Publication | Open Access
Selective Reduction of Esters to Aldehydes under the Catalysis of Well‐Defined NHC–Iron Complexes
159
Citations
56
References
2013
Year
Chemical EngineeringEngineeringNatural SciencesDiversity-oriented SynthesisTotal SynthesisOrganic ChemistryOrganometallic CatalysisCatalysisWell‐defined Nhc–iron ComplexesChemistryPrimary AlcoholsHomogeneous CatalysisCatalytic SynthesisSelective ReductionUv Irradiation
On a direct course to the aldehyde: Hydrosilylation catalyzed by a well-defined N-heterocyclic-carbene–iron complex under UV irradiation enabled the selective reduction of esters to aldehydes (see scheme; Bn=benzyl, Mes=mesityl). The low catalyst loading and very mild reaction conditions make this chemoselective transformation a promising alternative to the reduction of esters with diisobutylaluminum hydride. Aldehydes are important building blocks used as reactive intermediates in organic and total synthesis. Their preparation from primary alcohols through oxidative pathways or from carboxylic derivatives by reductive methods is still challenging, as they are usually more reactive than the starting materials towards both oxidation and reduction, and this higher reactivity often leads to chemoselectivity issues.1 For the specific reduction of esters to aldehydes, bulky reducing agents, such as diisobutylaluminum hydride (DIBAL-H) and lithium tri-tert-butoxyaluminum hydride, have been used at low temperature for this transformation.2 However, such reagents can be difficult to handle in air and are only moderately compatible with other functional groups. Alternative methodologies are reduction reactions catalyzed by transition-metal complexes. In particular, hydrosilylation generally leads to silylacetals, which are stable in situ but converted into aldehydes upon hydrolysis. Indeed, besides hydrogenation, the use of hydrosilanes as reductants is an area of growing interest for the mild and selective reduction of carboxylic acid derivatives.3 This strategy has been applied to the reduction of esters to alcohols4–6 or ethers.5e, 7, 8 However, examples of selective catalytic hydrosilylation reactions of esters to give aldehydes in the presence of transition-metal9 or Lewis acid10 catalysts are scarce. On the other hand, during the last decade, iron11 has become a useful substitute for precious transition metals, particularly for reduction,12 and has been studied extensively. Surprisingly, although reduction with iron as the catalyst is well-exemplified for the hydrosilylation of carbonyl derivatives,13 examples with amides,14 imines,15 and carboxylic acids16 are scarce. Notably, selective iron-catalyzed hydrosilylation reactions of esters to form alcohols6 or ethers8 were recently reported. To the best of our knowledge, no iron-catalyzed selective hydrosilylation of esters to give aldehydes has been reported previously. Herein, we report an efficient and selective method for the reduction of esters to aldehydes through hydrosilylation catalyzed by N-heterocyclic-carbene (NHC) iron(0) complexes [(NHC)Fe(CO)4].17 We started our investigation with the reduction of methyl phenylacetate (1 a) as the model substrate with diethylsilane (1.1 equiv). With commercially available [Fe(CO)5] and [Fe2(CO)9] or complexes already developed for the selective hydrosilylation of esters ([CpFe(CO)2(PCy3)][BF4])6 and carboxylic acids ([(cod)Fe(CO)3] and [(pbo)Fe(CO)3])16 at room temperature under UV irradiation (350 nm) for 16 h, only moderate conversion and selectivity were observed (see the Supporting Information; cod=1,5-cyclooctadiene, Cy=cyclohexyl, pbo=(E)-4-phenylbut-3-en-2-one). In our search for more reactive and selective catalysts, a new class of tetracarbonyl iron(0) complexes bearing NHC ligands (Scheme 1) showed promising results.18 The [(NHC)Fe(CO)4] complexes were prepared in one step from [Fe(CO)5] (2 equiv) by treatment with the freshly prepared NHC (1 equiv) in toluene at 70 °C overnight in the presence of [{Cp*Fe(CO)2}2] (2 mol %) as the catalyst and were isolated in moderate to good yields (36–63 %).19 The molecular structures of 4 b and 5 are depicted in Figure 1. The coordination geometry at the iron center is distorted trigonal bipyramidal, with the N-heterocyclic carbene ligand located in an axial position. The FeCcarbene bond length is 2.007(2) Å for 5 (for 4 a, 1.9988(17) Å and for 4 b, 2.0041(14) Å) and is comparable to those reported previously.17 Molecular structures of the complexes 4 b and 5. Thermal ellipsoids correspond to 50 % probability. Hydrogen atoms are omitted for clarity. Synthesis of the [(NHC)Fe(CO)4] complexes. Cp*=pentamethylcyclopentadienyl, Mes=mesityl (2,4,6-trimethylphenyl). We tested these complexes 4 a,b and 5 as catalysts in the hydrosilylation of esters (Table 1). Notably, with complex 4 a (2.5 mol %) in the presence of Et2SiH2 (1.1 equiv), both activity and selectivity were enhanced: full conversion was observed after 3 h at room temperature, and only the aldehyde product was detected (Table 1, entry 1). Although complexes 4 b and 5 gave the aldehyde 2 a selectively under similar conditions, the conversion after reaction for 1 h was lower (compare entries 3, 5, and 6 in Table 1). The superior efficiency of the iron complex 4 a was confirmed by lowering the catalyst loading to 0.5 mol % (Table 1, entry 4). Among a variety of silanes tested, PhSiH3 gave exclusively the alcohol 3 a, albeit with low conversion (25 %; Table 1, entry 10), whereas PH2SiH2 gave the aldehyde 2 a with 75 % conversion (Table 1, entry 7). Thus, the nature of the silane is of crucial importance for the selectivity. Importantly, the role of UV activation (350 nm) was also shown: no reaction was observed with Et2SiH2 in the absence of UV irradiation (Table 1, entry 12). Entry [Fe] (mol %) Silane (equiv) Conv. [%][b] 2 a/3 a [%][b] 1 4 a (2.5) Et2SiH2 (1.1) 98 98:0 2 4 a (1) Et2SiH2 (1.1) 96 95:1 3[c] 4 a (1) Et2SiH2 (1.1) 76 72:4 4 4 a (0.5) Et2SiH2 (1.1) 91 88:3 5[c) 4 b (1) Et2SiH2 (1.1) 68 65:3 6[c] 5 (1) Et2SiH2 (1.1) 54 53:1 7 4 a (2.5) Ph2SiH2 (1.1) 75 75:0 8 4 a (2.5) PhMeSiH2 (1.1) 22 18:2 9 4 a (2.5) Et3SiH (1.1) 0 – 10 4 a (2.5) PhSiH3 (1.1) 25 0:25 11 4 a (2.5) Ph3SiH (1.1) 0 – 12[d] 4 a (1) Et2SiH2 (1.1) 0 – Having optimized the reaction conditions (complex 4 a (1 mol %), Et2SiH2 (1.1 equiv), toluene, room temperature, UV irradiation at 350 nm, 3 h), we investigated the scope of this chemoselective reduction of esters to aldehydes (Table 2). Under these conditions, methyl, ethyl, and benzyl esters were converted selectively into the corresponding aldehydes: full conversion was observed after 3 h, and the aldehydes were obtained in good yields (83–95 %; Table 2, entries 1, 2, and 8–10). The reaction also proceeded well with methyl acetates substituted with para-, meta-, and ortho-tolyl groups, although ortho substitution seemed to inhibit the reactivity slightly (Table 2, entries 3–5). Electronic effects at the para position of the aromatic ring of the phenylacetate derivatives were limited, and the corresponding aldehydes were obtained selectively in good yields regardless of the para substituent (71–85 %; Table 2, entries 3, 6, and 7). More strikingly, aliphatic esters, such as ethyl acetate or methyl decanoate, could be reduced to the corresponding aliphatic aldehydes in good yields (up to 90 % yield; Table 2, entries 12 and 13). Moreover, methyl 10-undecenoate was transformed into 10-undecenal in 70 % yield; products resulting from the isomerization of the CC bond were obtained in only 10 % yield (Table 2, entry 14). Importantly, the reduction of dimethyl tridecanedioate led selectively to the corresponding dialdehyde, which was isolated in very good yield (90 %; Table 2, entry 15). Entry Ester Aldehyde Conv. [%] Yield [%][b] 1 2 R=Me R=Et 96[c] 97[c] 83 85[d] 3 4 5 6 7 R=p-Me R=m-Me R=o-Me R=p-Br R=p-OMe 99[e] 99[e] 92[e] 99[e] 83[e] 79 71 65[f] 85 77 8 9 10 R=Me R=Et R=Bn 92[d] 95[c] 92[c] 90 95[d] 83[d] 11[g] 76 65 12 97[e] 90 13 EtOAc CH3CHO 95[c] 85[f] 14 75[c] 70[f,h] 15[g] 99[e] 90 16[i] 17[i] 18[i] 19[i] R=p-Br R=p-Me R=p-CF3 R=H 99[c] 91[c] 40[e] 93[c,j] 86 80 – 83[d] 20 60[c] 55[f] Interestingly, this reaction was also successful with aromatic esters. When the reaction was performed with the catalyst 4 a (2.5 mol %) in the presence of Et2SiH2 (1.1 equiv) at room temperature under UV irradiation (350 nm) for 4 h, 58 % conversion of methyl p-bromobenzoate into the corresponding aldehyde occurred. Indeed, the use of diphenylsilane (1.5 equiv) led to full conversion, and p-bromobenzaldehyde was isolated in 86 % yield (Table 2, entry 16). Under similar conditions, p-methylbenzaldehyde was also formed in good yield (80 %), whereas the reduction of methyl p-trifluoromethylbenzoate occurred with only 40 % conversion, probably because of the strongly electron withdrawing substituent (Table 2, entries 17 and 18). Heteroaromatic esters can be reduced under similar conditions: methyl 5-bromofur-2-yl-methanoate was converted into the corresponding aldehyde, which was isolated as its hydrazone derivative in 55 % yield (Table 2, entry 20). We next focused our attention on the reduction of lactones to lactols (Scheme 2).9a, 20 Such reductions are usually performed with DIBAL-H at low temperature (−78 °C). Under the optimized conditions (complex 4 a (1 mol %), Et2SiH2 (1.1 equiv), room temperature, UV irradiation (350 nm), 3 h), the γ-lactones 5-methyltetrahydrofuran-2-one and 5-heptyltetrahydrofuran-2-one were fully converted into the corresponding γ-lactol 6 a,b. 5-Heptyltetrahydrofuran-2-ol 6 b was isolated in 83 % yield. Notably, δ-decalactone was converted selectively into δ-decalactol (6 c), which was isolated in 83 % yield. Similarly, with ε-decalactone as the starting substrate, the reduction proceeded with full conversion, and the hydroxyaldehyde 6 d was obtained selectively in 92 % yield. [(IMes)Fe(CO)4]-catalyzed hydrosilylation of lactones. Reaction conditions: 1) lactone (0.5 mmol), Et2SiH2 (1.1 equiv), complex 4 a (1 mol %), C6D6 (solvent), room temperature, UV irradiation (350 nm), 3 h; 2) hydrolysis with 1 M aqueous HCl (2 mL) and THF (2 mL), room temperature, 2 h. [a] Yield determined by 1H NMR spectroscopy for the silylacetal intermediate. To explore the mechanism of this iron-catalyzed hydrosilylation, we first investigated the stoichiometric reaction between 4 a and silanes. The reaction of 4 a with HSiPh3 under UV irradiation (350 nm) led to the formation of the complex [(IMes)Fe(H)(SiPh3)(CO)3] (8 a) resulting from the oxidative addition of the SiH bond of the silane to unsaturated [(IMes)Fe(CO)3] (7).21 This complex was fully characterized by elemental analysis, X-ray diffraction studies,22 and NMR spectroscopy: in the 1H NMR spectrum, the hydride exhibited a well-resolved singlet at δ=−8.74 ppm,23b, 25 and in the 29Si NMR spectrum, a singlet at δ=+28.19 ppm was observed.25 The phosphine analogue [(PMe2Ph)Fe(H)(SiPh3)(CO)3] (9) was also synthesized. Complex 9 showed similar spectroscopic features to those of 10 a: the hydride proton resonates at δ=−9.30 ppm, and the silicon resonance is at δ=+28.4 ppm. The molecular structures of 8 a and 9 are depicted in Figure 2. The coordination geometry at the iron center is distorted trigonal bipyramidal; the N-heterocyclic carbene ligand (or the phosphine ligand) and the silyl fragment are located in a trans axial arrangement with a CIMes-Fe-Si angle of 168.50(6)° (P-Fe-Si, 177.32(2)°). The CNHCFe bond length of 1.997(2) Å is similar to that found in the starting complex 4 a (1.9988(17) Å). Furthermore, the FeSiPh3 bond lengths (2.3488(9) and 2.3564 (6) Å for 8 a and 9, respectively) are consistent with those described previously.22a, 23 Notably, the hydride is located in the meridional plane, and the Fe–H distance of 1.43(2) Å for 8 a (1.46(2) Å for 9) is consistent with an iron–hydride bond.22a, 24 Molecular structures of the complexes 8 a and 9. Thermal ellipsoids correspond to 50 % probability. One molecule of toluene in 8 a and hydrogen atoms, except the iron hydride ligands, are omitted for clarity. The reaction of 4 a with Et2SiH2 in C6D6 led exclusively to the formation of [(IMes)Fe(H)(CO)3(SiHEt2)] (8 b), which was characterized by NMR spectroscopy (hydride signal at δ=−9.29 ppm in the 1H NMR spectrum, silicon resonance at δ=+28.2 ppm with a coupling constant 1JSi,H=178 Hz in the 29Si NMR spectrum). The addition of a stoichiometric amount of methyl phenylacetate to 8 b under UV irradiation26 led to the formation of several compounds, including several iron hydride species; as a result, we were unable to distinguish clearly the reactivity of the FeH bond or the FeSi bond towards the ester. On the basis of these observations, a plausible catalytic cycle is presented in Scheme 3 for the reduction of esters to aldehydes with Et2SiH2 in the presence of [(IMes)Fe(CO)4] as the catalyst precursor. In the initial step, UV activation of [(IMes)Fe(CO)4] generates a 16-electron species [(IMes)Fe(CO)3] (7), which then reacts with the hydrosilane to generate the oxidative-addition complex 8 b. At this stage, the insertion of the carbonyl moiety of the ester into the FeSi and/or FeH bond is likely. In the final step, reductive elimination affords the silylated acetal 10 and regenerates the active species 7. Proposed catalytic cycle for the iron-catalyzed hydrosilylation of esters to give aldehydes. In summary, we have developed simple methodology for the chemoselective reduction of esters to aldehydes with an N-heterocyclic-carbene–iron complex, such as [(IMes)Fe(CO)4], as the catalyst (1 mol %) in the presence of a secondary silane (diethylsilane or diphenylsilane) as the reducing agent. This reaction occurs at room temperature under UV irradiation with both aromatic and aliphatic esters. Notably, this catalytic system also permitted the efficient and selective reduction of lactones to lactols. We have provided experimental evidence that the hydrosilylation occurs by oxidative addition of the hydrosilane to an unsaturated NHC–Fe species to yield a silyl iron hydride complex. As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.
| Year | Citations | |
|---|---|---|
Page 1
Page 1