Introduction
One of the available solutions to address the world’s increasing energy consumption is the production of fatty acid methyl esters (FAMEs), also known as biodiesel, which are an alternative fuel similar to conventional diesel. They are usually produced from various triglycerides – components of vegetable or plant oils, animal fats and tallows [1]. Keeping in mind the critical aspect of food security (i.e., minimizing the use of edible feedstock types for fuel production) and the need for sustainability in the long run, recent developments in the use of algae as the feedstock have given the concepts of biofuel production renewed attention [2]. After several decades since the first reports on the usage of microalgae for biodiesel production [3,4], it has been determined that microalgal biofuels are promising candidates for the partial replacement of fossil fuels.
The production of microalgal biodiesel requires an efficient catalyst for initiation of the transesterification process that converts triglycerides into FAMEs. Other methods that do not require a catalyst, such as pyrolysis and utilization of supercritical fluid technology, are considered to be highly energy-intensive, inhibiting their practical implementation on the industrial scale [5].
The preparation and application of different types of catalysts for homogeneous, heterogeneous, and even enzymatic transesterification processes have been extensively investigated [6]. However, the most commonly used commercial process for biodiesel production is the homogenous transesterification of triglycerides with methanol in the presence of sodium hydroxide (NaOH) or potassium hydroxide (KOH), which is still in use on an industrial scale. Along with the obvious advantages of being highly active and relatively inexpensive (depending however on the number of washing steps and possible need for neutralization), the downsides of alkali hydroxides are long known – difficulties in separation of the K+/Na+ traces from the product, higher catalyst consumption compared to the solid ones and low reusability [7]. On the other hand, the reported solid catalysts can potentially resolve issues associated with catalysts used in the homogeneous process, that is, they can be prepared with a desired particle size for separation needs and can be reused. The main drawbacks are their lower activity compared to the catalysts used in homogeneous solutions (often associated with mass transfer issues [7]) and the leaching of the active phase (if supported catalysts are used [8]).
Thus, the search continues for a new generation of catalysts that ideally combine the advantages of both types of catalysts used in the homo- and heterogeneously catalyzed reactions [9-12]. In the context of biodiesel production via the transesterification of triglycerides with methanol, the challenge is to prepare a catalyst possessing the following characteristics: i) large particle size for convenient separation, ii) accessibility of the active sites for reactants, iii) minimized diffusion limitations for reaching effectiveness factors close to unity, iv) leaching resistance, v) activity at moderate temperatures (e.g., at the boiling point of methanol) being ideally comparable to KOH/NaOH catalysts and vi) stability for multiple reuse. Our approach to address this challenge is based on the preparation of bulk catalysts that are reported to be active in the transesterification reaction together with further (nano)structural modification aimed at enhancing accessibility to the active sites and improving the mass transfer characteristics for efficient reactant supply and product removal [13].
Amongst the prospective solid catalysts designed for transesterification reactions, such as calcium [14] and other metal oxides [15], metal-organic frameworks (MOFs) [10], silica-supported catalysts [16], biochar [17] and other biomass-derived catalysts [18], zeolites and molecular sieves [19,20] offer a combination of the possibility for the pore network modification (e.g., as a result of a post-synthetic treatment) and high stability (e.g., compared to MOFs) with active sites being part of a framework.
The microporous, titanosilicate ETS-10 catalyst was found to be one of the most active catalysts amongst the crystalline microporous molecular sieves (such as, e.g., zeolites) reported for the transesterification of triglycerides with methanol [20]. Its crystal structure is built up from orthogonal TiO6 octahedra and SiO4 tetrahedra sharing oxygen atoms and forming a three-dimensional interconnected pore system [21] consisting of channels with cross-sectional dimensions of 0.8 × 0.5 nm [22,23] (Figure 1A). Each Ti atom in a six-coordinated state bears two negative charges, which can be balanced by Na+ or K+ cations according to the following stoichiometry (Na,K)2TiSi5O13 [22]. The high activity of ETS-10 in the transesterification reactions compared to zeolites and other molecular sieves originates presumably from its pronounced Brønsted basicity of the shared oxygen atoms [20]. For example, it has been reported that the parent ETS-10 catalyst is approximately four times more basic than the NaX zeolite based on acetone selectivity in the conversion of isopropanol to acetone and propene [24]. However, due to its microporous nature, to the best of our knowledge, in all reported studies on the transesterification of triglycerides (single-component or in mixtures/oils) for biodiesel production, the reaction took place solely on the outer crystal surface. The smallest triglyceride, triacetin, used in these studies, has a reported critical diameter of ≈1 nm [25], which appears to be too large to diffuse into the micropores.
Several strategies have been attempted for a post-synthetic modification of the ETS-10: an ion-exchange method [27], acidic treatment [28,29], and treatment with hydrogen peroxide (H2O2) [26,30,31]. The latter approach resulted in a notable pore network modification by introducing larger micropores (≈1-2 nm) and mesopores (≈5-30 nm) in a controllable way, for example, by using different concentrations of the H2O2 solution and treatment times (Figure 1B). Herewith, the crystallinity and mechanical stability of the materials were well-preserved despite the experienced desilication and detitanation. This approach was followed in the present study to prepare hierarchically structured ETS-10 crystals (i.e., containing interconnected micro- and mesopores). In general, the micropores of the hierarchical catalysts mostly contribute to the catalytic process, while the function of the larger mesopores is the promotion of the reactant supply and product removal. However, for bulky triglycerides, the micropores are not expected to be accessible. In such a case, the mesopores will fulfil the task of providing access to the catalytically active sites located within them.
In the present contribution, we report the successful preparation of large crystallite (≈30 μm) ETS-10 titanosilicates with improved active site accessibility (achieved by post-synthetic treatment) for efficient transesterification of triglycerides into biodiesel. A triolein has been selected as the triglyceride due to its prevalence in, for example, microalgal oils typically used in production of 3rd and 4th generation biodiesel. Additionally, CaO catalysts, known to be highly active in the transesterification process, were prepared as references for comparison of the catalytic activity under identical reaction conditions. Prior to catalytic tests, the prepared CaO- and titanosilicate catalysts were characterized to obtain quantitative information on properties such as crystal structure by X-ray diffraction (XRD), crystal size by laser diffraction, crystal morphology by scanning electron microscopy (SEM) and transmission electron microscopy (TEM), pore width by N2 sorption and Hg intrusion, acid and basic site density by NH3 and CO2 temperature-programmed desorption (TPD), presence of hydroxyl groups by 29Si magic angle spinning nuclear magnetic resonance (MAS NMR), pore interconnectivity by hyperpolarized (HP) 129Xe NMR, pore accessibility for triolein by 1H pulsed field gradient (PFG) NMR, state of the Ti atoms before and after the treatment of titanosilicates by electron paramagnetic resonance (EPR) and thermal stability of the crystals by differential thermal analysis (DTA).
