M[TPP]Cl (M = Fe or Mn)-Catalyzed Oxidative Amination of Phenols by Primary and Secondary Anilines

Iron- and manganese-catalyzed para-selective oxidative amination of (4-R)phenols by primary and secondary anilines was developed. Depending on the identity of the R group, the products of this efficient reaction are either benzoquinone anils (C–N coupling) that are produced via a sequential oxidative amination/dehydrogenation (R = H), oxidative amination/elimination (R = OMe) steps, or N,O-biaryl compounds (C–C coupling) that are formed when R = alkyl through an oxidative amination/[3,3]-sigmatropic rearrangement (quinamine rearrangement) process.

* sı Supporting Information ABSTRACT: Iron-and manganese-catalyzed para-selective oxidative amination of (4-R)phenols by primary and secondary anilines was developed. Depending on the identity of the R group, the products of this efficient reaction are either benzoquinone anils (C−N coupling) that are produced via a sequential oxidative amination/dehydrogenation (R = H), oxidative amination/elimination (R = OMe) steps, or N,O-biaryl compounds (C−C coupling) that are formed when R = alkyl through an oxidative amination/ [3,3]-sigmatropic rearrangement (quinamine rearrangement) process.
O xidative cross-coupling reactions 1 between phenols and anilines are powerful methods for assembling Ncontaining phenolic compounds that are applied in a variety of applications, such as natural product synthesis and asymmetric catalysis. 2 Phenols and anilines are both strong π-nucleophiles and have low oxidation potentials. 3 Therefore, carrying out their oxidative coupling in a selective manner is a challenging synthetic task. The use of anilides, protected anilines, 4 or masked phenols 5 as coupling partners overcame part of the chemoselectivity challenges. However, these reactants do not necessarily react in the same fashion as their parent anilines and phenols.
Over the years, methods for mediating the oxidative coupling between tertiary or secondary aniline derivatives and phenols have been reported. Chandrasekharam 6 and Knolker 7 developed conditions for the iron-catalyzed oxidative cross-coupling between N,N-dialkylanilines and 2-naphthols, and Shindo reported that a heterogeneous Rh−C catalyst mediates the oxidative coupling between 1-(2-naphthalenyl)piperidine and phenol derivatives (Scheme 1A). 8 The accepted mechanism underlying these reactions, which affords the N,Obiaryl products, involves the coupling between the tertiary anilino radical and nucleophilic phenol(ate). On the other hand, Patureau, 9 Xia, 10 Lei, 11 and Antonchick 12 demonstrated that secondary diarylamines undergo oxidative amination of phenols under various metal-free oxidation conditions (Scheme 1B). However, primary anilines are considered to be more challenging coupling partners as they tend to form "aniline black" and azobenzenes under mild oxidation conditions. 13 One of the very few examples is Kocǒvsky's synthesis of 2-amino-2′-hydroxy-1,1′-binaphthyl (BINOL), an axially chiral ligand, by coupling 2-naphthol and 2-aminonaphthalene using redox copper(II)amine complexes. 14 Our study reveals that the chemical properties of 2-aminonaphthalenes and primary or secondary phenylamine deriva-tives are different; therefore, they react by distinct mechanisms and selectivity.
As part of our group's ongoing research program, which focuses on the development of metal-catalyzed oxidative phenol coupling reactions, we have developed a highly efficient FeCl 3 -catalyzed oxidative cross-coupling reaction between phenols and primary, secondary, and tertiary 2-aminonaphthalenes [FeCl 3 (10 mol %), TFA (1.25 equiv), t-BuOOt-Bu, and 1,1,1,3,3,3-hexafluoropropan-2-ol (HFIP) at rt]. 15 The mechanism underlying this reaction involves coupling between an iron-ligated phenoxyl radical and a neighboring nucleophilic 2-aminonaphthalene ligand (radical-anion/nucleophile coupling mechanism). 15 2-Aminonaphthalenes are much stronger nucleophiles and have oxidation potentials higher than those of their phenolic partners; therefore, a high level of cross-coupling selectivity is observed. However, the latter conditions were found to be limited to 2-aminonaphthalene derivatives. For example, no product was obtained when 2,6-dimethoxyphenol (1a) and 3,4-dimethoxyaniline (2a) coupled under the above conditions (Table 1, entries 1 and 2). Intrigued by these results, we examined alternative redox systems for mediating the oxidative cross-coupling between phenols and primary anilines.
Herein, an efficient M[TPP]Cl (M = Fe or Mn)-catalyzed para-selective oxidative amination of phenols by primary and secondary anilines is described. The products of this reaction are unstable quinamine intermediates (Scheme 1C) that, depending on the identity of the phenol's para-substituent, are rapidly converted into either benzoquinone anils (via sequential dehydrogenation/elimination) or N,O-biaryl compounds (through quinamine rearrangement).
Formation of benzoquinone anil 3 from 1a and 2a is a process that involves two sequential dehydrogenation steps. Our attempts to isolate reaction intermediates (such as I, Table 1) by reducing the amount of the peroxide to 1 equiv or by using the less acidic 2,2,2-trifluoroethanol (TFE) solvent were not successful. Under these conditions, product 3 was obtained in reduced yields, as the sole detected product (57 and 44%, respectively, entries 7 and 8). Finally, monitoring the progress of the reaction by HPLC analysis revealed that the coupling is completed within 30 min (76% yield, entry 9).
Further insights into the oxidative amination of phenols by anilines were obtained from a set of control experiments. In order to dismiss a nucleophilic addition mechanism, 1,4benzoquinone 1aa and 2,4-dimethoxyaniline (2b) were reacted under the coupling conditions (Scheme 2A). Indeed, HPLC analysis revealed that benzoquinone anil 4, which is obtained in 96% yield from the reaction between phenol 1a and aniline 2b ( Figure 1A), was not detected. Next, we examined the dehydrogenation reaction of anilinophenols. For that purpose, masked anilinophenol 13a was prepared according to the Buchwald−Hartwig amination protocol (see the Supporting Information). 19 The hydrogenolysis of the benzylic group from 13a (H 2 , Pd/C, EtOAc, rt) afforded air-sensitive anilinophenol 13b, which rapidly developed the characteristic red color of benzoquinone anils. Unfortunately, our attempts to isolate 13b in a pure form failed as it gradually oxidized into compound 13, indicating that the dehydrogenation of the anilinophenol intermediates takes place spontaneously under air atmosphere, even in the absence of a redox metal catalyst.
Mechanistically, based on the work of Groves and others that studied the chemistry of iron porphyrins, 20 our group's   Organic Letters pubs.acs.org/OrgLett Letter tained. These competitive binding experiments support our central hypothesis that the selective binding of the phenol to the axial ligand during the aniline dehydrogenation is the source of the cross-coupling selectivity. The outcome of this catalytic cycle is unstable anilinoquinone intermediate II (Scheme 2B) that further reacts via pathways that are related to the identity of the phenolic R group. Yet, it is not necessarily catalyzed by the iron porphyrin catalyst. To study the courses that para-substituted phenols take during the oxidative amination reaction, various phenols (1.5 equiv) were reacted with anilines (1 equiv) under our general conditions [Fe[TPP]Cl (1 mol %), t-BuOOH (2 equiv), HFIP, rt, Figure 1]. First, 2,6-dimethoxyphenol (1a), 2,6-dimethylphenol (1b), 2-methoxy-5-methylphenol (1c), and 2,5-dimethylphenol (1d), which have an available paraposition, reacted with a long list of aniline derivatives, affording, after a sequential dehydrogenation step, benzoquinone anils 3−19 in moderate to excellent yields ( Figure 1A). 25 Only anilines with strong electron-donating groups at the paraposition (such as methoxy, hydroxy, or amino) participate in the reaction. Aliphatic amines failed to react and seems to be beyond the scope of this reaction. To demonstrate the scalability of this reaction, benzoquinone anil 3 was prepared on a 3 mmol scale in 70% yield. The phenoxazinone is a common structural motif in natural products 26 and in pharmaceutically active compounds. 27 The homocoupling of 2-aminophenol (1e) and the cross-coupling with 2-amino-(4-t-Bu)phenol (1f) afforded 2-aminophenoxazinone 20 and 21 in 83 and 82% yields, respectively.
The oxidative amination of 4-alkylphenols led to a different class of compounds ( Figure 1C). The reaction between 2,4dimethylphenol (1i) and aniline 2a with Fe[TPP]Cl as the catalyst afforded N,O-biaryl product 24 in 39% yield. However, Mn[TPP]Cl, which was too reactive for the oxidative amination of phenol 1a by aniline 2a (Table 1, entry 11), showed improved results in the case of less oxidizable paraalkylphenols. Indeed, with Mn[TPP]Cl (1 mol %, UHP (1.5 equiv), HFIP, rt) as the catalyst, we were able to isolate compound 24 in an improved (52%) yield. The alternative conditions provided entry to N,O-biaryl compounds 25−30 in moderate to good yields ( Figure 1C). On the basis of the Miller studies, 28 it is postulated that the acid-sensitive quinamine intermediate II (R = alkyl) undergoes [3,3]sigmatropic rearrangement. 29 Importantly, despite our efforts to react other aniline derivatives with para-alkylphenols, only aniline 2a and its N-methyl 2a derivative (products 27−29) were found to be suitable coupling partners. This is probably because they have a suitable electronic structure for stabilizing the anilino radical in the first step and for facilitating the rearrangement in the second step.
Next, we studied the oxidative amination of phenols by secondary diarylamines. First, 4-(4-methoxyphenylamino)aniline (2c), which has both primary and secondary amine groups, was reacted with phenols 1a, 1b, 1g, and 1h (Scheme 3A). Although diarylamine 2c can react with both amine groups, selective coupling at the primary amine took place, affording benzoquinone anils 31−34 in moderate to good yields. On the other hand, the reaction of 4,4′-dimethoxydiphenylamine (2d) with phenol 1a led, after a sequential dehydrogenation steps, to the formation of catechol 35 (63% yield, Scheme 3B), presumably via intermediate III. In contrast, the reaction of 2d and 1b afforded benzaldehyde 36 in 43% yield, possibly through intermediates III and IV. These examples prove that secondary anilines are suitable coupling partners; however, they form highly unstable quinamine intermediates that further react under the reaction conditions.
Finally, in this work, we show that para-anilinophenols, such as 13b (Scheme 2A), are highly air sensitive; therefore, their isolation in a pure form is a challenging task. To access paraanilinophenols from benzoquinone anils, the phenolic group needs to be blocked under reduction conditions. For example, the hydrogenation of benzoquinone anil 3 (H 2 , Pd/C, EtOAc, rt) in the presence of di-tert-butyldicarbonate or acetic In conclusion, a highly efficient para-selective oxidative amination of phenols by primary or secondary anilines catalyzed by M[TPP]Cl (M = Fe or Mn) complexes was developed. This reaction provides a direct entry to benzoquinone anils and N,O-biaryl compounds that are not readily accessible by any other sustainable method. The postulated mechanism involves the coupling of a liberated anilino radical and an iron-ligated (4-R)phenoxyl radical. The C−N coupling step affords unstable anilinoquinone intermediates that undergo either dehydrogenation (R = H), elimination (R = OMe), or [3,3]-sigmatropic rearrangement (R = alkyl), depending on the identity of the phenolic parasubstituent. We intend to continue our group effort in developing novel catalytic systems for chemo-, regio-, and stereoselective oxidative aniline coupling reactions.
Experimental procedures and spectroscopic data of all new compounds (PDF)