Liquid fuel generation from algal biomass via a two-step process: effect of feedstocks

Materials

Ru/C (5 wt% Ru, surface area = 966 m²/g, metal dispersion = 23.2, and average particle size = 25 μm) was commercially available from Zhengzhou Alfachem Co., Ltd., and employed as the upgrading catalyst. Ru/C was directly mixed with crude bio-oil without undergoing prereduction under H₂ prior to use. Tetralin and dichloromethane (Shanghai Aladdin Biochemical Technology Co., Ltd., Shanghai, China) with a purity ≥ 99.9% were used as the hydrogen donor solvent and extraction solvent, respectively. Argon and helium (Changzhou Jinghua Industrial Gas Co., Ltd., Changzhou, China) were obtained with purities of 99.999%.

Two custom-made high-pressure and corrosion-resistant batch reactors (Zhengxin Instrument Factory, Yancheng, Jiangsu, China) were used to perform the experiments. The total internal volumes of these two reactors are 100 mL for the HTL reactions and 58 mL for the upgrading reactions, respectively. These two reactors were treated with supercritical water at 400 °C for 4 h to eliminate any lubricating oil during their manufacture and to oxidize the reactor wall for reduction of possible catalysis effect of the reactor wall. A custom-built molten-salt bath that consists of KNO₃ and NaNO₃ at a mass ratio of 5:4 was employed as a heat source to heat these two reactors.

HTL

The previous studies [3, 4, 5] suggested that the maximum bio-oil yield was typically obtained at 350 °C for 1 h when subjected algal biomass to the HTL process. Therefore, 350 °C and 1 h were selected as parameters for the HTL reactions. Aliquots of 20.0 g of dry algae powder and 40.0 mL of DI water were added into the 100 mL reactor. The air in the loaded reactor was eliminated by flushing the reactor head space with helium for roughly 20 min and further charged with helium at 0.01 MPa. This known amount of helium acted as an internal standard to determine the gas yield. The loaded and pressurized reactor was put into a molten-salt bath pretreated to 350 °C and maintained at 350 ± 2 °C using a temperature controller. About 18 min was cost as the temperature inside the reactor achieved at 350 °C, and the reaction time was set zero at this time. The reaction lasted for 60 min. The reactor was taken out of the molten-salt bath and immediately placed into a cool water bath to quench the reaction. The cooled reactor was placed at room temperature for 4 h to ensure the uniformity of gas in the reactor.

The outlet of the reactor was connected to a six-way valve inlet of gas chromatography, and the pressurized gases in the reactor were fed into a 1 mL sample loop as the reactor valve was cutoff slowly until carrier gas in the sample loop was completely exchanged with the sample gas. After analyzing the gas fraction, the reactor was opened to recover the liquid fraction. The gas weight was estimated by mass difference before and after exhausting the gas from the reactor. The reactor was charged with 20 mL dichloromethane and sealed and vigorously shaken by hand. The same operation was repeated for three times to maximize recovery of reactor contents. The solid residue in the reactor contents was isolated by filtration and dried in an oven at 110 °C for 12 h and then weighed. The separated liquid phase was further subjected to a separatory funnel to get aqueous phase and dichloromethane extracts. The dichloromethane was then evaporated to get the crude bio-oil. The yield of each product fraction was calculated as its mass divided by the mass of algae loaded into the reactor.

Hydro-upgrading

A previous research suggested that 400 °C was a more appropriate temperature for crude algal bio-oil upgrading [7]. Thus, in the present study, 400 °C was employed as the upgrading temperature. The mentioned above 58 mL batch reactor was used to perform the upgrading experiments. In a typical run, 5.0 g of crude bio-oil, 0.5 g of the Ru/C catalyst, and 15 g of tetralin were introduced into the reactor and tightly sealed. The air in the loaded reactor was removed by purging the reactor head space with helium for about 20 min and further charged with helium at 1.0 MPa. No additional hydrogen was added. The upgrading reaction was initiated by immerging the reactor body into a molten-salts tank preheated to 400 °C. Approximately 30 min elapsed as the temperature inside the reactor mounted from room temperature to 400 °C, and then, the reaction time was counted. After the upgrading reaction was lasted for 2 h, the reactor was taken out of the molten-salt tank and placed into an ice-water bath to stop the reaction. The same procedure for subsequent product separation as that of the HTL mentioned above was employed. Dichloromethane was also used as an extraction solvent. The noncatalyst solid residue was considered as the coke whose amount was calculated from the mass difference between the solid residue and the initial loaded catalyst. The mass ratio of each product fraction to the total mass of the loaded crude bio-oils and tetralin was defined as the yield of each product fraction.

The results presented in this study are all the average results of three repeated experiments. The results reported in this study represent the mean values for three independent trials. The uncertainties are reported as the experimentally determined standard deviations.

Product analysis

Bligh–Dyer method was used to quantify the crude fat content in algae [29]. The ash content was tested according to the ASTM [30]. The volatile matter content was quantified from the weight difference before and after heating a known amount of algae sample at 575 °C under a N₂ atmosphere. The fixed carbon content was calculated by the following equation: \( {\text{fixed carbon}}\% \, = \, 100\% \, - \,{\text{moisture}}\% \, - \,{\text{volatile}}\% \, - \,{\text{ash}}\% . \) Kjeldahl method [31] was applied to roughly estimate the protein content in algae. The carbohydrate content was calculated by the following equation: \( {\text{carbohydrate}}\% \, = \, 100\% \, - \,{\text{moisture}}\% \, - \,{\text{crude fat}}\% \, - \,{\text{protein}}\% . \) X-ray fluorescence (Bruker S8 TIGER, Karlsruhe, Germany) was used to analyze the inorganic composition of algae and Table 1 presents the results.

An organic element analyzer (Flash 2000) (Thermo Fisher Scientific, Waltham, MA, USA) was used to quantify C, H, O, N, and S in the sample. The total S and N contents in the bio-oil were determined based on ASTM D5453-12 [32] and D4629-12 [33], respectively, as their values in the bio-oil are below detection limit of the organic element analyzer. Dulong formula was employed to estimate the higher heating value (HHV) of the algae and bio-oils as follows:

A gas chromatography coupled with a mass spectrometry (7890A GC-5975C) (Agilent Technologies Co. Ltd., CA, Palo Alto, USA) was used to qualify molecular compositions of the bio-oil. More details are available in Ref. [21]. An automatic total-acidity determination instrument (YUS-A2 Zhengzhou Zhonggu Machinery Equipment Co., Ltd., China) was used to test the total acid number (TAN) of the bio-oil. Thermogravimetric analysis was employed to simulate the boiling-point distribution of the bio-oil. An SDT Q600 Simultaneous DSC-TGA instrument (TA Instruments Co. Ltd., New Castle, DE, USA) was used. A known amount of sample was heated to 780 °C at a heating rate of 10 °C/min in N₂ atmosphere at a flow rate of 10 mL/min.

Characterization of the algal biomass feedstocks

Table 1 shows the proximate and ultimate analyses of the algal biomass feedstocks used in the present study. The relative amount of crude fat, which contributes most significantly to the bio-oil production during the thermo-chemical process, was significantly higher in the microalgae than that in the macroalgae. The macroalgae contain higher amount of ash than the microalgae. Therefore, as expected, macroalgae contain less volatile than microalgae. The fixed carbon ranged from 17.26 to 23.92% for the microalgae and from 18.63 to 25.53% for the macroalgae. Based on differences in lipid, volatile matter, and fixed carbon, the yields of the HTL products can be expected to be different between microalgae and macroalgae. The differences in the mineral content of algae also depend on the geographical location, oceanic residence time, season, and the species of algae. The microalgae were usually cultured in fresh water, while the macroalgae were cultured in seawater. Therefore, the macroalgae (particular the UP) contained higher chlorine contents (1.58–19.90 wt%) than the microalgae because of their saline environment. Both the microalgae and macroalgae contained large amounts of macrominerals (Ca, K, Mg, Si, Al, P, and Na) and trace elements (Mo, Ti, Fe, Cr, Cu, Mn, Sr, Ni, Zn, and Ru). Biosorption might be responsible for the presence of trace elements in algae. Slagging and fouling usually happened during the thermo-chemical process of those biomasses that contain high amount of alkali metals. The product distribution during the HTL was also influenced by the ash contained in the biomass.

TG analysis of the algal biomass feedstocks

Detailed and accurate characterization of algae is a necessity for any biomass-to-biofuel conversion process. Understanding how the individual component of biomass and reaction products interacts at each stage in the process is important for researchers.

The thermal behavior of algae can be learned by subjecting it to a TG analysis under an inert atmosphere. Both TG and derivative thermogravimetric (DTG) analyses of the algae samples were performed by checking the mass loss with increasing time and temperature. Figure 1a–d shows all the results. This is the first study to compare the thermal decomposition behavior of a range of different algae under identical conditions. Algae mainly consist of proteins, carbohydrates, and lipids, among which the lipid content of microalgae is commonly higher than that of macroalgae. These components were transformed into bio-oil, gas, solid residue, and water-soluble products during HTL. The TG and DTG curves can be used to determine the temperature at which each biochemical component of algae becomes completely pyrolyzed. Of course, secondary reactions between biochemical compounds can also affect the yield of the product distribution. In general, the microalgae species showed similar TGA and DTG curves that were different from those of the macroalgae species. The TG curve revealed three stages of thermal degradation for the algae during pyrolysis. As the DTG curves shown in Fig. 1b, d, Stage I is the dehydration step at temperatures ranging from 30 to 150 °C during which the cellular water in the feedstock evaporated. Mass loss was observed for the both microalgae and macroalgae, indicating that the physically dried algae still contained some water. More water mass loss was observed for the macroalgae than the microalgae due to the higher carbohydrate content in the macroalgae. As the temperature rose to 200 °C, the mass loss entered the second stage, devolatilization, during which the crude lipids, proteins, and carbohydrates gradually decomposed, leading to a significant weight reduction and the production of main pyrolytic products. In Stage II, two shoulders were observed in the DTG curves that corresponded to the initial decomposition temperatures of the proteins and carbohydrates [34], respectively. In Stage III, within the temperature range of 500–800 °C, more volatiles were released and a char residue was formed. Since the macroalgae contained more ash (mainly consists of inorganic salts) than the microalgae, more weight loss was also observed for the macroalgae when the temperature rose by more than 600 °C due to the vaporization of the salts. The DTG curves of all algae indicated that 350 °C was the suitable temperature. Therefore, 350 °C was chosen as the HTL temperature for the subsequent study.

Open image in new window

Effect of algal biomass feedstocks on the yield of the crude bio-oil and solid residue

HTL can convert algae to crude bio-oil, solid residue, water-soluble products, and gases. In this present study, only the yield of the solid residue, crude bio-oil, and gaseous products was concerned. Figure 2 reveals the effects of the different algae on yields of the crude bio-oil and solid residue produced from HTL at 350 °C for 60 min. The yields of the bio-oil and solid residue were estimated based on the dry feedstocks. For comparison purpose, the bio-oil yield was also calculated based on the dry-ash free feedstocks. Figure 2 indicates that the crude bio-oil and crude bio-oil (dry-ash free, DAF) yields derived from microalgae were all larger than those of produced from macroalgae. The decomposition of any of the biochemical components, such as the proteins, lipids, and carbohydrates, in the feedstock could dedicate to the production of the crude bio-oil, among which carbohydrate decomposition mainly resulted in the formation of water-soluble products [14]. Therefore, the product yields and total lipid and protein content (TLP) value are also listed in Fig. 2. For microalgae, a clear trend was observed, that is, the larger the TLP content was, the larger the crude bio-oil yield_(DAF) was. For macroalgae, however, the crude bio-oil yield_(DAF) showed no clear trend. Table 1 shows that the macroalgae commonly contained a certain amount of ash that would affect the HTL process and, subsequently, the product distribution. Of course, this catalytic role was likely due to the different reactivities of the individual lipids, proteins, and carbohydrates. The crude bio-oil yields also showed significant differences within the microalgae or macroalgae species. For the microalgae, the crude bio-oi yield_(DAF) ranged from 36.50 to 58.04 wt%. For the macroalgae, the crude bio-oi yield_(DAF) ranged from 19.56 to 36.44 wt%. The compositions of various microalgae or macroalgae species generally exhibit pronounced seasonal variation. In addition, such species can be strongly affected by the temperature, geographical location [35], water salinity, and aqueous nutrient content [36]; hence, samples from the same species that are grown in alternative climates can differ substantially. For the NO, a 43 wt% crude bio-oil yield was obtained at 350 °C for 60 min by Brown et al. [5]. They used a feedstock concentration of 4.5 wt% compared to the 33 wt% feedstock used in the present study. Zou et al. [37] suggested that increasing the concentration of feedstock would cause a reduced bio-oil yield. Normally, a higher bio-oil yield was observed at a lower feedstock concentration; however, a large amount of water will be required at the same time, which will increase energy consumption and costs for handling the down-stream waste water. Zhou et al. [38] achieved a crude bio-oil yield of 23 wt% as they hydrothermally liquefied UP at 300 °C for 30 min with 5 wt% of added Na₂CO₃. Therefore, in view of the crude bio-oil yield, SL is the algal species with the greatest potential for the generation of crude bio-oil. Figure 2 also indicates that the solid residue yields produced from macroalgae are overall higher those obtained using microalgae; this result is consistent with the higher ash and carbohydrate contents from the macroalgae as carbohydrates are more difficult to convert than proteins and lipids [32]. HTL of the algal biomass would also produce a certain amount of gaseous products which mainly consisted of H₂, CO, CH₄, and CO₂ together with a small quantity of C₂–C₅ hydrocarbons. Figure 3 presents the amounts of gaseous products produced from the HTL of the eight algae. The yield of CO₂ was always the largest, which is consistent with the previous studies on the HTL of biomass [6, 9]. Two reactions such as water–gas shift and steam reforming are responsible for the formation of CO₂. In addition, the decarboxylation of fatty acids also favored the production of CO₂. Among the examined eight algae, SJ produced the highest CO₂ yield of 5.9 mmol/g. H₂ can also be generated during steam reforming and water–gas shift reactions; however, the H₂ yield was much lower than that of the CO₂, indicating that the production mechanisms of these two products were different. SJ provided the highest H₂ yield of 0.53 mmol/g. CO was always present in the lowest amount in the gaseous fraction as the initial formed CO would be consumed from reactions such as the water–gas shift and methanation as the reaction proceeds [6]. The production of CH₄ mainly occurred during the methanation reaction, and its yield was low for all algal biomass feedstocks.

Open image in new window

Open image in new window

Characterization of the crude bio-oil and solid residue

The heteroatoms likely evidenced the existence of the compounds derived from proteins, carbohydrates, and lipids in the algae and formed in reactions, such as depolymerization, decomposition, and addition reactions [36]. Table 2A shows that the C and H contents and energy densities of all crude bio-oils are significantly larger than those of their feedstocks; these results are consistent with those from the previous studies [4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16]. All the crude bio-oils showed similar C content that ranged from 73.79 to 76.40 wt%. The H content of the microalgae was larger than that of the macroalgae for all samples. Consequently, the H content in the crude bio-oils produced from the microalgae was higher than that produced from the macroalgae for all samples. However, within the microalgae or macroalgae groups, the feedstocks with higher hydrogen content did not produce a crude bio-oil with a higher hydrogen content. The O content of all crude bio-oils was significantly lower than that of their corresponding feedstocks, indicating in situ deoxygenation occurred simultaneously during HTL. This deoxygenation process could increase the energy density of the crude bio-oils and thus improve the properties of the crude bio-oils. Compared with the feedstocks, the N content in the crude bio-oils produced from the microalgae was generally reduced. However, contrary results were observed for the macroalgae and these results were also evidenced by Zhou et al. [34]. Hence, it appears to be more difficult to remove N from macroalgae compared to microalgae. The N in the crude bio-oil was mainly derived from the conversion of proteins; the larger the protein content in the algal biomass was, the larger the N in the crude bio-oil was. The S content in all crude bio-oils was generally reduced relative to the feedstock, but this reduction was small. The H/C molar ratio is an important indicator of the quality of a bio-oil; a higher H/C molar ratio indicates a higher quality bio-oil. The crude bio-oils produced from the microalgae have higher H/C molar ratios than those produced from the macroalgae. The energy density of the crude bio-oils showed little variation across the species, ranging from 36.94 to 37.67 MJ/kg for the microalgae and from 35.67 to 37.33 MJ/kg for the macroalgae, despite the significant variation in the feedstock biochemical compositions and crude bio-oil yields. Due to the higher yield and energy density of the crude bio-oil produced from the microalgae, a higher ER was also observed for the microalgae, among which SL presented the highest ER of 0.79. The total ER from the crude bio-oil, solid residue, and gas indicated that a massive energy loss occurred in the aqueous phase, particular for the macroalgae. The lost energy could be recovered using additional processing steps such as hydrothermal gasification. The TAN values of all crude bio-oils produced from the microalgae were similar to each other and ranged from 56.73 to 59.85. The TAN values of all crude bio-oils produced from the macroalgae were different from each other and ranged from 37.82 to 87.83, indicating the much different biochemical composition of the macroalgal crude oil. All the crude bio-oils produced either from the microalgae or macroalgae contained heteroatom concentrations that were significantly higher than the requirement of ASTM [39]. Therefore, further treatment is needed before considering the utilization of algal bio-oil. Although the crude bio-oils were obtained by the solvent extraction method, small amounts of impurities would still reside in the crude bio-oils and might interfere with the catalytic activity of the catalyst during the subsequent upgrading process.

Table 2B shows that C is the predominant element in all the solid residues, indicating that some of the organic matter in the algal biomass was not completely converted during the HTL, particularly the carbohydrates. The solid residue produced from PY contained the highest C content of 62.54 wt%, indicating that the organic matter in PY was the most difficult to convert. In contrast, the solid residue produced from ArP contained the lowest C content of 6.97 wt%, indicating that the organic matter in ArP was the easiest to convert. All the solid residues contained a certain amount of N and S, implying that some protein or protein derivatives resided in the solid residue.

To compare the differences among the molecular compositions of the crude bio-oils generated from the eight different algae, the individual components in the crude bio-oils were qualified by the aid of GC–MS. The temperature of the gasification chamber was set to 300 °C. The thermogravimetric analysis (TG) suggested that a minimum of 50–60 wt% of the material in the crude bio-oil samples was volatilized at this temperature, whereas the rest exhibited a higher boiling point. Because the bio-oils did not totally volatilize, the acquired data are relevant only to the volatile fraction and do not necessarily represent all the ingredients of the bio-oil. Figure 4a–h shows the total ion chromatograms (TICs) of the crude bio-oils derived from those eight different algae. Significant TIC differences in the crude bio-oils are clearly observed in Fig. 4a–h. In particular, clear differences exist between the crude bio-oils derived from the microalgae and macroalgae because of the different biochemical compositions among the algae. Only a few large peaks at retention times shorter than 40 min appeared for all of the crude bio-oils, among which the TICs of the crude bio-oils are centered within 20 min for the macroalgae and 20–40 min for the microalgae. That is, the crude bio-oils produced from macroalgae contained more light-end products than those of produced from microalgae. Computer matching and a mass spectral library were used to facilitate compound identification. Additional file 1: Table S1 summarizes the GS–MS identification of the major components in the crude bio-oils derived from the eight different algae. For all crude bio-oils, only those compounds whose area % in the TIC was over 1.0% were provided. For the microalgae, NO, AuP, and ArP produced crude bio-oils with a certain amount of phytane and phytene, but the crude bio-oils produced from AuP, SL, and ArP also contained significant quantities of palmitamide, among which that from ArP also contained a certain amount of pyrrolidinone derivatives. In contrast, benzene derivatives were the major products of the crude bio-oils produced from macroalgae. For example, the crude bio-oil produced from UP contained 17.4% benzene and 14.8% ethyl benzene, but the crude bio-oil produced from SJ contained 9.3% 2,3-dimethyl-2-cyclopenten-1-one. The GC–MS results suggested that biomasses with different biochemical components would lead to crude bio-oils with different molecular compositions.

Open image in new window

Effect of the crude bio-oils on the yields of the upgraded bio-oil and coke

The goal of the upgrading process is to reduce the viscosity and decrease the N, O, and S contents of crude bio-oils. The crude bio-oil was converted into three major fractions of upgraded bio-oil, solid residue, and gaseous products after it was subjected to treatment with tetralin and Ru/C at 400 °C for 2 h. The mass ratio of the crude bio-oil to tetralin was 1:3. This present study mainly centered on the effects of the crude bio-oils on the yields of the upgraded bio-oil, coke, and gaseous products.

Figure 5 shows the effects of crude bio-oils on the yields of the upgraded bio-oil and coke. A blank experiment with only tetralin (without the addition of crude bio-oil) was also performed under the same reaction conditions. The results show that 96 wt% liquid products were recovered and almost no coke was observed, suggesting that the tetralin was only slightly decomposed. Therefore, the upgraded bio-oil predominantly consisted of tetralin and its derivatives. For comparison purposes, the actual upgraded bio-oil and coke yields are also provided in Fig. 5. The yields of the upgraded bio-oil and the actual upgraded bio-oil are reported within a lower range as mass loss occurred during the sample handling process, particularly during the solvent elimination process. As shown in Fig. 5, the crude bio-oils only slightly affected the yield of the upgraded bio-oil due to the existence of a large amount of tetralin in the products, which ranged from 83.69 to 96.10 wt% for microalgae and 83.72 to 85.35 wt% for macroalgae, respectively. However, the crude bio-oil affected the actual yield of the upgraded bio-oil due to the large differences among the molecular compositions of the crude bio-oils. Due to the high viscosity and asphaltene content of the crude bio-oils, depolymerization and dehydrogenation generally occurred at severe temperatures during the upgrading process [21, 22], and coke was subsequently formed. Since only a minimal amount of coke was formed during the blank experiment, the coke yield was dominantly produced from the crude bio-oil. Figure 5 shows that the type of crude bio-oil used affected the coke yield during the crude bio-oil upgrading, particular the actual coke yield. Table 3 indicates that the heavy-oil fractions with boiling points > 450 °C in the crude bio-oil produced from ArP and AuP were higher than 10 wt%, particularly for the crude bio-oil produced from AuP (15.91 wt%). Therefore, higher coke yields were observed for those crude bio-oils produced from ArP and AuP, 10.11 and 9.54 wt%, which corresponded to actual coke yields of 40.44 and 38.16 wt%, respectively. The same coke yield trends were also observed for the crude bio-oils produced from macroalgae. A 17.3 wt% coke yield was observed as the pretreated AuP bio-oil was hydrothermally treated at 400 °C for 4 h with Ru/C and 6 MPa H₂ [22], and a 21.8 wt% coke yield was obtained when the pretreated AuP bio-oil was thermally treated at 400 °C for 1 h with Pt/γ-Al₂O₃ and 6 MPa H₂ [21]. In the present study, the actual coke yield could reach as high as 40.44 wt%, indicating that tetralin is not as effective as external hydrogen for suppressing coke formation or that the hydrogen provided was not sufficient to control the coke formation. These cokes would cover the catalyst surfaces during the upgrading process or enter the pores of the catalyst supports and decrease or deactivate the catalyst activity. Therefore, effective coke control is indispensable for maintaining the high activity of catalysts during crude bio-oil upgrading.

Open image in new window

The crude algal bio-oils upgrading also produced some gaseous products. The main identified gaseous products were H₂, CO, and CH₄. A similar set of gaseous products was reported as the crude bio-oil was treated in supercritical water [21, 22]. Only small quantities of CO₂ were obtained as the O in the crude bio-oils was eliminated in the form of H₂O and not from decarboxylation.

Characterization of the upgraded bio-oils and coke

Since the catalyst could not be isolated from the solid residue after the reaction, the solid residue (catalyst + coke) was directly analyzed using the elemental analyzer. Table 4B lists the elemental compositions of all solid resides. Since the coke and catalyst stayed together after the upgrading process, these values present only the averaged elemental compositions of the coke and catalyst. Table 4B shows that C (higher than 50 wt%) was the predominant element with Ru/C present in the solid residue. A certain amount of H was also detected in the solid residue, and the amount of H ranged from 2.40 wt% for NS to 3.20 wt% for ArP, indicating the existence of organic matter in the solid residue. Large amounts of S and N were also observed in the solid residue, indicating that Ru/C could adsorb N- and S-containing compounds or N- and S-containing polymers to form coke. Basically, the higher the N and S contents of the crude bio-oils were, the higher the N and S contents of the solid residues were. These N and S atoms could poison the active sites and deactivate the catalyst.