Mild and Regioselective Pd(OAc)2-Catalyzed C–H Arylation of Tryptophans by [ArN2]X, Promoted by Tosic Acid

A regioselective Pd-mediated C–H bond arylation methodology for tryptophans, utilizing stable aryldiazonium salts, affords C2-arylated tryptophan derivatives, in several cases quantitatively. The reactions proceed in air, without base, and at room temperature in EtOAc. The synthetic methodology has been evaluated and compared against other tryptophan derivative arylation methods using the CHEM21 green chemistry toolkit. The behavior of the Pd catalyst species has been probed in preliminary mechanistic studies, which indicate that the reaction is operating homogeneously, although Pd nanoparticles are formed during substrate turnover. The effects of these higher order Pd species on catalysis, under the reaction conditions examined, appear to be minimal: e.g., acting as a Pd reservoir in the latter stages of substrate turnover or as a moribund form (derived from catalyst deactivation). We have determined that TsOH shortens the induction period observed when [ArN2]BF4 salts are employed with Pd(OAc)2. Pd(OTs)2(MeCN)2 was found to be a superior precatalyst (confirmed by kinetic studies) in comparison to Pd(OAc)2.


General Experimental Details
Commercially-sourced solvents and reagents were purchased from Acros Organics, Alfa Aesar, Fisher Scientific, Fluorochem, Sigma-Aldrich or VWR and used as received unless otherwise noted. Petrol refers to the fraction of petroleum ether boiling in the range of 40-60 °C. Room temperature (RT) refers to reactions where no thermostatic control was applied and was recorded as 16-23 °C.
Thin layer chromatography (TLC) analysis was performed using Merck 5554 aluminium backed silica plates. Spots were visualised by the quenching of ultraviolet light (λmax = 254 nm). Retention factors (Rf) are quoted to two decimal places and reported along with the solvent system used in parentheses. All flash column chromatography was performed using either Merck 60 or Fluorochem 60 Å silica gel (particle size 40-63 µm) and the solvent system used is reported in parentheses.
Optical rotations were recorded using a digital polarimeter at 20 °C (using the sodium D line, 259 nm) with a path length of 100 mm, with the solvent and concentration used indicated in the text. The appropriate solvent was used as a background with ten readings taken for each sample and the average [α]ᴅ values in units of 10 −1 deg cm 3 g −1 quoted to one decimal place.
Melting points were recorded using a Stuart digital SMP3 machine using a temperature ramp of 3 °C min -1 and are quoted to the nearest whole number. Where applicable, decomposition (dec.) is noted.
Proton ( 1 H) spectra were typically recorded at 400 MHz. Chemical shifts are internally referenced to residual non-deuterated solvent (CHCl3 δH = 7.26 ppm), given to two decimal places.
Carbon-13 ( 13 C) spectra were recorded at 101 MHz. Chemical shifts are internally referenced to residual solvent (CDCl3 δC = 77.16 ppm) and given to one decimal place.
Boron-11 ( 11 B) spectra were recorded at 128 MHz and obtained with 1 H decoupling. Chemical shifts are externally referenced to BF3·OEt2 and given to one decimal place.
Fluorine-19 ( 19 F) spectra were recorded at 376 MHz and obtained with 1 H decoupling. Chemical shifts are externally referenced to CFCl3 and given to one decimal place.
Electrospray ionisation (ESI) mass spectrometry was performed using a Bruker Daltronics micrOTOF spectrometer. Electron impact (EI) mass spectrometry was performed using a Waters GCT Premier mass spectrometer. Mass to charge ratios (m/z) are reported in Daltons with percentage abundance in parentheses along with the corresponding fragment ion, where known. Where complex isotope patterns were observed, the most abundant ion is reported. High resolution mass spectra (HRMS) are reported with less than 5 ppm error.
Infrared spectra were recorded using a Bruker Alpha FT-IR spectrometer and were carried out as ATR. Absorption maxima (νmax) are reported in wavenumbers (cm -1 ) to the nearest whole number and described as weak (w), medium (m), strong (s) or broad (br).
UV-visible spectroscopy was performed on a Jasco V-560 spectrometer, with a background taken in the appropriate solvent prior to recording spectra, using a quartz cell with a path length of 1 cm. The wavelength of maximum absorption (λmax) is reported in nm along with the extinction coefficient (ε) in mol dm -3 cm -1 .

S3
Diffraction data were collected at 110 K on an Agilent SuperNova diffractometer MoKα radiation (λ = 0.71073 Å). Data collection, unit cell determination and frame integration were carried out with CrysalisPro. Absorption coefficients were applied using face indexing and the ABSPACK absorption correction software within CrysalisPro. Structures were solved and refined using Olex2 1 implementing SHELX algorithms and the Superflip 2 structure solution program. Structures were solved by charge flipping, Patterson or direct methods and refined with the ShelXL 3 package using full-matrix least squares minimisation. All non-hydrogen atoms were refined anisotropically. Where applicable, absolute configurations were established by anomalous dispersion.
Transmission electron microscopy was performed at the Department of Biology Technology Facility, University of York, using an FEI Technai 12 G2 BioTWIN microscope operating at 120 kV, and images were captured using an SIS Megaview III camera. Samples were prepared by suspending ca. 1 mg of material in reagent grade ethanol with vigorous shaking, applying a small amount to a TEM grid, and allowing the solvent to evaporate. The grids used were 200 mesh copper grids with a Formvar/carbon support film. The resulting images were enlarged and particle sizes measured manually.

General Procedures General Procedure A: Synthesis of aryldiazonium tetrafluoroborates 4
The appropriate aniline (1 eq.) was dissolved in ethanol and HBF4 (50 wt% in H2O, 2 eq.) before being cooled to 0 °C with stirring. A 90% solution of tert-butylnitrite (2 eq.) was then added dropwise and the mixture was allowed to warm to room temperature with stirring for 1 h. After 1 h Et2O was added to precipitate the aryldiazonium tetrafluoroborate which was collected by filtration through a glass sinter and washed with further Et2O until the filtrate ran clear, then dried in vacuo to afford the desired compound, which was subsequently stored at −18 °C.

General Procedure B: Direct arylation of tryptophan with Pd(OAc)2
To a microwave tube was added tryptophan 1 (50 mg, 0.192 mmol, 1 eq.), the appropriate aryldiazonium salt (0.192 mmol, 1 eq.), Pd(OAc)2 (2 mg, 9.6 μmol, 5 mol%) and EtOAc (5 mL). The reaction mixture was stirred at RT for 16 h. After 16 h the resulting brown reaction mixture was filtered through Celite then washed with sat. aq. NaHCO3. The organic layer was collected and dried over MgSO4, filtered and evaporated to give a brown solid. When purification was required, it was performed using dry-loaded flash column chromatography with a SiO2 stationary phase and the solvent system specified for each compound.

Green Metrics Data
All metrics were calculated using the Chem21 unified green metrics toolkit. 16 In addition to the reagent quantities stated in the experimental section in this publication and our previous publications, 13 the following values for workup reagents/solvents were used in all cases:

Kinetic Curves Using UV-Visible Spectroscopic Data
General procedure for kinetic measurements with Pd(OAc)2 To a microwave tube was added tryptophan 1 (50 mg, 0.192 mmol, 1 eq.), aryldiazonium salt 2a (37 mg, 0.192 mmol, 1 eq.), Pd(OAc)2 (2 mg, 9.6 μmol, 5 mol%) and EtOAc (5 mL). The reaction mixture was stirred at 37 °C, with aliquots of 100 µL taken every 5 min. The stirring was stopped for 10-15 s before each aliquot was taken. The aliquots were prepared for UV-visible spectroscopy by filtration through a Celite plug and dilution to 100 mL in EtOAc (1000-fold dilution). A UV-visible spectrum was then recorded, scanning between 400-256 nm. After the reaction had reached completion (or ceased as a result of catalyst poisoning), the resulting brown reaction mixture was filtered through Celite then washed with sat. aq. NaHCO3. The organic layer was collected and dried over MgSO4, filtered and evaporated to give a brown solid. 1 H NMR spectroscopic analysis of the crude material confirmed product conversion (3).

Catalyst poisoning tests
The general procedure above was followed, with addition of either PVPy (202 mg, 1.92 mmol, 10 eq., 200 eq. wrt Pd) or Hg (28 µL, 385 mg, 1.92 mmol, 10 eq., 200 eq. wrt Pd) to the reaction after 90 min. Alternatively, the reaction mixture was filtered through a pre-heated (ca. 37 °C) Celite TM plug after 90 min then recharged to a microwave vial and reaction continued.

Analysis of errors in kinetic measurements
The reaction between tryptophan 1 and aryldiazonium salt 2a was performed three times to evaluate the errors associated with each measurement. This indicated that the key source of error was irregularity in the length of the induction period (subsequently confirmed to be due to watersee studies by in situ IR in the main paper), which resulted in the data spread of kobs seen in the figure below.
The figure above shows the errors between three different runs, with associated kobs values.