Synthesis of Ethylene Glycol from Syngas via Oxidative Double Carbonylation of Ethanol to Diethyl Oxalate and Its Subsequent Hydrogenation.

This work reports a novel sustainable two-step method for the synthesis of ethylene glycol (EG) using syngas. In the first step, diethyl oxalate was selectively synthesized via oxidative double carbonylation of ethanol and carbon monoxide (CO) using a ligand-free, recyclable Pd/C catalyst. In the second step, the diethyl oxalate produced underwent subsequent hydrogenation using [2-(di-tert-butylphosphinomethyl)-6-(diethylaminomethyl)pyridine]ruthenium(II) chlorocarbonyl hydride to get EG and ethanol. Thus, the generated ethanol can be recycled back to the first step for double carbonylation. This method gives a sustainable route to manufacture EG using carbon monoxide and hydrogen.

Introduction

In the industrial technology perspective, production of chemicals based on C1 chemistry from abundant carbon sources such as coal, natural gas, biomass, and various solid waste is an ideal alternative to the petrochemical-based production due to its green nature.[1−3] Increase in the global demand of bulk chemicals has compelled the researchers to innovate various new and alternative strategies to prepare these bulk chemicals. Ethylene glycol (EG) synthesized using such easily available and cost-efficient chemical building blocks can substitute the production of EG derived from traditional petroleum sources.[4−6] EG is a much important commodity chemical because of its vast applications, such as antifreeze agents, solvents, in manufacturing of heat transfer agents, and as a precursor for the manufacture of polyester fiber, poly(ethylene terephthalate) (PET) resin.[2,7] Due to increase in population, the demand of PET resin and polyester fibers has increased significantly, which also leads to a constant growth of EG production.[8,9] Currently, hydrolysis of petroleum-based ethylene oxide derived from ethylene is used for the commercial production of ethylene glycol (Scheme a).[7,10,11] The major drawback associated with the process is the use of harsh reaction conditions and of large by-product generation. Although synthesis of EG from syngas under reductive conditions is an ideal method, but low reactivity and selectivity limit its applications. Syntheses of EG from methanol, formaldehyde, and methyl formate are some alternate methods, but again, low reactivity and selectivity limit their applications also.[12−15] In the current scenario, carbonylation chemistry is used in industry for producing many useful products.[16] Palladium-catalyzed oxidative carbonylation is of great interest, which applies different organic nucleophiles or electrophiles in the presence of carbon monoxide (CO) and oxidation reagents to prepare various carbonyl-containing compounds.[17,18]

Scheme 1

Previous Approach for EG: (a) Traditional Methods, (b) Carbonylation and Hydrogenation Method

Recently, Prof. Beller and his group have developed a novel method for the synthesis of EG by catalytic hydrogenation of oxamide using Ru- and Fe-based catalysts for the first time (Scheme b).[19] In our previous work, we have published a similar two-step approach for EG synthesis by oxidative cross double carbonylation of amines and alcohols followed by catalytic hydrogenation of oxamates using a highly active ruthenium PNN pincer catalyst prepared by Milstein and co-workers and used for amide and ester hydrogenation.[20,21] In our laboratory, we have synthesized oxamates successfully by oxidative cross double carbonylation of alcohols and amines using Pd/C as a heterogeneous and recyclable catalyst.[22] On the basis of the previous results with oxamide and oxamate, we thought that these processes can be further simplified using oxalates for such an application.[23,24] Traditionally, oxalates are prepared by esterification of oxalic acid or oxalyl chloride. The disadvantage of this method is the use of thermally unstable oxalyl chlorides. Other alternatives like nitric oxide-mediated carbonylation of alcohols to dialkyl oxalates, using palladium complexes, are preferred in industry because of its higher reactivity.[25] However, the major limitation of using nitric oxide is that it is not environmentally friendly due to its highly corrosive nature, and there is a requirement of a specific quality of the material, which adds complication to this process. Hence, this work is focused on the production of EG selectively using the process from CO without the use/formation of unstable reagents.

On the basis of the current progress in the field of carbonylation and hydrogenation reactions and to extend our interest in this area, diethyl oxalate was used as a key intermediate for the synthesis of EG via oxidative double carbonylation of ethanol using the Pd/C catalyst under ligand-free conditions and subsequent hydrogenation to EG using [2-(di-tert-butylphosphinomethyl)-6-(diethylaminomethyl)pyridine]ruthenium(II) chlorocarbonyl hydride (Milstein’s catalyst) (Scheme ).

Scheme 2

Oxalate-Mediated Approach to EG

Results and Discussion

In this work, diethyl oxalate was synthesized by oxidative double carbonylation of ethanol, using Pd/C as a heterogeneous catalyst (Scheme ).

Scheme 3

Oxidative Double Carbonylation of Ethanol Using the Pd/C Catalyst

Preliminary studies were carried out using the Pd/C catalyst (5 and 10% loading) for the double carbonylation reaction using ethanol in the presence of CO/O2 (25:6 ratio), as shown in Table . The catalyst (10 mol %, 10% Pd/C) provides excellent yield of the desired product (Table , entry 4). The lower yield of the oxalate was obtained due to low catalyst loading. Molecular oxygen and the iodide additive along with the Pd/C catalyst play a significant role in oxidative carbonylation reactions.[26] In the presence of iodide promoters such as sodium iodide, potassium iodide, and tetrabutylammonium iodide (TBAI), Pd/C was found to be an effective catalyst (Table , entries 1–3). The tetrabutylammonium iodide was found to be an excellent promoter for the present carbonylation reaction (Table , entry 3). Results in the presence of TBAI might be due to the “soft” binding nature of iodide and because it is electron-rich, polarizable, and a good nucleophile compared with the other halides. For the effective progress of the reaction, the iodide promoter is an essential requirement without which the reaction never proceeds[17b] (Table , entry 4). The plausible mechanism for double carbonylation in the presence of the TBAI promoter is shown in Figure . Initially, the iodides get adsorbed on the Pd surface and oxidative addition of the first molecule of ethanol takes place to generate metal hydride species. This metal hydride species undergoes double CO insertion. Finally, reductive elimination of diethyl oxalate from the Pd surface takes place in the presence of the second molecule of ethanol and oxygen. The iodides remain on the Pd surface as studied by X-ray photoelectron spectroscopy (XPS) analysis and continue the catalytic cycle.

Table 1

Effect of Dose of Pd/C on the Reactiona

entry	catalyst	catalyst loading (mol %)	yield (%)b
1	5% Pd/C	10	82
2	10% Pd/C	4	57
3	10% Pd/C	6	81
4	10% Pd/C	10	90
5	10% Pd/C	12	90

Reaction conditions: ethanol (10 g, 217 mmol), TBAI (0.2 mmol), CO/O2 (25:6 atm), temperature 70 °C, time 8 h.

Gas chromatography (GC) yield.

Table 2

Optimization of the Pd/C-Catalyzed Oxidative Double Carbonylation Reactiona

entry	additive (mmol)	temp (°C)	pressure	time (h)	yield (%)b
Effect of Additive
1	NaI	70	25	8	81
2	KI	70	25	8	83
3	TBAI	70	25	8	90
4		70	25	8
Effect of Temperature
5	TBAI	60	25	8	79
6	TBAI	80	25	8	90
Effect of Pressure
7	TBAI	70	5	8	15
8	TBAI	70	15	8	45
Effect of Time
9	TBAI	70	25	7	86
10	TBAI	70	25	9	90

Reaction conditions: ethanol (10 g, 217 mmol), 10% Pd/C (10 mol %), TBAI (0.2 mmol), CO/O2 (25:6 atm), temperature 70 °C, time 8 h.

GC yields.

Figure 1

Plausible mechanism of Pd/C-catalyzed double carbonylation in the presence of TBAI promoters.

In the Pd/C-catalyzed double carbonylation reaction, temperature plays an important role for the reaction to proceed (Table , entries 5 and 6). It was observed that at 70 °C the maximum yield of the desired product was obtained and the diethyl oxalate was obtained with 90% yield in 8 h (Table , entry 3). No significant effect on the yield of the product was observed by further increasing the reaction temperature (Table , entry 6). Pressure was found to be critical in this process. At 5 atm pressure, less yield of the oxalate was obtained, whereas on further increasing the pressure up to 25 atm, the product was obtained in 90% yield (Table , entries 3, 7, and 8). The reaction requires 8 h to get maximum conversion of ethanol into diethyl oxalate (Table , entry 3). At 7 h reaction time, the yield was reduced to 86% (Table , entry 9). Longer reaction time had no profound effect on the yield and quality (Table , entry 10).

To make a process more economical, the recyclability study of the catalyst plays an important role. In this protocol, under optimized reaction conditions, the recyclability of the Pd/C catalyst has been studied. After four consecutive recycles, the Pd/C catalyst was found to be effective without loss in performance activity (Scheme ). No significant leaching of the Pd metal was observed in the product mixture after completion of first and fourth recycle runs. Pd was found below the detectable limit (0.01 ppm) reveled by inductively coupled plasma atomic emission spectroscopy analysis. The transmission electron microscopy (TEM) analysis of fresh and recycled catalysts showed that the components of Pd were uniformly distributed over the carbon surface, and no agglomeration was seen. Structural changes of fresh, first, and fourth recycled Pd/C catalysts were studied by X-ray diffraction (XRD). The XRD pattern showed the diffracted peaks for the Pd/C catalyst, and no significant structural change was found in fresh and recycled catalysts. The composition of elements Pd, C, and O in the catalyst used in this double carbonylation reaction is analyzed by XPS (Figure a). The spectra of element Pd at 336 and 341 eV represent Pd2+ in the fresh Pd/C catalyst (Figure b). Shifting of Pd catalyst peaks to 334.4 and 339.6 eV after first and fourth recycles, which are assigned as 3d5/2 and 3d3/2 for Pd0 species, indicates that the catalyst is reduced but the activity of Pd species remains constant for the Pd reaction.

Scheme 4

Recycle Study of Pd/C-Catalyzed Synthesis of Diethyl Oxalate

Figure 2

Spectra of the fresh and reused Pd/C catalysts: (a) wide scan; (b) Pd 3d; (c) C 1s; (d) O 1s; and (e) I 3d.

This suggests that double carbonylation of ethanol was promoted by the Pd(0) species present on the carbon support. One would expect the activity of Pd to increase in the fourth cycle when there is more Pd0 in the system, but due to some handling loss of the catalyst during the recycle study, yields are got to somewhat decreased site. Figure c shows the presence of carbon species in fresh and reused catalysts. Figure d shows the presence of oxygen species. Generally, O2 adsorption on Pd happens as O2 or O species. Moreover, H2O also gets adsorbed on the Pd surface. Several types of O-containing species are likely to form. Some of them are easily removed, and Pd sites are available for catalytic reactions. The others may be difficult to remove and continue to stay on Pd. This might be the reason for the presence of two O species initially, and on continuous use of recycling, the fourth cycle showed one species. For the first and fourth recycled catalysts, two intensive peaks appear at 617.8 eV (I 3d5/2) and 629.2 eV (I 3d3/2), which were assigned to the I 3d region, indicating that adsorption of iodide atoms takes place on the surface of carbon (Figure e).

Furthermore, we have investigated the hydrogenation of diethyl oxalate using commercially available catalysts, which are efficient for the hydrogenation process. It is reported that for hydrogenation of esters, amides, and carboxylic acid, a Ru-based organometallic pincer complex is an efficient catalyst.[27] The oxalate contains ester functionalities. On the basis of the literature survey, different Ru-complex catalysts were screened for hydrogenation of oxalates (Table ). We have explored various Ru precursors and P-coordinating ligands such as triphos, PPh3, and 2,2′-bis(diphenylphosphino)-1,1′-binaphthyl (BINAP) in the present hydrogenation reaction, and the results are summarized in Table . Ru(OAc)2, RuCl2, and Ru(acac)3 were found to be ineffective catalysts along with the P-ligands (Table , entries 1–4). The pincer complexes such as carbonylhydrido(tetrahydroborato)[bis(2-diphenylphosphinoethyl)amino]ruthenium(II) (Ru-MACHO-BH) and [2-(di-tert-butylphosphinomethyl)-6-(diethylaminomethyl)pyridine]ruthenium(II) chlorocarbonyl hydride (Milstein’s catalyst) were found to be highly active catalysts for the present protocol. Lower yields of the desired product were observed using Ru-MACHO-BH and Ru-MACHO as catalysts (Table , entries 5 and 7), whereas the yield of EG was increased sharply to 92% when 1 mol % Milstein’s catalyst was added (Table , entry 6). The catalyst loading of 1 mol % was found to be optimal as the yield reduced to 68% by decreasing the catalyst loading to 0.5% (Table , entries 1 and 2). No profound effect was observed on the yield of the product with an increase in catalyst loading (Table , entry 3).

Table 3

Study of Catalystsa

entry	catalyst	conversion (%)	yield (%)b
1	Ru(OAc)2/BINAP	76	9
2	Ru(acac)3/triphos	70	6
3	Ru(OAc)2/PPh3	61	trace
4	RuCl2/triphos	71	trace
5	Ru-MACHO-BH	95	83
6	Milstein catalyst	100	92
7	Ru-MACHO	89	71

Reaction conditions: oxalate (1 mmol), catalyst (0.01 mmol), ligand (where appropriate; 0.02 mmol), KOBu (0.05 mmol), ethanol (10.0 mL), H2 (40 bar), 100 °C, 14 h.

GC yield of EG.

Table 4

Study of Catalyst Loadinga

entry	catalyst loading (mol %)	conversion (%)	glycol (%)b
1	0.5	81	68
2	1.0	100	92
3	2.0	100	92

Reaction conditions: oxalate (1 mmol), KOBu (0.05 mmol), ethanol (10.0 mL), H2 (40 bar), 100 °C, 14 h.

GC yield of EG.

The influence of different solvents was studied. Among several other selected solvents, ethanol was found to be the best solvent (Table , entries 1–3). The solvents like toluene and tetrahydrofuran (THF) provided moderate yield of the desired product. The selection of ethanol also suits to the entire process as it brings uniform solvent for both the steps and hence it can be recyclable. The role of bases in the present reaction was also studied, and it was found that replacement of KOBu with other bases such as KOH, NaOH, and K2CO3 drastically reduced the yield of the product (Table , entries 3, 4–6). Due to better solubility of KOBu, both in aqueous and organic media, and its higher basicity, the reaction proceeds well in the presence of this base. Other bases failed to give complete reduction of oxamate to ethylene glycol as most of the inorganic bases have poor solubility in organic solvents. The catalyst loading of 1.0% was optimal for getting high yield of the product. A 2% catalyst loading did not show further improvement in the yield. It might be due to the deactivation/degradation of the substrate (oxalate) at high catalyst loading.

Table 5

Study of Different Reaction Parametersa

entry	solvent	base	temperature (°C)	pressure (atm)	yield (%)b
Study of Solvents
1	THF	KOtBu	100	40	69
2	Toluene	KOtBu	100	40	90
3	EtOH	KOtBu	100	40	92
Study of Bases
4	EtOH	KOH	100	40	53
5	EtOH	NaOH	100	40	61
6	EtOH	K2CO3	100	40	trace
Study of Temperature
7	EtOH	KOtBu	80	40	79
8	EtOH	KOtBu	120	40	94
Study of Pressure
9	EtOH	KOtBu	100	30	77
10	EtOH	KOtBu	100	50	92

Reaction conditions: oxalate (1 mmol), Milstein’s catalyst (0.01 mmol), base (0.05 mmol), solvent (10.0 mL), H2 (40 bar), 14 h.

GC yield of EG.

The effect of temperature on the yield of the desired product has been studied. Experiments were carried out at different temperatures ranging from 80 to 120 °C, in which 100 °C reflects to be the most favorable temperature for the reaction (Table , entries 4, 7, and 8). No significant effect on the yield of the desired product was observed by increasing the temperature up to 120 °C (Table , entry 8). Pressure also plays an important role in the effective progress of reaction and yield of the product. A lower yield of the product was obtained when the reaction was carried out at a lower H2 pressure (30 atm) (Table , entry 9). No major impact on the yield of the desired product was observed on increasing the pressure (50 atm) (Table , entry 10). A H2 pressure of 40 atm was kept for all reactions at it provides the maximum yield of the desired product. The optimized time period was 14 h to achieve the maximum yield of the desired product.

Conclusions

In conclusion, an efficient two-step approach for the synthesis of EG from CO and H2 via oxalate-mediated hydrogenation has been developed. Notably, in the first step, the synthesis of diethyl oxalate was achieved using Pd/C as a heterogeneous catalyst. Pd/C was easily separated from the reaction mixture and reused four times, demonstrating the recyclable and inexpensive protocol. The hydrogenation of oxalate to EG was achieved under mild conditions compared with the previously reported oxamide- and oxamate-mediated hydrogenation.

Experimental Section

Materials and Methods

All reagents and materials were commercially available and were used as received. Pd/C (10 wt % loading, matrix: activated carbon support, product number: 205699, brand: Aldrich) was purchased from Sigma-Aldrich. Gas chromatography PerkinElmer Clarus 400 GC equipped with a flame ionization detector and capillary column (30 m × 0.25 mm × 0.25 μm) and thin layer chromatography using Merck silica gel 60 F254 plates were used for monitoring the progress of the reaction. Column chromatography on silica gel (100–200) mesh has been used for the purification of the product. However, known compounds were confirmed by comparing with their authentic samples on GC and GC–mass spectrometry (MS). Shimadzu LCMS-2010EV instrument (column length, 50 mm; internal diameter, 4.6 mm; particle size, 3 μm; nebulizing gap, 1.5 L/min; vacuum, 10–3 Pa) (column flow 1.2 mL/min; serial temperature, 260 °C; and heat block temperature, 200 °C) and GCMS-QP 2010 instrument (Rtx-17, 30 m × 25 mm ID; film thickness, 0.25 μm df) (column flow, 2 mL/min; 80–240 °C at 10°/min rise) were used to get mass spectra. The XPS of Pd/C was measured using a PHI5000 Versa Probe with a monochromatic focused (100 μm × 100 μm) Al Kα X-ray radiation (15 kV, 30 mA) and dual beam neutralization using a combination of argon ion gun and electron irradiation. X-ray diffraction (XRD) patterns were recorded on Shimadzu XRD-6100 using Cu Kα radiation, 1/4 1.5405 Å, with a scanning rate of 2°/min and 2θ angle ranging from 10 to 85 with current 30 mA and voltage 40 kV. A transmission electron microscope (JEOL JEM-2100) operating at 200 kV was used to obtain bright field images of the catalyst.

Two-Step Oxalate-Mediated Production of EG from CO

Step 1: A reaction mixture containing ethanol (10.0 g, 217 mmol), tetrabutylammonium iodide (0.2 mmol), and 10% Pd/C (10 mol %) was prepared and kept in a 100 mL stainless steel autoclave. The autoclave was closed and pressurized with oxygen (6 atm) and CO (25 atm) without flushing. The reaction mixture was stirred with a mechanical stirrer for 8 h at 70 °C. The pressure was released carefully after cooling to room temperature. The filtered catalyst was washed with ethanol (25.0 mL), vacuum-dried, and used for the next cycle. The yield (90%) of the corresponding oxalate formed was confirmed by GC analysis.

Step 2: To the reaction mass was added 80 mL of ethanol followed by stirring for 30 min. The resulting reaction mass was filtered through silica gel (2–3 cm pad). The residue was washed with ethanol (20 mL). The concentration of oxalate in ethanol was found to be 0.143 g/mL (0.01 M), confirmed by GC analysis. An ethanol solution (10 mL) of oxalate (10 mmol) was added to Milstein’s catalyst (0.01 mmol) and KOBu (0.05 mmol) under an argon atmosphere, and the reaction mixture was kept in an autoclave. To the reaction mixture, 15.0 mL of ethanol was added. The autoclave was flushed with H2 three times and then pressurized with H2 to 40 atm. The reaction was carried out at 100 °C for 14 h. Full conversion of oxalate was detected by GC analysis. The reaction mass was then cooled to room temperature, and pressure was released. The reaction mass was passed through the silica gel (2–3 cm pad) and washed with ethanol. The desired product was separated from ethanol by distillation with a purity of >92%. The recovered ethanol was used for the next cycle.