1 Introduction

Bio-oil from renewable biomass sources is the most promising replacement for the exhaustibility fossil fuels. Algae biomass, which own the characteristics of fast growth rate, low requirements on growth environment, inhibition of water pollution and greenhouse effect, are considered as sustainable development and environmental-friendly feedstock for bio-oil production (Biswas et al., 2017; Wang et al., 2017; Sharma et al., 2021). The mainly approaches to produce bio-oil from algae are transesterification conversion, biochemical conversion and thermochemical method (Ambaye et al., 2021). Compared with above methods, hydrothermal liquefaction (HTL) of algae belonging to the thermochemical method features priority for the following aspects: i) it is an economical process especially for the wet algae feedstock since it avoids the dewatering and drying steps (Saber et al., 2016; Sheng et al., 2018; Marzbali et al., 2021); ii) the components of algae can be fully utilized; iii) the oil products generally possess higher energy content and better stability properties than pyrolysis oil (López Barreiro et al., 2013; Hu et al., 2017; Xu et al., 2020); iv) the decomposition of biochemical compositions into aqueous-phase, heavy oil and light oil products requires quite low activation energy (Kim et al., 2013; Vo et al., 2017a, 2017b; SundarRajan et al., 2021).

In order to increase the yield and quality of bio-oil, the solvent water in HTL is heated and pressurized to a subcritical state, which has excellent solubility to biological macromolecules such as proteins and polysaccharides, and can promote the progress of acid-catalyzed reaction (Zhu et al., 2017; Xu et al., 2020). It has been found that the ionic product constant (K_w) of water increases from 10⁻¹⁴ to 10⁻¹² when the temperature increases from 25 to 350℃ (Toor et al., 2011). The resulting high levels of H⁺ and OH⁻ can promote the acid or base-catalyzed of HTL process (Cheng et al., 2017; Galadima and Muraza, 2018). In addition, the introducing of catalysts is an effective strategy to increase bio-oil yield and improve the quality. The application of various homogeneous and heterogeneous catalysts in algal HTL has attracted extensive attention (Gai et al., 2015; Saber et al., 2016). Yang et al. (2017) introduced sulfuric acid and acetic acid as the homogeneous catalysts in HTL and the bio-oil yield has been obviously increased. However, heterogeneous catalyst has received extensive attention for the advantages with environment-friendly, convenient recovery and cyclic utilization, etc. (Carrero et al., 2015; Navalon et al., 2016). Solid acid catalyst is a typical heterogeneous catalyst, which is widely used in alkylation, esterification, isomerization, HTL, etc. In order to achieve prominent catalytic effect, it should focus on the acidity and stability of the solid acid catalyst. Relevant research efforts have confirmed that increasing the acidic of catalyst can improve the fluidity and reduce the boiling point of bio-oil (Yang et al., 2017). Accordingly, solid superacid which provide strong acidity can be regarded as a recommendable kind of heterogeneous catalyst. SO₄²⁻/ZrO₂ with super acidity have been seen as a representative solid superacid for the advantages of easy preparation, high hydrothermal stability and strong activity. Therefore, the use of solid superacid catalyst SO₄²⁻/ZrO₂ in HTL of algae exhibits broad application prospects.

The biochemical composition of algae is also an important factor to influence the product yield and quality. Biller et al. (2011) preliminarily reported the related effects using soya oil and microalgae. However, it is still worth further exploring other model compounds such as polysaccharides and proteins. In addition, few studies have an insight into the effects of catalysts on the distribution and properties of HTL product for algae major model components (Peng et al., 2018). Considering comprehensively, Enteromorpha prolifera with high polysaccharide and Chlorella vulgaris with high proteins and lipids were selected as representative algae to produce bio-oil by HTL.

Herein, a series of solid superacid SO₄²⁻/ZrO₂ were prepared by calcining at different temperature to explore the effect of the catalyst acidity on its catalytic performance. Subsequently, it investigated the optimum temperature and dosage of the prepared catalyst on bio-oil production for the selected algae from HTL. Furthermore, it has been conducted the same performance for main components of algae including polysaccharides, proteins and lipids. The analysis of bio-oil product by means of higher heating values (HHVs), thermal gravimetric analysis (TGA) and gas chromatography-mass spectrometry (GCMS) was conductive to approve the effects of each biochemical composition for bio-oil and the formation pathways. The recycling of the SO₄²⁻/ZrO₂ catalyst used in the HTL confirmed the catalyst still exhibits good catalytic activity after 4 cycles. In this study, it firstly proposed the formation pathways of bio-oil from algae and model compounds with the solid superacid SO₄²⁻/ZrO₂ as heterogeneous catalyst. The results could provide a fundamental basis for the screening of catalyst in the HTL of algae.

2 Msterials and Methods 2.1 Materials

Crude polysaccharides, proteins and Chlorella vulgaris dry powder were purchased from Shanxi Pioneer Biotech Co., Ltd. Oleic acid and glycerol were mixed in a molar ratio of 1:3 representing a model lipid material since lipids are easily hydrolyzed to fatty acids and glycerol under hydrothermal conditions. Enteromorpha prolifera, which was collected from Stone old man area in Qingdao, China, was used as algae liquefaction raw materials. All chemicals and reagents are analytical grade without treatment. The proximate analysis of moisture, ash, volatile matter, fixed carbon and ultimate analysis of C, H, N of the same batch of algae have been previously reported by our group (Zhao et al., 2017).

2.2 Preparation of Catalyst

SO₄²⁻/ZrO₂ catalyst was prepared by precipitation-impregnation. Zirconium oxychloride solution at a concentration of 0.125 g mL⁻¹ was prepared and the added ammonium hydroxide drop by drop to adjust the pH value to 9 with rapidly stirring. After aging for 12 h at room temperature, the precipitate was filtered and washed with deionized water until chlorine-free. The filter cake was dried at 110℃ for 24 h and then grind. The resulting powder was immersed in 1 mol L⁻¹ dilute sulfuric acid for 12 h and filtered. The resulting solid was dried at 110℃ for 2 h and then calcined for 4 h.

2.3 Characterization of Catalyst

The method of back titration is selected to determine the total acid amount of SO₄²⁻/ZrO₂ catalyst (Yang et al., 2015). 0.1 g calcined SO₄²⁻/ZrO₂ catalyst was added to 20 mL 0.1 mol L⁻¹ sodium hydroxide standard solution and reacted for 1 h. The liquid was obtained by centrifugation and the remaining sodium hydroxide solution was titrated with hydrochloric acid standard solution calibrated by NaHCO₃. The formula for quantity of total acid amount of SO₄²⁻/ZrO₂ catalyst is as follows:

$ \left[\mathrm{H}^{+}\right]\left(\mathrm{mmolg}^{-1}\right)=\left(c_{\mathrm{NaOH}} \times V_{\mathrm{NaOH}}-c_{\mathrm{HCl}} \times V_{\mathrm{HCl}}\right) / m $

(1)

SO₄²⁻/ZrO₂ catalyst was characterized by D8-Advance X-ray diffractometer (XRD) from Bruker Company, Germany. Test conditions: anode target Cu target, diffraction wavelength 1.5406 Å, operating voltage 40 kV, current 40 mA, NaI scintillation counter, step length 0.05˚ step⁻¹, scanning speed 3˚ min⁻¹, scan angle 2θ = 10˚–70˚. The surface morphology and elements of SO₄²⁻/ZrO₂ were observed by scanning electron microscopy (SEM, Zeiss Supra 55) and energy X-ray dispersive spectrometer (EDS, X-MAX-20), respectively.

2.4 Apparatus and Experimental Procedure

The HTL process was performed in a stainless steel autoclave (CJK-(0.2)), equipped with an electrically heated furnace, a magnetic stirrer and a temperature controller. In a typical run, the autoclave was charged with suitable polysaccharides, protein powder, and oleic acidglycerol mixtures or algae. The ratio of material to liquid is 0.1 g mL⁻¹. The autoclave was sealed and purged with nitrogen for 5 min before heated from 200 to 300℃ with 20℃ intervals and held for 10 min at each temperature. Then autoclave was cooled down to room temperature. The gas products produced by HTL were released from the reactor and then the reactor was opened. Liquid products were separated by a series of filtration and extraction procedures.

Dichloromethane (CH₂Cl₂) was added after filtering out the residue to extract crude oil phase. The CH₂Cl₂ phase was evaporated at 35℃ under reduced pressure to remove the solvent, and the remaining liquid product is defined as bio-oil. The aqueous phase was vacuum evaporated at 60℃ to remove water and the residue was defined as water-soluble organics (WSOs). The residue was dried at 110℃ for 12 h and then weighed. Product yield was calculated by the following equations:

$ y_{b }= m_{b}/m_{f} × 100\%, $

(2)

$ y_{r }= m_{r}/m_{f} × 100\%, $

(3)

$ y_{w }= m_{w}/m_{f} × 100\%, $

(4)

$ y_{g }= 100\% − y_{b }− y_{r} − y_{w}, $

(5)

where y_b, y_r, y_w and y_g correspond to the yield of bio-oil, residue, water-soluble organics and gas, respectively. m_b, m_r, m_w and m_f refer to the mass of bio-oil, residue, water-soluble organics and feed, respectively.

2.5 Analysis of Bio-oil

Higher heating values (HHVs) of bio-oils were analyzed by using the ZR-3R combustion heat experimental device produced by Nanjing Dosuke Technology Development Co., Ltd. with 0.001℃ temperature difference resolution, and the temperature resolution is 0.01℃.

Gas chromatography-mass spectrometry (GC-MS) analyses of bio-oils were carried out using an Agilent 7890A/5975C (Santa Clara, CA, US) with a DB-5 column (30 m × 250 μm × 0.25 μm). The carrier gas was helium with a flow rate of 1.5 mL min⁻¹. 1 mL of CH₃OH solution of the sample (0.1–0.2 g mL⁻¹) was injected into the column. The injector was set at split mode with an inlet temperature of 280℃ and the diversion flow was 20 mL min⁻¹. The GC-MS oven temperature was held at 40℃ for 3 min and then raised to 300℃ with a heating rate of 10℃ min⁻¹. Compounds were identified by means of the National Institute of Standards and Technology (NIST) library of mass spectra.

Thermal gravimetric analysis (TGA) of bio-oils was performed in nitrogen atmosphere. Samples were purged to a constant weight at room temperature and then heated to 900℃ with a heating rate of 15℃ min⁻¹. The gas flow rate was 30 mL min⁻¹.

2.6 Reusability of Catalyst

The catalyst was used for successive batches. During each cycle, the residue obtained by HTL of C. vulgaris were collected and white powder was obtained after calcined at 550℃ for 4 h. Then the white powder was added into the HTL of C. vulgaris for the next cycle. The yields and HHVs of bio-oil were measured according to the assay described in the previous paragraph.

3 Results and Discussion 3.1 Characterization of Solid Superacid Catalyst SO₄²⁻/ZrO₂

Acid amount is an important property of acid catalyst, while the influence of total acid amount on the catalytic performance of SO₄²⁻/ZrO₂ catalyst is less studied. The calcination temperature has a particularly prominent effect on its acid amount. Therefore, it was measured the total acid amount of SO₄²⁻/ZrO₂ catalyst calcined at different temperature (Fig.1a). The total acid amount of SO₄²⁻/ZrO₂ catalyst fluctuates slightly but maintains above 2.5 mmol g⁻¹ when the calcination temperature ranges from 200 to 400℃. With the further increase of calcination temperature, the total acid amount of SO₄²⁻/ ZrO₂ catalyst shows a decreasing trend until it is close to zero at 800℃. The decrease of acid amount can be ascribed to the reduction of total hydroxyl content on the surface, which is influenced by the surface hydroxyl density and surface area of solid acid catalyst with the calcination temperature increases (Qin et al., 2004).

Acid strength is another fundamental property of solid acids. The acid strength of SO₄²⁻/ZrO₂ catalyst calcined at 400, 550 and 600℃ had been studied by our group (Zhao et al., 2017). The proportion of strong acidity of SO₄²⁻/ ZrO₂ catalyst calcined at 550℃ was more than 60% which is higher than 400 and 600℃. Herein, we further explore the essential reasons by XRD, which indicates the difference of acid strength can be ascribed to the different crystalline state of SO₄²⁻/ZrO₂ catalyst at different calcination temperature. Since sulfate trends to combine with materials with tetragonal crystal system, the SO₄²⁻/ZrO₂ catalyst with tetragonal crystal system possesses strong acidity and shows better catalytic performance than other types. As the XRD patterns in Fig.1b, the broad peak indicates the solid superacid catalyst SO₄²⁻/ZrO₂ calcined at 400℃ was amorphous. Therefore, the lower calcination temperature is not conducive to the formation of surface acid sites for SO₄²⁻/ZrO₂ catalyst. The peaks of tetragonal crystals appeared with the calcination temperature rising to 550℃, indicating that more tetragonal crystals appeared in the SO₄²⁻/ZrO₂ catalyst and the surface acidity enhanced. In SO₄²⁻/ZrO₂ solid superacid, sulfate acts in a double coordination form with metal ions on the surface of metal oxide. The formation of SO₄²⁻/ZrO₂ solid superacid center is mainly due to the coordination adsorption of SO₄²⁻ on the surface of metal oxide. However, the intensity of the peaks of tetragonal crystals decreased when the calcination temperature further increased to 600℃, which indicates that an excessively high calcination temperature would have an adverse effect on the formation of tetragonal crystals and reduced the acidity of SO₄²⁻/ZrO₂ catalyst. Overall, it can further conclude that 550℃ is the optimum calcination temperature for the preparation of SO₄²⁻/ZrO₂ catalyst. Fig.2 shows the surface morphology and elements of SO₄²⁻/ZrO₂ catalyst, which has a smooth surface, indicating that the material supported on the surface has not been shed by high temperature (Wang et al., 2009). Sulfur was detected in the EDS spectrum, suggesting that the sulfate ions had been successfully supported on ZrO₂.

3.2 Effect of Temperature and SO₄²⁻/ZrO₂ Catalyst Dosage on Bio-oil Yield from HTL of Algae

The HTL of algae is an extremely complex process that involves interaction among polysaccharides, proteins and lipids. In order to understand the effect of different biochemical composition of algae on the yield of bio-oil obtained from HTL, Enteromorpha prolifera with high polysaccharide and Chlorella vulgaris with high proteins and lipids are selected as representative algae.

In order to explore the optimum temperature for the HTL of Enteromorpha prolifera and Chlorella vulgaris, the experiment without catalysts were conducted firstly at different liquefaction temperature and the product distribution results were shown in Fig.3a. The optimum liquefaction temperatures for Enteromorpha prolifera and Chlorella vulgaris are 280 and 300℃ and the corresponding maximum yield of bio-oil are 12.63% and 23.39%, respectively (Fig.3b). The yields of residue from Enteromorpha prolifera HTL are higher than those of Chlorella vulgaris at different temperatures. The optimum temperature and maximum bio-oil yield for Enteromorpha prolifera are both lower than those for Chlorella vulgaris, which indicated that algae with higher proteins and lipids content required higher liquefaction temperature and higher bio-oil yield and lower residue yield can be obtained.

Keeping the optimum reaction temperature and adding SO₄²⁻/ZrO₂ catalyst, the bio-oil yield produced by Enteromorpha prolifera fluctuates slightly compared to which was catalytic with ZrO₂ carrier or no catalyst, and maintains at about 10% and the yield of residue is almost constant. Nevertheless, the yield of bio-oil for Chlorella vulgaris increases firstly and then decreases with the increase of SO₄²⁻/ZrO₂ catalyst dosage compared to which was catalytic with ZrO₂ carrier or no catalyst. It indicates that excessive use of SO₄²⁻/ZrO₂ catalyst would inhibit the formation of bio-oil. The residue yields of Enteromorpha prolifera are higher than those of Chlorella vulgaris at different SO₄²⁻/ZrO₂ dosage. As a result, SO₄²⁻/ ZrO₂ catalyst is more effective for the HTL of algae with higher proteins and lipids content, which can significantly increase the bio-oil yield and decrease the residue yield.

3.3 Effect of Temperature and SO₄²⁻/ZrO₂ Catalyst Dosage on Product Yields from HTL of Model Compounds of Algal Main Components

To further explore the influence of main composition of algae including polysaccharides, proteins and lipids on the yield and composition of bio-oil and understand the bio-oil formation pathway, the similar experimental were performed to study the effect of temperature and addition of SO₄²⁻/ZrO₂ catalyst on product distribution obtained from HTL of model compounds of algal main components.

HTL experiments were then conducted for three major model components at different temperatures (Fig.4a). The yield of bio-oil from lipids is significant higher than that of polysaccharides and proteins, since polysaccharides are freely soluble in water at room temperature and easily decomposed into water-soluble polar small molecules under the subcritical hydrothermal environment. Hence, polysaccharides had little contribution to the formation of biooil. In virtue of the corresponding optimal yield, 200, 300 and 280℃ were respectively selected as the optimal liquefaction temperature for polysaccharides, proteins and lipids.

In addition, it is also investigated the effect of SO₄²⁻/ ZrO₂ catalyst dosage on the HTL for three major model components at the corresponding optimal liquefaction temperature (Fig.4b). It can conclude that WSOs and gas products are the main products for polysaccharides in HTL. The bio-oil yield does not affect by the amount of SO₄²⁻/ZrO₂ catalyst, while the residue yield increases with SO₄²⁻/ZrO₂ catalyst dosage. This is due to the hydrolysis of polysaccharides in a subcritical water environment to produce 5-Hydroxymethylfurfural (5-HMF) which is prone to polymerize in an acidic environment to generate a coke-like substance. For proteins and lipids, only bio-oil yields are the subject of discussion. The bio-oil yields of proteins and lipids reach the highest when the amount of SO₄²⁻/ZrO₂ catalyst is 5% and 9%, respectively. By comparing the performance of three major model components in HTL, it is further verified that the contribution of main components of algae to bio-oil is lipids > proteins > polysaccharides.

3.4 Effect of Temperature and SO₄²⁻/ZrO₂ Catalyst Dosage on Bio-oil Yield from HTL of Binary Mixtures

The effects of reaction between polysaccharides and proteins on the HTL of algae can be roughly judged by the bio-oil yield distribution and bio-oil composition. A series of different proportions of polysaccharide and protein mixtures are hydrothermally liquefied at 200, 250 and 300℃, respectively (Fig.5a). When liquefaction temperature is 200℃, the increase of polysaccharides content contributes to the formation of bio-oil. However, the yield of bio-oil shows a decreasing trend in varying degrees with the increase of polysaccharides content at liquefaction temperature of 250 and 300℃. This is consistent with the higher bio-oil yield of proteins HTL at high temperatures and the higher bio-oil yield of polysaccharides HTL at low temperatures.

The influence of SO₄²⁻/ZrO₂ catalyst dosage with different proportion of polysaccharides-proteins binary mixtures on bio-oil yield is evaluated by the experimental bio-oil yield (solid line) and theoretical bio-oil yield (dash line) at the optimum liquefaction temperature of 300℃. As seen in Fig.5b, the bio-oil yield slightly increases with the addition of SO₄²⁻/ZrO₂ catalyst for the binary mixtures with 20% and 40% polysaccharide, while the yield of bio-oil decreases slightly with polysaccharides content of 80%. In addition, the experimental bio-oil yields with polysaccharides content of 80% are closer to the theoretical bio-oil yields comparing with the polysaccharide content of 20% and 40%.

HTL experiments are conducted for equal mass proportion of binary mixtures for polysaccharides-lipids and proteins-lipids, and the yields of bio-oil varying with liquefaction temperature are shown in Fig.5c. The yield of bio-oil from polysaccharides-lipids decreases with liquefaction temperature increasing which is opposite of the result of lipids HTL alone. It is speculated that the increase of liquefaction temperature is not conducive to polysaccharide HTL and even cause side reactions of polysaccharides and lipids, which led to the decrease of bio-oil yield. The similar results have also been found by Yang's group (Yang et al., 2015, 2019). On the contrary, the yield of bio-oil from proteins-lipids increases with the increase of liquefaction temperature and reaches the highest at 300℃.

HTL experiments with varying amount of SO₄²⁻/ZrO₂ catalyst are conducted on equal mass proportion of binary mixtures for polysaccharides-lipids and proteins-lipids (Fig.5d). The yield of bio-oil from proteins-lipids HTL reached the highest with 5% SO₄²⁻/ZrO₂ catalyst addition. On the contrary, the yield of bio-oil decreases for polysaccharides-lipids HTL with the addition of SO₄²⁻/ZrO₂ catalyst.

3.5 Effect of Temperature and SO₄²⁻/ZrO₂ Catalyst Dosage on Bio-oil Yield from HTL of Ternary Mixtures Simulated Algae

In order to make use of HTL results of model compounds more accurately to infer the actual liquefaction process of algae, three major components were mixed to simulate algae. Simulating Enteromorpha prolifera and Chlorella vulgaris were performed by mixing polysaccharides, proteins and lipids at a mass fraction of 80%, 18% and 2%, 10%, 60% and 30%, respectively. HTL experiments are conducted at different liquefaction temperatures and different addition of SO₄²⁻/ZrO₂ catalyst. The optimal bio-oil yield of simulated Enteromorpha prolifera is at 220℃ shown in Fig.6a. The trend of bio-oil yield varying with liquefaction temperature of simulated Enteromorpha prolifera is different with the real one for the complexity of algae structure. With respect to simulating Chlorella vulgaris, the bio-oil yield is significantly higher than that of simulating Enteromorpha prolifera at the optimal liquefaction temperature of 300℃, which is similar to the HTL results of real algae. This indicates that higher liquefaction temperature can promote HTL of algae with higher protein and lipids content and increase the yield of bio-oil.

The results of bio-oil distribution from simulated and real algae varying with the dosage of SO₄²⁻/ZrO₂ catalyst are shown in Fig.6b. For the simulation of Enteromorpha prolifera, the yield of bio-oil hardly changes with the amount of SO₄²⁻/ZrO₂ catalyst. The possible reason is that polysaccharides occupy a larger proportion in the ternary mixtures and the products distribution results are similar to that of polysaccharides HTL. Besides, as for the simulated Chlorella vulgaris, the yield of bio-oil first increases and then decreases with the increase of SO₄²⁻/ZrO₂ catalyst dosage.

3.6 Recycling of SO₄²⁻/ZrO₂ Catalyst

The reusability of the SO₄²⁻/ZrO₂ catalyst was studied by successive batches. As shown in Fig.7, the SO₄²⁻/ZrO₂ catalyst still exhibits 92.06% of its initial conversion after 4 cycles. From the XRD patterns of SO₄²⁻/ZrO₂ catalyst after each circle (Fig.8), the tetragonal peak of SO₄²⁻/ ZrO₂ catalyst is substantially the same as original SO₄²⁻/ ZrO₂ catalyst after calcination, indicating the good cyclic regeneration performance. Therefore, the solid superacid SO₄²⁻/ZrO₂ with high hydrothermal stability is considered to be a kind of good heterogeneous catalyst for regeneration recycling.

3.7 Elemental Analysis of Bio-oils

Elemental analysis results with HHVs of bio-oils produced from three major model compounds are shown in Table 1. It shows that hydrogen content for catalytic conditions of bio-oil increases significantly and carbon content decreases obviously, which resulted in an increase in the H/C ratio and HHVs of the bio-oil obtained in the presence of SO₄²⁻/ZrO₂ catalyst. High H/C ratios cause low molecular weights, which associates with TGA analysis, indicating that the light component proportion increases greatly when SO₄²⁻/ZrO₂ catalyst are employed. By using GC-MS, it can further conclude that the addition of SO₄²⁻/ZrO₂ catalyst actually promotes the decarboxylation of amino acids and causes higher H/C ratios, which is an efficient way to increase HHVs.

3.8 TGA of Bio-oils

Although thermal decomposition may occur during the heating process, TGA provides an estimate of the boiling point range of bio-oil (Yuan et al., 2010; Nazari et al., 2015; Yang et al., 2017). In this study, it carried out TGA of bio-oils obtained from catalytic and non-catalytic HTL of three major model compounds (Fig.9). The estimated boiling point distributions of bio-oils are shown in Table 2, which can be roughly divided into three temperature intervals. The results show that the addition of SO₄²⁻/ ZrO₂ catalyst slightly increases the content of light component for bio-oil obtained from polysaccharides and lipids. It also proves that SO₄²⁻/ZrO₂ catalyst possesses strongly acidity and the acid condition is beneficial to the cracking reaction to get smaller molecular weight organics. Among three model compounds, the highest content of light components in bio-oil is obtained from lipids, which reaches 73.68% with SO₄²⁻/ZrO₂ catalyst.

For bio-oil from HTL of polysaccharides, the medium component proportions decrease greatly and mainly convert into light components in the presence of SO₄²⁻/ZrO₂ catalyst. It indicates that the SO₄²⁻/ZrO₂ catalyst promotes algae to split into more small organic molecules and significantly narrow the boiling point range of bio-oil of polysaccharides.

3.9 GC-MS Analysis of Bio-oils and General Reaction Network

Compared with bio-oil, the composition of gas and aqueous phase products from HTL of algal major components is relatively simple. The gas phase products are mainly composed of CO₂, H₂ and hydrocarbons. The largest proportion of gas is CO₂ from HTL of polysaccharides and proteins analyzed by GC. The main components in the aqueous phase formed by HTL of polysaccharides are acids and alcohols, while that from HTL of proteins is Nitrogen-containing compounds such as pyrazine, piperazine and pyrrole (Yang et al., 2017). Herein, we emphatically analyze the composition of bio-oil from HTL of algal major components.

During the HTL, the biochemical composition of biooil is mainly influenced by catalyst. Therefore, the chemical composition of bio-oil obtained with and without SO₄²⁻/ZrO₂ catalyst was analyzed at the same time by GC-MS. The total ion chromatogram (TIC) of bio-oil obtained is shown in Fig.10. Although SO₄²⁻/ZrO₂ catalyst has little effect on the yield of bio-oil, it significantly changed the composition of bio-oil from HTL of polysaccharides. Organic products of polysaccharides liquefaction are mainly divided into five categories: cyclic ketones, furfural, furans, hydrocarbons and phenols (Table 3). Among these organic products, 5-HMF has the largest proportion without the catalyst, while the content of 5-HMF decreased and the content of furan increased significantly in the presence of SO₄²⁻/ZrO₂ catalyst. It is speculated that the addition of SO₄²⁻/ZrO₂ catalyst promotes the degradation of 5-HMF into furans in the HTL process. To some extent, furans are inclined to polymerize to undesired coke-like substances, which led to an increase in the yield of residue with the addition of SO₄²⁻/ZrO₂ catalyst improves. Parsa et al. (2017) reported that furfurals or furans are more unstable under alkali condition or neutral conditions. They tended to decompose by dehydration and partial degradation of furfural or furans to produce phenolic compounds. In this study, due to the strongly acidity of SO₄²⁻/ZrO₂ catalyst, the hydrolysis of furfural or furans to phenols is inhibited, resulting in a decrease in phenolic substances. Cyclic ketones can be formed by dehydration, isomerization, and cyclization of sugars and reducing sugars. The formation of cyclic ketones is inhibitd by SO₄²⁻/ZrO₂ catalyst. The way of bio-oil deoxidation during HTL is generally achieved by removing CO₂ or water molecules. The removal of oxygen in the form of water is inhibited, but the formation of CO₂ is promoted, and the H/C in bio-oil is improved, so the HHVs of bio-oil are promoted in the presence of SO₄²⁻/ZrO₂ catalyst.

Major bio-oils components from HTL of proteins can be divided into ketones, phenols, amines, amides, alcohol and various N-containing heterocyclic compounds (Table 4). The ketones in bio-oil are formed by the process of intermolecular condensation, hydrolysis and isomerization of amino acids produced by proteins. The results show that N-containing heterocyclic compounds possesses a larger proportion without catalyst and the total content decreases with the addition of catalyst, which proves that SO₄²⁻/ZrO₂ catalyst promotes the denitrification reaction. In addition, the reduction of amine compounds indicates that SO₄²⁻/ZrO₂ catalyst may inhibit the decarboxylation reaction of amino acid, resulting in the decrease of released CO₂, H/C and HHVs of bio-oil. The substantial increase of alcohols indicates the addition of SO₄²⁻/ZrO₂ catalyst promoted the deamination reaction of amino acids, while the main product carboxylic acids further reduced to alcohols. The hydrocarbon substance increases significantly, which is possibly due to SO₄²⁻/ ZrO₂ catalyst promote the decarboxylation of amino acids to generate alkanes and alkenes.

Lipids are the mixtures of glycerol and oleic acid in which glycerol is generally dissolved in the aqueous phase without being converted to oil phase products. The oleic acid is mainly reconstituted by HTL to form four kinds of fatty acids (Table 5). The addition of SO₄²⁻/ZrO₂ catalyst promoted the formation of linoleic acid, palmitic acid and n-Hexadecenoic acid but inhibited the formation of 6-Octadecenoic acid. In addition, a small amount of hydrocarbons was detected, which mainly derived from polysaccharides and proteins. The total content of fatty acid decreased, which is possibly due to the further decarboxylation of fatty acid to form hydrocarbon in the presence of SO₄²⁻/ZrO₂ catalyst.

The resulting bio-oil from HTL of polysaccharidesproteins binary mixtures is mainly composed of N-containing heterocyclic compounds, cyclic ketones, amides, phenols, alcohols and other substances (Table 6). Among them, N-containing heterocyclic compounds such as pyrazines and pyrroles have a large proportion. Yang et al. (2017) concluded that these N-containing heterocyclic compounds were formed by Maillard reactions between polysaccharides and proteins degradation products. Gai et al. (2015) reported that a part of N-containing heterocyclic compounds were formed by the reaction of proteinproduced ammonia and polysaccharide-produced cyclic oxygen compounds. By introducing SO₄²⁻/ZrO₂ catalyst, the pyrazines content in the bio-oil decreases which indicates that the catalyst might inhibit the above two reactions to increase the bio-oil yield. In addition, the content of cyclic ketones in the bio-oil also decreases with the addition of SO₄²⁻/ZrO₂ catalyst, since the cyclic ketones are the product of Maillard reaction in which are inhibited by the catalyst. However, the contents of phenols and alcohols increases obviously, indicating the SO₄²⁻/ZrO₂ catalyst promoted the deamination of amino acids and the generation of alcohols reduced from carboxylic acids. The resulting alcohols can reduce the viscosity of bio-oil to make it exhibits good fluidity.

Overall, the addition of SO₄²⁻/ZrO₂ catalyst can significantly regulate the composition and content of bio-oil and affect the degradation and interaction pathway of polysaccharides, proteins and lipids (Fig.11). During the HTL process, the reaction pathways can be divided into the following steps: 1) at low temperature stage, the main algae components polysaccharides, proteins and lipids hydrolyzed into small molecules; 2) with the increasing of temperature, the small molecules were further degraded, dehydrated, decarboxylated and deaminated to decompose into active intermediates; 3) as temperature continues to rise, Maillard reaction, amidation reaction, esterification reaction were tended to occur between the active intermediates (Galadima and Muraza, 2018; Wang et al., 2018; Marzbali et al., 2021; ). The coordination of sulfate radical with metal atoms on the oxide surface provides activity for the SO₄²⁻/ ZrO₂ solid acid (Li et al., 2018). The addition of solid acids has an important influence on macromolecular hydrolysis during the whole process of producing bio-oil, and controls the further reaction rate of bio-oil formation process, which is beneficial to accelerate the formation of bio-oil from cellular components (Yang et al., 2017).

4 Conclusions

Solid superacid SO₄²⁻/ZrO₂ has been prepared by precipitation-impregnation method and the acid amount showed a decreasing trend as the calcination temperature increased. The prepared SO₄²⁻/ZrO₂ can been seen as excellent catalyst in the HTL of Chlorella vulgaris and Enteromorpha prolifera, which promote the increasing of light component proportion and the HHVs, while inhibited the Maillard reaction between polysaccharides and proteins to upgrade the bio-oil producing. Based on the analysis of the properties of bio-oil produced by the major model components, it has put forward the formation pathways of bio-oils from HTL of real algae in the presence of solid superacid catalyst SO₄²⁻/ZrO₂.

Acknowledgements

This research was supported by the Shandong Provincial Natural Science Foundation, China (No. ZR2019BB 033) and the Fundamental Research Funds for the Central Universities of Ocean University of China (No. 201813 031).