Shravankumar Kankalaa,
Sreekantha B. Jonnalagadda*a and
Chandra Sekhar Vasam*bc
aSchool of Chemistry & Physics, University of Kwazulu-Natal, Westville Campus, Chiltern Hills, Durban-4000, South Africa. E-mail: jonnalagaddas@ukzn.ac.za; Fax: +27-31-260-3091; Tel: +27-31-260-7325, +27-31-260-3090
bDepartment of Chemistry, Satavahana University, Karimnagar, India
cDepartment of Pharmaceutical Chemistry, Telangana University, Nizamabad, India. E-mail: vasamcs@yahoo.co.in; Fax: +91-878-2255933; Tel: +91-9000285433
First published on 3rd September 2015
A nucleophilic organo N-heterocyclic carbene (NHC) catalysed click-type fast domino addition of allenyl-MgBr to aryl nitrile oxides to produce 5-butynylisoxazoles with excellent selectivity and good yields is reported. The unwanted protonation and subsequent formation of 5-methylisoxazole byproducts is successfully suppressed. Furthermore, a Pd/Ag catalysed protocol for Sonogashira cross-coupling of 5-butynylisoxazoles to obtain the corresponding internal alkynes with high selectivity and yields is developed by minimising the alkyne homo-coupling.
The biologically imperative isoxazoles are key synthones and synthetic intermediates3 and often used as masked β-diketones and β-aminoenones.4 To construct isoxazole ring, the 1,3-dipolar cycloaddition between nitrile oxides and alkynes,5 and domino addition between alkyne organometallics and nitrile oxides6 have been reported. The selective synthesis of terminal alkyne-tagged isoxazole ring via nucleophilic addition would be an important aspect, which allows many subsequent side chain elaborations on isoxazole ring.
Previously Sampath et al. reported6 the domino addition of allenyl-MgBr with nitrile oxide and observed the formation of mixture of 5-butynyl and 5-methylisoxazoles. It was described that during the second mole addition of allenyl-MgBr to Mg-cycloadduct, quenching of the reaction with aq. NH4Cl lead to proton capture and directed the formation of 5-methylisoxazole as byproduct along with 5-butynylisoxazole (Scheme 1). Although, the above domino addition is an elegant concerted synthesis for 5-butynylisoxazole, the subsequent side-chain elaborations in one-pot are not possible due to the presence of 5-methylisoxazole byproduct. Since, the carbanion of Mg-cycloadduct is very reactive and functions as an extremely strong base, it captures the proton not only from water, but also from other possible sources during the domino additions and form the byproducts. In this respect, finding a solution to evade the unwanted protonation using Grignard reagents not only expands their scope in domino additions, but also in many other nucleophilic additions.
While organocatalysed domino reactions have emerged as important tools in organic synthesis, the usefulness of Lewis base nucleophilic NHCs as organo catalysts in domino reactions has received less attention.7 It is also important to note that the NHCs and their precursors i.e. imidazolium salts are the good example for weak strong Bronsted conjugate acid–base pair. Therefore, organo-NHCs are expected to mediate an efficient and fast non-concerted domino addition of Grignard reagents via forming NHC-Grignard adduct, that would minimise the unwanted protonation and byproduct formation. Our objective is to enhance the reactivity of Grignard reagents with a substoichiometric amount of a Lewis base NHCs, whereby the resulting NHC-Grignard complex induces high regioselectivity.
At first, we have investigated the domino addition between nitrile oxides (1) and allenyl-MgBr (2) without a catalyst to find out the exact source of proton that leads to formation of unwanted 5-methylisoxazole byproduct along with the target 5-butynylisoxazole. The reactants nitrile oxides (1) and allenyl-MgBr (2) were generated in situ using the established procedures.8 Aq. NH4Cl solution that used to quench the domino addition has been reported to be the probable source for proton to interact with Mg-cycloadduct.6 During our experiments, we found that the Et3NHCl present in nitrile oxide solution could also be a potential proton source (that formed by treating chloro-oxime with Et3N organic base). The acidic pH of the solution was determined by simple litmus paper test. As the uncatalysed domino addition reaction needs ∼5–6 h, the unwanted protonation of Mg-cycloadduct is possible to form the byproducts.6
Recently we have reported the organo-NHC catalysed alkyne-nitrile oxide 1,3-dipolar cycloaddition for the regioselective synthesis of 3,5-disubstituted isoxazoles.5 Based on the above observations, we assessed the scope of Lewis base nucleophilic organo-NHC catalysts (Fig. 1) in domino addition. Interestingly, we found that the use of organo-NHC catalyst (1,3-di-tert-butylimidazolin-2-ylidene (i)) facilitated the domino addition very fast (<10 min). The reaction mixture turned colourless to green precipitate. Quenching of the reaction with aq. NH4Cl, gave 5-butynylisoxazoles (4) selectively in high yields without any 5-methylisoxazole (Scheme 2), indicating the protonation either by Et3NHCl or quenching agent aq. NH4Cl during domino addition was successfully minimised. The role of Et3NHCl as possible source for proton in other synthetic reactions was also reported previously.9
The reduced reaction time (<10 min) in the catalysed reaction relative to the uncatalysed process (∼6 h) has enhanced the product selectivity. Although Table 1 illustrates the results obtained with only organo-NHC catalyst of salt (i), the organo-NHC catalysts of salts ii–v (Fig. 1) can also perform equally well to produce selectively the 5-butynylisoxazole (ESI†). It is an evidence for the high efficacy of organo-NHC catalysts in domino addition. The pure 5-butynylisoxazole product was characterised by NMR and mass spectroscopic techniques (ESI†).
Entry | Ar | Products | Yieldsb (%) |
---|---|---|---|
a All products were characterised by 1H/13C NMR, mass spectral analysis.b Determined by GC.c The yield in the parenthesis is for 5-methylisoxazoles were isolated in 12% yields without organo-NHC catalyst. | |||
1 | 4-OMeC6H4 (1a) | 4a | 84 |
2 | 4-MeC6H4 (1b) | 4b | 85 |
3 | 2-NO2C6H4 (1c) | 4c | 91 |
4 | 2-NO2C6H4 (1c) | 4c | 81 (12)c |
5 | 4-CIC6H4 (1d) | 4d | 89 |
6 | C6H5 (1e) | 4e | 90 |
7 | 4-NO2C6H4 (1f) | 4f | 94 |
8 | 4-FC6H4 (1g) | 4g | 86 |
9 | 2,3-diOMeC6H3 (1h) | 4h | 87 |
10 | 2-Naphthyl (1i) | 4i | 92 |
11 | 4-CNC6H4 (1j) | 4j | 90 |
12 | 3-Br-4-OMeC6H3 (1k) | 4k | 88 |
13 | 4-OHC6H4 (1l) | 4l | 90 |
Based on the success of the title reaction, to further explore the robustness of the protocol, a series of nitrile oxides (1a–l) were subjected for domino addition with allenyl-MgBr in the presence of organo-NHC catalyst. All of these investigated domino additions were accomplished fast producing their corresponding 5-butynylisoxazoles (4a–l) selectively in high yields (Scheme 2, Table 1).
Taking into consideration the known the acid–base chemistry of interaction between Grignard reagents and imidazolium salts, a probable mechanism is postulated for the reaction as depicted in Scheme 3. It would be reasonable to propose that an organo-NHC generated during the catalytic cycle will stabilize the transient species and acts as charge carrier. Based on this hypothesis Scheme 3 will provide the possible sequential steps to complete the catalytic cycle.
![]() | ||
Scheme 3 A plausible mechanism of organo-NHC catalysed domino addition between allenyl-MgBr and nitrile oxide and selective formation of 5-butynylisoxazole. |
As illustrated in Scheme 3, the formation of 5-butynylisoxazoles could occur in a non-concerted catalytic intermolecular domino addition fashion. Lewis base activation of Grignard reagents by NHCs has been reported.10 The reactions of imidazolium salt with 2 moles of allenyl-MgBr will form the NHC–Mg-allenyl intermediate (3b) via the deprotonation of imidazolium salt1b and the interaction of (3b) with nitrile oxide leads to (3c) via 3 + 2 cycloaddition through a pseudo-envelop conformation, which explains the regioselective formation of 3,5-disubstituted isoxazole-Mg–NHC (3c). The domino addition of second mole of allenyl-MgBr to the intermediate (3c) leads to the selective formation of final product, 5-butynylisoxazole (4).
(1) Pd/Cu catalysed Sonogashira cross-coupling (route 1).
(2) Cu-free Pd catalysed Sonogashira cross-coupling (route 2).
(3) Pd/Ag catalysed Sonogashira cross-coupling (route 3).
The details of the reaction conditions, catalyst composition and products are illustrated in Scheme 4. Instead of commercial Pd(II) salt source, NHC-ligand bound Pd(II) was used as catalysts to improve the catalytic performance in all three routes.
The Pd/Cu catalysed Sonogashira coupling conducted between the alkyne (4a) and nitro substituted aryl iodide (5d) yielded cross-coupled and homo-coupled products (Scheme 4). The role of copper as catalyst for alkyne homo-coupling side reaction (Hay/Glaser-type) during Sonogashira cross-coupling was described in some previous works.11 A copper free Pd-catalysed Sonogashira coupling (Scheme 4, catalytic route 2) conducted using the same substrates i.e. (4a) and (5d) gave single cross-coupled product, but the reaction was very slow and yields were poor (∼34%).
Literature reports the use of Ag(I) and Au(I) salts/complexes also as co-catalysts with palladium to act as transmetallation agents.12 We explored with scope of Pd with Ag as co-catalyst for Sonogashira cross-coupling of 5-butynylisoxazole (Scheme 4, catalytic route 3) to reap the benefits of Ag(I) and Pd(II) properties. At this juncture, we attempted to combine the well-known transmetallation chemistry13 between Ag(I)–NHCs and Pd(II) with Sonogashira cross-coupling to introduce Ag(I) as co-catalyst.
Generally, the Pd(II)–NHC catalysts for cross-coupling can be generated in two ways, either by (i) direct metallation of Pd(II) source with imidazolium salt or by (ii) transmetallation of carbene from Ag(I)–NHCs to Pd(II) source.13 In this work, for the catalytic route 3 of Scheme 4, the Pd(II)–NHC catalyst was generated in situ from Ag(I)–NHCs via carbene transfer/transmetallation as shown in Scheme 5. This process usually accompanies Ag(I)Cl byproduct.13
![]() | ||
Scheme 5 Methodology for Pd/Ag catalysed Sonogashira coupling developed in this work via double transmetallation reactions of Ag(I). |
The reaction of Ag(I)Cl with terminal acetylenes to form Ag–acetylides intermediates those are useful in organic synthesis was also reported previously.12b,14 In this respect, we found that the Ag(I)Cl byproduct of Scheme 5 could be useful to design the Pd/Ag co-catalysed Sonogashira coupling of catalytic route 3 of Scheme 4 without any additional/external Ag(I)–halide. This strategy makes the catalytic route 3 of Scheme 4 more atom-economic with no need for addition of any external Ag(I) co-catalyst.
In this context, we noticed that a combination of in situ generated Pd(II)–NHC and Ag(I)Cl as catalysts worked well to accomplish a fast Sonogashira coupling between (4a) and (5d) to produce the corresponding internal alkyne derivative (6d) selectively and minimising the formation of unwanted homo-coupled products (Scheme 4, catalytic route 3). The effect of different bases and solvents in Sonogashira coupling between (4a) and (5d) was also investigated (ESI†). The use of pyrrolidine is found to be advantageous over the other bases (ESI†). The details of the spectral analysis of these products are given in ESI.†
The latitude of Pd/Ag catalysed Sonogashira coupling protocol was extended to other aryl halides (Scheme 6). The details of the reactants and corresponding product yields are summarised in Table 2. In less than 2 h all the reaction could be accomplished with good to excellent yields (Table 2). No homo-coupling was detected in any of these reactions. Nevertheless, the aryl halides bearing electron withdrawing groups provided slightly better yields of internal alkynes as compared to those with electron donating substituents concurring with earlier reports.15
Entry | Ar | R | Product | Yieldb (%) |
---|---|---|---|---|
a All products were characterised by IR, 1H/13C NMR and mass spectral analysis.b GC yields after 2 hour. | ||||
1 | 4-OMeC6H5 (4a) | Me (5a) | 6a | 84 |
2 | 4a | OMe (5b) | 6b | 82 |
3 | 4a | H (5c) | 6c | 90 |
4 | 4a | NO2 (5d) | 6d | 93 |
5 | 4a | COMe (5e) | 6e | 94 |
6 | C6H5 (4e) | 5a | 6f | 82 |
7 | 4e | 5b | 6g | 82 |
8 | 4e | 5c | 6h | 88 |
9 | 4e | 5d | 6i | 91 |
10 | 4e | 5e | 6j | 90 |
11 | 4-NO2C6H5 (4f) | 5a | 6k | 89 |
12 | 4f | 5b | 6l | 90 |
13 | 4f | 5c | 6m | 90 |
14 | 4f | 5d | 6n | 96 |
15 | 4f | 5e | 6o | 94 |
The above approach enabled us to perform side-chain elaboration directly on as-synthesised 5-butynylisoxazole via Sonogashira cross-coupling without the interference of 5-methylisoxazole. A new Pd/Ag catalysed system, which is much efficient and atom-economic than Pd/Cu catalysed and copper free Pd-catalysed systems is developed for the Sonogashira cross-coupling of 5-butynylisoxazole to give the corresponding internal alkynes as sole products in high yields, by suppressing the alkyne homo-coupling side reaction. Thus, the internal alkynes of 5-butynylisoxazoles act as useful starting materials to access 1,3-dicarbonyl compounds via hydrolysis of isoxazole ring of internal alkynes. We have also demonstrated that the benefit of Ag(I)Cl, a byproduct of the transmetallation-I (Scheme 5), as a co-catalyst for the subsequent Sonogashira coupling of 5-butynylisoxazoles (transmetallation-II, Scheme 5), which again highlights the atom efficiency of our Pd/Ag catalysed Sonogashira coupling protocol.
Footnote |
† Electronic supplementary information (ESI) available: The spectral data of 5-butynylisoxazoles (4a–l) and their internal alkynes (6a–o). See DOI: 10.1039/c5ra11947b |
This journal is © The Royal Society of Chemistry 2015 |