Synthesis of Arylamines via Aminium Radicals

Abstract Arylamines constitute the core structure of many therapeutic agents, agrochemicals, and organic materials. The development of methods for the efficient and selective construction of these structural motifs from simple building blocks is desirable but still challenging. We demonstrate that protonated electron‐poor O‐aryl hydroxylamines give aminium radicals in the presence of Ru(bpy)3Cl2. These highly electrophilic species undergo polarized radical addition to aromatic compounds in high yield and selectivity. We successfully applied this method to the late‐stage modification of chiral catalyst templates, therapeutic agents, and natural products.

All required fine chemicals were used directly without purification unless stated otherwise. All air and moisture sensitive reactions were carried out under nitrogen atmosphere using standard Schlenk manifold technique. THF was distilled from sodium/benzophenone, CH 2 Cl 2 and was distilled from CaH 2 , CH 3 CN was distilled from activated 4Å molecular sieves, Et 3 N was distilled over KOH. 1 H and 13 C Nuclear Magnetic Resonance (NMR) spectra were acquired at various field strengths as indicated and were referenced to CHCl 3 (7.27 and 77.0 ppm for 1 H and 13 C respectively). 1   SI-6 equiv.). The mixture was allowed to warm to room temperature overnight. The reaction was diluted with H 2 O and CH 2 Cl 2 and the layers were separated. The aqueous layer was extracted with EtOAc (x 3) and the combined organic layers were dried (MgSO 4 ), filtered and evaporated. Purification by column chromatography on silica gel eluting with petrol:EtOAc (9:1) gave the product.

SI-13
General Procedure for the Preparation of S10,11 -GP2 The carboxylic acid (1.0 equiv.) was dissolved in MeOH (0.05M) and a drop of H 2 SO 4conc was added. The mixture was refluxed for 1 hour, cooled to room temperature and evaporated. Et 2 O (10 mL) was added and the organic layer was washed with NaHCO 3sat (10 mL). The organic layer was dried (MgSO 4 ), filtered and evapoarted to give the methyl ester. No further purification was required.

Alternative Aminium Radical Precursors
We have evaluated the possibility of using alternative aminium radical precursors and subjected S8 and S9 to our reaction conditions. As reported in Table S2, none of them provided 3a and as a result cannot be considered as viable options in our reaction manifold.

General Procedure for the Amination Reaction -GP4
To a dry tube was added with the O-aryl hydroxylamine (1.0 equiv.) and Ru(bpy) 3  Purification by column chromatography on silica gel gave the product.
The structures of the unknown products have been determined by 2D analysis using COSY, HSQC, HMBC and NOESY spectroscopy.
The addition of strong Brönsted acids, like HClO 4 , has been known to increase dramatically the reduction potential of organic molecules and to accelerate the SET reduction from *Ru(bpy) 3

2+
. 16 In some cases, the reduction potential can even reach positive values (Scheme S3).

Mechanism Based on the Ru(II)-Catalysed Electron Relay
Scheme S5.

STEP 6:
The possibility of Ru(II) to act as an electron relay catalyst is supported by the evidence that the reaction between 1a and t-Bu-benzene leads to the formation of 3a in the absence of light (se also Table S1). As shown in Scheme S6, in the absence of Ru(bpy) 3 Cl 2 with or without blue LEDs irradiation, no product was formed (reactions 2 and 4). However, when the reaction was run with both Ru(bpy) 3 Cl 2 and HClO 4 in the dark, 3a was formed in 58% yield (reaction 3).

Detection of [Ru(bpy) 3 ] 3+
In order to get insights into the reactivity observed in the darkness, we studied the  Scheme S7.

Further Studies to Support the Electron Relay Mode
As this reactivity was not envisaged we have performed further studies in order to evaluate its feasibility. As shown in Table S3 we have evaluated several other electron relay catalysts in the reaction between 1a and t-Bu-benzene in the presence of HClO 4 in the dark. In general, 3a was obtained in low conversion, which supports the feasibility of reactivity (entries 2-9). In the case of Fe(bpy) 3 Cl 2 , which has been reported in the literature to be a competent electron relay catalyst, 21 3a was obtained in 40% yield but no improvement was observed under blue LEDs irradiation (entries 10 and 11). Furthermore, in the absence of HClO 4 no product was observed which supports protonated 1a to act as a strong oxidant (entry 12).

SI-53
We have then evaluated this dark electron relay reactivity with Ru(bpy) 3 Cl 2 in the presence of other aromatics and found that the yields were consistently lower when compared to the ones obtained upon blue LEDs irradiation (Table S4). In the case of cyano-naphtalene and strychnine (entries 3 and 4) the reactions were also run for 1h and overnight with no changes in the reaction yield. These study showed a significant difference in the reaction efficiency depending on blue LEDs irradiation. As a result, we believe that the electron-relay mechanism might be involved as part of a chain-propagation process but it does not account alone for the full formation of the reaction products.

STEP 7.
We have studied by DFT the feasibility of the chain-propagation and found that a SET between B and D/D' is thermodynamically feasible (Scheme S10). We believe that the strong ability of B to act as an oxidant is responsible for the success of this step. ii ii We have calculated the quantum yield for the process to be Φ = 44.

Conclusions
Overall, these mechanistic studies, revealed a complex interplay of three main mechanistic pathways all contributing to the success of the reaction. It is difficult at this stage to rule out any of these possible mechanisms as they all have supporting evidences. From these studies we also believe that, changing the nature of the aromatic coupling partner, can change the extent by which each mechanistic pathway contributes to the productive formation of the final product. This is visually represented in Scheme S11 with a generic benzene aromatic partner. The exact role of the blue LEDs irradiation in the improvement of the reaction performance depending on the aromatic partners remains unclear. Scheme S11.

Stability of Ru(bpy) 3 Cl 2 in the presence of HClO 4
The ability of *Ru(bpy) 3 Cl 2 to perform reductive SET in the presence of large amounts of HClO 4 has been already demonstrated in the literature whereby the rate of SET increases linearly with the HClO 4 concentration. 16,22 This means that decomposition pathways are not likely to happen. Nevertheless, we have evaluated Ru(bpy) 3 Cl 2 stability by monitoring its luminescence profile upon addition of HClO 4 .
This study revealed no change in the emission profile of the catalyst (see Section 4.8). SI-58

Arylation of Aminium Radicals
In all the possible productive mechanisms, the formation of the reaction product relies on the ability of the aminium radical to undergo polarised radical reaction with the aromatic partner. We have performed DFT studies to understand better this step.

Electrophilicity of Aminium Radicals
Aminium radicals are isoelectronic to alkyl radicals but carry a formal positive charge, which makes them powerful electrophilic species. According to our calculations, upon protonation of the aminyl radical there is a remarkable increase in electrophilicity (Scheme S12). This is in agreement with the radical reaction of aminium radicals with aromatics to be a highly polarized process. Scheme S12.

Reaction Selectivity -Fukui's Indices
In general our reactions lead to the formation of the para-aminated products as the major product. To rationalise this selectivity we have calculated the Fukui indices on several aromatics (Scheme S13). Our calculations show that the para-carbon is generally the most reactive and this is in accordance with our experimental results. This is in agreement with the radical reaction of aminium radicals with aromatics to be a highly polarized process.

Protonation Studies
In order to evaluate the ability of our nitrogen radical precursors to undergo protonation, we have performed 1 H NMR spectroscopy studies on the following aminyl radical precursors and Brönsted acids.

General Procedure for the Protonation Studies -GP5
A dry NMR was charge with a solution of 1a-S8,9 (0.06 mmol, 1.0 equiv.) in CD 3 CN (0.6 mL). The Brönsted acid (2.0 equiv.) was added, the NMR tube was shaken and the 1 H NMR spectrum was recorded.
The results of this study are collected in Table S6. SI-62

General Experimental Details
Cyclic voltammetry was conducted on an EmStat (PalmSens) potentiostat using a 3electrode cell configuration. A glassy carbon working electrode was employed alongside a platinum flag counter electrode and a silver pseudo-reference electrode.
All the solutions were degassed by bubbling N 2 prior to measurements. 5 mM solutions of the desired compounds were freshly prepared in dry acetonitrile along with 0.1 M of tetrabutylammonium hexafluorophosphate as supporting electrolyte and were examined at a scan rate of 0.1 V s -1 . Ferrocene (E 1/2 = +0.42 V vs SCE) 17,23 was added at the end of the measurements as an internal standard to determine the precise potential scale. Potential values are given versus the saturated calomel electrode (SCE). Irreversible reduction waves were obtained in all cases; therefore, the reduction potentials were estimated at half the maximum current, as previously described by Nicewicz. 17 SI-63

Emission Quenching Experiments
Stern-Volmer experiments for all the components of the reaction mixture were carried out monitoring the emission intensity of argon-degassed solutions of Ru(bpy) 3 Cl 2 (1 x 10 -5 M) containing variable amounts of the quencher in dry acetonitrile (Table S8 and Scheme S15). The reported excited-state lifetime for Ru(bpy) 3 Cl 2 in MeCN (0.855 µs) 24 was used for k q calculations (Table S9).