Silver-catalyzed furoquinolines synthesis: from nitrogen effects to the use of silver imidazolate polymer as a new and robust silver catalyst

Abstract

Silver-catalyzed tandem acetalization and cycloisomerization reactions were found to lead to various furoquinolines, and a nitrogen effect was noticed for AgOTf reactivity, since the cyclization mode switched from 6-endo-dig to 5-exo-dig; from these observations silver imidazolate polymer is proposed as a stable silver catalyst.

---

Of the coinage metals (Cu, Ag and Au), silver has received less attention as a catalyst,¹ since it has been mostly used in stoichiometric quantities in oxidation reactions, as an halogen scavenger, or for anion metathesis reactions (counterion exchange). Our main focus is directed towards silver-catalyzed cycloisomerization reactions.² However, based on Marshall’s work³ on allenyl ketones/aldehydes, only a few groups have contributed to the development of silver-catalyzed cycloisomerization reactions involving carbonyl groups and alkynyl moieties.⁴

We recently reported a versatile access to pyranoquinoline and furoquinoline cores^4d and found that the nature of the silver salt used (AgOTfvs. Ag₂O) could force the regioselectivity (6-endo-dig vs. 5-exo-dig) of the cycloisomerization reaction. During further experiments intended to study the role of traces of bases or acids in the reaction outcome, we found an interesting nitrogen effect (Table 1): the presence of a nitrogen group in the R or R¹ substituents or in the reaction media (as an additive) had no impact on Ag₂O catalysis, whereas it changed the behaviour of AgOTf dramatically. Indeed, compound 1, bearing a methoxymethyl substituent on the alkynyl group (entry 1, Table 1), gave exclusively selective 6-endo-dig cyclization, whereas the 5-exo-dig cyclization mode occurred preferentially for compound 2 (pyridinyl group, entry 2 (Table 1), 5-exo : 6-endo 85 : 15) or compound 3 (–CH₂–NMeBn, entry 3 (Table 1), 100 : 0). From compound 1 (entry 4, Table 1), the use of an aminoalcohol instead of methanol (as solvent and nucleophile) afforded opposite cycloisomerization selectivity (5-exo-dig, 100%) to entry 1 (6-endo-dig, 100%). Moreover, reacting compound 1 in the presence of various additives provided interesting information (Table 1). Several amine derivatives (1 equiv.) were added to the reaction mixture – one of these, quinoline (entry 6, Table 1), did not have any effect on 6-endo-dig cycloisomerization selectivity compared to entry 1 (Table 1), which was not surprising, since our starting materials all bear this heterocyclic nucleus. But additives in entries 7 and 8 (Table 1) were used as mimics of compounds 2 and 3, since the pyridine or triethylamine is closely related to the alkynyl substituent in compounds 2 and 3. Although the conversion was moderate in the case of triethylamine (55%), both reactions led preferentially to the 5-exo-dig product (5-exo : 6-endo ratio, 75 : 25 for pyridine and 85 : 15 for triethylamine).

Table 1 Nitrogen effects: influence of reagents and additives^a,b

Entry	R	R^1–OH	Additive	pKa^c	Conversion (%)^d	Selectivity^d (5-exo : 6-endo)
a Entries 1–4: the reaction was carried out for 2 h at room temperature (r.t.) in MeOH (500 equiv.) or Et2N(CH2)2OH (150 equiv.). b Entries 1 and 5–11: the reaction was carried out for 21 h with 1 (0.05 M) in dichoroethane (DCE), with MeOH (1.2 equiv.), at r.t., with AgOTf (10 mol%) and additive (1 equiv.). c Determined in water for deprotonation of conjugated acid^5 (values in parentheses were determined in acetonitrile^6). d Determined by ^1H NMR analysis. e DABCO = 1,4-diazabicyclo[2.2.2]octane; DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene; TTBP = 2,4,6-tri-tert-butylpyrimidine. f Instead of AgOTf catalysis, this reaction was performed with 10 mol% of triflic acid (TfOH).
1	1	MeOH	—	—	100 (2 h or 21 h)	0  : 100
2	2	MeOH	—	—	100	85 : 15
3	3	MeOH	—	—	100	100 : 0
4	1	Et2N(CH2)2OH	—	—	100	100 : 0
5	1	MeOH	DABCO ^e	8.82	0	—
6	1	MeOH	Quinoline	4.92	95	0 : 100
7	1	MeOH	Pyridine	5.21 (12.53)	100	75 : 25
8	1	MeOH	Et3N	10.75 (18.82)	55	85 : 15
9	1	MeOH	DBU ^e	(24.34)	10	90 : 10
10	1	MeOH	TTBP ^e	—	100	0 : 100
11	1	MeOH	TfOH ^f	—	5 (2 h), 50 (77 h)	0 : 100

---

Attempts at cyclization in the presence of DABCO (entry 5, Table 1) didn't afford any reaction. According to previously reported results,⁷ this could be due to insoluble polymeric [Ag_n-DABCO_m]_n, which would inhibit silver reactivity. Under the same reaction conditions DBU as an additive led to very low conversion (10%, entry 9, Table 1) but in this case some of the catalytic system formed⁸ could account for the reversal of selectivity (5-exo : 6-endo ratio, 90 : 10). Since traces of triflic acid (TfOH)⁹ could also explain the transformation studied here, the reaction was tested with and without a proton scavenger (TTBP,¹⁰ entries 1 vs. 10, Table 1). We were pleased to note by ¹H NMR analysis that the only product formed in both cases was the 6-endo-dig pyranoquinoline derivative, ruling out any possibility of Brønsted acid catalysis (TTBP is too hindered for complexing silver). Additional experiments where AgOTf was replaced by 10 mol% TfOH (entry 11, Table 1), led only to 5% conversion in favor of 6-endo-dig compound (vs. 100% conversion for AgOTf in 2 h). A 50% conversion was reached after 77 h with TfOH (10 mol%). TTBP and TfOH results confirm the critical need for the silver cation and therefore exclude the hypothesis of Brønsted acid catalysis.

The steric difference around the nitrogen lone pair between quinoline and pyridine does not explain the huge selectivity scrambling; rather, the strength of the complexation seems to be the best explanation.¹¹ Another way to rationalize these results would be ranking by the additive amine pK_a values. Further comparison of pK_a values (entries 7–9, Table 1) show that the greater the pK_a is, the better selectivity the reaction in favour of the 5-exo-dig product.

Inspired by the nitrogen effect described above, and also knowing that the same selectivity (5-exo-dig) was seen with silver salts having a counterion pK_a greater than 10 (Ag₂O, AgO, Ag₂CO₃),^4d we envisioned a new “Ag–nitrogen” catalyst. A known silver imidazolate polymer ([Ag(Im)]_n) had the necessary requirements (pK_a = 14.52 for imidazolate compared with pK_a = 6.92 for imidazole).¹² This was isolated in 1964 by Bauman and Wang¹³ and fully characterized only thirty years later.¹⁴ Since then, this polymer has been reported to be useful for its antimicrobial activities,¹⁵ as a latent heat-activatable crosslinking catalyst for epoxy resins/coating, and as a stoichiometric promoter in glycosidation reactions.¹⁶ But, to our knowledge, no further catalytic properties have been reported. This polymer is stable to light and moisture, making it very useful, since the main drawback mostly encountered with silver catalysts are their high sensitivity to light and/or moisture. Moreover, a mixture of [Ag(Im)]_n with various quantities of PPh₃ forms soluble complexes that have been characterized,^14b,c but that have found no use in catalysis. Therefore, we studied the catalytic properties of [Ag(Im)]_n in parallel with Ag₂O (Table 2). Ag₂O-catalyzed tandem acetalization and cycloisomerization reactions with compound 4^4d in methanol (entry 1, Table 2) gave furoquinoline 5 accompanied by traces of 6-endo-dig compound 6 (structure confirmed by 2D ¹⁵N–¹H NMR and also X-ray analysis; data not shown).

Table 2 Silver imidazolate as a catalyst^a

Entry	Catalyst	Additive^b	Time	Conversion (%)^c	Selectivity^c (5-exo : 6-endo)
a The reactions were carried out at 0.05 M in MeOH, at room temperature. b 5 mol%. c Determined by ^1H NMR analysis.
1	Ag2O	—	30 min	100	100 : trace
2	[Ag(Im)]n	—	40 h	100	100 : 0
3	[Ag(Im)]n	PPh3	10 min	100	100 : 0
4	[Ag(Im)]n	P(OPh)3	10 min	100	100 : 0
5	—	PPh3	48 h	0	—
6	AgNO3	—	12 h	66	trace : 100

---

Under [Ag(Im)]_n heterogeneous catalysis, a longer reaction time was required (40 h (entry 2, Table 2), compared with 30 min for Ag₂O (entry 1, Table 2) but the reaction was cleaner and selective for the 5-exo-dig compound 5. Then, PPh₃ or P(OPh)₃ were used as additives in the reaction with [Ag(Im)]_n (entries 3 and 4, Table 2), which dramatically shortened the reaction time (only 10 min in both cases) and led to furoquinoline 5 as the only product. These additives helped to generate a soluble catalyst and therefore, under homogeneous conditions, the reaction proceeded faster. A reaction with PPh₃ and no silver catalyst (entry 5, Table 2) gave no results. More interestingly, using AgNO₃ (the starting material for [Ag(Im)]_n synthesis), the major compound formed was the 6-endo-dig derivative 6 (entry 6, Table 1), proving that [Ag(Im)]_n was not contaminated with residual AgNO₃.

Having the optimized conditions in hand, we examined the reactivity of AgOTf, Ag₂O and [Ag(Im)]_n/PPh₃ on a set of functionalized quinolines (Table 3). In order to compare furoquinoline synthesis by these different silver catalysts, the reaction was performed in 2-(diethylamino)ethanol.

Table 3 AgOTf, Ag₂O and [Ag(Im)]_n/PPh₃ compared catalysis^a

Entry	Product	[Ag] (time, yield)^b
a Reaction with quinoline substrate (0.05 M) in 2-(diethylamino)ethanol at r.t., 0.2 M for [Ag(Im)]n/PPh3. b Isolated yields. c ^1H NMR analysis.
1		AgOTf (12 h, 40%)
		Ag2O (1 h, 95%)
		[Ag(Im)]n/PPh3 (12 h, 44%)
2		AgOTf (19 h, 23% conv.^c)
		Ag2O (3 h, 99%)
		[Ag(Im)]n/PPh3 (18 h, 29%)
3		AgOTf (5 h, 74%)
		Ag2O (1.5 h, quant.)
		[Ag(Im)]n/PPh3 (0.5 h, 99%)
4		AgOTf (6 h, 40%)
		Ag2O (4 h, 86%)
		[Ag(Im)]n/PPh3 (12 h, 50%)
5		AgOTf (12 h, 20%)
		Ag2O (4 h, 86%)
		[Ag(Im)]n/PPh3 (12 h, 92%)
6		AgOTf (12 h, 51%)
		Ag2O (3 h, 93%)
		[Ag(Im)]n/PPh3 (0.5 h, quant.)
7		AgOTf (2 h, 92%)
		Ag2O (2 h, 95%)
		[Ag(Im)]n/PPh3 (1.5 h, 71%)
8		AgOTf (12 h, 70%)
		Ag2O (3 h, 90%)
		[Ag(Im)]n/PPh3 (24 h, trace)

---

AgOTf catalysis usually needed a longer reaction time (2–12 h) compared to Ag₂O (1–4 h), but [Ag(Im)]_n/PPh₃ showed a broader range (0.5–18 h; see Table 3). The range of yields is pretty wide for AgOTf (20–92%) and [Ag(Im)]_n/PPh₃ (29%–quantitative), but the mean yield is more informative – 93% for Ag₂O, 70% for [Ag(Im)]/PPh₃ and 51% for AgOTf. This analysis shows us that [Ag(Im)]_n/PPh₃ is overall the most promising, although the reactivity is not as high as for Ag₂O. On the other hand, handling [Ag(Im)]_n (and even the reaction medium, [Ag(Im)]_n/PPh₃) is more practical regarding moisture, light and storage. With [Ag(Im)]_n/PPh₃, furoquinolines 7 and 8 bearing alkyl substituents (entries 1–2, Table 3) were obtained in fair yields (44 and 29%), but results were worse with the o-methoxyphenyl derivative 14, which was detected only in trace amounts (entry 8, Table 3). Steric hindrance could explain this behaviour, since having an extra methylene unit such as in furoquinolines 9–11 increases all the yields (99%, 50% and 92%, respectively).

With [Ag(Im)]_n/PPh₃, furoquinolines 12 and 13 were efficiently synthesized (yields being quantitative and 71%, and reaction times 0.5 and 1.5 h, respectively).

For a mechanistic point of view, AgNO₃ and AgOTf (without additives) are well known for interacting with alkynesvia π-complexes, and the reactions proceed via a pyrylium intermediate^4a,17a,b leading to the 6-endo-dig product. Based on these results and previous work,^4d we propose a different mechanism pathway for the other silver catalysts (Scheme 1). The reactivity of [Ag(Im)]_n/PPh₃, AgOTf/amine and Ag₂O would favor direct coordination to the oxygen atom of the carbonyl function (intermediate A); the silver salt acts as a Lewis acid,^17c,d promoting the formation of acetalB, which on cyclization would give complex C (anti addition)^4d and after [Ag] release would yield the 5-exo-dig product.

Scheme 1 Plausible mechanism.

In conclusion, we have shed light on the interesting behaviour of AgOTf with some amine additives – namely, an inversion of the cycloisomerization regioselectivity, from 6-endo-dig to 5-exo-dig. Driven by this observation, we have shifted [Ag(Im)]_n from its use in material sciences to organometallic catalysis. Indeed, this polymer is particularly attractive, since it is resistant to air exposure, moisture and light, thus distinguishing it from other silver derivatives. The use of triphenylphosphine for the homogeneous reaction is of interest and will be investigated for an enantioselective version.

We would like to thank the French Ministry of Research, INCa, CNRS, and the “Cluster Recherche Chimie de la Région Rhône-Alpes”. Thanks also to D. Bouchu, N. Henriques and C. Duchamp for mass spectrometry.

Notes and references

1. (a) B. H. Lipshutz and Y. Yamamoto, Chem. Rev., 2008, 108, 2793; (b) M. Naodovic and H. Yamamoto, Chem. Rev., 2008, 108, 3132; (c) J.-M. Weibel, A. Blanc and P. Pale, Chem. Rev., 2008, 108, 3149; (d) M. Álvarez-Corral, M. Muñoz-Dorado and I. Rodríguez-Garcia, Chem. Rev., 2008, 108, 3174; (e) Y. Yamamoto, Chem. Rev., 2008, 108, 3199–3222; (f) H. V. R. Dias and C. J. Lovely, Chem. Rev., 2008, 108, 3223; (g) N. T. Patil and Y. Yamamoto, Chem. Rev., 2008, 108, 3395.
2. P. Belmont and E. Parker, Eur. J. Org. Chem., 2009, 6075.
3. J. A. Marshall and E. D. Robinson, J. Org. Chem., 1990, 55, 3450.
4. (a) J. Zhu, A. R. Germain and J. A. Porco, Jr., Angew. Chem., Int. Ed., 2004, 43, 1239; (b) T. Yao, X. Zhang and R. C. Larock, J. Am. Chem. Soc., 2004, 126, 11164; (c) N. T. Patil, N. K. Pahadi and Y. Yamamoto, J. Org. Chem., 2005, 70, 10096; (d) T. Godet, C. Vaxelaire, C. Michel, A. Milet and P. Belmont, Chem.–Eur. J., 2007, 13, 5632; (e) I. Cikotiene, R. Buksnaitiene, S. Rudys, M. Morkunas and D. Motiejaitis, Tetrahedron, 2010, 66, 251; (f) X. Yu, Q. Ding, W. Wang and J. Wu, Tetrahedron Lett., 2008, 49, 4390.
5. M. B. Smith and J. March, March's Advanced Organic Chemistry Reactions, Mechanisms and Structure, 5th edn, 2000.
6. I. Kaljurand, I. Kütt, L. Sooväli, T. Rodima, V. Mäemets, I. Leito and I. A. Koppel, J. Org. Chem., 2005, 70, 1019.
7. (a) H. Krishna, S. S. Krishnamurthy and M. Nethaji, Inorg. Chim. Acta, 2009, 362, 38; (b) Y. Qu, L. Su and J. Wu, J. Chem. Crystallogr., 2007, 37, 579; (c) J. T. Sampanthar and J. J. Vittal, Cryst. Eng., 2000, 3, 117.
8. (a) C. Qi, L. Huang and H. Jiang, Synthesis, 2010, 1433; (b) W. Yamada, Y. Sugawara, H. M. Cheng, T. Ikeno and T. Yamada, Eur. J. Org. Chem., 2007, 2604; (c) Y. Sugawara, W. Yamada, S. Yoshida, H. M. Cheng, T. Ikeno and T. Yamada, J. Am. Chem. Soc., 2007, 129, 12902.
9. (a) A. Li, P. J. Kindelin and D. A. Klumpp, Org. Lett., 2006, 8, 1233; (b) A. S. K. Hashmi, Catal. Today, 2007, 122, 211; (c) J. U. Rhee and M. J. Krische, Org. Lett., 2005, 7, 2493.
10. T. Schwier, A. W. Sromek, D. M. L. Yap, D. Chernyak and V. Gevorgyan, J. Am. Chem. Soc., 2007, 129, 9868.
11. (a) N. K. Mills and A. H. White, J. Chem. Soc., Dalton Trans., 1984, 229; (b) S. Del Piero, R. Fedele, A. Melchior, R. Portanova, M. Tolazzi and E. Zangrando, Inorg. Chem., 2007, 46, 4683; (c) R. P. Feazell, C. E. Carson and K. K. Klausmeyer, Inorg. Chem., 2006, 45, 935.
12. T. Eicher, S. Haputmann and H. Suschitzky, The Chemistry of Heterocycles: Structure, Reactions, Syntheses and Applications, 2003.
13. J. E. Bauman, Jr. and J. C. Wang, Inorg. Chem., 1964, 3, 368.
14. (a) N. Masciocchi, M. Moret, P. Cairati, A. Sironi, G. Attilio Ardizzoia and G. La Monica, J. Chem. Soc., Dalton Trans., 1995, 1671; (b) K. Nomiya, K. Tsuda, Y. Tanabe and H. Nagano, J. Inorg. Biochem., 1998, 69, 9; (c) G. A. Ardizzoia, S. Brenna, F. Castelli, S. Galli, G. LaMonica, N. Masciocchi and A. Maspero, Polyhedron, 2004, 23, 3063.
15. K. Nomiya, K. Tsuda, T. Sudoh and M. Oda, J. Inorg. Biochem., 1997, 68, 39.
16. (a) P. J. Garegg and B. Samuelsson, Carbohydr. Res., 1980, 84, C1; (b) D. R. Rae, e-EROS – Encycl. Reagents Org. Synth., 2001.
17. (a) N. Asao, H. Aikawa and Y. Yamamoto, J. Am. Chem. Soc., 2004, 126, 7458; (b) N. Asao, K. Takahashi, S. Lee, T. Kasahara and Y. Yamamoto, J. Am. Chem. Soc., 2002, 124, 12650; (c) Y. Yamamoto, J. Org. Chem., 2007, 72, 7817; (d) N. Asao, T. Nogami, K. Takahashi and Y. Yamamoto, J. Am. Chem. Soc., 2002, 124, 764.

---

Footnotes

† Electronic supplementary information (ESI) available: Detailed experimental procedures and analytical data. See DOI: 10.1039/c0cc02623a

‡ This article is part of the ‘Emerging Investigators’ themed issue for ChemComm.

---