nanomaterials Article
Synthesis of Formate Esters and Formamides Using an Au/TiO2-Catalyzed Aerobic Oxidative Coupling of Paraformaldehyde Ioannis Metaxas, Eleni Vasilikogiannaki and Manolis Stratakis *
ID
Department of Chemistry, University of Crete, Voutes, 71003 Iraklion, Greece;
[email protected] (I.M.);
[email protected] (E.V.) * Correspondence:
[email protected]; Tel.: +30-2810-545-087 Received: 18 September 2017; Accepted: 6 December 2017; Published: 12 December 2017
Abstract: A simple method for the synthesis of formate esters and formamides is presented based on the Au/TiO2 -catalyzed aerobic oxidative coupling between alcohols or amines and formaldehyde. The suitable form of formaldehyde is paraformaldehyde, as cyclic trimeric 1,3,5-trioxane is inactive. The reaction proceeds via the formation of an intermediate hemiacetal or hemiaminal, respectively, followed by the Au nanoparticle-catalyzed aerobic oxidation of the intermediate. Typically, the oxidative coupling between formaldehyde (2 equiv) and amines occurs quantitatively at room temperature within 4 h, and there is no need to add a base as in analogous coupling reactions. The oxidative coupling between formaldehyde (typically 3 equiv) and alcohols is unprecedented and occurs more slowly, yet in good to excellent yields and selectivity. Minor side-products (2–12%) from the acetalization of formaldehyde by the alcohol are also formed. The catalyst is recyclable and can be reused after a simple filtration in five consecutive runs with a small loss of activity. Keywords: formates; formamides; paraformaldehyde; aerobic coupling; Au nanoparticles
1. Introduction Supported Au nanoparticles (Au NPs) have emerged during the past two decades as potent catalysts in several organic transformations. The surprising initial observation by Haruta’s group, that Au NPs are capable of oxidizing CO into CO2 even at −70 ◦ C by atmospheric air [1,2], a reaction with great environmental significance, was the starting point of the extended growth of the applications of Au NPs in catalysis. This first breakthrough observation established a metal (Au) that no one could foresee having any major catalytic activity, into the first row of exploration by many research groups worldwide. The reason for this catalytic activity is that the small clusters of Au deposited on several supports, such as metal oxides or polymeric materials, in contrast to bulk gold, which is formally inactive, exhibit unexpected properties. The key role of the support is not only to stabilize and protect Au clusters from leaking, aggregation, or against oxidation, which deactivates them in certain cases, but, in many reports, it is proposed that there is a synergy in the catalytic activity among metal oxide and Au NPs. Following the Au NPs-catalyzed aerobic oxidation of CO, the first major step towards their recognition as powerful catalysts was triggered by their application in the aerobic oxidation of organic compounds, such as alcohols, aldehydes, amines, alkenes, etc. [3–8]. Such transformations gained significant attention within the organic community, due to their green nature, as the oxidant is simply the atmospheric air. The second major step in the field of catalysis using Au NPs, which essentially expanded during the first decade of this century, was rather unexpected, as nanogold catalysts found catalytic applications in organic transformations beyond aerobic oxidations. Such transformations are
Nanomaterials 2017, 7, 440; doi:10.3390/nano7120440
www.mdpi.com/journal/nanomaterials
Nanomaterials 2017, 7, 440 Nanomaterials 2017, 7, 440
2 of 10 2 of 10
catalysts found catalytic applications in organic transformations beyond aerobic oxidations. Such transformations areeither typically catalyzed either by or ionic substances by Pd(0) andmostly other under noble typically catalyzed by ionic Au substances by Au Pd(0) and otherornoble metals, metals, mostly conditions under homogeneous conditions [9–13]. homogeneous [9–13]. The results that will be presented The results that will be presented and and analyzed analyzed below below were were triggered triggered during during our our recent recent studies studies on the difunctionalization of oxetanes with a silylborane catalyzed by supported Au nanoparticles on the difunctionalization of oxetanes with a silylborane catalyzed by supported Au nanoparticles [14]. [14]. was found that activated α,α-disubstituted oxetanes not undergo silaboration and, in the It wasItfound that activated α,α-disubstituted oxetanes do not do undergo silaboration and, in the absence absence of the silylborane, twoAu-nanoparticle-catalyzed competing Au-nanoparticle-catalyzed pathways take place in of the silylborane, two competing pathways take place in 1,2-dichloroethane 1,2-dichloroethane as a solvent: isomerization into or homoallylic alcohols or a retro [2+2] (DCE) as a solvent: (DCE) isomerization into homoallylic alcohols a retro [2+2] decomposition forming decomposition forming an alkene and formaldehyde. Generally, the first path is prevailing. A an alkene and formaldehyde. Generally, the first path is prevailing. A typical example is shown in typical example is shown in Scheme 1, regarding the case of 2-methyl-2-phenyloxetane (1). Scheme 1, regarding the case of 2-methyl-2-phenyloxetane (1). Surprisingly, though, one one of of the the minor minor side-products side-products was was the the formate formate of Surprisingly, though, of the the homoallylic homoallylic alcohol alcohol (1b), and we reasonably considered its formation as arising from the reaction of homoallylic alcohol (1b), and we reasonably considered its formation as arising from the reaction of homoallylic alcohol 1a 1a (isomerization product main path withformaldehyde, formaldehyde,which whichisisformed formedvia viathe the retro retro [2+2] [2+2] (isomerization product of of thethe main path a)a)with decomposition decomposition path path of of the the oxetane oxetane (path (path b). b). The The isomerization isomerization mechanism mechanism of of activated activated oxetanes oxetanes into into homoallylic alcohols catalyzed by Au/TiO 2 (path b) is obviously identical to the concerted process homoallylic alcohols catalyzed by Au/TiO2 (path b) is obviously identical to the concerted process for for isomerization of activated epoxides allylic alcohols, previously established by group our group the the isomerization of activated epoxides intointo allylic alcohols, previously established by our [15] [15] and requires the synergy among electrophilic Au nanoparticles and the basic oxygen atoms of and requires the synergy among electrophilic Au nanoparticles and the basic oxygen atoms of the the support. Additional experiments established that, under the reaction conditions, alkene 1c does support. Additional experiments established that, under the reaction conditions, alkene 1c does not not a retro carbonyl-ene reaction of homoallylic alcohol 1a,also, but 1a also, 1a is not formed via a arisearise via avia retro carbonyl-ene reaction of homoallylic alcohol 1a, but is not formed via a direct direct carbonyl-ene reaction among alkene 1c and formaldehyde. carbonyl-ene reaction among alkene 1c and formaldehyde.
The transformations transformations of of an an oxetane oxetane (1) in the presence and a proposed Scheme 1. The presence of of Au/TiO Au/TiO22 and mechanism, which triggered the current study of a methodology for the formation of formate esters and paraformaldehyde. paraformaldehyde. and formamides via the aerobic oxidative coupling between between alcohols/amines alcohols/amines and
We, therefore,focused focusedon onaapossible possiblemethodology methodology synthesis of formate esters, based on We, therefore, forfor thethe synthesis of formate esters, based on the the reaction between an alcohol and formaldehyde in the of presence Au/TiO 2 under aerobic reaction between an alcohol and formaldehyde in the presence Au/TiOof aerobic conditions. 2 under conditions. In essentially this concept, theoxidation in situ ofcatalytic oxidation the produced labile In this concept, the inessentially situ catalytic the produced labile of hemiacetal is anticipated. hemiacetal is anticipated. As a second task, we extended this concept to the mild and facile synthesis As a second task, we extended this concept to the mild and facile synthesis of formamides from the of formamides amines from the corresponding amines base without an external as in oxidative previous corresponding without using an external as inusing previous reported base analogous reported analogous oxidative coupling protocols. coupling protocols. 2. Results 2.1. Synthesis Synthesis of of Formate Formate Esters Esters 2.1. To test test this this concept, concept,we weinitially initially examined reaction a simple alcohol (1-octanol, 2) To examined thethe reaction of aofsimple alcohol (1-octanol, 2) with with formaldehyde in the presence of Au/TiO (2 mol %). There are two stable solid sources of formaldehyde in the presence of Au/TiO2 (22 mol %). There are two stable solid sources of formaldehyde: paraformaldehyde, a linear polymeric formform and delivers formaldehyde upon formaldehyde: paraformaldehyde,which whichis is a linear polymeric and delivers formaldehyde heating, and 1,3,5-trioxane, which is the cyclic trimeric acetal form. Treatment of 1-octanol with 3 molar upon heating, and 1,3,5-trioxane, which is the cyclic trimeric acetal form. Treatment of 1-octanol equivalents paraformaldehyde in 1,2-dichloroethane (DCE) and heating to and 65 ◦ C for 10 h, to the with 3 molarofequivalents of paraformaldehyde in 1,2-dichloroethane (DCE) heating to led 65 °C for quantitative consumption of the alcohol and to a mixture of formate ester 2a and acetal 2b in a relative 10 h, led to the quantitative consumption of the alcohol and to a mixture of formate ester 2a and
Nanomaterials 2017, 7, 440 Nanomaterials 2017, 2017, 7, 7, 440 440 Nanomaterials
3 of 10 of 10 10 33 of
acetal 2b in a relative ratio 2a/2b = 95/5 (Scheme 2). The mixture was chromatographed and 2a was acetal 2b in a relative ratio 2a/2b = 95/5 (Scheme 2). The mixture was chromatographed and 2a was isolated in 80% yield. Contrary tomixture paraformaldehyde, we could and not 2a achieve any reaction with ratio 2a/2b = 95/5 (Scheme 2). The was chromatographed was isolated in 80% yield. isolated in 80% yield. Contrary to paraformaldehyde, we could not achieve any reaction with 1,3,5-trioxane. Contrary to paraformaldehyde, we could not achieve any reaction with 1,3,5-trioxane. 1,3,5-trioxane.
Scheme 2. Aerobic Aerobic oxidative oxidative coupling coupling between between 1-octanol 1-octanol and paraformaldehyde paraformaldehyde forming forming 1-octyl Scheme 2. Aerobic oxidative coupling between 1-octanol and paraformaldehyde forming 1-octyl formate (2b) catalyzed by Au/TiO Au/TiO22and andthe theproposed proposedmechanism. mechanism. formate (2b) catalyzed by Au/TiO2 and the proposed mechanism.
Acetal 2b[16] [16] is anot a precursor of 2a, formate as it was separately treated with Acetal 2b is not precursor of formate as it was2a, separately treated with paraformaldehyde Acetal 2b [16] is not a precursor of formate 2a, as it was separately treated with paraformaldehyde under conditions identical the reaction, but no reaction was seen. under conditions identical to the reaction, butto reaction was seen. Apparently, from Apparently, the reaction paraformaldehyde under conditions identical tonothe reaction, but no reaction was seen. Apparently, from the reaction between 1-octanol and the formaldehyde, the labile hemiacetal 2c is reversibly formed, between 1-octanol and formaldehyde, labile hemiacetal 2c is reversibly formed, which then from the reaction between 1-octanol and formaldehyde, the labile hemiacetal 2c is reversibly formed, which then undergoes aerobic oxidation catalyzed by Au/TiO 2, leading to formate 2a. The aldehyde undergoes aerobic oxidation catalyzed by Au/TiO , leading to formate 2a. The aldehyde functionality which then undergoes aerobic oxidation catalyzed2 by Au/TiO2, leading to formate 2a. The aldehyde functionality of formate ester is deactivated further nucleophilic attackmolecule, by an alcohol of formate ester is deactivated toward further toward nucleophilic attack by an alcohol as the functionality of formate ester is deactivated toward further nucleophilic attack by an alcohol molecule, as the corresponding dioctyl carbonate ester was not detected by GC-MS. corresponding dioctyl carbonate ester was not detected by GC-MS. molecule, as the corresponding dioctyl carbonate ester was not detected by GC-MS. For catalysts For the the optimization optimization of of the the reaction, reaction, aaa series series of of solvents solvents and and Au Au nanoparticle nanoparticle catalysts [17] [17] was was For the optimization of the reaction, series of solvents and Au nanoparticle catalysts [17] was tested. Au/TiO 2 is more efficient compared to Au/Al2O3 or Au/ZnO. These commercially available tested. Au/TiO is more efficient compared to Au/Al O or Au/ZnO. These commercially available 2 2 3 tested. Au/TiO2 is more efficient compared to Au/Al2O3 or Au/ZnO. These commercially available catalysts havesimilar similar wt content in and Au and average nanoparticle As solvents, per the solvents, catalysts have %% content in Au nanoparticle size. Assize. per the although catalysts have similarwtwt % content in Au average and average nanoparticle size. As per the solvents, although in toluene the reaction is taking place slightly slowly, the product selectivity (2a/2b) is in tolueneinthe reaction taking place slightly slowly, the slowly, productthe selectivity is better than although toluene theisreaction is taking place slightly product(2a/2b) selectivity (2a/2b) is better than in DCE. All other solvents gave inferior results in terms of conversion rates or selectivity in DCE. All other solvents gave inferior results in terms of conversion rates or selectivity (Table 1). better than in DCE. All other solvents gave inferior results in terms of conversion rates or selectivity (Table 1). Theiscatalyst is easily recyclable andafter reusable afterand filtration drying the at290 ◦ C for The catalyst easily recyclable and reusable filtration dryingand in the ovenin 90oven h,°C in (Table 1). The catalyst is easily recyclable and reusable after filtration and drying inatthe oven at 90 °C for 2 h, in five consecutive runs (see footnote of Table 1). A small decline in activity was observed, five2consecutive runs (see footnote Table 1).ofATable small 1). decline in activity observed, for h, in five consecutive runs (seeoffootnote A small declinewas in activity wasapparently observed, apparently due toofsome loss during of material during the recycling process. due to some loss material the recycling process. apparently due to some loss of material during the recycling process. Table Optimization of conditions regarding the Au nanoparticle-catalyzed formylation of Table 1.1. of conditions regarding the Au nanoparticle-catalyzed formylation of 1-octanol Table 1.Optimization Optimization of conditions regarding the Au nanoparticle-catalyzed formylation of 1-octanol (2) with paraformaldehyde. (2) with paraformaldehyde. 1-octanol (2) with paraformaldehyde.
a
Entry Catalyst Solvent Conversion (%) Selectivity (2a/2b) Entry Catalyst Solvent Conversion (2a/2b) Entry Catalyst Solvent Conversion(%) (%) Selectivity Selectivity (2a/2b) 1 Au/TiO2 1,2-Dichloroethane 100 95/5 1 Au/TiO 2 1,2-Dichloroethane 100 95/5 12 Au/TiO 1,2-Dichloroethane 95/5 Au/TiO 2 2 Ethyl acetate 45100 92/8 Au/TiO 2 2 Ethyl acetate 4545 92/8 22 Au/TiO Ethyl acetate 92/8 3 Au/TiO2 Tetrahydrofuran 34 79/21 33 Au/TiO Tetrahydrofuran 34 79/21 Au/TiO 2 2 Tetrahydrofuran 34a,b 79/21 4 Au/TiO 2 Toluene 89 98/2 a,b 44 Au/TiO Toluene 98/2 89 a,b Au/TiO2 2 Toluene 89 98/2 Au/TiO 2 2 Hexane 3434 97/3 55 Au/TiO Hexane 97/3 5 Au/TiO 2 Hexane 34 97/3 66 Au/Al 1,2-Dichloroethane 96/4 Au/Al 2O 3 3 1,2-Dichloroethane 3535 96/4 2O 6 Au/Al 2O3 1,2-Dichloroethane 35 96/4 77 Au/ZnO 1,2-Dichloroethane 98/2 Au/ZnO 1,2-Dichloroethane 5050 98/2 87 Au/Al2 O3 1,2-Dichloroethane Toluene Not98/2 Determined Au/ZnO 505 Au/Al 2O3 Toluene 5 61 Not Determined 98 Au/ZnO Toluene 99/1 8 Au/Al2O3 Toluene 5 Not Determined 9 Au/ZnO Toluene 61 99/1 ◦ b Conversion after 12–13 h, or after 10 h at 80 C. Regarding the 2nd run: 9 100% Au/ZnO Toluene 61 recycling studies, in 99/1 b
100%
Conversion 100% 12–13 h, or after 10 hafter at 80 the run: 2nd 100% run: after 13 h;after 3rd run: 100% conversion 15°C. h; 4thRegarding run: 100% recycling conversionstudies, after 15 in h; 5th Conversion 12–13 h, or after 10 h at 80 °C. b Regarding recycling studies, in the 2nd run: conversion after100% 16 after h. after 100% conversion 13 h; 3rd run: 100% conversion after 15 h; 4th run: 100% conversion after 15 h; 100% conversion after 13 h; 3rd run: 100% conversion after 15 h; 4th run: 100% conversion after 15 h; 5th run: 100% conversion after 16 h. 5th run: 100% conversion after 16 h. a conversion a
Nanomaterials 2017, 7, 440
4 of 10
Subsequently, we tested a series of alcohols to examine the scope and limitations of their Subsequently, tested2 a(2series to examine the scope and limitations of their formylation, formylation, using we Au/TiO mol of %)alcohols as the catalyst, 3 equiv of paraformaldehyde, toluene as the using Au/TiO (2 mol %) as the catalyst, 3 equiv of paraformaldehyde, toluene as the solvent, and 2 solvent, and heating to 80 °C for 10 h (Figure 1). Primary alcohols are, in general, good substrates ◦ heating to 80the C corresponding for 10 h (Figureformate 1). Primary are, in92–99%. general,Primary good substrates provide and provide estersalcohols in selectivity benzylic and alcohols are the corresponding formate esters in selectivity 92–99%. Primary benzylic alcohols are competitively competitively oxidized into the corresponding benzaldehydes in a relative ratio of ~15–20%, while oxidized into of theother corresponding relative ratio of ~15–20%, theSecondary formation the formation minor side benzaldehydes products dropsin thea yield of formates to aroundwhile 50–55%. of other minor products drops yield reasons of formates around 50–55%. Secondary alcoholsand are alcohols are lessside prone to react due the to steric and to require 5 equiv of paraformaldehyde less prone to react due to steric reasons and require 5 equiv of paraformaldehyde and ~20 h to reach ~20 h to reach completion. Phenols are rather unreactive even using a large excess of completion. Phenols(10 areequiv) rather unreactive evenreaction using a large ofthese paraformaldehyde (10 of equiv) and paraformaldehyde and extended times;excess under conditions
