A novel kinetic model applied to heterogeneous fatty acid deoxygenation

Highlights

• A novel kinetic model for the deoxygenation of fatty acids was developed.

• AIC and BIC showed that the novel model was better than previous models.

• A novel, improved aerogel catalyst was tested and extensively characterized.

• Reaction rate and selectivity were mapped through a range of temperatures.

Abstract

An improved kinetic model for the deoxygenation of fatty acids was developed and compared with previously developed models. The models were compared using data collected from the deoxygenation of oleic acid using a novel, extensively characterized aerogel-based catalyst and previously published data of the deoxygenation of stearic acid, resulting in an extensive analysis ranging from 275 °C to 400 °C. The novel model was able to predict the behavior of reactants, intermediates and products for all cases analyzed in this paper. Akaike and Bayesian information criterions were applied, concluding that the novel kinetic model was indeed the best model to predict the behavior of both our original data and a previously published dataset. Relevant data regarding reaction rates and selectivity for the novel catalyst was given, showing a clear practical use of the results presented in this paper for the design and optimization of deoxygenation reactors.

Introduction

To overcome the challenge of the ever-increasing need for fuels and hydrocarbon-based feedstock, the deoxygenation of biomass was proposed. Fatty acid radicals are very similar to the hydrocarbons found in oil, distinctively those found in diesel (Wang et al., 2020). Due to the high energy content of vegetable oils (Gosselink et al., 2013), this type of biomass is especially suitable for the production of transportation fuels (Baharudin et al., 2019, Horáček et al., 2020, Pattanaik and Misra, 2017). As such, vegetable oils are one of the most promising types of biomass.

Another advantage to the use of vegetable oils as feedstock for the production of fuels is the decentralization of energy production brought by the fact that vegetable oils may be produced from several feedstock materials, such as sunflower, safflower, soybean, cottonseed, rapeseed, canola, corn and peanut (Huber and Corma, 2007).

So as not to confront the food production chain, non-edible crops may be used (Gosselink et al., 2013). Crops such as microalgae oil (Baharudin et al., 2019, Peng et al., 2012), macauba (Silva et al., 2016); jatropha (Gosselink et al., 2013) as well as waste cooking oil and fats (Gosselink et al., 2013, Baharudin et al., 2019, Huber and Corma, 2007) can serve for this purpose. Notwithstanding specific designs, vegetable oils are not compatible with common engines due to the presence of oxygenates, high acidity and viscosity, as well as low oxidative stability (Gosselink et al., 2013).

The transesterification of triglycerides into methyl esters solves some of those problems, as does the deoxygenation of vegetable oils (Gosselink et al., 2013, Baharudin et al., 2019, Pattanaik and Misra, 2017, Huber and Corma, 2007, Stepacheva et al.,2016). However, the transesterification produced fuel (biodiesel) has high oxygen content, low oxidative stability, high viscosity and has characteristics highly dependent on the feedstock (Baharudin et al., 2019, Pattanaik and Misra, 2017, Peng et al., 2012). On the other hand, the deoxygenation of vegetable oils is a promising alternative that is able to produce a hydrocarbon based fuel that is fully compatible with modern diesel engines (Baharudin et al., 2019, Pattanaik and Misra, 2017), exhibiting a high cetane number (Stepacheva et al.,2016), as well as high thermal and oxidation stabilities (Baharudin et al., 2019). Fuels produced by deoxygenation of vegetable oils have also been proved in 50:50 blends for aviation fuels, further increasing the interest over this method for the production of fuels.

The deoxygenation of fatty acids into hydrocarbons is carried out mainly through three reaction pathways: hydrodeoxygenations, decarbonylations and decarboxylations, which remove oxygen as H₂O, CO and CO₂, respectively (Jeništová et al., 2017, Imai et al., 2020, Pimenta et al., 2020). Parallel reactions include the cracking of heavier hydrocarbons, isomerizations, Diels-Alder reactions and scissions of fatty acids, which may alter the quality of the produced fuel (Baharudin et al., 2019, Peng et al., 2012, Adebanjo et al., 2005, Jeon et al., 2019).

The deoxygenation of fatty acids have been carried at a large range of temperatures and pressures, with previous reports ranging from 250 °C to 450 °C and hydrogen pressure from 1 bar to 300 bar (Baharudin et al., 2019, Peng et al., 2012, Scaldaferri and Pasa, 2019, Wang et al., 2013, Zandonai et al., 2016).

The knowledge about products, kinetics and yield of a reaction is strictly necessary for the development of mathematical models (Yang et al., 2019), which in turn will be used to project and optimize industrial processes. Only recently researchers have been able to confirm the reaction pathways involved in the deoxygenation of fatty acids.

Few studies aimed to develop a kinetic model for the deoxygenation of free fatty acids (Jeništová et al., 2017, Arora et al., 2019, Hachemi and Murzin, 2018). Those were mostly performed on idealistic conditions, such as using diluted fatty acids, while others explored effects of reaction parameters without exploring the reaction kinetics. Previous kinetic studies used mostly traditional catalysts (Jeništová et al., 2017, Imai et al., 2020, Arora et al., 2019) and only one recent study was able to provide information on reaction intermediates (Arora et al., 2019).

The few studies which presented kinetic models were developed using Langmuir-Hinshelwood kinetics (Jeništová et al., 2017, Arora et al., 2019, Hachemi and Murzin, 2018, Peroni et al., 2017). The Langmuir-Hinshelwood kinetic model is based on the assumption that the reactants are first adsorbed on the active sites of the catalyst. The bonding of the adsorbed molecules and active sites weakens the chemical bonds of the reactants, thus facilitating the reaction of molecules adsorbed on neighboring sites (Prins, 2018).

However, none of the previously mentioned studies were able to provide a model that qualitatively fits the behavior seen in the deoxygenation of fatty acids. In accordance to this need, we explore the use of a novel aerogel-based catalyst in practical conditions, using undiluted fatty acids and high temperatures. Our main objective was to achieve results with direct industrial application, as well as to develop a better model that explains the behavior of deoxygenation reactions. We explored the use of this novel kinetic model to explain the experimental data collected for this study and exhibited in previous literature, thus validating the model for the simulation, design and optimization of industrial reactors.

This is the very first study in which a kinetic model was able to qualitatively and quantitatively predict the concentration of all species seen in the deoxygenation of fatty acids. The model was validated for a wide range of concentrations, temperatures and conversions. It was also able to explain the zero-order behavior presented by our data. The results presented in this paper have a clear practical use regarding the design and optimization of deoxygenation reactors.