Hydrodeoxygenation of stearic acid for the production of “green” diesel

1 Introduction

Nowadays biomass conversion is of great interest. One of the ways to process biomass is biofuel production. Biogas, bioethanol and biodiesel are widely used types of biofuel [1]. Among the various fuels-from-biomass investigated, fatty acid methyl ester (FAME) obtained by transesterification of triglycerides (TG) from natural oils and fats with methanol have received considerable attention. They exhibit high cetane number and are considered to burn cleanly. However, there is growing concern about the fungibility of these fuels with conventional petroleum-derived diesel due to their oxidative and thermal instability. Reduction of the oxygen content of the fuel would readily improve the stability of the fuel and therefore its utilization potential [2]. In addition, biodiesel production requires the use of raw material with certain quality. For example, waste oil and frying fat are not desirable due to the high concentration of free fatty acids.

It is worth noting that biodiesel of the best quality should contain compounds with a certain degree of unsaturation, which is characterized by the iodine number (optimum iodine number should not exceed 120 g I₂/100 g). Biodiesel produced via transesterification of TG contains rather high level of unsaturated hydrocarbon chains in comparison with petrol diesel.

The most promising way for biodiesel synthesis is biomass hydrofining. Hydrofining includes the following processes: hydrodeoxygenation (HDO), hydrocracking, hydrogenation. Biofuel produced via HDO is usually called the second-generation of biodiesel or “green” diesel [3].

“Green” diesel has numerous advantages such as high cetane number, good low-temperature properties, superior thermal stability, storage stability, and materials compatibility in comparison with both petrol diesel and biodiesel [4]. Besides, in contrast to FAME, where fuel properties depend on the raw source and composition, “Green” diesel is a product which does not depend on raw materials origin, and it mixes easily with common diesel fuel.

Comparison of fuel characteristics for diesel, biodiesel and Green diesel is presented in Table 1.

The presented data clearly show that the second-generation biodiesel has fuel characteristics similar to petroleum diesel. Thus for fuel manufacturers “Green” diesel is a preferable diesel component for the mixing as it has a boiling point interval comparable to that of typical diesel products, as well as a significant high cetane number and low density [1].

Catalytic deoxygenation (DO) of fatty acids, obtained by hydrolysis of TG, and their alkyl esters is the new way of biodiesel production in the form of diesel-like hydrocarbons. DO process consists in the removing of the oxygen of fatty acid carboxyl group and the production of saturated or unsaturated hydrocarbons [3].

The DO of vegetable-based feeds is typically related to pyrolysis (cracking), where the hydrocarbon chain is broken. The drawback of this approach is the loss of carbon and the decreasing energy content of the produced fuel.

There are several possible reaction paths for the production of straight-chain hydrocarbons. Fatty acids can be directly decarboxylated or decarbonylated. Direct decarboxylation removes the carboxyl group by releasing carbon dioxide and producing a paraffinic hydrocarbon, while direct decarbonylation produces an olefinic hydrocarbon via the removal of the carboxyl group by forming carbon monoxide and water, as illustrated by reactions 1 and 2.

Additionally, the fatty acid can be deoxygenated by adding hydrogen; in this case, the production of linear hydrocarbon can occur via direct hydrogenation or indirect decarbonylation (reactions 3 and 4, respectively).

More than 80 years ago, Bertram [6] succeeded in decarboxylating stearic acid (SA) to form heptadecane by a homogeneous catalytic reaction over selenium. A paraffin yield of just 50% was, however, obtained, and simultaneous dehydrogenation of the produced paraffin to olefin was observed. Much later, Foglia and Barr [7] demonstrated the conversion of fatty acids to alkenes by a homogeneous catalytic reaction with complexes of palladium and rhodium.

Typically, HDO is conducted under the pressure 4–6 bar at a temperature of 250–300°C. Most of the catalysts used in HDO are Pd, Pt, Ni, Ru based on organic (carbon) or inorganic (alumina or silica) supports. For example, decarboxylation of aliphatic and aromatic carboxylic acids has been carried out in the gas phase over Pd/SiO₂ and Ni/Al₂O₃. The experimental results showed that Pd/SiO₂ catalyst gave a much higher yield in decarboxylating heptanoic and octanoic acid (98% and 97%, respectively) than that achieved over the Ni/Al₂O₃ catalyst (26% and 64%, respectively). The production of straight-chain olefins from saturated fatty acids and fatty acid esters over a nickel based catalyst promoted with either tin, germanium, or lead was the subject of a patent [8].

However, the catalysts described above have one significant disadvantage: the decrease of catalytic activity caused by metal leaching. The use of hypercrosslinked polymers, e.g., hypercrosslinked polystyrene (HPS), can resolve this problem. As it can be seen from previous works [9–12], the catalysts based on HPS showed good activity and selectivity in the processes of hydrogenation and oxidation of organic compounds such as acetylenic alcohols [9, 13], ketones [10, 13], phenol [11] and monosaccharides [12]. Furthermore these catalysts have high stability and keep their catalytic activity at least for 10 cycles. This fact suggests high efficiency of HPS-based catalysts in the HDO process for the production of hydrocarbons with superior yield.

2 Materials and methods

The SA (98.0%, ChimMedService, Russia) was used as a model substrate for catalytic HDO. n-Dodecane (99.9%, Sigma Aldrich, USA) was used as the solvent. Initial concentration of SA in n-dodecane was 0.1 mol/l. The process was carried out in a stainless steel batch reactor (PARR Instrument, USA) with total volume of liquid fraction of 30 ml under the pressure 0.2–1.8 MPa at the temperature of 230–260°C. Pd-containing catalysts with metal loading varying from 1% (wt.) up to 5% (wt.) based on polymeric matrix of HPS (Purolite Ltd., UK) were synthesized by the wet impregnation method as described elsewhere [13] and were tested under varying temperature and pressure. To prevent mass transfer resistances, a granules diameter equal to 70 μm was chosen. The metal-catalyst concentration used was 3.1×10^-4 mol/l.

The suspension of the catalyst in solvent (15 ml) was inputted into the reactor through the side fitting. Before the reaction the catalyst was saturated with hydrogen at a preselected temperature and pressure for 30 min, and then the substrate in solvent (15 ml) was injected.

Liquid samples were analyzed by the method of gaseous chromatography mass-spectrometry (GC-MS), using a chromatograph GC-2010 and a mass spectrometer GCMS-QP2010S (SHIMADZU, Japan). The following analysis conditions were used: run time: 81 min; the initial column temperature of 50°C was withstood for 1 min, then the temperature was smoothly increased up to 210°C with a heating rate 2°C/min; injector temperature: 280°C; automatic split; pressure of He 53.6 kPa; common stream of He 81.5 ml/min; linear gas velocity 36.3 sm³/s; chromatographic column type HP-1MS: L=30 m; d=0.25 mm; film thickness 0.25 μm; ion source temperature: 260°C; interface temperature: 280°C; scanning mode 10 up to 800 m/z; scanning rate: 1666; electron-impact ionization.

Synthesized HPS-based catalysts were characterized using low-temperature nitrogen physisorption, transmission electron microscopy (TEM), X-ray photoelectron spectroscopy (XPS), thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC).

3 Results and discussion

Table 2 presents the data for the catalytic testing after 220 min of the experiment. As it can be seen the increase of Pd loading from 1% up to 5% results in the slight increase of SA conversion for Pd/HPS catalysts at the same reaction time. However the turnover frequency (TOF) was decreased with the increase of metal loading. This can be explained by the metal distribution in the pores of HPS obtained by TEM (see Table 3). The smaller Pd particles in the case of 5%-Pd/HPS catalyst are distributed in the micropores of the polymeric matrix, whereas the Pd particles in 1%-Pd/HPS catalyst are located in the macro- and mesopores. Since the SA molecule is rather big it has better accessibility to the large metal particles.

The selectivity regarding n-heptadecane (target product) was slightly decreased with the increase of metal loading. It is worth noting that the only byproduct conducive to the decrease of selectivity was the n-pentadecane, which indicates that cracking also takes place.

In Table 3 the results of the physicochemical characterization of the HPS-based catalysts are shown. It was revealed that the increase of Pd loading causes the decrease of the specific surface area of the micropores, which likely corresponds to the observed decrease of catalytic activity. The results of TGA and DSC show that the polymeric matrix of HPS is stable at a temperature above 500°C.

However, selectivity with respect to n-heptadecane at rather high conversion was the key factor in the choice of most prospective catalytic system. Thus 1%-Pd/HPS was chosen for further investigation of the influence of the reaction temperature and pressure.

The dependence of the conversion of SA on the reaction time varying both temperature and hydrogen pressure is shown in Figure 1.

Figure 1

The conversion of SA in time varying temperature (A) and hydrogen pressure (B).

(A) Hydrogen pressure: 0.6 MPa; solvent: n-dodecane; stirring rate: 700 rpm; catalyst granules size: 70 μm; initial concentration of SA C_0: 0.1 mol/l; catalytic system: 1%-Pd/HPS; catalyst concentration C_Me: 3.1·10^-4 mol/l. (B) Temperature: 255°C; solvent: n-dodecane; stirring rate: 700 rpm; catalyst granules size: 70 μm; initial concentration of SA C_0:0.1 mol/l; catalytic system: 1%-Pd/HPS; catalyst concentration C_Me: 3.1×10^-4 mol/l.

As it can be seen from Figure 1A the increase of the temperature from 230 to 255°C leads to the significant increase of conversion, whereas at a temperature of 260°C aslight increase of conversion is observed. Besides, a negligible decrease of selectivity regarding the target product (∼0.2%) was observed at temperatures over 255°C. The apparent activation energy calculated in the temperature range 230–255°C at the linear curve part was equal to 85±5 kJ/mol.

Figure 1B shows that the increase of partial pressure of hydrogen from 0.2 to 1.4 MPa leads to the increase of the reaction rate. At a pressure of 1.4 MPa the reaction rate reaches the maximum and the reaction rate becomes constant. However at a pressure above 0.6 MPa the selectivity regarding n-heptadecane is decreased, as the side reactions rate, i.e., cracking, is increased.

As a result of the investigation, the optimal reaction conditions were found (temperature 255°C; partial H₂ pressure 0.6 MPa; initial concentration of SA C₀ of 0.1 mol/l; catalytic system1%-Pd/CΠC; catalyst concentration C_Me 3.1×10^-4 mol/l), which allow achieving the selectivity of the process up to 98.8% at 100% of SA conversion. Based on the data on the temperature effect, the apparent activation energies for the best catalytic systems were calculated to be 85±5 kJ/mol for 1%-Pd/HPS.

The catalyst chosen was also tested in multiple repeated uses. It was found that activity and selectivity of 1%-Pd/HPS sample remains the same after the fourth reuse (see Table 4).

The data of TGA (Figure 2) shows that the catalyst keeps its structure after multiple uses.

The additional front of the thermogramm for the catalyst after multiple uses (Figure 2B) corresponds to the evaporation of adsorbed n-dodecane at a temperature of 216.1°C. The endothermic effects on the DSC curves (1–4) are fully congruent for both catalysts.

Figure 2

TGA curves of 1%-Pd/HPS catalyst before (A) and after (B) usage.

Besides the analysis of reaction media at each reuse by the atomic absorption spectrometry method showing the deficiency of Pd in liquid samples, it indicates that the metal leaching does not take place.

The gas phase analysis showed the formation of significant amount of CO, whereas CO₂ was not observed. This fact suggests the possible mechanism of SA HDO process – decarbonylation of fatty acid with the following hydrogenation of obtained olefins.

4 Conclusions

In the current work the catalytic HDO of SA, which is a potential feedstock for the production of the second-generation of biodiesel, was investigated using Pd/HPS catalysts. It was revealed that the main product of the reaction was n-heptadecane. The selectivity of the process (regarding n-heptadecane) reached up to 98.8% at 100% of substrate conversion.

As the result of the investigation possible transformation paths were proposed. The HDO process was found to proceed via a decarbonylation mechanism. Besides, cracking of hydrocarbon products was found to occur.

Synthesized HPS-based catalysts were observed to be stable and selective, they provide constant activity and selectivity for up to four uses.