Mechanistic Investigations into Amination of Unactivated Arenes via Cation Radical Accelerated Nucleophilic Aromatic Substitution

Abstract

A mechanistic investigation into the amination of electron-neutral and electron-rich arenes using organic photoredox catalysis is presented. Kinetic and computational data support rate-limiting nucleophilic addition into an arene cation radical using both azole and primary amine nucleophiles. This finding is consistent with both fluoride and alkoxide nucleofuges, supporting a unified mechanistic picture using cation radical accelerated nucleophilic aromatic substitution (CRA-S_NAr). Electrochemistry and time-resolved fluorescence spectroscopy confirm the key role solvents play in enabling selective arene oxidation in the presence of amines. The synthetic limitations of xanthylium salts are elucidated via photophysical studies. An alternative catalyst scaffold with improved turnover numbers is presented.

Figure 1.

Strategies for S_NAr.

Figure 2.

Kinetic order determination of 1a (A) and 2a (B) using VTNA.

Figure 3.

Stern–Volmer plots for 1a and 2a using Catalyst A in HFIP.

Figure 4.

(A) Reaction coordinates for C─F and C─O functionalization using 2a as a nucleophile. Computations were carried out at the (U)M06-2X/6-311+G(d,p)/CPCM(dcm) level of theory. (B) Overall change in Gibbs free energy for fluoro- vs methoxy-substitution.

Figure 5.

Kinetic order determination of primary amine 2b using fluoroarenes (A) and alkoxyarenes (B).

Figure 6.

(A) Computation reaction coordinate C─O functionalization using 2b as a nucleophile. Computations were carried out at the (U)M06-2X/6-311+G(d,p)/CPCM(dcm) level of theory. All C─H bonds, except those associated with the nucleofuge, are omitted for clarity. (B) Overall change in Gibbs free energy for amination.

Figure 7.

Stern–Volmer plots for 1b and 2d using Catalyst C in TFE and DCE. Potentials are referenced to saturated calomel electrode (SCE).

Figure 8.

Emission spectra of Catalyst A and Catalyst D( A). Excitation spectra for Catalyst A (B) and Catalyst D (C). Individual spectra were obtained every 10 nm from 500 to 620 nm and normalized to 400 nm. All spectra were measured in HFIP and plotted as normalized intensity vs. wavelength (nm).

Figure 9.

Normalized fluorescence spectrum of Catalyst E (A). Excitation spectrum for Catalyst E (B). Individual spectra were normalized to 400 nm. All data were collected in HFIP and plotted as normalized intensity vs wavelength (nm).

Scheme 1.. Possible Mechanistic Pathways for the Model Reaction^a

^aAll electrochemical potentials were obtained in MeCN and are referenced to SCE.

Scheme 2.. Investigations into Catalyst Turnover via Possible Chain Propagation^a

^aIsolated yields and quantum yields are reported as an average of 2 individual trials.

Scheme 3.. Revised Mechanism for CRA-S_NAr

Scheme 4.. ¹³C KIE Studies at Natural Abundance^a

^aCrude conversion determined by GC using 1,2-dichlorobenzene as an internal standard. The error associated with ¹³C KIE measurements in the thousandths place is reported in parentheses.

Scheme 5.. Studies of Possible Racemization under CRA-S_NAr Conditions^a

^aAverage isolated yields are reported (n = 2). Enantiomeric excess values were determined via chiral HPLC.

Scheme 6.. Catalyst Degradation Studies^a,b,c

^aAverage yields are reported (n = 2). ^bYields determined by ¹H NMR using HMDSO as an internal standard. ^cIsolated yields are reported.

Scheme 7.. Synthetic Application of Catalysts D and E as Photooxidants^a,b

^aAverage isolated yields are reported (n = 2). For condition A, reactions were run on a 0.3 mmol scale while conditions B and C were conducted on a 1 mmol scale. TON = turnover number. ^b456 nm irradiation used in place of 427 nm