ZrCl4-promoted facile synthesis of indole derivatives

J. Tummatorn*a, M. P. Gleesonb, S. Krajangsria, C. Thongsornkleebac and S. Ruchirawatac
aChulabhorn Research Institute, 54 Kamphaeng Phet 6, Laksi, Bangkok 10210, Thailand. E-mail: jumreang@cri.or.th; Fax: +66 2 5538545
bDepartment of Chemistry, Faculty of Science, Kasetsart University, 50 Phahonyothin Road, Bangkok, Thailand 10900
cProgram on Chemical Biology, Chulabhorn Graduate Institute, Center of Excellence on Environmental Health and Toxicology, CHE, Ministry of Education, 54 Kamphaeng Phet 6, Laksi, Bangkok 10210, Thailand

Received 12th March 2014 , Accepted 17th April 2014

First published on 18th April 2014

Zirconium(IV) chloride effectively activates nitrogen (N2) extrusion from aryl azidoacrylates followed by annulation to provide the desired indole products in moderate to good yields. The reaction proceeds at low temperature and in short reaction time and is applicable to a variety of substrates.

Methodologies for the synthesis of indole derivatives have been continuously developed over the past several decades due to both naturally-occurring and synthetic indoles exhibiting a broad range of biological activities.1 Several strategies for indole construction have been reported including the synthesis of indole-2-carboxylate derivatives2 of which aryl azidoacrylates have been among the best precursors. Typically, the aryl azidoacrylate derivatives can be converted into indole-2-carboxylates using the Hemetsberger reaction which requires high temperature.3 Recently, Driver has developed a new catalyst, rhodium(II) perfluorobutyrate, which can activate aryl azidoacrylates for the formation of indole-2-carboxylates via C–H amination at relatively low temperatures (room temperature to 60 °C).4 Subsequently, commercially available iron(II) triflate was also reported as an effective catalyst for this transformation at 80 °C.5 However, both catalysts still require high reaction temperatures and long reaction times. Herein, we report for the first time that ZrCl4 is an effective reagent for such transformation at room temperature or lower in less than 30 minutes. We commenced the investigation using the unactivated phenyl azidoacrylate (1a) as a screening substrate and started our study with several Lewis acids including BF3·OEt2, AlCl3, In(OTf)3, Yb(OTf)3, HfCl4 and ZrCl4. We found that only ZrCl4 and HfCl4 were promising as the reactions went to completion within a few minutes at room temperature and provided the desired product (2a) in acceptable yields (entries 2–3, Table 1). Among metal catalysts reported to effect such a transformation, ZrCl4 is much less expensive, hence more economically advantageous. We then decided to select ZrCl4 as the reagent for this transformation.6 In our optimization process, we encountered a competitive reaction which resulted in the formation of by-product 3a in a significant amount. To reduce the formation of by-product 3a, a lower temperature (0 °C, entry 4) was employed and found to improve the yield to 46%. Due to the polymeric structure with octahedral coordination geometry around the metal centers,7 ZrCl4 and HfCl4 could not be dissolved in DCE and, as a result, the corresponding reactions were heterogeneous. To improve the efficiency of ZrCl4, we envisioned that breaking up the polymeric structure of ZrCl4 might enhance the efficiency of this transformation. As reported in the literature8 that AgNO3 could remove the chloride ligand from a metal center complex, we thought that the silver salt could abstract chloride ligand from ZrCl4 resulting in depolymerization of its polymeric structure into short chains thereby increasing the reactivity of ZrCl4 in the reaction. Therefore 0.1 equivalent of AgNO3 was added to the reaction mixture (Table 1, entry 5). Unfortunately, the result was an incomplete reaction after 20 min as monitored by TLC. When this reaction was allowed to proceed longer (60 min), side reactions were observed and we obtained only 28% yield of the desired indole. Then we attempted to enhance the conversion by increasing the amount of ZrCl4 to 1.4 equivalents and we found the reaction went to completion within 20 min at 0 °C to provide the desired product in 70% yield (entry 6). However, when 1.0 equivalent of AgNO3 was employed, the reaction provided a complex mixture (entry 7). In this entry, we initially observed the formation of the desired product by TLC but this product gradually disappeared from the decomposition under the reaction conditions. We believed this was due to the increased amount of AgNO3 which led to the excess amount of short chain ZrCl4 species from the depolymerization resulting in disintegration of the indole product. Furthermore, we also examined the reaction using a catalytic amount of ZrCl4 (0.1 equivalent) and AgNO3 (0.2 equivalent) (entry 8). Only trace amount of indole product 2a was observed attributing to incomplete conversion of the azidoacrylate precursor 1a. In addition, no conversion of compound 1a was obtained when ZrCl4 was excluded in the reaction mixture (entry 9). Therefore, stoichiometric amount of ZrCl4 was needed for this transformation.9 Additionally, the use of different additives such as Ag2O, KOAc, and K2CO3 (entries 10–12) were also explored. Unfortunately, these proved detrimental to the reactions, providing only mixtures of unidentifiable by-products. We also investigated the effect of solvents and found DCE to be well-suited for this transformation while there was no reaction in CH3CN, DCM, toluene or THF (entries 13–16). We therefore used 1.4 equivalent of ZrCl4 and 0.1 equivalent of AgNO3 in DCE as our optimal conditions.
Table 1 Optimization of the synthesis of indole-2-carboxylate

image file: c4ra02158d-u1.tif

Entry Reagent/additive (eq.) Solvent Time (min) Temp (°C) Yield of 2a (%)a
a Isolated yield.b 1 eq. of BF3·OEt2 = NR (rt, 6 h), 1 eq. of AlCl3 = complex mixture (rt, 30 min), 10 mol% of In(OTf)3 = 14% (reflux, O/N), 20 mol% of Yb(OTf)3 = trace (reflux, O/N).
1 Lewis acids DCE b b 0–14
2 HfCl4 (1.0) DCE 1 rt 42
3 ZrCl4 (1.1) DCE 1 rt 40
4 ZrCl4 (1.1) DCE 5 0 46
5 ZrCl4/AgNO3 (1.1[thin space (1/6-em)]:[thin space (1/6-em)]0.1) DCE 60 0 28
6 ZrCl4/AgNO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) DCE 20 0 70
7 ZrCl4/AgNO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]1.0) DCE 10 0
8 ZrCl4/AgNO3 (0.1[thin space (1/6-em)]:[thin space (1/6-em)]0.2) DCE 30 0 Trace
9 ZrCl4/Ag2O (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) DCE 20 0
10 ZrCl4/KOAc (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) DCE 30 0
11 ZrCl4/K2CO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) DCE 30 0
12 AgNO3 (1.0 eq.) DCE 30 rt NR
13 ZrCl4/AgNO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) CH3CN 25 0 NR
14 ZrCl4/AgNO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) DCM 25 0 NR
15 ZrCl4/AgNO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) PhCH3 25 0 NR
16 ZrCl4/AgNO3 (1.4[thin space (1/6-em)]:[thin space (1/6-em)]0.1) THF 25 0 NR

To study the scope of this reaction, we then applied the optimal conditions to a variety of aryl azidoacrylate derivatives and the results are summarized in Table 2. o-Tolyl azidoacrylate and o-ethylphenyl azidoacrylate (entries 2–3) provided the desired products in good yields. The reaction of p-tert-butylphenyl azidoacrylate (1d, entry 4) also proceeded well under these conditions and furnished the corresponding indole in 69% yield. Unfortunately, when p-biphenyl azidoacrylate was subjected to this transformation, the yield dropped to 44% and we observed the formation of by-product 3e in a significant amount (42%). The results implied that the p-phenyl substituent increased the electrophilicity of the benzylic position of carbocation intermediate 5 (Scheme 1) which induced the nucleophilic attack by a chlorine atom from the zirconium complex. α-Naphthyl azidoacrylate was successfully converted to indole 2f in 80% yield, whereas β-naphthyl azidoacrylate gave product 2g in lower yield (48%) as the only isolable regioisomer. In the latter case, we observed a purple-colored reaction which we suspected was a sign of decomposition. p-Halogen-substituted phenyl azidoacrylates (entries 8–10) were effectively transformed to indole products in good efficiency. By contrast, when a substrate containing an electron-withdrawing nitro group (entry 11) was attempted, a low conversion was observed and the corresponding product was obtained in low yield. This could be attributed to the strong electron-withdrawing effect of the p-nitro group which prevented the nitrogen extrusion. Subsequently, o-halogen-substituted phenyl azidoacrylates (entries 12–13) were examined and found to provide the corresponding indoles in good yields. In the cases of m-chlorophenyl azidoacrylate and m-fluoro-p-tolyl azidoacrylate (entries 14–15), we obtained a mixture of regioisomeric products in moderate to good overall yields. The products were obtained in approximately a 1[thin space (1/6-em)]:[thin space (1/6-em)]1 ratio for the chloro-substituted indoles whereas in the fluoro-substituted case, the desired products were obtained in a 2[thin space (1/6-em)]:[thin space (1/6-em)]1 ratio. Interestingly, m-methoxyphenyl azidoacrylate (entry 16) provided a single regioisomeric product. In this entry, the purple-colored reaction mixture was also observed and the reaction produced the desired product in moderate yield. The reaction of polysubstituted phenyl azidoacrylate (entry 17) proceeded well, affording the indole product in excellent yield. From these results, we can draw a trend that aryl azidoacrylates p-substituted with electron-withdrawing groups such as phenyl and nitro groups (entries 5 and 11), led to the formation of the indole product in low yields whereas substrates containing electron-donating substituents in the para-position (t-Bu, Br, and OCH3; entries 4, 8, and 17, respectively) could provide the desired indole products in higher yields.

Table 2 Scope of aryl azidoacrylates

image file: c4ra02158d-u2.tif

Entry Substrate Product Temp/Time Yielda (%)
a Isolated yield.b Combined yield of separable regioisomeric products; 2nX = 31%, 2nY = 30%, 2oX = 38%, and 2oY = 18%.
1 image file: c4ra02158d-u3.tif image file: c4ra02158d-u4.tif 0 °C (20 min) 70
2 image file: c4ra02158d-u5.tif image file: c4ra02158d-u6.tif 15 °C (3 min) 75
3 image file: c4ra02158d-u7.tif image file: c4ra02158d-u8.tif 15 °C (10 min) 69
4 image file: c4ra02158d-u9.tif image file: c4ra02158d-u10.tif 15 °C (10 min) 69
5 image file: c4ra02158d-u11.tif image file: c4ra02158d-u12.tif 0 °C → rt (5 min) 44
6 image file: c4ra02158d-u13.tif image file: c4ra02158d-u14.tif 0 °C (15 min) 80
7 image file: c4ra02158d-u15.tif image file: c4ra02158d-u16.tif 0 °C → rt (5 min) 48
8b image file: c4ra02158d-u17.tif image file: c4ra02158d-u18.tif 0 °C → rt (10 min) 79
9 image file: c4ra02158d-u19.tif image file: c4ra02158d-u20.tif 0 °C (10 min) 71
10 image file: c4ra02158d-u21.tif image file: c4ra02158d-u22.tif 0 °C (10 min) 65
11 image file: c4ra02158d-u23.tif image file: c4ra02158d-u24.tif rt (30 min) 22
12 image file: c4ra02158d-u25.tif image file: c4ra02158d-u26.tif 0 °C → rt (30 min) 63
13 image file: c4ra02158d-u27.tif image file: c4ra02158d-u28.tif 0 °C → rt (10 min) 63
14 image file: c4ra02158d-u29.tif image file: c4ra02158d-u30.tif 0 °C → rt (10 min) 61b
15 image file: c4ra02158d-u31.tif image file: c4ra02158d-u32.tif 15 °C (20 min) 56b
16 image file: c4ra02158d-u33.tif image file: c4ra02158d-u34.tif 0 °C (10 min) 54
17 image file: c4ra02158d-u35.tif image file: c4ra02158d-u36.tif 15 °C (1 min) 88

image file: c4ra02158d-s1.tif
Scheme 1 Proposed mechanism for the formation of indole 2 and by-product 3.

The proposed mechanism for the formation of the indole 2a and by-product 3a is shown in Scheme 1. Since Zr(IV) are highly oxophilic which can form a stable metal–oxygen complex7,10 and also have a good chelation with the nitrogen donor11 thus the coordinations of ZrCl4 with the nitrogen and carbonyl group of the aryl azidoacrylate occurred in the initial step to generate the hexacoordinate zirconium complex 4.12 This complex subsequently undergoes a slight rearrangement of bonds and a nitrogen extrusion to provide zirconium complex 5. This is the important key intermediate which could undergo two competing reaction pathways to form the indole product 2a (pathway a) or chlorinated-product 3a (pathway b). In pathway a, an intramolecular nucleophilic aromatic substitution-type reaction of complex 5 provides the cyclization intermediate complex 6 which rearomatizes and hydrolyzes upon workup to the desired indole 2a. For the formation of chlorinated-product 3a, the carbocation complex 5 could undergo an intramolecular chlorine atom transfer (pathway b) to form the pentacoordinate zirconium complex 8.13 This complex is then further hydrolyzed during workup to give by-product 3a.14 DFT calculations were performed to understand the reaction pathways of these transformation as shown in Fig. 1. The conversion of intermediate 5 to indole 2a is thermodynamically more favorable (pathway a) than the conversion to compound 3a via the chlorine atom transfer (pathway b) (Fig. 1). In addition, the activation energies of pathway b is significantly larger than pathway a (TS4–TS2 = 6.2 kcal mol−1), which explains the exclusive formation of indole 2a.

image file: c4ra02158d-f1.tif
Fig. 1 The calculated DFT energy profiles15 associated for the two mechanisms proposed in Scheme 1 (see ESI for further details).


In conclusion, a cost-effective synthesis of indole derivatives is presented. We have established the use of commercially available and inexpensive zirconium(IV) chloride as a novel reagent to promote the nitrogen extrusion of aryl azidoacrylates followed by an annulation process at low temperature and in short reaction time to give the desired indole products in moderate to good yields. This provides a fast and mild synthetic protocol for the synthesis of indole derivatives which is applicable to a variety of substrates.


This research work was supported in part by grants from Center of Excellence on Environmental Health and Toxicology, Science & Technology Postgraduate Education and Research Development Office (PERDO), Ministry of Education, and Chulabhorn Research Institute.

Notes and references

  1. (a) A. Arcadi, G. Bianchi and F. Marinelli, Synthesis, 2004, 610–618 CAS; (b) Y. Du, R. Liu, G. Linn and K. Zhao, Org. Lett., 2006, 8, 5919–5922 CrossRef CAS PubMed; (c) G. A. Kraus and H. Guo, Org. Lett., 2008, 10, 3061–3063 CrossRef CAS PubMed; (d) N. Sakai, K. Annaka, A. Fujita, A. Sato and T. Konakahara, J. Org. Chem., 2008, 73, 4160–4165 CrossRef CAS PubMed; (e) Z. He, H. Li and Z. Li, J. Org. Chem., 2010, 75, 4296–4299 CrossRef PubMed; (f) Q. Yan, J. Luo, D. Zhang-Negrerie, H. Li, X. Qi and K. Zhao, J. Org. Chem., 2011, 76, 8690–8697 CrossRef CAS PubMed; (g) Y. Wei, I. Deb and N. Yoshikai, J. Am. Chem. Soc., 2012, 134, 9098–9101 CrossRef CAS PubMed; (h) S. Gore, S. Baskaran and B. König, Org. Lett., 2012, 14, 4568–4571 CrossRef CAS PubMed; (i) M. Inman and C. J. Moody, J. Org. Chem., 2010, 75, 6023–6026 CrossRef CAS PubMed; (j) M. Inman and C. J. Moody, Chem. Sci., 2013, 4, 29–41 RSC; (k) Y. Zhang, S. Liu, W. Yu, M. Hu, G. Zhang and Y. Yu, Tetrahedron, 2013, 69, 2070–2074 CrossRef CAS PubMed.
  2. (a) D. J. Bentley, J. Fairhurst, P. T. Gallagher, A. K. Manteuffel, C. J. Moody and J. L. Pinder, Org. Biomol. Chem., 2004, 2, 701–708 RSC; (b) M. Inman and C. J. Moody, Chem. Commun., 2011, 47, 788–790 RSC; (c) M. Inman, A. Carbone and C. J. Moody, J. Org. Chem., 2012, 77, 1217–1232 CrossRef CAS PubMed.
  3. (a) H. Hemetsberger and D. Knittel, Monatsh. Chem., 1972, 103, 194–204 CrossRef CAS; (b) A. G. O'Brien, F. Levesque and P. H. Seeberger, Chem. Commun., 2011, 47, 2688–2690 RSC.
  4. B. J. Stokes, H. Dong, B. E. Leslie, A. L. Pumphrey and T. G. Driver, J. Am. Chem. Soc., 2007, 129, 7500–7501 CrossRef CAS PubMed.
  5. J. Bonnamour and C. Bolm, Org. Lett., 2011, 13, 2012–2014 CrossRef CAS PubMed.
  6. At the time of submission of this work, the prices of metal catalysts mentioned from Strem Chemicals, Inc. were as followed: ZrCl4 $82.00 (250 g), AgNO3 $42.00 (10 g), HfCl4 $299.00 (250 g) and Fe(OTf)2 $264.00 (100 g).
  7. N. N. Greenwood and A. Earnshaw, in Chemistry of the Elements, Butterworth-Heinemann, Oxford, 2nd edn, 1997 Search PubMed.
  8. J. Karthikeyan, R. Haridharan and C.-H. Cheng, Angew. Chem., Int. Ed., 2012, 51, 12343–12347 CrossRef CAS PubMed.
  9. (a) S. K. Guchait, M. Kashyap and H. Kamble, J. Org. Chem., 2011, 76, 4753–4758 CrossRef PubMed; (b) R. Lenarŝič, M. Kočevar and S. Polanc, J. Org. Chem., 1999, 64, 2558–2563 CrossRef.
  10. E. Samuel, J. F. Harrod, M. J. McGlinchey, C. Cabestaing and F. Robert, Inorg. Chem., 1994, 33, 1292–1296 CrossRef CAS.
  11. (a) C. Janiak, K. C. H. Lange and T. G. Scharmann, Appl. Organomet. Chem., 2000, 14, 316–324 CrossRef CAS; (b) M. Banerjee, R. W. Seidel, M. Winter, H. Becker, D. Rogalla and A. Devi, Dalton Trans., 2014, 43, 2384 RSC.
  12. (a) S. A. Scholl, G. T. Plundrich, H. Wadepohl and L. H. Gade, Inorg. Chem., 2013, 52, 10158–10166 CrossRef CAS PubMed; (b) T. Gehrmann, J. Lloret-Fillol, H. Herrmann, H. Wadepohl and L. H. Gade, Organometallics, 2013, 32, 3877–3889 CrossRef CAS; (c) S. A. Scholl, H. Wadepohl and L. H. Gade, Organometallics, 2013, 32, 937–940 CrossRef CAS; (d) S. Dagorne, S. Bellemin-Laponnaz and C. Romain, Organometallics, 2013, 32, 2736–2743 CrossRef CAS.
  13. (a) K. Manna, W. C. Everett, G. Schoendorff, A. Ellern, T. L. Windus and A. D. Sadow, J. Am. Chem. Soc., 2013, 135, 7235–7250 CrossRef CAS PubMed; (b) E. Y. Tshuva, I. Goldberg, M. Kol, H. Weitman and Z. Goldschmidt, Chem. Commun., 2000, 379–380 RSC.
  14. (a) V. A. Mamedov, E. A. Berdnikov, S. Tsuboi, H. Hamamoto, T. Komiyama, E. A. Gorbunova, A. T. Gubaidullin and I. A. Litvinova, Russ. Chem. Bull., 2006, 55, 1402–1410 Search PubMed; (b) S. Tsuboi, H. Furutani, A. Takeda, K. Kawazoe and S. Sato, Bull. Chem. Soc. Jpn., 1987, 60, 2475–2479 CrossRef CAS.
  15. Quantum chemical calculations were performed using the M06 density functional theory (DFT) method, with the LANL2DZ basis set in the Gaussian 09 program. Minima and transition states were confirmed as stationary points using the results from frequency calculations. Minima showed positive frequencies while transitions states showed a single negative frequency corresponding to the expected reaction coordinates. For more details, see the ESI..


Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra02158d

This journal is © The Royal Society of Chemistry 2014