), AgSbF6 (20 or 40 mol%), Ag2CO3 (1.5-3.0 equiv.) and a stainless steel ball (15 mm diameter). The vessel was mounted into the holding station of a mixer mill and milling was started at the indicated frequency (25 or 36 Hz) and continued for the indicated time. The crude reaction mixture was washed out with EtOAc or MeOH, then filtered through a thin layer of Celite (3 cm), eluting with EtOAc (200 mL) or an EtOAc/MeOH mixture (1:1, 200 mL). For indicated cases, 1,3,5trimethoxybenzene was added as a stock solution (0.1 M in EtOAc) for use as an NMR standard, the filtrate was concentrated under reduced pressure and analysed by 1 H NMR. The residue was further purified by silica-gel column chromatography to give the product.

A note about milling reaction temperature measurements:
The temperature of the interior wall of the milling vessels as well as that of the stainless steel balls were measured using a laser thermometer immediately after the cessation of several reactions. Representative values for Teflon TM and stainless steel balls are:  [22] According to general procedure 1, the reaction was stopped after 2 h. The mono/di-methylated product ratio was 44:1 according to crude 1 H NMR spectroscopy and after purification, the title compound was obtained as a colorless solid (0.058 g, 92%). 1 [23] According to general procedure 1, the reaction was stopped after 1 h. The mono/di-methylated product ratio is 49:1 according to 1 H NMR spectroscopy and after purification, the title compound was obtained as a light yellow solid (0.060 g, 84%).  [24] According to general procedure 1, the reaction was stopped after 1 h. The mono/di-methylated product ratio is 49/1 according to the crude 1 H NMR and after purification, the title compound was obtained as a colorless solid (0.063 g, 72%). 1

Synthesis of Organometallic Complexes
Based on the reported method [34] for synthesizing rhodacycle 4a, an optimized reaction condition (general procedure 3) could provide a better result. Rhodacycle 4a: [35] According to general procedure 4, 4a was obtained as an orange solid (0.071 g, 76%; general procedure 3: 96% for 24 hours

Mechanistic experiments
To test their competence in catalysis, C-H methylation reactions were performed using corresponding rhodacycles as the catalyst.

S32
In the first phenoxy pyridine competition experiment products 8c and 8d were formed in a 26% and 22% yield, respectively. In the second competition reaction, product 8h was formed in 16% spectroscopic yield, while product 8i was formed in 6% yield.

S33
6. Crystallographic data X-ray crystallography: All the measurements performed at 170 K using graphitemonochromatized Mo Kα radiation using a Bruker D8 APEX-II equipped with a CCD camera. Data reduction was performed with SAINT. Absorption corrections for the area detector were performed using SADABS. The structure was solved by direct methods (SHELXT) and refined by full-matrix least-squares techniques against F2 using all data (SHELXL) using the OLEX2 suit of programs. All non-hydrogen atoms were refined with anisotropic displacement parameters if not stated otherwise. Hydrogen atoms constrained in geometric positions to their parent atoms. CCDC 2017322-2017324 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/structures.