Xiao-Hui Hua,
Xiao-Qiu Pua,
Rui Liua,
Chun-Xiao Cuia,
Jun Yang*b and
Xian-Jin Yang*ab
aKey Lab for Advanced Material & Institute of Fine Chemicals, East China University of Science and Technology, 130 Meilong Road, Shanghai 2002 31, China. E-mail: yxj@ecust.edu.cn
bKey Laboratory of Organofluorine Chemistry, Shanghai Institute of Organic Chemistry, Chinese Academy of Science, 345 Ling-Ling Road, Shanghai 200232, China
First published on 3rd June 2016
A facile access to cis-dihydrofurobenzofuran from a Pd-catalyzed intramolecular oxidative C–O bond formation of 2-(2,5-dihydro-furan-2-yl)-phenol is reported. The combined use of Pd(OAc)2 as catalyst, acetonitrile as solvent and AgOAc or air as oxidant affords cis-dihydrofurobenzofurans in good (up to 83%) yields.
Several synthetic methods of benzofuran have been reported in the literature.6 A pioneering and elegant work on the synthesis of dihydrofurobenzofuran derivatives via hypervalent iodine reagent promoted formal [3 + 2] cycloaddition of phenol with furan was developed by Canesi and coworkers.7a,b In that report, 2-TMS-protected-phenol was utilized to achieve a multi-step conversion for controlling regioselectivity ((1) in Scheme 1).7c Recently, we have developed a one-pot amine-promoted Petasis reaction for the synthesis of 2-(2,5-dihydro-furan-2-yl)-phenols,8 which has been highlighted as a short, simple and cheap strategy for the construction of bioactive natural compounds.9,10 Thus, we were encouraged to broaden the application of this strategy11 ((2) in Scheme 1). In our investigation towards the complexation of 3a with metal salt, serendipitously, tricyclic product 4a was produced in acetonitrile at room temperature in 78% yield in the presence of one equivalent of Pd(OAc)2 ((3) in Scheme 1), therefore, we sought to explore this process to establish a convenient protocol for the synthesis of these compounds.
With 2-(4-phenyl-2,5-dihydrofuran-2-yl)-phenol 3a as the model substrate, we began with the optimization of reaction conditions. With 10 mol% Pd(OAc)2 as catalyst, a number of solvents was screened with the reaction mixture exposed to air. The results indicated that the reaction was sensitive to solvents. In methanol, acetone, 1,2-dichloroethane, chloroform, dichloro-methane and ethyl acetate, 4a was produced in 20–41% yields (entries 2, 6, 7, 9, 10, 11 in Table 1), while in toluene, xylene, DMF, DMSO, 1,4-dioxane and hexane, no product was produced with a large amount of starting material 3a remaining (entries 3, 4, 5, 8, 12, 13 in Table 1). CH3CN was found to be the best solvent giving a 53% yield (entry 1 in Table 1). When the temperature was elevated from room temperature to 80 °C, less reaction time was needed to complete the conversion of 3a, though the yield of 4a failed to be improved (entries 1 and 14 in Table 1). Subsequently, at 80 °C the reaction gave 4a in the presence of 5% and 20% Pd(OAc)2 in 42% and 60% yield, respectively.
Entrya | Solvent | Temp (°C) | Pd(OAc)2 (mol%) | Time (h) | Yieldb (%) |
---|---|---|---|---|---|
a The reaction was performed with 3a (0.1 mmol), Pd(OAc)2 (10 mol%) in solvents (2 mL).b Isolated yields.c Not detected. | |||||
1 | CH3CN | rt | 10 | 24 | 53 |
2 | CH3OH | rt | 10 | 24 | 41 |
3 | Toluene | rt | 10 | 24 | n.dc |
4 | DMF | rt | 10 | 24 | n.d |
5 | Xylene | rt | 10 | 24 | n.d |
6 | (CH3)2CO | rt | 10 | 24 | 32 |
7 | ClCH2CH2Cl | rt | 10 | 24 | 20 |
8 | DMSO | rt | 10 | 24 | n.d |
9 | CHCl3 | rt | 10 | 24 | 28 |
10 | CH2Cl2 | rt | 10 | 24 | 34 |
11 | EA | rt | 10 | 24 | 30 |
12 | 1,4-Dioxane | rt | 10 | 24 | n.d |
13 | Hexane | rt | 10 | 24 | n.d |
14 | CH3CN | 80 | 10 | 12 | 58 |
15 | CH3CN | 80 | 5 | 12 | 42 |
16 | CH3CN | 80 | 20 | 6 | 60 |
Then, with 10 mol% Pd(OAc)2 and CH3CN as solvent at 80 °C, a variety of oxidants were surveyed to improve the reaction. Pure oxygen instead of air did not change the reactivity (entry 1 Table 2). In the presence of Cs2CO3 or KOtBu as base, no product was given as determined by TLC (entries 4 and 5 in Table 2). To our delight, with AgOAc (2 equiv.) and Cu(OAc)2 (2 equiv.) as oxidant, the yield was increased to 72% and 68%, respectively (entries 6 and 7 in Table 2). Other oxidants such as DDQ, Selectfluor, PhI(OAc)2 and chloranil failed to give the desired product (entries 8–11 in Table 2). Accordingly, Pd(OAc)2 (10 mol%)/AgOAc (2 equiv.)/80 °C/CH3CN was chosen as the optimal reaction condition for further study.
Entrya | Oxidant (2 equiv.) | Base (2 equiv.) | Time (h) | Yield of 4ab (%) |
---|---|---|---|---|
a All the reactions were performed with 3a (0.1 mmol), Pd(OAc)2 (10 mol%) in MeCN (2 mL) with oxidants (2 equiv.), and if needed, base (2 equiv.) was added.b Isolated yields.c Not detected. | ||||
1 | O2 | — | 12 | 56 |
2 | Air | K2CO3 | 12 | 53 |
3 | Air | Na2CO3 | 12 | 50 |
4 | Air | Cs2CO3 | 12 | n.dc |
5 | Air | KOtBu | 12 | n.d |
6 | Cu(OAc)2 | — | 12 | 68 |
7 | AgOAc | — | 6 | 72 |
8 | DDQ | — | 12 | n.d |
9 | Selectfluor | — | 12 | n.d |
10 | PhI(OAc)2 | — | 12 | n.d |
11 | Chloranil | — | 12 | n.d |
With the optimized reaction conditions in hand, a diverse range of substrates were examined to broaden the substrate scope (Scheme 2). Substrates with electron withdrawing groups (R2) on aryl afforded 4 in higher yields than those with electron donating groups, and a substitution of R2 on aryls also had significantly impact on the reactivity. Higher yields (68–83%) were obtained for meta-R2 substitution on aryls (Scheme 2, 4b, 4e, 4h and 4l), while para- and ortho-R2 substitution gave 57–70% (Scheme 2, 4c, 4g, 4j and 4m) and 0–67% yields (Scheme 2, 4d, 4f and 4i), respectively. For the OMe-substituted substrates, no desired product was obtained for ortho OMe, while compound 4b with meta-OMe and compound 4c with para-OMe substitution were afforded in 81% and 57% yields, respectively. Similar results were obtained for fluoro-substituted substrates (R2 = F) (meta-F (4e, 76%), ortho-F (4f, 67%) and para-F (4g, 68%)). For meta and para-substituted products, these results were in agreement with the electronic effect of substituents i.e. the electron withdrawing induced effect: meta-OMe > para-OMe; meta-F > para-F; the electron donating conjugated effect: para-OMe > meta-OMe; para-F > meta-F. However, ortho-substituted products were hardly produced possibly due to more complicated electronic and steric effects.
While 4c was afforded with Cu(OAc)2 as an oxidant, no desired compound was detected when AgOAc was used as an oxidant. When R1 was para-methyl phenyl instead of phenyl, no change in reactivity was observed (Scheme 2, 4k and 4l). However, when R1 was 2-phenylethyl, no desired compound was obtained, which might be attributed to the structural instability of the corresponding tricycle compound 4 as a result of a loss of the conjugation between CC double bond and phenylethyl group. The structure of products 4 were further confirmed by the X-ray single crystal diffraction of 4j (Fig. 2). In this intramolecular cyclization, a strong steric strain needs to be overcome, and Pd(II) acted as a template to connect between C
C and OH groups in the oxidative cyclization process. This could be demonstrated by its easy ring-opening isomerization under acid catalysis at room temperature (Scheme 3).
A plausible mechanism for this reaction was proposed as shown in Scheme 4. Initial coordination and ligand exchange of Pd(OAc)2 with substrate 3 give intermediate I and II. Then intramolecular oxo-palladation with CC double bond gives intermediate III, followed by Pd(II) and β-hydrogen elimination to form the desired product 4 and Pd(0) species. Finally, Pd(II) was regenerated by the oxidation of Pd(0) with AgOAc, achieving the catalytic cycle.12,13
Footnote |
† Electronic supplementary information (ESI) available. CCDC 1447730. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra06840e |
This journal is © The Royal Society of Chemistry 2016 |