A practical Cu(I)-catalyzed domino approach to 1,2-disubstituted indoles and its application for the assembly of indolophenanthridines

,


Introduction
Many 1,2-disubstituted indoles are biologically active and pharmaceutically useful. 1,2The assembly of 1,2-disubstituted indoles have attracted great interest for their important applications in medicinal chemistry and materials science. 3,48][19] In spite of these remarkable progresses, many of the previous methods require expensive and/or sensitive catalysts (such as Pd and Au salts) or special substrates such as 1-or 2-substituted indoles and o-(gem-dihalovinyl)anilines.The annulation of N-substituted 2alkynylanilines [20][21][22] and the coupling/cyclization of 2-alkynyl halobenzenes with primary amines [23][24][25] can also generate 1,2-disubstituted indoles.Katz's group utilized the tandem substitution/cyclization to furnish 2-substituted N-arylindoles, but the protocol was limited to the reactions of 2-alkynylfluorobenzenes. 26 Our group has developed a Cu(II)-catalyzed domino synthesis of 1,2-disubstituted indoles using coupling/cyclization of o-alkynylanilines with aryl boronic acids. 27Compared with aryl boronic acids, aryl halides are relatively cheap and more easily accessible.So replacing the arylating reagents by aryl halides for the transformation may prove to be both practical and economic.

Scheme 1. Reports on the one-pot synthesis of indolo[1,2-f]phenanthridines.
Cu-mediated cross-coupling has aroused increasing interest for its low cost and high efficiency. 33,34And Cu-mediated domino transformation has been applied as an attractive methodology since it can conveniently and efficiently construct various heterocyclic moieties. 35or instance, Ackermann et al. found that 1,2-disubstituted indoles could be synthesized using Cu(I)-catalyzed coupling/cyclization of 2-alkynyl haloarenes with anilines. 36To our knowledge, there is no report on the Cu-catalyzed domino assembly of 1,2-disubstituted indole derivatives from o-alkynylanilines and aryl halides.Furthermore, a one-pot approach to indolo[1,2-f]phenanthridines from o-alkynylanilines and 1,2-dihaloarenes has not been achieved until now.
As part of our ongoing research efforts toward the domino synthesis of heterocycles under copper catalysis, 27,[37][38][39][40] we became interested in applying the one-pot protocol to the assembly of 1,2-disubstituted indoles and the tandem synthesis of indolo[1,2-f]phenanthridines.The findings are reported herein.

Results and Discussion
For a preliminary study we chose the reaction of 2-(phenylethynyl)aniline 1a with iodobenzene 2a as the model transformation.(Table 1) A good yield of the desired disubstituted indole 3a was obtained when the reaction was performed with CuI (10 mol%), 1,10-phenanthroline (1,10-phen, 20 mol%), and Cs 2 CO 3 (2 equiv) (entry 1).From the catalyst screen it was found that CuI was the best catalyst (compare entry 1 with entries 2-5).It was found that non-polar solvents performed better than polar ones, and toluene was chosen as the optimal solvent (compare entry 1 with entries 6-9).The study of the effect of ligands indicated that 1,10-phenanthroline was superior to others (compare entry 1 with entries 10-13).In a blank experiment 29% yield of the indole was obtained in the absence of any ligand (entry 14).Other bases were also tested, t-BuOK performing best (compare entry 18 with entries 1 and 15-17).The addition of ligand had little effect on the reaction when t-BuOK was used as the base (entries 18, 19).The amount of base could be reduced to 1.2 equiv without affecting the reaction efficiency (entry 20).
The scope of this Cu(I)-catalyzed domino protocol was then investigated using a wide range of 2-alkynylanilines and aryl halide (Table 2).Considering that milder conditions may be more compatible and favorable for some reactions, we also investigated several reactions with Cs 2 CO 3 as the base (Table 2, 3a-3c, 3e, 3h, 3i, 3k).In most cases, the performance of Cs 2 CO 3 was comparable to that of t-BuOK (3a-3c, 3e).However, when the alkynylanilines with functional groups (such as NO 2 and CF 3 ) were employed, the use of Cs 2 CO 3 afforded the corresponding indoles in higher yields (3h and 3i).Thus Cs 2 CO 3 may be appropriate for the reactions of the substrates bearing strong electron-withdrawing groups.On the other hand, when other aryl halides (such as ArBr and ArCl) or o-alkynylanilines with bulky substituents (e.g.t-Bu and o-Br) were employed, t-BuOK was much more efficient than Cs 2 CO 3 (3a, 3k and 3l).
It is noteworthy that the reactions with o-dihaloarenes could selectively afforded the desired N-(o-bromoaryl)indoles, which would provide an additional handle for further derivation of these products (Table 2, 3l-3q).Inspired by the hypothesis that the intramolecular direct sp 2 C-H arylation of the N-(o-bromoaryl)indole would result in the formation of the corresponding polycyclic indole derivative, we attempted to synthesize the indolo[1,2-f]phenanthridine from 1-(2-bromophenyl)-2-phenyl-1H-indole (3l) by using this type of intramolecular transformation.Gratifyingly, the desired indolo[1,2-f]phenanthridine 4a was smoothly generated under Pdcatalysis (Scheme 2).We then applied this protocol to the assembly of more indolo[1,2-f]phenanthridines from N-(o-bromoaryl)-2-arylindoles.The intramolecular C-H arylation of the indoles with different substituents (such as Me, i-Pr, Cl, and F) on the phenyls was also successfully achieved under Pd catalysis (Scheme 2).In order further to enhance the power and practicality of the method, we next tried to combine the above two steps in one pot.Fortunately, the one-pot tandem synthesis of the polycyclic indoles was also achieved (Scheme 3).Initially, the domino synthesis of 3l was chosen as the model reaction.After stirring for 18 h under the Cu(I) catalysis, Pd(OAc) 2 , P(p-Tol) 3 and Cs 2 CO 3 were directly added to the reaction mixture without isolating intermediate 3l.We were pleased to find that the desired indolo[1,2-f]phenanthridine 4a was isolated in 85% yield (Scheme 3, entry 1).The domino reactions with other o-bromo iodobenzenes were also successively investigated, and moderate to good yields of the desired polycyclic heteroarenes were assembled.Compared with previous reports 29,31,32 , the present methods provided a more convenient and versatile approach to these polycyclic heterocycles, and the starting materials were more easily accessible (see Scheme 1).2) the intramolecular cyclization / intermolecular coupling process (Eq.][43][44] We attempted to capture the intermediate N-phenyl-2-(phenylethynyl)aniline 5a or 2phenylindole 6a at a lower temperature (Eq.3).However, only 18% of N-phenylindole 3a was isolated in this case.Without the addition of any copper catalyst, neither the desired indole 3a nor any intermediate was detected (Eq. 4), indicating that the copper catalyst was indispensable for the domino reaction.Notably, we also synthesized the intermediate 5a 27 and then subjected it to the standard copper(I)-catalyzed conditions (Eq. 5).The intramolecular annulation efficiently occurred to afford the desired indole in excellent yield (Eq. 5), indicating that the reaction probably underwent a coupling/cyclization process.In order to further eliminate the possibility of the cyclization/coupling pathway under our conditions (Eq.1), two additional control experiments were also performed.With or without the promotion of the Cu catalyst, no cyclized product 6a was detected when iodobenzene was absent under the standard conditions (Eq. 6 and Eq. 7).It is worth noting that the weakly polar toluene was employed as the solvent in our transformation, and the results significantly differ from those methods [42][43][44] using strongly polar solvents. 45Based on these experiments and the relevant reports, 46 we tend to the opinion that the domino transformation probably underwent the intermolecular C-N coupling / intramolecular cyclization pathway (Eq.1).

Conclusions
In summary, we have demonstrated a copper(I)-catalyzed domino synthesis of 1,2-disubstituted indoles from 2-alkynylanilines and aryl halides.A broad range of the readily available starting materials can be efficiently and easily converted into the desired 1,2-disubstituted indole derivatives under the low-cost copper catalysis.Furthermore, the Cu(I)-mediated domino reactions with o-bromoiodoarenes selectively afforded the corresponding N-(o-bromoaryl)-indoles, which were further transformed to indolo[1,2-f]phenanthridine via Pd-catalyzed intramolecular direct C(sp 2 )-H arylation.The polycyclic indole derivatives were also smoothly assembled in one pot by sequential Cu/Pd catalysis.The high efficiency, the convenient procedures, and the wide application scope would make this protocol attractive for the assembly of heterocyclic molecules of biological or materials interest.

Experimental Section
General.Unless otherwise noted, all one-pot reactions were carried out in an over-dried Schlenk tube equipped with a magnetic stir bar under N 2 atmosphere.Toluene, o-xylene and dioxane were distilled from Na; DCE and DMF were distilled from CaH 2 .2-alkynylanilines 47 and aryl halides 48 were synthesized according to the known literature.All other reagents were obtained from commercial sources and used without further purification, unless stated otherwise.The NMR spectra were recorded in CDCl 3 on a 600 MHz instrument with TMS as internal standard.Recorded shifts were reported in parts per million (δ) downfield from TMS.Data are represented as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet, b = broad), coupling constant (J, Hz) and integration.TLC was carried out with 0.2 mm thick silica gel plates (GF254).Visualization was achieved by UV light.The columns were hand packed with silica gel 60 (150-200 mesh).Unknown products were additionally confirmed by HRMS.Mass spectra were obtained using ESI ionization.

Supplementary Material available
Experimental procedures, characterization data (for all the products), copies of 1 H and 13 C NMR spectra associated with this paper can be found in the online version.

Scheme 4 .
Scheme 4. Control experiments to probe the possible pathway of the one-pot indole synthesis.

Table 1 .
Optimization of the reaction conditions a b Isolated yield (%). c No ligand.d 1.2 equiv of t-BuOK was utilized as the base.