PdI2 as a Simple and Efficient Catalyst for the Hydroamination of Arylacetylenes with Anilines

The hydroamination reaction is a convenient alternative strategy for the formation of C– N bonds. Herein, we report a new versatile and convenient protocol for the hydroamination of arylacetylenes with anilines using palladium iodide in the absence of any added ligand as catalyst. Mild conditions, excellent regioand stereoselectivity, and high functional group tolerance are the main features of this methodology. A subsequent reduction step gives access to a wide variety of secondary aromatic amines.


Introduction
Amines, imines, and enamines are useful building blocks in organic synthesis and ubiquitous motifs in biologically active molecules, as exemplified in Figure 1. One of the most attractive strategies for their preparation is the catalytic hydroamination of unsaturated compounds, such as alkenes or alkynes [1][2][3][4][5][6][7][8]. The hydroamination reaction, consisting of the addition of an N-H bond to a C-C multiple bond, features a 100% atom efficiency associated with the large availability of the starting materials (these are alkynes, alkenes, amines). Nevertheless, in order to promote the addition of the amino group on the multiple C-C bond, a high activation barrier, resulting from the repulsion between the high electron density of the nucleophile and the πelectrons of the multiple bond, needs to be overcome. In this context, metal catalysis can serve as a powerful tool to reduce the activation barrier either by coordinating the multiple bond and thus subtracting the electron density or by N-H metalation, resulting in a strongly nucleophilic metal amide moiety. Amide intermediates can be also generated by the oxidative addition of the N-H bond to the metal. Several metals have been found to catalyze hydroamination reactions. Early transition metals, such as Ti or Zr, as well as lanthanides, display high activity [9][10][11][12][13][14], although they are poorly stable, owing to their oxophilicity that renders them air and moisture sensitive. Late transition metals [3,[15][16][17][18][19] feature a superior stability to moisture and air, together with broader functional group tolerance. Remarkably, cationic gold-based catalysts were highly active in the hydroamination of alkynes with amines [20][21]. The choice of the catalytic system is driven not only by the performance but also by its availability and cost. In this regard, palladium is clearly more attractive than rhodium, iridium, or gold.
Palladium-catalyzed hydroamination reactions have found several applications in the synthesis of biologically relevant heterocycles [22][23][24][25]. Alkynes [26], alkenes [27], dienes [28], and allenes [29] can serve as acceptor in combination with both aromatic and aliphatic amines. Despite the versatility of palladium-catalyzed hydroaminations, elaborate and expensive catalytic systems are often required. In particular, in the reaction between an alkyne and an aromatic amine, palladium has been employed in the presence of NHC [30] or NHCP [31] ligands. In both these cases, the catalytic precursor needs to be activated with a silver salt that acts as halogen scavenger. 3-Imino phosphine (3IP) ligands [32] also provide an efficient alternative but, similarly, a multistep synthesis for obtaining the catalytic active complex is required. Yamamoto and coworkers contributed significantly to the field by exploiting the possibility of using simple palladium salts such as palladium nitrate [33]. A dramatic rate enhancement effect was observed with 2-aminophenols, where the chelating OH group was crucial for the reaction outcome. Subsequently, the same research group was able to expand the reaction scope to variously substituted aryl amines resorting to an aqua palladium complex [Pd(dppe)(H2O)2](TfO)2 [34]. The Yamamoto's group also reported the use of Pd(PPh3)4 in combination with benzoic acid for the hydroamination of internal alkynes to allyl amines [35].
Owing to its flexibility and synthetic attractiveness, hydroamination reactions are still of high interest, and we recently reported the hydroamidation of propargylic ureas to give imidazolidin-2ones and imidazol-2-ones [36]. In this work, we report the use of a cheap and commercially available palladium catalyst (i.e., PdI2) for the hydroamination of arylacetylenes with anilines under mild reaction conditions and low catalyst loading.

Results and Discussion
The palladium-catalyzed intermolecular hydroamination reaction was initially investigated using aniline (1a, 1 equiv, 0.4 mmol) and phenylacetylene (2a, 1.2 equiv) as model substrates, in dioxane (0.2 M) as the solvent at 80 °C for 18 h, in the presence of catalytic amount of a palladium salt (2 mol%, Table 1). Palladium acetate did not provide any conversion, while palladium dichloride produced only traces of imine 3a (Table 1, Entries 1-2). Gratifyingly, palladium iodide resulted in 73% 1 HNMR yield of the hydroamination product, which was identified as (E)-N,1-diphenylethan-1imine 3a deriving from the Markovnikov addition to the triple bond followed by isomerization (  [37]. The excess of iodide anions can in this case be detrimental to the yield, since it can affect the equilibrium of species I and II in the catalytic cycle (see below). We then investigated the effect of concentration, temperature, catalyst loading, and solvent. At 0.4 M, compound 3a was obtained with 81% yield (Table 1, entries 6), while a further increase of the concentration to 0.8 M led to a less satisfactory result (Table 1, entries 7). As expected, temperature also plays a fundamental role in this transformation. In fact, when the reaction was performed at 100 °C, we observed a slightly decrease in the yield of 3a (Table 1, entry 8, 77%). Synthetically useful yields were still observed when reducing the catalyst loading to 1.0 mol% ( Table  1, entry 9) and to 0.5 mol% (Table 1, entry 10) with yields of 3a of 63% and 47%, respectively. Table 1. Optimization study for the hydroamination of phenylacetylene with aniline a .
PdI2 (2)  Lastly, different solvents were evaluated. Dimethylformamide (DMF) proved to be ineffective under these reaction conditions (entry 11), while acetonitrile and toluene allowed the formation of the desired imine 3a even though in lower yields (entries 12 and 13).
Once the best reaction conditions ( Table 1, entry 6) were determined, we then investigated the scope for variously substituted anilines (1) and arylacetylenes (2) (Scheme 1). In general, the yields attained using PdI2 as catalyst were comparable with the ones achieved with the previously employed more complex catalytic systems [30]. For example, different alkyl chains at the para position of the aniline provided good yields, irrespective of the length of the chain (yields of 3b, 67%; 3c, 79%; 3e, 67%) or the steric hindrance (3d, 80%). Interestingly, an isopropyl group at the ortho position of the aniline also allowed an excellent product yield (3f, 83%). Fluoro-containing groups were well tolerated under these reaction conditions, regardless of their position or electronic properties. With a CF3 group at the meta position, the expected imine 3g was formed in 78% yield, and the same result was observed starting from the aniline ring bearing an OCF3 substituent para to the NH2 group (3h, 78%). para-Substituted anilines with fluoro, chloro, or bromo smoothly underwent the hydroamination, resulting in 75% (3i), 70% (3j), and 60% (3k) yields of the desired imines, respectively. Even a disubstituted aniline, such as 2-methyl-3-chloroaniline, behaved nicely, providing the corresponding product 3l with 85% yield. The OMe substituent in para position on the aniline ring led to 95% yield of 3m, while the presence of two OMe groups caused a significant decrease of the yield (3n, 29%).
Other substrates were less tolerated (Scheme 2). For instance, 4-iodoaniline gave only 26% yield of the corresponding imine (3r). The iodide on the ring was highly reactive under palladium catalysis, leading to a complex mixture of byproducts. For the same reason, 2-bromoaniline in combination with phenyl acetylene gave only 30% of the expected 3s. The presence of a strongly electronwithdrawing group, such as the nitro group on the aniline ring, completely inhibited the process. Alkylacetylenes were considerably less reactive than arylacetylenes, as exemplified by the formation of 3v in 20% yield starting from 1-decyne. Primary alkylamines were not reactive under the optimized conditions, likely due to their pronounced coordinating nature. Not surprisingly, imine products 3 were unstable under acidic conditions and the isolated yields after column chromatography were very low. We then decided to reduce compounds 3 to the corresponding, more stable secondary amines 4, using simple and inexpensive NaBH4. The overall isolated yields for the two-step synthesis of secondary amines 4 (I step: PdI2-catalyzed hydroamination and II step: reduction) starting from anilines and terminal arylacetylenes, reported in Scheme 3, ranged from 17% to 90%. In some cases, the two-step yield was about 20% lower than the corresponding NMR hydroamination yield (4o-q). This was mainly due to a non-optimized reduction procedure, which can be improved with the use of alternative reduction reagents, such as NaBH3CN.
From a mechanistic point of view, coordination of the triple bond of 2 to PdI2 generates the πcomplex I (Scheme 4). This type of coordination makes the triple bond susceptible to undergo nucleophilic attack of the nitrogen atom of amine 1 resulting in the vinylpalladium iodide intermediate II. Then, protonolysis of II regenerates PdI2 and leads to enamine intermediate III, which readily tautomerizes to imine 3. The presence of the iodide counteranion in the process is crucial and we believe that its role may be connected, on one hand, to its higher ability to act as the leaving group in the nucleophilic attack step and, on the other hand, to its higher electron-releasing power, which tends to favor the protonolysis step [38]. Moreover, iodide anions display a superior ability to stabilize palladium(II) species, compared to other counteranions, such as acetates [39]. Efforts at understanding the effect of iodide anions in this transformation are underway in our laboratories [40].

Materials
All chemicals were purchased from commercial sources and used as received. Dioxane and methanol were dried and stored over molecular sieves previously activated in an oven at 300 °C overnight. Catalytic reactions were carried out under nitrogen using the standard Schlenk technique. Reaction mixtures were analyzed using a GC Agilent Tenchnologies 7820A GC System equipped with FID detector and column Agilent Technologies 19091J 413 (30mX0,32mm) and a gas chromatograph Agilent Technologies 6890N Network GC System equipped with quadrupole Agilent Technologies 5973 Network Mass Selective Detector and column Fused Silica Capillary Column (30mX0,25mm) for GC-MS analyses. 1 H NMR and 13 C NMR spectra were recorded at 300 K on a Bruker 400 MHz using the solvent as internal standard (7.26 ppm for 1 H NMR and 77.16 ppm for 13 C NMR for CDCl3). The terms m, s, d, t, q and quint represent multiplet, singlet, doublet, triplet, quadruplet, and quintuplet respectively, and the term br means a broad signal.

Hydroamination of Terminal Alkynes with Anilines
A flame-dried test tube of 10 mL was charged with palladium iodide (0.016 mmol, 2 mol%), aniline 1 (0.8 mmol, 1 equiv) and alkyne 2 (1 mmol, 1.2 equiv). The test tube was sealed with a rubber cup, filled with nitrogen, evacuated, and backfilled with nitrogen three times. Then, dry 1,4-dioxane (2 mL) was added. The tube was placed in an oil bath at 80 or 100 °C and stirred for 18 hours. The reaction crude was cooled to room temperature, filtered through celite, and dried under reduced pressure. The 1 H NMR analysis on the reaction crude using dimethyl maleate as the internal standard was immediately acquired to determine the yield of imine 3.

Reduction of Imine 3 to Secondary Amine 4
The hydroamination crude was transferred in a two-neck round-bottom flask and dissolved in dry methanol (2 mL). The flask was placed in an ice bath and sodium borohydride (2-5 equiv) was added portion wise. The mixture was stirred at 0 °C until a complete conversion of 3, detected by TLC analysis. The mixture was quenched with 2 mL of KOH (1M). The crude was diluted with 20 mL of EtOAc, washed with water and brine and dried over MgSO4. The pure amine 4 was obtained after flash column chromatography using a mixture of hexane and EtOAc as eluent.