An ultrasound-based approach for the synthesis of indoles under Pd/C catalysis

1 Introduction

The indole ring is considered as one of the privileged structures not only in the area of medicinal chemistry/drug discovery (Kaushik et al., 2013; Parrino et al., 2015a,b; Carbone et al., 2014; Gordon et al., 2013; Sindac et al., 2013) but also in other research areas such as agro chemistry and material science (Sundberg, 1970; Brown, 1972). Examples of some of the bioactive molecules and drugs containing indole ring are shown in Fig. 1 (Campbell et al., 2004; Hart and Boardman, 1963; Gastpar et al., 1998; Flynn et al., 2002). These include a cyclooxygenase-2 inhibitor (A), anti-inflammatory drug indomethacin (B), a microtubulin inhibitor (C) and an anticancer agent (D). All these molecules contain a 2,3-disubstituted indole framework highlighting the pharmacological importance of this scaffold.

Close

Figure 1

Examples of indole based bioactive molecules and drugs.

Export to PPT

Due to the high interest in indole structures numerous methodologies have been developed for the construction of indole ring for over last 100 years (Joule, 2000). While the traditional approaches based on condensation and cyclization sequences are still being utilized widely for the synthesis of indole derivatives, the methodologies based on transition metal catalyzed C—C and C—N bond forming reactions have become powerful tools for this purpose (Zeni and Larock, 2004; Cacchi and Fabrizi, 2005; Humphrey and Kuethe, 2006; Ackermann, 2007; Kruger et al., 2008; Patil and Yamamoto, 2008; Barluenga et al., 2009; Palmisano et al., 2010; Patil et al., 2011; Cacchi et al., 2011). Indeed, the alkyne based strategies catalyzed by palladium complexes or salts (Scheme 1) have attracted particular attention due to the simple one-pot procedure, wide substrate scope and functional group tolerance. Thus, the Larock indole synthesis reported in 1991 became an attractive and popular method for the preparation of 2,3-disubstituted indoles (Larock and Yum, 1991; Larock et al., 1998) which involved a Pd-catalyzed heteroannulation of internal alkynes with N-protected o-iodoanilines (cf path a, Scheme 1). A combination of Pd(OAc)₂, LiCl, and K₂CO₃ was used for this purpose. While being effective not only for the synthesis of indoles but also for a range of other similar N-heterocycles e.g. tryptophan derivatives (Jeschke et al., 1993), azaindoles (Park et al., 1998; Xu et al., 1998; Ujjainwalla and Warner, 1998; Roschangar et al., 2008; Koolman et al., 2009) the methodology required the use of an expensive Pd catalyst, longer reaction time and harsh reaction conditions. Due to the great variety of bioactive indole derivatives (Pal et al., 2004a,b; Rao et al., 2011; Nakhi et al., 2011; Kumar et al., 2012; Gorja et al., 2013; Dulla et al., 2014; Parrino et al., 2015; Carbone et al., 2014; Diana et al., 2011a,b; Inman et al., 2012) we required a more convenient and direct access to a library of 2-substituted or 2,3-disubstituted indoles for our in-house pharmacological screen.

Close

Scheme 1

Reported synthesis of indoles via alkyne based strategies catalyzed by palladium complexes or salts.

Export to PPT

The use of Pd/C as a heterogenous catalyst for the hydrogenation reaction is known for over 100 years (Nishimura, 2001). However, its uses for various types of other organic reactions involving C—C, C—N and C—O bond formations have recently been explored (Pal, 2009; Monguchi et al., 2010). Indeed, as an air-stable, easily separable (from the product), recoverable and recyclable catalyst the use of Pd/C is attractive from the viewpoint of green and sustainable chemistry. Moreover, it is cheaper than other traditional Pd-complexes and salt and avoids the requirement of extra precautions for long storage. Like Pd/C-mediated hydrogenation reactions the Pd/C catalyzed C—C bond forming reactions are also expected to be scalable for industrial application purpose.

Being considered as an important step toward green chemistry the ultrasound assisted organic reactions on the other hand have received enormous importance in recent time. Indeed, the use of ultrasound in organic reactions is beneficial in terms of energy conservation and waste minimization compared to conventional heating. Moreover, compared to the traditional methods, the ultrasound assisted reactions offer advantages such as shorter reaction time, milder conditions, and higher yields of products (Li et al., 2005; Ratoarinoro et al., 1992). Thus, it is not surprising that ultrasound assisted reactions have emerged as a powerful technique in present day organic synthesis (Cravotto and Cintas, 2006). In view of considerable advantages associated with the use of both ultrasound and Pd/C we decided to explore the reaction of 2-iodosulfanilide (1) with internal alkynes (2) under ultrasound irradiation leading to the synthesis of a library of indoles (3) (Scheme 2).

Close

Scheme 2

Pd/C catalyzed synthesis of indoles under ultrasound irradiation.

Export to PPT

2 Material and methods

3 Results and discussion

To establish the optimum reaction conditions of the present ultrasound assisted indole synthesis we first examined the ultrasound assisted reaction of N-(2-iodophenyl)-4-methylbenzenesulfonamide (1a) with hex-3-yne (2a) in the presence of 10% Pd/C (5 mol%) and LiCl (1 equiv) under various conditions. Each of these reactions was performed in a sealed tube under closed condition as well as under ultrasound irradiation. Results of this study are summarized in Table 1. Initially, a number of solvents including EtOH, DMF, MeCN, 1,4-dioxane and toluene were examined (entries 1–5, Table 1) maintaining the reaction temperature at 80 °C for 6 h. Among all these solvents tested, DMF was found to be the best as the desired indole 3a was obtained in 81% yield (entry 6, Table 1). A trace amount of compound 4 was also obtained in this case due to the deiodination of 1a as a side reaction under the condition employed. Dehalogenation of ArX (when X = I, Br and Cl) in the presence of Pd/C is a well-known reaction in the literature. Nevertheless, we were delighted to isolate the indole 3a and continued our study. We then examined the effect of change of base, temperature and time on the yield of 3a. Accordingly, K₂CO₃ was used in place of NaHCO₃ when the yield of 3a was decreased (entry 7, Table 1). The decrease in reaction temperature and increase in time did not improve the yield (entry 8 and 9, Table 1). To assess the role of ultrasound in the present indole synthesis we performed the reaction in the absence of ultrasound. Though the reaction proceeded, the desired product 3a was isolated in 67% yield after 16 h. This clearly indicated that the reaction rate and the product yield were improved significantly by using ultrasound. The role of catalyst and additive was also examined. The reaction did not proceed in the absence of Pd/C (entry 11, Table 1) whereas lower yield of 3a was obtained when the reaction was performed in the absence of LiCl (entry 6 vs 12, Table 1). Overall, the reaction condition of entry 6 was found to be the best for the synthesis of indole 3a and was used for further study.

Table 1 Effect of reaction conditions on the coupling of 1a with 2a.^a

Entry	Solvent/base	Temp (°C)/Time (h)	% yieldb
			3a	4
1.	EtOH/Et3N	80/6	0	25
2.	DMF/Et3N	80/6	21	19
3.	MeCN/Et3N	80/6	10	27
4.	1,4-Dioxane/Et3N	80/6	8	20
5.	Toluene/Et3N	80/6	0	0
6.	DMF/NaHCO3	80/6	81	5
7.	DMF/K2CO3	80/6	59	0
8.	DMF/NaHCO3	60/6	43	0
9.	DMF/NaHCO3	80/8	80	7
10.	DMF/NaHCO3	80/16	57c	6
11.	DMF/NaHCO3	80/6	0d	0
12.	DMF/NaHCO3	80/6	62e	8

a All reactions were performed using the sulfanilide 1a (0.5 mmol) alkyne 2a (0.6 mmol), 10% Pd/C (5 mol%), LiCl (0.5 mmol) and a base (0.6 mmol) in a solvent (3 mL).

b Isolated yield.

c The reaction was performed in the absence of ultrasound.

d The reaction was performed in the absence of Pd/C.

e The reaction was performed in the absence of LiCl.

To expand the scope and generality of the present ultrasound assisted one-pot synthesis of indoles the optimized reaction condition was used to prepare a large number of indole derivatives. Thus a number of 2-iodosufanilides (1a-f) were allowed to react with a number of alkynes (2a-h) under the optimized reaction conditions. The reactions proceeded well in all these cases affording the desired and corresponding indole derivatives in good to acceptable yields (Table 2). Both symmetrical (e.g. 2a-c) and unsymmetrical alkynes (e.g. 2d and 2e) were used in this coupling cyclization reaction. Notably, the product yield was decreased when unsymmetrical alkynes were used (entries 4 and 5, Table 2) under the studied condition. This was due to the formation of regioisomeric product as a side product in these cases indicating the lack of high regioselectivity of this process. This problem appeared to be less significant when terminal alkynes e.g. 2f-h were used where the 2-substituted indoles were isolated as the main product. Both aryl alkynes (e.g. 2f and 2g) and alkyl alkyne (e.g. 2h) were used and found to be effective. However, the use of methyl propiolate (HC≡CCO₂Me) was not successful under this reaction conditions as the alkyne underwent rapid polymerization leading to the formation of tar like material. An attempt to use propiolic acid and propiolamide was also not successful indicating ineffectiveness of the activated terminal alkynes in the present reaction. While the product yields were not particularly high in cases where terminal alkynes e.g. 2f-h were used (entries 6–12, Table 2) (due to the partial dimerization of terminal alkynes used during the reaction), the formation of traces quantity of regioisomeric product could not be ruled out completely. Nevertheless, it was satisfying to note that both internal and certain terminal alkynes were effective in these reactions though the high regioselectivity was not achieved. Notably, a general method involving the use of both internal and terminal alkyne is not common in the literature. The presence of substituents such as Cl, F, Me, COMe and CO₂Me on the sulfanilide ring e.g. 1b-f was also examined and found to be well tolerated under the reaction conditions employed. All the indole derivatives prepared were well characterized by spectral (¹HNMR, ¹³CNMR and MS) data. The structures of regioisomers were also established via isolation and spectral characterization of the individual isomers followed by comparison with the reported data. For example the Me group at indole C-3 position of 3d appeared as a singlet at δ 2.02 (s, 3H) and 9.4 ppm in the ¹H and ¹³CNMR spectra respectively (Fig. 2). However, Me group at indole C-2 position of the regioisomer 3dd appeared at a more downfield region i.e. at δ 2.35 (s, 3H) and 13.5 ppm in the ¹H and ¹³CNMR spectra respectively (Fig. 2). This is because the indole C-2 Me group of 3dd being at position next to the indole nitrogen is more deshielded than that of 3d. Moreover, as the regioisomer 3dd is a known compound (Furstner et al., 1994) its spectral data were compared with that reported in the literature and a good correlation was observed. A similar NMR data comparison and analysis can be performed for compound 3e and its regioisomer 3ee with respect to the CH₂ group (of butyl side chain) attached to the indole ring (Fig. 3).

Table 2 Ultrasound assisted synthesis of indoles under Pd/C catalysis.^a

Entry	Iodide 1; Z	Alkyne 2; R1 & R2	Product 3	%yieldb
1.	1a; H	2a; Et & Et		81
2.	1a	2b; n-Bu & n-Bu		76
3.	1a	2c; n-Pentyl & n-Pentyl		77
4.	1a	2d; Me & Ph		60
5.	1a	2e; n-Bu & Ph		63
6.	1a	2f; H & C6H4Me-m		62
7.	1a	2g; H & 6-methoxynaphthalen-2-yl		59
8.	1a	2h; H & CMe3		60
9.	1b; Cl	2h		63
10.	1c; F	2h		61
11.	1d; Me	2h		63
12.	1e; COMe	2h		60
13.	1f; CO2Me	2c		80

a The reaction was performed using the iodo compound 1 (0.5 mmol), alkyne 2 (0.6 mmol), 10% Pd/C (5 mol%), LiCl (0.5 mmol) and NaHCO₃ (0.6 mmol) in DMF (3 mL).

b Isolated yield.

Close

Figure 2

Comparison of ¹H and ¹³C NMR data (partial) of indole 3d and its regioisomer 3dd.

Export to PPT

Close

Figure 3

Comparison of ¹H and ¹³C NMR data (partial) of indole 3e and its regioisomer 3ee.

Export to PPT

Based on the pathway proposed by Larock and Yum (1991) and Larock et al. (1998) for the synthesis of 2,3-disubstituted indoles from the internal alkynes, a reaction mechanism is presented for the Pd/C-mediated synthesis of 3 under ultrasound (Scheme 3). The reaction appeared to proceed via (i) Pd leaching in solution which on coordination with DMF (Griffith et al., 1989; Khain and Val’kova, 1978) and chloride anion generated the actual catalytic species, (ii) oxidative addition of chloride ligated Pd(0) to the iodoarene (1) affording the organo-Pd species E-1, (iii) coordination of Pd of E-1 with the alkyne 2 to give E-2, (iv) syn addition of the Pd-aryl bond to the coordinated alkyne followed by the halide displacement on the Pd of the resulting vinyl-Pd intermediate by the proximate sulfonamide nitrogen in the presence of base to give E-3 and finally (iv) reductive elimination of Pd(0) to give the desired product 3. Thus an initial Pd leaching process (Chen et al., 2007) from the minor portion of the bound palladium (Pd/C) into the solution was facilitated by ultrasound radiation. The Pd-leaching process in the present reaction was indicated by the following experiment: 10% Pd/C (5 mol%), LiCl (0.5 mmol) and NaHCO₃ (0.6 mmol) in DMF (3 mL) (all mmol amount was calculated based on 0.5 mmol of 1a) stirred at 80 °C under ultrasound irradiation for 1 h. The mixture was then filtered while hot and the filtrate was used for a new reaction by adding 1a (0.5 mmol) and 2a (0.6 mmol) without any additional catalyst. The mixture was then stirred at 80 °C under ultrasound irradiation for 6 h when only 50% conversion was observed. The low conversion perhaps was due to the partial removal of NaHCO₃ from the reaction mixture during the filtration step. Thus a separate experiment was carried out by adding additional NaHCO₃ (0.6 mmol) along with 1a and 2a to the reaction mixture after filtration when ∼80% conversion was observed. The results of Table 1 clearly indicated that ultrasound played an important role in the present reaction. Perhaps it accelerated the Pd leaching process greatly thereby facilitating the subsequent steps. The ultrasound might have also facilitated the rapid formation of intermediate E-3 followed by reductive elimination of Pd(0) to give the indole 3. Overall, both catalyst and ultrasound played key roles in completing the reaction within 6–8 h. The observed regiochemistry in the present reaction can also be explained according to the theory proposed by Larock et al. (1998). Thus the orientation of the approaching alkyne in E-2 (Scheme 3) would be such that the bulkier group remained away from the sterically encumbered aryl moiety thereby causing less steric strain in the vinylic palladium intermediate E-3 (Scheme 3 and Fig. 4).

Close

Scheme 3

Proposed reaction mechanism for the Pd/C catalyzed synthesis of indoles (3) under ultrasound irradiation.

Export to PPT

Close

Figure 4

Favored form of vinylic palladium intermediate E-3.

Export to PPT

4 Conclusion

In conclusion, a faster and convenient method has been developed for the Pd-mediated synthesis of indole derivatives in good to acceptable yields. The methodology involved Pd/C-catalyzed coupling–cyclization of 2-iodosulfanilide with alkynes under ultrasound irradiation in the presence of LiCl and NaHCO₃ in DMF. A variety of internal/terminal alkynes was employed in this C—C followed by C—N bond forming reaction to afford the corresponding indoles (3). A proposed reaction mechanism for the Pd/C catalyzed synthesis of indoles under ultrasound irradiation is presented and discussed. Being faster and milder method the present ultrasound-based approach appeared to be an useful and cheaper alternative to the existing method. The methodology therefore may be used for constructing diversity based library of small molecules based on indole framework of potential medicinal value.