1. Introduction
The use of biomass as renewable energy source can reduce the dependence on fossil fuels as well as reduce the impacts of global warming. Microalgae to biofuel represents a sustainable pathway due to the microalgae capacity to grow in marginal lands, using wastewater and CO
2 as source of energy and nutrients [
1,
2]. Among microalgae,
Isochrysis sp. has shown to be a promising contender as feedstock for a biorefinery setting, due to the possibility to convert it in several bio-products. Microalgae with very high lipid contents are suitable for producing biodiesel through transesterification processes, but
Isochrysis microalgae, which is rich in carbohydrates and proteins, is more suitable for thermochemical conversion processes, such as pyrolysis [
3].
Only few studies are available on
Isochrysis sp. catalytic pyrolysis literature review. Wang et al. (2015) investigated the pyrolysis of defatted and not-treated
Isochrysis sp. [
4]. The defatted pyrolysis at 475 °C produced lower bio-oil yield (36.9 wt %) compared to the whole microalgae (41.3 wt %) and phenols (from 19.99% to 31.18%) enriched bio-oil [
4]. Catalytic pyrolysis of
Isochrysis sp. using seven ceria-based catalyst was investigated by Aysu et al. [
5], who obtained a significant increase in the bio-oil yield in the presence of Ni-Ce/Al
2O
3 and Ni-Ce/ZrO
2 (26 wt %) compared to the non-catalytic pyrolysis (15 wt %). In addition, the presence of catalyst increased the energy content and decreased oxygen and nitrogen content of the bio-oils. Moreover, the catalytic pyrolysis of
Isocrysis microalgae in the presence of Li-LSX-zeolite under different operating conditions was studied [
6]. This work showed that Li-LSX-zeolite promoted aromatisation, deoxygenation and denitrogenation of the bio-oil. Compared to the commonly used ZSM-5, Li-LSX zeolite gives rise to a higher bio-oil denitrification, principally as NH
3, but also HCN in the gas phase [
7,
8,
9]. However, the high-level of macro-minerals (Na, K, Ca) in
Isochrysis sp. ash affect the mechanism of pyrolysis and decreases the pyrolysis oil yield [
10,
11]. Ash content also affects the pyrolysis process design and operations (causing fouling, slagging and corrosion in the reactors), as well as the product purification process. As a result, the removal of inorganics from microalgae can benefit their intrinsic quality.
So far, only a limited number of works are available, in which the effect of chemical pre-treatment has been evaluated on the catalytic pyrolysis of microalgae. Bae et al. investigated the effect of treatment on bio-oil production by pyrolysis of macroalgae
Undaria pinnatifida, which has high ash content (38 wt % on dry basis) [
12]. Treatment by acid washing (2 M HCl, mix on hot stirrer at 60 °C for 6 h) was able to remove most of ash content to 0.76 wt %. As a result, the bio-oil yield increased after acid treatment from 40 to 46 wt % at 500 °C. Ross et al. studied the pyrolysis behaviour of 2 M acid (HCl) treated seaweeds (6 h at 60 °C) [
13]. Pre-treatment in acid removed over 90% of the Mg, K, Na and Ca and resulted in furfural reach bio-oil [
13]. Choi et al. showed that acid sulphuric treatment of brown microalgae (
Saccharina Japonica) was able to remove active inorganic minerals by reducing the ash content from 18.3 to 3.3 wt % [
14].
Catalyst deactivation is also a big concern in industrial catalytic processes. Oxygen-containing chemical species such as aromatic and nitrogenated compounds in the pyrolysis oil tend to form coke formation during the upgrading process [
15]. Fouling or coking is the main reason for zeolite deactivation in catalytic cracking [
16]. Catalyst deactivation on zeolite occurs due to coke formation and strong adsorption of oxygenates compounds on the surface of catalyst support [
17]. In order to improve catalyst lifetime and reduce operation cost on the catalyst, the regeneration or recycling of catalyst becomes essential. Zeolite catalyst can be recovered by oxidation regeneration at high temperature through coke combustion.
A lot of attention has been paid to the kinetic study of coke formation and catalyst regeneration in various processes [
18,
19]. Zhang et al. carried out a study on the fresh, spent and regenerated ZSM-5 catalyst during biomass catalytic pyrolysis [
17]. The study was conducted on the pyrolysis of corn stover using Py-GC/MS at 500 °C. The catalysts in this study were indicated as FZ (fresh catalyst), SZ (spent catalyst) and RZ (regenerated catalyst). From the catalyst characterisation, FZ had the highest value of total acid sites and BET surface area compared to other catalysts. The results show that the catalyst produced vapour yield in the following order: (FZ > RZ > SZ). Besides, the highest coke yield was obtained by FZ followed by RZ and SZ.
Despite numerous studies investigated the cyclic stability of ZSM-5 catalyst for biomass pyrolysis indicating loss of catalytic activity (denoted by a decrease in aromatics and Poly Aromatic Hydrocarbons (PAH) formation) [
17,
18,
19], to our knowledge, the cyclic stability and regeneration of Li-LSX zeolite and its behaviour in the presence of pre-treated microalgae has not been studied yet. Therefore, this work investigates the activity of Li-LSX-zeolite catalyst over three pyrolysis/regeneration cycles in the presence of non-treated and 1% H
2SO
4 acid treated
Isochrysis sp. microalgae. This work contributes to the understanding of the deactivation process over Li-LSX zeolite and in defining strategies to reduce it.
2. Materials and Methods
2.1. Materials
Isochrysis 1800 microalgae were purchased from Varicon Aqua Solutions Ltd (Hallow, Worcester, UK). The received microalgae were dried at 60 °C in an oven for 1 week to remove about 90 wt % of the moisture and then milled for 1 min using a Fristch Pulverisette 2 to a particle size less than 177 µm. Li-LSX-zeolites was acquired (in pellets form) from Shanghai Hengye Chemical Industry Co. and then grounded using a pestle and mortar. Sulphuric acid (96% extra pure) was purchased from Fisher Scientific UK Ltd (Loughborough, UK). The pre-treatment of microalgae was performed by adding the dried microalgae (3 g) to 30 mL of 1% H2SO4 solution and stirring for 30 min (350 rpm) at 25 °C. After the treatment, the mixture was rinsed with deionized water to achieve a pH of 7 and centrifuged for 3 h to separate out the leached microalgae. Since the remaining wastewater after the separation of the leached microalgae still contained some algae in suspension, a micro-filtration stage (22 µm) was carried out. The residual solid was then oven dried at 60 °C to obtain constant weight.
2.2. Characterisation Techniques
XRF analyses were carried out to quantify the elemental composition of raw and acid treated microalgae using a Philips PW1480 XRF spectrometer and SemiQ semi-quantitative analysis software. Approximately, 5 mg of sample was placed between two layers of mylar film, mounted into a two-part holder system that is normally used for liquid samples. The X-ray scans identified and quantified the elements phosphorus, sulphur, chlorine, potassium and calcium in all the samples.
The elemental analysis, EA, (C, H, N, S) of the biomass samples and the solid/liquid products from pyrolysis reaction was determined using an Exeter CE-440 Elemental analyser. The oxygen (O) content was determined by difference (O = 100 − C + H + N + S).
The higher heating values (HHV) of the feedstocks and liquid/solid products were calculated based on the Equation (1), which is a correlation reported to be valid for solid and liquid fuels [
20]:
GC-MS analysis was performed by a Shimadzu GCMS QP2010 SE equipped with a Restek RXI-5HT column [
6]. The column (length: 30 m, inner diameter: 0.250; film: 0.25 µm) had temperature limits between 40 and 300 °C. The oven was programmed to hold at 40 °C for 10 min, ramp at 5 °C/min to 200 °C and hold for 10 min, ramp at 10 °C/min to 250 °C and hold for 10 min and ramp at 10 °C/min to 295 °C and hold for 10 min. Helium was used as the carrier gas with a constant flow rate of 1.7 mL/min and injector split ratio at 1:20 ratio. The end of the column was directly introduced into the ion source detector of VG Trio 1000 series. Typical mass spectrometer operating conditions were as follows: transfer line 270 °C, ion source 250 °C and electron energy of 70 eV. The chromatographic peaks were identified according to the NIST library to identify bio-oil components.
Proton NMR (1H NMR) was selected to give an overall picture of the bio-oil composition in terms of the proton distribution in the different chemical functionalities using a Bruker Avance III operating at 400 MHz. The instrument was equipped with 60 samples position autosampler, with a 5 mm dual 1H/12C pyro probe. For samples preparation, bio-oils were diluted in 99.9% of dichloromethane (CDCl3) (Merck, Germany) with ratio 1:1 by volume and poured into 5 mm NMR tubes. All the acquired NMR spectra were processed through Topspin version 2.1 software.
Gas analyses were carried out using a Cirrus MKS Mass Spectrometer controlled by Process Eye view software. Before starting the analysis, the capillary heater and system heater were switched on at least 1 day in advance to achieve stable conditions and remove any potential moisture from the capillary.
Total surface area (BET), external surface area, micropore volume and micropore area were all calculated using the software supplied with the Micrometrics Gemini VII 2390 V3.03 surface area/porosity analyser. Firstly, the catalyst was degassed for 12 h at 200 °C under N2 gas using a Micromeritics Flowprep 060. About 0.2 to 0.4 g of materials were weighed before and after degassing. Then, the catalysts underwent analysis using nitrogen as an adsorption gas. Sample evacuation was conducted at a rate of 760.9 mmHg/min and equilibrated for 5 min. The BET surface area was analysed on the adsorption isotherm using ten data points within the P/P0 range of 0.05 to 0.3.
XRD analyses were carried out using a Bruker D8 Advance powder diffractometer, operating with Ge-monochromated copper Kα1 radiation with a wavelength of 0.15406 nm and a LynxEye linear detector in reflectance mode. Prior to the analysis, the catalyst sample were ground using pestle and mortar and oven-dried at 110 °C overnight. Data were collected over the angular range 5° to 85° in two-theta under atmospheric pressure.
SEM/EDS analyses were carried out using Carl Zeiss Sigma HD VP Field Emission SEM and Oxford Aztec ED X-ray analysis and electron backscatter Diffraction (EBSD) system. The patterns were imaged and analysed using an Oxford instrument software to perform the compositional analysis on the catalyst.
2.3. Pyrolysis Apparatus and Procedure
A down-stream vertical configuration pyrolysis setup having a reactor-tube (1.27 cm inner diameter and 15 cm length) inserted in a high temperature tube furnace (GVA/GVC from Carbolite; max. heating rate: 100 °C/min, max. temperature: 1000 °C) was used. The N2 flow rate was set at 345 mL/min (8 sec gas residence time) and temperature to 500 °C. The temperature inside the furnace was measured by a K-type thermocouple. The condensation system was made of three 125 mL Dreschel bottles connected with high temperature resistant Viton tubing and placed in a salt-ice bath.
The sample inside the reactor was held by a sample holder (stainless steel tube), a SS316 wire mesh (with 0.45 mm wire diameter) and quartz wool. The reactor was set-up for ex-situ pyrolysis experiments, where the metal mesh and quartz wool were alternated between samples and catalyst to avoid mixing of the two materials and allowing only the released volatiles pushed by the nitrogen stream to flow across the catalyst bed. A catalyst to microalgae weight ratio of 1:1 g/g was used in the experiments. A schematic diagram of the vertical pyrolysis set-up used in this work is presented in
Figure 1.
Before each experiment, the reactor was purged with nitrogen flow for 10 min in order to remove the remaining air impurities in the reactor. The reaction was run for 20 min to ensure maximum decomposition of all microalgae during pyrolysis.
The liquid product (bio-oil) was recovered from the Drechsel bottles by washing with 50 mL acetone. Then, the solvent was evaporated at room temperature for 20 h. The non-condensable gaseous were sampled in a 1 L gasbag and then analysed by mass spectrometry analysis. The bio-char left behind in the reactor was taken out, weighed and stored for further analysis.
The gas yield (wt %) was calculated by the difference from overall mass balance (Gas = 100 − (Bio-oil + Bio-char).
Pyrolysis experiments and products analyses (proximate and EA) were carried out by triplicates to measure the experimental error, which was assessed to be lower than 5%.
2.4. Catalysts Regeneration Procedure
The catalysts were regenerated to evaluate the activity and deactivation of the catalyst after a number of cycles. After the pyrolysis tests, the spent catalyst was recovered and a small fraction was submitted to SEM/EDS and XRD analyses; meanwhile the rest of the catalyst was calcined to remove the coke from the catalyst surface. The catalyst was heated up in the muffle furnace (Carbolite) at 500 °C for 1 h in the presence of air. Then, the catalyst was kept in the desiccator for the 2nd cycle of pyrolysis. The same method was applied to the 3rd cycle regeneration. Moreover, a set of calcinations at 700 and 950 °C were carried out to evaluate the maximum temperature for calcining the Li-LSX zeolite and their effect in removing coke. After calcination, the catalysts were sieved to remove the ash from coke combustion and characterised by SEM, XRD and EA.
4. Conclusions
The activity of Li-LSX-zeolite catalyst on the pyrolysis of non-treated and acid pre-treated Isochrysis sp. microalgae after three consecutive pyrolysis/regeneration cycles was investigated. Overall, a very different behaviour was noticed in the pyrolysis process when non- or pre-treated microalgae where used. For the pyrolysis of non-treated microalgae, the bio-oil yield slightly decreased after three cycles, while the bio-oil yield for the pre-treated microalgae increased at the expense of gas, due to the removal of alkali metals in the pre-treatment. The products’ distribution, 1H-NMR and the EA analyses showed that the catalyst maintained its catalytic activity for cracking and deoxygenation over three cycles in presence of non-treated microalgae, while strong deactivation occurred when pre-treated microalgae where processed due to fouling (70% surface lost), with trace amount of P, S, Na deposited on the regenerated catalyst surface. In summary, Li-LSX zeolite was effective in maintaining deoxygenation activity over three cycles in the pyrolysis of non-treated Isochrysis microalgae, while the algae pre-treatment with sulphuric acid was detrimental on the catalyst activity.