(679f) On the Influence of Trialkylamine Reduction Strategies in the Direct Hydrogenation of CO2 to Formic Acid

Nowadays a major search is on the way to develop a method to transform CO_2, which is often seen as a waste product due to anthropogenic emissions, into valuable chemicals [1]. Previous work has shown CO₂ can be transformed into a wide array of added-value chemicals by incorporation of alkanes, alkynes and epoxides [2].Alternatively, CO₂ can also be used to store renewable energy in liquid fuels using hydrogen as an energy carrier [1][3], with examples being dimethyl ether (DME), methanol and formic acid [1].

Formic acid is often seen as a promising organic molecule to store hydrogen, as it readily allows for a reversible transformation back to hydrogen and CO₂ besides being a valuable bulk chemical for preservatives and antibacterial agents [1]. Formic acid is known to store 4.3 wt% hydrogen and is liquid at ambient conditions, therefore allowing straightforward storage and transport [1]. Despite methanol having a higher gravimetric hydrogen density as well as a greater volumetric hydrogen density, formic acid is a less dangerous molecule due to it being much less flammable. Besides a safety factor, when synthesizing methanol from CO_2, part of the hydrogen is consumed due to water formation leading to decreased economic and ecological feasibility [1], thus making formic acid a highly promising candidate as an energy carrier.

Though the number of heterogeneous catalysts for CO₂ hydrogenation to formic acid reported is yet relatively limited. The most popular active metals studied are Pd, Au and Ru on supports such as activated carbon, alumina and titania. Pd and Au are the most used catalysts, with the catalytic performance showing a strong dependency on the support material. Here, the most promising supports were found to be dependent on the active metal, with Pd preferring hydrophobic carbon-based supports, while Au should be combined with a hydrophilic support such as Al₂O₃ and TiO₂ [1][5]. Recently, promising results showing selectivity’s > 99% were obtained by several research groups using gold based supported catalysts [6][7].

Even though the state-of-the-art catalysts seem sufficiently active and selective [1][4][5], the equilibrium concentrations reached remain low due to the reaction being thermodynamically unfavorable [4]. One of the commonly employed strategies is to reduce formic acid within the reaction mixture through the formation of adducts/complexes [4]. Often, nitrogenous bases or alkali is used for this purpose [4][5]. Promising results have been obtained using tertiary alkylamines [6][7].

While promising results were obtained for triethylamine (NEt₃) in previous works [6][7], the formed salt does not allow for direct thermal splitting [6][8]. Low-boiling amines such as NEt₃ result in stable azeotropes, rendering separation through (vacuum) distillation impossible, thereby significantly complicating separation processes [6]. Previous works have shown that the usage of NHex₃ (trihexylamine) allows for thermal cleavage under mild conditions (150 ℃, 150 mbar) [8], therefore enabling the formic acid to be readily separated from the reducing agent.

From here it readily becomes apparent that the ability to separate the produced formic acid and the tertiary alkylamine is crucial for the economic viability of the proposed process, providing the motivation for this work. Here, the intent is to allow for thermal cleavage of the produced adducts, while maintaining the promising yields found in previous works, such as [6][7]. To gain a deeper understanding into key reaction mechanisms, the aim is to elucidate the influence of the tertiary alkylamine hydrocarbon length on key reaction parameters such as CO₂ conversion and selectivity towards formic acid.

In order to gain more insight into the formation of tertiary alkylamine – formic acid adducts, 4 alkylamines having increasing hydrocarbon chain lengths were studied through the addition of different formic acid ratios. The tertiary alkylamines had a hydrocarbon chain length between 2 and 8, with the studied amines being: triethylamine (NEt₃), tributylamine (NBut₃), trihexylamine (NHex₃) and trioctylamine (TOA).

For all tertiary alkylamine – formic acid systems the critical molar fraction of acid was established for which phase separation occurs, with the system becoming biphasic. In order to determine the chemical composition of each phase ¹H-NMR was used. The acid- base ratio was calculated from the integral area ratio of two signals, attributed to the methyl groups in the hydrocarbon tail and the carboxylic acid hydrogen of formic acid. The chemical composition of the bottom layer was found to have a molar ratio of 1.5:1 acid:amine, with the equilibrium composition being independent of the hydrocarbon chain length. The upper layer consists purely out of tertiary amine, showing that the formic acid adduct is not soluble in free amine, which was also confirmed using ATR-FTIR.

When the system has a biphasic character, the carboxylic acid hydrogen peak has a constant chemical shift within the obtained ¹H-NMR spectra. Above the critical molar fraction, a linear relation between the obtained chemical shift and the amount of unbound formic acid in the system is present. The linear relation was used to quantify the amount of protonated species in each phase, as well as the equilibrium constants. Surprisingly, the obtained equilibrium constants were found to be independent of the hydrocarbon chain length, showing all studied amines have an identical ability to reduce formic acid.

The direct hydrogenation was studied at temperatures between 35-90°C and 10-20 bar using a 1:1 volume mixture of H₂ and CO₂ in an autoclave together with the aforementioned amines as a reduction base, with no other solvents added. The selected catalyst is commercially available Au/ TiO₂. The liquid reaction mixture is monitored using in-line ¹H-NMR (Magritek, 43 MHz). The reliability of in-line ¹H-NMR was verified through comparison with a high resolution 400 MHz off-line ¹H-NMR.

The biphasic nature of the tertiary alkylamine – formic acid system was found to have significant implications for the reaction performance of the proposed system. The inability of the adduct to dissolve in the available non-binding amine results in the catalyst pores readily reaching equilibrium, with the system unable to refresh the reagents in the catalyst pores. Besides the products being confined in the catalyst pores, the significant polarity difference between the biphasic layers results in the catalyst solely being suspended in the bottom layer, resulting in low CO2 conversions and limited formic acid production. This finding was confirmed by examining the formic acid concentrations in the catalytic pores after reaction, with the total molar amount of formic acid in the pores being orders of magnitude larger than those in the reactor fluid.

Previous works have shown that the addition of polar solvent results in a significant performance increase [7]. Through the addition of solvents such as 1-decanol and N,N- dimethylformamide a monophasic system is obtained, independent of the formic acid molar ratio. The addition of a solvent does not lead to a decrease in formic acid binding capacity, as shown by a combination of ¹H- NMR and ATR-FTIR. Besides the addition of a solvent, the biphasic nature can also be inhibited through the usage of ethoxylated/propoxylated amines, as the phase separation behavior of the proposed system is mostly a result of the polarity difference between the employed amine and the produced acid

Initial results have shown that by removing the biphasic nature of the reaction mixture through the addition of a solvent, a significant improvement in formic acid production is obtained.

[1] Álvarez, A. et al. Challenges in the Greener Production of Formates/Formic Acid, Methanol, and DME by Heterogeneously Catalyzed CO2 Hydrogenation Processes. Chem. Rev. 117, 9804–9838 (2017).

[2] Hietala, J. et al. Formic Acid. in Ullmann’s Encyclopedia of Industrial Chemistry 1, 1–22 (Wiley-VCH Verlag GmbH & Co. KGaA, 2016).

[3] Putten, R. Van, Wissink, T., Swinkels, T. & Pidko, E. A. ScienceDirect Fuelling the hydrogen economy : Scale-up of an integrated formic acid-to-power system. Int. J. Hydrogen Energy 1–9 (2019). doi:10.1016/j.ijhydene.2019.01.15

[4] Bulushev, D. A. & Ross, J. R. H. Heterogeneous catalysts for hydrogenation of CO2 and bicarbonates to formic acid and formates. Catal. Rev. - Sci. Eng. 60, 566–593 (2018)

[5] Gunasekar, G. H., Park, K., Jung, K. D. & Yoon, S. Recent developments in the catalytic hydrogenation of CO2 to formic acid/formate using heterogeneous catalysts. Inorg. Chem. Front. 3, 882–895 (2016)

[6] Preti, D., Resta, C., Squarcialupi, S. & Fachinetti, G. Carbon dioxide hydrogenation to formic acid by using a heterogeneous gold catalyst. Angew. Chemie - Int. Ed. 50, 12551–12554 (2011)

[7] Filonenko, G. A., Vrijburg, W. L., Hensen, E. J. M. & Pidko, E. A. On the activity of supported Au catalysts in the liquid phase hydrogenation of CO2 to formates. J. Catal. 343, 97–105 (2016).

[8] Schaub, T. & Paciello, R. A. A process for the synthesis of formic acid by CO2 hydrogenation: Thermodynamic aspects and the role of CO. Angew. Chemie - Int. Ed. 50, 7278–7282 (2011).