Catalytic Epoxidation of a Technical Mixture of Methyl Oleate and Methyl Linoleate in Ionic Liquids Using MoO(O2)2·2QOH (QOH = 8-quinilinol) as Catalyst and NaHCO3 as co-Catalyst

The oxo-diperoxo molybdenum(VI) complex MoO(O2)2·2QOH (QOH = 8-quinilinol) was prepared and characterized by elemental analysis, IR and UV-Vis spectra. The ionic liquids (ILs) [bmim][BF4], [hydemim][BF4], and [bmim][PF6] were characterized by 1H-NMR and UV-Vis spectra. The epoxidation of a technical mixture of methyl oleate and methyl linoleate with H2O2, in [bmim][BF4], [hydemim][BF4] and [bmim][PF6], catalyzed by MoO(O2)2·2QOH (QOH = 8-quinilinol) and with NaHCO3 as co-catalyst has been studied for the first time. It was found that high conversions of methyl oleate and methyl linoleate to their respective oxidation products, as well as the total selectivity of their oxidation products to oxirane in [hydemim][BF4] were obtained. Also, the IL phases containing the Mo(VI) catalyst can be readily recycled by washing with diethyl ether and drying, and the Mo(VI) catalyst can be reused at least five times.


Introduction
In recent years, much emphasis has been placed on the utilization of renewable resources as an alternative to petrochemical-based feedstocks. In this context, vegetable oils are undoubtedly promising candidates. Not only they are widespread, inexpensive, eco-compatible, renewable and non-

OPEN ACCESS
noxious, but, above all, several multi-functionalised molecules can be obtained by their chemical modification. For example, the unsaturated olefinic bonds in vegetable oils can be epoxidized to give epoxy oils, which have been already used widely as PVC-plasticizers and stabilizers, as reactive diluents for paints, as additives in lubricants, etc. [1] Traditionally, the only commercial source of epoxidized oils is based on the Prileshajev peracid process [2]. This procedure, however, has several drawbacks that require improvement: (i) selectivity to epoxidized products is relatively low due to acid-catalyzed oxirane ring opening side reactions; (ii) the separation of acidic by-products, whose presence may be detrimental for further applications, is not easy; (iii) the handling of highly-concentrated hydrogen peroxide and strong acids is dangerous and causes corrosion problems. For all these reasons there is considerable interest in finding an alternative epoxidation route for these substrates [3][4][5][6].
To this end, efficient metal-catalyzed epoxidations of vegetable oils have been investigated, especially using high-valence catalysts based on titanium [7], tungsten [8] and rhenium [9]. The use in oxidation of molybdenum(VI) catalysts [10] containing the stable oxo-peroxo core [11] is wellestablished, and with them a variety of organic substrates [12][13][14], including alkenes, alcohols, amides and nitro compounds, can be catalytically oxidized under homogeneous as well as heterogeneous conditions [15]. Recently, several interesting oxo-peroxo molybdenum(VI) complexes such as [MoO(O 2 ) 2 L 2 ] (L = 8-quinilinol) [16], have proven to be excellent room temperature (rt) catalysts for the epoxidation using the H 2 O 2 /NaHCO 3 system of many alkenes such as styrene, 1-octene, etc. in CH 3 CN solutions. Not only is this catalyst-oxidant system economical and the reaction conditions mild, but hydrogen peroxide is a readily-available and green oxidant [17], so, it would be undoubtedly noteworthy if this catalytic system could also be used in the epoxidation of vegetable oils, an area in which to our knowledge no work has been conducted.
Furthermore, increased awareness in recent years of the detrimental environmental effects of organic solvents has resulted in rapid growth of research on alternative reaction media, and more attention has been given to the reusability of solvents and catalysts for the development of costeffective protocols. Ionic liquids (ILs) are a promising alternative to classical solvents because ILs, especially those based on 1,3-dialkylimidazolium cations, such as 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim][BF 4 ]) and its hexafluorophosphate analog ([bmim][PF 6 ]), are air and moisture stable, and possess negligible vapor pressure, low viscosity, high thermal and chemical stability and a wide electrochemical window [18][19][20][21]. They can also be recycled in many cases. Various classical reactions such as Friedel-Crafts [22], Diels-Alder [23], alkylation [24], hydrogenation [25], Heck [26], hydroformylation [27] and oligomerisation [28] using ILs as solvents have been reported. They have also emerged as ideal immobilization media for catalysts, which combine the advantages of ILs and heterogeneous supports, such as increased reaction rates, enhanced catalytic efficiency, simplified product separation together with improved catalyst recyclability, therefore, the use of ILs as solvents in organic reactions has become a particular focus of research [29].
To our knowledge, however, there is no literature describing the catalytic epoxidation of vegetable oils using ILs as reaction media/immobilization media for transition metal complex catalysts. In this paper, we report the epoxidation of a technical mixture of methyl oleate and methyl linoleate with The UV-Vis spectra of 8-quinilinol in CH 2 Cl 2 displayed two shoulders at 242 and 311 nm ( Figure 1, curve a), while the oxo-diperoxo molybdenum(VI) complex [MoO(O 2 ) 2 ·2QOH] (QOH = 8-quinilinol) showed two strong absorption bands at 244 and 380 nm ( Figure 1, curve b), and a small absorption band at 315 nm. Further, it showed a very broad low intensity absorption ranging from 300 to 500 nm. The 380 nm absorption band could be assigned to the ligand-to-metal charge transfer [16]. [MoO(O 2 ) 2 ·2QOH] (QOH = 8-quinilinol) was found to be slightly soluble in the ILs used, but almost immiscible with substrate at rt. When hydrogen peroxide is added, interestingly, it dissolves immediately and gives an orange solution. To find out more about the mechanism of reaction and confirm the nature of the reactive intermediate, the UV-Vis spectrum of the reaction mixture resulting from the addition of substrate to a solution of catalyst and hydrogen peroxide was studied. Two strong absorption bands at 240 and 308 nm (Figure 1, curve c) were observed. Similarly, a small absorption band was found at 362 nm. Additionally, compared with [MoO(O 2 ) 2 ·2QOH] (QOH = 8-quinilinol), the absorbance of the peak in the 300 to 500 nm region was decreased considerably. It is believed that the new small peak appearing at 362 nm (lit. [31] 364 nm), corresponds to the active oxo-dioxo molybdenum species which has been proposed as the catalyst precursor [32], and is generated from the oxo-diperoxo molybdenum species. Each double bond can be epoxidized in the presence of monooxodiperoxo species, and the intermediate oxo-dioxo species formed regenerates the monooxodiperoxo centers.  Table 1 shows the results of epoxidation of methyl oleate and methyl linoleate. In all the cases, the conversion of the former to its oxidation products was higher than that of the latter, a result of the difference in the number of double bonds in these two unsaturated esters. Without any added solvent the total conversion of methyl oleate and methyl linoleate to their oxidation products was low (41%, Entry 1), and with the use of only NaHCO 3 without catalyst, no conversion of methyl linoleate or methyl linoleate was observed (Entry 2). Further, the catalytic efficiency when H 2 O 2 is used as a sole oxidant is poor (Entries 6 and 7). When NaHCO 3 is used together with H 2 O 2 , however, the efficiency of the system is greatly increased. The key aspect of such a reaction is that H 2 O 2 and NaHCO 3 react in an equilibrium process to produce peroxymonocarbonate HCO 4 − which is a more reactive nucleophile than H 2 O 2 and speeds up the epoxidation reaction (Scheme 3) [33,34].  Figure 2) were purchased from the Shanghai Chengjie Chemical Co., Ltd, and the mass fraction purities were all above 99%. The mass fraction of water in these ILs was less than 1 × 10 3 ppm. Fatty acid methyl esters (C. P.) was purchased from the Beijing Jinlong Chemical Co., Ltd., and contained methyl oleate (28.26%), methyl linoleate (19.43%), methyl palmitate (32.69%), methyl octadecanoate (7.28%), methyl dodecanoate (7.05%), methyl tetradecanoate (4.29%), iso-propyl myristate (0.53%) and decanoic acid 2-ethylhexyl ester (0.47%) according to GC-MS analysis. Standard samples of methyl oleate (100%, GC-MS) and methyl linoleate (99%) were purchased from Alfa Aesar and Sigma, respectively. Hydrogen peroxide (G. R., 30%) and the other reagents (A. R., >95%) were all provided by the Beijing Chemical Reagents Company and used without further purification.

Instruments
Chemical analyses for C, H, and N were performed by elemental microanalysis (Elementar Vario EL analyzer). IR spectra were measured on a Nicolet MAGNA 750 instrument fitted with a Nic-plan IR microscope. The electronic spectra were recorded on a 756 PC UV/VIS spectrometer. 1 H NMR spectra were recorded on a Varian Unity Inova-400 spectrometer. The reactions mixtures and their respective fatty acid methyl ester contents were analyzed by GC-MS [DSQ (Thermo Fisher)] with a capillary column (DB-5, 30 m × 0.25 µm × 0.25 mm), equipped with an FID detector. Oven program: 100 °C for 2 min, 5 °C/min to 300 °C and held for 10 min. Both the injector port temperature and the detector transfer line temperature were 250 °C, and helium was used as carrier gas, l mL/min. Spilt sampling, split ratio was 50:1. Sample volume was 0.1 µL. MS: electron ionization, 70 eV. The reaction products were analyzed on a GC2010 with a capillary column (AB-1, 30 m × 0.25 µm × 0.25 mm), equipped with an automatic sampler (AOC-20i) and an FID detector. Oven program: 100 °C for 2 min, 5 °C/min to 300 °C and held for 10 min. Both the injector port temperature and the detector transfer line temperature were 250 °C, and helium was used as carrier gas, 2 mL/min. Spilt sampling, split ratio was 50:1. Sample volume was 2 µL.

Epoxidation
Catalyst (0.265 µmol) and NaHCO 3 (7.950 mmol) were placed in a 50 mL round bottomed flask equipped with a magnetic stir bar. The catalyst was dissolved in the IL (2 mL) by heating and then cooled to rt. To this substrate (ca. 11.66 g) was added and the the reaction mixture stirred at 30 °C. To maintain a constant flow of peroxide during the entire period of reaction, 30 % H 2 O 2 (106 mmol) was added droppedwise periodically over 2 h, then the reaction mixture was cooled to rt and extracted with diethyl ether (3 × 20 mL). The diethyl ether layers were combined and concentrated under reduced pressure, and a sample was taken for GC analysis and measurement of oxirane oxygen (see Appendix). The IL phase including the catalyst was washed with diethyl ether, and dried prior to being recycled.

Determination of the conversion of methyl oleate or methyl linoleate
The relative percentage conversion of methyl oleate or methyl linoleate to its oxidation products, C 1 or C 2 , can be calculated by the following expressions: where m 1 ' and m 2 ' are the content of methyl oleate and methyl linoleate unconverted after the epoxidation reaction, respectively. By comparing the retention times of their standard samples with the reaction products, both the content of methyl oleate and methyl linoleate can be determined by GC, using the area normalization method. In addition, m 1 and m 2 are the initial content of methyl oleate and methyl linoleate in the substrate, respectively. They were known by GC-MS analysis (m 1 = 28.26% and m 2 = 19.43%, respectively). So: Since the number of double bonds for each mole of methyl oleate is one, while for methyl linoleate it is two, the total conversion of all the double bonds of these two esters, C t , can be calculated by the following expression: where M 1 ', M 2 ' are the amount of methyl oleate and methyl linoleate converted (mol), respectively. M 1 and M 2 are the initial amount of methyl oleate and methyl linoleate (mol), respectively. They can be calculated respectively by the following four expressions:

Determination of the total selectivity of oxidation products to epoxidized methyl oleate and methyl linoleate
The total yield of the different epoxides can be known by measuring the epoxy value (E) of reaction product, according to the hydrochloric acid -acetone method [35]. E is defined as the content of oxirane oxygen in a sample measured (g/100 g sample): The total percentage yield to oxirane, i.e., the total selectivity of reaction product to oxirane (S t ), can be calculated by following expression: where E e and E c are the experimentally obtained oxirane oxygen and the maximum oxirane oxygen calculated, respectively. E e can be obtained by the following expression: where V is the volume of sodium hydroxide standard solution consumed (mL) for measuring the oxirane oxygen of the sample; V 1 is the volume of sodium hydroxide standard solution consumed (mL) in the blank test for determinating the oxirane oxygen; V 2 is the volume of sodium hydroxide standard solution consumed (mL) for determinating the acid value of the sample; G is the mass of sample for determinating the acid value (g); N is the concentration of sodium hydroxide standard solution, mol/L; 0.016 is milligram equivalent of oxygen. The acid value of the sample, X (mg/g sample), is defined as the mass of potassium hydroxide consumed in neutralizing 1 g of sample, and determined by the addition of potassium hydroxideethanol standard solution to the sample [36]: where C is the concentration of potassium hydroxide-ethanol standard solution (mol/L); V 2 is the volume of potassium hydroxide-ethanol standard solution consumed (mL) for determinating the acid value of the sample; V 0 is the volume of potassium hydroxide-ethanol standard solution consumed (mL) in the blank test for determinating the acid value of the sample; M KOH is the molecular mass of potassium hydroxide, 56.11 g/mol.
Since the molecular weight of the oxygen in each mol of epoxidized methyl oleate is 16, while that in epoxidized methyl linoleate is 32, E c can be calculated by the following expression: