Oxidative Dearomatization of Phenols and Anilines via λ3- and λ5-Iodane-Mediated Phenylation and Oxygenation

Treatment of 2-methylphenols with chloro(diphenyl)-λ3-iodane led to their regioselective dearomatizing 2-phenylation into cyclohexa-2,4-dienone derivatives via a proposed ligand coupling reaction. In the same vein of investigation, treatment of 2-methylanilines with the λ5-iodane 2-iodoxybenzoic acid IBX reagent led to their regioselective dearomatization into previously undescribed ortho-quinol imines.


Results and Discussion
Our previous investigation on carbon-carbon (C-C) bond-forming and dearomatizing reaction of 2substituted arenols was accomplished by a process that was apparently intermolecular and oxidatively initiated by BTI in the presence of either an allylsilane or a silyl enol ether [8,10]. Intramolecular variants of such a transformation are known and the most effective ones rely on the use of either a lead(IV) [13,14] or a bismuth(V) [15,16] species bearing heteroatomic and aryl ligands. The starting 2substituted arenol first reacts with the aforementioned metallic species to exchange one of its labile heteroatomic ligands. Once the aryloxy unit of the substrate is mounted onto the metallic center, a reductive elimination step can take place with a concomitant C-C ligand coupling directed at the 2substituted position of the aryloxy unit to generate a cyclohexa-2,4-dienone product. Ligand coupling is also possible on an hypervalent iodine(III) center [17,18], and it has been proposed that diaryl-λ 3iodanes (i.e., Ar 2 IL) † do transfer, through this mechanistic path, an aryl group to enolates to furnish α-C-arylated carbonyl compounds [17,[19][20][21]. † Ar 2 IL-type compounds are usually referred to as diaryliodonium salts despite the fact that their geometry is not tetrahedral, but pseudotrigonal bipyramidal, as expected for hypervalent iodine(III) species. The lamba iodane terminology is thus used here to name hypervalent iodine compounds [17]. Phenolates being aromatic variants of generic enolates, we thus contemplated the idea of achieving phenol dearomatization leading to 6-arylcyclohexa-2,4-dienone derivatives of type 3 using diaryl-λ 3iodane species (Scheme 2). The desired ligand coupling reaction could imply either passage through an aryloxy-λ 3 -iodane intermediate of type 2c (i.e., ligand exchange path a), which would then tautomerize into a 6-λ 3 -iodanyl ketone of type 2d, or direct formation of 2d via participation of the 2-methylated carbanionic form of the starting phenolate 1b (i.e., ligand exchange path b). In any event and to the best of our knowledge, such an iodane-mediated reaction has never been described.
We commenced our investigation of this dearomatizing arylation process by identifying the most appropriate reaction conditions to promote C-C ligand coupling (i.e., via 2d), while minimizing the C-O ligand coupling alternative (i.e., via 2c) that leads to diaryl ethers of type 4 (Scheme 2). This preliminary examination was carried out using 2,3,5-trimethylphenol (1a) with the aim of directing the introduction of a phenyl group at its substituted 2-position to enable the desired dearomatization into the cyclohexa-2,4-dienone 3a. The presence of a methyl group at the 3-position of 1a was chosen in order to block [4+2] dimerization of 3a [5]. Indeed, the presence of such a small electron-releasing group at the corresponding 5-position of cyclohexa-2,4-dienones is known to block their dimerization [22,23]. Treatment of 1a with tetrafluoroboro(diphenyl)-λ 3 -iodane [24] in DMF at room temperature using weak bases such as pyridine or triethylamine failed to initiate any reaction (Table 1, entries 1 and 2). The use of a strong alkoxide base [20,25,26] was necessary to engage the reaction, but the only observed product was the diaryl ether 4a, which was isolated in a moderate yield of 41% (entry 3) [27,28]. Change of the diphenyl-λ 3 -iodane reagent to chloro(diphenyl)-λ 3 -iodane led to some improvement in the yield of 4a, but no formation of the desired dearomatized product 3a was observed (entry 4). The first glimpse of success was obtained by performing the reaction in a polar protic solvent (i.e., t-BuOH) at room temperature, in which case 3a was isolated in 6% or 8% using either a tetrafluoroborate unit or a chloride atom as the ligand L on the iodine(III) of the reagent. In both cases, the major product was again the diaryl ether 4a (entries 5 and 6). It is worth noting that no biaryl product of type 5, which could have resulted from introduction of a phenyl residue at the unsubstituted 6-position of the starting phenol 1a via an α-λ 3 -iodanyl ketone of type 2e (Scheme 2), was observed. It thus appears that a small alkyl group, such as a methyl group, at one phenoxy ortho-position plays a role in directing delivery of a phenyl residue at that substituted position and not at the unsubstituted position (vide infra). This observation is in agreement with a route via an α-λ 3 -iodanyl ketone intermediate of type 2d (Scheme 2). Neither performing the reaction at -20 °C (entry 7) nor using a smaller counter-cation such as Na + or Li + (entries 6 b and 8 c ) to promote a change of chemoselectivity in favor of C-C coupling resulted in any improvement of the yield of 3a. So, the use of chloro(diphenyl)-λ 3 -iodane in the presence of potassium tert-butoxide in tert-butanol at room temperature so far emerged as the best conditions to achieve the desired transformation, albeit in low yields.
We then carried out a series of reactions under these conditions using three differently substituted phenols 1b-d ( Table 2). The unsubstituted phenol 1b led exclusively to the formation of the diaryl ether 4b in a good yield of 75%. This observation corroborates that made with phenol 1a on the role of a small alkyl substituent at one phenoxy ortho-position (vide supra) not only in directing the introduction of the phenyl residue at that carbon position, but also in promoting C-phenylation. Hence, starting from phenols bearing the same alkyl substituent at each of two equivalent ortho-positions should significantly boost yields of C-phenylation over O-phenylation. This hypothesis was confirmed using the di-and trimethylphenols 1c and 1d, which gave rise to the formation of the desired cyclohexa-2,4-dienones 3c and 3d in 37% and 42%, respectively ( Table 2, entries 2 and 3). Moreover, these cyclohexa-2,4-dienones are stable compounds that do not spontaneously engage in [4+2] dimerization, despite the absence of an alkyl substituent at their 5position [22,23]. Interestingly, 2,4,6-trimethylphenol (1d) also led to phenylation at the methylated 4position (entry 3). Although the yield of the corresponding product 6d was only 7%, its formation is in accordance with a ligand coupling event from a γ-λ 3 -iodanyl ketone of type 2f that would be formed from a direct ligand exchange between the 4-methylated carbanionic form of the starting phenolate 1d and the chloro(diphenyl)-λ 3 -iodane reagent (Scheme 3). Further mechanistic investigations need to be performed to confirm that the observed products are generated through ligand coupling, but the reactions described here are the first examples of iodane-mediated dearomatizing phenylation of substituted phenols.  We then turned our attention to another iodane-mediated dearomatization reaction involving the oxygenation of anilines and anilides with the aim of preparing ortho-quinol imines. Quinonoid compounds are powerful intermediates for organic synthesis, and those bearing a nitrogen functionality obviously expand their versatility as potential precursors and building blocks for the synthesis of biologically relevant natural and non-natural substances [29]. Among those, quinonoid imines such as quinone imines, quinone imine ketals and quinol imines occupy a valuable position as potentially useful synthons, but their imine function is unfortunately very sensitive to hydrolysis, unless the imine nitrogen is substituted by an acyl or a sulphonyl group [30,31] Most previous studies on quinonoid imines and imides have focused on para-compounds, and their generation was based on electrochemical or chemical oxidation of aniline and anilide parents [30,31]. Iodane reagents have seldom been used for mediating such transformations [32][33][34], and the only examples in the orthoseries [35] are those recently reported by Nicolaou and his co-workers, who exploited Dess-Martin periodinane (DMP) generated ortho-quinone imides in Diels-Alder reactions [29,36,37]; DMP belongs to the ArIL 4 type of the λ 5 -iodane reagents. To the best of our knowledge, the formation of ortho-quinol imines by any means has never been reported. In order to achieve the preparation of such elusive species, we first went back to λ 3 -iodanes bearing two heteroatomic nucleofugal ligands (i.e., ArIL 2 type) in order to promote the desired oxidative oxygenation process [17] starting from 2substituted anilines or anilides.
With this consideration in mind, we attempted to perform an oxidative ortho-acetoxylation of aniline 7a with DIB either in pure CH 2 Cl 2 or in a 3:1 CH 2 Cl 2 /AcOH mixture, but only intractable mixtures were obtained. However, the use of the λ 5 -iodane of type ArIL 4 IBX at room temperature gave a 2:1 mixture of the para-quinone imine 10a and the ortho-quinol imine 11a, as evidenced by NMR analysis (Scheme 4). Silica gel chromatography of this mixture furnished 10a in 42% and 11a in 11.5%, as well as traces (i.e. < 2%) of the Diels-Alder dimer 12a derived from 11a (vide infra) [5]. Compound 10a probably results from an initial IBX-mediated formation of the ortho-quinone imine 8a (Schemes 4 and 7), which is rapidly trapped by some intact starting aniline 7a to furnish the Schiff base 9a. This ortho-quinone bisimine can then be quenched in a conjugate manner by the water released during the Schiff base formation and reoxidized to furnish the observed para-quinone imine 10a. Moreover, we were pleased to isolate 11a even in such a low yield, since, as mentioned above, such an ortho-quinol imine has never been described before.   Furthermore, this first reaction with aniline 7a showed us that IBX is capable of selectively delivering an oxygen at the positions adjacent to the amine function, in a manner probably similar to the one that we and others have suggested for analogous IBX-and SIBX-mediated orthooxygenations of phenols (Scheme 7) [4,5,12]. Having thus successfully tested the feasibility of dearomatizing an aniline into an ortho-quinol imine, the same reaction was performed using the anilide 7b in the hope of improving its yield, for quinonoid imides are known to be stable enough for isolation as opposed to imines that are sensitive to hydrolysis and polymerization [29][30][31]37]. In addition, the lower nucleophilicity of the anilide nitrogen should preclude nucleophilic trapping of the orthoquinone imine formed. For the same reason, the anilide 7b turned out to be a very robust starting material, which necessitated a treatment with IBX for 48 h in refluxing THF to be entirely transformed, as indicated by TLC monitoring [CH 2 Cl 2 /MeOH (40:1)] of the reaction progress (Scheme 5). Now IBX does react with hot THF [4,[38][39][40], therefore additional quantities of IBX were added to the reaction mixture every 12 h (see Experimental Section). These conditions led to the isolation of the ortho-quinone imide 8b and the ortho-quinol imine 11b in 6% and 8% yield, respectively. The acetyl group did not protect the imine double bond from hydrolysis, since dimer 12a was again obtained, this time in 35% yield. No dimer with the two imide functions still intact was observed. However, the formation of both 11b and its derived dimer 12a after treating 7b with IBX remains an interesting observation, since treatment of analogous 2-substituted anilides using DMP in the presence of water only led to ortho-quinone imides [29,37]. The mechanism proposed by Nicolaou and his co-workers for this ortho-oxygenation implies the intermolecular participation of the DMP-derived acetate of IBX (Ac-IBX) as the source of the oxygen atom [36,39], so the oxygen delivery by the bulky Ac-IBX may be indeed blocked at the 2-substituted position because of steric impediment. However, when the same authors used IBX in a mixed solvent system of THF/DMSO (10:1) at 90 °C to transform anilide derivatives tethered with an olefin unit, they only observed nitrogen/olefin cyclization products derived from one-electron oxidation at the nitrogen lone pair [39,40]. It is somewhat puzzling that they did not observe any formation of ortho-oxygenation products, like in the examples reported herein for which we propose that IBX acts as the oxygen source (Scheme 7).

Scheme 5.
NH O In any event, with this second dearomatization of a 2-substituted aniline derivative into an orthoquinol imine accomplished (Scheme 5), we then moved on to treat 2,4,6-trimethylaniline (7c), having two equivalent 2-substituted ortho-positions, with IBX in order to boost the yield of the corresponding ortho-quinol imine product. This reaction led to a very clean crude product showing NMR signals assignable only to the expected ortho-quinol imine 11c (Scheme 6). This quinol imine did not survive silica gel chromatography and was isolated in only 33.5% yield, together with the Diels-Alder cycloadduct 12c in 13% yield. The use of the stabilized version of IBX (i.e., SIBX) [4] also led to the isolation of 11c in 35% yield, again in concert with that of the dimer 12c (22%) (Scheme 6). This dimer has been previously prepared from 2,4,6-trimethylphenol via sodium periodate-mediated Adler oxidation [41], and via the benzeneseleninic anhydride-mediated Barton oxidation [42]. IBX is the only obvious source of oxygen in these ortho-oxygenations of anilines 7a and 7c and anilide 7b. In analogy with our previous mechanistic proposal for related oxygenations of phenols [5,12], an ionic pathway is here depicted to rationalize these oxidative dearomatizing transformations leading to the isolation of previously undescribed ortho-quinol imines such as 11a-c (Scheme 7). Thus, once the amine or amide nitrogen of 7 has attacked the IBX iodine to give 13, reduction of this iodine(V) center could proceed via an ionic and concerted path to yield 11 and o-iodosobenzoic acid (IBA).

OH
Of course, a single electron transfer (SET) mechanism [43] leading to the nitrogen radical cation 14, followed by delivery of an hydroxyl radical to the substrate aromatic ring to give the intermediate 15, that would then fragment into 11 and IBA, cannot be excluded at this point. Further work is in progress to delineate these mechanistic possibilities.
We wish to thank SIMAFEX (www.simafex.com) and the Association Nationale de la Recherche Technique (CIFRE n° 301/2002) for A.O.'s graduate research assistantship, and the Ministère de la Recherche for J.G.'s graduate research assistantship. We are also grateful to SIMAFEX for their financial support and their generous gifts of DIB and SIBX reagents. We also wish to thank the Centre Régional de Mesures Physiques de l'Ouest, Université de Rennes 1, for electrospray high resolution mass spectrometry data.

Experimental
General Diacetoxyiodobenzene (DIB) and stabilized o-iodoxybenzoic acid (SIBX) [4] were obtained from SIMAFEX. Chloro(diphenyl)-λ 3 -iodane (Ph 2 ICl, 97%) was purchased from Aldrich. Tetrafluoroboro(diphenyl)-λ 3 -iodane (Ph 2 IBF 4 ) [20,24] and o-iodoxybenzoic acid (IBX) [44] were prepared from DIB and 2-iodobenzoic acid, respectively. Dichloromethane (CH 2 Cl 2 ) and tert-butyl alcohol (t-BuOH) were used as received. Tetrahydrofuran (THF) was purified by filtration through alumina under N 2 immediately before use. Moisture and oxygen sensitive reactions were carried out in flamedried glassware under N 2 . Evaporations were conducted under reduced pressure at temperatures less than 30 °C unless otherwise noted. Column chromatography was carried out under positive pressure using 40-63 µm silica gel (Merck) and the indicated solvents. Further drying of the residues was accomplished under high vacuum. Melting points are uncorrected. NMR spectra of samples in the indicated solvent were run at 250, 300 or 400 MHz. Carbon multiplicities were determined by DEPT-135 and J-MOD experiments. Electron impact (70 eV), chemical ionization and electrospray mass spectrometry low and/or high resolution (EIMS, CIMS, ESIMS and HRMS) were obtained from the mass spectrometry laboratory at the CESAMO, Université Bordeaux 1, from the mass spectrometry laboratory at the Institut Européen de Chimie et Biologie, Pessac, and from the mass spectrometry laboratory at the Centre Régional de Mesures Physiques de l'Ouest, Université de Rennes 1.

General Procedure for λ 3 -Iodane-Mediated Phenylation of Phenols
To a stirred solution of KOt-Bu (1.4 mmol, 1.05 equiv) in t-BuOH (4 mL, ca. 0.35 M) was added the phenol 1a-d (1.3 mmol) at room temperature, and the mixture was stirred for 1 h. The diphenyl-λ 3iodane reagent, i.e., Ph 2 IBF 4 [24] or Ph 2 ICl (1.5 mmol, 1.1 equiv), was added in one portion to this mixture. The resulting suspension was stirred for 20 h at room temperature, after which time water (20 mL) was added, and the mixture was extracted with CH 2 Cl 2 (3 × 5 mL). The combined organic extracts were washed once with water (5 mL), dried over Na 2 SO 4 , filtered, evaporated, and further dried under high vacuum.